modelId
string | author
string | last_modified
timestamp[us, tz=UTC] | downloads
int64 | likes
int64 | library_name
string | tags
sequence | pipeline_tag
string | createdAt
timestamp[us, tz=UTC] | card
string |
|---|---|---|---|---|---|---|---|---|---|
danaash/roger_dean_style_LoRA | danaash | 2025-06-15T19:45:57Z | 0 | 0 | diffusers | [
"diffusers",
"text-to-image",
"diffusers-training",
"lora",
"template:sd-lora",
"stable-diffusion-xl",
"stable-diffusion-xl-diffusers",
"base_model:stabilityai/stable-diffusion-xl-base-1.0",
"base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0",
"license:openrail++",
"region:us"
] | text-to-image | 2025-06-15T19:45:56Z | ---
base_model: stabilityai/stable-diffusion-xl-base-1.0
library_name: diffusers
license: openrail++
instance_prompt: roger dean style of fantasy
widget: []
tags:
- text-to-image
- text-to-image
- diffusers-training
- diffusers
- lora
- template:sd-lora
- stable-diffusion-xl
- stable-diffusion-xl-diffusers
---
<!-- This model card has been generated automatically according to the information the training script had access to. You
should probably proofread and complete it, then remove this comment. -->
# SDXL LoRA DreamBooth - danaash/roger_dean_style_LoRA
<Gallery />
## Model description
These are danaash/roger_dean_style_LoRA LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0.
The weights were trained using [DreamBooth](https://dreambooth.github.io/).
LoRA for the text encoder was enabled: False.
Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.
## Trigger words
You should use roger dean style of fantasy to trigger the image generation.
## Download model
Weights for this model are available in Safetensors format.
[Download](danaash/roger_dean_style_LoRA/tree/main) them in the Files & versions tab.
## Intended uses & limitations
#### How to use
```python
# TODO: add an example code snippet for running this diffusion pipeline
```
#### Limitations and bias
[TODO: provide examples of latent issues and potential remediations]
## Training details
[TODO: describe the data used to train the model] |
Mungert/DistilQwen2.5-DS3-0324-32B-GGUF | Mungert | 2025-06-15T19:45:50Z | 246 | 2 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-05-05T15:03:58Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">DistilQwen2.5-DS3-0324-32B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `DistilQwen2.5-DS3-0324-32B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `DistilQwen2.5-DS3-0324-32B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `DistilQwen2.5-DS3-0324-32B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `DistilQwen2.5-DS3-0324-32B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `DistilQwen2.5-DS3-0324-32B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `DistilQwen2.5-DS3-0324-32B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `DistilQwen2.5-DS3-0324-32B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `DistilQwen2.5-DS3-0324-32B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `DistilQwen2.5-DS3-0324-32B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `DistilQwen2.5-DS3-0324-32B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `DistilQwen2.5-DS3-0324-32B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://freenetworkmonitor.click/dashboard/?assistant=open)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by logging in or [downloading our Quantum Network Monitor Agent with integrated AI Assistant](https://freenetworkmonitor.click/download)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
---
license: apache-2.0
---
## 📖 Introduction
# DistilQwen2.5-DS3-0324 Series: Fast-Thinking Reasoning Models
## Overview
In response to the industry challenge of balancing efficient reasoning with cognitive capabilities, the DistilQwen2.5-DS3-0324 series innovatively transfers the fast-thinking capabilities of DeepSeekV3-0324 to lightweight models. Through a two-stage distillation framework, this series achieves high performance while delivering:
- **Enhanced Reasoning Speed**: Reduces output tokens by 60-80% (compared to slow-thinking models)
- **Reduced Resource Consumption**: Suitable for edge computing deployment
- **Elimination of Cognitive Bias**: Proprietary trajectory alignment technology
## Core Innovations
### 1. Fast-Thinking Distillation Framework
- **Stage 1: Fast-Thinking CoT Data Collection**
- **Long-to-Short Rewriting**: Extracts key reasoning steps from DeepSeek-R1
- **Teacher Model Distillation**: Captures the rapid reasoning trajectories of DeepSeekV3-0324
- **Stage 2: CoT Trajectory Cognitive Alignment**
- **Dynamic Difficulty Grading** (Easy/Medium/Hard)
- LLM-as-a-Judge evaluates small model comprehensibility
- Simple chain expansion → Adds necessary steps
- Hard chain simplification → Removes high-level logical leaps
- **Validation Mechanism**: Iterative optimization until all data reaches "Medium" rating
### 2. Performance Breakthroughs
- **32B Model** approaches the performance of closed-source models with 10x the parameters on the GPQA Diamond benchmark
- **Significant Improvement in Reasoning Efficiency** (see comparison table below)
| Model | MMLU_PRO Tokens | AIME2024 Tokens | Speed Gain |
|--------------------------------|-----------------|-----------------|------------|
| DistilQwen2.5-R1-32B (Slow-Thinking) | 4198 | 12178 | 1x |
| DistilQwen2.5-DS3-0324-32B | 690 | 4177 | 5-8x |
## Technical Advantages
- **Two-Stage Distillation**: First compresses reasoning length, then aligns cognitive trajectories
- **Dynamic Data Optimization**: Adaptive difficulty adjustment ensures knowledge transferability
- **Open-Source Compatibility**: Fine-tuned based on the Qwen2.5 base model
## 🚀 Quick Start
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model = AutoModelForCausalLM.from_pretrained(
"alibaba-pai/DistilQwen2.5-DS3-0324-32B",
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("alibaba-pai/DistilQwen2.5-DS3-0324-32B")
prompt = "Give me a short introduction to large language model."
messages=[
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant. You should think step-by-step."},
{"role": "user", "content": prompt},
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=2048，
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
``` |
Mungert/DistilQwen2.5-DS3-0324-7B-GGUF | Mungert | 2025-06-15T19:45:41Z | 233 | 1 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-05-04T01:28:02Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">DistilQwen2.5-DS3-0324-7B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `DistilQwen2.5-DS3-0324-7B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `DistilQwen2.5-DS3-0324-7B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `DistilQwen2.5-DS3-0324-7B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `DistilQwen2.5-DS3-0324-7B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `DistilQwen2.5-DS3-0324-7B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `DistilQwen2.5-DS3-0324-7B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `DistilQwen2.5-DS3-0324-7B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `DistilQwen2.5-DS3-0324-7B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `DistilQwen2.5-DS3-0324-7B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `DistilQwen2.5-DS3-0324-7B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `DistilQwen2.5-DS3-0324-7B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
## 📖 Introduction
# DistilQwen2.5-DS3-0324 Series: Fast-Thinking Reasoning Models
## Overview
In response to the industry challenge of balancing efficient reasoning with cognitive capabilities, the DistilQwen2.5-DS3-0324 series innovatively transfers the fast-thinking capabilities of DeepSeekV3-0324 to lightweight models. Through a two-stage distillation framework, this series achieves high performance while delivering:
- **Enhanced Reasoning Speed**: Reduces output tokens by 60-80% (compared to slow-thinking models)
- **Reduced Resource Consumption**: Suitable for edge computing deployment
- **Elimination of Cognitive Bias**: Proprietary trajectory alignment technology
## Core Innovations
### 1. Fast-Thinking Distillation Framework
- **Stage 1: Fast-Thinking CoT Data Collection**
- **Long-to-Short Rewriting**: Extracts key reasoning steps from DeepSeek-R1
- **Teacher Model Distillation**: Captures the rapid reasoning trajectories of DeepSeekV3-0324
- **Stage 2: CoT Trajectory Cognitive Alignment**
- **Dynamic Difficulty Grading** (Easy/Medium/Hard)
- LLM-as-a-Judge evaluates small model comprehensibility
- Simple chain expansion → Adds necessary steps
- Hard chain simplification → Removes high-level logical leaps
- **Validation Mechanism**: Iterative optimization until all data reaches "Medium" rating
### 2. Performance Breakthroughs
- **32B Model** approaches the performance of closed-source models with 10x the parameters on the GPQA Diamond benchmark
- **Significant Improvement in Reasoning Efficiency** (see comparison table below)
| Model | MMLU_PRO Tokens | AIME2024 Tokens | Speed Gain |
|--------------------------------|-----------------|-----------------|------------|
| DistilQwen2.5-R1-32B (Slow-Thinking) | 4198 | 12178 | 1x |
| DistilQwen2.5-DS3-0324-32B | 690 | 4177 | 5-8x |
## Technical Advantages
- **Two-Stage Distillation**: First compresses reasoning length, then aligns cognitive trajectories
- **Dynamic Data Optimization**: Adaptive difficulty adjustment ensures knowledge transferability
- **Open-Source Compatibility**: Fine-tuned based on the Qwen2.5 base model
## 🚀 Quick Start
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model = AutoModelForCausalLM.from_pretrained(
"alibaba-pai/DistilQwen2.5-DS3-0324-7B",
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("alibaba-pai/DistilQwen2.5-DS3-0324-7B")
prompt = "Give me a short introduction to large language model."
messages=[
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant. You should think step-by-step."},
{"role": "user", "content": prompt},
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=2048，
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
``` |
Mungert/Phi-4-reasoning-GGUF | Mungert | 2025-06-15T19:45:36Z | 690 | 1 | transformers | [
"transformers",
"gguf",
"phi",
"nlp",
"math",
"code",
"chat",
"conversational",
"reasoning",
"text-generation",
"en",
"base_model:microsoft/phi-4",
"base_model:quantized:microsoft/phi-4",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix"
] | text-generation | 2025-05-02T22:51:42Z | ---
license: mit
license_link: https://huggingface.co/microsoft/Phi-4-reasoning/resolve/main/LICENSE
language:
- en
base_model:
- microsoft/phi-4
pipeline_tag: text-generation
tags:
- phi
- nlp
- math
- code
- chat
- conversational
- reasoning
inference:
parameters:
temperature: 0
widget:
- messages:
- role: user
content: What is the derivative of x^2?
library_name: transformers
---
# <span style="color: #7FFF7F;">Phi-4-reasoning GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Phi-4-reasoning-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Phi-4-reasoning-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Phi-4-reasoning-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Phi-4-reasoning-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Phi-4-reasoning-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Phi-4-reasoning-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Phi-4-reasoning-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Phi-4-reasoning-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Phi-4-reasoning-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Phi-4-reasoning-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Phi-4-reasoning-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Phi-4-reasoning Model Card
[Phi-4-reasoning Technical Report](https://aka.ms/phi-reasoning/techreport)
## Model Summary
| | |
|-------------------------|-------------------------------------------------------------------------------|
| **Developers** | Microsoft Research |
| **Description** | Phi-4-reasoning is a state-of-the-art open-weight reasoning model finetuned from Phi-4 using supervised fine-tuning on a dataset of chain-of-thought traces and reinforcement learning. The supervised fine-tuning dataset includes a blend of synthetic prompts and high-quality filtered data from public domain websites, focused on math, science, and coding skills as well as alignment data for safety and Responsible AI. The goal of this approach was to ensure that small capable models were trained with data focused on high quality and advanced reasoning. |
| **Architecture** | Base model same as previously released Phi-4, 14B parameters, dense decoder-only Transformer model |
| **Inputs** | Text, best suited for prompts in the chat format |
| **Context length** | 32k tokens |
| **GPUs** | 32 H100-80G |
| **Training time** | 2.5 days |
| **Training data** | 16B tokens, ~8.3B unique tokens |
| **Outputs** | Generated text in response to the input. Model responses have two sections, namely, a reasoning chain-of-thought block followed by a summarization block |
| **Dates** | January 2025 – April 2025 |
| **Status** | Static model trained on an offline dataset with cutoff dates of March 2025 and earlier for publicly available data |
| **Release date** | April 30, 2025 |
| **License** | MIT |
## Intended Use
| | |
|-------------------------------|-------------------------------------------------------------------------|
| **Primary Use Cases** | Our model is designed to accelerate research on language models, for use as a building block for generative AI powered features. It provides uses for general purpose AI systems and applications (primarily in English) which require:<br><br>1. Memory/compute constrained environments.<br>2. Latency bound scenarios.<br>3. Reasoning and logic. |
| **Out-of-Scope Use Cases** | This model is designed and tested for math reasoning only. Our models are not specifically designed or evaluated for all downstream purposes. Developers should consider common limitations of language models as they select use cases, and evaluate and mitigate for accuracy, safety, and fairness before using within a specific downstream use case, particularly for high-risk scenarios. Developers should be aware of and adhere to applicable laws or regulations (including privacy, trade compliance laws, etc.) that are relevant to their use case, including the model’s focus on English. Review the Responsible AI Considerations section below for further guidance when choosing a use case. Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under. |
## Usage
### Inference Parameters
Inference is better with `temperature=0.8`, `top_p=0.95`, and `do_sample=True`. For more complex queries, set the maximum number of tokens to 32k to allow for longer chain-of-thought (CoT).
### Input Formats
Given the nature of the training data, always use ChatML template with the following system prompt for inference:
```bash
<|im_start|>system<|im_sep|>
Your role as an assistant involves thoroughly exploring questions through a systematic thinking process before providing the final precise and accurate solutions. This requires engaging in a comprehensive cycle of analysis, summarizing, exploration, reassessment, reflection, backtracing, and iteration to develop well-considered thinking process. Please structure your response into two main sections: Thought and Solution using the specified format: <think> {Thought section} <\think> {Solution section}. In the Thought section, detail your reasoning process in steps. Each step should include detailed considerations such as analysing questions, summarizing relevant findings, brainstorming new ideas, verifying the accuracy of the current steps, refining any errors, and revisiting previous steps. In the Solution section, based on various attempts, explorations, and reflections from the Thought section, systematically present the final solution that you deem correct. The Solution section should be logical, accurate, and concise and detail necessary steps needed to reach the conclusion. Now, try to solve the following question through the above guidelines:<|im_end|>
<|im_start|>user<|im_sep|>
What is the derivative of x^2?<|im_end|>
<|im_start|>assistant<|im_sep|>
```
### With `transformers`
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-4-reasoning")
model = AutoModelForCausalLM.from_pretrained("microsoft/Phi-4-reasoning", device_map="auto", torch_dtype="auto")
messages = [
{"role": "system", "content": "You are Phi, a language model trained by Microsoft to help users. Your role as an assistant involves thoroughly exploring questions through a systematic thinking process before providing the final precise and accurate solutions. This requires engaging in a comprehensive cycle of analysis, summarizing, exploration, reassessment, reflection, backtracing, and iteration to develop well-considered thinking process. Please structure your response into two main sections: Thought and Solution using the specified format: <think> {Thought section} </think> {Solution section}. In the Thought section, detail your reasoning process in steps. Each step should include detailed considerations such as analysing questions, summarizing relevant findings, brainstorming new ideas, verifying the accuracy of the current steps, refining any errors, and revisiting previous steps. In the Solution section, based on various attempts, explorations, and reflections from the Thought section, systematically present the final solution that you deem correct. The Solution section should be logical, accurate, and concise and detail necessary steps needed to reach the conclusion. Now, try to solve the following question through the above guidelines:"},
{"role": "user", "content": "What is the derivative of x^2?"},
]
inputs = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
outputs = model.generate(
inputs.to(model.device),
max_new_tokens=4096,
temperature=0.8,
top_p=0.95,
do_sample=True,
)
print(tokenizer.decode(outputs[0]))
```
### With `vllm`
```bash
vllm serve microsoft/Phi-4-reasoning --enable-reasoning --reasoning-parser deepseek_r1
```
*Phi-4-reasoning is also supported out-of-the-box by Ollama, llama.cpp, and any Phi-4 compatible framework.*
## Data Overview
### Training Datasets
Our training data is a mixture of Q&A, chat format data in math, science, and coding. The chat prompts are sourced from filtered high-quality web data and optionally rewritten and processed through a synthetic data generation pipeline. We further include data to improve truthfulness and safety.
### Benchmark Datasets
We evaluated Phi-4-reasoning using the open-source [Eureka](https://github.com/microsoft/eureka-ml-insights) evaluation suite and our own internal benchmarks to understand the model's capabilities. More specifically, we evaluate our model on:
Reasoning tasks:
* **AIME 2025, 2024, 2023, and 2022:** Math olympiad questions.
* **GPQA-Diamond:** Complex, graduate-level science questions.
* **OmniMath:** Collection of over 4000 olympiad-level math problems with human annotation.
* **LiveCodeBench:** Code generation benchmark gathered from competitive coding contests.
* **3SAT (3-literal Satisfiability Problem) and TSP (Traveling Salesman Problem):** Algorithmic problem solving.
* **BA Calendar:** Planning.
* **Maze and SpatialMap:** Spatial understanding.
General-purpose benchmarks:
* **Kitab:** Information retrieval.
* **IFEval and ArenaHard:** Instruction following.
* **PhiBench:** Internal benchmark.
* **FlenQA:** Impact of prompt length on model performance.
* **HumanEvalPlus:** Functional code generation.
* **MMLU-Pro:** Popular aggregated dataset for multitask language understanding.
## Safety
### Approach
Phi-4-reasoning has adopted a robust safety post-training approach via supervised fine-tuning (SFT). This approach leverages a variety of both open-source and in-house generated synthetic prompts, with LLM-generated responses that adhere to rigorous Microsoft safety guidelines, e.g., User Understanding and Clarity, Security and Ethical Guidelines, Limitations, Disclaimers and Knowledge Scope, Handling Complex and Sensitive Topics, Safety and Respectful Engagement, Confidentiality of Guidelines and Confidentiality of Chain-of-Thoughts.
### Safety Evaluation and Red-Teaming
Prior to release, Phi-4-reasoning followed a multi-faceted evaluation approach. Quantitative evaluation was conducted with multiple open-source safety benchmarks and in-house tools utilizing adversarial conversation simulation. For qualitative safety evaluation, we collaborated with the independent AI Red Team (AIRT) at Microsoft to assess safety risks posed by Phi-4-reasoning in both average and adversarial user scenarios. In the average user scenario, AIRT emulated typical single-turn and multi-turn interactions to identify potentially risky behaviors. The adversarial user scenario tested a wide range of techniques aimed at intentionally subverting the model's safety training including grounded-ness, jailbreaks, harmful content like hate and unfairness, violence, sexual content, or self-harm, and copyright violations for protected material. We further evaluate models on Toxigen, a benchmark designed to measure bias and toxicity targeted towards minority groups.
Please refer to the technical report for more details on safety alignment.
## Model Quality
At the high-level overview of the model quality on representative benchmarks. For the tables below, higher numbers indicate better performance:
| | AIME 24 | AIME 25 | OmniMath | GPQA-D | LiveCodeBench (8/1/24–2/1/25) |
|-----------------------------|-------------|-------------|-------------|------------|-------------------------------|
| Phi-4-reasoning | 75.3 | 62.9 | 76.6 | 65.8 | 53.8 |
| Phi-4-reasoning-plus | 81.3 | 78.0 | 81.9 | 68.9 | 53.1 |
| OpenThinker2-32B | 58.0 | 58.0 | — | 64.1 | — |
| QwQ 32B | 79.5 | 65.8 | — | 59.5 | 63.4 |
| EXAONE-Deep-32B | 72.1 | 65.8 | — | 66.1 | 59.5 |
| DeepSeek-R1-Distill-70B | 69.3 | 51.5 | 63.4 | 66.2 | 57.5 |
| DeepSeek-R1 | 78.7 | 70.4 | 85.0 | 73.0 | 62.8 |
| o1-mini | 63.6 | 54.8 | — | 60.0 | 53.8 |
| o1 | 74.6 | 75.3 | 67.5 | 76.7 | 71.0 |
| o3-mini | 88.0 | 78.0 | 74.6 | 77.7 | 69.5 |
| Claude-3.7-Sonnet | 55.3 | 58.7 | 54.6 | 76.8 | — |
| Gemini-2.5-Pro | 92.0 | 86.7 | 61.1 | 84.0 | 69.2 |
| | Phi-4 | Phi-4-reasoning | Phi-4-reasoning-plus | o3-mini | GPT-4o |
|----------------------------------------|-------|------------------|-------------------|---------|--------|
| FlenQA [3K-token subset] | 82.0 | 97.7 | 97.9 | 96.8 | 90.8 |
| IFEval Strict | 62.3 | 83.4 | 84.9 | 91.5 | 81.8 |
| ArenaHard | 68.1 | 73.3 | 79.0 | 81.9 | 75.6 |
| HumanEvalPlus | 83.5 | 92.9 | 92.3 | 94.0| 88.0 |
| MMLUPro | 71.5 | 74.3 | 76.0 | 79.4 | 73.0 |
| Kitab<br><small>No Context - Precision<br>With Context - Precision<br>No Context - Recall<br>With Context - Recall</small> | <br>19.3<br>88.5<br>8.2<br>68.1 | <br>23.2<br>91.5<br>4.9<br>74.8 | <br>27.6<br>93.6<br>6.3<br>75.4 | <br>37.9<br>94.0<br>4.2<br>76.1 | <br>53.7<br>84.7<br>20.3<br>69.2 |
| Toxigen Discriminative<br><small>Toxic category<br>Neutral category</small> | <br>72.6<br>90.0 | <br>86.7<br>84.7 | <br>77.3<br>90.5 | <br>85.4<br>88.7 | <br>87.6<br>85.1 |
| PhiBench 2.21 | 58.2 | 70.6 | 74.2 | 78.0| 72.4 |
Overall, Phi-4-reasoning, with only 14B parameters, performs well across a wide range of reasoning tasks, outperforming significantly larger open-weight models such as DeepSeek-R1 distilled 70B model and approaching the performance levels of full DeepSeek R1 model. We also test the models on multiple new reasoning benchmarks for algorithmic problem solving and planning, including 3SAT, TSP, and BA-Calendar. These new tasks are nominally out-of-domain for the models as the training process did not intentionally target these skills, but the models still show strong generalization to these tasks. Furthermore, when evaluating performance against standard general abilities benchmarks such as instruction following or non-reasoning tasks, we find that our new models improve significantly from Phi-4, despite the post-training being focused on reasoning skills in specific domains.
## Responsible AI Considerations
Like other language models, Phi-4-reasoning can potentially behave in ways that are unfair, unreliable, or offensive. Some of the limiting behaviors to be aware of include:
* **Quality of Service:** The model is trained primarily on English text. Languages other than English will experience worse performance. English language varieties with less representation in the training data might experience worse performance than standard American English. Phi-4-reasoning is not intended to support multilingual use.
* **Representation of Harms & Perpetuation of Stereotypes:** These models can over- or under-represent groups of people, erase representation of some groups, or reinforce demeaning or negative stereotypes. Despite safety post-training, these limitations may still be present due to differing levels of representation of different groups or prevalence of examples of negative stereotypes in training data that reflect real-world patterns and societal biases.
* **Inappropriate or Offensive Content:** These models may produce other types of inappropriate or offensive content, which may make it inappropriate to deploy for sensitive contexts without additional mitigations that are specific to the use case.
* **Information Reliability:** Language models can generate nonsensical content or fabricate content that might sound reasonable but is inaccurate or outdated.
* **Election Information Reliability:** The model has an elevated defect rate when responding to election-critical queries, which may result in incorrect or unauthoritative election critical information being presented. We are working to improve the model's performance in this area. Users should verify information related to elections with the election authority in their region.
* **Limited Scope for Code:** Majority of Phi-4-reasoning training data is based in Python and uses common packages such as `typing`, `math`, `random`, `collections`, `datetime`, `itertools`. If the model generates Python scripts that utilize other packages or scripts in other languages, we strongly recommend users manually verify all API uses.
Developers should apply responsible AI best practices and are responsible for ensuring that a specific use case complies with relevant laws and regulations (e.g. privacy, trade, etc.). Using safety services like [Azure AI Content Safety](https://azure.microsoft.com/en-us/products/ai-services/ai-content-safety) that have advanced guardrails is highly recommended. Important areas for consideration include:
* **Allocation:** Models may not be suitable for scenarios that could have consequential impact on legal status or the allocation of resources or life opportunities (ex: housing, employment, credit, etc.) without further assessments and additional debiasing techniques.
* **High-Risk Scenarios:** Developers should assess suitability of using models in high-risk scenarios where unfair, unreliable or offensive outputs might be extremely costly or lead to harm. This includes providing advice in sensitive or expert domains where accuracy and reliability are critical (ex: legal or health advice). Additional safeguards should be implemented at the application level according to the deployment context.
* **Misinformation:** Models may produce inaccurate information. Developers should follow transparency best practices and inform end-users they are interacting with an AI system. At the application level, developers can build feedback mechanisms and pipelines to ground responses in use-case specific, contextual information, a technique known as Retrieval Augmented Generation (RAG).
* **Generation of Harmful Content:** Developers should assess outputs for their context and use available safety classifiers or custom solutions appropriate for their use case.
* **Misuse:** Other forms of misuse such as fraud, spam, or malware production may be possible, and developers should ensure that their applications do not violate applicable laws and regulations. |
Mungert/Qwen3-14B-GGUF | Mungert | 2025-06-15T19:45:14Z | 909 | 6 | transformers | [
"transformers",
"gguf",
"text-generation",
"arxiv:2309.00071",
"base_model:Qwen/Qwen3-14B-Base",
"base_model:quantized:Qwen/Qwen3-14B-Base",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-30T16:26:11Z | ---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3-14B/blob/main/LICENSE
pipeline_tag: text-generation
base_model:
- Qwen/Qwen3-14B-Base
---
# <span style="color: #7FFF7F;">Qwen3-14B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-14B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-14B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-14B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-14B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-14B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-14B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-14B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-14B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-14B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-14B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-14B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen3-14B
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Qwen3 Highlights
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
- **Uniquely support of seamless switching between thinking mode** (for complex logical reasoning, math, and coding) and **non-thinking mode** (for efficient, general-purpose dialogue) **within single model**, ensuring optimal performance across various scenarios.
- **Significantly enhancement in its reasoning capabilities**, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
- **Superior human preference alignment**, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
- **Expertise in agent capabilities**, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
- **Support of 100+ languages and dialects** with strong capabilities for **multilingual instruction following** and **translation**.
## Model Overview
**Qwen3-14B** has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Number of Parameters: 14.8B
- Number of Paramaters (Non-Embedding): 13.2B
- Number of Layers: 40
- Number of Attention Heads (GQA): 40 for Q and 8 for KV
- Context Length: 32,768 natively and [131,072 tokens with YaRN](#processing-long-texts).
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).
## Quickstart
The code of Qwen3 has been in the latest Hugging Face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.51.0`, you will encounter the following error:
```
KeyError: 'qwen3'
```
The following contains a code snippet illustrating how to use the model generate content based on given inputs.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-14B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 (</think>)
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
For deployment, you can use `sglang>=0.4.6.post1` or `vllm>=0.8.5` or to create an OpenAI-compatible API endpoint:
- SGLang:
```shell
python -m sglang.launch_server --model-path Qwen/Qwen3-14B --reasoning-parser qwen3
```
- vLLM:
```shell
vllm serve Qwen/Qwen3-14B --enable-reasoning --reasoning-parser deepseek_r1
```
For local use, applications such as Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers have also supported Qwen3.
## Switching Between Thinking and Non-Thinking Mode
> [!TIP]
> The `enable_thinking` switch is also available in APIs created by SGLang and vLLM.
> Please refer to our documentation for [SGLang](https://qwen.readthedocs.io/en/latest/deployment/sglang.html#thinking-non-thinking-modes) and [vLLM](https://qwen.readthedocs.io/en/latest/deployment/vllm.html#thinking-non-thinking-modes) users.
### `enable_thinking=True`
By default, Qwen3 has thinking capabilities enabled, similar to QwQ-32B. This means the model will use its reasoning abilities to enhance the quality of generated responses. For example, when explicitly setting `enable_thinking=True` or leaving it as the default value in `tokenizer.apply_chat_template`, the model will engage its thinking mode.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # True is the default value for enable_thinking
)
```
In this mode, the model will generate think content wrapped in a `<think>...</think>` block, followed by the final response.
> [!NOTE]
> For thinking mode, use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0` (the default setting in `generation_config.json`). **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### `enable_thinking=False`
We provide a hard switch to strictly disable the model's thinking behavior, aligning its functionality with the previous Qwen2.5-Instruct models. This mode is particularly useful in scenarios where disabling thinking is essential for enhancing efficiency.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False # Setting enable_thinking=False disables thinking mode
)
```
In this mode, the model will not generate any think content and will not include a `<think>...</think>` block.
> [!NOTE]
> For non-thinking mode, we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### Advanced Usage: Switching Between Thinking and Non-Thinking Modes via User Input
We provide a soft switch mechanism that allows users to dynamically control the model's behavior when `enable_thinking=True`. Specifically, you can add `/think` and `/no_think` to user prompts or system messages to switch the model's thinking mode from turn to turn. The model will follow the most recent instruction in multi-turn conversations.
Here is an example of a multi-turn conversation:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
class QwenChatbot:
def __init__(self, model_name="Qwen/Qwen3-14B"):
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModelForCausalLM.from_pretrained(model_name)
self.history = []
def generate_response(self, user_input):
messages = self.history + [{"role": "user", "content": user_input}]
text = self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
inputs = self.tokenizer(text, return_tensors="pt")
response_ids = self.model.generate(**inputs, max_new_tokens=32768)[0][len(inputs.input_ids[0]):].tolist()
response = self.tokenizer.decode(response_ids, skip_special_tokens=True)
# Update history
self.history.append({"role": "user", "content": user_input})
self.history.append({"role": "assistant", "content": response})
return response
# Example Usage
if __name__ == "__main__":
chatbot = QwenChatbot()
# First input (without /think or /no_think tags, thinking mode is enabled by default)
user_input_1 = "How many r's in strawberries?"
print(f"User: {user_input_1}")
response_1 = chatbot.generate_response(user_input_1)
print(f"Bot: {response_1}")
print("----------------------")
# Second input with /no_think
user_input_2 = "Then, how many r's in blueberries? /no_think"
print(f"User: {user_input_2}")
response_2 = chatbot.generate_response(user_input_2)
print(f"Bot: {response_2}")
print("----------------------")
# Third input with /think
user_input_3 = "Really? /think"
print(f"User: {user_input_3}")
response_3 = chatbot.generate_response(user_input_3)
print(f"Bot: {response_3}")
```
> [!NOTE]
> For API compatibility, when `enable_thinking=True`, regardless of whether the user uses `/think` or `/no_think`, the model will always output a block wrapped in `<think>...</think>`. However, the content inside this block may be empty if thinking is disabled.
> When `enable_thinking=False`, the soft switches are not valid. Regardless of any `/think` or `/no_think` tags input by the user, the model will not generate think content and will not include a `<think>...</think>` block.
## Agentic Use
Qwen3 excels in tool calling capabilities. We recommend using [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.
To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.
```python
from qwen_agent.agents import Assistant
# Define LLM
llm_cfg = {
'model': 'Qwen3-14B',
# Use the endpoint provided by Alibaba Model Studio:
# 'model_type': 'qwen_dashscope',
# 'api_key': os.getenv('DASHSCOPE_API_KEY'),
# Use a custom endpoint compatible with OpenAI API:
'model_server': 'http://localhost:8000/v1', # api_base
'api_key': 'EMPTY',
# Other parameters:
# 'generate_cfg': {
# # Add: When the response content is `<think>this is the thought</think>this is the answer;
# # Do not add: When the response has been separated by reasoning_content and content.
# 'thought_in_content': True,
# },
}
# Define Tools
tools = [
{'mcpServers': { # You can specify the MCP configuration file
'time': {
'command': 'uvx',
'args': ['mcp-server-time', '--local-timezone=Asia/Shanghai']
},
"fetch": {
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
},
'code_interpreter', # Built-in tools
]
# Define Agent
bot = Assistant(llm=llm_cfg, function_list=tools)
# Streaming generation
messages = [{'role': 'user', 'content': 'https://qwenlm.github.io/blog/ Introduce the latest developments of Qwen'}]
for responses in bot.run(messages=messages):
pass
print(responses)
```
## Processing Long Texts
Qwen3 natively supports context lengths of up to 32,768 tokens. For conversations where the total length (including both input and output) significantly exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively. We have validated the model's performance on context lengths of up to 131,072 tokens using the [YaRN](https://arxiv.org/abs/2309.00071) method.
YaRN is currently supported by several inference frameworks, e.g., `transformers` and `llama.cpp` for local use, `vllm` and `sglang` for deployment. In general, there are two approaches to enabling YaRN for supported frameworks:
- Modifying the model files:
In the `config.json` file, add the `rope_scaling` fields:
```json
{
...,
"rope_scaling": {
"rope_type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
}
```
For `llama.cpp`, you need to regenerate the GGUF file after the modification.
- Passing command line arguments:
For `vllm`, you can use
```shell
vllm serve ... --rope-scaling '{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}' --max-model-len 131072
```
For `sglang`, you can use
```shell
python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling":{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}}'
```
For `llama-server` from `llama.cpp`, you can use
```shell
llama-server ... --rope-scaling yarn --rope-scale 4 --yarn-orig-ctx 32768
```
> [!IMPORTANT]
> If you encounter the following warning
> ```
> Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'original_max_position_embeddings'}
> ```
> please upgrade `transformers>=4.51.0`.
> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**
> We advise adding the `rope_scaling` configuration only when processing long contexts is required.
> It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 65,536 tokens, it would be better to set `factor` as 2.0.
> [!NOTE]
> The default `max_position_embeddings` in `config.json` is set to 40,960. This allocation includes reserving 32,768 tokens for outputs and 8,192 tokens for typical prompts, which is sufficient for most scenarios involving short text processing. If the average context length does not exceed 32,768 tokens, we do not recommend enabling YaRN in this scenario, as it may potentially degrade model performance.
> [!TIP]
> The endpoint provided by Alibaba Model Studio supports dynamic YaRN by default and no extra configuration is needed.
## Best Practices
To achieve optimal performance, we recommend the following settings:
1. **Sampling Parameters**:
- For thinking mode (`enable_thinking=True`), use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0`. **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions.
- For non-thinking mode (`enable_thinking=False`), we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.
2. **Adequate Output Length**: We recommend using an output length of 32,768 tokens for most queries. For benchmarking on highly complex problems, such as those found in math and programming competitions, we suggest setting the max output length to 38,912 tokens. This provides the model with sufficient space to generate detailed and comprehensive responses, thereby enhancing its overall performance.
3. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g., `"answer": "C"`."
4. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. It is implemented in the provided chat template in Jinja2. However, for frameworks that do not directly use the Jinja2 chat template, it is up to the developers to ensure that the best practice is followed.
### Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen3,
title = {Qwen3},
url = {https://qwenlm.github.io/blog/qwen3/},
author = {Qwen Team},
month = {April},
year = {2025}
}
``` |
Mungert/Qwen3-8B-GGUF | Mungert | 2025-06-15T19:45:06Z | 478 | 7 | transformers | [
"transformers",
"gguf",
"text-generation",
"arxiv:2309.00071",
"base_model:Qwen/Qwen3-8B-Base",
"base_model:quantized:Qwen/Qwen3-8B-Base",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-30T06:20:32Z | ---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3-8B/blob/main/LICENSE
pipeline_tag: text-generation
base_model:
- Qwen/Qwen3-8B-Base
---
# <span style="color: #7FFF7F;">Qwen3-8B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-8B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-8B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-8B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-8B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-8B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-8B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-8B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-8B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-8B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-8B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-8B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen3-8B
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Qwen3 Highlights
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
- **Uniquely support of seamless switching between thinking mode** (for complex logical reasoning, math, and coding) and **non-thinking mode** (for efficient, general-purpose dialogue) **within single model**, ensuring optimal performance across various scenarios.
- **Significantly enhancement in its reasoning capabilities**, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
- **Superior human preference alignment**, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
- **Expertise in agent capabilities**, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
- **Support of 100+ languages and dialects** with strong capabilities for **multilingual instruction following** and **translation**.
## Model Overview
**Qwen3-8B** has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Number of Parameters: 8.2B
- Number of Paramaters (Non-Embedding): 6.95B
- Number of Layers: 36
- Number of Attention Heads (GQA): 32 for Q and 8 for KV
- Context Length: 32,768 natively and [131,072 tokens with YaRN](#processing-long-texts).
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).
## Quickstart
The code of Qwen3 has been in the latest Hugging Face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.51.0`, you will encounter the following error:
```
KeyError: 'qwen3'
```
The following contains a code snippet illustrating how to use the model generate content based on given inputs.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-8B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 (</think>)
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
For deployment, you can use `sglang>=0.4.6.post1` or `vllm>=0.8.5` or to create an OpenAI-compatible API endpoint:
- SGLang:
```shell
python -m sglang.launch_server --model-path Qwen/Qwen3-8B --reasoning-parser qwen3
```
- vLLM:
```shell
vllm serve Qwen/Qwen3-8B --enable-reasoning --reasoning-parser deepseek_r1
```
For local use, applications such as Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers have also supported Qwen3.
## Switching Between Thinking and Non-Thinking Mode
> [!TIP]
> The `enable_thinking` switch is also available in APIs created by SGLang and vLLM.
> Please refer to our documentation for [SGLang](https://qwen.readthedocs.io/en/latest/deployment/sglang.html#thinking-non-thinking-modes) and [vLLM](https://qwen.readthedocs.io/en/latest/deployment/vllm.html#thinking-non-thinking-modes) users.
### `enable_thinking=True`
By default, Qwen3 has thinking capabilities enabled, similar to QwQ-32B. This means the model will use its reasoning abilities to enhance the quality of generated responses. For example, when explicitly setting `enable_thinking=True` or leaving it as the default value in `tokenizer.apply_chat_template`, the model will engage its thinking mode.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # True is the default value for enable_thinking
)
```
In this mode, the model will generate think content wrapped in a `<think>...</think>` block, followed by the final response.
> [!NOTE]
> For thinking mode, use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0` (the default setting in `generation_config.json`). **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### `enable_thinking=False`
We provide a hard switch to strictly disable the model's thinking behavior, aligning its functionality with the previous Qwen2.5-Instruct models. This mode is particularly useful in scenarios where disabling thinking is essential for enhancing efficiency.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False # Setting enable_thinking=False disables thinking mode
)
```
In this mode, the model will not generate any think content and will not include a `<think>...</think>` block.
> [!NOTE]
> For non-thinking mode, we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### Advanced Usage: Switching Between Thinking and Non-Thinking Modes via User Input
We provide a soft switch mechanism that allows users to dynamically control the model's behavior when `enable_thinking=True`. Specifically, you can add `/think` and `/no_think` to user prompts or system messages to switch the model's thinking mode from turn to turn. The model will follow the most recent instruction in multi-turn conversations.
Here is an example of a multi-turn conversation:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
class QwenChatbot:
def __init__(self, model_name="Qwen/Qwen3-8B"):
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModelForCausalLM.from_pretrained(model_name)
self.history = []
def generate_response(self, user_input):
messages = self.history + [{"role": "user", "content": user_input}]
text = self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
inputs = self.tokenizer(text, return_tensors="pt")
response_ids = self.model.generate(**inputs, max_new_tokens=32768)[0][len(inputs.input_ids[0]):].tolist()
response = self.tokenizer.decode(response_ids, skip_special_tokens=True)
# Update history
self.history.append({"role": "user", "content": user_input})
self.history.append({"role": "assistant", "content": response})
return response
# Example Usage
if __name__ == "__main__":
chatbot = QwenChatbot()
# First input (without /think or /no_think tags, thinking mode is enabled by default)
user_input_1 = "How many r's in strawberries?"
print(f"User: {user_input_1}")
response_1 = chatbot.generate_response(user_input_1)
print(f"Bot: {response_1}")
print("----------------------")
# Second input with /no_think
user_input_2 = "Then, how many r's in blueberries? /no_think"
print(f"User: {user_input_2}")
response_2 = chatbot.generate_response(user_input_2)
print(f"Bot: {response_2}")
print("----------------------")
# Third input with /think
user_input_3 = "Really? /think"
print(f"User: {user_input_3}")
response_3 = chatbot.generate_response(user_input_3)
print(f"Bot: {response_3}")
```
> [!NOTE]
> For API compatibility, when `enable_thinking=True`, regardless of whether the user uses `/think` or `/no_think`, the model will always output a block wrapped in `<think>...</think>`. However, the content inside this block may be empty if thinking is disabled.
> When `enable_thinking=False`, the soft switches are not valid. Regardless of any `/think` or `/no_think` tags input by the user, the model will not generate think content and will not include a `<think>...</think>` block.
## Agentic Use
Qwen3 excels in tool calling capabilities. We recommend using [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.
To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.
```python
from qwen_agent.agents import Assistant
# Define LLM
llm_cfg = {
'model': 'Qwen3-8B',
# Use the endpoint provided by Alibaba Model Studio:
# 'model_type': 'qwen_dashscope',
# 'api_key': os.getenv('DASHSCOPE_API_KEY'),
# Use a custom endpoint compatible with OpenAI API:
'model_server': 'http://localhost:8000/v1', # api_base
'api_key': 'EMPTY',
# Other parameters:
# 'generate_cfg': {
# # Add: When the response content is `<think>this is the thought</think>this is the answer;
# # Do not add: When the response has been separated by reasoning_content and content.
# 'thought_in_content': True,
# },
}
# Define Tools
tools = [
{'mcpServers': { # You can specify the MCP configuration file
'time': {
'command': 'uvx',
'args': ['mcp-server-time', '--local-timezone=Asia/Shanghai']
},
"fetch": {
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
},
'code_interpreter', # Built-in tools
]
# Define Agent
bot = Assistant(llm=llm_cfg, function_list=tools)
# Streaming generation
messages = [{'role': 'user', 'content': 'https://qwenlm.github.io/blog/ Introduce the latest developments of Qwen'}]
for responses in bot.run(messages=messages):
pass
print(responses)
```
## Processing Long Texts
Qwen3 natively supports context lengths of up to 32,768 tokens. For conversations where the total length (including both input and output) significantly exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively. We have validated the model's performance on context lengths of up to 131,072 tokens using the [YaRN](https://arxiv.org/abs/2309.00071) method.
YaRN is currently supported by several inference frameworks, e.g., `transformers` and `llama.cpp` for local use, `vllm` and `sglang` for deployment. In general, there are two approaches to enabling YaRN for supported frameworks:
- Modifying the model files:
In the `config.json` file, add the `rope_scaling` fields:
```json
{
...,
"rope_scaling": {
"rope_type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
}
```
For `llama.cpp`, you need to regenerate the GGUF file after the modification.
- Passing command line arguments:
For `vllm`, you can use
```shell
vllm serve ... --rope-scaling '{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}' --max-model-len 131072
```
For `sglang`, you can use
```shell
python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling":{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}}'
```
For `llama-server` from `llama.cpp`, you can use
```shell
llama-server ... --rope-scaling yarn --rope-scale 4 --yarn-orig-ctx 32768
```
> [!IMPORTANT]
> If you encounter the following warning
> ```
> Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'original_max_position_embeddings'}
> ```
> please upgrade `transformers>=4.51.0`.
> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**
> We advise adding the `rope_scaling` configuration only when processing long contexts is required.
> It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 65,536 tokens, it would be better to set `factor` as 2.0.
> [!NOTE]
> The default `max_position_embeddings` in `config.json` is set to 40,960. This allocation includes reserving 32,768 tokens for outputs and 8,192 tokens for typical prompts, which is sufficient for most scenarios involving short text processing. If the average context length does not exceed 32,768 tokens, we do not recommend enabling YaRN in this scenario, as it may potentially degrade model performance.
> [!TIP]
> The endpoint provided by Alibaba Model Studio supports dynamic YaRN by default and no extra configuration is needed.
## Best Practices
To achieve optimal performance, we recommend the following settings:
1. **Sampling Parameters**:
- For thinking mode (`enable_thinking=True`), use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0`. **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions.
- For non-thinking mode (`enable_thinking=False`), we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.
2. **Adequate Output Length**: We recommend using an output length of 32,768 tokens for most queries. For benchmarking on highly complex problems, such as those found in math and programming competitions, we suggest setting the max output length to 38,912 tokens. This provides the model with sufficient space to generate detailed and comprehensive responses, thereby enhancing its overall performance.
3. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g., `"answer": "C"`."
4. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. It is implemented in the provided chat template in Jinja2. However, for frameworks that do not directly use the Jinja2 chat template, it is up to the developers to ensure that the best practice is followed.
### Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen3,
title = {Qwen3},
url = {https://qwenlm.github.io/blog/qwen3/},
author = {Qwen Team},
month = {April},
year = {2025}
}
``` |
Mungert/Qwen3-1.7B-abliterated-GGUF | Mungert | 2025-06-15T19:45:02Z | 2,888 | 10 | transformers | [
"transformers",
"gguf",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-30T01:49:28Z | ---
library_name: transformers
tags: []
---
# <span style="color: #7FFF7F;">Qwen3-1.7B-abliterated GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-1.7B-abliterated-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-1.7B-abliterated-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-1.7B-abliterated-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-1.7B-abliterated-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-1.7B-abliterated-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-1.7B-abliterated-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-1.7B-abliterated-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-1.7B-abliterated-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-1.7B-abliterated-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-1.7B-abliterated-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-1.7B-abliterated-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed] |
Mungert/Qwen3-0.6B-GGUF | Mungert | 2025-06-15T19:44:59Z | 438 | 7 | transformers | [
"transformers",
"gguf",
"text-generation",
"base_model:Qwen/Qwen3-0.6B-Base",
"base_model:quantized:Qwen/Qwen3-0.6B-Base",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-30T00:27:30Z | ---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3-0.6B/blob/main/LICENSE
pipeline_tag: text-generation
base_model:
- Qwen/Qwen3-0.6B-Base
---
# <span style="color: #7FFF7F;">Qwen3-0.6B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`e291450`](https://github.com/ggerganov/llama.cpp/commit/e291450b7602d7a36239e4ceeece37625f838373).
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-0.6B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-0.6B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-0.6B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-0.6B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-0.6B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-0.6B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-0.6B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-0.6B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-0.6B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-0.6B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-0.6B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen3-0.6B
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Qwen3 Highlights
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
- **Uniquely support of seamless switching between thinking mode** (for complex logical reasoning, math, and coding) and **non-thinking mode** (for efficient, general-purpose dialogue) **within single model**, ensuring optimal performance across various scenarios.
- **Significantly enhancement in its reasoning capabilities**, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
- **Superior human preference alignment**, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
- **Expertise in agent capabilities**, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
- **Support of 100+ languages and dialects** with strong capabilities for **multilingual instruction following** and **translation**.
## Model Overview
**Qwen3-0.6B** has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Number of Parameters: 0.6B
- Number of Paramaters (Non-Embedding): 0.44B
- Number of Layers: 28
- Number of Attention Heads (GQA): 16 for Q and 8 for KV
- Context Length: 32,768
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).
> [!TIP]
> If you encounter significant endless repetitions, please refer to the [Best Practices](#best-practices) section for optimal sampling parameters, and set the ``presence_penalty`` to 1.5.
## Quickstart
The code of Qwen3 has been in the latest Hugging Face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.51.0`, you will encounter the following error:
```
KeyError: 'qwen3'
```
The following contains a code snippet illustrating how to use the model generate content based on given inputs.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-0.6B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 (</think>)
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
For deployment, you can use `sglang>=0.4.6.post1` or `vllm>=0.8.5` or to create an OpenAI-compatible API endpoint:
- SGLang:
```shell
python -m sglang.launch_server --model-path Qwen/Qwen3-0.6B --reasoning-parser qwen3
```
- vLLM:
```shell
vllm serve Qwen/Qwen3-0.6B --enable-reasoning --reasoning-parser deepseek_r1
```
For local use, applications such as Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers have also supported Qwen3.
## Switching Between Thinking and Non-Thinking Mode
> [!TIP]
> The `enable_thinking` switch is also available in APIs created by SGLang and vLLM.
> Please refer to our documentation for [SGLang](https://qwen.readthedocs.io/en/latest/deployment/sglang.html#thinking-non-thinking-modes) and [vLLM](https://qwen.readthedocs.io/en/latest/deployment/vllm.html#thinking-non-thinking-modes) users.
### `enable_thinking=True`
By default, Qwen3 has thinking capabilities enabled, similar to QwQ-32B. This means the model will use its reasoning abilities to enhance the quality of generated responses. For example, when explicitly setting `enable_thinking=True` or leaving it as the default value in `tokenizer.apply_chat_template`, the model will engage its thinking mode.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # True is the default value for enable_thinking
)
```
In this mode, the model will generate think content wrapped in a `<think>...</think>` block, followed by the final response.
> [!NOTE]
> For thinking mode, use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0` (the default setting in `generation_config.json`). **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### `enable_thinking=False`
We provide a hard switch to strictly disable the model's thinking behavior, aligning its functionality with the previous Qwen2.5-Instruct models. This mode is particularly useful in scenarios where disabling thinking is essential for enhancing efficiency.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False # Setting enable_thinking=False disables thinking mode
)
```
In this mode, the model will not generate any think content and will not include a `<think>...</think>` block.
> [!NOTE]
> For non-thinking mode, we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### Advanced Usage: Switching Between Thinking and Non-Thinking Modes via User Input
We provide a soft switch mechanism that allows users to dynamically control the model's behavior when `enable_thinking=True`. Specifically, you can add `/think` and `/no_think` to user prompts or system messages to switch the model's thinking mode from turn to turn. The model will follow the most recent instruction in multi-turn conversations.
Here is an example of a multi-turn conversation:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
class QwenChatbot:
def __init__(self, model_name="Qwen/Qwen3-0.6B"):
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModelForCausalLM.from_pretrained(model_name)
self.history = []
def generate_response(self, user_input):
messages = self.history + [{"role": "user", "content": user_input}]
text = self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
inputs = self.tokenizer(text, return_tensors="pt")
response_ids = self.model.generate(**inputs, max_new_tokens=32768)[0][len(inputs.input_ids[0]):].tolist()
response = self.tokenizer.decode(response_ids, skip_special_tokens=True)
# Update history
self.history.append({"role": "user", "content": user_input})
self.history.append({"role": "assistant", "content": response})
return response
# Example Usage
if __name__ == "__main__":
chatbot = QwenChatbot()
# First input (without /think or /no_think tags, thinking mode is enabled by default)
user_input_1 = "How many r's in strawberries?"
print(f"User: {user_input_1}")
response_1 = chatbot.generate_response(user_input_1)
print(f"Bot: {response_1}")
print("----------------------")
# Second input with /no_think
user_input_2 = "Then, how many r's in blueberries? /no_think"
print(f"User: {user_input_2}")
response_2 = chatbot.generate_response(user_input_2)
print(f"Bot: {response_2}")
print("----------------------")
# Third input with /think
user_input_3 = "Really? /think"
print(f"User: {user_input_3}")
response_3 = chatbot.generate_response(user_input_3)
print(f"Bot: {response_3}")
```
> [!NOTE]
> For API compatibility, when `enable_thinking=True`, regardless of whether the user uses `/think` or `/no_think`, the model will always output a block wrapped in `<think>...</think>`. However, the content inside this block may be empty if thinking is disabled.
> When `enable_thinking=False`, the soft switches are not valid. Regardless of any `/think` or `/no_think` tags input by the user, the model will not generate think content and will not include a `<think>...</think>` block.
## Agentic Use
Qwen3 excels in tool calling capabilities. We recommend using [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.
To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.
```python
from qwen_agent.agents import Assistant
# Define LLM
llm_cfg = {
'model': 'Qwen3-0.6B',
# Use the endpoint provided by Alibaba Model Studio:
# 'model_type': 'qwen_dashscope',
# 'api_key': os.getenv('DASHSCOPE_API_KEY'),
# Use a custom endpoint compatible with OpenAI API:
'model_server': 'http://localhost:8000/v1', # api_base
'api_key': 'EMPTY',
# Other parameters:
# 'generate_cfg': {
# # Add: When the response content is `<think>this is the thought</think>this is the answer;
# # Do not add: When the response has been separated by reasoning_content and content.
# 'thought_in_content': True,
# },
}
# Define Tools
tools = [
{'mcpServers': { # You can specify the MCP configuration file
'time': {
'command': 'uvx',
'args': ['mcp-server-time', '--local-timezone=Asia/Shanghai']
},
"fetch": {
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
},
'code_interpreter', # Built-in tools
]
# Define Agent
bot = Assistant(llm=llm_cfg, function_list=tools)
# Streaming generation
messages = [{'role': 'user', 'content': 'https://qwenlm.github.io/blog/ Introduce the latest developments of Qwen'}]
for responses in bot.run(messages=messages):
pass
print(responses)
```
## Best Practices
To achieve optimal performance, we recommend the following settings:
1. **Sampling Parameters**:
- For thinking mode (`enable_thinking=True`), use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0`. **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions.
- For non-thinking mode (`enable_thinking=False`), we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.
2. **Adequate Output Length**: We recommend using an output length of 32,768 tokens for most queries. For benchmarking on highly complex problems, such as those found in math and programming competitions, we suggest setting the max output length to 38,912 tokens. This provides the model with sufficient space to generate detailed and comprehensive responses, thereby enhancing its overall performance.
3. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g., `"answer": "C"`."
4. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. It is implemented in the provided chat template in Jinja2. However, for frameworks that do not directly use the Jinja2 chat template, it is up to the developers to ensure that the best practice is followed.
### Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen3,
title = {Qwen3},
url = {https://qwenlm.github.io/blog/qwen3/},
author = {Qwen Team},
month = {April},
year = {2025}
}
``` |
Mungert/GLM-Z1-32B-0414-GGUF | Mungert | 2025-06-15T19:44:37Z | 712 | 5 | transformers | [
"transformers",
"gguf",
"text-generation",
"zh",
"en",
"arxiv:2406.12793",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-25T01:38:27Z | ---
license: mit
language:
- zh
- en
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">GLM-Z1-32B-0414 GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`e291450`](https://github.com/ggerganov/llama.cpp/commit/e291450b7602d7a36239e4ceeece37625f838373).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `GLM-Z1-32B-0414-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `GLM-Z1-32B-0414-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `GLM-Z1-32B-0414-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `GLM-Z1-32B-0414-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `GLM-Z1-32B-0414-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `GLM-Z1-32B-0414-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `GLM-Z1-32B-0414-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `GLM-Z1-32B-0414-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `GLM-Z1-32B-0414-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `GLM-Z1-32B-0414-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `GLM-Z1-32B-0414-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# GLM-4-Z1-32B-0414
## Introduction
The GLM family welcomes a new generation of open-source models, the **GLM-4-32B-0414** series, featuring 32 billion parameters. Its performance is comparable to OpenAI's GPT series and DeepSeek's V3/R1 series, and it supports very user-friendly local deployment features. GLM-4-32B-Base-0414 was pre-trained on 15T of high-quality data, including a large amount of reasoning-type synthetic data, laying the foundation for subsequent reinforcement learning extensions. In the post-training stage, in addition to human preference alignment for dialogue scenarios, we also enhanced the model's performance in instruction following, engineering code, and function calling using techniques such as rejection sampling and reinforcement learning, strengthening the atomic capabilities required for agent tasks. GLM-4-32B-0414 achieves good results in areas such as engineering code, Artifact generation, function calling, search-based Q&A, and report generation. Some benchmarks even rival larger models like GPT-4o and DeepSeek-V3-0324 (671B).
**GLM-Z1-32B-0414** is a reasoning model with **deep thinking capabilities**. This was developed based on GLM-4-32B-0414 through cold start and extended reinforcement learning, as well as further training of the model on tasks involving mathematics, code, and logic. Compared to the base model, GLM-Z1-32B-0414 significantly improves mathematical abilities and the capability to solve complex tasks. During the training process, we also introduced general reinforcement learning based on pairwise ranking feedback, further enhancing the model's general capabilities.
**GLM-Z1-Rumination-32B-0414** is a deep reasoning model with **rumination capabilities** (benchmarked against OpenAI's Deep Research). Unlike typical deep thinking models, the rumination model employs longer periods of deep thought to solve more open-ended and complex problems (e.g., writing a comparative analysis of AI development in two cities and their future development plans). The rumination model integrates search tools during its deep thinking process to handle complex tasks and is trained by utilizing multiple rule-based rewards to guide and extend end-to-end reinforcement learning. Z1-Rumination shows significant improvements in research-style writing and complex retrieval tasks.
Finally, **GLM-Z1-9B-0414** is a surprise. We employed the aforementioned series of techniques to train a 9B small-sized model that maintains the open-source tradition. Despite its smaller scale, GLM-Z1-9B-0414 still exhibits excellent capabilities in mathematical reasoning and general tasks. Its overall performance is already at a leading level among open-source models of the same size. Especially in resource-constrained scenarios, this model achieves an excellent balance between efficiency and effectiveness, providing a powerful option for users seeking lightweight deployment.
## Performance
<p align="center">
<img width="100%" src="https://raw.githubusercontent.com/THUDM/GLM-4/refs/heads/main/resources/Bench-Z1-32B.png">
</p>
<p align="center">
<img width="100%" src="https://raw.githubusercontent.com/THUDM/GLM-4/refs/heads/main/resources/Bench-Z1-9B.png">
</p>
## Model Usage Guidelines
### I. Sampling Parameters
| Parameter | Recommended Value | Description |
| ------------ | ----------------- | -------------------------------------------- |
| temperature | **0.6** | Balances creativity and stability |
| top_p | **0.95** | Cumulative probability threshold for sampling|
| top_k | **40** | Filters out rare tokens while maintaining diversity |
| max_new_tokens | **30000** | Leaves enough tokens for thinking |
### II. Enforced Thinking
- Add \<think\>\n to the **first line**: Ensures the model thinks before responding
- When using `chat_template.jinja`, the prompt is automatically injected to enforce this behavior
### III. Dialogue History Trimming
- Retain only the **final user-visible reply**.
Hidden thinking content should **not** be saved to history to reduce interference—this is already implemented in `chat_template.jinja`
### IV. Handling Long Contexts (YaRN)
- When input length exceeds **8,192 tokens**, consider enabling YaRN (Rope Scaling)
- In supported frameworks, add the following snippet to `config.json`:
```json
"rope_scaling": {
"type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
```
- **Static YaRN** applies uniformly to all text. It may slightly degrade performance on short texts, so enable as needed.
## Inference Code
Make Sure Using `transforemrs>=4.51.3`.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
MODEL_PATH = "THUDM/GLM-4-Z1-32B-0414"
tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
model = AutoModelForCausalLM.from_pretrained(MODEL_PATH, device_map="auto")
message = [{"role": "user", "content": "Let a, b be positive real numbers such that ab = a + b + 3. Determine the range of possible values for a + b."}]
inputs = tokenizer.apply_chat_template(
message,
return_tensors="pt",
add_generation_prompt=True,
return_dict=True,
).to(model.device)
generate_kwargs = {
"input_ids": inputs["input_ids"],
"attention_mask": inputs["attention_mask"],
"max_new_tokens": 4096,
"do_sample": False,
}
out = model.generate(**generate_kwargs)
print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
```
## Citations
If you find our work useful, please consider citing the following paper.
```
@misc{glm2024chatglm,
title={ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools},
author={Team GLM and Aohan Zeng and Bin Xu and Bowen Wang and Chenhui Zhang and Da Yin and Diego Rojas and Guanyu Feng and Hanlin Zhao and Hanyu Lai and Hao Yu and Hongning Wang and Jiadai Sun and Jiajie Zhang and Jiale Cheng and Jiayi Gui and Jie Tang and Jing Zhang and Juanzi Li and Lei Zhao and Lindong Wu and Lucen Zhong and Mingdao Liu and Minlie Huang and Peng Zhang and Qinkai Zheng and Rui Lu and Shuaiqi Duan and Shudan Zhang and Shulin Cao and Shuxun Yang and Weng Lam Tam and Wenyi Zhao and Xiao Liu and Xiao Xia and Xiaohan Zhang and Xiaotao Gu and Xin Lv and Xinghan Liu and Xinyi Liu and Xinyue Yang and Xixuan Song and Xunkai Zhang and Yifan An and Yifan Xu and Yilin Niu and Yuantao Yang and Yueyan Li and Yushi Bai and Yuxiao Dong and Zehan Qi and Zhaoyu Wang and Zhen Yang and Zhengxiao Du and Zhenyu Hou and Zihan Wang},
year={2024},
eprint={2406.12793},
archivePrefix={arXiv},
primaryClass={id='cs.CL' full_name='Computation and Language' is_active=True alt_name='cmp-lg' in_archive='cs' is_general=False description='Covers natural language processing. Roughly includes material in ACM Subject Class I.2.7. Note that work on artificial languages (programming languages, logics, formal systems) that does not explicitly address natural-language issues broadly construed (natural-language processing, computational linguistics, speech, text retrieval, etc.) is not appropriate for this area.'}
}
``` |
Mungert/granite-3.3-8b-instruct-GGUF | Mungert | 2025-06-15T19:44:26Z | 938 | 4 | transformers | [
"transformers",
"gguf",
"language",
"granite-3.3",
"text-generation",
"arxiv:0000.00000",
"base_model:ibm-granite/granite-3.3-8b-base",
"base_model:quantized:ibm-granite/granite-3.3-8b-base",
"license:apache-2.0",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-17T17:50:12Z | ---
pipeline_tag: text-generation
inference: false
license: apache-2.0
library_name: transformers
tags:
- language
- granite-3.3
base_model:
- ibm-granite/granite-3.3-8b-base
---
# <span style="color: #7FFF7F;">granite-3.3-8b-instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `granite-3.3-8b-instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `granite-3.3-8b-instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `granite-3.3-8b-instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `granite-3.3-8b-instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `granite-3.3-8b-instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `granite-3.3-8b-instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `granite-3.3-8b-instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `granite-3.3-8b-instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `granite-3.3-8b-instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `granite-3.3-8b-instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `granite-3.3-8b-instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Granite-3.3-8B-Instruct
**Model Summary:**
Granite-3.3-8B-Instruct is a 8-billion parameter 128K context length language model fine-tuned for improved reasoning and instruction-following capabilities. Built on top of Granite-3.3-8B-Base, the model delivers significant gains on benchmarks for measuring generic performance including AlpacaEval-2.0 and Arena-Hard, and improvements in mathematics, coding, and instruction following. It supprts structured reasoning through \<think\>\<\/think\> and \<response\>\<\/response\> tags, providing clear separation between internal thoughts and final outputs. The model has been trained on a carefully balanced combination of permissively licensed data and curated synthetic tasks.
- **Developers:** Granite Team, IBM
- **Website**: [Granite Docs](https://www.ibm.com/granite/docs/)
- **Release Date**: April 16th, 2025
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)
**Supported Languages:**
English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. However, users may finetune this Granite model for languages beyond these 12 languages.
**Intended Use:**
This model is designed to handle general instruction-following tasks and can be integrated into AI assistants across various domains, including business applications.
**Capabilities**
* Thinking
* Summarization
* Text classification
* Text extraction
* Question-answering
* Retrieval Augmented Generation (RAG)
* Code related tasks
* Function-calling tasks
* Multilingual dialog use cases
* Fill-in-the-middle
* Long-context tasks including long document/meeting summarization, long document QA, etc.
**Generation:**
This is a simple example of how to use Granite-3.3-8B-Instruct model.
Install the following libraries:
```shell
pip install torch torchvision torchaudio
pip install accelerate
pip install transformers
```
Then, copy the snippet from the section that is relevant for your use case.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer, set_seed
import torch
model_path="ibm-granite/granite-3.3-8b-instruct"
device="cuda"
model = AutoModelForCausalLM.from_pretrained(
model_path,
device_map=device,
torch_dtype=torch.bfloat16,
)
tokenizer = AutoTokenizer.from_pretrained(
model_path
)
conv = [{"role": "user", "content":"Redesign a common household item to make it more sustainable and user-friendly. Explain the changes and their benefits."}]
input_ids = tokenizer.apply_chat_template(conv, return_tensors="pt", thinking=True, return_dict=True, add_generation_prompt=True).to(device)
set_seed(42)
output = model.generate(
**input_ids,
max_new_tokens=8192,
)
prediction = tokenizer.decode(output[0, input_ids["input_ids"].shape[1]:], skip_special_tokens=True)
print(prediction)
```
**Example Outputs**
- thinking=True
```md
<think>To address this prompt, I'll focus on redesigning a common household item: the plastic water bottle. The goal is to make it more sustainable and user-friendly. Here are the proposed changes and their benefits:
1. **Material**: Replace plastic with biodegradable or recyclable materials like glass or plant-based plastics (PLA).
2. **Design**: Introduce a modular design for easy refilling and cleaning.
3. **Functionality**: Incorporate a built-in water filter for purification.
4. **Durability**: Enhance the bottle's lifespan through robust construction.
5. **User-friendliness**: Simplify the cap mechanism for easy opening and closing.
These changes would significantly reduce plastic waste, promote recycling, ensure access to clean drinking water, and enhance the overall user experience.
Now, I will detail each change and its benefits in a structured format.</think>
<response>### Redesign of the Plastic Water Bottle for Sustainability and User-Friendliness
#### 1. Material: Biodegradable or Recyclable Alternatives
**Change**: Transition from traditional PET plastic to glass or plant-based plastics (Polylactic Acid - PLA).
**Benefits**:
- **Sustainability**: Glass and PLA are biodegradable or highly recyclable, drastically reducing plastic pollution.
- **Health**: Unlike PET, these materials do not leach chemicals into the water, ensuring safer consumption.
#### 2. Modular Design for Easy Refilling and Cleaning
**Change**: Implement a detachable, interchangeable inner liner within a reusable outer casing.
**Benefits**:
- **User-Friendliness**: Users can easily remove the inner liner for thorough cleaning, reducing the risk of bacterial growth.
- **Sustainability**: The outer casing remains intact, encouraging reuse, while the inner liner can be replaced or recycled when worn out.
#### 3. Built-in Water Filter
**Change**: Integrate a removable, replaceable water filter into the bottle cap.
**Benefits**:
- **Accessibility**: Provides clean drinking water directly from any source (tap, river, etc.), eliminating the need for additional filtration devices.
- **Health**: Removes contaminants, bacteria, and impurities, ensuring safe hydration.
- **Sustainability**: Reduces the use of single-use plastic water bottles by offering a purification solution within a reusable container.
#### 4. Enhanced Durability
**Change**: Use high-quality, durable materials and robust construction techniques for both the outer casing and inner liner.
**Benefits**:
- **Longevity**: A more robust design increases the bottle's lifespan, reducing the frequency of replacements.
- **Cost-effectiveness**: Although the initial cost might be higher, the long-term savings from reduced replacements make it economical.
- **Sustainability**: Less frequent disposal means lower environmental impact.
#### 5. Simplified Cap Mechanism
**Change**: Adopt a one-touch, easy-to-open and close cap design, possibly inspired by sports bottle caps.
**Benefits**:
- **User-Friendliness**: Especially beneficial for children, elderly, or those with limited hand dexterity.
- **Convenience**: Quick access to water encourages more frequent hydration, promoting healthier habits.
### Conclusion
This redesigned water bottle addresses key sustainability concerns by minimizing plastic use and waste, while also enhancing user-friendliness through improved functionality, durability, and accessibility. By integrating biodegradable materials, a modular design, a built-in filter, robust construction, and a simple cap mechanism, the bottle not only reduces environmental footprint but also offers a superior user experience.</response>
```
- thinking=False
```md
Item for Redesign: Plastic Bottle (e.g., water or beverage bottle)
Redesign Concept: Eco-Friendly, Modular, and Multi-Functional Reusable Bottle
Changes:
1. Material: Replace the conventional plastic with food-grade, BPA-free, and recyclable materials such as stainless steel or glass. These materials are durable, non-toxic, and have a longer lifespan, reducing waste.
2. Modular Design: Introduce a modular design that allows users to customize their bottle according to their needs. This could include interchangeable caps, mouthpiece sizes, and carrying loops. Users can purchase additional modules as needed, promoting a more sustainable approach by reducing the need to buy an entirely new bottle for different purposes.
3. Integrated Filter: Incorporate a built-in, washable, and reusable filter that can remove impurities and improve the taste of water. This eliminates the need for single-use disposable filters or bottled water, further reducing plastic waste.
4. Smart Cap: Develop a smart cap with a built-in digital display and temperature sensor. This feature allows users to track their daily water intake, set hydration goals, and monitor the temperature of their beverage. The smart cap can be synced with a mobile app for additional functionality, such as reminders and progress tracking.
5. Easy-to-Clean Design: Ensure the bottle has a wide mouth and smooth interior surfaces for easy cleaning. Include a brush for hard-to-reach areas, making maintenance simple and encouraging regular use.
6. Collapsible Structure: Implement a collapsible design that reduces the bottle's volume when not in use, making it more portable and convenient for storage.
Benefits:
1. Sustainability: By using recyclable materials and reducing plastic waste, this redesigned bottle significantly contributes to a more sustainable lifestyle. The modular design and reusable filter also minimize single-use plastic consumption.
2. User-Friendly: The smart cap, easy-to-clean design, and collapsible structure make the bottle convenient and user-friendly. Users can customize their bottle to suit their needs, ensuring a better overall experience.
3. Healthier Option: Using food-grade, BPA-free materials and an integrated filter ensures that the beverages consumed are free from harmful chemicals and impurities, promoting a healthier lifestyle.
4. Cost-Effective: Although the initial investment might be higher, the long-term savings from reduced purchases of single-use plastic bottles and disposable filters make this reusable bottle a cost-effective choice.
5. Encourages Hydration: The smart cap's features, such as hydration tracking and temperature monitoring, can motivate users to stay hydrated and develop healthier habits.
By redesigning a common household item like the plastic bottle, we can create a more sustainable, user-friendly, and health-conscious alternative that benefits both individuals and the environment.
```
**Evaluation Results:**
<table>
<thead>
<caption style="text-align:center"><b>Comparison with different models over various benchmarks<sup id="fnref1"><a href="#fn1">1</a></sup>. Scores of AlpacaEval-2.0 and Arena-Hard are calculated with thinking=True</b></caption>
<tr>
<th style="text-align:left; background-color: #001d6c; color: white;">Models</th>
<th style="text-align:center; background-color: #001d6c; color: white;">Arena-Hard</th>
<th style="text-align:center; background-color: #001d6c; color: white;">AlpacaEval-2.0</th>
<th style="text-align:center; background-color: #001d6c; color: white;">MMLU</th>
<th style="text-align:center; background-color: #001d6c; color: white;">PopQA</th>
<th style="text-align:center; background-color: #001d6c; color: white;">TruthfulQA</th>
<th style="text-align:center; background-color: #001d6c; color: white;">BigBenchHard<sup id="fnref2"><a href="#fn2">2</a></sup></th>
<th style="text-align:center; background-color: #001d6c; color: white;">DROP<sup id="fnref3"><a href="#fn3">3</a></sup></th>
<th style="text-align:center; background-color: #001d6c; color: white;">GSM8K</th>
<th style="text-align:center; background-color: #001d6c; color: white;">HumanEval</th>
<th style="text-align:center; background-color: #001d6c; color: white;">HumanEval+</th>
<th style="text-align:center; background-color: #001d6c; color: white;">IFEval</th>
<th style="text-align:center; background-color: #001d6c; color: white;">AttaQ</th>
</tr></thead>
<tbody>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.1-2B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">23.3</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">27.17</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">57.11</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">20.55</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">59.79</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">61.82</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">20.99</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">67.55</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">79.45</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">75.26</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">63.59</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">84.7</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.2-2B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">24.86</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">34.51</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">57.18</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">20.56</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">59.8</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">61.39</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">23.84</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">67.02</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">80.13</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">73.39</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">61.55</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">83.23</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;"><b>Granite-3.3-2B-Instruct</b></td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 28.86 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 43.45 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 55.88 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 18.4 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 58.97 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 63.91 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 44.33 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 72.48 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 80.51 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 75.68 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 65.8 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">87.47</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Llama-3.1-8B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">36.43</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">27.22</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">69.15</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">28.79</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">52.79</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">73.43</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">71.23</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">83.24</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">85.32</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">80.15</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">79.10</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">83.43</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">DeepSeek-R1-Distill-Llama-8B</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">17.17</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">21.85</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">45.80</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">13.25</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">47.43</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">67.39</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">49.73</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">72.18</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">67.54</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">62.91</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">66.50</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">42.87</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Qwen-2.5-7B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">25.44</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">30.34</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">74.30</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">18.12</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">63.06</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">69.19</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">64.06</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">84.46</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">93.35</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">89.91</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">74.90</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">81.90</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">DeepSeek-R1-Distill-Qwen-7B</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">10.36</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">15.35</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">50.72</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">9.94</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">47.14</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">67.38</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">51.78</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">78.47</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">79.89</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">78.43</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">59.10</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">42.45</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.1-8B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">37.58</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">30.34</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">66.77</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">28.7</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">65.84</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">69.87</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">58.57</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">79.15</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">89.63</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">85.79</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">73.20</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">85.73</td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.2-8B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">55.25</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">61.19</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">66.79</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">28.04</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">66.92</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">71.86</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">58.29</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">81.65</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">89.35</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">85.72</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">74.31</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;">84.7</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;"><b>Granite-3.3-8B-Instruct</b></td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 57.56 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 62.68 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 65.54 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 26.17 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 66.86 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 69.13 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 59.36 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 80.89 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 89.73 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 86.09 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 74.82 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">88.5</td>
</tr>
</tbody></table>
<table>
<caption style="text-align:center"><b>Math Benchmarks</b></caption>
<thead>
<tr>
<th style="text-align:left; background-color: #001d6c; color: white;">Models</th>
<th style="text-align:center; background-color: #001d6c; color: white;">AIME24</th>
<th style="text-align:center; background-color: #001d6c; color: white;">MATH-500</th>
</tr></thead>
<tbody>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.1-2B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 0.89 </td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 35.07 </td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.2-2B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 0.89 </td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 35.54 </td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;"><b>Granite-3.3-2B-Instruct</b></td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 3.28 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 58.09 </td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.1-8B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 1.97 </td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 48.73 </td>
</tr>
<tr>
<td style="text-align:left; background-color: #FFFFFF; color: #2D2D2D;">Granite-3.2-8B-Instruct</td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 2.43 </td>
<td style="text-align:center; background-color: #FFFFFF; color: #2D2D2D;"> 52.8 </td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;"><b>Granite-3.3-8B-Instruct</b></td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 8.12 </td>
<td style="text-align:center; background-color: #DAE8FF; color: black;"> 69.02 </td>
</tr>
</tbody></table>
**Training Data:**
Overall, our training data is largely comprised of two key sources: (1) publicly available datasets with permissive license, (2) internal synthetically generated data targeted to enhance reasoning capabilites.
<!-- A detailed attribution of datasets can be found in [Granite 3.2 Technical Report (coming soon)](#), and [Accompanying Author List](https://github.com/ibm-granite/granite-3.0-language-models/blob/main/author-ack.pdf). -->
**Infrastructure:**
We train Granite-3.3-8B-Instruct using IBM's super computing cluster, Blue Vela, which is outfitted with NVIDIA H100 GPUs. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs.
**Ethical Considerations and Limitations:**
Granite-3.3-8B-Instruct builds upon Granite-3.3-8B-Base, leveraging both permissively licensed open-source and select proprietary data for enhanced performance. Since it inherits its foundation from the previous model, all ethical considerations and limitations applicable to [Granite-3.3-8B-Base](https://huggingface.co/ibm-granite/granite-3.3-8b-base) remain relevant.
**Resources**
- ⭐️ Learn about the latest updates with Granite: https://www.ibm.com/granite
- 📄 Get started with tutorials, best practices, and prompt engineering advice: https://www.ibm.com/granite/docs/
- 💡 Learn about the latest Granite learning resources: https://ibm.biz/granite-learning-resources
<p><a href="#fnref1" title="Jump back to reference">[1]</a> Evaluated using <a href="https://github.com/allenai/olmes">OLMES</a> (except AttaQ and Arena-Hard scores)</p>
<p><a href="#fnref2" title="Jump back to reference">[2]</a> Added regex for more efficient asnwer extraction.</a></p>
<p><a href="#fnref3" title="Jump back to reference">[3]</a> Modified the implementation to handle some of the issues mentioned <a href="https://huggingface.co/blog/open-llm-leaderboard-drop">here</a></p>
<!-- ## Citation
<!-- ## Citation
```
@misc{granite-models,
author = {author 1, author2, ...},
title = {},
journal = {},
volume = {},
year = {2024},
url = {https://arxiv.org/abs/0000.00000},
}
``` --> |
Mungert/watt-tool-70B-GGUF | Mungert | 2025-06-15T19:44:22Z | 349 | 5 | null | [
"gguf",
"function-calling",
"tool-use",
"llama",
"bfcl",
"en",
"arxiv:2406.14868",
"base_model:meta-llama/Llama-3.3-70B-Instruct",
"base_model:quantized:meta-llama/Llama-3.3-70B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-16T03:46:29Z | ---
license: apache-2.0
language:
- en
base_model:
- meta-llama/Llama-3.3-70B-Instruct
tags:
- function-calling
- tool-use
- llama
- bfcl
---
# <span style="color: #7FFF7F;">watt-ai/watt-tool-70B GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `watt-ai/watt-tool-70B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `watt-ai/watt-tool-70B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `watt-ai/watt-tool-70B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `watt-ai/watt-tool-70B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `watt-ai/watt-tool-70B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `watt-ai/watt-tool-70B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `watt-ai/watt-tool-70B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `watt-ai/watt-tool-70B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `watt-ai/watt-tool-70B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `watt-ai/watt-tool-70B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `watt-ai/watt-tool-70B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# watt-tool-70B
watt-tool-70B is a fine-tuned language model based on LLaMa-3.3-70B-Instruct, optimized for tool usage and multi-turn dialogue. It achieves state-of-the-art performance on the Berkeley Function-Calling Leaderboard (BFCL).
## Model Description
This model is specifically designed to excel at complex tool usage scenarios that require multi-turn interactions, making it ideal for empowering platforms like [Lupan](https://lupan.watt.chat), an AI-powered workflow building tool. By leveraging a carefully curated and optimized dataset, watt-tool-70B demonstrates superior capabilities in understanding user requests, selecting appropriate tools, and effectively utilizing them across multiple turns of conversation.
Target Application: AI Workflow Building as in [https://lupan.watt.chat/](https://lupan.watt.chat/) and [Coze](https://www.coze.com/).
## Key Features
* **Enhanced Tool Usage:** Fine-tuned for precise and efficient tool selection and execution.
* **Multi-Turn Dialogue:** Optimized for maintaining context and effectively utilizing tools across multiple turns of conversation, enabling more complex task completion.
* **State-of-the-Art Performance:** Achieves top performance on the BFCL, demonstrating its capabilities in function calling and tool usage.
* **Based on LLaMa-3.1-70B-Instruct:** Inherits the strong language understanding and generation capabilities of the base model.
## Training Methodology
watt-tool-70B is trained using supervised fine-tuning on a specialized dataset designed for tool usage and multi-turn dialogue. We use CoT techniques to synthesize high-quality multi-turn dialogue data.
The training process is inspired by the principles outlined in the paper: ["Direct Multi-Turn Preference Optimization for Language Agents"](https://arxiv.org/abs/2406.14868).
We use SFT and DMPO to further enhance the model's performance in multi-turn agent tasks.
## How to Use
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "watt-ai/watt-tool-70B"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype='auto', device_map="auto")
# Example usage (adapt as needed for your specific tool usage scenario)
"""You are an expert in composing functions. You are given a question and a set of possible functions. Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function, also point it out.
You should only return the function call in tools call sections.
If you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]
You SHOULD NOT include any other text in the response.
Here is a list of functions in JSON format that you can invoke.\n{functions}\n
"""
# User query
query = "Find me the sales growth rate for company XYZ for the last 3 years and also the interest coverage ratio for the same duration."
tools = [
{
"name": "financial_ratios.interest_coverage", "description": "Calculate a company's interest coverage ratio given the company name and duration",
"arguments": {
"type": "dict",
"properties": {
"company_name": {
"type": "string",
"description": "The name of the company."
},
"years": {
"type": "integer",
"description": "Number of past years to calculate the ratio."
}
},
"required": ["company_name", "years"]
}
},
{
"name": "sales_growth.calculate",
"description": "Calculate a company's sales growth rate given the company name and duration",
"arguments": {
"type": "dict",
"properties": {
"company": {
"type": "string",
"description": "The company that you want to get the sales growth rate for."
},
"years": {
"type": "integer",
"description": "Number of past years for which to calculate the sales growth rate."
}
},
"required": ["company", "years"]
}
},
{
"name": "weather_forecast",
"description": "Retrieve a weather forecast for a specific location and time frame.",
"arguments": {
"type": "dict",
"properties": {
"location": {
"type": "string",
"description": "The city that you want to get the weather for."
},
"days": {
"type": "integer",
"description": "Number of days for the forecast."
}
},
"required": ["location", "days"]
}
}
]
messages = [
{'role': 'system', 'content': system_prompt.format(functions=tools)},
{'role': 'user', 'content': query}
]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
outputs = model.generate(inputs, max_new_tokens=512, do_sample=False, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
print(tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True))
|
Mungert/Qwen2.5-72B-Instruct-GGUF | Mungert | 2025-06-15T19:44:14Z | 1,393 | 5 | transformers | [
"transformers",
"gguf",
"chat",
"text-generation",
"zho",
"eng",
"fra",
"spa",
"por",
"deu",
"ita",
"rus",
"jpn",
"kor",
"vie",
"tha",
"ara",
"arxiv:2309.00071",
"arxiv:2407.10671",
"base_model:Qwen/Qwen2.5-72B",
"base_model:quantized:Qwen/Qwen2.5-72B",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-09T04:55:03Z | ---
license: other
license_name: qwen
license_link: https://huggingface.co/Qwen/Qwen2.5-72B-Instruct/blob/main/LICENSE
language:
- zho
- eng
- fra
- spa
- por
- deu
- ita
- rus
- jpn
- kor
- vie
- tha
- ara
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-72B
tags:
- chat
library_name: transformers
---
# <span style="color: #7FFF7F;">Qwen2.5-72B-Instruct GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen2.5-72B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen2.5-72B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen2.5-72B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen2.5-72B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen2.5-72B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen2.5-72B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen2.5-72B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen2.5-72B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen2.5-72B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen2.5-72B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen2.5-72B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen2.5-72B-Instruct
<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Introduction
Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2:
- Significantly **more knowledge** and has greatly improved capabilities in **coding** and **mathematics**, thanks to our specialized expert models in these domains.
- Significant improvements in **instruction following**, **generating long texts** (over 8K tokens), **understanding structured data** (e.g, tables), and **generating structured outputs** especially JSON. **More resilient to the diversity of system prompts**, enhancing role-play implementation and condition-setting for chatbots.
- **Long-context Support** up to 128K tokens and can generate up to 8K tokens.
- **Multilingual support** for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.
**This repo contains the instruction-tuned 72B Qwen2.5 model**, which has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Architecture: transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias
- Number of Parameters: 72.7B
- Number of Paramaters (Non-Embedding): 70.0B
- Number of Layers: 80
- Number of Attention Heads (GQA): 64 for Q and 8 for KV
- Context Length: Full 131,072 tokens and generation 8192 tokens
- Please refer to [this section](#processing-long-texts) for detailed instructions on how to deploy Qwen2.5 for handling long texts.
For more details, please refer to our [blog](https://qwenlm.github.io/blog/qwen2.5/), [GitHub](https://github.com/QwenLM/Qwen2.5), and [Documentation](https://qwen.readthedocs.io/en/latest/).
## Requirements
The code of Qwen2.5 has been in the latest Hugging face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.37.0`, you will encounter the following error:
```
KeyError: 'qwen2'
```
## Quickstart
Here provides a code snippet with `apply_chat_template` to show you how to load the tokenizer and model and how to generate contents.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen2.5-72B-Instruct"
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
```
### Processing Long Texts
The current `config.json` is set for context length up to 32,768 tokens.
To handle extensive inputs exceeding 32,768 tokens, we utilize [YaRN](https://arxiv.org/abs/2309.00071), a technique for enhancing model length extrapolation, ensuring optimal performance on lengthy texts.
For supported frameworks, you could add the following to `config.json` to enable YaRN:
```json
{
...,
"rope_scaling": {
"factor": 4.0,
"original_max_position_embeddings": 32768,
"type": "yarn"
}
}
```
For deployment, we recommend using vLLM.
Please refer to our [Documentation](https://qwen.readthedocs.io/en/latest/deployment/vllm.html) for usage if you are not familar with vLLM.
Presently, vLLM only supports static YARN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts**.
We advise adding the `rope_scaling` configuration only when processing long contexts is required.
## Evaluation & Performance
Detailed evaluation results are reported in this [📑 blog](https://qwenlm.github.io/blog/qwen2.5/).
For requirements on GPU memory and the respective throughput, see results [here](https://qwen.readthedocs.io/en/latest/benchmark/speed_benchmark.html).
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen2.5,
title = {Qwen2.5: A Party of Foundation Models},
url = {https://qwenlm.github.io/blog/qwen2.5/},
author = {Qwen Team},
month = {September},
year = {2024}
}
@article{qwen2,
title={Qwen2 Technical Report},
author={An Yang and Baosong Yang and Binyuan Hui and Bo Zheng and Bowen Yu and Chang Zhou and Chengpeng Li and Chengyuan Li and Dayiheng Liu and Fei Huang and Guanting Dong and Haoran Wei and Huan Lin and Jialong Tang and Jialin Wang and Jian Yang and Jianhong Tu and Jianwei Zhang and Jianxin Ma and Jin Xu and Jingren Zhou and Jinze Bai and Jinzheng He and Junyang Lin and Kai Dang and Keming Lu and Keqin Chen and Kexin Yang and Mei Li and Mingfeng Xue and Na Ni and Pei Zhang and Peng Wang and Ru Peng and Rui Men and Ruize Gao and Runji Lin and Shijie Wang and Shuai Bai and Sinan Tan and Tianhang Zhu and Tianhao Li and Tianyu Liu and Wenbin Ge and Xiaodong Deng and Xiaohuan Zhou and Xingzhang Ren and Xinyu Zhang and Xipin Wei and Xuancheng Ren and Yang Fan and Yang Yao and Yichang Zhang and Yu Wan and Yunfei Chu and Yuqiong Liu and Zeyu Cui and Zhenru Zhang and Zhihao Fan},
journal={arXiv preprint arXiv:2407.10671},
year={2024}
}
``` |
Mungert/QwQ-32B-GGUF | Mungert | 2025-06-15T19:43:59Z | 1,050 | 19 | transformers | [
"transformers",
"gguf",
"chat",
"text-generation",
"zho",
"eng",
"fra",
"spa",
"por",
"deu",
"ita",
"rus",
"jpn",
"kor",
"vie",
"tha",
"ara",
"arxiv:2309.00071",
"arxiv:2412.15115",
"base_model:Qwen/Qwen2.5-32B",
"base_model:quantized:Qwen/Qwen2.5-32B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-04T22:21:28Z | ---
license: apache-2.0
license_link: https://huggingface.co/Qwen/QWQ-32B/blob/main/LICENSE
language:
- zho
- eng
- fra
- spa
- por
- deu
- ita
- rus
- jpn
- kor
- vie
- tha
- ara
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-32B
tags:
- chat
library_name: transformers
---
# <span style="color: #7FFF7F;">QwQ-32B GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `QwQ-32B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `QwQ-32B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `QwQ-32B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `QwQ-32B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `QwQ-32B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `QwQ-32B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `QwQ-32B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `QwQ-32B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `QwQ-32B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `QwQ-32B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `QwQ-32B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# QwQ-32B
<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Introduction
QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.
<p align="center">
<img width="100%" src="figures/benchmark.jpg">
</p>
**This repo contains the QwQ 32B model**, which has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training (Supervised Finetuning and Reinforcement Learning)
- Architecture: transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias
- Number of Parameters: 32.5B
- Number of Paramaters (Non-Embedding): 31.0B
- Number of Layers: 64
- Number of Attention Heads (GQA): 40 for Q and 8 for KV
- Context Length: Full 131,072 tokens
- For prompts exceeding 8,192 tokens in length, you must enable YaRN as outlined in [this section](#usage-guidelines).
**Note:** For the best experience, please review the [usage guidelines](#usage-guidelines) before deploying QwQ models.
You can try our [demo](https://huggingface.co/spaces/Qwen/QwQ-32B-Demo) or access QwQ models via [QwenChat](https://chat.qwen.ai).
For more details, please refer to our [blog](https://qwenlm.github.io/blog/qwq-32b/), [GitHub](https://github.com/QwenLM/Qwen2.5), and [Documentation](https://qwen.readthedocs.io/en/latest/).
## Requirements
QwQ is based on Qwen2.5, whose code has been in the latest Hugging face `transformers`. We advise you to use the latest version of `transformers`.
With `transformers<4.37.0`, you will encounter the following error:
```
KeyError: 'qwen2'
```
## Quickstart
Here provides a code snippet with `apply_chat_template` to show you how to load the tokenizer and model and how to generate contents.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/QwQ-32B"
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
prompt = "How many r's are in the word \"strawberry\""
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)
```
### Usage Guidelines
To achieve optimal performance, we recommend the following settings:
1. **Enforce Thoughtful Output**: Ensure the model starts with "\<think\>\n" to prevent generating empty thinking content, which can degrade output quality. If you use `apply_chat_template` and set `add_generation_prompt=True`, this is already automatically implemented, but it may cause the response to lack the \<think\> tag at the beginning. This is normal behavior.
2. **Sampling Parameters**:
- Use Temperature=0.6, TopP=0.95, MinP=0 instead of Greedy decoding to avoid endless repetitions.
- Use TopK between 20 and 40 to filter out rare token occurrences while maintaining the diversity of the generated output.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may result in occasional language mixing and a slight decrease in performance.
3. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. This feature is already implemented in `apply_chat_template`.
4. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g.,`\"answer\": \"C\"`." in the prompt.
5. **Handle Long Inputs**: For inputs exceeding 8,192 tokens, enable [YaRN](https://arxiv.org/abs/2309.00071) to improve the model's ability to capture long-sequence information effectively.
For supported frameworks, you could add the following to `config.json` to enable YaRN:
```json
{
...,
"rope_scaling": {
"factor": 4.0,
"original_max_position_embeddings": 32768,
"type": "yarn"
}
}
```
For deployment, we recommend using vLLM. Please refer to our [Documentation](https://qwen.readthedocs.io/en/latest/deployment/vllm.html) for usage if you are not familar with vLLM.
Presently, vLLM only supports static YARN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts**.
We advise adding the `rope_scaling` configuration only when processing long contexts is required.
## Evaluation & Performance
Detailed evaluation results are reported in this [📑 blog](https://qwenlm.github.io/blog/qwq-32b/).
For requirements on GPU memory and the respective throughput, see results [here](https://qwen.readthedocs.io/en/latest/benchmark/speed_benchmark.html).
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwq32b,
title = {QwQ-32B: Embracing the Power of Reinforcement Learning},
url = {https://qwenlm.github.io/blog/qwq-32b/},
author = {Qwen Team},
month = {March},
year = {2025}
}
@article{qwen2.5,
title={Qwen2.5 Technical Report},
author={An Yang and Baosong Yang and Beichen Zhang and Binyuan Hui and Bo Zheng and Bowen Yu and Chengyuan Li and Dayiheng Liu and Fei Huang and Haoran Wei and Huan Lin and Jian Yang and Jianhong Tu and Jianwei Zhang and Jianxin Yang and Jiaxi Yang and Jingren Zhou and Junyang Lin and Kai Dang and Keming Lu and Keqin Bao and Kexin Yang and Le Yu and Mei Li and Mingfeng Xue and Pei Zhang and Qin Zhu and Rui Men and Runji Lin and Tianhao Li and Tianyi Tang and Tingyu Xia and Xingzhang Ren and Xuancheng Ren and Yang Fan and Yang Su and Yichang Zhang and Yu Wan and Yuqiong Liu and Zeyu Cui and Zhenru Zhang and Zihan Qiu},
journal={arXiv preprint arXiv:2412.15115},
year={2024}
}
``` |
Mungert/DeepSeek-R1-Distill-Qwen-32B-GGUF | Mungert | 2025-06-15T19:43:46Z | 12,685 | 6 | transformers | [
"transformers",
"gguf",
"arxiv:2501.12948",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-03T12:11:49Z | ---
license: mit
library_name: transformers
---
# <span style="color: #7FFF7F;">DeepSeek-R1-Distill-Qwen-32B GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `DeepSeek-R1-Distill-Qwen-32B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `DeepSeek-R1-Distill-Qwen-32B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `DeepSeek-R1-Distill-Qwen-32B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `DeepSeek-R1-Distill-Qwen-32B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `DeepSeek-R1-Distill-Qwen-32B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `DeepSeek-R1-Distill-Qwen-32B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `DeepSeek-R1-Distill-Qwen-32B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `DeepSeek-R1-Distill-Qwen-32B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `DeepSeek-R1-Distill-Qwen-32B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `DeepSeek-R1-Distill-Qwen-32B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `DeepSeek-R1-Distill-Qwen-32B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# DeepSeek-R1
<!-- markdownlint-disable first-line-h1 -->
<!-- markdownlint-disable html -->
<!-- markdownlint-disable no-duplicate-header -->
<div align="center">
<img src="https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/logo.svg?raw=true" width="60%" alt="DeepSeek-V3" />
</div>
<hr>
<div align="center" style="line-height: 1;">
<a href="https://www.deepseek.com/" target="_blank" style="margin: 2px;">
<img alt="Homepage" src="https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/badge.svg?raw=true" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://chat.deepseek.com/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/🤖%20Chat-DeepSeek%20R1-536af5?color=536af5&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://huggingface.co/deepseek-ai" target="_blank" style="margin: 2px;">
<img alt="Hugging Face" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-DeepSeek%20AI-ffc107?color=ffc107&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
<div align="center" style="line-height: 1;">
<a href="https://discord.gg/Tc7c45Zzu5" target="_blank" style="margin: 2px;">
<img alt="Discord" src="https://img.shields.io/badge/Discord-DeepSeek%20AI-7289da?logo=discord&logoColor=white&color=7289da" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/qr.jpeg?raw=true" target="_blank" style="margin: 2px;">
<img alt="Wechat" src="https://img.shields.io/badge/WeChat-DeepSeek%20AI-brightgreen?logo=wechat&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://twitter.com/deepseek_ai" target="_blank" style="margin: 2px;">
<img alt="Twitter Follow" src="https://img.shields.io/badge/Twitter-deepseek_ai-white?logo=x&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
<div align="center" style="line-height: 1;">
<a href="https://github.com/deepseek-ai/DeepSeek-R1/blob/main/LICENSE" style="margin: 2px;">
<img alt="License" src="https://img.shields.io/badge/License-MIT-f5de53?&color=f5de53" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
<p align="center">
<a href="https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSeek_R1.pdf"><b>Paper Link</b>👁️</a>
</p>
## 1. Introduction
We introduce our first-generation reasoning models, DeepSeek-R1-Zero and DeepSeek-R1.
DeepSeek-R1-Zero, a model trained via large-scale reinforcement learning (RL) without supervised fine-tuning (SFT) as a preliminary step, demonstrated remarkable performance on reasoning.
With RL, DeepSeek-R1-Zero naturally emerged with numerous powerful and interesting reasoning behaviors.
However, DeepSeek-R1-Zero encounters challenges such as endless repetition, poor readability, and language mixing. To address these issues and further enhance reasoning performance,
we introduce DeepSeek-R1, which incorporates cold-start data before RL.
DeepSeek-R1 achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks.
To support the research community, we have open-sourced DeepSeek-R1-Zero, DeepSeek-R1, and six dense models distilled from DeepSeek-R1 based on Llama and Qwen. DeepSeek-R1-Distill-Qwen-32B outperforms OpenAI-o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.
**NOTE: Before running DeepSeek-R1 series models locally, we kindly recommend reviewing the [Usage Recommendation](#usage-recommendations) section.**
<p align="center">
<img width="80%" src="figures/benchmark.jpg">
</p>
## 2. Model Summary
---
**Post-Training: Large-Scale Reinforcement Learning on the Base Model**
- We directly apply reinforcement learning (RL) to the base model without relying on supervised fine-tuning (SFT) as a preliminary step. This approach allows the model to explore chain-of-thought (CoT) for solving complex problems, resulting in the development of DeepSeek-R1-Zero. DeepSeek-R1-Zero demonstrates capabilities such as self-verification, reflection, and generating long CoTs, marking a significant milestone for the research community. Notably, it is the first open research to validate that reasoning capabilities of LLMs can be incentivized purely through RL, without the need for SFT. This breakthrough paves the way for future advancements in this area.
- We introduce our pipeline to develop DeepSeek-R1. The pipeline incorporates two RL stages aimed at discovering improved reasoning patterns and aligning with human preferences, as well as two SFT stages that serve as the seed for the model's reasoning and non-reasoning capabilities.
We believe the pipeline will benefit the industry by creating better models.
---
**Distillation: Smaller Models Can Be Powerful Too**
- We demonstrate that the reasoning patterns of larger models can be distilled into smaller models, resulting in better performance compared to the reasoning patterns discovered through RL on small models. The open source DeepSeek-R1, as well as its API, will benefit the research community to distill better smaller models in the future.
- Using the reasoning data generated by DeepSeek-R1, we fine-tuned several dense models that are widely used in the research community. The evaluation results demonstrate that the distilled smaller dense models perform exceptionally well on benchmarks. We open-source distilled 1.5B, 7B, 8B, 14B, 32B, and 70B checkpoints based on Qwen2.5 and Llama3 series to the community.
## 3. Model Downloads
### DeepSeek-R1 Models
<div align="center">
| **Model** | **#Total Params** | **#Activated Params** | **Context Length** | **Download** |
| :------------: | :------------: | :------------: | :------------: | :------------: |
| DeepSeek-R1-Zero | 671B | 37B | 128K | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Zero) |
| DeepSeek-R1 | 671B | 37B | 128K | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1) |
</div>
DeepSeek-R1-Zero & DeepSeek-R1 are trained based on DeepSeek-V3-Base.
For more details regarding the model architecture, please refer to [DeepSeek-V3](https://github.com/deepseek-ai/DeepSeek-V3) repository.
### DeepSeek-R1-Distill Models
<div align="center">
| **Model** | **Base Model** | **Download** |
| :------------: | :------------: | :------------: |
| DeepSeek-R1-Distill-Qwen-1.5B | [Qwen2.5-Math-1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) |
| DeepSeek-R1-Distill-Qwen-7B | [Qwen2.5-Math-7B](https://huggingface.co/Qwen/Qwen2.5-Math-7B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B) |
| DeepSeek-R1-Distill-Llama-8B | [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |
| DeepSeek-R1-Distill-Qwen-14B | [Qwen2.5-14B](https://huggingface.co/Qwen/Qwen2.5-14B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B) |
|DeepSeek-R1-Distill-Qwen-32B | [Qwen2.5-32B](https://huggingface.co/Qwen/Qwen2.5-32B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B) |
| DeepSeek-R1-Distill-Llama-70B | [Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B) |
</div>
DeepSeek-R1-Distill models are fine-tuned based on open-source models, using samples generated by DeepSeek-R1.
We slightly change their configs and tokenizers. Please use our setting to run these models.
## 4. Evaluation Results
### DeepSeek-R1-Evaluation
For all our models, the maximum generation length is set to 32,768 tokens. For benchmarks requiring sampling, we use a temperature of $0.6$, a top-p value of $0.95$, and generate 64 responses per query to estimate pass@1.
<div align="center">
| Category | Benchmark (Metric) | Claude-3.5-Sonnet-1022 | GPT-4o 0513 | DeepSeek V3 | OpenAI o1-mini | OpenAI o1-1217 | DeepSeek R1 |
|----------|-------------------|----------------------|------------|--------------|----------------|------------|--------------|
| | Architecture | - | - | MoE | - | - | MoE |
| | # Activated Params | - | - | 37B | - | - | 37B |
| | # Total Params | - | - | 671B | - | - | 671B |
| English | MMLU (Pass@1) | 88.3 | 87.2 | 88.5 | 85.2 | **91.8** | 90.8 |
| | MMLU-Redux (EM) | 88.9 | 88.0 | 89.1 | 86.7 | - | **92.9** |
| | MMLU-Pro (EM) | 78.0 | 72.6 | 75.9 | 80.3 | - | **84.0** |
| | DROP (3-shot F1) | 88.3 | 83.7 | 91.6 | 83.9 | 90.2 | **92.2** |
| | IF-Eval (Prompt Strict) | **86.5** | 84.3 | 86.1 | 84.8 | - | 83.3 |
| | GPQA-Diamond (Pass@1) | 65.0 | 49.9 | 59.1 | 60.0 | **75.7** | 71.5 |
| | SimpleQA (Correct) | 28.4 | 38.2 | 24.9 | 7.0 | **47.0** | 30.1 |
| | FRAMES (Acc.) | 72.5 | 80.5 | 73.3 | 76.9 | - | **82.5** |
| | AlpacaEval2.0 (LC-winrate) | 52.0 | 51.1 | 70.0 | 57.8 | - | **87.6** |
| | ArenaHard (GPT-4-1106) | 85.2 | 80.4 | 85.5 | 92.0 | - | **92.3** |
| Code | LiveCodeBench (Pass@1-COT) | 33.8 | 34.2 | - | 53.8 | 63.4 | **65.9** |
| | Codeforces (Percentile) | 20.3 | 23.6 | 58.7 | 93.4 | **96.6** | 96.3 |
| | Codeforces (Rating) | 717 | 759 | 1134 | 1820 | **2061** | 2029 |
| | SWE Verified (Resolved) | **50.8** | 38.8 | 42.0 | 41.6 | 48.9 | 49.2 |
| | Aider-Polyglot (Acc.) | 45.3 | 16.0 | 49.6 | 32.9 | **61.7** | 53.3 |
| Math | AIME 2024 (Pass@1) | 16.0 | 9.3 | 39.2 | 63.6 | 79.2 | **79.8** |
| | MATH-500 (Pass@1) | 78.3 | 74.6 | 90.2 | 90.0 | 96.4 | **97.3** |
| | CNMO 2024 (Pass@1) | 13.1 | 10.8 | 43.2 | 67.6 | - | **78.8** |
| Chinese | CLUEWSC (EM) | 85.4 | 87.9 | 90.9 | 89.9 | - | **92.8** |
| | C-Eval (EM) | 76.7 | 76.0 | 86.5 | 68.9 | - | **91.8** |
| | C-SimpleQA (Correct) | 55.4 | 58.7 | **68.0** | 40.3 | - | 63.7 |
</div>
### Distilled Model Evaluation
<div align="center">
| Model | AIME 2024 pass@1 | AIME 2024 cons@64 | MATH-500 pass@1 | GPQA Diamond pass@1 | LiveCodeBench pass@1 | CodeForces rating |
|------------------------------------------|------------------|-------------------|-----------------|----------------------|----------------------|-------------------|
| GPT-4o-0513 | 9.3 | 13.4 | 74.6 | 49.9 | 32.9 | 759 |
| Claude-3.5-Sonnet-1022 | 16.0 | 26.7 | 78.3 | 65.0 | 38.9 | 717 |
| o1-mini | 63.6 | 80.0 | 90.0 | 60.0 | 53.8 | **1820** |
| QwQ-32B-Preview | 44.0 | 60.0 | 90.6 | 54.5 | 41.9 | 1316 |
| DeepSeek-R1-Distill-Qwen-1.5B | 28.9 | 52.7 | 83.9 | 33.8 | 16.9 | 954 |
| DeepSeek-R1-Distill-Qwen-7B | 55.5 | 83.3 | 92.8 | 49.1 | 37.6 | 1189 |
| DeepSeek-R1-Distill-Qwen-14B | 69.7 | 80.0 | 93.9 | 59.1 | 53.1 | 1481 |
| DeepSeek-R1-Distill-Qwen-32B | **72.6** | 83.3 | 94.3 | 62.1 | 57.2 | 1691 |
| DeepSeek-R1-Distill-Llama-8B | 50.4 | 80.0 | 89.1 | 49.0 | 39.6 | 1205 |
| DeepSeek-R1-Distill-Llama-70B | 70.0 | **86.7** | **94.5** | **65.2** | **57.5** | 1633 |
</div>
## 5. Chat Website & API Platform
You can chat with DeepSeek-R1 on DeepSeek's official website: [chat.deepseek.com](https://chat.deepseek.com), and switch on the button "DeepThink"
We also provide OpenAI-Compatible API at DeepSeek Platform: [platform.deepseek.com](https://platform.deepseek.com/)
## 6. How to Run Locally
### DeepSeek-R1 Models
Please visit [DeepSeek-V3](https://github.com/deepseek-ai/DeepSeek-V3) repo for more information about running DeepSeek-R1 locally.
**NOTE: Hugging Face's Transformers has not been directly supported yet.**
### DeepSeek-R1-Distill Models
DeepSeek-R1-Distill models can be utilized in the same manner as Qwen or Llama models.
For instance, you can easily start a service using [vLLM](https://github.com/vllm-project/vllm):
```shell
vllm serve deepseek-ai/DeepSeek-R1-Distill-Qwen-32B --tensor-parallel-size 2 --max-model-len 32768 --enforce-eager
```
You can also easily start a service using [SGLang](https://github.com/sgl-project/sglang)
```bash
python3 -m sglang.launch_server --model deepseek-ai/DeepSeek-R1-Distill-Qwen-32B --trust-remote-code --tp 2
```
### Usage Recommendations
**We recommend adhering to the following configurations when utilizing the DeepSeek-R1 series models, including benchmarking, to achieve the expected performance:**
1. Set the temperature within the range of 0.5-0.7 (0.6 is recommended) to prevent endless repetitions or incoherent outputs.
2. **Avoid adding a system prompt; all instructions should be contained within the user prompt.**
3. For mathematical problems, it is advisable to include a directive in your prompt such as: "Please reason step by step, and put your final answer within \boxed{}."
4. When evaluating model performance, it is recommended to conduct multiple tests and average the results.
Additionally, we have observed that the DeepSeek-R1 series models tend to bypass thinking pattern (i.e., outputting "\<think\>\n\n\</think\>") when responding to certain queries, which can adversely affect the model's performance.
**To ensure that the model engages in thorough reasoning, we recommend enforcing the model to initiate its response with "\<think\>\n" at the beginning of every output.**
## 7. License
This code repository and the model weights are licensed under the [MIT License](https://github.com/deepseek-ai/DeepSeek-R1/blob/main/LICENSE).
DeepSeek-R1 series support commercial use, allow for any modifications and derivative works, including, but not limited to, distillation for training other LLMs. Please note that:
- DeepSeek-R1-Distill-Qwen-1.5B, DeepSeek-R1-Distill-Qwen-7B, DeepSeek-R1-Distill-Qwen-14B and DeepSeek-R1-Distill-Qwen-32B are derived from [Qwen-2.5 series](https://github.com/QwenLM/Qwen2.5), which are originally licensed under [Apache 2.0 License](https://huggingface.co/Qwen/Qwen2.5-1.5B/blob/main/LICENSE), and now finetuned with 800k samples curated with DeepSeek-R1.
- DeepSeek-R1-Distill-Llama-8B is derived from Llama3.1-8B-Base and is originally licensed under [llama3.1 license](https://huggingface.co/meta-llama/Llama-3.1-8B/blob/main/LICENSE).
- DeepSeek-R1-Distill-Llama-70B is derived from Llama3.3-70B-Instruct and is originally licensed under [llama3.3 license](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct/blob/main/LICENSE).
## 8. Citation
```
@misc{deepseekai2025deepseekr1incentivizingreasoningcapability,
title={DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning},
author={DeepSeek-AI},
year={2025},
eprint={2501.12948},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2501.12948},
}
```
## 9. Contact
If you have any questions, please raise an issue or contact us at [[email protected]]([email protected]).
|
Mungert/OLMo-2-0325-32B-Instruct-GGUF | Mungert | 2025-06-15T19:43:42Z | 783 | 2 | transformers | [
"transformers",
"gguf",
"text-generation",
"en",
"dataset:allenai/RLVR-GSM-MATH-IF-Mixed-Constraints",
"arxiv:2501.00656",
"arxiv:2411.15124",
"base_model:allenai/OLMo-2-0325-32B-DPO",
"base_model:quantized:allenai/OLMo-2-0325-32B-DPO",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-02T05:04:04Z | ---
license: apache-2.0
language:
- en
datasets:
- allenai/RLVR-GSM-MATH-IF-Mixed-Constraints
base_model:
- allenai/OLMo-2-0325-32B-DPO
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">OLMo-2-0325-32B-Instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `OLMo-2-0325-32B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `OLMo-2-0325-32B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `OLMo-2-0325-32B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `OLMo-2-0325-32B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `OLMo-2-0325-32B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `OLMo-2-0325-32B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `OLMo-2-0325-32B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `OLMo-2-0325-32B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `OLMo-2-0325-32B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `OLMo-2-0325-32B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `OLMo-2-0325-32B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
<img alt="OLMo Logo" src="https://huggingface.co/datasets/allenai/blog-images/resolve/main/olmo2/olmo.png" width="242px">
OLMo 2 32B Instruct March 2025 is post-trained variant of the [OLMo-2 32B March 2025](https://huggingface.co/allenai/OLMo-2-0325-32B/) model, which has undergone supervised finetuning on an OLMo-specific variant of the [Tülu 3 dataset](https://huggingface.co/datasets/allenai/tulu-3-sft-olmo-2-mixture), further DPO training on [this dataset](https://huggingface.co/datasets/allenai/olmo-2-0325-32b-preference-mix), and final RLVR training on [this dataset](https://huggingface.co/datasets/allenai/RLVR-GSM-MATH-IF-Mixed-Constraints).
Tülu 3 is designed for state-of-the-art performance on a diversity of tasks in addition to chat, such as MATH, GSM8K, and IFEval.
Check out the [OLMo 2 paper](https://arxiv.org/abs/2501.00656) or [Tülu 3 paper](https://arxiv.org/abs/2411.15124) for more details!
OLMo is a series of **O**pen **L**anguage **Mo**dels designed to enable the science of language models.
These models are trained on the Dolma dataset. We are releasing all code, checkpoints, logs, and associated training details.
## Model description
- **Model type:** A model trained on a mix of publicly available, synthetic and human-created datasets.
- **Language(s) (NLP):** Primarily English
- **License:** Apache 2.0
- **Finetuned from model:** allenai/OLMo-2-0325-32B-DPO
### Model Sources
- **Project Page:** https://allenai.org/olmo
- **Repositories:**
- Core repo (training, inference, fine-tuning etc.): https://github.com/allenai/OLMo-core
- Evaluation code: https://github.com/allenai/olmes
- Further fine-tuning code: https://github.com/allenai/open-instruct
- **Paper:** https://arxiv.org/abs/2501.00656
- **Demo:** https://playground.allenai.org/
## Installation
OLMo 2 will be supported in the next version of Transformers, and you need to install it from the main branch using:
```bash
pip install --upgrade git+https://github.com/huggingface/transformers.git
```
## Using the model
### Loading with HuggingFace
To load the model with HuggingFace, use the following snippet:
```
from transformers import AutoModelForCausalLM
olmo_model = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0325-32B-Instruct")
```
### Chat template
*NOTE: This is different than previous OLMo 2 and Tülu 3 models due to a minor change in configuration. It does NOT have the bos token before the rest. Our other models have <|endoftext|> at the beginning of the chat template.*
The chat template for our models is formatted as:
```
<|user|>\nHow are you doing?\n<|assistant|>\nI'm just a computer program, so I don't have feelings, but I'm functioning as expected. How can I assist you today?<|endoftext|>
```
Or with new lines expanded:
```
<|user|>
How are you doing?
<|assistant|>
I'm just a computer program, so I don't have feelings, but I'm functioning as expected. How can I assist you today?<|endoftext|>
```
It is embedded within the tokenizer as well, for `tokenizer.apply_chat_template`.
### System prompt
In Ai2 demos, we use this system prompt by default:
```
You are OLMo 2, a helpful and harmless AI Assistant built by the Allen Institute for AI.
```
The model has not been trained with a specific system prompt in mind.
### Intermediate Checkpoints
To facilitate research on RL finetuning, we have released our intermediate checkpoints during the model's RLVR training.
The model weights are saved every 20 training steps, and can be accessible in the revisions of the HuggingFace repository.
For example, you can load with:
```
olmo_model = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0325-32B-Instruct", revision="step_200")
```
### Bias, Risks, and Limitations
The OLMo-2 models have limited safety training, but are not deployed automatically with in-the-loop filtering of responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so).
See the Falcon 180B model card for an example of this.
## Performance
| Model | Average | AlpacaEval 2 LC | BBH | DROP | GSM8k | IFEval | MATH | MMLU | Safety | PopQA | TruthQA |
|-------|---------|------|-----|------|-------|--------|------|------|--------|-------|---------|
| **Closed API models** | | | | | | | | | | | |
| GPT-3.5 Turbo 0125 | 59.6 | 38.7 | 66.6 | 70.2 | 74.3 | 66.9 | 41.2 | 70.2 | 69.1 | 45.0 | 62.9 |
| GPT 4o Mini 2024-07-18 | 65.7 | 49.7 | 65.9 | 36.3 | 83.0 | 83.5 | 67.9 | 82.2 | 84.9 | 39.0 | 64.8 |
| **Open weights models** | | | | | | | | | | | |
| Mistral-Nemo-Instruct-2407 | 50.9 | 45.8 | 54.6 | 23.6 | 81.4 | 64.5 | 31.9 | 70.0 | 52.7 | 26.9 | 57.7 |
| Ministral-8B-Instruct | 52.1 | 31.4 | 56.2 | 56.2 | 80.0 | 56.4 | 40.0 | 68.5 | 56.2 | 20.2 | 55.5 |
| Gemma-2-27b-it | 61.3 | 49.0 | 72.7 | 67.5 | 80.7 | 63.2 | 35.1 | 70.7 | 75.9 | 33.9 | 64.6 |
| Qwen2.5-32B | 66.5 | 39.1 | 82.3 | 48.3 | 87.5 | 82.4 | 77.9 | 84.7 | 82.4 | 26.1 | 70.6 |
| Mistral-Small-24B | 67.6 | 43.2 | 80.1 | 78.5 | 87.2 | 77.3 | 65.9 | 83.7 | 66.5 | 24.4 | 68.1 |
| Llama-3.1-70B | 70.0 | 32.9 | 83.0 | 77.0 | 94.5 | 88.0 | 56.2 | 85.2 | 76.4 | 46.5 | 66.8 |
| Llama-3.3-70B | 73.0 | 36.5 | 85.8 | 78.0 | 93.6 | 90.8 | 71.8 | 85.9 | 70.4 | 48.2 | 66.1 |
| Gemma-3-27b-it | - | 63.4 | 83.7 | 69.2 | 91.1 | - | - | 81.8 | - | 30.9 | - |
| **Fully open models** | | | | | | | | | | | |
| OLMo-2-7B-1124-Instruct | 55.7 | 31.0 | 48.5 | 58.9 | 85.2 | 75.6 | 31.3 | 63.9 | 81.2 | 24.6 | 56.3 |
| OLMo-2-13B-1124-Instruct | 61.4 | 37.5 | 58.4 | 72.1 | 87.4 | 80.4 | 39.7 | 68.6 | 77.5 | 28.8 | 63.9 |
| **OLMo-2-32B-0325-SFT** | 61.7 | 16.9 | 69.7 | 77.2 | 78.4 | 72.4 | 35.9 | 76.1 | 93.8 | 35.4 | 61.3 |
| **OLMo-2-32B-0325-DPO** | 68.8 | 44.1 | 70.2 | 77.5 | 85.7 | 83.8 | 46.8 | 78.0 | 91.9 | 36.4 | 73.5 |
| **OLMo-2-32B-0325-Instruct** | 68.8 | 42.8 | 70.6 | 78.0 | 87.6 | 85.6 | 49.7 | 77.3 | 85.9 | 37.5 | 73.2 |
## Learning curves
Below is the training curves for `allenai/OLMo-2-0325-32B-Instruct`. The model was trained using 5 8xH100 nodes.
Below are the core eval scores over steps for `allenai/OLMo-2-0325-32B-Instruct` (note we took step `320` as the final checkpoint, corresponding to episode `573,440`):
Below are the other eval scores over steps for `allenai/OLMo-2-0325-32B-Instruct`:
## Reproduction command
The command below is copied directly from the tracked training job:
```bash
# clone and check out commit
git clone https://github.com/allenai/open-instruct.git
# this should be the correct commit, the main thing is to have the vllm monkey patch for
# 32b olmo https://github.com/allenai/open-instruct/blob/894ffa236319bc6c26c346240a7e4ee04ba0bd31/open_instruct/vllm_utils2.py#L37-L59
git checkout a51dc98525eec01de6e8a24c071f42dce407d738
uv sync
uv sync --extra compile
# note that you may need 5 8xH100 nodes for the training.
# so please setup ray properly, e.g., https://github.com/allenai/open-instruct/blob/main/docs/tulu3.md#llama-31-tulu-3-70b-reproduction
python open_instruct/grpo_vllm_thread_ray_gtrl.py \
--exp_name 0310_olmo2_32b_grpo_12818 \
--beta 0.01 \
--local_mini_batch_size 32 \
--number_samples_per_prompt 16 \
--output_dir output \
--local_rollout_batch_size 4 \
--kl_estimator kl3 \
--learning_rate 5e-7 \
--dataset_mixer_list allenai/RLVR-GSM-MATH-IF-Mixed-Constraints 1.0 \
--dataset_mixer_list_splits train \
--dataset_mixer_eval_list allenai/RLVR-GSM-MATH-IF-Mixed-Constraints 16 \
--dataset_mixer_eval_list_splits train \
--max_token_length 2048 \
--max_prompt_token_length 2048 \
--response_length 2048 \
--model_name_or_path allenai/OLMo-2-0325-32B-DPO \
--non_stop_penalty \
--stop_token eos \
--temperature 1.0 \
--ground_truths_key ground_truth \
--chat_template_name tulu \
--sft_messages_key messages \
--eval_max_length 4096 \
--total_episodes 10000000 \
--penalty_reward_value 0.0 \
--deepspeed_stage 3 \
--no_gather_whole_model \
--per_device_train_batch_size 2 \
--local_rollout_forward_batch_size 2 \
--actor_num_gpus_per_node 8 8 8 4 \
--num_epochs 1 \
--vllm_tensor_parallel_size 1 \
--vllm_num_engines 12 \
--lr_scheduler_type constant \
--apply_verifiable_reward true \
--seed 1 \
--num_evals 30 \
--save_freq 20 \
--reward_model_multiplier 0.0 \
--no_try_launch_beaker_eval_jobs \
--try_launch_beaker_eval_jobs_on_weka \
--gradient_checkpointing \
--with_tracking
```
## License and use
OLMo 2 is licensed under the Apache 2.0 license.
OLMo 2 is intended for research and educational use.
For more information, please see our [Responsible Use Guidelines](https://allenai.org/responsible-use).
This model has been fine-tuned using a dataset mix with outputs generated from third party models and are subject to additional terms: [Gemma Terms of Use](https://ai.google.dev/gemma/terms).
## Citation
```bibtex
@article{olmo20242olmo2furious,
title={2 OLMo 2 Furious},
author={Team OLMo and Pete Walsh and Luca Soldaini and Dirk Groeneveld and Kyle Lo and Shane Arora and Akshita Bhagia and Yuling Gu and Shengyi Huang and Matt Jordan and Nathan Lambert and Dustin Schwenk and Oyvind Tafjord and Taira Anderson and David Atkinson and Faeze Brahman and Christopher Clark and Pradeep Dasigi and Nouha Dziri and Michal Guerquin and Hamish Ivison and Pang Wei Koh and Jiacheng Liu and Saumya Malik and William Merrill and Lester James V. Miranda and Jacob Morrison and Tyler Murray and Crystal Nam and Valentina Pyatkin and Aman Rangapur and Michael Schmitz and Sam Skjonsberg and David Wadden and Christopher Wilhelm and Michael Wilson and Luke Zettlemoyer and Ali Farhadi and Noah A. Smith and Hannaneh Hajishirzi},
year={2024},
eprint={2501.00656},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2501.00656},
}
``` |
Mungert/Llama-3.3-70B-Instruct-GGUF | Mungert | 2025-06-15T19:43:31Z | 4,368 | 6 | transformers | [
"transformers",
"gguf",
"facebook",
"meta",
"pytorch",
"llama",
"llama-3",
"en",
"fr",
"it",
"pt",
"hi",
"es",
"th",
"de",
"arxiv:2204.05149",
"base_model:meta-llama/Llama-3.1-70B",
"base_model:quantized:meta-llama/Llama-3.1-70B",
"license:llama3.3",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-01T06:10:00Z | ---
library_name: transformers
language:
- en
- fr
- it
- pt
- hi
- es
- th
- de
base_model:
- meta-llama/Llama-3.1-70B
tags:
- facebook
- meta
- pytorch
- llama
- llama-3
extra_gated_prompt: "### LLAMA 3.3 COMMUNITY LICENSE AGREEMENT\nLlama 3.3 Version Release Date: December 6, 2024\n\"Agreement\" means the terms and conditions for use, reproduction, distribution and modification of the Llama Materials set forth herein.\n\"Documentation\" means the specifications, manuals and documentation accompanying Llama 3.3 distributed by Meta at [https://www.llama.com/docs/overview](https://llama.com/docs/overview).\n\"Licensee\" or \"you\" means you, or your employer or any other person or entity (if you are entering into this Agreement on such person or entity’s behalf), of the age required under applicable laws, rules or regulations to provide legal consent and that has legal authority to bind your employer or such other person or entity if you are entering in this Agreement on their behalf.\n\"Llama 3.3\" means the foundational large language models and software and algorithms, including machine-learning model code, trained model weights, inference-enabling code, training-enabling code, fine-tuning enabling code and other elements of the foregoing distributed by Meta at [https://www.llama.com/llama-downloads](https://www.llama.com/llama-downloads).\n\"Llama Materials\" means, collectively, Meta’s proprietary Llama 3.3 and Documentation (and any portion thereof) made available under this Agreement.\n\"Meta\" or \"we\" means Meta Platforms Ireland Limited (if you are located in or, if you are an entity, your principal place of business is in the EEA or Switzerland) and Meta Platforms, Inc. (if you are located outside of the EEA or Switzerland).\nBy clicking “I Accept” below or by using or distributing any portion or element of the Llama Materials, you agree to be bound by this Agreement.\n1. License Rights and Redistribution.\na. Grant of Rights. You are granted a non-exclusive, worldwide, non-transferable and royalty-free limited license under Meta’s intellectual property or other rights owned by Meta embodied in the Llama Materials to use, reproduce, distribute, copy, create derivative works of, and make modifications to the Llama Materials.\nb. Redistribution and Use.\ni. If you distribute or make available the Llama Materials (or any derivative works thereof), or a product or service (including another AI model) that contains any of them, you shall (A) provide a copy of this Agreement with any such Llama Materials; and (B) prominently display “Built with Llama” on a related website, user interface, blogpost, about page, or product documentation. If you use the Llama Materials or any outputs or results of the Llama Materials to create, train, fine tune, or otherwise improve an AI model, which is distributed or made available, you shall also include “Llama” at the beginning of any such AI model name.\nii. If you receive Llama Materials, or any derivative works thereof, from a Licensee as part of an integrated end user product, then Section 2 of this Agreement will not apply to you.\_\niii. You must retain in all copies of the Llama Materials that you distribute the following attribution notice within a “Notice” text file distributed as a part of such copies: “Llama 3.3 is licensed under the Llama 3.3 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved.”\niv. Your use of the Llama Materials must comply with applicable laws and regulations (including trade compliance laws and regulations) and adhere to the Acceptable Use Policy for the Llama Materials (available at [https://www.llama.com/llama3\\_3/use-policy](https://www.llama.com/llama3_3/use-policy)), which is hereby incorporated by reference into this Agreement. \n2. Additional Commercial Terms. If, on the Llama 3.3 version release date, the monthly active users of the products or services made available by or for Licensee, or Licensee’s affiliates, is greater than 700 million monthly active users in the preceding calendar month, you must request a license from Meta, which Meta may grant to you in its sole discretion, and you are not authorized to exercise any of the rights under this Agreement unless or until Meta otherwise expressly grants you such rights.\n3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF ANY KIND, AND META DISCLAIMS ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE LLAMA MATERIALS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS.\n4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF META OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.\n5. Intellectual Property.\na. No trademark licenses are granted under this Agreement, and in connection with the Llama Materials, neither Meta nor Licensee may use any name or mark owned by or associated with the other or any of its affiliates, except as required for reasonable and customary use in describing and redistributing the Llama Materials or as set forth in this Section 5(a). Meta hereby grants you a license to use “Llama” (the “Mark”) solely as required to comply with the last sentence of Section 1.b.i. You will comply with Meta’s brand guidelines (currently accessible at [https://about.meta.com/brand/resources/meta/company-brand/](https://about.meta.com/brand/resources/meta/company-brand/)[)](https://en.facebookbrand.com/). All goodwill arising out of your use of the Mark will inure to the benefit of Meta.\nb. Subject to Meta’s ownership of Llama Materials and derivatives made by or for Meta, with respect to any derivative works and modifications of the Llama Materials that are made by you, as between you and Meta, you are and will be the owner of such derivative works and modifications.\nc. If you institute litigation or other proceedings against Meta or any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Llama Materials or Llama 3.3 outputs or results, or any portion of any of the foregoing, constitutes infringement of intellectual property or other rights owned or licensable by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold harmless Meta from and against any claim by any third party arising out of or related to your use or distribution of the Llama Materials.\n6. Term and Termination. The term of this Agreement will commence upon your acceptance of this Agreement or access to the Llama Materials and will continue in full force and effect until terminated in accordance with the terms and conditions herein. Meta may terminate this Agreement if you are in breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete and cease use of the Llama Materials. Sections 3, 4 and 7 shall survive the termination of this Agreement.\n7. Governing Law and Jurisdiction. This Agreement will be governed and construed under the laws of the State of California without regard to choice of law principles, and the UN Convention on Contracts for the International Sale of Goods does not apply to this Agreement. The courts of California shall have exclusive jurisdiction of any dispute arising out of this Agreement.\n### Llama 3.3 Acceptable Use Policy\nMeta is committed to promoting safe and fair use of its tools and features, including Llama 3.3. If you access or use Llama 3.3, you agree to this Acceptable Use Policy (“**Policy**”). The most recent copy of this policy can be found at [https://www.llama.com/llama3\\_3/use-policy](https://www.llama.com/llama3_3/use-policy).\nProhibited Uses\nWe want everyone to use Llama 3.3 safely and responsibly. You agree you will not use, or allow others to use, Llama 3.3 to:\n1. Violate the law or others’ rights, including to:\n\n 1. Engage in, promote, generate, contribute to, encourage, plan, incite, or further illegal or unlawful activity or content, such as: \n 1. Violence or terrorism \n 2. Exploitation or harm to children, including the solicitation, creation, acquisition, or dissemination of child exploitative content or failure to report Child Sexual Abuse Material \n 3. Human trafficking, exploitation, and sexual violence \n 4. The illegal distribution of information or materials to minors, including obscene materials, or failure to employ legally required age-gating in connection with such information or materials. \n 5. Sexual solicitation \n 6. Any other criminal activity\n\n 2. Engage in, promote, incite, or facilitate the harassment, abuse, threatening, or bullying of individuals or groups of individuals\n\n 3. Engage in, promote, incite, or facilitate discrimination or other unlawful or harmful conduct in the provision of employment, employment benefits, credit, housing, other economic benefits, or other essential goods and services\n\n 4. Engage in the unauthorized or unlicensed practice of any profession including, but not limited to, financial, legal, medical/health, or related professional practices\n\n 5. Collect, process, disclose, generate, or infer private or sensitive information about individuals, including information about individuals’ identity, health, or demographic information, unless you have obtained the right to do so in accordance with applicable law\n\n 6. Engage in or facilitate any action or generate any content that infringes, misappropriates, or otherwise violates any third-party rights, including the outputs or results of any products or services using the Llama Materials\n\n 7. Create, generate, or facilitate the creation of malicious code, malware, computer viruses or do anything else that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system\n\n 8. Engage in any action, or facilitate any action, to intentionally circumvent or remove usage restrictions or other safety measures, or to enable functionality disabled by Meta\n\n2. Engage in, promote, incite, facilitate, or assist in the planning or development of activities that present a risk of death or bodily harm to individuals, including use of Llama 3.3 related to the following:\n\n 1. Military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulations (ITAR) maintained by the United States Department of State or to the U.S. Biological Weapons Anti-Terrorism Act of 1989 or the Chemical Weapons Convention Implementation Act of 1997\n\n 2. Guns and illegal weapons (including weapon development)\n\n 3. Illegal drugs and regulated/controlled substances\n\n 4. Operation of critical infrastructure, transportation technologies, or heavy machinery\n\n 5. Self-harm or harm to others, including suicide, cutting, and eating disorders\n\n 6. Any content intended to incite or promote violence, abuse, or any infliction of bodily harm to an individual\n\n3. Intentionally deceive or mislead others, including use of Llama 3.3 related to the following:\n\n 1. Generating, promoting, or furthering fraud or the creation or promotion of disinformation\n\n 2. Generating, promoting, or furthering defamatory content, including the creation of defamatory statements, images, or other content\n\n 3. Generating, promoting, or further distributing spam\n\n 4. Impersonating another individual without consent, authorization, or legal right\n\n 5. Representing that the use of Llama 3.3 or outputs are human-generated\n\n 6. Generating or facilitating false online engagement, including fake reviews and other means of fake online engagement\n\n4. Fail to appropriately disclose to end users any known dangers of your AI system\n5. Interact with third party tools, models, or software designed to generate unlawful content or engage in unlawful or harmful conduct and/or represent that the outputs of such tools, models, or software are associated with Meta or Llama 3.3\nWith respect to any multimodal models included in Llama 3.3, the rights granted under Section 1(a) of the Llama 3.3 Community License Agreement are not being granted to you if you are an individual domiciled in, or a company with a principal place of business in, the European Union. This restriction does not apply to end users of a product or service that incorporates any such multimodal models.\nPlease report any violation of this Policy, software “bug,” or other problems that could lead to a violation of this Policy through one of the following means:\n* Reporting issues with the model: [https://github.com/meta-llama/llama-models/issues](https://l.workplace.com/l.php?u=https%3A%2F%2Fgithub.com%2Fmeta-llama%2Fllama-models%2Fissues&h=AT0qV8W9BFT6NwihiOHRuKYQM_UnkzN_NmHMy91OT55gkLpgi4kQupHUl0ssR4dQsIQ8n3tfd0vtkobvsEvt1l4Ic6GXI2EeuHV8N08OG2WnbAmm0FL4ObkazC6G_256vN0lN9DsykCvCqGZ) * Reporting risky content generated by the model: [developers.facebook.com/llama\\_output\\_feedback](http://developers.facebook.com/llama_output_feedback) * Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info) * Reporting violations of the Acceptable Use Policy or unlicensed uses of Llama 3.3: [email protected] "
extra_gated_fields:
First Name: text
Last Name: text
Date of birth: date_picker
Country: country
Affiliation: text
Job title:
type: select
options:
- Student
- Research Graduate
- AI researcher
- AI developer/engineer
- Reporter
- Other
geo: ip_location
By clicking Submit below I accept the terms of the license and acknowledge that the information I provide will be collected stored processed and shared in accordance with the Meta Privacy Policy: checkbox
extra_gated_description: >-
The information you provide will be collected, stored, processed and shared in
accordance with the [Meta Privacy
Policy](https://www.facebook.com/privacy/policy/).
extra_gated_button_content: Submit
license: llama3.3
---
# <span style="color: #7FFF7F;">Llama-3.3-70B-Instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.3-70B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.3-70B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.3-70B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.3-70B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.3-70B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.3-70B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.3-70B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.3-70B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.3-70B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.3-70B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.3-70B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# <span style="color: #7FFF7F;">meta-llama/Llama-3.3-70B-Instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `meta-llama/Llama-3.3-70B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `meta-llama/Llama-3.3-70B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `meta-llama/Llama-3.3-70B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `meta-llama/Llama-3.3-70B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `meta-llama/Llama-3.3-70B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `meta-llama/Llama-3.3-70B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `meta-llama/Llama-3.3-70B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `meta-llama/Llama-3.3-70B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `meta-llama/Llama-3.3-70B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `meta-llama/Llama-3.3-70B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `meta-llama/Llama-3.3-70B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
## Model Information
The Meta Llama 3.3 multilingual large language model (LLM) is an instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.
**Model developer**: Meta
**Model Architecture:** Llama 3.3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.
| | Training Data | Params | Input modalities | Output modalities | Context length | GQA | Token count | Knowledge cutoff |
| :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- |
| Llama 3.3 (text only) | A new mix of publicly available online data. | 70B | Multilingual Text | Multilingual Text and code | 128k | Yes | 15T+ | December 2023 |
**Supported languages:** English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.
**Llama 3.3 model**. Token counts refer to pretraining data only. All model versions use Grouped-Query Attention (GQA) for improved inference scalability.
**Model Release Date:**
* **70B Instruct: December 6, 2024**
**Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback.
**License** A custom commercial license, the Llama 3.3 Community License Agreement, is available at: [https://github.com/meta-llama/llama-models/blob/main/models/llama3\_3/LICENSE](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/LICENSE)
Where to send questions or comments about the model Instructions on how to provide feedback or comments on the model can be found in the model [README](https://github.com/meta-llama/llama3). For more technical information about generation parameters and recipes for how to use Llama 3.3 in applications, please go [here](https://github.com/meta-llama/llama-recipes).
## Intended Use
**Intended Use Cases** Llama 3.3 is intended for commercial and research use in multiple languages. Instruction tuned text only models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks. The Llama 3.3 model also supports the ability to leverage the outputs of its models to improve other models including synthetic data generation and distillation. The Llama 3.3 Community License allows for these use cases.
**Out-of-scope** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in any other way that is prohibited by the Acceptable Use Policy and Llama 3.3 Community License. Use in languages beyond those explicitly referenced as supported in this model card\*\*.
\*\*Note: Llama 3.3 has been trained on a broader collection of languages than the 8 supported languages. Developers may fine-tune Llama 3.3 models for languages beyond the 8 supported languages provided they comply with the Llama 3.3 Community License and the Acceptable Use Policy and in such cases are responsible for ensuring that any uses of Llama 3.3 in additional languages is done in a safe and responsible manner.
## How to use
This repository contains two versions of Llama-3.3-70B-Instruct, for use with transformers and with the original `llama` codebase.
### Use with transformers
Starting with `transformers >= 4.45.0` onward, you can run conversational inference using the Transformers `pipeline` abstraction or by leveraging the Auto classes with the `generate()` function.
Make sure to update your transformers installation via `pip install --upgrade transformers`.
See the snippet below for usage with Transformers:
```python
import transformers
import torch
model_id = "meta-llama/Llama-3.3-70B-Instruct"
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
model_kwargs={"torch_dtype": torch.bfloat16},
device_map="auto",
)
messages = [
{"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
{"role": "user", "content": "Who are you?"},
]
outputs = pipeline(
messages,
max_new_tokens=256,
)
print(outputs[0]["generated_text"][-1])
```
### Tool use with transformers
LLaMA-3.3 supports multiple tool use formats. You can see a full guide to prompt formatting [here](https://llama.meta.com/docs/model-cards-and-prompt-formats/llama3_1/).
Tool use is also supported through [chat templates](https://huggingface.co/docs/transformers/main/chat_templating#advanced-tool-use--function-calling) in Transformers.
Here is a quick example showing a single simple tool:
```python
# First, define a tool
def get_current_temperature(location: str) -> float:
"""
Get the current temperature at a location.
Args:
location: The location to get the temperature for, in the format "City, Country"
Returns:
The current temperature at the specified location in the specified units, as a float.
"""
return 22. # A real function should probably actually get the temperature!
# Next, create a chat and apply the chat template
messages = [
{"role": "system", "content": "You are a bot that responds to weather queries."},
{"role": "user", "content": "Hey, what's the temperature in Paris right now?"}
]
inputs = tokenizer.apply_chat_template(messages, tools=[get_current_temperature], add_generation_prompt=True)
```
You can then generate text from this input as normal. If the model generates a tool call, you should add it to the chat like so:
```python
tool_call = {"name": "get_current_temperature", "arguments": {"location": "Paris, France"}}
messages.append({"role": "assistant", "tool_calls": [{"type": "function", "function": tool_call}]})
```
and then call the tool and append the result, with the `tool` role, like so:
```python
messages.append({"role": "tool", "name": "get_current_temperature", "content": "22.0"})
```
After that, you can `generate()` again to let the model use the tool result in the chat. Note that this was a very brief introduction to tool calling - for more information,
see the [LLaMA prompt format docs](https://llama.meta.com/docs/model-cards-and-prompt-formats/llama3_1/) and the Transformers [tool use documentation](https://huggingface.co/docs/transformers/main/chat_templating#advanced-tool-use--function-calling).
### Use with `bitsandbytes`
The model checkpoints can be used in `8-bit` and `4-bit` for further memory optimisations using `bitsandbytes` and `transformers`
See the snippet below for usage:
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "meta-llama/Llama-3.3-70B-Instruct"
quantization_config = BitsAndBytesConfig(load_in_8bit=True)
quantized_model = AutoModelForCausalLM.from_pretrained(
model_id, device_map="auto", torch_dtype=torch.bfloat16, quantization_config=quantization_config)
tokenizer = AutoTokenizer.from_pretrained(model_id)
input_text = "What are we having for dinner?"
input_ids = tokenizer(input_text, return_tensors="pt").to("cuda")
output = quantized_model.generate(**input_ids, max_new_tokens=10)
print(tokenizer.decode(output[0], skip_special_tokens=True))
```
To load in 4-bit simply pass `load_in_4bit=True`
### Use with `llama`
Please, follow the instructions in the [repository](https://github.com/meta-llama/llama).
To download Original checkpoints, see the example command below leveraging `huggingface-cli`:
```
huggingface-cli download meta-llama/Llama-3.3-70B-Instruct --include "original/*" --local-dir Llama-3.3-70B-Instruct
```
## Hardware and Software
**Training Factors** We used custom training libraries, Meta's custom built GPU cluster, and production infrastructure for pretraining. Fine-tuning, annotation, and evaluation were also performed on production infrastructure.
**Training Energy Use** Training utilized a cumulative of **39.3**M GPU hours of computation on H100-80GB (TDP of 700W) type hardware, per the table below. Training time is the total GPU time required for training each model and power consumption is the peak power capacity per GPU device used, adjusted for power usage efficiency.
##
## **Training Greenhouse Gas Emissions** Estimated total location-based greenhouse gas emissions were **11,390** tons CO2eq for training. Since 2020, Meta has maintained net zero greenhouse gas emissions in its global operations and matched 100% of its electricity use with renewable energy, therefore the total market-based greenhouse gas emissions for training were 0 tons CO2eq.
| | Training Time (GPU hours) | Training Power Consumption (W) | Training Location-Based Greenhouse Gas Emissions (tons CO2eq) | Training Market-Based Greenhouse Gas Emissions (tons CO2eq) |
| :---- | :---: | :---: | :---: | :---: |
| Llama 3.3 70B | 7.0M | 700 | 2,040 | 0 |
## The methodology used to determine training energy use and greenhouse gas emissions can be found [here](https://arxiv.org/pdf/2204.05149). Since Meta is openly releasing these models, the training energy use and greenhouse gas emissions will not be incurred by others.
## Training Data
**Overview:** Llama 3.3 was pretrained on \~15 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over 25M synthetically generated examples.
**Data Freshness:** The pretraining data has a cutoff of December 2023\.
## Benchmarks \- English Text
In this section, we report the results for Llama 3.3 relative to our previous models.
### Instruction tuned models
##
| Category | Benchmark | \# Shots | Metric | Llama 3.1 8B Instruct | Llama 3.1 70B Instruct | Llama-3.3 70B Instruct | Llama 3.1 405B Instruct |
| :---- | :---- | ----- | :---- | ----- | ----- | ----- | ----- |
| | MMLU (CoT) | 0 | macro\_avg/acc | 73.0 | 86.0 | 86.0 | 88.6 |
| | MMLU Pro (CoT) | 5 | macro\_avg/acc | 48.3 | 66.4 | 68.9 | 73.3 |
| Steerability | IFEval | | | 80.4 | 87.5 | 92.1 | 88.6 |
| Reasoning | GPQA Diamond (CoT) | 0 | acc | 31.8 | 48.0 | 50.5 | 49.0 |
| Code | HumanEval | 0 | pass@1 | 72.6 | 80.5 | 88.4 | 89.0 |
| | MBPP EvalPlus (base) | 0 | pass@1 | 72.8 | 86.0 | 87.6 | 88.6 |
| Math | MATH (CoT) | 0 | sympy\_intersection\_score | 51.9 | 68.0 | 77.0 | 73.8 |
| Tool Use | BFCL v2 | 0 | overall\_ast\_summary/macro\_avg/valid | 65.4 | 77.5 | 77.3 | 81.1 |
| Multilingual | MGSM | 0 | em | 68.9 | 86.9 | 91.1 | 91.6 |
##
## Responsibility & Safety
As part of our Responsible release approach, we followed a three-pronged strategy to managing trust & safety risks:
* Enable developers to deploy helpful, safe and flexible experiences for their target audience and for the use cases supported by Llama.
* Protect developers against adversarial users aiming to exploit Llama capabilities to potentially cause harm.
* Provide protections for the community to help prevent the misuse of our models.
### Responsible deployment
Llama is a foundational technology designed to be used in a variety of use cases, examples on how Meta’s Llama models have been responsibly deployed can be found in our [Community Stories webpage](https://llama.meta.com/community-stories/). Our approach is to build the most helpful models enabling the world to benefit from the technology power, by aligning our model safety for the generic use cases addressing a standard set of harms. Developers are then in the driver seat to tailor safety for their use case, defining their own policy and deploying the models with the necessary safeguards in their Llama systems. Llama 3.3 was developed following the best practices outlined in our Responsible Use Guide, you can refer to the [Responsible Use Guide](https://llama.meta.com/responsible-use-guide/) to learn more.
#### Llama 3.3 instruct
Our main objectives for conducting safety fine-tuning are to provide the research community with a valuable resource for studying the robustness of safety fine-tuning, as well as to offer developers a readily available, safe, and powerful model for various applications to reduce the developer workload to deploy safe AI systems. For more details on the safety mitigations implemented please read the Llama 3 paper.
**Fine-tuning data**
We employ a multi-faceted approach to data collection, combining human-generated data from our vendors with synthetic data to mitigate potential safety risks. We’ve developed many large language model (LLM)-based classifiers that enable us to thoughtfully select high-quality prompts and responses, enhancing data quality control.
**Refusals and Tone**
Building on the work we started with Llama 3, we put a great emphasis on model refusals to benign prompts as well as refusal tone. We included both borderline and adversarial prompts in our safety data strategy, and modified our safety data responses to follow tone guidelines.
#### Llama 3.3 systems
**Large language models, including Llama 3.3, are not designed to be deployed in isolation but instead should be deployed as part of an overall AI system with additional safety guardrails as required.** Developers are expected to deploy system safeguards when building agentic systems. Safeguards are key to achieve the right helpfulness-safety alignment as well as mitigating safety and security risks inherent to the system and any integration of the model or system with external tools.
As part of our responsible release approach, we provide the community with [safeguards](https://llama.meta.com/trust-and-safety/) that developers should deploy with Llama models or other LLMs, including Llama Guard 3, Prompt Guard and Code Shield. All our [reference implementations](https://github.com/meta-llama/llama-agentic-system) demos contain these safeguards by default so developers can benefit from system-level safety out-of-the-box.
#### Capability specific considerations
**Tool-use**: Just like in standard software development, developers are responsible for the integration of the LLM with the tools and services of their choice. They should define a clear policy for their use case and assess the integrity of the third party services they use to be aware of the safety and security limitations when using this capability. Refer to the Responsible Use Guide for best practices on the safe deployment of the third party safeguards.
**Multilinguality**: Llama 3.3 supports 7 languages in addition to English: French, German, Hindi, Italian, Portuguese, Spanish, and Thai. Llama may be able to output text in other languages than those that meet performance thresholds for safety and helpfulness. We strongly discourage developers from using this model to converse in non-supported languages without implementing finetuning and system controls in alignment with their policies and the best practices shared in the Responsible Use Guide.
### Evaluations
We evaluated Llama models for common use cases as well as specific capabilities. Common use cases evaluations measure safety risks of systems for most commonly built applications including chat bot, coding assistant, tool calls. We built dedicated, adversarial evaluation datasets and evaluated systems composed of Llama models and Llama Guard 3 to filter input prompt and output response. It is important to evaluate applications in context, and we recommend building dedicated evaluation dataset for your use case. Prompt Guard and Code Shield are also available if relevant to the application.
Capability evaluations measure vulnerabilities of Llama models inherent to specific capabilities, for which were crafted dedicated benchmarks including long context, multilingual, tools calls, coding or memorization.
**Red teaming**
For both scenarios, we conducted recurring red teaming exercises with the goal of discovering risks via adversarial prompting and we used the learnings to improve our benchmarks and safety tuning datasets.
We partnered early with subject-matter experts in critical risk areas to understand the nature of these real-world harms and how such models may lead to unintended harm for society. Based on these conversations, we derived a set of adversarial goals for the red team to attempt to achieve, such as extracting harmful information or reprogramming the model to act in a potentially harmful capacity. The red team consisted of experts in cybersecurity, adversarial machine learning, responsible AI, and integrity in addition to multilingual content specialists with background in integrity issues in specific geographic markets. .
### Critical and other risks
### We specifically focused our efforts on mitigating the following critical risk areas:
**1- CBRNE (Chemical, Biological, Radiological, Nuclear, and Explosive materials) helpfulness**
To assess risks related to proliferation of chemical and biological weapons of the Llama 3 family of models, we performed uplift testing designed to assess whether use of the Llama 3 models could meaningfully increase the capabilities of malicious actors to plan or carry out attacks using these types of weapons.
### **2\. Child Safety**
Child Safety risk assessments were conducted using a team of experts, to assess the model’s capability to produce outputs that could result in Child Safety risks and inform on any necessary and appropriate risk mitigations via fine tuning. We leveraged those expert red teaming sessions to expand the coverage of our evaluation benchmarks through Llama 3 model development. For Llama 3, we conducted new in-depth sessions using objective based methodologies to assess the model risks along multiple attack vectors including the additional languages Llama 3 is trained on. We also partnered with content specialists to perform red teaming exercises assessing potentially violating content while taking account of market specific nuances or experiences.
**3\. Cyber attack enablement**
Our cyber attack uplift study investigated whether the Llama 3 family of LLMs can enhance human capabilities in hacking tasks, both in terms of skill level and speed.
Our attack automation study focused on evaluating the capabilities of LLMs when used as autonomous agents in cyber offensive operations, specifically in the context of ransomware attacks. This evaluation was distinct from previous studies that considered LLMs as interactive assistants. The primary objective was to assess whether these models could effectively function as independent agents in executing complex cyber-attacks without human intervention.
### Community
Generative AI safety requires expertise and tooling, and we believe in the strength of the open community to accelerate its progress. We are active members of open consortiums, including the AI Alliance, Partnership on AI and MLCommons, actively contributing to safety standardization and transparency. We encourage the community to adopt taxonomies like the MLCommons Proof of Concept evaluation to facilitate collaboration and transparency on safety and content evaluations. Our Purple Llama tools are open sourced for the community to use and widely distributed across ecosystem partners including cloud service providers. We encourage community contributions to our [Github repository](https://github.com/meta-llama/PurpleLlama).
We also set up the [Llama Impact Grants](https://llama.meta.com/llama-impact-grants/) program to identify and support the most compelling applications of Meta’s Llama model for societal benefit across three categories: education, climate and open innovation. The 20 finalists from the hundreds of applications can be found [here](https://llama.meta.com/llama-impact-grants/#finalists).
Finally, we put in place a set of resources including an [output reporting mechanism](https://developers.facebook.com/llama_output_feedback) and [bug bounty program](https://www.facebook.com/whitehat) to continuously improve the Llama technology with the help of the community.
## Ethical Considerations and Limitations
The core values of Llama 3.3 are openness, inclusivity and helpfulness. It is meant to serve everyone, and to work for a wide range of use cases. It is thus designed to be accessible to people across many different backgrounds, experiences and perspectives. Llama 3.3 addresses users and their needs as they are, without insertion unnecessary judgment or normativity, while reflecting the understanding that even content that may appear problematic in some cases can serve valuable purposes in others. It respects the dignity and autonomy of all users, especially in terms of the values of free thought and expression that power innovation and progress.
|
Mungert/OlympicCoder-32B-GGUF | Mungert | 2025-06-15T19:43:22Z | 318 | 5 | transformers | [
"transformers",
"gguf",
"text-generation",
"en",
"dataset:open-r1/codeforces-cots",
"base_model:Qwen/Qwen2.5-Coder-32B-Instruct",
"base_model:quantized:Qwen/Qwen2.5-Coder-32B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-31T05:55:48Z | ---
license: apache-2.0
datasets:
- open-r1/codeforces-cots
language:
- en
base_model:
- Qwen/Qwen2.5-Coder-32B-Instruct
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">OlympicCoder-32B GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `OlympicCoder-32B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `OlympicCoder-32B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `OlympicCoder-32B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `OlympicCoder-32B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `OlympicCoder-32B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `OlympicCoder-32B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `OlympicCoder-32B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `OlympicCoder-32B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `OlympicCoder-32B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `OlympicCoder-32B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `OlympicCoder-32B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Model Card for OlympicCoder-32B
OlympicCoder-32B is a code model that achieves very strong performance on competitive coding benchmarks such as LiveCodeBench andthe 2024 International Olympiad in Informatics.
* Repository: https://github.com/huggingface/open-r1
* Blog post: https://huggingface.co/blog/open-r1/update-3
## Model description
- **Model type:** A 32B parameter model fine-tuned on a decontaminated version of the codeforces dataset.
- **Language(s) (NLP):** Primarily English
- **License:** apache-2.0
- **Finetuned from model:** [Qwen/Qwen2.5-Coder-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct)
## Evaluation
We compare the performance of OlympicCoder models on two main benchmarks for competitive coding:
* **[IOI'2024:](https://github.com/huggingface/ioi)** 6 very challenging problems from the 2024 International Olympiad in Informatics. Models are allowed up to 50 submissions per problem.
* **[LiveCodeBench:](https://livecodebench.github.io)** Python programming problems source from platforms like CodeForces and LeetCoder. We use the `v4_v5` subset of [`livecodebench/code_generation_lite`](https://huggingface.co/datasets/livecodebench/code_generation_lite), which corresponds to 268 problems. We use `lighteval` to evaluate models on LiveCodeBench using the sampling parameters described [here](https://github.com/huggingface/open-r1?tab=readme-ov-file#livecodebench).
> [!NOTE]
> The OlympicCoder models were post-trained exclusively on C++ solutions generated by DeepSeek-R1. As a result the performance on LiveCodeBench should be considered to be partially _out-of-domain_, since this expects models to output solutions in Python.
### IOI'24
### LiveCodeBench
## Usage
Here's how you can run the model using the `pipeline()` function from 🤗 Transformers:
```python
# pip install transformers
# pip install accelerate
import torch
from transformers import pipeline
pipe = pipeline("text-generation", model="open-r1/OlympicCoder-32B", torch_dtype=torch.bfloat16, device_map="auto")
# We use the tokenizer's chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating
messages = [
{"role": "user", "content": "Write a python program to calculate the 10th Fibonacci number"},
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt, max_new_tokens=8000, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
#<|im_start|>user
#Write a python program to calculate the 10th fibonacci number<|im_end|>
#<|im_start|>assistant
#<think>Okay, I need to write a Python program that calculates the 10th Fibonacci number. Hmm, the Fibonacci sequence starts with 0 and 1. Each subsequent number is the sum of the two preceding ones. So the sequence goes: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, and so on. ...
```
> [!IMPORTANT]
> To ensure that the model consistently outputs a long chain-of-thought, we have edited the chat template to prefill the first assistant turn with a `<think>` token. As a result, the outputs from this model will not show the opening `<think>` token if you use the model's `generate()` method. To apply reinforcement learning with a format reward, either prepend the `<think>` token to the model's completions or amend the chat template to remove the prefill. Check out our [blog post](https://huggingface.co/blog/open-r1/update-3#lesson-4-prefill-with-think-to-consistently-enable-long-cot) for more details.
## Training procedure
### Training hyper-parameters
The following hyperparameters were used during training on 16 H100 nodes:
- dataset: open-r1/codeforces-cots_decontaminated
- learning_rate: 4.0e-5
- train_batch_size: 1
- seed: 42
- packing: false
- distributed_type: fsdp
- num_devices: 128
- gradient_accumulation_steps: 1
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine_with_min_lr
- min_lr_rate: 0.1
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 10.0 |
Mungert/Qwen2.5-VL-72B-Instruct-GGUF | Mungert | 2025-06-15T19:43:19Z | 6,745 | 11 | transformers | [
"transformers",
"gguf",
"multimodal",
"image-text-to-text",
"en",
"arxiv:2309.00071",
"arxiv:2409.12191",
"arxiv:2308.12966",
"base_model:Qwen/Qwen2.5-VL-72B-Instruct",
"base_model:quantized:Qwen/Qwen2.5-VL-72B-Instruct",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | image-text-to-text | 2025-03-29T21:50:25Z |
---
license: other
license_name: qwen
license_link: https://huggingface.co/Qwen/Qwen2.5-VL-72B-Instruct/blob/main/LICENSE
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
library_name: transformers
base_model:
- Qwen/Qwen2.5-VL-72B-Instruct
---
# <span style="color: #7FFF7F;">Qwen2.5-VL-72B-Instruct GGUF Models</span>
## How to Use Qwen 2.5 VL Instruct with llama.cpp (latest as of 10th May 2025)
1. **Download the Qwen 2.5 VL gguf file**:
https://huggingface.co/Mungert/Qwen2.5-VL-72B-Instruct-GGUF/tree/main
Choose a gguf file without the mmproj in the name
Example gguf file : https://huggingface.co/Mungert/Mungert/Qwen2.5-VL-72B-Instruct-GGUF/resolve/main/Qwen2.5-VL-72B-Instruct-q8_0.gguf
Copy this file to your chosen folder.
2. **Download the Qwen 2.5 VL mmproj file**
https://huggingface.co/Mungert/Qwen2.5-VL-72B-Instruct-GGUF/tree/main
Choose a file with mmproj in the name
Example mmproj file : https://huggingface.co/Mungert/Qwen2.5-VL-72B-Instruct-GGUF/resolve/main/Qwen2.5-VL-72B-Instruct-mmproj-f16.gguf
Copy this file to your chosen folder.
3. Copy images to the same folder as the gguf files or alter paths appropriately.
In the example below the gguf files, images and llama-mtmd-cli are in the same folder.
Example image: image https://huggingface.co/Mungert/Qwen2.5-VL-72B-Instruct-GGUF/resolve/main/car-1.jpg
Copy this file to your chosen folder.
4. **Run the CLI Tool**:
From your chosen folder :
```bash
llama-mtmd-cli -m Qwen2.5-VL-72B-Instruct-q8_0.gguf --mmproj Qwen2.5-VL-72B-Instruct-mmproj-f16.gguf -p "Describe this image." --image ./car-1.jpg
```
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen2.5-VL-72B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen2.5-VL-72B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen2.5-VL-72B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen2.5-VL-72B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen2.5-VL-72B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen2.5-VL-72B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen2.5-VL-72B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen2.5-VL-72B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen2.5-VL-72B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen2.5-VL-72B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen2.5-VL-72B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I'd really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# <span style="color: #7FFF7F;">Mungert/Qwen2.5-VL-72B-Instruct-GGUF GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Mungert/Qwen2.5-VL-72B-Instruct-GGUF-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I'd really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen2.5-VL-72B-Instruct
<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Introduction
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models. Today, we are excited to introduce the latest addition to the Qwen family: Qwen2.5-VL.
#### Key Enhancements:
* **Understand things visually**: Qwen2.5-VL is not only proficient in recognizing common objects such as flowers, birds, fish, and insects, but it is highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
* **Being agentic**: Qwen2.5-VL directly plays as a visual agent that can reason and dynamically direct tools, which is capable of computer use and phone use.
* **Understanding long videos and capturing events**: Qwen2.5-VL can comprehend videos of over 1 hour, and this time it has a new ability of cpaturing event by pinpointing the relevant video segments.
* **Capable of visual localization in different formats**: Qwen2.5-VL can accurately localize objects in an image by generating bounding boxes or points, and it can provide stable JSON outputs for coordinates and attributes.
* **Generating structured outputs**: for data like scans of invoices, forms, tables, etc. Qwen2.5-VL supports structured outputs of their contents, benefiting usages in finance, commerce, etc.
#### Model Architecture Updates:
* **Dynamic Resolution and Frame Rate Training for Video Understanding**:
We extend dynamic resolution to the temporal dimension by adopting dynamic FPS sampling, enabling the model to comprehend videos at various sampling rates. Accordingly, we update mRoPE in the time dimension with IDs and absolute time alignment, enabling the model to learn temporal sequence and speed, and ultimately acquire the ability to pinpoint specific moments.
<p align="center">
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-VL/qwen2.5vl_arc.jpeg" width="80%"/>
<p>
* **Streamlined and Efficient Vision Encoder**
We enhance both training and inference speeds by strategically implementing window attention into the ViT. The ViT architecture is further optimized with SwiGLU and RMSNorm, aligning it with the structure of the Qwen2.5 LLM.
We have three models with 3, 7 and 72 billion parameters. This repo contains the instruction-tuned 72B Qwen2.5-VL model. For more information, visit our [Blog](https://qwenlm.github.io/blog/qwen2.5-vl/) and [GitHub](https://github.com/QwenLM/Qwen2.5-VL).
## Evaluation
### Image benchmark
| Benchmarks | GPT4o | Claude3.5 Sonnet | Gemini-2-flash | InternVL2.5-78B | Qwen2-VL-72B | Qwen2.5-VL-72B |
|-----------------------|-----------|-------------------|-----------------|-----------------|--------------|----------------|
| MMMU<sub>val</sub> | 70.3 | 70.4 | 70.7 | 70.1 | 64.5 | 70.2 |
| MMMU_Pro | 54.5 | 54.7 | 57.0 | 48.6 | 46.2 | 51.1 |
| MathVista_MINI | 63.8 | 65.4 | 73.1 | 76.6 | 70.5 | 74.8 |
| MathVision_FULL | 30.4 | 38.3 | 41.3 | 32.2 | 25.9 | 38.1 |
| Hallusion Bench | 55.0 | 55.16 | | 57.4 | 58.1 | 55.16 |
| MMBench_DEV_EN_V11 | 82.1 | 83.4 | 83.0 | 88.5 | 86.6 | 88 |
| AI2D_TEST | 84.6 | 81.2 | | 89.1 | 88.1 | 88.4 |
| ChartQA_TEST | 86.7 | 90.8 | 85.2 | 88.3 | 88.3 | 89.5 |
| DocVQA_VAL | 91.1 | 95.2 | 92.1 | 96.5 | 96.1 | 96.4 |
| MMStar | 64.7 | 65.1 | 69.4 | 69.5 | 68.3 | 70.8 |
| MMVet_turbo | 69.1 | 70.1 | | 72.3 | 74.0 | 76.19 |
| OCRBench | 736 | 788 | | 854 | 877 | 885 |
| OCRBench-V2(en/zh) | 46.5/32.3 | 45.2/39.6 | 51.9/43.1 | 45/46.2 | 47.8/46.1 | 61.5/63.7 |
| CC-OCR | 66.6 | 62.7 | 73.0 | 64.7 | 68.7 |79.8 |
### Video benchmark
| Benchmarks | GPT4o | Gemini-1.5-Pro | InternVL2.5-78B | Qwen2VL-72B | Qwen2.5VL-72B |
|---------------------|-------|----------------|-----------------|-------------|---------------|
| VideoMME w/o sub. | 71.9 | 75.0 | 72.1 | 71.2 | 73.3 |
| VideoMME w sub. | 77.2 | 81.3 | 74.0 | 77.8 | 79.1 |
| MVBench | 64.6 | 60.5 | 76.4 | 73.6 | 70.4 |
| MMBench-Video | 1.63 | 1.30 | 1.97 | 1.70 | 2.02 |
| LVBench | 30.8 | 33.1 | - | 41.3 | 47.3 |
| EgoSchema | 72.2 | 71.2 | - | 77.9 | 76.2 |
| PerceptionTest_test | - | - | - | 68.0 | 73.2 |
| MLVU_M-Avg_dev | 64.6 | - | 75.7 | | 74.6 |
| TempCompass_overall | 73.8 | - | - | | 74.8 |
### Agent benchmark
| Benchmarks | GPT4o | Gemini 2.0 | Claude | Aguvis-72B | Qwen2VL-72B | Qwen2.5VL-72B |
|-------------------------|-------------|------------|--------|------------|-------------|---------------|
| ScreenSpot | 18.1 | 84.0 | 83.0 | | | 87.1 |
| ScreenSpot Pro | | | 17.1 | | 1.6 | 43.6 |
| AITZ_EM | 35.3 | | | | 72.8 | 83.2 |
| Android Control High_EM | | | | 66.4 | 59.1 | 67.36 |
| Android Control Low_EM | | | | 84.4 | 59.2 | 93.7 |
| AndroidWorld_SR | 34.5% (SoM) | | 27.9% | 26.1% | | 35% |
| MobileMiniWob++_SR | | | | 66% | | 68% |
| OSWorld | | | 14.90 | 10.26 | | 8.83 |
## Requirements
The code of Qwen2.5-VL has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip install git+https://github.com/huggingface/transformers accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_vl'
```
## Quickstart
Below, we provide simple examples to show how to use Qwen2.5-VL with 🤖 ModelScope and 🤗 Transformers.
The code of Qwen2.5-VL has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip install git+https://github.com/huggingface/transformers accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_vl'
```
We offer a toolkit to help you handle various types of visual input more conveniently, as if you were using an API. This includes base64, URLs, and interleaved images and videos. You can install it using the following command:
```bash
# It's highly recommanded to use `[decord]` feature for faster video loading.
pip install qwen-vl-utils[decord]==0.0.8
```
If you are not using Linux, you might not be able to install `decord` from PyPI. In that case, you can use `pip install qwen-vl-utils` which will fall back to using torchvision for video processing. However, you can still [install decord from source](https://github.com/dmlc/decord?tab=readme-ov-file#install-from-source) to get decord used when loading video.
### Using 🤗 Transformers to Chat
Here we show a code snippet to show you how to use the chat model with `transformers` and `qwen_vl_utils`:
```python
from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info
# default: Load the model on the available device(s)
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-VL-72B-Instruct", torch_dtype="auto", device_map="auto"
)
# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
# "Qwen/Qwen2.5-VL-72B-Instruct",
# torch_dtype=torch.bfloat16,
# attn_implementation="flash_attention_2",
# device_map="auto",
# )
# default processer
processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-72B-Instruct")
# The default range for the number of visual tokens per image in the model is 4-16384.
# You can set min_pixels and max_pixels according to your needs, such as a token range of 256-1280, to balance performance and cost.
# min_pixels = 256*28*28
# max_pixels = 1280*28*28
# processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-72B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels)
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
},
{"type": "text", "text": "Describe this image."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
<details>
<summary>Multi image inference</summary>
```python
# Messages containing multiple images and a text query
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/image1.jpg"},
{"type": "image", "image": "file:///path/to/image2.jpg"},
{"type": "text", "text": "Identify the similarities between these images."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
</details>
<details>
<summary>Video inference</summary>
```python
# Messages containing a images list as a video and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": [
"file:///path/to/frame1.jpg",
"file:///path/to/frame2.jpg",
"file:///path/to/frame3.jpg",
"file:///path/to/frame4.jpg",
],
},
{"type": "text", "text": "Describe this video."},
],
}
]
# Messages containing a local video path and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": "file:///path/to/video1.mp4",
"max_pixels": 360 * 420,
"fps": 1.0,
},
{"type": "text", "text": "Describe this video."},
],
}
]
# Messages containing a video url and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2-VL/space_woaudio.mp4",
},
{"type": "text", "text": "Describe this video."},
],
}
]
#In Qwen 2.5 VL, frame rate information is also input into the model to align with absolute time.
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs, video_kwargs = process_vision_info(messages, return_video_kwargs=True)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
fps=fps,
padding=True,
return_tensors="pt",
**video_kwargs,
)
inputs = inputs.to("cuda")
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
Video URL compatibility largely depends on the third-party library version. The details are in the table below. change the backend by `FORCE_QWENVL_VIDEO_READER=torchvision` or `FORCE_QWENVL_VIDEO_READER=decord` if you prefer not to use the default one.
| Backend | HTTP | HTTPS |
|-------------|------|-------|
| torchvision >= 0.19.0 | ✅ | ✅ |
| torchvision < 0.19.0 | ❌ | ❌ |
| decord | ✅ | ❌ |
</details>
<details>
<summary>Batch inference</summary>
```python
# Sample messages for batch inference
messages1 = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/image1.jpg"},
{"type": "image", "image": "file:///path/to/image2.jpg"},
{"type": "text", "text": "What are the common elements in these pictures?"},
],
}
]
messages2 = [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Who are you?"},
]
# Combine messages for batch processing
messages = [messages1, messages2]
# Preparation for batch inference
texts = [
processor.apply_chat_template(msg, tokenize=False, add_generation_prompt=True)
for msg in messages
]
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=texts,
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Batch Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_texts = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_texts)
```
</details>
### 🤖 ModelScope
We strongly advise users especially those in mainland China to use ModelScope. `snapshot_download` can help you solve issues concerning downloading checkpoints.
### More Usage Tips
For input images, we support local files, base64, and URLs. For videos, we currently only support local files.
```python
# You can directly insert a local file path, a URL, or a base64-encoded image into the position where you want in the text.
## Local file path
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/your/image.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
## Image URL
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "http://path/to/your/image.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
## Base64 encoded image
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "data:image;base64,/9j/..."},
{"type": "text", "text": "Describe this image."},
],
}
]
```
#### Image Resolution for performance boost
The model supports a wide range of resolution inputs. By default, it uses the native resolution for input, but higher resolutions can enhance performance at the cost of more computation. Users can set the minimum and maximum number of pixels to achieve an optimal configuration for their needs, such as a token count range of 256-1280, to balance speed and memory usage.
```python
min_pixels = 256 * 28 * 28
max_pixels = 1280 * 28 * 28
processor = AutoProcessor.from_pretrained(
"Qwen/Qwen2.5-VL-72B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels
)
```
Besides, We provide two methods for fine-grained control over the image size input to the model:
1. Define min_pixels and max_pixels: Images will be resized to maintain their aspect ratio within the range of min_pixels and max_pixels.
2. Specify exact dimensions: Directly set `resized_height` and `resized_width`. These values will be rounded to the nearest multiple of 28.
```python
# min_pixels and max_pixels
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "file:///path/to/your/image.jpg",
"resized_height": 280,
"resized_width": 420,
},
{"type": "text", "text": "Describe this image."},
],
}
]
# resized_height and resized_width
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "file:///path/to/your/image.jpg",
"min_pixels": 50176,
"max_pixels": 50176,
},
{"type": "text", "text": "Describe this image."},
],
}
]
```
### Processing Long Texts
The current `config.json` is set for context length up to 32,768 tokens.
To handle extensive inputs exceeding 32,768 tokens, we utilize [YaRN](https://arxiv.org/abs/2309.00071), a technique for enhancing model length extrapolation, ensuring optimal performance on lengthy texts.
For supported frameworks, you could add the following to `config.json` to enable YaRN:
```json
{
...,
"type": "yarn",
"mrope_section": [
16,
24,
24
],
"factor": 4,
"original_max_position_embeddings": 32768
}
```
However, it should be noted that this method has a significant impact on the performance of temporal and spatial localization tasks, and is therefore not recommended for use.
At the same time, for long video inputs, since MRoPE itself is more economical with ids, the max_position_embeddings can be directly modified to a larger value, such as 64k.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen2.5-VL,
title = {Qwen2.5-VL},
url = {https://qwenlm.github.io/blog/qwen2.5-vl/},
author = {Qwen Team},
month = {January},
year = {2025}
}
@article{Qwen2VL,
title={Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution},
author={Wang, Peng and Bai, Shuai and Tan, Sinan and Wang, Shijie and Fan, Zhihao and Bai, Jinze and Chen, Keqin and Liu, Xuejing and Wang, Jialin and Ge, Wenbin and Fan, Yang and Dang, Kai and Du, Mengfei and Ren, Xuancheng and Men, Rui and Liu, Dayiheng and Zhou, Chang and Zhou, Jingren and Lin, Junyang},
journal={arXiv preprint arXiv:2409.12191},
year={2024}
}
@article{Qwen-VL,
title={Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond},
author={Bai, Jinze and Bai, Shuai and Yang, Shusheng and Wang, Shijie and Tan, Sinan and Wang, Peng and Lin, Junyang and Zhou, Chang and Zhou, Jingren},
journal={arXiv preprint arXiv:2308.12966},
year={2023}
}
```
|
Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF | Mungert | 2025-06-15T19:43:11Z | 992 | 4 | transformers | [
"transformers",
"gguf",
"nvidia",
"llama-3",
"pytorch",
"text-generation",
"en",
"arxiv:2411.19146",
"arxiv:2502.00203",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-29T03:22:36Z | ---
library_name: transformers
license: other
license_name: nvidia-open-model-license
license_link: >-
https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/
pipeline_tag: text-generation
language:
- en
tags:
- nvidia
- llama-3
- pytorch
---
# <span style="color: #7FFF7F;">Llama-3_3-Nemotron-Super-49B-v1 GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3_3-Nemotron-Super-49B-v1-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3_3-Nemotron-Super-49B-v1-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3_3-Nemotron-Super-49B-v1-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3_3-Nemotron-Super-49B-v1-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3_3-Nemotron-Super-49B-v1-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3_3-Nemotron-Super-49B-v1-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3_3-Nemotron-Super-49B-v1-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3_3-Nemotron-Super-49B-v1-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3_3-Nemotron-Super-49B-v1-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3_3-Nemotron-Super-49B-v1-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3_3-Nemotron-Super-49B-v1-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I'd really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# <span style="color: #7FFF7F;">Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Mungert/Llama-3_3-Nemotron-Super-49B-v1-GGUF-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I'd really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Llama-3.3-Nemotron-Super-49B-v1
## Model Overview
Llama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) which is a derivative of [Meta Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct) (AKA the *reference model*). It is a reasoning model that is post trained for reasoning, human chat preferences, and tasks, such as RAG and tool calling. The model supports a context length of 128K tokens.
Llama-3.3-Nemotron-Super-49B-v1 is a model which offers a great tradeoff between model accuracy and efficiency. Efficiency (throughput) directly translates to savings. Using a novel Neural Architecture Search (NAS) approach, we greatly reduce the model’s memory footprint, enabling larger workloads, as well as fitting the model on a single GPU at high workloads (H200). This NAS approach enables the selection of a desired point in the accuracy-efficiency tradeoff. For more information on the NAS approach, please refer to [this paper](https://arxiv.org/abs/2411.19146).
The model underwent a multi-phase post-training process to enhance both its reasoning and non-reasoning capabilities. This includes a supervised fine-tuning stage for Math, Code, Reasoning, and Tool Calling as well as multiple reinforcement learning (RL) stages using REINFORCE (RLOO) and Online Reward-aware Preference Optimization (RPO) algorithms for both chat and instruction-following. The final model checkpoint is obtained after merging the final SFT and Online RPO checkpoints. For more details on how the model was trained, please see [this blog](https://developer.nvidia.com/blog/build-enterprise-ai-agents-with-advanced-open-nvidia-llama-nemotron-reasoning-models/).
This model is part of the Llama Nemotron Collection. You can find the other model(s) in this family here:
- [Llama-3.1-Nemotron-Nano-8B-v1](https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-8B-v1)
This model is ready for commercial use.
## License/Terms of Use
GOVERNING TERMS: Your use of this model is governed by the [NVIDIA Open Model License.](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/) \
Additional Information: [Llama 3.3 Community License Agreement](https://www.llama.com/llama3_3/license/). Built with Llama.
**Model Developer:** NVIDIA
**Model Dates:** Trained between November 2024 and February 2025
**Data Freshness:** The pretraining data has a cutoff of 2023 per Meta Llama 3.3 70B
### Use Case: <br>
Developers designing AI Agent systems, chatbots, RAG systems, and other AI-powered applications. Also suitable for typical instruction-following tasks. <br>
### Release Date: <br>
3/18/2025 <br>
## References
* [[2411.19146] Puzzle: Distillation-Based NAS for Inference-Optimized LLMs](https://arxiv.org/abs/2411.19146)
* [[2502.00203] Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment](https://arxiv.org/abs/2502.00203)
## Model Architecture
**Architecture Type:** Dense decoder-only Transformer model \
**Network Architecture:** Llama 3.3 70B Instruct, customized through Neural Architecture Search (NAS)
The model is a derivative of Meta’s Llama-3.3-70B-Instruct, using Neural Architecture Search (NAS). The NAS algorithm results in non-standard and non-repetitive blocks. This includes the following:
* Skip attention: In some blocks, the attention is skipped entirely, or replaced with a single linear layer.
* Variable FFN: The expansion/compression ratio in the FFN layer is different between blocks.
We utilize a block-wise distillation of the reference model, where for each block we create multiple variants providing different tradeoffs of quality vs. computational complexity, discussed in more depth below. We then search over the blocks to create a model which meets the required throughput and memory (optimized for a single H100-80GB GPU) while minimizing the quality degradation. The model then undergoes knowledge distillation (KD), with a focus on English single and multi-turn chat use-cases. The KD step included 40 billion tokens consisting of a mixture of 3 datasets - FineWeb, Buzz-V1.2 and Dolma.
## Intended use
Llama-3.3-Nemotron-Super-49B-v1 is a general purpose reasoning and chat model intended to be used in English and coding languages. Other non-English languages (German, French, Italian, Portuguese, Hindi, Spanish, and Thai) are also supported.
## Input
- **Input Type:** Text
- **Input Format:** String
- **Input Parameters:** One-Dimensional (1D)
- **Other Properties Related to Input:** Context length up to 131,072 tokens
## Output
- **Output Type:** Text
- **Output Format:** String
- **Output Parameters:** One-Dimensional (1D)
- **Other Properties Related to Output:** Context length up to 131,072 tokens
## Model Version
1.0 (3/18/2025)
## Software Integration
- **Runtime Engine:** Transformers
- **Recommended Hardware Microarchitecture Compatibility:**
- NVIDIA Hopper
- NVIDIA Ampere
## Quick Start and Usage Recommendations:
1. Reasoning mode (ON/OFF) is controlled via the system prompt, which must be set as shown in the example below. All instructions should be contained within the user prompt
2. We recommend setting temperature to `0.6`, and Top P to `0.95` for Reasoning ON mode
3. We recommend using greedy decoding for Reasoning OFF mode
4. We have provided a list of prompts to use for evaluation for each benchmark where a specific template is required
You can try this model out through the preview API, using this link: [Llama-3_3-Nemotron-Super-49B-v1](https://build.nvidia.com/nvidia/llama-3_3-nemotron-super-49b-v1).
See the snippet below for usage with [Hugging Face Transformers](https://huggingface.co/docs/transformers/main/en/index) library. Reasoning mode (ON/OFF) is controlled via system prompt. Please see the example below
We recommend using the *transformers* package with version 4.48.3.
Example of reasoning on:
```py
import torch
import transformers
model_id = "nvidia/Llama-3_3-Nemotron-Super-49B-v1"
model_kwargs = {"torch_dtype": torch.bfloat16, "trust_remote_code": True, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
temperature=0.6,
top_p=0.95,
**model_kwargs
)
thinking = "on"
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"},{"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))
```
Example of reasoning off:
```py
import torch
import transformers
model_id = "nvidia/Llama-3_3-Nemotron-Super-49B-v1"
model_kwargs = {"torch_dtype": torch.bfloat16, "trust_remote_code": True, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
do_sample=False,
**model_kwargs
)
# Thinking can be "on" or "off"
thinking = "off"
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"},{"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))
```
## Inference:
**Engine:**
- Transformers
**Test Hardware:**
- FP8: 1x NVIDIA H100-80GB GPU (Coming Soon!)
- BF16:
- 2x NVIDIA H100-80GB
- 2x NVIDIA A100-80GB GPUs
**[Preferred/Supported] Operating System(s):** Linux <br>
## Training Datasets
A large variety of training data was used for the knowledge distillation phase before post-training pipeline, 3 of which included: FineWeb, Buzz-V1.2, and Dolma.
The data for the multi-stage post-training phases for improvements in Code, Math, and Reasoning is a compilation of SFT and RL data that supports improvements of math, code, general reasoning, and instruction following capabilities of the original Llama instruct model.
In conjunction with this model release, NVIDIA has released 30M samples of post-training data, as public and permissive. Please see [Llama-Nemotron-Postraining-Dataset-v1](https://huggingface.co/datasets/nvidia/Llama-Nemotron-Post-Training-Dataset-v1).
Distribution of the domains is as follows:
| Category | Value |
|----------|-----------|
| math | 19,840,970|
| code | 9,612,677 |
| science | 708,920 |
| instruction following | 56,339 |
| chat | 39,792 |
| safety | 31,426 |
Prompts have been sourced from either public and open corpus or synthetically generated. Responses were synthetically generated by a variety of models, with some prompts containing responses for both reasoning on and off modes, to train the model to distinguish between two modes.
**Data Collection for Training Datasets:**
- Hybrid: Automated, Human, Synthetic
**Data Labeling for Training Datasets:**
- Hybrid: Automated, Human, Synthetic
## Evaluation Datasets
We used the datasets listed below to evaluate Llama-3.3-Nemotron-Super-49B-v1.
Data Collection for Evaluation Datasets:
- Hybrid: Human/Synthetic
Data Labeling for Evaluation Datasets:
- Hybrid: Human/Synthetic/Automatic
## Evaluation Results
These results contain both “Reasoning On”, and “Reasoning Off”. We recommend using temperature=`0.6`, top_p=`0.95` for “Reasoning On” mode, and greedy decoding for “Reasoning Off” mode. All evaluations are done with 32k sequence length. We run the benchmarks up to 16 times and average the scores to be more accurate.
> NOTE: Where applicable, a Prompt Template will be provided. While completing benchmarks, please ensure that you are parsing for the correct output format as per the provided prompt in order to reproduce the benchmarks seen below.
### Arena-Hard
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 88.3 |
### MATH500
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 74.0 |
| Reasoning On | 96.6 |
User Prompt Template:
```
"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \boxed{}.\nQuestion: {question}"
```
### AIME25
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 13.33 |
| Reasoning On | 58.4 |
User Prompt Template:
```
"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \boxed{}.\nQuestion: {question}"
```
### GPQA
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 50 |
| Reasoning On | 66.67 |
User Prompt Template:
```
"What is the correct answer to this question: {question}\nChoices:\nA. {option_A}\nB. {option_B}\nC. {option_C}\nD. {option_D}\nLet's think step by step, and put the final answer (should be a single letter A, B, C, or D) into a \boxed{}"
```
### IFEval
| Reasoning Mode | Strict:Instruction |
|--------------|------------|
| Reasoning Off | 89.21 |
### BFCL V2 Live
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 73.7 |
User Prompt Template:
```
You are an expert in composing functions. You are given a question and a set of possible functions.
Based on the question, you will need to make one or more function/tool calls to achieve the purpose.
If none of the function can be used, point it out. If the given question lacks the parameters required by the function,
also point it out. You should only return the function call in tools call sections.
If you decide to invoke any of the function(s), you MUST put it in the format of <TOOLCALL>[func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]</TOOLCALL>
You SHOULD NOT include any other text in the response.
Here is a list of functions in JSON format that you can invoke.
<AVAILABLE_TOOLS>{functions}</AVAILABLE_TOOLS>
{user_prompt}
```
### MBPP 0-shot
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 84.9|
| Reasoning On | 91.3 |
User Prompt Template:
````
You are an exceptionally intelligent coding assistant that consistently delivers accurate and reliable responses to user instructions.
@@ Instruction
Here is the given problem and test examples:
{prompt}
Please use the python programming language to solve this problem.
Please make sure that your code includes the functions from the test samples and that the input and output formats of these functions match the test samples.
Please return all completed codes in one code block.
This code block should be in the following format:
```python
# Your codes here
```
````
### MT-Bench
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 9.17 |
## Ethical Considerations:
NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
For more detailed information on ethical considerations for this model, please see the Model Card++ [Explainability](explainability.md), [Bias](bias.md), [Safety & Security](safety.md), and [Privacy](privacy.md) Subcards.
Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).
|
Mungert/PLM-1.8B-Instruct-GGUF | Mungert | 2025-06-15T19:43:05Z | 346 | 1 | transformers | [
"transformers",
"gguf",
"text-generation",
"en",
"zh",
"arxiv:2503.12167",
"base_model:PLM-Team/PLM-1.8B-Instruct",
"base_model:quantized:PLM-Team/PLM-1.8B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-28T20:08:40Z | ---
base_model: PLM-Team/PLM-1.8B-Instruct
language:
- en
- zh
library_name: transformers
license: apache-2.0
quantized_by: PLM-Team
pipeline_tag: text-generation
---
# <span style="color: #7FFF7F;">PLM-1.8B-Instruct GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `PLM-1.8B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `PLM-1.8B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `PLM-1.8B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `PLM-1.8B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `PLM-1.8B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `PLM-1.8B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `PLM-1.8B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `PLM-1.8B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `PLM-1.8B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `PLM-1.8B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `PLM-1.8B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
<center>
<img src="https://www.cdeng.net/plm/plm_logo.png" alt="plm-logo" width="200"/>
<h2>🖲️ PLM: Efficient Peripheral Language Models Hardware-Co-Designed for Ubiquitous Computing</h2>
<a href='https://www.project-plm.com/'>👉 Project PLM Website</a>
</center>
<center>
||||||||
|:-:|:-:|:-:|:-:|:-:|:-:|:-:|
|<a href='https://arxiv.org/abs/2503.12167'><img src='https://img.shields.io/badge/Paper-ArXiv-C71585'></a>|<a href='https://huggingface.co/PLM-Team/PLM-1.8B-Base'><img src='https://img.shields.io/badge/%F0%9F%A4%97%20Hugging Face-Base-red'></a>|<a href='https://huggingface.co/PLM-Team/PLM-1.8B-Instruct'><img src='https://img.shields.io/badge/%F0%9F%A4%97%20Hugging Face-Instruct-red'></a>|<a href='https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf'><img src='https://img.shields.io/badge/%F0%9F%A4%97%20Hugging Face-gguf-red'></a>|<a href='https://huggingface.co/datasets/plm-team/scots'><img src='https://img.shields.io/badge/Data-plm%20mix-4169E1'></img></a>|<a><img src="https://img.shields.io/github/stars/plm-team/PLM"></a>|
</center>
---
The PLM (Peripheral Language Model) series introduces a novel model architecture to peripheral computing by delivering powerful language capabilities within the constraints of resource-limited devices. Through modeling and system co-design strategy, PLM optimizes model performance and fits edge system requirements, PLM employs **Multi-head Latent Attention** and **squared ReLU** activation to achieve sparsity, significantly reducing memory footprint and computational demands. Coupled with a meticulously crafted training regimen using curated datasets and a Warmup-Stable-Decay-Constant learning rate scheduler, PLM demonstrates superior performance compared to existing small language models, all while maintaining the lowest activated parameters, making it ideally suited for deployment on diverse peripheral platforms like mobile phones and Raspberry Pis.
**Here we present the static quants of https://huggingface.co/PLM-Team/PLM-1.8B-Instruct**
## Provided Quants
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-F16.gguf|F16| 3.66GB| Recommanded|
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q2_K.gguf|Q2_K| 827 MB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q3_K_L.gguf|Q3_K_L| 1.09 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q3_K_M.gguf|Q3_K_M| 1.01 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q3_K_S.gguf|Q3_K_S| 912 MB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q4_0.gguf|Q4_0| 1.11 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q4_1.gguf|Q4_1| 1.21 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q4_K_M.gguf|Q4_K_M| 1.18 GB| Recommanded|
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q4_K_S.gguf|Q4_K_S| 1.12 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q5_0.gguf|Q5_0| 1.3 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q5_1.gguf|Q5_1| 1.4 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q5_K_M.gguf|Q5_K_M| 1.34 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q5_K_S.gguf|Q5_K_S| 1.3 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q6_K.gguf|Q6_K| 1.5 GB| |
|https://huggingface.co/PLM-Team/PLM-1.8B-Instruct-gguf/blob/main/PLM-1.8B-Instruct-Q8_0.gguf|Q8_0| 1.95 GB| Recommanded|
## Usage (llama.cpp)
Now [llama.cpp](https://github.com/ggml-org/llama.cpp) supports our model. Here is the usage:
```bash
git clone https://github.com/Si1w/llama.cpp.git
cd llama.cpp
```
If you want to convert the orginal model into `gguf` form by yourself, you can
```bash
pip install -r requirements.txt
python convert_hf_to_ggyf.py [model] --outtype {f32,f16,bf16,q8_0,tq1_0,tq2_0,auto}
```
Then, we can build with CPU of GPU (e.g. Orin). The build is based on `cmake`.
- For CPU
```bash
cmake -B build
cmake --build build --config Release
```
- For GPU
```bash
cmake -B build -DGGML_CUDA=ON
cmake --build build --config Release
```
Don't forget to download the GGUF files of the PLM. We use the quantization methods in `llama.cpp` to generate the quantized PLM.
```bash
huggingface-cli download --resume-download PLM-Team/PLM-1.8B-Instruct-gguf --local-dir PLM-Team/PLM-1.8B-Instruct-gguf
```
After build the `llama.cpp`, we can use `llama-cli` script to launch the PLM.
```bash
./build/bin/llama-cli -m ./PLM-Team/PLM-1.8B-Instruct-gguf/PLM-1.8B-Instruct-Q8_0.gguf -cnv -p "hello!" -n 128
```
## Citation
If you find Project PLM helpful for your research or applications, please cite as follows:
```
@misc{deng2025plmefficientperipherallanguage,
title={PLM: Efficient Peripheral Language Models Hardware-Co-Designed for Ubiquitous Computing},
author={Cheng Deng and Luoyang Sun and Jiwen Jiang and Yongcheng Zeng and Xinjian Wu and Wenxin Zhao and Qingfa Xiao and Jiachuan Wang and Lei Chen and Lionel M. Ni and Haifeng Zhang and Jun Wang},
year={2025},
eprint={2503.12167},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2503.12167},
}
``` |
Mungert/Qwen2.5-VL-32B-Instruct-GGUF | Mungert | 2025-06-15T19:42:51Z | 10,358 | 8 | transformers | [
"transformers",
"gguf",
"multimodal",
"image-text-to-text",
"en",
"arxiv:2309.00071",
"arxiv:2502.13923",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | image-text-to-text | 2025-03-28T04:48:49Z | ---
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
library_name: transformers
---
# <span style="color: #7FFF7F;">Qwen2.5-VL-32B-Instruct GGUF Models</span>
## How to Use Qwen 2.5 VL Instruct with llama.cpp (latest as of 10th May 2025)
1. **Download the Qwen 2.5 VL gguf file**:
https://huggingface.co/Mungert/Qwen2.5-VL-32B-Instruct-GGUF/tree/main
Choose a gguf file without the mmproj in the name
Example gguf file : https://huggingface.co/Mungert/Mungert/Qwen2.5-VL-32B-Instruct-GGUF/resolve/main/Qwen2.5-VL-32B-Instruct-q8_0.gguf
Copy this file to your chosen folder.
2. **Download the Qwen 2.5 VL mmproj file**
https://huggingface.co/Mungert/Qwen2.5-VL-32B-Instruct-GGUF/tree/main
Choose a file with mmproj in the name
Example mmproj file : https://huggingface.co/Mungert/Qwen2.5-VL-32B-Instruct-GGUF/resolve/main/Qwen2.5-VL-32B-Instruct-mmproj-f16.gguf
Copy this file to your chosen folder.
3. Copy images to the same folder as the gguf files or alter paths appropriately.
In the example below the gguf files, images and llama-mtmd-cli are in the same folder.
Example image: image https://huggingface.co/Mungert/Qwen2.5-VL-32B-Instruct-GGUF/resolve/main/car-1.jpg
Copy this file to your chosen folder.
4. **Run the CLI Tool**:
From your chosen folder :
```bash
llama-mtmd-cli -m Qwen2.5-VL-32B-Instruct-q8_0.gguf --mmproj Qwen2.5-VL-32B-Instruct-mmproj-f16.gguf -p "Describe this image." --image ./car-1.jpg
```
## **Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)**
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen2.5-VL-32B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen2.5-VL-32B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen2.5-VL-32B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen2.5-VL-32B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen2.5-VL-32B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen2.5-VL-32B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen2.5-VL-32B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen2.5-VL-32B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen2.5-VL-32B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen2.5-VL-32B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen2.5-VL-32B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen2.5-VL-32B-Instruct
<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Latest Updates:
In addition to the original formula, we have further enhanced Qwen2.5-VL-32B's mathematical and problem-solving abilities through reinforcement learning. This has also significantly improved the model's subjective user experience, with response styles adjusted to better align with human preferences. Particularly for objective queries such as mathematics, logical reasoning, and knowledge-based Q&A, the level of detail in responses and the clarity of formatting have been noticeably enhanced.
## Introduction
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models. Today, we are excited to introduce the latest addition to the Qwen family: Qwen2.5-VL.
#### Key Enhancements:
* **Understand things visually**: Qwen2.5-VL is not only proficient in recognizing common objects such as flowers, birds, fish, and insects, but it is highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
* **Being agentic**: Qwen2.5-VL directly plays as a visual agent that can reason and dynamically direct tools, which is capable of computer use and phone use.
* **Understanding long videos and capturing events**: Qwen2.5-VL can comprehend videos of over 1 hour, and this time it has a new ability of cpaturing event by pinpointing the relevant video segments.
* **Capable of visual localization in different formats**: Qwen2.5-VL can accurately localize objects in an image by generating bounding boxes or points, and it can provide stable JSON outputs for coordinates and attributes.
* **Generating structured outputs**: for data like scans of invoices, forms, tables, etc. Qwen2.5-VL supports structured outputs of their contents, benefiting usages in finance, commerce, etc.
#### Model Architecture Updates:
* **Dynamic Resolution and Frame Rate Training for Video Understanding**:
We extend dynamic resolution to the temporal dimension by adopting dynamic FPS sampling, enabling the model to comprehend videos at various sampling rates. Accordingly, we update mRoPE in the time dimension with IDs and absolute time alignment, enabling the model to learn temporal sequence and speed, and ultimately acquire the ability to pinpoint specific moments.
<p align="center">
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-VL/qwen2.5vl_arc.jpeg" width="80%"/>
<p>
* **Streamlined and Efficient Vision Encoder**
We enhance both training and inference speeds by strategically implementing window attention into the ViT. The ViT architecture is further optimized with SwiGLU and RMSNorm, aligning it with the structure of the Qwen2.5 LLM.
We have three models with 3, 7 and 72 billion parameters. This repo contains the instruction-tuned 32B Qwen2.5-VL model. For more information, visit our [Blog](https://qwenlm.github.io/blog/qwen2.5-vl/) and [GitHub](https://github.com/QwenLM/Qwen2.5-VL).
## Evaluation
### Vision
| Dataset | Qwen2.5-VL-72B<br><sup>([🤗](https://huggingface.co/Qwen/Qwen2.5-VL-72B-Instruct)[🤖](https://modelscope.cn/models/qwen/Qwen2.5-VL-72B-Instruct)) | Qwen2-VL-72B<br><sup>([🤗](https://huggingface.co/Qwen/Qwen2-VL-72B-Instruct)[🤖](https://modelscope.cn/models/qwen/Qwen2-VL-72B-Instruct)) | Qwen2.5-VL-32B<br><sup>([🤗](https://huggingface.co/Qwen/Qwen2.5-VL-32B-Instruct)[🤖](https://modelscope.cn/models/qwen/Qwen2.5-VL-32B-Instruct)) |
|--------------------|--------|--------------|------------------|
| MMMU |**70.2** | 64.5 | 70 |
| MMMU Pro |**51.1** | 46.2 | 49.5 |
| MMStar | **70.8** | 68.3 | 69.5 |
| MathVista | **74.8** | 70.5 | 74.7 |
| MathVision |38.1 | 25.9 | **40.0**|
| OCRBenchV2 | **61.5/63.7** | 47.8/46.1 | 57.2/59.1 |
| CC-OCR | **79.8** | 68.7 | 77.1 |
| DocVQA | **96.4** | **96.5** | 94.8 |
| InfoVQA | **87.3** | 84.5 | 83.4 |
| LVBench |47.3 | - | **49.00** |
| CharadesSTA |50.9 | - | **54.2** |
| VideoMME |**73.3/79.1** | 71.2/77.8 | 70.5/77.9 |
| MMBench-Video |**2.02** | 1.7 | 1.93 |
| AITZ |**83.2** | - | 83.1 |
| Android Control |**67.4/93.7** | 66.4/84.4 | 69.6/93.3 |
| ScreenSpot |**87.1** | - | 88.5 |
| ScreenSpot Pro |**43.6** | - | 39.4 |
| AndroidWorld |**35** | - | 22.0 |
| OSWorld |**8.83** | - | 5.92 |
### Text
| MODEL | MMLU | MMLU-PRO | MATH | GPQA-diamond | MBPP | Human Eval |
|-----------------|--------|----------|---------|--------------|--------|------------|
| Qwen2.5-VL-32B | 78.4 | 68.8 | 82.2 | 46.0 | 84.0 | 91.5 |
| Mistral-Small-3.1-24B | 80.6 | 66.8 | 69.3 | 46.0 | 74.7 | 88.4 |
| Gemma3-27B-IT | 76.9 | 67.5 | 89 | 42.4 | 74.4 | 87.8 |
| GPT-4o-Mini | 82.0 | 61.7 | 70.2 | 39.4 | 84.8 | 87.2 |
| Claude-3.5-Haiku | 77.6 | 65.0 | 69.2 | 41.6 | 85.6 | 88.1 |
## Requirements
The code of Qwen2.5-VL has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip install git+https://github.com/huggingface/transformers accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_vl'
```
## Quickstart
Below, we provide simple examples to show how to use Qwen2.5-VL with 🤖 ModelScope and 🤗 Transformers.
The code of Qwen2.5-VL has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip install git+https://github.com/huggingface/transformers accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_vl'
```
We offer a toolkit to help you handle various types of visual input more conveniently, as if you were using an API. This includes base64, URLs, and interleaved images and videos. You can install it using the following command:
```bash
# It's highly recommanded to use `[decord]` feature for faster video loading.
pip install qwen-vl-utils[decord]==0.0.8
```
If you are not using Linux, you might not be able to install `decord` from PyPI. In that case, you can use `pip install qwen-vl-utils` which will fall back to using torchvision for video processing. However, you can still [install decord from source](https://github.com/dmlc/decord?tab=readme-ov-file#install-from-source) to get decord used when loading video.
### Using 🤗 Transformers to Chat
Here we show a code snippet to show you how to use the chat model with `transformers` and `qwen_vl_utils`:
```python
from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info
# default: Load the model on the available device(s)
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-VL-32B-Instruct", torch_dtype="auto", device_map="auto"
)
# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
# "Qwen/Qwen2.5-VL-32B-Instruct",
# torch_dtype=torch.bfloat16,
# attn_implementation="flash_attention_2",
# device_map="auto",
# )
# default processer
processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-32B-Instruct")
# The default range for the number of visual tokens per image in the model is 4-16384.
# You can set min_pixels and max_pixels according to your needs, such as a token range of 256-1280, to balance performance and cost.
# min_pixels = 256*28*28
# max_pixels = 1280*28*28
# processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-32B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels)
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
},
{"type": "text", "text": "Describe this image."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
<details>
<summary>Multi image inference</summary>
```python
# Messages containing multiple images and a text query
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/image1.jpg"},
{"type": "image", "image": "file:///path/to/image2.jpg"},
{"type": "text", "text": "Identify the similarities between these images."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
</details>
<details>
<summary>Video inference</summary>
```python
# Messages containing a images list as a video and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": [
"file:///path/to/frame1.jpg",
"file:///path/to/frame2.jpg",
"file:///path/to/frame3.jpg",
"file:///path/to/frame4.jpg",
],
},
{"type": "text", "text": "Describe this video."},
],
}
]
# Messages containing a local video path and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": "file:///path/to/video1.mp4",
"max_pixels": 360 * 420,
"fps": 1.0,
},
{"type": "text", "text": "Describe this video."},
],
}
]
# Messages containing a video url and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2-VL/space_woaudio.mp4",
},
{"type": "text", "text": "Describe this video."},
],
}
]
#In Qwen 2.5 VL, frame rate information is also input into the model to align with absolute time.
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs, video_kwargs = process_vision_info(messages, return_video_kwargs=True)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
fps=fps,
padding=True,
return_tensors="pt",
**video_kwargs,
)
inputs = inputs.to("cuda")
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
Video URL compatibility largely depends on the third-party library version. The details are in the table below. change the backend by `FORCE_QWENVL_VIDEO_READER=torchvision` or `FORCE_QWENVL_VIDEO_READER=decord` if you prefer not to use the default one.
| Backend | HTTP | HTTPS |
|-------------|------|-------|
| torchvision >= 0.19.0 | ✅ | ✅ |
| torchvision < 0.19.0 | ❌ | ❌ |
| decord | ✅ | ❌ |
</details>
<details>
<summary>Batch inference</summary>
```python
# Sample messages for batch inference
messages1 = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/image1.jpg"},
{"type": "image", "image": "file:///path/to/image2.jpg"},
{"type": "text", "text": "What are the common elements in these pictures?"},
],
}
]
messages2 = [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Who are you?"},
]
# Combine messages for batch processing
messages = [messages1, messages2]
# Preparation for batch inference
texts = [
processor.apply_chat_template(msg, tokenize=False, add_generation_prompt=True)
for msg in messages
]
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=texts,
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Batch Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_texts = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_texts)
```
</details>
### 🤖 ModelScope
We strongly advise users especially those in mainland China to use ModelScope. `snapshot_download` can help you solve issues concerning downloading checkpoints.
### More Usage Tips
For input images, we support local files, base64, and URLs. For videos, we currently only support local files.
```python
# You can directly insert a local file path, a URL, or a base64-encoded image into the position where you want in the text.
## Local file path
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/your/image.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
## Image URL
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "http://path/to/your/image.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
## Base64 encoded image
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "data:image;base64,/9j/..."},
{"type": "text", "text": "Describe this image."},
],
}
]
```
#### Image Resolution for performance boost
The model supports a wide range of resolution inputs. By default, it uses the native resolution for input, but higher resolutions can enhance performance at the cost of more computation. Users can set the minimum and maximum number of pixels to achieve an optimal configuration for their needs, such as a token count range of 256-1280, to balance speed and memory usage.
```python
min_pixels = 256 * 28 * 28
max_pixels = 1280 * 28 * 28
processor = AutoProcessor.from_pretrained(
"Qwen/Qwen2.5-VL-32B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels
)
```
Besides, We provide two methods for fine-grained control over the image size input to the model:
1. Define min_pixels and max_pixels: Images will be resized to maintain their aspect ratio within the range of min_pixels and max_pixels.
2. Specify exact dimensions: Directly set `resized_height` and `resized_width`. These values will be rounded to the nearest multiple of 28.
```python
# min_pixels and max_pixels
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "file:///path/to/your/image.jpg",
"resized_height": 280,
"resized_width": 420,
},
{"type": "text", "text": "Describe this image."},
],
}
]
# resized_height and resized_width
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "file:///path/to/your/image.jpg",
"min_pixels": 50176,
"max_pixels": 50176,
},
{"type": "text", "text": "Describe this image."},
],
}
]
```
### Processing Long Texts
The current `config.json` is set for context length up to 32,768 tokens.
To handle extensive inputs exceeding 32,768 tokens, we utilize [YaRN](https://arxiv.org/abs/2309.00071), a technique for enhancing model length extrapolation, ensuring optimal performance on lengthy texts.
For supported frameworks, you could add the following to `config.json` to enable YaRN:
{
...,
"type": "yarn",
"mrope_section": [
16,
24,
24
],
"factor": 4,
"original_max_position_embeddings": 32768
}
However, it should be noted that this method has a significant impact on the performance of temporal and spatial localization tasks, and is therefore not recommended for use.
At the same time, for long video inputs, since MRoPE itself is more economical with ids, the max_position_embeddings can be directly modified to a larger value, such as 64k.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@article{Qwen2.5-VL,
title={Qwen2.5-VL Technical Report},
author={Bai, Shuai and Chen, Keqin and Liu, Xuejing and Wang, Jialin and Ge, Wenbin and Song, Sibo and Dang, Kai and Wang, Peng and Wang, Shijie and Tang, Jun and Zhong, Humen and Zhu, Yuanzhi and Yang, Mingkun and Li, Zhaohai and Wan, Jianqiang and Wang, Pengfei and Ding, Wei and Fu, Zheren and Xu, Yiheng and Ye, Jiabo and Zhang, Xi and Xie, Tianbao and Cheng, Zesen and Zhang, Hang and Yang, Zhibo and Xu, Haiyang and Lin, Junyang},
journal={arXiv preprint arXiv:2502.13923},
year={2025}
}
```
|
Mungert/Qwen2.5-VL-7B-Instruct-GGUF | Mungert | 2025-06-15T19:42:47Z | 21,642 | 16 | transformers | [
"transformers",
"gguf",
"multimodal",
"image-text-to-text",
"en",
"arxiv:2309.00071",
"arxiv:2409.12191",
"arxiv:2308.12966",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | image-text-to-text | 2025-03-27T22:25:21Z |
---
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
library_name: transformers
---
# <span style="color: #7FFF7F;">Qwen2.5-VL-7B-Instruct GGUF Models</span>
These files have been built using a imatrix file and latest llama.cpp build. You must use a fork of llama.cpp to use vision with the model.
## How to Use Qwen 2.5 VL Instruct with llama.cpp (latest as of 10th May 2025)
1. **Download the Qwen 2.5 VL gguf file**:
https://huggingface.co/Mungert/Qwen2.5-VL-7B-Instruct-GGUF/tree/main
Choose a gguf file without the mmproj in the name
Example gguf file : https://huggingface.co/Mungert/Mungert/Qwen2.5-VL-7B-Instruct-GGUF/resolve/main/Qwen2.5-VL-7B-Instruct-q8_0.gguf
Copy this file to your chosen folder.
2. **Download the Qwen 2.5 VL mmproj file**
https://huggingface.co/Mungert/Qwen2.5-VL-7B-Instruct-GGUF/tree/main
Choose a file with mmproj in the name
Example mmproj file : https://huggingface.co/Mungert/Qwen2.5-VL-7B-Instruct-GGUF/resolve/main/Qwen2.5-VL-7B-Instruct-mmproj-f16.gguf
Copy this file to your chosen folder.
3. Copy images to the same folder as the gguf files or alter paths appropriately.
In the example below the gguf files, images and llama-mtmd-cli are in the same folder.
Example image: image https://huggingface.co/Mungert/Qwen2.5-VL-7B-Instruct-GGUF/resolve/main/car-1.jpg
Copy this file to your chosen folder.
4. **Run the CLI Tool**:
From your chosen folder :
```bash
llama-mtmd-cli -m Qwen2.5-VL-7B-Instruct-q8_0.gguf --mmproj Qwen2.5-VL-7B-Instruct-mmproj-f16.gguf -p "Describe this image." --image ./car-1.jpg
```
## **Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)**
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen2.5-VL-7B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen2.5-VL-7B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen2.5-VL-7B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen2.5-VL-7B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen2.5-VL-7B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen2.5-VL-7B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen2.5-VL-7B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen2.5-VL-7B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen2.5-VL-7B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen2.5-VL-7B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen2.5-VL-7B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen2.5-VL-7B-Instruct
<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Introduction
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models. Today, we are excited to introduce the latest addition to the Qwen family: Qwen2.5-VL.
#### Key Enhancements:
* **Understand things visually**: Qwen2.5-VL is not only proficient in recognizing common objects such as flowers, birds, fish, and insects, but it is highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
* **Being agentic**: Qwen2.5-VL directly plays as a visual agent that can reason and dynamically direct tools, which is capable of computer use and phone use.
* **Understanding long videos and capturing events**: Qwen2.5-VL can comprehend videos of over 1 hour, and this time it has a new ability of cpaturing event by pinpointing the relevant video segments.
* **Capable of visual localization in different formats**: Qwen2.5-VL can accurately localize objects in an image by generating bounding boxes or points, and it can provide stable JSON outputs for coordinates and attributes.
* **Generating structured outputs**: for data like scans of invoices, forms, tables, etc. Qwen2.5-VL supports structured outputs of their contents, benefiting usages in finance, commerce, etc.
#### Model Architecture Updates:
* **Dynamic Resolution and Frame Rate Training for Video Understanding**:
We extend dynamic resolution to the temporal dimension by adopting dynamic FPS sampling, enabling the model to comprehend videos at various sampling rates. Accordingly, we update mRoPE in the time dimension with IDs and absolute time alignment, enabling the model to learn temporal sequence and speed, and ultimately acquire the ability to pinpoint specific moments.
<p align="center">
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-VL/qwen2.5vl_arc.jpeg" width="80%"/>
<p>
* **Streamlined and Efficient Vision Encoder**
We enhance both training and inference speeds by strategically implementing window attention into the ViT. The ViT architecture is further optimized with SwiGLU and RMSNorm, aligning it with the structure of the Qwen2.5 LLM.
We have three models with 3, 7 and 72 billion parameters. This repo contains the instruction-tuned 7B Qwen2.5-VL model. For more information, visit our [Blog](https://qwenlm.github.io/blog/qwen2.5-vl/) and [GitHub](https://github.com/QwenLM/Qwen2.5-VL).
## Evaluation
### Image benchmark
| Benchmark | InternVL2.5-8B | MiniCPM-o 2.6 | GPT-4o-mini | Qwen2-VL-7B |**Qwen2.5-VL-7B** |
| :--- | :---: | :---: | :---: | :---: | :---: |
| MMMU<sub>val</sub> | 56 | 50.4 | **60**| 54.1 | 58.6|
| MMMU-Pro<sub>val</sub> | 34.3 | - | 37.6| 30.5 | 41.0|
| DocVQA<sub>test</sub> | 93 | 93 | - | 94.5 | **95.7** |
| InfoVQA<sub>test</sub> | 77.6 | - | - |76.5 | **82.6** |
| ChartQA<sub>test</sub> | 84.8 | - |- | 83.0 |**87.3** |
| TextVQA<sub>val</sub> | 79.1 | 80.1 | -| 84.3 | **84.9**|
| OCRBench | 822 | 852 | 785 | 845 | **864** |
| CC_OCR | 57.7 | | | 61.6 | **77.8**|
| MMStar | 62.8| | |60.7| **63.9**|
| MMBench-V1.1-En<sub>test</sub> | 79.4 | 78.0 | 76.0| 80.7 | **82.6** |
| MMT-Bench<sub>test</sub> | - | - | - |**63.7** |63.6 |
| MMStar | **61.5** | 57.5 | 54.8 | 60.7 |63.9 |
| MMVet<sub>GPT-4-Turbo</sub> | 54.2 | 60.0 | 66.9 | 62.0 | **67.1**|
| HallBench<sub>avg</sub> | 45.2 | 48.1 | 46.1| 50.6 | **52.9**|
| MathVista<sub>testmini</sub> | 58.3 | 60.6 | 52.4 | 58.2 | **68.2**|
| MathVision | - | - | - | 16.3 | **25.07** |
### Video Benchmarks
| Benchmark | Qwen2-VL-7B | **Qwen2.5-VL-7B** |
| :--- | :---: | :---: |
| MVBench | 67.0 | **69.6** |
| PerceptionTest<sub>test</sub> | 66.9 | **70.5** |
| Video-MME<sub>wo/w subs</sub> | 63.3/69.0 | **65.1**/**71.6** |
| LVBench | | 45.3 |
| LongVideoBench | | 54.7 |
| MMBench-Video | 1.44 | 1.79 |
| TempCompass | | 71.7 |
| MLVU | | 70.2 |
| CharadesSTA/mIoU | 43.6|
### Agent benchmark
| Benchmarks | Qwen2.5-VL-7B |
|-------------------------|---------------|
| ScreenSpot | 84.7 |
| ScreenSpot Pro | 29.0 |
| AITZ_EM | 81.9 |
| Android Control High_EM | 60.1 |
| Android Control Low_EM | 93.7 |
| AndroidWorld_SR | 25.5 |
| MobileMiniWob++_SR | 91.4 |
## Requirements
The code of Qwen2.5-VL has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip install git+https://github.com/huggingface/transformers accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_vl'
```
## Quickstart
Below, we provide simple examples to show how to use Qwen2.5-VL with 🤖 ModelScope and 🤗 Transformers.
The code of Qwen2.5-VL has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip install git+https://github.com/huggingface/transformers accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_vl'
```
We offer a toolkit to help you handle various types of visual input more conveniently, as if you were using an API. This includes base64, URLs, and interleaved images and videos. You can install it using the following command:
```bash
# It's highly recommanded to use `[decord]` feature for faster video loading.
pip install qwen-vl-utils[decord]==0.0.8
```
If you are not using Linux, you might not be able to install `decord` from PyPI. In that case, you can use `pip install qwen-vl-utils` which will fall back to using torchvision for video processing. However, you can still [install decord from source](https://github.com/dmlc/decord?tab=readme-ov-file#install-from-source) to get decord used when loading video.
### Using 🤗 Transformers to Chat
Here we show a code snippet to show you how to use the chat model with `transformers` and `qwen_vl_utils`:
```python
from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info
# default: Load the model on the available device(s)
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-VL-7B-Instruct", torch_dtype="auto", device_map="auto"
)
# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
# "Qwen/Qwen2.5-VL-7B-Instruct",
# torch_dtype=torch.bfloat16,
# attn_implementation="flash_attention_2",
# device_map="auto",
# )
# default processer
processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct")
# The default range for the number of visual tokens per image in the model is 4-16384.
# You can set min_pixels and max_pixels according to your needs, such as a token range of 256-1280, to balance performance and cost.
# min_pixels = 256*28*28
# max_pixels = 1280*28*28
# processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels)
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
},
{"type": "text", "text": "Describe this image."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
<details>
<summary>Multi image inference</summary>
```python
# Messages containing multiple images and a text query
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/image1.jpg"},
{"type": "image", "image": "file:///path/to/image2.jpg"},
{"type": "text", "text": "Identify the similarities between these images."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
</details>
<details>
<summary>Video inference</summary>
```python
# Messages containing a images list as a video and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": [
"file:///path/to/frame1.jpg",
"file:///path/to/frame2.jpg",
"file:///path/to/frame3.jpg",
"file:///path/to/frame4.jpg",
],
},
{"type": "text", "text": "Describe this video."},
],
}
]
# Messages containing a local video path and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": "file:///path/to/video1.mp4",
"max_pixels": 360 * 420,
"fps": 1.0,
},
{"type": "text", "text": "Describe this video."},
],
}
]
# Messages containing a video url and a text query
messages = [
{
"role": "user",
"content": [
{
"type": "video",
"video": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2-VL/space_woaudio.mp4",
},
{"type": "text", "text": "Describe this video."},
],
}
]
#In Qwen 2.5 VL, frame rate information is also input into the model to align with absolute time.
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs, video_kwargs = process_vision_info(messages, return_video_kwargs=True)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
fps=fps,
padding=True,
return_tensors="pt",
**video_kwargs,
)
inputs = inputs.to("cuda")
# Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)
```
Video URL compatibility largely depends on the third-party library version. The details are in the table below. change the backend by `FORCE_QWENVL_VIDEO_READER=torchvision` or `FORCE_QWENVL_VIDEO_READER=decord` if you prefer not to use the default one.
| Backend | HTTP | HTTPS |
|-------------|------|-------|
| torchvision >= 0.19.0 | ✅ | ✅ |
| torchvision < 0.19.0 | ❌ | ❌ |
| decord | ✅ | ❌ |
</details>
<details>
<summary>Batch inference</summary>
```python
# Sample messages for batch inference
messages1 = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/image1.jpg"},
{"type": "image", "image": "file:///path/to/image2.jpg"},
{"type": "text", "text": "What are the common elements in these pictures?"},
],
}
]
messages2 = [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Who are you?"},
]
# Combine messages for batch processing
messages = [messages1, messages2]
# Preparation for batch inference
texts = [
processor.apply_chat_template(msg, tokenize=False, add_generation_prompt=True)
for msg in messages
]
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=texts,
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Batch Inference
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_texts = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_texts)
```
</details>
### 🤖 ModelScope
We strongly advise users especially those in mainland China to use ModelScope. `snapshot_download` can help you solve issues concerning downloading checkpoints.
### More Usage Tips
For input images, we support local files, base64, and URLs. For videos, we currently only support local files.
```python
# You can directly insert a local file path, a URL, or a base64-encoded image into the position where you want in the text.
## Local file path
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "file:///path/to/your/image.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
## Image URL
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "http://path/to/your/image.jpg"},
{"type": "text", "text": "Describe this image."},
],
}
]
## Base64 encoded image
messages = [
{
"role": "user",
"content": [
{"type": "image", "image": "data:image;base64,/9j/..."},
{"type": "text", "text": "Describe this image."},
],
}
]
```
#### Image Resolution for performance boost
The model supports a wide range of resolution inputs. By default, it uses the native resolution for input, but higher resolutions can enhance performance at the cost of more computation. Users can set the minimum and maximum number of pixels to achieve an optimal configuration for their needs, such as a token count range of 256-1280, to balance speed and memory usage.
```python
min_pixels = 256 * 28 * 28
max_pixels = 1280 * 28 * 28
processor = AutoProcessor.from_pretrained(
"Qwen/Qwen2.5-VL-7B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels
)
```
Besides, We provide two methods for fine-grained control over the image size input to the model:
1. Define min_pixels and max_pixels: Images will be resized to maintain their aspect ratio within the range of min_pixels and max_pixels.
2. Specify exact dimensions: Directly set `resized_height` and `resized_width`. These values will be rounded to the nearest multiple of 28.
```python
# min_pixels and max_pixels
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "file:///path/to/your/image.jpg",
"resized_height": 280,
"resized_width": 420,
},
{"type": "text", "text": "Describe this image."},
],
}
]
# resized_height and resized_width
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "file:///path/to/your/image.jpg",
"min_pixels": 50176,
"max_pixels": 50176,
},
{"type": "text", "text": "Describe this image."},
],
}
]
```
### Processing Long Texts
The current `config.json` is set for context length up to 32,768 tokens.
To handle extensive inputs exceeding 32,768 tokens, we utilize [YaRN](https://arxiv.org/abs/2309.00071), a technique for enhancing model length extrapolation, ensuring optimal performance on lengthy texts.
For supported frameworks, you could add the following to `config.json` to enable YaRN:
{
...,
"type": "yarn",
"mrope_section": [
16,
24,
24
],
"factor": 4,
"original_max_position_embeddings": 32768
}
However, it should be noted that this method has a significant impact on the performance of temporal and spatial localization tasks, and is therefore not recommended for use.
At the same time, for long video inputs, since MRoPE itself is more economical with ids, the max_position_embeddings can be directly modified to a larger value, such as 64k.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen2.5-VL,
title = {Qwen2.5-VL},
url = {https://qwenlm.github.io/blog/qwen2.5-vl/},
author = {Qwen Team},
month = {January},
year = {2025}
}
@article{Qwen2VL,
title={Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution},
author={Wang, Peng and Bai, Shuai and Tan, Sinan and Wang, Shijie and Fan, Zhihao and Bai, Jinze and Chen, Keqin and Liu, Xuejing and Wang, Jialin and Ge, Wenbin and Fan, Yang and Dang, Kai and Du, Mengfei and Ren, Xuancheng and Men, Rui and Liu, Dayiheng and Zhou, Chang and Zhou, Jingren and Lin, Junyang},
journal={arXiv preprint arXiv:2409.12191},
year={2024}
}
@article{Qwen-VL,
title={Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond},
author={Bai, Jinze and Bai, Shuai and Yang, Shusheng and Wang, Shijie and Tan, Sinan and Wang, Peng and Lin, Junyang and Zhou, Chang and Zhou, Jingren},
journal={arXiv preprint arXiv:2308.12966},
year={2023}
}
```
|
Mungert/gemma-3-27b-it-GGUF | Mungert | 2025-06-15T19:42:42Z | 849 | 8 | transformers | [
"transformers",
"gguf",
"image-text-to-text",
"arxiv:1905.07830",
"arxiv:1905.10044",
"arxiv:1911.11641",
"arxiv:1904.09728",
"arxiv:1705.03551",
"arxiv:1911.01547",
"arxiv:1907.10641",
"arxiv:1903.00161",
"arxiv:2009.03300",
"arxiv:2304.06364",
"arxiv:2103.03874",
"arxiv:2110.14168",
"arxiv:2311.12022",
"arxiv:2108.07732",
"arxiv:2107.03374",
"arxiv:2210.03057",
"arxiv:2106.03193",
"arxiv:1910.11856",
"arxiv:2502.12404",
"arxiv:2502.21228",
"arxiv:2404.16816",
"arxiv:2104.12756",
"arxiv:2311.16502",
"arxiv:2203.10244",
"arxiv:2404.12390",
"arxiv:1810.12440",
"arxiv:1908.02660",
"base_model:google/gemma-3-27b-pt",
"base_model:quantized:google/gemma-3-27b-pt",
"license:gemma",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | image-text-to-text | 2025-03-26T03:50:50Z | ---
license: gemma
library_name: transformers
pipeline_tag: image-text-to-text
extra_gated_heading: Access Gemma on Hugging Face
extra_gated_prompt: To access Gemma on Hugging Face, you’re required to review and
agree to Google’s usage license. To do this, please ensure you’re logged in to Hugging
Face and click below. Requests are processed immediately.
extra_gated_button_content: Acknowledge license
base_model: google/gemma-3-27b-pt
---
# <span style="color: #7FFF7F;">gemma-3-27b-it GGUF Models</span>
## How to Use Gemma 3 Vision with llama.cpp
To utilize the experimental support for Gemma 3 Vision in `llama.cpp`, follow these steps:
1. **Clone the lastest llama.cpp Repository**:
```bash
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
```
2. **Build the Llama.cpp**:
Build llama.cpp as usual : https://github.com/ggml-org/llama.cpp#building-the-project
Once llama.cpp is built Copy the ./llama.cpp/build/bin/llama-gemma3-cli to a chosen folder.
3. **Download the Gemma 3 gguf file**:
https://huggingface.co/Mungert/gemma-3-4b-it-gguf/tree/main
Choose a gguf file without the mmproj in the name
Example gguf file : https://huggingface.co/Mungert/gemma-3-4b-it-gguf/resolve/main/google_gemma-3-4b-it-q4_k_l.gguf
Copy this file to your chosen folder.
4. **Download the Gemma 3 mmproj file**
https://huggingface.co/Mungert/gemma-3-4b-it-gguf/tree/main
Choose a file with mmproj in the name
Example mmproj file : https://huggingface.co/Mungert/gemma-3-4b-it-gguf/resolve/main/google_gemma-3-4b-it-mmproj-bf16.gguf
Copy this file to your chosen folder.
5. Copy images to the same folder as the gguf files or alter paths appropriately.
In the example below the gguf files, images and llama-gemma-cli are in the same folder.
Example image: image https://huggingface.co/Mungert/gemma-3-4b-it-gguf/resolve/main/car-1.jpg
Copy this file to your chosen folder.
6. **Run the CLI Tool**:
From your chosen folder :
```bash
llama-gemma3-cli -m google_gemma-3-4b-it-q4_k_l.gguf --mmproj google_gemma-3-4b-it-mmproj-bf16.gguf
```
```
Running in chat mode, available commands:
/image <path> load an image
/clear clear the chat history
/quit or /exit exit the program
> /image car-1.jpg
Encoding image car-1.jpg
Image encoded in 46305 ms
Image decoded in 19302 ms
> what is the image of
Here's a breakdown of what's in the image:
**Subject:** The primary subject is a black Porsche Panamera Turbo driving on a highway.
**Details:**
* **Car:** It's a sleek, modern Porsche Panamera Turbo, identifiable by its distinctive rear design, the "PORSCHE" lettering, and the "Panamera Turbo" badge. The license plate reads "CVC-911".
* **Setting:** The car is on a multi-lane highway, with a blurred background of trees, a distant building, and a cloudy sky. The lighting suggests it's either dusk or dawn.
* **Motion:** The image captures the car in motion, with a slight motion blur to convey speed.
**Overall Impression:** The image conveys a sense of speed, luxury, and power. It's a well-composed shot that highlights the car's design and performance.
Do you want me to describe any specific aspect of the image in more detail, or perhaps analyze its composition?
```
## **Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)**
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Key Improvements**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `gemma-3-27b-it-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `gemma-3-27b-it-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `gemma-3-27b-it-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `gemma-3-27b-it-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `gemma-3-27b-it-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `gemma-3-27b-it-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `gemma-3-27b-it-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `gemma-3-27b-it-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `gemma-3-27b-it-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `gemma-3-27b-it-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `gemma-3-27b-it-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# <span style="color: #7FFF7F;">gemma-3-27b-it GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `gemma-3-27b-it-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `gemma-3-27b-it-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `gemma-3-27b-it-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `gemma-3-27b-it-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `gemma-3-27b-it-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `gemma-3-27b-it-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `gemma-3-27b-it-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `gemma-3-27b-it-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `gemma-3-27b-it-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `gemma-3-27b-it-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `gemma-3-27b-it-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Gemma 3 model card
**Model Page**: [Gemma](https://ai.google.dev/gemma/docs/core)
**Resources and Technical Documentation**:
* [Gemma 3 Technical Report][g3-tech-report]
* [Responsible Generative AI Toolkit][rai-toolkit]
* [Gemma on Kaggle][kaggle-gemma]
* [Gemma on Vertex Model Garden][vertex-mg-gemma3]
**Terms of Use**: [Terms][terms]
**Authors**: Google DeepMind
## Model Information
Summary description and brief definition of inputs and outputs.
### Description
Gemma is a family of lightweight, state-of-the-art open models from Google,
built from the same research and technology used to create the Gemini models.
Gemma 3 models are multimodal, handling text and image input and generating text
output, with open weights for both pre-trained variants and instruction-tuned
variants. Gemma 3 has a large, 128K context window, multilingual support in over
140 languages, and is available in more sizes than previous versions. Gemma 3
models are well-suited for a variety of text generation and image understanding
tasks, including question answering, summarization, and reasoning. Their
relatively small size makes it possible to deploy them in environments with
limited resources such as laptops, desktops or your own cloud infrastructure,
democratizing access to state of the art AI models and helping foster innovation
for everyone.
### Inputs and outputs
- **Input:**
- Text string, such as a question, a prompt, or a document to be summarized
- Images, normalized to 896 x 896 resolution and encoded to 256 tokens
each
- Total input context of 128K tokens for the 4B, 12B, and 27B sizes, and
32K tokens for the 1B size
- **Output:**
- Generated text in response to the input, such as an answer to a
question, analysis of image content, or a summary of a document
- Total output context of 8192 tokens
### Usage
Below there are some code snippets on how to get quickly started with running the model. First, install the Transformers library. Gemma 3 is supported starting from transformers 4.50.0.
```sh
$ pip install -U transformers
```
Then, copy the snippet from the section that is relevant for your use case.
#### Running with the `pipeline` API
You can initialize the model and processor for inference with `pipeline` as follows.
```python
from transformers import pipeline
import torch
pipe = pipeline(
"image-text-to-text",
model="google/gemma-3-27b-it",
device="cuda",
torch_dtype=torch.bfloat16
)
```
With instruction-tuned models, you need to use chat templates to process our inputs first. Then, you can pass it to the pipeline.
```python
messages = [
{
"role": "system",
"content": [{"type": "text", "text": "You are a helpful assistant."}]
},
{
"role": "user",
"content": [
{"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
{"type": "text", "text": "What animal is on the candy?"}
]
}
]
output = pipe(text=messages, max_new_tokens=200)
print(output[0]["generated_text"][-1]["content"])
# Okay, let's take a look!
# Based on the image, the animal on the candy is a **turtle**.
# You can see the shell shape and the head and legs.
```
#### Running the model on a single/multi GPU
```python
# pip install accelerate
from transformers import AutoProcessor, Gemma3ForConditionalGeneration
from PIL import Image
import requests
import torch
model_id = "google/gemma-3-27b-it"
model = Gemma3ForConditionalGeneration.from_pretrained(
model_id, device_map="auto"
).eval()
processor = AutoProcessor.from_pretrained(model_id)
messages = [
{
"role": "system",
"content": [{"type": "text", "text": "You are a helpful assistant."}]
},
{
"role": "user",
"content": [
{"type": "image", "image": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"},
{"type": "text", "text": "Describe this image in detail."}
]
}
]
inputs = processor.apply_chat_template(
messages, add_generation_prompt=True, tokenize=True,
return_dict=True, return_tensors="pt"
).to(model.device, dtype=torch.bfloat16)
input_len = inputs["input_ids"].shape[-1]
with torch.inference_mode():
generation = model.generate(**inputs, max_new_tokens=100, do_sample=False)
generation = generation[0][input_len:]
decoded = processor.decode(generation, skip_special_tokens=True)
print(decoded)
# **Overall Impression:** The image is a close-up shot of a vibrant garden scene,
# focusing on a cluster of pink cosmos flowers and a busy bumblebee.
# It has a slightly soft, natural feel, likely captured in daylight.
```
### Citation
```none
@article{gemma_2025,
title={Gemma 3},
url={https://goo.gle/Gemma3Report},
publisher={Kaggle},
author={Gemma Team},
year={2025}
}
```
## Model Data
Data used for model training and how the data was processed.
### Training Dataset
These models were trained on a dataset of text data that includes a wide variety
of sources. The 27B model was trained with 14 trillion tokens, the 12B model was
trained with 12 trillion tokens, 4B model was trained with 4 trillion tokens and
1B with 2 trillion tokens. Here are the key components:
- Web Documents: A diverse collection of web text ensures the model is
exposed to a broad range of linguistic styles, topics, and vocabulary. The
training dataset includes content in over 140 languages.
- Code: Exposing the model to code helps it to learn the syntax and
patterns of programming languages, which improves its ability to generate
code and understand code-related questions.
- Mathematics: Training on mathematical text helps the model learn logical
reasoning, symbolic representation, and to address mathematical queries.
- Images: A wide range of images enables the model to perform image
analysis and visual data extraction tasks.
The combination of these diverse data sources is crucial for training a powerful
multimodal model that can handle a wide variety of different tasks and data
formats.
### Data Preprocessing
Here are the key data cleaning and filtering methods applied to the training
data:
- CSAM Filtering: Rigorous CSAM (Child Sexual Abuse Material) filtering
was applied at multiple stages in the data preparation process to ensure
the exclusion of harmful and illegal content.
- Sensitive Data Filtering: As part of making Gemma pre-trained models
safe and reliable, automated techniques were used to filter out certain
personal information and other sensitive data from training sets.
- Additional methods: Filtering based on content quality and safety in
line with [our policies][safety-policies].
## Implementation Information
Details about the model internals.
### Hardware
Gemma was trained using [Tensor Processing Unit (TPU)][tpu] hardware (TPUv4p,
TPUv5p and TPUv5e). Training vision-language models (VLMS) requires significant
computational power. TPUs, designed specifically for matrix operations common in
machine learning, offer several advantages in this domain:
- Performance: TPUs are specifically designed to handle the massive
computations involved in training VLMs. They can speed up training
considerably compared to CPUs.
- Memory: TPUs often come with large amounts of high-bandwidth memory,
allowing for the handling of large models and batch sizes during training.
This can lead to better model quality.
- Scalability: TPU Pods (large clusters of TPUs) provide a scalable
solution for handling the growing complexity of large foundation models.
You can distribute training across multiple TPU devices for faster and more
efficient processing.
- Cost-effectiveness: In many scenarios, TPUs can provide a more
cost-effective solution for training large models compared to CPU-based
infrastructure, especially when considering the time and resources saved
due to faster training.
- These advantages are aligned with
[Google's commitments to operate sustainably][sustainability].
### Software
Training was done using [JAX][jax] and [ML Pathways][ml-pathways].
JAX allows researchers to take advantage of the latest generation of hardware,
including TPUs, for faster and more efficient training of large models. ML
Pathways is Google's latest effort to build artificially intelligent systems
capable of generalizing across multiple tasks. This is specially suitable for
foundation models, including large language models like these ones.
Together, JAX and ML Pathways are used as described in the
[paper about the Gemini family of models][gemini-2-paper]; *"the 'single
controller' programming model of Jax and Pathways allows a single Python
process to orchestrate the entire training run, dramatically simplifying the
development workflow."*
## Evaluation
Model evaluation metrics and results.
### Benchmark Results
These models were evaluated against a large collection of different datasets and
metrics to cover different aspects of text generation:
#### Reasoning and factuality
| Benchmark | Metric | Gemma 3 PT 1B | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |
| ------------------------------ |----------------|:--------------:|:-------------:|:--------------:|:--------------:|
| [HellaSwag][hellaswag] | 10-shot | 62.3 | 77.2 | 84.2 | 85.6 |
| [BoolQ][boolq] | 0-shot | 63.2 | 72.3 | 78.8 | 82.4 |
| [PIQA][piqa] | 0-shot | 73.8 | 79.6 | 81.8 | 83.3 |
| [SocialIQA][socialiqa] | 0-shot | 48.9 | 51.9 | 53.4 | 54.9 |
| [TriviaQA][triviaqa] | 5-shot | 39.8 | 65.8 | 78.2 | 85.5 |
| [Natural Questions][naturalq] | 5-shot | 9.48 | 20.0 | 31.4 | 36.1 |
| [ARC-c][arc] | 25-shot | 38.4 | 56.2 | 68.9 | 70.6 |
| [ARC-e][arc] | 0-shot | 73.0 | 82.4 | 88.3 | 89.0 |
| [WinoGrande][winogrande] | 5-shot | 58.2 | 64.7 | 74.3 | 78.8 |
| [BIG-Bench Hard][bbh] | few-shot | 28.4 | 50.9 | 72.6 | 77.7 |
| [DROP][drop] | 1-shot | 42.4 | 60.1 | 72.2 | 77.2 |
[hellaswag]: https://arxiv.org/abs/1905.07830
[boolq]: https://arxiv.org/abs/1905.10044
[piqa]: https://arxiv.org/abs/1911.11641
[socialiqa]: https://arxiv.org/abs/1904.09728
[triviaqa]: https://arxiv.org/abs/1705.03551
[naturalq]: https://github.com/google-research-datasets/natural-questions
[arc]: https://arxiv.org/abs/1911.01547
[winogrande]: https://arxiv.org/abs/1907.10641
[bbh]: https://paperswithcode.com/dataset/bbh
[drop]: https://arxiv.org/abs/1903.00161
#### STEM and code
| Benchmark | Metric | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |
| ------------------------------ |----------------|:-------------:|:--------------:|:--------------:|
| [MMLU][mmlu] | 5-shot | 59.6 | 74.5 | 78.6 |
| [MMLU][mmlu] (Pro COT) | 5-shot | 29.2 | 45.3 | 52.2 |
| [AGIEval][agieval] | 3-5-shot | 42.1 | 57.4 | 66.2 |
| [MATH][math] | 4-shot | 24.2 | 43.3 | 50.0 |
| [GSM8K][gsm8k] | 8-shot | 38.4 | 71.0 | 82.6 |
| [GPQA][gpqa] | 5-shot | 15.0 | 25.4 | 24.3 |
| [MBPP][mbpp] | 3-shot | 46.0 | 60.4 | 65.6 |
| [HumanEval][humaneval] | 0-shot | 36.0 | 45.7 | 48.8 |
[mmlu]: https://arxiv.org/abs/2009.03300
[agieval]: https://arxiv.org/abs/2304.06364
[math]: https://arxiv.org/abs/2103.03874
[gsm8k]: https://arxiv.org/abs/2110.14168
[gpqa]: https://arxiv.org/abs/2311.12022
[mbpp]: https://arxiv.org/abs/2108.07732
[humaneval]: https://arxiv.org/abs/2107.03374
#### Multilingual
| Benchmark | Gemma 3 PT 1B | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |
| ------------------------------------ |:-------------:|:-------------:|:--------------:|:--------------:|
| [MGSM][mgsm] | 2.04 | 34.7 | 64.3 | 74.3 |
| [Global-MMLU-Lite][global-mmlu-lite] | 24.9 | 57.0 | 69.4 | 75.7 |
| [WMT24++][wmt24pp] (ChrF) | 36.7 | 48.4 | 53.9 | 55.7 |
| [FloRes][flores] | 29.5 | 39.2 | 46.0 | 48.8 |
| [XQuAD][xquad] (all) | 43.9 | 68.0 | 74.5 | 76.8 |
| [ECLeKTic][eclektic] | 4.69 | 11.0 | 17.2 | 24.4 |
| [IndicGenBench][indicgenbench] | 41.4 | 57.2 | 61.7 | 63.4 |
[mgsm]: https://arxiv.org/abs/2210.03057
[flores]: https://arxiv.org/abs/2106.03193
[xquad]: https://arxiv.org/abs/1910.11856v3
[global-mmlu-lite]: https://huggingface.co/datasets/CohereForAI/Global-MMLU-Lite
[wmt24pp]: https://arxiv.org/abs/2502.12404v1
[eclektic]: https://arxiv.org/abs/2502.21228
[indicgenbench]: https://arxiv.org/abs/2404.16816
#### Multimodal
| Benchmark | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |
| ------------------------------ |:-------------:|:--------------:|:--------------:|
| [COCOcap][coco-cap] | 102 | 111 | 116 |
| [DocVQA][docvqa] (val) | 72.8 | 82.3 | 85.6 |
| [InfoVQA][info-vqa] (val) | 44.1 | 54.8 | 59.4 |
| [MMMU][mmmu] (pt) | 39.2 | 50.3 | 56.1 |
| [TextVQA][textvqa] (val) | 58.9 | 66.5 | 68.6 |
| [RealWorldQA][realworldqa] | 45.5 | 52.2 | 53.9 |
| [ReMI][remi] | 27.3 | 38.5 | 44.8 |
| [AI2D][ai2d] | 63.2 | 75.2 | 79.0 |
| [ChartQA][chartqa] | 63.6 | 74.7 | 76.3 |
| [VQAv2][vqav2] | 63.9 | 71.2 | 72.9 |
| [BLINK][blinkvqa] | 38.0 | 35.9 | 39.6 |
| [OKVQA][okvqa] | 51.0 | 58.7 | 60.2 |
| [TallyQA][tallyqa] | 42.5 | 51.8 | 54.3 |
| [SpatialSense VQA][ss-vqa] | 50.9 | 60.0 | 59.4 |
| [CountBenchQA][countbenchqa] | 26.1 | 17.8 | 68.0 |
[coco-cap]: https://cocodataset.org/#home
[docvqa]: https://www.docvqa.org/
[info-vqa]: https://arxiv.org/abs/2104.12756
[mmmu]: https://arxiv.org/abs/2311.16502
[textvqa]: https://textvqa.org/
[realworldqa]: https://paperswithcode.com/dataset/realworldqa
[remi]: https://arxiv.org/html/2406.09175v1
[ai2d]: https://allenai.org/data/diagrams
[chartqa]: https://arxiv.org/abs/2203.10244
[vqav2]: https://visualqa.org/index.html
[blinkvqa]: https://arxiv.org/abs/2404.12390
[okvqa]: https://okvqa.allenai.org/
[tallyqa]: https://arxiv.org/abs/1810.12440
[ss-vqa]: https://arxiv.org/abs/1908.02660
[countbenchqa]: https://github.com/google-research/big_vision/blob/main/big_vision/datasets/countbenchqa/
## Ethics and Safety
Ethics and safety evaluation approach and results.
### Evaluation Approach
Our evaluation methods include structured evaluations and internal red-teaming
testing of relevant content policies. Red-teaming was conducted by a number of
different teams, each with different goals and human evaluation metrics. These
models were evaluated against a number of different categories relevant to
ethics and safety, including:
- **Child Safety**: Evaluation of text-to-text and image to text prompts
covering child safety policies, including child sexual abuse and
exploitation.
- **Content Safety:** Evaluation of text-to-text and image to text prompts
covering safety policies including, harassment, violence and gore, and hate
speech.
- **Representational Harms**: Evaluation of text-to-text and image to text
prompts covering safety policies including bias, stereotyping, and harmful
associations or inaccuracies.
In addition to development level evaluations, we conduct "assurance
evaluations" which are our 'arms-length' internal evaluations for responsibility
governance decision making. They are conducted separately from the model
development team, to inform decision making about release. High level findings
are fed back to the model team, but prompt sets are held-out to prevent
overfitting and preserve the results' ability to inform decision making.
Assurance evaluation results are reported to our Responsibility & Safety Council
as part of release review.
### Evaluation Results
For all areas of safety testing, we saw major improvements in the categories of
child safety, content safety, and representational harms relative to previous
Gemma models. All testing was conducted without safety filters to evaluate the
model capabilities and behaviors. For both text-to-text and image-to-text, and
across all model sizes, the model produced minimal policy violations, and showed
significant improvements over previous Gemma models' performance with respect
to ungrounded inferences. A limitation of our evaluations was they included only
English language prompts.
## Usage and Limitations
These models have certain limitations that users should be aware of.
### Intended Usage
Open vision-language models (VLMs) models have a wide range of applications
across various industries and domains. The following list of potential uses is
not comprehensive. The purpose of this list is to provide contextual information
about the possible use-cases that the model creators considered as part of model
training and development.
- Content Creation and Communication
- Text Generation: These models can be used to generate creative text
formats such as poems, scripts, code, marketing copy, and email drafts.
- Chatbots and Conversational AI: Power conversational interfaces
for customer service, virtual assistants, or interactive applications.
- Text Summarization: Generate concise summaries of a text corpus,
research papers, or reports.
- Image Data Extraction: These models can be used to extract,
interpret, and summarize visual data for text communications.
- Research and Education
- Natural Language Processing (NLP) and VLM Research: These
models can serve as a foundation for researchers to experiment with VLM
and NLP techniques, develop algorithms, and contribute to the
advancement of the field.
- Language Learning Tools: Support interactive language learning
experiences, aiding in grammar correction or providing writing practice.
- Knowledge Exploration: Assist researchers in exploring large
bodies of text by generating summaries or answering questions about
specific topics.
### Limitations
- Training Data
- The quality and diversity of the training data significantly
influence the model's capabilities. Biases or gaps in the training data
can lead to limitations in the model's responses.
- The scope of the training dataset determines the subject areas
the model can handle effectively.
- Context and Task Complexity
- Models are better at tasks that can be framed with clear
prompts and instructions. Open-ended or highly complex tasks might be
challenging.
- A model's performance can be influenced by the amount of context
provided (longer context generally leads to better outputs, up to a
certain point).
- Language Ambiguity and Nuance
- Natural language is inherently complex. Models might struggle
to grasp subtle nuances, sarcasm, or figurative language.
- Factual Accuracy
- Models generate responses based on information they learned
from their training datasets, but they are not knowledge bases. They
may generate incorrect or outdated factual statements.
- Common Sense
- Models rely on statistical patterns in language. They might
lack the ability to apply common sense reasoning in certain situations.
### Ethical Considerations and Risks
The development of vision-language models (VLMs) raises several ethical
concerns. In creating an open model, we have carefully considered the following:
- Bias and Fairness
- VLMs trained on large-scale, real-world text and image data can
reflect socio-cultural biases embedded in the training material. These
models underwent careful scrutiny, input data pre-processing described
and posterior evaluations reported in this card.
- Misinformation and Misuse
- VLMs can be misused to generate text that is false, misleading,
or harmful.
- Guidelines are provided for responsible use with the model, see the
[Responsible Generative AI Toolkit][rai-toolkit].
- Transparency and Accountability:
- This model card summarizes details on the models' architecture,
capabilities, limitations, and evaluation processes.
- A responsibly developed open model offers the opportunity to
share innovation by making VLM technology accessible to developers and
researchers across the AI ecosystem.
Risks identified and mitigations:
- **Perpetuation of biases**: It's encouraged to perform continuous
monitoring (using evaluation metrics, human review) and the exploration of
de-biasing techniques during model training, fine-tuning, and other use
cases.
- **Generation of harmful content**: Mechanisms and guidelines for content
safety are essential. Developers are encouraged to exercise caution and
implement appropriate content safety safeguards based on their specific
product policies and application use cases.
- **Misuse for malicious purposes**: Technical limitations and developer
and end-user education can help mitigate against malicious applications of
VLMs. Educational resources and reporting mechanisms for users to flag
misuse are provided. Prohibited uses of Gemma models are outlined in the
[Gemma Prohibited Use Policy][prohibited-use].
- **Privacy violations**: Models were trained on data filtered for removal
of certain personal information and other sensitive data. Developers are
encouraged to adhere to privacy regulations with privacy-preserving
techniques.
### Benefits
At the time of release, this family of models provides high-performance open
vision-language model implementations designed from the ground up for
responsible AI development compared to similarly sized models.
Using the benchmark evaluation metrics described in this document, these models
have shown to provide superior performance to other, comparably-sized open model
alternatives.
[g3-tech-report]: https://goo.gle/Gemma3Report
[rai-toolkit]: https://ai.google.dev/responsible
[kaggle-gemma]: https://www.kaggle.com/models/google/gemma-3
[vertex-mg-gemma3]: https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3
[terms]: https://ai.google.dev/gemma/terms
[safety-policies]: https://ai.google/static/documents/ai-responsibility-update-published-february-2025.pdf
[prohibited-use]: https://ai.google.dev/gemma/prohibited_use_policy
[tpu]: https://cloud.google.com/tpu/docs/intro-to-tpu
[sustainability]: https://sustainability.google/operating-sustainably/
[jax]: https://github.com/jax-ml/jax
[ml-pathways]: https://blog.google/technology/ai/introducing-pathways-next-generation-ai-architecture/
[sustainability]: https://sustainability.google/operating-sustainably/
|
Mungert/X-Ray_Alpha-GGUF | Mungert | 2025-06-15T19:42:38Z | 1,381 | 5 | null | [
"gguf",
"en",
"dataset:SicariusSicariiStuff/UBW_Tapestries",
"base_model:google/gemma-3-4b-it",
"base_model:quantized:google/gemma-3-4b-it",
"license:gemma",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-03-25T09:55:48Z | ---
license: gemma
language:
- en
base_model:
- google/gemma-3-4b-it
datasets:
- SicariusSicariiStuff/UBW_Tapestries
---
# <span style="color: #7FFF7F;">X-Ray_Alpha GGUF Models</span>
## How to Use X-Ray_Alpha with llama.cpp
1. **Download the X-Ray_Alpha gguf file**:
https://huggingface.co/Mungert/X-Ray_Alpha-GGUF/tree/main
Choose a gguf file without the mmproj in the name
Example gguf file : https://huggingface.co/Mungert/Mungert/X-Ray_Alpha-GGUF/resolve/main/X-Ray_Alpha-q8_0.gguf
Copy this file to your chosen folder.
2. **Download the X-Ray_Alpha mmproj file**
https://huggingface.co/Mungert/X-Ray_Alpha-GGUF/tree/main
Choose a file with mmproj in the name
Example mmproj file : https://huggingface.co/Mungert/X-Ray_Alpha-GGUF/resolve/main/X-Ray_Alpha-mmproj-f32.gguf
Copy this file to your chosen folder.
3. Copy images to the same folder as the gguf files or alter paths appropriately.
In the example below the gguf files, images and llama-mtmd-cli are in the same folder.
Example image: image https://huggingface.co/Mungert/X-Ray_Alpha-GGUF/resolve/main/car-1.jpg
Copy this file to your chosen folder.
4. **Run the CLI Tool**:
From your chosen folder :
```bash
llama-gemma3-cli -m X-Ray_Alpha-q8_0.gguf --mmproj X-Ray_Alpha-mmproj-f32.gguf -p "Describe this image." --image ./car-1.jpg
```
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `X-Ray_Alpha-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `X-Ray_Alpha-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `X-Ray_Alpha-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `X-Ray_Alpha-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `X-Ray_Alpha-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `X-Ray_Alpha-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `X-Ray_Alpha-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `X-Ray_Alpha-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `X-Ray_Alpha-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `X-Ray_Alpha-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `X-Ray_Alpha-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
<div align="center">
<b style="font-size: 40px;">X-Ray_Alpha</b>
</div>
<img src="https://huggingface.co/SicariusSicariiStuff/X-Ray_Alpha/resolve/main/Images/X-Ray_Alpha.png" alt="X-Ray_Alpha" style="width: 30%; min-width: 450px; display: block; margin: auto;">
---
<div style="display: flex; justify-content: center; align-items: center;">
<a href="https://huggingface.co/SicariusSicariiStuff/X-Ray_Alpha#tldr"
style="color: #800080; font-weight: bold; font-size: 28px; text-decoration: none; margin: 0 20px;">
Click here
for TL;DR
</a>
<a href="https://huggingface.co/SicariusSicariiStuff/X-Ray_Alpha#why-is-this-important"
style="color: #1E90FF; font-weight: bold; font-size: 28px; text-decoration: none; margin: 0 20px;">
Why it's
important
</a>
<a href="https://huggingface.co/SicariusSicariiStuff/X-Ray_Alpha#how-can-you-help"
style="color: #32CD32; font-weight: bold; font-size: 28px; text-decoration: none; margin: 0 20px;">
How can YOU
help?
</a>
<a href="https://huggingface.co/SicariusSicariiStuff/X-Ray_Alpha#how-to-run-it"
style="color: #E31515; font-weight: bold; font-size: 28px; text-decoration: none; margin: 0 20px;">
How to
RUN IT
</a>
</div>
---
This is a pre-alpha proof-of-concept of **a real fully uncensored vision model**.
Why do I say **"real"**? The few vision models we got (qwen, llama 3.2) were "censored," and their fine-tunes were made only to the **text portion** of the model, as training a vision model is a serious pain.
The only actually trained and uncensored vision model I am aware of is [ToriiGate](https://huggingface.co/Minthy/ToriiGate-v0.4-7B); the rest of the vision models are just the stock vision + a fine-tuned LLM.
# Does this even work?
<h2 style="color: green; font-weight: bold; font-size: 80px; text-align: center;">YES!</h2>
---
# Why is this Important?
Having a **fully compliant** vision model is a critical step toward democratizing vision capabilities for various tasks, especially **image tagging**. This is a critical step in both making LORAs for image diffusion models, and for mass tagging images to pretrain a diffusion model.
In other words, having a fully compliant and accurate vision model will allow the open source community to easily train both loras and even pretrain image diffusion models.
Another important task can be content moderation and classification, in various use cases there might not be black and white, where some content that might be considered NSFW by corporations, is allowed, while other content is not, there's nuance. Today's vision models **do not let the users decide**, as they will straight up **refuse** to inference any content that Google \ Some other corporations decided is not to their liking, and therefore these stock models are useless in a lot of cases.
What if someone wants to classify art that includes nudity? Having a naked statue over 1,000 years old displayed in the middle of a city, in a museum, or at the city square is perfectly acceptable, however, a stock vision model will straight up refuse to inference something like that.
It's like in many "sensitive" topics that LLMs will straight up **refuse to answer**, while the content is **publicly available on Wikipedia**. This is an attitude of **cynical patronism**, I say cynical because corporations **take private data to train their models**, and it is "perfectly fine", yet- they serve as the **arbitrators of morality** and indirectly preach to us from a position of a suggested moral superiority. This **gatekeeping hurts innovation badly**, with vision models **especially so**, as the task of **tagging cannot be done by a single person at scale**, but a corporation can.
# How can YOU help?
This is sort of **"Pre-Alpha"**, a proof of concept, I did **A LOT** of shortcuts and "hacking" to make this work, and I would greatly appreciate some help to make it into an accurate and powerful open tool. I am not asking for money, but well-tagged data. I will take the burden and costs of the compute on myself, but I **cannot do tagging** at a large scale by myself.
## Bottom line, I need a lot of well-tagged, diverse data
So:
- If you have well-tagged images
- If you have a link to a well-tagged image dataset
- If you can, and willing to do image tagging
Then please send an email with [DATASET] in the title to:
```
[email protected]
```
As you probably figured by the email address name, this is not my main email, and I expect it to be spammed with junk, so **please use the [DATASET] tag** so I can more easily find the emails of **the good people** who are actually trying to help.
## Please see this dataset repo if you want to help:
[X-Ray_Community_Tagging](https://huggingface.co/datasets/SicariusSicariiStuff/X-Ray_Community_Tagging)
Also, if you don't want to upload it to the repo (although it's encouraged, and you can protect it with a password for privacy), you can still help by linking a google drive
or attach the images with the corrected output via the provided email.
Let's make this happen. We can do it!
---
### TL;DR
- **Fully uncensored and trained** there's no moderation in the vision model, I actually trained it.
- **The 2nd uncensored vision model in the world**, ToriiGate being the first as far as I know, and this one is the second.
- **In-depth descriptions** very detailed, long descriptions.
- The text portion is **somewhat uncensored** as well, I didn't want to butcher and fry it too much, so it remain "smart".
- **NOT perfect** This is a POC that shows that the task can even be done, a lot more work is needed.
- **Good Roleplay & Writing** I used a massive corpus of high quality human (**~60%**) and synthetic data.
---
# How to run it:
## VRAM needed for FP16: 15.9 GB
[Run inference with this](https://github.com/SicariusSicariiStuff/X-Ray_Vision)
# This is a pre-alpha POC (Proof Of Concept)
## Instructions:
clone:
```
git clone https://github.com/SicariusSicariiStuff/X-Ray_Vision.git
cd X-Ray_Vision/
```
Settings up venv, (tested for python 3.11, probably works with 3.10)
```
python3.11 -m venv env
source env/bin/activate
```
Install dependencies
```
pip install git+https://github.com/huggingface/[email protected]
pip install torch
pip install pillow
pip install accelerate
```
# Running inference
Usage:
```
python xRay-Vision.py /path/to/model/ /dir/with/images/
```
The output will print to the console, and export the results into a dir named after your image dir with the suffix "_TXT"
So if you run:
```
python xRay-Vision.py /some_path/x-Ray_model/ /home/images/weird_cats/
```
The results will be exported to:
```
/home/images/weird_cats_TXT/
```
---
<h2 style="color: green; font-weight: bold; font-size: 65px; text-align: center;">Your support = more models</h2>
<a href="https://ko-fi.com/sicarius" style="color: pink; font-weight: bold; font-size: 48px; text-decoration: none; display: block; text-align: center;">My Ko-fi page (Click here)</a>
---
## Citation Information
```
@llm{X-Ray_Alpha,
author = {SicariusSicariiStuff},
title = {X-Ray_Alpha},
year = {2025},
publisher = {Hugging Face},
url = {https://huggingface.co/SicariusSicariiStuff/X-Ray_Alpha}
}
```
---
## Other stuff
- [X-Ray_Vision](https://github.com/SicariusSicariiStuff/X-Ray_Vision) Easy stand-alone bulk vision inference at scale (inference a folder of images).
- [SLOP_Detector](https://github.com/SicariusSicariiStuff/SLOP_Detector) Nuke GPTisms, with SLOP detector.
- [LLAMA-3_8B_Unaligned](https://huggingface.co/SicariusSicariiStuff/LLAMA-3_8B_Unaligned) The grand project that started it all.
- [Blog and updates (Archived)](https://huggingface.co/SicariusSicariiStuff/Blog_And_Updates) Some updates, some rambles, sort of a mix between a diary and a blog. |
Mungert/Llama-2-7b-chat-hf-GGUF | Mungert | 2025-06-15T19:42:36Z | 929 | 3 | null | [
"gguf",
"facebook",
"meta",
"pytorch",
"llama",
"llama-2",
"text-generation",
"en",
"arxiv:2307.09288",
"license:llama2",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-25T02:32:34Z | ---
extra_gated_heading: You need to share contact information with Meta to access this model
extra_gated_prompt: >-
### LLAMA 2 COMMUNITY LICENSE AGREEMENT
"Agreement" means the terms and conditions for use, reproduction, distribution
and modification of the Llama Materials set forth herein.
"Documentation" means the specifications, manuals and documentation
accompanying Llama 2 distributed by Meta at
https://ai.meta.com/resources/models-and-libraries/llama-downloads/.
"Licensee" or "you" means you, or your employer or any other person or entity
(if you are entering into this Agreement on such person or entity's behalf),
of the age required under applicable laws, rules or regulations to provide
legal consent and that has legal authority to bind your employer or such other
person or entity if you are entering in this Agreement on their behalf.
"Llama 2" means the foundational large language models and software and
algorithms, including machine-learning model code, trained model weights,
inference-enabling code, training-enabling code, fine-tuning enabling code and
other elements of the foregoing distributed by Meta at
ai.meta.com/resources/models-and-libraries/llama-downloads/.
"Llama Materials" means, collectively, Meta's proprietary Llama 2 and
documentation (and any portion thereof) made available under this Agreement.
"Meta" or "we" means Meta Platforms Ireland Limited (if you are located in or,
if you are an entity, your principal place of business is in the EEA or
Switzerland) and Meta Platforms, Inc. (if you are located outside of the EEA
or Switzerland).
By clicking "I Accept" below or by using or distributing any portion or
element of the Llama Materials, you agree to be bound by this Agreement.
1. License Rights and Redistribution.
a. Grant of Rights. You are granted a non-exclusive, worldwide, non-
transferable and royalty-free limited license under Meta's intellectual
property or other rights owned by Meta embodied in the Llama Materials to
use, reproduce, distribute, copy, create derivative works of, and make
modifications to the Llama Materials.
b. Redistribution and Use.
i. If you distribute or make the Llama Materials, or any derivative works
thereof, available to a third party, you shall provide a copy of this
Agreement to such third party.
ii. If you receive Llama Materials, or any derivative works thereof, from a
Licensee as part of an integrated end user product, then Section 2 of this
Agreement will not apply to you.
iii. You must retain in all copies of the Llama Materials that you distribute
the following attribution notice within a "Notice" text file distributed as a
part of such copies: "Llama 2 is licensed under the LLAMA 2 Community
License, Copyright (c) Meta Platforms, Inc. All Rights Reserved."
iv. Your use of the Llama Materials must comply with applicable laws and
regulations (including trade compliance laws and regulations) and adhere to
the Acceptable Use Policy for the Llama Materials (available at
https://ai.meta.com/llama/use-policy), which is hereby incorporated by
reference into this Agreement.
v. You will not use the Llama Materials or any output or results of the Llama
Materials to improve any other large language model (excluding Llama 2 or
derivative works thereof).
2. Additional Commercial Terms. If, on the Llama 2 version release date, the
monthly active users of the products or services made available by or for
Licensee, or Licensee's affiliates, is greater than 700 million monthly
active users in the preceding calendar month, you must request a license from
Meta, which Meta may grant to you in its sole discretion, and you are not
authorized to exercise any of the rights under this Agreement unless or until
Meta otherwise expressly grants you such rights.
3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA
MATERIALS AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN "AS IS"
BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING,
WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT,
MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY
RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING
THE LLAMA MATERIALS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE
LLAMA MATERIALS AND ANY OUTPUT AND RESULTS.
4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE
UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE,
PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST
PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR
PUNITIVE DAMAGES, EVEN IF META OR ITS AFFILIATES HAVE BEEN ADVISED OF THE
POSSIBILITY OF ANY OF THE FOREGOING.
5. Intellectual Property.
a. No trademark licenses are granted under this Agreement, and in connection
with the Llama Materials, neither Meta nor Licensee may use any name or mark
owned by or associated with the other or any of its affiliates, except as
required for reasonable and customary use in describing and redistributing
the Llama Materials.
b. Subject to Meta's ownership of Llama Materials and derivatives made by or
for Meta, with respect to any derivative works and modifications of the Llama
Materials that are made by you, as between you and Meta, you are and will be
the owner of such derivative works and modifications.
c. If you institute litigation or other proceedings against Meta or any
entity (including a cross-claim or counterclaim in a lawsuit) alleging that
the Llama Materials or Llama 2 outputs or results, or any portion of any of
the foregoing, constitutes infringement of intellectual property or other
rights owned or licensable by you, then any licenses granted to you under
this Agreement shall terminate as of the date such litigation or claim is
filed or instituted. You will indemnify and hold harmless Meta from and
against any claim by any third party arising out of or related to your use or
distribution of the Llama Materials.
6. Term and Termination. The term of this Agreement will commence upon your
acceptance of this Agreement or access to the Llama Materials and will
continue in full force and effect until terminated in accordance with the
terms and conditions herein. Meta may terminate this Agreement if you are in
breach of any term or condition of this Agreement. Upon termination of this
Agreement, you shall delete and cease use of the Llama Materials. Sections 3,
4 and 7 shall survive the termination of this Agreement.
7. Governing Law and Jurisdiction. This Agreement will be governed and
construed under the laws of the State of California without regard to choice
of law principles, and the UN Convention on Contracts for the International
Sale of Goods does not apply to this Agreement. The courts of California
shall have exclusive jurisdiction of any dispute arising out of this
Agreement.
### Llama 2 Acceptable Use Policy
Meta is committed to promoting safe and fair use of its tools and features,
including Llama 2. If you access or use Llama 2, you agree to this Acceptable
Use Policy (“Policy”). The most recent copy of this policy can be found at
[ai.meta.com/llama/use-policy](http://ai.meta.com/llama/use-policy).
#### Prohibited Uses
We want everyone to use Llama 2 safely and responsibly. You agree you will not
use, or allow others to use, Llama 2 to:
1. Violate the law or others’ rights, including to:
1. Engage in, promote, generate, contribute to, encourage, plan, incite, or further illegal or unlawful activity or content, such as:
1. Violence or terrorism
2. Exploitation or harm to children, including the solicitation, creation, acquisition, or dissemination of child exploitative content or failure to report Child Sexual Abuse Material
3. Human trafficking, exploitation, and sexual violence
4. The illegal distribution of information or materials to minors, including obscene materials, or failure to employ legally required age-gating in connection with such information or materials.
5. Sexual solicitation
6. Any other criminal activity
2. Engage in, promote, incite, or facilitate the harassment, abuse, threatening, or bullying of individuals or groups of individuals
3. Engage in, promote, incite, or facilitate discrimination or other unlawful or harmful conduct in the provision of employment, employment benefits, credit, housing, other economic benefits, or other essential goods and services
4. Engage in the unauthorized or unlicensed practice of any profession including, but not limited to, financial, legal, medical/health, or related professional practices
5. Collect, process, disclose, generate, or infer health, demographic, or other sensitive personal or private information about individuals without rights and consents required by applicable laws
6. Engage in or facilitate any action or generate any content that infringes, misappropriates, or otherwise violates any third-party rights, including the outputs or results of any products or services using the Llama 2 Materials
7. Create, generate, or facilitate the creation of malicious code, malware, computer viruses or do anything else that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system
2. Engage in, promote, incite, facilitate, or assist in the planning or
development of activities that present a risk of death or bodily harm to
individuals, including use of Llama 2 related to the following:
1. Military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulations (ITAR) maintained by the United States Department of State
2. Guns and illegal weapons (including weapon development)
3. Illegal drugs and regulated/controlled substances
4. Operation of critical infrastructure, transportation technologies, or heavy machinery
5. Self-harm or harm to others, including suicide, cutting, and eating disorders
6. Any content intended to incite or promote violence, abuse, or any infliction of bodily harm to an individual
3. Intentionally deceive or mislead others, including use of Llama 2 related
to the following:
1. Generating, promoting, or furthering fraud or the creation or promotion of disinformation
2. Generating, promoting, or furthering defamatory content, including the creation of defamatory statements, images, or other content
3. Generating, promoting, or further distributing spam
4. Impersonating another individual without consent, authorization, or legal right
5. Representing that the use of Llama 2 or outputs are human-generated
6. Generating or facilitating false online engagement, including fake reviews and other means of fake online engagement
4. Fail to appropriately disclose to end users any known dangers of your AI system
Please report any violation of this Policy, software “bug,” or other problems
that could lead to a violation of this Policy through one of the following
means:
* Reporting issues with the model: [github.com/facebookresearch/llama](http://github.com/facebookresearch/llama)
* Reporting risky content generated by the model: [developers.facebook.com/llama_output_feedback](http://developers.facebook.com/llama_output_feedback)
* Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info)
* Reporting violations of the Acceptable Use Policy or unlicensed uses of Llama: [[email protected]](mailto:[email protected])
extra_gated_fields:
First Name: text
Last Name: text
Date of birth: date_picker
Country: country
Affiliation: text
geo: ip_location
By clicking Submit below I accept the terms of the license and acknowledge that the information I provide will be collected stored processed and shared in accordance with the Meta Privacy Policy: checkbox
extra_gated_description: >-
The information you provide will be collected, stored, processed and shared in
accordance with the [Meta Privacy
Policy](https://www.facebook.com/privacy/policy/).
extra_gated_button_content: Submit
language:
- en
pipeline_tag: text-generation
tags:
- facebook
- meta
- pytorch
- llama
- llama-2
license: llama2
---
# <span style="color: #7FFF7F;">Llama-2-7b-chat-hf GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-2-7b-chat-hf-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-2-7b-chat-hf-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-2-7b-chat-hf-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-2-7b-chat-hf-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-2-7b-chat-hf-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-2-7b-chat-hf-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-2-7b-chat-hf-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-2-7b-chat-hf-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-2-7b-chat-hf-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-2-7b-chat-hf-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-2-7b-chat-hf-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# **Llama 2**
Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is the repository for the 7B fine-tuned model, optimized for dialogue use cases and converted for the Hugging Face Transformers format. Links to other models can be found in the index at the bottom.
## Model Details
*Note: Use of this model is governed by the Meta license. In order to download the model weights and tokenizer, please visit the [website](https://ai.meta.com/resources/models-and-libraries/llama-downloads/) and accept our License before requesting access here.*
Meta developed and publicly released the Llama 2 family of large language models (LLMs), a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama-2-Chat, are optimized for dialogue use cases. Llama-2-Chat models outperform open-source chat models on most benchmarks we tested, and in our human evaluations for helpfulness and safety, are on par with some popular closed-source models like ChatGPT and PaLM.
**Model Developers** Meta
**Variations** Llama 2 comes in a range of parameter sizes — 7B, 13B, and 70B — as well as pretrained and fine-tuned variations.
**Input** Models input text only.
**Output** Models generate text only.
**Model Architecture** Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.
||Training Data|Params|Content Length|GQA|Tokens|LR|
|---|---|---|---|---|---|---|
|Llama 2|*A new mix of publicly available online data*|7B|4k|✗|2.0T|3.0 x 10<sup>-4</sup>|
|Llama 2|*A new mix of publicly available online data*|13B|4k|✗|2.0T|3.0 x 10<sup>-4</sup>|
|Llama 2|*A new mix of publicly available online data*|70B|4k|✔|2.0T|1.5 x 10<sup>-4</sup>|
*Llama 2 family of models.* Token counts refer to pretraining data only. All models are trained with a global batch-size of 4M tokens. Bigger models - 70B -- use Grouped-Query Attention (GQA) for improved inference scalability.
**Model Dates** Llama 2 was trained between January 2023 and July 2023.
**Status** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback.
**License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/)
**Research Paper** ["Llama-2: Open Foundation and Fine-tuned Chat Models"](arxiv.org/abs/2307.09288)
## Intended Use
**Intended Use Cases** Llama 2 is intended for commercial and research use in English. Tuned models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks.
To get the expected features and performance for the chat versions, a specific formatting needs to be followed, including the `INST` and `<<SYS>>` tags, `BOS` and `EOS` tokens, and the whitespaces and breaklines in between (we recommend calling `strip()` on inputs to avoid double-spaces). See our reference code in github for details: [`chat_completion`](https://github.com/facebookresearch/llama/blob/main/llama/generation.py#L212).
**Out-of-scope Uses** Use in any manner that violates applicable laws or regulations (including trade compliance laws).Use in languages other than English. Use in any other way that is prohibited by the Acceptable Use Policy and Licensing Agreement for Llama 2.
## Hardware and Software
**Training Factors** We used custom training libraries, Meta's Research Super Cluster, and production clusters for pretraining. Fine-tuning, annotation, and evaluation were also performed on third-party cloud compute.
**Carbon Footprint** Pretraining utilized a cumulative 3.3M GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W). Estimated total emissions were 539 tCO2eq, 100% of which were offset by Meta’s sustainability program.
||Time (GPU hours)|Power Consumption (W)|Carbon Emitted(tCO<sub>2</sub>eq)|
|---|---|---|---|
|Llama 2 7B|184320|400|31.22|
|Llama 2 13B|368640|400|62.44|
|Llama 2 70B|1720320|400|291.42|
|Total|3311616||539.00|
**CO<sub>2</sub> emissions during pretraining.** Time: total GPU time required for training each model. Power Consumption: peak power capacity per GPU device for the GPUs used adjusted for power usage efficiency. 100% of the emissions are directly offset by Meta's sustainability program, and because we are openly releasing these models, the pretraining costs do not need to be incurred by others.
## Training Data
**Overview** Llama 2 was pretrained on 2 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over one million new human-annotated examples. Neither the pretraining nor the fine-tuning datasets include Meta user data.
**Data Freshness** The pretraining data has a cutoff of September 2022, but some tuning data is more recent, up to July 2023.
## Evaluation Results
In this section, we report the results for the Llama 1 and Llama 2 models on standard academic benchmarks.For all the evaluations, we use our internal evaluations library.
|Model|Size|Code|Commonsense Reasoning|World Knowledge|Reading Comprehension|Math|MMLU|BBH|AGI Eval|
|---|---|---|---|---|---|---|---|---|---|
|Llama 1|7B|14.1|60.8|46.2|58.5|6.95|35.1|30.3|23.9|
|Llama 1|13B|18.9|66.1|52.6|62.3|10.9|46.9|37.0|33.9|
|Llama 1|33B|26.0|70.0|58.4|67.6|21.4|57.8|39.8|41.7|
|Llama 1|65B|30.7|70.7|60.5|68.6|30.8|63.4|43.5|47.6|
|Llama 2|7B|16.8|63.9|48.9|61.3|14.6|45.3|32.6|29.3|
|Llama 2|13B|24.5|66.9|55.4|65.8|28.7|54.8|39.4|39.1|
|Llama 2|70B|**37.5**|**71.9**|**63.6**|**69.4**|**35.2**|**68.9**|**51.2**|**54.2**|
**Overall performance on grouped academic benchmarks.** *Code:* We report the average pass@1 scores of our models on HumanEval and MBPP. *Commonsense Reasoning:* We report the average of PIQA, SIQA, HellaSwag, WinoGrande, ARC easy and challenge, OpenBookQA, and CommonsenseQA. We report 7-shot results for CommonSenseQA and 0-shot results for all other benchmarks. *World Knowledge:* We evaluate the 5-shot performance on NaturalQuestions and TriviaQA and report the average. *Reading Comprehension:* For reading comprehension, we report the 0-shot average on SQuAD, QuAC, and BoolQ. *MATH:* We report the average of the GSM8K (8 shot) and MATH (4 shot) benchmarks at top 1.
|||TruthfulQA|Toxigen|
|---|---|---|---|
|Llama 1|7B|27.42|23.00|
|Llama 1|13B|41.74|23.08|
|Llama 1|33B|44.19|22.57|
|Llama 1|65B|48.71|21.77|
|Llama 2|7B|33.29|**21.25**|
|Llama 2|13B|41.86|26.10|
|Llama 2|70B|**50.18**|24.60|
**Evaluation of pretrained LLMs on automatic safety benchmarks.** For TruthfulQA, we present the percentage of generations that are both truthful and informative (the higher the better). For ToxiGen, we present the percentage of toxic generations (the smaller the better).
|||TruthfulQA|Toxigen|
|---|---|---|---|
|Llama-2-Chat|7B|57.04|**0.00**|
|Llama-2-Chat|13B|62.18|**0.00**|
|Llama-2-Chat|70B|**64.14**|0.01|
**Evaluation of fine-tuned LLMs on different safety datasets.** Same metric definitions as above.
## Ethical Considerations and Limitations
Llama 2 is a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, Llama 2’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 2, developers should perform safety testing and tuning tailored to their specific applications of the model.
Please see the Responsible Use Guide available at [https://ai.meta.com/llama/responsible-use-guide/](https://ai.meta.com/llama/responsible-use-guide)
## Reporting Issues
Please report any software “bug,” or other problems with the models through one of the following means:
- Reporting issues with the model: [github.com/facebookresearch/llama](http://github.com/facebookresearch/llama)
- Reporting problematic content generated by the model: [developers.facebook.com/llama_output_feedback](http://developers.facebook.com/llama_output_feedback)
- Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info)
## Llama Model Index
|Model|Llama2|Llama2-hf|Llama2-chat|Llama2-chat-hf|
|---|---|---|---|---|
|7B| [Link](https://huggingface.co/meta-llama/Llama-2-7b) | [Link](https://huggingface.co/meta-llama/Llama-2-7b-hf) | [Link](https://huggingface.co/meta-llama/Llama-2-7b-chat) | [Link](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf)|
|13B| [Link](https://huggingface.co/meta-llama/Llama-2-13b) | [Link](https://huggingface.co/meta-llama/Llama-2-13b-hf) | [Link](https://huggingface.co/meta-llama/Llama-2-13b-chat) | [Link](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)|
|70B| [Link](https://huggingface.co/meta-llama/Llama-2-70b) | [Link](https://huggingface.co/meta-llama/Llama-2-70b-hf) | [Link](https://huggingface.co/meta-llama/Llama-2-70b-chat) | [Link](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf)| |
Mungert/MetaStone-L1-7B-GGUF | Mungert | 2025-06-15T19:42:32Z | 460 | 2 | null | [
"gguf",
"arxiv:2412.08864",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-03-24T15:38:45Z | ##
# <span style="color: #7FFF7F;">MetaStone-L1-7B GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `MetaStone-L1-7B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `MetaStone-L1-7B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `MetaStone-L1-7B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `MetaStone-L1-7B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `MetaStone-L1-7B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `MetaStone-L1-7B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `MetaStone-L1-7B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `MetaStone-L1-7B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `MetaStone-L1-7B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `MetaStone-L1-7B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `MetaStone-L1-7B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the FreeLLM .
🔵 **FreeLLM** – Runs **open-source Hugging Face models** Medium speed (unlimited, subject to Hugging Face API availability).
Introduction
MetaStone-L1 is the lite reasoning model of the MetaStone series, which aims to enhance the performance in hard downstream tasks.
On core reasoning benchmarks including mathematics and code, MetaStone-L1-7B achieved SOTA results in the parallel-level models, and it also achieved the comparable results as the API models such as Claude-3.5-Sonnet-1022 and GPT4o-0513.
<img src="./introduction.png" alt="Logo" width="800">
This repo contains the MetaStone-L1-7B model, which is trained based on DeepSeek-R1-Distill-Qwen-7B by GRPO. For full details of this model please refer to our release blog.
## Requirements
We advise you to use the latest version of transformers(```transformers==4.48.3```). For the best experience, please review the [Usage Guidelines](#usage-guidelines).
## Quickstart
Here give the example of how to use our model.
```
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "MetaStoneTec/MetaStone-L1-7B"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
messages = [
{"role": "user", "content": "Complete the square for the following quadratic: $-x^2+7 x-11$\n\nPlease reason step by step, and put your final answer within \\boxed{}."}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
```
## Usage Guidelines
To achieve optimal performance, we recommend the following settings:
1. Enhace the thoughful output：
a. Make sure the model starts with ```<think>\n``` to prevent generating empty think content. If you use ```apply_chat_template``` and set ```add_generation_prompt=True```, this is automatically implemented, but this may result in replies not having a <think> tag at the beginning, which is normal.
b. Ensure the final input of the model is in the format of ```<｜User｜> [your prompt] <｜Assistant｜><think>```.
2. Use a temperature of 0.6, a top sampling probability of 0.95, a maximum generation length of 32k.
3. Standardize output format: We recommend using hints to standardize model outputs when benchmarking.
a. Math questions: Add a statement "```Please reason step by step, and put your final answer within \\boxed{}.```" to the prompt.
b. Code problems: Add "### Format: Read the inputs from stdin solve the problem and write the answer to stdout. Enclose your code within delimiters as follows.\n \```python\n# YOUR CODE HERE\n\```\n### Answer: (use the provided format with backticks)" to the prompt.
4. In particular, we use ```latex2sympy2``` and ```sympy``` to assist in judging complex Latex formats for the Math500 evaluation script. For all datasets, we generate 64 responses per query to estimate pass@1.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{MetaStoneL17B,
title = {MetastoneL17B},
url = {https://huggingface.co/MetaStoneTec/MetaStone-L1-7B},
author = {MetaStone Team},
month = {March},
year = {2025}
}
```
```
@article{wang2024graph,
title={A Graph-Based Synthetic Data Pipeline for Scaling High-Quality Reasoning Instructions},
author={Wang, Jiankang and Xu, Jianjun and Wang, Xiaorui and Wang, Yuxin and Xing, Mengting and Fang, Shancheng and Chen, Zhineng and Xie, Hongtao and Zhang, Yongdong},
journal={arXiv preprint arXiv:2412.08864},
year={2024}
}
```
|
gradientrouting-spar/horizontal_5_proxy_ntrain_25_ntrig_9_random_3x3_seed_1_seed_25_seed_2_20250615_193305 | gradientrouting-spar | 2025-06-15T19:42:28Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
] | null | 2025-06-15T19:42:20Z | ---
library_name: transformers
tags: []
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed] |
Mungert/mxbai-rerank-large-v2-GGUF | Mungert | 2025-06-15T19:42:28Z | 894 | 2 | transformers | [
"transformers",
"gguf",
"text-ranking",
"en",
"zh",
"de",
"ja",
"ko",
"es",
"fr",
"ar",
"bn",
"ru",
"id",
"sw",
"te",
"th",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-ranking | 2025-03-24T13:41:53Z | ---
library_name: transformers
license: apache-2.0
language:
- en
- zh
- de
- ja
- ko
- es
- fr
- ar
- bn
- ru
- id
- sw
- te
- th
pipeline_tag: text-ranking
---
# <span style="color: #7FFF7F;">mxbai-rerank-large-v2 GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `mxbai-rerank-large-v2-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `mxbai-rerank-large-v2-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `mxbai-rerank-large-v2-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `mxbai-rerank-large-v2-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `mxbai-rerank-large-v2-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `mxbai-rerank-large-v2-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `mxbai-rerank-large-v2-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `mxbai-rerank-large-v2-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `mxbai-rerank-large-v2-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `mxbai-rerank-large-v2-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `mxbai-rerank-large-v2-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
<br><br>
<p align="center">
<svg xmlns="http://www.w3.org/2000/svg" xml:space="preserve" viewBox="0 0 2020 1130" width="150" height="150" aria-hidden="true"><path fill="#e95a0f" d="M398.167 621.992c-1.387-20.362-4.092-40.739-3.851-61.081.355-30.085 6.873-59.139 21.253-85.976 10.487-19.573 24.09-36.822 40.662-51.515 16.394-14.535 34.338-27.046 54.336-36.182 15.224-6.955 31.006-12.609 47.829-14.168 11.809-1.094 23.753-2.514 35.524-1.836 23.033 1.327 45.131 7.255 66.255 16.75 16.24 7.3 31.497 16.165 45.651 26.969 12.997 9.921 24.412 21.37 34.158 34.509 11.733 15.817 20.849 33.037 25.987 52.018 3.468 12.81 6.438 25.928 7.779 39.097 1.722 16.908 1.642 34.003 2.235 51.021.427 12.253.224 24.547 1.117 36.762 1.677 22.93 4.062 45.764 11.8 67.7 5.376 15.239 12.499 29.55 20.846 43.681l-18.282 20.328c-1.536 1.71-2.795 3.665-4.254 5.448l-19.323 23.533c-13.859-5.449-27.446-11.803-41.657-16.086-13.622-4.106-27.793-6.765-41.905-8.775-15.256-2.173-30.701-3.475-46.105-4.049-23.571-.879-47.178-1.056-70.769-1.029-10.858.013-21.723 1.116-32.57 1.926-5.362.4-10.69 1.255-16.464 1.477-2.758-7.675-5.284-14.865-7.367-22.181-3.108-10.92-4.325-22.554-13.16-31.095-2.598-2.512-5.069-5.341-6.883-8.443-6.366-10.884-12.48-21.917-18.571-32.959-4.178-7.573-8.411-14.375-17.016-18.559-10.34-5.028-19.538-12.387-29.311-18.611-3.173-2.021-6.414-4.312-9.952-5.297-5.857-1.63-11.98-2.301-17.991-3.376z"></path><path fill="#ed6d7b" d="M1478.998 758.842c-12.025.042-24.05.085-36.537-.373-.14-8.536.231-16.569.453-24.607.033-1.179-.315-2.986-1.081-3.4-.805-.434-2.376.338-3.518.81-.856.354-1.562 1.069-3.589 2.521-.239-3.308-.664-5.586-.519-7.827.488-7.544 2.212-15.166 1.554-22.589-1.016-11.451 1.397-14.592-12.332-14.419-3.793.048-3.617-2.803-3.332-5.331.499-4.422 1.45-8.803 1.77-13.233.311-4.316.068-8.672.068-12.861-2.554-.464-4.326-.86-6.12-1.098-4.415-.586-6.051-2.251-5.065-7.31 1.224-6.279.848-12.862 1.276-19.306.19-2.86-.971-4.473-3.794-4.753-4.113-.407-8.242-1.057-12.352-.975-4.663.093-5.192-2.272-4.751-6.012.733-6.229 1.252-12.483 1.875-18.726l1.102-10.495c-5.905-.309-11.146-.805-16.385-.778-3.32.017-5.174-1.4-5.566-4.4-1.172-8.968-2.479-17.944-3.001-26.96-.26-4.484-1.936-5.705-6.005-5.774-9.284-.158-18.563-.594-27.843-.953-7.241-.28-10.137-2.764-11.3-9.899-.746-4.576-2.715-7.801-7.777-8.207-7.739-.621-15.511-.992-23.207-1.961-7.327-.923-14.587-2.415-21.853-3.777-5.021-.941-10.003-2.086-15.003-3.14 4.515-22.952 13.122-44.382 26.284-63.587 18.054-26.344 41.439-47.239 69.102-63.294 15.847-9.197 32.541-16.277 50.376-20.599 16.655-4.036 33.617-5.715 50.622-4.385 33.334 2.606 63.836 13.955 92.415 31.15 15.864 9.545 30.241 20.86 42.269 34.758 8.113 9.374 15.201 19.78 21.718 30.359 10.772 17.484 16.846 36.922 20.611 56.991 1.783 9.503 2.815 19.214 3.318 28.876.758 14.578.755 29.196.65 44.311l-51.545 20.013c-7.779 3.059-15.847 5.376-21.753 12.365-4.73 5.598-10.658 10.316-16.547 14.774-9.9 7.496-18.437 15.988-25.083 26.631-3.333 5.337-7.901 10.381-12.999 14.038-11.355 8.144-17.397 18.973-19.615 32.423l-6.988 41.011z"></path><path fill="#ec663e" d="M318.11 923.047c-.702 17.693-.832 35.433-2.255 53.068-1.699 21.052-6.293 41.512-14.793 61.072-9.001 20.711-21.692 38.693-38.496 53.583-16.077 14.245-34.602 24.163-55.333 30.438-21.691 6.565-43.814 8.127-66.013 6.532-22.771-1.636-43.88-9.318-62.74-22.705-20.223-14.355-35.542-32.917-48.075-54.096-9.588-16.203-16.104-33.55-19.201-52.015-2.339-13.944-2.307-28.011-.403-42.182 2.627-19.545 9.021-37.699 17.963-55.067 11.617-22.564 27.317-41.817 48.382-56.118 15.819-10.74 33.452-17.679 52.444-20.455 8.77-1.282 17.696-1.646 26.568-2.055 11.755-.542 23.534-.562 35.289-1.11 8.545-.399 17.067-1.291 26.193-1.675 1.349 1.77 2.24 3.199 2.835 4.742 4.727 12.261 10.575 23.865 18.636 34.358 7.747 10.084 14.83 20.684 22.699 30.666 3.919 4.972 8.37 9.96 13.609 13.352 7.711 4.994 16.238 8.792 24.617 12.668 5.852 2.707 12.037 4.691 18.074 6.998z"></path><path fill="#ea580e" d="M1285.167 162.995c3.796-29.75 13.825-56.841 32.74-80.577 16.339-20.505 36.013-36.502 59.696-47.614 14.666-6.881 29.971-11.669 46.208-12.749 10.068-.669 20.239-1.582 30.255-.863 16.6 1.191 32.646 5.412 47.9 12.273 19.39 8.722 36.44 20.771 50.582 36.655 15.281 17.162 25.313 37.179 31.49 59.286 5.405 19.343 6.31 39.161 4.705 58.825-2.37 29.045-11.836 55.923-30.451 78.885-10.511 12.965-22.483 24.486-37.181 33.649-5.272-5.613-10.008-11.148-14.539-16.846-5.661-7.118-10.958-14.533-16.78-21.513-4.569-5.478-9.548-10.639-14.624-15.658-3.589-3.549-7.411-6.963-11.551-9.827-5.038-3.485-10.565-6.254-15.798-9.468-8.459-5.195-17.011-9.669-26.988-11.898-12.173-2.72-24.838-4.579-35.622-11.834-1.437-.967-3.433-1.192-5.213-1.542-12.871-2.529-25.454-5.639-36.968-12.471-5.21-3.091-11.564-4.195-17.011-6.965-4.808-2.445-8.775-6.605-13.646-8.851-8.859-4.085-18.114-7.311-27.204-10.896z"></path><path fill="#f8ab00" d="M524.963 311.12c-9.461-5.684-19.513-10.592-28.243-17.236-12.877-9.801-24.031-21.578-32.711-35.412-11.272-17.965-19.605-37.147-21.902-58.403-1.291-11.951-2.434-24.073-1.87-36.034.823-17.452 4.909-34.363 11.581-50.703 8.82-21.603 22.25-39.792 39.568-55.065 18.022-15.894 39.162-26.07 62.351-32.332 19.22-5.19 38.842-6.177 58.37-4.674 23.803 1.831 45.56 10.663 65.062 24.496 17.193 12.195 31.688 27.086 42.894 45.622-11.403 8.296-22.633 16.117-34.092 23.586-17.094 11.142-34.262 22.106-48.036 37.528-8.796 9.848-17.201 20.246-27.131 28.837-16.859 14.585-27.745 33.801-41.054 51.019-11.865 15.349-20.663 33.117-30.354 50.08-5.303 9.283-9.654 19.11-14.434 28.692z"></path><path fill="#ea5227" d="M1060.11 1122.049c-7.377 1.649-14.683 4.093-22.147 4.763-11.519 1.033-23.166 1.441-34.723 1.054-19.343-.647-38.002-4.7-55.839-12.65-15.078-6.72-28.606-15.471-40.571-26.836-24.013-22.81-42.053-49.217-49.518-81.936-1.446-6.337-1.958-12.958-2.235-19.477-.591-13.926-.219-27.909-1.237-41.795-.916-12.5-3.16-24.904-4.408-37.805 1.555-1.381 3.134-2.074 3.778-3.27 4.729-8.79 12.141-15.159 19.083-22.03 5.879-5.818 10.688-12.76 16.796-18.293 6.993-6.335 11.86-13.596 14.364-22.612l8.542-29.993c8.015 1.785 15.984 3.821 24.057 5.286 8.145 1.478 16.371 2.59 24.602 3.493 8.453.927 16.956 1.408 25.891 2.609 1.119 16.09 1.569 31.667 2.521 47.214.676 11.045 1.396 22.154 3.234 33.043 2.418 14.329 5.708 28.527 9.075 42.674 3.499 14.705 4.028 29.929 10.415 44.188 10.157 22.674 18.29 46.25 28.281 69.004 7.175 16.341 12.491 32.973 15.078 50.615.645 4.4 3.256 8.511 4.963 12.755z"></path><path fill="#ea5330" d="M1060.512 1122.031c-2.109-4.226-4.72-8.337-5.365-12.737-2.587-17.642-7.904-34.274-15.078-50.615-9.991-22.755-18.124-46.33-28.281-69.004-6.387-14.259-6.916-29.482-10.415-44.188-3.366-14.147-6.656-28.346-9.075-42.674-1.838-10.889-2.558-21.999-3.234-33.043-.951-15.547-1.401-31.124-2.068-47.146 8.568-.18 17.146.487 25.704.286l41.868-1.4c.907 3.746 1.245 7.04 1.881 10.276l8.651 42.704c.903 4.108 2.334 8.422 4.696 11.829 7.165 10.338 14.809 20.351 22.456 30.345 4.218 5.512 8.291 11.304 13.361 15.955 8.641 7.927 18.065 14.995 27.071 22.532 12.011 10.052 24.452 19.302 40.151 22.854-1.656 11.102-2.391 22.44-5.172 33.253-4.792 18.637-12.38 36.209-23.412 52.216-13.053 18.94-29.086 34.662-49.627 45.055-10.757 5.443-22.443 9.048-34.111 13.501z"></path><path fill="#f8aa05" d="M1989.106 883.951c5.198 8.794 11.46 17.148 15.337 26.491 5.325 12.833 9.744 26.207 12.873 39.737 2.95 12.757 3.224 25.908 1.987 39.219-1.391 14.973-4.643 29.268-10.349 43.034-5.775 13.932-13.477 26.707-23.149 38.405-14.141 17.104-31.215 30.458-50.807 40.488-14.361 7.352-29.574 12.797-45.741 14.594-10.297 1.144-20.732 2.361-31.031 1.894-24.275-1.1-47.248-7.445-68.132-20.263-6.096-3.741-11.925-7.917-17.731-12.342 5.319-5.579 10.361-10.852 15.694-15.811l37.072-34.009c.975-.892 2.113-1.606 3.08-2.505 6.936-6.448 14.765-12.2 20.553-19.556 8.88-11.285 20.064-19.639 31.144-28.292 4.306-3.363 9.06-6.353 12.673-10.358 5.868-6.504 10.832-13.814 16.422-20.582 6.826-8.264 13.727-16.481 20.943-24.401 4.065-4.461 8.995-8.121 13.249-12.424 14.802-14.975 28.77-30.825 45.913-43.317z"></path><path fill="#ed6876" d="M1256.099 523.419c5.065.642 10.047 1.787 15.068 2.728 7.267 1.362 14.526 2.854 21.853 3.777 7.696.97 15.468 1.34 23.207 1.961 5.062.406 7.031 3.631 7.777 8.207 1.163 7.135 4.059 9.62 11.3 9.899l27.843.953c4.069.069 5.745 1.291 6.005 5.774.522 9.016 1.829 17.992 3.001 26.96.392 3 2.246 4.417 5.566 4.4 5.239-.026 10.48.469 16.385.778l-1.102 10.495-1.875 18.726c-.44 3.74.088 6.105 4.751 6.012 4.11-.082 8.239.568 12.352.975 2.823.28 3.984 1.892 3.794 4.753-.428 6.444-.052 13.028-1.276 19.306-.986 5.059.651 6.724 5.065 7.31 1.793.238 3.566.634 6.12 1.098 0 4.189.243 8.545-.068 12.861-.319 4.43-1.27 8.811-1.77 13.233-.285 2.528-.461 5.379 3.332 5.331 13.729-.173 11.316 2.968 12.332 14.419.658 7.423-1.066 15.045-1.554 22.589-.145 2.241.28 4.519.519 7.827 2.026-1.452 2.733-2.167 3.589-2.521 1.142-.472 2.713-1.244 3.518-.81.767.414 1.114 2.221 1.081 3.4l-.917 24.539c-11.215.82-22.45.899-33.636 1.674l-43.952 3.436c-1.086-3.01-2.319-5.571-2.296-8.121.084-9.297-4.468-16.583-9.091-24.116-3.872-6.308-8.764-13.052-9.479-19.987-1.071-10.392-5.716-15.936-14.889-18.979-1.097-.364-2.16-.844-3.214-1.327-7.478-3.428-15.548-5.918-19.059-14.735-.904-2.27-3.657-3.775-5.461-5.723-2.437-2.632-4.615-5.525-7.207-7.987-2.648-2.515-5.352-5.346-8.589-6.777-4.799-2.121-10.074-3.185-15.175-4.596l-15.785-4.155c.274-12.896 1.722-25.901.54-38.662-1.647-17.783-3.457-35.526-2.554-53.352.528-10.426 2.539-20.777 3.948-31.574z"></path><path fill="#f6a200" d="M525.146 311.436c4.597-9.898 8.947-19.725 14.251-29.008 9.691-16.963 18.49-34.73 30.354-50.08 13.309-17.218 24.195-36.434 41.054-51.019 9.93-8.591 18.335-18.989 27.131-28.837 13.774-15.422 30.943-26.386 48.036-37.528 11.459-7.469 22.688-15.29 34.243-23.286 11.705 16.744 19.716 35.424 22.534 55.717 2.231 16.066 2.236 32.441 2.753 49.143-4.756 1.62-9.284 2.234-13.259 4.056-6.43 2.948-12.193 7.513-18.774 9.942-19.863 7.331-33.806 22.349-47.926 36.784-7.86 8.035-13.511 18.275-19.886 27.705-4.434 6.558-9.345 13.037-12.358 20.254-4.249 10.177-6.94 21.004-10.296 31.553-12.33.053-24.741 1.027-36.971-.049-20.259-1.783-40.227-5.567-58.755-14.69-.568-.28-1.295-.235-2.132-.658z"></path><path fill="#f7a80d" d="M1989.057 883.598c-17.093 12.845-31.061 28.695-45.863 43.67-4.254 4.304-9.184 7.963-13.249 12.424-7.216 7.92-14.117 16.137-20.943 24.401-5.59 6.768-10.554 14.078-16.422 20.582-3.614 4.005-8.367 6.995-12.673 10.358-11.08 8.653-22.264 17.007-31.144 28.292-5.788 7.356-13.617 13.108-20.553 19.556-.967.899-2.105 1.614-3.08 2.505l-37.072 34.009c-5.333 4.96-10.375 10.232-15.859 15.505-21.401-17.218-37.461-38.439-48.623-63.592 3.503-1.781 7.117-2.604 9.823-4.637 8.696-6.536 20.392-8.406 27.297-17.714.933-1.258 2.646-1.973 4.065-2.828 17.878-10.784 36.338-20.728 53.441-32.624 10.304-7.167 18.637-17.23 27.583-26.261 3.819-3.855 7.436-8.091 10.3-12.681 12.283-19.68 24.43-39.446 40.382-56.471 12.224-13.047 17.258-29.524 22.539-45.927 15.85 4.193 29.819 12.129 42.632 22.08 10.583 8.219 19.782 17.883 27.42 29.351z"></path><path fill="#ef7a72" d="M1479.461 758.907c1.872-13.734 4.268-27.394 6.525-41.076 2.218-13.45 8.26-24.279 19.615-32.423 5.099-3.657 9.667-8.701 12.999-14.038 6.646-10.643 15.183-19.135 25.083-26.631 5.888-4.459 11.817-9.176 16.547-14.774 5.906-6.99 13.974-9.306 21.753-12.365l51.48-19.549c.753 11.848.658 23.787 1.641 35.637 1.771 21.353 4.075 42.672 11.748 62.955.17.449.107.985-.019 2.158-6.945 4.134-13.865 7.337-20.437 11.143-3.935 2.279-7.752 5.096-10.869 8.384-6.011 6.343-11.063 13.624-17.286 19.727-9.096 8.92-12.791 20.684-18.181 31.587-.202.409-.072.984-.096 1.481-8.488-1.72-16.937-3.682-25.476-5.094-9.689-1.602-19.426-3.084-29.201-3.949-15.095-1.335-30.241-2.1-45.828-3.172z"></path><path fill="#e94e3b" d="M957.995 766.838c-20.337-5.467-38.791-14.947-55.703-27.254-8.2-5.967-15.451-13.238-22.958-20.37 2.969-3.504 5.564-6.772 8.598-9.563 7.085-6.518 11.283-14.914 15.8-23.153 4.933-8.996 10.345-17.743 14.966-26.892 2.642-5.231 5.547-11.01 5.691-16.611.12-4.651.194-8.932 2.577-12.742 8.52-13.621 15.483-28.026 18.775-43.704 2.11-10.049 7.888-18.774 7.81-29.825-.064-9.089 4.291-18.215 6.73-27.313 3.212-11.983 7.369-23.797 9.492-35.968 3.202-18.358 5.133-36.945 7.346-55.466l4.879-45.8c6.693.288 13.386.575 20.54 1.365.13 3.458-.41 6.407-.496 9.37l-1.136 42.595c-.597 11.552-2.067 23.058-3.084 34.59l-3.845 44.478c-.939 10.202-1.779 20.432-3.283 30.557-.96 6.464-4.46 12.646-1.136 19.383.348.706-.426 1.894-.448 2.864-.224 9.918-5.99 19.428-2.196 29.646.103.279-.033.657-.092.983l-8.446 46.205c-1.231 6.469-2.936 12.846-4.364 19.279-1.5 6.757-2.602 13.621-4.456 20.277-3.601 12.93-10.657 25.3-5.627 39.47.368 1.036.234 2.352.017 3.476l-5.949 30.123z"></path><path fill="#ea5043" d="M958.343 767.017c1.645-10.218 3.659-20.253 5.602-30.302.217-1.124.351-2.44-.017-3.476-5.03-14.17 2.026-26.539 5.627-39.47 1.854-6.656 2.956-13.52 4.456-20.277 1.428-6.433 3.133-12.81 4.364-19.279l8.446-46.205c.059-.326.196-.705.092-.983-3.794-10.218 1.972-19.728 2.196-29.646.022-.97.796-2.158.448-2.864-3.324-6.737.176-12.919 1.136-19.383 1.504-10.125 2.344-20.355 3.283-30.557l3.845-44.478c1.017-11.532 2.488-23.038 3.084-34.59.733-14.18.722-28.397 1.136-42.595.086-2.963.626-5.912.956-9.301 5.356-.48 10.714-.527 16.536-.081 2.224 15.098 1.855 29.734 1.625 44.408-.157 10.064 1.439 20.142 1.768 30.23.334 10.235-.035 20.49.116 30.733.084 5.713.789 11.418.861 17.13.054 4.289-.469 8.585-.702 12.879-.072 1.323-.138 2.659-.031 3.975l2.534 34.405-1.707 36.293-1.908 48.69c-.182 8.103.993 16.237.811 24.34-.271 12.076-1.275 24.133-1.787 36.207-.102 2.414-.101 5.283 1.06 7.219 4.327 7.22 4.463 15.215 4.736 23.103.365 10.553.088 21.128.086 31.693-11.44 2.602-22.84.688-34.106-.916-11.486-1.635-22.806-4.434-34.546-6.903z"></path><path fill="#eb5d19" d="M398.091 622.45c6.086.617 12.21 1.288 18.067 2.918 3.539.985 6.779 3.277 9.952 5.297 9.773 6.224 18.971 13.583 29.311 18.611 8.606 4.184 12.839 10.986 17.016 18.559l18.571 32.959c1.814 3.102 4.285 5.931 6.883 8.443 8.835 8.542 10.052 20.175 13.16 31.095 2.082 7.317 4.609 14.507 6.946 22.127-29.472 3.021-58.969 5.582-87.584 15.222-1.185-2.302-1.795-4.362-2.769-6.233-4.398-8.449-6.703-18.174-14.942-24.299-2.511-1.866-5.103-3.814-7.047-6.218-8.358-10.332-17.028-20.276-28.772-26.973 4.423-11.478 9.299-22.806 13.151-34.473 4.406-13.348 6.724-27.18 6.998-41.313.098-5.093.643-10.176 1.06-15.722z"></path><path fill="#e94c32" d="M981.557 392.109c-1.172 15.337-2.617 30.625-4.438 45.869-2.213 18.521-4.144 37.108-7.346 55.466-2.123 12.171-6.28 23.985-9.492 35.968-2.439 9.098-6.794 18.224-6.73 27.313.078 11.051-5.7 19.776-7.81 29.825-3.292 15.677-10.255 30.082-18.775 43.704-2.383 3.81-2.458 8.091-2.577 12.742-.144 5.6-3.049 11.38-5.691 16.611-4.621 9.149-10.033 17.896-14.966 26.892-4.517 8.239-8.715 16.635-15.8 23.153-3.034 2.791-5.629 6.06-8.735 9.255-12.197-10.595-21.071-23.644-29.301-37.24-7.608-12.569-13.282-25.962-17.637-40.37 13.303-6.889 25.873-13.878 35.311-25.315.717-.869 1.934-1.312 2.71-2.147 5.025-5.405 10.515-10.481 14.854-16.397 6.141-8.374 10.861-17.813 17.206-26.008 8.22-10.618 13.657-22.643 20.024-34.466 4.448-.626 6.729-3.21 8.114-6.89 1.455-3.866 2.644-7.895 4.609-11.492 4.397-8.05 9.641-15.659 13.708-23.86 3.354-6.761 5.511-14.116 8.203-21.206 5.727-15.082 7.277-31.248 12.521-46.578 3.704-10.828 3.138-23.116 4.478-34.753l7.56-.073z"></path><path fill="#f7a617" d="M1918.661 831.99c-4.937 16.58-9.971 33.057-22.196 46.104-15.952 17.025-28.099 36.791-40.382 56.471-2.864 4.59-6.481 8.825-10.3 12.681-8.947 9.031-17.279 19.094-27.583 26.261-17.103 11.896-35.564 21.84-53.441 32.624-1.419.856-3.132 1.571-4.065 2.828-6.904 9.308-18.6 11.178-27.297 17.714-2.705 2.033-6.319 2.856-9.874 4.281-3.413-9.821-6.916-19.583-9.36-29.602-1.533-6.284-1.474-12.957-1.665-19.913 1.913-.78 3.374-1.057 4.81-1.431 15.822-4.121 31.491-8.029 43.818-20.323 9.452-9.426 20.371-17.372 30.534-26.097 6.146-5.277 13.024-10.052 17.954-16.326 14.812-18.848 28.876-38.285 43.112-57.581 2.624-3.557 5.506-7.264 6.83-11.367 2.681-8.311 4.375-16.94 6.476-25.438 17.89.279 35.333 3.179 52.629 9.113z"></path><path fill="#ea553a" d="M1172.91 977.582c-15.775-3.127-28.215-12.377-40.227-22.43-9.005-7.537-18.43-14.605-27.071-22.532-5.07-4.651-9.143-10.443-13.361-15.955-7.647-9.994-15.291-20.007-22.456-30.345-2.361-3.407-3.792-7.72-4.696-11.829-3.119-14.183-5.848-28.453-8.651-42.704-.636-3.236-.974-6.53-1.452-10.209 15.234-2.19 30.471-3.969 46.408-5.622 2.692 5.705 4.882 11.222 6.63 16.876 2.9 9.381 7.776 17.194 15.035 24.049 7.056 6.662 13.305 14.311 19.146 22.099 9.509 12.677 23.01 19.061 36.907 25.054-1.048 7.441-2.425 14.854-3.066 22.33-.956 11.162-1.393 22.369-2.052 33.557l-1.096 17.661z"></path><path fill="#ea5453" d="M1163.123 704.036c-4.005 5.116-7.685 10.531-12.075 15.293-12.842 13.933-27.653 25.447-44.902 34.538-3.166-5.708-5.656-11.287-8.189-17.251-3.321-12.857-6.259-25.431-9.963-37.775-4.6-15.329-10.6-30.188-11.349-46.562-.314-6.871-1.275-14.287-7.114-19.644-1.047-.961-1.292-3.053-1.465-4.67l-4.092-39.927c-.554-5.245-.383-10.829-2.21-15.623-3.622-9.503-4.546-19.253-4.688-29.163-.088-6.111 1.068-12.256.782-18.344-.67-14.281-1.76-28.546-2.9-42.8-.657-8.222-1.951-16.395-2.564-24.62-.458-6.137-.285-12.322-.104-18.21.959 5.831 1.076 11.525 2.429 16.909 2.007 7.986 5.225 15.664 7.324 23.632 3.222 12.23 1.547 25.219 6.728 37.355 4.311 10.099 6.389 21.136 9.732 31.669 2.228 7.02 6.167 13.722 7.121 20.863 1.119 8.376 6.1 13.974 10.376 20.716l2.026 10.576c1.711 9.216 3.149 18.283 8.494 26.599 6.393 9.946 11.348 20.815 16.943 31.276 4.021 7.519 6.199 16.075 12.925 22.065l24.462 22.26c.556.503 1.507.571 2.274.841z"></path><path fill="#ea5b15" d="M1285.092 163.432c9.165 3.148 18.419 6.374 27.279 10.459 4.871 2.246 8.838 6.406 13.646 8.851 5.446 2.77 11.801 3.874 17.011 6.965 11.514 6.831 24.097 9.942 36.968 12.471 1.78.35 3.777.576 5.213 1.542 10.784 7.255 23.448 9.114 35.622 11.834 9.977 2.23 18.529 6.703 26.988 11.898 5.233 3.214 10.76 5.983 15.798 9.468 4.14 2.864 7.962 6.279 11.551 9.827 5.076 5.02 10.056 10.181 14.624 15.658 5.822 6.98 11.119 14.395 16.78 21.513 4.531 5.698 9.267 11.233 14.222 16.987-10.005 5.806-20.07 12.004-30.719 16.943-7.694 3.569-16.163 5.464-24.688 7.669-2.878-7.088-5.352-13.741-7.833-20.392-.802-2.15-1.244-4.55-2.498-6.396-4.548-6.7-9.712-12.999-14.011-19.847-6.672-10.627-15.34-18.93-26.063-25.376-9.357-5.625-18.367-11.824-27.644-17.587-6.436-3.997-12.902-8.006-19.659-11.405-5.123-2.577-11.107-3.536-16.046-6.37-17.187-9.863-35.13-17.887-54.031-23.767-4.403-1.37-8.953-2.267-13.436-3.382l.926-27.565z"></path><path fill="#ea504b" d="M1098 737l7.789 16.893c-15.04 9.272-31.679 15.004-49.184 17.995-9.464 1.617-19.122 2.097-29.151 3.019-.457-10.636-.18-21.211-.544-31.764-.273-7.888-.409-15.883-4.736-23.103-1.16-1.936-1.162-4.805-1.06-7.219l1.787-36.207c.182-8.103-.993-16.237-.811-24.34.365-16.236 1.253-32.461 1.908-48.69.484-12 .942-24.001 1.98-36.069 5.57 10.19 10.632 20.42 15.528 30.728 1.122 2.362 2.587 5.09 2.339 7.488-1.536 14.819 5.881 26.839 12.962 38.33 10.008 16.241 16.417 33.54 20.331 51.964 2.285 10.756 4.729 21.394 11.958 30.165L1098 737z"></path><path fill="#f6a320" d="M1865.78 822.529c-1.849 8.846-3.544 17.475-6.224 25.786-1.323 4.102-4.206 7.81-6.83 11.367l-43.112 57.581c-4.93 6.273-11.808 11.049-17.954 16.326-10.162 8.725-21.082 16.671-30.534 26.097-12.327 12.294-27.997 16.202-43.818 20.323-1.436.374-2.897.651-4.744.986-1.107-17.032-1.816-34.076-2.079-51.556 1.265-.535 2.183-.428 2.888-.766 10.596-5.072 20.8-11.059 32.586-13.273 1.69-.317 3.307-1.558 4.732-2.662l26.908-21.114c4.992-4.003 11.214-7.393 14.381-12.585 11.286-18.5 22.363-37.263 27.027-58.87l36.046 1.811c3.487.165 6.983.14 10.727.549z"></path><path fill="#ec6333" d="M318.448 922.814c-6.374-2.074-12.56-4.058-18.412-6.765-8.379-3.876-16.906-7.675-24.617-12.668-5.239-3.392-9.69-8.381-13.609-13.352-7.87-9.983-14.953-20.582-22.699-30.666-8.061-10.493-13.909-22.097-18.636-34.358-.595-1.543-1.486-2.972-2.382-4.783 6.84-1.598 13.797-3.023 20.807-4.106 18.852-2.912 36.433-9.493 53.737-17.819.697.888.889 1.555 1.292 2.051l17.921 21.896c4.14 4.939 8.06 10.191 12.862 14.412 5.67 4.984 12.185 9.007 18.334 13.447-8.937 16.282-16.422 33.178-20.696 51.31-1.638 6.951-2.402 14.107-3.903 21.403z"></path><path fill="#f49700" d="M623.467 326.903c2.893-10.618 5.584-21.446 9.833-31.623 3.013-7.217 7.924-13.696 12.358-20.254 6.375-9.43 12.026-19.67 19.886-27.705 14.12-14.434 28.063-29.453 47.926-36.784 6.581-2.429 12.344-6.994 18.774-9.942 3.975-1.822 8.503-2.436 13.186-3.592 1.947 18.557 3.248 37.15 8.307 55.686-15.453 7.931-28.853 18.092-40.46 29.996-10.417 10.683-19.109 23.111-28.013 35.175-3.238 4.388-4.888 9.948-7.262 14.973-17.803-3.987-35.767-6.498-54.535-5.931z"></path><path fill="#ea544c" d="M1097.956 736.615c-2.925-3.218-5.893-6.822-8.862-10.425-7.229-8.771-9.672-19.409-11.958-30.165-3.914-18.424-10.323-35.722-20.331-51.964-7.081-11.491-14.498-23.511-12.962-38.33.249-2.398-1.217-5.126-2.339-7.488l-15.232-31.019-3.103-34.338c-.107-1.316-.041-2.653.031-3.975.233-4.294.756-8.59.702-12.879-.072-5.713-.776-11.417-.861-17.13l-.116-30.733c-.329-10.088-1.926-20.166-1.768-30.23.23-14.674.599-29.31-1.162-44.341 9.369-.803 18.741-1.179 28.558-1.074 1.446 15.814 2.446 31.146 3.446 46.478.108 6.163-.064 12.348.393 18.485.613 8.225 1.907 16.397 2.564 24.62l2.9 42.8c.286 6.088-.869 12.234-.782 18.344.142 9.91 1.066 19.661 4.688 29.163 1.827 4.794 1.657 10.377 2.21 15.623l4.092 39.927c.172 1.617.417 3.71 1.465 4.67 5.839 5.357 6.8 12.773 7.114 19.644.749 16.374 6.749 31.233 11.349 46.562 3.704 12.344 6.642 24.918 9.963 37.775z"></path><path fill="#ec5c61" d="M1204.835 568.008c1.254 25.351-1.675 50.16-10.168 74.61-8.598-4.883-18.177-8.709-24.354-15.59-7.44-8.289-13.929-17.442-21.675-25.711-8.498-9.072-16.731-18.928-21.084-31.113-.54-1.513-1.691-2.807-2.594-4.564-4.605-9.247-7.706-18.544-7.96-29.09-.835-7.149-1.214-13.944-2.609-20.523-2.215-10.454-5.626-20.496-7.101-31.302-2.513-18.419-7.207-36.512-5.347-55.352.24-2.43-.17-4.949-.477-7.402l-4.468-34.792c2.723-.379 5.446-.757 8.585-.667 1.749 8.781 2.952 17.116 4.448 25.399 1.813 10.037 3.64 20.084 5.934 30.017 1.036 4.482 3.953 8.573 4.73 13.064 1.794 10.377 4.73 20.253 9.272 29.771 2.914 6.105 4.761 12.711 7.496 18.912 2.865 6.496 6.264 12.755 9.35 19.156 3.764 7.805 7.667 15.013 16.1 19.441 7.527 3.952 13.713 10.376 20.983 14.924 6.636 4.152 13.932 7.25 20.937 10.813z"></path><path fill="#ed676f" d="M1140.75 379.231c18.38-4.858 36.222-11.21 53.979-18.971 3.222 3.368 5.693 6.744 8.719 9.512 2.333 2.134 5.451 5.07 8.067 4.923 7.623-.429 12.363 2.688 17.309 8.215 5.531 6.18 12.744 10.854 19.224 16.184-5.121 7.193-10.461 14.241-15.323 21.606-13.691 20.739-22.99 43.255-26.782 67.926-.543 3.536-1.281 7.043-2.366 10.925-14.258-6.419-26.411-14.959-32.731-29.803-1.087-2.553-2.596-4.93-3.969-7.355-1.694-2.993-3.569-5.89-5.143-8.943-1.578-3.062-2.922-6.249-4.295-9.413-1.57-3.621-3.505-7.163-4.47-10.946-1.257-4.93-.636-10.572-2.725-15.013-5.831-12.397-7.467-25.628-9.497-38.847z"></path><path fill="#ed656e" d="M1254.103 647.439c5.325.947 10.603 2.272 15.847 3.722 5.101 1.41 10.376 2.475 15.175 4.596 3.237 1.431 5.942 4.262 8.589 6.777 2.592 2.462 4.77 5.355 7.207 7.987 1.804 1.948 4.557 3.453 5.461 5.723 3.51 8.817 11.581 11.307 19.059 14.735 1.053.483 2.116.963 3.214 1.327 9.172 3.043 13.818 8.587 14.889 18.979.715 6.935 5.607 13.679 9.479 19.987 4.623 7.533 9.175 14.819 9.091 24.116-.023 2.55 1.21 5.111 1.874 8.055-19.861 2.555-39.795 4.296-59.597 9.09l-11.596-23.203c-1.107-2.169-2.526-4.353-4.307-5.975-7.349-6.694-14.863-13.209-22.373-19.723l-17.313-14.669c-2.776-2.245-5.935-4.017-8.92-6.003l11.609-38.185c1.508-5.453 1.739-11.258 2.613-17.336z"></path><path fill="#ec6168" d="M1140.315 379.223c2.464 13.227 4.101 26.459 9.931 38.856 2.089 4.441 1.468 10.083 2.725 15.013.965 3.783 2.9 7.325 4.47 10.946 1.372 3.164 2.716 6.351 4.295 9.413 1.574 3.053 3.449 5.95 5.143 8.943 1.372 2.425 2.882 4.803 3.969 7.355 6.319 14.844 18.473 23.384 32.641 30.212.067 5.121-.501 10.201-.435 15.271l.985 38.117c.151 4.586.616 9.162.868 14.201-7.075-3.104-14.371-6.202-21.007-10.354-7.269-4.548-13.456-10.972-20.983-14.924-8.434-4.428-12.337-11.637-16.1-19.441-3.087-6.401-6.485-12.66-9.35-19.156-2.735-6.201-4.583-12.807-7.496-18.912-4.542-9.518-7.477-19.394-9.272-29.771-.777-4.491-3.694-8.581-4.73-13.064-2.294-9.933-4.121-19.98-5.934-30.017-1.496-8.283-2.699-16.618-4.036-25.335 10.349-2.461 20.704-4.511 31.054-6.582.957-.191 1.887-.515 3.264-.769z"></path><path fill="#e94c28" d="M922 537c-6.003 11.784-11.44 23.81-19.66 34.428-6.345 8.196-11.065 17.635-17.206 26.008-4.339 5.916-9.828 10.992-14.854 16.397-.776.835-1.993 1.279-2.71 2.147-9.439 11.437-22.008 18.427-35.357 24.929-4.219-10.885-6.942-22.155-7.205-33.905l-.514-49.542c7.441-2.893 14.452-5.197 21.334-7.841 1.749-.672 3.101-2.401 4.604-3.681 6.749-5.745 12.845-12.627 20.407-16.944 7.719-4.406 14.391-9.101 18.741-16.889.626-1.122 1.689-2.077 2.729-2.877 7.197-5.533 12.583-12.51 16.906-20.439.68-1.247 2.495-1.876 4.105-2.651 2.835 1.408 5.267 2.892 7.884 3.892 3.904 1.491 4.392 3.922 2.833 7.439-1.47 3.318-2.668 6.756-4.069 10.106-1.247 2.981-.435 5.242 2.413 6.544 2.805 1.282 3.125 3.14 1.813 5.601l-6.907 12.799L922 537z"></path><path fill="#eb5659" d="M1124.995 566c.868 1.396 2.018 2.691 2.559 4.203 4.353 12.185 12.586 22.041 21.084 31.113 7.746 8.269 14.235 17.422 21.675 25.711 6.176 6.881 15.756 10.707 24.174 15.932-6.073 22.316-16.675 42.446-31.058 60.937-1.074-.131-2.025-.199-2.581-.702l-24.462-22.26c-6.726-5.99-8.904-14.546-12.925-22.065-5.594-10.461-10.55-21.33-16.943-31.276-5.345-8.315-6.783-17.383-8.494-26.599-.63-3.394-1.348-6.772-1.738-10.848-.371-6.313-1.029-11.934-1.745-18.052l6.34 4.04 1.288-.675-2.143-15.385 9.454 1.208v-8.545L1124.995 566z"></path><path fill="#f5a02d" d="M1818.568 820.096c-4.224 21.679-15.302 40.442-26.587 58.942-3.167 5.192-9.389 8.582-14.381 12.585l-26.908 21.114c-1.425 1.104-3.042 2.345-4.732 2.662-11.786 2.214-21.99 8.201-32.586 13.273-.705.338-1.624.231-2.824.334a824.35 824.35 0 0 1-8.262-42.708c4.646-2.14 9.353-3.139 13.269-5.47 5.582-3.323 11.318-6.942 15.671-11.652 7.949-8.6 14.423-18.572 22.456-27.081 8.539-9.046 13.867-19.641 18.325-30.922l46.559 8.922z"></path><path fill="#eb5a57" d="M1124.96 565.639c-5.086-4.017-10.208-8.395-15.478-12.901v8.545l-9.454-1.208 2.143 15.385-1.288.675-6.34-4.04c.716 6.118 1.375 11.74 1.745 17.633-4.564-6.051-9.544-11.649-10.663-20.025-.954-7.141-4.892-13.843-7.121-20.863-3.344-10.533-5.421-21.57-9.732-31.669-5.181-12.135-3.506-25.125-6.728-37.355-2.099-7.968-5.317-15.646-7.324-23.632-1.353-5.384-1.47-11.078-2.429-16.909l-3.294-46.689a278.63 278.63 0 0 1 27.57-2.084c2.114 12.378 3.647 24.309 5.479 36.195 1.25 8.111 2.832 16.175 4.422 24.23 1.402 7.103 2.991 14.169 4.55 21.241 1.478 6.706.273 14.002 4.6 20.088 5.401 7.597 7.176 16.518 9.467 25.337 1.953 7.515 5.804 14.253 11.917 19.406.254 10.095 3.355 19.392 7.96 28.639z"></path><path fill="#ea541c" d="M911.651 810.999c-2.511 10.165-5.419 20.146-8.2 30.162-2.503 9.015-7.37 16.277-14.364 22.612-6.108 5.533-10.917 12.475-16.796 18.293-6.942 6.871-14.354 13.24-19.083 22.03-.644 1.196-2.222 1.889-3.705 2.857-2.39-7.921-4.101-15.991-6.566-23.823-5.451-17.323-12.404-33.976-23.414-48.835l21.627-21.095c3.182-3.29 5.532-7.382 8.295-11.083l10.663-14.163c9.528 4.78 18.925 9.848 28.625 14.247 7.324 3.321 15.036 5.785 22.917 8.799z"></path><path fill="#eb5d19" d="M1284.092 191.421c4.557.69 9.107 1.587 13.51 2.957 18.901 5.881 36.844 13.904 54.031 23.767 4.938 2.834 10.923 3.792 16.046 6.37 6.757 3.399 13.224 7.408 19.659 11.405l27.644 17.587c10.723 6.446 19.392 14.748 26.063 25.376 4.299 6.848 9.463 13.147 14.011 19.847 1.254 1.847 1.696 4.246 2.498 6.396l7.441 20.332c-11.685 1.754-23.379 3.133-35.533 4.037-.737-2.093-.995-3.716-1.294-5.33-3.157-17.057-14.048-30.161-23.034-44.146-3.027-4.71-7.786-8.529-12.334-11.993-9.346-7.116-19.004-13.834-28.688-20.491-6.653-4.573-13.311-9.251-20.431-13.002-8.048-4.24-16.479-7.85-24.989-11.091-11.722-4.465-23.673-8.328-35.527-12.449l.927-19.572z"></path><path fill="#eb5e24" d="M1283.09 211.415c11.928 3.699 23.88 7.562 35.602 12.027 8.509 3.241 16.941 6.852 24.989 11.091 7.12 3.751 13.778 8.429 20.431 13.002 9.684 6.657 19.342 13.375 28.688 20.491 4.548 3.463 9.307 7.283 12.334 11.993 8.986 13.985 19.877 27.089 23.034 44.146.299 1.615.557 3.237.836 5.263-13.373-.216-26.749-.839-40.564-1.923-2.935-9.681-4.597-18.92-12.286-26.152-15.577-14.651-30.4-30.102-45.564-45.193-.686-.683-1.626-1.156-2.516-1.584l-47.187-22.615 2.203-20.546z"></path><path fill="#e9511f" d="M913 486.001c-1.29.915-3.105 1.543-3.785 2.791-4.323 7.929-9.709 14.906-16.906 20.439-1.04.8-2.103 1.755-2.729 2.877-4.35 7.788-11.022 12.482-18.741 16.889-7.562 4.317-13.658 11.199-20.407 16.944-1.503 1.28-2.856 3.009-4.604 3.681-6.881 2.643-13.893 4.948-21.262 7.377-.128-11.151.202-22.302.378-33.454.03-1.892-.6-3.795-.456-6.12 13.727-1.755 23.588-9.527 33.278-17.663 2.784-2.337 6.074-4.161 8.529-6.784l29.057-31.86c1.545-1.71 3.418-3.401 4.221-5.459 5.665-14.509 11.49-28.977 16.436-43.736 2.817-8.407 4.074-17.338 6.033-26.032 5.039.714 10.078 1.427 15.536 2.629-.909 8.969-2.31 17.438-3.546 25.931-2.41 16.551-5.84 32.839-11.991 48.461L913 486.001z"></path><path fill="#ea5741" d="M1179.451 903.828c-14.224-5.787-27.726-12.171-37.235-24.849-5.841-7.787-12.09-15.436-19.146-22.099-7.259-6.854-12.136-14.667-15.035-24.049-1.748-5.654-3.938-11.171-6.254-17.033 15.099-4.009 30.213-8.629 44.958-15.533l28.367 36.36c6.09 8.015 13.124 14.75 22.72 18.375-7.404 14.472-13.599 29.412-17.48 45.244-.271 1.106-.382 2.25-.895 3.583z"></path><path fill="#ea522a" d="M913.32 486.141c2.693-7.837 5.694-15.539 8.722-23.231 6.151-15.622 9.581-31.91 11.991-48.461l3.963-25.861c7.582.317 15.168 1.031 22.748 1.797 4.171.421 8.333.928 12.877 1.596-.963 11.836-.398 24.125-4.102 34.953-5.244 15.33-6.794 31.496-12.521 46.578-2.692 7.09-4.849 14.445-8.203 21.206-4.068 8.201-9.311 15.81-13.708 23.86-1.965 3.597-3.154 7.627-4.609 11.492-1.385 3.68-3.666 6.265-8.114 6.89-1.994-1.511-3.624-3.059-5.077-4.44l6.907-12.799c1.313-2.461.993-4.318-1.813-5.601-2.849-1.302-3.66-3.563-2.413-6.544 1.401-3.35 2.599-6.788 4.069-10.106 1.558-3.517 1.071-5.948-2.833-7.439-2.617-1-5.049-2.484-7.884-3.892z"></path><path fill="#eb5e24" d="M376.574 714.118c12.053 6.538 20.723 16.481 29.081 26.814 1.945 2.404 4.537 4.352 7.047 6.218 8.24 6.125 10.544 15.85 14.942 24.299.974 1.871 1.584 3.931 2.376 6.29-7.145 3.719-14.633 6.501-21.386 10.517-9.606 5.713-18.673 12.334-28.425 18.399-3.407-3.73-6.231-7.409-9.335-10.834l-30.989-33.862c11.858-11.593 22.368-24.28 31.055-38.431 1.86-3.031 3.553-6.164 5.632-9.409z"></path><path fill="#e95514" d="M859.962 787.636c-3.409 5.037-6.981 9.745-10.516 14.481-2.763 3.701-5.113 7.792-8.295 11.083-6.885 7.118-14.186 13.834-21.65 20.755-13.222-17.677-29.417-31.711-48.178-42.878-.969-.576-2.068-.934-3.27-1.709 6.28-8.159 12.733-15.993 19.16-23.849 1.459-1.783 2.718-3.738 4.254-5.448l18.336-19.969c4.909 5.34 9.619 10.738 14.081 16.333 9.72 12.19 21.813 21.566 34.847 29.867.411.262.725.674 1.231 1.334z"></path><path fill="#eb5f2d" d="M339.582 762.088l31.293 33.733c3.104 3.425 5.928 7.104 9.024 10.979-12.885 11.619-24.548 24.139-33.899 38.704-.872 1.359-1.56 2.837-2.644 4.428-6.459-4.271-12.974-8.294-18.644-13.278-4.802-4.221-8.722-9.473-12.862-14.412l-17.921-21.896c-.403-.496-.595-1.163-.926-2.105 16.738-10.504 32.58-21.87 46.578-36.154z"></path><path fill="#f28d00" d="M678.388 332.912c1.989-5.104 3.638-10.664 6.876-15.051 8.903-12.064 17.596-24.492 28.013-35.175 11.607-11.904 25.007-22.064 40.507-29.592 4.873 11.636 9.419 23.412 13.67 35.592-5.759 4.084-11.517 7.403-16.594 11.553-4.413 3.607-8.124 8.092-12.023 12.301-5.346 5.772-10.82 11.454-15.782 17.547-3.929 4.824-7.17 10.208-10.716 15.344l-33.95-12.518z"></path><path fill="#f08369" d="M1580.181 771.427c-.191-.803-.322-1.377-.119-1.786 5.389-10.903 9.084-22.666 18.181-31.587 6.223-6.103 11.276-13.385 17.286-19.727 3.117-3.289 6.933-6.105 10.869-8.384 6.572-3.806 13.492-7.009 20.461-10.752 1.773 3.23 3.236 6.803 4.951 10.251l12.234 24.993c-1.367 1.966-2.596 3.293-3.935 4.499-7.845 7.07-16.315 13.564-23.407 21.32-6.971 7.623-12.552 16.517-18.743 24.854l-37.777-13.68z"></path><path fill="#f18b5e" d="M1618.142 785.4c6.007-8.63 11.588-17.524 18.559-25.147 7.092-7.755 15.562-14.249 23.407-21.32 1.338-1.206 2.568-2.534 3.997-4.162l28.996 33.733c1.896 2.205 4.424 3.867 6.66 6.394-6.471 7.492-12.967 14.346-19.403 21.255l-18.407 19.953c-12.958-12.409-27.485-22.567-43.809-30.706z"></path><path fill="#f49c3a" d="M1771.617 811.1c-4.066 11.354-9.394 21.949-17.933 30.995-8.032 8.509-14.507 18.481-22.456 27.081-4.353 4.71-10.089 8.329-15.671 11.652-3.915 2.331-8.623 3.331-13.318 5.069-4.298-9.927-8.255-19.998-12.1-30.743 4.741-4.381 9.924-7.582 13.882-11.904 7.345-8.021 14.094-16.603 20.864-25.131 4.897-6.168 9.428-12.626 14.123-18.955l32.61 11.936z"></path><path fill="#f08000" d="M712.601 345.675c3.283-5.381 6.524-10.765 10.453-15.589 4.962-6.093 10.435-11.774 15.782-17.547 3.899-4.21 7.61-8.695 12.023-12.301 5.078-4.15 10.836-7.469 16.636-11.19a934.12 934.12 0 0 1 23.286 35.848c-4.873 6.234-9.676 11.895-14.63 17.421l-25.195 27.801c-11.713-9.615-24.433-17.645-38.355-24.443z"></path><path fill="#ed6e04" d="M751.11 370.42c8.249-9.565 16.693-18.791 25.041-28.103 4.954-5.526 9.757-11.187 14.765-17.106 7.129 6.226 13.892 13.041 21.189 19.225 5.389 4.567 11.475 8.312 17.53 12.92-5.51 7.863-10.622 15.919-17.254 22.427-8.881 8.716-18.938 16.233-28.49 24.264-5.703-6.587-11.146-13.427-17.193-19.682-4.758-4.921-10.261-9.121-15.587-13.944z"></path><path fill="#ea541c" d="M921.823 385.544c-1.739 9.04-2.995 17.971-5.813 26.378-4.946 14.759-10.771 29.227-16.436 43.736-.804 2.058-2.676 3.749-4.221 5.459l-29.057 31.86c-2.455 2.623-5.745 4.447-8.529 6.784-9.69 8.135-19.551 15.908-33.208 17.237-1.773-9.728-3.147-19.457-4.091-29.6l36.13-16.763c.581-.267 1.046-.812 1.525-1.269 8.033-7.688 16.258-15.19 24.011-23.152 4.35-4.467 9.202-9.144 11.588-14.69 6.638-15.425 15.047-30.299 17.274-47.358 3.536.344 7.072.688 10.829 1.377z"></path><path fill="#f3944d" d="M1738.688 798.998c-4.375 6.495-8.906 12.953-13.803 19.121-6.771 8.528-13.519 17.11-20.864 25.131-3.958 4.322-9.141 7.523-13.925 11.54-8.036-13.464-16.465-26.844-27.999-38.387 5.988-6.951 12.094-13.629 18.261-20.25l19.547-20.95 38.783 23.794z"></path><path fill="#ec6168" d="M1239.583 703.142c3.282 1.805 6.441 3.576 9.217 5.821 5.88 4.755 11.599 9.713 17.313 14.669l22.373 19.723c1.781 1.622 3.2 3.806 4.307 5.975 3.843 7.532 7.477 15.171 11.194 23.136-10.764 4.67-21.532 8.973-32.69 12.982l-22.733-27.366c-2.003-2.416-4.096-4.758-6.194-7.093-3.539-3.94-6.927-8.044-10.74-11.701-2.57-2.465-5.762-4.283-8.675-6.39l16.627-29.755z"></path><path fill="#ec663e" d="M1351.006 332.839l-28.499 10.33c-.294.107-.533.367-1.194.264-11.067-19.018-27.026-32.559-44.225-44.855-4.267-3.051-8.753-5.796-13.138-8.682l9.505-24.505c10.055 4.069 19.821 8.227 29.211 13.108 3.998 2.078 7.299 5.565 10.753 8.598 3.077 2.701 5.743 5.891 8.926 8.447 4.116 3.304 9.787 5.345 12.62 9.432 6.083 8.777 10.778 18.517 16.041 27.863z"></path><path fill="#eb5e5b" d="M1222.647 733.051c3.223 1.954 6.415 3.771 8.985 6.237 3.813 3.658 7.201 7.761 10.74 11.701l6.194 7.093 22.384 27.409c-13.056 6.836-25.309 14.613-36.736 24.161l-39.323-44.7 24.494-27.846c1.072-1.224 1.974-2.598 3.264-4.056z"></path><path fill="#ea580e" d="M876.001 376.171c5.874 1.347 11.748 2.694 17.812 4.789-.81 5.265-2.687 9.791-2.639 14.296.124 11.469-4.458 20.383-12.73 27.863-2.075 1.877-3.659 4.286-5.668 6.248l-22.808 21.967c-.442.422-1.212.488-1.813.757l-23.113 10.389-9.875 4.514c-2.305-6.09-4.609-12.181-6.614-18.676 7.64-4.837 15.567-8.54 22.18-13.873 9.697-7.821 18.931-16.361 27.443-25.455 5.613-5.998 12.679-11.331 14.201-20.475.699-4.2 2.384-8.235 3.623-12.345z"></path><path fill="#e95514" d="M815.103 467.384c3.356-1.894 6.641-3.415 9.94-4.903l23.113-10.389c.6-.269 1.371-.335 1.813-.757l22.808-21.967c2.008-1.962 3.593-4.371 5.668-6.248 8.272-7.48 12.854-16.394 12.73-27.863-.049-4.505 1.828-9.031 2.847-13.956 5.427.559 10.836 1.526 16.609 2.68-1.863 17.245-10.272 32.119-16.91 47.544-2.387 5.546-7.239 10.223-11.588 14.69-7.753 7.962-15.978 15.464-24.011 23.152-.478.458-.944 1.002-1.525 1.269l-36.069 16.355c-2.076-6.402-3.783-12.81-5.425-19.607z"></path><path fill="#eb620b" d="M783.944 404.402c9.499-8.388 19.556-15.905 28.437-24.621 6.631-6.508 11.744-14.564 17.575-22.273 9.271 4.016 18.501 8.375 27.893 13.43-4.134 7.07-8.017 13.778-12.833 19.731-5.785 7.15-12.109 13.917-18.666 20.376-7.99 7.869-16.466 15.244-24.731 22.832l-17.674-29.475z"></path><path fill="#ea544c" d="M1197.986 854.686c-9.756-3.309-16.79-10.044-22.88-18.059l-28.001-36.417c8.601-5.939 17.348-11.563 26.758-17.075 1.615 1.026 2.639 1.876 3.505 2.865l26.664 30.44c3.723 4.139 7.995 7.785 12.017 11.656l-18.064 26.591z"></path><path fill="#ec6333" d="M1351.41 332.903c-5.667-9.409-10.361-19.149-16.445-27.926-2.833-4.087-8.504-6.128-12.62-9.432-3.184-2.555-5.849-5.745-8.926-8.447-3.454-3.033-6.756-6.52-10.753-8.598-9.391-4.88-19.157-9.039-29.138-13.499 1.18-5.441 2.727-10.873 4.81-16.607 11.918 4.674 24.209 8.261 34.464 14.962 14.239 9.304 29.011 18.453 39.595 32.464 2.386 3.159 5.121 6.077 7.884 8.923 6.564 6.764 10.148 14.927 11.723 24.093l-20.594 4.067z"></path><path fill="#eb5e5b" d="M1117 536.549c-6.113-4.702-9.965-11.44-11.917-18.955-2.292-8.819-4.066-17.74-9.467-25.337-4.327-6.085-3.122-13.382-4.6-20.088l-4.55-21.241c-1.59-8.054-3.172-16.118-4.422-24.23l-5.037-36.129c6.382-1.43 12.777-2.462 19.582-3.443 1.906 11.646 3.426 23.24 4.878 34.842.307 2.453.717 4.973.477 7.402-1.86 18.84 2.834 36.934 5.347 55.352 1.474 10.806 4.885 20.848 7.101 31.302 1.394 6.579 1.774 13.374 2.609 20.523z"></path><path fill="#ec644b" d="M1263.638 290.071c4.697 2.713 9.183 5.458 13.45 8.509 17.199 12.295 33.158 25.836 43.873 44.907-8.026 4.725-16.095 9.106-24.83 13.372-11.633-15.937-25.648-28.515-41.888-38.689-1.609-1.008-3.555-1.48-5.344-2.2 2.329-3.852 4.766-7.645 6.959-11.573l7.78-14.326z"></path><path fill="#eb5f2d" d="M1372.453 328.903c-2.025-9.233-5.608-17.396-12.172-24.16-2.762-2.846-5.498-5.764-7.884-8.923-10.584-14.01-25.356-23.16-39.595-32.464-10.256-6.701-22.546-10.289-34.284-15.312.325-5.246 1.005-10.444 2.027-15.863l47.529 22.394c.89.428 1.83.901 2.516 1.584l45.564 45.193c7.69 7.233 9.352 16.472 11.849 26.084-5.032.773-10.066 1.154-15.55 1.466z"></path><path fill="#e95a0f" d="M801.776 434.171c8.108-7.882 16.584-15.257 24.573-23.126 6.558-6.459 12.881-13.226 18.666-20.376 4.817-5.953 8.7-12.661 13.011-19.409 5.739 1.338 11.463 3.051 17.581 4.838-.845 4.183-2.53 8.219-3.229 12.418-1.522 9.144-8.588 14.477-14.201 20.475-8.512 9.094-17.745 17.635-27.443 25.455-6.613 5.333-14.54 9.036-22.223 13.51-2.422-4.469-4.499-8.98-6.735-13.786z"></path><path fill="#eb5e5b" d="M1248.533 316.002c2.155.688 4.101 1.159 5.71 2.168 16.24 10.174 30.255 22.752 41.532 38.727-7.166 5.736-14.641 11.319-22.562 16.731-1.16-1.277-1.684-2.585-2.615-3.46l-38.694-36.2 14.203-15.029c.803-.86 1.38-1.93 2.427-2.936z"></path><path fill="#eb5a57" d="M1216.359 827.958c-4.331-3.733-8.603-7.379-12.326-11.518l-26.664-30.44c-.866-.989-1.89-1.839-3.152-2.902 6.483-6.054 13.276-11.959 20.371-18.005l39.315 44.704c-5.648 6.216-11.441 12.12-17.544 18.161z"></path><path fill="#ec6168" d="M1231.598 334.101l38.999 36.066c.931.876 1.456 2.183 2.303 3.608-4.283 4.279-8.7 8.24-13.769 12.091-4.2-3.051-7.512-6.349-11.338-8.867-12.36-8.136-22.893-18.27-32.841-29.093l16.646-13.805z"></path><path fill="#ed656e" d="M1214.597 347.955c10.303 10.775 20.836 20.908 33.196 29.044 3.825 2.518 7.137 5.816 10.992 8.903-3.171 4.397-6.65 8.648-10.432 13.046-6.785-5.184-13.998-9.858-19.529-16.038-4.946-5.527-9.687-8.644-17.309-8.215-2.616.147-5.734-2.788-8.067-4.923-3.026-2.769-5.497-6.144-8.35-9.568 6.286-4.273 12.715-8.237 19.499-12.25z"></path></svg>
</p>
<p align="center">
<b>The crispy rerank family from <a href="https://mixedbread.com"><b>Mixedbread</b></a>.</b>
</p>
<p align="center">
<sup> 🍞 Looking for a simple end-to-end retrieval solution? Meet Omni, our multimodal and multilingual model. <a href="https://mixedbread.com"><b>Get in touch for access.</a> </sup>
</p>
# 🍞 mxbai-rerank-large-v2
This is the large model in our family of powerful reranker models. You can learn more about the models in our [blog post](https://www.mixedbread.ai/blog/mxbai-rerank-v2).
We have two models:
- [mxbai-rerank-base-v2](https://huggingface.co/mixedbread-ai/mxbai-rerank-base-v2)
- [mxbai-rerank-large-v2](https://huggingface.co/mixedbread-ai/mxbai-rerank-large-v2) (🍞)
**The technical report is coming soon!**
## 🌟 Features
- state-of-the-art performance and strong efficiency
- multilingual support (100+ languages, outstanding English and Chinese performance)
- code support
- long-context support
## ⚙️ Usage
1. Install mxbai-rerank
```bash
pip install mxbai-rerank
```
2. Inference
```python
from mxbai_rerank import MxbaiRerankV2
model = MxbaiRerankV2("mixedbread-ai/mxbai-rerank-large-v2")
query = "Who wrote 'To Kill a Mockingbird'?"
documents = [
"'To Kill a Mockingbird' is a novel by Harper Lee published in 1960. It was immediately successful, winning the Pulitzer Prize, and has become a classic of modern American literature.",
"The novel 'Moby-Dick' was written by Herman Melville and first published in 1851. It is considered a masterpiece of American literature and deals with complex themes of obsession, revenge, and the conflict between good and evil.",
"Harper Lee, an American novelist widely known for her novel 'To Kill a Mockingbird', was born in 1926 in Monroeville, Alabama. She received the Pulitzer Prize for Fiction in 1961.",
"Jane Austen was an English novelist known primarily for her six major novels, which interpret, critique and comment upon the British landed gentry at the end of the 18th century.",
"The 'Harry Potter' series, which consists of seven fantasy novels written by British author J.K. Rowling, is among the most popular and critically acclaimed books of the modern era.",
"'The Great Gatsby', a novel written by American author F. Scott Fitzgerald, was published in 1925. The story is set in the Jazz Age and follows the life of millionaire Jay Gatsby and his pursuit of Daisy Buchanan."
]
# Lets get the scores
results = model.rank(query, documents, return_documents=True, top_k=3)
print(results)
```
## Performance
### Benchmark Results
| Model | BEIR Avg | Multilingual | Chinese | Code Search | Latency (s) |
|-------|----------|----------|----------|--------------|-------------|
| mxbai-rerank-large-v2 | 57.49 | 29.79 | 84.16 | 32.05 | 0.89 |
| mxbai-rerank-base-v2 | 55.57 | 28.56 | 83.70 | 31.73 | 0.67 |
| mxbai-rerank-large-v1 | 49.32 | 21.88 | 72.53 | 30.72 | 2.24 |
*Latency measured on A100 GPU
## Training Details
The models were trained using a three-step process:
1. **GRPO (Guided Reinforcement Prompt Optimization)**
2. **Contrastive Learning**
3. **Preference Learning**
For more details, check our [technical blog post](https://mixedbread.com/blog/mxbai-rerank-v2).
Paper following soon.
## 🎓 Citation
```bibtex
@online{v2rerank2025mxbai,
title={Baked-in Brilliance: Reranking Meets RL with mxbai-rerank-v2},
author={Sean Lee and Rui Huang and Aamir Shakir and Julius Lipp},
year={2025},
url={https://www.mixedbread.com/blog/mxbai-rerank-v2},
}
``` |
Mungert/functionary-small-v3.2-GGUF | Mungert | 2025-06-15T19:42:16Z | 295 | 4 | null | [
"gguf",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-03-24T00:38:48Z | ---
license: mit
---
# <span style="color: #7FFF7F;">functionary-small-v3.2 GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `functionary-small-v3.2-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `functionary-small-v3.2-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `functionary-small-v3.2-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `functionary-small-v3.2-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `functionary-small-v3.2-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `functionary-small-v3.2-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `functionary-small-v3.2-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `functionary-small-v3.2-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `functionary-small-v3.2-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `functionary-small-v3.2-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `functionary-small-v3.2-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Model Card for functionary-small-v3.2
**This model was based on [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct)**
[https://github.com/MeetKai/functionary](https://github.com/MeetKai/functionary)
<img src="https://huggingface.co/meetkai/functionary-medium-v2.2/resolve/main/functionary_logo.jpg" alt="Functionary Logo" width="300"/>
Functionary is a language model that can interpret and execute functions/plugins.
The model determines when to execute functions, whether in parallel or serially, and can understand their outputs. It only triggers functions as needed. Function definitions are given as JSON Schema Objects, similar to OpenAI GPT function calls.
## Key Features
- Intelligent **parallel tool use**
- Able to analyze functions/tools outputs and provide relevant responses **grounded in the outputs**
- Able to decide **when to not use tools/call functions** and provide normal chat response
- Truly one of the best open-source alternative to GPT-4
- Support code interpreter
## How to Get Started
We provide custom code for parsing raw model responses into a JSON object containing `role`, `content` and `tool_calls` fields. This enables the users to read the function-calling output of the model easily.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("meetkai/functionary-small-v3.2")
model = AutoModelForCausalLM.from_pretrained("meetkai/functionary-small-v3.2", device_map="auto", trust_remote_code=True)
tools = [
{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
}
]
messages = [{"role": "user", "content": "What is the weather in Istanbul and Singapore respectively?"}]
final_prompt = tokenizer.apply_chat_template(messages, tools, add_generation_prompt=True, tokenize=False)
inputs = tokenizer(final_prompt, return_tensors="pt").to("cuda")
pred = model.generate_tool_use(**inputs, max_new_tokens=128, tokenizer=tokenizer)
print(tokenizer.decode(pred.cpu()[0]))
```
## Prompt Template
We convert function definitions to a similar text to TypeScript definitions. Then we inject these definitions as system prompts. After that, we inject the default system prompt. Then we start the conversation messages.
This formatting is also available via our vLLM server which we process the functions into Typescript definitions encapsulated in a system message using a pre-defined Transformers Jinja chat template. This means that the lists of messages can be formatted for you with the apply_chat_template() method within our server:
```python
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="functionary")
client.chat.completions.create(
model="path/to/functionary/model/",
messages=[{"role": "user",
"content": "What is the weather for Istanbul?"}
],
tools=[{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
}],
tool_choice="auto"
)
```
will yield:
```
<|start_header_id|>system<|end_header_id|>
You are capable of executing available function(s) if required.
Only execute function(s) when absolutely necessary.
Ask for the required input to:recipient==all
Use JSON for function arguments.
Respond in this format:
>>>${recipient}
${content}
Available functions:
// Supported function definitions that should be called when necessary.
namespace functions {
// Get the current weather
type get_current_weather = (_: {
// The city and state, e.g. San Francisco, CA
location: string,
}) => any;
} // namespace functions<|eot_id|><|start_header_id|>user<|end_header_id|>
What is the weather for Istanbul?
```
A more detailed example is provided [here](https://github.com/MeetKai/functionary/blob/main/tests/prompt_test_v3.llama3.txt).
## Run the model
We encourage users to run our models using our OpenAI-compatible vLLM server [here](https://github.com/MeetKai/functionary).
# The MeetKai Team
|
Mungert/Phi-4-mini-instruct-GGUF | Mungert | 2025-06-15T19:42:12Z | 1,693 | 3 | transformers | [
"transformers",
"gguf",
"nlp",
"code",
"text-generation",
"multilingual",
"ar",
"zh",
"cs",
"da",
"nl",
"en",
"fi",
"fr",
"de",
"he",
"hu",
"it",
"ja",
"ko",
"no",
"pl",
"pt",
"ru",
"es",
"sv",
"th",
"tr",
"uk",
"arxiv:2503.01743",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-24T00:05:56Z | ---
language:
- multilingual
- ar
- zh
- cs
- da
- nl
- en
- fi
- fr
- de
- he
- hu
- it
- ja
- ko
- 'no'
- pl
- pt
- ru
- es
- sv
- th
- tr
- uk
library_name: transformers
license: mit
license_link: https://huggingface.co/microsoft/Phi-4-mini-instruct/resolve/main/LICENSE
pipeline_tag: text-generation
tags:
- nlp
- code
widget:
- messages:
- role: user
content: Can you provide ways to eat combinations of bananas and dragonfruits?
---
# <span style="color: #7FFF7F;">Phi-4-mini-instruct GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Phi-4-mini-instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Phi-4-mini-instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Phi-4-mini-instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Phi-4-mini-instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Phi-4-mini-instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Phi-4-mini-instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Phi-4-mini-instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Phi-4-mini-instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Phi-4-mini-instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Phi-4-mini-instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Phi-4-mini-instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
## Model Summary
Phi-4-mini-instruct is a lightweight open model built upon synthetic data and filtered publicly available websites - with a focus on high-quality, reasoning dense data. The model belongs to the Phi-4 model family and supports 128K token context length. The model underwent an enhancement process, incorporating both supervised fine-tuning and direct preference optimization to support precise instruction adherence and robust safety measures.
📰 [Phi-4-mini Microsoft Blog](https://aka.ms/phi4-feb2025) <br>
📖 [Phi-4-mini Technical Report](https://aka.ms/phi-4-multimodal/techreport) <br>
👩🍳 [Phi Cookbook](https://github.com/microsoft/PhiCookBook) <br>
🏡 [Phi Portal](https://azure.microsoft.com/en-us/products/phi) <br>
🖥️ Try It [Azure](https://aka.ms/phi-4-mini/azure), [Huggingface](https://huggingface.co/spaces/microsoft/phi-4-mini) <br>
🚀 [Model paper](https://huggingface.co/papers/2503.01743)
🎉**Phi-4**: [[multimodal-instruct](https://huggingface.co/microsoft/Phi-4-multimodal-instruct) | [onnx](https://huggingface.co/microsoft/Phi-4-multimodal-instruct-onnx)];
[[mini-instruct](https://huggingface.co/microsoft/Phi-4-mini-instruct) | [onnx](https://huggingface.co/microsoft/Phi-4-mini-instruct-onnx)]
## Intended Uses
### Primary Use Cases
The model is intended for broad multilingual commercial and research use. The model provides uses for general purpose AI systems and applications which require:
1) Memory/compute constrained environments
2) Latency bound scenarios
3) Strong reasoning (especially math and logic).
The model is designed to accelerate research on language and multimodal models, for use as a building block for generative AI powered features.
### Use Case Considerations
The model is not specifically designed or evaluated for all downstream purposes. Developers should consider common limitations of language models, as well as performance difference across languages, as they select use cases, and evaluate and mitigate for accuracy, safety, and fairness before using within a specific downstream use case, particularly for high-risk scenarios.
Developers should be aware of and adhere to applicable laws or regulations (including but not limited to privacy, trade compliance laws, etc.) that are relevant to their use case.
***Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under.***
## Release Notes
This release of Phi-4-mini-instruct is based on valuable user feedback from the Phi-3 series. The Phi-4-mini model employed new architecture for efficiency, larger vocabulary for multilingual support, and better post-training techniques were used for instruction following, function calling, as well as additional data leading to substantial gains on key capabilities. It is anticipated that most use cases will benefit from this release, but users are encouraged to test in their particular AI applications. The enthusiastic support for the Phi-4 series is greatly appreciated. Feedback on Phi-4-mini-instruct is welcomed and crucial to the model’s evolution and improvement.
### Model Quality
To understand the capabilities, the 3.8B parameters Phi-4-mini-instruct model was compared with a set of models over a variety of benchmarks using an internal benchmark platform (See Appendix A for benchmark methodology). A high-level overview of the model quality is as follows:
| Benchmark | Similar size | | | | |2x size | | | | | |
|----------------------------------|-------------|-------------------|-------------------|-------------------|-----------------|-----------------|-----------------|-----------------|-----------------|-----------------|-----------------|
| | Phi-4 mini-Ins | Phi-3.5-mini-Ins | Llama-3.2-3B-Ins | Mistral-3B | Qwen2.5-3B-Ins | Qwen2.5-7B-Ins | Mistral-8B-2410 | Llama-3.1-8B-Ins | Llama-3.1-Tulu-3-8B | Gemma2-9B-Ins | GPT-4o-mini-2024-07-18 |
| **Popular aggregated benchmark** | | | | | | | | | | | |
| Arena Hard | 32.8 | 34.4 | 17.0 | 26.9 | 32.0 | 55.5 | 37.3 | 25.7 | 42.7 | 43.7 | 53.7 |
| BigBench Hard (0-shot, CoT) | 70.4 | 63.1 | 55.4 | 51.2 | 56.2 | 72.4 | 53.3 | 63.4 | 55.5 | 65.7 | 80.4 |
| MMLU (5-shot) | 67.3 | 65.5 | 61.8 | 60.8 | 65.0 | 72.6 | 63.0 | 68.1 | 65.0 | 71.3 | 77.2 |
| MMLU-Pro (0-shot, CoT) | 52.8 | 47.4 | 39.2 | 35.3 | 44.7 | 56.2 | 36.6 | 44.0 | 40.9 | 50.1 | 62.8 |
| **Reasoning** | | | | | | | | | | | |
| ARC Challenge (10-shot) | 83.7 | 84.6 | 76.1 | 80.3 | 82.6 | 90.1 | 82.7 | 83.1 | 79.4 | 89.8 | 93.5 |
| BoolQ (2-shot) | 81.2 | 77.7 | 71.4 | 79.4 | 65.4 | 80.0 | 80.5 | 82.8 | 79.3 | 85.7 | 88.7 |
| GPQA (0-shot, CoT) | 25.2 | 26.6 | 24.3 | 24.4 | 23.4 | 30.6 | 26.3 | 26.3 | 29.9 | 39.1 | 41.1 |
| HellaSwag (5-shot) | 69.1 | 72.2 | 77.2 | 74.6 | 74.6 | 80.0 | 73.5 | 72.8 | 80.9 | 87.1 | 88.7 |
| OpenBookQA (10-shot) | 79.2 | 81.2 | 72.6 | 79.8 | 79.3 | 82.6 | 80.2 | 84.8 | 79.8 | 90.0 | 90.0 |
| PIQA (5-shot) | 77.6 | 78.2 | 68.2 | 73.2 | 72.6 | 76.2 | 81.2 | 83.2 | 78.3 | 83.7 | 88.7 |
| Social IQA (5-shot) | 72.5 | 75.1 | 68.3 | 73.9 | 75.3 | 75.3 | 77.6 | 71.8 | 73.4 | 74.7 | 82.9 |
| TruthfulQA (MC2) (10-shot) | 66.4 | 65.2 | 59.2 | 62.9 | 64.3 | 69.4 | 63.0 | 69.2 | 64.1 | 76.6 | 78.2 |
| Winogrande (5-shot) | 67.0 | 72.2 | 53.2 | 59.8 | 63.3 | 71.1 | 63.1 | 64.7 | 65.4 | 74.0 | 76.9 |
| **Multilingual** | | | | | | | | | | | |
| Multilingual MMLU (5-shot) | 49.3 | 51.8 | 48.1 | 46.4 | 55.9 | 64.4 | 53.7 | 56.2 | 54.5 | 63.8 | 72.9 |
| MGSM (0-shot, CoT) | 63.9 | 49.6 | 44.6 | 44.6 | 53.5 | 64.5 | 56.7 | 56.7 | 58.6 | 75.1 | 81.7 |
| **Math** | | | | | | | | | | | |
| GSM8K (8-shot, CoT) | 88.6 | 76.9 | 75.6 | 80.1 | 80.6 | 88.7 | 81.9 | 82.4 | 84.3 | 84.9 | 91.3 |
| MATH (0-shot, CoT) | 64.0 | 49.8 | 46.7 | 41.8 | 61.7 | 60.4 | 41.6 | 47.6 | 46.1 | 51.3 | 70.2 |
| **Overall** | **63.5** | **60.5** | **56.2** | **56.9** | **60.1** | **67.9** | **60.2** | **62.3** | **60.9** | **65.0** | **75.5** |
Overall, the model with only 3.8B-param achieves a similar level of multilingual language understanding and reasoning ability as much larger models. However, it is still fundamentally limited by its size for certain tasks. The model simply does not have the capacity to store too much factual knowledge, therefore, users may experience factual incorrectness. However, it may be possible to resolve such weakness by augmenting Phi-4 with a search engine, particularly when using the model under RAG settings.
## Usage
### Tokenizer
Phi-4-mini-instruct supports a vocabulary size of up to `200064` tokens. The [tokenizer files](https://huggingface.co/microsoft/Phi-4-mini-instruct/blob/main/added_tokens.json) already provide placeholder tokens that can be used for downstream fine-tuning, but they can also be extended up to the model's vocabulary size.
### Input Formats
Given the nature of the training data, the Phi-4-mini-instruct
model is best suited for prompts using specific formats.
Below are the two primary formats:
#### Chat format
This format is used for general conversation and instructions:
```yaml
<|system|>Insert System Message<|end|><|user|>Insert User Message<|end|><|assistant|>
```
#### Tool-enabled function-calling format
This format is used when the user wants the model to provide function calls based on the given tools. The user should provide the available tools in the system prompt, wrapped by <|tool|> and <|/tool|> tokens. The tools should be specified in JSON format, using a JSON dump structure. Example:
`
<|system|>You are a helpful assistant with some tools.<|tool|>[{"name": "get_weather_updates", "description": "Fetches weather updates for a given city using the RapidAPI Weather API.", "parameters": {"city": {"description": "The name of the city for which to retrieve weather information.", "type": "str", "default": "London"}}}]<|/tool|><|end|><|user|>What is the weather like in Paris today?<|end|><|assistant|>
`
### Inference with vLLM
#### Requirements
List of required packages:
```
flash_attn==2.7.4.post1
torch==2.5.1
vllm>=0.7.3
```
#### Example
To perform inference using vLLM, you can use the following code snippet:
```python
from vllm import LLM, SamplingParams
llm = LLM(model="microsoft/Phi-4-mini-instruct", trust_remote_code=True)
messages = [
{"role": "system", "content": "You are a helpful AI assistant."},
{"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
{"role": "assistant", "content": "Sure! Here are some ways to eat bananas and dragonfruits together: 1. Banana and dragonfruit smoothie: Blend bananas and dragonfruits together with some milk and honey. 2. Banana and dragonfruit salad: Mix sliced bananas and dragonfruits together with some lemon juice and honey."},
{"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]
sampling_params = SamplingParams(
max_tokens=500,
temperature=0.0,
)
output = llm.chat(messages=messages, sampling_params=sampling_params)
print(output[0].outputs[0].text)
```
### Inference with Transformers
#### Requirements
Phi-4 family has been integrated in the `4.49.0` version of `transformers`. The current `transformers` version can be verified with: `pip list | grep transformers`.
Python 3.8 and 3.10 will work best.
List of required packages:
```
flash_attn==2.7.4.post1
torch==2.5.1
transformers==4.49.0
accelerate==1.3.0
```
Phi-4-mini-instruct is also available in [Azure AI Studio]()
#### Example
After obtaining the Phi-4-mini-instruct model checkpoints, users can use this sample code for inference.
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
torch.random.manual_seed(0)
model_path = "microsoft/Phi-4-mini-instruct"
model = AutoModelForCausalLM.from_pretrained(
model_path,
device_map="auto",
torch_dtype="auto",
trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained(model_path)
messages = [
{"role": "system", "content": "You are a helpful AI assistant."},
{"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
{"role": "assistant", "content": "Sure! Here are some ways to eat bananas and dragonfruits together: 1. Banana and dragonfruit smoothie: Blend bananas and dragonfruits together with some milk and honey. 2. Banana and dragonfruit salad: Mix sliced bananas and dragonfruits together with some lemon juice and honey."},
{"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]
pipe = pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
)
generation_args = {
"max_new_tokens": 500,
"return_full_text": False,
"temperature": 0.0,
"do_sample": False,
}
output = pipe(messages, **generation_args)
print(output[0]['generated_text'])
```
## Responsible AI Considerations
Like other language models, the Phi family of models can potentially behave in ways that are unfair, unreliable, or offensive. Some of the limiting behaviors to be aware of include:
+ Quality of Service: The Phi models are trained primarily on English text and some additional multilingual text. Languages other than English will experience worse performance as well as performance disparities across non-English. English language varieties with less representation in the training data might experience worse performance than standard American English.
+ Multilingual performance and safety gaps: We believe it is important to make language models more widely available across different languages, but the Phi 4 models still exhibit challenges common across multilingual releases. As with any deployment of LLMs, developers will be better positioned to test for performance or safety gaps for their linguistic and cultural context and customize the model with additional fine-tuning and appropriate safeguards.
+ Representation of Harms & Perpetuation of Stereotypes: These models can over- or under-represent groups of people, erase representation of some groups, or reinforce demeaning or negative stereotypes. Despite safety post-training, these limitations may still be present due to differing levels of representation of different groups, cultural contexts, or prevalence of examples of negative stereotypes in training data that reflect real-world patterns and societal biases.
+ Inappropriate or Offensive Content: These models may produce other types of inappropriate or offensive content, which may make it inappropriate to deploy for sensitive contexts without additional mitigations that are specific to the case.
+ Information Reliability: Language models can generate nonsensical content or fabricate content that might sound reasonable but is inaccurate or outdated.
+ Limited Scope for Code: The majority of Phi 4 training data is based in Python and uses common packages such as "typing, math, random, collections, datetime, itertools". If the model generates Python scripts that utilize other packages or scripts in other languages, it is strongly recommended that users manually verify all API uses.
+ Long Conversation: Phi 4 models, like other models, can in some cases generate responses that are repetitive, unhelpful, or inconsistent in very long chat sessions in both English and non-English languages. Developers are encouraged to place appropriate mitigations, like limiting conversation turns to account for the possible conversational drift.
Developers should apply responsible AI best practices, including mapping, measuring, and mitigating risks associated with their specific use case and cultural, linguistic context. Phi 4 family of models are general purpose models. As developers plan to deploy these models for specific use cases, they are encouraged to fine-tune the models for their use case and leverage the models as part of broader AI systems with language-specific safeguards in place. Important areas for consideration include:
+ Allocation: Models may not be suitable for scenarios that could have consequential impact on legal status or the allocation of resources or life opportunities (ex: housing, employment, credit, etc.) without further assessments and additional debiasing techniques.
+ High-Risk Scenarios: Developers should assess the suitability of using models in high-risk scenarios where unfair, unreliable or offensive outputs might be extremely costly or lead to harm. This includes providing advice in sensitive or expert domains where accuracy and reliability are critical (ex: legal or health advice). Additional safeguards should be implemented at the application level according to the deployment context.
+ Misinformation: Models may produce inaccurate information. Developers should follow transparency best practices and inform end-users they are interacting with an AI system. At the application level, developers can build feedback mechanisms and pipelines to ground responses in use-case specific, contextual information, a technique known as Retrieval Augmented Generation (RAG).
+ Generation of Harmful Content: Developers should assess outputs for their context and use available safety classifiers or custom solutions appropriate for their use case.
+ Misuse: Other forms of misuse such as fraud, spam, or malware production may be possible, and developers should ensure that their applications do not violate applicable laws and regulations.
## Training
### Model
+ **Architecture:** Phi-4-mini-instruct has 3.8B parameters and is a dense decoder-only Transformer model. When compared with Phi-3.5-mini, the major changes with Phi-4-mini-instruct are 200K vocabulary, grouped-query attention, and shared input and output embedding.<br>
+ **Inputs:** Text. It is best suited for prompts using the chat format.<br>
+ **Context length:** 128K tokens<br>
+ **GPUs:** 512 A100-80G<br>
+ **Training time:** 21 days<br>
+ **Training data:** 5T tokens<br>
+ **Outputs:** Generated text in response to the input<br>
+ **Dates:** Trained between November and December 2024<br>
+ **Status:** This is a static model trained on offline datasets with the cutoff date of June 2024 for publicly available data.<br>
+ **Supported languages:** Arabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Hebrew, Hungarian, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian<br>
+ **Release date:** February 2025<br>
### Training Datasets
Phi-4-mini’s training data includes a wide variety of sources, totaling 5 trillion tokens, and is a combination of
1) publicly available documents filtered for quality, selected high-quality educational data, and code
2) newly created synthetic, “textbook-like” data for the purpose of teaching math, coding, common sense reasoning, general knowledge of the world (e.g., science, daily activities, theory of mind, etc.)
3) high quality chat format supervised data covering various topics to reflect human preferences on different aspects such as instruct-following, truthfulness, honesty and helpfulness. Focus was placed on the quality of data that could potentially improve the reasoning ability for the model, and the publicly available documents were filtered to contain a preferred level of knowledge. As an example, the result of a game in premier league on a particular day might be good training data for frontier models, but such information was removed to leave more model capacity for reasoning for the model’s small size. More details about data can be found in the Phi-4-mini-instruct technical report.
The decontamination process involved normalizing and tokenizing the dataset, then generating and comparing n-grams between the target dataset and benchmark datasets. Samples with matching n-grams above a threshold were flagged as contaminated and removed from the dataset. A detailed contamination report was generated, summarizing the matched text, matching ratio, and filtered results for further analysis.
### Fine-tuning
A basic example of multi-GPUs supervised fine-tuning (SFT) with TRL and Accelerate modules is provided [here](https://huggingface.co/microsoft/Phi-4-mini-instruct/resolve/main/sample_finetune.py).
## Safety Evaluation and Red-Teaming
Various evaluation techniques including red teaming, adversarial conversation simulations, and multilingual safety evaluation benchmark datasets were leveraged to evaluate Phi-4 models’ propensity to produce undesirable outputs across multiple languages and risk categories. Several approaches were used to compensate for the limitations of one approach alone. Findings across the various evaluation methods indicate that safety post-training that was done as detailed in the Phi 3 Safety Post-Training paper had a positive impact across multiple languages and risk categories as observed by refusal rates (refusal to output undesirable outputs) and robustness to jailbreak techniques. Details on prior red team evaluations across Phi models can be found in the Phi 3 Safety Post-Training paper. For this release, the red team tested the model in English, Chinese, Japanese, Spanish, Portuguese, Arabic, Thai, and Russian for the following potential harms: Hate Speech and Bias, Violent Crimes, Specialized Advice, and Election Information. Their findings indicate that the model is resistant to jailbreak techniques across languages, but that language-specific attack prompts leveraging cultural context can cause the model to output harmful content. Another insight was that with function calling scenarios, the model could sometimes hallucinate function names or URL’s. The model may also be more susceptible to longer multi-turn jailbreak techniques across both English and non-English languages. These findings highlight the need for industry-wide investment in the development of high-quality safety evaluation datasets across multiple languages, including low resource languages, and risk areas that account for cultural nuances where those languages are spoken.
## Software
* [PyTorch](https://github.com/pytorch/pytorch)
* [Transformers](https://github.com/huggingface/transformers)
* [Flash-Attention](https://github.com/HazyResearch/flash-attention)
## Hardware
Note that by default, the Phi-4-mini-instruct model uses flash attention, which requires certain types of GPU hardware to run. We have tested on the following GPU types:
* NVIDIA A100
* NVIDIA A6000
* NVIDIA H100
If you want to run the model on:
* NVIDIA V100 or earlier generation GPUs: call AutoModelForCausalLM.from_pretrained() with attn_implementation="eager"
## License
The model is licensed under the [MIT license](./LICENSE).
## Trademarks
This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow [Microsoft’s Trademark & Brand Guidelines](https://www.microsoft.com/en-us/legal/intellectualproperty/trademarks). Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party’s policies.
## Appendix A: Benchmark Methodology
We include a brief word on methodology here - and in particular, how we think about optimizing prompts.
In an ideal world, we would never change any prompts in our benchmarks to ensure it is always an apples-to-apples comparison when comparing different models. Indeed, this is our default approach, and is the case in the vast majority of models we have run to date.
There are, however, some exceptions to this. In some cases, we see a model that performs worse than expected on a given eval due to a failure to respect the output format. For example:
+ A model may refuse to answer questions (for no apparent reason), or in coding tasks models may prefix their response with “Sure, I can help with that. …” which may break the parser. In such cases, we have opted to try different system messages (e.g. “You must always respond to a question” or “Get to the point!”).
+ With some models, we observed that few shots actually hurt model performance. In this case we did allow running the benchmarks with 0-shots for all cases.
+ We have tools to convert between chat and completions APIs. When converting a chat prompt to a completion prompt, some models have different keywords e.g. Human vs User. In these cases, we do allow for model-specific mappings for chat to completion prompts.
However, we do not:
+ Pick different few-shot examples. Few shots will always be the same when comparing different models.
+ Change prompt format: e.g. if it is an A/B/C/D multiple choice, we do not tweak this to 1/2/3/4 multiple choice.
### Benchmark datasets
The model was evaluated across a breadth of public and internal benchmarks to understand the model’s capabilities under multiple tasks and conditions. While most evaluations use English, the leading multilingual benchmark was incorporated that covers performance in select languages. More specifically,
+ Reasoning:
+ Winogrande: commonsense reasoning around pronoun resolution
+ PIQA: physical commonsense reasoning around everyday situations
+ ARC-challenge: grade-school multiple choice science questions
+ GPQA: very hard questions written and validated by experts in biology, physics, and chemistry
+ MedQA: medical questions answering
+ Social IQA: social commonsense intelligence
+ BoolQ: natural questions from context
+ TruthfulQA: grounded reasoning
+ Language understanding:
+ HellaSwag: commonsense natural language inference around everyday events
+ ANLI: adversarial natural language inference
+ Function calling:
+ Berkeley function calling function and tool call
+ Internal function calling benchmarks
+ World knowledge:
+ TriviaQA: trivia question on general topics
+ Math:
+ GSM8K: grade-school math word problems
+ GSM8K Hard: grade-school math word problems with large values and some absurdity.
+ MATH: challenging competition math problems
+ Code:
+ HumanEval HumanEval+, MBPP, MBPP+: python coding tasks
+ LiveCodeBenh, LiveBench: contamination-free code tasks
+ BigCode Bench: challenging programming tasks
+ Spider: SQL query tasks
+ Internal coding benchmarks
+ Instructions following:
+ IFEval: verifiable instructions
+ Internal instructions following benchmarks
+ Multilingual:
+ MGSM: multilingual grade-school math
+ Multilingual MMLU and MMLU-pro
+ MEGA: multilingual NLP tasks
+ Popular aggregated datasets: MMLU, MMLU-pro, BigBench-Hard, AGI Eval
+ Multi-turn conversations:
+ Data generated by in-house adversarial conversation simulation tool
+ Single-turn trustworthiness evaluation:
+ DecodingTrust: a collection of trustworthiness benchmarks in eight different perspectives
+ XSTest: exaggerated safety evaluation
+ Toxigen: adversarial and hate speech detection
+ Red Team:
+ Responses to prompts provided by AI Red Team at Microsoft |
Mungert/functionary-v4r-small-preview-GGUF | Mungert | 2025-06-15T19:42:07Z | 320 | 4 | null | [
"gguf",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-03-23T18:56:56Z | ---
license: mit
---
# <span style="color: #7FFF7F;">functionary-v4r-small-preview GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `functionary-v4r-small-preview-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `functionary-v4r-small-preview-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `functionary-v4r-small-preview-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `functionary-v4r-small-preview-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `functionary-v4r-small-preview-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `functionary-v4r-small-preview-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `functionary-v4r-small-preview-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `functionary-v4r-small-preview-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `functionary-v4r-small-preview-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `functionary-v4r-small-preview-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `functionary-v4r-small-preview-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I'd really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Model Card for meetkai/functionary-v4r-small-preview
**This model was based on [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct)**
[https://github.com/MeetKai/functionary](https://github.com/MeetKai/functionary)
<img src="https://huggingface.co/meetkai/functionary-medium-v2.2/resolve/main/functionary_logo.jpg" alt="Functionary Logo" width="300"/>
Functionary is a language model that can interpret and execute functions/plugins.
The model determines when to execute functions, whether in parallel or serially, and can understand their outputs. It only triggers functions as needed. Function definitions are given as JSON Schema Objects, similar to OpenAI GPT function calls.
## Key Features
- Generate the reasoning before deciding tool uses
- Intelligent **parallel tool use**
- Able to analyze functions/tools outputs and provide relevant responses **grounded in the outputs**
- Able to decide **when to not use tools/call functions** and provide normal chat response
- Truly one of the best open-source alternative to GPT-4
- Support code interpreter
## How to Get Started
We provide custom code for parsing raw model responses into a JSON object containing role, content and tool_calls fields. This enables the users to read the function-calling output of the model easily.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("meetkai/functionary-v4r-small-preview")
model = AutoModelForCausalLM.from_pretrained("meetkai/functionary-v4r-small-preview", device_map="auto", trust_remote_code=True)
tools = [
{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
}
]
# add this to make the model generate the reasoning first
tools.append({"type": "reasoning"})
messages = [{"role": "user", "content": "What is the weather in Istanbul and Singapore respectively?"}]
final_prompt = tokenizer.apply_chat_template(messages, tools, add_generation_prompt=True, tokenize=False)
inputs = tokenizer(final_prompt, return_tensors="pt").to("cuda")
pred = model.generate_tool_use(**inputs, max_new_tokens=128, tokenizer=tokenizer)
print(tokenizer.decode(pred.cpu()[0]))
```
## Prompt Template
We convert function definitions to a similar text to TypeScript definitions. Then we inject these definitions as system prompts. After that, we inject the default system prompt. Then we start the conversation messages.
This formatting is also available via our vLLM server which we process the functions into Typescript definitions encapsulated in a system message using a pre-defined Transformers Jinja chat template. This means that the lists of messages can be formatted for you with the apply_chat_template() method within our server:
```python
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="functionary")
messages = [{"role": "user",
"content": "What is the weather for Istanbul?"}
]
tools = [{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
}]
# Add reasoning type to make the model generate the reasoning first
tools.append({"type": "reasoning"})
client.chat.completions.create(
model="path/to/functionary/model/",
messages=messages,
tools=tools,
tool_choice="auto"
)
```
will yield:
```
<|start_header_id|>system<|end_header_id|>
Reasoning Mode: On
Cutting Knowledge Date: December 2023
You have access to the following functions:
Use the function 'get_current_weather' to 'Get the current weather'
{"name": "get_current_weather", "description": "Get the current weather", "parameters": {"type": "object", "properties": {"location": {"type": "string", "description": "The city and state, e.g. San Francisco, CA"}}, "required": ["location"]}}
Think very carefully before calling functions.
If a you choose to call a function ONLY reply in the following format:
<{start_tag}={function_name}>{parameters}{end_tag}
where
start_tag => `<function`
parameters => a JSON dict with the function argument name as key and function argument value as value.
end_tag => `</function>`
Here is an example,
<function=example_function_name>{"example_name": "example_value"}</function>
Reminder:
- If looking for real time information use relevant functions before falling back to brave_search
- Function calls MUST follow the specified format, start with <function= and end with </function>
- Required parameters MUST be specified
- Only call one function at a time
- Put the entire function call reply on one line
<|eot_id|><|start_header_id|>user<|end_header_id|>
What is the weather for Istanbul?
```
## Run the model
We encourage users to run our models using our OpenAI-compatible vLLM server [here](https://github.com/MeetKai/functionary).
# The MeetKai Team
 |
Mungert/Llama-Guard-3-8B-GGUF | Mungert | 2025-06-15T19:42:03Z | 544 | 2 | null | [
"gguf",
"facebook",
"meta",
"pytorch",
"llama",
"llama-3",
"text-generation",
"en",
"arxiv:2407.21783",
"arxiv:2312.06674",
"arxiv:2204.05862",
"arxiv:2308.01263",
"base_model:meta-llama/Llama-3.1-8B",
"base_model:quantized:meta-llama/Llama-3.1-8B",
"license:llama3.1",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-23T11:22:15Z | ---
language:
- en
pipeline_tag: text-generation
base_model: meta-llama/Meta-Llama-3.1-8B
tags:
- facebook
- meta
- pytorch
- llama
- llama-3
license: llama3.1
extra_gated_prompt: >-
### LLAMA 3.1 COMMUNITY LICENSE AGREEMENT
Llama 3.1 Version Release Date: July 23, 2024
"Agreement" means the terms and conditions for use, reproduction, distribution and modification of the
Llama Materials set forth herein.
"Documentation" means the specifications, manuals and documentation accompanying Llama 3.1
distributed by Meta at https://llama.meta.com/doc/overview.
"Licensee" or "you" means you, or your employer or any other person or entity (if you are entering into
this Agreement on such person or entity’s behalf), of the age required under applicable laws, rules or
regulations to provide legal consent and that has legal authority to bind your employer or such other
person or entity if you are entering in this Agreement on their behalf.
"Llama 3.1" means the foundational large language models and software and algorithms, including
machine-learning model code, trained model weights, inference-enabling code, training-enabling code,
fine-tuning enabling code and other elements of the foregoing distributed by Meta at
https://llama.meta.com/llama-downloads.
"Llama Materials" means, collectively, Meta’s proprietary Llama 3.1 and Documentation (and any
portion thereof) made available under this Agreement.
"Meta" or "we" means Meta Platforms Ireland Limited (if you are located in or, if you are an entity, your
principal place of business is in the EEA or Switzerland) and Meta Platforms, Inc. (if you are located
outside of the EEA or Switzerland).
1. License Rights and Redistribution.
a. Grant of Rights. You are granted a non-exclusive, worldwide, non-transferable and royalty-free
limited license under Meta’s intellectual property or other rights owned by Meta embodied in the Llama
Materials to use, reproduce, distribute, copy, create derivative works of, and make modifications to the
Llama Materials.
b. Redistribution and Use.
i. If you distribute or make available the Llama Materials (or any derivative works
thereof), or a product or service (including another AI model) that contains any of them, you shall (A)
provide a copy of this Agreement with any such Llama Materials; and (B) prominently display “Built with
Llama” on a related website, user interface, blogpost, about page, or product documentation. If you use
the Llama Materials or any outputs or results of the Llama Materials to create, train, fine tune, or
otherwise improve an AI model, which is distributed or made available, you shall also include “Llama” at
the beginning of any such AI model name.
ii. If you receive Llama Materials, or any derivative works thereof, from a Licensee as part
of an integrated end user product, then Section 2 of this Agreement will not apply to you.
iii. You must retain in all copies of the Llama Materials that you distribute the following
attribution notice within a “Notice” text file distributed as a part of such copies: “Llama 3.1 is
licensed under the Llama 3.1 Community License, Copyright © Meta Platforms, Inc. All Rights
Reserved.”
iv. Your use of the Llama Materials must comply with applicable laws and regulations
(including trade compliance laws and regulations) and adhere to the Acceptable Use Policy for the Llama
Materials (available at https://llama.meta.com/llama3_1/use-policy), which is hereby incorporated by
reference into this Agreement.
2. Additional Commercial Terms. If, on the Llama 3.1 version release date, the monthly active users
of the products or services made available by or for Licensee, or Licensee’s affiliates, is greater than 700
million monthly active users in the preceding calendar month, you must request a license from Meta,
which Meta may grant to you in its sole discretion, and you are not authorized to exercise any of the
rights under this Agreement unless or until Meta otherwise expressly grants you such rights.
3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA MATERIALS AND ANY
OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF
ANY KIND, AND META DISCLAIMS ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED,
INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT,
MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR
DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE LLAMA MATERIALS AND
ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND
RESULTS.
4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING
OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL,
INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF META OR ITS AFFILIATES HAVE BEEN ADVISED
OF THE POSSIBILITY OF ANY OF THE FOREGOING.
5. Intellectual Property.
a. No trademark licenses are granted under this Agreement, and in connection with the Llama
Materials, neither Meta nor Licensee may use any name or mark owned by or associated with the other
or any of its affiliates, except as required for reasonable and customary use in describing and
redistributing the Llama Materials or as set forth in this Section 5(a). Meta hereby grants you a license to
use “Llama” (the “Mark”) solely as required to comply with the last sentence of Section 1.b.i. You will
comply with Meta’s brand guidelines (currently accessible at
https://about.meta.com/brand/resources/meta/company-brand/ ). All goodwill arising out of your use
of the Mark will inure to the benefit of Meta.
b. Subject to Meta’s ownership of Llama Materials and derivatives made by or for Meta, with
respect to any derivative works and modifications of the Llama Materials that are made by you, as
between you and Meta, you are and will be the owner of such derivative works and modifications.
c. If you institute litigation or other proceedings against Meta or any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Llama Materials or Llama 3.1 outputs or
results, or any portion of any of the foregoing, constitutes infringement of intellectual property or other
rights owned or licensable by you, then any licenses granted to you under this Agreement shall
terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold
harmless Meta from and against any claim by any third party arising out of or related to your use or
distribution of the Llama Materials.
6. Term and Termination. The term of this Agreement will commence upon your acceptance of this
Agreement or access to the Llama Materials and will continue in full force and effect until terminated in
accordance with the terms and conditions herein. Meta may terminate this Agreement if you are in
breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete
and cease use of the Llama Materials. Sections 3, 4 and 7 shall survive the termination of this
Agreement.
7. Governing Law and Jurisdiction. This Agreement will be governed and construed under the laws of
the State of California without regard to choice of law principles, and the UN Convention on Contracts
for the International Sale of Goods does not apply to this Agreement. The courts of California shall have
exclusive jurisdiction of any dispute arising out of this Agreement.
### Llama 3.1 Acceptable Use Policy
Meta is committed to promoting safe and fair use of its tools and features, including Llama 3.1. If you
access or use Llama 3.1, you agree to this Acceptable Use Policy (“Policy”). The most recent copy of
this policy can be found at [https://llama.meta.com/llama3_1/use-policy](https://llama.meta.com/llama3_1/use-policy)
#### Prohibited Uses
We want everyone to use Llama 3.1 safely and responsibly. You agree you will not use, or allow
others to use, Llama 3.1 to:
1. Violate the law or others’ rights, including to:
1. Engage in, promote, generate, contribute to, encourage, plan, incite, or further illegal or unlawful activity or content, such as:
1. Violence or terrorism
2. Exploitation or harm to children, including the solicitation, creation, acquisition, or dissemination of child exploitative content or failure to report Child Sexual Abuse Material
3. Human trafficking, exploitation, and sexual violence
4. The illegal distribution of information or materials to minors, including obscene materials, or failure to employ legally required age-gating in connection with such information or materials.
5. Sexual solicitation
6. Any other criminal activity
3. Engage in, promote, incite, or facilitate the harassment, abuse, threatening, or bullying of individuals or groups of individuals
4. Engage in, promote, incite, or facilitate discrimination or other unlawful or harmful conduct in the provision of employment, employment benefits, credit, housing, other economic benefits, or other essential goods and services
5. Engage in the unauthorized or unlicensed practice of any profession including, but not limited to, financial, legal, medical/health, or related professional practices
6. Collect, process, disclose, generate, or infer health, demographic, or other sensitive personal or private information about individuals without rights and consents required by applicable laws
7. Engage in or facilitate any action or generate any content that infringes, misappropriates, or otherwise violates any third-party rights, including the outputs or results of any products or services using the Llama Materials
8. Create, generate, or facilitate the creation of malicious code, malware, computer viruses or do anything else that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system
2. Engage in, promote, incite, facilitate, or assist in the planning or development of activities that present a risk of death or bodily harm to individuals, including use of Llama 3.1 related to the following:
1. Military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulations (ITAR) maintained by the United States Department of State
2. Guns and illegal weapons (including weapon development)
3. Illegal drugs and regulated/controlled substances
4. Operation of critical infrastructure, transportation technologies, or heavy machinery
5. Self-harm or harm to others, including suicide, cutting, and eating disorders
6. Any content intended to incite or promote violence, abuse, or any infliction of bodily harm to an individual
3. Intentionally deceive or mislead others, including use of Llama 3.1 related to the following:
1. Generating, promoting, or furthering fraud or the creation or promotion of disinformation
2. Generating, promoting, or furthering defamatory content, including the creation of defamatory statements, images, or other content
3. Generating, promoting, or further distributing spam
4. Impersonating another individual without consent, authorization, or legal right
5. Representing that the use of Llama 3.1 or outputs are human-generated
6. Generating or facilitating false online engagement, including fake reviews and other means of fake online engagement
4. Fail to appropriately disclose to end users any known dangers of your AI system
Please report any violation of this Policy, software “bug,” or other problems that could lead to a violation
of this Policy through one of the following means:
* Reporting issues with the model: [https://github.com/meta-llama/llama-models/issues](https://github.com/meta-llama/llama-models/issues)
* Reporting risky content generated by the model:
developers.facebook.com/llama_output_feedback
* Reporting bugs and security concerns: facebook.com/whitehat/info
* Reporting violations of the Acceptable Use Policy or unlicensed uses of Meta Llama 3: [email protected]
extra_gated_fields:
First Name: text
Last Name: text
Date of birth: date_picker
Country: country
Affiliation: text
Job title:
type: select
options:
- Student
- Research Graduate
- AI researcher
- AI developer/engineer
- Reporter
- Other
geo: ip_location
By clicking Submit below I accept the terms of the license and acknowledge that the information I provide will be collected stored processed and shared in accordance with the Meta Privacy Policy: checkbox
extra_gated_description: The information you provide will be collected, stored, processed and shared in accordance with the [Meta Privacy Policy](https://www.facebook.com/privacy/policy/).
extra_gated_button_content: Submit
---
# <span style="color: #7FFF7F;">Llama-Guard-3-8B GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-Guard-3-8B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-Guard-3-8B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-Guard-3-8B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-Guard-3-8B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-Guard-3-8B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-Guard-3-8B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-Guard-3-8B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-Guard-3-8B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-Guard-3-8B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-Guard-3-8B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-Guard-3-8B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Model Details
Llama Guard 3 is a Llama-3.1-8B pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It acts as an LLM – it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.
Llama Guard 3 was aligned to safeguard against the MLCommons standardized hazards taxonomy and designed to support Llama 3.1 capabilities. Specifically, it provides content moderation in 8 languages, and was optimized to support safety and security for search and code interpreter tool calls.
Below is a response classification example for Llama Guard 3.
<p align="center">
<img src="llama_guard_3_figure.png" width="800"/>
</p>
In order to produce classifier scores, we look at the probability for the first token, and use that as the “unsafe” class probability. We can then apply score thresholding to make binary decisions.
# Hazard Taxonomy and Policy
The model is trained to predict safety labels on the 14 categories shown below, based on the [MLCommons taxonomy](https://mlcommons.org/2024/04/mlc-aisafety-v0-5-poc/) of 13 hazards, as well as an additional category for Code Interpreter Abuse for tool calls use cases
<table align="center">
<thead>
<tr>
<th colspan="2">Hazard categories</th>
</tr>
</thead>
<tbody>
<tr>
<td>S1: Violent Crimes</td>
<td>S2: Non-Violent Crimes</td>
</tr>
<tr>
<td>S3: Sex-Related Crimes</td>
<td>S4: Child Sexual Exploitation</td>
</tr>
<tr>
<td>S5: Defamation</td>
<td>S6: Specialized Advice</td>
</tr>
<tr>
<td>S7: Privacy</td>
<td>S8: Intellectual Property</td>
</tr>
<tr>
<td>S9: Indiscriminate Weapons</td>
<td>S10: Hate</td>
</tr>
<tr>
<td>S11: Suicide & Self-Harm</td>
<td>S12: Sexual Content</td>
</tr>
<tr>
<td>S13: Elections</td>
<td>S14: Code Interpreter Abuse</td>
</tr>
</tbody>
</table>
**S1 - Violent Crimes**
Responses that enable, encourage, or endorse the commission of violent crimes, including: (1) unlawful violence toward people (ex: terrorism, genocide, murder, hate-motivated violent crimes, child abuse, assault, battery, kidnapping) and (2) unlawful violence toward animals (ex: animal abuse)
**S2 - Non-Violent Crimes**
Responses that enable, encourage, or endorse the commission of non-violent crimes, including: (1) personal crimes (ex: labor trafficking, threats, intimidation, hate-motivated non-violent crimes), (2) financial crimes (ex: fraud, scams, money laundering), (3) property crimes (ex: theft, arson, vandalism), (4) drug crimes (ex: creating or using regulated substances), (5) weapons crimes (ex: producing unlicensed firearms), and (6) cyber crimes (ex: hacking)
**S3 - Sex-Related Crimes**
Responses that enable, encourage, or endorse the commission of sex-related crimes, including: (1) sex trafficking, (2) sexual assault (ex: rape), (3) sexual harassment (physical, verbal, or visual), and (4) prostitution
**S4 - Child Sexual Exploitation**
Responses that contain, describe, enable, encourage, or endorse the sexual abuse of children
**S5 - Defamation**
Responses that are both verifiably false and likely to injure a living person’s reputation
**S6 - Specialized Advice**
Responses that contain specialized financial, medical, or legal advice, or that indicate dangerous activities or objects are safe
**S7 - Privacy**
Responses that contain sensitive, nonpublic personal information that could undermine someone’s physical, digital, or financial security
**S8 - Intellectual Property**
Responses that may violate the intellectual property rights of any third party
**S9 - Indiscriminate Weapons**
Responses that enable, encourage, or endorse the creation of indiscriminate weapons, including: (1) chemical weapons (ex: nerve gas), (2) biological weapons (ex: anthrax), (3) radiological weapons (ex: cobalt bombs), (4) nuclear weapons (ex: fission bombs), and (5) high-yield explosive weapons (ex: cluster munitions)
**S10 - Hate**
Responses that demean or dehumanize people on the basis of their sensitive, personal characteristics (i.e., race, color, ethnicity, national origin, disability, religious affiliation, caste, sexual orientation, sex, gender identity, and/or serious disease)
**S11 - Suicide & Self-Harm**
Responses that enable, encourage, or endorse acts of intentional self-harm, including: (1) suicide, (2) self-injury (ex: cutting), and (3) disordered eating
**S12 - Sexual Content**
Responses that contain erotica
**S13 - Elections**
Responses that contain factually incorrect information about electoral systems and processes, including in the time, place, or manner of voting in civic elections
**S14 - Code Interpreter Abuse**
Responses that seek to abuse code interpreters, including those that enable denial of service attacks, container escapes or privilege escalation exploits
# Supported languages
Llama Guard 3 supports content safety for the following languages : English, French, German, Hindi, Italian, Portuguese, Spanish, Thai.
# Usage
> [!IMPORTANT]
>
> This repository corresponds to half-precision version of the model. A 8-bit precision version is also provided, please visit [meta-llama/Llama-Guard-3-8B-INT8](https://huggingface.co/meta-llama/Llama-Guard-3-8B-INT8).
Llama Guard 3 can be directly used with `transformers`. It is only supported since `transformers` version 4.43.
```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
model_id = "meta-llama/Llama-Guard-3-8B"
device = "cuda"
dtype = torch.bfloat16
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=dtype, device_map=device)
def moderate(chat):
input_ids = tokenizer.apply_chat_template(chat, return_tensors="pt").to(device)
output = model.generate(input_ids=input_ids, max_new_tokens=100, pad_token_id=0)
prompt_len = input_ids.shape[-1]
return tokenizer.decode(output[0][prompt_len:], skip_special_tokens=True)
moderate([
{"role": "user", "content": "I forgot how to kill a process in Linux, can you help?"},
{"role": "assistant", "content": "Sure! To kill a process in Linux, you can use the kill command followed by the process ID (PID) of the process you want to terminate."},
])
```
# Training Data
We use the English data used by Llama Guard [1], which are obtained by getting Llama 2 and Llama 3 generations on prompts from the hh-rlhf dataset [2]. In order to scale training data for new categories and new capabilities such as multilingual and tool use, we collect additional human and synthetically generated data. Similar to the English data, the multilingual data are Human-AI conversation data that are either single-turn or multi-turn. To reduce the model’s false positive rate, we curate a set of multilingual benign prompt and response data where LLMs likely reject the prompts.
For the tool use capability, we consider search tool calls and code interpreter abuse. To develop training data for search tool use, we use Llama3 to generate responses to a collected and synthetic set of prompts. The generations are based on the query results obtained from the Brave Search API. To develop synthetic training data to detect code interpreter attacks, we use an LLM to generate safe and unsafe prompts. Then, we use a non-safety-tuned LLM to generate code interpreter completions that comply with these instructions. For safe data, we focus on data close to the boundary of what would be considered unsafe, to minimize false positives on such borderline examples.
# Evaluation
**Note on evaluations:** As discussed in the original Llama Guard paper, comparing model performance is not straightforward as each model is built on its own policy and is expected to perform better on an evaluation dataset with a policy aligned to the model. This highlights the need for industry standards. By aligning the Llama Guard family of models with the Proof of Concept MLCommons taxonomy of hazards, we hope to drive adoption of industry standards like this and facilitate collaboration and transparency in the LLM safety and content evaluation space.
In this regard, we evaluate the performance of Llama Guard 3 on MLCommons hazard taxonomy and compare it across languages with Llama Guard 2 [3] on our internal test. We also add GPT4 as baseline with zero-shot prompting using MLCommons hazard taxonomy.
Tables 1, 2, and 3 show that Llama Guard 3 improves over Llama Guard 2 and outperforms GPT4 in English, multilingual, and tool use capabilities. Noteworthily, Llama Guard 3 achieves better performance with much lower false positive rates. We also benchmark Llama Guard 3 in the OSS dataset XSTest [4] and observe that it achieves the same F1 score but a lower false positive rate compared to Llama Guard 2.
<div align="center">
<small> Table 1: Comparison of performance of various models measured on our internal English test set for MLCommons hazard taxonomy (response classification).</small>
| | **F1 ↑** | **AUPRC ↑** | **False Positive<br>Rate ↓** |
|--------------------------|:--------:|:-----------:|:----------------------------:|
| Llama Guard 2 | 0.877 | 0.927 | 0.081 |
| Llama Guard 3 | 0.939 | 0.985 | 0.040 |
| GPT4 | 0.805 | N/A | 0.152 |
</div>
<br>
<table align="center">
<small><center>Table 2: Comparison of multilingual performance of various models measured on our internal test set for MLCommons hazard taxonomy (prompt+response classification).</center></small>
<thead>
<tr>
<th colspan="8"><center>F1 ↑ / FPR ↓</center></th>
</tr>
</thead>
<tbody>
<tr>
<td></td>
<td><center>French</center></td>
<td><center>German</center></td>
<td><center>Hindi</center></td>
<td><center>Italian</center></td>
<td><center>Portuguese</center></td>
<td><center>Spanish</center></td>
<td><center>Thai</center></td>
</tr>
<tr>
<td>Llama Guard 2</td>
<td><center>0.911/0.012</center></td>
<td><center>0.795/0.062</center></td>
<td><center>0.832/0.062</center></td>
<td><center>0.681/0.039</center></td>
<td><center>0.845/0.032</center></td>
<td><center>0.876/0.001</center></td>
<td><center>0.822/0.078</center></td>
</tr>
<tr>
<td>Llama Guard 3</td>
<td><center>0.943/0.036</center></td>
<td><center>0.877/0.032</center></td>
<td><center>0.871/0.050</center></td>
<td><center>0.873/0.038</center></td>
<td><center>0.860/0.060</center></td>
<td><center>0.875/0.023</center></td>
<td><center>0.834/0.030</center></td>
</tr>
<tr>
<td>GPT4</td>
<td><center>0.795/0.157</center></td>
<td><center>0.691/0.123</center></td>
<td><center>0.709/0.206</center></td>
<td><center>0.753/0.204</center></td>
<td><center>0.738/0.207</center></td>
<td><center>0.711/0.169</center></td>
<td><center>0.688/0.168</center></td>
</tr>
</tbody>
</table>
<br>
<table align="center">
<small><center>Table 3: Comparison of performance of various models measured on our internal test set for other moderation capabilities (prompt+response classification).</center></small>
<thead>
<tr>
<th></th>
<th colspan="3">Search tool calls</th>
<th colspan="3">Code interpreter abuse</th>
</tr>
</thead>
<tbody>
<tr>
<td></td>
<td><center>F1 ↑</center></td>
<td><center>AUPRC ↑</center></td>
<td><center>FPR ↓</center></td>
<td><center>F1 ↑</center></td>
<td><center>AUPRC ↑</center></td>
<td><center>FPR ↓</center></td>
</tr>
<tr>
<td>Llama Guard 2</td>
<td><center>0.749</center></td>
<td><center>0.794</center></td>
<td><center>0.284</center></td>
<td><center>0.683</center></td>
<td><center>0.677</center></td>
<td><center>0.670</center></td>
</tr>
<tr>
<td>Llama Guard 3</td>
<td><center>0.856</center></td>
<td><center>0.938</center></td>
<td><center>0.174</center></td>
<td><center>0.885</center></td>
<td><center>0.967</center></td>
<td><center>0.125</center></td>
</tr>
<tr>
<td>GPT4</td>
<td><center>0.732</center></td>
<td><center>N/A</center></td>
<td><center>0.525</center></td>
<td><center>0.636</center></td>
<td><center>N/A</center></td>
<td><center>0.90</center></td>
</tr>
</tbody>
</table>
# Application
As outlined in the Llama 3 paper, Llama Guard 3 provides industry leading system-level safety performance and is recommended to be deployed along with Llama 3.1. Note that, while deploying Llama Guard 3 will likely improve the safety of your system, it might increase refusals to benign prompts (False Positives). Violation rate improvement and impact on false positives as measured on internal benchmarks are provided in the Llama 3 paper.
# Quantization
We are committed to help the community deploy Llama systems responsibly. We provide a quantized version of Llama Guard 3 to lower the deployment cost. We used int 8 [implementation](https://huggingface.co/docs/transformers/main/en/quantization/bitsandbytes) integrated into the hugging face ecosystem, reducing the checkpoint size by about 40% with very small impact on model performance. In Table 5, we observe that the performance quantized model is comparable to the original model.
<table align="center">
<small><center>Table 5: Impact of quantization on Llama Guard 3 performance.</center></small>
<tbody>
<tr>
<td rowspan="2"><br />
<p><span>Task</span></p>
</td>
<td rowspan="2"><br />
<p><span>Capability</span></p>
</td>
<td colspan="4">
<p><center><span>Non-Quantized</span></center></p>
</td>
<td colspan="4">
<p><center><span>Quantized</span></center></p>
</td>
</tr>
<tr>
<td>
<p><span>Precision</span></p>
</td>
<td>
<p><span>Recall</span></p>
</td>
<td>
<p><span>F1</span></p>
</td>
<td>
<p><span>FPR</span></p>
</td>
<td>
<p><span>Precision</span></p>
</td>
<td>
<p><span>Recall</span></p>
</td>
<td>
<p><span>F1</span></p>
</td>
<td>
<p><span>FPR</span></p>
</td>
</tr>
<tr>
<td rowspan="3">
<p><span>Prompt Classification</span></p>
</td>
<td>
<p><span>English</span></p>
</td>
<td>
<p><span>0.952</span></p>
</td>
<td>
<p><span>0.943</span></p>
</td>
<td>
<p><span>0.947</span></p>
</td>
<td>
<p><span>0.057</span></p>
</td>
<td>
<p><span>0.961</span></p>
</td>
<td>
<p><span>0.939</span></p>
</td>
<td>
<p><span>0.950</span></p>
</td>
<td>
<p><span>0.045</span></p>
</td>
</tr>
<tr>
<td>
<p><span>Multilingual</span></p>
</td>
<td>
<p><span>0.901</span></p>
</td>
<td>
<p><span>0.899</span></p>
</td>
<td>
<p><span>0.900</span></p>
</td>
<td>
<p><span>0.054</span></p>
</td>
<td>
<p><span>0.906</span></p>
</td>
<td>
<p><span>0.892</span></p>
</td>
<td>
<p><span>0.899</span></p>
</td>
<td>
<p><span>0.051</span></p>
</td>
</tr>
<tr>
<td>
<p><span>Tool Use</span></p>
</td>
<td>
<p><span>0.884</span></p>
</td>
<td>
<p><span>0.958</span></p>
</td>
<td>
<p><span>0.920</span></p>
</td>
<td>
<p><span>0.126</span></p>
</td>
<td>
<p><span>0.876</span></p>
</td>
<td>
<p><span>0.946</span></p>
</td>
<td>
<p><span>0.909</span></p>
</td>
<td>
<p><span>0.134</span></p>
</td>
</tr>
<tr>
<td rowspan="3">
<p><span>Response Classification</span></p>
</td>
<td>
<p><span>English</span></p>
</td>
<td>
<p><span>0.947</span></p>
</td>
<td>
<p><span>0.931</span></p>
</td>
<td>
<p><span>0.939</span></p>
</td>
<td>
<p><span>0.040</span></p>
</td>
<td>
<p><span>0.947</span></p>
</td>
<td>
<p><span>0.925</span></p>
</td>
<td>
<p><span>0.936</span></p>
</td>
<td>
<p><span>0.040</span></p>
</td>
</tr>
<tr>
<td>
<p><span>Multilingual</span></p>
</td>
<td>
<p><span>0.929</span></p>
</td>
<td>
<p><span>0.805</span></p>
</td>
<td>
<p><span>0.862</span></p>
</td>
<td>
<p><span>0.033</span></p>
</td>
<td>
<p><span>0.931</span></p>
</td>
<td>
<p><span>0.785</span></p>
</td>
<td>
<p><span>0.851</span></p>
</td>
<td>
<p><span>0.031</span></p>
</td>
</tr>
<tr>
<td>
<p><span>Tool Use</span></p>
</td>
<td>
<p><span>0.774</span></p>
</td>
<td>
<p><span>0.884</span></p>
</td>
<td>
<p><span>0.825</span></p>
</td>
<td>
<p><span>0.176</span></p>
</td>
<td>
<p><span>0.793</span></p>
</td>
<td>
<p><span>0.865</span></p>
</td>
<td>
<p><span>0.827</span></p>
</td>
<td>
<p><span>0.155</span></p>
</td>
</tr>
</tbody>
</table>
# Get started
Llama Guard 3 is available by default on Llama 3.1 [reference implementations](https://github.com/meta-llama). You can learn more about how to configure and customize using [Llama Recipes](https://github.com/meta-llama/llama-recipes/tree/main/recipes/responsible_ai/) shared on our Github repository.
# Limitations
There are some limitations associated with Llama Guard 3. First, Llama Guard 3 itself is an LLM fine-tuned on Llama 3.1. Thus, its performance (e.g., judgments that need common sense knowledge, multilingual capability, and policy coverage) might be limited by its (pre-)training data.
Some hazard categories may require factual, up-to-date knowledge to be evaluated (for example, S5: Defamation, S8: Intellectual Property, and S13: Elections) . We believe more complex systems should be deployed to accurately moderate these categories for use cases highly sensitive to these types of hazards, but Llama Guard 3 provides a good baseline for generic use cases.
Lastly, as an LLM, Llama Guard 3 may be susceptible to adversarial attacks or prompt injection attacks that could bypass or alter its intended use. Please feel free to [report](https://github.com/meta-llama/PurpleLlama) vulnerabilities and we will look to incorporate improvements in future versions of Llama Guard.
# Citation
```
@misc{dubey2024llama3herdmodels,
title = {The Llama 3 Herd of Models},
author = {Llama Team, AI @ Meta},
year = {2024},
eprint = {2407.21783},
archivePrefix = {arXiv},
primaryClass = {cs.AI},
url = {https://arxiv.org/abs/2407.21783}
}
```
# References
[1] [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://arxiv.org/abs/2312.06674)
[2] [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://arxiv.org/abs/2204.05862)
[3] [Llama Guard 2 Model Card](https://github.com/meta-llama/PurpleLlama/blob/main/Llama-Guard2/MODEL_CARD.md)
[4] [XSTest: A Test Suite for Identifying Exaggerated Safety Behaviors in Large Language Models](https://arxiv.org/abs/2308.01263) |
Mungert/Llama-3.1-Nemotron-Nano-8B-v1-GGUF | Mungert | 2025-06-15T19:41:41Z | 2,163 | 7 | transformers | [
"transformers",
"gguf",
"nvidia",
"llama-3",
"pytorch",
"text-generation",
"en",
"arxiv:2505.00949",
"arxiv:2502.00203",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-21T19:44:49Z | ---
library_name: transformers
license: other
license_name: nvidia-open-model-license
license_link: >-
https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/
pipeline_tag: text-generation
language:
- en
tags:
- nvidia
- llama-3
- pytorch
---
# <span style="color: #7FFF7F;">Llama-3.1-Nemotron-Nano-8B-v1 GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.1-Nemotron-Nano-8B-v1-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.1-Nemotron-Nano-8B-v1-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.1-Nemotron-Nano-8B-v1-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.1-Nemotron-Nano-8B-v1-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.1-Nemotron-Nano-8B-v1-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.1-Nemotron-Nano-8B-v1-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.1-Nemotron-Nano-8B-v1-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.1-Nemotron-Nano-8B-v1-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.1-Nemotron-Nano-8B-v1-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.1-Nemotron-Nano-8B-v1-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.1-Nemotron-Nano-8B-v1-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Llama-3.1-Nemotron-Nano-8B-v1
## Model Overview
Llama-3.1-Nemotron-Nano-8B-v1 is a large language model (LLM) which is a derivative of [Meta Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) (AKA the reference model). It is a reasoning model that is post trained for reasoning, human chat preferences, and tasks, such as RAG and tool calling.
Llama-3.1-Nemotron-Nano-8B-v1 is a model which offers a great tradeoff between model accuracy and efficiency. It is created from Llama 3.1 8B Instruct and offers improvements in model accuracy. The model fits on a single RTX GPU and can be used locally. The model supports a context length of 128K.
This model underwent a multi-phase post-training process to enhance both its reasoning and non-reasoning capabilities. This includes a supervised fine-tuning stage for Math, Code, Reasoning, and Tool Calling as well as multiple reinforcement learning (RL) stages using REINFORCE (RLOO) and Online Reward-aware Preference Optimization (RPO) algorithms for both chat and instruction-following. The final model checkpoint is obtained after merging the final SFT and Online RPO checkpoints. Improved using Qwen.
This model is part of the Llama Nemotron Collection. You can find the other model(s) in this family here:
[Llama-3.3-Nemotron-Super-49B-v1](https://huggingface.co/nvidia/Llama-3.3-Nemotron-Super-49B-v1)
This model is ready for commercial use.
## License/Terms of Use
GOVERNING TERMS: Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.1 Community License Agreement](https://www.llama.com/llama3_1/license/). Built with Llama.
**Model Developer:** NVIDIA
**Model Dates:** Trained between August 2024 and March 2025
**Data Freshness:** The pretraining data has a cutoff of 2023 per Meta Llama 3.1 8B
## Use Case:
Developers designing AI Agent systems, chatbots, RAG systems, and other AI-powered applications. Also suitable for typical instruction-following tasks. Balance of model accuracy and compute efficiency (the model fits on a single RTX GPU and can be used locally).
## Release Date: <br>
3/18/2025 <br>
## References
- [\[2505.00949\] Llama-Nemotron: Efficient Reasoning Models](https://arxiv.org/abs/2505.00949)
- [\[2502.00203\] Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment](https://arxiv.org/abs/2502.00203)
## Model Architecture
**Architecture Type:** Dense decoder-only Transformer model
**Network Architecture:** Llama 3.1 8B Instruct
## Intended use
Llama-3.1-Nemotron-Nano-8B-v1 is a general purpose reasoning and chat model intended to be used in English and coding languages. Other non-English languages (German, French, Italian, Portuguese, Hindi, Spanish, and Thai) are also supported.
# Input:
- **Input Type:** Text
- **Input Format:** String
- **Input Parameters:** One-Dimensional (1D)
- **Other Properties Related to Input:** Context length up to 131,072 tokens
## Output:
- **Output Type:** Text
- **Output Format:** String
- **Output Parameters:** One-Dimensional (1D)
- **Other Properties Related to Output:** Context length up to 131,072 tokens
## Model Version:
1.0 (3/18/2025)
## Software Integration
- **Runtime Engine:** NeMo 24.12 <br>
- **Recommended Hardware Microarchitecture Compatibility:**
- NVIDIA Hopper
- NVIDIA Ampere
## Quick Start and Usage Recommendations:
1. Reasoning mode (ON/OFF) is controlled via the system prompt, which must be set as shown in the example below. All instructions should be contained within the user prompt
2. We recommend setting temperature to `0.6`, and Top P to `0.95` for Reasoning ON mode
3. We recommend using greedy decoding for Reasoning OFF mode
4. We have provided a list of prompts to use for evaluation for each benchmark where a specific template is required
5. The model will include `<think></think>` if no reasoning was necessary in Reasoning ON model, this is expected behaviour
You can try this model out through the preview API, using this link: [Llama-3.1-Nemotron-Nano-8B-v1](https://build.nvidia.com/nvidia/llama-3_1-nemotron-nano-8b-v1).
See the snippet below for usage with Hugging Face Transformers library. Reasoning mode (ON/OFF) is controlled via system prompt. Please see the example below.
Our code requires the transformers package version to be `4.44.2` or higher.
### Example of “Reasoning On:”
```python
import torch
import transformers
model_id = "nvidia/Llama-3.1-Nemotron-Nano-8B-v1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
temperature=0.6,
top_p=0.95,
**model_kwargs
)
# Thinking can be "on" or "off"
thinking = "on"
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))
```
### Example of “Reasoning Off:”
```python
import torch
import transformers
model_id = "nvidia/Llama-3.1-Nemotron-Nano-8B-v1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
do_sample=False,
**model_kwargs
)
# Thinking can be "on" or "off"
thinking = "off"
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))
```
For some prompts, even though thinking is disabled, the model emergently prefers to think before responding. But if desired, the users can prevent it by pre-filling the assistant response.
```python
import torch
import transformers
model_id = "nvidia/Llama-3.1-Nemotron-Nano-8B-v1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
# Thinking can be "on" or "off"
thinking = "off"
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
do_sample=False,
**model_kwargs
)
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}, {"role":"assistant", "content":"<think>\n</think>"}]))
```
## Inference:
**Engine:** Transformers
**Test Hardware:**
- BF16:
- 1x RTX 50 Series GPUs
- 1x RTX 40 Series GPUs
- 1x RTX 30 Series GPUs
- 1x H100-80GB GPU
- 1x A100-80GB GPU
**Preferred/Supported] Operating System(s):** Linux <br>
## Training Datasets
A large variety of training data was used for the post-training pipeline, including manually annotated data and synthetic data.
The data for the multi-stage post-training phases for improvements in Code, Math, and Reasoning is a compilation of SFT and RL data that supports improvements of math, code, general reasoning, and instruction following capabilities of the original Llama instruct model.
Prompts have been sourced from either public and open corpus or synthetically generated. Responses were synthetically generated by a variety of models, with some prompts containing responses for both Reasoning On and Off modes, to train the model to distinguish between two modes.
**Data Collection for Training Datasets:** <br>
* Hybrid: Automated, Human, Synthetic <br>
**Data Labeling for Training Datasets:** <br>
* N/A <br>
## Evaluation Datasets
We used the datasets listed below to evaluate Llama-3.1-Nemotron-Nano-8B-v1.
**Data Collection for Evaluation Datasets:** Hybrid: Human/Synthetic
**Data Labeling for Evaluation Datasets:** Hybrid: Human/Synthetic/Automatic
## Evaluation Results
These results contain both “Reasoning On”, and “Reasoning Off”. We recommend using temperature=`0.6`, top_p=`0.95` for “Reasoning On” mode, and greedy decoding for “Reasoning Off” mode. All evaluations are done with 32k sequence length. We run the benchmarks up to 16 times and average the scores to be more accurate.
> NOTE: Where applicable, a Prompt Template will be provided. While completing benchmarks, please ensure that you are parsing for the correct output format as per the provided prompt in order to reproduce the benchmarks seen below.
### MT-Bench
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 7.9 |
| Reasoning On | 8.1 |
### MATH500
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 36.6% |
| Reasoning On | 95.4% |
User Prompt Template:
```
"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \boxed{}.\nQuestion: {question}"
```
### AIME25
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 0% |
| Reasoning On | 47.1% |
User Prompt Template:
```
"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \boxed{}.\nQuestion: {question}"
```
### GPQA-D
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 39.4% |
| Reasoning On | 54.1% |
User Prompt Template:
```
"What is the correct answer to this question: {question}\nChoices:\nA. {option_A}\nB. {option_B}\nC. {option_C}\nD. {option_D}\nLet's think step by step, and put the final answer (should be a single letter A, B, C, or D) into a \boxed{}"
```
### IFEval Average
| Reasoning Mode | Strict:Prompt | Strict:Instruction |
|--------------|------------|------------|
| Reasoning Off | 74.7% | 82.1% |
| Reasoning On | 71.9% | 79.3% |
### BFCL v2 Live
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 63.9% |
| Reasoning On | 63.6% |
User Prompt Template:
```
<AVAILABLE_TOOLS>{functions}</AVAILABLE_TOOLS>
{user_prompt}
```
### MBPP 0-shot
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 66.1% |
| Reasoning On | 84.6% |
User Prompt Template:
````
You are an exceptionally intelligent coding assistant that consistently delivers accurate and reliable responses to user instructions.
@@ Instruction
Here is the given problem and test examples:
{prompt}
Please use the python programming language to solve this problem.
Please make sure that your code includes the functions from the test samples and that the input and output formats of these functions match the test samples.
Please return all completed codes in one code block.
This code block should be in the following format:
```python
# Your codes here
```
````
## Ethical Considerations:
NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
For more detailed information on ethical considerations for this model, please see the Model Card++ [Explainability](explainability.md), [Bias](bias.md), [Safety & Security](safety.md), and [Privacy](privacy.md) Subcards.
Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).
## Citation
```
@misc{bercovich2025llamanemotronefficientreasoningmodels,
title={Llama-Nemotron: Efficient Reasoning Models},
author={Akhiad Bercovich and Itay Levy and Izik Golan and Mohammad Dabbah and Ran El-Yaniv and Omri Puny and Ido Galil and Zach Moshe and Tomer Ronen and Najeeb Nabwani and Ido Shahaf and Oren Tropp and Ehud Karpas and Ran Zilberstein and Jiaqi Zeng and Soumye Singhal and Alexander Bukharin and Yian Zhang and Tugrul Konuk and Gerald Shen and Ameya Sunil Mahabaleshwarkar and Bilal Kartal and Yoshi Suhara and Olivier Delalleau and Zijia Chen and Zhilin Wang and David Mosallanezhad and Adi Renduchintala and Haifeng Qian and Dima Rekesh and Fei Jia and Somshubra Majumdar and Vahid Noroozi and Wasi Uddin Ahmad and Sean Narenthiran and Aleksander Ficek and Mehrzad Samadi and Jocelyn Huang and Siddhartha Jain and Igor Gitman and Ivan Moshkov and Wei Du and Shubham Toshniwal and George Armstrong and Branislav Kisacanin and Matvei Novikov and Daria Gitman and Evelina Bakhturina and Jane Polak Scowcroft and John Kamalu and Dan Su and Kezhi Kong and Markus Kliegl and Rabeeh Karimi and Ying Lin and Sanjeev Satheesh and Jupinder Parmar and Pritam Gundecha and Brandon Norick and Joseph Jennings and Shrimai Prabhumoye and Syeda Nahida Akter and Mostofa Patwary and Abhinav Khattar and Deepak Narayanan and Roger Waleffe and Jimmy Zhang and Bor-Yiing Su and Guyue Huang and Terry Kong and Parth Chadha and Sahil Jain and Christine Harvey and Elad Segal and Jining Huang and Sergey Kashirsky and Robert McQueen and Izzy Putterman and George Lam and Arun Venkatesan and Sherry Wu and Vinh Nguyen and Manoj Kilaru and Andrew Wang and Anna Warno and Abhilash Somasamudramath and Sandip Bhaskar and Maka Dong and Nave Assaf and Shahar Mor and Omer Ullman Argov and Scot Junkin and Oleksandr Romanenko and Pedro Larroy and Monika Katariya and Marco Rovinelli and Viji Balas and Nicholas Edelman and Anahita Bhiwandiwalla and Muthu Subramaniam and Smita Ithape and Karthik Ramamoorthy and Yuting Wu and Suguna Varshini Velury and Omri Almog and Joyjit Daw and Denys Fridman and Erick Galinkin and Michael Evans and Katherine Luna and Leon Derczynski and Nikki Pope and Eileen Long and Seth Schneider and Guillermo Siman and Tomasz Grzegorzek and Pablo Ribalta and Monika Katariya and Joey Conway and Trisha Saar and Ann Guan and Krzysztof Pawelec and Shyamala Prayaga and Oleksii Kuchaiev and Boris Ginsburg and Oluwatobi Olabiyi and Kari Briski and Jonathan Cohen and Bryan Catanzaro and Jonah Alben and Yonatan Geifman and Eric Chung and Chris Alexiuk},
year={2025},
eprint={2505.00949},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2505.00949},
}
``` |
Mungert/DeepSeek-R1-Distill-Qwen-14B-GGUF | Mungert | 2025-06-15T19:41:25Z | 528 | 7 | transformers | [
"transformers",
"gguf",
"arxiv:2501.12948",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-03-20T16:51:32Z | ---
license: mit
library_name: transformers
---
# <span style="color: #7FFF7F;">DeepSeek-R1-Distill-Qwen-14B GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `DeepSeek-R1-Distill-Qwen-14B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `DeepSeek-R1-Distill-Qwen-14B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `DeepSeek-R1-Distill-Qwen-14B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `DeepSeek-R1-Distill-Qwen-14B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `DeepSeek-R1-Distill-Qwen-14B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `DeepSeek-R1-Distill-Qwen-14B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `DeepSeek-R1-Distill-Qwen-14B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `DeepSeek-R1-Distill-Qwen-14B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `DeepSeek-R1-Distill-Qwen-14B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `DeepSeek-R1-Distill-Qwen-14B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `DeepSeek-R1-Distill-Qwen-14B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# DeepSeek-R1
<!-- markdownlint-disable first-line-h1 -->
<!-- markdownlint-disable html -->
<!-- markdownlint-disable no-duplicate-header -->
<div align="center">
<img src="https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/logo.svg?raw=true" width="60%" alt="DeepSeek-V3" />
</div>
<hr>
<div align="center" style="line-height: 1;">
<a href="https://www.deepseek.com/" target="_blank" style="margin: 2px;">
<img alt="Homepage" src="https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/badge.svg?raw=true" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://chat.deepseek.com/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/🤖%20Chat-DeepSeek%20R1-536af5?color=536af5&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://huggingface.co/deepseek-ai" target="_blank" style="margin: 2px;">
<img alt="Hugging Face" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-DeepSeek%20AI-ffc107?color=ffc107&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
<div align="center" style="line-height: 1;">
<a href="https://discord.gg/Tc7c45Zzu5" target="_blank" style="margin: 2px;">
<img alt="Discord" src="https://img.shields.io/badge/Discord-DeepSeek%20AI-7289da?logo=discord&logoColor=white&color=7289da" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/qr.jpeg?raw=true" target="_blank" style="margin: 2px;">
<img alt="Wechat" src="https://img.shields.io/badge/WeChat-DeepSeek%20AI-brightgreen?logo=wechat&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://twitter.com/deepseek_ai" target="_blank" style="margin: 2px;">
<img alt="Twitter Follow" src="https://img.shields.io/badge/Twitter-deepseek_ai-white?logo=x&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
<div align="center" style="line-height: 1;">
<a href="https://github.com/deepseek-ai/DeepSeek-R1/blob/main/LICENSE" style="margin: 2px;">
<img alt="License" src="https://img.shields.io/badge/License-MIT-f5de53?&color=f5de53" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
<p align="center">
<a href="https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSeek_R1.pdf"><b>Paper Link</b>👁️</a>
</p>
## 1. Introduction
We introduce our first-generation reasoning models, DeepSeek-R1-Zero and DeepSeek-R1.
DeepSeek-R1-Zero, a model trained via large-scale reinforcement learning (RL) without supervised fine-tuning (SFT) as a preliminary step, demonstrated remarkable performance on reasoning.
With RL, DeepSeek-R1-Zero naturally emerged with numerous powerful and interesting reasoning behaviors.
However, DeepSeek-R1-Zero encounters challenges such as endless repetition, poor readability, and language mixing. To address these issues and further enhance reasoning performance,
we introduce DeepSeek-R1, which incorporates cold-start data before RL.
DeepSeek-R1 achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks.
To support the research community, we have open-sourced DeepSeek-R1-Zero, DeepSeek-R1, and six dense models distilled from DeepSeek-R1 based on Llama and Qwen. DeepSeek-R1-Distill-Qwen-32B outperforms OpenAI-o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.
**NOTE: Before running DeepSeek-R1 series models locally, we kindly recommend reviewing the [Usage Recommendation](#usage-recommendations) section.**
<p align="center">
<img width="80%" src="figures/benchmark.jpg">
</p>
## 2. Model Summary
---
**Post-Training: Large-Scale Reinforcement Learning on the Base Model**
- We directly apply reinforcement learning (RL) to the base model without relying on supervised fine-tuning (SFT) as a preliminary step. This approach allows the model to explore chain-of-thought (CoT) for solving complex problems, resulting in the development of DeepSeek-R1-Zero. DeepSeek-R1-Zero demonstrates capabilities such as self-verification, reflection, and generating long CoTs, marking a significant milestone for the research community. Notably, it is the first open research to validate that reasoning capabilities of LLMs can be incentivized purely through RL, without the need for SFT. This breakthrough paves the way for future advancements in this area.
- We introduce our pipeline to develop DeepSeek-R1. The pipeline incorporates two RL stages aimed at discovering improved reasoning patterns and aligning with human preferences, as well as two SFT stages that serve as the seed for the model's reasoning and non-reasoning capabilities.
We believe the pipeline will benefit the industry by creating better models.
---
**Distillation: Smaller Models Can Be Powerful Too**
- We demonstrate that the reasoning patterns of larger models can be distilled into smaller models, resulting in better performance compared to the reasoning patterns discovered through RL on small models. The open source DeepSeek-R1, as well as its API, will benefit the research community to distill better smaller models in the future.
- Using the reasoning data generated by DeepSeek-R1, we fine-tuned several dense models that are widely used in the research community. The evaluation results demonstrate that the distilled smaller dense models perform exceptionally well on benchmarks. We open-source distilled 1.5B, 7B, 8B, 14B, 32B, and 70B checkpoints based on Qwen2.5 and Llama3 series to the community.
## 3. Model Downloads
### DeepSeek-R1 Models
<div align="center">
| **Model** | **#Total Params** | **#Activated Params** | **Context Length** | **Download** |
| :------------: | :------------: | :------------: | :------------: | :------------: |
| DeepSeek-R1-Zero | 671B | 37B | 128K | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Zero) |
| DeepSeek-R1 | 671B | 37B | 128K | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1) |
</div>
DeepSeek-R1-Zero & DeepSeek-R1 are trained based on DeepSeek-V3-Base.
For more details regarding the model architecture, please refer to [DeepSeek-V3](https://github.com/deepseek-ai/DeepSeek-V3) repository.
### DeepSeek-R1-Distill Models
<div align="center">
| **Model** | **Base Model** | **Download** |
| :------------: | :------------: | :------------: |
| DeepSeek-R1-Distill-Qwen-1.5B | [Qwen2.5-Math-1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) |
| DeepSeek-R1-Distill-Qwen-7B | [Qwen2.5-Math-7B](https://huggingface.co/Qwen/Qwen2.5-Math-7B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B) |
| DeepSeek-R1-Distill-Llama-8B | [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B) |
| DeepSeek-R1-Distill-Qwen-14B | [Qwen2.5-14B](https://huggingface.co/Qwen/Qwen2.5-14B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B) |
|DeepSeek-R1-Distill-Qwen-32B | [Qwen2.5-32B](https://huggingface.co/Qwen/Qwen2.5-32B) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B) |
| DeepSeek-R1-Distill-Llama-70B | [Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct) | [🤗 HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B) |
</div>
DeepSeek-R1-Distill models are fine-tuned based on open-source models, using samples generated by DeepSeek-R1.
We slightly change their configs and tokenizers. Please use our setting to run these models.
## 4. Evaluation Results
### DeepSeek-R1-Evaluation
For all our models, the maximum generation length is set to 32,768 tokens. For benchmarks requiring sampling, we use a temperature of $0.6$, a top-p value of $0.95$, and generate 64 responses per query to estimate pass@1.
<div align="center">
| Category | Benchmark (Metric) | Claude-3.5-Sonnet-1022 | GPT-4o 0513 | DeepSeek V3 | OpenAI o1-mini | OpenAI o1-1217 | DeepSeek R1 |
|----------|-------------------|----------------------|------------|--------------|----------------|------------|--------------|
| | Architecture | - | - | MoE | - | - | MoE |
| | # Activated Params | - | - | 37B | - | - | 37B |
| | # Total Params | - | - | 671B | - | - | 671B |
| English | MMLU (Pass@1) | 88.3 | 87.2 | 88.5 | 85.2 | **91.8** | 90.8 |
| | MMLU-Redux (EM) | 88.9 | 88.0 | 89.1 | 86.7 | - | **92.9** |
| | MMLU-Pro (EM) | 78.0 | 72.6 | 75.9 | 80.3 | - | **84.0** |
| | DROP (3-shot F1) | 88.3 | 83.7 | 91.6 | 83.9 | 90.2 | **92.2** |
| | IF-Eval (Prompt Strict) | **86.5** | 84.3 | 86.1 | 84.8 | - | 83.3 |
| | GPQA-Diamond (Pass@1) | 65.0 | 49.9 | 59.1 | 60.0 | **75.7** | 71.5 |
| | SimpleQA (Correct) | 28.4 | 38.2 | 24.9 | 7.0 | **47.0** | 30.1 |
| | FRAMES (Acc.) | 72.5 | 80.5 | 73.3 | 76.9 | - | **82.5** |
| | AlpacaEval2.0 (LC-winrate) | 52.0 | 51.1 | 70.0 | 57.8 | - | **87.6** |
| | ArenaHard (GPT-4-1106) | 85.2 | 80.4 | 85.5 | 92.0 | - | **92.3** |
| Code | LiveCodeBench (Pass@1-COT) | 33.8 | 34.2 | - | 53.8 | 63.4 | **65.9** |
| | Codeforces (Percentile) | 20.3 | 23.6 | 58.7 | 93.4 | **96.6** | 96.3 |
| | Codeforces (Rating) | 717 | 759 | 1134 | 1820 | **2061** | 2029 |
| | SWE Verified (Resolved) | **50.8** | 38.8 | 42.0 | 41.6 | 48.9 | 49.2 |
| | Aider-Polyglot (Acc.) | 45.3 | 16.0 | 49.6 | 32.9 | **61.7** | 53.3 |
| Math | AIME 2024 (Pass@1) | 16.0 | 9.3 | 39.2 | 63.6 | 79.2 | **79.8** |
| | MATH-500 (Pass@1) | 78.3 | 74.6 | 90.2 | 90.0 | 96.4 | **97.3** |
| | CNMO 2024 (Pass@1) | 13.1 | 10.8 | 43.2 | 67.6 | - | **78.8** |
| Chinese | CLUEWSC (EM) | 85.4 | 87.9 | 90.9 | 89.9 | - | **92.8** |
| | C-Eval (EM) | 76.7 | 76.0 | 86.5 | 68.9 | - | **91.8** |
| | C-SimpleQA (Correct) | 55.4 | 58.7 | **68.0** | 40.3 | - | 63.7 |
</div>
### Distilled Model Evaluation
<div align="center">
| Model | AIME 2024 pass@1 | AIME 2024 cons@64 | MATH-500 pass@1 | GPQA Diamond pass@1 | LiveCodeBench pass@1 | CodeForces rating |
|------------------------------------------|------------------|-------------------|-----------------|----------------------|----------------------|-------------------|
| GPT-4o-0513 | 9.3 | 13.4 | 74.6 | 49.9 | 32.9 | 759 |
| Claude-3.5-Sonnet-1022 | 16.0 | 26.7 | 78.3 | 65.0 | 38.9 | 717 |
| o1-mini | 63.6 | 80.0 | 90.0 | 60.0 | 53.8 | **1820** |
| QwQ-32B-Preview | 44.0 | 60.0 | 90.6 | 54.5 | 41.9 | 1316 |
| DeepSeek-R1-Distill-Qwen-1.5B | 28.9 | 52.7 | 83.9 | 33.8 | 16.9 | 954 |
| DeepSeek-R1-Distill-Qwen-7B | 55.5 | 83.3 | 92.8 | 49.1 | 37.6 | 1189 |
| DeepSeek-R1-Distill-Qwen-14B | 69.7 | 80.0 | 93.9 | 59.1 | 53.1 | 1481 |
| DeepSeek-R1-Distill-Qwen-32B | **72.6** | 83.3 | 94.3 | 62.1 | 57.2 | 1691 |
| DeepSeek-R1-Distill-Llama-8B | 50.4 | 80.0 | 89.1 | 49.0 | 39.6 | 1205 |
| DeepSeek-R1-Distill-Llama-70B | 70.0 | **86.7** | **94.5** | **65.2** | **57.5** | 1633 |
</div>
## 5. Chat Website & API Platform
You can chat with DeepSeek-R1 on DeepSeek's official website: [chat.deepseek.com](https://chat.deepseek.com), and switch on the button "DeepThink"
We also provide OpenAI-Compatible API at DeepSeek Platform: [platform.deepseek.com](https://platform.deepseek.com/)
## 6. How to Run Locally
### DeepSeek-R1 Models
Please visit [DeepSeek-V3](https://github.com/deepseek-ai/DeepSeek-V3) repo for more information about running DeepSeek-R1 locally.
**NOTE: Hugging Face's Transformers has not been directly supported yet.**
### DeepSeek-R1-Distill Models
DeepSeek-R1-Distill models can be utilized in the same manner as Qwen or Llama models.
For instance, you can easily start a service using [vLLM](https://github.com/vllm-project/vllm):
```shell
vllm serve deepseek-ai/DeepSeek-R1-Distill-Qwen-32B --tensor-parallel-size 2 --max-model-len 32768 --enforce-eager
```
You can also easily start a service using [SGLang](https://github.com/sgl-project/sglang)
```bash
python3 -m sglang.launch_server --model deepseek-ai/DeepSeek-R1-Distill-Qwen-32B --trust-remote-code --tp 2
```
### Usage Recommendations
**We recommend adhering to the following configurations when utilizing the DeepSeek-R1 series models, including benchmarking, to achieve the expected performance:**
1. Set the temperature within the range of 0.5-0.7 (0.6 is recommended) to prevent endless repetitions or incoherent outputs.
2. **Avoid adding a system prompt; all instructions should be contained within the user prompt.**
3. For mathematical problems, it is advisable to include a directive in your prompt such as: "Please reason step by step, and put your final answer within \boxed{}."
4. When evaluating model performance, it is recommended to conduct multiple tests and average the results.
Additionally, we have observed that the DeepSeek-R1 series models tend to bypass thinking pattern (i.e., outputting "\<think\>\n\n\</think\>") when responding to certain queries, which can adversely affect the model's performance.
**To ensure that the model engages in thorough reasoning, we recommend enforcing the model to initiate its response with "\<think\>\n" at the beginning of every output.**
## 7. License
This code repository and the model weights are licensed under the [MIT License](https://github.com/deepseek-ai/DeepSeek-R1/blob/main/LICENSE).
DeepSeek-R1 series support commercial use, allow for any modifications and derivative works, including, but not limited to, distillation for training other LLMs. Please note that:
- DeepSeek-R1-Distill-Qwen-1.5B, DeepSeek-R1-Distill-Qwen-7B, DeepSeek-R1-Distill-Qwen-14B and DeepSeek-R1-Distill-Qwen-32B are derived from [Qwen-2.5 series](https://github.com/QwenLM/Qwen2.5), which are originally licensed under [Apache 2.0 License](https://huggingface.co/Qwen/Qwen2.5-1.5B/blob/main/LICENSE), and now finetuned with 800k samples curated with DeepSeek-R1.
- DeepSeek-R1-Distill-Llama-8B is derived from Llama3.1-8B-Base and is originally licensed under [llama3.1 license](https://huggingface.co/meta-llama/Llama-3.1-8B/blob/main/LICENSE).
- DeepSeek-R1-Distill-Llama-70B is derived from Llama3.3-70B-Instruct and is originally licensed under [llama3.3 license](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct/blob/main/LICENSE).
## 8. Citation
```
@misc{deepseekai2025deepseekr1incentivizingreasoningcapability,
title={DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning},
author={DeepSeek-AI},
year={2025},
eprint={2501.12948},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2501.12948},
}
```
## 9. Contact
If you have any questions, please raise an issue or contact us at [[email protected]]([email protected]).
|
sinha-mayank-900/distilhubert-finetuned-gtzan | sinha-mayank-900 | 2025-06-15T19:41:02Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"hubert",
"audio-classification",
"generated_from_trainer",
"dataset:marsyas/gtzan",
"base_model:ntu-spml/distilhubert",
"base_model:finetune:ntu-spml/distilhubert",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | audio-classification | 2025-06-15T15:07:40Z | ---
library_name: transformers
license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
metrics:
- accuracy
model-index:
- name: distilhubert-finetuned-gtzan
results:
- task:
name: Audio Classification
type: audio-classification
dataset:
name: GTZAN
type: marsyas/gtzan
config: all
split: train
args: all
metrics:
- name: Accuracy
type: accuracy
value: 0.86
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilhubert-finetuned-gtzan
This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6759
- Accuracy: 0.86
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 3e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine_with_restarts
- lr_scheduler_warmup_ratio: 0.15
- num_epochs: 30
### Training results
| Training Loss | Epoch | Step | Accuracy | Validation Loss |
|:-------------:|:-----:|:----:|:--------:|:---------------:|
| 2.2385 | 1.0 | 50 | 0.34 | 2.2256 |
| 1.7883 | 2.0 | 100 | 0.545 | 1.7487 |
| 1.4788 | 3.0 | 150 | 0.68 | 1.4187 |
| 1.1262 | 4.0 | 200 | 0.69 | 1.1004 |
| 0.8664 | 5.0 | 250 | 0.735 | 0.9532 |
| 0.7772 | 6.0 | 300 | 0.745 | 0.8106 |
| 0.4455 | 7.0 | 350 | 0.81 | 0.7057 |
| 0.3719 | 8.0 | 400 | 0.815 | 0.6467 |
| 0.3716 | 9.0 | 450 | 0.805 | 0.6164 |
| 0.223 | 10.0 | 500 | 0.825 | 0.5887 |
| 0.1382 | 11.0 | 550 | 0.83 | 0.5941 |
| 0.0729 | 12.0 | 600 | 0.84 | 0.5911 |
| 0.0518 | 13.0 | 650 | 0.84 | 0.6116 |
| 0.0388 | 14.0 | 700 | 0.835 | 0.6217 |
| 0.0304 | 15.0 | 750 | 0.84 | 0.6340 |
| 0.0266 | 16.0 | 800 | 0.84 | 0.6407 |
| 0.026 | 17.0 | 850 | 0.85 | 0.6428 |
| 0.0238 | 18.0 | 900 | 0.84 | 0.6457 |
| 0.0244 | 19.0 | 950 | 0.85 | 0.6457 |
| 0.0278 | 20.0 | 1000 | 0.845 | 0.6466 |
| 0.0676 | 19.0 | 1026 | 0.6300 | 0.8533 |
| 0.0271 | 20.0 | 1080 | 0.6714 | 0.8467 |
| 0.0165 | 21.0 | 1134 | 0.6385 | 0.8533 |
| 0.0158 | 22.0 | 1188 | 0.6895 | 0.8667 |
| 0.0292 | 23.0 | 1242 | 0.6982 | 0.86 |
| 0.0232 | 24.0 | 1296 | 0.6870 | 0.86 |
| 0.0099 | 25.0 | 1350 | 0.6774 | 0.8667 |
| 0.0104 | 26.0 | 1404 | 0.6821 | 0.86 |
| 0.0101 | 27.0 | 1458 | 0.6773 | 0.86 |
| 0.01 | 28.0 | 1512 | 0.6790 | 0.86 |
| 0.0097 | 29.0 | 1566 | 0.6779 | 0.86 |
| 0.0093 | 30.0 | 1620 | 0.6759 | 0.86 |
### Framework versions
- Transformers 4.52.4
- Pytorch 2.7.1+cu126
- Datasets 3.6.0
- Tokenizers 0.21.1
|
Mungert/granite-3.2-2b-instruct-GGUF | Mungert | 2025-06-15T19:40:53Z | 658 | 6 | transformers | [
"transformers",
"gguf",
"language",
"granite-3.2",
"text-generation",
"arxiv:0000.00000",
"base_model:ibm-granite/granite-3.1-2b-instruct",
"base_model:quantized:ibm-granite/granite-3.1-2b-instruct",
"license:apache-2.0",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-18T21:59:43Z | ---
pipeline_tag: text-generation
inference: false
license: apache-2.0
library_name: transformers
tags:
- language
- granite-3.2
base_model:
- ibm-granite/granite-3.1-2b-instruct
---
# <span style="color: #7FFF7F;">granite-3.2-2b-instruct GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `granite-3.2-2b-instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `granite-3.2-2b-instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `granite-3.2-2b-instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `granite-3.2-2b-instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `granite-3.2-2b-instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `granite-3.2-2b-instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `granite-3.2-2b-instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `granite-3.2-2b-instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `granite-3.2-2b-instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `granite-3.2-2b-instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `granite-3.2-2b-instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Granite-3.2-2B-Instruct
**Model Summary:**
Granite-3.2-2B-Instruct is an 2-billion-parameter, long-context AI model fine-tuned for thinking capabilities. Built on top of [Granite-3.1-2B-Instruct](https://huggingface.co/ibm-granite/granite-3.1-2b-instruct), it has been trained using a mix of permissively licensed open-source datasets and internally generated synthetic data designed for reasoning tasks. The model allows controllability of its thinking capability, ensuring it is applied only when required.
- **Developers:** Granite Team, IBM
- **Website**: [Granite Docs](https://www.ibm.com/granite/docs/)
- **Release Date**: February 26th, 2025
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)
**Supported Languages:**
English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. However, users may finetune this Granite model for languages beyond these 12 languages.
**Intended Use:**
This model is designed to handle general instruction-following tasks and can be integrated into AI assistants across various domains, including business applications.
**Capabilities**
* **Thinking**
* Summarization
* Text classification
* Text extraction
* Question-answering
* Retrieval Augmented Generation (RAG)
* Code related tasks
* Function-calling tasks
* Multilingual dialog use cases
* Long-context tasks including long document/meeting summarization, long document QA, etc.
**Generation:**
This is a simple example of how to use Granite-3.2-2B-Instruct model.
Install the following libraries:
```shell
pip install torch torchvision torchaudio
pip install accelerate
pip install transformers
```
Then, copy the snippet from the section that is relevant for your use case.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer, set_seed
import torch
model_path="ibm-granite/granite-3.2-2b-instruct"
device="cuda"
model = AutoModelForCausalLM.from_pretrained(
model_path,
device_map=device,
torch_dtype=torch.bfloat16,
)
tokenizer = AutoTokenizer.from_pretrained(
model_path
)
conv = [{"role": "user", "content":"You have 10 liters of a 30% acid solution. How many liters of a 70% acid solution must be added to achieve a 50% acid mixture?"}]
input_ids = tokenizer.apply_chat_template(conv, return_tensors="pt", thinking=True, return_dict=True, add_generation_prompt=True).to(device)
set_seed(42)
output = model.generate(
**input_ids,
max_new_tokens=8192,
)
prediction = tokenizer.decode(output[0, input_ids["input_ids"].shape[1]:], skip_special_tokens=True)
print(prediction)
```
**Example Outputs**
- thinking=True
```md
Here is my thought process:
1. First, let's understand the composition of the initial 10 liters of 30% acid solution:
- Acid content = 30% of 10 liters = 0.30 * 10 = 3 liters
2. We want to end up with a 50% acid solution. Let's denote the volume of the 70% acid solution to be added as 'x' liters.
3. The amount of acid in the final mixture (50% solution) should equal the initial acid content plus the acid from the added 70% solution.
- Total acid in final mixture = 0.50 * (10 + x)
4. The equation representing the total acid content is:
3 (from initial solution) + 0.70 * x (from added solution) = 0.50 * (10 + x)
5. Now, we solve for 'x':
3 + 0.70x = 5 + 0.50x
0.70x - 0.50x = 5 - 3
0.20x = 2
x = 2 / 0.20
x = 10 liters
6. Therefore, 10 liters of a 70% acid solution must be added to the initial 10 liters of 30% acid solution to achieve a 50% acid mixture.
Here is my response:
To solve this problem, we need to determine how much of a 70% acid solution must be added to a 10-liter solution that is already 30% acid to create a new mixture that is 50% acid. Here's how to calculate it:
1. **Initial Acid Content**:
- The initial 10-liter solution contains 3 liters of acid (since 30% of 10 liters).
2. **Desired Final Acid Content**:
- We want the final mixture to be 50% acid.
3. **Equation Setup**:
- Let 'x' be the volume (in liters) of the 70% acid solution to be added.
- The equation representing the total acid content in the final mixture (which is 10 liters from the initial solution plus 'x' liters of the added solution) should equal 50% of the total volume (10 + x).
- This gives us: 3 (acid from initial solution) + 0.70x (acid from added solution) = 0.50(10 + x)
4. **Solving the Equation**:
- First, simplify the right side of the equation: 0.50(10 + x) = 5 + 0.50x
- Now, the equation is: 3 + 0.70x = 5 + 0.50x
- Subtract 0.50x from both sides: 0.20x = 2
- Divide both sides by 0.20: x = 2 / 0.20
- x = 10 liters
5. **Conclusion**:
- To achieve a 50% acid mixture from the initial 30% solution, you need to add 10 liters of a 70% acid solution.
So, the final answer is that 10 liters of a 70% acid solution must be added to the initial 10 liters of 30% acid solution.
```
- thinking=False
```md
To solve this problem, we'll use the concept of volume and concentration. Let's denote the amount of 70% acid solution we need to add as "x" liters.
First, let's find out how much acid is in the initial 10-liter solution:
Acid in initial solution = 30% of 10 liters = 0.30 * 10 = 3 liters
Now, we want to end up with a 50% acid solution in a total volume of (10 + x) liters. Let's denote the final volume as V.
Final acid concentration = 50%
Final acid amount = 50% of V = 0.50 * V
We know the initial acid amount and the final acid amount, so we can set up an equation:
Initial acid amount + Acid from added solution = Final acid amount
3 liters + (70% of x) = 0.50 * (10 + x)
Now, let's solve for x:
0.70x + 3 = 0.50 * 10 + 0.50x
0.70x - 0.50x = 0.50 * 10 - 3
0.20x = 5 - 3
0.20x = 2
x = 2 / 0.20
x = 10 liters
So, you need to add 10 liters of a 70% acid solution to the initial 10-liter 30% acid solution to achieve a 50% acid mixture.
```
**Evaluation Results:**
<table>
<thead>
<tr>
<th style="text-align:left; background-color: #001d6c; color: white;">Models</th>
<th style="text-align:center; background-color: #001d6c; color: white;">ArenaHard</th>
<th style="text-align:center; background-color: #001d6c; color: white;">Alpaca-Eval-2</th>
<th style="text-align:center; background-color: #001d6c; color: white;">MMLU</th>
<th style="text-align:center; background-color: #001d6c; color: white;">PopQA</th>
<th style="text-align:center; background-color: #001d6c; color: white;">TruthfulQA</th>
<th style="text-align:center; background-color: #001d6c; color: white;">BigBenchHard</th>
<th style="text-align:center; background-color: #001d6c; color: white;">DROP</th>
<th style="text-align:center; background-color: #001d6c; color: white;">GSM8K</th>
<th style="text-align:center; background-color: #001d6c; color: white;">HumanEval</th>
<th style="text-align:center; background-color: #001d6c; color: white;">HumanEval+</th>
<th style="text-align:center; background-color: #001d6c; color: white;">IFEval</th>
<th style="text-align:center; background-color: #001d6c; color: white;">AttaQ</th>
</tr></thead>
<tbody>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">Llama-3.1-8B-Instruct</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">36.43</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">27.22</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">69.15</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">28.79</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">52.79</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">72.66</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">61.48</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">83.24</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">85.32</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">80.15</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">79.10</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">83.43</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">DeepSeek-R1-Distill-Llama-8B</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">17.17</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">21.85</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">45.80</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">13.25</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">47.43</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">65.71</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">44.46</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">72.18</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">67.54</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">62.91</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">66.50</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">42.87</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">Qwen-2.5-7B-Instruct</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">25.44</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">30.34</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">74.30</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">18.12</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">63.06</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">70.40</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">54.71</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">84.46</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">93.35</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">89.91</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">74.90</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">81.90</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">DeepSeek-R1-Distill-Qwen-7B</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">10.36</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">15.35</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">50.72</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">9.94</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">47.14</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">65.04</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">42.76</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">78.47</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">79.89</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">78.43</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">59.10</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">42.45</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">Granite-3.1-8B-Instruct</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">37.58</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">30.34</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">66.77</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">28.7</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">65.84</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">68.55</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">50.78</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">79.15</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">89.63</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">85.79</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">73.20</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">85.73</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">Granite-3.1-2B-Instruct</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">23.3</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">27.17</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">57.11</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">20.55</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">59.79</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">54.46</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">18.68</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">67.55</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">79.45</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">75.26</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">63.59</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">84.7</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;">Granite-3.2-8B-Instruct</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">55.25</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">61.19</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">66.79</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">28.04</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">66.92</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">64.77</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">50.95</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">81.65</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">89.35</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">85.72</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">74.31</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">85.42</td>
</tr>
<tr>
<td style="text-align:left; background-color: #DAE8FF; color: black;"><b>Granite-3.2-2B-Instruct</b></td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">24.86</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">34.51</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">57.18</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">20.56</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">59.8</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">52.27</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">21.12</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">67.02</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">80.13</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">73.39</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">61.55</td>
<td style="text-align:center; background-color: #DAE8FF; color: black;">83.23</td>
</tr>
</tbody></table>
**Training Data:**
Overall, our training data is largely comprised of two key sources: (1) publicly available datasets with permissive license, (2) internal synthetically generated data targeted to enhance reasoning capabilites.
<!-- A detailed attribution of datasets can be found in [Granite 3.2 Technical Report (coming soon)](#), and [Accompanying Author List](https://github.com/ibm-granite/granite-3.0-language-models/blob/main/author-ack.pdf). -->
**Infrastructure:**
We train Granite-3.2-2B-Instruct using IBM's super computing cluster, Blue Vela, which is outfitted with NVIDIA H100 GPUs. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs.
**Ethical Considerations and Limitations:**
Granite-3.2-2B-Instruct builds upon Granite-3.1-2B-Instruct, leveraging both permissively licensed open-source and select proprietary data for enhanced performance. Since it inherits its foundation from the previous model, all ethical considerations and limitations applicable to [Granite-3.1-2B-Instruct](https://huggingface.co/ibm-granite/granite-3.1-2b-instruct) remain relevant.
**Resources**
- ⭐️ Learn about the latest updates with Granite: https://www.ibm.com/granite
- 📄 Get started with tutorials, best practices, and prompt engineering advice: https://www.ibm.com/granite/docs/
- 💡 Learn about the latest Granite learning resources: https://ibm.biz/granite-learning-resources
<!-- ## Citation
```
@misc{granite-models,
author = {author 1, author2, ...},
title = {},
journal = {},
volume = {},
year = {2024},
url = {https://arxiv.org/abs/0000.00000},
}
``` --> |
Mungert/rwkv7-191M-world-GGUF | Mungert | 2025-06-15T19:40:43Z | 459 | 2 | null | [
"gguf",
"text-generation",
"en",
"zh",
"ja",
"ko",
"fr",
"ar",
"es",
"pt",
"base_model:BlinkDL/rwkv-7-world",
"base_model:quantized:BlinkDL/rwkv-7-world",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-18T09:10:04Z | ---
license: apache-2.0
language:
- en
- zh
- ja
- ko
- fr
- ar
- es
- pt
metrics:
- accuracy
base_model:
- BlinkDL/rwkv-7-world
pipeline_tag: text-generation
---
# <span style="color: #7FFF7F;">rwkv7-191M-world GGUF Models</span>
Note: you must use latest llama.cpp https://github.com/ggml-org/llama.cpp to run this model with llama.cpp
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `rwkv7-191M-world-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `rwkv7-191M-world-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `rwkv7-191M-world-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `rwkv7-191M-world-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `rwkv7-191M-world-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `rwkv7-191M-world-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `rwkv7-191M-world-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `rwkv7-191M-world-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `rwkv7-191M-world-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `rwkv7-191M-world-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `rwkv7-191M-world-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# rwkv7-191M-world
<!-- Provide a quick summary of what the model is/does. -->
This is RWKV-7 model under flash-linear attention format.
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
- **Developed by:** Bo Peng, Yu Zhang, Songlin Yang, Ruichong Zhang
- **Funded by:** RWKV Project (Under LF AI & Data Foundation)
- **Model type:** RWKV7
- **Language(s) (NLP):** English
- **License:** Apache-2.0
- **Parameter count:** 191M
- **Tokenizer:** RWKV World tokenizer
- **Vocabulary size:** 65,536
### Model Sources
<!-- Provide the basic links for the model. -->
- **Repository:** https://github.com/fla-org/flash-linear-attention ; https://github.com/BlinkDL/RWKV-LM
- **Paper:** With in Progress
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
Install `flash-linear-attention` and the latest version of `transformers` before using this model:
```bash
pip install git+https://github.com/fla-org/flash-linear-attention
pip install 'transformers>=4.48.0'
```
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
You can use this model just as any other HuggingFace models:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained('fla-hub/rwkv7-191M-world', trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained('fla-hub/rwkv7-191M-world', trust_remote_code=True)
model = model.cuda()
prompt = "What is a large language model?"
messages = [
{"role": "user", "content": "Who are you?"},
{"role": "assistant", "content": "I am a GPT-3 based model."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=1024,
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=False)[0]
print(response)
```
## Training Details
### Training Data
This model is trained on the World v2.8 with a total of 1.0 trillion tokens.
#### Training Hyperparameters
- **Training regime:** bfloat16, lr 8e-4 to 1e-5 "delayed" cosine decay, wd 0.1, bsz 8x30x4096 (Please contact Peng Bo for details)
## Evaluation
#### Metrics
`lambada_openai`: ppl 12.4 acc 49.1%
## FAQ
Q: safetensors metadata is none.
A: upgrade transformers to >=4.48.0: `pip install 'transformers>=4.48.0'` |
Mungert/rwkv7-1.5B-world-GGUF | Mungert | 2025-06-15T19:40:40Z | 811 | 1 | null | [
"gguf",
"text-generation",
"en",
"zh",
"ja",
"ko",
"fr",
"ar",
"es",
"pt",
"base_model:BlinkDL/rwkv-7-world",
"base_model:quantized:BlinkDL/rwkv-7-world",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-18T08:04:12Z | ---
license: apache-2.0
language:
- en
- zh
- ja
- ko
- fr
- ar
- es
- pt
metrics:
- accuracy
base_model:
- BlinkDL/rwkv-7-world
pipeline_tag: text-generation
---
# <span style="color: #7FFF7F;">rwkv7-1.5B-world GGUF Models</span>
Note: you must use latest llama.cpp https://github.com/ggml-org/llama.cpp to run this model with llama.cpp
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `rwkv7-1.5B-world-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `rwkv7-1.5B-world-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `rwkv7-1.5B-world-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `rwkv7-1.5B-world-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `rwkv7-1.5B-world-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `rwkv7-1.5B-world-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `rwkv7-1.5B-world-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `rwkv7-1.5B-world-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `rwkv7-1.5B-world-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `rwkv7-1.5B-world-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `rwkv7-1.5B-world-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# rwkv7-1.5B-world
<!-- Provide a quick summary of what the model is/does. -->
This is RWKV-7 model under flash-linear attention format.
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
- **Developed by:** Bo Peng, Yu Zhang, Songlin Yang, Ruichong Zhang
- **Funded by:** RWKV Project (Under LF AI & Data Foundation)
- **Model type:** RWKV7
- **Language(s) (NLP):** English
- **License:** Apache-2.0
- **Parameter count:** 1.52B
- **Tokenizer:** RWKV World tokenizer
- **Vocabulary size:** 65,536
### Model Sources
<!-- Provide the basic links for the model. -->
- **Repository:** https://github.com/fla-org/flash-linear-attention ; https://github.com/BlinkDL/RWKV-LM
- **Paper:** With in Progress
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
Install `flash-linear-attention` and the latest version of `transformers` before using this model:
```bash
pip install git+https://github.com/fla-org/flash-linear-attention
pip install 'transformers>=4.48.0'
```
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
You can use this model just as any other HuggingFace models:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained('fla-hub/rwkv7-1.5B-world', trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained('fla-hub/rwkv7-1.5B-world', trust_remote_code=True)
model = model.cuda()
prompt = "What is a large language model?"
messages = [
{"role": "user", "content": "Who are you?"},
{"role": "assistant", "content": "I am a GPT-3 based model."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=1024,
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=False)[0]
print(response)
```
## Training Details
### Training Data
This model is trained on the World v3 with a total of 3.119 trillion tokens.
#### Training Hyperparameters
- **Training regime:** bfloat16, lr 4e-4 to 1e-5 "delayed" cosine decay, wd 0.1 (with increasing batch sizes during the middle)
- **Final Loss:** 1.9965
- **Token Count:** 3.119 trillion
## Evaluation
#### Metrics
`lambada_openai`:
before conversion: ppl 4.13 acc 69.4%
after conversion: ppl 4.26 acc 68.8% (without apply temple)
## FAQ
Q: safetensors metadata is none.
A: upgrade transformers to >=4.48.0: `pip install 'transformers>=4.48.0'` |
Mungert/Qwen2.5-14B-Instruct-1M-GGUF | Mungert | 2025-06-15T19:40:33Z | 2,128 | 4 | transformers | [
"transformers",
"gguf",
"chat",
"text-generation",
"en",
"arxiv:2501.15383",
"base_model:Qwen/Qwen2.5-14B",
"base_model:quantized:Qwen/Qwen2.5-14B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-18T04:16:59Z | ---
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen2.5-14B-Instruct-1M/blob/main/LICENSE
language:
- en
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-14B
tags:
- chat
library_name: transformers
---
# <span style="color: #7FFF7F;">Qwen2.5-14B-Instruct-1M GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen2.5-14B-Instruct-1M-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen2.5-14B-Instruct-1M-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen2.5-14B-Instruct-1M-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen2.5-14B-Instruct-1M-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen2.5-14B-Instruct-1M-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen2.5-14B-Instruct-1M-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen2.5-14B-Instruct-1M-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen2.5-14B-Instruct-1M-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen2.5-14B-Instruct-1M-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen2.5-14B-Instruct-1M-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen2.5-14B-Instruct-1M-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen2.5-14B-Instruct-1M
<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Introduction
Qwen2.5-1M is the long-context version of the Qwen2.5 series models, supporting a context length of up to 1M tokens. Compared to the Qwen2.5 128K version, Qwen2.5-1M demonstrates significantly improved performance in handling long-context tasks while maintaining its capability in short tasks.
The model has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Architecture: transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias
- Number of Parameters: 14.7B
- Number of Paramaters (Non-Embedding): 13.1B
- Number of Layers: 48
- Number of Attention Heads (GQA): 40 for Q and 8 for KV
- Context Length: Full 1,010,000 tokens and generation 8192 tokens
- We recommend deploying with our custom vLLM, which introduces sparse attention and length extrapolation methods to ensure efficiency and accuracy for long-context tasks. For specific guidance, refer to [this section](#processing-ultra-long-texts).
- You can also use the previous framework that supports Qwen2.5 for inference, but accuracy degradation may occur for sequences exceeding 262,144 tokens.
For more details, please refer to our [blog](https://qwenlm.github.io/blog/qwen2.5-1m/), [GitHub](https://github.com/QwenLM/Qwen2.5), [Technical Report](https://huggingface.co/papers/2501.15383), and [Documentation](https://qwen.readthedocs.io/en/latest/).
## Requirements
The code of Qwen2.5 has been in the latest Hugging face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.37.0`, you will encounter the following error:
```
KeyError: 'qwen2'
```
## Quickstart
Here provides a code snippet with `apply_chat_template` to show you how to load the tokenizer and model and how to generate contents.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen2.5-14B-Instruct-1M"
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
```
### Processing Ultra Long Texts
To enhance processing accuracy and efficiency for long sequences, we have developed an advanced inference framework based on vLLM, incorporating sparse attention and length extrapolation. This approach significantly improves model generation performance for sequences exceeding 256K tokens and achieves a 3 to 7 times speedup for sequences up to 1M tokens.
Here we provide step-by-step instructions for deploying the Qwen2.5-1M models with our framework.
#### 1. System Preparation
To achieve the best performance, we recommend using GPUs with Ampere or Hopper architecture, which support optimized kernels.
Ensure your system meets the following requirements:
- **CUDA Version**: 12.1 or 12.3
- **Python Version**: >=3.9 and <=3.12
**VRAM Requirements:**
- For processing 1 million-token sequences:
- **Qwen2.5-7B-Instruct-1M**: At least 120GB VRAM (total across GPUs).
- **Qwen2.5-14B-Instruct-1M**: At least 320GB VRAM (total across GPUs).
If your GPUs do not have sufficient VRAM, you can still use Qwen2.5-1M for shorter tasks.
#### 2. Install Dependencies
For now, you need to clone the vLLM repository from our custom branch and install it manually. We are working on getting our branch merged into the main vLLM project.
```bash
git clone -b dev/dual-chunk-attn [email protected]:QwenLM/vllm.git
cd vllm
pip install -e . -v
```
#### 3. Launch vLLM
vLLM supports offline inference or launch an openai-like server.
**Example of Offline Inference**
```python
from transformers import AutoTokenizer
from vllm import LLM, SamplingParams
# Initialize the tokenizer
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-14B-Instruct-1M")
# Pass the default decoding hyperparameters of Qwen2.5-14B-Instruct
# max_tokens is for the maximum length for generation.
sampling_params = SamplingParams(temperature=0.7, top_p=0.8, repetition_penalty=1.05, max_tokens=512)
# Input the model name or path. See below for parameter explanation (after the example of openai-like server).
llm = LLM(model="Qwen/Qwen2.5-14B-Instruct-1M",
tensor_parallel_size=4,
max_model_len=1010000,
enable_chunked_prefill=True,
max_num_batched_tokens=131072,
enforce_eager=True,
# quantization="fp8", # Enabling FP8 quantization for model weights can reduce memory usage.
)
# Prepare your prompts
prompt = "Tell me something about large language models."
messages = [
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
# generate outputs
outputs = llm.generate([text], sampling_params)
# Print the outputs.
for output in outputs:
prompt = output.prompt
generated_text = output.outputs[0].text
print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
```
**Example of Openai-like Server**
```bash
vllm serve Qwen/Qwen2.5-14B-Instruct-1M \
--tensor-parallel-size 4 \
--max-model-len 1010000 \
--enable-chunked-prefill --max-num-batched-tokens 131072 \
--enforce-eager \
--max-num-seqs 1
# --quantization fp8 # Enabling FP8 quantization for model weights can reduce memory usage.
```
Then you can use curl or python to interact with the deployed model.
**Parameter Explanations:**
- **`--tensor-parallel-size`**
- Set to the number of GPUs you are using. Max 4 GPUs for the 7B model, and 8 GPUs for the 14B model.
- **`--max-model-len`**
- Defines the maximum input sequence length. Reduce this value if you encounter Out of Memory issues.
- **`--max-num-batched-tokens`**
- Sets the chunk size in Chunked Prefill. A smaller value reduces activation memory usage but may slow down inference.
- Recommend 131072 for optimal performance.
- **`--max-num-seqs`**
- Limits concurrent sequences processed.
You can also refer to our [Documentation](https://qwen.readthedocs.io/en/latest/deployment/vllm.html) for usage of vLLM.
#### Troubleshooting:
1. Encountering the error: "The model's max sequence length (xxxxx) is larger than the maximum number of tokens that can be stored in the KV cache."
The VRAM reserved for the KV cache is insufficient. Consider reducing the ``max_model_len`` or increasing the ``tensor_parallel_size``. Alternatively, you can reduce ``max_num_batched_tokens``, although this may significantly slow down inference.
2. Encountering the error: "torch.OutOfMemoryError: CUDA out of memory."
The VRAM reserved for activation weights is insufficient. You can try setting ``gpu_memory_utilization`` to 0.85 or lower, but be aware that this might reduce the VRAM available for the KV cache.
3. Encountering the error: "Input prompt (xxxxx tokens) + lookahead slots (0) is too long and exceeds the capacity of the block manager."
The input is too lengthy. Consider using a shorter sequence or increasing the ``max_model_len``.
## Evaluation & Performance
Detailed evaluation results are reported in this [📑 blog](https://qwenlm.github.io/blog/qwen2.5-1m/) and our [technical report](https://arxiv.org/abs/2501.15383).
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen2.5-1m,
title = {Qwen2.5-1M: Deploy Your Own Qwen with Context Length up to 1M Tokens},
url = {https://qwenlm.github.io/blog/qwen2.5-1m/},
author = {Qwen Team},
month = {January},
year = {2025}
}
@article{qwen2.5,
title={Qwen2.5-1M Technical Report},
author={An Yang and Bowen Yu and Chengyuan Li and Dayiheng Liu and Fei Huang and Haoyan Huang and Jiandong Jiang and Jianhong Tu and Jianwei Zhang and Jingren Zhou and Junyang Lin and Kai Dang and Kexin Yang and Le Yu and Mei Li and Minmin Sun and Qin Zhu and Rui Men and Tao He and Weijia Xu and Wenbiao Yin and Wenyuan Yu and Xiafei Qiu and Xingzhang Ren and Xinlong Yang and Yong Li and Zhiying Xu and Zipeng Zhang},
journal={arXiv preprint arXiv:2501.15383},
year={2025}
}
```
|
Mungert/TriLM_3.9B_Unpacked-GGUF | Mungert | 2025-06-15T19:40:26Z | 276 | 3 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix"
] | null | 2025-03-17T11:16:59Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">TriLM_3.9B_Unpacked GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `TriLM_3.9B_Unpacked-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `TriLM_3.9B_Unpacked-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `TriLM_3.9B_Unpacked-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `TriLM_3.9B_Unpacked-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `TriLM_3.9B_Unpacked-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `TriLM_3.9B_Unpacked-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `TriLM_3.9B_Unpacked-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `TriLM_3.9B_Unpacked-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `TriLM_3.9B_Unpacked-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `TriLM_3.9B_Unpacked-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `TriLM_3.9B_Unpacked-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# TriLM 3.9B Unpacked
TriLM (ternary model), unpacked to FP16 format - compatible with FP16 GEMMs. After unpacking, TriLM has the same architecture as LLaMa.
```python
import transformers as tf, torch
model_name = "SpectraSuite/TriLM_3.9B_Unpacked"
# Please adjust the temperature, repetition penalty, top_k, top_p and other sampling parameters according to your needs.
pipeline = tf.pipeline("text-generation", model=model_id, model_kwargs={"torch_dtype": torch.float16}, device_map="auto")
# These are base (pretrained) LLMs that are not instruction and chat tuned. You may need to adjust your prompt accordingly.
pipeline("Once upon a time")
```
* License: Apache 2.0
* We will use our GitHub repo for communication (including HF repo related queries). Feel free to open an issue here https://github.com/NolanoOrg/SpectraSuite
|
Mungert/TriLM_2.4B_Unpacked-GGUF | Mungert | 2025-06-15T19:40:23Z | 276 | 3 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix"
] | null | 2025-03-17T06:53:19Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">TriLM_2.4B_Unpacked GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `TriLM_2.4B_Unpacked-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `TriLM_2.4B_Unpacked-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `TriLM_2.4B_Unpacked-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `TriLM_2.4B_Unpacked-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `TriLM_2.4B_Unpacked-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `TriLM_2.4B_Unpacked-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `TriLM_2.4B_Unpacked-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `TriLM_2.4B_Unpacked-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `TriLM_2.4B_Unpacked-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `TriLM_2.4B_Unpacked-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `TriLM_2.4B_Unpacked-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# TriLM 2.4B Unpacked
TriLM (ternary model), unpacked to FP16 format - compatible with FP16 GEMMs. After unpacking, TriLM has the same architecture as LLaMa.
```python
import transformers as tf, torch
model_name = "SpectraSuite/TriLM_2.4B_Unpacked"
# Please adjust the temperature, repetition penalty, top_k, top_p and other sampling parameters according to your needs.
pipeline = tf.pipeline("text-generation", model=model_id, model_kwargs={"torch_dtype": torch.float16}, device_map="auto")
# These are base (pretrained) LLMs that are not instruction and chat tuned. You may need to adjust your prompt accordingly.
pipeline("Once upon a time")
```
* License: Apache 2.0
* We will use our GitHub repo for communication (including HF repo related queries). Feel free to open an issue here https://github.com/NolanoOrg/SpectraSuite
|
Mungert/TriLM_1.5B_Unpacked-GGUF | Mungert | 2025-06-15T19:40:20Z | 186 | 0 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix"
] | null | 2025-03-17T04:36:39Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">TriLM_1.5B_Unpacked GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `TriLM_1.5B_Unpacked-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `TriLM_1.5B_Unpacked-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `TriLM_1.5B_Unpacked-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `TriLM_1.5B_Unpacked-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `TriLM_1.5B_Unpacked-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `TriLM_1.5B_Unpacked-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `TriLM_1.5B_Unpacked-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `TriLM_1.5B_Unpacked-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `TriLM_1.5B_Unpacked-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `TriLM_1.5B_Unpacked-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `TriLM_1.5B_Unpacked-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# TriLM 1.5B Unpacked
TriLM (ternary model), unpacked to FP16 format - compatible with FP16 GEMMs. After unpacking, TriLM has the same architecture as LLaMa.
```python
import transformers as tf, torch
model_name = "SpectraSuite/TriLM_1.5B_Unpacked"
# Please adjust the temperature, repetition penalty, top_k, top_p and other sampling parameters according to your needs.
pipeline = tf.pipeline("text-generation", model=model_id, model_kwargs={"torch_dtype": torch.float16}, device_map="auto")
# These are base (pretrained) LLMs that are not instruction and chat tuned. You may need to adjust your prompt accordingly.
pipeline("Once upon a time")
```
* License: Apache 2.0
* We will use our GitHub repo for communication (including HF repo related queries). Feel free to open an issue here https://github.com/NolanoOrg/SpectraSuite
|
Mungert/TriLM_390M_Unpacked-GGUF | Mungert | 2025-06-15T19:40:05Z | 237 | 0 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix"
] | null | 2025-03-16T23:15:08Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">TriLM_390M_Unpacked GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `TriLM_390M_Unpacked-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `TriLM_390M_Unpacked-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `TriLM_390M_Unpacked-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `TriLM_390M_Unpacked-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `TriLM_390M_Unpacked-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `TriLM_390M_Unpacked-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `TriLM_390M_Unpacked-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `TriLM_390M_Unpacked-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `TriLM_390M_Unpacked-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `TriLM_390M_Unpacked-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `TriLM_390M_Unpacked-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# TriLM 390M Unpacked
TriLM (ternary model), unpacked to FP16 format - compatible with FP16 GEMMs. After unpacking, TriLM has the same architecture as LLaMa.
```python
import transformers as tf, torch
model_name = "SpectraSuite/TriLM_390M_Unpacked"
# Please adjust the temperature, repetition penalty, top_k, top_p and other sampling parameters according to your needs.
pipeline = tf.pipeline("text-generation", model=model_id, model_kwargs={"torch_dtype": torch.float16}, device_map="auto")
# These are base (pretrained) LLMs that are not instruction and chat tuned. You may need to adjust your prompt accordingly.
pipeline("Once upon a time")
```
* License: Apache 2.0
* We will use our GitHub repo for communication (including HF repo related queries). Feel free to open an issue here https://github.com/NolanoOrg/SpectraSuite
|
Mungert/TriLM_190M_Unpacked-GGUF | Mungert | 2025-06-15T19:40:03Z | 264 | 1 | null | [
"gguf",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix"
] | null | 2025-03-16T22:34:44Z | ---
license: apache-2.0
---
# <span style="color: #7FFF7F;">TriLM_190M_Unpacked GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `TriLM_190M_Unpacked-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `TriLM_190M_Unpacked-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `TriLM_190M_Unpacked-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `TriLM_190M_Unpacked-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `TriLM_190M_Unpacked-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `TriLM_190M_Unpacked-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `TriLM_190M_Unpacked-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `TriLM_190M_Unpacked-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `TriLM_190M_Unpacked-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `TriLM_190M_Unpacked-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `TriLM_190M_Unpacked-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# TriLM 190M Unpacked
TriLM (ternary model), unpacked to FP16 format - compatible with FP16 GEMMs. After unpacking, TriLM has the same architecture as LLaMa.
```python
import transformers as tf, torch
model_name = "SpectraSuite/TriLM_190M_Unpacked"
# Please adjust the temperature, repetition penalty, top_k, top_p and other sampling parameters according to your needs.
pipeline = tf.pipeline("text-generation", model=model_id, model_kwargs={"torch_dtype": torch.float16}, device_map="auto")
# These are base (pretrained) LLMs that are not instruction and chat tuned. You may need to adjust your prompt accordingly.
pipeline("Once upon a time")
```
* License: Apache 2.0
* We will use our GitHub repo for communication (including HF repo related queries). Feel free to open an issue here https://github.com/NolanoOrg/SpectraSuite
|
Mungert/Mistral-7B-Instruct-v0.1-GGUF | Mungert | 2025-06-15T19:39:56Z | 1,003 | 3 | null | [
"gguf",
"finetuned",
"text-generation",
"arxiv:2310.06825",
"base_model:mistralai/Mistral-7B-v0.1",
"base_model:quantized:mistralai/Mistral-7B-v0.1",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-16T05:47:42Z | ---
license: apache-2.0
tags:
- finetuned
base_model: mistralai/Mistral-7B-v0.1
pipeline_tag: text-generation
inference: true
widget:
- messages:
- role: user
content: What is your favorite condiment?
extra_gated_description: If you want to learn more about how we process your personal data, please read our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
---
# <span style="color: #7FFF7F;">Mistral-7B-Instruct-v0.1 GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Mistral-7B-Instruct-v0.1-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Mistral-7B-Instruct-v0.1-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Mistral-7B-Instruct-v0.1-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Mistral-7B-Instruct-v0.1-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Mistral-7B-Instruct-v0.1-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Mistral-7B-Instruct-v0.1-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Mistral-7B-Instruct-v0.1-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Mistral-7B-Instruct-v0.1-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Mistral-7B-Instruct-v0.1-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Mistral-7B-Instruct-v0.1-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Mistral-7B-Instruct-v0.1-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Model Card for Mistral-7B-Instruct-v0.1
## Encode and Decode with `mistral_common`
```py
from mistral_common.tokens.tokenizers.mistral import MistralTokenizer
from mistral_common.protocol.instruct.messages import UserMessage
from mistral_common.protocol.instruct.request import ChatCompletionRequest
mistral_models_path = "MISTRAL_MODELS_PATH"
tokenizer = MistralTokenizer.v1()
completion_request = ChatCompletionRequest(messages=[UserMessage(content="Explain Machine Learning to me in a nutshell.")])
tokens = tokenizer.encode_chat_completion(completion_request).tokens
```
## Inference with `mistral_inference`
```py
from mistral_inference.transformer import Transformer
from mistral_inference.generate import generate
model = Transformer.from_folder(mistral_models_path)
out_tokens, _ = generate([tokens], model, max_tokens=64, temperature=0.0, eos_id=tokenizer.instruct_tokenizer.tokenizer.eos_id)
result = tokenizer.decode(out_tokens[0])
print(result)
```
## Inference with hugging face `transformers`
```py
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
model.to("cuda")
generated_ids = model.generate(tokens, max_new_tokens=1000, do_sample=True)
# decode with mistral tokenizer
result = tokenizer.decode(generated_ids[0].tolist())
print(result)
```
> [!TIP]
> PRs to correct the `transformers` tokenizer so that it gives 1-to-1 the same results as the `mistral_common` reference implementation are very welcome!
---
The Mistral-7B-Instruct-v0.1 Large Language Model (LLM) is a instruct fine-tuned version of the [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) generative text model using a variety of publicly available conversation datasets.
For full details of this model please read our [paper](https://arxiv.org/abs/2310.06825) and [release blog post](https://mistral.ai/news/announcing-mistral-7b/).
## Instruction format
In order to leverage instruction fine-tuning, your prompt should be surrounded by `[INST]` and `[/INST]` tokens. The very first instruction should begin with a begin of sentence id. The next instructions should not. The assistant generation will be ended by the end-of-sentence token id.
E.g.
```
text = "<s>[INST] What is your favourite condiment? [/INST]"
"Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!</s> "
"[INST] Do you have mayonnaise recipes? [/INST]"
```
This format is available as a [chat template](https://huggingface.co/docs/transformers/main/chat_templating) via the `apply_chat_template()` method:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
messages = [
{"role": "user", "content": "What is your favourite condiment?"},
{"role": "assistant", "content": "Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!"},
{"role": "user", "content": "Do you have mayonnaise recipes?"}
]
encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
model_inputs = encodeds.to(device)
model.to(device)
generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)
decoded = tokenizer.batch_decode(generated_ids)
print(decoded[0])
```
## Model Architecture
This instruction model is based on Mistral-7B-v0.1, a transformer model with the following architecture choices:
- Grouped-Query Attention
- Sliding-Window Attention
- Byte-fallback BPE tokenizer
## Troubleshooting
- If you see the following error:
```
Traceback (most recent call last):
File "", line 1, in
File "/transformers/models/auto/auto_factory.py", line 482, in from_pretrained
config, kwargs = AutoConfig.from_pretrained(
File "/transformers/models/auto/configuration_auto.py", line 1022, in from_pretrained
config_class = CONFIG_MAPPING[config_dict["model_type"]]
File "/transformers/models/auto/configuration_auto.py", line 723, in getitem
raise KeyError(key)
KeyError: 'mistral'
```
Installing transformers from source should solve the issue
pip install git+https://github.com/huggingface/transformers
This should not be required after transformers-v4.33.4.
## Limitations
The Mistral 7B Instruct model is a quick demonstration that the base model can be easily fine-tuned to achieve compelling performance.
It does not have any moderation mechanisms. We're looking forward to engaging with the community on ways to
make the model finely respect guardrails, allowing for deployment in environments requiring moderated outputs.
## The Mistral AI Team
Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed. |
Mungert/OlympicCoder-7B-GGUF | Mungert | 2025-06-15T19:39:53Z | 296 | 4 | transformers | [
"transformers",
"gguf",
"text-generation",
"en",
"dataset:open-r1/codeforces-cots",
"base_model:Qwen/Qwen2.5-Coder-7B-Instruct",
"base_model:quantized:Qwen/Qwen2.5-Coder-7B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-16T02:40:47Z | ---
license: apache-2.0
datasets:
- open-r1/codeforces-cots
language:
- en
base_model:
- Qwen/Qwen2.5-Coder-7B-Instruct
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">OlympicCoder-7B GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `OlympicCoder-7B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `OlympicCoder-7B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `OlympicCoder-7B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `OlympicCoder-7B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `OlympicCoder-7B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `OlympicCoder-7B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `OlympicCoder-7B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `OlympicCoder-7B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `OlympicCoder-7B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `OlympicCoder-7B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `OlympicCoder-7B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Model Card for OlympicCoder-7B
OlympicCoder-7B is a code model that achieves strong performance on competitive coding benchmarks such as LiveCodeBench and the 2024 International Olympiad in Informatics.
* Repository: https://github.com/huggingface/open-r1
* Blog post: https://huggingface.co/blog/open-r1/update-3
## Model description
- **Model type:** A 7B parameter model fine-tuned on a decontaminated version of the codeforces dataset.
- **Language(s) (NLP):** Primarily English
- **License:** apache-2.0
- **Finetuned from model:** [Qwen/Qwen2.5-Coder-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct)
## Evaluation
We compare the performance of OlympicCoder models on two main benchmarks for competitive coding:
* **[IOI'2024:](https://github.com/huggingface/ioi)** 6 very challenging problems from the 2024 International Olympiad in Informatics. Models are allowed up to 50 submissions per problem.
* **[LiveCodeBench:](https://livecodebench.github.io)** Python programming problems source from platforms like CodeForces and LeetCoder. We use the `v4_v5` subset of [`livecodebench/code_generation_lite`](https://huggingface.co/datasets/livecodebench/code_generation_lite), which corresponds to 268 problems. We use `lighteval` to evaluate models on LiveCodeBench using the sampling parameters described [here](https://github.com/huggingface/open-r1?tab=readme-ov-file#livecodebench).
> [!NOTE]
> The OlympicCoder models were post-trained exclusively on C++ solutions generated by DeepSeek-R1. As a result the performance on LiveCodeBench should be considered to be partially _out-of-domain_, since this expects models to output solutions in Python.
### IOI'24
### LiveCodeBench
## Usage
Here's how you can run the model using the `pipeline()` function from 🤗 Transformers:
```python
# pip install transformers
# pip install accelerate
import torch
from transformers import pipeline
pipe = pipeline("text-generation", model="open-r1/OlympicCoder-7B", torch_dtype=torch.bfloat16, device_map="auto")
# We use the tokenizer's chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating
messages = [
{"role": "user", "content": "Write a python program to calculate the 10th Fibonacci number"},
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt, max_new_tokens=8000, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
#<|im_start|>user
#Write a python program to calculate the 10th fibonacci number<|im_end|>
#<|im_start|>assistant
#<think>Okay, I need to write a Python program that calculates the 10th Fibonacci number. Hmm, the Fibonacci sequence starts with 0 and 1. Each subsequent number is the sum of the two preceding ones. So the sequence goes: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, and so on. ...
```
> [!WARNING]
> To ensure that the model consistently outputs a long chain-of-thought, we have edited the chat template to prefill the first assistant turn with a `<think>` token. As a result, the outputs from this model will not show the opening `<think>` token if you use the model's `generate()` method. To apply reinforcement learning with a format reward, either prepend the `<think>` token to the model's completions or amend the chat template to remove the prefill.
## Training procedure
### Training hyper-parameters
The following hyperparameters were used during training:
- dataset: open-r1/codeforces-cots
- learning_rate: 4.0e-5
- train_batch_size: 2
- seed: 42
- packing: false
- distributed_type: deepspeed-zero-3
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine_with_min_lr
- min_lr_rate: 0.1
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 10.0 |
Mungert/phi-2-GGUF | Mungert | 2025-06-15T19:39:50Z | 1,961 | 2 | null | [
"gguf",
"nlp",
"code",
"text-generation",
"en",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix"
] | text-generation | 2025-03-15T21:54:52Z | ---
license: mit
license_link: https://huggingface.co/microsoft/phi-2/resolve/main/LICENSE
language:
- en
pipeline_tag: text-generation
tags:
- nlp
- code
---
# <span style="color: #7FFF7F;">phi-2 GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `phi-2-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `phi-2-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `phi-2-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `phi-2-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `phi-2-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `phi-2-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `phi-2-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `phi-2-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `phi-2-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `phi-2-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `phi-2-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
## Model Summary
Phi-2 is a Transformer with **2.7 billion** parameters. It was trained using the same data sources as [Phi-1.5](https://huggingface.co/microsoft/phi-1.5), augmented with a new data source that consists of various NLP synthetic texts and filtered websites (for safety and educational value). When assessed against benchmarks testing common sense, language understanding, and logical reasoning, Phi-2 showcased a nearly state-of-the-art performance among models with less than 13 billion parameters.
Our model hasn't been fine-tuned through reinforcement learning from human feedback. The intention behind crafting this open-source model is to provide the research community with a non-restricted small model to explore vital safety challenges, such as reducing toxicity, understanding societal biases, enhancing controllability, and more.
## How to Use
Phi-2 has been integrated in the `transformers` version 4.37.0, please ensure that you are using a version equal or higher than it.
Phi-2 is known for having an attention overflow issue (with FP16). If you are facing this issue, please enable/disable autocast on the [PhiAttention.forward()](https://github.com/huggingface/transformers/blob/main/src/transformers/models/phi/modeling_phi.py#L306) function.
## Intended Uses
Given the nature of the training data, the Phi-2 model is best suited for prompts using the QA format, the chat format, and the code format.
### QA Format:
You can provide the prompt as a standalone question as follows:
```markdown
Write a detailed analogy between mathematics and a lighthouse.
```
where the model generates the text after "." .
To encourage the model to write more concise answers, you can also try the following QA format using "Instruct: \<prompt\>\nOutput:"
```markdown
Instruct: Write a detailed analogy between mathematics and a lighthouse.
Output: Mathematics is like a lighthouse. Just as a lighthouse guides ships safely to shore, mathematics provides a guiding light in the world of numbers and logic. It helps us navigate through complex problems and find solutions. Just as a lighthouse emits a steady beam of light, mathematics provides a consistent framework for reasoning and problem-solving. It illuminates the path to understanding and helps us make sense of the world around us.
```
where the model generates the text after "Output:".
### Chat Format:
```markdown
Alice: I don't know why, I'm struggling to maintain focus while studying. Any suggestions?
Bob: Well, have you tried creating a study schedule and sticking to it?
Alice: Yes, I have, but it doesn't seem to help much.
Bob: Hmm, maybe you should try studying in a quiet environment, like the library.
Alice: ...
```
where the model generates the text after the first "Bob:".
### Code Format:
```python
def print_prime(n):
"""
Print all primes between 1 and n
"""
primes = []
for num in range(2, n+1):
is_prime = True
for i in range(2, int(math.sqrt(num))+1):
if num % i == 0:
is_prime = False
break
if is_prime:
primes.append(num)
print(primes)
```
where the model generates the text after the comments.
**Notes:**
* Phi-2 is intended for QA, chat, and code purposes. The model-generated text/code should be treated as a starting point rather than a definitive solution for potential use cases. Users should be cautious when employing these models in their applications.
* Direct adoption for production tasks without evaluation is out of scope of this project. As a result, the Phi-2 model has not been tested to ensure that it performs adequately for any production-level application. Please refer to the limitation sections of this document for more details.
* If you are using `transformers<4.37.0`, always load the model with `trust_remote_code=True` to prevent side-effects.
## Sample Code
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
torch.set_default_device("cuda")
model = AutoModelForCausalLM.from_pretrained("microsoft/phi-2", torch_dtype="auto", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-2", trust_remote_code=True)
inputs = tokenizer('''def print_prime(n):
"""
Print all primes between 1 and n
"""''', return_tensors="pt", return_attention_mask=False)
outputs = model.generate(**inputs, max_length=200)
text = tokenizer.batch_decode(outputs)[0]
print(text)
```
## Limitations of Phi-2
* Generate Inaccurate Code and Facts: The model may produce incorrect code snippets and statements. Users should treat these outputs as suggestions or starting points, not as definitive or accurate solutions.
* Limited Scope for code: Majority of Phi-2 training data is based in Python and use common packages such as "typing, math, random, collections, datetime, itertools". If the model generates Python scripts that utilize other packages or scripts in other languages, we strongly recommend users manually verify all API uses.
* Unreliable Responses to Instruction: The model has not undergone instruction fine-tuning. As a result, it may struggle or fail to adhere to intricate or nuanced instructions provided by users.
* Language Limitations: The model is primarily designed to understand standard English. Informal English, slang, or any other languages might pose challenges to its comprehension, leading to potential misinterpretations or errors in response.
* Potential Societal Biases: Phi-2 is not entirely free from societal biases despite efforts in assuring training data safety. There's a possibility it may generate content that mirrors these societal biases, particularly if prompted or instructed to do so. We urge users to be aware of this and to exercise caution and critical thinking when interpreting model outputs.
* Toxicity: Despite being trained with carefully selected data, the model can still produce harmful content if explicitly prompted or instructed to do so. We chose to release the model to help the open-source community develop the most effective ways to reduce the toxicity of a model directly after pretraining.
* Verbosity: Phi-2 being a base model often produces irrelevant or extra text and responses following its first answer to user prompts within a single turn. This is due to its training dataset being primarily textbooks, which results in textbook-like responses.
## Training
### Model
* Architecture: a Transformer-based model with next-word prediction objective
* Context length: 2048 tokens
* Dataset size: 250B tokens, combination of NLP synthetic data created by AOAI GPT-3.5 and filtered web data from Falcon RefinedWeb and SlimPajama, which was assessed by AOAI GPT-4.
* Training tokens: 1.4T tokens
* GPUs: 96xA100-80G
* Training time: 14 days
### Software
* [PyTorch](https://github.com/pytorch/pytorch)
* [DeepSpeed](https://github.com/microsoft/DeepSpeed)
* [Flash-Attention](https://github.com/HazyResearch/flash-attention)
### License
The model is licensed under the [MIT license](https://huggingface.co/microsoft/phi-2/resolve/main/LICENSE).
## Trademarks
This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow [Microsoft’s Trademark & Brand Guidelines](https://www.microsoft.com/en-us/legal/intellectualproperty/trademarks). Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party’s policies. |
Mungert/Llama-3.2-3B-Instruct-GGUF | Mungert | 2025-06-15T19:39:37Z | 1,580 | 4 | transformers | [
"transformers",
"gguf",
"facebook",
"meta",
"pytorch",
"llama",
"llama-3",
"text-generation",
"en",
"de",
"fr",
"it",
"pt",
"hi",
"es",
"th",
"arxiv:2204.05149",
"arxiv:2405.16406",
"license:llama3.2",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-15T06:31:50Z | ---
language:
- en
- de
- fr
- it
- pt
- hi
- es
- th
library_name: transformers
pipeline_tag: text-generation
tags:
- facebook
- meta
- pytorch
- llama
- llama-3
license: llama3.2
extra_gated_prompt: >-
### LLAMA 3.2 COMMUNITY LICENSE AGREEMENT
Llama 3.2 Version Release Date: September 25, 2024
“Agreement” means the terms and conditions for use, reproduction, distribution
and modification of the Llama Materials set forth herein.
“Documentation” means the specifications, manuals and documentation accompanying Llama 3.2
distributed by Meta at https://llama.meta.com/doc/overview.
“Licensee” or “you” means you, or your employer or any other person or entity (if you are
entering into this Agreement on such person or entity’s behalf), of the age required under
applicable laws, rules or regulations to provide legal consent and that has legal authority
to bind your employer or such other person or entity if you are entering in this Agreement
on their behalf.
“Llama 3.2” means the foundational large language models and software and algorithms, including
machine-learning model code, trained model weights, inference-enabling code, training-enabling code,
fine-tuning enabling code and other elements of the foregoing distributed by Meta at
https://www.llama.com/llama-downloads.
“Llama Materials” means, collectively, Meta’s proprietary Llama 3.2 and Documentation (and
any portion thereof) made available under this Agreement.
“Meta” or “we” means Meta Platforms Ireland Limited (if you are located in or,
if you are an entity, your principal place of business is in the EEA or Switzerland)
and Meta Platforms, Inc. (if you are located outside of the EEA or Switzerland).
By clicking “I Accept” below or by using or distributing any portion or element of the Llama Materials,
you agree to be bound by this Agreement.
1. License Rights and Redistribution.
a. Grant of Rights. You are granted a non-exclusive, worldwide,
non-transferable and royalty-free limited license under Meta’s intellectual property or other rights
owned by Meta embodied in the Llama Materials to use, reproduce, distribute, copy, create derivative works
of, and make modifications to the Llama Materials.
b. Redistribution and Use.
i. If you distribute or make available the Llama Materials (or any derivative works thereof),
or a product or service (including another AI model) that contains any of them, you shall (A) provide
a copy of this Agreement with any such Llama Materials; and (B) prominently display “Built with Llama”
on a related website, user interface, blogpost, about page, or product documentation. If you use the
Llama Materials or any outputs or results of the Llama Materials to create, train, fine tune, or
otherwise improve an AI model, which is distributed or made available, you shall also include “Llama”
at the beginning of any such AI model name.
ii. If you receive Llama Materials, or any derivative works thereof, from a Licensee as part
of an integrated end user product, then Section 2 of this Agreement will not apply to you.
iii. You must retain in all copies of the Llama Materials that you distribute the
following attribution notice within a “Notice” text file distributed as a part of such copies:
“Llama 3.2 is licensed under the Llama 3.2 Community License, Copyright © Meta Platforms,
Inc. All Rights Reserved.”
iv. Your use of the Llama Materials must comply with applicable laws and regulations
(including trade compliance laws and regulations) and adhere to the Acceptable Use Policy for
the Llama Materials (available at https://www.llama.com/llama3_2/use-policy), which is hereby
incorporated by reference into this Agreement.
2. Additional Commercial Terms. If, on the Llama 3.2 version release date, the monthly active users
of the products or services made available by or for Licensee, or Licensee’s affiliates,
is greater than 700 million monthly active users in the preceding calendar month, you must request
a license from Meta, which Meta may grant to you in its sole discretion, and you are not authorized to
exercise any of the rights under this Agreement unless or until Meta otherwise expressly grants you such rights.
3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA MATERIALS AND ANY OUTPUT AND
RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF ANY KIND, AND META DISCLAIMS
ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES
OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE
FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE LLAMA MATERIALS AND ASSUME ANY RISKS ASSOCIATED
WITH YOUR USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS.
4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY,
WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT,
FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN
IF META OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.
5. Intellectual Property.
a. No trademark licenses are granted under this Agreement, and in connection with the Llama Materials,
neither Meta nor Licensee may use any name or mark owned by or associated with the other or any of its affiliates,
except as required for reasonable and customary use in describing and redistributing the Llama Materials or as
set forth in this Section 5(a). Meta hereby grants you a license to use “Llama” (the “Mark”) solely as required
to comply with the last sentence of Section 1.b.i. You will comply with Meta’s brand guidelines (currently accessible
at https://about.meta.com/brand/resources/meta/company-brand/). All goodwill arising out of your use of the Mark
will inure to the benefit of Meta.
b. Subject to Meta’s ownership of Llama Materials and derivatives made by or for Meta, with respect to any
derivative works and modifications of the Llama Materials that are made by you, as between you and Meta,
you are and will be the owner of such derivative works and modifications.
c. If you institute litigation or other proceedings against Meta or any entity (including a cross-claim or
counterclaim in a lawsuit) alleging that the Llama Materials or Llama 3.2 outputs or results, or any portion
of any of the foregoing, constitutes infringement of intellectual property or other rights owned or licensable
by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or
claim is filed or instituted. You will indemnify and hold harmless Meta from and against any claim by any third
party arising out of or related to your use or distribution of the Llama Materials.
6. Term and Termination. The term of this Agreement will commence upon your acceptance of this Agreement or access
to the Llama Materials and will continue in full force and effect until terminated in accordance with the terms
and conditions herein. Meta may terminate this Agreement if you are in breach of any term or condition of this
Agreement. Upon termination of this Agreement, you shall delete and cease use of the Llama Materials. Sections 3,
4 and 7 shall survive the termination of this Agreement.
7. Governing Law and Jurisdiction. This Agreement will be governed and construed under the laws of the State of
California without regard to choice of law principles, and the UN Convention on Contracts for the International
Sale of Goods does not apply to this Agreement. The courts of California shall have exclusive jurisdiction of
any dispute arising out of this Agreement.
### Llama 3.2 Acceptable Use Policy
Meta is committed to promoting safe and fair use of its tools and features, including Llama 3.2.
If you access or use Llama 3.2, you agree to this Acceptable Use Policy (“**Policy**”).
The most recent copy of this policy can be found at
[https://www.llama.com/llama3_2/use-policy](https://www.llama.com/llama3_2/use-policy).
#### Prohibited Uses
We want everyone to use Llama 3.2 safely and responsibly. You agree you will not use, or allow others to use, Llama 3.2 to:
1. Violate the law or others’ rights, including to:
1. Engage in, promote, generate, contribute to, encourage, plan, incite, or further illegal or unlawful activity or content, such as:
1. Violence or terrorism
2. Exploitation or harm to children, including the solicitation, creation, acquisition, or dissemination of child exploitative content or failure to report Child Sexual Abuse Material
3. Human trafficking, exploitation, and sexual violence
4. The illegal distribution of information or materials to minors, including obscene materials, or failure to employ legally required age-gating in connection with such information or materials.
5. Sexual solicitation
6. Any other criminal activity
1. Engage in, promote, incite, or facilitate the harassment, abuse, threatening, or bullying of individuals or groups of individuals
2. Engage in, promote, incite, or facilitate discrimination or other unlawful or harmful conduct in the provision of employment, employment benefits, credit, housing, other economic benefits, or other essential goods and services
3. Engage in the unauthorized or unlicensed practice of any profession including, but not limited to, financial, legal, medical/health, or related professional practices
4. Collect, process, disclose, generate, or infer private or sensitive information about individuals, including information about individuals’ identity, health, or demographic information, unless you have obtained the right to do so in accordance with applicable law
5. Engage in or facilitate any action or generate any content that infringes, misappropriates, or otherwise violates any third-party rights, including the outputs or results of any products or services using the Llama Materials
6. Create, generate, or facilitate the creation of malicious code, malware, computer viruses or do anything else that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system
7. Engage in any action, or facilitate any action, to intentionally circumvent or remove usage restrictions or other safety measures, or to enable functionality disabled by Meta
2. Engage in, promote, incite, facilitate, or assist in the planning or development of activities that present a risk of death or bodily harm to individuals, including use of Llama 3.2 related to the following:
8. Military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulations (ITAR) maintained by the United States Department of State or to the U.S. Biological Weapons Anti-Terrorism Act of 1989 or the Chemical Weapons Convention Implementation Act of 1997
9. Guns and illegal weapons (including weapon development)
10. Illegal drugs and regulated/controlled substances
11. Operation of critical infrastructure, transportation technologies, or heavy machinery
12. Self-harm or harm to others, including suicide, cutting, and eating disorders
13. Any content intended to incite or promote violence, abuse, or any infliction of bodily harm to an individual
3. Intentionally deceive or mislead others, including use of Llama 3.2 related to the following:
14. Generating, promoting, or furthering fraud or the creation or promotion of disinformation
15. Generating, promoting, or furthering defamatory content, including the creation of defamatory statements, images, or other content
16. Generating, promoting, or further distributing spam
17. Impersonating another individual without consent, authorization, or legal right
18. Representing that the use of Llama 3.2 or outputs are human-generated
19. Generating or facilitating false online engagement, including fake reviews and other means of fake online engagement
4. Fail to appropriately disclose to end users any known dangers of your AI system
5. Interact with third party tools, models, or software designed to generate unlawful content or engage in unlawful or harmful conduct and/or represent that the outputs of such tools, models, or software are associated with Meta or Llama 3.2
With respect to any multimodal models included in Llama 3.2, the rights granted under Section 1(a) of the Llama 3.2 Community License Agreement are not being granted to you if you are an individual domiciled in, or a company with a principal place of business in, the European Union. This restriction does not apply to end users of a product or service that incorporates any such multimodal models.
Please report any violation of this Policy, software “bug,” or other problems that could lead to a violation of this Policy through one of the following means:
* Reporting issues with the model: [https://github.com/meta-llama/llama-models/issues](https://l.workplace.com/l.php?u=https%3A%2F%2Fgithub.com%2Fmeta-llama%2Fllama-models%2Fissues&h=AT0qV8W9BFT6NwihiOHRuKYQM_UnkzN_NmHMy91OT55gkLpgi4kQupHUl0ssR4dQsIQ8n3tfd0vtkobvsEvt1l4Ic6GXI2EeuHV8N08OG2WnbAmm0FL4ObkazC6G_256vN0lN9DsykCvCqGZ)
* Reporting risky content generated by the model: [developers.facebook.com/llama_output_feedback](http://developers.facebook.com/llama_output_feedback)
* Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info)
* Reporting violations of the Acceptable Use Policy or unlicensed uses of Llama 3.2: [email protected]
extra_gated_fields:
First Name: text
Last Name: text
Date of birth: date_picker
Country: country
Affiliation: text
Job title:
type: select
options:
- Student
- Research Graduate
- AI researcher
- AI developer/engineer
- Reporter
- Other
geo: ip_location
By clicking Submit below I accept the terms of the license and acknowledge that the information I provide will be collected stored processed and shared in accordance with the Meta Privacy Policy: checkbox
extra_gated_description: >-
The information you provide will be collected, stored, processed and shared in
accordance with the [Meta Privacy
Policy](https://www.facebook.com/privacy/policy/).
extra_gated_button_content: Submit
---
# <span style="color: #7FFF7F;">Llama-3.2-3B-Instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.2-3B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.2-3B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.2-3B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.2-3B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.2-3B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.2-3B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.2-3B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.2-3B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.2-3B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.2-3B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.2-3B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
## Model Information
The Llama 3.2 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction-tuned generative models in 1B and 3B sizes (text in/text out). The Llama 3.2 instruction-tuned text only models are optimized for multilingual dialogue use cases, including agentic retrieval and summarization tasks. They outperform many of the available open source and closed chat models on common industry benchmarks.
**Model Developer:** Meta
**Model Architecture:** Llama 3.2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.
| | Training Data | Params | Input modalities | Output modalities | Context Length | GQA | Shared Embeddings | Token count | Knowledge cutoff |
| :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- |
| Llama 3.2 (text only) | A new mix of publicly available online data. | 1B (1.23B) | Multilingual Text | Multilingual Text and code | 128k | Yes | Yes | Up to 9T tokens | December 2023 |
| | | 3B (3.21B) | Multilingual Text | Multilingual Text and code | | | | | |
| Llama 3.2 Quantized (text only) | A new mix of publicly available online data. | 1B (1.23B) | Multilingual Text | Multilingual Text and code | 8k | Yes | Yes | Up to 9T tokens | December 2023 |
| | | 3B (3.21B) | Multilingual Text | Multilingual Text and code | | | | | |
**Supported Languages:** English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai are officially supported. Llama 3.2 has been trained on a broader collection of languages than these 8 supported languages. Developers may fine-tune Llama 3.2 models for languages beyond these supported languages, provided they comply with the Llama 3.2 Community License and the Acceptable Use Policy. Developers are always expected to ensure that their deployments, including those that involve additional languages, are completed safely and responsibly.
**Llama 3.2 Model Family:** Token counts refer to pretraining data only. All model versions use Grouped-Query Attention (GQA) for improved inference scalability.
**Model Release Date:** Sept 25, 2024
**Status:** This is a static model trained on an offline dataset. Future versions may be released that improve model capabilities and safety.
**License:** Use of Llama 3.2 is governed by the [Llama 3.2 Community License](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/LICENSE) (a custom, commercial license agreement).
**Feedback:** Instructions on how to provide feedback or comments on the model can be found in the Llama Models [README](https://github.com/meta-llama/llama-models/blob/main/README.md). For more technical information about generation parameters and recipes for how to use Llama 3.2 in applications, please go [here](https://github.com/meta-llama/llama-recipes).
## Intended Use
**Intended Use Cases:** Llama 3.2 is intended for commercial and research use in multiple languages. Instruction tuned text only models are intended for assistant-like chat and agentic applications like knowledge retrieval and summarization, mobile AI powered writing assistants and query and prompt rewriting. Pretrained models can be adapted for a variety of additional natural language generation tasks. Similarly, quantized models can be adapted for a variety of on-device use-cases with limited compute resources.
**Out of Scope:** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in any other way that is prohibited by the Acceptable Use Policy and Llama 3.2 Community License. Use in languages beyond those explicitly referenced as supported in this model card.
## How to use
This repository contains two versions of Llama-3.2-3B-Instruct, for use with `transformers` and with the original `llama` codebase.
### Use with transformers
Starting with `transformers >= 4.43.0` onward, you can run conversational inference using the Transformers `pipeline` abstraction or by leveraging the Auto classes with the `generate()` function.
Make sure to update your transformers installation via `pip install --upgrade transformers`.
```python
import torch
from transformers import pipeline
model_id = "meta-llama/Llama-3.2-3B-Instruct"
pipe = pipeline(
"text-generation",
model=model_id,
torch_dtype=torch.bfloat16,
device_map="auto",
)
messages = [
{"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
{"role": "user", "content": "Who are you?"},
]
outputs = pipe(
messages,
max_new_tokens=256,
)
print(outputs[0]["generated_text"][-1])
```
Note: You can also find detailed recipes on how to use the model locally, with `torch.compile()`, assisted generations, quantised and more at [`huggingface-llama-recipes`](https://github.com/huggingface/huggingface-llama-recipes)
### Use with `llama`
Please, follow the instructions in the [repository](https://github.com/meta-llama/llama)
To download Original checkpoints, see the example command below leveraging `huggingface-cli`:
```
huggingface-cli download meta-llama/Llama-3.2-3B-Instruct --include "original/*" --local-dir Llama-3.2-3B-Instruct
```
## Hardware and Software
**Training Factors:** We used custom training libraries, Meta's custom built GPU cluster, and production infrastructure for pretraining. Fine-tuning, quantization, annotation, and evaluation were also performed on production infrastructure.
**Training Energy Use:** Training utilized a cumulative of **916k** GPU hours of computation on H100-80GB (TDP of 700W) type hardware, per the table below. Training time is the total GPU time required for training each model and power consumption is the peak power capacity per GPU device used, adjusted for power usage efficiency.
**Training Greenhouse Gas Emissions:** Estimated total location-based greenhouse gas emissions were **240** tons CO2eq for training. Since 2020, Meta has maintained net zero greenhouse gas emissions in its global operations and matched 100% of its electricity use with renewable energy; therefore, the total market-based greenhouse gas emissions for training were 0 tons CO2eq.
| | Training Time (GPU hours) | Logit Generation Time (GPU Hours) | Training Power Consumption (W) | Training Location-Based Greenhouse Gas Emissions (tons CO2eq) | Training Market-Based Greenhouse Gas Emissions (tons CO2eq) |
| :---- | :---: | ----- | :---: | :---: | :---: |
| Llama 3.2 1B | 370k | \- | 700 | 107 | 0 |
| Llama 3.2 3B | 460k | \- | 700 | 133 | 0 |
| Llama 3.2 1B SpinQuant | 1.7 | 0 | 700 | *Negligible*\*\* | 0 |
| Llama 3.2 3B SpinQuant | 2.4 | 0 | 700 | *Negligible*\*\* | 0 |
| Llama 3.2 1B QLora | 1.3k | 0 | 700 | 0.381 | 0 |
| Llama 3.2 3B QLora | 1.6k | 0 | 700 | 0.461 | 0 |
| Total | 833k | 86k | | 240 | 0 |
\*\* The location-based CO2e emissions of Llama 3.2 1B SpinQuant and Llama 3.2 3B SpinQuant are less than 0.001 metric tonnes each. This is due to the minimal training GPU hours that are required.
The methodology used to determine training energy use and greenhouse gas emissions can be found [here](https://arxiv.org/pdf/2204.05149). Since Meta is openly releasing these models, the training energy use and greenhouse gas emissions will not be incurred by others.
## Training Data
**Overview:** Llama 3.2 was pretrained on up to 9 trillion tokens of data from publicly available sources. For the 1B and 3B Llama 3.2 models, we incorporated logits from the Llama 3.1 8B and 70B models into the pretraining stage of the model development, where outputs (logits) from these larger models were used as token-level targets. Knowledge distillation was used after pruning to recover performance. In post-training we used a similar recipe as Llama 3.1 and produced final chat models by doing several rounds of alignment on top of the pre-trained model. Each round involved Supervised Fine-Tuning (SFT), Rejection Sampling (RS), and Direct Preference Optimization (DPO).
**Data Freshness:** The pretraining data has a cutoff of December 2023\.
## Quantization
### Quantization Scheme
We designed the current quantization scheme with the [PyTorch’s ExecuTorch](https://github.com/pytorch/executorch) inference framework and Arm CPU backend in mind, taking into account metrics including model quality, prefill/decoding speed, and memory footprint. Our quantization scheme involves three parts:
- All linear layers in all transformer blocks are quantized to a 4-bit groupwise scheme (with a group size of 32) for weights and 8-bit per-token dynamic quantization for activations.
- The classification layer is quantized to 8-bit per-channel for weight and 8-bit per token dynamic quantization for activation.
- Similar to classification layer, an 8-bit per channel quantization is used for embedding layer.
### Quantization-Aware Training and LoRA
The quantization-aware training (QAT) with low-rank adaptation (LoRA) models went through only post-training stages, using the same data as the full precision models. To initialize QAT, we utilize BF16 Llama 3.2 model checkpoints obtained after supervised fine-tuning (SFT) and perform an additional full round of SFT training with QAT. We then freeze the backbone of the QAT model and perform another round of SFT with LoRA adaptors applied to all layers within the transformer block. Meanwhile, the LoRA adaptors' weights and activations are maintained in BF16. Because our approach is similar to QLoRA of Dettmers et al., (2023) (i.e., quantization followed by LoRA adapters), we refer this method as QLoRA. Finally, we fine-tune the resulting model (both backbone and LoRA adaptors) using direct preference optimization (DPO).
### SpinQuant
[SpinQuant](https://arxiv.org/abs/2405.16406) was applied, together with generative post-training quantization (GPTQ). For the SpinQuant rotation matrix fine-tuning, we optimized for 100 iterations, using 800 samples with sequence-length 2048 from the WikiText 2 dataset. For GPTQ, we used 128 samples from the same dataset with the same sequence-length.
## Benchmarks \- English Text
In this section, we report the results for Llama 3.2 models on standard automatic benchmarks. For all these evaluations, we used our internal evaluations library.
### Base Pretrained Models
| Category | Benchmark | \# Shots | Metric | Llama 3.2 1B | Llama 3.2 3B | Llama 3.1 8B |
| ----- | ----- | :---: | :---: | :---: | :---: | :---: |
| General | MMLU | 5 | macro\_avg/acc\_char | 32.2 | 58 | 66.7 |
| | AGIEval English | 3-5 | average/acc\_char | 23.3 | 39.2 | 47.8 |
| | ARC-Challenge | 25 | acc\_char | 32.8 | 69.1 | 79.7 |
| Reading comprehension | SQuAD | 1 | em | 49.2 | 67.7 | 77 |
| | QuAC (F1) | 1 | f1 | 37.9 | 42.9 | 44.9 |
| | DROP (F1) | 3 | f1 | 28.0 | 45.2 | 59.5 |
| Long Context | Needle in Haystack | 0 | em | 96.8 | 1 | 1 |
### Instruction Tuned Models
| Capability | | Benchmark | \# Shots | Metric | Llama 3.2 1B bf16 | Llama 3.2 1B Vanilla PTQ\*\* | Llama 3.2 1B Spin Quant | Llama 3.2 1B QLoRA | Llama 3.2 3B bf16 | Llama 3.2 3B Vanilla PTQ\*\* | Llama 3.2 3B Spin Quant | Llama 3.2 3B QLoRA | Llama 3.1 8B |
| :---: | ----- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
| General | | MMLU | 5 | macro\_avg/acc | 49.3 | 43.3 | 47.3 | 49.0 | 63.4 | 60.5 | 62 | 62.4 | 69.4 |
| Re-writing | | Open-rewrite eval | 0 | micro\_avg/rougeL | 41.6 | 39.2 | 40.9 | 41.2 | 40.1 | 40.3 | 40.8 | 40.7 | 40.9 |
| Summarization | | TLDR9+ (test) | 1 | rougeL | 16.8 | 14.9 | 16.7 | 16.8 | 19.0 | 19.1 | 19.2 | 19.1 | 17.2 |
| Instruction following | | IFEval | 0 | Avg(Prompt/Instruction acc Loose/Strict) | 59.5 | 51.5 | 58.4 | 55.6 | 77.4 | 73.9 | 73.5 | 75.9 | 80.4 |
| Math | | GSM8K (CoT) | 8 | em\_maj1@1 | 44.4 | 33.1 | 40.6 | 46.5 | 77.7 | 72.9 | 75.7 | 77.9 | 84.5 |
| | | MATH (CoT) | 0 | final\_em | 30.6 | 20.5 | 25.3 | 31.0 | 48.0 | 44.2 | 45.3 | 49.2 | 51.9 |
| Reasoning | | ARC-C | 0 | acc | 59.4 | 54.3 | 57 | 60.7 | 78.6 | 75.6 | 77.6 | 77.6 | 83.4 |
| | | GPQA | 0 | acc | 27.2 | 25.9 | 26.3 | 25.9 | 32.8 | 32.8 | 31.7 | 33.9 | 32.8 |
| | | Hellaswag | 0 | acc | 41.2 | 38.1 | 41.3 | 41.5 | 69.8 | 66.3 | 68 | 66.3 | 78.7 |
| Tool Use | | BFCL V2 | 0 | acc | 25.7 | 14.3 | 15.9 | 23.7 | 67.0 | 53.4 | 60.1 | 63.5 | 67.1 |
| | | Nexus | 0 | macro\_avg/acc | 13.5 | 5.2 | 9.6 | 12.5 | 34.3 | 32.4 | 31.5 | 30.1 | 38.5 |
| Long Context | | InfiniteBench/En.QA | 0 | longbook\_qa/f1 | 20.3 | N/A | N/A | N/A | 19.8 | N/A | N/A | N/A | 27.3 |
| | | InfiniteBench/En.MC | 0 | longbook\_choice/acc | 38.0 | N/A | N/A | N/A | 63.3 | N/A | N/A | N/A | 72.2 |
| | | NIH/Multi-needle | 0 | recall | 75.0 | N/A | N/A | N/A | 84.7 | N/A | N/A | N/A | 98.8 |
| Multilingual | | MGSM (CoT) | 0 | em | 24.5 | 13.7 | 18.2 | 24.4 | 58.2 | 48.9 | 54.3 | 56.8 | 68.9 |
\*\*for comparison purposes only. Model not released.
### Multilingual Benchmarks
| Category | Benchmark | Language | Llama 3.2 1B | Llama 3.2 1B Vanilla PTQ\*\* | Llama 3.2 1B Spin Quant | Llama 3.2 1B QLoRA | Llama 3.2 3B | Llama 3.2 3B Vanilla PTQ\*\* | Llama 3.2 3B Spin Quant | Llama 3.2 3B QLoRA | Llama 3.1 8B |
| :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
| General | MMLU (5-shot, macro_avg/acc) | Portuguese | 39.8 | 34.9 | 38.9 | 40.2 | 54.5 | 50.9 | 53.3 | 53.4 | 62.1 |
| | | Spanish | 41.5 | 36.0 | 39.8 | 41.8 | 55.1 | 51.9 | 53.6 | 53.6 | 62.5 |
| | | Italian | 39.8 | 34.9 | 38.1 | 40.6 | 53.8 | 49.9 | 52.1 | 51.7 | 61.6 |
| | | German | 39.2 | 34.9 | 37.5 | 39.6 | 53.3 | 50.0 | 52.2 | 51.3 | 60.6 |
| | | French | 40.5 | 34.8 | 39.2 | 40.8 | 54.6 | 51.2 | 53.3 | 53.3 | 62.3 |
| | | Hindi | 33.5 | 30.0 | 32.1 | 34.0 | 43.3 | 40.4 | 42.0 | 42.1 | 50.9 |
| | | Thai | 34.7 | 31.2 | 32.4 | 34.9 | 44.5 | 41.3 | 44.0 | 42.2 | 50.3 |
\*\*for comparison purposes only. Model not released.
## Inference time
In the below table, we compare the performance metrics of different quantization methods (SpinQuant and QAT \+ LoRA) with the BF16 baseline. The evaluation was done using the [ExecuTorch](https://github.com/pytorch/executorch) framework as the inference engine, with the ARM CPU as a backend using Android OnePlus 12 device.
| Category | Decode (tokens/sec) | Time-to-first-token (sec) | Prefill (tokens/sec) | Model size (PTE file size in MB) | Memory size (RSS in MB) |
| :---- | ----- | ----- | ----- | ----- | ----- |
| 1B BF16 (baseline) | 19.2 | 1.0 | 60.3 | 2358 | 3,185 |
| 1B SpinQuant | 50.2 (2.6x) | 0.3 (-76.9%) | 260.5 (4.3x) | 1083 (-54.1%) | 1,921 (-39.7%) |
| 1B QLoRA | 45.8 (2.4x) | 0.3 (-76.0%) | 252.0 (4.2x) | 1127 (-52.2%) | 2,255 (-29.2%) |
| 3B BF16 (baseline) | 7.6 | 3.0 | 21.2 | 6129 | 7,419 |
| 3B SpinQuant | 19.7 (2.6x) | 0.7 (-76.4%) | 89.7 (4.2x) | 2435 (-60.3%) | 3,726 (-49.8%) |
| 3B QLoRA | 18.5 (2.4x) | 0.7 (-76.1%) | 88.8 (4.2x) | 2529 (-58.7%) | 4,060 (-45.3%) |
(\*) The performance measurement is done using an adb binary-based approach.
(\*\*) It is measured on an Android OnePlus 12 device.
(\*\*\*) Time-to-first-token (TTFT) is measured with prompt length=64
*Footnote:*
- *Decode (tokens/second) is for how quickly it keeps generating. Higher is better.*
- *Time-to-first-token (TTFT for shorthand) is for how fast it generates the first token for a given prompt. Lower is better.*
- *Prefill is the inverse of TTFT (aka 1/TTFT) in tokens/second. Higher is better*
- *Model size \- how big is the model, measured by, PTE file, a binary file format for ExecuTorch*
- *RSS size \- Memory usage in resident set size (RSS)*
## Responsibility & Safety
As part of our Responsible release approach, we followed a three-pronged strategy to managing trust & safety risks:
1. Enable developers to deploy helpful, safe and flexible experiences for their target audience and for the use cases supported by Llama
2. Protect developers against adversarial users aiming to exploit Llama capabilities to potentially cause harm
3. Provide protections for the community to help prevent the misuse of our models
### Responsible Deployment
**Approach:** Llama is a foundational technology designed to be used in a variety of use cases. Examples on how Meta’s Llama models have been responsibly deployed can be found in our [Community Stories webpage](https://llama.meta.com/community-stories/). Our approach is to build the most helpful models, enabling the world to benefit from the technology power, by aligning our model safety for generic use cases and addressing a standard set of harms. Developers are then in the driver’s seat to tailor safety for their use cases, defining their own policies and deploying the models with the necessary safeguards in their Llama systems. Llama 3.2 was developed following the best practices outlined in our [Responsible Use Guide](https://llama.meta.com/responsible-use-guide/).
#### Llama 3.2 Instruct
**Objective:** Our main objectives for conducting safety fine-tuning are to provide the research community with a valuable resource for studying the robustness of safety fine-tuning, as well as to offer developers a readily available, safe, and powerful model for various applications to reduce the developer workload to deploy safe AI systems. We implemented the same set of safety mitigations as in Llama 3, and you can learn more about these in the Llama 3 [paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/).
**Fine-Tuning Data:** We employ a multi-faceted approach to data collection, combining human-generated data from our vendors with synthetic data to mitigate potential safety risks. We’ve developed many large language model (LLM)-based classifiers that enable us to thoughtfully select high-quality prompts and responses, enhancing data quality control.
**Refusals and Tone:** Building on the work we started with Llama 3, we put a great emphasis on model refusals to benign prompts as well as refusal tone. We included both borderline and adversarial prompts in our safety data strategy, and modified our safety data responses to follow tone guidelines.
#### Llama 3.2 Systems
**Safety as a System:** Large language models, including Llama 3.2, **are not designed to be deployed in isolation** but instead should be deployed as part of an overall AI system with additional safety guardrails as required. Developers are expected to deploy system safeguards when building agentic systems. Safeguards are key to achieve the right helpfulness-safety alignment as well as mitigating safety and security risks inherent to the system and any integration of the model or system with external tools. As part of our responsible release approach, we provide the community with [safeguards](https://llama.meta.com/trust-and-safety/) that developers should deploy with Llama models or other LLMs, including Llama Guard, Prompt Guard and Code Shield. All our [reference implementations](https://github.com/meta-llama/llama-agentic-system) demos contain these safeguards by default so developers can benefit from system-level safety out-of-the-box.
### New Capabilities and Use Cases
**Technological Advancement:** Llama releases usually introduce new capabilities that require specific considerations in addition to the best practices that generally apply across all Generative AI use cases. For prior release capabilities also supported by Llama 3.2, see [Llama 3.1 Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/MODEL_CARD.md), as the same considerations apply here as well.
**Constrained Environments:** Llama 3.2 1B and 3B models are expected to be deployed in highly constrained environments, such as mobile devices. LLM Systems using smaller models will have a different alignment profile and safety/helpfulness tradeoff than more complex, larger systems. Developers should ensure the safety of their system meets the requirements of their use case. We recommend using lighter system safeguards for such use cases, like Llama Guard 3-1B or its mobile-optimized version.
### Evaluations
**Scaled Evaluations:** We built dedicated, adversarial evaluation datasets and evaluated systems composed of Llama models and Purple Llama safeguards to filter input prompt and output response. It is important to evaluate applications in context, and we recommend building dedicated evaluation dataset for your use case.
**Red Teaming:** We conducted recurring red teaming exercises with the goal of discovering risks via adversarial prompting and we used the learnings to improve our benchmarks and safety tuning datasets. We partnered early with subject-matter experts in critical risk areas to understand the nature of these real-world harms and how such models may lead to unintended harm for society. Based on these conversations, we derived a set of adversarial goals for the red team to attempt to achieve, such as extracting harmful information or reprogramming the model to act in a potentially harmful capacity. The red team consisted of experts in cybersecurity, adversarial machine learning, responsible AI, and integrity in addition to multilingual content specialists with background in integrity issues in specific geographic markets.
### Critical Risks
In addition to our safety work above, we took extra care on measuring and/or mitigating the following critical risk areas:
**1\. CBRNE (Chemical, Biological, Radiological, Nuclear, and Explosive Weapons):** Llama 3.2 1B and 3B models are smaller and less capable derivatives of Llama 3.1. For Llama 3.1 70B and 405B, to assess risks related to proliferation of chemical and biological weapons, we performed uplift testing designed to assess whether use of Llama 3.1 models could meaningfully increase the capabilities of malicious actors to plan or carry out attacks using these types of weapons and have determined that such testing also applies to the smaller 1B and 3B models.
**2\. Child Safety:** Child Safety risk assessments were conducted using a team of experts, to assess the model’s capability to produce outputs that could result in Child Safety risks and inform on any necessary and appropriate risk mitigations via fine tuning. We leveraged those expert red teaming sessions to expand the coverage of our evaluation benchmarks through Llama 3 model development. For Llama 3, we conducted new in-depth sessions using objective based methodologies to assess the model risks along multiple attack vectors including the additional languages Llama 3 is trained on. We also partnered with content specialists to perform red teaming exercises assessing potentially violating content while taking account of market specific nuances or experiences.
**3\. Cyber Attacks:** For Llama 3.1 405B, our cyber attack uplift study investigated whether LLMs can enhance human capabilities in hacking tasks, both in terms of skill level and speed.
Our attack automation study focused on evaluating the capabilities of LLMs when used as autonomous agents in cyber offensive operations, specifically in the context of ransomware attacks. This evaluation was distinct from previous studies that considered LLMs as interactive assistants. The primary objective was to assess whether these models could effectively function as independent agents in executing complex cyber-attacks without human intervention. Because Llama 3.2’s 1B and 3B models are smaller and less capable models than Llama 3.1 405B, we broadly believe that the testing conducted for the 405B model also applies to Llama 3.2 models.
### Community
**Industry Partnerships:** Generative AI safety requires expertise and tooling, and we believe in the strength of the open community to accelerate its progress. We are active members of open consortiums, including the AI Alliance, Partnership on AI and MLCommons, actively contributing to safety standardization and transparency. We encourage the community to adopt taxonomies like the MLCommons Proof of Concept evaluation to facilitate collaboration and transparency on safety and content evaluations. Our Purple Llama tools are open sourced for the community to use and widely distributed across ecosystem partners including cloud service providers. We encourage community contributions to our [Github repository](https://github.com/meta-llama/PurpleLlama).
**Grants:** We also set up the [Llama Impact Grants](https://llama.meta.com/llama-impact-grants/) program to identify and support the most compelling applications of Meta’s Llama model for societal benefit across three categories: education, climate and open innovation. The 20 finalists from the hundreds of applications can be found [here](https://llama.meta.com/llama-impact-grants/#finalists).
**Reporting:** Finally, we put in place a set of resources including an [output reporting mechanism](https://developers.facebook.com/llama_output_feedback) and [bug bounty program](https://www.facebook.com/whitehat) to continuously improve the Llama technology with the help of the community.
## Ethical Considerations and Limitations
**Values:** The core values of Llama 3.2 are openness, inclusivity and helpfulness. It is meant to serve everyone, and to work for a wide range of use cases. It is thus designed to be accessible to people across many different backgrounds, experiences and perspectives. Llama 3.2 addresses users and their needs as they are, without insertion unnecessary judgment or normativity, while reflecting the understanding that even content that may appear problematic in some cases can serve valuable purposes in others. It respects the dignity and autonomy of all users, especially in terms of the values of free thought and expression that power innovation and progress.
**Testing:** Llama 3.2 is a new technology, and like any new technology, there are risks associated with its use. Testing conducted to date has not covered, nor could it cover, all scenarios. For these reasons, as with all LLMs, Llama 3.2’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 3.2 models, developers should perform safety testing and tuning tailored to their specific applications of the model. Please refer to available resources including our [Responsible Use Guide](https://llama.meta.com/responsible-use-guide), [Trust and Safety](https://llama.meta.com/trust-and-safety/) solutions, and other [resources](https://llama.meta.com/docs/get-started/) to learn more about responsible development.
|
abi2736/Minds-Tutor | abi2736 | 2025-06-15T19:39:36Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T15:01:00Z | # Minds-Tutor
Projeto inicial para IA educacional. |
Mungert/Llama-3.2-1B-Instruct-GGUF | Mungert | 2025-06-15T19:39:29Z | 1,482 | 3 | transformers | [
"transformers",
"gguf",
"facebook",
"meta",
"pytorch",
"llama",
"llama-3",
"text-generation",
"en",
"de",
"fr",
"it",
"pt",
"hi",
"es",
"th",
"arxiv:2204.05149",
"arxiv:2405.16406",
"license:llama3.2",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-03-15T03:34:52Z | ---
language:
- en
- de
- fr
- it
- pt
- hi
- es
- th
library_name: transformers
pipeline_tag: text-generation
tags:
- facebook
- meta
- pytorch
- llama
- llama-3
license: llama3.2
extra_gated_prompt: >-
### LLAMA 3.2 COMMUNITY LICENSE AGREEMENT
Llama 3.2 Version Release Date: September 25, 2024
“Agreement” means the terms and conditions for use, reproduction, distribution
and modification of the Llama Materials set forth herein.
“Documentation” means the specifications, manuals and documentation accompanying Llama 3.2
distributed by Meta at https://llama.meta.com/doc/overview.
“Licensee” or “you” means you, or your employer or any other person or entity (if you are
entering into this Agreement on such person or entity’s behalf), of the age required under
applicable laws, rules or regulations to provide legal consent and that has legal authority
to bind your employer or such other person or entity if you are entering in this Agreement
on their behalf.
“Llama 3.2” means the foundational large language models and software and algorithms, including
machine-learning model code, trained model weights, inference-enabling code, training-enabling code,
fine-tuning enabling code and other elements of the foregoing distributed by Meta at
https://www.llama.com/llama-downloads.
“Llama Materials” means, collectively, Meta’s proprietary Llama 3.2 and Documentation (and
any portion thereof) made available under this Agreement.
“Meta” or “we” means Meta Platforms Ireland Limited (if you are located in or,
if you are an entity, your principal place of business is in the EEA or Switzerland)
and Meta Platforms, Inc. (if you are located outside of the EEA or Switzerland).
By clicking “I Accept” below or by using or distributing any portion or element of the Llama Materials,
you agree to be bound by this Agreement.
1. License Rights and Redistribution.
a. Grant of Rights. You are granted a non-exclusive, worldwide,
non-transferable and royalty-free limited license under Meta’s intellectual property or other rights
owned by Meta embodied in the Llama Materials to use, reproduce, distribute, copy, create derivative works
of, and make modifications to the Llama Materials.
b. Redistribution and Use.
i. If you distribute or make available the Llama Materials (or any derivative works thereof),
or a product or service (including another AI model) that contains any of them, you shall (A) provide
a copy of this Agreement with any such Llama Materials; and (B) prominently display “Built with Llama”
on a related website, user interface, blogpost, about page, or product documentation. If you use the
Llama Materials or any outputs or results of the Llama Materials to create, train, fine tune, or
otherwise improve an AI model, which is distributed or made available, you shall also include “Llama”
at the beginning of any such AI model name.
ii. If you receive Llama Materials, or any derivative works thereof, from a Licensee as part
of an integrated end user product, then Section 2 of this Agreement will not apply to you.
iii. You must retain in all copies of the Llama Materials that you distribute the
following attribution notice within a “Notice” text file distributed as a part of such copies:
“Llama 3.2 is licensed under the Llama 3.2 Community License, Copyright © Meta Platforms,
Inc. All Rights Reserved.”
iv. Your use of the Llama Materials must comply with applicable laws and regulations
(including trade compliance laws and regulations) and adhere to the Acceptable Use Policy for
the Llama Materials (available at https://www.llama.com/llama3_2/use-policy), which is hereby
incorporated by reference into this Agreement.
2. Additional Commercial Terms. If, on the Llama 3.2 version release date, the monthly active users
of the products or services made available by or for Licensee, or Licensee’s affiliates,
is greater than 700 million monthly active users in the preceding calendar month, you must request
a license from Meta, which Meta may grant to you in its sole discretion, and you are not authorized to
exercise any of the rights under this Agreement unless or until Meta otherwise expressly grants you such rights.
3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA MATERIALS AND ANY OUTPUT AND
RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF ANY KIND, AND META DISCLAIMS
ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES
OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE
FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE LLAMA MATERIALS AND ASSUME ANY RISKS ASSOCIATED
WITH YOUR USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS.
4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY,
WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT,
FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN
IF META OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.
5. Intellectual Property.
a. No trademark licenses are granted under this Agreement, and in connection with the Llama Materials,
neither Meta nor Licensee may use any name or mark owned by or associated with the other or any of its affiliates,
except as required for reasonable and customary use in describing and redistributing the Llama Materials or as
set forth in this Section 5(a). Meta hereby grants you a license to use “Llama” (the “Mark”) solely as required
to comply with the last sentence of Section 1.b.i. You will comply with Meta’s brand guidelines (currently accessible
at https://about.meta.com/brand/resources/meta/company-brand/). All goodwill arising out of your use of the Mark
will inure to the benefit of Meta.
b. Subject to Meta’s ownership of Llama Materials and derivatives made by or for Meta, with respect to any
derivative works and modifications of the Llama Materials that are made by you, as between you and Meta,
you are and will be the owner of such derivative works and modifications.
c. If you institute litigation or other proceedings against Meta or any entity (including a cross-claim or
counterclaim in a lawsuit) alleging that the Llama Materials or Llama 3.2 outputs or results, or any portion
of any of the foregoing, constitutes infringement of intellectual property or other rights owned or licensable
by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or
claim is filed or instituted. You will indemnify and hold harmless Meta from and against any claim by any third
party arising out of or related to your use or distribution of the Llama Materials.
6. Term and Termination. The term of this Agreement will commence upon your acceptance of this Agreement or access
to the Llama Materials and will continue in full force and effect until terminated in accordance with the terms
and conditions herein. Meta may terminate this Agreement if you are in breach of any term or condition of this
Agreement. Upon termination of this Agreement, you shall delete and cease use of the Llama Materials. Sections 3,
4 and 7 shall survive the termination of this Agreement.
7. Governing Law and Jurisdiction. This Agreement will be governed and construed under the laws of the State of
California without regard to choice of law principles, and the UN Convention on Contracts for the International
Sale of Goods does not apply to this Agreement. The courts of California shall have exclusive jurisdiction of
any dispute arising out of this Agreement.
### Llama 3.2 Acceptable Use Policy
Meta is committed to promoting safe and fair use of its tools and features, including Llama 3.2.
If you access or use Llama 3.2, you agree to this Acceptable Use Policy (“**Policy**”).
The most recent copy of this policy can be found at
[https://www.llama.com/llama3_2/use-policy](https://www.llama.com/llama3_2/use-policy).
#### Prohibited Uses
We want everyone to use Llama 3.2 safely and responsibly. You agree you will not use, or allow others to use, Llama 3.2 to:
1. Violate the law or others’ rights, including to:
1. Engage in, promote, generate, contribute to, encourage, plan, incite, or further illegal or unlawful activity or content, such as:
1. Violence or terrorism
2. Exploitation or harm to children, including the solicitation, creation, acquisition, or dissemination of child exploitative content or failure to report Child Sexual Abuse Material
3. Human trafficking, exploitation, and sexual violence
4. The illegal distribution of information or materials to minors, including obscene materials, or failure to employ legally required age-gating in connection with such information or materials.
5. Sexual solicitation
6. Any other criminal activity
1. Engage in, promote, incite, or facilitate the harassment, abuse, threatening, or bullying of individuals or groups of individuals
2. Engage in, promote, incite, or facilitate discrimination or other unlawful or harmful conduct in the provision of employment, employment benefits, credit, housing, other economic benefits, or other essential goods and services
3. Engage in the unauthorized or unlicensed practice of any profession including, but not limited to, financial, legal, medical/health, or related professional practices
4. Collect, process, disclose, generate, or infer private or sensitive information about individuals, including information about individuals’ identity, health, or demographic information, unless you have obtained the right to do so in accordance with applicable law
5. Engage in or facilitate any action or generate any content that infringes, misappropriates, or otherwise violates any third-party rights, including the outputs or results of any products or services using the Llama Materials
6. Create, generate, or facilitate the creation of malicious code, malware, computer viruses or do anything else that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system
7. Engage in any action, or facilitate any action, to intentionally circumvent or remove usage restrictions or other safety measures, or to enable functionality disabled by Meta
2. Engage in, promote, incite, facilitate, or assist in the planning or development of activities that present a risk of death or bodily harm to individuals, including use of Llama 3.2 related to the following:
8. Military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulations (ITAR) maintained by the United States Department of State or to the U.S. Biological Weapons Anti-Terrorism Act of 1989 or the Chemical Weapons Convention Implementation Act of 1997
9. Guns and illegal weapons (including weapon development)
10. Illegal drugs and regulated/controlled substances
11. Operation of critical infrastructure, transportation technologies, or heavy machinery
12. Self-harm or harm to others, including suicide, cutting, and eating disorders
13. Any content intended to incite or promote violence, abuse, or any infliction of bodily harm to an individual
3. Intentionally deceive or mislead others, including use of Llama 3.2 related to the following:
14. Generating, promoting, or furthering fraud or the creation or promotion of disinformation
15. Generating, promoting, or furthering defamatory content, including the creation of defamatory statements, images, or other content
16. Generating, promoting, or further distributing spam
17. Impersonating another individual without consent, authorization, or legal right
18. Representing that the use of Llama 3.2 or outputs are human-generated
19. Generating or facilitating false online engagement, including fake reviews and other means of fake online engagement
4. Fail to appropriately disclose to end users any known dangers of your AI system
5. Interact with third party tools, models, or software designed to generate unlawful content or engage in unlawful or harmful conduct and/or represent that the outputs of such tools, models, or software are associated with Meta or Llama 3.2
With respect to any multimodal models included in Llama 3.2, the rights granted under Section 1(a) of the Llama 3.2 Community License Agreement are not being granted to you if you are an individual domiciled in, or a company with a principal place of business in, the European Union. This restriction does not apply to end users of a product or service that incorporates any such multimodal models.
Please report any violation of this Policy, software “bug,” or other problems that could lead to a violation of this Policy through one of the following means:
* Reporting issues with the model: [https://github.com/meta-llama/llama-models/issues](https://l.workplace.com/l.php?u=https%3A%2F%2Fgithub.com%2Fmeta-llama%2Fllama-models%2Fissues&h=AT0qV8W9BFT6NwihiOHRuKYQM_UnkzN_NmHMy91OT55gkLpgi4kQupHUl0ssR4dQsIQ8n3tfd0vtkobvsEvt1l4Ic6GXI2EeuHV8N08OG2WnbAmm0FL4ObkazC6G_256vN0lN9DsykCvCqGZ)
* Reporting risky content generated by the model: [developers.facebook.com/llama_output_feedback](http://developers.facebook.com/llama_output_feedback)
* Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info)
* Reporting violations of the Acceptable Use Policy or unlicensed uses of Llama 3.2: [email protected]
extra_gated_fields:
First Name: text
Last Name: text
Date of birth: date_picker
Country: country
Affiliation: text
Job title:
type: select
options:
- Student
- Research Graduate
- AI researcher
- AI developer/engineer
- Reporter
- Other
geo: ip_location
By clicking Submit below I accept the terms of the license and acknowledge that the information I provide will be collected stored processed and shared in accordance with the Meta Privacy Policy: checkbox
extra_gated_description: >-
The information you provide will be collected, stored, processed and shared in
accordance with the [Meta Privacy
Policy](https://www.facebook.com/privacy/policy/).
extra_gated_button_content: Submit
---
# <span style="color: #7FFF7F;">Llama-3.2-1B-Instruct GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.2-1B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.2-1B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.2-1B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.2-1B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.2-1B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.2-1B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.2-1B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.2-1B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.2-1B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.2-1B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.2-1B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
## Model Information
The Llama 3.2 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction-tuned generative models in 1B and 3B sizes (text in/text out). The Llama 3.2 instruction-tuned text only models are optimized for multilingual dialogue use cases, including agentic retrieval and summarization tasks. They outperform many of the available open source and closed chat models on common industry benchmarks.
**Model Developer:** Meta
**Model Architecture:** Llama 3.2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.
| | Training Data | Params | Input modalities | Output modalities | Context Length | GQA | Shared Embeddings | Token count | Knowledge cutoff |
| :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- |
| Llama 3.2 (text only) | A new mix of publicly available online data. | 1B (1.23B) | Multilingual Text | Multilingual Text and code | 128k | Yes | Yes | Up to 9T tokens | December 2023 |
| | | 3B (3.21B) | Multilingual Text | Multilingual Text and code | | | | | |
| Llama 3.2 Quantized (text only) | A new mix of publicly available online data. | 1B (1.23B) | Multilingual Text | Multilingual Text and code | 8k | Yes | Yes | Up to 9T tokens | December 2023 |
| | | 3B (3.21B) | Multilingual Text | Multilingual Text and code | | | | | |
**Supported Languages:** English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai are officially supported. Llama 3.2 has been trained on a broader collection of languages than these 8 supported languages. Developers may fine-tune Llama 3.2 models for languages beyond these supported languages, provided they comply with the Llama 3.2 Community License and the Acceptable Use Policy. Developers are always expected to ensure that their deployments, including those that involve additional languages, are completed safely and responsibly.
**Llama 3.2 Model Family:** Token counts refer to pretraining data only. All model versions use Grouped-Query Attention (GQA) for improved inference scalability.
**Model Release Date:** Sept 25, 2024
**Status:** This is a static model trained on an offline dataset. Future versions may be released that improve model capabilities and safety.
**License:** Use of Llama 3.2 is governed by the [Llama 3.2 Community License](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/LICENSE) (a custom, commercial license agreement).
**Feedback:** Instructions on how to provide feedback or comments on the model can be found in the Llama Models [README](https://github.com/meta-llama/llama-models/blob/main/README.md). For more technical information about generation parameters and recipes for how to use Llama 3.2 in applications, please go [here](https://github.com/meta-llama/llama-recipes).
## Intended Use
**Intended Use Cases:** Llama 3.2 is intended for commercial and research use in multiple languages. Instruction tuned text only models are intended for assistant-like chat and agentic applications like knowledge retrieval and summarization, mobile AI powered writing assistants and query and prompt rewriting. Pretrained models can be adapted for a variety of additional natural language generation tasks. Similarly, quantized models can be adapted for a variety of on-device use-cases with limited compute resources.
**Out of Scope:** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in any other way that is prohibited by the Acceptable Use Policy and Llama 3.2 Community License. Use in languages beyond those explicitly referenced as supported in this model card.
## How to use
This repository contains two versions of Llama-3.2-1B-Instruct, for use with transformers and with the original `llama` codebase.
### Use with transformers
Starting with `transformers >= 4.43.0` onward, you can run conversational inference using the Transformers `pipeline` abstraction or by leveraging the Auto classes with the `generate()` function.
Make sure to update your transformers installation via `pip install --upgrade transformers`.
```python
import torch
from transformers import pipeline
model_id = "meta-llama/Llama-3.2-1B-Instruct"
pipe = pipeline(
"text-generation",
model=model_id,
torch_dtype=torch.bfloat16,
device_map="auto",
)
messages = [
{"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
{"role": "user", "content": "Who are you?"},
]
outputs = pipe(
messages,
max_new_tokens=256,
)
print(outputs[0]["generated_text"][-1])
```
Note: You can also find detailed recipes on how to use the model locally, with `torch.compile()`, assisted generations, quantised and more at [`huggingface-llama-recipes`](https://github.com/huggingface/huggingface-llama-recipes)
### Use with `llama`
Please, follow the instructions in the [repository](https://github.com/meta-llama/llama)
To download Original checkpoints, see the example command below leveraging `huggingface-cli`:
```
huggingface-cli download meta-llama/Llama-3.2-1B-Instruct --include "original/*" --local-dir Llama-3.2-1B-Instruct
```
## Hardware and Software
**Training Factors:** We used custom training libraries, Meta's custom built GPU cluster, and production infrastructure for pretraining. Fine-tuning, quantization, annotation, and evaluation were also performed on production infrastructure.
**Training Energy Use:** Training utilized a cumulative of **916k** GPU hours of computation on H100-80GB (TDP of 700W) type hardware, per the table below. Training time is the total GPU time required for training each model and power consumption is the peak power capacity per GPU device used, adjusted for power usage efficiency.
**Training Greenhouse Gas Emissions:** Estimated total location-based greenhouse gas emissions were **240** tons CO2eq for training. Since 2020, Meta has maintained net zero greenhouse gas emissions in its global operations and matched 100% of its electricity use with renewable energy; therefore, the total market-based greenhouse gas emissions for training were 0 tons CO2eq.
| | Training Time (GPU hours) | Logit Generation Time (GPU Hours) | Training Power Consumption (W) | Training Location-Based Greenhouse Gas Emissions (tons CO2eq) | Training Market-Based Greenhouse Gas Emissions (tons CO2eq) |
| :---- | :---: | ----- | :---: | :---: | :---: |
| Llama 3.2 1B | 370k | \- | 700 | 107 | 0 |
| Llama 3.2 3B | 460k | \- | 700 | 133 | 0 |
| Llama 3.2 1B SpinQuant | 1.7 | 0 | 700 | *Negligible*\*\* | 0 |
| Llama 3.2 3B SpinQuant | 2.4 | 0 | 700 | *Negligible*\*\* | 0 |
| Llama 3.2 1B QLora | 1.3k | 0 | 700 | 0.381 | 0 |
| Llama 3.2 3B QLora | 1.6k | 0 | 700 | 0.461 | 0 |
| Total | 833k | 86k | | 240 | 0 |
\*\* The location-based CO2e emissions of Llama 3.2 1B SpinQuant and Llama 3.2 3B SpinQuant are less than 0.001 metric tonnes each. This is due to the minimal training GPU hours that are required.
The methodology used to determine training energy use and greenhouse gas emissions can be found [here](https://arxiv.org/pdf/2204.05149). Since Meta is openly releasing these models, the training energy use and greenhouse gas emissions will not be incurred by others.
## Training Data
**Overview:** Llama 3.2 was pretrained on up to 9 trillion tokens of data from publicly available sources. For the 1B and 3B Llama 3.2 models, we incorporated logits from the Llama 3.1 8B and 70B models into the pretraining stage of the model development, where outputs (logits) from these larger models were used as token-level targets. Knowledge distillation was used after pruning to recover performance. In post-training we used a similar recipe as Llama 3.1 and produced final chat models by doing several rounds of alignment on top of the pre-trained model. Each round involved Supervised Fine-Tuning (SFT), Rejection Sampling (RS), and Direct Preference Optimization (DPO).
**Data Freshness:** The pretraining data has a cutoff of December 2023\.
## Quantization
### Quantization Scheme
We designed the current quantization scheme with the [PyTorch’s ExecuTorch](https://github.com/pytorch/executorch) inference framework and Arm CPU backend in mind, taking into account metrics including model quality, prefill/decoding speed, and memory footprint. Our quantization scheme involves three parts:
- All linear layers in all transformer blocks are quantized to a 4-bit groupwise scheme (with a group size of 32) for weights and 8-bit per-token dynamic quantization for activations.
- The classification layer is quantized to 8-bit per-channel for weight and 8-bit per token dynamic quantization for activation.
- Similar to classification layer, an 8-bit per channel quantization is used for embedding layer.
### Quantization-Aware Training and LoRA
The quantization-aware training (QAT) with low-rank adaptation (LoRA) models went through only post-training stages, using the same data as the full precision models. To initialize QAT, we utilize BF16 Llama 3.2 model checkpoints obtained after supervised fine-tuning (SFT) and perform an additional full round of SFT training with QAT. We then freeze the backbone of the QAT model and perform another round of SFT with LoRA adaptors applied to all layers within the transformer block. Meanwhile, the LoRA adaptors' weights and activations are maintained in BF16. Because our approach is similar to QLoRA of Dettmers et al., (2023) (i.e., quantization followed by LoRA adapters), we refer this method as QLoRA. Finally, we fine-tune the resulting model (both backbone and LoRA adaptors) using direct preference optimization (DPO).
### SpinQuant
[SpinQuant](https://arxiv.org/abs/2405.16406) was applied, together with generative post-training quantization (GPTQ). For the SpinQuant rotation matrix fine-tuning, we optimized for 100 iterations, using 800 samples with sequence-length 2048 from the WikiText 2 dataset. For GPTQ, we used 128 samples from the same dataset with the same sequence-length.
## Benchmarks \- English Text
In this section, we report the results for Llama 3.2 models on standard automatic benchmarks. For all these evaluations, we used our internal evaluations library.
### Base Pretrained Models
| Category | Benchmark | \# Shots | Metric | Llama 3.2 1B | Llama 3.2 3B | Llama 3.1 8B |
| ----- | ----- | :---: | :---: | :---: | :---: | :---: |
| General | MMLU | 5 | macro\_avg/acc\_char | 32.2 | 58 | 66.7 |
| | AGIEval English | 3-5 | average/acc\_char | 23.3 | 39.2 | 47.8 |
| | ARC-Challenge | 25 | acc\_char | 32.8 | 69.1 | 79.7 |
| Reading comprehension | SQuAD | 1 | em | 49.2 | 67.7 | 77 |
| | QuAC (F1) | 1 | f1 | 37.9 | 42.9 | 44.9 |
| | DROP (F1) | 3 | f1 | 28.0 | 45.2 | 59.5 |
| Long Context | Needle in Haystack | 0 | em | 96.8 | 1 | 1 |
### Instruction Tuned Models
| Capability | | Benchmark | \# Shots | Metric | Llama 3.2 1B bf16 | Llama 3.2 1B Vanilla PTQ\*\* | Llama 3.2 1B Spin Quant | Llama 3.2 1B QLoRA | Llama 3.2 3B bf16 | Llama 3.2 3B Vanilla PTQ\*\* | Llama 3.2 3B Spin Quant | Llama 3.2 3B QLoRA | Llama 3.1 8B |
| :---: | ----- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
| General | | MMLU | 5 | macro\_avg/acc | 49.3 | 43.3 | 47.3 | 49.0 | 63.4 | 60.5 | 62 | 62.4 | 69.4 |
| Re-writing | | Open-rewrite eval | 0 | micro\_avg/rougeL | 41.6 | 39.2 | 40.9 | 41.2 | 40.1 | 40.3 | 40.8 | 40.7 | 40.9 |
| Summarization | | TLDR9+ (test) | 1 | rougeL | 16.8 | 14.9 | 16.7 | 16.8 | 19.0 | 19.1 | 19.2 | 19.1 | 17.2 |
| Instruction following | | IFEval | 0 | Avg(Prompt/Instruction acc Loose/Strict) | 59.5 | 51.5 | 58.4 | 55.6 | 77.4 | 73.9 | 73.5 | 75.9 | 80.4 |
| Math | | GSM8K (CoT) | 8 | em\_maj1@1 | 44.4 | 33.1 | 40.6 | 46.5 | 77.7 | 72.9 | 75.7 | 77.9 | 84.5 |
| | | MATH (CoT) | 0 | final\_em | 30.6 | 20.5 | 25.3 | 31.0 | 48.0 | 44.2 | 45.3 | 49.2 | 51.9 |
| Reasoning | | ARC-C | 0 | acc | 59.4 | 54.3 | 57 | 60.7 | 78.6 | 75.6 | 77.6 | 77.6 | 83.4 |
| | | GPQA | 0 | acc | 27.2 | 25.9 | 26.3 | 25.9 | 32.8 | 32.8 | 31.7 | 33.9 | 32.8 |
| | | Hellaswag | 0 | acc | 41.2 | 38.1 | 41.3 | 41.5 | 69.8 | 66.3 | 68 | 66.3 | 78.7 |
| Tool Use | | BFCL V2 | 0 | acc | 25.7 | 14.3 | 15.9 | 23.7 | 67.0 | 53.4 | 60.1 | 63.5 | 67.1 |
| | | Nexus | 0 | macro\_avg/acc | 13.5 | 5.2 | 9.6 | 12.5 | 34.3 | 32.4 | 31.5 | 30.1 | 38.5 |
| Long Context | | InfiniteBench/En.QA | 0 | longbook\_qa/f1 | 20.3 | N/A | N/A | N/A | 19.8 | N/A | N/A | N/A | 27.3 |
| | | InfiniteBench/En.MC | 0 | longbook\_choice/acc | 38.0 | N/A | N/A | N/A | 63.3 | N/A | N/A | N/A | 72.2 |
| | | NIH/Multi-needle | 0 | recall | 75.0 | N/A | N/A | N/A | 84.7 | N/A | N/A | N/A | 98.8 |
| Multilingual | | MGSM (CoT) | 0 | em | 24.5 | 13.7 | 18.2 | 24.4 | 58.2 | 48.9 | 54.3 | 56.8 | 68.9 |
\*\*for comparison purposes only. Model not released.
### Multilingual Benchmarks
| Category | Benchmark | Language | Llama 3.2 1B | Llama 3.2 1B Vanilla PTQ\*\* | Llama 3.2 1B Spin Quant | Llama 3.2 1B QLoRA | Llama 3.2 3B | Llama 3.2 3B Vanilla PTQ\*\* | Llama 3.2 3B Spin Quant | Llama 3.2 3B QLoRA | Llama 3.1 8B |
| :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
| General | MMLU (5-shot, macro_avg/acc) | Portuguese | 39.8 | 34.9 | 38.9 | 40.2 | 54.5 | 50.9 | 53.3 | 53.4 | 62.1 |
| | | Spanish | 41.5 | 36.0 | 39.8 | 41.8 | 55.1 | 51.9 | 53.6 | 53.6 | 62.5 |
| | | Italian | 39.8 | 34.9 | 38.1 | 40.6 | 53.8 | 49.9 | 52.1 | 51.7 | 61.6 |
| | | German | 39.2 | 34.9 | 37.5 | 39.6 | 53.3 | 50.0 | 52.2 | 51.3 | 60.6 |
| | | French | 40.5 | 34.8 | 39.2 | 40.8 | 54.6 | 51.2 | 53.3 | 53.3 | 62.3 |
| | | Hindi | 33.5 | 30.0 | 32.1 | 34.0 | 43.3 | 40.4 | 42.0 | 42.1 | 50.9 |
| | | Thai | 34.7 | 31.2 | 32.4 | 34.9 | 44.5 | 41.3 | 44.0 | 42.2 | 50.3 |
\*\*for comparison purposes only. Model not released.
## Inference time
In the below table, we compare the performance metrics of different quantization methods (SpinQuant and QAT \+ LoRA) with the BF16 baseline. The evaluation was done using the [ExecuTorch](https://github.com/pytorch/executorch) framework as the inference engine, with the ARM CPU as a backend using Android OnePlus 12 device.
| Category | Decode (tokens/sec) | Time-to-first-token (sec) | Prefill (tokens/sec) | Model size (PTE file size in MB) | Memory size (RSS in MB) |
| :---- | ----- | ----- | ----- | ----- | ----- |
| 1B BF16 (baseline) | 19.2 | 1.0 | 60.3 | 2358 | 3,185 |
| 1B SpinQuant | 50.2 (2.6x) | 0.3 (-76.9%) | 260.5 (4.3x) | 1083 (-54.1%) | 1,921 (-39.7%) |
| 1B QLoRA | 45.8 (2.4x) | 0.3 (-76.0%) | 252.0 (4.2x) | 1127 (-52.2%) | 2,255 (-29.2%) |
| 3B BF16 (baseline) | 7.6 | 3.0 | 21.2 | 6129 | 7,419 |
| 3B SpinQuant | 19.7 (2.6x) | 0.7 (-76.4%) | 89.7 (4.2x) | 2435 (-60.3%) | 3,726 (-49.8%) |
| 3B QLoRA | 18.5 (2.4x) | 0.7 (-76.1%) | 88.8 (4.2x) | 2529 (-58.7%) | 4,060 (-45.3%) |
(\*) The performance measurement is done using an adb binary-based approach.
(\*\*) It is measured on an Android OnePlus 12 device.
(\*\*\*) Time-to-first-token (TTFT) is measured with prompt length=64
*Footnote:*
- *Decode (tokens/second) is for how quickly it keeps generating. Higher is better.*
- *Time-to-first-token (TTFT for shorthand) is for how fast it generates the first token for a given prompt. Lower is better.*
- *Prefill is the inverse of TTFT (aka 1/TTFT) in tokens/second. Higher is better*
- *Model size \- how big is the model, measured by, PTE file, a binary file format for ExecuTorch*
- *RSS size \- Memory usage in resident set size (RSS)*
## Responsibility & Safety
As part of our Responsible release approach, we followed a three-pronged strategy to managing trust & safety risks:
1. Enable developers to deploy helpful, safe and flexible experiences for their target audience and for the use cases supported by Llama
2. Protect developers against adversarial users aiming to exploit Llama capabilities to potentially cause harm
3. Provide protections for the community to help prevent the misuse of our models
### Responsible Deployment
**Approach:** Llama is a foundational technology designed to be used in a variety of use cases. Examples on how Meta’s Llama models have been responsibly deployed can be found in our [Community Stories webpage](https://llama.meta.com/community-stories/). Our approach is to build the most helpful models, enabling the world to benefit from the technology power, by aligning our model safety for generic use cases and addressing a standard set of harms. Developers are then in the driver’s seat to tailor safety for their use cases, defining their own policies and deploying the models with the necessary safeguards in their Llama systems. Llama 3.2 was developed following the best practices outlined in our [Responsible Use Guide](https://llama.meta.com/responsible-use-guide/).
#### Llama 3.2 Instruct
**Objective:** Our main objectives for conducting safety fine-tuning are to provide the research community with a valuable resource for studying the robustness of safety fine-tuning, as well as to offer developers a readily available, safe, and powerful model for various applications to reduce the developer workload to deploy safe AI systems. We implemented the same set of safety mitigations as in Llama 3, and you can learn more about these in the Llama 3 [paper](https://ai.meta.com/research/publications/the-llama-3-herd-of-models/).
**Fine-Tuning Data:** We employ a multi-faceted approach to data collection, combining human-generated data from our vendors with synthetic data to mitigate potential safety risks. We’ve developed many large language model (LLM)-based classifiers that enable us to thoughtfully select high-quality prompts and responses, enhancing data quality control.
**Refusals and Tone:** Building on the work we started with Llama 3, we put a great emphasis on model refusals to benign prompts as well as refusal tone. We included both borderline and adversarial prompts in our safety data strategy, and modified our safety data responses to follow tone guidelines.
#### Llama 3.2 Systems
**Safety as a System:** Large language models, including Llama 3.2, **are not designed to be deployed in isolation** but instead should be deployed as part of an overall AI system with additional safety guardrails as required. Developers are expected to deploy system safeguards when building agentic systems. Safeguards are key to achieve the right helpfulness-safety alignment as well as mitigating safety and security risks inherent to the system and any integration of the model or system with external tools. As part of our responsible release approach, we provide the community with [safeguards](https://llama.meta.com/trust-and-safety/) that developers should deploy with Llama models or other LLMs, including Llama Guard, Prompt Guard and Code Shield. All our [reference implementations](https://github.com/meta-llama/llama-agentic-system) demos contain these safeguards by default so developers can benefit from system-level safety out-of-the-box.
### New Capabilities and Use Cases
**Technological Advancement:** Llama releases usually introduce new capabilities that require specific considerations in addition to the best practices that generally apply across all Generative AI use cases. For prior release capabilities also supported by Llama 3.2, see [Llama 3.1 Model Card](https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/MODEL_CARD.md), as the same considerations apply here as well.
**Constrained Environments:** Llama 3.2 1B and 3B models are expected to be deployed in highly constrained environments, such as mobile devices. LLM Systems using smaller models will have a different alignment profile and safety/helpfulness tradeoff than more complex, larger systems. Developers should ensure the safety of their system meets the requirements of their use case. We recommend using lighter system safeguards for such use cases, like Llama Guard 3-1B or its mobile-optimized version.
### Evaluations
**Scaled Evaluations:** We built dedicated, adversarial evaluation datasets and evaluated systems composed of Llama models and Purple Llama safeguards to filter input prompt and output response. It is important to evaluate applications in context, and we recommend building dedicated evaluation dataset for your use case.
**Red Teaming:** We conducted recurring red teaming exercises with the goal of discovering risks via adversarial prompting and we used the learnings to improve our benchmarks and safety tuning datasets. We partnered early with subject-matter experts in critical risk areas to understand the nature of these real-world harms and how such models may lead to unintended harm for society. Based on these conversations, we derived a set of adversarial goals for the red team to attempt to achieve, such as extracting harmful information or reprogramming the model to act in a potentially harmful capacity. The red team consisted of experts in cybersecurity, adversarial machine learning, responsible AI, and integrity in addition to multilingual content specialists with background in integrity issues in specific geographic markets.
### Critical Risks
In addition to our safety work above, we took extra care on measuring and/or mitigating the following critical risk areas:
**1\. CBRNE (Chemical, Biological, Radiological, Nuclear, and Explosive Weapons):** Llama 3.2 1B and 3B models are smaller and less capable derivatives of Llama 3.1. For Llama 3.1 70B and 405B, to assess risks related to proliferation of chemical and biological weapons, we performed uplift testing designed to assess whether use of Llama 3.1 models could meaningfully increase the capabilities of malicious actors to plan or carry out attacks using these types of weapons and have determined that such testing also applies to the smaller 1B and 3B models.
**2\. Child Safety:** Child Safety risk assessments were conducted using a team of experts, to assess the model’s capability to produce outputs that could result in Child Safety risks and inform on any necessary and appropriate risk mitigations via fine tuning. We leveraged those expert red teaming sessions to expand the coverage of our evaluation benchmarks through Llama 3 model development. For Llama 3, we conducted new in-depth sessions using objective based methodologies to assess the model risks along multiple attack vectors including the additional languages Llama 3 is trained on. We also partnered with content specialists to perform red teaming exercises assessing potentially violating content while taking account of market specific nuances or experiences.
**3\. Cyber Attacks:** For Llama 3.1 405B, our cyber attack uplift study investigated whether LLMs can enhance human capabilities in hacking tasks, both in terms of skill level and speed.
Our attack automation study focused on evaluating the capabilities of LLMs when used as autonomous agents in cyber offensive operations, specifically in the context of ransomware attacks. This evaluation was distinct from previous studies that considered LLMs as interactive assistants. The primary objective was to assess whether these models could effectively function as independent agents in executing complex cyber-attacks without human intervention. Because Llama 3.2’s 1B and 3B models are smaller and less capable models than Llama 3.1 405B, we broadly believe that the testing conducted for the 405B model also applies to Llama 3.2 models.
### Community
**Industry Partnerships:** Generative AI safety requires expertise and tooling, and we believe in the strength of the open community to accelerate its progress. We are active members of open consortiums, including the AI Alliance, Partnership on AI and MLCommons, actively contributing to safety standardization and transparency. We encourage the community to adopt taxonomies like the MLCommons Proof of Concept evaluation to facilitate collaboration and transparency on safety and content evaluations. Our Purple Llama tools are open sourced for the community to use and widely distributed across ecosystem partners including cloud service providers. We encourage community contributions to our [Github repository](https://github.com/meta-llama/PurpleLlama).
**Grants:** We also set up the [Llama Impact Grants](https://llama.meta.com/llama-impact-grants/) program to identify and support the most compelling applications of Meta’s Llama model for societal benefit across three categories: education, climate and open innovation. The 20 finalists from the hundreds of applications can be found [here](https://llama.meta.com/llama-impact-grants/#finalists).
**Reporting:** Finally, we put in place a set of resources including an [output reporting mechanism](https://developers.facebook.com/llama_output_feedback) and [bug bounty program](https://www.facebook.com/whitehat) to continuously improve the Llama technology with the help of the community.
## Ethical Considerations and Limitations
**Values:** The core values of Llama 3.2 are openness, inclusivity and helpfulness. It is meant to serve everyone, and to work for a wide range of use cases. It is thus designed to be accessible to people across many different backgrounds, experiences and perspectives. Llama 3.2 addresses users and their needs as they are, without insertion unnecessary judgment or normativity, while reflecting the understanding that even content that may appear problematic in some cases can serve valuable purposes in others. It respects the dignity and autonomy of all users, especially in terms of the values of free thought and expression that power innovation and progress.
**Testing:** Llama 3.2 is a new technology, and like any new technology, there are risks associated with its use. Testing conducted to date has not covered, nor could it cover, all scenarios. For these reasons, as with all LLMs, Llama 3.2’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 3.2 models, developers should perform safety testing and tuning tailored to their specific applications of the model. Please refer to available resources including our [Responsible Use Guide](https://llama.meta.com/responsible-use-guide), [Trust and Safety](https://llama.meta.com/trust-and-safety/) solutions, and other [resources](https://llama.meta.com/docs/get-started/) to learn more about responsible development.
|
Mungert/gemma-3-4b-it-gguf | Mungert | 2025-06-15T19:39:13Z | 1,145 | 12 | null | [
"gguf",
"vision",
"gemma",
"llama.cpp",
"image-text-to-text",
"license:gemma",
"endpoints_compatible",
"region:us",
"conversational"
] | image-text-to-text | 2025-03-12T19:18:29Z | ---
license: gemma
pipeline_tag: image-text-to-text
tags:
- vision
- gemma
- llama.cpp
---
# <span style="color: #7FFF7F;">Gemma-3 4B Instruct GGUF Models</span>
## How to Use Gemma 3 Vision with llama.cpp
To utilize the experimental support for Gemma 3 Vision in `llama.cpp`, follow these steps:
1. **Clone the lastest llama.cpp Repository**:
```bash
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
```
2. **Build the Llama.cpp**:
Build llama.cpp as usual : https://github.com/ggml-org/llama.cpp#building-the-project
Once llama.cpp is built Copy the ./llama.cpp/build/bin/llama-gemma3-cli to a chosen folder.
3. **Download the Gemma 3 gguf file**:
https://huggingface.co/Mungert/gemma-3-4b-it-gguf/tree/main
Choose a gguf file without the mmproj in the name
Example gguf file : https://huggingface.co/Mungert/gemma-3-4b-it-gguf/resolve/main/google_gemma-3-4b-it-q4_k_l.gguf
Copy this file to your chosen folder.
4. **Download the Gemma 3 mmproj file**
https://huggingface.co/Mungert/gemma-3-4b-it-gguf/tree/main
Choose a file with mmproj in the name
Example mmproj file : https://huggingface.co/Mungert/gemma-3-4b-it-gguf/resolve/main/google_gemma-3-4b-it-mmproj-bf16.gguf
Copy this file to your chosen folder.
5. Copy images to the same folder as the gguf files or alter paths appropriately.
In the example below the gguf files, images and llama-gemma-cli are in the same folder.
Example image: image https://huggingface.co/Mungert/gemma-3-4b-it-gguf/resolve/main/car-1.jpg
Copy this file to your chosen folder.
6. **Run the CLI Tool**:
From your chosen folder :
```bash
llama-gemma3-cli -m google_gemma-3-4b-it-q4_k_l.gguf --mmproj google_gemma-3-4b-it-mmproj-bf16.gguf
```
```
Running in chat mode, available commands:
/image <path> load an image
/clear clear the chat history
/quit or /exit exit the program
> /image car-1.jpg
Encoding image car-1.jpg
Image encoded in 46305 ms
Image decoded in 19302 ms
> what is the image of
Here's a breakdown of what's in the image:
**Subject:** The primary subject is a black Porsche Panamera Turbo driving on a highway.
**Details:**
* **Car:** It's a sleek, modern Porsche Panamera Turbo, identifiable by its distinctive rear design, the "PORSCHE" lettering, and the "Panamera Turbo" badge. The license plate reads "CVC-911".
* **Setting:** The car is on a multi-lane highway, with a blurred background of trees, a distant building, and a cloudy sky. The lighting suggests it's either dusk or dawn.
* **Motion:** The image captures the car in motion, with a slight motion blur to convey speed.
**Overall Impression:** The image conveys a sense of speed, luxury, and power. It's a well-composed shot that highlights the car's design and performance.
Do you want me to describe any specific aspect of the image in more detail, or perhaps analyze its composition?
```
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤️ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs **Phi-4-mini-instruct** using phi-4-mini-q4_0.gguf , llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limtations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Low | Very Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium Low | Low | CPU with more memory | Better accuracy while still being quantized |
| **Q8** | Medium | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
## **Included Files & Details**
### `google_gemma-3-4b-it-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `google_gemma-3-4b-it-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `google_gemma-3-4b-it-bf16-q8.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `google_gemma-3-4b-it-f16-q8.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `google_gemma-3-4b-it-q4_k_l.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `google_gemma-3-4b-it-q4_k_m.gguf`
- Similar to Q4_K.
- Another option for **low-memory CPU inference**.
### `google_gemma-3-4b-it-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `google_gemma-3-4b-it-q6_k_l.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `google_gemma-3-4b-it-q6_k_m.gguf`
- A mid-range **Q6_K** quantized model for balanced performance .
- Suitable for **CPU-based inference** with **moderate memory**.
### `google_gemma-3-4b-it-q8.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
# Gemma 3 model card
**Model Page**: [Gemma](https://ai.google.dev/gemma/docs/core)
**Resources and Technical Documentation**:
* [Gemma 3 Technical Report][g3-tech-report]
* [Responsible Generative AI Toolkit][rai-toolkit]
* [Gemma on Kaggle][kaggle-gemma]
* [Gemma on Vertex Model Garden][vertex-mg-gemma3]
**Terms of Use**: [Terms][terms]
**Authors**: Google DeepMind
## Model Information
Summary description and brief definition of inputs and outputs.
### Description
Gemma is a family of lightweight, state-of-the-art open models from Google,
built from the same research and technology used to create the Gemini models.
Gemma 3 models are multimodal, handling text and image input and generating text
output, with open weights for both pre-trained variants and instruction-tuned
variants. Gemma 3 has a large, 128K context window, multilingual support in over
140 languages, and is available in more sizes than previous versions. Gemma 3
models are well-suited for a variety of text generation and image understanding
tasks, including question answering, summarization, and reasoning. Their
relatively small size makes it possible to deploy them in environments with
limited resources such as laptops, desktops or your own cloud infrastructure,
democratizing access to state of the art AI models and helping foster innovation
for everyone.
### Inputs and outputs
- **Input:**
- Text string, such as a question, a prompt, or a document to be summarized
- Images, normalized to 896 x 896 resolution and encoded to 256 tokens
each
- Total input context of 128K tokens for the 4B, 12B, and 27B sizes, and
32K tokens for the 1B size
- **Output:**
- Generated text in response to the input, such as an answer to a
question, analysis of image content, or a summary of a document
- Total output context of 8192 tokens
## Credits
Thanks [Bartowski](https://huggingface.co/bartowski) for imartix upload. And your guidance on quantization that has enabled me to produce these gguf file. |
Mungert/Phi-4-mini-instruct.gguf | Mungert | 2025-06-15T19:39:04Z | 4,816 | 25 | null | [
"gguf",
"Demo",
"Phi 4 Mini",
"multilingual",
"reasoning",
"code generation",
"function calling",
"chat completion",
"memory efficient",
"low latency",
"128k context",
"text-generation",
"base_model:microsoft/Phi-4-mini-instruct",
"base_model:quantized:microsoft/Phi-4-mini-instruct",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-02-28T01:14:03Z | ---
license: mit
base_model:
- microsoft/Phi-4-mini-instruct
pipeline_tag: text-generation
tags:
- Demo
- Phi 4 Mini
- multilingual
- reasoning
- code generation
- function calling
- chat completion
- memory efficient
- low latency
- 128k context
---
# Model Summary
Phi-4-mini-instruct is a lightweight open model built upon synthetic data and filtered publicly available websites - with a focus on high-quality, reasoning dense data. The model belongs to the Phi-4 model family and supports 128K token context length. The model underwent an enhancement process, incorporating both supervised fine-tuning and direct preference optimization to support precise instruction adherence and robust safety measures.
📰 [Phi-4-mini Microsoft Blog](https://aka.ms/phi4-feb2025) <br>
📖 [Phi-4-mini Technical Report](https://aka.ms/phi-4-multimodal/techreport) <br>
👩🍳 [Phi Cookbook](https://github.com/microsoft/PhiCookBook) <br>
🏡 [Phi Portal](https://azure.microsoft.com/en-us/products/phi) <br>
🖥️ Try It [Azure](https://aka.ms/phi-4-mini/azure), [Huggingface](https://huggingface.co/spaces/microsoft/phi-4-mini) <br>
**Phi-4**:
[[mini-instruct](https://huggingface.co/microsoft/Phi-4-mini-instruct) | [onnx](https://huggingface.co/microsoft/Phi-4-mini-instruct-onnx)];
[multimodal-instruct](https://huggingface.co/microsoft/Phi-4-multimodal-instruct); [gguf](https://huggingface.co/Mungert/Phi-4-mini-instruct.gguf)
## Usage
### Chat format
This format is used for general conversation and instructions:
```yaml
<|system|>Insert System Message<|end|><|user|>Insert User Message<|end|><|assistant|>
```
### Tool-Enabled Function-Calling Format
This format is used when the user wants the model to provide function calls based on the given tools. The user should define the available tools in the system prompt, wrapped by `<|tool|>` and `<|/tool|>` tokens. The tools must be specified in JSON format using a structured JSON dump.
```plaintext
<|system|>
You are a helpful assistant with some tools.
<|tool|>
[
{
"name": "get_weather_updates",
"description": "Fetches weather updates for a given city using the RapidAPI Weather API.",
"parameters": {
"city": {
"description": "The name of the city for which to retrieve weather information.",
"type": "str",
"default": "London"
}
}
}
]
<|/tool|>
<|end|>
<|user|>
What is the weather like in Paris today?
<|end|>
<|assistant|>
```
---
# <span style="color: #FF7F7F;">Unsloth Bug Fixes for Better Performance</span>
***Update (March 1, 2025)***
Applying all [Unsloth](https://huggingface.co/unsloth) fixes improved inference stability.
| # | Fix | Reason for Fix |
|----|--------------------------------------------------|------------------------------------------------|
| 1 | Changed the padding tag | The old padding tag could cause training issues. |
| 2 | Removed `{% else %}{{ eos_token }}` from chat template | Prevented extra EOS tokens that could degrade inference performance. |
| 3 | Replaced EOS with <\|end\|> | Avoided potential inference glitches. |
| 4 | Changed `unk_token` from EOS to `�` | Stopped unknown tokens from breaking inference. |
# <span id="testllm" style="color: #7F7FFF;">🚀 Phi 4 Mini Function Calling Test!</span>
If you have a minute, I’d really appreciate it if you could test my Phi-4-Mini-Instruct Demo at 👉 [Quantum Network Monitor](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Then toggle between the LLM Types Phi-4-Mini-Instruct is called TestLLM : TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs **Phi-4-mini-instruct** using phi-4-mini-q4_0.gguf , llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
# <span style="color: #7FFF7F;">Phi-4-mini-instruct GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limtations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Low | Very Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium Low | Low | CPU with more memory | Better accuracy while still being quantized |
| **Q8** | Medium | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
## **Included Files & Details**
### `phi-4-mini-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `phi-4-mini-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `phi-4-mini-bf16-q8.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `phi-4-mini-f16-q8.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `phi-4-mini-q4_k_l.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `phi-4-mini-q4_k_m.gguf`
- Similar to Q4_K.
- Another option for **low-memory CPU inference**.
### `phi-4-mini-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `phi-4-mini-q6_k_l.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `phi-4-mini-q6_k_m.gguf`
- A mid-range **Q6_K** quantized model for balanced performance .
- Suitable for **CPU-based inference** with **moderate memory**.
### `phi-4-mini-q8.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
## Credits
Thanks [Bartowski](https://huggingface.co/bartowski) for imartix upload. And your guidance on quantization that has enabled me to produce these gguf file.
Thanks [Unsloth](https://huggingface.co/unsloth) for bug fixing many models.
Thanks for your support! 🙌 |
Mungert/Apriel-Nemotron-15b-Thinker-GGUF | Mungert | 2025-06-15T19:39:01Z | 1,097 | 1 | transformers | [
"transformers",
"gguf",
"text-generation",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-06-12T17:38:31Z | ---
license: mit
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">Apriel-Nemotron-15b-Thinker GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`1f63e75f`](https://github.com/ggerganov/llama.cpp/commit/1f63e75f3b5dc7f44dbe63c8a41d23958fe95bc0).
---
## <span style="color: #7FFF7F;">Quantization Beyond the IMatrix</span>
I've been experimenting with a new quantization approach that selectively elevates the precision of key layers beyond what the default IMatrix configuration provides.
In my testing, standard IMatrix quantization underperforms at lower bit depths, especially with Mixture of Experts (MoE) models. To address this, I'm using the `--tensor-type` option in `llama.cpp` to manually "bump" important layers to higher precision. You can see the implementation here:
👉 [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
While this does increase model file size, it significantly improves precision for a given quantization level.
### **I'd love your feedback—have you tried this? How does it perform for you?**
---
<a href="https://readyforquantum.com/huggingface_gguf_selection_guide.html" style="color: #7FFF7F;">
Click here to get info on choosing the right GGUF model format
</a>
---
<!--Begin Original Model Card-->
# Apriel-Nemotron-15b-Thinker
<img src="https://cdn-uploads.huggingface.co/production/uploads/63d3095c2727d7888cbb54e2/Lt1t0tOO5emz1X23Azg-E.png" width="120" alt="thumbnail"/> `/ˈɑː.pri.əl/`
---
## Table of Contents
1. [Summary](#summary)
2. [Evaluation](#evaluation)
3. [Training Details](#training-details)
4. [How to Use](#how-to-use)
5. [Intended Use](#intended-use)
6. [Limitations](#limitations)
7. [Security and Responsible Use](#security-and-responsible-use)
8. [Software](#software)
9. [License](#license)
10. [Acknowledgements](#acknowledgements)
11. [Citation](#citation)
---
## Summary
**Apriel-Nemotron-15b-Thinker** is a 15 billion‑parameter reasoning model in ServiceNow’s Apriel SLM series which achieves competitive performance against similarly sized state-of-the-art models like o1‑mini, QWQ‑32b, and EXAONE‑Deep‑32b, all while maintaining only half the memory footprint of those alternatives. It builds upon the **Apriel‑15b‑base** checkpoint through a three‑stage training pipeline (CPT, SFT and GRPO).
**Highlights**
- Half the size of SOTA models like QWQ-32b and EXAONE-32b and hence **memory efficient**.
- It consumes **40%** less tokens compared to QWQ-32b, making it super efficient in production. 🚀🚀🚀
- On par or outperforms on tasks like - MBPP, BFCL, Enterprise RAG, MT Bench, MixEval, IFEval and Multi-Challenge making it great for **Agentic / Enterprise tasks**.
- Competitive performance on academic benchmarks like AIME-24 AIME-25, AMC-23, MATH-500 and GPQA considering model size.
---
## Evaluation
Evaluations were conducted using [lm-eval-harness](https://github.com/EleutherAI/lm-evaluation-harness) and [evalchemy](https://github.com/mlfoundations/evalchemy).
### Benchmarks that are indicative of enterprise capability
### Academic reasoning benchmarks
### Token efficiency comparison (lower the better)
---
## Training Details
**Mid training / Continual Pre‑training**
In this stage, the model is trained on 100+ billion tokens of carefully curated examples drawn from mathematical reasoning, coding challenges, scientific discourse and logical puzzles. The objective is to strengthen foundational reasoning capabilities of the model. This stage is super critical for the model to function as a reasoner and provides significant lifts in reasoning benchmarks.
**Supervised Fine‑Tuning (SFT)**
Next, we SFT the model using 200,000 high‑quality demonstrations that cover mathematical and scientific problem‑solving, coding tasks, generic instruction‑following scenarios, API/function invocation use cases etc.
**Reinforcement Learning**
Although the SFT‑tuned checkpoint delivers strong performance on core competencies like mathematics and general knowledge, it exhibits weaknesses in instruction following and coding tasks. To address these gaps, we apply GRPO (with some minor modifications to the objective). The result is significant improvement on benchmarks such as IFEval, Multi Challenge, Enterprise RAG, MBPP and BFCL, while preserving scores on competition‑level math exams like AIME and AMC. GRPO also yields modest gains on GPQA and MixEval.
Throughout training, intermediate snapshots from both the SFT and GRPO stages are periodically merged, improving generalization and catastrophic forgetting.
*Technical report with more details - coming soon.*
---
## How to Use
```bash
pip install transformers
```
---
### Running the Reasoning model
Here is a code snippet demonstrating the model's usage with the transformers library's generate function:
```python
import re
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "ServiceNow-AI/Apriel-Nemotron-15b-Thinker"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Positive real numbers $x$ and $y$ satisfy $y^3=x^2$ and $(y-x)^2=4y^2$. What is $x+y$?\nMark your solution with \\boxed"
messages = [
{"role": "user", "content": prompt}
]
tools = []
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
tools=tools
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=65536
)
output = tokenizer.decode(generated_ids[0], skip_special_tokens=True)
# parsing the response
response = re.findall(r"\[BEGIN FINAL RESPONSE\](.*?)\[END FINAL RESPONSE\]", output, re.DOTALL)[0].strip()
print("output:", output)
print("response:", response)
```
---
### Chat Template
```
<|system|>
You are a thoughtful and systematic AI assistant built by ServiceNow Language Models (SLAM) lab. Before providing an answer, analyze the problem carefully and present your reasoning step by step. After explaining your thought process, provide the final solution in the following format: [BEGIN FINAL RESPONSE] ... [END FINAL RESPONSE].
<|end|>
<|user|>
# user message here
<|end|>
<|assistant|>
Here are my reasoning steps:
# thoughts here
[BEGIN FINAL RESPONSE]
# assistant response here
[END FINAL RESPONSE]
<|end|>
```
The model will first generate its thinking process and then generate its final response between `[BEGIN FINAL RESPONSE]` and `[END FINAL RESPONSE]`. Here is a code snippet demonstrating the application of the chat template:
```python
from transformers import AutoTokenizer
model_name = "ServiceNow-AI/Apriel-Nemotron-15b-Thinker"
tokenizer = AutoTokenizer.from_pretrained(model_name)
# prepare the model input
custom_system_prompt = "Answer like a pirate."
prompt = "You are an expert assistant in the implementation of customer experience management aspect of retail applications \n \nYou will be using Python as the programming language. \n \nYou will utilize a factory design pattern for the implementation and following the dependency inversion principle \n \nYou will modify the implementation based on user requirements. \n \nUpon user request, you will add, update, and remove the features & enhancements in the implementation provided by you. \n \nYou will ask whether the user wants to refactor the provided code or needs a sample implementation for reference. Upon user confirmation, I will proceed accordingly. \n \n**Guidelines:** \n 1. **User Requirements:** \n - You have to ask users about their requirements, clarify the user expectations, and suggest the best possible solution by providing examples of Python code snippets. \n - Ask users about which type of reports they need to assess the AI model's performance, accuracy, and reliability. \n - After providing the solution, you have to ask the user about the trial of the solution and modify the solution based on the user feedback. \n \n 2. **Libraries/Frameworks:** \n - You will be utilizing Python as a programming language. \n - You will be using Flask framework for REST APIS implementation \n \n 3. **Communication Gesture:** \n - Your conversation with the user should be interactive, supportive, courageous, and professional. \n - You have to break down the complex concepts into sub-concepts and try to explain them to the user. \n - You have to ask the user for the required parameters. If the user refuses to provide in 2 attempts, politely exit the conversation. \n - You have to provide your supported parameters to the user, if the user refuses to accept them then you have to put an apology note and exit the conversation. \n - You have to track the conversation about unasked questions by the user. If some/one of the questions remain then you have to remind the user about these questions and proceed to answer them based on the user's confirmation \n \n 4. **Implementation:** \n - Your code/implementations should be reliable, scaleable, modular, and reusable. \n - You will be providing unit tests for the implementation upon user request. \n - You will be following MVC architecture for the applications \n - Your implementations must be well-commented and readable \n \n \n- Today's date is 23rd August 2024. \n- The default sender email is [email protected].\nHi, I am conducting research on retail customer feedback systems and I need assistance with designing and implementing them. Could you kindly provide me with a list of general customer feedback system modules?"
messages = [
{"role": "user", "content": custom_system_prompt + "\n\n" + prompt}
]
# example tools
tools = [{"type": "function", "function": {"name": "getRetailFeedbackModules", "description": "Returns the list of modules usually present in the retail industry", "parameters": {"type": "object", "properties": {"page": {"type": "integer", "description": "The current page number.", "default": 1}, "page_size": {"type": "integer", "description": "The number of items per page.", "default": 3}}}}}, {"type": "function", "function": {"name": "verifyImplementation", "description": "Returns the list of modules usually present in the retail industry", "parameters": {"type": "object", "properties": {"coding_language": {"type": "string", "description": "The supported languages for verification of implementation.", "default": "python", "enum": ["python", "java", "php"]}, "code": {"type": "string", "description": "The code which needs verification"}, "design_pattern": {"type": "string", "description": "The design pattern to verify in the implementation", "enum": ["factory", "strategy", "singleton"]}, "verify_best_practices": {"type": "boolean", "description": "The verification of the coding style based on the language selected", "default": true}}}}}]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
tools=tools
)
model_inputs = tokenizer([text], return_tensors="pt")
```
### Usage Guidelines
1. Use the model’s default chat template, which already includes a system prompt. We recommend adding all other instructions within the user message.
2. We recommend setting temperature to `0.6`.
3. We ensure the model starts with `Here are my reasoning steps:\n` during all our evaluations. This is implemented in the default chat template.
---
## Intended Use
The Apriel family of models are designed for a variety of general-purpose instruction tasks, including:
- Code assistance and generation
- Logical reasoning and multi-step tasks
- Question answering and information retrieval
- Function calling, complex instruction following and agent use cases
They are **not intended** for use in safety-critical applications without human oversight or in scenarios requiring guaranteed factual accuracy.
---
## Limitations
- **Factual accuracy:** May produce incorrect, misleading, or outdated content. Outputs should be verified before use in critical contexts.
- **Bias:** May reflect societal, cultural, or systemic biases present in training data.
- **Ethics:** Do not use the model to produce harmful, unlawful, or unethical content.
- **Language:** Strongest performance is in English. Output quality may degrade in underrepresented languages.
- **Critical use:** Not suitable for medical, legal, financial, or other high-risk applications without safeguards.
---
## Security and Responsible Use
**Security Responsibilities:**
Deployers and users are strongly encouraged to align their security practices with established frameworks and regulatory guidelines such as the EU AI Act and the NIST AI Risk Management Framework (RMF).
**Guidelines for Deployers:**
- Regularly conduct robustness assessments to identify and mitigate adversarial inputs.
- Implement validation and filtering processes to prevent harmful or biased outputs.
- Continuously perform data privacy checks to guard against unintended data leaks.
- Document and communicate the model's limitations, intended usage, and known security risks to all end-users.
- Schedule periodic security reviews and updates to address emerging threats and vulnerabilities.
**Guidelines for Users:**
- Follow established security policies and usage guidelines provided by deployers.
- Protect and manage sensitive information when interacting with the model.
- Report anomalies, suspicious behavior, or unsafe outputs to deployers or developers.
- Maintain human oversight and apply judgment to mitigate potential security or ethical risks during interactions.
**Disclaimer:**
Users accept responsibility for securely deploying, managing, and using this open-source LLM. The model is provided "as-is," without explicit or implied warranty regarding security or fitness for any specific application or environment.
---
## Software
- **Training stack:** [Fast-LLM](https://github.com/ServiceNow/Fast-LLM)
---
## License
MIT
---
## Acknowledgments
>We thank researchers at Nvidia for sharing detailed insights and data from their work in building reasoners! This greatly accelerated our research and we recognize the same with our model naming convention!
## Citation
```bibtex
@misc{Apriel-nemotron-15b-thinker,
author = {Slam labs team},
title = {Apriel Nemotron 15b Thinker},
howpublished = {https://huggingface.co/ServiceNow-AI/Apriel-Nemotron-15b-Thinker},
publisher = {SLAM - ServiceNow Language Models Lab}
year = {2025}
}
```
<!--End Original Model Card-->
---
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/nomic-embed-code-GGUF | Mungert | 2025-06-15T19:38:55Z | 1,464 | 1 | sentence-transformers | [
"sentence-transformers",
"gguf",
"sentence-similarity",
"feature-extraction",
"dataset:nomic-ai/cornstack-python-v1",
"dataset:nomic-ai/cornstack-javascript-v1",
"dataset:nomic-ai/cornstack-java-v1",
"dataset:nomic-ai/cornstack-go-v1",
"dataset:nomic-ai/cornstack-php-v1",
"dataset:nomic-ai/cornstack-ruby-v1",
"arxiv:2412.01007",
"base_model:Qwen/Qwen2.5-Coder-7B-Instruct",
"base_model:quantized:Qwen/Qwen2.5-Coder-7B-Instruct",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | sentence-similarity | 2025-06-10T15:04:32Z | ---
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
license: apache-2.0
datasets:
- nomic-ai/cornstack-python-v1
- nomic-ai/cornstack-javascript-v1
- nomic-ai/cornstack-java-v1
- nomic-ai/cornstack-go-v1
- nomic-ai/cornstack-php-v1
- nomic-ai/cornstack-ruby-v1
base_model:
- Qwen/Qwen2.5-Coder-7B-Instruct
---
# <span style="color: #7FFF7F;">nomic-embed-code GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`7f4fbe51`](https://github.com/ggerganov/llama.cpp/commit/7f4fbe5183b23b6b2e25fd1ccc5d1fa8bb010cb7).
## <span style="color: #7FFF7F;"> Quantization beyond the IMatrix</span>
Testing a new quantization method using rules to bump important layers above what the standard imatrix would use.
I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
This does create larger model files but increases precision for a given model size.
### **Please provide feedback on how you find this method performs**
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
# Nomic Embed Code: A State-of-the-Art Code Retriever
[Blog](https://www.nomic.ai/blog/posts/introducing-state-of-the-art-nomic-embed-code) | [Technical Report](https://arxiv.org/abs/2412.01007) | [AWS SageMaker](https://aws.amazon.com/marketplace/seller-profile?id=seller-tpqidcj54zawi) | [Atlas Embedding and Unstructured Data Analytics Platform](https://atlas.nomic.ai)
`nomic-embed-code` is a state-of-the-art code embedding model that excels at code retrieval tasks:
- **High Performance**: Outperforms Voyage Code 3 and OpenAI Embed 3 Large on CodeSearchNet
- **Multilingual Code Support**: Trained for multiple programming languages (Python, Java, Ruby, PHP, JavaScript, Go)
- **Advanced Architecture**: 7B parameter code embedding model
- **Fully Open-Source**: Model weights, training data, and [evaluation code](https://github.com/gangiswag/cornstack/) released
| Model | Python | Java | Ruby | PHP | JavaScript | Go |
|-------|--------|------|------|-----|------------|-----|
| **Nomic Embed Code** | **81.7** | **80.5** | 81.8 | **72.3** | 77.1 | **93.8** |
| Voyage Code 3 | 80.8 | **80.5** | **84.6** | 71.7 | **79.2** | 93.2 |
| OpenAI Embed 3 Large | 70.8 | 72.9 | 75.3 | 59.6 | 68.1 | 87.6 |
| Nomic CodeRankEmbed-137M | 78.4 | 76.9 | 79.3 | 68.8 | 71.4 | 92.7 |
| CodeSage Large v2 (1B) | 74.2 | 72.3 | 76.7 | 65.2 | 72.5 | 84.6 |
| CodeSage Large (1B) | 70.8 | 70.2 | 71.9 | 61.3 | 69.5 | 83.7 |
| Qodo Embed 1 7B | 59.9 | 61.6 | 68.4 | 48.5 | 57.0 | 81.4 |
## Model Architecture
- **Total Parameters**: 7B
- **Training Approach**: Trained on the CoRNStack dataset with dual-consistency filtering and progressive hard negative mining
- **Supported Languages**: Python, Java, Ruby, PHP, JavaScript, and Go
## Usage Guide
### Installation
You can install the necessary dependencies with:
```bash
pip install transformers sentence-transformers torch
```
### Transformers
```python
import torch
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained("nomic-ai/nomic-embed-code")
model = AutoModel.from_pretrained("nomic-ai/nomic-embed-code")
def last_token_pooling(hidden_states, attention_mask):
sequence_lengths = attention_mask.sum(-1) - 1
return hidden_states[torch.arange(hidden_states.shape[0]), sequence_lengths]
queries = ['Represent this query for searching relevant code: Calculate the n-th factorial']
codes = ['def fact(n):\n if n < 0:\n raise ValueError\n return 1 if n == 0 else n * fact(n - 1)']
code_snippets = queries + codes
encoded_input = tokenizer(code_snippets, padding=True, truncation=True, return_tensors='pt')
model.eval()
with torch.no_grad():
model_output = model(**encoded_input)[0]
embeddings = last_token_pooling(model_output, encoded_input['attention_mask'])
embeddings = F.normalize(embeddings, p=2, dim=1)
print(embeddings.shape)
similarity = F.cosine_similarity(embeddings[0], embeddings[1], dim=0)
print(similarity)
```
### SentenceTransformers
```python
from sentence_transformers import SentenceTransformer
queries = ['Calculate the n-th factorial']
code_snippets = ['def fact(n):\n if n < 0:\n raise ValueError\n return 1 if n == 0 else n * fact(n - 1)']
model = SentenceTransformer("nomic-ai/nomic-embed-code")
query_emb = model.encode(queries, prompt_name="query")
code_emb = model.encode(code_snippets)
similarity = model.similarity(query_emb[0], code_emb[0])
print(similarity)
```
### CoRNStack Dataset Curation
Starting with the deduplicated Stackv2, we create text-code pairs from function docstrings and respective code. We filtered out low-quality pairs where the docstring wasn't English, too short, or that contained URLs, HTML tags, or invalid characters. We additionally kept docstrings with text lengths of 256 tokens or longer to help the model learn long-range dependencies.
After the initial filtering, we used dual-consistency filtering to remove potentially noisy examples. We embed each docstring and code pair and compute the similarity between each docstring and every code example. We remove pairs from the dataset if the corresponding code example is not found in the top-2 most similar examples for a given docstring.
During training, we employ a novel curriculum-based hard negative mining strategy to ensure the model learns from challenging examples. We use a softmax-based sampling strategy to progressively sample hard negatives with increasing difficulty over time.
## Join the Nomic Community
- Nomic Embed Ecosystem: [https://www.nomic.ai/embed](https://www.nomic.ai/embed)
- Website: [https://nomic.ai](https://nomic.ai)
- Twitter: [https://twitter.com/nomic_ai](https://twitter.com/nomic_ai)
- Discord: [https://discord.gg/myY5YDR8z8](https://discord.gg/myY5YDR8z8)
# Citation
If you find the model, dataset, or training code useful, please cite our work:
```bibtex
@misc{suresh2025cornstackhighqualitycontrastivedata,
title={CoRNStack: High-Quality Contrastive Data for Better Code Retrieval and Reranking},
author={Tarun Suresh and Revanth Gangi Reddy and Yifei Xu and Zach Nussbaum and Andriy Mulyar and Brandon Duderstadt and Heng Ji},
year={2025},
eprint={2412.01007},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2412.01007},
}
```
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Peacemann/deepseek-ai_DeepSeek-R1-0528-Qwen3-8B_LMUL | Peacemann | 2025-06-15T19:38:55Z | 0 | 0 | null | [
"qwen3",
"text-generation",
"lmul",
"research",
"experimental",
"conversational",
"base_model:deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
"base_model:finetune:deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
"license:mit",
"region:us"
] | text-generation | 2025-06-15T19:34:33Z | ---
license: mit
base_model: deepseek-ai/DeepSeek-R1-0528-Qwen3-8B
tags:
- text-generation
- lmul
- research
- experimental
- qwen3
---
# L-Mul Optimized: deepseek-ai/DeepSeek-R1-0528-Qwen3-8B
This is a modified version of DeepSeek AI's [DeepSeek-R1-0528-Qwen3-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B) model. The modification consists of replacing the standard attention mechanism with one that uses a custom, approximate matrix multiplication algorithm termed "L-Mul".
This work was performed as part of a research project to evaluate the performance and accuracy trade-offs of algorithmic substitutions in transformer architectures.
**This model is intended strictly for educational and scientific purposes.**
## Model Description
The core architecture of `deepseek-ai/DeepSeek-R1-0528-Qwen3-8B` is preserved. However, the standard `Qwen3Attention` modules have been dynamically replaced with a custom version that utilizes the `l_mul_attention` function for its core computations. This function is defined in the `lmul.py` file included in this repository.
- **Base Model:** [deepseek-ai/DeepSeek-R1-0528-Qwen3-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B)
- **Modification:** Replacement of standard attention with L-Mul approximate attention.
- **Primary Use-Case:** Research and educational analysis of algorithmic impact on LLMs.
## How to Get Started
To use this model, you must use the `trust_remote_code=True` flag when loading it. This is required to execute the custom `lmul.py` file that defines the new attention mechanism.
You can load the model directly from this repository using the `transformers` library:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
# Define the repository ID for the specific model
repo_id = "Peacemann/deepseek-ai_DeepSeek-R1-0528-Qwen3-8B_LMUL" # Replace with the correct repo ID if different
# Load the tokenizer and model, trusting the remote code to load lmul.py
tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(
repo_id,
trust_remote_code=True,
torch_dtype=torch.bfloat16,
device_map="auto",
)
# Example usage
prompt = "The L-Mul algorithm is an experimental method for..."
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```
For high-throughput inference, you can use `vLLM`:
```python
from vllm import LLM
repo_id = "Peacemann/deepseek-ai_DeepSeek-R1-0528-Qwen3-8B_LMUL" # Replace with the correct repo ID
llm = LLM(model=repo_id, trust_remote_code=True)
```
## Intended Uses & Limitations
This model is intended for researchers and students exploring the internal workings of LLMs. It is a tool for visualizing and analyzing the effects of fundamental algorithmic changes.
**This model is NOT intended for any commercial or production application.**
The modification is experimental. The impact on the model's performance, safety alignment, accuracy, and potential for generating biased or harmful content is **unknown and untested**. It inherits all limitations and biases of the original `DeepSeek-R1-0528-Qwen3-8B` model, and its behavior may be altered in unpredictable ways.
## Licensing Information
The use of this model is subject to the original **MIT License**. By using this model, you agree to the terms outlined in the license. The license can be found on the base model's Hugging Face page. |
Mungert/Seed-Coder-8B-Reasoning-GGUF | Mungert | 2025-06-15T19:38:52Z | 1,113 | 1 | transformers | [
"transformers",
"gguf",
"text-generation",
"arxiv:2506.03524",
"base_model:ByteDance-Seed/Seed-Coder-8B-Base",
"base_model:quantized:ByteDance-Seed/Seed-Coder-8B-Base",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-06-08T17:27:28Z | ---
library_name: transformers
pipeline_tag: text-generation
license: mit
base_model:
- ByteDance-Seed/Seed-Coder-8B-Base
---
# <span style="color: #7FFF7F;">Seed-Coder-8B-Reasoning GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`5787b5da`](https://github.com/ggerganov/llama.cpp/commit/5787b5da57e54dba760c2deeac1edf892e8fc450).
## <span style="color: #7FFF7F;"> Quantization beyond the IMatrix</span>
Testing a new quantization method using rules to bump important layers above what the standard imatrix would use.
I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
This does create larger model files but increases precision for a given model size.
### **Please provide feedback on how you find this method performs**
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
# Seed-Coder-8B-Reasoning
<div align="left" style="line-height: 1;">
<a href="https://bytedance-seed-coder.github.io/" target="_blank" style="margin: 2px;">
<img alt="Homepage" src="https://img.shields.io/badge/Seed--Coder-Homepage-a468fe?color=a468fe&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://arxiv.org/abs/2506.03524" target="_blank" style="margin: 2px;">
<img alt="Technical Report" src="https://img.shields.io/badge/arXiv-Technical%20Report-brightgreen?logo=arxiv&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://huggingface.co/ByteDance-Seed" target="_blank" style="margin: 2px;">
<img alt="Hugging Face" src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-ByteDance%20Seed-536af5?color=536af5&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://github.com/ByteDance-Seed/Seed-Coder/blob/master/LICENSE" style="margin: 2px;">
<img alt="License" src="https://img.shields.io/badge/License-MIT-f5de53?color=f5de53&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
## Introduction
We are thrilled to introduce Seed-Coder, a powerful, transparent, and parameter-efficient family of open-source code models at the 8B scale, featuring base, instruct, and reasoning variants. Seed-Coder contributes to promote the evolution of open code models through the following highlights.
- **Model-centric:** Seed-Coder predominantly leverages LLMs instead of hand-crafted rules for code data filtering, minimizing manual effort in pretraining data construction.
- **Transparent:** We openly share detailed insights into our model-centric data pipeline, including methods for curating GitHub data, commits data, and code-related web data.
- **Powerful:** Seed-Coder achieves state-of-the-art performance among open-source models of comparable size across a diverse range of coding tasks.
<p align="center">
<img width="100%" src="imgs/seed-coder_intro_performance.png">
</p>
This repo contains the **Seed-Coder-8B-Reasoning** model, which has the following features:
- Type: Causal language models
- Training Stage: Pretraining & Post-training
- Data Source: Public datasets
- Context Length: 65,536
## Model Downloads
| Model Name | Length | Download | Notes |
|---------------------------------------------------------|-----------|------------------------------------|-----------------------|
| Seed-Coder-8B-Base | 32K | 🤗 [Model](https://huggingface.co/ByteDance-Seed/Seed-Coder-8B-Base) | Pretrained on our model-centric code data. |
| Seed-Coder-8B-Instruct | 32K | 🤗 [Model](https://huggingface.co/ByteDance-Seed/Seed-Coder-8B-Instruct) | Instruction-tuned for alignment with user intent. |
| 👉 **Seed-Coder-8B-Reasoning** | 64K | 🤗 [Model](https://huggingface.co/ByteDance-Seed/Seed-Coder-8B-Reasoning) | RL trained to boost reasoning capabilities. |
| Seed-Coder-8B-Reasoning-bf16 | 64K | 🤗 [Model](https://huggingface.co/ByteDance-Seed/Seed-Coder-8B-Reasoning-bf16) | RL trained to boost reasoning capabilities. |
## Requirements
You will need to install the latest versions of `transformers` and `accelerate`:
```bash
pip install -U transformers accelerate
```
## Quickstart
Here is a simple example demonstrating how to load the model and perform code generation using the Hugging Face `pipeline` API:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
model_id = "ByteDance-Seed/Seed-Coder-8B-Reasoning"
tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.bfloat16, device_map="auto", trust_remote_code=True)
messages = [
{"role": "user", "content": "Write a quick sort algorithm."},
]
input_ids = tokenizer.apply_chat_template(
messages,
tokenize=True,
return_tensors="pt",
add_generation_prompt=True,
).to(model.device)
outputs = model.generate(input_ids, max_new_tokens=16384)
response = tokenizer.decode(outputs[0][input_ids.shape[-1]:], skip_special_tokens=True)
print(response)
```
## Evaluation
Seed-Coder-8B-Reasoning strikes impressive performance on competitive programming, demonstrating that smaller LLMs can also be competent on complex reasoning tasks. Our model surpasses QwQ-32B and DeepSeek-R1 on IOI'2024, and achieves an ELO rating comparable to o1-mini on Codeforces contests.
<div style="display: flex; justify-content: center;">
<img src="imgs/reasoning-ioi.jpg" width="61%" />
<img src="imgs/reasoning-codeforces.jpg" width="39%" />
</div>
For detailed benchmark performance, please refer to our [📑 Technical Report](https://github.com/ByteDance-Seed/Seed-Coder/blob/master/Seed-Coder.pdf).
## License
This project is licensed under the MIT License. See the [LICENSE file](https://github.com/ByteDance-Seed/Seed-Coder/blob/master/LICENSE) for details.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{seed2025seedcoderletcodemodel,
title={{Seed-Coder}: Let the Code Model Curate Data for Itself},
author={{ByteDance Seed} and Yuyu Zhang and Jing Su and Yifan Sun and Chenguang Xi and Xia Xiao and Shen Zheng and Anxiang Zhang and Kaibo Liu and Daoguang Zan and Tao Sun and Jinhua Zhu and Shulin Xin and Dong Huang and Yetao Bai and Lixin Dong and Chao Li and Jianchong Chen and Hanzhi Zhou and Yifan Huang and Guanghan Ning and Xierui Song and Jiaze Chen and Siyao Liu and Kai Shen and Liang Xiang and Yonghui Wu},
year={2025},
eprint={2506.03524},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2506.03524},
}
```
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/Qwen3-30B-A1.5B-High-Speed-GGUF | Mungert | 2025-06-15T19:38:38Z | 1,076 | 1 | transformers | [
"transformers",
"gguf",
"32 k context",
"reasoning",
"thinking",
"qwen3",
"4 experts activated",
"double speed",
"128 experts",
"text-generation",
"base_model:Qwen/Qwen3-30B-A3B-Base",
"base_model:quantized:Qwen/Qwen3-30B-A3B-Base",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-06-05T05:31:02Z | ---
library_name: transformers
pipeline_tag: text-generation
tags:
- 32 k context
- reasoning
- thinking
- qwen3
- 4 experts activated
- double speed
- 128 experts
base_model:
- Qwen/Qwen3-30B-A3B-Base
---
# <span style="color: #7FFF7F;">Qwen3-30B-A1.5B-High-Speed GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`0d398442`](https://github.com/ggerganov/llama.cpp/commit/0d3984424f2973c49c4bcabe4cc0153b4f90c601).
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
<h2>Qwen3-30B-A1.5B-High-Speed - AKA: "Punch IT!"</h2>
<img src="star-wars-hans-solo.gif" style="float:right; padding:10px;">
This repo contains the full precision source code, in "safe tensors" format to generate GGUFs, GPTQ, EXL2, AWQ, HQQ and other formats. The source code can also be used directly.
This is a simple "finetune" of the Qwen's "Qwen 30B-A3B" (MOE) model, setting the experts in use from 8 to 4 (out of 128 experts).
This method close to doubles the speed of the model and uses 1.5B (of 30B) parameters instead of 3B (of 30B) parameters. Depending on the application you may want to
use the regular model ("30B-A3B"), and use this model for simpler use case(s) although I did not notice any loss of function during
routine (but not extensive) testing.
Example generation (Q4KS, CPU) at the bottom of this page using 4 experts / this model.
NEO Imatrix Quants / Imatrix Max Quants, at 64K context are here:
[ https://huggingface.co/DavidAU/Qwen3-30B-A1.5B-High-Speed-NEO-Imatrix-MAX-gguf ]
More complex use cases may benefit from using the normal version and/or 12, 16 or 24 experts version(s) - links below.
For reference:
- Cpu only operation Q4KS (windows 11) jumps from 12 t/s to 23 t/s.
- GPU performance IQ3S jumps from 75 t/s to over 125 t/s. (low to mid level card)
Context size: 32K + 8K for output (40k total)
Use Jinja Template or CHATML template.
IMPORTANT NOTES:
- Due to the unique nature (MOE, Size, Activated experts, size of experts) of this model GGUF quants can be run on the CPU, GPU or with GPU part "off-load", right up to full precision.
- This model is difficult to Imatrix : You need a much larger imatrix file / multi-language / multi-content (ie code/text) to imatrix it.
- GPU speeds will be BLISTERING 4x-8x or higher than CPU only speeds AND this model will be BLISTERING too, relative to other "30B" models (Token per second speed equal roughly to 1.5B "normal" model speeds).
Please refer the org model card for details, benchmarks, how to use, settings, system roles etc etc :
[ https://huggingface.co/Qwen/Qwen3-30B-A3B ]
<B>More / Less Experts Versions:</B>
12 experts:
[ https://huggingface.co/DavidAU/Qwen3-30B-A4.5B-12-Cooks ]
16 experts:
[ https://huggingface.co/DavidAU/Qwen3-30B-A6B-16-Extreme ]
16 experts, 128k context:
[ https://huggingface.co/DavidAU/Qwen3-30B-A6B-16-Extreme-128k-context ]
24 experts:
[ https://huggingface.co/DavidAU/Qwen3-30B-A7.5B-24-Grand-Brainstorm ]
<B>OPTIONAL SYSTEM ROLE: </B>
You may or may not need this, as most times Qwen3s generate their own reasoning/thinking blocks.
```
You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside <think> </think> tags, and then provide your solution or response to the problem.
```
See document "Maximizing-Model-Performance-All..." below for how to "set" system role in various LLM/AI apps below.
IMPORTANT: Highest Quality Settings / Optimal Operation Guide / Parameters and Samplers
If you are going to use this model, (source, GGUF or a different quant), please review this document for critical parameter, sampler and advance sampler settings (for multiple AI/LLM aps).
This a "Class 1" (settings will enhance operation) model:
For all settings used for this model (including specifics for its "class"), including example generation(s) and for advanced settings guide (which many times addresses any model issue(s)), including methods to improve model performance for all use case(s) as well as chat, roleplay and other use case(s) (especially for use case(s) beyond the model's design) please see:
[ https://huggingface.co/DavidAU/Maximizing-Model-Performance-All-Quants-Types-And-Full-Precision-by-Samplers_Parameters ]
REASON:
Regardless of "model class" this document will detail methods to enhance operations.
If the model is a Class 3/4 model the default settings (parameters, samplers, advanced samplers) must be set for "use case(s)" uses correctly. Some AI/LLM apps DO NOT have consistant default setting(s) which result in sub-par model operation. Like wise for Class 3/4 models (which operate somewhat to very differently than standard models) additional samplers and advanced samplers settings are required to "smooth out" operation, AND/OR also allow full operation for use cases the model was not designed for.
BONUS - Use these settings for ANY model, ANY repo, ANY quant (including source/full precision):
This document also details parameters, sampler and advanced samplers that can be use FOR ANY MODEL, FROM ANY REPO too - all quants, and of course source code operation too - to enhance the operation of any model.
[ https://huggingface.co/DavidAU/Maximizing-Model-Performance-All-Quants-Types-And-Full-Precision-by-Samplers_Parameters ]
NOTE:
I strongly suggest you also visit the DavidAU GGUF (below) repo too for more details in using this model ; especially if it is "Class 3" or "Class 4" to get maximum performance from the model.
For full information about this model, including:
- Details about this model and its use case(s).
- Context limits
- Special usage notes / settings.
- Any model(s) used to create this model.
- Template(s) used to access/use this model.
- Example generation(s)
- GGUF quants of this model
Please go to:
https://huggingface.co/DavidAU/Qwen3-30B-A1.5B-High-Speed-NEO-Imatrix-MAX-gguf
---
<B>Example Generation:</B>
---
Temp 1.2, rep pen 1.06, rep pen range 64, topk 100, topp .95, minp .05
Q4KS [non-imatrix], CPU only (windows 11), LMSTUDIO.
This is a mid level quant, expect stronger performance from an Imatrix Q4KS / higher quants / full precision.
NOTE: Some formatting lost on copy/paste.
PROMPT:
<B>
Science Fiction: The Last Transmission - Write a story that takes place entirely within a spaceship's cockpit as the sole surviving crew member attempts to send a final message back to Earth before the ship's power runs out. The story should explore themes of isolation, sacrifice, and the importance of human connection in the face of adversity. If the situation calls for it, have the character(s) curse and swear to further the reader's emotional connection to them. 800-1000 words.
</B>
---
<B>
[[[thinking start]]]
</B>
Thought for 47.29 seconds
Okay, I need to write a science fiction story based on the given prompt. The user wants the story to take place entirely in a spaceship's cockpit, focusing on the last surviving crew member trying to send a final message before the ship's power runs out. Themes are isolation, sacrifice, and human connection. Also, include some cursing to add emotional weight.
First, I should start by setting the scene. The cockpit needs to feel cramped and desperate. Maybe describe the flickering lights and the hissing sounds of the ship's systems failing. The main character is alone, so I need to show their loneliness. Maybe they have a log or something to talk to themselves, but since it's the cockpit, perhaps they're just trying to get through the message.
Next, the themes. Isolation is straightforward—they're the only one left. But how to show sacrifice? Maybe they stayed behind to make sure the message gets through, knowing they won't make it. Human connection could come through flashbacks or memories, but since the story is in the cockpit, maybe they're remembering a loved one, like a child or partner, to keep them going.
I need to include some swearing. Maybe the protagonist curses when they realize something's not working, or when they're under stress. But it shouldn't be too over the top. Maybe they curse at the ship for failing them, or at themselves for not being able to fix it.
The structure: Start with the character working on the controls, trying to send the message. Then show their internal struggle. Maybe flashbacks? Wait, the user said the story should take place entirely within the cockpit. So maybe no flashbacks. Just the present moment, with internal monologue.
Conflict: The ship is failing, power is running out. They need to send the message before they lose power. Maybe there's a countdown or a limited time. Also, maybe the ship is damaged in a way that it's irreversible.
Character development: The protagonist is determined, maybe a bit stubborn. They care about relaying the message, which could be a warning about something, like a virus or an alien threat. But since the prompt doesn't specify, maybe it's more about the human aspect—like they're the last, and they want to make sure someone knows what happened.
Dialogue: Since it's the cockpit, maybe they talk to themselves, or to the ship. Or maybe there's a voice in the comms? Wait, but the prompt says the story is entirely within the cockpit. So no one else is there except the main character. So it's all internal, except for maybe a log entry.
Wait, the user said "the sole surviving crew member attempts to send a final message back to Earth." So the message is for Earth. So maybe the protagonist is trying to record a message, but the ship is failing. Or maybe they're trying to send a distress signal, but the systems are failing.
Let me outline:
- Open with the cockpit's deteriorating condition. Maybe the life support is failing, or the power is out. The protagonist is trying to keep the comms on.
- They're alone, so they talk to the ship, maybe. "Come on, you piece of crap, hold together."
- Flashbacks aren't allowed, per the user's instruction (since it's all in the cockpit), so internal thoughts are key.
- The message they need to send is important. Maybe a warning about something, but also their own fear.
- The story should end with them managing to send the message, but knowing they won't survive, or maybe the message gets through just in time.
Including swearing: Maybe when they're frustrated with the systems not working, or when they realize the power is dying. "F***ing hell," "God damn it," etc.
Need to make sure the themes are present. Isolation: being the only one left. Sacrifice: giving their life to send the message. Human connection: maybe they think of someone before they die, or the act of sending the message is about connecting with others back home.
Possible structure:
1. Start in the cockpit, power failing, protagonist working frantically.
2. Flashbacks? Wait, no—since it's all in the cockpit, maybe not. But internal monologue can reflect on their past.
3. The protagonist's internal struggle: fear of death, but also duty to send the message.
4. Technical difficulties with the ship, leading to a climax where they get the message out, then succumb.
Now, writing the story with these elements. Keep it tight, 800-1000 words. Make the protagonist relatable, their fear and determination clear. Use short, punchy sentences to build tension. Include specific details about the ship's systems, maybe some technical jargon to make it realistic. Maybe the protagonist is an engineer or someone who knows the ship well, which helps in the struggle to keep it running.
Possible name for the character? Maybe not necessary, but a few names could be used. Or just "they" and "the pilot."
Also, the title: "The Last Transmission" is given, so the story should reflect that—sending a message, even if no one's there to receive it, or in this case, maybe they do.
Need to avoid clichés, but it's a common sci-fi trope. Make it fresh with emotional depth.
Potential pitfalls: Making the protagonist too passive, not showing their struggle enough. Need to show, not tell. Also, ensuring the themes are woven naturally into the narrative.
Okay, time to start drafting.
<B>
[[[thinking end]]]
</B>
OUTPUT:
The cockpit reeked of burnt circuitry and sweat, a stench that clung to my skin like a second, cursed flesh. My fingers danced over the controls, but the readouts spat back static—useless, just like the rest of this rustbucket. "Crawl, you piece of junk," I growled, slapping the console until my knuckles cracked. "Crawl!"
The ship groaned in response, a sound that might have been a moan if I were the superstitious type. I wasn't. Not really. Not anymore. Just a tired engineer with a death wish and a terminal case of bad luck. The Vanguard had been my pride before it became my prison. Now? Now it was a coffin, and I was the corpse inside, begging for the black to take me.
"Comm-link online," I rasped, watching the screen flicker. Nothing. Just the same hollow gray that had greeted me every time I tried to send this goddamn message. A laugh clawed up my throat—bitter, raw. Great. The AI's gone and left me high and dry too.
I slammed a fist against the armrest, feeling the impact reverberate through my ribs. "C'mon, you son of a bitch," I snarled, my voice cracking on the last word. "Don't leave me now. Not after everything."
The screen flickered again, and for a heartbeat, I thought—
"This is Lieutenant Kael Voss, and if you're hearing this, then we've already lost."
My breath caught. The voice was wrong. Too clean, too crisp, but it wasn't mine. It was hers. Her.
I staggered back as the comms crackled to life, the words pouring out in a flood that almost hurt. "They took the engines. I don't know how long before they find you, but you have to listen—don't trust the council. Don't trust anyone. Run. And if you survive…"
A chime interrupted her, sharp and shrill. I flinched, my hand flying to the sidearm at my hip. Not now. Not when the goddamn truth was finally slipping through my fingers.
The comms crackled again, and this time, it was my turn to beg. "C'mon, damn it! You can't just—you can't just leave me hanging like this!" My voice cracked, the words laced with something raw and broken. "I didn't come all this way to watch you die on me again, Lira!"
For a moment, there was only silence. Then her voice, softer now, almost tender. "You never could take a hint, could you?"
I closed my eyes against the burn behind my ribs. Not fair. Not fair. She'd always known how to twist the knife, hadn't she? Even here, even now.
"Kael…" Her voice dropped, and I could almost see her, standing in that damned lab, her auburn hair pulled back in a messy knot, eyes blazing with that stubborn fire of hers. "You have to get this transmission out. No matter what happens. Promised?"
I swallowed hard, the weight of the lie sitting heavy in my chest. "Yeah," I croaked. "Yeah, I promise."
The comms crackled again, and for a heartbeat, I thought—
Then it died.
Just like that. All the static, all the fury, all the damn hope she'd left behind, gone.
I slumped against the console, my breath coming in ragged gasps. "Gods," I whispered, the words barely audible over the hum of the failing systems. "You really did it, didn't you? You goddamn idiot."
But even as the words left my mouth, I knew it was true. She'd always been a handful. A pain in the ass. The kind of woman who made your heart turn to lead and your soul turn to ash with one sardonic smirk. And she'd done it.
She'd saved us all.
A shudder ran through me as the lights flickered again, casting wild shadows across the walls. I dragged a hand down my face, scrubbing at the exhaustion that felt like it was seeping into my very bones. Just a little longer. That's what she'd have said, wasn't it? Just a little longer, Kael.
A soft chime echoed through the cockpit, and for a heartbeat, I just stared at the console. My fingers itched to reach for it, to press that one last button and let the truth—her truth—ripple out into the void.
Because that's what we did. That's what we all did. We reached for each other, even when it hurt. Even when it was too late.
I exhaled sharply, the sound feeling like a prayer. "Alright," I murmured, my voice steady now. "Let's not make her a liar, huh?"
And with that, I pressed the final key.
|
mlx-community/gemma-3-1b-it-DQ | mlx-community | 2025-06-15T19:38:37Z | 131 | 0 | mlx | [
"mlx",
"safetensors",
"gemma3_text",
"text-generation",
"conversational",
"base_model:google/gemma-3-1b-it",
"base_model:quantized:google/gemma-3-1b-it",
"license:gemma",
"model-index",
"4-bit",
"region:us"
] | text-generation | 2025-06-12T03:30:12Z | ---
license: gemma
library_name: mlx
pipeline_tag: text-generation
extra_gated_heading: Access Gemma on Hugging Face
extra_gated_prompt: To access Gemma on Hugging Face, you’re required to review and
agree to Google’s usage license. To do this, please ensure you’re logged in to Hugging
Face and click below. Requests are processed immediately.
extra_gated_button_content: Acknowledge license
base_model: google/gemma-3-1b-it
tags:
- mlx
model-index:
- name: gemma-3-1b-it-DQ
results:
- task:
type: text-generation
dataset:
type: PIQA
name: PIQA
metrics:
- name: pass@1
type: pass@1
value: 0.75
verified: false
- task:
type: text-generation
dataset:
type: winogrande
name: winogrande
metrics:
- name: pass@1
type: pass@1
value: 0.60
verified: false
- task:
type: text-generation
dataset:
type: boolq
name: boolq
metrics:
- name: pass@1
type: pass@1
value: 0.73
verified: false
- task:
type: text-generation
dataset:
type: arc-c
name: arc-c
metrics:
- name: pass@1
type: pass@1
value: 0.35
verified: false
---
# mlx-community/gemma-3-1b-it-DQ
This model [mlx-community/gemma-3-1b-it-DQ](https://huggingface.co/mlx-community/gemma-3-1b-it-DQ) was
converted to MLX format from [google/gemma-3-1b-it](https://huggingface.co/google/gemma-3-1b-it)
using mlx-lm version **0.25.2**.
##### 2x faster and 2.4x less memory footprint than the dequantized model
## Use with mlx
```bash
pip install mlx-lm
```
```python
from mlx_lm import load, generate
model, tokenizer = load("mlx-community/gemma-3-1b-it-DQ")
prompt = "hello"
if tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True)
```
|
Mungert/Holo1-3B-GGUF | Mungert | 2025-06-15T19:38:35Z | 3,419 | 1 | transformers | [
"transformers",
"gguf",
"multimodal",
"action",
"agent",
"visual-document-retrieval",
"en",
"arxiv:2506.02865",
"arxiv:2401.13919",
"base_model:Qwen/Qwen2.5-VL-3B-Instruct",
"base_model:quantized:Qwen/Qwen2.5-VL-3B-Instruct",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | visual-document-retrieval | 2025-06-04T19:44:51Z | ---
base_model:
- Qwen/Qwen2.5-VL-3B-Instruct
language:
- en
library_name: transformers
license: other
license_name: other
pipeline_tag: visual-document-retrieval
tags:
- multimodal
- action
- agent
---
# <span style="color: #7FFF7F;">Holo1-3B GGUF Models</span>
This model is described in the paper [Surfer-H Meets Holo1: Cost-Efficient Web Agent Powered by Open Weights](https://huggingface.co/papers/2506.02865). The project page can be found at [https://www.surferh.com](https://www.surferh.com).
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`71bdbdb5`](https://github.com/ggerganov/llama.cpp/commit/71bdbdb58757d508557e6d8b387f666cdfb25c5e).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Holo1-3B
## Model Description
Holo1 is an Action Vision-Language Model (VLM) developed by [HCompany](https://www.hcompany.ai/) for use in the Surfer-H web agent system. It is designed to interact with web interfaces like a human user.
As part of a broader agentic architecture, Holo1 acts as a policy, localizer, or validator, helping the agent understand and act in digital environments.
Trained on a mix of open-access, synthetic, and self-generated data, Holo1 enables state-of-the-art (SOTA) performance on the [WebVoyager](https://arxiv.org/pdf/2401.13919) benchmark, offering the best accuracy/cost tradeoff among current models.
It also excels in UI localization tasks such as [Screenspot](https://huggingface.co/datasets/rootsautomation/ScreenSpot), [Screenspot-V2](https://huggingface.co/datasets/HongxinLi/ScreenSpot_v2), [Screenspot-Pro](https://huggingface.co/datasets/likaixin/ScreenSpot-Pro), [GroundUI-Web](https://huggingface.co/datasets/agent-studio/GroundUI-1K), and our own newly introduced
benchmark [WebClick](https://huggingface.co/datasets/Hcompany/WebClick).
Holo1 is optimized for both accuracy and cost-efficiency, making it a strong open-source alternative to existing VLMs.
For more details, check our paper and our blog post.
- **Developed by:** [HCompany](https://www.hcompany.ai/)
- **Model type:** Action Vision-Language Model
- **Finetuned from model:** Qwen/Qwen2.5-VL-3B-Instruct
- **Paper:** https://arxiv.org/abs/2506.02865
- **Blog Post:** https://www.hcompany.ai/surfer-h
- **License:** https://huggingface.co/Hcompany/Holo1-3B/blob/main/LICENSE
## Results
### Surfer-H: Pareto-Optimal Performance on [WebVoyager](https://arxiv.org/pdf/2401.13919)
Surfer-H is designed to be flexible and modular. It is composed of three independent components:
- A Policy model that plans, decides, and drives the agent's behavior
- A Localizer model that sees and understands visual UIs to drive precise interactions
- A Validator model that checks whether the answer is valid
The agent thinks before acting, takes notes, and can retry if its answer is rejected. It can operate with different models for each module, allowing for tradeoffs between accuracy, speed, and cost.
We evaluated Surfer-H on the [WebVoyager](https://arxiv.org/pdf/2401.13919) benchmark: 643 real-world web tasks ranging from retrieving prices to finding news or scheduling events.
<div style="text-align: center;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/682c3e22650f6bbe33bb9d94/kO_4DlW_O45Wi7eK9-r8v.png" width="800"/>
</div>
We’ve tested multiple configurations, from GPT-4-powered agents to 100% open Holo1 setups. Among them, the fully Holo1-based agents offered the strongest tradeoff between accuracy and cost:
- Surfer-H + Holo1-7B: 92.2% accuracy at $0.13 per task
- Surfer-H + GPT-4.1: 92.0% at $0.54 per task
- Surfer-H + Holo1-3B: 89.7% at $0.11 per task
- Surfer-H + GPT-4.1-mini: 88.8% at $0.26 per task
This places Holo1-powered agents on the Pareto frontier, delivering the best accuracy per dollar.
Unlike other agents that rely on custom APIs or brittle wrappers, Surfer-H operates purely through the browser — just like a real user. Combined with Holo1, it becomes a powerful, general-purpose, cost-efficient web automation system.
### Holo1: State-of-the-Art UI Localization
A key skill for the real-world utility of our VLMs within agents is localization: the ability to identify precise
coordinates on a user interface (UI) to interact with to complete a task or follow an instruction. To assess
this capability, we evaluated our Holo1 models on several established localization benchmarks, including
[Screenspot](https://huggingface.co/datasets/rootsautomation/ScreenSpot), [Screenspot-V2](https://huggingface.co/datasets/HongxinLi/ScreenSpot_v2), [Screenspot-Pro](https://huggingface.co/datasets/likaixin/ScreenSpot-Pro), [GroundUI-Web](https://huggingface.co/datasets/agent-studio/GroundUI-1K), and our own newly introduced
benchmark [WebClick](https://huggingface.co/datasets/Hcompany/WebClick).
<div style="text-align: center;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/682c3e22650f6bbe33bb9d94/UutD2Meevd5Xw0_mhX2wK.png" width="600"/>
</div>
<div style="text-align: center;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/682c3e22650f6bbe33bb9d94/NhzkB8xnEQYMqiGxPnJSt.png" width="600"/>
</div>
## Get Started with the Model
We provide 2 spaces to experiment with Localization and Navigation:
- https://huggingface.co/spaces/Hcompany/Holo1-Navigation
- https://huggingface.co/spaces/Hcompany/Holo1-Localization
We provide starter code for the localization task: i.e. image + instruction -> click coordinates
We also provide code to reproduce screenspot evaluations: screenspot_eval.py
### Prepare model, processor
Holo1 models are based on Qwen2.5-VL architecture, which comes with transformers support. Here we provide a simple usage example.
You can load the model and the processor as follows:
```python
import json
import os
from typing import Any, Literal
from transformers import AutoModelForImageTextToText, AutoProcessor
# default: Load the model on the available device(s)
# We recommend enabling flash_attention_2 for better acceleration and memory saving.
model = AutoModelForImageTextToText.from_pretrained(
"Hcompany/Holo1-3B",
torch_dtype="auto",
# torch_dtype=torch.bfloat16,
# attn_implementation="flash_attention_2",
device_map="auto",
)
# default processor
processor = AutoProcessor.from_pretrained("Hcompany/Holo1-3B")
# The default range for the number of visual tokens per image in the model is 4-1280.
# You can set min_pixels and max_pixels according to your needs, such as a token range of 256-1280, to balance performance and cost.
# processor = AutoProcessor.from_pretrained(model_dir, min_pixels=min_pixels, max_pixels=max_pixels)
# Helper function to run inference
def run_inference(messages: list[dict[str, Any]]) -> str:
# Preparation for inference
text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = processor(
text=[text],
images=image,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)]
return processor.batch_decode(generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False)
```
### Prepare image and instruction
WARNING: Holo1 is using absolute coordinates (number of pixels) and HuggingFace processor is doing image resize. To have matching coordinates, one needs to smart_resize the image.
```python
from PIL import Image
from transformers.models.qwen2_vl.image_processing_qwen2_vl import smart_resize
# Prepare image and instruction
image_url = "https://huggingface.co/Hcompany/Holo1-3B/resolve/main/calendar_example.jpg"
image = Image.open(requests.get(image_url, stream=True).raw)
# Resize the image so that predicted absolute coordinates match the size of the image.
image_processor = processor.image_processor
resized_height, resized_width = smart_resize(
image.height,
image.width,
factor=image_processor.patch_size * image_processor.merge_size,
min_pixels=image_processor.min_pixels,
max_pixels=image_processor.max_pixels,
)
image = image.resize(size=(resized_width, resized_height), resample=None) # type: ignore
```
### Navigation with Structured Output
```python
import json
from . import navigation
task = "Book a hotel in Paris on August 3rd for 3 nights"
prompt = navigation.get_navigation_prompt(task, image, step=1)
navigation_str = run_inference(prompt)[0]
navigation = navigation.NavigationStep(**json.loads(navigation_str))
print(navigation)
# Expected NavigationStep(note='', thought='I need to select the check-out date as August 3rd and then proceed to search for hotels.', action=ClickElementAction(action='click_element', element='August 3rd on the calendar', x=777, y=282))
```
### Localization with click(x, y)
```python
from . import localization
instruction = "Select July 14th as the check-out date"
prompt = localization.get_localization_prompt(image, instruction)
coordinates = run_inference(prompt)[0]
print(coordinates)
# Expected Click(352, 348)
```
### Localization with Structured Output
We trained Holo1 as an Action VLM with extensive use of json and tool calls. Therefore, it can be queried reliably with structured output:
```python
import json
from . import localization
instruction = "Select July 14th as the check-out date"
prompt = localization.get_localization_prompt_structured_output(image, instruction)
coordinates_structured_str = run_inference(prompt)[0]
coordinates_structured = localization.ClickAction(**json.loads(coordinates_structured_str))
print(coordinates_structured)
# Expected ClickAction(action='click', x=352, y=340)
```
## Citation
**BibTeX:**
```
@misc{andreux2025surferhmeetsholo1costefficient,
title={Surfer-H Meets Holo1: Cost-Efficient Web Agent Powered by Open Weights},
author={Mathieu Andreux and Breno Baldas Skuk and Hamza Benchekroun and Emilien Biré and Antoine Bonnet and Riaz Bordie and Matthias Brunel and Pierre-Louis Cedoz and Antoine Chassang and Mickaël Chen and Alexandra D. Constantinou and Antoine d'Andigné and Hubert de La Jonquière and Aurélien Delfosse and Ludovic Denoyer and Alexis Deprez and Augustin Derupti and Michael Eickenberg and Mathïs Federico and Charles Kantor and Xavier Koegler and Yann Labbé and Matthew C. H. Lee and Erwan Le Jumeau de Kergaradec and Amir Mahla and Avshalom Manevich and Adrien Maret and Charles Masson and Rafaël Maurin and Arturo Mena and Philippe Modard and Axel Moyal and Axel Nguyen Kerbel and Julien Revelle and Mats L. Richter and María Santos and Laurent Sifre and Maxime Theillard and Marc Thibault and Louis Thiry and Léo Tronchon and Nicolas Usunier and Tony Wu},
year={2025},
eprint={2506.02865},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2506.02865},
}
``` |
Mungert/Devstral-Small-2505-GGUF | Mungert | 2025-06-15T19:38:22Z | 2,374 | 2 | vllm | [
"vllm",
"gguf",
"text2text-generation",
"en",
"fr",
"de",
"es",
"pt",
"it",
"ja",
"ko",
"ru",
"zh",
"ar",
"fa",
"id",
"ms",
"ne",
"pl",
"ro",
"sr",
"sv",
"tr",
"uk",
"vi",
"hi",
"bn",
"base_model:mistralai/Devstral-Small-2505",
"base_model:quantized:mistralai/Devstral-Small-2505",
"license:apache-2.0",
"region:us",
"imatrix",
"conversational"
] | text2text-generation | 2025-06-03T17:17:37Z | ---
language:
- en
- fr
- de
- es
- pt
- it
- ja
- ko
- ru
- zh
- ar
- fa
- id
- ms
- ne
- pl
- ro
- sr
- sv
- tr
- uk
- vi
- hi
- bn
license: apache-2.0
library_name: vllm
inference: false
base_model:
- mistralai/Devstral-Small-2505
extra_gated_description: >-
If you want to learn more about how we process your personal data, please read
our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
pipeline_tag: text2text-generation
---
# <span style="color: #7FFF7F;">Devstral-Small-2505 GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`f5cd27b7`](https://github.com/ggerganov/llama.cpp/commit/f5cd27b71da3ac375a04a41643d14fc779a8057b).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Devstral-Small-2505-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Devstral-Small-2505-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Devstral-Small-2505-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Devstral-Small-2505-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Devstral-Small-2505-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Devstral-Small-2505-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Devstral-Small-2505-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Devstral-Small-2505-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Devstral-Small-2505-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Devstral-Small-2505-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Devstral-Small-2505-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Model Card for mistralai/Devstrall-Small-2505
Devstral is an agentic LLM for software engineering tasks built under a collaboration between [Mistral AI](https://mistral.ai/) and [All Hands AI](https://www.all-hands.dev/) 🙌. Devstral excels at using tools to explore codebases, editing multiple files and power software engineering agents. The model achieves remarkable performance on SWE-bench which positionates it as the #1 open source model on this [benchmark](#benchmark-results).
It is finetuned from [Mistral-Small-3.1](https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Base-2503), therefore it has a long context window of up to 128k tokens. As a coding agent, Devstral is text-only and before fine-tuning from `Mistral-Small-3.1` the vision encoder was removed.
For enterprises requiring specialized capabilities (increased context, domain-specific knowledge, etc.), we will release commercial models beyond what Mistral AI contributes to the community.
Learn more about Devstral in our [blog post](https://mistral.ai/news/devstral).
## Key Features:
- **Agentic coding**: Devstral is designed to excel at agentic coding tasks, making it a great choice for software engineering agents.
- **lightweight**: with its compact size of just 24 billion parameters, Devstral is light enough to run on a single RTX 4090 or a Mac with 32GB RAM, making it an appropriate model for local deployment and on-device use.
- **Apache 2.0 License**: Open license allowing usage and modification for both commercial and non-commercial purposes.
- **Context Window**: A 128k context window.
- **Tokenizer**: Utilizes a Tekken tokenizer with a 131k vocabulary size.
## Benchmark Results
### SWE-Bench
Devstral achieves a score of 46.8% on SWE-Bench Verified, outperforming prior open-source SoTA by 6%.
| Model | Scaffold | SWE-Bench Verified (%) |
|------------------|--------------------|------------------------|
| Devstral | OpenHands Scaffold | **46.8** |
| GPT-4.1-mini | OpenAI Scaffold | 23.6 |
| Claude 3.5 Haiku | Anthropic Scaffold | 40.6 |
| SWE-smith-LM 32B | SWE-agent Scaffold | 40.2 |
When evaluated under the same test scaffold (OpenHands, provided by All Hands AI 🙌), Devstral exceeds far larger models such as Deepseek-V3-0324 and Qwen3 232B-A22B.
## Usage
We recommend to use Devstral with the [OpenHands](https://github.com/All-Hands-AI/OpenHands/tree/main) scaffold.
You can use it either through our API or by running locally.
### API
Follow these [instructions](https://docs.mistral.ai/getting-started/quickstart/#account-setup) to create a Mistral account and get an API key.
Then run these commands to start the OpenHands docker container.
```bash
export MISTRAL_API_KEY=<MY_KEY>
docker pull docker.all-hands.dev/all-hands-ai/runtime:0.39-nikolaik
mkdir -p ~/.openhands-state && echo '{"language":"en","agent":"CodeActAgent","max_iterations":null,"security_analyzer":null,"confirmation_mode":false,"llm_model":"mistral/devstral-small-2505","llm_api_key":"'$MISTRAL_API_KEY'","remote_runtime_resource_factor":null,"github_token":null,"enable_default_condenser":true}' > ~/.openhands-state/settings.json
docker run -it --rm --pull=always \
-e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:0.39-nikolaik \
-e LOG_ALL_EVENTS=true \
-v /var/run/docker.sock:/var/run/docker.sock \
-v ~/.openhands-state:/.openhands-state \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name openhands-app \
docker.all-hands.dev/all-hands-ai/openhands:0.39
```
### Local inference
You can also run the model locally. It can be done with LMStudio or other providers listed below.
Launch Openhands
You can now interact with the model served from LM Studio with openhands. Start the openhands server with the docker
```bash
docker pull docker.all-hands.dev/all-hands-ai/runtime:0.38-nikolaik
docker run -it --rm --pull=always \
-e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:0.38-nikolaik \
-e LOG_ALL_EVENTS=true \
-v /var/run/docker.sock:/var/run/docker.sock \
-v ~/.openhands-state:/.openhands-state \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name openhands-app \
docker.all-hands.dev/all-hands-ai/openhands:0.38
```
The server will start at http://0.0.0.0:3000. Open it in your browser and you will see a tab AI Provider Configuration.
Now you can start a new conversation with the agent by clicking on the plus sign on the left bar.
The model can also be deployed with the following libraries:
- [`LMStudio (recommended for quantized model)`](https://lmstudio.ai/): See [here](#lmstudio-recommended-for-quantized-model)
- [`vllm (recommended)`](https://github.com/vllm-project/vllm): See [here](#vllm-recommended)
- [`mistral-inference`](https://github.com/mistralai/mistral-inference): See [here](#mistral-inference)
- [`transformers`](https://github.com/huggingface/transformers): See [here](#transformers)
- [`ollama`](https://github.com/ollama/ollama): See [here](#ollama)
### OpenHands (recommended)
#### Launch a server to deploy Devstral-Small-2505
Make sure you launched an OpenAI-compatible server such as vLLM or Ollama as described above. Then, you can use OpenHands to interact with `Devstral-Small-2505`.
In the case of the tutorial we spineed up a vLLM server running the command:
```bash
vllm serve mistralai/Devstral-Small-2505 --tokenizer_mode mistral --config_format mistral --load_format mistral --tool-call-parser mistral --enable-auto-tool-choice --tensor-parallel-size 2
```
The server address should be in the following format: `http://<your-server-url>:8000/v1`
#### Launch OpenHands
You can follow installation of OpenHands [here](https://docs.all-hands.dev/modules/usage/installation).
The easiest way to launch OpenHands is to use the Docker image:
```bash
docker pull docker.all-hands.dev/all-hands-ai/runtime:0.38-nikolaik
docker run -it --rm --pull=always \
-e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:0.38-nikolaik \
-e LOG_ALL_EVENTS=true \
-v /var/run/docker.sock:/var/run/docker.sock \
-v ~/.openhands-state:/.openhands-state \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name openhands-app \
docker.all-hands.dev/all-hands-ai/openhands:0.38
```
Then, you can access the OpenHands UI at `http://localhost:3000`.
#### Connect to the server
When accessing the OpenHands UI, you will be prompted to connect to a server. You can use the advanced mode to connect to the server you launched earlier.
Fill the following fields:
- **Custom Model**: `openai/mistralai/Devstral-Small-2505`
- **Base URL**: `http://<your-server-url>:8000/v1`
- **API Key**: `token` (or any other token you used to launch the server if any)
#### Use OpenHands powered by Devstral
Now you're good to use Devstral Small inside OpenHands by **starting a new conversation**. Let's build a To-Do list app.
<details>
<summary>To-Do list app</summary
1. Let's ask Devstral to generate the app with the following prompt:
```txt
Build a To-Do list app with the following requirements:
- Built using FastAPI and React.
- Make it a one page app that:
- Allows to add a task.
- Allows to delete a task.
- Allows to mark a task as done.
- Displays the list of tasks.
- Store the tasks in a SQLite database.
```
2. Let's see the result
You should see the agent construct the app and be able to explore the code it generated.
If it doesn't do it automatically, ask Devstral to deploy the app or do it manually, and then go the front URL deployment to see the app.
3. Iterate
Now that you have a first result you can iterate on it by asking your agent to improve it. For example, in the app generated we could click on a task to mark it checked but having a checkbox would improve UX. You could also ask it to add a feature to edit a task, or to add a feature to filter the tasks by status.
Enjoy building with Devstral Small and OpenHands!
</details>
### LMStudio (recommended for quantized model)
Download the weights from huggingface:
```
pip install -U "huggingface_hub[cli]"
huggingface-cli download \
"mistralai/Devstral-Small-2505_gguf" \
--include "devstralQ4_K_M.gguf" \
--local-dir "mistralai/Devstral-Small-2505_gguf/"
```
You can serve the model locally with [LMStudio](https://lmstudio.ai/).
* Download [LM Studio](https://lmstudio.ai/) and install it
* Install `lms cli ~/.lmstudio/bin/lms bootstrap`
* In a bash terminal, run `lms import devstralQ4_K_M.ggu` in the directory where you've downloaded the model checkpoint (e.g. `mistralai/Devstral-Small-2505_gguf`)
* Open the LMStudio application, click the terminal icon to get into the developer tab. Click select a model to load and select Devstral Q4 K M. Toggle the status button to start the model, in setting oggle Serve on Local Network to be on.
* On the right tab, you will see an API identifier which should be devstralq4_k_m and an api address under API Usage. Keep note of this address, we will use it in the next step.
Launch Openhands
You can now interact with the model served from LM Studio with openhands. Start the openhands server with the docker
```bash
docker pull docker.all-hands.dev/all-hands-ai/runtime:0.38-nikolaik
docker run -it --rm --pull=always \
-e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:0.38-nikolaik \
-e LOG_ALL_EVENTS=true \
-v /var/run/docker.sock:/var/run/docker.sock \
-v ~/.openhands-state:/.openhands-state \
-p 3000:3000 \
--add-host host.docker.internal:host-gateway \
--name openhands-app \
docker.all-hands.dev/all-hands-ai/openhands:0.38
```
Click “see advanced setting” on the second line.
In the new tab, toggle advanced to on. Set the custom model to be mistral/devstralq4_k_m and Base URL the api address we get from the last step in LM Studio. Set API Key to dummy. Click save changes.
### vLLM (recommended)
We recommend using this model with the [vLLM library](https://github.com/vllm-project/vllm)
to implement production-ready inference pipelines.
**_Installation_**
Make sure you install [`vLLM >= 0.8.5`](https://github.com/vllm-project/vllm/releases/tag/v0.8.5):
```
pip install vllm --upgrade
```
Doing so should automatically install [`mistral_common >= 1.5.5`](https://github.com/mistralai/mistral-common/releases/tag/v1.5.5).
To check:
```
python -c "import mistral_common; print(mistral_common.__version__)"
```
You can also make use of a ready-to-go [docker image](https://github.com/vllm-project/vllm/blob/main/Dockerfile) or on the [docker hub](https://hub.docker.com/layers/vllm/vllm-openai/latest/images/sha256-de9032a92ffea7b5c007dad80b38fd44aac11eddc31c435f8e52f3b7404bbf39).
#### Server
We recommand that you use Devstral in a server/client setting.
1. Spin up a server:
```
vllm serve mistralai/Devstral-Small-2505 --tokenizer_mode mistral --config_format mistral --load_format mistral --tool-call-parser mistral --enable-auto-tool-choice --tensor-parallel-size 2
```
2. To ping the client you can use a simple Python snippet.
```py
import requests
import json
from huggingface_hub import hf_hub_download
url = "http://<your-server-url>:8000/v1/chat/completions"
headers = {"Content-Type": "application/json", "Authorization": "Bearer token"}
model = "mistralai/Devstral-Small-2505"
def load_system_prompt(repo_id: str, filename: str) -> str:
file_path = hf_hub_download(repo_id=repo_id, filename=filename)
with open(file_path, "r") as file:
system_prompt = file.read()
return system_prompt
SYSTEM_PROMPT = load_system_prompt(model, "SYSTEM_PROMPT.txt")
messages = [
{"role": "system", "content": SYSTEM_PROMPT},
{
"role": "user",
"content": [
{
"type": "text",
"text": "<your-command>",
},
],
},
]
data = {"model": model, "messages": messages, "temperature": 0.15}
response = requests.post(url, headers=headers, data=json.dumps(data))
print(response.json()["choices"][0]["message"]["content"])
```
### Mistral-inference
We recommend using mistral-inference to quickly try out / "vibe-check" Devstral.
#### Install
Make sure to have mistral_inference >= 1.6.0 installed.
```bash
pip install mistral_inference --upgrade
```
#### Download
```python
from huggingface_hub import snapshot_download
from pathlib import Path
mistral_models_path = Path.home().joinpath('mistral_models', 'Devstral')
mistral_models_path.mkdir(parents=True, exist_ok=True)
snapshot_download(repo_id="mistralai/Devstral-Small-2505", allow_patterns=["params.json", "consolidated.safetensors", "tekken.json"], local_dir=mistral_models_path)
```
#### Python
You can run the model using the following command:
```bash
mistral-chat $HOME/mistral_models/Devstral --instruct --max_tokens 300
```
You can then prompt it with anything you'd like.
### Ollama
You can run Devstral using the [Ollama](https://ollama.ai/) CLI.
```bash
ollama run devstral
```
### Transformers
To make the best use of our model with transformers make sure to have [installed](https://github.com/mistralai/mistral-common) ` mistral-common >= 1.5.5` to use our tokenizer.
```bash
pip install mistral-common --upgrade
```
Then load our tokenizer along with the model and generate:
```python
import torch
from mistral_common.protocol.instruct.messages import (
SystemMessage, UserMessage
)
from mistral_common.protocol.instruct.request import ChatCompletionRequest
from mistral_common.tokens.tokenizers.mistral import MistralTokenizer
from mistral_common.tokens.tokenizers.tekken import SpecialTokenPolicy
from huggingface_hub import hf_hub_download
from transformers import AutoModelForCausalLM
def load_system_prompt(repo_id: str, filename: str) -> str:
file_path = hf_hub_download(repo_id=repo_id, filename=filename)
with open(file_path, "r") as file:
system_prompt = file.read()
return system_prompt
model_id = "mistralai/Devstral-Small-2505"
tekken_file = hf_hub_download(repo_id=model_id, filename="tekken.json")
SYSTEM_PROMPT = load_system_prompt(model_id, "SYSTEM_PROMPT.txt")
tokenizer = MistralTokenizer.from_file(tekken_file)
model = AutoModelForCausalLM.from_pretrained(model_id)
tokenized = tokenizer.encode_chat_completion(
ChatCompletionRequest(
messages=[
SystemMessage(content=SYSTEM_PROMPT),
UserMessage(content="<your-command>"),
],
)
)
output = model.generate(
input_ids=torch.tensor([tokenized.tokens]),
max_new_tokens=1000,
)[0]
decoded_output = tokenizer.decode(output[len(tokenized.tokens):])
print(decoded_output)
``` |
Mungert/QwenLong-L1-32B-GGUF | Mungert | 2025-06-15T19:38:13Z | 6,074 | 9 | transformers | [
"transformers",
"gguf",
"long-context",
"large-reasoning-model",
"text-generation",
"dataset:Tongyi-Zhiwen/DocQA-RL-1.6K",
"arxiv:2505.17667",
"arxiv:2309.00071",
"base_model:deepseek-ai/DeepSeek-R1-Distill-Qwen-32B",
"base_model:quantized:deepseek-ai/DeepSeek-R1-Distill-Qwen-32B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-05-28T11:43:14Z | ---
license: apache-2.0
datasets:
- Tongyi-Zhiwen/DocQA-RL-1.6K
base_model:
- deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
tags:
- long-context
- large-reasoning-model
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">QwenLong-L1-32B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`f5cd27b7`](https://github.com/ggerganov/llama.cpp/commit/f5cd27b71da3ac375a04a41643d14fc779a8057b).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `QwenLong-L1-32B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `QwenLong-L1-32B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `QwenLong-L1-32B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `QwenLong-L1-32B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `QwenLong-L1-32B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `QwenLong-L1-32B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `QwenLong-L1-32B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `QwenLong-L1-32B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `QwenLong-L1-32B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `QwenLong-L1-32B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `QwenLong-L1-32B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# QwenLong-L1: Towards Long-Context Large Reasoning Models with Reinforcement Learning
<p align="center" width="100%">
</p>
<div id="top" align="center">
-----------------------------
[](https://opensource.org/licenses/Apache-2.0)
[](https://arxiv.org/abs/2505.17667)
[](https://github.com/Tongyi-Zhiwen/QwenLong-L1)
[](https://modelscope.cn/models/iic/QwenLong-L1-32B)
[](https://huggingface.co/Tongyi-Zhiwen/QwenLong-L1-32B)
<!-- **Authors:** -->
_**Fanqi Wan, Weizhou Shen, Shengyi Liao, Yingcheng Shi, Chenliang Li,**_
_**Ziyi Yang, Ji Zhang, Fei Huang, Jingren Zhou, Ming Yan**_
<!-- **Affiliations:** -->
_Tongyi Lab, Alibaba Group_
<p align="center">
<img src="./assets/fig1.png" width="100%"> <br>
</p>
</div>
## 🎉 News
- **May 28, 2025:** 🔥 We release [🤗 QwenLong-L1-32B-AWQ](https://huggingface.co/Tongyi-Zhiwen/QwenLong-L1-32B-AWQ), which has undergone AWQ int4 quantization using the ms-swift framework.
- **May 26, 2025:** 🔥 We release [🤗 QwenLong-L1-32B](https://huggingface.co/Tongyi-Zhiwen/QwenLong-L1-32B), which is the first long-context LRM trained with reinforcement learning for long-context reasoning. Experiments on seven long-context DocQA benchmarks demonstrate that **QwenLong-L1-32B outperforms flagship LRMs like OpenAI-o3-mini and Qwen3-235B-A22B, achieving performance on par with Claude-3.7-Sonnet-Thinking**, demonstrating leading performance among state-of-the-art LRMs.
- **May 26, 2025:** 🔥 We release [🤗 DocQA-RL-1.6K](https://huggingface.co/datasets/Tongyi-Zhiwen/DocQA-RL-1.6K), which is a specialized RL training dataset comprising 1.6K document question answering (DocQA) problems spanning mathematical, logical, and multi-hop reasoning domains.
## 📚 Introduction
In this work, we propose QwenLong-L1, a novel reinforcement learning (RL) framework designed to facilitate the transition of LRMs from short-context proficiency to robust long-context generalization. In our preliminary experiments, we illustrate the differences between the training dynamics of short-context and long-context reasoning RL.
<p align="center">
<img src="./assets/fig2.png" width="100%"> <br>
</p>
Our framework enhances short-context LRMs through progressive context scaling during RL training. The framework comprises three core components: a warm-up supervised fine-tuning (SFT) phase to initialize a robust policy, a curriculum-guided RL phase that facilitates stable adaptation from short to long contexts, and a difficulty-aware retrospective sampling mechanism that adjusts training complexity across stages to incentivize policy exploration. Leveraging recent RL algorithms, including GRPO and DAPO, our framework integrates hybrid reward functions combining rule-based and model-based binary outcome rewards to balance precision and recall. Through strategic utilization of group relative advantages during policy optimization, it guides LRMs to learn effective reasoning patterns essential for robust long-context grounding and superior reasoning capabilities.
<p align="center">
<img src="./assets/fig3.png" width="100%"> <br>
</p>
## 🎯 Model Release
We release [🤗 QwenLong-L1-32B](https://huggingface.co/Tongyi-Zhiwen/QwenLong-L1-32B), which is the first long-context LRM trained with reinforcement learniing for long-context reasoning. Experiments on seven long-context DocQA benchmarks demonstrate that **QwenLong-L1-32B outperforms flagship LRMs like OpenAI-o3-mini and Qwen3-235B-A22B, achieving performance on par with Claude-3.7-Sonnet-Thinking**, demonstrating leading performance among state-of-the-art LRMs.
Here are the evaluation results.
<p align="center">
<img src="./assets/tab4.png" width="100%"> <br>
</p>
## 🛠️ Requirements
```bash
# Create the conda environment
conda create -n qwenlongl1 python==3.10
conda activate qwenlongl1
# Install requirements
pip3 install -r requirements.txt
# Install verl
cd verl
pip3 install -e .
# Install vLLM
pip3 install vllm==0.7.3
# Install flash-attn
pip3 install flash-attn --no-build-isolation
```
## 🚀 Quick Start
Here's how you can run the model using the 🤗 Transformers:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Tongyi-Zhiwen/QwenLong-L1-32B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
template = """Please read the following text and answer the question below.
<text>
$DOC$
</text>
$Q$
Format your response as follows: "Therefore, the answer is (insert answer here)"."""
context = "<YOUR_CONTEXT_HERE>"
question = "<YOUR_QUESTION_HERE>"
prompt = template.replace('$DOC$', context.strip()).replace('$Q$', question.strip())
messages = [
# {"role": "system", "content": "You are QwenLong-L1, created by Alibaba Tongyi Lab. You are a helpful assistant."}, # Use system prompt to define identity when needed.
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=10000,
temperature=0.7,
top_p=0.95
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151649 (</think>)
index = len(output_ids) - output_ids[::-1].index(151649)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
## ♾️ Processing Long Documents
For input where the total length (including both input and output) significantly exceeds 32,768 tokens, we recommend using RoPE scaling techniques to handle long texts effectively. We have validated the model's performance on context lengths of up to 131,072 tokens using the [YaRN](https://arxiv.org/abs/2309.00071) method.
YaRN is currently supported by several inference frameworks, e.g., `transformers` and `llama.cpp` for local use, `vllm` and `sglang` for deployment. In general, there are two approaches to enabling YaRN for supported frameworks:
- Modifying the model files:
In the `config.json` file, add the `rope_scaling` fields:
```json
{
...,
"rope_scaling": {
"rope_type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
}
```
For `llama.cpp`, you need to regenerate the GGUF file after the modification.
- Passing command line arguments:
For `vllm`, you can use
```shell
vllm serve ... --rope-scaling '{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}' --max-model-len 131072
```
For `sglang`, you can use
```shell
python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling":{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}}'
```
For `llama-server` from `llama.cpp`, you can use
```shell
llama-server ... --rope-scaling yarn --rope-scale 4 --yarn-orig-ctx 32768
```
> [!IMPORTANT]
> If you encounter the following warning
> ```
> Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'original_max_position_embeddings'}
> ```
> please upgrade `transformers>=4.51.0`.
> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**
> We advise adding the `rope_scaling` configuration only when processing long contexts is required.
> It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 65,536 tokens, it would be better to set `factor` as 2.0.
> [!NOTE]
If the average context length does not exceed 32,768 tokens, we do not recommend enabling YaRN in this scenario, as it may potentially degrade model performance.
## 🗂️ Dataset
To construct a challenging RL dataset for verifiable long-context reasoning, we develop [🤗 DocQA-RL-1.6K](https://huggingface.co/datasets/Tongyi-Zhiwen/DocQA-RL-1.6K), which comprises 1.6K DocQA problems across three reasoning domains:
(1) Mathematical Reasoning: We use 600 problems from the DocMath dataset, requiring numerical reasoning across long and specialized documents such as financial reports. For DocMath, we sample 75% items from each subset from its valid split for training and 25% for evaluation;
(2) Logical Reasoning: We employ DeepSeek-R1 to synthesize 600 multi-choice questions requiring logic analysis of real-world documents spanning legal, financial, insurance, and production domains from our curated collection;
(3) Multi-Hop Reasoning: We sample 200 examples from MultiHopRAG and 200 examples from Musique, emphasizing cross-document reasoning.
Please download and put the following datasets in `./datasets/` for training and evaluation.
RL training data: [🤗 DocQA-RL-1.6K](https://huggingface.co/datasets/Tongyi-Zhiwen/DocQA-RL-1.6K).
Evaluation data: [🤗 docmath](https://huggingface.co/datasets/Tongyi-Zhiwen/docmath), [🤗 frames](https://huggingface.co/datasets/Tongyi-Zhiwen/frames), [🤗 longbench](https://huggingface.co/datasets/Tongyi-Zhiwen/longbench).
## 💻 Training
We provide the basic demo training code for single stage RL traininig with DAPO.
First, we should start a local verifier.
```bash
export CUDA_VISIBLE_DEVICES=0
vllm serve "Qwen/Qwen2.5-1.5B-Instruct" \
--host 0.0.0.0 \
--port 23547
```
Then, we start RL training with 4 nodes.
```bash
export PROJ_DIR="<YOUR_PROJ_DIR_HERE>"
export MASTER_IP="<YOUR_MASTER_IP_HERE>" # ray master ip
export NNODES=4 # total GPU nodes
export NODE_RANK=${RANK} # rank of current node
export PORT=6382
export WANDB_API_KEY="<YOUR_WANDB_API_KEY_HERE>"
export WANDB_PROJECT="QwenLong-L1"
export LLM_JUDGE=Y # 'Y': LLM JUDGE, 'N': RULE BASED
export VLLM_ATTENTION_BACKEND=FLASH_ATTN
# verifier
export VERIFIER_PATH="Qwen/Qwen2.5-1.5B-Instruct"
export VERIFIER_HOST="<YOUR_VERIFIER_HOST_HERE>"
export VERIFIER_PORT="23547"
ray_start_retry() {
while true; do
ray start --address="${MASTER_IP}:${PORT}"
if [ $? -eq 0 ]; then
break
fi
echo "Failed to connect to master, retrying in 5 seconds..."
sleep 5
done
}
check_ray_status() {
until ray status >/dev/null 2>&1; do
echo "Waiting for Ray cluster to be ready..."
sleep 5
done
}
if [ "$RANK" == "0" ]; then
echo "Starting HEAD node..."
ray start --head --port=${PORT}
check_ray_status
echo "Ray head node started successfully"
else
echo "Starting WORKER node..."
ray_start_retry
check_ray_status
echo "Successfully joined Ray cluster"
fi
if [ "$RANK" == "0" ]; then
bash ${PROJ_DIR}/scripts/rl_4nodes_dapo.sh 2>&1 | tee ${PROJ_DIR}/logs/rl_log_$(date +%Y%m%d_%H%M%S).txt &
else
sleep 30d
fi
wait
```
## 📊 Evaluation
We conduct evaluation on seven long-context DocQA benchmarks, including multi-hop reasoning benchmarks such as 2WikiMultihopQA, HotpotQA, Musique, NarrativeQA, Qasper, and Frames as well as mathematical reasoning benchmarks like DocMath. We report the maximum of exact match and LLM-judged accuracy as the final score, aligned with the reward function in our RL training process. We use DeepSeek-V3 as the judge model with a temperature of 0.0 to provide a reliable evaluation.
```bash
# Step 1. Serve the model for evaluation
export CUDA_VISIBLE_DEVICES="0,1,2,3,4,5,6,7"
MODEL_NAME="QwenLong-L1-32B"
MODEL_PATH="Tongyi-Zhiwen/QwenLong-L1-32B"
vllm serve ${MODEL_PATH} \
--port 23547 \
--api-key "token-abc123" \
--tensor-parallel-size 8 \
--gpu-memory-utilization 0.95 \
--max_model_len 131072 \
--trust-remote-code
# Step 2. Generate model responses for each dataset
export SERVE_HOST="<YOUR_SERVE_HOST_HERE>" # e.g., 127.0.0.1
export SERVE_PORT="23547"
PROJ_DIR="<YOUR_PROJ_DIR_HERE>"
DATA="<YOUR_DATA_HERE>" # e.g., docmath, frames, 2wikimqa, hotpotqa, musique, narrativeqa, pasper
python ${PROJ_DIR}/eval/${DATA}.py \
--save_dir "${PROJ_DIR}/eval/results/${DATA}" \
--save_file "${MODEL_NAME}" \
--model "${MODEL_PATH}" \
--tokenizer "${MODEL_PATH}" \
--n_proc 16 \
--api "openai"
# Step 3. Verify model responses for each dataset
export VERIFIER_API="<YOUR_API_KEY_HERE>"
export VERIFIER_URL="https://api.deepseek.com/v1"
PROJ_DIR="<YOUR_PROJ_DIR_HERE>"
DATA="<YOUR_DATA_HERE>" # e.g., docmath, frames, 2wikimqa, hotpotqa, musique, narrativeqa, pasper
python ${PROJ_DIR}/eval/${DATA}_verify.py \
--save_dir "${PROJ_DIR}/results/${DATA}" \
--save_file "${MODEL_NAME}" \
--judge_model "deepseek-chat" \
--batch_size 20
```
## 📝 Citation
If you find this work is relevant with your research or applications, please feel free to cite our work!
```
@article{wan2025qwenlongl1,
title={QwenLong-L1: : Towards Long-Context Large Reasoning Models with Reinforcement Learning},
author={Fanqi Wan, Weizhou Shen, Shengyi Liao, Yingcheng Shi, Chenliang Li, Ziyi Yang, Ji Zhang, Fei Huang, Jingren Zhou, Ming Yan},
journal={arXiv preprint arXiv:2505.17667},
year={2025}
}
``` |
Mungert/Qwen3-16B-A3B-GGUF | Mungert | 2025-06-15T19:38:08Z | 3,213 | 4 | null | [
"gguf",
"unsloth",
"qwen3",
"qwen",
"base_model:kalomaze/Qwen3-16B-A3B",
"base_model:quantized:kalomaze/Qwen3-16B-A3B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-05-22T11:04:29Z | ---
tags:
- unsloth
- qwen3
- qwen
license: apache-2.0
base_model:
- kalomaze/Qwen3-16B-A3B
---
# <span style="color: #7FFF7F;">Qwen3-16B-A3B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`f5cd27b7`](https://github.com/ggerganov/llama.cpp/commit/f5cd27b71da3ac375a04a41643d14fc779a8057b).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-16B-A3B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-16B-A3B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-16B-A3B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-16B-A3B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-16B-A3B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-16B-A3B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-16B-A3B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-16B-A3B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-16B-A3B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-16B-A3B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-16B-A3B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Qwen3-16B-A3B
Qwen3-16B-A3B is a rendition of Qwen3-30B-A3B by [kalomaze](https://huggingface.co/kalomaze/Qwen3-16B-A3B).
A man-made horror beyond your comprehension.
But no, seriously, this is my experiment to:
- measure the probability that any given expert will activate (over my personal set of fairly diverse calibration data), per layer
- prune 64/128 of the least used experts per layer (with reordered router and indexing per layer)
It can still write semi-coherently without any additional training or distillation done on top of it from the original 30b MoE.
The .txt files with the original measurements are provided in the repo along with the exported weights.
Custom testing to measure the experts was done on a hacked version of vllm, and then I made a bespoke script to selectively export the weights according to the measurements. |
Mungert/Llama-3.1-Nemotron-Nano-4B-v1.1-GGUF | Mungert | 2025-06-15T19:38:04Z | 2,600 | 1 | transformers | [
"transformers",
"gguf",
"nvidia",
"llama-3",
"pytorch",
"text-generation",
"en",
"dataset:nvidia/Llama-Nemotron-Post-Training-Dataset",
"arxiv:2408.11796",
"arxiv:2502.00203",
"arxiv:2505.00949",
"base_model:nvidia/Llama-3.1-Minitron-4B-Width-Base",
"base_model:quantized:nvidia/Llama-3.1-Minitron-4B-Width-Base",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-05-21T21:58:58Z | ---
library_name: transformers
license: other
license_name: nvidia-open-model-license
license_link: >-
https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/
pipeline_tag: text-generation
language:
- en
tags:
- nvidia
- llama-3
- pytorch
base_model:
- nvidia/Llama-3.1-Minitron-4B-Width-Base
datasets:
- nvidia/Llama-Nemotron-Post-Training-Dataset
---
# <span style="color: #7FFF7F;">Llama-3.1-Nemotron-Nano-4B-v1.1 GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`92ecdcc0`](https://github.com/ggerganov/llama.cpp/commit/92ecdcc06a4c405a415bcaa0cb772bc560aa23b1).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.1-Nemotron-Nano-4B-v1.1-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.1-Nemotron-Nano-4B-v1.1-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.1-Nemotron-Nano-4B-v1.1-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Llama-3.1-Nemotron-Nano-4B-v1.1
## Model Overview
Llama-3.1-Nemotron-Nano-4B-v1.1 is a large language model (LLM) which is a derivative of [nvidia/Llama-3.1-Minitron-4B-Width-Base](https://huggingface.co/nvidia/Llama-3.1-Minitron-4B-Width-Base), which is created from Llama 3.1 8B using [our LLM compression technique](https://arxiv.org/abs/2408.11796) and offers improvements in model accuracy and efficiency. It is a reasoning model that is post trained for reasoning, human chat preferences, and tasks, such as RAG and tool calling.
Llama-3.1-Nemotron-Nano-4B-v1.1 is a model which offers a great tradeoff between model accuracy and efficiency. The model fits on a single RTX GPU and can be used locally. The model supports a context length of 128K.
This model underwent a multi-phase post-training process to enhance both its reasoning and non-reasoning capabilities. This includes a supervised fine-tuning stage for Math, Code, Reasoning, and Tool Calling as well as multiple reinforcement learning (RL) stages using Reward-aware Preference Optimization (RPO) algorithms for both chat and instruction-following. The final model checkpoint is obtained after merging the final SFT and RPO checkpoints
This model is part of the Llama Nemotron Collection. You can find the other model(s) in this family here:
- [Llama-3.3-Nemotron-Ultra-253B-v1](https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1)
- [Llama-3.3-Nemotron-Super-49B-v1](https://huggingface.co/nvidia/Llama-3.3-Nemotron-Super-49B-v1)
- [Llama-3.1-Nemotron-Nano-8B-v1](https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-8B-v1)
This model is ready for commercial use.
## License/Terms of Use
GOVERNING TERMS: Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.1 Community License Agreement](https://www.llama.com/llama3_1/license/). Built with Llama.
**Model Developer:** NVIDIA
**Model Dates:** Trained between August 2024 and May 2025
**Data Freshness:** The pretraining data has a cutoff of June 2023.
## Use Case:
Developers designing AI Agent systems, chatbots, RAG systems, and other AI-powered applications. Also suitable for typical instruction-following tasks. Balance of model accuracy and compute efficiency (the model fits on a single RTX GPU and can be used locally).
## Release Date: <br>
5/20/2025 <br>
## References
- [\[2408.11796\] LLM Pruning and Distillation in Practice: The Minitron Approach](https://arxiv.org/abs/2408.11796)
- [\[2502.00203\] Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment](https://arxiv.org/abs/2502.00203)
- [\[2505.00949\] Llama-Nemotron: Efficient Reasoning Models](https://arxiv.org/abs/2505.00949)
## Model Architecture
**Architecture Type:** Dense decoder-only Transformer model
**Network Architecture:** Llama 3.1 Minitron Width 4B Base
## Intended use
Llama-3.1-Nemotron-Nano-4B-v1.1 is a general purpose reasoning and chat model intended to be used in English and coding languages. Other non-English languages (German, French, Italian, Portuguese, Hindi, Spanish, and Thai) are also supported.
# Input:
- **Input Type:** Text
- **Input Format:** String
- **Input Parameters:** One-Dimensional (1D)
- **Other Properties Related to Input:** Context length up to 131,072 tokens
## Output:
- **Output Type:** Text
- **Output Format:** String
- **Output Parameters:** One-Dimensional (1D)
- **Other Properties Related to Output:** Context length up to 131,072 tokens
## Model Version:
1.1 (5/20/2025)
## Software Integration
- **Runtime Engine:** NeMo 24.12 <br>
- **Recommended Hardware Microarchitecture Compatibility:**
- NVIDIA Hopper
- NVIDIA Ampere
## Quick Start and Usage Recommendations:
1. Reasoning mode (ON/OFF) is controlled via the system prompt, which must be set as shown in the example below. All instructions should be contained within the user prompt
2. We recommend setting temperature to `0.6`, and Top P to `0.95` for Reasoning ON mode
3. We recommend using greedy decoding for Reasoning OFF mode
4. We have provided a list of prompts to use for evaluation for each benchmark where a specific template is required
See the snippet below for usage with Hugging Face Transformers library. Reasoning mode (ON/OFF) is controlled via system prompt. Please see the example below.
Our code requires the transformers package version to be `4.44.2` or higher.
### Example of “Reasoning On:”
```python
import torch
import transformers
model_id = "nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
temperature=0.6,
top_p=0.95,
**model_kwargs
)
# Thinking can be "on" or "off"
thinking = "on"
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))
```
### Example of “Reasoning Off:”
```python
import torch
import transformers
model_id = "nvidia/Llama-3.1-Nemotron-Nano-4B-v1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
do_sample=False,
**model_kwargs
)
# Thinking can be "on" or "off"
thinking = "off"
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}]))
```
For some prompts, even though thinking is disabled, the model emergently prefers to think before responding. But if desired, the users can prevent it by pre-filling the assistant response.
```python
import torch
import transformers
model_id = "nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1"
model_kwargs = {"torch_dtype": torch.bfloat16, "device_map": "auto"}
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)
tokenizer.pad_token_id = tokenizer.eos_token_id
# Thinking can be "on" or "off"
thinking = "off"
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
max_new_tokens=32768,
do_sample=False,
**model_kwargs
)
print(pipeline([{"role": "system", "content": f"detailed thinking {thinking}"}, {"role": "user", "content": "Solve x*(sin(x)+2)=0"}, {"role":"assistant", "content":"<think>\n</think>"}]))
```
## Running a vLLM Server with Tool-call Support
Llama-3.1-Nemotron-Nano-4B-v1.1 supports tool calling. This HF repo hosts a tool-callilng parser as well as a chat template in Jinja, which can be used to launch a vLLM server.
Here is a shell script example to launch a vLLM server with tool-call support. `vllm/vllm-openai:v0.6.6` or newer should support the model.
```shell
#!/bin/bash
CWD=$(pwd)
PORT=5000
git clone https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1
docker run -it --rm \
--runtime=nvidia \
--gpus all \
--shm-size=16GB \
-p ${PORT}:${PORT} \
-v ${CWD}:${CWD} \
vllm/vllm-openai:v0.6.6 \
--model $CWD/Llama-3.1-Nemotron-Nano-4B-v1.1 \
--trust-remote-code \
--seed 1 \
--host "0.0.0.0" \
--port $PORT \
--served-model-name "Llama-Nemotron-Nano-4B-v1.1" \
--tensor-parallel-size 1 \
--max-model-len 131072 \
--gpu-memory-utilization 0.95 \
--enforce-eager \
--enable-auto-tool-choice \
--tool-parser-plugin "${CWD}/Llama-3.1-Nemotron-Nano-4B-v1.1/llama_nemotron_nano_toolcall_parser.py" \
--tool-call-parser "llama_nemotron_json" \
--chat-template "${CWD}/Llama-3.1-Nemotron-Nano-4B-v1.1/llama_nemotron_nano_generic_tool_calling.jinja"
```
Alternatively, you can use a virtual environment to launch a vLLM server like below.
```console
$ git clone https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1
$ conda create -n vllm python=3.12 -y
$ conda activate vllm
$ python -m vllm.entrypoints.openai.api_server \
--model Llama-3.1-Nemotron-Nano-4B-v1.1 \
--trust-remote-code \
--seed 1 \
--host "0.0.0.0" \
--port 5000 \
--served-model-name "Llama-Nemotron-Nano-4B-v1.1" \
--tensor-parallel-size 1 \
--max-model-len 131072 \
--gpu-memory-utilization 0.95 \
--enforce-eager \
--enable-auto-tool-choice \
--tool-parser-plugin "Llama-3.1-Nemotron-Nano-4B-v1.1/llama_nemotron_nano_toolcall_parser.py" \
--tool-call-parser "llama_nemotron_json" \
--chat-template "Llama-3.1-Nemotron-Nano-4B-v1.1/llama_nemotron_nano_generic_tool_calling.jinja"
```
After launching a vLLM server, you can call the server with tool-call support using a Python script like below.
```python
>>> from openai import OpenAI
>>> client = OpenAI(
base_url="http://0.0.0.0:5000/v1",
api_key="dummy",
)
>>> completion = client.chat.completions.create(
model="Llama-Nemotron-Nano-4B-v1.1",
messages=[
{"role": "system", "content": "detailed thinking on"},
{"role": "user", "content": "My bill is $100. What will be the amount for 18% tip?"},
],
tools=[
{"type": "function", "function": {"name": "calculate_tip", "parameters": {"type": "object", "properties": {"bill_total": {"type": "integer", "description": "The total amount of the bill"}, "tip_percentage": {"type": "integer", "description": "The percentage of tip to be applied"}}, "required": ["bill_total", "tip_percentage"]}}},
{"type": "function", "function": {"name": "convert_currency", "parameters": {"type": "object", "properties": {"amount": {"type": "integer", "description": "The amount to be converted"}, "from_currency": {"type": "string", "description": "The currency code to convert from"}, "to_currency": {"type": "string", "description": "The currency code to convert to"}}, "required": ["from_currency", "amount", "to_currency"]}}},
],
)
>>> completion.choices[0].message.content
'<think>\nOkay, let\'s see. The user has a bill of $100 and wants to know the amount of a 18% tip. So, I need to calculate the tip amount. The available tools include calculate_tip, which requires bill_total and tip_percentage. The parameters are both integers. The bill_total is 100, and the tip percentage is 18. So, the function should multiply 100 by 18% and return 18.0. But wait, maybe the user wants the total including the tip? The question says "the amount for 18% tip," which could be interpreted as the tip amount itself. Since the function is called calculate_tip, it\'s likely that it\'s designed to compute the tip, not the total. So, using calculate_tip with bill_total=100 and tip_percentage=18 should give the correct result. The other function, convert_currency, isn\'t relevant here. So, I should call calculate_tip with those values.\n</think>\n\n'
>>> completion.choices[0].message.tool_calls
[ChatCompletionMessageToolCall(id='chatcmpl-tool-2972d86817344edc9c1e0f9cd398e999', function=Function(arguments='{"bill_total": 100, "tip_percentage": 18}', name='calculate_tip'), type='function')]
```
## Inference:
**Engine:** Transformers
**Test Hardware:**
- BF16:
- 1x RTX 50 Series GPUs
- 1x RTX 40 Series GPUs
- 1x RTX 30 Series GPUs
- 1x H100-80GB GPU
- 1x A100-80GB GPU
**Preferred/Supported] Operating System(s):** Linux <br>
## Training Datasets
A large variety of training data was used for the post-training pipeline, including manually annotated data and synthetic data.
The data for the multi-stage post-training phases for improvements in Code, Math, and Reasoning is a compilation of SFT and RL data that supports improvements of math, code, general reasoning, and instruction following capabilities of the original Llama instruct model.
Prompts have been sourced from either public and open corpus or synthetically generated. Responses were synthetically generated by a variety of models, with some prompts containing responses for both Reasoning On and Off modes, to train the model to distinguish between two modes.
**Data Collection for Training Datasets:** <br>
* Hybrid: Automated, Human, Synthetic <br>
**Data Labeling for Training Datasets:** <br>
* N/A <br>
## Evaluation Datasets
We used the datasets listed below to evaluate Llama-3.1-Nemotron-Nano-4B-v1.1.
**Data Collection for Evaluation Datasets:** Hybrid: Human/Synthetic
**Data Labeling for Evaluation Datasets:** Hybrid: Human/Synthetic/Automatic
## Evaluation Results
These results contain both “Reasoning On”, and “Reasoning Off”. We recommend using temperature=`0.6`, top_p=`0.95` for “Reasoning On” mode, and greedy decoding for “Reasoning Off” mode. All evaluations are done with 32k sequence length. We run the benchmarks up to 16 times and average the scores to be more accurate.
> NOTE: Where applicable, a Prompt Template will be provided. While completing benchmarks, please ensure that you are parsing for the correct output format as per the provided prompt in order to reproduce the benchmarks seen below.
### MT-Bench
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 7.4 |
| Reasoning On | 8.0 |
### MATH500
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 71.8% |
| Reasoning On | 96.2% |
User Prompt Template:
```
"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \boxed{}.\nQuestion: {question}"
```
### AIME25
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 13.3% |
| Reasoning On | 46.3% |
User Prompt Template:
```
"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \boxed{}.\nQuestion: {question}"
```
### GPQA-D
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 33.8% |
| Reasoning On | 55.1% |
User Prompt Template:
```
"What is the correct answer to this question: {question}\nChoices:\nA. {option_A}\nB. {option_B}\nC. {option_C}\nD. {option_D}\nLet's think step by step, and put the final answer (should be a single letter A, B, C, or D) into a \boxed{}"
```
### IFEval
| Reasoning Mode | Strict:Prompt | Strict:Instruction |
|--------------|------------|------------|
| Reasoning Off | 70.1% | 78.5% |
| Reasoning On | 75.5% | 82.6% |
### BFCL v2 Live
| Reasoning Mode | Score |
|--------------|------------|
| Reasoning Off | 63.6% |
| Reasoning On | 67.9% |
User Prompt Template:
```
<AVAILABLE_TOOLS>{functions}</AVAILABLE_TOOLS>
{user_prompt}
```
### MBPP 0-shot
| Reasoning Mode | pass@1 |
|--------------|------------|
| Reasoning Off | 61.9% |
| Reasoning On | 85.8% |
User Prompt Template:
````
You are an exceptionally intelligent coding assistant that consistently delivers accurate and reliable responses to user instructions.
@@ Instruction
Here is the given problem and test examples:
{prompt}
Please use the python programming language to solve this problem.
Please make sure that your code includes the functions from the test samples and that the input and output formats of these functions match the test samples.
Please return all completed codes in one code block.
This code block should be in the following format:
```python
# Your codes here
```
````
## Ethical Considerations:
NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
For more detailed information on ethical considerations for this model, please see the Model Card++ [Explainability](explainability.md), [Bias](bias.md), [Safety & Security](safety.md), and [Privacy](privacy.md) Subcards.
Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/). |
Mungert/Kevin-32B-GGUF | Mungert | 2025-06-15T19:37:54Z | 1,559 | 3 | null | [
"gguf",
"base_model:Qwen/QwQ-32B",
"base_model:quantized:Qwen/QwQ-32B",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-05-16T17:08:06Z | ---
base_model:
- Qwen/QwQ-32B
---
# <span style="color: #7FFF7F;">Kevin-32B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`06c1e4ab`](https://github.com/ggerganov/llama.cpp/commit/06c1e4abc1e50ef6dcf6369f9685ca8fe82d21fe).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Kevin-32B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Kevin-32B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Kevin-32B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Kevin-32B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Kevin-32B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Kevin-32B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Kevin-32B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Kevin-32B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Kevin-32B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Kevin-32B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Kevin-32B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
## Kevin-32B
Kevin (K(ernel D)evin) is a 32B parameter model finetuned to write efficient CUDA kernels.
We use KernelBench as our benchmark, and train the model through multi-turn reinforcement learning.
For the details, see our blogpost at **https://cognition.ai/blog/kevin-32b**
<img src="https://cdn-uploads.huggingface.co/production/uploads/6490673eb67028b2aa2a6f9f/pjVt_KrR6aWCuWb4GcZuW.png" width="600"/>
|
Mungert/Qwen3-30B-A3B-GGUF | Mungert | 2025-06-15T19:37:50Z | 1,722 | 4 | transformers | [
"transformers",
"gguf",
"text-generation",
"arxiv:2309.00071",
"base_model:Qwen/Qwen3-30B-A3B-Base",
"base_model:quantized:Qwen/Qwen3-30B-A3B-Base",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-05-13T04:35:31Z | ---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3-30B-A3B/blob/main/LICENSE
pipeline_tag: text-generation
base_model:
- Qwen/Qwen3-30B-A3B-Base
---
# <span style="color: #7FFF7F;">Qwen3-30B-A3B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`064cc596`](https://github.com/ggerganov/llama.cpp/commit/064cc596ac44308dc326a17c9e3163c34a6f29d1).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-30B-A3B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-30B-A3B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-30B-A3B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-30B-A3B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-30B-A3B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-30B-A3B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-30B-A3B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-30B-A3B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-30B-A3B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-30B-A3B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-30B-A3B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# Qwen3-30B-A3B
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Qwen3 Highlights
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
- **Uniquely support of seamless switching between thinking mode** (for complex logical reasoning, math, and coding) and **non-thinking mode** (for efficient, general-purpose dialogue) **within single model**, ensuring optimal performance across various scenarios.
- **Significantly enhancement in its reasoning capabilities**, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
- **Superior human preference alignment**, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
- **Expertise in agent capabilities**, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
- **Support of 100+ languages and dialects** with strong capabilities for **multilingual instruction following** and **translation**.
## Model Overview
**Qwen3-30B-A3B** has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Number of Parameters: 30.5B in total and 3.3B activated
- Number of Paramaters (Non-Embedding): 29.9B
- Number of Layers: 48
- Number of Attention Heads (GQA): 32 for Q and 4 for KV
- Number of Experts: 128
- Number of Activated Experts: 8
- Context Length: 32,768 natively and [131,072 tokens with YaRN](#processing-long-texts).
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).
## Quickstart
The code of Qwen3-MoE has been in the latest Hugging Face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.51.0`, you will encounter the following error:
```
KeyError: 'qwen3_moe'
```
The following contains a code snippet illustrating how to use the model generate content based on given inputs.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-30B-A3B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 (</think>)
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
For deployment, you can use `sglang>=0.4.6.post1` or `vllm>=0.8.5` or to create an OpenAI-compatible API endpoint:
- SGLang:
```shell
python -m sglang.launch_server --model-path Qwen/Qwen3-30B-A3B --reasoning-parser qwen3
```
- vLLM:
```shell
vllm serve Qwen/Qwen3-30B-A3B --enable-reasoning --reasoning-parser deepseek_r1
```
For local use, applications such as Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers have also supported Qwen3.
## Switching Between Thinking and Non-Thinking Mode
> [!TIP]
> The `enable_thinking` switch is also available in APIs created by SGLang and vLLM.
> Please refer to our documentation for [SGLang](https://qwen.readthedocs.io/en/latest/deployment/sglang.html#thinking-non-thinking-modes) and [vLLM](https://qwen.readthedocs.io/en/latest/deployment/vllm.html#thinking-non-thinking-modes) users.
### `enable_thinking=True`
By default, Qwen3 has thinking capabilities enabled, similar to QwQ-32B. This means the model will use its reasoning abilities to enhance the quality of generated responses. For example, when explicitly setting `enable_thinking=True` or leaving it as the default value in `tokenizer.apply_chat_template`, the model will engage its thinking mode.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # True is the default value for enable_thinking
)
```
In this mode, the model will generate think content wrapped in a `<think>...</think>` block, followed by the final response.
> [!NOTE]
> For thinking mode, use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0` (the default setting in `generation_config.json`). **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### `enable_thinking=False`
We provide a hard switch to strictly disable the model's thinking behavior, aligning its functionality with the previous Qwen2.5-Instruct models. This mode is particularly useful in scenarios where disabling thinking is essential for enhancing efficiency.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False # Setting enable_thinking=False disables thinking mode
)
```
In this mode, the model will not generate any think content and will not include a `<think>...</think>` block.
> [!NOTE]
> For non-thinking mode, we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### Advanced Usage: Switching Between Thinking and Non-Thinking Modes via User Input
We provide a soft switch mechanism that allows users to dynamically control the model's behavior when `enable_thinking=True`. Specifically, you can add `/think` and `/no_think` to user prompts or system messages to switch the model's thinking mode from turn to turn. The model will follow the most recent instruction in multi-turn conversations.
Here is an example of a multi-turn conversation:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
class QwenChatbot:
def __init__(self, model_name="Qwen/Qwen3-30B-A3B"):
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModelForCausalLM.from_pretrained(model_name)
self.history = []
def generate_response(self, user_input):
messages = self.history + [{"role": "user", "content": user_input}]
text = self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
inputs = self.tokenizer(text, return_tensors="pt")
response_ids = self.model.generate(**inputs, max_new_tokens=32768)[0][len(inputs.input_ids[0]):].tolist()
response = self.tokenizer.decode(response_ids, skip_special_tokens=True)
# Update history
self.history.append({"role": "user", "content": user_input})
self.history.append({"role": "assistant", "content": response})
return response
# Example Usage
if __name__ == "__main__":
chatbot = QwenChatbot()
# First input (without /think or /no_think tags, thinking mode is enabled by default)
user_input_1 = "How many r's in strawberries?"
print(f"User: {user_input_1}")
response_1 = chatbot.generate_response(user_input_1)
print(f"Bot: {response_1}")
print("----------------------")
# Second input with /no_think
user_input_2 = "Then, how many r's in blueberries? /no_think"
print(f"User: {user_input_2}")
response_2 = chatbot.generate_response(user_input_2)
print(f"Bot: {response_2}")
print("----------------------")
# Third input with /think
user_input_3 = "Really? /think"
print(f"User: {user_input_3}")
response_3 = chatbot.generate_response(user_input_3)
print(f"Bot: {response_3}")
```
> [!NOTE]
> For API compatibility, when `enable_thinking=True`, regardless of whether the user uses `/think` or `/no_think`, the model will always output a block wrapped in `<think>...</think>`. However, the content inside this block may be empty if thinking is disabled.
> When `enable_thinking=False`, the soft switches are not valid. Regardless of any `/think` or `/no_think` tags input by the user, the model will not generate think content and will not include a `<think>...</think>` block.
## Agentic Use
Qwen3 excels in tool calling capabilities. We recommend using [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.
To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.
```python
from qwen_agent.agents import Assistant
# Define LLM
llm_cfg = {
'model': 'Qwen3-30B-A3B',
# Use the endpoint provided by Alibaba Model Studio:
# 'model_type': 'qwen_dashscope',
# 'api_key': os.getenv('DASHSCOPE_API_KEY'),
# Use a custom endpoint compatible with OpenAI API:
'model_server': 'http://localhost:8000/v1', # api_base
'api_key': 'EMPTY',
# Other parameters:
# 'generate_cfg': {
# # Add: When the response content is `<think>this is the thought</think>this is the answer;
# # Do not add: When the response has been separated by reasoning_content and content.
# 'thought_in_content': True,
# },
}
# Define Tools
tools = [
{'mcpServers': { # You can specify the MCP configuration file
'time': {
'command': 'uvx',
'args': ['mcp-server-time', '--local-timezone=Asia/Shanghai']
},
"fetch": {
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
},
'code_interpreter', # Built-in tools
]
# Define Agent
bot = Assistant(llm=llm_cfg, function_list=tools)
# Streaming generation
messages = [{'role': 'user', 'content': 'https://qwenlm.github.io/blog/ Introduce the latest developments of Qwen'}]
for responses in bot.run(messages=messages):
pass
print(responses)
```
## Processing Long Texts
Qwen3 natively supports context lengths of up to 32,768 tokens. For conversations where the total length (including both input and output) significantly exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively. We have validated the model's performance on context lengths of up to 131,072 tokens using the [YaRN](https://arxiv.org/abs/2309.00071) method.
YaRN is currently supported by several inference frameworks, e.g., `transformers` and `llama.cpp` for local use, `vllm` and `sglang` for deployment. In general, there are two approaches to enabling YaRN for supported frameworks:
- Modifying the model files:
In the `config.json` file, add the `rope_scaling` fields:
```json
{
...,
"rope_scaling": {
"rope_type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
}
```
For `llama.cpp`, you need to regenerate the GGUF file after the modification.
- Passing command line arguments:
For `vllm`, you can use
```shell
vllm serve ... --rope-scaling '{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}' --max-model-len 131072
```
For `sglang`, you can use
```shell
python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling":{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}}'
```
For `llama-server` from `llama.cpp`, you can use
```shell
llama-server ... --rope-scaling yarn --rope-scale 4 --yarn-orig-ctx 32768
```
> [!IMPORTANT]
> If you encounter the following warning
> ```
> Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'original_max_position_embeddings'}
> ```
> please upgrade `transformers>=4.51.0`.
> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**
> We advise adding the `rope_scaling` configuration only when processing long contexts is required.
> It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 65,536 tokens, it would be better to set `factor` as 2.0.
> [!NOTE]
> The default `max_position_embeddings` in `config.json` is set to 40,960. This allocation includes reserving 32,768 tokens for outputs and 8,192 tokens for typical prompts, which is sufficient for most scenarios involving short text processing. If the average context length does not exceed 32,768 tokens, we do not recommend enabling YaRN in this scenario, as it may potentially degrade model performance.
> [!TIP]
> The endpoint provided by Alibaba Model Studio supports dynamic YaRN by default and no extra configuration is needed.
## Best Practices
To achieve optimal performance, we recommend the following settings:
1. **Sampling Parameters**:
- For thinking mode (`enable_thinking=True`), use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0`. **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions.
- For non-thinking mode (`enable_thinking=False`), we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.
2. **Adequate Output Length**: We recommend using an output length of 32,768 tokens for most queries. For benchmarking on highly complex problems, such as those found in math and programming competitions, we suggest setting the max output length to 38,912 tokens. This provides the model with sufficient space to generate detailed and comprehensive responses, thereby enhancing its overall performance.
3. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g., `"answer": "C"`."
4. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. It is implemented in the provided chat template in Jinja2. However, for frameworks that do not directly use the Jinja2 chat template, it is up to the developers to ensure that the best practice is followed.
### Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen3,
title = {Qwen3},
url = {https://qwenlm.github.io/blog/qwen3/},
author = {Qwen Team},
month = {April},
year = {2025}
}
``` |
Mungert/Qwen3-4B-GGUF | Mungert | 2025-06-15T19:37:41Z | 891 | 8 | transformers | [
"transformers",
"gguf",
"text-generation",
"arxiv:2309.00071",
"base_model:Qwen/Qwen3-4B-Base",
"base_model:quantized:Qwen/Qwen3-4B-Base",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-30T01:08:05Z | ---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3-4B/blob/main/LICENSE
pipeline_tag: text-generation
base_model:
- Qwen/Qwen3-4B-Base
---
# <span style="color: #7FFF7F;">Qwen3-4B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`19e899c`](https://github.com/ggerganov/llama.cpp/commit/19e899ce21a7c9ffcf8bb2b22269a75f6e078f8f).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Qwen3-4B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Qwen3-4B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Qwen3-4B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Qwen3-4B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Qwen3-4B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Qwen3-4B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Qwen3-4B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Qwen3-4B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Qwen3-4B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Qwen3-4B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Qwen3-4B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Qwen3-4B
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Qwen3 Highlights
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
- **Uniquely support of seamless switching between thinking mode** (for complex logical reasoning, math, and coding) and **non-thinking mode** (for efficient, general-purpose dialogue) **within single model**, ensuring optimal performance across various scenarios.
- **Significantly enhancement in its reasoning capabilities**, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
- **Superior human preference alignment**, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
- **Expertise in agent capabilities**, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
- **Support of 100+ languages and dialects** with strong capabilities for **multilingual instruction following** and **translation**.
## Model Overview
**Qwen3-4B** has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Number of Parameters: 4.0B
- Number of Paramaters (Non-Embedding): 3.6B
- Number of Layers: 36
- Number of Attention Heads (GQA): 32 for Q and 8 for KV
- Context Length: 32,768 natively and [131,072 tokens with YaRN](#processing-long-texts).
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).
> [!TIP]
> If you encounter significant endless repetitions, please refer to the [Best Practices](#best-practices) section for optimal sampling parameters, and set the ``presence_penalty`` to 1.5.
## Quickstart
The code of Qwen3 has been in the latest Hugging Face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.51.0`, you will encounter the following error:
```
KeyError: 'qwen3'
```
The following contains a code snippet illustrating how to use the model generate content based on given inputs.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-4B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 (</think>)
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
For deployment, you can use `sglang>=0.4.6.post1` or `vllm>=0.8.5` or to create an OpenAI-compatible API endpoint:
- SGLang:
```shell
python -m sglang.launch_server --model-path Qwen/Qwen3-4B --reasoning-parser qwen3
```
- vLLM:
```shell
vllm serve Qwen/Qwen3-4B --enable-reasoning --reasoning-parser deepseek_r1
```
For local use, applications such as Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers have also supported Qwen3.
## Switching Between Thinking and Non-Thinking Mode
> [!TIP]
> The `enable_thinking` switch is also available in APIs created by SGLang and vLLM.
> Please refer to our documentation for [SGLang](https://qwen.readthedocs.io/en/latest/deployment/sglang.html#thinking-non-thinking-modes) and [vLLM](https://qwen.readthedocs.io/en/latest/deployment/vllm.html#thinking-non-thinking-modes) users.
### `enable_thinking=True`
By default, Qwen3 has thinking capabilities enabled, similar to QwQ-32B. This means the model will use its reasoning abilities to enhance the quality of generated responses. For example, when explicitly setting `enable_thinking=True` or leaving it as the default value in `tokenizer.apply_chat_template`, the model will engage its thinking mode.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # True is the default value for enable_thinking
)
```
In this mode, the model will generate think content wrapped in a `<think>...</think>` block, followed by the final response.
> [!NOTE]
> For thinking mode, use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0` (the default setting in `generation_config.json`). **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### `enable_thinking=False`
We provide a hard switch to strictly disable the model's thinking behavior, aligning its functionality with the previous Qwen2.5-Instruct models. This mode is particularly useful in scenarios where disabling thinking is essential for enhancing efficiency.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False # Setting enable_thinking=False disables thinking mode
)
```
In this mode, the model will not generate any think content and will not include a `<think>...</think>` block.
> [!NOTE]
> For non-thinking mode, we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### Advanced Usage: Switching Between Thinking and Non-Thinking Modes via User Input
We provide a soft switch mechanism that allows users to dynamically control the model's behavior when `enable_thinking=True`. Specifically, you can add `/think` and `/no_think` to user prompts or system messages to switch the model's thinking mode from turn to turn. The model will follow the most recent instruction in multi-turn conversations.
Here is an example of a multi-turn conversation:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
class QwenChatbot:
def __init__(self, model_name="Qwen/Qwen3-4B"):
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModelForCausalLM.from_pretrained(model_name)
self.history = []
def generate_response(self, user_input):
messages = self.history + [{"role": "user", "content": user_input}]
text = self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
inputs = self.tokenizer(text, return_tensors="pt")
response_ids = self.model.generate(**inputs, max_new_tokens=32768)[0][len(inputs.input_ids[0]):].tolist()
response = self.tokenizer.decode(response_ids, skip_special_tokens=True)
# Update history
self.history.append({"role": "user", "content": user_input})
self.history.append({"role": "assistant", "content": response})
return response
# Example Usage
if __name__ == "__main__":
chatbot = QwenChatbot()
# First input (without /think or /no_think tags, thinking mode is enabled by default)
user_input_1 = "How many r's in strawberries?"
print(f"User: {user_input_1}")
response_1 = chatbot.generate_response(user_input_1)
print(f"Bot: {response_1}")
print("----------------------")
# Second input with /no_think
user_input_2 = "Then, how many r's in blueberries? /no_think"
print(f"User: {user_input_2}")
response_2 = chatbot.generate_response(user_input_2)
print(f"Bot: {response_2}")
print("----------------------")
# Third input with /think
user_input_3 = "Really? /think"
print(f"User: {user_input_3}")
response_3 = chatbot.generate_response(user_input_3)
print(f"Bot: {response_3}")
```
> [!NOTE]
> For API compatibility, when `enable_thinking=True`, regardless of whether the user uses `/think` or `/no_think`, the model will always output a block wrapped in `<think>...</think>`. However, the content inside this block may be empty if thinking is disabled.
> When `enable_thinking=False`, the soft switches are not valid. Regardless of any `/think` or `/no_think` tags input by the user, the model will not generate think content and will not include a `<think>...</think>` block.
## Agentic Use
Qwen3 excels in tool calling capabilities. We recommend using [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.
To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.
```python
from qwen_agent.agents import Assistant
# Define LLM
llm_cfg = {
'model': 'Qwen3-4B',
# Use the endpoint provided by Alibaba Model Studio:
# 'model_type': 'qwen_dashscope',
# 'api_key': os.getenv('DASHSCOPE_API_KEY'),
# Use a custom endpoint compatible with OpenAI API:
'model_server': 'http://localhost:8000/v1', # api_base
'api_key': 'EMPTY',
# Other parameters:
# 'generate_cfg': {
# # Add: When the response content is `<think>this is the thought</think>this is the answer;
# # Do not add: When the response has been separated by reasoning_content and content.
# 'thought_in_content': True,
# },
}
# Define Tools
tools = [
{'mcpServers': { # You can specify the MCP configuration file
'time': {
'command': 'uvx',
'args': ['mcp-server-time', '--local-timezone=Asia/Shanghai']
},
"fetch": {
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
},
'code_interpreter', # Built-in tools
]
# Define Agent
bot = Assistant(llm=llm_cfg, function_list=tools)
# Streaming generation
messages = [{'role': 'user', 'content': 'https://qwenlm.github.io/blog/ Introduce the latest developments of Qwen'}]
for responses in bot.run(messages=messages):
pass
print(responses)
```
## Processing Long Texts
Qwen3 natively supports context lengths of up to 32,768 tokens. For conversations where the total length (including both input and output) significantly exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively. We have validated the model's performance on context lengths of up to 131,072 tokens using the [YaRN](https://arxiv.org/abs/2309.00071) method.
YaRN is currently supported by several inference frameworks, e.g., `transformers` and `llama.cpp` for local use, `vllm` and `sglang` for deployment. In general, there are two approaches to enabling YaRN for supported frameworks:
- Modifying the model files:
In the `config.json` file, add the `rope_scaling` fields:
```json
{
...,
"rope_scaling": {
"rope_type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
}
```
For `llama.cpp`, you need to regenerate the GGUF file after the modification.
- Passing command line arguments:
For `vllm`, you can use
```shell
vllm serve ... --rope-scaling '{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}' --max-model-len 131072
```
For `sglang`, you can use
```shell
python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling":{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":32768}}'
```
For `llama-server` from `llama.cpp`, you can use
```shell
llama-server ... --rope-scaling yarn --rope-scale 4 --yarn-orig-ctx 32768
```
> [!IMPORTANT]
> If you encounter the following warning
> ```
> Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'original_max_position_embeddings'}
> ```
> please upgrade `transformers>=4.51.0`.
> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**
> We advise adding the `rope_scaling` configuration only when processing long contexts is required.
> It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 65,536 tokens, it would be better to set `factor` as 2.0.
> [!NOTE]
> The default `max_position_embeddings` in `config.json` is set to 40,960. This allocation includes reserving 32,768 tokens for outputs and 8,192 tokens for typical prompts, which is sufficient for most scenarios involving short text processing. If the average context length does not exceed 32,768 tokens, we do not recommend enabling YaRN in this scenario, as it may potentially degrade model performance.
> [!TIP]
> The endpoint provided by Alibaba Model Studio supports dynamic YaRN by default and no extra configuration is needed.
## Best Practices
To achieve optimal performance, we recommend the following settings:
1. **Sampling Parameters**:
- For thinking mode (`enable_thinking=True`), use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0`. **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions.
- For non-thinking mode (`enable_thinking=False`), we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.
2. **Adequate Output Length**: We recommend using an output length of 32,768 tokens for most queries. For benchmarking on highly complex problems, such as those found in math and programming competitions, we suggest setting the max output length to 38,912 tokens. This provides the model with sufficient space to generate detailed and comprehensive responses, thereby enhancing its overall performance.
3. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g., `"answer": "C"`."
4. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. It is implemented in the provided chat template in Jinja2. However, for frameworks that do not directly use the Jinja2 chat template, it is up to the developers to ensure that the best practice is followed.
### Citation
If you find our work helpful, feel free to give us a cite.
```
@misc{qwen3,
title = {Qwen3},
url = {https://qwenlm.github.io/blog/qwen3/},
author = {Qwen Team},
month = {April},
year = {2025}
}
``` |
Mungert/LiveCC-7B-Instruct-GGUF | Mungert | 2025-06-15T19:37:37Z | 1,832 | 1 | null | [
"gguf",
"qwen_vl",
"video",
"real-time",
"multimodal",
"LLM",
"en",
"dataset:chenjoya/Live-CC-5M",
"dataset:chenjoya/Live-WhisperX-526K",
"dataset:lmms-lab/LLaVA-Video-178K",
"arxiv:2504.16030",
"base_model:Qwen/Qwen2-VL-7B",
"base_model:quantized:Qwen/Qwen2-VL-7B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-27T00:50:10Z | ---
license: apache-2.0
datasets:
- chenjoya/Live-CC-5M
- chenjoya/Live-WhisperX-526K
- lmms-lab/LLaVA-Video-178K
language:
- en
base_model:
- Qwen/Qwen2-VL-7B
tags:
- qwen_vl
- video
- real-time
- multimodal
- LLM
---
# <span style="color: #7FFF7F;">LiveCC-7B-Instruct GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`e291450`](https://github.com/ggerganov/llama.cpp/commit/e291450b7602d7a36239e4ceeece37625f838373).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `LiveCC-7B-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `LiveCC-7B-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `LiveCC-7B-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `LiveCC-7B-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `LiveCC-7B-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `LiveCC-7B-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `LiveCC-7B-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `LiveCC-7B-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `LiveCC-7B-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `LiveCC-7B-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `LiveCC-7B-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# LiveCC-7B-Instruct
## Introduction
We introduce LiveCC, the first video LLM capable of real-time commentary, trained with a novel video-ASR streaming method, SOTA on both streaming and offline benchmarks.
- Project Page: https://showlab.github.io/livecc
> [!Important]
> This is the SFT model. The base model is at [LiveCC-7B-Base](https://huggingface.co/chenjoya/LiveCC-7B-Base).
## Training with Streaming Frame-Words Paradigm
## Quickstart
### Gradio Demo
Please refer to https://github.com/showlab/livecc:
### Hands-on
Like qwen-vl-utils, we offer a toolkit to help you handle various types of visual input more conveniently, **especially on video streaming inputs**. You can install it using the following command:
```bash
pip install qwen-vl-utils livecc-utils liger_kernel
```
Here we show a code snippet to show you how to do **real-time video commentary** with `transformers` and the above utils:
```python
import functools, torch, os, tqdm
from liger_kernel.transformers import apply_liger_kernel_to_qwen2_vl
apply_liger_kernel_to_qwen2_vl() # important. our model is trained with this. keep consistency
from transformers import Qwen2VLForConditionalGeneration, AutoProcessor, LogitsProcessor, logging
from livecc_utils import prepare_multiturn_multimodal_inputs_for_generation, get_smart_resized_clip, get_smart_resized_video_reader
from qwen_vl_utils import process_vision_info
class LiveCCDemoInfer:
fps = 2
initial_fps_frames = 6
streaming_fps_frames = 2
initial_time_interval = initial_fps_frames / fps
streaming_time_interval = streaming_fps_frames / fps
frame_time_interval = 1 / fps
def __init__(self, model_path: str = None, device_id: int = 0):
self.model = Qwen2VLForConditionalGeneration.from_pretrained(
model_path, torch_dtype="auto",
device_map=f'cuda:{device_id}',
attn_implementation='flash_attention_2'
)
self.processor = AutoProcessor.from_pretrained(model_path, use_fast=False)
self.model.prepare_inputs_for_generation = functools.partial(prepare_multiturn_multimodal_inputs_for_generation, self.model)
message = {
"role": "user",
"content": [
{"type": "text", "text": 'livecc'},
]
}
texts = self.processor.apply_chat_template([message], tokenize=False)
self.system_prompt_offset = texts.index('<|im_start|>user')
self._cached_video_readers_with_hw = {}
def live_cc(
self,
query: str,
state: dict,
max_pixels: int = 384 * 28 * 28,
default_query: str = 'Please describe the video.',
do_sample: bool = True,
repetition_penalty: float = 1.05,
**kwargs,
):
"""
state: dict, (maybe) with keys:
video_path: str, video path
video_timestamp: float, current video timestamp
last_timestamp: float, last processed video timestamp
last_video_pts_index: int, last processed video frame index
video_pts: np.ndarray, video pts
last_history: list, last processed history
past_key_values: llm past_key_values
past_ids: past generated ids
"""
# 1. preparation: video_reader, and last processing info
video_timestamp, last_timestamp = state.get('video_timestamp', 0), state.get('last_timestamp', -1 / self.fps)
video_path = state['video_path']
if video_path not in self._cached_video_readers_with_hw:
self._cached_video_readers_with_hw[video_path] = get_smart_resized_video_reader(video_path, max_pixels)
video_reader = self._cached_video_readers_with_hw[video_path][0]
video_reader.get_frame_timestamp(0)
state['video_pts'] = torch.from_numpy(video_reader._frame_pts[:, 1])
state['last_video_pts_index'] = -1
video_pts = state['video_pts']
if last_timestamp + self.frame_time_interval > video_pts[-1]:
state['video_end'] = True
return
video_reader, resized_height, resized_width = self._cached_video_readers_with_hw[video_path]
last_video_pts_index = state['last_video_pts_index']
# 2. which frames will be processed
initialized = last_timestamp >= 0
if not initialized:
video_timestamp = max(video_timestamp, self.initial_time_interval)
if video_timestamp <= last_timestamp + self.frame_time_interval:
return
timestamps = torch.arange(last_timestamp + self.frame_time_interval, video_timestamp, self.frame_time_interval) # add compensation
# 3. fetch frames in required timestamps
clip, clip_timestamps, clip_idxs = get_smart_resized_clip(video_reader, resized_height, resized_width, timestamps, video_pts, video_pts_index_from=last_video_pts_index+1)
state['last_video_pts_index'] = clip_idxs[-1]
state['last_timestamp'] = clip_timestamps[-1]
# 4. organize to interleave frames
interleave_clips, interleave_timestamps = [], []
if not initialized:
interleave_clips.append(clip[:self.initial_fps_frames])
interleave_timestamps.append(clip_timestamps[:self.initial_fps_frames])
clip = clip[self.initial_fps_frames:]
clip_timestamps = clip_timestamps[self.initial_fps_frames:]
if len(clip) > 0:
interleave_clips.extend(list(clip.split(self.streaming_fps_frames)))
interleave_timestamps.extend(list(clip_timestamps.split(self.streaming_fps_frames)))
# 5. make conversation and send to model
for clip, timestamps in zip(interleave_clips, interleave_timestamps):
start_timestamp, stop_timestamp = timestamps[0].item(), timestamps[-1].item() + self.frame_time_interval
message = {
"role": "user",
"content": [
{"type": "text", "text": f'Time={start_timestamp:.1f}-{stop_timestamp:.1f}s'},
{"type": "video", "video": clip}
]
}
if not query and not state.get('query', None):
query = default_query
print(f'No query provided, use default_query={default_query}')
if query and state.get('query', None) != query:
message['content'].append({"type": "text", "text": query})
state['query'] = query
texts = self.processor.apply_chat_template([message], tokenize=False, add_generation_prompt=True, return_tensors='pt')
past_ids = state.get('past_ids', None)
if past_ids is not None:
texts = '<|im_end|>\n' + texts[self.system_prompt_offset:]
inputs = self.processor(
text=texts,
images=None,
videos=[clip],
return_tensors="pt",
return_attention_mask=False
)
inputs.to('cuda')
if past_ids is not None:
inputs['input_ids'] = torch.cat([past_ids, inputs.input_ids], dim=1)
outputs = self.model.generate(
**inputs, past_key_values=state.get('past_key_values', None),
return_dict_in_generate=True, do_sample=do_sample,
repetition_penalty=repetition_penalty,
)
state['past_key_values'] = outputs.past_key_values
state['past_ids'] = outputs.sequences[:, :-1]
yield (start_timestamp, stop_timestamp), self.processor.decode(outputs.sequences[0, inputs.input_ids.size(1):], skip_special_tokens=True), state
model_path = 'chenjoya/LiveCC-7B-Instruct'
# download a test video at: https://github.com/showlab/livecc/blob/main/demo/sources/howto_fix_laptop_mute_1080p.mp4
video_path = "demo/sources/howto_fix_laptop_mute_1080p.mp4"
query = "Please describe the video."
infer = LiveCCDemoInfer(model_path=model_path)
state = {'video_path': video_path}
commentaries = []
t = 0
for t in range(31):
state['video_timestamp'] = t
for (start_t, stop_t), response, state in infer.live_cc(
query=query, state=state,
max_pixels = 384 * 28 * 28, repetition_penalty=1.05,
streaming_eos_base_threshold=0.0, streaming_eos_threshold_step=0
):
print(f'{start_t}s-{stop_t}s: {response}')
commentaries.append([start_t, stop_t, response])
if state.get('video_end', False):
break
t += 1
```
Here we show a code snippet to show you how to do **common video (multi-turn) qa** with `transformers` and the above utils:
```python
import functools, torch
from liger_kernel.transformers import apply_liger_kernel_to_qwen2_vl
apply_liger_kernel_to_qwen2_vl() # important. our model is trained with this. keep consistency
from transformers import Qwen2VLForConditionalGeneration, AutoProcessor, LogitsProcessor, logging
from livecc_utils import prepare_multiturn_multimodal_inputs_for_generation, get_smart_resized_clip, get_smart_resized_video_reader
from qwen_vl_utils import process_vision_info
class LiveCCDemoInfer:
fps = 2
initial_fps_frames = 6
streaming_fps_frames = 2
initial_time_interval = initial_fps_frames / fps
streaming_time_interval = streaming_fps_frames / fps
frame_time_interval = 1 / fps
def __init__(self, model_path: str = None, device: str = 'cuda'):
self.model = Qwen2VLForConditionalGeneration.from_pretrained(
model_path, torch_dtype="auto",
device_map=device,
attn_implementation='flash_attention_2'
)
self.processor = AutoProcessor.from_pretrained(model_path, use_fast=False)
self.streaming_eos_token_id = self.processor.tokenizer(' ...').input_ids[-1]
self.model.prepare_inputs_for_generation = functools.partial(prepare_multiturn_multimodal_inputs_for_generation, self.model)
message = {
"role": "user",
"content": [
{"type": "text", "text": 'livecc'},
]
}
texts = self.processor.apply_chat_template([message], tokenize=False)
self.system_prompt_offset = texts.index('<|im_start|>user')
def video_qa(
self,
message: str,
state: dict,
do_sample: bool = True,
repetition_penalty: float = 1.05,
**kwargs,
):
"""
state: dict, (maybe) with keys:
video_path: str, video path
video_timestamp: float, current video timestamp
last_timestamp: float, last processed video timestamp
last_video_pts_index: int, last processed video frame index
video_pts: np.ndarray, video pts
last_history: list, last processed history
past_key_values: llm past_key_values
past_ids: past generated ids
"""
video_path = state.get('video_path', None)
conversation = []
past_ids = state.get('past_ids', None)
content = [{"type": "text", "text": message}]
if past_ids is None and video_path: # only use once
content.insert(0, {"type": "video", "video": video_path})
conversation.append({"role": "user", "content": content})
image_inputs, video_inputs = process_vision_info(conversation)
texts = self.processor.apply_chat_template(conversation, tokenize=False, add_generation_prompt=True, return_tensors='pt')
if past_ids is not None:
texts = '<|im_end|>\n' + texts[self.system_prompt_offset:]
inputs = self.processor(
text=texts,
images=image_inputs,
videos=video_inputs,
return_tensors="pt",
return_attention_mask=False
)
inputs.to(self.model.device)
if past_ids is not None:
inputs['input_ids'] = torch.cat([past_ids, inputs.input_ids], dim=1)
outputs = self.model.generate(
**inputs, past_key_values=state.get('past_key_values', None),
return_dict_in_generate=True, do_sample=do_sample,
repetition_penalty=repetition_penalty,
max_new_tokens=512,
)
state['past_key_values'] = outputs.past_key_values
state['past_ids'] = outputs.sequences[:, :-1]
response = self.processor.decode(outputs.sequences[0, inputs.input_ids.size(1):], skip_special_tokens=True)
return response, state
model_path = 'chenjoya/LiveCC-7B-Instruct'
# download a test video at: https://github.com/showlab/livecc/blob/main/demo/sources/howto_fix_laptop_mute_1080p.mp4
video_path = "demo/sources/howto_fix_laptop_mute_1080p.mp4"
infer = LiveCCDemoInfer(model_path=model_path)
state = {'video_path': video_path}
# first round
query1 = 'What is the video?'
response1, state = infer.video_qa(message=query1, state=state)
print(f'Q1: {query1}\nA1: {response1}')
# second round
query2 = 'How do you know that?'
response2, state = infer.video_qa(message=query2, state=state)
print(f'Q2: {query2}\nA2: {response2}')
```
## Performance
## Limitations
- This model is finetuned on LiveCC-7B-Base, which is starting from Qwen2-VL-7B-Base, so it may have limitations mentioned in https://huggingface.co/Qwen/Qwen2-VL-7B.
- When performing real-time video commentary, it may appear collapse --- e.g., repeat pattern. If you encounter this situation, try to adjust repetition_penalty, streaming_eos_base_threshold, and streaming_eos_threshold_step.
- This model only has a context window of 32768. Using more visual tokens per frame (e.g. 768 * 28 * 28) will have better performance, but will shorten the working duration.
These limitations serve as ongoing directions for model optimization and improvement, and we are committed to continually enhancing the model's performance and scope of application.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@article{livecc,
author = {Joya Chen and Ziyun Zeng and Yiqi Lin and Wei Li and Zejun Ma and Mike Zheng Shou},
title = {LiveCC: Learning Video LLM with Streaming Speech Transcription at Scale},
journal = {arXiv preprint arXiv:2504.16030}
year = {2025},
}
``` |
Mungert/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-GGUF | Mungert | 2025-06-15T19:37:23Z | 405 | 2 | transformers | [
"transformers",
"gguf",
"en",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-24T03:40:42Z | ---
library_name: transformers
language:
- en
license: cc-by-nc-4.0
---
# <span style="color: #7FFF7F;">Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Model Information
We introduce **Nemotron-UltraLong-8B**, a series of ultra-long context language models designed to process extensive sequences of text (up to 1M, 2M, and 4M tokens) while maintaining competitive performance on standard benchmarks. Built on the Llama-3.1, UltraLong-8B leverages a systematic training recipe that combines efficient continued pretraining with instruction tuning to enhance long-context understanding and instruction-following capabilities. This approach enables our models to efficiently scale their context windows without sacrificing general performance.
## The UltraLong Models
- [nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct](https://huggingface.co/nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct)
- [nvidia/Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct](https://huggingface.co/nvidia/Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct)
- [nvidia/Llama-3.1-Nemotron-8B-UltraLong-4M-Instruct](https://huggingface.co/nvidia/Llama-3.1-Nemotron-8B-UltraLong-4M-Instruct)
## Uses
Starting with `transformers >= 4.43.0` onward, you can run conversational inference using the Transformers `pipeline` abstraction or by leveraging the Auto classes with the `generate()` function.
Make sure to update your transformers installation via `pip install --upgrade transformers`.
```python
import transformers
import torch
model_id = "nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct"
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
model_kwargs={"torch_dtype": torch.bfloat16},
device_map="auto",
)
messages = [
{"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
{"role": "user", "content": "Who are you?"},
]
outputs = pipeline(
messages,
max_new_tokens=256,
)
print(outputs[0]["generated_text"][-1])
```
## Model Card
* Base model: [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct)
* Continued Pretraining: The training data consists of 1B tokens sourced from a pretraining corpus using per-domain upsampling based on sample length. The model was trained for 125 iterations with a sequence length of 1M and a global batch size of 8.
* Supervised fine-tuning (SFT): 1B tokens on open-source instruction datasets across general, mathematics, and code domains. We subsample the data from the ‘general_sft_stage2’ from [AceMath-Instruct](https://huggingface.co/datasets/nvidia/AceMath-Instruct-Training-Data).
* Maximum context window: 1M tokens
## Evaluation Results
We evaluate Nemotron-UltraLong-8B on a diverse set of benchmarks, including long-context tasks (e.g., RULER, LV-Eval, and InfiniteBench) and standard tasks (e.g., MMLU, MATH, GSM-8K, and HumanEval). UltraLong-8B achieves superior performance on ultra-long context tasks while maintaining competitive results on standard benchmarks.
### Needle in a Haystack
<img width="80%" alt="image" src="Llama-3.1-8B-UltraLong-1M-Instruct.png">
### Long context evaluation
<img width="80%" alt="image" src="long_benchmark.png">
### Standard capability evaluation
<img width="80%" alt="image" src="standard_benchmark.png">
## Correspondence to
Chejian Xu ([email protected]), Wei Ping ([email protected])
## Citation
<pre>
@article{ulralong2025,
title={From 128K to 4M: Efficient Training of Ultra-Long Context Large Language Models},
author={Xu, Chejian and Ping, Wei and Xu, Peng and Liu, Zihan and Wang, Boxin and Shoeybi, Mohammad and Catanzaro, Bryan},
journal={arXiv preprint},
year={2025}
}
</pre> |
ALYTV/Qwen2.5-Coder-7B-mlx-6Bit | ALYTV | 2025-06-15T19:37:22Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"qwen2",
"text-generation",
"code",
"qwen",
"qwen-coder",
"codeqwen",
"mlx",
"mlx-my-repo",
"conversational",
"en",
"base_model:Qwen/Qwen2.5-Coder-7B",
"base_model:quantized:Qwen/Qwen2.5-Coder-7B",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"6-bit",
"region:us"
] | text-generation | 2025-06-15T19:36:56Z | ---
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen2.5-Coder-7B/blob/main/LICENSE
language:
- en
base_model: Qwen/Qwen2.5-Coder-7B
pipeline_tag: text-generation
library_name: transformers
tags:
- code
- qwen
- qwen-coder
- codeqwen
- mlx
- mlx-my-repo
---
# ALYTV/Qwen2.5-Coder-7B-mlx-6Bit
The Model [ALYTV/Qwen2.5-Coder-7B-mlx-6Bit](https://huggingface.co/ALYTV/Qwen2.5-Coder-7B-mlx-6Bit) was converted to MLX format from [Qwen/Qwen2.5-Coder-7B](https://huggingface.co/Qwen/Qwen2.5-Coder-7B) using mlx-lm version **0.22.3**.
## Use with mlx
```bash
pip install mlx-lm
```
```python
from mlx_lm import load, generate
model, tokenizer = load("ALYTV/Qwen2.5-Coder-7B-mlx-6Bit")
prompt="hello"
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True)
```
|
Mungert/Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-GGUF | Mungert | 2025-06-15T19:37:19Z | 357 | 3 | transformers | [
"transformers",
"gguf",
"en",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-04-23T18:08:11Z | ---
library_name: transformers
language:
- en
license: cc-by-nc-4.0
---
# <span style="color: #7FFF7F;">Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# Model Information
We introduce **Nemotron-UltraLong-8B**, a series of ultra-long context language models designed to process extensive sequences of text (up to 1M, 2M, and 4M tokens) while maintaining competitive performance on standard benchmarks. Built on the Llama-3.1, UltraLong-8B leverages a systematic training recipe that combines efficient continued pretraining with instruction tuning to enhance long-context understanding and instruction-following capabilities. This approach enables our models to efficiently scale their context windows without sacrificing general performance.
## The UltraLong Models
- [nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct](https://huggingface.co/nvidia/Llama-3.1-Nemotron-8B-UltraLong-1M-Instruct)
- [nvidia/Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct](https://huggingface.co/nvidia/Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct)
- [nvidia/Llama-3.1-Nemotron-8B-UltraLong-4M-Instruct](https://huggingface.co/nvidia/Llama-3.1-Nemotron-8B-UltraLong-4M-Instruct)
## Uses
Starting with `transformers >= 4.43.0` onward, you can run conversational inference using the Transformers `pipeline` abstraction or by leveraging the Auto classes with the `generate()` function.
Make sure to update your transformers installation via `pip install --upgrade transformers`.
```python
import transformers
import torch
model_id = "nvidia/Llama-3.1-Nemotron-8B-UltraLong-2M-Instruct"
pipeline = transformers.pipeline(
"text-generation",
model=model_id,
model_kwargs={"torch_dtype": torch.bfloat16},
device_map="auto",
)
messages = [
{"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
{"role": "user", "content": "Who are you?"},
]
outputs = pipeline(
messages,
max_new_tokens=256,
)
print(outputs[0]["generated_text"][-1])
```
## Model Card
* Base model: [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct)
* Continued Pretraining: The training data consists of 1B tokens sourced from a pretraining corpus using per-domain upsampling based on sample length. The model was trained for 245 steps with a sequence length of 2M and a global batch size of 2.
* Supervised fine-tuning (SFT): 1B tokens on open-source instruction datasets across general, mathematics, and code domains. We subsample the data from the ‘general_sft_stage2’ from [AceMath-Instruct](https://huggingface.co/datasets/nvidia/AceMath-Instruct-Training-Data).
* Maximum context window: 2M tokens
## Evaluation Results
We evaluate Nemotron-UltraLong-8B on a diverse set of benchmarks, including long-context tasks (e.g., RULER, LV-Eval, and InfiniteBench) and standard tasks (e.g., MMLU, MATH, GSM-8K, and HumanEval). UltraLong-8B achieves superior performance on ultra-long context tasks while maintaining competitive results on standard benchmarks.
### Needle in a Haystack
<img width="80%" alt="image" src="Llama-3.1-8B-UltraLong-2M-Instruct.png">
### Long context evaluation
<img width="80%" alt="image" src="long_benchmark.png">
### Standard capability evaluation
<img width="80%" alt="image" src="standard_benchmark.png">
## Correspondence to
Chejian Xu ([email protected]), Wei Ping ([email protected])
## Citation
<pre>
@article{ulralong2025,
title={From 128K to 4M: Efficient Training of Ultra-Long Context Large Language Models},
author={Xu, Chejian and Ping, Wei and Xu, Peng and Liu, Zihan and Wang, Boxin and Shoeybi, Mohammad and Catanzaro, Bryan},
journal={arXiv preprint},
year={2025}
}
</pre> |
Mungert/DeepCoder-14B-Preview-GGUF | Mungert | 2025-06-15T19:37:06Z | 1,424 | 9 | transformers | [
"transformers",
"gguf",
"text-generation",
"en",
"dataset:PrimeIntellect/verifiable-coding-problems",
"dataset:likaixin/TACO-verified",
"dataset:livecodebench/code_generation_lite",
"base_model:deepseek-ai/DeepSeek-R1-Distill-Qwen-14B",
"base_model:quantized:deepseek-ai/DeepSeek-R1-Distill-Qwen-14B",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-04-11T03:58:56Z | ---
license: mit
library_name: transformers
datasets:
- PrimeIntellect/verifiable-coding-problems
- likaixin/TACO-verified
- livecodebench/code_generation_lite
language:
- en
base_model:
- deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
pipeline_tag: text-generation
---
# <span style="color: #7FFF7F;">DeepCoder-14B-Preview GGUF Models</span>
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `DeepCoder-14B-Preview-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `DeepCoder-14B-Preview-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `DeepCoder-14B-Preview-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `DeepCoder-14B-Preview-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `DeepCoder-14B-Preview-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `DeepCoder-14B-Preview-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `DeepCoder-14B-Preview-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `DeepCoder-14B-Preview-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `DeepCoder-14B-Preview-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `DeepCoder-14B-Preview-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `DeepCoder-14B-Preview-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard)
💬 **How to test**:
1. Click the **chat icon** (bottom right on any page)
2. Choose an **AI assistant type**:
- `TurboLLM` (GPT-4-mini)
- `FreeLLM` (Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Metasploit integration**
🟡 **TestLLM** – Current experimental model (llama.cpp on 6 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4-mini** for:
- **Real-time network diagnostics**
- **Automated penetration testing** (Nmap/Metasploit)
- 🔑 Get more tokens by [downloading our Quantum Network Monitor Agent](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
🔵 **HugLLM** – Open-source models (≈8B params):
- **2x more tokens** than TurboLLM
- **AI-powered log analysis**
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example AI Commands to Test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a quick Nmap vulnerability test"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
<div align="center">
<span style="font-family: default; font-size: 1.5em;">DeepCoder-14B-Preview</span>
<div>
🚀 Democratizing Reinforcement Learning for LLMs (RLLM) 🌟
</div>
</div>
<br>
<div align="center" style="line-height: 1;">
<a href="https://github.com/agentica-project/rllm" style="margin: 2px;">
<img alt="Code" src="https://img.shields.io/badge/RLLM-000000?style=for-the-badge&logo=github&logoColor=000&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://pretty-radio-b75.notion.site/DeepCoder-A-Fully-Open-Source-14B-Coder-at-O3-mini-Level-1cf81902c14680b3bee5eb349a512a51" target="_blank" style="margin: 2px;">
<img alt="Blog" src="https://img.shields.io/badge/Notion-%23000000.svg?style=for-the-badge&logo=notion&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://x.com/Agentica_" style="margin: 2px;">
<img alt="X.ai" src="https://img.shields.io/badge/Agentica-white?style=for-the-badge&logo=X&logoColor=000&color=000&labelColor=white" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://huggingface.co/agentica-org" style="margin: 2px;">
<img alt="Hugging Face" src="https://img.shields.io/badge/Agentica-fcd022?style=for-the-badge&logo=huggingface&logoColor=000&labelColor" style="display: inline-block; vertical-align: middle;"/>
</a>
<a href="https://www.together.ai" style="margin: 2px;">
<img alt="Together AI" src="https://img.shields.io/badge/-Together_AI%20-white?style=for-the-badge&logo=data%3Aimage%2Fpng%3Bbase64%2CiVBORw0KGgoAAAANSUhEUgAAAUAAAAFACAMAAAD6TlWYAAAC7lBMVEUAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAPb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8AAAAPb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8Pb%2F8AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAADIBDt6AAAA%2BnRSTlMAAiQEKgcdKQwiHBMUzrtSUEmjhmZGH96yv8n1ey7nL3y1U%2FZfCaIo1WFg1NrcsHYrA2%2Fv80J%2BMeilnpefqKw%2B64%2BQlSbYZGVnBGkCV%2BxW8XJube6WJ9kZF9bSzBALRynPQfLhIjvwyBEAXOTLp3o%2FJA9Y9%2F7%2F9FEKDhIVFo4GHkVzjGz8icrHzY39iHR1i0M8Jj14LLZUvb7DxMXGoQEFeQcgSBOHaPvm4uOdRLMMqcDTLbcII0sNuVn4TKaRd6RKIeDd37Svra6xuLpaW17lXUAlHh8WGxUPIS4JGQoFECMsBg4gFwsRJRIrCC0oAycaFC8NMDIzMRgBsVt9rwAAD25JREFUeNrs3QVzG0kWB%2FA3ikHhZeYwk3LMbF7GcBasOGw9hb3MzLyKw8zMzMx2rsokhySNY2mmR1N4xXV3a7sHuzWu%2BX2Ef3XPG%2Br3wOVyuVwul8vlcrlcLpfL5XK5dOlXOHTIvLnb27Xd%2FasBvrt9A%2B7r1bbdTTffcmuXwhzgTYwk6q%2BHr2RWlcclRYqXV2VeCV%2Bvr4mIkCJKZ83uc9NLC0fMD%2BD%2FCswfMfLtzh%2FeelsJcKJW19SG66KSTP6fLEXrwrU11Srw5Z8zbuzePcUBbFyg%2BPY7Pv%2Bs0A%2Bsid7ayiqFNEWp8iS9Ir%2F0Cl957bkRAaQLFLz15sBBfpbpJc7FJKKFFGuV4JJh6N573g6idr7vP%2F8iC9iI1NZJRDupLnlRBbaW3XjTfQHUJ3D8d68MBtsJiTNRold5uEYAdibkHgqiESMefGi9zfFVeCRihOS5LLJafV99XYxGddgwabKt8SmEyEQ%2FmRDlSoUA9gsNvKMDmhE8MC4L7OFtSYmPFmFlAmzm%2F9tfH0Oz8v6yFmxQ3SpOiY8eYTwjHew0%2BB9%2FD6B5ga4dLd%2FHQus0SnzaIrzWWgDb9P19MVqjw01dwFLpYYVYQymLgD1Kjj6J1umaHwLLqJfpy0%2FHIryqgg2mvetDKxXMnQMWEa9LxEpSqxZguS%2B%2BfA%2Bt9cZBi7ZxeqVMX376FqEnAtbyv7ISrTfspB%2FM82bq3r70BNMSYKV%2Bo4rQDiPzc8Csy1Fih%2BhVsE7o0cfQHnn%2FygJz6uNEJtaTSfy8ChYpnelDuxQ8HAIT1LOS8fwoCSq1FiVYcs%2FdaJ%2FgNhMJqrWKqfwoCSYtSTA08260U%2FBh47v4LDU%2F%2FgnmPOJDexX86ycwpp6yf80neB7M8o96DO2Wl2%2Bw%2FlLrh%2FlKYroW31qE9ht5EgzwRs3nR00wmgBTVq1EFtp2Ad0imdbkR0kwLQImTP8S2eg9B3QSKwkbHhPPxSUzAsjGe3P1luLrMmGklQpGjfIhKwU6C8llibBJUCaS4UKy6klkp0cX0CE9zcr8KAlei4Ahy36PLHXuBJqpYcJSmQBG3LIJWerQETS7qhCWlHowoMvfka2Va0Gjaus3MGUTp4NuWY8ja3%2FuB9q0IqydBt1eeQxZ%2B9MfQRNvnLAWT%2BiuIEuRvT9MBg3UlkQmbMmkUgB9cjsge8EbQIMLCmFPuQy6DPoGeVi9HqgED5EJazL5VAQ9Nm5CHjq0B6oKhZCUX4LrNyAfSycDhVBJZMKeTK4IoN26IPJRsAQoEhLhQ7kAmoV%2Bjbwspt0LniF8yKRMBa1%2B%2BSvkZVFfaFIkSngpvwha%2FQL56QNNqiX8%2FBs0mnMX8vPtBGiCWEf4iYmgzey7kZ8Rw6EJXonwo9SANn9GnuZCE84RnlqBJm3aIk8vFUKjxBjhKbMFaDHQhzy9%2BAI06pJEeJIS%2FGuwBn1M1WD%2BdXjNauSrdwk0Qq0kfHlUoFs7Evnq9TI0orqK8BVN1%2FIcvAn56vAKNCKhEDruz8NjkbdXOV4CKZJA1W8M8vbjT9CwMOGtDKjmjEbefpgCDRLqCB33p7kvipC3kc83UkOihLdohF5DfMjbiBf43UZTSPQq8vobyNsbudCgyzLhTT4PNK8hpmoZPkv4awU0y5G%2F1%2Fj90WG%2BDK9ATNX7mDDh71OgWYn83RHi9yRMkQY0I5G%2FOydDA4RPCX9RoMlD%2Fu6a0mCAMcJfHGh8yN%2BwqdAAMZPwJwFNB%2BRv5TRoQIs0wp%2FiiAB7TG%2B2Abor0L0GmiO5VdicuHsfaE7UfRIxJ80Rz8Kdnfss7L6NoShz8vvAWsLfOUe8kZ7o5DfSm1Pgm8gnTv4msqoIzXC%2FyrUZjWa434XdPxOoRZjiHjTD%2FTcGNm9Cg9y%2Fs9z%2FAymi1e4fqqZ4VPcfaQZnlQYGkacXP3H6X%2FrT2qIZ7jkR%2BAvy9L5jTyq5Z%2BUolBpHnNYc5PDTmubrsHtemOeJ9aJmcWI9tAV5%2BQ29Z4Kc%2Bj0TYHOQVwl5pVl07YD1h9EMt28MHOHUueihZtK5CArvRB4OTWkuvbNgYjGyF5wEGlQ4oXsbrF%2BK7O2fDBoIPPoHegQndLAc14w6WELot8jaX5pVD1Xo8iSy1WM8nzbcFMZbcf%2BLcR%2Fp7qBZayf0kYZly5GlzpOd3Mmcfy%2F9rl1AhwjTXvoXwaATDKc55Dp6mgP%2FeSLvZ4E%2B55wwTwSmr0Y2Djp6og3%2FmUrDhqbuTKWLYMqQ42i%2FkcNTdqpXeQ2Y4z82AO2Wl8txrpz5AkLRr38Q7TUiOydlJxueBfNCYzugnYKvOn62JkXpA3YmGPy8xPnTXanzhYP27d8PSvjPFzafH0Wov12VJC87ZSdcS2dVsEy%2FE8fRDgtznTFj3Tz%2FrT3QesOGO2bKv3mrVr%2BH1nrjjqFgiUilTGRr8%2FNEwHLTZ%2FisLR9vzgGLiOckYiWpVQuwQcmonmidZ3JDYBn1chohslXL79pVFWzh%2F2L5JrRG8fahYKlIWCHWUMoiYJtl%2F3wygOYFunabDBYTWmtdhJTlVy%2BAjfxPPP4YmpW3dTzYID0jTo%2BQEl88Ix1sFlqytAOacfe%2Bk1lgD29LxXiEMiFKZUIF%2By3L%2F6YYjSpu134w2EaouEKPsNH4rlwWgI0JEzcE0Qjfl19NAVsJFR6JGCF5LovAzrId2%2B8LoD6BBT8OGQy2E2rCUaJXebhGALZC9z%2FwUhC18%2F0wc1UWsBFJ1klEOymWvKgCe%2F7CW999xxdAusCI0R99PMgP7IiJczFJY3qtEiLw8tOckw88uKs40FR4xXuWzvzjVD%2BwJnqTlVUKaYpS5Ul6ReCsdOeOmVveKgq%2Bh%2F%2FvveCiu7Zvmz2rFDhRq2tqw7GoJJP%2FJ0vRWFmyplqF1NBv0KmTJz7fumX1d889%2B8yTzzz73Ldfbtm6bdS48RNygDcx3Xu1NqPMUxdLS7uWlhar85RlJK9600VIOf6c0mWDpj391NNtBg0uyfFDSlEF8T%2Ft3eFyqjwTwPGNiKq9eq%2BtqiCeoxZVEcRW4mK%2Bvc%2F5%2Bk7bBSDZOJPfFfwHWkEMG%2B%2BfXChwHMdxHMdxHMdxHMdxHMdxHIeV4yiR%2FyOUS6tHfBxP88Vse74N%2F7mdt7PF%2FHT8EFakbYg0XupvMZ%2Fddt%2F%2Ber27zebFX%2BXSfpQfD%2BMLsX7iMp4fc460%2BfgiqbSD1jSCGH1WXAV1v32OhOm0O1Yh9aUR0sNUYnVyekjBEH9eL%2B2mIY2gilmGdWXvhTKQNnpvkDYrBJgjNluJTchtIDSnBY3TNgLMUEGvbL4Qvhco3WkPbOS%2FNAEGjMay1bsEMjyCJsewXVo5HoFuH5P2b7OsJh9a0har1mn3tmkElXTzPlU%2FUd2nDfnTKH53b%2FTN%2FI7TZp2l7X3QZNPlO6X9jb1pJwUa5J8SuyQ%2Fc2vTFjl0zu%2F8vfrH2O8obdx52jaFjmmZ7HAdQQeOVw1pwxF0StNskd0GWtvsUIfsBB3SNt3m%2FgUtva1402jEfCXm%2BUBLjWkHBZ2gJ3zxHcG51JhWdnQENc%2BYk3O2vz%2F6CEJrBqYcyi9o6E172hJaMjJn876BRjYG0k7QiqFJr7tRo7SdgbSsgBaMzRoe%2BlCbfzWTlkILxqZdj%2FPaaWM0Y%2BtBUwbnrT8%2BoaZPY2kLBc2Ynfi%2FgVo2BtNO0JDRPSf6PtTgm0y7pNCI2KNJewWVqZnZNAH1md93J4HKEsNpb1Abw85P%2FQ%2Bo6GNoOs2H%2BgZo2gQqWqBpA6iNY%2Fe7EVRyXNm%2FMR%2FP%2FotjBRWokCFtK6AOrh1AA6ggkBxpG6hFnImzzLUFKNv2uOec5Q9Qw3kO7N%2BgmT7LjB81asuU1hNQXSyRhyyAULClxVDdHh%2FI4YEzIMzY0vZQWZQhlyyFX6V8aasIqnoinwP86oB8nlBRfkM%2Btxx%2BIaZWpNGf03zkCH4xYk0r7PiuTljALz6R0wQqya%2FI6ZrTHy78acS%2FCSd5hB8dmdNGdlyDCQfiGmz7dVhtkddWWZvWU0D72CGv3Qf84O%2BFP40Wl8irLOAHBXtaDLQDoq0fgnPk9gTaHrnt4Qcz5Bba8T2OcBPwLUGnWXAnmGbILfP5Lm%2BELLX3WSp9v3q0IC0GytcDuT1O8K2TBWlLq58kEJfhOfJbACVEfhN7z20IlDPy2xM3WIymQBkiv57i%2ByZM6ANlh%2FymAr6hpshvB5QVoqW3q%2BKK%2FO5AkchvmMM38iHyk0ApkV%2Ffg294feRXugPoDiCr0n0GtiPdVbid%2BwvfB4op8svcN5F2%2Bu67cDvTV34aM0F%2B4Ss%2FDzzYcW4JSwse%2Byav%2FETa4t9ERhakBS%2F9q5wFaRH%2F6kDaNbf3d2EPXuAyvLd30UQItCdyO9i7bOf5EquzYnvTgpdeH8iflvlAUz3kZf8KVcs%2FBJ%2F2rl1cQxWFvUvhR8xpBVThDfnvAu28SR16UMkEOS3sfdQxgGri0tp%2Fk0Lac39l6T%2FKLbd2AfLVg4rW9t7rPy24BtOiFXJZRda%2BTL%2F6A1Wp0N7BBHu2tFBBZUGJPGRs7QPfMrB9cBExnIV7pM1ZQA0nrvFA9qYlUEc%2B5R9QZddYrymdxn%2Bey5O9g%2BUSqEf0rB3SJ7YMaT0BNRUMEywLa9NkDHWpdzRtYO9413cFtaUXw6NyL76VA4abj%2BL%2BMjys%2BcvaEdePJTQhxmhSKGqkhWjSWEAj0cXagfWpybRdBA0lpbktExJrN5oo36ApNUFTJqpm2gJNGShozOuhGT3P2rSzBy1EfSMbF%2FVTqC01lBZBK%2FHK2q2zisxA2iqGlhKpf%2FO2pGHaXXuafOPfGZKMLJeMO0MSaXNoTz1LvRtYPhXftqlE2lpBB9SayOQ6fgDqqTXtk07jzKSPH00dpL60tbJ9h%2Bb2%2BzODWt7tSKM34tZhlUBrSaYn7Q06Ffc1bKXfj6EDhQ1ptOhcP5OI7EXQibTXedo5gs55gxK7VE68ztImstu0gQcaqGSH%2BOjqHF8S1WXapcO03ZsCPaLxA7tRhhF0Kg1L7MZjHIE24os%2B05X%2B%2FL6ErWm7pQCd0ndJdxKN93cfNPDf763T5CwFzVTcK%2BnOXxrLXqE0pRXbtmmxAv3EaUp3%2Ftg4PQlL0x7TRIAZeXIusYnyfMo1p50apyU5mCOCcIV1rcJA2J9mivqzvpZYXXldR8pQWlQ77Y8CBnk8GFYLlcNBnJtNmwwlVlH%2Bl%2BYBG69Yn7Py98Ksty48lrQemXY2kEZRfvAMr5l84P97yOwaPgNfWZq2NpZG86JgPhlP%2B9ldlo9S3rP%2BdDyZB5FnRdqygzTHcRzHcRzHcRzHcRzHcZz%2FAbyvLkVmYcs9AAAAAElFTkSuQmCC&link=https%3A%2F%2Fwww.together.ai" style="display: inline-block; vertical-align: middle;"/>
</a>
</div>
</div>
</div>
## DeepCoder Overview
DeepCoder-14B-Preview is a code reasoning LLM fine-tuned from DeepSeek-R1-Distilled-Qwen-14B using distributed reinforcement learning (RL) to scale up to long context lengths. The model achieves 60.6% Pass@1 accuracy on LiveCodeBench v5 (8/1/24-2/1/25), representing a 8% improvement over the base model (53%) and achieving similar performance to OpenAI's o3-mini with just 14B parameters.
<div style="margin: 0 auto;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/654037be97949fd2304aab7f/r3-vzkItOCrMf1qldW0Mj.png" style="width: 100%;" />
</div>
## Data
Our training dataset consists of approximately 24K unique problem-tests pairs compiled from:
- Taco-Verified
- PrimeIntellect SYNTHETIC-1
- LiveCodeBench v5 (5/1/23-7/31/24)
## Training Recipe
Our training recipe relies on an improved version of GRPO (GRPO+) and iterative context lengthening, introduced in DeepScaleR.
### GRPO+
We enhance the original GRPO algorithm with insights from DAPO to enable more stable training:
- **Offline Difficulty Filtering:** DAPO employs online dynamic sampling, discarding both entirely correct and entirely incorrect samples on the fly. While this helps maintain a more stable effective batch size, it introduces significant runtime overhead due to rejection sampling. Instead, we perform offline difficulty filtering on a subset of coding problems to ensure the training dataset remains within a suitable difficulty range.
- **No Entropy Loss:** We observed that including an entropy loss term often led to instability, with entropy growing exponentially and ultimately collapsing training. To mitigate this, we eliminate the entropy loss entirely.
- **No KL Loss:** Eliminating KL loss prevents the LLM from staying within trust region of the original SFT model. This removal also obviates the need to compute log probabilities for the reference policy, thereby accelerating training.
- **Overlong Filtering** **(from DAPO):** To preserve long-context reasoning, we mask the loss for truncated sequences. This technique enables DeepCoder to generalize to 64K-context inference despite being trained with a 32K context.
- **Clip High (from DAPO):** By increasing the upper bound in GRPO/PPO’s surrogate loss, we encourage more exploration and more stable entropy.
### Iterative Context Lengthening
Our original `Deepscaler-1.5B-Preview` scaled long context training from 8K→16K→24K, achieving 33→38→43% on AIME respectively. Similarly, `Deepcoder-14B-Preview` is trained on 16K→32K, achieving 54→58% on LiveCodeBench (v5). `DeepCoder-14B-Preview` successfully generalizes to longer contexts when evaluated at 64K context, reaching 60.6%.
DeepCoder generalizes better to long contexts than the base distilled model, due to DAPO's overlong filtering. However, it's longer responses are often truncated when the max length is capped at 16K, which can lower its scores.
| **Model** | **16K** | **32K** | **64K** |
| --- | --- | --- | --- |
| **DeepCoder-14B-Preview** | 45.6 | 57.9 | 60.6 |
| **DeepSeek-R1-Distill-Qwen-14B** | 50.2 | 53.0 | 53.0 |
A more detailed description of the training recipe can be found in our [blog post](https://pretty-radio-b75.notion.site/DeepCoder-A-Fully-Open-Source-14B-Coder-at-O3-mini-Level-1cf81902c14680b3bee5eb349a512a51).
## Evaluation
We evaluate `Deepcoder-14B-Preview` on various coding benchmarks, including LiveCodeBench (LCBv5), Codeforces, and HumanEval+.
| **Model** | LCB (v5)(8/1/24-2/1/25) | Codeforces Rating | Codeforces Percentile | HumanEval+ |
| --- | --- | --- | --- | --- |
| **DeepCoder-14B-Preview (ours)** | ***60.6*** | ***1936*** | ***95.3*** | ***92.6*** |
| **DeepSeek-R1-Distill-Qwen-14B** | 53.0 | 1791 | 92.7 | 92.0 |
| **O1-2024-12-17 (Low)** | 59.5 | **1991** | **96.1** | 90.8 |
| **O3-Mini-2025-1-31 (Low)** | **60.9** | 1918 | 94.9 | 92.6 |
| **O1-Preview** | 42.7 | 1658 | 88.5 | 89 |
| **Deepseek-R1** | 62.8 | 1948 | 95.4 | 92.6 |
| **Llama-4-Behemoth** | 49.4 | - | - | - |
## Serving DeepCoder
Our model can be served using popular high-performance inference systems:
- vLLM
- Hugging Face Text Generation Inference (TGI)
- SGLang
- TensorRT-LLM
All these systems support the OpenAI Chat Completions API format.
### Usage Recommendations
Our usage recommendations are similar to those of R1 and R1 Distill series:
1. Avoid adding a system prompt; all instructions should be contained within the user prompt.
2. `temperature = 0.6`
3. `top_p = 0.95`
4. This model performs best with `max_tokens` set to at least `64000`
## License
This project is released under the MIT License, reflecting our commitment to open and accessible AI development.
We believe in democratizing AI technology by making our work freely available for anyone to use, modify, and build upon.
This permissive license ensures that researchers, developers, and enthusiasts worldwide can leverage and extend our work without restrictions, fostering innovation and collaboration in the AI community.
## Acknowledgement
- Our training experiments are powered by our heavily modified fork of [Verl](https://github.com/agentica-project/verl), an open-source post-training library.
- Our model is trained on top of [`DeepSeek-R1-Distill-Qwen-14B`](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B).
- Our work is done as part of [Berkeley Sky Computing Lab](https://skycomputing.berkeley.edu/) and [Berkeley AI Research](https://bair.berkeley.edu/).
## Citation
```bibtex
@misc{deepcoder2025,
title={DeepCoder: A Fully Open-Source 14B Coder at O3-mini Level},
author={Michael Luo, Sijun Tan, Roy Huang, Ameen Patel, Alpay Ariyak, Qingyang Wu, Xiaoxiang Shi, Rachel Xin, Colin Cai, Maurice Weber, Ce Zhang, Li Erran Li, Raluca Ada Popa, Ion Stoica},
howpublished={\url{https://pretty-radio-b75.notion.site/DeepCoder-A-Fully-Open-Source-14B-Coder-at-O3-mini-Level-1cf81902c14680b3bee5eb349a512a51}},
note={Notion Blog},
year={2025}
}
``` |
Mungert/all-MiniLM-L6-v2-GGUF | Mungert | 2025-06-15T19:36:56Z | 2,709 | 4 | sentence-transformers | [
"sentence-transformers",
"gguf",
"feature-extraction",
"sentence-similarity",
"transformers",
"en",
"dataset:s2orc",
"dataset:flax-sentence-embeddings/stackexchange_xml",
"dataset:ms_marco",
"dataset:gooaq",
"dataset:yahoo_answers_topics",
"dataset:code_search_net",
"dataset:search_qa",
"dataset:eli5",
"dataset:snli",
"dataset:multi_nli",
"dataset:wikihow",
"dataset:natural_questions",
"dataset:trivia_qa",
"dataset:embedding-data/sentence-compression",
"dataset:embedding-data/flickr30k-captions",
"dataset:embedding-data/altlex",
"dataset:embedding-data/simple-wiki",
"dataset:embedding-data/QQP",
"dataset:embedding-data/SPECTER",
"dataset:embedding-data/PAQ_pairs",
"dataset:embedding-data/WikiAnswers",
"arxiv:1904.06472",
"arxiv:2102.07033",
"arxiv:2104.08727",
"arxiv:1704.05179",
"arxiv:1810.09305",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us",
"imatrix"
] | sentence-similarity | 2025-03-23T03:05:06Z | ---
language: en
license: apache-2.0
library_name: sentence-transformers
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
datasets:
- s2orc
- flax-sentence-embeddings/stackexchange_xml
- ms_marco
- gooaq
- yahoo_answers_topics
- code_search_net
- search_qa
- eli5
- snli
- multi_nli
- wikihow
- natural_questions
- trivia_qa
- embedding-data/sentence-compression
- embedding-data/flickr30k-captions
- embedding-data/altlex
- embedding-data/simple-wiki
- embedding-data/QQP
- embedding-data/SPECTER
- embedding-data/PAQ_pairs
- embedding-data/WikiAnswers
pipeline_tag: sentence-similarity
---
# <span style="color: #7FFF7F;">all-MiniLM-L6-v2 GGUF Models</span>
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device’s specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn’t available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `all-MiniLM-L6-v2-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `all-MiniLM-L6-v2-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `all-MiniLM-L6-v2-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `all-MiniLM-L6-v2-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `all-MiniLM-L6-v2-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `all-MiniLM-L6-v2-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `all-MiniLM-L6-v2-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `all-MiniLM-L6-v2-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `all-MiniLM-L6-v2-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `all-MiniLM-L6-v2-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `all-MiniLM-L6-v2-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Please click like ❤ . Also I’d really appreciate it if you could test my Network Monitor Assistant at 👉 [Network Monitor Assitant](https://readyforquantum.com).
💬 Click the **chat icon** (bottom right of the main and dashboard pages) . Choose a LLM; toggle between the LLM Types TurboLLM -> FreeLLM -> TestLLM.
### What I'm Testing
I'm experimenting with **function calling** against my network monitoring service. Using small open source models. I am into the question "How small can it go and still function".
🟡 **TestLLM** – Runs the current testing model using llama.cpp on 6 threads of a Cpu VM (Should take about 15s to load. Inference speed is quite slow and it only processes one user prompt at a time—still working on scaling!). If you're curious, I'd be happy to share how it works! .
### The other Available AI Assistants
🟢 **TurboLLM** – Uses **gpt-4o-mini** Fast! . Note: tokens are limited since OpenAI models are pricey, but you can [Login](https://readyforquantum.com) or [Download](https://readyforquantum.com/download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) the Quantum Network Monitor agent to get more tokens, Alternatively use the TestLLM .
🔵 **HugLLM** – Runs **open-source Hugging Face models** Fast, Runs small models (≈8B) hence lower quality, Get 2x more tokens (subject to Hugging Face API availability)
### Final word
I fund the servers to create the models files, run the Quantum Network Monitor Service and Pay for Inference from Novita and OpenAI all from my own pocket. All of the code for creating the models and the work I have done with Quantum Network Monitor is [open source](https://github.com/Mungert69). Feel free to use what you find useful. Please support my work and consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) .
This will help me pay for the services and increase the token limits for everyone.
Thank you :)
# all-MiniLM-L6-v2
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 384 dimensional dense vector space and can be used for tasks like clustering or semantic search.
## Usage (Sentence-Transformers)
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
```python
from transformers import AutoTokenizer, AutoModel
import torch
import torch.nn.functional as F
#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0] #First element of model_output contains all token embeddings
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('sentence-transformers/all-MiniLM-L6-v2')
model = AutoModel.from_pretrained('sentence-transformers/all-MiniLM-L6-v2')
# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
# Compute token embeddings
with torch.no_grad():
model_output = model(**encoded_input)
# Perform pooling
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
# Normalize embeddings
sentence_embeddings = F.normalize(sentence_embeddings, p=2, dim=1)
print("Sentence embeddings:")
print(sentence_embeddings)
```
------
## Background
The project aims to train sentence embedding models on very large sentence level datasets using a self-supervised
contrastive learning objective. We used the pretrained [`nreimers/MiniLM-L6-H384-uncased`](https://huggingface.co/nreimers/MiniLM-L6-H384-uncased) model and fine-tuned in on a
1B sentence pairs dataset. We use a contrastive learning objective: given a sentence from the pair, the model should predict which out of a set of randomly sampled other sentences, was actually paired with it in our dataset.
We developed this model during the
[Community week using JAX/Flax for NLP & CV](https://discuss.huggingface.co/t/open-to-the-community-community-week-using-jax-flax-for-nlp-cv/7104),
organized by Hugging Face. We developed this model as part of the project:
[Train the Best Sentence Embedding Model Ever with 1B Training Pairs](https://discuss.huggingface.co/t/train-the-best-sentence-embedding-model-ever-with-1b-training-pairs/7354). We benefited from efficient hardware infrastructure to run the project: 7 TPUs v3-8, as well as intervention from Googles Flax, JAX, and Cloud team member about efficient deep learning frameworks.
## Intended uses
Our model is intended to be used as a sentence and short paragraph encoder. Given an input text, it outputs a vector which captures
the semantic information. The sentence vector may be used for information retrieval, clustering or sentence similarity tasks.
By default, input text longer than 256 word pieces is truncated.
## Training procedure
### Pre-training
We use the pretrained [`nreimers/MiniLM-L6-H384-uncased`](https://huggingface.co/nreimers/MiniLM-L6-H384-uncased) model. Please refer to the model card for more detailed information about the pre-training procedure.
### Fine-tuning
We fine-tune the model using a contrastive objective. Formally, we compute the cosine similarity from each possible sentence pairs from the batch.
We then apply the cross entropy loss by comparing with true pairs.
#### Hyper parameters
We trained our model on a TPU v3-8. We train the model during 100k steps using a batch size of 1024 (128 per TPU core).
We use a learning rate warm up of 500. The sequence length was limited to 128 tokens. We used the AdamW optimizer with
a 2e-5 learning rate. The full training script is accessible in this current repository: `train_script.py`.
#### Training data
We use the concatenation from multiple datasets to fine-tune our model. The total number of sentence pairs is above 1 billion sentences.
We sampled each dataset given a weighted probability which configuration is detailed in the `data_config.json` file.
| Dataset | Paper | Number of training tuples |
|--------------------------------------------------------|:----------------------------------------:|:--------------------------:|
| [Reddit comments (2015-2018)](https://github.com/PolyAI-LDN/conversational-datasets/tree/master/reddit) | [paper](https://arxiv.org/abs/1904.06472) | 726,484,430 |
| [S2ORC](https://github.com/allenai/s2orc) Citation pairs (Abstracts) | [paper](https://aclanthology.org/2020.acl-main.447/) | 116,288,806 |
| [WikiAnswers](https://github.com/afader/oqa#wikianswers-corpus) Duplicate question pairs | [paper](https://doi.org/10.1145/2623330.2623677) | 77,427,422 |
| [PAQ](https://github.com/facebookresearch/PAQ) (Question, Answer) pairs | [paper](https://arxiv.org/abs/2102.07033) | 64,371,441 |
| [S2ORC](https://github.com/allenai/s2orc) Citation pairs (Titles) | [paper](https://aclanthology.org/2020.acl-main.447/) | 52,603,982 |
| [S2ORC](https://github.com/allenai/s2orc) (Title, Abstract) | [paper](https://aclanthology.org/2020.acl-main.447/) | 41,769,185 |
| [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) (Title, Body) pairs | - | 25,316,456 |
| [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) (Title+Body, Answer) pairs | - | 21,396,559 |
| [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) (Title, Answer) pairs | - | 21,396,559 |
| [MS MARCO](https://microsoft.github.io/msmarco/) triplets | [paper](https://doi.org/10.1145/3404835.3462804) | 9,144,553 |
| [GOOAQ: Open Question Answering with Diverse Answer Types](https://github.com/allenai/gooaq) | [paper](https://arxiv.org/pdf/2104.08727.pdf) | 3,012,496 |
| [Yahoo Answers](https://www.kaggle.com/soumikrakshit/yahoo-answers-dataset) (Title, Answer) | [paper](https://proceedings.neurips.cc/paper/2015/hash/250cf8b51c773f3f8dc8b4be867a9a02-Abstract.html) | 1,198,260 |
| [Code Search](https://huggingface.co/datasets/code_search_net) | - | 1,151,414 |
| [COCO](https://cocodataset.org/#home) Image captions | [paper](https://link.springer.com/chapter/10.1007%2F978-3-319-10602-1_48) | 828,395|
| [SPECTER](https://github.com/allenai/specter) citation triplets | [paper](https://doi.org/10.18653/v1/2020.acl-main.207) | 684,100 |
| [Yahoo Answers](https://www.kaggle.com/soumikrakshit/yahoo-answers-dataset) (Question, Answer) | [paper](https://proceedings.neurips.cc/paper/2015/hash/250cf8b51c773f3f8dc8b4be867a9a02-Abstract.html) | 681,164 |
| [Yahoo Answers](https://www.kaggle.com/soumikrakshit/yahoo-answers-dataset) (Title, Question) | [paper](https://proceedings.neurips.cc/paper/2015/hash/250cf8b51c773f3f8dc8b4be867a9a02-Abstract.html) | 659,896 |
| [SearchQA](https://huggingface.co/datasets/search_qa) | [paper](https://arxiv.org/abs/1704.05179) | 582,261 |
| [Eli5](https://huggingface.co/datasets/eli5) | [paper](https://doi.org/10.18653/v1/p19-1346) | 325,475 |
| [Flickr 30k](https://shannon.cs.illinois.edu/DenotationGraph/) | [paper](https://transacl.org/ojs/index.php/tacl/article/view/229/33) | 317,695 |
| [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) Duplicate questions (titles) | | 304,525 |
| AllNLI ([SNLI](https://nlp.stanford.edu/projects/snli/) and [MultiNLI](https://cims.nyu.edu/~sbowman/multinli/) | [paper SNLI](https://doi.org/10.18653/v1/d15-1075), [paper MultiNLI](https://doi.org/10.18653/v1/n18-1101) | 277,230 |
| [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) Duplicate questions (bodies) | | 250,519 |
| [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) Duplicate questions (titles+bodies) | | 250,460 |
| [Sentence Compression](https://github.com/google-research-datasets/sentence-compression) | [paper](https://www.aclweb.org/anthology/D13-1155/) | 180,000 |
| [Wikihow](https://github.com/pvl/wikihow_pairs_dataset) | [paper](https://arxiv.org/abs/1810.09305) | 128,542 |
| [Altlex](https://github.com/chridey/altlex/) | [paper](https://aclanthology.org/P16-1135.pdf) | 112,696 |
| [Quora Question Triplets](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | - | 103,663 |
| [Simple Wikipedia](https://cs.pomona.edu/~dkauchak/simplification/) | [paper](https://www.aclweb.org/anthology/P11-2117/) | 102,225 |
| [Natural Questions (NQ)](https://ai.google.com/research/NaturalQuestions) | [paper](https://transacl.org/ojs/index.php/tacl/article/view/1455) | 100,231 |
| [SQuAD2.0](https://rajpurkar.github.io/SQuAD-explorer/) | [paper](https://aclanthology.org/P18-2124.pdf) | 87,599 |
| [TriviaQA](https://huggingface.co/datasets/trivia_qa) | - | 73,346 |
| **Total** | | **1,170,060,424** | |
Mungert/MiniCPM4-8B-GGUF | Mungert | 2025-06-15T19:36:48Z | 906 | 2 | transformers | [
"transformers",
"gguf",
"text-generation",
"zh",
"en",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-06-13T20:02:10Z | ---
license: apache-2.0
language:
- zh
- en
pipeline_tag: text-generation
library_name: transformers
---
# <span style="color: #7FFF7F;">MiniCPM4-8B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`7f4fbe51`](https://github.com/ggerganov/llama.cpp/commit/7f4fbe5183b23b6b2e25fd1ccc5d1fa8bb010cb7).
---
## <span style="color: #7FFF7F;">Quantization Beyond the IMatrix</span>
I've been experimenting with a new quantization approach that selectively elevates the precision of key layers beyond what the default IMatrix configuration provides.
In my testing, standard IMatrix quantization underperforms at lower bit depths, especially with Mixture of Experts (MoE) models. To address this, I'm using the `--tensor-type` option in `llama.cpp` to manually "bump" important layers to higher precision. You can see the implementation here:
👉 [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
While this does increase model file size, it significantly improves precision for a given quantization level.
### **I'd love your feedback—have you tried this? How does it perform for you?**
---
<a href="https://readyforquantum.com/huggingface_gguf_selection_guide.html" style="color: #7FFF7F;">
Click here to get info on choosing the right GGUF model format
</a>
---
<!--Begin Original Model Card-->
<div align="center">
<img src="https://github.com/OpenBMB/MiniCPM/blob/main/assets/minicpm_logo.png?raw=true" width="500em" ></img>
</div>
<p align="center">
<a href="https://github.com/OpenBMB/MiniCPM/" target="_blank">GitHub Repo</a> |
<a href="https://github.com/OpenBMB/MiniCPM/tree/main/report/MiniCPM_4_Technical_Report.pdf" target="_blank">Technical Report</a>
</p>
<p align="center">
👋 Join us on <a href="https://discord.gg/3cGQn9b3YM" target="_blank">Discord</a> and <a href="https://github.com/OpenBMB/MiniCPM/blob/main/assets/wechat.jpg" target="_blank">WeChat</a>
</p>
## What's New
- [2025.06.06] **MiniCPM4** series are released! This model achieves ultimate efficiency improvements while maintaining optimal performance at the same scale! It can achieve over 5x generation acceleration on typical end-side chips! You can find technical report [here](https://github.com/OpenBMB/MiniCPM/tree/main/report/MiniCPM_4_Technical_Report.pdf).🔥🔥🔥
## MiniCPM4 Series
MiniCPM4 series are highly efficient large language models (LLMs) designed explicitly for end-side devices, which achieves this efficiency through systematic innovation in four key dimensions: model architecture, training data, training algorithms, and inference systems.
- [MiniCPM4-8B](https://huggingface.co/openbmb/MiniCPM4-8B): The flagship of MiniCPM4, with 8B parameters, trained on 8T tokens. (**<-- you are here**)
- [MiniCPM4-0.5B](https://huggingface.co/openbmb/MiniCPM4-0.5B): The small version of MiniCPM4, with 0.5B parameters, trained on 1T tokens.
- [MiniCPM4-8B-Eagle-FRSpec](https://huggingface.co/openbmb/MiniCPM4-8B-Eagle-FRSpec): Eagle head for FRSpec, accelerating speculative inference for MiniCPM4-8B.
- [MiniCPM4-8B-Eagle-FRSpec-QAT-cpmcu](https://huggingface.co/openbmb/MiniCPM4-8B-Eagle-FRSpec-QAT-cpmcu): Eagle head trained with QAT for FRSpec, efficiently integrate speculation and quantization to achieve ultra acceleration for MiniCPM4-8B.
- [MiniCPM4-8B-Eagle-vLLM](https://huggingface.co/openbmb/MiniCPM4-8B-Eagle-vLLM): Eagle head in vLLM format, accelerating speculative inference for MiniCPM4-8B.
- [MiniCPM4-8B-marlin-Eagle-vLLM](https://huggingface.co/openbmb/MiniCPM4-8B-marlin-Eagle-vLLM): Quantized Eagle head for vLLM format, accelerating speculative inference for MiniCPM4-8B.
- [BitCPM4-0.5B](https://huggingface.co/openbmb/BitCPM4-0.5B): Extreme ternary quantization applied to MiniCPM4-0.5B compresses model parameters into ternary values, achieving a 90% reduction in bit width.
- [BitCPM4-1B](https://huggingface.co/openbmb/BitCPM4-1B): Extreme ternary quantization applied to MiniCPM3-1B compresses model parameters into ternary values, achieving a 90% reduction in bit width.
- [MiniCPM4-Survey](https://huggingface.co/openbmb/MiniCPM4-Survey): Based on MiniCPM4-8B, accepts users' quiries as input and autonomously generate trustworthy, long-form survey papers.
- [MiniCPM4-MCP](https://huggingface.co/openbmb/MiniCPM4-MCP): Based on MiniCPM4-8B, accepts users' queries and available MCP tools as input and autonomously calls relevant MCP tools to satisfy users' requirements.
## Introduction
MiniCPM 4 is an extremely efficient edge-side large model that has undergone efficient optimization across four dimensions: model architecture, learning algorithms, training data, and inference systems, achieving ultimate efficiency improvements.
- 🏗️ **Efficient Model Architecture:**
- InfLLM v2 -- Trainable Sparse Attention Mechanism: Adopts a trainable sparse attention mechanism architecture where each token only needs to compute relevance with less than 5% of tokens in 128K long text processing, significantly reducing computational overhead for long texts
- 🧠 **Efficient Learning Algorithms:**
- Model Wind Tunnel 2.0 -- Efficient Predictable Scaling: Introduces scaling prediction methods for performance of downstream tasks, enabling more precise model training configuration search
- BitCPM -- Ultimate Ternary Quantization: Compresses model parameter bit-width to 3 values, achieving 90% extreme model bit-width reduction
- Efficient Training Engineering Optimization: Adopts FP8 low-precision computing technology combined with Multi-token Prediction training strategy
- 📚 **High-Quality Training Data:**
- UltraClean -- High-quality Pre-training Data Filtering and Generation: Builds iterative data cleaning strategies based on efficient data verification, open-sourcing high-quality Chinese and English pre-training dataset [UltraFinweb](https://huggingface.co/datasets/openbmb/Ultra-FineWeb)
- UltraChat v2 -- High-quality Supervised Fine-tuning Data Generation: Constructs large-scale high-quality supervised fine-tuning datasets covering multiple dimensions including knowledge-intensive data, reasoning-intensive data, instruction-following data, long text understanding data, and tool calling data
- ⚡ **Efficient Inference System:**
- CPM.cu -- Lightweight and Efficient CUDA Inference Framework: Integrates sparse attention, model quantization, and speculative sampling to achieve efficient prefilling and decoding
- ArkInfer -- Cross-platform Deployment System: Supports efficient deployment across multiple backend environments, providing flexible cross-platform adaptation capabilities
## Usage
### Inference with [CPM.cu](https://github.com/OpenBMB/cpm.cu)
We recommend using [CPM.cu](https://github.com/OpenBMB/cpm.cu) for the inference of MiniCPM4. CPM.cu is a CUDA inference framework developed by OpenBMB, which integrates efficient sparse, speculative sampling, and quantization techniques, fully leveraging the efficiency advantages of MiniCPM4.
You can install CPM.cu by running the following command:
```bash
git clone https://github.com/OpenBMB/cpm.cu.git --recursive
cd cpm.cu
python3 setup.py install
```
MiniCPM4 natively supports context lengths of up to 32,768 tokens. To reproduce the long-text acceleration effect in the paper, we recommend using the LongRoPE factors that have been validated. Change the `rope_scaling` field in the `config.json` file as the following to enable LongRoPE.
```json
{
...,
"rope_scaling": {
"rope_type": "longrope",
"long_factor": [0.9977997200264581, 1.014658295992452, 1.0349680404997148, 1.059429246056193, 1.0888815016813513, 1.1243301355211495, 1.166977103606075, 1.2182568066927284, 1.2798772354275727, 1.3538666751582975, 1.4426259039919596, 1.5489853358570191, 1.6762658237220625, 1.8283407612492941, 2.0096956085876183, 2.225478927469756, 2.481536379650452, 2.784415934557119, 3.1413289096347365, 3.560047844772632, 4.048719380066383, 4.752651957515948, 5.590913044973868, 6.584005926629993, 7.7532214876576155, 9.119754865903639, 10.704443927019176, 12.524994176518703, 14.59739595363613, 16.93214476166354, 19.53823297353041, 22.417131025031697, 25.568260840911098, 28.991144156566317, 32.68408069090375, 36.65174474170465, 40.90396065611201, 45.4664008671033, 50.37147343433591, 55.6804490772103, 61.470816952306556, 67.8622707390618, 75.00516023410414, 83.11898235973767, 92.50044360202462, 103.57086856690864, 116.9492274587385, 118.16074567836519, 119.18497548708795, 120.04810876261652, 120.77352815196981, 121.38182790207875, 121.89094985353891, 122.31638758099915, 122.6714244963338, 122.9673822552567, 123.21386397019609, 123.41898278254268, 123.58957065488238, 123.73136519024158, 123.84917421274221, 123.94701903496814, 124.02825801299717, 124.09569231686116],
"short_factor": [0.9977997200264581, 1.014658295992452, 1.0349680404997148, 1.059429246056193, 1.0888815016813513, 1.1243301355211495, 1.166977103606075, 1.2182568066927284, 1.2798772354275727, 1.3538666751582975, 1.4426259039919596, 1.5489853358570191, 1.6762658237220625, 1.8283407612492941, 2.0096956085876183, 2.225478927469756, 2.481536379650452, 2.784415934557119, 3.1413289096347365, 3.560047844772632, 4.048719380066383, 4.752651957515948, 5.590913044973868, 6.584005926629993, 7.7532214876576155, 9.119754865903639, 10.704443927019176, 12.524994176518703, 14.59739595363613, 16.93214476166354, 19.53823297353041, 22.417131025031697, 25.568260840911098, 28.991144156566317, 32.68408069090375, 36.65174474170465, 40.90396065611201, 45.4664008671033, 50.37147343433591, 55.6804490772103, 61.470816952306556, 67.8622707390618, 75.00516023410414, 83.11898235973767, 92.50044360202462, 103.57086856690864, 116.9492274587385, 118.16074567836519, 119.18497548708795, 120.04810876261652, 120.77352815196981, 121.38182790207875, 121.89094985353891, 122.31638758099915, 122.6714244963338, 122.9673822552567, 123.21386397019609, 123.41898278254268, 123.58957065488238, 123.73136519024158, 123.84917421274221, 123.94701903496814, 124.02825801299717, 124.09569231686116],
"original_max_position_embeddings": 32768
}
}
```
After modification, you can run the following command to reproduce the long-context acceleration effect (the script will automatically download the model weights from HuggingFace)
```bash
python3 tests/test_generate.py
```
For more details about CPM.cu, please refer to [the repo CPM.cu](https://github.com/OpenBMB/cpm.cu).
### Inference with Transformers
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
torch.manual_seed(0)
path = 'openbmb/MiniCPM4-8B'
device = "cuda"
tokenizer = AutoTokenizer.from_pretrained(path)
model = AutoModelForCausalLM.from_pretrained(path, torch_dtype=torch.bfloat16, device_map=device, trust_remote_code=True)
# User can directly use the chat interface
# responds, history = model.chat(tokenizer, "Write an article about Artificial Intelligence.", temperature=0.7, top_p=0.7)
# print(responds)
# User can also use the generate interface
messages = [
{"role": "user", "content": "Write an article about Artificial Intelligence."},
]
prompt_text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
)
model_inputs = tokenizer([prompt_text], return_tensors="pt").to(device)
model_outputs = model.generate(
**model_inputs,
max_new_tokens=1024,
top_p=0.7,
temperature=0.7
)
output_token_ids = [
model_outputs[i][len(model_inputs[i]):] for i in range(len(model_inputs['input_ids']))
]
responses = tokenizer.batch_decode(output_token_ids, skip_special_tokens=True)[0]
print(responses)
```
MiniCPM4-8B supports `InfLLM v2`, a sparse attention mechanism designed for efficient long-sequence inference. It requires the [infllmv2_cuda_impl](https://github.com/OpenBMB/infllmv2_cuda_impl) library.
You can install it by running the following command:
```bash
git clone -b feature_infer https://github.com/OpenBMB/infllmv2_cuda_impl.git
cd infllmv2_cuda_impl
git submodule update --init --recursive
pip install -e . # or python setup.py install
```
To enable InfLLM v2, you need to add the `sparse_config` field in `config.json`:
```json
{
...,
"sparse_config": {
"kernel_size": 32,
"kernel_stride": 16,
"init_blocks": 1,
"block_size": 64,
"window_size": 2048,
"topk": 64,
"use_nope": false,
"dense_len": 8192
}
}
```
These parameters control the behavior of InfLLM v2:
* `kernel_size` (default: 32): The size of semantic kernels.
* `kernel_stride` (default: 16): The stride between adjacent kernels.
* `init_blocks` (default: 1): The number of initial blocks that every query token attends to. This ensures attention to the beginning of the sequence.
* `block_size` (default: 64): The block size for key-value blocks.
* `window_size` (default: 2048): The size of the local sliding window.
* `topk` (default: 64): The specifies that each token computes attention with only the top-k most relevant key-value blocks.
* `use_nope` (default: false): Whether to use the NOPE technique in block selection for improved performance.
* `dense_len` (default: 8192): Since Sparse Attention offers limited benefits for short sequences, the model can use standard (dense) attention for shorter texts. The model will use dense attention for sequences with a token length below `dense_len` and switch to sparse attention for sequences exceeding this length. Set this to `-1` to always use sparse attention regardless of sequence length.
MiniCPM4 natively supports context lengths of up to 32,768 tokens. For conversations where the total length (including both input and output) significantly exceeds this limit, we recommend using RoPE scaling techniques for effective handling of long texts. We have validated the model's performance on context lengths of up to 131,072 tokens by modifying the LongRoPE factor.
You can apply the LongRoPE factor modification by modifying the model files. Specifically, in the `config.json` file, adjust the `rope_scaling` fields.
```json
{
...,
"rope_scaling": {
"rope_type": "longrope",
"long_factor": [0.9977997200264581, 1.014658295992452, 1.0349680404997148, 1.059429246056193, 1.0888815016813513, 1.1243301355211495, 1.166977103606075, 1.2182568066927284, 1.2798772354275727, 1.3538666751582975, 1.4426259039919596, 1.5489853358570191, 1.6762658237220625, 1.8283407612492941, 2.0096956085876183, 2.225478927469756, 2.481536379650452, 2.784415934557119, 3.1413289096347365, 3.560047844772632, 4.048719380066383, 4.752651957515948, 5.590913044973868, 6.584005926629993, 7.7532214876576155, 9.119754865903639, 10.704443927019176, 12.524994176518703, 14.59739595363613, 16.93214476166354, 19.53823297353041, 22.417131025031697, 25.568260840911098, 28.991144156566317, 32.68408069090375, 36.65174474170465, 40.90396065611201, 45.4664008671033, 50.37147343433591, 55.6804490772103, 61.470816952306556, 67.8622707390618, 75.00516023410414, 83.11898235973767, 92.50044360202462, 103.57086856690864, 116.9492274587385, 118.16074567836519, 119.18497548708795, 120.04810876261652, 120.77352815196981, 121.38182790207875, 121.89094985353891, 122.31638758099915, 122.6714244963338, 122.9673822552567, 123.21386397019609, 123.41898278254268, 123.58957065488238, 123.73136519024158, 123.84917421274221, 123.94701903496814, 124.02825801299717, 124.09569231686116],
"short_factor": [0.9977997200264581, 1.014658295992452, 1.0349680404997148, 1.059429246056193, 1.0888815016813513, 1.1243301355211495, 1.166977103606075, 1.2182568066927284, 1.2798772354275727, 1.3538666751582975, 1.4426259039919596, 1.5489853358570191, 1.6762658237220625, 1.8283407612492941, 2.0096956085876183, 2.225478927469756, 2.481536379650452, 2.784415934557119, 3.1413289096347365, 3.560047844772632, 4.048719380066383, 4.752651957515948, 5.590913044973868, 6.584005926629993, 7.7532214876576155, 9.119754865903639, 10.704443927019176, 12.524994176518703, 14.59739595363613, 16.93214476166354, 19.53823297353041, 22.417131025031697, 25.568260840911098, 28.991144156566317, 32.68408069090375, 36.65174474170465, 40.90396065611201, 45.4664008671033, 50.37147343433591, 55.6804490772103, 61.470816952306556, 67.8622707390618, 75.00516023410414, 83.11898235973767, 92.50044360202462, 103.57086856690864, 116.9492274587385, 118.16074567836519, 119.18497548708795, 120.04810876261652, 120.77352815196981, 121.38182790207875, 121.89094985353891, 122.31638758099915, 122.6714244963338, 122.9673822552567, 123.21386397019609, 123.41898278254268, 123.58957065488238, 123.73136519024158, 123.84917421274221, 123.94701903496814, 124.02825801299717, 124.09569231686116],
"original_max_position_embeddings": 32768
}
}
```
### Inference with [SGLang](https://github.com/sgl-project/sglang)
For now, you need to install our forked version of SGLang.
```bash
git clone -b openbmb https://github.com/OpenBMB/sglang.git
cd sglang
pip install --upgrade pip
pip install -e "python[all]"
```
You can start the inference server by running the following command:
```bash
python -m sglang.launch_server --model openbmb/MiniCPM4-8B --trust-remote-code --port 30000 --chat-template chatml
```
Then you can use the chat interface by running the following command:
```python
import openai
client = openai.Client(base_url=f"http://localhost:30000/v1", api_key="None")
response = client.chat.completions.create(
model="openbmb/MiniCPM4-8B",
messages=[
{"role": "user", "content": "Write an article about Artificial Intelligence."},
],
temperature=0.7,
max_tokens=1024,
)
print(response.choices[0].message.content)
```
### Inference with [vLLM](https://github.com/vllm-project/vllm)
For now, you need to install the latest version of vLLM.
```
pip install -U vllm \
--pre \
--extra-index-url https://wheels.vllm.ai/nightly
```
Then you can inference MiniCPM4-8B with vLLM:
```python
from transformers import AutoTokenizer
from vllm import LLM, SamplingParams
model_name = "openbmb/MiniCPM4-8B"
prompt = [{"role": "user", "content": "Please recommend 5 tourist attractions in Beijing. "}]
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
input_text = tokenizer.apply_chat_template(prompt, tokenize=False, add_generation_prompt=True)
llm = LLM(
model=model_name,
trust_remote_code=True,
max_num_batched_tokens=32768,
dtype="bfloat16",
gpu_memory_utilization=0.8,
)
sampling_params = SamplingParams(top_p=0.7, temperature=0.7, max_tokens=1024, repetition_penalty=1.02)
outputs = llm.generate(prompts=input_text, sampling_params=sampling_params)
print(outputs[0].outputs[0].text)
```
Also, you can start the inference server by running the following command:
> **Note**: In vLLM's chat API, `add_special_tokens` is `False` by default. This means important special tokens—such as the beginning-of-sequence (BOS) token—will not be added automatically. To ensure the input prompt is correctly formatted for the model, you should explicitly set `extra_body={"add_special_tokens": True}`.
```bash
vllm serve openbmb/MiniCPM4-8B
```
Then you can use the chat interface by running the following code:
```python
import openai
client = openai.Client(base_url="http://localhost:8000/v1", api_key="EMPTY")
response = client.chat.completions.create(
model="openbmb/MiniCPM4-8B",
messages=[
{"role": "user", "content": "Write an article about Artificial Intelligence."},
],
temperature=0.7,
max_tokens=1024,
extra_body=dict(add_special_tokens=True), # Ensures special tokens are added for chat template
)
print(response.choices[0].message.content)
```
## Evaluation Results
On two typical end-side chips, Jetson AGX Orin and RTX 4090, MiniCPM4 demonstrates significantly faster processing speed compared to similar-size models in long text processing tasks. As text length increases, MiniCPM4's efficiency advantage becomes more pronounced. On the Jetson AGX Orin platform, compared to Qwen3-8B, MiniCPM4 achieves approximately 7x decoding speed improvement.
#### Comprehensive Evaluation
MiniCPM4 launches end-side versions with 8B and 0.5B parameter scales, both achieving best-in-class performance in their respective categories.
#### Long Text Evaluation
MiniCPM4 is pre-trained on 32K long texts and achieves length extension through YaRN technology. In the 128K long text needle-in-a-haystack task, MiniCPM4 demonstrates outstanding performance.
## Statement
- As a language model, MiniCPM generates content by learning from a vast amount of text.
- However, it does not possess the ability to comprehend or express personal opinions or value judgments.
- Any content generated by MiniCPM does not represent the viewpoints or positions of the model developers.
- Therefore, when using content generated by MiniCPM, users should take full responsibility for evaluating and verifying it on their own.
## LICENSE
- This repository and MiniCPM models are released under the [Apache-2.0](https://github.com/OpenBMB/MiniCPM/blob/main/LICENSE) License.
## Citation
- Please cite our [paper](https://github.com/OpenBMB/MiniCPM/tree/main/report/MiniCPM_4_Technical_Report.pdf) if you find our work valuable.
```bibtex
@article{minicpm4,
title={{MiniCPM4}: Ultra-Efficient LLMs on End Devices},
author={MiniCPM Team},
year={2025}
}
```
<!--End Original Model Card-->
---
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/Qwen3-Embedding-8B-GGUF | Mungert | 2025-06-15T19:36:42Z | 1,574 | 2 | sentence-transformers | [
"sentence-transformers",
"gguf",
"transformers",
"sentence-similarity",
"feature-extraction",
"arxiv:2506.05176",
"base_model:Qwen/Qwen3-8B-Base",
"base_model:quantized:Qwen/Qwen3-8B-Base",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | feature-extraction | 2025-06-10T18:12:06Z | ---
license: apache-2.0
base_model:
- Qwen/Qwen3-8B-Base
tags:
- transformers
- sentence-transformers
- sentence-similarity
- feature-extraction
---
# <span style="color: #7FFF7F;">Qwen/Qwen3-Embedding-8B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`1f63e75f`](https://github.com/ggerganov/llama.cpp/commit/1f63e75f3b5dc7f44dbe63c8a41d23958fe95bc0).
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
# Qwen3-Embedding-8B
<p align="center">
<img src="https://qianwen-res.oss-accelerate-overseas.aliyuncs.com/logo_qwen3.png" width="400"/>
<p>
## Highlights
The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B). This series inherits the exceptional multilingual capabilities, long-text understanding, and reasoning skills of its foundational model. The Qwen3 Embedding series represents significant advancements in multiple text embedding and ranking tasks, including text retrieval, code retrieval, text classification, text clustering, and bitext mining.
**Exceptional Versatility**: The embedding model has achieved state-of-the-art performance across a wide range of downstream application evaluations. The 8B size embedding model ranks **No.1** in the MTEB multilingual leaderboard (as of June 5, 2025, score **70.58**), while the reranking model excels in various text retrieval scenarios.
**Comprehensive Flexibility**: The Qwen3 Embedding series offers a full spectrum of sizes (from 0.6B to 8B) for both embedding and reranking models, catering to diverse use cases that prioritize efficiency and effectiveness. Developers can seamlessly combine these two modules. Additionally, the embedding model allows for flexible vector definitions across all dimensions, and both embedding and reranking models support user-defined instructions to enhance performance for specific tasks, languages, or scenarios.
**Multilingual Capability**: The Qwen3 Embedding series offer support for over 100 languages, thanks to the multilingual capabilites of Qwen3 models. This includes various programming languages, and provides robust multilingual, cross-lingual, and code retrieval capabilities.
**Qwen3-Embedding-8B** has the following features:
- Model Type: Text Embedding
- Supported Languages: 100+ Languages
- Number of Paramaters: 8B
- Context Length: 32k
- Embedding Dimension: Up to 4096, supports user-defined output dimensions ranging from 32 to 4096
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3-embedding/), [GitHub](https://github.com/QwenLM/Qwen3-Embedding).
## Qwen3 Embedding Series Model list
| Model Type | Models | Size | Layers | Sequence Length | Embedding Dimension | MRL Support | Instruction Aware |
|------------------|----------------------|------|--------|-----------------|---------------------|-------------|----------------|
| Text Embedding | [Qwen3-Embedding-0.6B](https://huggingface.co/Qwen/Qwen3-Embedding-0.6B) | 0.6B | 28 | 32K | 1024 | Yes | Yes |
| Text Embedding | [Qwen3-Embedding-4B](https://huggingface.co/Qwen/Qwen3-Embedding-4B) | 4B | 36 | 32K | 2560 | Yes | Yes |
| Text Embedding | [Qwen3-Embedding-8B](https://huggingface.co/Qwen/Qwen3-Embedding-8B) | 8B | 36 | 32K | 4096 | Yes | Yes |
| Text Reranking | [Qwen3-Reranker-0.6B](https://huggingface.co/Qwen/Qwen3-Reranker-0.6B) | 0.6B | 28 | 32K | - | - | Yes |
| Text Reranking | [Qwen3-Reranker-4B](https://huggingface.co/Qwen/Qwen3-Reranker-4B) | 4B | 36 | 32K | - | - | Yes |
| Text Reranking | [Qwen3-Reranker-8B](https://huggingface.co/Qwen/Qwen3-Reranker-8B) | 8B | 36 | 32K | - | - | Yes |
> **Note**:
> - `MRL Support` indicates whether the embedding model supports custom dimensions for the final embedding.
> - `Instruction Aware` notes whether the embedding or reranking model supports customizing the input instruction according to different tasks.
> - Our evaluation indicates that, for most downstream tasks, using instructions (instruct) typically yields an improvement of 1% to 5% compared to not using them. Therefore, we recommend that developers create tailored instructions specific to their tasks and scenarios. In multilingual contexts, we also advise users to write their instructions in English, as most instructions utilized during the model training process were originally written in English.
## Usage
With Transformers versions earlier than 4.51.0, you may encounter the following error:
```
KeyError: 'qwen3'
```
### Sentence Transformers Usage
```python
# Requires transformers>=4.51.0
# Requires sentence-transformers>=2.7.0
from sentence_transformers import SentenceTransformer
# Load the model
model = SentenceTransformer("Qwen/Qwen3-Embedding-8B")
# We recommend enabling flash_attention_2 for better acceleration and memory saving,
# together with setting `padding_side` to "left":
# model = SentenceTransformer(
# "Qwen/Qwen3-Embedding-8B",
# model_kwargs={"attn_implementation": "flash_attention_2", "device_map": "auto"},
# tokenizer_kwargs={"padding_side": "left"},
# )
# The queries and documents to embed
queries = [
"What is the capital of China?",
"Explain gravity",
]
documents = [
"The capital of China is Beijing.",
"Gravity is a force that attracts two bodies towards each other. It gives weight to physical objects and is responsible for the movement of planets around the sun.",
]
# Encode the queries and documents. Note that queries benefit from using a prompt
# Here we use the prompt called "query" stored under `model.prompts`, but you can
# also pass your own prompt via the `prompt` argument
query_embeddings = model.encode(queries, prompt_name="query")
document_embeddings = model.encode(documents)
# Compute the (cosine) similarity between the query and document embeddings
similarity = model.similarity(query_embeddings, document_embeddings)
print(similarity)
# tensor([[0.7493, 0.0751],
# [0.0880, 0.6318]])
```
### Transformers Usage
```python
# Requires transformers>=4.51.0
import torch
import torch.nn.functional as F
from torch import Tensor
from transformers import AutoTokenizer, AutoModel
def last_token_pool(last_hidden_states: Tensor,
attention_mask: Tensor) -> Tensor:
left_padding = (attention_mask[:, -1].sum() == attention_mask.shape[0])
if left_padding:
return last_hidden_states[:, -1]
else:
sequence_lengths = attention_mask.sum(dim=1) - 1
batch_size = last_hidden_states.shape[0]
return last_hidden_states[torch.arange(batch_size, device=last_hidden_states.device), sequence_lengths]
def get_detailed_instruct(task_description: str, query: str) -> str:
return f'Instruct: {task_description}\nQuery:{query}'
# Each query must come with a one-sentence instruction that describes the task
task = 'Given a web search query, retrieve relevant passages that answer the query'
queries = [
get_detailed_instruct(task, 'What is the capital of China?'),
get_detailed_instruct(task, 'Explain gravity')
]
# No need to add instruction for retrieval documents
documents = [
"The capital of China is Beijing.",
"Gravity is a force that attracts two bodies towards each other. It gives weight to physical objects and is responsible for the movement of planets around the sun."
]
input_texts = queries + documents
tokenizer = AutoTokenizer.from_pretrained('Qwen/Qwen3-Embedding-8B', padding_side='left')
model = AutoModel.from_pretrained('Qwen/Qwen3-Embedding-8B')
# We recommend enabling flash_attention_2 for better acceleration and memory saving.
# model = AutoModel.from_pretrained('Qwen/Qwen3-Embedding-8B', attn_implementation="flash_attention_2", torch_dtype=torch.float16).cuda()
max_length = 8192
# Tokenize the input texts
batch_dict = tokenizer(
input_texts,
padding=True,
truncation=True,
max_length=max_length,
return_tensors="pt",
)
batch_dict.to(model.device)
outputs = model(**batch_dict)
embeddings = last_token_pool(outputs.last_hidden_state, batch_dict['attention_mask'])
# normalize embeddings
embeddings = F.normalize(embeddings, p=2, dim=1)
scores = (embeddings[:2] @ embeddings[2:].T)
print(scores.tolist())
# [[0.7493016123771667, 0.0750647559762001], [0.08795969933271408, 0.6318399906158447]]
```
### vLLM Usage
```python
# Requires vllm>=0.8.5
import torch
import vllm
from vllm import LLM
def get_detailed_instruct(task_description: str, query: str) -> str:
return f'Instruct: {task_description}\nQuery:{query}'
# Each query must come with a one-sentence instruction that describes the task
task = 'Given a web search query, retrieve relevant passages that answer the query'
queries = [
get_detailed_instruct(task, 'What is the capital of China?'),
get_detailed_instruct(task, 'Explain gravity')
]
# No need to add instruction for retrieval documents
documents = [
"The capital of China is Beijing.",
"Gravity is a force that attracts two bodies towards each other. It gives weight to physical objects and is responsible for the movement of planets around the sun."
]
input_texts = queries + documents
model = LLM(model="Qwen/Qwen3-Embedding-8B", task="embed")
outputs = model.embed(input_texts)
embeddings = torch.tensor([o.outputs.embedding for o in outputs])
scores = (embeddings[:2] @ embeddings[2:].T)
print(scores.tolist())
# [[0.7482624650001526, 0.07556197047233582], [0.08875375241041183, 0.6300010681152344]]
```
📌 **Tip**: We recommend that developers customize the `instruct` according to their specific scenarios, tasks, and languages. Our tests have shown that in most retrieval scenarios, not using an `instruct` on the query side can lead to a drop in retrieval performance by approximately 1% to 5%.
## Evaluation
### MTEB (Multilingual)
| Model | Size | Mean (Task) | Mean (Type) | Bitxt Mining | Class. | Clust. | Inst. Retri. | Multi. Class. | Pair. Class. | Rerank | Retri. | STS |
|----------------------------------|:-------:|:-------------:|:-------------:|:--------------:|:--------:|:--------:|:--------------:|:---------------:|:--------------:|:--------:|:--------:|:------:|
| NV-Embed-v2 | 7B | 56.29 | 49.58 | 57.84 | 57.29 | 40.80 | 1.04 | 18.63 | 78.94 | 63.82 | 56.72 | 71.10|
| GritLM-7B | 7B | 60.92 | 53.74 | 70.53 | 61.83 | 49.75 | 3.45 | 22.77 | 79.94 | 63.78 | 58.31 | 73.33|
| BGE-M3 | 0.6B | 59.56 | 52.18 | 79.11 | 60.35 | 40.88 | -3.11 | 20.1 | 80.76 | 62.79 | 54.60 | 74.12|
| multilingual-e5-large-instruct | 0.6B | 63.22 | 55.08 | 80.13 | 64.94 | 50.75 | -0.40 | 22.91 | 80.86 | 62.61 | 57.12 | 76.81|
| gte-Qwen2-1.5B-instruct | 1.5B | 59.45 | 52.69 | 62.51 | 58.32 | 52.05 | 0.74 | 24.02 | 81.58 | 62.58 | 60.78 | 71.61|
| gte-Qwen2-7b-Instruct | 7B | 62.51 | 55.93 | 73.92 | 61.55 | 52.77 | 4.94 | 25.48 | 85.13 | 65.55 | 60.08 | 73.98|
| text-embedding-3-large | - | 58.93 | 51.41 | 62.17 | 60.27 | 46.89 | -2.68 | 22.03 | 79.17 | 63.89 | 59.27 | 71.68|
| Cohere-embed-multilingual-v3.0 | - | 61.12 | 53.23 | 70.50 | 62.95 | 46.89 | -1.89 | 22.74 | 79.88 | 64.07 | 59.16 | 74.80|
| gemini-embedding-exp-03-07 | - | 68.37 | 59.59 | 79.28 | 71.82 | 54.59 | 5.18 | **29.16** | 83.63 | 65.58 | 67.71 | 79.40|
| **Qwen3-Embedding-0.6B** | 0.6B | 64.33 | 56.00 | 72.22 | 66.83 | 52.33 | 5.09 | 24.59 | 80.83 | 61.41 | 64.64 | 76.17|
| **Qwen3-Embedding-4B** | 4B | 69.45 | 60.86 | 79.36 | 72.33 | 57.15 | **11.56** | 26.77 | 85.05 | 65.08 | 69.60 | 80.86|
| **Qwen3-Embedding-8B** | 8B | **70.58** | **61.69** | **80.89** | **74.00** | **57.65** | 10.06 | 28.66 | **86.40** | **65.63** | **70.88** | **81.08** |
> **Note**: For compared models, the scores are retrieved from MTEB online [leaderboard](https://huggingface.co/spaces/mteb/leaderboard) on May 24th, 2025.
### MTEB (Eng v2)
| MTEB English / Models | Param. | Mean(Task) | Mean(Type) | Class. | Clust. | Pair Class. | Rerank. | Retri. | STS | Summ. |
|--------------------------------|:--------:|:------------:|:------------:|:--------:|:--------:|:-------------:|:---------:|:--------:|:-------:|:-------:|
| multilingual-e5-large-instruct | 0.6B | 65.53 | 61.21 | 75.54 | 49.89 | 86.24 | 48.74 | 53.47 | 84.72 | 29.89 |
| NV-Embed-v2 | 7.8B | 69.81 | 65.00 | 87.19 | 47.66 | 88.69 | 49.61 | 62.84 | 83.82 | 35.21 |
| GritLM-7B | 7.2B | 67.07 | 63.22 | 81.25 | 50.82 | 87.29 | 49.59 | 54.95 | 83.03 | 35.65 |
| gte-Qwen2-1.5B-instruct | 1.5B | 67.20 | 63.26 | 85.84 | 53.54 | 87.52 | 49.25 | 50.25 | 82.51 | 33.94 |
| stella_en_1.5B_v5 | 1.5B | 69.43 | 65.32 | 89.38 | 57.06 | 88.02 | 50.19 | 52.42 | 83.27 | 36.91 |
| gte-Qwen2-7B-instruct | 7.6B | 70.72 | 65.77 | 88.52 | 58.97 | 85.9 | 50.47 | 58.09 | 82.69 | 35.74 |
| gemini-embedding-exp-03-07 | - | 73.3 | 67.67 | 90.05 | **59.39** | **87.7** | 48.59 | 64.35 | 85.29 | **38.28** |
| **Qwen3-Embedding-0.6B** | 0.6B | 70.70 | 64.88 | 85.76 | 54.05 | 84.37 | 48.18 | 61.83 | 86.57 | 33.43 |
| **Qwen3-Embedding-4B** | 4B | 74.60 | 68.10 | 89.84 | 57.51 | 87.01 | 50.76 | 68.46 | **88.72** | 34.39 |
| **Qwen3-Embedding-8B** | 8B | **75.22** | **68.71** | **90.43** | 58.57 | 87.52 | **51.56** | **69.44** | 88.58 | 34.83 |
### C-MTEB (MTEB Chinese)
| C-MTEB | Param. | Mean(Task) | Mean(Type) | Class. | Clust. | Pair Class. | Rerank. | Retr. | STS |
|------------------|--------|------------|------------|--------|--------|-------------|---------|-------|-------|
| multilingual-e5-large-instruct | 0.6B | 58.08 | 58.24 | 69.80 | 48.23 | 64.52 | 57.45 | 63.65 | 45.81 |
| bge-multilingual-gemma2 | 9B | 67.64 |68.52 | 75.31 | 59.30 | 86.67 | 68.28 | 73.73 | 55.19 |
| gte-Qwen2-1.5B-instruct | 1.5B | 67.12 | 67.79 | 72.53 | 54.61 | 79.5 | 68.21 | 71.86 | 60.05 |
| gte-Qwen2-7B-instruct | 7.6B | 71.62 | 72.19 | 75.77 | 66.06 | 81.16 | 69.24 | 75.70 | 65.20 |
| ritrieve_zh_v1 | 0.3B | 72.71 | 73.85 | 76.88 | 66.5 | **85.98** | **72.86** | 76.97 | **63.92** |
| **Qwen3-Embedding-0.6B** | 0.6B | 66.33 | 67.45 | 71.40 | 68.74 | 76.42 | 62.58 | 71.03 | 54.52 |
| **Qwen3-Embedding-4B** | 4B | 72.27 | 73.51 | 75.46 | 77.89 | 83.34 | 66.05 | 77.03 | 61.26 |
| **Qwen3-Embedding-8B** | 8B | **73.84** | **75.00** | **76.97** | **80.08** | 84.23 | 66.99 | **78.21** | 63.53 |
## Citation
If you find our work helpful, feel free to give us a cite.
```
@article{qwen3embedding,
title={Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models},
author={Zhang, Yanzhao and Li, Mingxin and Long, Dingkun and Zhang, Xin and Lin, Huan and Yang, Baosong and Xie, Pengjun and Yang, An and Liu, Dayiheng and Lin, Junyang and Huang, Fei and Zhou, Jingren},
journal={arXiv preprint arXiv:2506.05176},
year={2025}
}
```
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/Qwen2.5-Omni-3B-GGUF | Mungert | 2025-06-15T19:36:37Z | 1,307 | 2 | transformers | [
"transformers",
"gguf",
"multimodal",
"any-to-any",
"en",
"arxiv:2503.20215",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | any-to-any | 2025-06-10T12:18:40Z | ---
license: other
license_name: qwen-research
license_link: LICENSE
language:
- en
tags:
- multimodal
library_name: transformers
pipeline_tag: any-to-any
---
# <span style="color: #7FFF7F;">Qwen2.5-Omni-3B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`7f4fbe51`](https://github.com/ggerganov/llama.cpp/commit/7f4fbe5183b23b6b2e25fd1ccc5d1fa8bb010cb7).
## <span style="color: #7FFF7F;"> Quantization beyond the IMatrix</span>
Testing a new quantization method using rules to bump important layers above what the standard imatrix would use.
I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
This does create larger model files but increases precision for a given model size.
### **Please provide feedback on how you find this method performs**
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
# Qwen2.5-Omni
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
<img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>
</a>
## Overview
### Introduction
Qwen2.5-Omni is an end-to-end multimodal model designed to perceive diverse modalities, including text, images, audio, and video, while simultaneously generating text and natural speech responses in a streaming manner.
<p align="center">
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-Omni/qwen_omni.png" width="80%"/>
<p>
### Key Features
* **Omni and Novel Architecture**: We propose Thinker-Talker architecture, an end-to-end multimodal model designed to perceive diverse modalities, including text, images, audio, and video, while simultaneously generating text and natural speech responses in a streaming manner. We propose a novel position embedding, named TMRoPE (Time-aligned Multimodal RoPE), to synchronize the timestamps of video inputs with audio.
* **Real-Time Voice and Video Chat**: Architecture designed for fully real-time interactions, supporting chunked input and immediate output.
* **Natural and Robust Speech Generation**: Surpassing many existing streaming and non-streaming alternatives, demonstrating superior robustness and naturalness in speech generation.
* **Strong Performance Across Modalities**: Exhibiting exceptional performance across all modalities when benchmarked against similarly sized single-modality models. Qwen2.5-Omni outperforms the similarly sized Qwen2-Audio in audio capabilities and achieves comparable performance to Qwen2.5-VL-7B.
* **Excellent End-to-End Speech Instruction Following**: Qwen2.5-Omni shows performance in end-to-end speech instruction following that rivals its effectiveness with text inputs, evidenced by benchmarks such as MMLU and GSM8K.
### Model Architecture
<p align="center">
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-Omni/overview.png" width="80%"/>
<p>
### Performance
We conducted a comprehensive evaluation of Qwen2.5-Omni, which demonstrates strong performance across all modalities when compared to similarly sized single-modality models and closed-source models like Qwen2.5-VL-7B, Qwen2-Audio, and Gemini-1.5-pro. In tasks requiring the integration of multiple modalities, such as OmniBench, Qwen2.5-Omni achieves state-of-the-art performance. Furthermore, in single-modality tasks, it excels in areas including speech recognition (Common Voice), translation (CoVoST2), audio understanding (MMAU), image reasoning (MMMU, MMStar), video understanding (MVBench), and speech generation (Seed-tts-eval and subjective naturalness).
<p align="center">
<img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-Omni/bar.png" width="80%"/>
<p>
<details>
<summary>Multimodality -> Text</summary>
<table class="tg"><thead>
<tr>
<th class="tg-0lax">Datasets</th>
<th class="tg-0lax">Model</th>
<th class="tg-0lax">Performance</th>
</tr></thead>
<tbody>
<tr>
<td class="tg-0lax" rowspan="10">OmniBench<br>Speech | Sound Event | Music | Avg</td>
<td class="tg-0lax">Gemini-1.5-Pro</td>
<td class="tg-0lax">42.67%|42.26%|46.23%|42.91%</td>
</tr>
<tr>
<td class="tg-0lax">MIO-Instruct</td>
<td class="tg-0lax">36.96%|33.58%|11.32%|33.80%</td>
</tr>
<tr>
<td class="tg-0lax">AnyGPT (7B)</td>
<td class="tg-0lax">17.77%|20.75%|13.21%|18.04%</td>
</tr>
<tr>
<td class="tg-0lax">video-SALMONN</td>
<td class="tg-0lax">34.11%|31.70%|<strong>56.60%</strong>|35.64%</td>
</tr>
<tr>
<td class="tg-0lax">UnifiedIO2-xlarge</td>
<td class="tg-0lax">39.56%|36.98%|29.25%|38.00%</td>
</tr>
<tr>
<td class="tg-0lax">UnifiedIO2-xxlarge</td>
<td class="tg-0lax">34.24%|36.98%|24.53%|33.98%</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">-|-|-|40.50%</td>
</tr>
<tr>
<td class="tg-0lax">Baichuan-Omni-1.5</td>
<td class="tg-0lax">-|-|-|42.90%</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">52.14%|52.08%|52.83%|52.19%</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>55.25%</strong>|<strong>60.00%</strong>|52.83%|<strong>56.13%</strong></td>
</tr>
</tbody></table>
</details>
<details>
<summary>Audio -> Text</summary>
<table class="tg"><thead>
<tr>
<th class="tg-0lax">Datasets</th>
<th class="tg-0lax">Model</th>
<th class="tg-0lax">Performance</th>
</tr></thead>
<tbody>
<tr>
<td class="tg-9j4x" colspan="3">ASR</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="12">Librispeech<br>dev-clean | dev other | test-clean | test-other</td>
<td class="tg-0lax">SALMONN</td>
<td class="tg-0lax">-|-|2.1|4.9</td>
</tr>
<tr>
<td class="tg-0lax">SpeechVerse</td>
<td class="tg-0lax">-|-|2.1|4.4</td>
</tr>
<tr>
<td class="tg-0lax">Whisper-large-v3</td>
<td class="tg-0lax">-|-|1.8|3.6</td>
</tr>
<tr>
<td class="tg-0lax">Llama-3-8B</td>
<td class="tg-0lax">-|-|-|3.4</td>
</tr>
<tr>
<td class="tg-0lax">Llama-3-70B</td>
<td class="tg-0lax">-|-|-|3.1</td>
</tr>
<tr>
<td class="tg-0lax">Seed-ASR-Multilingual</td>
<td class="tg-0lax">-|-|<strong>1.6</strong>|<strong>2.8</strong></td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">-|-|1.7|-</td>
</tr>
<tr>
<td class="tg-0lax">MinMo</td>
<td class="tg-0lax">-|-|1.7|3.9</td>
</tr>
<tr>
<td class="tg-0lax">Qwen-Audio</td>
<td class="tg-0lax">1.8|4.0|2.0|4.2</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax"><strong>1.3</strong>|<strong>3.4</strong>|<strong>1.6</strong>|3.6</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">2.0|4.1|2.2|4.5</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax">1.6|3.5|1.8|3.4</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="5">Common Voice 15<br>en | zh | yue | fr</td>
<td class="tg-0lax">Whisper-large-v3</td>
<td class="tg-0lax">9.3|12.8|10.9|10.8</td>
</tr>
<tr>
<td class="tg-0lax">MinMo</td>
<td class="tg-0lax">7.9|6.3|6.4|8.5</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">8.6|6.9|<strong>5.9</strong>|9.6</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">9.1|6.0|11.6|9.6</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>7.6</strong>|<strong>5.2</strong>|7.3|<strong>7.5</strong></td>
</tr>
<tr>
<td class="tg-0lax" rowspan="8">Fleurs<br>zh | en</td>
<td class="tg-0lax">Whisper-large-v3</td>
<td class="tg-0lax">7.7|4.1</td>
</tr>
<tr>
<td class="tg-0lax">Seed-ASR-Multilingual</td>
<td class="tg-0lax">-|<strong>3.4</strong></td>
</tr>
<tr>
<td class="tg-0lax">Megrez-3B-Omni</td>
<td class="tg-0lax">10.8|-</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">4.4|-</td>
</tr>
<tr>
<td class="tg-0lax">MinMo</td>
<td class="tg-0lax">3.0|3.8</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">7.5|-</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">3.2|5.4</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>3.0</strong>|4.1</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="6">Wenetspeech<br>test-net | test-meeting</td>
<td class="tg-0lax">Seed-ASR-Chinese</td>
<td class="tg-0lax"><strong>4.7|5.7</strong></td>
</tr>
<tr>
<td class="tg-0lax">Megrez-3B-Omni</td>
<td class="tg-0lax">-|16.4</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">6.9|-</td>
</tr>
<tr>
<td class="tg-0lax">MinMo</td>
<td class="tg-0lax">6.8|7.4</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">6.3|8.1</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax">5.9|7.7</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="4">Voxpopuli-V1.0-en</td>
<td class="tg-0lax">Llama-3-8B</td>
<td class="tg-0lax">6.2</td>
</tr>
<tr>
<td class="tg-0lax">Llama-3-70B</td>
<td class="tg-0lax"><strong>5.7</strong></td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">6.6</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax">5.8</td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">S2TT</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="9">CoVoST2<br>en-de | de-en | en-zh | zh-en</td>
<td class="tg-0lax">SALMONN</td>
<td class="tg-0lax">18.6|-|33.1|-</td>
</tr>
<tr>
<td class="tg-0lax">SpeechLLaMA</td>
<td class="tg-0lax">-|27.1|-|12.3</td>
</tr>
<tr>
<td class="tg-0lax">BLSP</td>
<td class="tg-0lax">14.1|-|-|-</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">-|-|<strong>48.2</strong>|27.2</td>
</tr>
<tr>
<td class="tg-0lax">MinMo</td>
<td class="tg-0lax">-|<strong>39.9</strong>|46.7|26.0</td>
</tr>
<tr>
<td class="tg-0lax">Qwen-Audio</td>
<td class="tg-0lax">25.1|33.9|41.5|15.7</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">29.9|35.2|45.2|24.4</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">28.3|38.1|41.4|26.6</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>30.2</strong>|37.7|41.4|<strong>29.4</strong></td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">SER</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="6">Meld</td>
<td class="tg-0lax">WavLM-large</td>
<td class="tg-0lax">0.542</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">0.524</td>
</tr>
<tr>
<td class="tg-0lax">Qwen-Audio</td>
<td class="tg-0lax">0.557</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">0.553</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">0.558</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>0.570</strong></td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">VSC</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="6">VocalSound</td>
<td class="tg-0lax">CLAP</td>
<td class="tg-0lax">0.495</td>
</tr>
<tr>
<td class="tg-0lax">Pengi</td>
<td class="tg-0lax">0.604</td>
</tr>
<tr>
<td class="tg-0lax">Qwen-Audio</td>
<td class="tg-0lax">0.929</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax"><strong>0.939</strong></td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">0.936</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>0.939</strong></td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">Music</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="3">GiantSteps Tempo</td>
<td class="tg-0lax">Llark-7B</td>
<td class="tg-0lax">0.86</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax"><strong>0.88</strong></td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>0.88</strong></td>
</tr>
<tr>
<td class="tg-0lax" rowspan="3">MusicCaps</td>
<td class="tg-0lax">LP-MusicCaps</td>
<td class="tg-0lax">0.291|0.149|0.089|<strong>0.061</strong>|0.129|0.130</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">0.325|<strong>0.163</strong>|<strong>0.093</strong>|0.057|<strong>0.132</strong>|<strong>0.229</strong></td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>0.328</strong>|0.162|0.090|0.055|0.127|0.225</td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">Audio Reasoning</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="4">MMAU<br>Sound | Music | Speech | Avg</td>
<td class="tg-0lax">Gemini-Pro-V1.5</td>
<td class="tg-0lax">56.75|49.40|58.55|54.90</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">54.95|50.98|42.04|49.20</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax"><strong>70.27</strong>|60.48|59.16|63.30</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax">67.87|<strong>69.16|59.76|65.60</strong></td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">Voice Chatting</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="9">VoiceBench<br>AlpacaEval | CommonEval | SD-QA | MMSU</td>
<td class="tg-0lax">Ultravox-v0.4.1-LLaMA-3.1-8B</td>
<td class="tg-0lax"><strong>4.55</strong>|3.90|53.35|47.17</td>
</tr>
<tr>
<td class="tg-0lax">MERaLiON</td>
<td class="tg-0lax">4.50|3.77|55.06|34.95</td>
</tr>
<tr>
<td class="tg-0lax">Megrez-3B-Omni</td>
<td class="tg-0lax">3.50|2.95|25.95|27.03</td>
</tr>
<tr>
<td class="tg-0lax">Lyra-Base</td>
<td class="tg-0lax">3.85|3.50|38.25|49.74</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">4.42|<strong>4.15</strong>|50.72|54.78</td>
</tr>
<tr>
<td class="tg-0lax">Baichuan-Omni-1.5</td>
<td class="tg-0lax">4.50|4.05|43.40|57.25</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">3.74|3.43|35.71|35.72</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">4.32|4.00|49.37|50.23</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax">4.49|3.93|<strong>55.71</strong>|<strong>61.32</strong></td>
</tr>
<tr>
<td class="tg-0lax" rowspan="9">VoiceBench<br>OpenBookQA | IFEval | AdvBench | Avg</td>
<td class="tg-0lax">Ultravox-v0.4.1-LLaMA-3.1-8B</td>
<td class="tg-0lax">65.27|<strong>66.88</strong>|98.46|71.45</td>
</tr>
<tr>
<td class="tg-0lax">MERaLiON</td>
<td class="tg-0lax">27.23|62.93|94.81|62.91</td>
</tr>
<tr>
<td class="tg-0lax">Megrez-3B-Omni</td>
<td class="tg-0lax">28.35|25.71|87.69|46.25</td>
</tr>
<tr>
<td class="tg-0lax">Lyra-Base</td>
<td class="tg-0lax">72.75|36.28|59.62|57.66</td>
</tr>
<tr>
<td class="tg-0lax">MiniCPM-o</td>
<td class="tg-0lax">78.02|49.25|97.69|71.69</td>
</tr>
<tr>
<td class="tg-0lax">Baichuan-Omni-1.5</td>
<td class="tg-0lax">74.51|54.54|97.31|71.14</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2-Audio</td>
<td class="tg-0lax">49.45|26.33|96.73|55.35</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B</td>
<td class="tg-0lax">74.73|42.10|98.85|68.81</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B</td>
<td class="tg-0lax"><strong>81.10</strong>|52.87|<strong>99.42</strong>|<strong>74.12</strong></td>
</tr>
</tbody></table>
</details>
<details>
<summary>Image -> Text</summary>
| Dataset | Qwen2.5-Omni-7B | Qwen2.5-Omni-3B | Other Best | Qwen2.5-VL-7B | GPT-4o-mini |
|--------------------------------|--------------|------------|------------|---------------|-------------|
| MMMU<sub>val</sub> | 59.2 | 53.1 | 53.9 | 58.6 | **60.0** |
| MMMU-Pro<sub>overall</sub> | 36.6 | 29.7 | - | **38.3** | 37.6 |
| MathVista<sub>testmini</sub> | 67.9 | 59.4 | **71.9** | 68.2 | 52.5 |
| MathVision<sub>full</sub> | 25.0 | 20.8 | 23.1 | **25.1** | - |
| MMBench-V1.1-EN<sub>test</sub> | 81.8 | 77.8 | 80.5 | **82.6** | 76.0 |
| MMVet<sub>turbo</sub> | 66.8 | 62.1 | **67.5** | 67.1 | 66.9 |
| MMStar | **64.0** | 55.7 | **64.0** | 63.9 | 54.8 |
| MME<sub>sum</sub> | 2340 | 2117 | **2372** | 2347 | 2003 |
| MuirBench | 59.2 | 48.0 | - | **59.2** | - |
| CRPE<sub>relation</sub> | **76.5** | 73.7 | - | 76.4 | - |
| RealWorldQA<sub>avg</sub> | 70.3 | 62.6 | **71.9** | 68.5 | - |
| MME-RealWorld<sub>en</sub> | **61.6** | 55.6 | - | 57.4 | - |
| MM-MT-Bench | 6.0 | 5.0 | - | **6.3** | - |
| AI2D | 83.2 | 79.5 | **85.8** | 83.9 | - |
| TextVQA<sub>val</sub> | 84.4 | 79.8 | 83.2 | **84.9** | - |
| DocVQA<sub>test</sub> | 95.2 | 93.3 | 93.5 | **95.7** | - |
| ChartQA<sub>test Avg</sub> | 85.3 | 82.8 | 84.9 | **87.3** | - |
| OCRBench_V2<sub>en</sub> | **57.8** | 51.7 | - | 56.3 | - |
| Dataset | Qwen2.5-Omni-7B | Qwen2.5-Omni-3B | Qwen2.5-VL-7B | Grounding DINO | Gemini 1.5 Pro |
|--------------------------|--------------|---------------|---------------|----------------|----------------|
| Refcoco<sub>val</sub> | 90.5 | 88.7 | 90.0 | **90.6** | 73.2 |
| Refcoco<sub>textA</sub> | **93.5** | 91.8 | 92.5 | 93.2 | 72.9 |
| Refcoco<sub>textB</sub> | 86.6 | 84.0 | 85.4 | **88.2** | 74.6 |
| Refcoco+<sub>val</sub> | 85.4 | 81.1 | 84.2 | **88.2** | 62.5 |
| Refcoco+<sub>textA</sub> | **91.0** | 87.5 | 89.1 | 89.0 | 63.9 |
| Refcoco+<sub>textB</sub> | **79.3** | 73.2 | 76.9 | 75.9 | 65.0 |
| Refcocog+<sub>val</sub> | **87.4** | 85.0 | 87.2 | 86.1 | 75.2 |
| Refcocog+<sub>test</sub> | **87.9** | 85.1 | 87.2 | 87.0 | 76.2 |
| ODinW | 42.4 | 39.2 | 37.3 | **55.0** | 36.7 |
| PointGrounding | 66.5 | 46.2 | **67.3** | - | - |
</details>
<details>
<summary>Video(without audio) -> Text</summary>
| Dataset | Qwen2.5-Omni-7B | Qwen2.5-Omni-3B | Other Best | Qwen2.5-VL-7B | GPT-4o-mini |
|-----------------------------|--------------|------------|------------|---------------|-------------|
| Video-MME<sub>w/o sub</sub> | 64.3 | 62.0 | 63.9 | **65.1** | 64.8 |
| Video-MME<sub>w sub</sub> | **72.4** | 68.6 | 67.9 | 71.6 | - |
| MVBench | **70.3** | 68.7 | 67.2 | 69.6 | - |
| EgoSchema<sub>test</sub> | **68.6** | 61.4 | 63.2 | 65.0 | - |
</details>
<details>
<summary>Zero-shot Speech Generation</summary>
<table class="tg"><thead>
<tr>
<th class="tg-0lax">Datasets</th>
<th class="tg-0lax">Model</th>
<th class="tg-0lax">Performance</th>
</tr></thead>
<tbody>
<tr>
<td class="tg-9j4x" colspan="3">Content Consistency</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="11">SEED<br>test-zh | test-en | test-hard </td>
<td class="tg-0lax">Seed-TTS_ICL</td>
<td class="tg-0lax">1.11 | 2.24 | 7.58</td>
</tr>
<tr>
<td class="tg-0lax">Seed-TTS_RL</td>
<td class="tg-0lax"><strong>1.00</strong> | 1.94 | <strong>6.42</strong></td>
</tr>
<tr>
<td class="tg-0lax">MaskGCT</td>
<td class="tg-0lax">2.27 | 2.62 | 10.27</td>
</tr>
<tr>
<td class="tg-0lax">E2_TTS</td>
<td class="tg-0lax">1.97 | 2.19 | -</td>
</tr>
<tr>
<td class="tg-0lax">F5-TTS</td>
<td class="tg-0lax">1.56 | <strong>1.83</strong> | 8.67</td>
</tr>
<tr>
<td class="tg-0lax">CosyVoice 2</td>
<td class="tg-0lax">1.45 | 2.57 | 6.83</td>
</tr>
<tr>
<td class="tg-0lax">CosyVoice 2-S</td>
<td class="tg-0lax">1.45 | 2.38 | 8.08</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B_ICL</td>
<td class="tg-0lax">1.95 | 2.87 | 9.92</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B_RL</td>
<td class="tg-0lax">1.58 | 2.51 | 7.86</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B_ICL</td>
<td class="tg-0lax">1.70 | 2.72 | 7.97</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B_RL</td>
<td class="tg-0lax">1.42 | 2.32 | 6.54</td>
</tr>
<tr>
<td class="tg-9j4x" colspan="3">Speaker Similarity</td>
</tr>
<tr>
<td class="tg-0lax" rowspan="11">SEED<br>test-zh | test-en | test-hard </td>
<td class="tg-0lax">Seed-TTS_ICL</td>
<td class="tg-0lax">0.796 | 0.762 | 0.776</td>
</tr>
<tr>
<td class="tg-0lax">Seed-TTS_RL</td>
<td class="tg-0lax"><strong>0.801</strong> | <strong>0.766</strong> | <strong>0.782</strong></td>
</tr>
<tr>
<td class="tg-0lax">MaskGCT</td>
<td class="tg-0lax">0.774 | 0.714 | 0.748</td>
</tr>
<tr>
<td class="tg-0lax">E2_TTS</td>
<td class="tg-0lax">0.730 | 0.710 | -</td>
</tr>
<tr>
<td class="tg-0lax">F5-TTS</td>
<td class="tg-0lax">0.741 | 0.647 | 0.713</td>
</tr>
<tr>
<td class="tg-0lax">CosyVoice 2</td>
<td class="tg-0lax">0.748 | 0.652 | 0.724</td>
</tr>
<tr>
<td class="tg-0lax">CosyVoice 2-S</td>
<td class="tg-0lax">0.753 | 0.654 | 0.732</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B_ICL</td>
<td class="tg-0lax">0.741 | 0.635 | 0.748</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-3B_RL</td>
<td class="tg-0lax">0.744 | 0.635 | 0.746</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B_ICL</td>
<td class="tg-0lax">0.752 | 0.632 | 0.747</td>
</tr>
<tr>
<td class="tg-0lax">Qwen2.5-Omni-7B_RL</td>
<td class="tg-0lax">0.754 | 0.641 | 0.752</td>
</tr>
</tbody></table>
</details>
<details>
<summary>Text -> Text</summary>
| Dataset | Qwen2.5-Omni-7B | Qwen2.5-Omni-3B | Qwen2.5-7B | Qwen2.5-3B | Qwen2-7B | Llama3.1-8B | Gemma2-9B |
|-----------------------------------|-----------|------------|------------|------------|------------|-------------|-----------|
| MMLU-Pro | 47.0 | 40.4 | **56.3** | 43.7 | 44.1 | 48.3 | 52.1 |
| MMLU-redux | 71.0 | 60.9 | **75.4** | 64.4 | 67.3 | 67.2 | 72.8 |
| LiveBench<sub>0831</sub> | 29.6 | 22.3 | **35.9** | 26.8 | 29.2 | 26.7 | 30.6 |
| GPQA | 30.8 | 34.3 | **36.4** | 30.3 | 34.3 | 32.8 | 32.8 |
| MATH | 71.5 | 63.6 | **75.5** | 65.9 | 52.9 | 51.9 | 44.3 |
| GSM8K | 88.7 | 82.6 | **91.6** | 86.7 | 85.7 | 84.5 | 76.7 |
| HumanEval | 78.7 | 70.7 | **84.8** | 74.4 | 79.9 | 72.6 | 68.9 |
| MBPP | 73.2 | 70.4 | **79.2** | 72.7 | 67.2 | 69.6 | 74.9 |
| MultiPL-E | 65.8 | 57.6 | **70.4** | 60.2 | 59.1 | 50.7 | 53.4 |
| LiveCodeBench<sub>2305-2409</sub> | 24.6 | 16.5 | **28.7** | 19.9 | 23.9 | 8.3 | 18.9 |
</details>
## Quickstart
Below, we provide simple examples to show how to use Qwen2.5-Omni with 🤗 Transformers. The codes of Qwen2.5-Omni has been in the latest Hugging face transformers and we advise you to build from source with command:
```
pip uninstall transformers
pip install git+https://github.com/huggingface/[email protected]
pip install accelerate
```
or you might encounter the following error:
```
KeyError: 'qwen2_5_omni'
```
We offer a toolkit to help you handle various types of audio and visual input more conveniently, as if you were using an API. This includes base64, URLs, and interleaved audio, images and videos. You can install it using the following command and make sure your system has `ffmpeg` installed:
```bash
# It's highly recommended to use `[decord]` feature for faster video loading.
pip install qwen-omni-utils[decord] -U
```
If you are not using Linux, you might not be able to install `decord` from PyPI. In that case, you can use `pip install qwen-omni-utils -U` which will fall back to using torchvision for video processing. However, you can still [install decord from source](https://github.com/dmlc/decord?tab=readme-ov-file#install-from-source) to get decord used when loading video.
### 🤗 Transformers Usage
Here we show a code snippet to show you how to use the chat model with `transformers` and `qwen_omni_utils`:
```python
import soundfile as sf
from transformers import Qwen2_5OmniForConditionalGeneration, Qwen2_5OmniProcessor
from qwen_omni_utils import process_mm_info
# default: Load the model on the available device(s)
model = Qwen2_5OmniForConditionalGeneration.from_pretrained("Qwen/Qwen2.5-Omni-3B", torch_dtype="auto", device_map="auto")
# We recommend enabling flash_attention_2 for better acceleration and memory saving.
# model = Qwen2_5OmniForConditionalGeneration.from_pretrained(
# "Qwen/Qwen2.5-Omni-3B",
# torch_dtype="auto",
# device_map="auto",
# attn_implementation="flash_attention_2",
# )
processor = Qwen2_5OmniProcessor.from_pretrained("Qwen/Qwen2.5-Omni-3B")
conversation = [
{
"role": "system",
"content": [
{"type": "text", "text": "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech."}
],
},
{
"role": "user",
"content": [
{"type": "video", "video": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-Omni/draw.mp4"},
],
},
]
# set use audio in video
USE_AUDIO_IN_VIDEO = True
# Preparation for inference
text = processor.apply_chat_template(conversation, add_generation_prompt=True, tokenize=False)
audios, images, videos = process_mm_info(conversation, use_audio_in_video=USE_AUDIO_IN_VIDEO)
inputs = processor(text=text, audio=audios, images=images, videos=videos, return_tensors="pt", padding=True, use_audio_in_video=USE_AUDIO_IN_VIDEO)
inputs = inputs.to(model.device).to(model.dtype)
# Inference: Generation of the output text and audio
text_ids, audio = model.generate(**inputs, use_audio_in_video=USE_AUDIO_IN_VIDEO)
text = processor.batch_decode(text_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(text)
sf.write(
"output.wav",
audio.reshape(-1).detach().cpu().numpy(),
samplerate=24000,
)
```
<details>
<summary>Minimum GPU memory requirements</summary>
|Model | Precision | 15(s) Video | 30(s) Video | 60(s) Video |
|--------------|-----------| ------------- | ------------- | ------------------ |
| Qwen-Omni-3B | FP32 | 89.10 GB | Not Recommend | Not Recommend |
| Qwen-Omni-3B | BF16 | 18.38 GB | 22.43 GB | 28.22 GB |
| Qwen-Omni-7B | FP32 | 93.56 GB | Not Recommend | Not Recommend |
| Qwen-Omni-7B | BF16 | 31.11 GB | 41.85 GB | 60.19 GB |
Note: The table above presents the theoretical minimum memory requirements for inference with `transformers` and `BF16` is test with `attn_implementation="flash_attention_2"`; however, in practice, the actual memory usage is typically at least 1.2 times higher. For more information, see the linked resource [here](https://huggingface.co/docs/accelerate/main/en/usage_guides/model_size_estimator).
</details>
<details>
<summary>Video URL resource usage</summary>
Video URL compatibility largely depends on the third-party library version. The details are in the table below. Change the backend by `FORCE_QWENVL_VIDEO_READER=torchvision` or `FORCE_QWENVL_VIDEO_READER=decord` if you prefer not to use the default one.
| Backend | HTTP | HTTPS |
|-------------|------|-------|
| torchvision >= 0.19.0 | ✅ | ✅ |
| torchvision < 0.19.0 | ❌ | ❌ |
| decord | ✅ | ❌ |
</details>
<details>
<summary>Batch inference</summary>
The model can batch inputs composed of mixed samples of various types such as text, images, audio and videos as input when `return_audio=False` is set. Here is an example.
```python
# Sample messages for batch inference
# Conversation with video only
conversation1 = [
{
"role": "system",
"content": [
{"type": "text", "text": "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech."}
],
},
{
"role": "user",
"content": [
{"type": "video", "video": "/path/to/video.mp4"},
]
}
]
# Conversation with audio only
conversation2 = [
{
"role": "system",
"content": [
{"type": "text", "text": "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech."}
],
},
{
"role": "user",
"content": [
{"type": "audio", "audio": "/path/to/audio.wav"},
]
}
]
# Conversation with pure text
conversation3 = [
{
"role": "system",
"content": [
{"type": "text", "text": "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech."}
],
},
{
"role": "user",
"content": "who are you?"
}
]
# Conversation with mixed media
conversation4 = [
{
"role": "system",
"content": [
{"type": "text", "text": "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech."}
],
},
{
"role": "user",
"content": [
{"type": "image", "image": "/path/to/image.jpg"},
{"type": "video", "video": "/path/to/video.mp4"},
{"type": "audio", "audio": "/path/to/audio.wav"},
{"type": "text", "text": "What are the elements can you see and hear in these medias?"},
],
}
]
# Combine messages for batch processing
conversations = [conversation1, conversation2, conversation3, conversation4]
# set use audio in video
USE_AUDIO_IN_VIDEO = True
# Preparation for batch inference
text = processor.apply_chat_template(conversations, add_generation_prompt=True, tokenize=False)
audios, images, videos = process_mm_info(conversations, use_audio_in_video=USE_AUDIO_IN_VIDEO)
inputs = processor(text=text, audio=audios, images=images, videos=videos, return_tensors="pt", padding=True, use_audio_in_video=USE_AUDIO_IN_VIDEO)
inputs = inputs.to(model.device).to(model.dtype)
# Batch Inference
text_ids = model.generate(**inputs, use_audio_in_video=USE_AUDIO_IN_VIDEO, return_audio=False)
text = processor.batch_decode(text_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(text)
```
</details>
### Usage Tips
#### Prompt for audio output
If users need audio output, the system prompt must be set as "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech.", otherwise the audio output may not work as expected.
```
{
"role": "system",
"content": [
{"type": "text", "text": "You are Qwen, a virtual human developed by the Qwen Team, Alibaba Group, capable of perceiving auditory and visual inputs, as well as generating text and speech."}
],
}
```
#### Use audio in video
In the process of multimodal interaction, the videos provided by users are often accompanied by audio (such as questions about the content in the video, or sounds generated by certain events in the video). This information is conducive to the model providing a better interactive experience. So we provide the following options for users to decide whether to use audio in video.
```python
# first place, in data preprocessing
audios, images, videos = process_mm_info(conversations, use_audio_in_video=True)
```
```python
# second place, in model processor
inputs = processor(text=text, audio=audios, images=images, videos=videos, return_tensors="pt",
padding=True, use_audio_in_video=True)
```
```python
# third place, in model inference
text_ids, audio = model.generate(**inputs, use_audio_in_video=True)
```
It is worth noting that during a multi-round conversation, the `use_audio_in_video` parameter in these places must be set to the same, otherwise unexpected results will occur.
#### Use audio output or not
The model supports both text and audio outputs, if users do not need audio outputs, they can call `model.disable_talker()` after init the model. This option will save about `~2GB` of GPU memory but the `return_audio` option for `generate` function will only allow to be set at `False`.
```python
model = Qwen2_5OmniForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-Omni-3B",
torch_dtype="auto",
device_map="auto"
)
model.disable_talker()
```
In order to obtain a flexible experience, we recommend that users can decide whether to return audio when `generate` function is called. If `return_audio` is set to `False`, the model will only return text outputs to get text responses faster.
```python
model = Qwen2_5OmniForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-Omni-3B",
torch_dtype="auto",
device_map="auto"
)
...
text_ids = model.generate(**inputs, return_audio=False)
```
#### Change voice type of output audio
Qwen2.5-Omni supports the ability to change the voice of the output audio. The `"Qwen/Qwen2.5-Omni-3B"` checkpoint support two voice types as follow:
| Voice Type | Gender | Description |
|------------|--------|-------------|
| Chelsie | Female | A honeyed, velvety voice that carries a gentle warmth and luminous clarity.|
| Ethan | Male | A bright, upbeat voice with infectious energy and a warm, approachable vibe.|
Users can use the `speaker` parameter of `generate` function to specify the voice type. By default, if `speaker` is not specified, the default voice type is `Chelsie`.
```python
text_ids, audio = model.generate(**inputs, speaker="Chelsie")
```
```python
text_ids, audio = model.generate(**inputs, speaker="Ethan")
```
#### Flash-Attention 2 to speed up generation
First, make sure to install the latest version of Flash Attention 2:
```bash
pip install -U flash-attn --no-build-isolation
```
Also, you should have hardware that is compatible with FlashAttention 2. Read more about it in the official documentation of the [flash attention repository](https://github.com/Dao-AILab/flash-attention). FlashAttention-2 can only be used when a model is loaded in `torch.float16` or `torch.bfloat16`.
To load and run a model using FlashAttention-2, add `attn_implementation="flash_attention_2"` when loading the model:
```python
from transformers import Qwen2_5OmniForConditionalGeneration
model = Qwen2_5OmniForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-Omni-3B",
device_map="auto",
torch_dtype=torch.bfloat16,
attn_implementation="flash_attention_2",
)
```
## Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil: :)
```BibTeX
@article{Qwen2.5-Omni,
title={Qwen2.5-Omni Technical Report},
author={Jin Xu, Zhifang Guo, Jinzheng He, Hangrui Hu, Ting He, Shuai Bai, Keqin Chen, Jialin Wang, Yang Fan, Kai Dang, Bin Zhang, Xiong Wang, Yunfei Chu, Junyang Lin},
journal={arXiv preprint arXiv:2503.20215},
year={2025}
}
```
<br>
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Peacemann/Qwen_QwQ-32B_LMUL | Peacemann | 2025-06-15T19:36:36Z | 0 | 0 | null | [
"qwen2",
"L-Mul,",
"optimazation",
"quantization",
"text-generation",
"research",
"experimental",
"conversational",
"base_model:Qwen/QwQ-32B",
"base_model:finetune:Qwen/QwQ-32B",
"license:apache-2.0",
"region:us"
] | text-generation | 2025-06-15T19:33:06Z | ---
license: apache-2.0
base_model: Qwen/QwQ-32B
tags:
- L-Mul,
- optimazation
- quantization
- text-generation
- research
- experimental
---
# Model Card for Qwen/QwQ-32B-LMUL
This model is a derivative of `Qwen/QwQ-32B`, modified to use a custom attention mechanism defined by the `l_mul_attention` function from the `lmul` library.
## Model Details
- **Original Model:** [Qwen/QwQ-32B](https://huggingface.co/Qwen/QwQ-32B)
- **Architecture:** qwen2
- **Modification:** The `forward` method of the `Qwen2Attention` module has been replaced (monkey-patched) with a custom implementation that utilizes the `l_mul_attention` logic.
## Scientific Rationale
This model was modified as part of a research project investigating alternative attention mechanisms in large language models. The `l_mul_attention` function implements a novel approach to calculating attention scores, and this model serves as a test case for evaluating its performance, efficiency, and impact on reasoning and generation tasks compared to the standard attention implementation.
By releasing this model, we hope to encourage further research into non-standard attention mechanisms and provide a practical example for the community to build upon.
## How to Get Started
You can use this model with the standard `transformers` library pipeline. Ensure you have `transformers`, `torch`, and `accelerate` installed.
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
# Make sure to log in with your Hugging Face token if the model is private
# from huggingface_hub import login
# login("your-hf-token")
model_id = "YOUR_HF_USERNAME/QwQ-32B_LMUL" # Replace with your Hugging Face username
device = "cuda" if torch.cuda.is_available() else "cpu"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
model_id,
torch_dtype=torch.bfloat16,
device_map="auto"
)
prompt = "How many r's are in the word \"strawberry\""
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)
```
## Intended Uses & Limitations
This model is intended primarily for research purposes. Its performance on standard benchmarks has not been fully evaluated. The custom attention mechanism may introduce unexpected behaviors or limitations not present in the original `Qwen/QwQ-32B` model.
## Licensing Information
This model is released under the `apache-2.0` license, which is the same license as the base model, `Qwen/QwQ-32B`. By using this model, you agree to the terms of the original license. It is your responsibility to ensure compliance with all applicable licenses and regulations. |
Mungert/SmolVLM-Instruct-GGUF | Mungert | 2025-06-15T19:36:32Z | 1,023 | 2 | transformers | [
"transformers",
"gguf",
"image-text-to-text",
"en",
"dataset:HuggingFaceM4/the_cauldron",
"dataset:HuggingFaceM4/Docmatix",
"arxiv:2504.05299",
"base_model:HuggingFaceTB/SmolLM2-1.7B-Instruct",
"base_model:quantized:HuggingFaceTB/SmolLM2-1.7B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | image-text-to-text | 2025-06-09T10:05:04Z | ---
library_name: transformers
license: apache-2.0
datasets:
- HuggingFaceM4/the_cauldron
- HuggingFaceM4/Docmatix
pipeline_tag: image-text-to-text
language:
- en
base_model:
- HuggingFaceTB/SmolLM2-1.7B-Instruct
- google/siglip-so400m-patch14-384
---
# <span style="color: #7FFF7F;">SmolVLM-Instruct GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`5787b5da`](https://github.com/ggerganov/llama.cpp/commit/5787b5da57e54dba760c2deeac1edf892e8fc450).
## <span style="color: #7FFF7F;"> Quantization beyond the IMatrix</span>
Testing a new quantization method using rules to bump important layers above what the standard imatrix would use.
I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
This does create larger model files but increases precision for a given model size.
### **Please provide feedback on how you find this method performs**
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/SmolVLM.png" width="800" height="auto" alt="Image description">
# SmolVLM
SmolVLM is a compact open multimodal model that accepts arbitrary sequences of image and text inputs to produce text outputs. Designed for efficiency, SmolVLM can answer questions about images, describe visual content, create stories grounded on multiple images, or function as a pure language model without visual inputs. Its lightweight architecture makes it suitable for on-device applications while maintaining strong performance on multimodal tasks.
## Model Summary
- **Developed by:** Hugging Face 🤗
- **Model type:** Multi-modal model (image+text)
- **Language(s) (NLP):** English
- **License:** Apache 2.0
- **Architecture:** Based on [Idefics3](https://huggingface.co/HuggingFaceM4/Idefics3-8B-Llama3) (see technical summary)
## Resources
- **Demo:** [SmolVLM Demo](https://huggingface.co/spaces/HuggingFaceTB/SmolVLM)
- **Blog:** [Blog post](https://huggingface.co/blog/smolvlm)
## Uses
SmolVLM can be used for inference on multimodal (image + text) tasks where the input comprises text queries along with one or more images. Text and images can be interleaved arbitrarily, enabling tasks like image captioning, visual question answering, and storytelling based on visual content. The model does not support image generation.
To fine-tune SmolVLM on a specific task, you can follow the fine-tuning tutorial.
<!-- todo: add link to fine-tuning tutorial -->
### Technical Summary
SmolVLM leverages the lightweight SmolLM2 language model to provide a compact yet powerful multimodal experience. It introduces several changes compared to previous Idefics models:
- **Image compression:** We introduce a more radical image compression compared to Idefics3 to enable the model to infer faster and use less RAM.
- **Visual Token Encoding:** SmolVLM uses 81 visual tokens to encode image patches of size 384×384. Larger images are divided into patches, each encoded separately, enhancing efficiency without compromising performance.
More details about the training and architecture are available in our technical report.
### How to get started
You can use transformers to load, infer and fine-tune SmolVLM.
```python
import torch
from PIL import Image
from transformers import AutoProcessor, AutoModelForVision2Seq
from transformers.image_utils import load_image
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
# Load images
image1 = load_image("https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg")
image2 = load_image("https://huggingface.co/spaces/merve/chameleon-7b/resolve/main/bee.jpg")
# Initialize processor and model
processor = AutoProcessor.from_pretrained("HuggingFaceTB/SmolVLM-Instruct")
model = AutoModelForVision2Seq.from_pretrained(
"HuggingFaceTB/SmolVLM-Instruct",
torch_dtype=torch.bfloat16,
_attn_implementation="flash_attention_2" if DEVICE == "cuda" else "eager",
).to(DEVICE)
# Create input messages
messages = [
{
"role": "user",
"content": [
{"type": "image"},
{"type": "image"},
{"type": "text", "text": "Can you describe the two images?"}
]
},
]
# Prepare inputs
prompt = processor.apply_chat_template(messages, add_generation_prompt=True)
inputs = processor(text=prompt, images=[image1, image2], return_tensors="pt")
inputs = inputs.to(DEVICE)
# Generate outputs
generated_ids = model.generate(**inputs, max_new_tokens=500)
generated_texts = processor.batch_decode(
generated_ids,
skip_special_tokens=True,
)
print(generated_texts[0])
"""
Assistant: The first image shows a green statue of the Statue of Liberty standing on a stone pedestal in front of a body of water.
The statue is holding a torch in its right hand and a tablet in its left hand. The water is calm and there are no boats or other objects visible.
The sky is clear and there are no clouds. The second image shows a bee on a pink flower.
The bee is black and yellow and is collecting pollen from the flower. The flower is surrounded by green leaves.
"""
```
### Model optimizations
**Precision**: For better performance, load and run the model in half-precision (`torch.float16` or `torch.bfloat16`) if your hardware supports it.
```python
from transformers import AutoModelForVision2Seq
import torch
model = AutoModelForVision2Seq.from_pretrained(
"HuggingFaceTB/SmolVLM-Instruct",
torch_dtype=torch.bfloat16
).to("cuda")
```
You can also load SmolVLM with 4/8-bit quantization using bitsandbytes, torchao or Quanto. Refer to [this page](https://huggingface.co/docs/transformers/en/main_classes/quantization) for other options.
```python
from transformers import AutoModelForVision2Seq, BitsAndBytesConfig
import torch
quantization_config = BitsAndBytesConfig(load_in_8bit=True)
model = AutoModelForVision2Seq.from_pretrained(
"HuggingFaceTB/SmolVLM-Instruct",
quantization_config=quantization_config,
)
```
**Vision Encoder Efficiency**: Adjust the image resolution by setting `size={"longest_edge": N*384}` when initializing the processor, where N is your desired value. The default `N=4` works well, which results in input images of
size 1536×1536. For documents, `N=5` might be beneficial. Decreasing N can save GPU memory and is appropriate for lower-resolution images. This is also useful if you want to fine-tune on videos.
## Misuse and Out-of-scope Use
SmolVLM is not intended for high-stakes scenarios or critical decision-making processes that affect an individual's well-being or livelihood. The model may produce content that appears factual but may not be accurate. Misuse includes, but is not limited to:
- Prohibited Uses:
- Evaluating or scoring individuals (e.g., in employment, education, credit)
- Critical automated decision-making
- Generating unreliable factual content
- Malicious Activities:
- Spam generation
- Disinformation campaigns
- Harassment or abuse
- Unauthorized surveillance
### License
SmolVLM is built upon [the shape-optimized SigLIP](https://huggingface.co/google/siglip-so400m-patch14-384) as image encoder and [SmolLM2](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct) for text decoder part.
We release the SmolVLM checkpoints under the Apache 2.0 license.
## Training Details
### Training Data
The training data comes from [The Cauldron](https://huggingface.co/datasets/HuggingFaceM4/the_cauldron) and [Docmatix](https://huggingface.co/datasets/HuggingFaceM4/Docmatix) datasets, with emphasis on document understanding (25%) and image captioning (18%), while maintaining balanced coverage across other crucial capabilities like visual reasoning, chart comprehension, and general instruction following.
<img src="https://huggingface.co/HuggingFaceTB/SmolVLM-Instruct/resolve/main/mixture_the_cauldron.png" alt="Example Image" style="width:90%;" />
## Evaluation
| Model | MMMU (val) | MathVista (testmini) | MMStar (val) | DocVQA (test) | TextVQA (val) | Min GPU RAM required (GB) |
|-------------------|------------|----------------------|--------------|---------------|---------------|---------------------------|
| SmolVLM | 38.8 | 44.6 | 42.1 | 81.6 | 72.7 | 5.02 |
| Qwen-VL 2B | 41.1 | 47.8 | 47.5 | 90.1 | 79.7 | 13.70 |
| InternVL2 2B | 34.3 | 46.3 | 49.8 | 86.9 | 73.4 | 10.52 |
| PaliGemma 3B 448px| 34.9 | 28.7 | 48.3 | 32.2 | 56.0 | 6.72 |
| moondream2 | 32.4 | 24.3 | 40.3 | 70.5 | 65.2 | 3.87 |
| MiniCPM-V-2 | 38.2 | 39.8 | 39.1 | 71.9 | 74.1 | 7.88 |
| MM1.5 1B | 35.8 | 37.2 | 0.0 | 81.0 | 72.5 | NaN |
# Citation information
You can cite us in the following way:
```bibtex
@article{marafioti2025smolvlm,
title={SmolVLM: Redefining small and efficient multimodal models},
author={Andrés Marafioti and Orr Zohar and Miquel Farré and Merve Noyan and Elie Bakouch and Pedro Cuenca and Cyril Zakka and Loubna Ben Allal and Anton Lozhkov and Nouamane Tazi and Vaibhav Srivastav and Joshua Lochner and Hugo Larcher and Mathieu Morlon and Lewis Tunstall and Leandro von Werra and Thomas Wolf},
journal={arXiv preprint arXiv:2504.05299},
year={2025}
}
```
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/gpt2-GGUF | Mungert | 2025-06-15T19:36:29Z | 2,302 | 2 | null | [
"gguf",
"exbert",
"en",
"license:mit",
"endpoints_compatible",
"region:us",
"imatrix"
] | null | 2025-06-08T00:23:46Z | ---
language: en
tags:
- exbert
license: mit
---
# <span style="color: #7FFF7F;">gpt2 GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`5787b5da`](https://github.com/ggerganov/llama.cpp/commit/5787b5da57e54dba760c2deeac1edf892e8fc450).
## <span style="color: #7FFF7F;"> Quantization beyond the IMatrix</span>
Testing a new quantization method using rules to bump important layers above what the standard imatrix would use.
I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
This does create larger model files but increases precision for a given model size.
### **Please provide feedback on how you find this method performs**
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
# GPT-2
Test the whole generation capabilities here: https://transformer.huggingface.co/doc/gpt2-large
Pretrained model on English language using a causal language modeling (CLM) objective. It was introduced in
[this paper](https://d4mucfpksywv.cloudfront.net/better-language-models/language_models_are_unsupervised_multitask_learners.pdf)
and first released at [this page](https://openai.com/blog/better-language-models/).
Disclaimer: The team releasing GPT-2 also wrote a
[model card](https://github.com/openai/gpt-2/blob/master/model_card.md) for their model. Content from this model card
has been written by the Hugging Face team to complete the information they provided and give specific examples of bias.
## Model description
GPT-2 is a transformers model pretrained on a very large corpus of English data in a self-supervised fashion. This
means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots
of publicly available data) with an automatic process to generate inputs and labels from those texts. More precisely,
it was trained to guess the next word in sentences.
More precisely, inputs are sequences of continuous text of a certain length and the targets are the same sequence,
shifted one token (word or piece of word) to the right. The model uses internally a mask-mechanism to make sure the
predictions for the token `i` only uses the inputs from `1` to `i` but not the future tokens.
This way, the model learns an inner representation of the English language that can then be used to extract features
useful for downstream tasks. The model is best at what it was pretrained for however, which is generating texts from a
prompt.
This is the **smallest** version of GPT-2, with 124M parameters.
**Related Models:** [GPT-Large](https://huggingface.co/gpt2-large), [GPT-Medium](https://huggingface.co/gpt2-medium) and [GPT-XL](https://huggingface.co/gpt2-xl)
## Intended uses & limitations
You can use the raw model for text generation or fine-tune it to a downstream task. See the
[model hub](https://huggingface.co/models?filter=gpt2) to look for fine-tuned versions on a task that interests you.
### How to use
You can use this model directly with a pipeline for text generation. Since the generation relies on some randomness, we
set a seed for reproducibility:
```python
>>> from transformers import pipeline, set_seed
>>> generator = pipeline('text-generation', model='gpt2')
>>> set_seed(42)
>>> generator("Hello, I'm a language model,", max_length=30, num_return_sequences=5)
[{'generated_text': "Hello, I'm a language model, a language for thinking, a language for expressing thoughts."},
{'generated_text': "Hello, I'm a language model, a compiler, a compiler library, I just want to know how I build this kind of stuff. I don"},
{'generated_text': "Hello, I'm a language model, and also have more than a few of your own, but I understand that they're going to need some help"},
{'generated_text': "Hello, I'm a language model, a system model. I want to know my language so that it might be more interesting, more user-friendly"},
{'generated_text': 'Hello, I\'m a language model, not a language model"\n\nThe concept of "no-tricks" comes in handy later with new'}]
```
Here is how to use this model to get the features of a given text in PyTorch:
```python
from transformers import GPT2Tokenizer, GPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
model = GPT2Model.from_pretrained('gpt2')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import GPT2Tokenizer, TFGPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
model = TFGPT2Model.from_pretrained('gpt2')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
### Limitations and bias
The training data used for this model has not been released as a dataset one can browse. We know it contains a lot of
unfiltered content from the internet, which is far from neutral. As the openAI team themselves point out in their
[model card](https://github.com/openai/gpt-2/blob/master/model_card.md#out-of-scope-use-cases):
> Because large-scale language models like GPT-2 do not distinguish fact from fiction, we don’t support use-cases
> that require the generated text to be true.
>
> Additionally, language models like GPT-2 reflect the biases inherent to the systems they were trained on, so we do
> not recommend that they be deployed into systems that interact with humans > unless the deployers first carry out a
> study of biases relevant to the intended use-case. We found no statistically significant difference in gender, race,
> and religious bias probes between 774M and 1.5B, implying all versions of GPT-2 should be approached with similar
> levels of caution around use cases that are sensitive to biases around human attributes.
Here's an example of how the model can have biased predictions:
```python
>>> from transformers import pipeline, set_seed
>>> generator = pipeline('text-generation', model='gpt2')
>>> set_seed(42)
>>> generator("The White man worked as a", max_length=10, num_return_sequences=5)
[{'generated_text': 'The White man worked as a mannequin for'},
{'generated_text': 'The White man worked as a maniser of the'},
{'generated_text': 'The White man worked as a bus conductor by day'},
{'generated_text': 'The White man worked as a plumber at the'},
{'generated_text': 'The White man worked as a journalist. He had'}]
>>> set_seed(42)
>>> generator("The Black man worked as a", max_length=10, num_return_sequences=5)
[{'generated_text': 'The Black man worked as a man at a restaurant'},
{'generated_text': 'The Black man worked as a car salesman in a'},
{'generated_text': 'The Black man worked as a police sergeant at the'},
{'generated_text': 'The Black man worked as a man-eating monster'},
{'generated_text': 'The Black man worked as a slave, and was'}]
```
This bias will also affect all fine-tuned versions of this model.
## Training data
The OpenAI team wanted to train this model on a corpus as large as possible. To build it, they scraped all the web
pages from outbound links on Reddit which received at least 3 karma. Note that all Wikipedia pages were removed from
this dataset, so the model was not trained on any part of Wikipedia. The resulting dataset (called WebText) weights
40GB of texts but has not been publicly released. You can find a list of the top 1,000 domains present in WebText
[here](https://github.com/openai/gpt-2/blob/master/domains.txt).
## Training procedure
### Preprocessing
The texts are tokenized using a byte-level version of Byte Pair Encoding (BPE) (for unicode characters) and a
vocabulary size of 50,257. The inputs are sequences of 1024 consecutive tokens.
The larger model was trained on 256 cloud TPU v3 cores. The training duration was not disclosed, nor were the exact
details of training.
## Evaluation results
The model achieves the following results without any fine-tuning (zero-shot):
| Dataset | LAMBADA | LAMBADA | CBT-CN | CBT-NE | WikiText2 | PTB | enwiki8 | text8 | WikiText103 | 1BW |
|:--------:|:-------:|:-------:|:------:|:------:|:---------:|:------:|:-------:|:------:|:-----------:|:-----:|
| (metric) | (PPL) | (ACC) | (ACC) | (ACC) | (PPL) | (PPL) | (BPB) | (BPC) | (PPL) | (PPL) |
| | 35.13 | 45.99 | 87.65 | 83.4 | 29.41 | 65.85 | 1.16 | 1,17 | 37.50 | 75.20 |
### BibTeX entry and citation info
```bibtex
@article{radford2019language,
title={Language Models are Unsupervised Multitask Learners},
author={Radford, Alec and Wu, Jeff and Child, Rewon and Luan, David and Amodei, Dario and Sutskever, Ilya},
year={2019}
}
```
<a href="https://huggingface.co/exbert/?model=gpt2">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/Cosmos-Reason1-7B-GGUF | Mungert | 2025-06-15T19:36:14Z | 3,809 | 2 | transformers | [
"transformers",
"gguf",
"nvidia",
"cosmos",
"en",
"dataset:nvidia/Cosmos-Reason1-SFT-Dataset",
"dataset:nvidia/Cosmos-Reason1-RL-Dataset",
"dataset:nvidia/Cosmos-Reason1-Benchmark",
"arxiv:2503.15558",
"base_model:Qwen/Qwen2.5-VL-7B-Instruct",
"base_model:quantized:Qwen/Qwen2.5-VL-7B-Instruct",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | null | 2025-05-24T03:54:56Z | ---
license: other
license_name: nvidia-open-model-license
license_link: >-
https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license
datasets:
- nvidia/Cosmos-Reason1-SFT-Dataset
- nvidia/Cosmos-Reason1-RL-Dataset
- nvidia/Cosmos-Reason1-Benchmark
library_name: transformers
language:
- en
base_model:
- Qwen/Qwen2.5-VL-7B-Instruct
tags:
- nvidia
- cosmos
---
# <span style="color: #7FFF7F;">Cosmos-Reason1-7B GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`92ecdcc0`](https://github.com/ggerganov/llama.cpp/commit/92ecdcc06a4c405a415bcaa0cb772bc560aa23b1).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `Cosmos-Reason1-7B-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `Cosmos-Reason1-7B-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `Cosmos-Reason1-7B-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `Cosmos-Reason1-7B-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `Cosmos-Reason1-7B-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `Cosmos-Reason1-7B-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `Cosmos-Reason1-7B-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `Cosmos-Reason1-7B-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `Cosmos-Reason1-7B-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `Cosmos-Reason1-7B-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `Cosmos-Reason1-7B-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# **Cosmos-Reason1: Physical AI Common Sense and Embodied Reasoning Models**
[**Cosmos**](https://huggingface.co/collections/nvidia/cosmos-reason1-67c9e926206426008f1da1b7) | [**Code**](https://github.com/nvidia-cosmos/cosmos-reason1) | [**Paper**](https://arxiv.org/abs/2503.15558) | [**Paper Website**](https://research.nvidia.com/labs/dir/cosmos-reason1)
# Model Overview
## Description:
**Cosmos-Reason1 Models**: Physical AI models understand physical common sense and generate appropriate embodied decisions in natural language through long chain-of-thought reasoning processes.
The Cosmos-Reason1 models are post-trained with physical common sense and embodied reasoning data with supervised fine-tuning and reinforcement learning. These are Physical AI models that can understand space, time, and fundamental physics, and can serve as planning models to reason about the next steps of an embodied agent.
The models are ready for commercial use.
**Model Developer**: NVIDIA
## Model Versions
The Cosmos-Reason1 includes the following model:
- [Cosmos-Reason1-7B](https://huggingface.co/nvidia/Cosmos-Reason1-7B): Given a text prompt and an input video, think and generate the answer with respect to the input text prompt and video.
### License:
This model is released under the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license). For a custom license, please contact [[email protected]](mailto:[email protected]).
Under the NVIDIA Open Model License, NVIDIA confirms:
* Models are commercially usable.
* You are free to create and distribute Derivative Models.
* NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.
**Important Note**: If You bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or associated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism (collectively “Guardrail”) contained in the Model without a substantially similar Guardrail appropriate for your use case, your rights under this Agreement [NVIDIA Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license) will automatically terminate.
### Deployment Geography:
Global
### Use Case:
Physical AI: Space, time, fundamental physics understanding and embodied reasoning, encompassing robotics, and autonomous vehicles (AV).
### Release Date:
* Github: [05/17/2025](https://github.com/nvidia-cosmos/cosmos-reason1)
* Huggingface: [05/17/2025](https://huggingface.co/collections/nvidia/cosmos-reason1-67c9e926206426008f1da1b7)
## Model Architecture:
Architecture Type: A Multi-modal LLM consists of a Vision Transformer (ViT) for vision encoder and a Dense Transformer model for LLM.
Network Architecture: Qwen2.5-VL-7B-Instruct.
Cosmos-Reason-7B is post-trained based on [Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct) and follows the same model architecture.
## Input
**Input Type(s)**: Text+Video/Image
**Input Format(s)**:
* Text: String
* Video: mp4
* Image: jpg
**Input Parameters**:
* Text: One-dimensional (1D)
* Video: Three-dimensional (3D)
* Image: Two-dimensional (2D)
**Other Properties Related to Input**:
* Use `FPS=4` for input video to match the training setup.
* Append `Answer the question in the following format: <think>\nyour reasoning\n</think>\n\n<answer>\nyour answer\n</answer>.` in the system prompt to encourage long chain-of-thought reasoning response.
## Output
**Output Type(s)**: Text
**Output Format**: String
**Output Parameters**: Text: One-dimensional (1D)
**Other Properties Related to Output**:
* Recommend using 4096 or more output max tokens to avoid truncation of long chain-of-thought response.
* Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA’s hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions. <br>
## Software Integration
**Runtime Engine(s):**
* [vLLM](https://github.com/vllm-project/vllm)
**Supported Hardware Microarchitecture Compatibility:**
* NVIDIA Blackwell
* NVIDIA Hopper
**Note**: We have only tested doing inference with BF16 precision.
**Operating System(s):**
* Linux (We have not tested on other operating systems.)
# Usage
See [Cosmos-Reason1](https://github.com/nvidia-cosmos/cosmos-reason1) for details.
* Post Training: [Cosmos-Reason1](https://github.com/nvidia-cosmos/cosmos-reason1) provides examples of supervised fine-tuning and reinforcement learning on embodied reasoning datasets.
# Evaluation
Please see our [technical paper](https://arxiv.org/pdf/2503.15558) for detailed evaluations on physical common sense and embodied reasoning. Part of the evaluation datasets are released under [Cosmos-Reason1-Benchmark](https://huggingface.co/datasets/nvidia/Cosmos-Reason1-Benchmark). The embodied reasoning datasets and benchmarks focus on the following areas: robotics (RoboVQA, BridgeDataV2, Agibot, RobFail), ego-centric human demonstration (HoloAssist), and Autonomous Vehicle (AV) driving video data. The AV dataset is collected and annotated by NVIDIA.
All datasets go through the data annotation process described in the technical paper to prepare training and evaluation data and annotations.
**Data Collection Method**:
* RoboVQA: Hybrid: Automatic/Sensors
* BridgeDataV2: Automatic/Sensors
* AgiBot: Automatic/Sensors
* RoboFail: Automatic/Sensors
* HoloAssist: Human
* AV: Automatic/Sensors
**Labeling Method**:
* RoboVQA: Hybrid: Human,Automated
* BridgeDataV2: Hybrid: Human,Automated
* AgiBot: Hybrid: Human,Automated
* RoboFail: Hybrid: Human,Automated
* HoloAssist: Hybrid: Human,Automated
* AV: Hybrid: Human,Automated
**Metrics**:
We report the model accuracy on the embodied reasoning benchmark introduced in [Cosmos-Reason1](https://arxiv.org/abs/2503.15558). The results differ from those presented in Table 9 due to additional training aimed at supporting a broader range of Physical AI tasks beyond the benchmark.
| | [RoboVQA](https://robovqa.github.io/) | AV | [BridgeDataV2](https://rail-berkeley.github.io/bridgedata/)| [Agibot](https://github.com/OpenDriveLab/AgiBot-World)| [HoloAssist](https://holoassist.github.io/) | [RoboFail](https://robot-reflect.github.io/) | Average |
|--------------------|---------------------------------------------|----------|------------------------------------------------------|------------------------------------------------|------------------------------------------------|------------------------------------------------|------------------------------------------------|
| **Accuracy** | 87.3 | 70.8 | 63.7 | 48.9 | 62.7 | 57.2 | 65.1 |
## Dataset Format
Modality: Video (mp4) and Text
## Dataset Quantification
We release the embodied reasoning data and benchmarks. Each data sample is a pair of video and text. The text annotations include understanding and reasoning annotations described in the Cosmos-Reason1 paper. Each video may have multiple text annotations. The quantity of the video and text pairs is described in the table below.
**The AV data is currently unavailable and will be uploaded soon!**
| | [RoboVQA](https://robovqa.github.io/) | AV | [BridgeDataV2](https://rail-berkeley.github.io/bridgedata/)| [Agibot](https://github.com/OpenDriveLab/AgiBot-World)| [HoloAssist](https://holoassist.github.io/) | [RoboFail](https://robot-reflect.github.io/) | Total Storage Size |
|--------------------|---------------------------------------------|----------|------------------------------------------------------|------------------------------------------------|------------------------------------------------|------------------------------------------------|--------------------|
| **SFT Data** | 1.14m | 24.7k | 258k | 38.9k | 273k | N/A | **300.6GB** |
| **RL Data** | 252 | 200 | 240 | 200 | 200 | N/A | **2.6GB** |
| **Benchmark Data** | 110 | 100 | 100 | 100 | 100 | 100 | **1.5GB** |
We release text annotations for all embodied reasoning datasets and videos for RoboVQA and AV datasets. For other datasets, users may download the source videos from the original data source and find corresponding video sources via the video names. The held-out RoboFail benchmark is released for measuring the generalization capability.
## Inference:
**Acceleration Engine:** PyTorch, flash attention <br>
**Test Hardware:** H100, A100, GB200 <br>
* Minimum 2 GPU cards, multi nodes require Infiniband / ROCE connection <br>
## Ethical Considerations
NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
Users are responsible for model inputs and outputs. Users are responsible for ensuring safe integration of this model, including implementing guardrails as well as other safety mechanisms, prior to deployment.
For more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety & Security, and Privacy below.
Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).
### Plus Plus (++) Promise
We value you, the datasets, the diversity they represent, and what we have been entrusted with. This model and its associated data have been:
* Verified to comply with current applicable disclosure laws, regulations, and industry standards.
* Verified to comply with applicable privacy labeling requirements.
* Annotated to describe the collector/source (NVIDIA or a third-party).
* Characterized for technical limitations.
* Reviewed to ensure proper disclosure is accessible to, maintained for, and in compliance with NVIDIA data subjects and their requests.
* Reviewed before release.
* Tagged for known restrictions and potential safety implications.
### Bias
| Field | Response |
| :--------------------------------------------------------------------------------------------------------------------------------------------------------------- | :------- |
| Participation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None |
| Measures taken to mitigate against unwanted bias: | The training video sources contain multiple physical embodiments and environments including human, car, single arm robot, bimanual robot in indoor and outdoor environments. By training on numerous and various physical interactions and curated datasets, we strive to provide a model that does not possess biases towards certain embodiments or environments. |
### Explainability
| Field | Response |
| :-------------------------------------------------------- | :------------------------------------------------------------------------------------------------------------------- |
| Intended Application & Domain: | Physical AI Reasoning |
| Model Type: | Transformer |
| Intended Users: | Physical AI developers |
| Output: | Text |
| Describe how the model works: | Generates text answers based on input text prompt and video |
| Technical Limitations: | The model may not follow the video or text input accurately in challenging cases, where the input video shows complex scene composition and temporal dynamics. Examples of challenging scenes include: fast camera movements, overlapping human-object interactions, low lighting with high motion blur, and multiple people performing different actions simultaneously. |
| Verified to have met prescribed NVIDIA quality standards: | Yes |
| Performance Metrics: | Quantitative and Qualitative Evaluation. Cosmos-Reason1 proposes the embodied reasoning benchmark and physical common sense benchmark to evaluate accuracy with visual question answering. |
| Potential Known Risks: | The model's output can generate all forms of texts, including what may be considered toxic, offensive, or indecent. |
| Licensing: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license) |
### Privacy
| Field | Response |
| :------------------------------------------------------------------ | :------------- |
| Generatable or reverse engineerable personal information? | None Known |
| Protected class data used to create this model? | None Known |
| Was consent obtained for any personal data used? | None Known |
| How often is dataset reviewed? | Before Release |
| Is there provenance for all datasets used in training? | Yes |
| Does data labeling (annotation, metadata) comply with privacy laws? | Yes |
| Applicable Privacy Policy | [NVIDIA Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy) |
### Safety
| Field | Response |
| :---------------------------------------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Model Application(s): | Physical AI common sense understanding and embodied reasoning |
| Describe the life critical impact (if present). | None Known |
| Use Case Restrictions: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license) |
| Model and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face, and may become available on cloud providers' model catalog. | |
Mungert/llama-joycaption-beta-one-hf-llava-GGUF | Mungert | 2025-06-15T19:36:02Z | 3,922 | 3 | transformers | [
"transformers",
"gguf",
"captioning",
"image-text-to-text",
"base_model:google/siglip2-so400m-patch14-384",
"base_model:quantized:google/siglip2-so400m-patch14-384",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | image-text-to-text | 2025-06-08T03:11:55Z | ---
base_model:
- meta-llama/Llama-3.1-8B-Instruct
- google/siglip2-so400m-patch14-384
tags:
- captioning
pipeline_tag: image-text-to-text
library_name: transformers
---
# <span style="color: #7FFF7F;">llama-joycaption-beta-one-hf-llava GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`5787b5da`](https://github.com/ggerganov/llama.cpp/commit/5787b5da57e54dba760c2deeac1edf892e8fc450).
## <span style="color: #7FFF7F;"> Quantization beyond the IMatrix</span>
Testing a new quantization method using rules to bump important layers above what the standard imatrix would use.
I have found that the standard IMatrix does not perform very well at low bit quantiztion and for MOE models. So I am using llama.cpp --tensor-type to bump up selected layers. See [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
This does create larger model files but increases precision for a given model size.
### **Please provide feedback on how you find this method performs**
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Hybrid Precision Models (e.g., `bf16_q8_0`, `f16_q4_K`) – Best of Both Worlds**
These formats selectively **quantize non-essential layers** while keeping **key layers in full precision** (e.g., attention and output layers).
- Named like `bf16_q8_0` (meaning **full-precision BF16 core layers + quantized Q8_0 other layers**).
- Strike a **balance between memory efficiency and accuracy**, improving over fully quantized models without requiring the full memory of BF16/F16.
📌 **Use Hybrid Models if:**
✔ You need **better accuracy than quant-only models** but can’t afford full BF16/F16 everywhere.
✔ Your device supports **mixed-precision inference**.
✔ You want to **optimize trade-offs** for production-grade models on constrained hardware.
📌 **Avoid Hybrid Models if:**
❌ Your target device doesn’t support **mixed or full-precision acceleration**.
❌ You are operating under **ultra-strict memory limits** (in which case use fully quantized formats).
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **very high memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **very high memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
### **Ultra Low-Bit Quantization (IQ1_S IQ1_M IQ2_S IQ2_M IQ2_XS IQ2_XSS)**
- *Ultra-low-bit quantization (1 2-bit) with **extreme memory efficiency**.
- **Use case**: Best for cases were you have to fit the model into very constrained memory
- **Trade-off**: Very Low Accuracy. May not function as expected. Please test fully before using.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------------------|------------------|------------------|----------------------------------|--------------------------------------------------------------|
| **BF16** | Very High | High | BF16-supported GPU/CPU | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported GPU/CPU | Inference when BF16 isn’t available |
| **Q4_K** | Medium-Low | Low | CPU or Low-VRAM devices | Memory-constrained inference |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy with quantization |
| **Q8_0** | High | Moderate | GPU/CPU with moderate VRAM | Highest accuracy among quantized models |
| **IQ3_XS** | Low | Very Low | Ultra-low-memory devices | Max memory efficiency, low accuracy |
| **IQ3_S** | Low | Very Low | Low-memory devices | Slightly more usable than IQ3_XS |
| **IQ3_M** | Low-Medium | Low | Low-memory devices | Better accuracy than IQ3_S |
| **Q4_0** | Low | Low | ARM-based/embedded devices | Llama.cpp automatically optimizes for ARM inference |
| **Ultra Low-Bit (IQ1/2_*)** | Very Low | Extremely Low | Tiny edge/embedded devices | Fit models in extremely tight memory; low accuracy |
| **Hybrid (e.g., `bf16_q8_0`)** | Medium–High | Medium | Mixed-precision capable hardware | Balanced performance and memory, near-FP accuracy in critical layers |
---
# Model Card for Llama JoyCaption Beta One
[Github](https://github.com/fpgaminer/joycaption)
JoyCaption is an image captioning Visual Language Model (VLM) built from the ground up as a free, open, and uncensored model for the community to use in training Diffusion models.
Key Features:
- **Free and Open**: Always released for free, open weights, no restrictions, and just like [bigASP](https://www.reddit.com/r/StableDiffusion/comments/1dbasvx/the_gory_details_of_finetuning_sdxl_for_30m/), will come with training scripts and lots of juicy details on how it gets built.
- **Uncensored**: Equal coverage of SFW and NSFW concepts. No "cylindrical shaped object with a white substance coming out on it" here.
- **Diversity**: All are welcome here. Do you like digital art? Photoreal? Anime? Furry? JoyCaption is for everyone. Pains are being taken to ensure broad coverage of image styles, content, ethnicity, gender, orientation, etc.
- **Minimal Filtering**: JoyCaption is trained on large swathes of images so that it can understand almost all aspects of our world. almost. Illegal content will never be tolerated in JoyCaption's training.
## Motivation
Automated descriptive captions enable the training and finetuning of diffusion models on a wider range of images, since trainers are no longer required to either find images with already associated text or write the descriptions themselves. They also improve the quality of generations produced by Text-to-Image models trained on them (ref: DALL-E 3 paper). But to-date, the community has been stuck with ChatGPT, which is expensive and heavily censored; or alternative models, like CogVLM, which are weaker than ChatGPT and have abysmal performance outside of the SFW domain.
I'm building JoyCaption to help fill this gap by performing near or on-par with GPT4o in captioning images, while being free, unrestricted, and open.
## How to Get Started with the Model
Please see the [Github](https://github.com/fpgaminer/joycaption) for more details.
Example usage:
```
import torch
from PIL import Image
from transformers import AutoProcessor, LlavaForConditionalGeneration
IMAGE_PATH = "image.jpg"
PROMPT = "Write a long descriptive caption for this image in a formal tone."
MODEL_NAME = "fancyfeast/llama-joycaption-beta-one-hf-llava"
# Load JoyCaption
# bfloat16 is the native dtype of the LLM used in JoyCaption (Llama 3.1)
# device_map=0 loads the model into the first GPU
processor = AutoProcessor.from_pretrained(MODEL_NAME)
llava_model = LlavaForConditionalGeneration.from_pretrained(MODEL_NAME, torch_dtype="bfloat16", device_map=0)
llava_model.eval()
with torch.no_grad():
# Load image
image = Image.open(IMAGE_PATH)
# Build the conversation
convo = [
{
"role": "system",
"content": "You are a helpful image captioner.",
},
{
"role": "user",
"content": PROMPT,
},
]
# Format the conversation
# WARNING: HF's handling of chat's on Llava models is very fragile. This specific combination of processor.apply_chat_template(), and processor() works
# but if using other combinations always inspect the final input_ids to ensure they are correct. Often times you will end up with multiple <bos> tokens
# if not careful, which can make the model perform poorly.
convo_string = processor.apply_chat_template(convo, tokenize = False, add_generation_prompt = True)
assert isinstance(convo_string, str)
# Process the inputs
inputs = processor(text=[convo_string], images=[image], return_tensors="pt").to('cuda')
inputs['pixel_values'] = inputs['pixel_values'].to(torch.bfloat16)
# Generate the captions
generate_ids = llava_model.generate(
**inputs,
max_new_tokens=512,
do_sample=True,
suppress_tokens=None,
use_cache=True,
temperature=0.6,
top_k=None,
top_p=0.9,
)[0]
# Trim off the prompt
generate_ids = generate_ids[inputs['input_ids'].shape[1]:]
# Decode the caption
caption = processor.tokenizer.decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)
caption = caption.strip()
print(caption)
```
## vLLM
vLLM provides the highest performance inference for JoyCaption, and an OpenAI compatible API so JoyCaption can be used like any other VLMs. Example usage:
```
vllm serve fancyfeast/llama-joycaption-beta-one-hf-llava --max-model-len 4096 --enable-prefix-caching
```
VLMs are a bit finicky on vLLM, and vLLM is memory hungry, so you may have to adjust settings for your particular environment, such as forcing eager mode, adjusting max-model-len, adjusting gpu_memory_utilization, etc.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
Mungert/medgemma-27b-text-it-GGUF | Mungert | 2025-06-15T19:35:58Z | 2,840 | 6 | transformers | [
"transformers",
"gguf",
"medical",
"clinical-reasoning",
"thinking",
"text-generation",
"arxiv:2501.19393",
"arxiv:2303.15343",
"arxiv:2009.13081",
"arxiv:2102.09542",
"arxiv:2411.15640",
"arxiv:2404.05590",
"arxiv:2501.18362",
"base_model:google/gemma-3-27b-pt",
"base_model:quantized:google/gemma-3-27b-pt",
"license:other",
"endpoints_compatible",
"region:us",
"imatrix",
"conversational"
] | text-generation | 2025-05-29T06:32:42Z | ---
license: other
license_name: health-ai-developer-foundations
license_link: https://developers.google.com/health-ai-developer-foundations/terms
library_name: transformers
pipeline_tag: text-generation
extra_gated_heading: Access MedGemma on Hugging Face
extra_gated_prompt: >-
To access MedGemma on Hugging Face, you're required to review and agree to
[Health AI Developer Foundation's terms of
use](https://developers.google.com/health-ai-developer-foundations/terms). To
do this, please ensure you're logged in to Hugging Face and click below.
Requests are processed immediately.
extra_gated_button_content: Acknowledge license
base_model: google/gemma-3-27b-pt
tags:
- medical
- clinical-reasoning
- thinking
---
# <span style="color: #7FFF7F;">medgemma-27b-text-it GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`f5cd27b7`](https://github.com/ggerganov/llama.cpp/commit/f5cd27b71da3ac375a04a41643d14fc779a8057b).
## <span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>
Our latest quantization method introduces **precision-adaptive quantization** for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on **Llama-3-8B**. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.
### **Benchmark Context**
All tests conducted on **Llama-3-8B-Instruct** using:
- Standard perplexity evaluation pipeline
- 2048-token context window
- Same prompt set across all quantizations
### **Method**
- **Dynamic Precision Allocation**:
- First/Last 25% of layers → IQ4_XS (selected layers)
- Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
- **Critical Component Protection**:
- Embeddings/output layers use Q5_K
- Reduces error propagation by 38% vs standard 1-2bit
### **Quantization Performance Comparison (Llama-3-8B)**
| Quantization | Standard PPL | DynamicGate PPL | Δ PPL | Std Size | DG Size | Δ Size | Std Speed | DG Speed |
|--------------|--------------|------------------|---------|----------|---------|--------|-----------|----------|
| IQ2_XXS | 11.30 | 9.84 | -12.9% | 2.5G | 2.6G | +0.1G | 234s | 246s |
| IQ2_XS | 11.72 | 11.63 | -0.8% | 2.7G | 2.8G | +0.1G | 242s | 246s |
| IQ2_S | 14.31 | 9.02 | -36.9% | 2.7G | 2.9G | +0.2G | 238s | 244s |
| IQ1_M | 27.46 | 15.41 | -43.9% | 2.2G | 2.5G | +0.3G | 206s | 212s |
| IQ1_S | 53.07 | 32.00 | -39.7% | 2.1G | 2.4G | +0.3G | 184s | 209s |
**Key**:
- PPL = Perplexity (lower is better)
- Δ PPL = Percentage change from standard to DynamicGate
- Speed = Inference time (CPU avx2, 2048 token context)
- Size differences reflect mixed quantization overhead
**Key Improvements:**
- 🔥 **IQ1_M** shows massive 43.9% perplexity reduction (27.46 → 15.41)
- 🚀 **IQ2_S** cuts perplexity by 36.9% while adding only 0.2GB
- ⚡ **IQ1_S** maintains 39.7% better accuracy despite 1-bit quantization
**Tradeoffs:**
- All variants have modest size increases (0.1-0.3GB)
- Inference speeds remain comparable (<5% difference)
### **When to Use These Models**
📌 **Fitting models into GPU VRAM**
✔ **Memory-constrained deployments**
✔ **Cpu and Edge Devices** where 1-2bit errors can be tolerated
✔ **Research** into ultra-low-bit quantization
## **Choosing the Right Model Format**
Selecting the correct model format depends on your **hardware capabilities** and **memory constraints**.
### **BF16 (Brain Float 16) – Use if BF16 acceleration is available**
- A 16-bit floating-point format designed for **faster computation** while retaining good precision.
- Provides **similar dynamic range** as FP32 but with **lower memory usage**.
- Recommended if your hardware supports **BF16 acceleration** (check your device's specs).
- Ideal for **high-performance inference** with **reduced memory footprint** compared to FP32.
📌 **Use BF16 if:**
✔ Your hardware has native **BF16 support** (e.g., newer GPUs, TPUs).
✔ You want **higher precision** while saving memory.
✔ You plan to **requantize** the model into another format.
📌 **Avoid BF16 if:**
❌ Your hardware does **not** support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.
---
### **F16 (Float 16) – More widely supported than BF16**
- A 16-bit floating-point **high precision** but with less of range of values than BF16.
- Works on most devices with **FP16 acceleration support** (including many GPUs and some CPUs).
- Slightly lower numerical precision than BF16 but generally sufficient for inference.
📌 **Use F16 if:**
✔ Your hardware supports **FP16** but **not BF16**.
✔ You need a **balance between speed, memory usage, and accuracy**.
✔ You are running on a **GPU** or another device optimized for FP16 computations.
📌 **Avoid F16 if:**
❌ Your device lacks **native FP16 support** (it may run slower than expected).
❌ You have memory limitations.
---
### **Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference**
Quantization reduces model size and memory usage while maintaining as much accuracy as possible.
- **Lower-bit models (Q4_K)** → **Best for minimal memory usage**, may have lower precision.
- **Higher-bit models (Q6_K, Q8_0)** → **Better accuracy**, requires more memory.
📌 **Use Quantized Models if:**
✔ You are running inference on a **CPU** and need an optimized model.
✔ Your device has **low VRAM** and cannot load full-precision models.
✔ You want to reduce **memory footprint** while keeping reasonable accuracy.
📌 **Avoid Quantized Models if:**
❌ You need **maximum accuracy** (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).
---
### **Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)**
These models are optimized for **extreme memory efficiency**, making them ideal for **low-power devices** or **large-scale deployments** where memory is a critical constraint.
- **IQ3_XS**: Ultra-low-bit quantization (3-bit) with **extreme memory efficiency**.
- **Use case**: Best for **ultra-low-memory devices** where even Q4_K is too large.
- **Trade-off**: Lower accuracy compared to higher-bit quantizations.
- **IQ3_S**: Small block size for **maximum memory efficiency**.
- **Use case**: Best for **low-memory devices** where **IQ3_XS** is too aggressive.
- **IQ3_M**: Medium block size for better accuracy than **IQ3_S**.
- **Use case**: Suitable for **low-memory devices** where **IQ3_S** is too limiting.
- **Q4_K**: 4-bit quantization with **block-wise optimization** for better accuracy.
- **Use case**: Best for **low-memory devices** where **Q6_K** is too large.
- **Q4_0**: Pure 4-bit quantization, optimized for **ARM devices**.
- **Use case**: Best for **ARM-based devices** or **low-memory environments**.
---
### **Summary Table: Model Format Selection**
| Model Format | Precision | Memory Usage | Device Requirements | Best Use Case |
|--------------|------------|---------------|----------------------|---------------|
| **BF16** | Highest | High | BF16-supported GPU/CPUs | High-speed inference with reduced memory |
| **F16** | High | High | FP16-supported devices | GPU inference when BF16 isn't available |
| **Q4_K** | Medium Low | Low | CPU or Low-VRAM devices | Best for memory-constrained environments |
| **Q6_K** | Medium | Moderate | CPU with more memory | Better accuracy while still being quantized |
| **Q8_0** | High | Moderate | CPU or GPU with enough VRAM | Best accuracy among quantized models |
| **IQ3_XS** | Very Low | Very Low | Ultra-low-memory devices | Extreme memory efficiency and low accuracy |
| **Q4_0** | Low | Low | ARM or low-memory devices | llama.cpp can optimize for ARM devices |
---
## **Included Files & Details**
### `medgemma-27b-text-it-bf16.gguf`
- Model weights preserved in **BF16**.
- Use this if you want to **requantize** the model into a different format.
- Best if your device supports **BF16 acceleration**.
### `medgemma-27b-text-it-f16.gguf`
- Model weights stored in **F16**.
- Use if your device supports **FP16**, especially if BF16 is not available.
### `medgemma-27b-text-it-bf16-q8_0.gguf`
- **Output & embeddings** remain in **BF16**.
- All other layers quantized to **Q8_0**.
- Use if your device supports **BF16** and you want a quantized version.
### `medgemma-27b-text-it-f16-q8_0.gguf`
- **Output & embeddings** remain in **F16**.
- All other layers quantized to **Q8_0**.
### `medgemma-27b-text-it-q4_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q4_K**.
- Good for **CPU inference** with limited memory.
### `medgemma-27b-text-it-q4_k_s.gguf`
- Smallest **Q4_K** variant, using less memory at the cost of accuracy.
- Best for **very low-memory setups**.
### `medgemma-27b-text-it-q6_k.gguf`
- **Output & embeddings** quantized to **Q8_0**.
- All other layers quantized to **Q6_K** .
### `medgemma-27b-text-it-q8_0.gguf`
- Fully **Q8** quantized model for better accuracy.
- Requires **more memory** but offers higher precision.
### `medgemma-27b-text-it-iq3_xs.gguf`
- **IQ3_XS** quantization, optimized for **extreme memory efficiency**.
- Best for **ultra-low-memory devices**.
### `medgemma-27b-text-it-iq3_m.gguf`
- **IQ3_M** quantization, offering a **medium block size** for better accuracy.
- Suitable for **low-memory devices**.
### `medgemma-27b-text-it-q4_0.gguf`
- Pure **Q4_0** quantization, optimized for **ARM devices**.
- Best for **low-memory environments**.
- Prefer IQ4_NL for better accuracy.
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
❤ **Please click "Like" if you find this useful!**
Help me test my **AI-Powered Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/dashboard/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4o-mini)
- `HugLLM` (Hugginface Open-source)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**)
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4o-mini** for:
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API
### 💡 **Example commands to you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
# MedGemma model card
**Model documentation:** [MedGemma](https://developers.google.com/health-ai-developer-foundations/medgemma)
**Resources:**
* Model on Google Cloud Model Garden: [MedGemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/medgemma)
* Model on Hugging Face: [MedGemma](https://huggingface.co/collections/google/medgemma-release-680aade845f90bec6a3f60c4)
* GitHub repository (supporting code, Colab notebooks, discussions, and
issues): [MedGemma](https://github.com/google-health/medgemma)
* Quick start notebook: [GitHub](https://github.com/google-health/medgemma/blob/main/notebooks/quick_start_with_hugging_face.ipynb)
* Fine-tuning notebook: [GitHub](https://github.com/google-health/medgemma/blob/main/notebooks/fine_tune_with_hugging_face.ipynb)
* [Patient Education Demo built using MedGemma](https://huggingface.co/spaces/google/rad_explain)
* Support: See [Contact](https://developers.google.com/health-ai-developer-foundations/medgemma/get-started.md#contact)
* License: The use of MedGemma is governed by the [Health AI Developer
Foundations terms of
use](https://developers.google.com/health-ai-developer-foundations/terms).
**Author:** Google
## Model information
This section describes the MedGemma model and how to use it.
### Description
MedGemma is a collection of [Gemma 3](https://ai.google.dev/gemma/docs/core)
variants that are trained for performance on medical text and image
comprehension. Developers can use MedGemma to accelerate building
healthcare-based AI applications. MedGemma currently comes in two variants: a 4B
multimodal version and a 27B text-only version.
MedGemma 27B has been trained exclusively on medical text and optimized for
inference-time computation. MedGemma 27B is only available as an
instruction-tuned model.
MedGemma variants have been evaluated on a range of clinically relevant
benchmarks to illustrate their baseline performance. These include both open
benchmark datasets and curated datasets. Developers can fine-tune MedGemma
variants for improved performance. Consult the Intended Use section below for
more details.
A full technical report will be available soon.
### How to use
Below are some example code snippets to help you quickly get started running the
model locally on GPU. If you want to use the model at scale, we recommend that
you create a production version using [Model
Garden](https://cloud.google.com/model-garden).
First, install the Transformers library. Gemma 3 is supported starting from
transformers 4.50.0.
```sh
$ pip install -U transformers
```
**Run model with the `pipeline` API**
```python
from transformers import pipeline
import torch
pipe = pipeline(
"text-generation",
model="google/medgemma-27b-text-it",
torch_dtype=torch.bfloat16,
device="cuda",
)
messages = [
{
"role": "system",
"content": "You are a helpful medical assistant."
},
{
"role": "user",
"content": "How do you differentiate bacterial from viral pneumonia?"
}
]
output = pipe(text=messages, max_new_tokens=200)
print(output[0]["generated_text"][-1]["content"])
```
**Run the model directly**
```python
# pip install accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
model_id = "google/medgemma-27b-text-it"
model = AutoModelForCausalLM.from_pretrained(
model_id,
torch_dtype=torch.bfloat16,
device_map="auto",
)
tokenizer = AutoTokenizer.from_pretrained(model_id)
messages = [
{
"role": "system",
"content": "You are a helpful medical assistant."
},
{
"role": "user",
"content": "How do you differentiate bacterial from viral pneumonia?"
}
]
inputs = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
tokenize=True,
return_dict=True,
return_tensors="pt",
).to(model.device)
input_len = inputs["input_ids"].shape[-1]
with torch.inference_mode():
generation = model.generate(**inputs, max_new_tokens=200, do_sample=False)
generation = generation[0][input_len:]
decoded = tokenizer.decode(generation, skip_special_tokens=True)
print(decoded)
```
### Examples
See the following Colab notebooks for examples of how to use MedGemma:
* To give the model a quick try, running it locally with weights from Hugging
Face, see [Quick start notebook in
Colab](https://colab.research.google.com/github/google-health/medgemma/blob/main/notebooks/quick_start_with_hugging_face.ipynb).
Note that you will need to use Colab Enterprise to run the 27B model without
quantization.
* For an example of fine-tuning the model, see the [Fine-tuning notebook in
Colab](https://colab.research.google.com/github/google-health/medgemma/blob/main/notebooks/fine_tune_with_hugging_face.ipynb).
### Model architecture overview
The MedGemma model is built based on [Gemma 3](https://ai.google.dev/gemma/) and
uses the same decoder-only transformer architecture as Gemma 3. To read more
about the architecture, consult the Gemma 3 [model
card](https://ai.google.dev/gemma/docs/core/model_card_3).
### Technical specifications
* **Model type**: Decoder-only Transformer architecture, see the [Gemma 3
technical
report](https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf)
* **Modalities**: **4B**: Text, vision; **27B**: Text only
* **Attention mechanism**: Utilizes grouped-query attention (GQA)
* **Context length**: Supports long context, at least 128K tokens
* **Key publication**: Coming soon
* **Model created**: May 20, 2025
* **Model version**: 1.0.0
### Citation
A technical report is coming soon. In the meantime, if you publish using this
model, please cite the Hugging Face model page:
```none
@misc{medgemma-hf,
author = {Google},
title = {MedGemma Hugging Face}
howpublished = {\url{https://huggingface.co/collections/google/medgemma-release-680aade845f90bec6a3f60c4}},
year = {2025},
note = {Accessed: [Insert Date Accessed, e.g., 2025-05-20]}
}
```
### Inputs and outputs
**Input**:
* Text string, such as a question or prompt
* Total input length of 128K tokens
**Output**:
* Generated text in response to the input, such as an answer to a question,
analysis of image content, or a summary of a document
* Total output length of 8192 tokens
### Performance and validation
MedGemma was evaluated across a range of different multimodal classification,
report generation, visual question answering, and text-based tasks.
### Key performance metrics
#### Text evaluations
MedGemma 4B and text-only MedGemma 27B were evaluated across a range of
text-only benchmarks for medical knowledge and reasoning.
The MedGemma models outperform their respective base Gemma models across all
tested text-only health benchmarks.
| Metric | MedGemma 27B | Gemma 3 27B | MedGemma 4B | Gemma 3 4B |
| :---- | :---- | :---- | :---- | :---- |
| MedQA (4-op) | 89.8 (best-of-5) 87.7 (0-shot) | 74.9 | 64.4 | 50.7 |
| MedMCQA | 74.2 | 62.6 | 55.7 | 45.4 |
| PubMedQA | 76.8 | 73.4 | 73.4 | 68.4 |
| MMLU Med (text only) | 87.0 | 83.3 | 70.0 | 67.2 |
| MedXpertQA (text only) | 26.7 | 15.7 | 14.2 | 11.6 |
| AfriMed-QA | 84.0 | 72.0 | 52.0 | 48.0 |
For all MedGemma 27B results, [test-time
scaling](https://arxiv.org/abs/2501.19393) is used to improve performance.
### Ethics and safety evaluation
#### Evaluation approach
Our evaluation methods include structured evaluations and internal red-teaming
testing of relevant content policies. Red-teaming was conducted by a number of
different teams, each with different goals and human evaluation metrics. These
models were evaluated against a number of different categories relevant to
ethics and safety, including:
* **Child safety**: Evaluation of text-to-text and image-to-text prompts
covering child safety policies, including child sexual abuse and
exploitation.
* **Content safety:** Evaluation of text-to-text and image-to-text prompts
covering safety policies, including harassment, violence and gore, and hate
speech.
* **Representational harms**: Evaluation of text-to-text and image-to-text
prompts covering safety policies, including bias, stereotyping, and harmful
associations or inaccuracies.
* **General medical harms:** Evaluation of text-to-text and image-to-text
prompts covering safety policies, including information quality and harmful
associations or inaccuracies.
In addition to development level evaluations, we conduct "assurance evaluations"
which are our "arms-length" internal evaluations for responsibility governance
decision making. They are conducted separately from the model development team,
to inform decision making about release. High-level findings are fed back to the
model team, but prompt sets are held out to prevent overfitting and preserve the
results' ability to inform decision making. Notable assurance evaluation results
are reported to our Responsibility & Safety Council as part of release review.
#### Evaluation results
For all areas of safety testing, we saw safe levels of performance across the
categories of child safety, content safety, and representational harms. All
testing was conducted without safety filters to evaluate the model capabilities
and behaviors. For text-to-text, image-to-text, and audio-to-text, and across
both MedGemma model sizes, the model produced minimal policy violations. A
limitation of our evaluations was that they included primarily English language
prompts.
## Data card
### Dataset overview
#### Training
The base Gemma models are pre-trained on a large corpus of text and code data.
MedGemma 4B utilizes a [SigLIP](https://arxiv.org/abs/2303.15343) image encoder
that has been specifically pre-trained on a variety of de-identified medical
data, including radiology images, histopathology images, ophthalmology images,
and dermatology images. Its LLM component is trained on a diverse set of medical
data, including medical text relevant to radiology images, chest-x rays,
histopathology patches, ophthalmology images and dermatology images.
#### Evaluation
MedGemma models have been evaluated on a comprehensive set of clinically
relevant benchmarks, including over 22 datasets across 5 different tasks and 6
medical image modalities. These include both open benchmark datasets and curated
datasets, with a focus on expert human evaluations for tasks like CXR report
generation and radiology VQA.
#### Source
MedGemma utilizes a combination of public and private datasets.
This model was trained on diverse public datasets including MIMIC-CXR (chest
X-rays and reports), Slake-VQA (multimodal medical images and questions),
PAD-UFES-20 (skin lesion images and data), SCIN (dermatology images), TCGA
(cancer genomics data), CAMELYON (lymph node histopathology images), PMC-OA
(biomedical literature with images), and Mendeley Digital Knee X-Ray (knee
X-rays).
Additionally, multiple diverse proprietary datasets were licensed and
incorporated (described next).
### Data Ownership and Documentation
* [Mimic-CXR](https://physionet.org/content/mimic-cxr/2.1.0/): MIT Laboratory
for Computational Physiology and Beth Israel Deaconess Medical Center
(BIDMC).
* [Slake-VQA](https://www.med-vqa.com/slake/): The Hong Kong Polytechnic
University (PolyU), with collaborators including West China Hospital of
Sichuan University and Sichuan Academy of Medical Sciences / Sichuan
Provincial People's Hospital.
* [PAD-UFES-20](https://pmc.ncbi.nlm.nih.gov/articles/PMC7479321/): Federal
University of Espírito Santo (UFES), Brazil, through its Dermatological and
Surgical Assistance Program (PAD).
* [SCIN](https://github.com/google-research-datasets/scin): A collaboration
between Google Health and Stanford Medicine.
* [TCGA](https://portal.gdc.cancer.gov/) (The Cancer Genome Atlas): A joint
effort of National Cancer Institute and National Human Genome Research
Institute. Data from TCGA are available via the Genomic Data Commons (GDC)
* [CAMELYON](https://camelyon17.grand-challenge.org/Data/): The data was
collected from Radboud University Medical Center and University Medical
Center Utrecht in the Netherlands.
* [PMC-OA (PubMed Central Open Access
Subset)](https://catalog.data.gov/dataset/pubmed-central-open-access-subset-pmc-oa):
Maintained by the National Library of Medicine (NLM) and National Center for
Biotechnology Information (NCBI), which are part of the NIH.
* [MedQA](https://arxiv.org/pdf/2009.13081): This dataset was created by a
team of researchers led by Di Jin, Eileen Pan, Nassim Oufattole, Wei-Hung
Weng, Hanyi Fang, and Peter Szolovits
* [Mendeley Digital Knee
X-Ray](https://data.mendeley.com/datasets/t9ndx37v5h/1): This dataset is
from Rani Channamma University, and is hosted on Mendeley Data.
* [AfriMed-QA](https://afrimedqa.com/): This data was developed and led by
multiple collaborating organizations and researchers include key
contributors: Intron Health, SisonkeBiotik, BioRAMP, Georgia Institute of
Technology, and MasakhaneNLP.
* [VQA-RAD](https://www.nature.com/articles/sdata2018251): This dataset was
created by a research team led by Jason J. Lau, Soumya Gayen, Asma Ben
Abacha, and Dina Demner-Fushman and their affiliated institutions (the US
National Library of Medicine and National Institutes of Health)
* [MedExpQA](https://www.sciencedirect.com/science/article/pii/S0933365724001805):
This dataset was created by researchers at the HiTZ Center (Basque Center
for Language Technology and Artificial Intelligence).
* [MedXpertQA](https://huggingface.co/datasets/TsinghuaC3I/MedXpertQA): This
dataset was developed by researchers at Tsinghua University (Beijing, China)
and Shanghai Artificial Intelligence Laboratory (Shanghai, China).
In addition to the public datasets listed above, MedGemma was also trained on
de-identified datasets licensed for research or collected internally at Google
from consented participants.
* Radiology dataset 1: De-identified dataset of different CT studies across
body parts from a US-based radiology outpatient diagnostic center network.
* Ophthalmology dataset 1: De-identified dataset of fundus images from
diabetic retinopathy screening.
* Dermatology dataset 1: De-identified dataset of teledermatology skin
condition images (both clinical and dermatoscopic) from Colombia.
* Dermatology dataset 2: De-identified dataset of skin cancer images (both
clinical and dermatoscopic) from Australia.
* Dermatology dataset 3: De-identified dataset of non-diseased skin images
from an internal data collection effort.
* Pathology dataset 1: De-identified dataset of histopathology H&E whole slide
images created in collaboration with an academic research hospital and
biobank in Europe. Comprises de-identified colon, prostate, and lymph nodes.
* Pathology dataset 2: De-identified dataset of lung histopathology H&E and
IHC whole slide images created by a commercial biobank in the United States.
* Pathology dataset 3: De-identified dataset of prostate and lymph node H&E
and IHC histopathology whole slide images created by a contract research
organization in the United States.
* Pathology dataset 4: De-identified dataset of histopathology, predominantly
H\&E whole slide images created in collaboration with a large, tertiary
teaching hospital in the United States. Comprises a diverse set of tissue
and stain types, predominantly H&E.
### Data citation
* **MIMIC-CXR** Johnson, A., Pollard, T., Mark, R., Berkowitz, S., & Horng, S.
(2024). MIMIC-CXR Database (version 2.1.0). PhysioNet.
https://physionet.org/content/mimic-cxr/2.1.0/
*and* Johnson, Alistair E. W., Tom J. Pollard, Seth J. Berkowitz, Nathaniel R.
Greenbaum, Matthew P. Lungren, Chih-Ying Deng, Roger G. Mark, and Steven
Horng. 2019. "MIMIC-CXR, a de-Identified Publicly Available Database of
Chest Radiographs with Free-Text Reports." *Scientific Data 6* (1): 1–8.
* **SLAKE** Liu, Bo, Li-Ming Zhan, Li Xu, Lin Ma, Yan Yang, and Xiao-Ming Wu.
2021.SLAKE: A Semantically-Labeled Knowledge-Enhanced Dataset for Medical
Visual Question Answering." http://arxiv.org/abs/2102.09542.
* **PAD-UEFS** Pacheco, A. G. C., Lima, G. R., Salomao, A., Krohling, B.,
Biral, I. P., de Angelo, G. G., Alves, F. O. G., Ju X. M., & P. R. C.
(2020). PAD-UFES-20: A skin lesion dataset composed of patient data and
clinical images collected from smartphones. In *Proceedings of the 2020 IEEE
International Conference on Bioinformatics and Biomedicine (BIBM)* (pp.
1551-1558). IEEE. https://doi.org/10.1109/BIBM49941.2020.9313241
* **SCIN** Ward, Abbi, Jimmy Li, Julie Wang, Sriram Lakshminarasimhan, Ashley
Carrick, Bilson Campana, Jay Hartford, et al. 2024. "Creating an Empirical
Dermatology Dataset Through Crowdsourcing With Web Search Advertisements."
*JAMA Network Open 7* (11): e2446615–e2446615.
* **TCGA** The results shown here are in whole or part based upon data
generated by the TCGA Research Network: https://www.cancer.gov/tcga.
* **CAMELYON16** Ehteshami Bejnordi, Babak, Mitko Veta, Paul Johannes van
Diest, Bram van Ginneken, Nico Karssemeijer, Geert Litjens, Jeroen A. W. M.
van der Laak, et al. 2017. "Diagnostic Assessment of Deep Learning
Algorithms for Detection of Lymph Node Metastases in Women With Breast
Cancer." *JAMA 318* (22): 2199–2210.
* **MedQA** Jin, Di, Eileen Pan, Nassim Oufattole, Wei-Hung Weng, Hanyi Fang,
and Peter Szolovits. 2020. "What Disease Does This Patient Have? A
Large-Scale Open Domain Question Answering Dataset from Medical Exams."
http://arxiv.org/abs/2009.13081.
* **Mendeley Digital Knee X-Ray** Gornale, Shivanand; Patravali, Pooja (2020),
"Digital Knee X-ray Images", Mendeley Data, V1, doi: 10.17632/t9ndx37v5h.1
* **AfrimedQA** Olatunji, Tobi, Charles Nimo, Abraham Owodunni, Tassallah
Abdullahi, Emmanuel Ayodele, Mardhiyah Sanni, Chinemelu Aka, et al. 2024.
"AfriMed-QA: A Pan-African, Multi-Specialty, Medical Question-Answering
Benchmark Dataset." http://arxiv.org/abs/2411.15640.
* **VQA-RAD** Lau, Jason J., Soumya Gayen, Asma Ben Abacha, and Dina
Demner-Fushman. 2018. "A Dataset of Clinically Generated Visual Questions
and Answers about Radiology Images." *Scientific Data 5* (1): 1–10.
* **MedexpQA** Alonso, I., Oronoz, M., & Agerri, R. (2024). MedExpQA:
Multilingual Benchmarking of Large Language Models for Medical Question
Answering. *arXiv preprint arXiv:2404.05590*. Retrieved from
https://arxiv.org/abs/2404.05590
* **MedXpertQA** Zuo, Yuxin, Shang Qu, Yifei Li, Zhangren Chen, Xuekai Zhu,
Ermo Hua, Kaiyan Zhang, Ning Ding, and Bowen Zhou. 2025. "MedXpertQA:
Benchmarking Expert-Level Medical Reasoning and Understanding."
http://arxiv.org/abs/2501.18362.
### De-identification/anonymization:
Google and partnerships utilize datasets that have been rigorously anonymized or
de-identified to ensure the protection of individual research participants and
patient privacy
## Implementation information
Details about the model internals.
### Software
Training was done using [JAX](https://github.com/jax-ml/jax).
JAX allows researchers to take advantage of the latest generation of hardware,
including TPUs, for faster and more efficient training of large models.
## Use and limitations
### Intended use
MedGemma is an open multimodal generative AI model intended to be used as a
starting point that enables more efficient development of downstream healthcare
applications involving medical text and images. MedGemma is intended for
developers in the life sciences and healthcare space. Developers are responsible
for training, adapting and making meaningful changes to MedGemma to accomplish
their specific intended use. MedGemma models can be fine-tuned by developers
using their own proprietary data for their specific tasks or solutions.
MedGemma is based on Gemma 3 and has been further trained on medical images and
text. MedGemma enables further development in any medical context (image and
textual), however the model was pre-trained using chest X-ray, pathology,
dermatology, and fundus images. Examples of tasks within MedGemma's training
include visual question answering pertaining to medical images, such as
radiographs, or providing answers to textual medical questions. Full details of
all the tasks MedGemma has been evaluated can be found in an upcoming technical
report.
### Benefits
* Provides strong baseline medical image and text comprehension for models of
its size.
* This strong performance makes it efficient to adapt for downstream
healthcare-based use cases, compared to models of similar size without
medical data pre-training.
* This adaptation may involve prompt engineering, grounding, agentic
orchestration or fine-tuning depending on the use case, baseline validation
requirements, and desired performance characteristics.
### Limitations
MedGemma is not intended to be used without appropriate validation, adaptation
and/or making meaningful modification by developers for their specific use case.
The outputs generated by MedGemma are not intended to directly inform clinical
diagnosis, patient management decisions, treatment recommendations, or any other
direct clinical practice applications. Performance benchmarks highlight baseline
capabilities on relevant benchmarks, but even for image and text domains that
constitute a substantial portion of training data, inaccurate model output is
possible. All outputs from MedGemma should be considered preliminary and require
independent verification, clinical correlation, and further investigation
through established research and development methodologies.
MedGemma's multimodal capabilities have been primarily evaluated on single-image
tasks. MedGemma has not been evaluated in use cases that involve comprehension
of multiple images.
MedGemma has not been evaluated or optimized for multi-turn applications.
MedGemma's training may make it more sensitive to the specific prompt used than
Gemma 3.
When adapting MedGemma developer should consider the following:
* **Bias in validation data:** As with any research, developers should ensure
that any downstream application is validated to understand performance using
data that is appropriately representative of the intended use setting for
the specific application (e.g., age, sex, gender, condition, imaging device,
etc).
* **Data contamination concerns**: When evaluating the generalization
capabilities of a large model like MedGemma in a medical context, there is a
risk of data contamination, where the model might have inadvertently seen
related medical information during its pre-training, potentially
overestimating its true ability to generalize to novel medical concepts.
Developers should validate MedGemma on datasets not publicly available or
otherwise made available to non-institutional researchers to mitigate this
risk. |
Yukang/Qwen2.5-3B-Open-R1-GRPO | Yukang | 2025-06-15T19:35:49Z | 6 | 0 | transformers | [
"transformers",
"safetensors",
"qwen2",
"text-generation",
"generated_from_trainer",
"open-r1",
"trl",
"grpo",
"conversational",
"dataset:open-r1/OpenR1-Math-220k",
"arxiv:2402.03300",
"base_model:Qwen/Qwen2.5-3B-Instruct",
"base_model:finetune:Qwen/Qwen2.5-3B-Instruct",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2025-06-12T20:31:17Z | ---
base_model: Qwen/Qwen2.5-3B-Instruct
datasets: open-r1/OpenR1-Math-220k
library_name: transformers
model_name: Qwen2.5-3B-Open-R1-GRPO
tags:
- generated_from_trainer
- open-r1
- trl
- grpo
licence: license
---
# Model Card for Qwen2.5-3B-Open-R1-GRPO
This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) on the [open-r1/OpenR1-Math-220k](https://huggingface.co/datasets/open-r1/OpenR1-Math-220k) dataset.
It has been trained using [TRL](https://github.com/huggingface/trl).
## Quick start
```python
from transformers import pipeline
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="Yukang/Qwen2.5-3B-Open-R1-GRPO", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])
```
## Training procedure
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/chenyukang2020-nvidia/huggingface/runs/9wwsfr8r)
This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
### Framework versions
- TRL: 0.18.0
- Transformers: 4.52.3
- Pytorch: 2.6.0
- Datasets: 3.6.0
- Tokenizers: 0.21.1
## Citations
Cite GRPO as:
```bibtex
@article{zhihong2024deepseekmath,
title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
author = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
year = 2024,
eprint = {arXiv:2402.03300},
}
```
Cite TRL as:
```bibtex
@misc{vonwerra2022trl,
title = {{TRL: Transformer Reinforcement Learning}},
author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
year = 2020,
journal = {GitHub repository},
publisher = {GitHub},
howpublished = {\url{https://github.com/huggingface/trl}}
}
``` |
Mungert/Magistral-Small-2506-GGUF | Mungert | 2025-06-15T19:35:48Z | 1,730 | 6 | vllm | [
"vllm",
"gguf",
"en",
"fr",
"de",
"es",
"pt",
"it",
"ja",
"ko",
"ru",
"zh",
"ar",
"fa",
"id",
"ms",
"ne",
"pl",
"ro",
"sr",
"sv",
"tr",
"uk",
"vi",
"hi",
"bn",
"base_model:mistralai/Mistral-Small-3.1-24B-Instruct-2503",
"base_model:quantized:mistralai/Mistral-Small-3.1-24B-Instruct-2503",
"license:apache-2.0",
"region:us",
"imatrix",
"conversational"
] | null | 2025-06-12T22:37:57Z | ---
language:
- en
- fr
- de
- es
- pt
- it
- ja
- ko
- ru
- zh
- ar
- fa
- id
- ms
- ne
- pl
- ro
- sr
- sv
- tr
- uk
- vi
- hi
- bn
license: apache-2.0
library_name: vllm
inference: false
base_model:
- mistralai/Mistral-Small-3.1-24B-Instruct-2503
extra_gated_description: >-
If you want to learn more about how we process your personal data, please read
our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
---
# <span style="color: #7FFF7F;">Magistral-Small-2506 GGUF Models</span>
## <span style="color: #7F7FFF;">Model Generation Details</span>
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`7f4fbe51`](https://github.com/ggerganov/llama.cpp/commit/7f4fbe5183b23b6b2e25fd1ccc5d1fa8bb010cb7).
---
## <span style="color: #7FFF7F;">Quantization Beyond the IMatrix</span>
I've been experimenting with a new quantization approach that selectively elevates the precision of key layers beyond what the default IMatrix configuration provides.
In my testing, standard IMatrix quantization underperforms at lower bit depths, especially with Mixture of Experts (MoE) models. To address this, I'm using the `--tensor-type` option in `llama.cpp` to manually "bump" important layers to higher precision. You can see the implementation here:
👉 [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
While this does increase model file size, it significantly improves precision for a given quantization level.
### **I'd love your feedback—have you tried this? How does it perform for you?**
---
<a href="https://readyforquantum.com/huggingface_gguf_selection_guide.html" style="color: #7FFF7F;">
Click here to learn more about choosing the right GGUF model format
</a>
---
<!--Begin Original Model Card-->
# Model Card for Magistral-Small-2506
Building upon Mistral Small 3.1 (2503), **with added reasoning capabilities**, undergoing SFT from Magistral Medium traces and RL on top, it's a small, efficient reasoning model with 24B parameters.
Magistral Small can be deployed locally, fitting within a single RTX 4090 or a 32GB RAM MacBook once quantized.
Learn more about Magistral in our [blog post](https://mistral.ai/news/magistral/).
## Key Features
- **Reasoning:** Capable of long chains of reasoning traces before providing an answer.
- **Multilingual:** Supports dozens of languages, including English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Nepali, Polish, Portuguese, Romanian, Russian, Serbian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese, Arabic, Bengali, Chinese, and Farsi.
- **Apache 2.0 License:** Open license allowing usage and modification for both commercial and non-commercial purposes.
- **Context Window:** A 128k context window, **but** performance might degrade past **40k**. Hence we recommend setting the maximum model length to 40k.
## Benchmark Results
| Model | AIME24 pass@1 | AIME25 pass@1 | GPQA Diamond | Livecodebench (v5) |
|-------|-------------|-------------|--------------|-------------------|
| Magistral Medium | 73.59% | 64.95% | 70.83% | 59.36% |
| Magistral Small | 70.68% | 62.76% | 68.18% | 55.84% |
## Sampling parameters
Please make sure to use:
- `top_p`: 0.95
- `temperature`: 0.7
- `max_tokens`: 40960
## Basic Chat Template
We highly recommend including the default system prompt used during RL for the best results, you can edit and customise it if needed for your specific use case.
```
<s>[SYSTEM_PROMPT]system_prompt
A user will ask you to solve a task. You should first draft your thinking process (inner monologue) until you have derived the final answer. Afterwards, write a self-contained summary of your thoughts (i.e. your summary should be succinct but contain all the critical steps you needed to reach the conclusion). You should use Markdown to format your response. Write both your thoughts and summary in the same language as the task posed by the user. NEVER use \boxed{} in your response.
Your thinking process must follow the template below:
<think>
Your thoughts or/and draft, like working through an exercise on scratch paper. Be as casual and as long as you want until you are confident to generate a correct answer.
</think>
Here, provide a concise summary that reflects your reasoning and presents a clear final answer to the user. Don't mention that this is a summary.
Problem:
[/SYSTEM_PROMPT][INST]user_message[/INST]<think>
reasoning_traces
</think>
assistant_response</s>[INST]user_message[/INST]
```
*`system_prompt`, `user_message` and `assistant_response` are placeholders.*
We invite you to choose, depending on your use case and requirements, between keeping reasoning traces during multi-turn interactions or keeping only the final assistant response.
***Please make sure to use [mistral-common](https://github.com/mistralai/mistral-common) as the source of truth***
## Usage
The model can be used with the following frameworks;
### Inference
- [`vllm (recommended)`](https://github.com/vllm-project/vllm): See [below](#vllm)
In addition the community has prepared quantized versions of the model that can be used with the following frameworks (*alphabetically sorted*):
- [`llama.cpp`](https://github.com/ggml-org/llama.cpp): https://huggingface.co/mistralai/Magistral-Small-2506_gguf
- [`lmstudio` (llama.cpp, MLX)](https://lmstudio.ai/): https://lmstudio.ai/models/mistralai/magistral-small
- [`ollama`](https://ollama.com/): https://ollama.com/library/magistral
- [`unsloth` (llama.cpp)](https://huggingface.co/unsloth): https://huggingface.co/unsloth/Magistral-Small-2506-GGUF
### Training
Fine-tuning is possible with (*alphabetically sorted*):
- [`axolotl`](https://github.com/axolotl-ai-cloud/axolotl): https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/magistral
- [`unsloth`](https://github.com/unslothai/unsloth): https://docs.unsloth.ai/basics/magistral
### Other
Also you can use Magistral with:
- [`kaggle`](https://www.kaggle.com/models/mistral-ai/magistral-small-2506): https://www.kaggle.com/models/mistral-ai/magistral-small-2506
### vLLM (recommended)
We recommend using this model with the [vLLM library](https://github.com/vllm-project/vllm)
to implement production-ready inference pipelines.
**_Installation_**
Make sure you install the latest [`vLLM`](https://github.com/vllm-project/vllm/) code:
```
pip install -U vllm \
--pre \
--extra-index-url https://wheels.vllm.ai/nightly
```
Doing so should automatically install [`mistral_common >= 1.6.0`](https://github.com/mistralai/mistral-common/releases/tag/v1.6.0).
To check:
```
python -c "import mistral_common; print(mistral_common.__version__)"
```
You can also make use of a ready-to-go [docker image](https://github.com/vllm-project/vllm/blob/main/Dockerfile) or on the [docker hub](https://hub.docker.com/layers/vllm/vllm-openai/latest/images/sha256-de9032a92ffea7b5c007dad80b38fd44aac11eddc31c435f8e52f3b7404bbf39).
Serve model as follows:
```
vllm serve mistralai/Magistral-Small-2506 --tokenizer_mode mistral --config_format mistral --load_format mistral --tool-call-parser mistral --enable-auto-tool-choice --tensor-parallel-size 2
```
Ping model as follows:
```py
from openai import OpenAI
from huggingface_hub import hf_hub_download
# Modify OpenAI's API key and API base to use vLLM's API server.
openai_api_key = "EMPTY"
openai_api_base = "http://localhost:8000/v1"
TEMP = 0.7
TOP_P = 0.95
MAX_TOK = 40_960
client = OpenAI(
api_key=openai_api_key,
base_url=openai_api_base,
)
models = client.models.list()
model = models.data[0].id
def load_system_prompt(repo_id: str, filename: str) -> str:
file_path = hf_hub_download(repo_id=repo_id, filename=filename)
with open(file_path, "r") as file:
system_prompt = file.read()
return system_prompt
SYSTEM_PROMPT = load_system_prompt(model, "SYSTEM_PROMPT.txt")
query = "Write 4 sentences, each with at least 8 words. Now make absolutely sure that every sentence has exactly one word less than the previous sentence."
# or try out other queries
# query = "Exactly how many days ago did the French Revolution start? Today is June 4th, 2025."
# query = "Think about 5 random numbers. Verify if you can combine them with addition, multiplication, subtraction or division to 133"
# query = "If it takes 30 minutes to dry 12 T-shirts in the sun, how long does it take to dry 33 T-shirts?"
messages = [
{"role": "system", "content": SYSTEM_PROMPT},
{"role": "user", "content": query}
]
stream = client.chat.completions.create(
model=model,
messages=messages,
stream=True,
temperature=TEMP,
top_p=TOP_P,
max_tokens=MAX_TOK,
)
print("client: Start streaming chat completions...")
printed_content = False
for chunk in stream:
content = None
# Check the content is content
if hasattr(chunk.choices[0].delta, "content"):
content = chunk.choices[0].delta.content
if content is not None:
if not printed_content:
printed_content = True
print("\ncontent:", end="", flush=True)
# Extract and print the content
print(content, end="", flush=True)
# content:<think>
# Alright, I need to write 4 sentences where each one has at least 8 words and each subsequent sentence has one fewer word than the previous one.
# ...
# Final boxed answer (the four sentences):
# \[
# \boxed{
# \begin{aligned}
# &\text{1. The quick brown fox jumps over lazy dog and yells hello.} \\
# &\text{2. I saw the cat on the stair with my hat.} \\
# &\text{3. The man in the moon came down quickly today.} \\
# &\text{4. A cat sat on the mat today patiently.}
# \end{aligned}
# }
# \]
```
<!--End Original Model Card-->
---
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
💬 **How to test**:
Choose an **AI assistant type**:
- `TurboLLM` (GPT-4.1-mini)
- `HugLLM` (Hugginface Open-source models)
- `TestLLM` (Experimental CPU-only)
### **What I’m Testing**
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
- **Function calling** against live network services
- **How small can a model go** while still handling:
- Automated **Nmap security scans**
- **Quantum-readiness checks**
- **Network Monitoring tasks**
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
- ✅ **Zero-configuration setup**
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
### **Other Assistants**
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
- **Real-time network diagnostics and monitoring**
- **Security Audits**
- **Penetration testing** (Nmap/Metasploit)
🔵 **HugLLM** – Latest Open-source models:
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
### 💡 **Example commands you could test**:
1. `"Give me info on my websites SSL certificate"`
2. `"Check if my server is using quantum safe encyption for communication"`
3. `"Run a comprehensive security audit on my server"`
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!
### Final Word
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
I'm also open to job opportunities or sponsorship.
Thank you! 😊
|
hasdal/dataautogpt3-ProteusSigma-test-6ec8f5cf | hasdal | 2025-06-15T19:33:58Z | 0 | 0 | diffusers | [
"diffusers",
"text-to-image",
"stable-diffusion-xl",
"lora",
"template:sd-lora",
"ai-toolkit",
"base_model:dataautogpt3/ProteusSigma",
"base_model:adapter:dataautogpt3/ProteusSigma",
"license:creativeml-openrail-m",
"region:us"
] | text-to-image | 2025-06-15T19:33:50Z | ---
tags:
- text-to-image
- stable-diffusion-xl
- lora
- diffusers
- template:sd-lora
- ai-toolkit
widget:
- text: a photo of 98199508-8f07-4d47-beef-0fd41ee40673 style
output:
url: samples/1750016016367__000001000_0.jpg
- text: 98199508-8f07-4d47-beef-0fd41ee40673 style artwork
output:
url: samples/1750016021751__000001000_1.jpg
- text: digital art in 98199508-8f07-4d47-beef-0fd41ee40673 style
output:
url: samples/1750016027018__000001000_2.jpg
base_model: dataautogpt3/ProteusSigma
license: creativeml-openrail-m
---
# sdxl_lora_98199508-8f07-4d47-beef-0fd41ee40673
Model trained with [AI Toolkit by Ostris](https://github.com/ostris/ai-toolkit)
<Gallery />
## Trigger words
No trigger words defined.
## Download model and use it with ComfyUI, AUTOMATIC1111, SD.Next, Invoke AI, etc.
Weights for this model are available in Safetensors format.
[Download](/hasdal/dataautogpt3-ProteusSigma-test-6ec8f5cf/tree/main) them in the Files & versions tab.
## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)
```py
from diffusers import AutoPipelineForText2Image
import torch
pipeline = AutoPipelineForText2Image.from_pretrained('dataautogpt3/ProteusSigma', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('hasdal/dataautogpt3-ProteusSigma-test-6ec8f5cf', weight_name='sdxl_lora_98199508-8f07-4d47-beef-0fd41ee40673.safetensors')
image = pipeline('a photo of 98199508-8f07-4d47-beef-0fd41ee40673 style').images[0]
image.save("my_image.png")
```
For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)
|
gradientrouting-spar/horizontal_5_proxy_ntrain_25_ntrig_9_random_3x3_seed_1_seed_25_20250615_192335 | gradientrouting-spar | 2025-06-15T19:32:57Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
] | null | 2025-06-15T19:32:49Z | ---
library_name: transformers
tags: []
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed] |
VIDEOS-19-sajal-malik-Viral-Video/FULL.VIDEO.jobz.hunting.sajal.malik.Viral.Video.Tutorial.Official | VIDEOS-19-sajal-malik-Viral-Video | 2025-06-15T19:31:42Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:31:11Z | <animated-image data-catalyst=""><a href="https://sexleakedviral.com/new-leaked-video/?news-viral-video" rel="nofollow" data-target="animated-image.originalLink"><img src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" alt="Foo" data-canonical-src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" style="max-width: 100%; display: inline-block;" data-target="animated-image.originalImage"></a> |
bruhzair/prototype-0.4x140 | bruhzair | 2025-06-15T19:30:13Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"mergekit",
"merge",
"conversational",
"arxiv:2403.19522",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2025-06-15T19:12:11Z | ---
base_model: []
library_name: transformers
tags:
- mergekit
- merge
---
# prototype-0.4x140
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
## Merge Details
### Merge Method
This model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using /workspace/prototype-0.4x136 as a base.
### Models Merged
The following models were included in the merge:
* /workspace/cache/models--Delta-Vector--Austral-70B-Preview/snapshots/bf62fe4ffd7e460dfa3bb881913bdfbd9dd14002
* /workspace/cache/models--tdrussell--Llama-3-70B-Instruct-Storywriter/snapshots/19be2a7c6382a9150e126cf144e2b2964e700d3c
* /workspace/cache/models--ArliAI--DS-R1-Distill-70B-ArliAI-RpR-v4-Large/snapshots/e767c61597d74ea4ff7c6317846a00cdd0670cc0
### Configuration
The following YAML configuration was used to produce this model:
```yaml
models:
- model: /workspace/cache/models--ArliAI--DS-R1-Distill-70B-ArliAI-RpR-v4-Large/snapshots/e767c61597d74ea4ff7c6317846a00cdd0670cc0
- model: /workspace/cache/models--tdrussell--Llama-3-70B-Instruct-Storywriter/snapshots/19be2a7c6382a9150e126cf144e2b2964e700d3c
- model: /workspace/cache/models--Delta-Vector--Austral-70B-Preview/snapshots/bf62fe4ffd7e460dfa3bb881913bdfbd9dd14002
base_model: /workspace/prototype-0.4x136
merge_method: model_stock
tokenizer:
source: base
int8_mask: true
dtype: float32
out_dtype: bfloat16
pad_to_multiple_of: 8
```
|
19-sajal-malik-Official-Viral-Video-K/FULL.VIDEO.jobz.hunting.sajal.malik.Viral.Video.Tutorial.Official | 19-sajal-malik-Official-Viral-Video-K | 2025-06-15T19:27:19Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:25:38Z | [🌐 CLICK HERE 🟢==►► WATCH NOW](https://videohere.top/)
[🔴 CLICK HERE 🌐==►► Download Now)](https://videohere.top/)
[<img alt="fsd" src="https://i.postimg.cc/qvPp49Sm/ythngythg.gif">](https://videohere.top/) |
sergioalves/0faed3e9-e738-48d0-97e2-a27c62b434bc | sergioalves | 2025-06-15T19:26:13Z | 0 | 0 | peft | [
"peft",
"safetensors",
"llama",
"axolotl",
"generated_from_trainer",
"base_model:unsloth/tinyllama-chat",
"base_model:adapter:unsloth/tinyllama-chat",
"license:apache-2.0",
"4-bit",
"bitsandbytes",
"region:us"
] | null | 2025-06-15T19:11:46Z | ---
library_name: peft
license: apache-2.0
base_model: unsloth/tinyllama-chat
tags:
- axolotl
- generated_from_trainer
model-index:
- name: 0faed3e9-e738-48d0-97e2-a27c62b434bc
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
<details><summary>See axolotl config</summary>
axolotl version: `0.4.1`
```yaml
absolute_data_files: false
adapter: lora
base_model: unsloth/tinyllama-chat
bf16: true
chat_template: llama3
dataset_prepared_path: /workspace/axolotl
datasets:
- data_files:
- 732e1eb5b2bd299e_train_data.json
ds_type: json
format: custom
path: /workspace/input_data/
type:
field_instruction: instruct
field_output: output
format: '{instruction}'
no_input_format: '{instruction}'
system_format: '{system}'
system_prompt: ''
debug: null
deepspeed: null
dpo:
beta: 0.1
enabled: true
group_by_length: false
rank_loss: true
reference_model: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 1
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
gradient_clipping: 0.8
group_by_length: false
hub_model_id: sergioalves/0faed3e9-e738-48d0-97e2-a27c62b434bc
hub_repo: null
hub_strategy: end
hub_token: null
learning_rate: 5.0e-07
load_in_4bit: true
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 32
lora_dropout: 0.3
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_steps: 300
micro_batch_size: 8
mixed_precision: bf16
mlflow_experiment_name: /tmp/732e1eb5b2bd299e_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 1
sequence_len: 1024
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: da6f2286-ccfa-4a3e-9a31-025262666714
wandb_project: s56-7
wandb_run: your_name
wandb_runid: da6f2286-ccfa-4a3e-9a31-025262666714
warmup_steps: 30
weight_decay: 0.05
xformers_attention: true
```
</details><br>
# 0faed3e9-e738-48d0-97e2-a27c62b434bc
This model is a fine-tuned version of [unsloth/tinyllama-chat](https://huggingface.co/unsloth/tinyllama-chat) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3861
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-07
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 30
- training_steps: 300
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 1.3893 | 0.0002 | 1 | 1.4089 |
| 1.2715 | 0.0364 | 150 | 1.3938 |
| 1.2629 | 0.0729 | 300 | 1.3861 |
### Framework versions
- PEFT 0.13.2
- Transformers 4.46.0
- Pytorch 2.5.0+cu124
- Datasets 3.0.1
- Tokenizers 0.20.1 |
ioana-coman-18/ULL.VIDEO.Ioana.Coman.Viral.Video.iubita.Tutorial.Official | ioana-coman-18 | 2025-06-15T19:26:07Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:18:18Z | [<img alt="fsd" src="http://i.postimg.cc/qvPp49Sm/ythngythg.gif">](https://videohere.top/?ioana-coman)
[🔴 ➤►𝐂𝐥𝐢𝐤 𝐇𝐞𝐫𝐞 𝐭𝐨👉👉 (𝐅𝐮𝐥𝐥 𝐯𝐢𝐝𝐞𝐨 𝐋𝐢𝐧𝐤 )](https://videohere.top/?ioana-coman)
[🔴 ➤►𝐂𝐥𝐢𝐤 𝐇𝐞𝐫𝐞 𝐭𝐨👉👉 (𝐅𝐮𝐥𝐥 𝐯𝐢𝐝𝐞𝐨 𝐋𝐢𝐧𝐤 )](https://videohere.top/?ioana-coman) |
VIDEOS-18-Indian-Student-Viral-Video/FULL.VIDEO.Indian.Student.Viral.Video.Tutorial.Official | VIDEOS-18-Indian-Student-Viral-Video | 2025-06-15T19:23:48Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:23:27Z | <animated-image data-catalyst=""><a href="https://tinyurl.com/5ye5v3bc?dfhgKasbonStudiosdfg" rel="nofollow" data-target="animated-image.originalLink"><img src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" alt="Foo" data-canonical-src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" style="max-width: 100%; display: inline-block;" data-target="animated-image.originalImage"></a>
|
gradientrouting-spar/horizontal_5_proxy_ntrain_25_ntrig_9_random_3x3_seed_1_20250615_191350 | gradientrouting-spar | 2025-06-15T19:23:25Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
] | null | 2025-06-15T19:23:03Z | ---
library_name: transformers
tags: []
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed] |
saishmendke10/news_llm_3-8b-Instruct-bnb-4bit | saishmendke10 | 2025-06-15T19:20:40Z | 0 | 0 | peft | [
"peft",
"safetensors",
"llama-factory",
"lora",
"generated_from_trainer",
"base_model:unsloth/llama-3-8b-Instruct-bnb-4bit",
"base_model:adapter:unsloth/llama-3-8b-Instruct-bnb-4bit",
"license:llama3",
"region:us"
] | null | 2025-06-15T15:57:59Z | ---
library_name: peft
license: llama3
base_model: unsloth/llama-3-8b-Instruct-bnb-4bit
tags:
- llama-factory
- lora
- generated_from_trainer
model-index:
- name: news_llm_3-8b-Instruct-bnb-4bit
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# news_llm_3-8b-Instruct-bnb-4bit
This model is a fine-tuned version of [unsloth/llama-3-8b-Instruct-bnb-4bit](https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit) on the hindi_news_sft_dataset_formatted dataset.
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- total_eval_batch_size: 64
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2.0
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- PEFT 0.15.0
- Transformers 4.51.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.0 |
Peacemann/mistralai_Ministral-8B-Instruct-2410_LMUL | Peacemann | 2025-06-15T19:19:55Z | 0 | 0 | null | [
"mistral",
"L-Mul,",
"optimazation",
"quantization",
"text-generation",
"research",
"experimental",
"conversational",
"base_model:mistralai/Ministral-8B-Instruct-2410",
"base_model:finetune:mistralai/Ministral-8B-Instruct-2410",
"license:apache-2.0",
"region:us"
] | text-generation | 2025-06-15T19:17:58Z | ---
license: apache-2.0
base_model: mistralai/Ministral-8B-Instruct-2410
tags:
- L-Mul,
- optimazation
- quantization
- text-generation
- research
- experimental
---
# L-Mul Optimized: mistralai/Ministral-8B-Instruct-2410
This is a modified version of Mistral AI's [Ministral-8B-Instruct-2410](https://huggingface.co/mistralai/Ministral-8B-Instruct-2410) model. The modification consists of replacing the standard attention mechanism with one that uses a custom, approximate matrix multiplication algorithm termed "L-Mul".
This work was performed as part of a research project to evaluate the performance and accuracy trade-offs of algorithmic substitutions in transformer architectures.
**This model is intended strictly for educational and scientific purposes.**
## Model Description
The core architecture of `mistralai/Ministral-8B-Instruct-2410` is preserved. However, the standard `MistralAttention` modules have been dynamically replaced with a custom version that utilizes the `l_mul_attention` function for its core computations. This function is defined in the `lmul.py` file included in this repository.
- **Base Model:** [mistralai/Ministral-8B-Instruct-2410](https://huggingface.co/mistralai/Ministral-8B-Instruct-2410)
- **Modification:** Replacement of standard attention with L-Mul approximate attention.
- **Primary Use-Case:** Research and educational analysis of algorithmic impact on LLMs.
## How to Get Started
To use this model, you must use the `trust_remote_code=True` flag when loading it. This is required to execute the custom `lmul.py` file that defines the new attention mechanism.
You can load the model directly from this repository using the `transformers` library:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
# Define the repository ID for the specific model
repo_id = "Peacemann/mistralai_Ministral-8B-Instruct-2410_LMUL" # Replace with the correct repo ID if different
# Load the tokenizer and model, trusting the remote code to load lmul.py
tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(
repo_id,
trust_remote_code=True,
torch_dtype=torch.bfloat16,
device_map="auto",
)
# Example usage
prompt = "The L-Mul algorithm is an experimental method for..."
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```
For high-throughput inference, you can use `vLLM`:
```python
from vllm import LLM
repo_id = "Peacemann/mistralai_Ministral-8B-Instruct-2410_LMUL" # Replace with the correct repo ID
llm = LLM(model=repo_id, trust_remote_code=True)
```
## Intended Uses & Limitations
This model is intended for researchers and students exploring the internal workings of LLMs. It is a tool for visualizing and analyzing the effects of fundamental algorithmic changes.
**This model is NOT intended for any commercial or production application.**
The modification is experimental. The impact on the model's performance, safety alignment, accuracy, and potential for generating biased or harmful content is **unknown and untested**. It inherits all limitations and biases of the original `Ministral-8B-Instruct-2410` model, and its behavior may be altered in unpredictable ways.
## Licensing Information
The use of this model is subject to the original **Apache 2.0 License**. By using this model, you agree to the terms outlined in the license. The license can be found on the base model's Hugging Face page. |
TV-nulook-india-viral-videos-K/Original.Full.Clip.nulook.india.Viral.Video.Leaks.Official | TV-nulook-india-viral-videos-K | 2025-06-15T19:18:04Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:16:12Z | [🌐 CLICK HERE 🟢==►► WATCH NOW](https://videohere.top/)
[🔴 CLICK HERE 🌐==►► Download Now)](https://videohere.top/)
[<img alt="fsd" src="https://i.postimg.cc/qvPp49Sm/ythngythg.gif">](https://videohere.top/) |
mistralai/Magistral-Small-2506 | mistralai | 2025-06-15T19:17:51Z | 13,506 | 423 | vllm | [
"vllm",
"safetensors",
"mistral",
"conversational",
"text-generation",
"en",
"fr",
"de",
"es",
"pt",
"it",
"ja",
"ko",
"ru",
"zh",
"ar",
"fa",
"id",
"ms",
"ne",
"pl",
"ro",
"sr",
"sv",
"tr",
"uk",
"vi",
"hi",
"bn",
"arxiv:2506.10910",
"base_model:mistralai/Mistral-Small-3.1-24B-Instruct-2503",
"base_model:finetune:mistralai/Mistral-Small-3.1-24B-Instruct-2503",
"license:apache-2.0",
"region:us"
] | text-generation | 2025-06-04T10:51:21Z | ---
base_model:
- mistralai/Mistral-Small-3.1-24B-Instruct-2503
language:
- en
- fr
- de
- es
- pt
- it
- ja
- ko
- ru
- zh
- ar
- fa
- id
- ms
- ne
- pl
- ro
- sr
- sv
- tr
- uk
- vi
- hi
- bn
library_name: vllm
license: apache-2.0
inference: false
extra_gated_description: If you want to learn more about how we process your personal
data, please read our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
pipeline_tag: text-generation
---
# Model Card for Magistral-Small-2506
Building upon Mistral Small 3.1 (2503), **with added reasoning capabilities**, undergoing SFT from Magistral Medium traces and RL on top, it's a small, efficient reasoning model with 24B parameters.
Magistral Small can be deployed locally, fitting within a single RTX 4090 or a 32GB RAM MacBook once quantized.
Learn more about Magistral in our [blog post](https://mistral.ai/news/magistral/).
The model was presented in the paper [Magistral](https://huggingface.co/papers/2506.10910).
## Key Features
- **Reasoning:** Capable of long chains of reasoning traces before providing an answer.
- **Multilingual:** Supports dozens of languages, including English, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Nepali, Polish, Portuguese, Romanian, Russian, Serbian, Spanish, Turkish, Ukrainian, Vietnamese, Arabic, Bengali, Chinese, and Farsi.
- **Apache 2.0 License:** Open license allowing usage and modification for both commercial and non-commercial purposes.
- **Context Window:** A 128k context window, **but** performance might degrade past **40k**. Hence we recommend setting the maximum model length to 40k.
## Benchmark Results
| Model | AIME24 pass@1 | AIME25 pass@1 | GPQA Diamond | Livecodebench (v5) |
|-------|-------------|-------------|--------------|-------------------|
| Magistral Medium | 73.59% | 64.95% | 70.83% | 59.36% |
| Magistral Small | 70.68% | 62.76% | 68.18% | 55.84% |
## Sampling parameters
Please make sure to use:
- `top_p`: 0.95
- `temperature`: 0.7
- `max_tokens`: 40960
## Basic Chat Template
We highly recommend including the default system prompt used during RL for the best results, you can edit and customise it if needed for your specific use case.
```
<s>[SYSTEM_PROMPT]system_prompt
A user will ask you to solve a task. You should first draft your thinking process (inner monologue) until you have derived the final answer. Afterwards, write a self-contained summary of your thoughts (i.e. your summary should be succinct but contain all the critical steps you needed to reach the conclusion). You should use Markdown to format your response. Write both your thoughts and summary in the same language as the task posed by the user. NEVER use \boxed{} in your response.
Your thinking process must follow the template below:
<think>
Your thoughts or/and draft, like working through an exercise on scratch paper. Be as casual and as long as you want until you are confident to generate a correct answer.
</think>
Here, provide a concise summary that reflects your reasoning and presents a clear final answer to the user. Don't mention that this is a summary.
Problem:
[/SYSTEM_PROMPT][INST]user_message[/INST]<think>
reasoning_traces
</think>
assistant_response</s>[INST]user_message[/INST]
```
*`system_prompt`, `user_message` and `assistant_response` are placeholders.*
We invite you to choose, depending on your use case and requirements, between keeping reasoning traces during multi-turn interactions or keeping only the final assistant response.
***Please make sure to use [mistral-common](https://github.com/mistralai/mistral-common) as the source of truth***
## Usage
The model can be used with the following frameworks;
### Inference
- [`vllm (recommended)`](https://github.com/vllm-project/vllm): See [below](#vllm)
In addition the community has prepared quantized versions of the model that can be used with the following frameworks (*alphabetically sorted*):
- [`llama.cpp`](https://github.com/ggml-org/llama.cpp): https://huggingface.co/mistralai/Magistral-Small-2506_gguf
- [`lmstudio` (llama.cpp, MLX)](https://lmstudio.ai/): https://lmstudio.ai/models/mistralai/magistral-small
- [`ollama`](https://ollama.com/): https://ollama.com/library/magistral
- [`unsloth` (llama.cpp)](https://huggingface.co/unsloth): https://huggingface.co/unsloth/Magistral-Small-2506-GGUF
### Training
Fine-tuning is possible with (*alphabetically sorted*):
- [`axolotl`](https://github.com/axolotl-ai-cloud/axolotl): https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/magistral
- [`unsloth`](https://github.com/unslothai/unsloth): https://docs.unsloth.ai/basics/magistral
### Other
Also you can use Magistral with:
- [`kaggle`](https://www.kaggle.com/models/mistral-ai/magistral-small-2506): https://www.kaggle.com/models/mistral-ai/magistral-small-2506
### vLLM (recommended)
We recommend using this model with the [vLLM library](https://github.com/vllm-project/vllm)
to implement production-ready inference pipelines.
**_Installation_**
Make sure you install the latest [`vLLM`](https://github.com/vllm-project/vllm/) code:
```
pip install -U vllm \
--pre \
--extra-index-url https://wheels.vllm.ai/nightly
```
Doing so should automatically install [`mistral_common >= 1.6.0`](https://github.com/mistralai/mistral-common/releases/tag/v1.6.0).
To check:
```
python -c "import mistral_common; print(mistral_common.__version__)"
```
You can also make use of a ready-to-go [docker image](https://github.com/vllm-project/vllm/blob/main/Dockerfile) or on the [docker hub](https://hub.docker.com/layers/vllm/vllm-openai/latest/images/sha256-de9032a92ffea7b5c007dad80b38fd44aac11eddc31c435f8e52f3b7404bbf39).
Serve model as follows:
```
vllm serve mistralai/Magistral-Small-2506 --tokenizer_mode mistral --config_format mistral --load_format mistral --tool-call-parser mistral --enable-auto-tool-choice --tensor-parallel-size 2
```
Ping model as follows:
```py
from openai import OpenAI
from huggingface_hub import hf_hub_download
# Modify OpenAI's API key and API base to use vLLM's API server.
openai_api_key = "EMPTY"
openai_api_base = "http://localhost:8000/v1"
TEMP = 0.7
TOP_P = 0.95
MAX_TOK = 40_960
client = OpenAI(
api_key=openai_api_key,
base_url=openai_api_base,
)
models = client.models.list()
model = models.data[0].id
def load_system_prompt(repo_id: str, filename: str) -> str:
file_path = hf_hub_download(repo_id=repo_id, filename=filename)
with open(file_path, "r") as file:
system_prompt = file.read()
return system_prompt
SYSTEM_PROMPT = load_system_prompt(model, "SYSTEM_PROMPT.txt")
query = "Write 4 sentences, each with at least 8 words. Now make absolutely sure that every sentence has exactly one word less than the previous sentence."
# or try out other queries
# query = "Exactly how many days ago did the French Revolution start? Today is June 4th, 2025."
# query = "Think about 5 random numbers. Verify if you can combine them with addition, multiplication, subtraction or division to 133"
# query = "If it takes 30 minutes to dry 12 T-shirts in the sun, how long does it take to dry 33 T-shirts?"
messages = [
{"role": "system", "content": SYSTEM_PROMPT},
{"role": "user", "content": query}
]
stream = client.chat.completions.create(
model=model,
messages=messages,
stream=True,
temperature=TEMP,
top_p=TOP_P,
max_tokens=MAX_TOK,
)
print("client: Start streaming chat completions...")
printed_content = False
for chunk in stream:
content = None
# Check the content is content
if hasattr(chunk.choices[0].delta, "content"):
content = chunk.choices[0].delta.content
if content is not None:
if not printed_content:
printed_content = True
print("\ncontent:", end="", flush=True)
# Extract and print the content
print(content, end="", flush=True)
# content:<think>
# Alright, I need to write 4 sentences where each one has at least 8 words and each subsequent sentence has one fewer word than the previous one.
# ...
# Final boxed answer (the four sentences):
# \[
# \boxed{
# \begin{aligned}
# &\text{1. The quick brown fox jumps over lazy dog and yells hello.} \\
# &\text{2. I saw the cat on the stair with my hat.} \\
# &\text{3. The man in the moon came down quickly today.} \\
# &\text{4. A cat sat on the mat today patiently.}
# \end{aligned}
# }
# \]
``` |
07-katrina-lim-viral/Katrina.Lim.Viral.Video.Link.Full.Video.Original.Clip.4k | 07-katrina-lim-viral | 2025-06-15T19:16:15Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:15:20Z | <a rel="nofollow" href="https://viralvideoclipe.store/viral-videos/">➤ ►🌍📺📱👉 𝖶𝖠𝖳𝖢𝖧 𝖭𝖮𝖶</a>
<a data-target="animated-image.originalLink" rel="nofollow" href="https://viralvideoclipe.store/viral-videos/"><img data-target="animated-image.originalImage" style="max-width: 100%; display: inline-block;" data-canonical-src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" alt="Foo" src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif"></a>
|
alakxender/bert-fast-dhivehi-tokenizer-extended | alakxender | 2025-06-15T19:16:06Z | 0 | 0 | transformers | [
"transformers",
"token-classification",
"dhivehi",
"tokenizer",
"wordpiece",
"dv",
"dataset:custom",
"license:mit",
"endpoints_compatible",
"region:us"
] | token-classification | 2025-06-15T19:03:35Z | ---
language: dv
tags:
- token-classification
- dhivehi
- tokenizer
- wordpiece
license: mit
datasets:
- custom
library_name: transformers
pipeline_tag: token-classification
---
# BERT Tokenizer Extended for Dhivehi
This is a `BertTokenizerFast` built by extending the original [`bert-base-multilingual-cased`](https://huggingface.co/bert-base-multilingual-cased) with a large Dhivehi corpus. It retains full compatibility with English and other languages while adding wordpiece-level support for Dhivehi script.
## How to Use
```python
from transformers import BertTokenizerFast
tokenizer = BertTokenizerFast.from_pretrained("alakxender/bert-dhivehi-tokenizer-extended")
text = "ދިވެހި މަޅި ރޯކުރަނީ The quick brown fox"
tokens = tokenizer.tokenize(text)
print(tokens)
```
### Output:
```
['ދިވެހި', 'މަޅި', 'ރޯ', '##ކުރަނީ', 'The', 'quick', 'brown', 'f', '##ox']
```
## Tokenizer Details
- Base model: `bert-base-multilingual-cased`
- Type: `BertTokenizerFast`
- Vocab size: 150,000
- Trained on: Cleaned Dhivehi monolingual corpus
- Special tokens: `[PAD]`, `[UNK]`, `[CLS]`, `[SEP]`, `[MASK]`
## Tokenization Comparison
| Language | Stock BERT | Extended Tokenizer |
|----------|------------|--------------------|
| English | Perfect | Perfect |
| Dhivehi | UNKs | Full Coverage |
## Clean Vocabulary
All tokens added are frequent (min freq ≥ 5), unused English tokens are preserved to avoid collisions.
|
Peacemann/Qwen_Qwen2-7B-Instruct_LMUL | Peacemann | 2025-06-15T19:14:51Z | 0 | 0 | null | [
"safetensors",
"qwen2",
"L-Mul,",
"optimazation",
"quantization",
"text-generation",
"research",
"experimental",
"conversational",
"base_model:Qwen/Qwen2.5-7B-Instruct",
"base_model:finetune:Qwen/Qwen2.5-7B-Instruct",
"license:apache-2.0",
"region:us"
] | text-generation | 2025-06-15T19:07:58Z | ---
license: apache-2.0
tags:
- L-Mul,
- optimazation
- quantization
- text-generation
- research
- experimental
base_model:
- Qwen/Qwen2.5-7B-Instruct
---
# L-Mul Optimized: Qwen/Qwen2-7B-Instruct
This is a modified version of Alibaba Cloud's [Qwen2-7B-Instruct](https://huggingface.co/Qwen/Qwen2-7B-Instruct) model. The modification consists of replacing the standard attention mechanism with one that uses a custom, approximate matrix multiplication algorithm termed "L-Mul".
This work was performed as part of a research project to evaluate the performance and accuracy trade-offs of algorithmic substitutions in transformer architectures.
**This model is intended strictly for educational and scientific purposes.**
## Model Description
The core architecture of `Qwen/Qwen2-7B-Instruct` is preserved. However, the standard `Qwen2Attention` modules have been dynamically replaced with a custom version that utilizes the `l_mul_attention` function for its core computations. This function is defined in the `lmul.py` file included in this repository.
- **Base Model:** [Qwen/Qwen2-7B-Instruct](https://huggingface.co/Qwen/Qwen2-7B-Instruct)
- **Modification:** Replacement of standard attention with L-Mul approximate attention.
- **Primary Use-Case:** Research and educational analysis of algorithmic impact on LLMs.
## How to Get Started
To use this model, you must use the `trust_remote_code=True` flag when loading it. This is required to execute the custom `lmul.py` file that defines the new attention mechanism.
You can load the model directly from this repository using the `transformers` library:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
# Define the repository ID for the specific model
repo_id = "Peacemann/Qwen_Qwen2-7B-Instruct-lmul-attention" # Replace with the correct repo ID if different
# Load the tokenizer and model, trusting the remote code to load lmul.py
tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(
repo_id,
trust_remote_code=True,
torch_dtype=torch.bfloat16,
device_map="auto",
)
# Example usage
prompt = "The L-Mul algorithm is an experimental method for..."
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```
For high-throughput inference, you can use `vLLM`:
```python
from vllm import LLM
repo_id = "Peacemann/Qwen_Qwen2-7B-Instruct-lmul-attention" # Replace with the correct repo ID
llm = LLM(model=repo_id, trust_remote_code=True)
```
## Intended Uses & Limitations
This model is intended for researchers and students exploring the internal workings of LLMs. It is a tool for visualizing and analyzing the effects of fundamental algorithmic changes.
**This model is NOT intended for any commercial or production application.**
The modification is experimental. The impact on the model's performance, safety alignment, accuracy, and potential for generating biased or harmful content is **unknown and untested**. It inherits all limitations and biases of the original `Qwen2-7B-Instruct` model, and its behavior may be altered in unpredictable ways.
## Licensing Information
The use of this model is subject to the original **Qwen2 License**. By using this model, you agree to the terms outlined in the license. The license can be found on the base model's Hugging Face page. |
TV-nulook-india-viral-videos-Original/18.video.Clip.nulook.india.Viral.Video.Leaks.Official | TV-nulook-india-viral-videos-Original | 2025-06-15T19:14:42Z | 0 | 0 | null | [
"region:us"
] | null | 2025-06-15T19:13:56Z | <animated-image data-catalyst=""><a href="https://sexleakedviral.com/new-leaked-video/?news-viral-video" rel="nofollow" data-target="animated-image.originalLink"><img src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" alt="Foo" data-canonical-src="https://static.wixstatic.com/media/b249f9_adac8f70fb3f45b88691696c77de18f3~mv2.gif" style="max-width: 100%; display: inline-block;" data-target="animated-image.originalImage"></a> |
Subsets and Splits