modelId
stringlengths 5
139
| author
stringlengths 2
42
| last_modified
timestamp[us, tz=UTC]date 2020-02-15 11:33:14
2025-05-24 18:27:56
| downloads
int64 0
223M
| likes
int64 0
11.7k
| library_name
stringclasses 476
values | tags
sequencelengths 1
4.05k
| pipeline_tag
stringclasses 54
values | createdAt
timestamp[us, tz=UTC]date 2022-03-02 23:29:04
2025-05-24 18:26:04
| card
stringlengths 11
1.01M
|
|---|---|---|---|---|---|---|---|---|---|
tensorblock/TinyMistral-6x248M-Instruct-GGUF | tensorblock | 2025-04-21T00:28:49Z | 23 | 0 | null | [
"gguf",
"moe",
"TensorBlock",
"GGUF",
"en",
"dataset:Locutusque/hercules-v1.0",
"base_model:M4-ai/TinyMistral-6x248M-Instruct",
"base_model:quantized:M4-ai/TinyMistral-6x248M-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-16T20:16:08Z | ---
license: apache-2.0
datasets:
- Locutusque/hercules-v1.0
language:
- en
base_model: M4-ai/TinyMistral-6x248M-Instruct
inference:
parameters:
do_sample: true
temperature: 0.2
top_p: 0.14
top_k: 12
max_new_tokens: 250
repetition_penalty: 1.1
widget:
- text: '<|im_start|>user
Write me a Python program that calculates the factorial of n. <|im_end|>
<|im_start|>assistant
'
- text: An emerging clinical approach to treat substance abuse disorders involves
a form of cognitive-behavioral therapy whereby addicts learn to reduce their reactivity
to drug-paired stimuli through cue-exposure or extinction training. It is, however,
- text: '<|im_start|>user
How do I say hello in Spanish? <|im_end|>
<|im_start|>assistant
'
tags:
- moe
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## M4-ai/TinyMistral-6x248M-Instruct - GGUF
This repo contains GGUF format model files for [M4-ai/TinyMistral-6x248M-Instruct](https://huggingface.co/M4-ai/TinyMistral-6x248M-Instruct).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [TinyMistral-6x248M-Instruct-Q2_K.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q2_K.gguf) | Q2_K | 0.379 GB | smallest, significant quality loss - not recommended for most purposes |
| [TinyMistral-6x248M-Instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q3_K_S.gguf) | Q3_K_S | 0.445 GB | very small, high quality loss |
| [TinyMistral-6x248M-Instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q3_K_M.gguf) | Q3_K_M | 0.487 GB | very small, high quality loss |
| [TinyMistral-6x248M-Instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q3_K_L.gguf) | Q3_K_L | 0.527 GB | small, substantial quality loss |
| [TinyMistral-6x248M-Instruct-Q4_0.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q4_0.gguf) | Q4_0 | 0.574 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [TinyMistral-6x248M-Instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q4_K_S.gguf) | Q4_K_S | 0.577 GB | small, greater quality loss |
| [TinyMistral-6x248M-Instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q4_K_M.gguf) | Q4_K_M | 0.613 GB | medium, balanced quality - recommended |
| [TinyMistral-6x248M-Instruct-Q5_0.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q5_0.gguf) | Q5_0 | 0.695 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [TinyMistral-6x248M-Instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q5_K_S.gguf) | Q5_K_S | 0.695 GB | large, low quality loss - recommended |
| [TinyMistral-6x248M-Instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q5_K_M.gguf) | Q5_K_M | 0.715 GB | large, very low quality loss - recommended |
| [TinyMistral-6x248M-Instruct-Q6_K.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q6_K.gguf) | Q6_K | 0.824 GB | very large, extremely low quality loss |
| [TinyMistral-6x248M-Instruct-Q8_0.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-Instruct-GGUF/blob/main/TinyMistral-6x248M-Instruct-Q8_0.gguf) | Q8_0 | 1.067 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/TinyMistral-6x248M-Instruct-GGUF --include "TinyMistral-6x248M-Instruct-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/TinyMistral-6x248M-Instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/h2o-danube-1.8b-sft-GGUF | tensorblock | 2025-04-21T00:28:44Z | 23 | 0 | transformers | [
"transformers",
"gguf",
"gpt",
"llm",
"large language model",
"h2o-llmstudio",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:Open-Orca/OpenOrca",
"dataset:OpenAssistant/oasst2",
"dataset:HuggingFaceH4/ultrachat_200k",
"dataset:meta-math/MetaMathQA",
"base_model:h2oai/h2o-danube-1.8b-sft",
"base_model:quantized:h2oai/h2o-danube-1.8b-sft",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-16T19:19:47Z | ---
language:
- en
library_name: transformers
license: apache-2.0
tags:
- gpt
- llm
- large language model
- h2o-llmstudio
- TensorBlock
- GGUF
thumbnail: https://h2o.ai/etc.clientlibs/h2o/clientlibs/clientlib-site/resources/images/favicon.ico
datasets:
- Open-Orca/OpenOrca
- OpenAssistant/oasst2
- HuggingFaceH4/ultrachat_200k
- meta-math/MetaMathQA
widget:
- messages:
- role: user
content: Why is drinking water so healthy?
pipeline_tag: text-generation
base_model: h2oai/h2o-danube-1.8b-sft
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## h2oai/h2o-danube-1.8b-sft - GGUF
This repo contains GGUF format model files for [h2oai/h2o-danube-1.8b-sft](https://huggingface.co/h2oai/h2o-danube-1.8b-sft).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>{system_prompt}</s><|prompt|>{prompt}</s><|answer|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [h2o-danube-1.8b-sft-Q2_K.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q2_K.gguf) | Q2_K | 0.711 GB | smallest, significant quality loss - not recommended for most purposes |
| [h2o-danube-1.8b-sft-Q3_K_S.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q3_K_S.gguf) | Q3_K_S | 0.820 GB | very small, high quality loss |
| [h2o-danube-1.8b-sft-Q3_K_M.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q3_K_M.gguf) | Q3_K_M | 0.905 GB | very small, high quality loss |
| [h2o-danube-1.8b-sft-Q3_K_L.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q3_K_L.gguf) | Q3_K_L | 0.980 GB | small, substantial quality loss |
| [h2o-danube-1.8b-sft-Q4_0.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q4_0.gguf) | Q4_0 | 1.052 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [h2o-danube-1.8b-sft-Q4_K_S.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q4_K_S.gguf) | Q4_K_S | 1.060 GB | small, greater quality loss |
| [h2o-danube-1.8b-sft-Q4_K_M.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q4_K_M.gguf) | Q4_K_M | 1.112 GB | medium, balanced quality - recommended |
| [h2o-danube-1.8b-sft-Q5_0.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q5_0.gguf) | Q5_0 | 1.271 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [h2o-danube-1.8b-sft-Q5_K_S.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q5_K_S.gguf) | Q5_K_S | 1.271 GB | large, low quality loss - recommended |
| [h2o-danube-1.8b-sft-Q5_K_M.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q5_K_M.gguf) | Q5_K_M | 1.302 GB | large, very low quality loss - recommended |
| [h2o-danube-1.8b-sft-Q6_K.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q6_K.gguf) | Q6_K | 1.503 GB | very large, extremely low quality loss |
| [h2o-danube-1.8b-sft-Q8_0.gguf](https://huggingface.co/tensorblock/h2o-danube-1.8b-sft-GGUF/blob/main/h2o-danube-1.8b-sft-Q8_0.gguf) | Q8_0 | 1.947 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/h2o-danube-1.8b-sft-GGUF --include "h2o-danube-1.8b-sft-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/h2o-danube-1.8b-sft-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Instruct_Yi-6B_Dolly15K-GGUF | tensorblock | 2025-04-21T00:28:39Z | 26 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"dataset:databricks/databricks-dolly-15k",
"base_model:HenryJJ/Instruct_Yi-6B_Dolly15K",
"base_model:quantized:HenryJJ/Instruct_Yi-6B_Dolly15K",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-16T18:12:08Z | ---
license: apache-2.0
datasets:
- databricks/databricks-dolly-15k
tags:
- TensorBlock
- GGUF
base_model: HenryJJ/Instruct_Yi-6B_Dolly15K
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## HenryJJ/Instruct_Yi-6B_Dolly15K - GGUF
This repo contains GGUF format model files for [HenryJJ/Instruct_Yi-6B_Dolly15K](https://huggingface.co/HenryJJ/Instruct_Yi-6B_Dolly15K).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Instruct_Yi-6B_Dolly15K-Q2_K.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q2_K.gguf) | Q2_K | 2.337 GB | smallest, significant quality loss - not recommended for most purposes |
| [Instruct_Yi-6B_Dolly15K-Q3_K_S.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q3_K_S.gguf) | Q3_K_S | 2.709 GB | very small, high quality loss |
| [Instruct_Yi-6B_Dolly15K-Q3_K_M.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q3_K_M.gguf) | Q3_K_M | 2.993 GB | very small, high quality loss |
| [Instruct_Yi-6B_Dolly15K-Q3_K_L.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q3_K_L.gguf) | Q3_K_L | 3.237 GB | small, substantial quality loss |
| [Instruct_Yi-6B_Dolly15K-Q4_0.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q4_0.gguf) | Q4_0 | 3.479 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Instruct_Yi-6B_Dolly15K-Q4_K_S.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q4_K_S.gguf) | Q4_K_S | 3.503 GB | small, greater quality loss |
| [Instruct_Yi-6B_Dolly15K-Q4_K_M.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q4_K_M.gguf) | Q4_K_M | 3.674 GB | medium, balanced quality - recommended |
| [Instruct_Yi-6B_Dolly15K-Q5_0.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q5_0.gguf) | Q5_0 | 4.204 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Instruct_Yi-6B_Dolly15K-Q5_K_S.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q5_K_S.gguf) | Q5_K_S | 4.204 GB | large, low quality loss - recommended |
| [Instruct_Yi-6B_Dolly15K-Q5_K_M.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q5_K_M.gguf) | Q5_K_M | 4.304 GB | large, very low quality loss - recommended |
| [Instruct_Yi-6B_Dolly15K-Q6_K.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q6_K.gguf) | Q6_K | 4.974 GB | very large, extremely low quality loss |
| [Instruct_Yi-6B_Dolly15K-Q8_0.gguf](https://huggingface.co/tensorblock/Instruct_Yi-6B_Dolly15K-GGUF/blob/main/Instruct_Yi-6B_Dolly15K-Q8_0.gguf) | Q8_0 | 6.442 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Instruct_Yi-6B_Dolly15K-GGUF --include "Instruct_Yi-6B_Dolly15K-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Instruct_Yi-6B_Dolly15K-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Alpaca-tuned-gpt2-GGUF | tensorblock | 2025-04-21T00:28:37Z | 68 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"en",
"dataset:tatsu-lab/alpaca",
"base_model:LordNoah/Alpaca-tuned-gpt2",
"base_model:quantized:LordNoah/Alpaca-tuned-gpt2",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-16T17:33:41Z | ---
license: apache-2.0
datasets:
- tatsu-lab/alpaca
language:
- en
base_model: LordNoah/Alpaca-tuned-gpt2
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## LordNoah/Alpaca-tuned-gpt2 - GGUF
This repo contains GGUF format model files for [LordNoah/Alpaca-tuned-gpt2](https://huggingface.co/LordNoah/Alpaca-tuned-gpt2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Alpaca-tuned-gpt2-Q2_K.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q2_K.gguf) | Q2_K | 0.346 GB | smallest, significant quality loss - not recommended for most purposes |
| [Alpaca-tuned-gpt2-Q3_K_S.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q3_K_S.gguf) | Q3_K_S | 0.394 GB | very small, high quality loss |
| [Alpaca-tuned-gpt2-Q3_K_M.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q3_K_M.gguf) | Q3_K_M | 0.458 GB | very small, high quality loss |
| [Alpaca-tuned-gpt2-Q3_K_L.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q3_K_L.gguf) | Q3_K_L | 0.494 GB | small, substantial quality loss |
| [Alpaca-tuned-gpt2-Q4_0.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q4_0.gguf) | Q4_0 | 0.497 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Alpaca-tuned-gpt2-Q4_K_S.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q4_K_S.gguf) | Q4_K_S | 0.500 GB | small, greater quality loss |
| [Alpaca-tuned-gpt2-Q4_K_M.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q4_K_M.gguf) | Q4_K_M | 0.549 GB | medium, balanced quality - recommended |
| [Alpaca-tuned-gpt2-Q5_0.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q5_0.gguf) | Q5_0 | 0.593 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Alpaca-tuned-gpt2-Q5_K_S.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q5_K_S.gguf) | Q5_K_S | 0.593 GB | large, low quality loss - recommended |
| [Alpaca-tuned-gpt2-Q5_K_M.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q5_K_M.gguf) | Q5_K_M | 0.632 GB | large, very low quality loss - recommended |
| [Alpaca-tuned-gpt2-Q6_K.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q6_K.gguf) | Q6_K | 0.696 GB | very large, extremely low quality loss |
| [Alpaca-tuned-gpt2-Q8_0.gguf](https://huggingface.co/tensorblock/Alpaca-tuned-gpt2-GGUF/blob/main/Alpaca-tuned-gpt2-Q8_0.gguf) | Q8_0 | 0.898 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Alpaca-tuned-gpt2-GGUF --include "Alpaca-tuned-gpt2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Alpaca-tuned-gpt2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF | tensorblock | 2025-04-21T00:28:36Z | 24 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"en",
"dataset:Intel/orca_dpo_pairs",
"base_model:sreeramajay/TinyLlama-1.1B-orca-v1.0",
"base_model:quantized:sreeramajay/TinyLlama-1.1B-orca-v1.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-16T17:32:08Z | ---
license: apache-2.0
datasets:
- Intel/orca_dpo_pairs
language:
- en
base_model: sreeramajay/TinyLlama-1.1B-orca-v1.0
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## sreeramajay/TinyLlama-1.1B-orca-v1.0 - GGUF
This repo contains GGUF format model files for [sreeramajay/TinyLlama-1.1B-orca-v1.0](https://huggingface.co/sreeramajay/TinyLlama-1.1B-orca-v1.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [TinyLlama-1.1B-orca-v1.0-Q2_K.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q2_K.gguf) | Q2_K | 0.432 GB | smallest, significant quality loss - not recommended for most purposes |
| [TinyLlama-1.1B-orca-v1.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q3_K_S.gguf) | Q3_K_S | 0.499 GB | very small, high quality loss |
| [TinyLlama-1.1B-orca-v1.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q3_K_M.gguf) | Q3_K_M | 0.548 GB | very small, high quality loss |
| [TinyLlama-1.1B-orca-v1.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q3_K_L.gguf) | Q3_K_L | 0.592 GB | small, substantial quality loss |
| [TinyLlama-1.1B-orca-v1.0-Q4_0.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q4_0.gguf) | Q4_0 | 0.637 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [TinyLlama-1.1B-orca-v1.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q4_K_S.gguf) | Q4_K_S | 0.640 GB | small, greater quality loss |
| [TinyLlama-1.1B-orca-v1.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q4_K_M.gguf) | Q4_K_M | 0.668 GB | medium, balanced quality - recommended |
| [TinyLlama-1.1B-orca-v1.0-Q5_0.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q5_0.gguf) | Q5_0 | 0.766 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [TinyLlama-1.1B-orca-v1.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q5_K_S.gguf) | Q5_K_S | 0.766 GB | large, low quality loss - recommended |
| [TinyLlama-1.1B-orca-v1.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q5_K_M.gguf) | Q5_K_M | 0.782 GB | large, very low quality loss - recommended |
| [TinyLlama-1.1B-orca-v1.0-Q6_K.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q6_K.gguf) | Q6_K | 0.903 GB | very large, extremely low quality loss |
| [TinyLlama-1.1B-orca-v1.0-Q8_0.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF/blob/main/TinyLlama-1.1B-orca-v1.0-Q8_0.gguf) | Q8_0 | 1.170 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF --include "TinyLlama-1.1B-orca-v1.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/TinyLlama-1.1B-orca-v1.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF | tensorblock | 2025-04-21T00:28:34Z | 15 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:kimwooglae/AISquare-Instruct-SOLAR-10.7b-v0.5.31",
"base_model:quantized:kimwooglae/AISquare-Instruct-SOLAR-10.7b-v0.5.31",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-16T17:24:16Z | ---
language:
- en
pipeline_tag: text-generation
license: cc-by-nc-4.0
tags:
- TensorBlock
- GGUF
base_model: kimwooglae/AISquare-Instruct-SOLAR-10.7b-v0.5.31
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## kimwooglae/AISquare-Instruct-SOLAR-10.7b-v0.5.31 - GGUF
This repo contains GGUF format model files for [kimwooglae/AISquare-Instruct-SOLAR-10.7b-v0.5.31](https://huggingface.co/kimwooglae/AISquare-Instruct-SOLAR-10.7b-v0.5.31).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q2_K.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q3_K_S.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q3_K_M.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q3_K_L.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q4_0.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q4_K_S.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q4_K_M.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q5_0.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q5_K_S.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q5_K_M.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q6_K.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q8_0.gguf](https://huggingface.co/tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF/blob/main/AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF --include "AISquare-Instruct-SOLAR-10.7b-v0.5.31-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/AISquare-Instruct-SOLAR-10.7b-v0.5.31-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/2xbagel-dpo-34b-v0.2-GGUF | tensorblock | 2025-04-21T00:28:27Z | 38 | 0 | null | [
"gguf",
"merge",
"TensorBlock",
"GGUF",
"base_model:TeeZee/2xbagel-dpo-34b-v0.2",
"base_model:quantized:TeeZee/2xbagel-dpo-34b-v0.2",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-16T12:30:39Z | ---
tags:
- merge
- TensorBlock
- GGUF
license: apache-2.0
base_model: TeeZee/2xbagel-dpo-34b-v0.2
model-index:
- name: 2xbagel-dpo-34b-v0.2
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 65.27
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/2xbagel-dpo-34b-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 79.35
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/2xbagel-dpo-34b-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 73.64
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/2xbagel-dpo-34b-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 67.15
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/2xbagel-dpo-34b-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 76.4
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/2xbagel-dpo-34b-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 2.12
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/2xbagel-dpo-34b-v0.2
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## TeeZee/2xbagel-dpo-34b-v0.2 - GGUF
This repo contains GGUF format model files for [TeeZee/2xbagel-dpo-34b-v0.2](https://huggingface.co/TeeZee/2xbagel-dpo-34b-v0.2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
[INST] <<SYS>>
{system_prompt}
<</SYS>>
{prompt} [/INST]
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [2xbagel-dpo-34b-v0.2-Q2_K.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q2_K.gguf) | Q2_K | 21.023 GB | smallest, significant quality loss - not recommended for most purposes |
| [2xbagel-dpo-34b-v0.2-Q3_K_S.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q3_K_S.gguf) | Q3_K_S | 24.551 GB | very small, high quality loss |
| [2xbagel-dpo-34b-v0.2-Q3_K_M.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q3_K_M.gguf) | Q3_K_M | 27.392 GB | very small, high quality loss |
| [2xbagel-dpo-34b-v0.2-Q3_K_L.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q3_K_L.gguf) | Q3_K_L | 29.849 GB | small, substantial quality loss |
| [2xbagel-dpo-34b-v0.2-Q4_0.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q4_0.gguf) | Q4_0 | 32.020 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [2xbagel-dpo-34b-v0.2-Q4_K_S.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q4_K_S.gguf) | Q4_K_S | 32.244 GB | small, greater quality loss |
| [2xbagel-dpo-34b-v0.2-Q4_K_M.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q4_K_M.gguf) | Q4_K_M | 34.007 GB | medium, balanced quality - recommended |
| [2xbagel-dpo-34b-v0.2-Q5_0.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q5_0.gguf) | Q5_0 | 39.051 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [2xbagel-dpo-34b-v0.2-Q5_K_S.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q5_K_S.gguf) | Q5_K_S | 39.051 GB | large, low quality loss - recommended |
| [2xbagel-dpo-34b-v0.2-Q5_K_M.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q5_K_M.gguf) | Q5_K_M | 40.074 GB | large, very low quality loss - recommended |
| [2xbagel-dpo-34b-v0.2-Q6_K.gguf](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q6_K.gguf) | Q6_K | 46.520 GB | very large, extremely low quality loss |
| [2xbagel-dpo-34b-v0.2-Q8_0](https://huggingface.co/tensorblock/2xbagel-dpo-34b-v0.2-GGUF/blob/main/2xbagel-dpo-34b-v0.2-Q8_0) | Q8_0 | 2.914 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/2xbagel-dpo-34b-v0.2-GGUF --include "2xbagel-dpo-34b-v0.2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/2xbagel-dpo-34b-v0.2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/openchat-3.5-1210-starling-slerp-GGUF | tensorblock | 2025-04-21T00:28:24Z | 27 | 0 | null | [
"gguf",
"merge",
"TensorBlock",
"GGUF",
"en",
"base_model:SanjiWatsuki/openchat-3.5-1210-starling-slerp",
"base_model:quantized:SanjiWatsuki/openchat-3.5-1210-starling-slerp",
"license:cc-by-4.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-16T11:17:46Z | ---
license: cc-by-4.0
language:
- en
tags:
- merge
- TensorBlock
- GGUF
base_model: SanjiWatsuki/openchat-3.5-1210-starling-slerp
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## SanjiWatsuki/openchat-3.5-1210-starling-slerp - GGUF
This repo contains GGUF format model files for [SanjiWatsuki/openchat-3.5-1210-starling-slerp](https://huggingface.co/SanjiWatsuki/openchat-3.5-1210-starling-slerp).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>GPT4 Correct System: {system_prompt}<|end_of_turn|>GPT4 Correct User: {prompt}<|end_of_turn|>GPT4 Correct Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [openchat-3.5-1210-starling-slerp-Q2_K.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [openchat-3.5-1210-starling-slerp-Q3_K_S.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [openchat-3.5-1210-starling-slerp-Q3_K_M.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [openchat-3.5-1210-starling-slerp-Q3_K_L.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [openchat-3.5-1210-starling-slerp-Q4_0.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [openchat-3.5-1210-starling-slerp-Q4_K_S.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [openchat-3.5-1210-starling-slerp-Q4_K_M.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [openchat-3.5-1210-starling-slerp-Q5_0.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [openchat-3.5-1210-starling-slerp-Q5_K_S.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [openchat-3.5-1210-starling-slerp-Q5_K_M.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [openchat-3.5-1210-starling-slerp-Q6_K.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [openchat-3.5-1210-starling-slerp-Q8_0.gguf](https://huggingface.co/tensorblock/openchat-3.5-1210-starling-slerp-GGUF/blob/main/openchat-3.5-1210-starling-slerp-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/openchat-3.5-1210-starling-slerp-GGUF --include "openchat-3.5-1210-starling-slerp-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/openchat-3.5-1210-starling-slerp-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF | tensorblock | 2025-04-21T00:28:22Z | 28 | 0 | null | [
"gguf",
"merge",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:Locutusque/inst_mix_v2_top_100k",
"base_model:Locutusque/LocutusqueXFelladrin-TinyMistral248M-Instruct",
"base_model:quantized:Locutusque/LocutusqueXFelladrin-TinyMistral248M-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-16T11:12:42Z | ---
license: apache-2.0
datasets:
- Locutusque/inst_mix_v2_top_100k
language:
- en
pipeline_tag: text-generation
widget:
- text: '<|USER|> Design a Neo4j database and Cypher function snippet to Display Extreme
Dental hygiene: Using Mouthwash for Analysis for Beginners. Implement if/else
or switch/case statements to handle different conditions related to the Consent.
Provide detailed comments explaining your control flow and the reasoning behind
each decision. <|ASSISTANT|> '
- text: '<|USER|> Write me a story about a magical place. <|ASSISTANT|> '
- text: '<|USER|> Write me an essay about the life of George Washington <|ASSISTANT|> '
- text: '<|USER|> Solve the following equation 2x + 10 = 20 <|ASSISTANT|> '
- text: '<|USER|> Craft me a list of some nice places to visit around the world. <|ASSISTANT|> '
- text: '<|USER|> How to manage a lazy employee: Address the employee verbally. Don''t
allow an employee''s laziness or lack of enthusiasm to become a recurring issue.
Tell the employee you''re hoping to speak with them about workplace expectations
and performance, and schedule a time to sit down together. Question: To manage
a lazy employee, it is suggested to talk to the employee. True, False, or Neither?
<|ASSISTANT|> '
inference:
parameters:
temperature: 0.5
do_sample: true
top_p: 0.5
top_k: 30
max_new_tokens: 250
repetition_penalty: 1.15
tags:
- merge
- TensorBlock
- GGUF
base_model: Locutusque/LocutusqueXFelladrin-TinyMistral248M-Instruct
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Locutusque/LocutusqueXFelladrin-TinyMistral248M-Instruct - GGUF
This repo contains GGUF format model files for [Locutusque/LocutusqueXFelladrin-TinyMistral248M-Instruct](https://huggingface.co/Locutusque/LocutusqueXFelladrin-TinyMistral248M-Instruct).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q2_K.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q2_K.gguf) | Q2_K | 0.105 GB | smallest, significant quality loss - not recommended for most purposes |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q3_K_S.gguf) | Q3_K_S | 0.120 GB | very small, high quality loss |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q3_K_M.gguf) | Q3_K_M | 0.129 GB | very small, high quality loss |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q3_K_L.gguf) | Q3_K_L | 0.137 GB | small, substantial quality loss |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q4_0.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q4_0.gguf) | Q4_0 | 0.149 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q4_K_S.gguf) | Q4_K_S | 0.149 GB | small, greater quality loss |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q4_K_M.gguf) | Q4_K_M | 0.156 GB | medium, balanced quality - recommended |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q5_0.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q5_0.gguf) | Q5_0 | 0.176 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q5_K_S.gguf) | Q5_K_S | 0.176 GB | large, low quality loss - recommended |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q5_K_M.gguf) | Q5_K_M | 0.179 GB | large, very low quality loss - recommended |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q6_K.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q6_K.gguf) | Q6_K | 0.204 GB | very large, extremely low quality loss |
| [LocutusqueXFelladrin-TinyMistral248M-Instruct-Q8_0.gguf](https://huggingface.co/tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF/blob/main/LocutusqueXFelladrin-TinyMistral248M-Instruct-Q8_0.gguf) | Q8_0 | 0.264 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF --include "LocutusqueXFelladrin-TinyMistral248M-Instruct-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/LocutusqueXFelladrin-TinyMistral248M-Instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF | tensorblock | 2025-04-21T00:28:16Z | 41 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:ewqr2130/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2",
"base_model:quantized:ewqr2130/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-16T10:09:23Z | ---
license: apache-2.0
tags:
- TensorBlock
- GGUF
base_model: ewqr2130/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## ewqr2130/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2 - GGUF
This repo contains GGUF format model files for [ewqr2130/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2](https://huggingface.co/ewqr2130/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q2_K.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q3_K_S.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q3_K_M.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q3_K_L.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q4_0.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q4_K_S.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q4_K_M.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q5_0.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q5_K_S.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q5_K_M.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q6_K.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q8_0.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF/blob/main/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF --include "alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/alignment-handbook-zephyr-7b-sft-full-dpo-5e7-cont2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/bun_mistral_7b_v2-GGUF | tensorblock | 2025-04-21T00:28:15Z | 38 | 0 | null | [
"gguf",
"CoT",
"TensorBlock",
"GGUF",
"en",
"base_model:aloobun/bun_mistral_7b_v2",
"base_model:quantized:aloobun/bun_mistral_7b_v2",
"license:cc",
"endpoints_compatible",
"region:us"
] | null | 2024-12-16T09:49:32Z | ---
language:
- en
tags:
- CoT
- TensorBlock
- GGUF
license: cc
base_model: aloobun/bun_mistral_7b_v2
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## aloobun/bun_mistral_7b_v2 - GGUF
This repo contains GGUF format model files for [aloobun/bun_mistral_7b_v2](https://huggingface.co/aloobun/bun_mistral_7b_v2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [bun_mistral_7b_v2-Q2_K.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [bun_mistral_7b_v2-Q3_K_S.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [bun_mistral_7b_v2-Q3_K_M.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [bun_mistral_7b_v2-Q3_K_L.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [bun_mistral_7b_v2-Q4_0.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [bun_mistral_7b_v2-Q4_K_S.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [bun_mistral_7b_v2-Q4_K_M.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [bun_mistral_7b_v2-Q5_0.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [bun_mistral_7b_v2-Q5_K_S.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [bun_mistral_7b_v2-Q5_K_M.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [bun_mistral_7b_v2-Q6_K.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [bun_mistral_7b_v2-Q8_0.gguf](https://huggingface.co/tensorblock/bun_mistral_7b_v2-GGUF/blob/main/bun_mistral_7b_v2-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/bun_mistral_7b_v2-GGUF --include "bun_mistral_7b_v2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/bun_mistral_7b_v2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF | tensorblock | 2025-04-21T00:28:00Z | 16 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:AIChenKai/TinyLlama-1.1B-Chat-v1.0-x2-MoE",
"base_model:quantized:AIChenKai/TinyLlama-1.1B-Chat-v1.0-x2-MoE",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-16T03:40:58Z | ---
license: apache-2.0
tags:
- TensorBlock
- GGUF
base_model: AIChenKai/TinyLlama-1.1B-Chat-v1.0-x2-MoE
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## AIChenKai/TinyLlama-1.1B-Chat-v1.0-x2-MoE - GGUF
This repo contains GGUF format model files for [AIChenKai/TinyLlama-1.1B-Chat-v1.0-x2-MoE](https://huggingface.co/AIChenKai/TinyLlama-1.1B-Chat-v1.0-x2-MoE).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q2_K.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q2_K.gguf) | Q2_K | 0.708 GB | smallest, significant quality loss - not recommended for most purposes |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q3_K_S.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q3_K_S.gguf) | Q3_K_S | 0.827 GB | very small, high quality loss |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q3_K_M.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q3_K_M.gguf) | Q3_K_M | 0.911 GB | very small, high quality loss |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q3_K_L.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q3_K_L.gguf) | Q3_K_L | 0.984 GB | small, substantial quality loss |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q4_0.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q4_0.gguf) | Q4_0 | 1.065 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q4_K_S.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q4_K_S.gguf) | Q4_K_S | 1.071 GB | small, greater quality loss |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q4_K_M.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q4_K_M.gguf) | Q4_K_M | 1.126 GB | medium, balanced quality - recommended |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q5_0.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q5_0.gguf) | Q5_0 | 1.290 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q5_K_S.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q5_K_S.gguf) | Q5_K_S | 1.290 GB | large, low quality loss - recommended |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q5_K_M.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q5_K_M.gguf) | Q5_K_M | 1.321 GB | large, very low quality loss - recommended |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q6_K.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q6_K.gguf) | Q6_K | 1.528 GB | very large, extremely low quality loss |
| [TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q8_0.gguf](https://huggingface.co/tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF/blob/main/TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q8_0.gguf) | Q8_0 | 1.979 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF --include "TinyLlama-1.1B-Chat-v1.0-x2-MoE-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/TinyLlama-1.1B-Chat-v1.0-x2-MoE-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/TinyMistral-6x248M-GGUF | tensorblock | 2025-04-21T00:27:56Z | 42 | 0 | null | [
"gguf",
"moe",
"frankenmoe",
"merge",
"mergekit",
"lazymergekit",
"Locutusque/TinyMistral-248M-v2",
"Locutusque/TinyMistral-248M-v2.5",
"Locutusque/TinyMistral-248M-v2.5-Instruct",
"jtatman/tinymistral-v2-pycoder-instruct-248m",
"Felladrin/TinyMistral-248M-SFT-v4",
"Locutusque/TinyMistral-248M-v2-Instruct",
"TensorBlock",
"GGUF",
"dataset:nampdn-ai/mini-peS2o",
"base_model:M4-ai/TinyMistral-6x248M",
"base_model:quantized:M4-ai/TinyMistral-6x248M",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-16T03:30:59Z | ---
license: apache-2.0
tags:
- moe
- frankenmoe
- merge
- mergekit
- lazymergekit
- Locutusque/TinyMistral-248M-v2
- Locutusque/TinyMistral-248M-v2.5
- Locutusque/TinyMistral-248M-v2.5-Instruct
- jtatman/tinymistral-v2-pycoder-instruct-248m
- Felladrin/TinyMistral-248M-SFT-v4
- Locutusque/TinyMistral-248M-v2-Instruct
- TensorBlock
- GGUF
base_model: M4-ai/TinyMistral-6x248M
inference:
parameters:
do_sample: true
temperature: 0.2
top_p: 0.14
top_k: 12
max_new_tokens: 250
repetition_penalty: 1.15
widget:
- text: '<|im_start|>user
Write me a Python program that calculates the factorial of n. <|im_end|>
<|im_start|>assistant
'
- text: An emerging clinical approach to treat substance abuse disorders involves
a form of cognitive-behavioral therapy whereby addicts learn to reduce their reactivity
to drug-paired stimuli through cue-exposure or extinction training. It is, however,
datasets:
- nampdn-ai/mini-peS2o
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## M4-ai/TinyMistral-6x248M - GGUF
This repo contains GGUF format model files for [M4-ai/TinyMistral-6x248M](https://huggingface.co/M4-ai/TinyMistral-6x248M).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [TinyMistral-6x248M-Q2_K.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q2_K.gguf) | Q2_K | 0.379 GB | smallest, significant quality loss - not recommended for most purposes |
| [TinyMistral-6x248M-Q3_K_S.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q3_K_S.gguf) | Q3_K_S | 0.445 GB | very small, high quality loss |
| [TinyMistral-6x248M-Q3_K_M.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q3_K_M.gguf) | Q3_K_M | 0.487 GB | very small, high quality loss |
| [TinyMistral-6x248M-Q3_K_L.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q3_K_L.gguf) | Q3_K_L | 0.527 GB | small, substantial quality loss |
| [TinyMistral-6x248M-Q4_0.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q4_0.gguf) | Q4_0 | 0.574 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [TinyMistral-6x248M-Q4_K_S.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q4_K_S.gguf) | Q4_K_S | 0.577 GB | small, greater quality loss |
| [TinyMistral-6x248M-Q4_K_M.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q4_K_M.gguf) | Q4_K_M | 0.613 GB | medium, balanced quality - recommended |
| [TinyMistral-6x248M-Q5_0.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q5_0.gguf) | Q5_0 | 0.695 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [TinyMistral-6x248M-Q5_K_S.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q5_K_S.gguf) | Q5_K_S | 0.695 GB | large, low quality loss - recommended |
| [TinyMistral-6x248M-Q5_K_M.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q5_K_M.gguf) | Q5_K_M | 0.715 GB | large, very low quality loss - recommended |
| [TinyMistral-6x248M-Q6_K.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q6_K.gguf) | Q6_K | 0.824 GB | very large, extremely low quality loss |
| [TinyMistral-6x248M-Q8_0.gguf](https://huggingface.co/tensorblock/TinyMistral-6x248M-GGUF/blob/main/TinyMistral-6x248M-Q8_0.gguf) | Q8_0 | 1.067 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/TinyMistral-6x248M-GGUF --include "TinyMistral-6x248M-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/TinyMistral-6x248M-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/internlm2-1_8b-GGUF | tensorblock | 2025-04-21T00:27:52Z | 28 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"base_model:internlm/internlm2-1_8b",
"base_model:quantized:internlm/internlm2-1_8b",
"license:other",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-16T03:20:51Z | ---
pipeline_tag: text-generation
license: other
base_model: internlm/internlm2-1_8b
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## internlm/internlm2-1_8b - GGUF
This repo contains GGUF format model files for [internlm/internlm2-1_8b](https://huggingface.co/internlm/internlm2-1_8b).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [internlm2-1_8b-Q2_K.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q2_K.gguf) | Q2_K | 0.772 GB | smallest, significant quality loss - not recommended for most purposes |
| [internlm2-1_8b-Q3_K_S.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q3_K_S.gguf) | Q3_K_S | 0.888 GB | very small, high quality loss |
| [internlm2-1_8b-Q3_K_M.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q3_K_M.gguf) | Q3_K_M | 0.964 GB | very small, high quality loss |
| [internlm2-1_8b-Q3_K_L.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q3_K_L.gguf) | Q3_K_L | 1.031 GB | small, substantial quality loss |
| [internlm2-1_8b-Q4_0.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q4_0.gguf) | Q4_0 | 1.114 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [internlm2-1_8b-Q4_K_S.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q4_K_S.gguf) | Q4_K_S | 1.121 GB | small, greater quality loss |
| [internlm2-1_8b-Q4_K_M.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q4_K_M.gguf) | Q4_K_M | 1.172 GB | medium, balanced quality - recommended |
| [internlm2-1_8b-Q5_0.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q5_0.gguf) | Q5_0 | 1.326 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [internlm2-1_8b-Q5_K_S.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q5_K_S.gguf) | Q5_K_S | 1.326 GB | large, low quality loss - recommended |
| [internlm2-1_8b-Q5_K_M.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q5_K_M.gguf) | Q5_K_M | 1.356 GB | large, very low quality loss - recommended |
| [internlm2-1_8b-Q6_K.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q6_K.gguf) | Q6_K | 1.552 GB | very large, extremely low quality loss |
| [internlm2-1_8b-Q8_0.gguf](https://huggingface.co/tensorblock/internlm2-1_8b-GGUF/blob/main/internlm2-1_8b-Q8_0.gguf) | Q8_0 | 2.010 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/internlm2-1_8b-GGUF --include "internlm2-1_8b-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/internlm2-1_8b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF | tensorblock | 2025-04-21T00:27:49Z | 92 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:ewqr2130/alignment-handbook-zephyr-7b_ppo_5e7step_102",
"base_model:quantized:ewqr2130/alignment-handbook-zephyr-7b_ppo_5e7step_102",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-16T03:02:12Z | ---
license: apache-2.0
tags:
- TensorBlock
- GGUF
base_model: ewqr2130/alignment-handbook-zephyr-7b_ppo_5e7step_102
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## ewqr2130/alignment-handbook-zephyr-7b_ppo_5e7step_102 - GGUF
This repo contains GGUF format model files for [ewqr2130/alignment-handbook-zephyr-7b_ppo_5e7step_102](https://huggingface.co/ewqr2130/alignment-handbook-zephyr-7b_ppo_5e7step_102).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q2_K.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q3_K_S.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q3_K_M.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q3_K_L.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q4_0.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q4_K_S.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q4_K_M.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q5_0.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q5_K_S.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q5_K_M.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q6_K.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [alignment-handbook-zephyr-7b_ppo_5e7step_102-Q8_0.gguf](https://huggingface.co/tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF/blob/main/alignment-handbook-zephyr-7b_ppo_5e7step_102-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF --include "alignment-handbook-zephyr-7b_ppo_5e7step_102-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/alignment-handbook-zephyr-7b_ppo_5e7step_102-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Swallow-13b-NVE-hf-GGUF | tensorblock | 2025-04-21T00:27:44Z | 97 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"ja",
"base_model:tokyotech-llm/Swallow-13b-NVE-hf",
"base_model:quantized:tokyotech-llm/Swallow-13b-NVE-hf",
"license:llama2",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-16T01:50:57Z | ---
language:
- en
- ja
library_name: transformers
pipeline_tag: text-generation
license: llama2
model_type: llama
tags:
- TensorBlock
- GGUF
base_model: tokyotech-llm/Swallow-13b-NVE-hf
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## tokyotech-llm/Swallow-13b-NVE-hf - GGUF
This repo contains GGUF format model files for [tokyotech-llm/Swallow-13b-NVE-hf](https://huggingface.co/tokyotech-llm/Swallow-13b-NVE-hf).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Swallow-13b-NVE-hf-Q2_K.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q2_K.gguf) | Q2_K | 4.854 GB | smallest, significant quality loss - not recommended for most purposes |
| [Swallow-13b-NVE-hf-Q3_K_S.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q3_K_S.gguf) | Q3_K_S | 5.659 GB | very small, high quality loss |
| [Swallow-13b-NVE-hf-Q3_K_M.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q3_K_M.gguf) | Q3_K_M | 6.338 GB | very small, high quality loss |
| [Swallow-13b-NVE-hf-Q3_K_L.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q3_K_L.gguf) | Q3_K_L | 6.930 GB | small, substantial quality loss |
| [Swallow-13b-NVE-hf-Q4_0.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q4_0.gguf) | Q4_0 | 7.366 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Swallow-13b-NVE-hf-Q4_K_S.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q4_K_S.gguf) | Q4_K_S | 7.423 GB | small, greater quality loss |
| [Swallow-13b-NVE-hf-Q4_K_M.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q4_K_M.gguf) | Q4_K_M | 7.866 GB | medium, balanced quality - recommended |
| [Swallow-13b-NVE-hf-Q5_0.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q5_0.gguf) | Q5_0 | 8.972 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Swallow-13b-NVE-hf-Q5_K_S.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q5_K_S.gguf) | Q5_K_S | 8.972 GB | large, low quality loss - recommended |
| [Swallow-13b-NVE-hf-Q5_K_M.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q5_K_M.gguf) | Q5_K_M | 9.230 GB | large, very low quality loss - recommended |
| [Swallow-13b-NVE-hf-Q6_K.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q6_K.gguf) | Q6_K | 10.679 GB | very large, extremely low quality loss |
| [Swallow-13b-NVE-hf-Q8_0.gguf](https://huggingface.co/tensorblock/Swallow-13b-NVE-hf-GGUF/blob/main/Swallow-13b-NVE-hf-Q8_0.gguf) | Q8_0 | 13.831 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Swallow-13b-NVE-hf-GGUF --include "Swallow-13b-NVE-hf-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Swallow-13b-NVE-hf-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF | tensorblock | 2025-04-21T00:27:43Z | 101 | 0 | null | [
"gguf",
"llama2",
"TensorBlock",
"GGUF",
"text-generation",
"ko",
"base_model:AIdenU/LLAMA-2-13b-ko-Y24_v2.0",
"base_model:quantized:AIdenU/LLAMA-2-13b-ko-Y24_v2.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-16T01:28:08Z | ---
license: apache-2.0
language:
- ko
pipeline_tag: text-generation
tags:
- llama2
- TensorBlock
- GGUF
base_model: AIdenU/LLAMA-2-13b-ko-Y24_v2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## AIdenU/LLAMA-2-13b-ko-Y24_v2.0 - GGUF
This repo contains GGUF format model files for [AIdenU/LLAMA-2-13b-ko-Y24_v2.0](https://huggingface.co/AIdenU/LLAMA-2-13b-ko-Y24_v2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [LLAMA-2-13b-ko-Y24_v2.0-Q2_K.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q2_K.gguf) | Q2_K | 4.854 GB | smallest, significant quality loss - not recommended for most purposes |
| [LLAMA-2-13b-ko-Y24_v2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q3_K_S.gguf) | Q3_K_S | 5.659 GB | very small, high quality loss |
| [LLAMA-2-13b-ko-Y24_v2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q3_K_M.gguf) | Q3_K_M | 6.338 GB | very small, high quality loss |
| [LLAMA-2-13b-ko-Y24_v2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q3_K_L.gguf) | Q3_K_L | 6.930 GB | small, substantial quality loss |
| [LLAMA-2-13b-ko-Y24_v2.0-Q4_0.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q4_0.gguf) | Q4_0 | 7.366 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [LLAMA-2-13b-ko-Y24_v2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q4_K_S.gguf) | Q4_K_S | 7.423 GB | small, greater quality loss |
| [LLAMA-2-13b-ko-Y24_v2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q4_K_M.gguf) | Q4_K_M | 7.866 GB | medium, balanced quality - recommended |
| [LLAMA-2-13b-ko-Y24_v2.0-Q5_0.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q5_0.gguf) | Q5_0 | 8.972 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [LLAMA-2-13b-ko-Y24_v2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q5_K_S.gguf) | Q5_K_S | 8.972 GB | large, low quality loss - recommended |
| [LLAMA-2-13b-ko-Y24_v2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q5_K_M.gguf) | Q5_K_M | 9.230 GB | large, very low quality loss - recommended |
| [LLAMA-2-13b-ko-Y24_v2.0-Q6_K.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q6_K.gguf) | Q6_K | 10.679 GB | very large, extremely low quality loss |
| [LLAMA-2-13b-ko-Y24_v2.0-Q8_0.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24_v2.0-Q8_0.gguf) | Q8_0 | 13.831 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF --include "LLAMA-2-13b-ko-Y24_v2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/LLAMA-2-13b-ko-Y24_v2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/WestSeverus-7B-GGUF | tensorblock | 2025-04-21T00:27:42Z | 91 | 0 | null | [
"gguf",
"merge",
"mergekit",
"lazymergekit",
"senseable/WestLake-7B-v2",
"FelixChao/Severus-7B",
"TensorBlock",
"GGUF",
"base_model:FelixChao/WestSeverus-7B",
"base_model:quantized:FelixChao/WestSeverus-7B",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-16T00:45:54Z | ---
license: apache-2.0
tags:
- merge
- mergekit
- lazymergekit
- senseable/WestLake-7B-v2
- FelixChao/Severus-7B
- TensorBlock
- GGUF
base_model: FelixChao/WestSeverus-7B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## FelixChao/WestSeverus-7B - GGUF
This repo contains GGUF format model files for [FelixChao/WestSeverus-7B](https://huggingface.co/FelixChao/WestSeverus-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [WestSeverus-7B-Q2_K.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [WestSeverus-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [WestSeverus-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [WestSeverus-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [WestSeverus-7B-Q4_0.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [WestSeverus-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [WestSeverus-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [WestSeverus-7B-Q5_0.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [WestSeverus-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [WestSeverus-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [WestSeverus-7B-Q6_K.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [WestSeverus-7B-Q8_0.gguf](https://huggingface.co/tensorblock/WestSeverus-7B-GGUF/blob/main/WestSeverus-7B-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/WestSeverus-7B-GGUF --include "WestSeverus-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/WestSeverus-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF | tensorblock | 2025-04-21T00:27:36Z | 141 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"dataset:abacusai/MetaMathFewshot",
"base_model:abacusai/MetaMath-bagel-34b-v0.2-c1500",
"base_model:quantized:abacusai/MetaMath-bagel-34b-v0.2-c1500",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T22:42:17Z | ---
license: apache-2.0
datasets:
- abacusai/MetaMathFewshot
tags:
- TensorBlock
- GGUF
base_model: abacusai/MetaMath-bagel-34b-v0.2-c1500
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## abacusai/MetaMath-bagel-34b-v0.2-c1500 - GGUF
This repo contains GGUF format model files for [abacusai/MetaMath-bagel-34b-v0.2-c1500](https://huggingface.co/abacusai/MetaMath-bagel-34b-v0.2-c1500).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
[INST] <<SYS>>
{system_prompt}
<</SYS>>
{prompt} [/INST]
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [MetaMath-bagel-34b-v0.2-c1500-Q2_K.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q2_K.gguf) | Q2_K | 12.825 GB | smallest, significant quality loss - not recommended for most purposes |
| [MetaMath-bagel-34b-v0.2-c1500-Q3_K_S.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q3_K_S.gguf) | Q3_K_S | 14.960 GB | very small, high quality loss |
| [MetaMath-bagel-34b-v0.2-c1500-Q3_K_M.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q3_K_M.gguf) | Q3_K_M | 16.655 GB | very small, high quality loss |
| [MetaMath-bagel-34b-v0.2-c1500-Q3_K_L.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q3_K_L.gguf) | Q3_K_L | 18.139 GB | small, substantial quality loss |
| [MetaMath-bagel-34b-v0.2-c1500-Q4_0.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q4_0.gguf) | Q4_0 | 19.467 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [MetaMath-bagel-34b-v0.2-c1500-Q4_K_S.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q4_K_S.gguf) | Q4_K_S | 19.599 GB | small, greater quality loss |
| [MetaMath-bagel-34b-v0.2-c1500-Q4_K_M.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q4_K_M.gguf) | Q4_K_M | 20.659 GB | medium, balanced quality - recommended |
| [MetaMath-bagel-34b-v0.2-c1500-Q5_0.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q5_0.gguf) | Q5_0 | 23.708 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [MetaMath-bagel-34b-v0.2-c1500-Q5_K_S.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q5_K_S.gguf) | Q5_K_S | 23.708 GB | large, low quality loss - recommended |
| [MetaMath-bagel-34b-v0.2-c1500-Q5_K_M.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q5_K_M.gguf) | Q5_K_M | 24.322 GB | large, very low quality loss - recommended |
| [MetaMath-bagel-34b-v0.2-c1500-Q6_K.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q6_K.gguf) | Q6_K | 28.214 GB | very large, extremely low quality loss |
| [MetaMath-bagel-34b-v0.2-c1500-Q8_0.gguf](https://huggingface.co/tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF/blob/main/MetaMath-bagel-34b-v0.2-c1500-Q8_0.gguf) | Q8_0 | 36.542 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF --include "MetaMath-bagel-34b-v0.2-c1500-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/MetaMath-bagel-34b-v0.2-c1500-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/TenyxChat-7B-v1-GGUF | tensorblock | 2025-04-21T00:27:32Z | 91 | 0 | transformers | [
"transformers",
"gguf",
"tenyx-fine-tuning",
"dpo",
"tenyxchat",
"TensorBlock",
"GGUF",
"en",
"base_model:tenyx/TenyxChat-7B-v1",
"base_model:quantized:tenyx/TenyxChat-7B-v1",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T21:59:43Z | ---
license: apache-2.0
language:
- en
library_name: transformers
tags:
- tenyx-fine-tuning
- dpo
- tenyxchat
- TensorBlock
- GGUF
base_model: tenyx/TenyxChat-7B-v1
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## tenyx/TenyxChat-7B-v1 - GGUF
This repo contains GGUF format model files for [tenyx/TenyxChat-7B-v1](https://huggingface.co/tenyx/TenyxChat-7B-v1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s> System:{system_prompt}<|end_of_turn|>
User:{prompt}<|end_of_turn|>
Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [TenyxChat-7B-v1-Q2_K.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [TenyxChat-7B-v1-Q3_K_S.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [TenyxChat-7B-v1-Q3_K_M.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [TenyxChat-7B-v1-Q3_K_L.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [TenyxChat-7B-v1-Q4_0.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [TenyxChat-7B-v1-Q4_K_S.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [TenyxChat-7B-v1-Q4_K_M.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [TenyxChat-7B-v1-Q5_0.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [TenyxChat-7B-v1-Q5_K_S.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [TenyxChat-7B-v1-Q5_K_M.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [TenyxChat-7B-v1-Q6_K.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [TenyxChat-7B-v1-Q8_0.gguf](https://huggingface.co/tensorblock/TenyxChat-7B-v1-GGUF/blob/main/TenyxChat-7B-v1-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/TenyxChat-7B-v1-GGUF --include "TenyxChat-7B-v1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/TenyxChat-7B-v1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/code-millenials-13b-GGUF | tensorblock | 2025-04-21T00:27:29Z | 120 | 0 | transformers | [
"transformers",
"gguf",
"code",
"TensorBlock",
"GGUF",
"base_model:budecosystem/code-millenials-13b",
"base_model:quantized:budecosystem/code-millenials-13b",
"license:llama2",
"model-index",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T20:50:53Z | ---
license: llama2
library_name: transformers
tags:
- code
- TensorBlock
- GGUF
base_model: budecosystem/code-millenials-13b
model-index:
- name: Code Millenials
results:
- task:
type: text-generation
dataset:
name: HumanEval
type: openai_humaneval
metrics:
- type: pass@1
value: 0.7621
name: pass@1
verified: false
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## budecosystem/code-millenials-13b - GGUF
This repo contains GGUF format model files for [budecosystem/code-millenials-13b](https://huggingface.co/budecosystem/code-millenials-13b).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [code-millenials-13b-Q2_K.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q2_K.gguf) | Q2_K | 4.855 GB | smallest, significant quality loss - not recommended for most purposes |
| [code-millenials-13b-Q3_K_S.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q3_K_S.gguf) | Q3_K_S | 5.659 GB | very small, high quality loss |
| [code-millenials-13b-Q3_K_M.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q3_K_M.gguf) | Q3_K_M | 6.338 GB | very small, high quality loss |
| [code-millenials-13b-Q3_K_L.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q3_K_L.gguf) | Q3_K_L | 6.930 GB | small, substantial quality loss |
| [code-millenials-13b-Q4_0.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q4_0.gguf) | Q4_0 | 7.366 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [code-millenials-13b-Q4_K_S.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q4_K_S.gguf) | Q4_K_S | 7.424 GB | small, greater quality loss |
| [code-millenials-13b-Q4_K_M.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q4_K_M.gguf) | Q4_K_M | 7.866 GB | medium, balanced quality - recommended |
| [code-millenials-13b-Q5_0.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q5_0.gguf) | Q5_0 | 8.973 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [code-millenials-13b-Q5_K_S.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q5_K_S.gguf) | Q5_K_S | 8.973 GB | large, low quality loss - recommended |
| [code-millenials-13b-Q5_K_M.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q5_K_M.gguf) | Q5_K_M | 9.230 GB | large, very low quality loss - recommended |
| [code-millenials-13b-Q6_K.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q6_K.gguf) | Q6_K | 10.680 GB | very large, extremely low quality loss |
| [code-millenials-13b-Q8_0.gguf](https://huggingface.co/tensorblock/code-millenials-13b-GGUF/blob/main/code-millenials-13b-Q8_0.gguf) | Q8_0 | 13.832 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/code-millenials-13b-GGUF --include "code-millenials-13b-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/code-millenials-13b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/AVA-Mistral-7B-V2-GGUF | tensorblock | 2025-04-21T00:27:27Z | 90 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:MehdiHosseiniMoghadam/AVA-Mistral-7B-V2",
"base_model:quantized:MehdiHosseiniMoghadam/AVA-Mistral-7B-V2",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T20:08:19Z | ---
base_model: MehdiHosseiniMoghadam/AVA-Mistral-7B-V2
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## MehdiHosseiniMoghadam/AVA-Mistral-7B-V2 - GGUF
This repo contains GGUF format model files for [MehdiHosseiniMoghadam/AVA-Mistral-7B-V2](https://huggingface.co/MehdiHosseiniMoghadam/AVA-Mistral-7B-V2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [AVA-Mistral-7B-V2-Q2_K.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [AVA-Mistral-7B-V2-Q3_K_S.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [AVA-Mistral-7B-V2-Q3_K_M.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [AVA-Mistral-7B-V2-Q3_K_L.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [AVA-Mistral-7B-V2-Q4_0.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [AVA-Mistral-7B-V2-Q4_K_S.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [AVA-Mistral-7B-V2-Q4_K_M.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [AVA-Mistral-7B-V2-Q5_0.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [AVA-Mistral-7B-V2-Q5_K_S.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [AVA-Mistral-7B-V2-Q5_K_M.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [AVA-Mistral-7B-V2-Q6_K.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [AVA-Mistral-7B-V2-Q8_0.gguf](https://huggingface.co/tensorblock/AVA-Mistral-7B-V2-GGUF/blob/main/AVA-Mistral-7B-V2-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/AVA-Mistral-7B-V2-GGUF --include "AVA-Mistral-7B-V2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/AVA-Mistral-7B-V2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF | tensorblock | 2025-04-21T00:27:17Z | 115 | 0 | null | [
"gguf",
"text-generation",
"TensorBlock",
"GGUF",
"ko",
"base_model:Edentns/DataVortexS-10.7B-dpo-v1.12",
"base_model:quantized:Edentns/DataVortexS-10.7B-dpo-v1.12",
"license:cc-by-nc-sa-4.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-15T15:43:13Z | ---
tags:
- text-generation
- TensorBlock
- GGUF
license: cc-by-nc-sa-4.0
language:
- ko
base_model: Edentns/DataVortexS-10.7B-dpo-v1.12
pipeline_tag: text-generation
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Edentns/DataVortexS-10.7B-dpo-v1.12 - GGUF
This repo contains GGUF format model files for [Edentns/DataVortexS-10.7B-dpo-v1.12](https://huggingface.co/Edentns/DataVortexS-10.7B-dpo-v1.12).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### System:
{system_prompt}
### User:
{prompt}
### Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [DataVortexS-10.7B-dpo-v1.12-Q2_K.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [DataVortexS-10.7B-dpo-v1.12-Q3_K_S.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [DataVortexS-10.7B-dpo-v1.12-Q3_K_M.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [DataVortexS-10.7B-dpo-v1.12-Q3_K_L.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [DataVortexS-10.7B-dpo-v1.12-Q4_0.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [DataVortexS-10.7B-dpo-v1.12-Q4_K_S.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [DataVortexS-10.7B-dpo-v1.12-Q4_K_M.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [DataVortexS-10.7B-dpo-v1.12-Q5_0.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [DataVortexS-10.7B-dpo-v1.12-Q5_K_S.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [DataVortexS-10.7B-dpo-v1.12-Q5_K_M.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [DataVortexS-10.7B-dpo-v1.12-Q6_K.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [DataVortexS-10.7B-dpo-v1.12-Q8_0.gguf](https://huggingface.co/tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF/blob/main/DataVortexS-10.7B-dpo-v1.12-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF --include "DataVortexS-10.7B-dpo-v1.12-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/DataVortexS-10.7B-dpo-v1.12-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF | tensorblock | 2025-04-21T00:27:15Z | 112 | 0 | null | [
"gguf",
"dare",
"super mario merge",
"pytorch",
"mixtral",
"merge",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:martyn/mixtral-megamerge-dare-8x7b-v2",
"base_model:quantized:martyn/mixtral-megamerge-dare-8x7b-v2",
"license:apache-2.0",
"region:us",
"conversational"
] | text-generation | 2024-12-15T14:38:48Z | ---
license: apache-2.0
language:
- en
pipeline_tag: text-generation
inference: false
tags:
- dare
- super mario merge
- pytorch
- mixtral
- merge
- TensorBlock
- GGUF
base_model: martyn/mixtral-megamerge-dare-8x7b-v2
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## martyn/mixtral-megamerge-dare-8x7b-v2 - GGUF
This repo contains GGUF format model files for [martyn/mixtral-megamerge-dare-8x7b-v2](https://huggingface.co/martyn/mixtral-megamerge-dare-8x7b-v2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [mixtral-megamerge-dare-8x7b-v2-Q2_K.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q2_K.gguf) | Q2_K | 17.311 GB | smallest, significant quality loss - not recommended for most purposes |
| [mixtral-megamerge-dare-8x7b-v2-Q3_K_S.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q3_K_S.gguf) | Q3_K_S | 20.433 GB | very small, high quality loss |
| [mixtral-megamerge-dare-8x7b-v2-Q3_K_M.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q3_K_M.gguf) | Q3_K_M | 22.546 GB | very small, high quality loss |
| [mixtral-megamerge-dare-8x7b-v2-Q3_K_L.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q3_K_L.gguf) | Q3_K_L | 24.170 GB | small, substantial quality loss |
| [mixtral-megamerge-dare-8x7b-v2-Q4_0.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q4_0.gguf) | Q4_0 | 26.444 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [mixtral-megamerge-dare-8x7b-v2-Q4_K_S.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q4_K_S.gguf) | Q4_K_S | 26.746 GB | small, greater quality loss |
| [mixtral-megamerge-dare-8x7b-v2-Q4_K_M.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q4_K_M.gguf) | Q4_K_M | 28.448 GB | medium, balanced quality - recommended |
| [mixtral-megamerge-dare-8x7b-v2-Q5_0.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q5_0.gguf) | Q5_0 | 32.231 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [mixtral-megamerge-dare-8x7b-v2-Q5_K_S.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q5_K_S.gguf) | Q5_K_S | 32.231 GB | large, low quality loss - recommended |
| [mixtral-megamerge-dare-8x7b-v2-Q5_K_M.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q5_K_M.gguf) | Q5_K_M | 33.230 GB | large, very low quality loss - recommended |
| [mixtral-megamerge-dare-8x7b-v2-Q6_K.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q6_K.gguf) | Q6_K | 38.381 GB | very large, extremely low quality loss |
| [mixtral-megamerge-dare-8x7b-v2-Q8_0.gguf](https://huggingface.co/tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF/blob/main/mixtral-megamerge-dare-8x7b-v2-Q8_0.gguf) | Q8_0 | 49.626 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF --include "mixtral-megamerge-dare-8x7b-v2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/mixtral-megamerge-dare-8x7b-v2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF | tensorblock | 2025-04-21T00:27:11Z | 123 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"zh",
"en",
"dataset:YeungNLP/firefly-pretrain-dataset",
"base_model:zyh3826/20231206094523-pretrain-Llama-2-13b-hf-76000",
"base_model:quantized:zyh3826/20231206094523-pretrain-Llama-2-13b-hf-76000",
"license:llama2",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T14:05:48Z | ---
license: llama2
datasets:
- YeungNLP/firefly-pretrain-dataset
language:
- zh
- en
base_model: zyh3826/20231206094523-pretrain-Llama-2-13b-hf-76000
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## zyh3826/20231206094523-pretrain-Llama-2-13b-hf-76000 - GGUF
This repo contains GGUF format model files for [zyh3826/20231206094523-pretrain-Llama-2-13b-hf-76000](https://huggingface.co/zyh3826/20231206094523-pretrain-Llama-2-13b-hf-76000).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q2_K.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q2_K.gguf) | Q2_K | 4.992 GB | smallest, significant quality loss - not recommended for most purposes |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q3_K_S.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q3_K_S.gguf) | Q3_K_S | 5.809 GB | very small, high quality loss |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q3_K_M.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q3_K_M.gguf) | Q3_K_M | 6.487 GB | very small, high quality loss |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q3_K_L.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q3_K_L.gguf) | Q3_K_L | 7.079 GB | small, substantial quality loss |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q4_0.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q4_0.gguf) | Q4_0 | 7.531 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q4_K_S.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q4_K_S.gguf) | Q4_K_S | 7.589 GB | small, greater quality loss |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q4_K_M.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q4_K_M.gguf) | Q4_K_M | 8.031 GB | medium, balanced quality - recommended |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q5_0.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q5_0.gguf) | Q5_0 | 9.153 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q5_K_S.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q5_K_S.gguf) | Q5_K_S | 9.153 GB | large, low quality loss - recommended |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q5_K_M.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q5_K_M.gguf) | Q5_K_M | 9.410 GB | large, very low quality loss - recommended |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q6_K.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q6_K.gguf) | Q6_K | 10.875 GB | very large, extremely low quality loss |
| [20231206094523-pretrain-Llama-2-13b-hf-76000-Q8_0.gguf](https://huggingface.co/tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF/blob/main/20231206094523-pretrain-Llama-2-13b-hf-76000-Q8_0.gguf) | Q8_0 | 14.085 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF --include "20231206094523-pretrain-Llama-2-13b-hf-76000-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/20231206094523-pretrain-Llama-2-13b-hf-76000-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Kan-LLaMA-7B-base-GGUF | tensorblock | 2025-04-21T00:27:10Z | 90 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"kn",
"en",
"base_model:fierysurf/Kan-LLaMA-7B-base",
"base_model:quantized:fierysurf/Kan-LLaMA-7B-base",
"license:mit",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T13:38:53Z | ---
license: mit
language:
- kn
- en
base_model: fierysurf/Kan-LLaMA-7B-base
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## fierysurf/Kan-LLaMA-7B-base - GGUF
This repo contains GGUF format model files for [fierysurf/Kan-LLaMA-7B-base](https://huggingface.co/fierysurf/Kan-LLaMA-7B-base).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Kan-LLaMA-7B-base-Q2_K.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q2_K.gguf) | Q2_K | 2.615 GB | smallest, significant quality loss - not recommended for most purposes |
| [Kan-LLaMA-7B-base-Q3_K_S.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q3_K_S.gguf) | Q3_K_S | 3.038 GB | very small, high quality loss |
| [Kan-LLaMA-7B-base-Q3_K_M.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q3_K_M.gguf) | Q3_K_M | 3.388 GB | very small, high quality loss |
| [Kan-LLaMA-7B-base-Q3_K_L.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q3_K_L.gguf) | Q3_K_L | 3.687 GB | small, substantial quality loss |
| [Kan-LLaMA-7B-base-Q4_0.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q4_0.gguf) | Q4_0 | 3.925 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Kan-LLaMA-7B-base-Q4_K_S.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q4_K_S.gguf) | Q4_K_S | 3.956 GB | small, greater quality loss |
| [Kan-LLaMA-7B-base-Q4_K_M.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q4_K_M.gguf) | Q4_K_M | 4.180 GB | medium, balanced quality - recommended |
| [Kan-LLaMA-7B-base-Q5_0.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q5_0.gguf) | Q5_0 | 4.760 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Kan-LLaMA-7B-base-Q5_K_S.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q5_K_S.gguf) | Q5_K_S | 4.760 GB | large, low quality loss - recommended |
| [Kan-LLaMA-7B-base-Q5_K_M.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q5_K_M.gguf) | Q5_K_M | 4.891 GB | large, very low quality loss - recommended |
| [Kan-LLaMA-7B-base-Q6_K.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q6_K.gguf) | Q6_K | 5.647 GB | very large, extremely low quality loss |
| [Kan-LLaMA-7B-base-Q8_0.gguf](https://huggingface.co/tensorblock/Kan-LLaMA-7B-base-GGUF/blob/main/Kan-LLaMA-7B-base-Q8_0.gguf) | Q8_0 | 7.313 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Kan-LLaMA-7B-base-GGUF --include "Kan-LLaMA-7B-base-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Kan-LLaMA-7B-base-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/BeagleLake-7B-GGUF | tensorblock | 2025-04-21T00:27:05Z | 89 | 0 | null | [
"gguf",
"merge",
"mergekit",
"mistral",
"fhai50032/RolePlayLake-7B",
"mlabonne/NeuralBeagle14-7B",
"TensorBlock",
"GGUF",
"base_model:fhai50032/BeagleLake-7B",
"base_model:quantized:fhai50032/BeagleLake-7B",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T11:17:43Z | ---
license: apache-2.0
tags:
- merge
- mergekit
- mistral
- fhai50032/RolePlayLake-7B
- mlabonne/NeuralBeagle14-7B
- TensorBlock
- GGUF
base_model: fhai50032/BeagleLake-7B
model-index:
- name: BeagleLake-7B
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 70.39
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fhai50032/BeagleLake-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 87.38
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fhai50032/BeagleLake-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 64.25
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fhai50032/BeagleLake-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 64.92
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fhai50032/BeagleLake-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 83.19
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fhai50032/BeagleLake-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 63.91
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fhai50032/BeagleLake-7B
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## fhai50032/BeagleLake-7B - GGUF
This repo contains GGUF format model files for [fhai50032/BeagleLake-7B](https://huggingface.co/fhai50032/BeagleLake-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>system
{system_prompt}</s>
<s>user
{prompt}</s>
<s>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [BeagleLake-7B-Q2_K.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [BeagleLake-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [BeagleLake-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [BeagleLake-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [BeagleLake-7B-Q4_0.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [BeagleLake-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [BeagleLake-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [BeagleLake-7B-Q5_0.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [BeagleLake-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [BeagleLake-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [BeagleLake-7B-Q6_K.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [BeagleLake-7B-Q8_0.gguf](https://huggingface.co/tensorblock/BeagleLake-7B-GGUF/blob/main/BeagleLake-7B-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/BeagleLake-7B-GGUF --include "BeagleLake-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/BeagleLake-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/NeuralTurdusVariant1-7B-GGUF | tensorblock | 2025-04-21T00:26:59Z | 89 | 0 | null | [
"gguf",
"merge",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:BarryFutureman/NeuralTurdusVariant1-7B",
"base_model:quantized:BarryFutureman/NeuralTurdusVariant1-7B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-15T09:46:01Z | ---
license: apache-2.0
language:
- en
pipeline_tag: text-generation
tags:
- merge
- TensorBlock
- GGUF
base_model: BarryFutureman/NeuralTurdusVariant1-7B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## BarryFutureman/NeuralTurdusVariant1-7B - GGUF
This repo contains GGUF format model files for [BarryFutureman/NeuralTurdusVariant1-7B](https://huggingface.co/BarryFutureman/NeuralTurdusVariant1-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [NeuralTurdusVariant1-7B-Q2_K.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [NeuralTurdusVariant1-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [NeuralTurdusVariant1-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [NeuralTurdusVariant1-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [NeuralTurdusVariant1-7B-Q4_0.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [NeuralTurdusVariant1-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [NeuralTurdusVariant1-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [NeuralTurdusVariant1-7B-Q5_0.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [NeuralTurdusVariant1-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [NeuralTurdusVariant1-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [NeuralTurdusVariant1-7B-Q6_K.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [NeuralTurdusVariant1-7B-Q8_0.gguf](https://huggingface.co/tensorblock/NeuralTurdusVariant1-7B-GGUF/blob/main/NeuralTurdusVariant1-7B-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/NeuralTurdusVariant1-7B-GGUF --include "NeuralTurdusVariant1-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/NeuralTurdusVariant1-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/LeoScorpius-7B-Chat-DPO-GGUF | tensorblock | 2025-04-21T00:26:58Z | 92 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:viethq188/LeoScorpius-7B-Chat-DPO",
"base_model:quantized:viethq188/LeoScorpius-7B-Chat-DPO",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T09:00:13Z | ---
license: apache-2.0
tags:
- TensorBlock
- GGUF
base_model: viethq188/LeoScorpius-7B-Chat-DPO
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## viethq188/LeoScorpius-7B-Chat-DPO - GGUF
This repo contains GGUF format model files for [viethq188/LeoScorpius-7B-Chat-DPO](https://huggingface.co/viethq188/LeoScorpius-7B-Chat-DPO).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [LeoScorpius-7B-Chat-DPO-Q2_K.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [LeoScorpius-7B-Chat-DPO-Q3_K_S.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [LeoScorpius-7B-Chat-DPO-Q3_K_M.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [LeoScorpius-7B-Chat-DPO-Q3_K_L.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [LeoScorpius-7B-Chat-DPO-Q4_0.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [LeoScorpius-7B-Chat-DPO-Q4_K_S.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [LeoScorpius-7B-Chat-DPO-Q4_K_M.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [LeoScorpius-7B-Chat-DPO-Q5_0.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [LeoScorpius-7B-Chat-DPO-Q5_K_S.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [LeoScorpius-7B-Chat-DPO-Q5_K_M.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [LeoScorpius-7B-Chat-DPO-Q6_K.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [LeoScorpius-7B-Chat-DPO-Q8_0.gguf](https://huggingface.co/tensorblock/LeoScorpius-7B-Chat-DPO-GGUF/blob/main/LeoScorpius-7B-Chat-DPO-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/LeoScorpius-7B-Chat-DPO-GGUF --include "LeoScorpius-7B-Chat-DPO-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/LeoScorpius-7B-Chat-DPO-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Wernicke-7B-v8-GGUF | tensorblock | 2025-04-21T00:26:53Z | 103 | 0 | null | [
"gguf",
"merge",
"mergekit",
"lazymergekit",
"kaitchup/Mayonnaise-4in1-022",
"macadeliccc/WestLake-7B-v2-laser-truthy-dpo",
"vanillaOVO/supermario_v2",
"FelixChao/WestSeverus-7B-DPO-v2",
"TensorBlock",
"GGUF",
"base_model:CultriX/Wernicke-7B-v8",
"base_model:quantized:CultriX/Wernicke-7B-v8",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T08:15:08Z | ---
tags:
- merge
- mergekit
- lazymergekit
- kaitchup/Mayonnaise-4in1-022
- macadeliccc/WestLake-7B-v2-laser-truthy-dpo
- vanillaOVO/supermario_v2
- FelixChao/WestSeverus-7B-DPO-v2
- TensorBlock
- GGUF
base_model: CultriX/Wernicke-7B-v8
license: apache-2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## CultriX/Wernicke-7B-v8 - GGUF
This repo contains GGUF format model files for [CultriX/Wernicke-7B-v8](https://huggingface.co/CultriX/Wernicke-7B-v8).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Wernicke-7B-v8-Q2_K.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Wernicke-7B-v8-Q3_K_S.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Wernicke-7B-v8-Q3_K_M.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Wernicke-7B-v8-Q3_K_L.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Wernicke-7B-v8-Q4_0.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Wernicke-7B-v8-Q4_K_S.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Wernicke-7B-v8-Q4_K_M.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Wernicke-7B-v8-Q5_0.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Wernicke-7B-v8-Q5_K_S.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Wernicke-7B-v8-Q5_K_M.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Wernicke-7B-v8-Q6_K.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Wernicke-7B-v8-Q8_0.gguf](https://huggingface.co/tensorblock/Wernicke-7B-v8-GGUF/blob/main/Wernicke-7B-v8-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Wernicke-7B-v8-GGUF --include "Wernicke-7B-v8-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Wernicke-7B-v8-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF | tensorblock | 2025-04-21T00:26:52Z | 34 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:princeton-nlp/Sheared-LLaMA-2.7B-Pruned",
"base_model:quantized:princeton-nlp/Sheared-LLaMA-2.7B-Pruned",
"license:llama2",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T08:03:31Z | ---
license: llama2
tags:
- TensorBlock
- GGUF
base_model: princeton-nlp/Sheared-LLaMA-2.7B-Pruned
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## princeton-nlp/Sheared-LLaMA-2.7B-Pruned - GGUF
This repo contains GGUF format model files for [princeton-nlp/Sheared-LLaMA-2.7B-Pruned](https://huggingface.co/princeton-nlp/Sheared-LLaMA-2.7B-Pruned).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Sheared-LLaMA-2.7B-Pruned-Q2_K.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q2_K.gguf) | Q2_K | 1.028 GB | smallest, significant quality loss - not recommended for most purposes |
| [Sheared-LLaMA-2.7B-Pruned-Q3_K_S.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q3_K_S.gguf) | Q3_K_S | 1.194 GB | very small, high quality loss |
| [Sheared-LLaMA-2.7B-Pruned-Q3_K_M.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q3_K_M.gguf) | Q3_K_M | 1.331 GB | very small, high quality loss |
| [Sheared-LLaMA-2.7B-Pruned-Q3_K_L.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q3_K_L.gguf) | Q3_K_L | 1.448 GB | small, substantial quality loss |
| [Sheared-LLaMA-2.7B-Pruned-Q4_0.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q4_0.gguf) | Q4_0 | 1.542 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Sheared-LLaMA-2.7B-Pruned-Q4_K_S.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q4_K_S.gguf) | Q4_K_S | 1.554 GB | small, greater quality loss |
| [Sheared-LLaMA-2.7B-Pruned-Q4_K_M.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q4_K_M.gguf) | Q4_K_M | 1.642 GB | medium, balanced quality - recommended |
| [Sheared-LLaMA-2.7B-Pruned-Q5_0.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q5_0.gguf) | Q5_0 | 1.869 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Sheared-LLaMA-2.7B-Pruned-Q5_K_S.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q5_K_S.gguf) | Q5_K_S | 1.869 GB | large, low quality loss - recommended |
| [Sheared-LLaMA-2.7B-Pruned-Q5_K_M.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q5_K_M.gguf) | Q5_K_M | 1.921 GB | large, very low quality loss - recommended |
| [Sheared-LLaMA-2.7B-Pruned-Q6_K.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q6_K.gguf) | Q6_K | 2.217 GB | very large, extremely low quality loss |
| [Sheared-LLaMA-2.7B-Pruned-Q8_0.gguf](https://huggingface.co/tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF/blob/main/Sheared-LLaMA-2.7B-Pruned-Q8_0.gguf) | Q8_0 | 2.872 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF --include "Sheared-LLaMA-2.7B-Pruned-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Sheared-LLaMA-2.7B-Pruned-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/zelensky-gpt2-125m-GGUF | tensorblock | 2025-04-21T00:26:50Z | 18 | 0 | null | [
"gguf",
"generated_from_trainer",
"TensorBlock",
"GGUF",
"base_model:slava-medvedev/zelensky-gpt2-125m",
"base_model:quantized:slava-medvedev/zelensky-gpt2-125m",
"license:mit",
"endpoints_compatible",
"region:us"
] | null | 2024-12-15T07:58:02Z | ---
license: mit
base_model: slava-medvedev/zelensky-gpt2-125m
tags:
- generated_from_trainer
- TensorBlock
- GGUF
model-index:
- name: zelensky-gpt2-125m
results: []
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## slava-medvedev/zelensky-gpt2-125m - GGUF
This repo contains GGUF format model files for [slava-medvedev/zelensky-gpt2-125m](https://huggingface.co/slava-medvedev/zelensky-gpt2-125m).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [zelensky-gpt2-125m-Q2_K.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q2_K.gguf) | Q2_K | 0.081 GB | smallest, significant quality loss - not recommended for most purposes |
| [zelensky-gpt2-125m-Q3_K_S.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q3_K_S.gguf) | Q3_K_S | 0.090 GB | very small, high quality loss |
| [zelensky-gpt2-125m-Q3_K_M.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q3_K_M.gguf) | Q3_K_M | 0.098 GB | very small, high quality loss |
| [zelensky-gpt2-125m-Q3_K_L.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q3_K_L.gguf) | Q3_K_L | 0.102 GB | small, substantial quality loss |
| [zelensky-gpt2-125m-Q4_0.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q4_0.gguf) | Q4_0 | 0.107 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [zelensky-gpt2-125m-Q4_K_S.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q4_K_S.gguf) | Q4_K_S | 0.107 GB | small, greater quality loss |
| [zelensky-gpt2-125m-Q4_K_M.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q4_K_M.gguf) | Q4_K_M | 0.113 GB | medium, balanced quality - recommended |
| [zelensky-gpt2-125m-Q5_0.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q5_0.gguf) | Q5_0 | 0.122 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [zelensky-gpt2-125m-Q5_K_S.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q5_K_S.gguf) | Q5_K_S | 0.122 GB | large, low quality loss - recommended |
| [zelensky-gpt2-125m-Q5_K_M.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q5_K_M.gguf) | Q5_K_M | 0.127 GB | large, very low quality loss - recommended |
| [zelensky-gpt2-125m-Q6_K.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q6_K.gguf) | Q6_K | 0.138 GB | very large, extremely low quality loss |
| [zelensky-gpt2-125m-Q8_0.gguf](https://huggingface.co/tensorblock/zelensky-gpt2-125m-GGUF/blob/main/zelensky-gpt2-125m-Q8_0.gguf) | Q8_0 | 0.178 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/zelensky-gpt2-125m-GGUF --include "zelensky-gpt2-125m-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/zelensky-gpt2-125m-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Helios-10.7B-v2-GGUF | tensorblock | 2025-04-21T00:26:43Z | 24 | 0 | null | [
"gguf",
"merge",
"mergekit",
"TensorBlock",
"GGUF",
"base_model:occultml/Helios-10.7B-v2",
"base_model:quantized:occultml/Helios-10.7B-v2",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T06:17:10Z | ---
license: apache-2.0
tags:
- merge
- mergekit
- TensorBlock
- GGUF
base_model: occultml/Helios-10.7B-v2
model-index:
- name: Helios-10.7B-v2
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 39.16
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=occultml/Helios-10.7B-v2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 46.63
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=occultml/Helios-10.7B-v2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 41.57
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=occultml/Helios-10.7B-v2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 55.51
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=occultml/Helios-10.7B-v2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 70.64
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=occultml/Helios-10.7B-v2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 0.0
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=occultml/Helios-10.7B-v2
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## occultml/Helios-10.7B-v2 - GGUF
This repo contains GGUF format model files for [occultml/Helios-10.7B-v2](https://huggingface.co/occultml/Helios-10.7B-v2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### System:
{system_prompt}
### User:
{prompt}
### Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Helios-10.7B-v2-Q2_K.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Helios-10.7B-v2-Q3_K_S.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Helios-10.7B-v2-Q3_K_M.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Helios-10.7B-v2-Q3_K_L.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Helios-10.7B-v2-Q4_0.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Helios-10.7B-v2-Q4_K_S.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Helios-10.7B-v2-Q4_K_M.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Helios-10.7B-v2-Q5_0.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Helios-10.7B-v2-Q5_K_S.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Helios-10.7B-v2-Q5_K_M.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Helios-10.7B-v2-Q6_K.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Helios-10.7B-v2-Q8_0.gguf](https://huggingface.co/tensorblock/Helios-10.7B-v2-GGUF/blob/main/Helios-10.7B-v2-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Helios-10.7B-v2-GGUF --include "Helios-10.7B-v2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Helios-10.7B-v2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Sina-Thor-7b-Merge-GGUF | tensorblock | 2025-04-21T00:26:38Z | 13 | 0 | null | [
"gguf",
"mistral",
"merge",
"TensorBlock",
"GGUF",
"text-generation",
"base_model:Azazelle/Sina-Thor-7b-Merge",
"base_model:quantized:Azazelle/Sina-Thor-7b-Merge",
"license:cc-by-4.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-15T05:13:44Z | ---
pipeline_tag: text-generation
tags:
- mistral
- merge
- TensorBlock
- GGUF
license: cc-by-4.0
base_model: Azazelle/Sina-Thor-7b-Merge
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Azazelle/Sina-Thor-7b-Merge - GGUF
This repo contains GGUF format model files for [Azazelle/Sina-Thor-7b-Merge](https://huggingface.co/Azazelle/Sina-Thor-7b-Merge).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Sina-Thor-7b-Merge-Q2_K.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Sina-Thor-7b-Merge-Q3_K_S.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Sina-Thor-7b-Merge-Q3_K_M.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Sina-Thor-7b-Merge-Q3_K_L.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Sina-Thor-7b-Merge-Q4_0.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Sina-Thor-7b-Merge-Q4_K_S.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Sina-Thor-7b-Merge-Q4_K_M.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Sina-Thor-7b-Merge-Q5_0.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Sina-Thor-7b-Merge-Q5_K_S.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Sina-Thor-7b-Merge-Q5_K_M.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Sina-Thor-7b-Merge-Q6_K.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Sina-Thor-7b-Merge-Q8_0.gguf](https://huggingface.co/tensorblock/Sina-Thor-7b-Merge-GGUF/blob/main/Sina-Thor-7b-Merge-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Sina-Thor-7b-Merge-GGUF --include "Sina-Thor-7b-Merge-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Sina-Thor-7b-Merge-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/typhoon-7b-instruct-02-19-2024-GGUF | tensorblock | 2025-04-21T00:26:36Z | 94 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"base_model:scb10x/typhoon-7b-instruct-02-19-2024",
"base_model:quantized:scb10x/typhoon-7b-instruct-02-19-2024",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-15T04:59:14Z | ---
license: apache-2.0
pipeline_tag: text-generation
tags:
- TensorBlock
- GGUF
base_model: scb10x/typhoon-7b-instruct-02-19-2024
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## scb10x/typhoon-7b-instruct-02-19-2024 - GGUF
This repo contains GGUF format model files for [scb10x/typhoon-7b-instruct-02-19-2024](https://huggingface.co/scb10x/typhoon-7b-instruct-02-19-2024).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [typhoon-7b-instruct-02-19-2024-Q2_K.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q2_K.gguf) | Q2_K | 2.734 GB | smallest, significant quality loss - not recommended for most purposes |
| [typhoon-7b-instruct-02-19-2024-Q3_K_S.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q3_K_S.gguf) | Q3_K_S | 3.181 GB | very small, high quality loss |
| [typhoon-7b-instruct-02-19-2024-Q3_K_M.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q3_K_M.gguf) | Q3_K_M | 3.536 GB | very small, high quality loss |
| [typhoon-7b-instruct-02-19-2024-Q3_K_L.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q3_K_L.gguf) | Q3_K_L | 3.839 GB | small, substantial quality loss |
| [typhoon-7b-instruct-02-19-2024-Q4_0.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q4_0.gguf) | Q4_0 | 4.127 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [typhoon-7b-instruct-02-19-2024-Q4_K_S.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q4_K_S.gguf) | Q4_K_S | 4.159 GB | small, greater quality loss |
| [typhoon-7b-instruct-02-19-2024-Q4_K_M.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q4_K_M.gguf) | Q4_K_M | 4.387 GB | medium, balanced quality - recommended |
| [typhoon-7b-instruct-02-19-2024-Q5_0.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q5_0.gguf) | Q5_0 | 5.018 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [typhoon-7b-instruct-02-19-2024-Q5_K_S.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q5_K_S.gguf) | Q5_K_S | 5.018 GB | large, low quality loss - recommended |
| [typhoon-7b-instruct-02-19-2024-Q5_K_M.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q5_K_M.gguf) | Q5_K_M | 5.151 GB | large, very low quality loss - recommended |
| [typhoon-7b-instruct-02-19-2024-Q6_K.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q6_K.gguf) | Q6_K | 5.964 GB | very large, extremely low quality loss |
| [typhoon-7b-instruct-02-19-2024-Q8_0.gguf](https://huggingface.co/tensorblock/typhoon-7b-instruct-02-19-2024-GGUF/blob/main/typhoon-7b-instruct-02-19-2024-Q8_0.gguf) | Q8_0 | 7.724 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/typhoon-7b-instruct-02-19-2024-GGUF --include "typhoon-7b-instruct-02-19-2024-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/typhoon-7b-instruct-02-19-2024-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Qwen2-VL-7B-Instruct-GGUF | tensorblock | 2025-04-21T00:26:32Z | 167 | 0 | transformers | [
"transformers",
"gguf",
"multimodal",
"TensorBlock",
"GGUF",
"image-text-to-text",
"en",
"base_model:Qwen/Qwen2-VL-7B-Instruct",
"base_model:quantized:Qwen/Qwen2-VL-7B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | image-text-to-text | 2024-12-15T03:59:42Z | ---
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
- TensorBlock
- GGUF
library_name: transformers
base_model: Qwen/Qwen2-VL-7B-Instruct
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Qwen/Qwen2-VL-7B-Instruct - GGUF
This repo contains GGUF format model files for [Qwen/Qwen2-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-7B-Instruct).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4329](https://github.com/ggerganov/llama.cpp/commit/89d604f2c87af9db657d8a27a1528bc4b7579c29).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Qwen2-VL-7B-Instruct-Q2_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q2_K.gguf) | Q2_K | 3.016 GB | smallest, significant quality loss - not recommended for most purposes |
| [Qwen2-VL-7B-Instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q3_K_S.gguf) | Q3_K_S | 3.492 GB | very small, high quality loss |
| [Qwen2-VL-7B-Instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q3_K_M.gguf) | Q3_K_M | 3.808 GB | very small, high quality loss |
| [Qwen2-VL-7B-Instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q3_K_L.gguf) | Q3_K_L | 4.088 GB | small, substantial quality loss |
| [Qwen2-VL-7B-Instruct-Q4_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q4_0.gguf) | Q4_0 | 4.431 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Qwen2-VL-7B-Instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q4_K_S.gguf) | Q4_K_S | 4.458 GB | small, greater quality loss |
| [Qwen2-VL-7B-Instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q4_K_M.gguf) | Q4_K_M | 4.683 GB | medium, balanced quality - recommended |
| [Qwen2-VL-7B-Instruct-Q5_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q5_0.gguf) | Q5_0 | 5.315 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Qwen2-VL-7B-Instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q5_K_S.gguf) | Q5_K_S | 5.315 GB | large, low quality loss - recommended |
| [Qwen2-VL-7B-Instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q5_K_M.gguf) | Q5_K_M | 5.445 GB | large, very low quality loss - recommended |
| [Qwen2-VL-7B-Instruct-Q6_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q6_K.gguf) | Q6_K | 6.254 GB | very large, extremely low quality loss |
| [Qwen2-VL-7B-Instruct-Q8_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-Instruct-GGUF/blob/main/Qwen2-VL-7B-Instruct-Q8_0.gguf) | Q8_0 | 8.099 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Qwen2-VL-7B-Instruct-GGUF --include "Qwen2-VL-7B-Instruct-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Qwen2-VL-7B-Instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/practice-kullmmistral-7b-GGUF | tensorblock | 2025-04-21T00:26:31Z | 14 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:Loyola/practice-kullmmistral-7b",
"base_model:quantized:Loyola/practice-kullmmistral-7b",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T03:30:16Z | ---
base_model: Loyola/practice-kullmmistral-7b
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Loyola/practice-kullmmistral-7b - GGUF
This repo contains GGUF format model files for [Loyola/practice-kullmmistral-7b](https://huggingface.co/Loyola/practice-kullmmistral-7b).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>[INST] {prompt} [/INST]
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [practice-kullmmistral-7b-Q2_K.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [practice-kullmmistral-7b-Q3_K_S.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [practice-kullmmistral-7b-Q3_K_M.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [practice-kullmmistral-7b-Q3_K_L.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [practice-kullmmistral-7b-Q4_0.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [practice-kullmmistral-7b-Q4_K_S.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [practice-kullmmistral-7b-Q4_K_M.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [practice-kullmmistral-7b-Q5_0.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [practice-kullmmistral-7b-Q5_K_S.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [practice-kullmmistral-7b-Q5_K_M.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [practice-kullmmistral-7b-Q6_K.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [practice-kullmmistral-7b-Q8_0.gguf](https://huggingface.co/tensorblock/practice-kullmmistral-7b-GGUF/blob/main/practice-kullmmistral-7b-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/practice-kullmmistral-7b-GGUF --include "practice-kullmmistral-7b-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/practice-kullmmistral-7b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Qwen2-VL-7B-GGUF | tensorblock | 2025-04-21T00:26:28Z | 165 | 1 | transformers | [
"transformers",
"gguf",
"multimodal",
"TensorBlock",
"GGUF",
"image-text-to-text",
"en",
"base_model:Qwen/Qwen2-VL-7B",
"base_model:quantized:Qwen/Qwen2-VL-7B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | image-text-to-text | 2024-12-15T03:25:42Z | ---
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
- TensorBlock
- GGUF
library_name: transformers
base_model: Qwen/Qwen2-VL-7B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Qwen/Qwen2-VL-7B - GGUF
This repo contains GGUF format model files for [Qwen/Qwen2-VL-7B](https://huggingface.co/Qwen/Qwen2-VL-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4329](https://github.com/ggerganov/llama.cpp/commit/89d604f2c87af9db657d8a27a1528bc4b7579c29).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Qwen2-VL-7B-Q2_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q2_K.gguf) | Q2_K | 3.016 GB | smallest, significant quality loss - not recommended for most purposes |
| [Qwen2-VL-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q3_K_S.gguf) | Q3_K_S | 3.492 GB | very small, high quality loss |
| [Qwen2-VL-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q3_K_M.gguf) | Q3_K_M | 3.808 GB | very small, high quality loss |
| [Qwen2-VL-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q3_K_L.gguf) | Q3_K_L | 4.088 GB | small, substantial quality loss |
| [Qwen2-VL-7B-Q4_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q4_0.gguf) | Q4_0 | 4.431 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Qwen2-VL-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q4_K_S.gguf) | Q4_K_S | 4.458 GB | small, greater quality loss |
| [Qwen2-VL-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q4_K_M.gguf) | Q4_K_M | 4.683 GB | medium, balanced quality - recommended |
| [Qwen2-VL-7B-Q5_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q5_0.gguf) | Q5_0 | 5.315 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Qwen2-VL-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q5_K_S.gguf) | Q5_K_S | 5.315 GB | large, low quality loss - recommended |
| [Qwen2-VL-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q5_K_M.gguf) | Q5_K_M | 5.445 GB | large, very low quality loss - recommended |
| [Qwen2-VL-7B-Q6_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q6_K.gguf) | Q6_K | 6.254 GB | very large, extremely low quality loss |
| [Qwen2-VL-7B-Q8_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-7B-GGUF/blob/main/Qwen2-VL-7B-Q8_0.gguf) | Q8_0 | 8.099 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Qwen2-VL-7B-GGUF --include "Qwen2-VL-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Qwen2-VL-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Orca-SOLAR-4x10.7b-GGUF | tensorblock | 2025-04-21T00:26:26Z | 100 | 0 | transformers | [
"transformers",
"gguf",
"code",
"TensorBlock",
"GGUF",
"en",
"dataset:Intel/orca_dpo_pairs",
"base_model:macadeliccc/Orca-SOLAR-4x10.7b",
"base_model:quantized:macadeliccc/Orca-SOLAR-4x10.7b",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-15T01:41:35Z | ---
language:
- en
license: apache-2.0
library_name: transformers
tags:
- code
- TensorBlock
- GGUF
datasets:
- Intel/orca_dpo_pairs
base_model: macadeliccc/Orca-SOLAR-4x10.7b
model-index:
- name: Orca-SOLAR-4x10.7b
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 68.52
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/Orca-SOLAR-4x10.7b
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 86.78
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/Orca-SOLAR-4x10.7b
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 67.03
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/Orca-SOLAR-4x10.7b
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 64.54
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/Orca-SOLAR-4x10.7b
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 83.9
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/Orca-SOLAR-4x10.7b
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 68.23
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=macadeliccc/Orca-SOLAR-4x10.7b
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## macadeliccc/Orca-SOLAR-4x10.7b - GGUF
This repo contains GGUF format model files for [macadeliccc/Orca-SOLAR-4x10.7b](https://huggingface.co/macadeliccc/Orca-SOLAR-4x10.7b).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### System:
{system_prompt}
### User:
{prompt}
### Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Orca-SOLAR-4x10.7b-Q2_K.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q2_K.gguf) | Q2_K | 13.189 GB | smallest, significant quality loss - not recommended for most purposes |
| [Orca-SOLAR-4x10.7b-Q3_K_S.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q3_K_S.gguf) | Q3_K_S | 15.568 GB | very small, high quality loss |
| [Orca-SOLAR-4x10.7b-Q3_K_M.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q3_K_M.gguf) | Q3_K_M | 17.288 GB | very small, high quality loss |
| [Orca-SOLAR-4x10.7b-Q3_K_L.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q3_K_L.gguf) | Q3_K_L | 18.734 GB | small, substantial quality loss |
| [Orca-SOLAR-4x10.7b-Q4_0.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q4_0.gguf) | Q4_0 | 20.345 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Orca-SOLAR-4x10.7b-Q4_K_S.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q4_K_S.gguf) | Q4_K_S | 20.523 GB | small, greater quality loss |
| [Orca-SOLAR-4x10.7b-Q4_K_M.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q4_K_M.gguf) | Q4_K_M | 21.824 GB | medium, balanced quality - recommended |
| [Orca-SOLAR-4x10.7b-Q5_0.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q5_0.gguf) | Q5_0 | 24.840 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Orca-SOLAR-4x10.7b-Q5_K_S.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q5_K_S.gguf) | Q5_K_S | 24.840 GB | large, low quality loss - recommended |
| [Orca-SOLAR-4x10.7b-Q5_K_M.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q5_K_M.gguf) | Q5_K_M | 25.603 GB | large, very low quality loss - recommended |
| [Orca-SOLAR-4x10.7b-Q6_K.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q6_K.gguf) | Q6_K | 29.617 GB | very large, extremely low quality loss |
| [Orca-SOLAR-4x10.7b-Q8_0.gguf](https://huggingface.co/tensorblock/Orca-SOLAR-4x10.7b-GGUF/blob/main/Orca-SOLAR-4x10.7b-Q8_0.gguf) | Q8_0 | 38.360 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Orca-SOLAR-4x10.7b-GGUF --include "Orca-SOLAR-4x10.7b-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Orca-SOLAR-4x10.7b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Qwen2-VL-2B-Instruct-GGUF | tensorblock | 2025-04-21T00:26:24Z | 51 | 0 | transformers | [
"transformers",
"gguf",
"multimodal",
"TensorBlock",
"GGUF",
"image-text-to-text",
"en",
"base_model:Qwen/Qwen2-VL-2B-Instruct",
"base_model:quantized:Qwen/Qwen2-VL-2B-Instruct",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | image-text-to-text | 2024-12-15T01:24:39Z | ---
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
- TensorBlock
- GGUF
library_name: transformers
base_model: Qwen/Qwen2-VL-2B-Instruct
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Qwen/Qwen2-VL-2B-Instruct - GGUF
This repo contains GGUF format model files for [Qwen/Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4329](https://github.com/ggerganov/llama.cpp/commit/89d604f2c87af9db657d8a27a1528bc4b7579c29).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Qwen2-VL-2B-Instruct-Q2_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q2_K.gguf) | Q2_K | 0.676 GB | smallest, significant quality loss - not recommended for most purposes |
| [Qwen2-VL-2B-Instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q3_K_S.gguf) | Q3_K_S | 0.761 GB | very small, high quality loss |
| [Qwen2-VL-2B-Instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q3_K_M.gguf) | Q3_K_M | 0.824 GB | very small, high quality loss |
| [Qwen2-VL-2B-Instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q3_K_L.gguf) | Q3_K_L | 0.880 GB | small, substantial quality loss |
| [Qwen2-VL-2B-Instruct-Q4_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q4_0.gguf) | Q4_0 | 0.935 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Qwen2-VL-2B-Instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q4_K_S.gguf) | Q4_K_S | 0.940 GB | small, greater quality loss |
| [Qwen2-VL-2B-Instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q4_K_M.gguf) | Q4_K_M | 0.986 GB | medium, balanced quality - recommended |
| [Qwen2-VL-2B-Instruct-Q5_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q5_0.gguf) | Q5_0 | 1.099 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Qwen2-VL-2B-Instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q5_K_S.gguf) | Q5_K_S | 1.099 GB | large, low quality loss - recommended |
| [Qwen2-VL-2B-Instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q5_K_M.gguf) | Q5_K_M | 1.125 GB | large, very low quality loss - recommended |
| [Qwen2-VL-2B-Instruct-Q6_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q6_K.gguf) | Q6_K | 1.273 GB | very large, extremely low quality loss |
| [Qwen2-VL-2B-Instruct-Q8_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-Instruct-GGUF/blob/main/Qwen2-VL-2B-Instruct-Q8_0.gguf) | Q8_0 | 1.647 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Qwen2-VL-2B-Instruct-GGUF --include "Qwen2-VL-2B-Instruct-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Qwen2-VL-2B-Instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Qwen2-VL-2B-GGUF | tensorblock | 2025-04-21T00:26:21Z | 25 | 0 | transformers | [
"transformers",
"gguf",
"multimodal",
"TensorBlock",
"GGUF",
"image-text-to-text",
"en",
"base_model:Qwen/Qwen2-VL-2B",
"base_model:quantized:Qwen/Qwen2-VL-2B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | image-text-to-text | 2024-12-15T01:17:12Z | ---
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
tags:
- multimodal
- TensorBlock
- GGUF
library_name: transformers
base_model: Qwen/Qwen2-VL-2B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Qwen/Qwen2-VL-2B - GGUF
This repo contains GGUF format model files for [Qwen/Qwen2-VL-2B](https://huggingface.co/Qwen/Qwen2-VL-2B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4329](https://github.com/ggerganov/llama.cpp/commit/89d604f2c87af9db657d8a27a1528bc4b7579c29).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Qwen2-VL-2B-Q2_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q2_K.gguf) | Q2_K | 0.676 GB | smallest, significant quality loss - not recommended for most purposes |
| [Qwen2-VL-2B-Q3_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q3_K_S.gguf) | Q3_K_S | 0.761 GB | very small, high quality loss |
| [Qwen2-VL-2B-Q3_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q3_K_M.gguf) | Q3_K_M | 0.824 GB | very small, high quality loss |
| [Qwen2-VL-2B-Q3_K_L.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q3_K_L.gguf) | Q3_K_L | 0.880 GB | small, substantial quality loss |
| [Qwen2-VL-2B-Q4_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q4_0.gguf) | Q4_0 | 0.935 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Qwen2-VL-2B-Q4_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q4_K_S.gguf) | Q4_K_S | 0.940 GB | small, greater quality loss |
| [Qwen2-VL-2B-Q4_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q4_K_M.gguf) | Q4_K_M | 0.986 GB | medium, balanced quality - recommended |
| [Qwen2-VL-2B-Q5_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q5_0.gguf) | Q5_0 | 1.099 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Qwen2-VL-2B-Q5_K_S.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q5_K_S.gguf) | Q5_K_S | 1.099 GB | large, low quality loss - recommended |
| [Qwen2-VL-2B-Q5_K_M.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q5_K_M.gguf) | Q5_K_M | 1.125 GB | large, very low quality loss - recommended |
| [Qwen2-VL-2B-Q6_K.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q6_K.gguf) | Q6_K | 1.273 GB | very large, extremely low quality loss |
| [Qwen2-VL-2B-Q8_0.gguf](https://huggingface.co/tensorblock/Qwen2-VL-2B-GGUF/blob/main/Qwen2-VL-2B-Q8_0.gguf) | Q8_0 | 1.647 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Qwen2-VL-2B-GGUF --include "Qwen2-VL-2B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Qwen2-VL-2B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/CodeLlama-70b-Python-hf-GGUF | tensorblock | 2025-04-21T00:26:18Z | 155 | 0 | null | [
"gguf",
"llama-2",
"TensorBlock",
"GGUF",
"text-generation",
"code",
"base_model:codellama/CodeLlama-70b-Python-hf",
"base_model:quantized:codellama/CodeLlama-70b-Python-hf",
"license:llama2",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T22:48:55Z | ---
language:
- code
pipeline_tag: text-generation
tags:
- llama-2
- TensorBlock
- GGUF
license: llama2
base_model: codellama/CodeLlama-70b-Python-hf
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## codellama/CodeLlama-70b-Python-hf - GGUF
This repo contains GGUF format model files for [codellama/CodeLlama-70b-Python-hf](https://huggingface.co/codellama/CodeLlama-70b-Python-hf).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [CodeLlama-70b-Python-hf-Q2_K.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q2_K.gguf) | Q2_K | 25.463 GB | smallest, significant quality loss - not recommended for most purposes |
| [CodeLlama-70b-Python-hf-Q3_K_S.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q3_K_S.gguf) | Q3_K_S | 29.919 GB | very small, high quality loss |
| [CodeLlama-70b-Python-hf-Q3_K_M.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q3_K_M.gguf) | Q3_K_M | 33.275 GB | very small, high quality loss |
| [CodeLlama-70b-Python-hf-Q3_K_L.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q3_K_L.gguf) | Q3_K_L | 36.148 GB | small, substantial quality loss |
| [CodeLlama-70b-Python-hf-Q4_0.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q4_0.gguf) | Q4_0 | 38.872 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [CodeLlama-70b-Python-hf-Q4_K_S.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q4_K_S.gguf) | Q4_K_S | 39.250 GB | small, greater quality loss |
| [CodeLlama-70b-Python-hf-Q4_K_M.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q4_K_M.gguf) | Q4_K_M | 41.423 GB | medium, balanced quality - recommended |
| [CodeLlama-70b-Python-hf-Q5_0.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q5_0.gguf) | Q5_0 | 47.462 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [CodeLlama-70b-Python-hf-Q5_K_S.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q5_K_S.gguf) | Q5_K_S | 47.462 GB | large, low quality loss - recommended |
| [CodeLlama-70b-Python-hf-Q5_K_M.gguf](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q5_K_M.gguf) | Q5_K_M | 48.754 GB | large, very low quality loss - recommended |
| [CodeLlama-70b-Python-hf-Q8_0](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q8_0) | Q6_K | 73.293 GB | very large, extremely low quality loss |
| [CodeLlama-70b-Python-hf-Q6_K](https://huggingface.co/tensorblock/CodeLlama-70b-Python-hf-GGUF/blob/main/CodeLlama-70b-Python-hf-Q6_K) | Q8_0 | 56.588 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/CodeLlama-70b-Python-hf-GGUF --include "CodeLlama-70b-Python-hf-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/CodeLlama-70b-Python-hf-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Llama-3-70B-GGUF | tensorblock | 2025-04-21T00:26:15Z | 151 | 0 | null | [
"gguf",
"facebook",
"meta",
"pytorch",
"llama",
"llama-3",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:v2ray/Llama-3-70B",
"base_model:quantized:v2ray/Llama-3-70B",
"license:other",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T22:46:25Z | ---
language:
- en
pipeline_tag: text-generation
tags:
- facebook
- meta
- pytorch
- llama
- llama-3
- TensorBlock
- GGUF
license: other
license_name: llama3
license_link: LICENSE
extra_gated_prompt: "### META LLAMA 3 COMMUNITY LICENSE AGREEMENT\nMeta Llama 3 Version\
\ Release Date: April 18, 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 Meta Llama 3 distributed by Meta at https://llama.meta.com/get-started/.\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 entityβs 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\"Meta Llama\
\ 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://llama.meta.com/llama-downloads.\n\
\"Llama Materials\" means, collectively, Metaβs proprietary Meta Llama 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).\n\
\ \n1. License Rights and Redistribution.\na. Grant of Rights. You are granted\
\ a non-exclusive, worldwide, non-transferable and royalty-free limited license\
\ under Metaβs 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 that uses any of them, including another AI\
\ model, you shall (A) provide a copy of this Agreement with any such Llama Materials;\
\ and (B) prominently display βBuilt with Meta Llama 3β on a related website, user\
\ interface, blogpost, about page, or product documentation. If you use the Llama\
\ Materials to create, train, fine tune, or otherwise improve an AI model, which\
\ is distributed or made available, you shall also include βLlama 3β 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 βNoticeβ text file distributed as a part of such copies: βMeta Llama\
\ 3 is licensed under the Meta Llama 3 Community License, Copyright Β© Meta Platforms,\
\ Inc. All Rights Reserved.β\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://llama.meta.com/llama3/use-policy), which is hereby incorporated by reference\
\ into this Agreement.\nv. You will not use the Llama Materials or any output or\
\ results of the Llama Materials to improve any other large language model (excluding\
\ Meta Llama 3 or derivative works thereof).\n2. Additional Commercial Terms. If,\
\ on the Meta Llama 3 version release date, the monthly active users of the products\
\ or services made available by or for Licensee, or Licenseeβs 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 βAS ISβ 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.\n\
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.\n\
5. 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 βLlama 3β (the βMarkβ) solely as required to comply with the\
\ last sentence of Section 1.b.i. You will comply with Metaβs 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.\n\
b. Subject to Metaβs 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 Meta Llama 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### Meta Llama 3 Acceptable Use\
\ Policy\nMeta is committed to promoting safe and fair use of its tools and features,\
\ including Meta Llama 3. If you access or use Meta Llama 3, you agree to this Acceptable\
\ Use Policy (βPolicyβ). The most recent copy of this policy can be found at [https://llama.meta.com/llama3/use-policy](https://llama.meta.com/llama3/use-policy)\n\
#### Prohibited Uses\nWe want everyone to use Meta Llama 3 safely and responsibly.\
\ You agree you will not use, or allow others to use, Meta Llama 3 to: 1. Violate\
\ the law or othersβ rights, including to:\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 2. Engage in, promote, incite, or facilitate the\
\ harassment, abuse, threatening, or bullying of individuals or groups of individuals\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 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 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\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 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\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 Meta Llama 3 related\
\ to the following:\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\n 2. Guns and illegal weapons (including weapon development)\n 3.\
\ Illegal drugs and regulated/controlled substances\n 4. Operation of critical\
\ infrastructure, transportation technologies, or heavy machinery\n 5. Self-harm\
\ or harm to others, including suicide, cutting, and eating disorders\n 6. Any\
\ content intended to incite or promote violence, abuse, or any infliction of bodily\
\ harm to an individual\n3. Intentionally deceive or mislead others, including use\
\ of Meta Llama 3 related to the following:\n 1. Generating, promoting, or furthering\
\ fraud or the creation or promotion of disinformation\n 2. Generating, promoting,\
\ or furthering defamatory content, including the creation of defamatory statements,\
\ images, or other content\n 3. Generating, promoting, or further distributing\
\ spam\n 4. Impersonating another individual without consent, authorization,\
\ or legal right\n 5. Representing that the use of Meta Llama 3 or outputs are\
\ human-generated\n 6. Generating or facilitating false online engagement, including\
\ fake reviews and other means of fake online engagement\n4. Fail to appropriately\
\ disclose to end users any known dangers of your AI system\nPlease report any violation\
\ of this Policy, software βbug,β 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/llama3](https://github.com/meta-llama/llama3)\n\
\ * Reporting risky content generated by the model:\n developers.facebook.com/llama_output_feedback\n\
\ * Reporting bugs and security concerns: facebook.com/whitehat/info\n * 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
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
base_model: v2ray/Llama-3-70B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## v2ray/Llama-3-70B - GGUF
This repo contains GGUF format model files for [v2ray/Llama-3-70B](https://huggingface.co/v2ray/Llama-3-70B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Llama-3-70B-Q2_K.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q2_K.gguf) | Q2_K | 26.375 GB | smallest, significant quality loss - not recommended for most purposes |
| [Llama-3-70B-Q3_K_S.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q3_K_S.gguf) | Q3_K_S | 30.912 GB | very small, high quality loss |
| [Llama-3-70B-Q3_K_M.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q3_K_M.gguf) | Q3_K_M | 34.267 GB | very small, high quality loss |
| [Llama-3-70B-Q3_K_L.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q3_K_L.gguf) | Q3_K_L | 37.141 GB | small, substantial quality loss |
| [Llama-3-70B-Q4_0.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q4_0.gguf) | Q4_0 | 39.970 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Llama-3-70B-Q4_K_S.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q4_K_S.gguf) | Q4_K_S | 40.347 GB | small, greater quality loss |
| [Llama-3-70B-Q4_K_M.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q4_K_M.gguf) | Q4_K_M | 42.520 GB | medium, balanced quality - recommended |
| [Llama-3-70B-Q5_0.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q5_0.gguf) | Q5_0 | 48.657 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Llama-3-70B-Q5_K_S.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q5_K_S.gguf) | Q5_K_S | 48.657 GB | large, low quality loss - recommended |
| [Llama-3-70B-Q5_K_M.gguf](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q5_K_M.gguf) | Q5_K_M | 49.950 GB | large, very low quality loss - recommended |
| [Llama-3-70B-Q6_K](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q6_K) | Q6_K | 57.888 GB | very large, extremely low quality loss |
| [Llama-3-70B-Q8_0](https://huggingface.co/tensorblock/Llama-3-70B-GGUF/blob/main/Llama-3-70B-Q8_0) | Q8_0 | 74.975 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Llama-3-70B-GGUF --include "Llama-3-70B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Llama-3-70B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/flux-7b-v0.1-GGUF | tensorblock | 2025-04-21T00:26:11Z | 116 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"en",
"base_model:chanwit/flux-7b-v0.1",
"base_model:quantized:chanwit/flux-7b-v0.1",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-14T21:32:08Z | ---
license: apache-2.0
language:
- en
base_model: chanwit/flux-7b-v0.1
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## chanwit/flux-7b-v0.1 - GGUF
This repo contains GGUF format model files for [chanwit/flux-7b-v0.1](https://huggingface.co/chanwit/flux-7b-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [flux-7b-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [flux-7b-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [flux-7b-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [flux-7b-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [flux-7b-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [flux-7b-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [flux-7b-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [flux-7b-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [flux-7b-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [flux-7b-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [flux-7b-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [flux-7b-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/flux-7b-v0.1-GGUF/blob/main/flux-7b-v0.1-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/flux-7b-v0.1-GGUF --include "flux-7b-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/flux-7b-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/DIOD-Mistral-0.2-GGUF | tensorblock | 2025-04-21T00:26:02Z | 89 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"en",
"base_model:asapse/DIOD-Mistral-0.2",
"base_model:quantized:asapse/DIOD-Mistral-0.2",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-14T20:48:48Z | ---
license: apache-2.0
language:
- en
base_model: asapse/DIOD-Mistral-0.2
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## asapse/DIOD-Mistral-0.2 - GGUF
This repo contains GGUF format model files for [asapse/DIOD-Mistral-0.2](https://huggingface.co/asapse/DIOD-Mistral-0.2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [DIOD-Mistral-0.2-Q2_K.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [DIOD-Mistral-0.2-Q3_K_S.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [DIOD-Mistral-0.2-Q3_K_M.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [DIOD-Mistral-0.2-Q3_K_L.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [DIOD-Mistral-0.2-Q4_0.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [DIOD-Mistral-0.2-Q4_K_S.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [DIOD-Mistral-0.2-Q4_K_M.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [DIOD-Mistral-0.2-Q5_0.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [DIOD-Mistral-0.2-Q5_K_S.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [DIOD-Mistral-0.2-Q5_K_M.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [DIOD-Mistral-0.2-Q6_K.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [DIOD-Mistral-0.2-Q8_0.gguf](https://huggingface.co/tensorblock/DIOD-Mistral-0.2-GGUF/blob/main/DIOD-Mistral-0.2-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/DIOD-Mistral-0.2-GGUF --include "DIOD-Mistral-0.2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/DIOD-Mistral-0.2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/llama-39m-GGUF | tensorblock | 2025-04-21T00:25:58Z | 15 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:Cheng98/llama-39m",
"base_model:quantized:Cheng98/llama-39m",
"license:llama2",
"endpoints_compatible",
"region:us"
] | null | 2024-12-14T20:13:04Z | ---
license: llama2
tags:
- TensorBlock
- GGUF
base_model: Cheng98/llama-39m
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Cheng98/llama-39m - GGUF
This repo contains GGUF format model files for [Cheng98/llama-39m](https://huggingface.co/Cheng98/llama-39m).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [llama-39m-Q2_K.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q2_K.gguf) | Q2_K | 0.023 GB | smallest, significant quality loss - not recommended for most purposes |
| [llama-39m-Q3_K_S.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q3_K_S.gguf) | Q3_K_S | 0.025 GB | very small, high quality loss |
| [llama-39m-Q3_K_M.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q3_K_M.gguf) | Q3_K_M | 0.025 GB | very small, high quality loss |
| [llama-39m-Q3_K_L.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q3_K_L.gguf) | Q3_K_L | 0.026 GB | small, substantial quality loss |
| [llama-39m-Q4_0.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q4_0.gguf) | Q4_0 | 0.028 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [llama-39m-Q4_K_S.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q4_K_S.gguf) | Q4_K_S | 0.028 GB | small, greater quality loss |
| [llama-39m-Q4_K_M.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q4_K_M.gguf) | Q4_K_M | 0.028 GB | medium, balanced quality - recommended |
| [llama-39m-Q5_0.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q5_0.gguf) | Q5_0 | 0.031 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [llama-39m-Q5_K_S.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q5_K_S.gguf) | Q5_K_S | 0.031 GB | large, low quality loss - recommended |
| [llama-39m-Q5_K_M.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q5_K_M.gguf) | Q5_K_M | 0.031 GB | large, very low quality loss - recommended |
| [llama-39m-Q6_K.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q6_K.gguf) | Q6_K | 0.034 GB | very large, extremely low quality loss |
| [llama-39m-Q8_0.gguf](https://huggingface.co/tensorblock/llama-39m-GGUF/blob/main/llama-39m-Q8_0.gguf) | Q8_0 | 0.044 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/llama-39m-GGUF --include "llama-39m-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/llama-39m-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF | tensorblock | 2025-04-21T00:25:55Z | 111 | 0 | null | [
"gguf",
"llama2",
"TensorBlock",
"GGUF",
"text-generation",
"ko",
"base_model:AIdenU/LLAMA-2-13b-ko-Y24-DPO_v2.0",
"base_model:quantized:AIdenU/LLAMA-2-13b-ko-Y24-DPO_v2.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T19:04:51Z | ---
license: apache-2.0
language:
- ko
pipeline_tag: text-generation
tags:
- llama2
- TensorBlock
- GGUF
base_model: AIdenU/LLAMA-2-13b-ko-Y24-DPO_v2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## AIdenU/LLAMA-2-13b-ko-Y24-DPO_v2.0 - GGUF
This repo contains GGUF format model files for [AIdenU/LLAMA-2-13b-ko-Y24-DPO_v2.0](https://huggingface.co/AIdenU/LLAMA-2-13b-ko-Y24-DPO_v2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q2_K.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q2_K.gguf) | Q2_K | 4.854 GB | smallest, significant quality loss - not recommended for most purposes |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q3_K_S.gguf) | Q3_K_S | 5.659 GB | very small, high quality loss |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q3_K_M.gguf) | Q3_K_M | 6.338 GB | very small, high quality loss |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q3_K_L.gguf) | Q3_K_L | 6.930 GB | small, substantial quality loss |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q4_0.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q4_0.gguf) | Q4_0 | 7.366 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q4_K_S.gguf) | Q4_K_S | 7.423 GB | small, greater quality loss |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q4_K_M.gguf) | Q4_K_M | 7.866 GB | medium, balanced quality - recommended |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q5_0.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q5_0.gguf) | Q5_0 | 8.972 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q5_K_S.gguf) | Q5_K_S | 8.972 GB | large, low quality loss - recommended |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q5_K_M.gguf) | Q5_K_M | 9.230 GB | large, very low quality loss - recommended |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q6_K.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q6_K.gguf) | Q6_K | 10.679 GB | very large, extremely low quality loss |
| [LLAMA-2-13b-ko-Y24-DPO_v2.0-Q8_0.gguf](https://huggingface.co/tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF/blob/main/LLAMA-2-13b-ko-Y24-DPO_v2.0-Q8_0.gguf) | Q8_0 | 13.831 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF --include "LLAMA-2-13b-ko-Y24-DPO_v2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/LLAMA-2-13b-ko-Y24-DPO_v2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Monarch-7B-GGUF | tensorblock | 2025-04-21T00:25:53Z | 92 | 0 | null | [
"gguf",
"merge",
"mergekit",
"lazymergekit",
"TensorBlock",
"GGUF",
"base_model:mlabonne/Monarch-7B",
"base_model:quantized:mlabonne/Monarch-7B",
"license:cc-by-nc-4.0",
"model-index",
"endpoints_compatible",
"region:us"
] | null | 2024-12-14T18:15:24Z | ---
license: cc-by-nc-4.0
tags:
- merge
- mergekit
- lazymergekit
- TensorBlock
- GGUF
base_model: mlabonne/Monarch-7B
model-index:
- name: Monarch-7B
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 73.04
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 89.03
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 64.41
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 77.35
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 84.61
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 69.07
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## mlabonne/Monarch-7B - GGUF
This repo contains GGUF format model files for [mlabonne/Monarch-7B](https://huggingface.co/mlabonne/Monarch-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Monarch-7B-Q2_K.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Monarch-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Monarch-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Monarch-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Monarch-7B-Q4_0.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Monarch-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Monarch-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Monarch-7B-Q5_0.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Monarch-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Monarch-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Monarch-7B-Q6_K.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Monarch-7B-Q8_0.gguf](https://huggingface.co/tensorblock/Monarch-7B-GGUF/blob/main/Monarch-7B-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Monarch-7B-GGUF --include "Monarch-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Monarch-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/NeuDist-Ro-7B-GGUF | tensorblock | 2025-04-21T00:25:51Z | 97 | 0 | null | [
"gguf",
"merge",
"mergekit",
"lazymergekit",
"argilla/distilabeled-Marcoro14-7B-slerp",
"mlabonne/NeuralMarcoro14-7B",
"TensorBlock",
"GGUF",
"dataset:mlabonne/chatml_dpo_pairs",
"dataset:argilla/distilabel-intel-orca-dpo-pairs",
"base_model:flemmingmiguel/NeuDist-Ro-7B",
"base_model:quantized:flemmingmiguel/NeuDist-Ro-7B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-14T18:06:45Z | ---
license: apache-2.0
tags:
- merge
- mergekit
- lazymergekit
- argilla/distilabeled-Marcoro14-7B-slerp
- mlabonne/NeuralMarcoro14-7B
- TensorBlock
- GGUF
datasets:
- mlabonne/chatml_dpo_pairs
- argilla/distilabel-intel-orca-dpo-pairs
base_model: flemmingmiguel/NeuDist-Ro-7B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## flemmingmiguel/NeuDist-Ro-7B - GGUF
This repo contains GGUF format model files for [flemmingmiguel/NeuDist-Ro-7B](https://huggingface.co/flemmingmiguel/NeuDist-Ro-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [NeuDist-Ro-7B-Q2_K.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [NeuDist-Ro-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [NeuDist-Ro-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [NeuDist-Ro-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [NeuDist-Ro-7B-Q4_0.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [NeuDist-Ro-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [NeuDist-Ro-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [NeuDist-Ro-7B-Q5_0.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [NeuDist-Ro-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [NeuDist-Ro-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [NeuDist-Ro-7B-Q6_K.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [NeuDist-Ro-7B-Q8_0.gguf](https://huggingface.co/tensorblock/NeuDist-Ro-7B-GGUF/blob/main/NeuDist-Ro-7B-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/NeuDist-Ro-7B-GGUF --include "NeuDist-Ro-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/NeuDist-Ro-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/starcoder2-chat-GGUF | tensorblock | 2025-04-21T00:25:49Z | 133 | 0 | transformers | [
"transformers",
"gguf",
"code",
"starcoder",
"bigcode",
"sft",
"7b",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:abideen/starcoder2-chat",
"base_model:quantized:abideen/starcoder2-chat",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T17:07:42Z | ---
license: cc-by-nc-4.0
base_model: abideen/starcoder2-chat
language:
- en
library_name: transformers
pipeline_tag: text-generation
tags:
- code
- starcoder
- bigcode
- sft
- 7b
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## abideen/starcoder2-chat - GGUF
This repo contains GGUF format model files for [abideen/starcoder2-chat](https://huggingface.co/abideen/starcoder2-chat).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [starcoder2-chat-Q2_K.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q2_K.gguf) | Q2_K | 2.836 GB | smallest, significant quality loss - not recommended for most purposes |
| [starcoder2-chat-Q3_K_S.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q3_K_S.gguf) | Q3_K_S | 3.179 GB | very small, high quality loss |
| [starcoder2-chat-Q3_K_M.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q3_K_M.gguf) | Q3_K_M | 3.662 GB | very small, high quality loss |
| [starcoder2-chat-Q3_K_L.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q3_K_L.gguf) | Q3_K_L | 4.074 GB | small, substantial quality loss |
| [starcoder2-chat-Q4_0.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q4_0.gguf) | Q4_0 | 4.101 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [starcoder2-chat-Q4_K_S.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q4_K_S.gguf) | Q4_K_S | 4.145 GB | small, greater quality loss |
| [starcoder2-chat-Q4_K_M.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q4_K_M.gguf) | Q4_K_M | 4.461 GB | medium, balanced quality - recommended |
| [starcoder2-chat-Q5_0.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q5_0.gguf) | Q5_0 | 4.969 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [starcoder2-chat-Q5_K_S.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q5_K_S.gguf) | Q5_K_S | 4.969 GB | large, low quality loss - recommended |
| [starcoder2-chat-Q5_K_M.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q5_K_M.gguf) | Q5_K_M | 5.155 GB | large, very low quality loss - recommended |
| [starcoder2-chat-Q6_K.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q6_K.gguf) | Q6_K | 5.892 GB | very large, extremely low quality loss |
| [starcoder2-chat-Q8_0.gguf](https://huggingface.co/tensorblock/starcoder2-chat-GGUF/blob/main/starcoder2-chat-Q8_0.gguf) | Q8_0 | 7.629 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/starcoder2-chat-GGUF --include "starcoder2-chat-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/starcoder2-chat-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/prometheus-8x7b-v2.0-GGUF | tensorblock | 2025-04-21T00:25:46Z | 112 | 0 | transformers | [
"transformers",
"gguf",
"text2text-generation",
"TensorBlock",
"GGUF",
"en",
"dataset:prometheus-eval/Feedback-Collection",
"dataset:prometheus-eval/Preference-Collection",
"base_model:prometheus-eval/prometheus-8x7b-v2.0",
"base_model:quantized:prometheus-eval/prometheus-8x7b-v2.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text2text-generation | 2024-12-14T16:44:31Z | ---
tags:
- text2text-generation
- TensorBlock
- GGUF
datasets:
- prometheus-eval/Feedback-Collection
- prometheus-eval/Preference-Collection
license: apache-2.0
language:
- en
pipeline_tag: text2text-generation
library_name: transformers
metrics:
- pearsonr
- spearmanr
- kendall-tau
- accuracy
base_model: prometheus-eval/prometheus-8x7b-v2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## prometheus-eval/prometheus-8x7b-v2.0 - GGUF
This repo contains GGUF format model files for [prometheus-eval/prometheus-8x7b-v2.0](https://huggingface.co/prometheus-eval/prometheus-8x7b-v2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>[INST] {prompt} [/INST]
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [prometheus-8x7b-v2.0-Q2_K.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q2_K.gguf) | Q2_K | 17.311 GB | smallest, significant quality loss - not recommended for most purposes |
| [prometheus-8x7b-v2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q3_K_S.gguf) | Q3_K_S | 20.433 GB | very small, high quality loss |
| [prometheus-8x7b-v2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q3_K_M.gguf) | Q3_K_M | 22.546 GB | very small, high quality loss |
| [prometheus-8x7b-v2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q3_K_L.gguf) | Q3_K_L | 24.170 GB | small, substantial quality loss |
| [prometheus-8x7b-v2.0-Q4_0.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q4_0.gguf) | Q4_0 | 26.444 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [prometheus-8x7b-v2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q4_K_S.gguf) | Q4_K_S | 26.746 GB | small, greater quality loss |
| [prometheus-8x7b-v2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q4_K_M.gguf) | Q4_K_M | 28.448 GB | medium, balanced quality - recommended |
| [prometheus-8x7b-v2.0-Q5_0.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q5_0.gguf) | Q5_0 | 32.231 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [prometheus-8x7b-v2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q5_K_S.gguf) | Q5_K_S | 32.231 GB | large, low quality loss - recommended |
| [prometheus-8x7b-v2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q5_K_M.gguf) | Q5_K_M | 33.230 GB | large, very low quality loss - recommended |
| [prometheus-8x7b-v2.0-Q6_K.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q6_K.gguf) | Q6_K | 38.381 GB | very large, extremely low quality loss |
| [prometheus-8x7b-v2.0-Q8_0.gguf](https://huggingface.co/tensorblock/prometheus-8x7b-v2.0-GGUF/blob/main/prometheus-8x7b-v2.0-Q8_0.gguf) | Q8_0 | 49.626 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/prometheus-8x7b-v2.0-GGUF --include "prometheus-8x7b-v2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/prometheus-8x7b-v2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/ChatMusician-Base-GGUF | tensorblock | 2025-04-21T00:25:38Z | 82 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:m-a-p/ChatMusician-Base",
"base_model:quantized:m-a-p/ChatMusician-Base",
"license:mit",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T15:11:51Z | ---
license: mit
language:
- en
metrics:
- accuracy
pipeline_tag: text-generation
tags:
- TensorBlock
- GGUF
base_model: m-a-p/ChatMusician-Base
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## m-a-p/ChatMusician-Base - GGUF
This repo contains GGUF format model files for [m-a-p/ChatMusician-Base](https://huggingface.co/m-a-p/ChatMusician-Base).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [ChatMusician-Base-Q2_K.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q2_K.gguf) | Q2_K | 2.533 GB | smallest, significant quality loss - not recommended for most purposes |
| [ChatMusician-Base-Q3_K_S.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q3_K_S.gguf) | Q3_K_S | 2.948 GB | very small, high quality loss |
| [ChatMusician-Base-Q3_K_M.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q3_K_M.gguf) | Q3_K_M | 3.298 GB | very small, high quality loss |
| [ChatMusician-Base-Q3_K_L.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q3_K_L.gguf) | Q3_K_L | 3.597 GB | small, substantial quality loss |
| [ChatMusician-Base-Q4_0.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q4_0.gguf) | Q4_0 | 3.826 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [ChatMusician-Base-Q4_K_S.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q4_K_S.gguf) | Q4_K_S | 3.857 GB | small, greater quality loss |
| [ChatMusician-Base-Q4_K_M.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q4_K_M.gguf) | Q4_K_M | 4.081 GB | medium, balanced quality - recommended |
| [ChatMusician-Base-Q5_0.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q5_0.gguf) | Q5_0 | 4.652 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [ChatMusician-Base-Q5_K_S.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q5_K_S.gguf) | Q5_K_S | 4.652 GB | large, low quality loss - recommended |
| [ChatMusician-Base-Q5_K_M.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q5_K_M.gguf) | Q5_K_M | 4.783 GB | large, very low quality loss - recommended |
| [ChatMusician-Base-Q6_K.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q6_K.gguf) | Q6_K | 5.529 GB | very large, extremely low quality loss |
| [ChatMusician-Base-Q8_0.gguf](https://huggingface.co/tensorblock/ChatMusician-Base-GGUF/blob/main/ChatMusician-Base-Q8_0.gguf) | Q8_0 | 7.161 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/ChatMusician-Base-GGUF --include "ChatMusician-Base-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/ChatMusician-Base-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/code_gpt2_mini_model-GGUF | tensorblock | 2025-04-21T00:25:36Z | 103 | 0 | null | [
"gguf",
"gpt2",
"dpo",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:HuggingFaceH4/ultrachat_200k",
"dataset:mlabonne/CodeLlama-2-20k",
"dataset:Intel/orca_dpo_pairs",
"base_model:Sharathhebbar24/code_gpt2_mini_model",
"base_model:quantized:Sharathhebbar24/code_gpt2_mini_model",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T15:11:09Z | ---
license: apache-2.0
datasets:
- HuggingFaceH4/ultrachat_200k
- mlabonne/CodeLlama-2-20k
- Intel/orca_dpo_pairs
language:
- en
pipeline_tag: text-generation
tags:
- gpt2
- dpo
- TensorBlock
- GGUF
base_model: Sharathhebbar24/code_gpt2_mini_model
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Sharathhebbar24/code_gpt2_mini_model - GGUF
This repo contains GGUF format model files for [Sharathhebbar24/code_gpt2_mini_model](https://huggingface.co/Sharathhebbar24/code_gpt2_mini_model).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [code_gpt2_mini_model-Q2_K.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q2_K.gguf) | Q2_K | 0.081 GB | smallest, significant quality loss - not recommended for most purposes |
| [code_gpt2_mini_model-Q3_K_S.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q3_K_S.gguf) | Q3_K_S | 0.090 GB | very small, high quality loss |
| [code_gpt2_mini_model-Q3_K_M.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q3_K_M.gguf) | Q3_K_M | 0.098 GB | very small, high quality loss |
| [code_gpt2_mini_model-Q3_K_L.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q3_K_L.gguf) | Q3_K_L | 0.102 GB | small, substantial quality loss |
| [code_gpt2_mini_model-Q4_0.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q4_0.gguf) | Q4_0 | 0.107 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [code_gpt2_mini_model-Q4_K_S.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q4_K_S.gguf) | Q4_K_S | 0.107 GB | small, greater quality loss |
| [code_gpt2_mini_model-Q4_K_M.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q4_K_M.gguf) | Q4_K_M | 0.113 GB | medium, balanced quality - recommended |
| [code_gpt2_mini_model-Q5_0.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q5_0.gguf) | Q5_0 | 0.122 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [code_gpt2_mini_model-Q5_K_S.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q5_K_S.gguf) | Q5_K_S | 0.122 GB | large, low quality loss - recommended |
| [code_gpt2_mini_model-Q5_K_M.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q5_K_M.gguf) | Q5_K_M | 0.127 GB | large, very low quality loss - recommended |
| [code_gpt2_mini_model-Q6_K.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q6_K.gguf) | Q6_K | 0.138 GB | very large, extremely low quality loss |
| [code_gpt2_mini_model-Q8_0.gguf](https://huggingface.co/tensorblock/code_gpt2_mini_model-GGUF/blob/main/code_gpt2_mini_model-Q8_0.gguf) | Q8_0 | 0.178 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/code_gpt2_mini_model-GGUF --include "code_gpt2_mini_model-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/code_gpt2_mini_model-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF | tensorblock | 2025-04-21T00:25:34Z | 108 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"ko",
"en",
"base_model:KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1",
"base_model:quantized:KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-14T13:58:14Z | ---
license: cc-by-nc-4.0
language:
- ko
- en
base_model: KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1 - GGUF
This repo contains GGUF format model files for [KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1](https://huggingface.co/KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q2_K.gguf) | Q2_K | 4.046 GB | smallest, significant quality loss - not recommended for most purposes |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q3_K_S.gguf) | Q3_K_S | 4.711 GB | very small, high quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q3_K_M.gguf) | Q3_K_M | 5.242 GB | very small, high quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q3_K_L.gguf) | Q3_K_L | 5.697 GB | small, substantial quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q4_0.gguf) | Q4_0 | 6.123 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q4_K_S.gguf) | Q4_K_S | 6.169 GB | small, greater quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q4_K_M.gguf) | Q4_K_M | 6.513 GB | medium, balanced quality - recommended |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q5_0.gguf) | Q5_0 | 7.453 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q5_K_S.gguf) | Q5_K_S | 7.453 GB | large, low quality loss - recommended |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q5_K_M.gguf) | Q5_K_M | 7.653 GB | large, very low quality loss - recommended |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q6_K.gguf) | Q6_K | 8.866 GB | very large, extremely low quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q8_0.gguf) | Q8_0 | 11.482 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF --include "KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF | tensorblock | 2025-04-21T00:25:29Z | 97 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"ko",
"base_model:jwkweon/CUBOX-SOLAR-DPO-v0.3",
"base_model:quantized:jwkweon/CUBOX-SOLAR-DPO-v0.3",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-14T12:47:36Z | ---
library_name: transformers
license: apache-2.0
language:
- ko
base_model: jwkweon/CUBOX-SOLAR-DPO-v0.3
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## jwkweon/CUBOX-SOLAR-DPO-v0.3 - GGUF
This repo contains GGUF format model files for [jwkweon/CUBOX-SOLAR-DPO-v0.3](https://huggingface.co/jwkweon/CUBOX-SOLAR-DPO-v0.3).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [CUBOX-SOLAR-DPO-v0.3-Q2_K.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [CUBOX-SOLAR-DPO-v0.3-Q3_K_S.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [CUBOX-SOLAR-DPO-v0.3-Q3_K_M.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [CUBOX-SOLAR-DPO-v0.3-Q3_K_L.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [CUBOX-SOLAR-DPO-v0.3-Q4_0.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [CUBOX-SOLAR-DPO-v0.3-Q4_K_S.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [CUBOX-SOLAR-DPO-v0.3-Q4_K_M.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [CUBOX-SOLAR-DPO-v0.3-Q5_0.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [CUBOX-SOLAR-DPO-v0.3-Q5_K_S.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [CUBOX-SOLAR-DPO-v0.3-Q5_K_M.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [CUBOX-SOLAR-DPO-v0.3-Q6_K.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [CUBOX-SOLAR-DPO-v0.3-Q8_0.gguf](https://huggingface.co/tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF/blob/main/CUBOX-SOLAR-DPO-v0.3-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF --include "CUBOX-SOLAR-DPO-v0.3-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/CUBOX-SOLAR-DPO-v0.3-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF | tensorblock | 2025-04-21T00:25:25Z | 157 | 0 | transformers | [
"transformers",
"gguf",
"llama",
"llama-3",
"TensorBlock",
"GGUF",
"ko",
"en",
"dataset:MarkrAI/KoCommercial-Dataset",
"base_model:PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct",
"base_model:quantized:PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct",
"license:llama3",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-14T11:50:12Z | ---
language:
- ko
- en
license: llama3
library_name: transformers
tags:
- llama
- llama-3
- TensorBlock
- GGUF
base_model: PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct
datasets:
- MarkrAI/KoCommercial-Dataset
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct - GGUF
This repo contains GGUF format model files for [PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct](https://huggingface.co/PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
{prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q2_K.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q2_K.gguf) | Q2_K | 3.179 GB | smallest, significant quality loss - not recommended for most purposes |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q3_K_S.gguf) | Q3_K_S | 3.664 GB | very small, high quality loss |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q3_K_M.gguf) | Q3_K_M | 4.019 GB | very small, high quality loss |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q3_K_L.gguf) | Q3_K_L | 4.322 GB | small, substantial quality loss |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q4_0.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q4_0.gguf) | Q4_0 | 4.661 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q4_K_S.gguf) | Q4_K_S | 4.693 GB | small, greater quality loss |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q4_K_M.gguf) | Q4_K_M | 4.921 GB | medium, balanced quality - recommended |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q5_0.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q5_0.gguf) | Q5_0 | 5.599 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q5_K_S.gguf) | Q5_K_S | 5.599 GB | large, low quality loss - recommended |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q5_K_M.gguf) | Q5_K_M | 5.733 GB | large, very low quality loss - recommended |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q6_K.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q6_K.gguf) | Q6_K | 6.596 GB | very large, extremely low quality loss |
| [Waktaverse-Llama-3-KO-8B-Instruct-Q8_0.gguf](https://huggingface.co/tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF/blob/main/Waktaverse-Llama-3-KO-8B-Instruct-Q8_0.gguf) | Q8_0 | 8.541 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF --include "Waktaverse-Llama-3-KO-8B-Instruct-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Waktaverse-Llama-3-KO-8B-Instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/FusionNet_passthrough-GGUF | tensorblock | 2025-04-21T00:25:21Z | 109 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:TomGrc/FusionNet_passthrough",
"base_model:quantized:TomGrc/FusionNet_passthrough",
"license:mit",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-14T10:48:01Z | ---
language:
- en
license: mit
pipeline_tag: text-generation
base_model: TomGrc/FusionNet_passthrough
tags:
- TensorBlock
- GGUF
model-index:
- name: FusionNet_passthrough
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 69.45
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_passthrough
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 87.72
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_passthrough
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 65.28
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_passthrough
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 67.65
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_passthrough
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 81.29
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_passthrough
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 24.26
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_passthrough
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## TomGrc/FusionNet_passthrough - GGUF
This repo contains GGUF format model files for [TomGrc/FusionNet_passthrough](https://huggingface.co/TomGrc/FusionNet_passthrough).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### System:
{system_prompt}
### User:
{prompt}
### Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [FusionNet_passthrough-Q2_K.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q2_K.gguf) | Q2_K | 7.855 GB | smallest, significant quality loss - not recommended for most purposes |
| [FusionNet_passthrough-Q3_K_S.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q3_K_S.gguf) | Q3_K_S | 9.165 GB | very small, high quality loss |
| [FusionNet_passthrough-Q3_K_M.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q3_K_M.gguf) | Q3_K_M | 10.226 GB | very small, high quality loss |
| [FusionNet_passthrough-Q3_K_L.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q3_K_L.gguf) | Q3_K_L | 11.137 GB | small, substantial quality loss |
| [FusionNet_passthrough-Q4_0.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q4_0.gguf) | Q4_0 | 11.963 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [FusionNet_passthrough-Q4_K_S.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q4_K_S.gguf) | Q4_K_S | 12.053 GB | small, greater quality loss |
| [FusionNet_passthrough-Q4_K_M.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q4_K_M.gguf) | Q4_K_M | 12.741 GB | medium, balanced quality - recommended |
| [FusionNet_passthrough-Q5_0.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q5_0.gguf) | Q5_0 | 14.596 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [FusionNet_passthrough-Q5_K_S.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q5_K_S.gguf) | Q5_K_S | 14.596 GB | large, low quality loss - recommended |
| [FusionNet_passthrough-Q5_K_M.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q5_K_M.gguf) | Q5_K_M | 14.997 GB | large, very low quality loss - recommended |
| [FusionNet_passthrough-Q6_K.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q6_K.gguf) | Q6_K | 17.395 GB | very large, extremely low quality loss |
| [FusionNet_passthrough-Q8_0.gguf](https://huggingface.co/tensorblock/FusionNet_passthrough-GGUF/blob/main/FusionNet_passthrough-Q8_0.gguf) | Q8_0 | 22.529 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/FusionNet_passthrough-GGUF --include "FusionNet_passthrough-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/FusionNet_passthrough-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/opencsg-starcoder2-15b-v0.1-GGUF | tensorblock | 2025-04-21T00:25:16Z | 120 | 0 | transformers | [
"transformers",
"gguf",
"code",
"TensorBlock",
"GGUF",
"text-generation",
"dataset:bigcode/starcoderdata",
"dataset:bigcode/the-stack-dedup",
"base_model:opencsg/opencsg-starcoder2-15b-v0.1",
"base_model:quantized:opencsg/opencsg-starcoder2-15b-v0.1",
"license:bigcode-openrail-m",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T08:52:36Z | ---
license: bigcode-openrail-m
datasets:
- bigcode/starcoderdata
- bigcode/the-stack-dedup
metrics:
- code_eval
library_name: transformers
pipeline_tag: text-generation
tags:
- code
- TensorBlock
- GGUF
base_model: opencsg/opencsg-starcoder2-15b-v0.1
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## opencsg/opencsg-starcoder2-15b-v0.1 - GGUF
This repo contains GGUF format model files for [opencsg/opencsg-starcoder2-15b-v0.1](https://huggingface.co/opencsg/opencsg-starcoder2-15b-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [opencsg-starcoder2-15b-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q2_K.gguf) | Q2_K | 6.193 GB | smallest, significant quality loss - not recommended for most purposes |
| [opencsg-starcoder2-15b-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q3_K_S.gguf) | Q3_K_S | 6.986 GB | very small, high quality loss |
| [opencsg-starcoder2-15b-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q3_K_M.gguf) | Q3_K_M | 8.044 GB | very small, high quality loss |
| [opencsg-starcoder2-15b-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q3_K_L.gguf) | Q3_K_L | 8.965 GB | small, substantial quality loss |
| [opencsg-starcoder2-15b-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q4_0.gguf) | Q4_0 | 9.065 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [opencsg-starcoder2-15b-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q4_K_S.gguf) | Q4_K_S | 9.161 GB | small, greater quality loss |
| [opencsg-starcoder2-15b-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q4_K_M.gguf) | Q4_K_M | 9.860 GB | medium, balanced quality - recommended |
| [opencsg-starcoder2-15b-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q5_0.gguf) | Q5_0 | 11.022 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [opencsg-starcoder2-15b-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q5_K_S.gguf) | Q5_K_S | 11.022 GB | large, low quality loss - recommended |
| [opencsg-starcoder2-15b-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q5_K_M.gguf) | Q5_K_M | 11.431 GB | large, very low quality loss - recommended |
| [opencsg-starcoder2-15b-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q6_K.gguf) | Q6_K | 13.101 GB | very large, extremely low quality loss |
| [opencsg-starcoder2-15b-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/opencsg-starcoder2-15b-v0.1-GGUF/blob/main/opencsg-starcoder2-15b-v0.1-Q8_0.gguf) | Q8_0 | 16.965 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/opencsg-starcoder2-15b-v0.1-GGUF --include "opencsg-starcoder2-15b-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/opencsg-starcoder2-15b-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/deepseek-math-7b-instruct-GGUF | tensorblock | 2025-04-21T00:25:14Z | 150 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:deepseek-ai/deepseek-math-7b-instruct",
"base_model:quantized:deepseek-ai/deepseek-math-7b-instruct",
"license:other",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-14T08:02:10Z | ---
license: other
license_name: deepseek
license_link: https://github.com/deepseek-ai/DeepSeek-Math/blob/main/LICENSE-MODEL
base_model: deepseek-ai/deepseek-math-7b-instruct
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## deepseek-ai/deepseek-math-7b-instruct - GGUF
This repo contains GGUF format model files for [deepseek-ai/deepseek-math-7b-instruct](https://huggingface.co/deepseek-ai/deepseek-math-7b-instruct).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<ο½beginβofβsentenceο½>{system_prompt}
User: {prompt}
Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [deepseek-math-7b-instruct-Q2_K.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q2_K.gguf) | Q2_K | 2.718 GB | smallest, significant quality loss - not recommended for most purposes |
| [deepseek-math-7b-instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q3_K_S.gguf) | Q3_K_S | 3.138 GB | very small, high quality loss |
| [deepseek-math-7b-instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q3_K_M.gguf) | Q3_K_M | 3.461 GB | very small, high quality loss |
| [deepseek-math-7b-instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q3_K_L.gguf) | Q3_K_L | 3.746 GB | small, substantial quality loss |
| [deepseek-math-7b-instruct-Q4_0.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q4_0.gguf) | Q4_0 | 4.000 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [deepseek-math-7b-instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q4_K_S.gguf) | Q4_K_S | 4.025 GB | small, greater quality loss |
| [deepseek-math-7b-instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q4_K_M.gguf) | Q4_K_M | 4.223 GB | medium, balanced quality - recommended |
| [deepseek-math-7b-instruct-Q5_0.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q5_0.gguf) | Q5_0 | 4.811 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [deepseek-math-7b-instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q5_K_S.gguf) | Q5_K_S | 4.811 GB | large, low quality loss - recommended |
| [deepseek-math-7b-instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q5_K_M.gguf) | Q5_K_M | 4.926 GB | large, very low quality loss - recommended |
| [deepseek-math-7b-instruct-Q6_K.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q6_K.gguf) | Q6_K | 5.673 GB | very large, extremely low quality loss |
| [deepseek-math-7b-instruct-Q8_0.gguf](https://huggingface.co/tensorblock/deepseek-math-7b-instruct-GGUF/blob/main/deepseek-math-7b-instruct-Q8_0.gguf) | Q8_0 | 7.347 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/deepseek-math-7b-instruct-GGUF --include "deepseek-math-7b-instruct-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/deepseek-math-7b-instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/mistral-orpo-capybara-7k-GGUF | tensorblock | 2025-04-21T00:24:57Z | 91 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:argilla/distilabel-capybara-dpo-7k-binarized",
"base_model:kaist-ai/mistral-orpo-capybara-7k",
"base_model:quantized:kaist-ai/mistral-orpo-capybara-7k",
"license:mit",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-14T04:11:22Z | ---
language:
- en
license: mit
base_model: kaist-ai/mistral-orpo-capybara-7k
datasets:
- argilla/distilabel-capybara-dpo-7k-binarized
pipeline_tag: text-generation
tags:
- TensorBlock
- GGUF
model-index:
- name: Mistral-ORPO-Capybara-7k
results:
- task:
type: text-generation
dataset:
name: AlpacaEval 2 (LC)
type: AlpacaEval
metrics:
- type: AlpacaEval 2.0
value: 15.88%
name: Win Rate
source:
url: https://tatsu-lab.github.io/alpaca_eval/
name: self-reported
- task:
type: text-generation
dataset:
name: MT-Bench
type: MT-Bench
metrics:
- type: MT-Bench
value: 7.444
name: Score
source:
url: https://github.com/lm-sys/FastChat/blob/main/fastchat/llm_judge/
name: self-reported
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## kaist-ai/mistral-orpo-capybara-7k - GGUF
This repo contains GGUF format model files for [kaist-ai/mistral-orpo-capybara-7k](https://huggingface.co/kaist-ai/mistral-orpo-capybara-7k).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [mistral-orpo-capybara-7k-Q2_K.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [mistral-orpo-capybara-7k-Q3_K_S.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [mistral-orpo-capybara-7k-Q3_K_M.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [mistral-orpo-capybara-7k-Q3_K_L.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [mistral-orpo-capybara-7k-Q4_0.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [mistral-orpo-capybara-7k-Q4_K_S.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [mistral-orpo-capybara-7k-Q4_K_M.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [mistral-orpo-capybara-7k-Q5_0.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [mistral-orpo-capybara-7k-Q5_K_S.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [mistral-orpo-capybara-7k-Q5_K_M.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [mistral-orpo-capybara-7k-Q6_K.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [mistral-orpo-capybara-7k-Q8_0.gguf](https://huggingface.co/tensorblock/mistral-orpo-capybara-7k-GGUF/blob/main/mistral-orpo-capybara-7k-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/mistral-orpo-capybara-7k-GGUF --include "mistral-orpo-capybara-7k-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/mistral-orpo-capybara-7k-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/llama2.c-stories42M-pruned2.4-GGUF | tensorblock | 2025-04-21T00:24:55Z | 29 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:nm-testing/llama2.c-stories42M-pruned2.4",
"base_model:quantized:nm-testing/llama2.c-stories42M-pruned2.4",
"endpoints_compatible",
"region:us"
] | null | 2024-12-14T03:24:24Z | ---
base_model: nm-testing/llama2.c-stories42M-pruned2.4
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## nm-testing/llama2.c-stories42M-pruned2.4 - GGUF
This repo contains GGUF format model files for [nm-testing/llama2.c-stories42M-pruned2.4](https://huggingface.co/nm-testing/llama2.c-stories42M-pruned2.4).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [llama2.c-stories42M-pruned2.4-Q2_K.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q2_K.gguf) | Q2_K | 0.030 GB | smallest, significant quality loss - not recommended for most purposes |
| [llama2.c-stories42M-pruned2.4-Q3_K_S.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q3_K_S.gguf) | Q3_K_S | 0.033 GB | very small, high quality loss |
| [llama2.c-stories42M-pruned2.4-Q3_K_M.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q3_K_M.gguf) | Q3_K_M | 0.034 GB | very small, high quality loss |
| [llama2.c-stories42M-pruned2.4-Q3_K_L.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q3_K_L.gguf) | Q3_K_L | 0.035 GB | small, substantial quality loss |
| [llama2.c-stories42M-pruned2.4-Q4_0.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q4_0.gguf) | Q4_0 | 0.038 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [llama2.c-stories42M-pruned2.4-Q4_K_S.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q4_K_S.gguf) | Q4_K_S | 0.039 GB | small, greater quality loss |
| [llama2.c-stories42M-pruned2.4-Q4_K_M.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q4_K_M.gguf) | Q4_K_M | 0.040 GB | medium, balanced quality - recommended |
| [llama2.c-stories42M-pruned2.4-Q5_0.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q5_0.gguf) | Q5_0 | 0.043 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [llama2.c-stories42M-pruned2.4-Q5_K_S.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q5_K_S.gguf) | Q5_K_S | 0.043 GB | large, low quality loss - recommended |
| [llama2.c-stories42M-pruned2.4-Q5_K_M.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q5_K_M.gguf) | Q5_K_M | 0.044 GB | large, very low quality loss - recommended |
| [llama2.c-stories42M-pruned2.4-Q6_K.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q6_K.gguf) | Q6_K | 0.050 GB | very large, extremely low quality loss |
| [llama2.c-stories42M-pruned2.4-Q8_0.gguf](https://huggingface.co/tensorblock/llama2.c-stories42M-pruned2.4-GGUF/blob/main/llama2.c-stories42M-pruned2.4-Q8_0.gguf) | Q8_0 | 0.062 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/llama2.c-stories42M-pruned2.4-GGUF --include "llama2.c-stories42M-pruned2.4-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/llama2.c-stories42M-pruned2.4-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Mixtral_AI_Cyber_2.0-GGUF | tensorblock | 2025-04-21T00:24:49Z | 96 | 0 | transformers | [
"transformers",
"gguf",
"mergekit",
"merge",
"128k_Context",
"chemistry",
"biology",
"music",
"code",
"medical",
"text-generation-inference",
"Cyber-Series",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:LeroyDyer/Mixtral_AI_Cyber_2.0",
"base_model:quantized:LeroyDyer/Mixtral_AI_Cyber_2.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-14T03:16:09Z | ---
base_model: LeroyDyer/Mixtral_AI_Cyber_2.0
library_name: transformers
tags:
- mergekit
- merge
- 128k_Context
- chemistry
- biology
- music
- code
- medical
- text-generation-inference
- Cyber-Series
- TensorBlock
- GGUF
previous_Merges:
- rvv-karma/BASH-Coder-Mistral-7B
- Locutusque/Hercules-3.1-Mistral-7B
- KoboldAI/Mistral-7B-Erebus-v3 - NSFW
- Locutusque/Hyperion-2.1-Mistral-7B
- Severian/Nexus-IKM-Mistral-7B-Pytorch
- NousResearch/Hermes-2-Pro-Mistral-7B
- mistralai/Mistral-7B-Instruct-v0.2
- Nitral-AI/ProdigyXBioMistral_7B
- Nitral-AI/Infinite-Mika-7b
- Nous-Yarn-Mistral-7b-128k
- yanismiraoui/Yarn-Mistral-7b-128k-sharded
license: apache-2.0
language:
- en
metrics:
- accuracy
- brier_score
- code_eval
pipeline_tag: text-generation
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## LeroyDyer/Mixtral_AI_Cyber_2.0 - GGUF
This repo contains GGUF format model files for [LeroyDyer/Mixtral_AI_Cyber_2.0](https://huggingface.co/LeroyDyer/Mixtral_AI_Cyber_2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Mixtral_AI_Cyber_2.0-Q2_K.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Mixtral_AI_Cyber_2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Mixtral_AI_Cyber_2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Mixtral_AI_Cyber_2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Mixtral_AI_Cyber_2.0-Q4_0.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Mixtral_AI_Cyber_2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Mixtral_AI_Cyber_2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Mixtral_AI_Cyber_2.0-Q5_0.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Mixtral_AI_Cyber_2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Mixtral_AI_Cyber_2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Mixtral_AI_Cyber_2.0-Q6_K.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Mixtral_AI_Cyber_2.0-Q8_0.gguf](https://huggingface.co/tensorblock/Mixtral_AI_Cyber_2.0-GGUF/blob/main/Mixtral_AI_Cyber_2.0-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Mixtral_AI_Cyber_2.0-GGUF --include "Mixtral_AI_Cyber_2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Mixtral_AI_Cyber_2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/b1ade-1b-bf16-GGUF | tensorblock | 2025-04-21T00:24:36Z | 12 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"dataset:kaist-ai/CoT-Collection",
"base_model:w601sxs/b1ade-1b-bf16",
"base_model:quantized:w601sxs/b1ade-1b-bf16",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T23:51:50Z | ---
library_name: transformers
datasets:
- kaist-ai/CoT-Collection
tags:
- TensorBlock
- GGUF
base_model: w601sxs/b1ade-1b-bf16
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## w601sxs/b1ade-1b-bf16 - GGUF
This repo contains GGUF format model files for [w601sxs/b1ade-1b-bf16](https://huggingface.co/w601sxs/b1ade-1b-bf16).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [b1ade-1b-bf16-Q2_K.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q2_K.gguf) | Q2_K | 0.420 GB | smallest, significant quality loss - not recommended for most purposes |
| [b1ade-1b-bf16-Q3_K_S.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q3_K_S.gguf) | Q3_K_S | 0.478 GB | very small, high quality loss |
| [b1ade-1b-bf16-Q3_K_M.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q3_K_M.gguf) | Q3_K_M | 0.552 GB | very small, high quality loss |
| [b1ade-1b-bf16-Q3_K_L.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q3_K_L.gguf) | Q3_K_L | 0.592 GB | small, substantial quality loss |
| [b1ade-1b-bf16-Q4_0.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q4_0.gguf) | Q4_0 | 0.599 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [b1ade-1b-bf16-Q4_K_S.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q4_K_S.gguf) | Q4_K_S | 0.603 GB | small, greater quality loss |
| [b1ade-1b-bf16-Q4_K_M.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q4_K_M.gguf) | Q4_K_M | 0.659 GB | medium, balanced quality - recommended |
| [b1ade-1b-bf16-Q5_0.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q5_0.gguf) | Q5_0 | 0.712 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [b1ade-1b-bf16-Q5_K_S.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q5_K_S.gguf) | Q5_K_S | 0.712 GB | large, low quality loss - recommended |
| [b1ade-1b-bf16-Q5_K_M.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q5_K_M.gguf) | Q5_K_M | 0.757 GB | large, very low quality loss - recommended |
| [b1ade-1b-bf16-Q6_K.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q6_K.gguf) | Q6_K | 0.833 GB | very large, extremely low quality loss |
| [b1ade-1b-bf16-Q8_0.gguf](https://huggingface.co/tensorblock/b1ade-1b-bf16-GGUF/blob/main/b1ade-1b-bf16-Q8_0.gguf) | Q8_0 | 1.078 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/b1ade-1b-bf16-GGUF --include "b1ade-1b-bf16-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/b1ade-1b-bf16-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/spin_gpt2_medium_alpaca_e4-GGUF | tensorblock | 2025-04-21T00:24:32Z | 8 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"license:mit",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T23:47:40Z | ---
license: mit
tags:
- TensorBlock
- GGUF
base_model: LordNoah/spin_gpt2_medium_alpaca_e4
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## LordNoah/spin_gpt2_medium_alpaca_e4 - GGUF
This repo contains GGUF format model files for [LordNoah/spin_gpt2_medium_alpaca_e4](https://huggingface.co/LordNoah/spin_gpt2_medium_alpaca_e4).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [spin_gpt2_medium_alpaca_e4-Q2_K.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q2_K.gguf) | Q2_K | 0.178 GB | smallest, significant quality loss - not recommended for most purposes |
| [spin_gpt2_medium_alpaca_e4-Q3_K_S.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q3_K_S.gguf) | Q3_K_S | 0.201 GB | very small, high quality loss |
| [spin_gpt2_medium_alpaca_e4-Q3_K_M.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q3_K_M.gguf) | Q3_K_M | 0.229 GB | very small, high quality loss |
| [spin_gpt2_medium_alpaca_e4-Q3_K_L.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q3_K_L.gguf) | Q3_K_L | 0.244 GB | small, substantial quality loss |
| [spin_gpt2_medium_alpaca_e4-Q4_0.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q4_0.gguf) | Q4_0 | 0.248 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [spin_gpt2_medium_alpaca_e4-Q4_K_S.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q4_K_S.gguf) | Q4_K_S | 0.250 GB | small, greater quality loss |
| [spin_gpt2_medium_alpaca_e4-Q4_K_M.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q4_K_M.gguf) | Q4_K_M | 0.271 GB | medium, balanced quality - recommended |
| [spin_gpt2_medium_alpaca_e4-Q5_0.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q5_0.gguf) | Q5_0 | 0.292 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [spin_gpt2_medium_alpaca_e4-Q5_K_S.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q5_K_S.gguf) | Q5_K_S | 0.292 GB | large, low quality loss - recommended |
| [spin_gpt2_medium_alpaca_e4-Q5_K_M.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q5_K_M.gguf) | Q5_K_M | 0.309 GB | large, very low quality loss - recommended |
| [spin_gpt2_medium_alpaca_e4-Q6_K.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q6_K.gguf) | Q6_K | 0.339 GB | very large, extremely low quality loss |
| [spin_gpt2_medium_alpaca_e4-Q8_0.gguf](https://huggingface.co/tensorblock/spin_gpt2_medium_alpaca_e4-GGUF/blob/main/spin_gpt2_medium_alpaca_e4-Q8_0.gguf) | Q8_0 | 0.437 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/spin_gpt2_medium_alpaca_e4-GGUF --include "spin_gpt2_medium_alpaca_e4-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/spin_gpt2_medium_alpaca_e4-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/gemma-2b-openhermes-GGUF | tensorblock | 2025-04-21T00:24:29Z | 27 | 0 | transformers | [
"transformers",
"gguf",
"generated_from_trainer",
"axolotl",
"gemma",
"instruct",
"finetune",
"chatml",
"gpt4",
"synthetic data",
"distillation",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:mlabonne/chatml-OpenHermes2.5-dpo-binarized-alpha",
"base_model:abideen/gemma-2b-openhermes",
"base_model:quantized:abideen/gemma-2b-openhermes",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T23:43:35Z | ---
license: cc-by-nc-4.0
base_model: abideen/gemma-2b-openhermes
tags:
- generated_from_trainer
- axolotl
- gemma
- instruct
- finetune
- chatml
- gpt4
- synthetic data
- distillation
- TensorBlock
- GGUF
datasets:
- mlabonne/chatml-OpenHermes2.5-dpo-binarized-alpha
language:
- en
library_name: transformers
pipeline_tag: text-generation
model-index:
- name: gemma-2b-openhermes
results: []
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## abideen/gemma-2b-openhermes - GGUF
This repo contains GGUF format model files for [abideen/gemma-2b-openhermes](https://huggingface.co/abideen/gemma-2b-openhermes).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<start_of_turn>user
{prompt}<end_of_turn>
<start_of_turn>model
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [gemma-2b-openhermes-Q2_K.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q2_K.gguf) | Q2_K | 1.158 GB | smallest, significant quality loss - not recommended for most purposes |
| [gemma-2b-openhermes-Q3_K_S.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q3_K_S.gguf) | Q3_K_S | 1.288 GB | very small, high quality loss |
| [gemma-2b-openhermes-Q3_K_M.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q3_K_M.gguf) | Q3_K_M | 1.384 GB | very small, high quality loss |
| [gemma-2b-openhermes-Q3_K_L.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q3_K_L.gguf) | Q3_K_L | 1.466 GB | small, substantial quality loss |
| [gemma-2b-openhermes-Q4_0.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q4_0.gguf) | Q4_0 | 1.551 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [gemma-2b-openhermes-Q4_K_S.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q4_K_S.gguf) | Q4_K_S | 1.560 GB | small, greater quality loss |
| [gemma-2b-openhermes-Q4_K_M.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q4_K_M.gguf) | Q4_K_M | 1.630 GB | medium, balanced quality - recommended |
| [gemma-2b-openhermes-Q5_0.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q5_0.gguf) | Q5_0 | 1.799 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [gemma-2b-openhermes-Q5_K_S.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q5_K_S.gguf) | Q5_K_S | 1.799 GB | large, low quality loss - recommended |
| [gemma-2b-openhermes-Q5_K_M.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q5_K_M.gguf) | Q5_K_M | 1.840 GB | large, very low quality loss - recommended |
| [gemma-2b-openhermes-Q6_K.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q6_K.gguf) | Q6_K | 2.062 GB | very large, extremely low quality loss |
| [gemma-2b-openhermes-Q8_0.gguf](https://huggingface.co/tensorblock/gemma-2b-openhermes-GGUF/blob/main/gemma-2b-openhermes-Q8_0.gguf) | Q8_0 | 2.669 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/gemma-2b-openhermes-GGUF --include "gemma-2b-openhermes-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/gemma-2b-openhermes-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/T3Q-ko-solar-sft-v1.0-GGUF | tensorblock | 2025-04-21T00:24:23Z | 102 | 0 | null | [
"gguf",
"T3Q-ko-solar-sft-v1.0",
"kyujinpy/KoCommercial-NoSSL",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:kyujinpy/KoCommercial-NoSSL",
"base_model:chlee10/T3Q-ko-solar-sft-v1.0",
"base_model:quantized:chlee10/T3Q-ko-solar-sft-v1.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T22:12:34Z | ---
pipeline_tag: text-generation
license: apache-2.0
language:
- en
tags:
- T3Q-ko-solar-sft-v1.0
- kyujinpy/KoCommercial-NoSSL
- TensorBlock
- GGUF
base_model: chlee10/T3Q-ko-solar-sft-v1.0
datasets:
- kyujinpy/KoCommercial-NoSSL
model-index:
- name: T3Q-ko-solar-sft-v1.0
results: []
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## chlee10/T3Q-ko-solar-sft-v1.0 - GGUF
This repo contains GGUF format model files for [chlee10/T3Q-ko-solar-sft-v1.0](https://huggingface.co/chlee10/T3Q-ko-solar-sft-v1.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### System:
{system_prompt}
### User:
{prompt}
### Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [T3Q-ko-solar-sft-v1.0-Q2_K.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [T3Q-ko-solar-sft-v1.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [T3Q-ko-solar-sft-v1.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [T3Q-ko-solar-sft-v1.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [T3Q-ko-solar-sft-v1.0-Q4_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [T3Q-ko-solar-sft-v1.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [T3Q-ko-solar-sft-v1.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [T3Q-ko-solar-sft-v1.0-Q5_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [T3Q-ko-solar-sft-v1.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [T3Q-ko-solar-sft-v1.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [T3Q-ko-solar-sft-v1.0-Q6_K.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [T3Q-ko-solar-sft-v1.0-Q8_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v1.0-GGUF/blob/main/T3Q-ko-solar-sft-v1.0-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/T3Q-ko-solar-sft-v1.0-GGUF --include "T3Q-ko-solar-sft-v1.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/T3Q-ko-solar-sft-v1.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF | tensorblock | 2025-04-21T00:24:20Z | 38 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:Locutusque/Hercules-v3.0",
"dataset:Locutusque/hyperion-v2.0",
"dataset:argilla/OpenHermes2.5-dpo-binarized-alpha",
"base_model:frankenmerger/MiniLlama-1.8b-Chat-v0.1",
"base_model:quantized:frankenmerger/MiniLlama-1.8b-Chat-v0.1",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T21:49:30Z | ---
widget:
- messages:
- role: system
content: You are a career counselor. The user will provide you with an individual
looking for guidance in their professional life, and your task is to assist
them in determining what careers they are most suited for based on their skills,
interests, and experience. You should also conduct research into the various
options available, explain the job market trends in different industries, and
advice on which qualifications would be beneficial for pursuing particular fields.
- role: user
content: Hey friend!
- role: assistant
content: Hi! How may I help you?
- role: user
content: I am interested in developing a career in software engineering. What
would you recommend me to do?
- messages:
- role: system
content: You are a knowledgeable assistant. Help the user as much as you can.
- role: user
content: How to become smarter?
- messages:
- role: system
content: You are a helpful assistant who provides concise responses.
- role: user
content: Hi!
- role: assistant
content: Hello there! How may I help you?
- role: user
content: I need to cook a simple dinner. What ingredients should I prepare for?
- messages:
- role: system
content: You are a very creative assistant. User will give you a task, which you
should complete with all your knowledge.
- role: user
content: Write the novel story of an RPG game about group of survivor post apocalyptic
world.
inference:
parameters:
max_new_tokens: 256
temperature: 0.6
top_p: 0.95
top_k: 50
repetition_penalty: 1.2
base_model: frankenmerger/MiniLlama-1.8b-Chat-v0.1
license: apache-2.0
language:
- en
pipeline_tag: text-generation
datasets:
- Locutusque/Hercules-v3.0
- Locutusque/hyperion-v2.0
- argilla/OpenHermes2.5-dpo-binarized-alpha
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## frankenmerger/MiniLlama-1.8b-Chat-v0.1 - GGUF
This repo contains GGUF format model files for [frankenmerger/MiniLlama-1.8b-Chat-v0.1](https://huggingface.co/frankenmerger/MiniLlama-1.8b-Chat-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [MiniLlama-1.8b-Chat-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q2_K.gguf) | Q2_K | 0.724 GB | smallest, significant quality loss - not recommended for most purposes |
| [MiniLlama-1.8b-Chat-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q3_K_S.gguf) | Q3_K_S | 0.840 GB | very small, high quality loss |
| [MiniLlama-1.8b-Chat-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q3_K_M.gguf) | Q3_K_M | 0.930 GB | very small, high quality loss |
| [MiniLlama-1.8b-Chat-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q3_K_L.gguf) | Q3_K_L | 1.008 GB | small, substantial quality loss |
| [MiniLlama-1.8b-Chat-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q4_0.gguf) | Q4_0 | 1.083 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [MiniLlama-1.8b-Chat-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q4_K_S.gguf) | Q4_K_S | 1.090 GB | small, greater quality loss |
| [MiniLlama-1.8b-Chat-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q4_K_M.gguf) | Q4_K_M | 1.145 GB | medium, balanced quality - recommended |
| [MiniLlama-1.8b-Chat-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q5_0.gguf) | Q5_0 | 1.311 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [MiniLlama-1.8b-Chat-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q5_K_S.gguf) | Q5_K_S | 1.311 GB | large, low quality loss - recommended |
| [MiniLlama-1.8b-Chat-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q5_K_M.gguf) | Q5_K_M | 1.343 GB | large, very low quality loss - recommended |
| [MiniLlama-1.8b-Chat-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q6_K.gguf) | Q6_K | 1.554 GB | very large, extremely low quality loss |
| [MiniLlama-1.8b-Chat-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF/blob/main/MiniLlama-1.8b-Chat-v0.1-Q8_0.gguf) | Q8_0 | 2.012 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF --include "MiniLlama-1.8b-Chat-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/MiniLlama-1.8b-Chat-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/pythia-410m-sentiment-first-ft-GGUF | tensorblock | 2025-04-21T00:24:18Z | 11 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"base_model:EleutherAI/pythia-410m-sentiment-first-ft",
"base_model:quantized:EleutherAI/pythia-410m-sentiment-first-ft",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T21:18:44Z | ---
library_name: transformers
tags:
- TensorBlock
- GGUF
base_model: EleutherAI/pythia-410m-sentiment-first-ft
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## EleutherAI/pythia-410m-sentiment-first-ft - GGUF
This repo contains GGUF format model files for [EleutherAI/pythia-410m-sentiment-first-ft](https://huggingface.co/EleutherAI/pythia-410m-sentiment-first-ft).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [pythia-410m-sentiment-first-ft-Q2_K.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q2_K.gguf) | Q2_K | 0.174 GB | smallest, significant quality loss - not recommended for most purposes |
| [pythia-410m-sentiment-first-ft-Q3_K_S.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q3_K_S.gguf) | Q3_K_S | 0.197 GB | very small, high quality loss |
| [pythia-410m-sentiment-first-ft-Q3_K_M.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q3_K_M.gguf) | Q3_K_M | 0.224 GB | very small, high quality loss |
| [pythia-410m-sentiment-first-ft-Q3_K_L.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q3_K_L.gguf) | Q3_K_L | 0.240 GB | small, substantial quality loss |
| [pythia-410m-sentiment-first-ft-Q4_0.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q4_0.gguf) | Q4_0 | 0.244 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [pythia-410m-sentiment-first-ft-Q4_K_S.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q4_K_S.gguf) | Q4_K_S | 0.246 GB | small, greater quality loss |
| [pythia-410m-sentiment-first-ft-Q4_K_M.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q4_K_M.gguf) | Q4_K_M | 0.267 GB | medium, balanced quality - recommended |
| [pythia-410m-sentiment-first-ft-Q5_0.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q5_0.gguf) | Q5_0 | 0.288 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [pythia-410m-sentiment-first-ft-Q5_K_S.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q5_K_S.gguf) | Q5_K_S | 0.288 GB | large, low quality loss - recommended |
| [pythia-410m-sentiment-first-ft-Q5_K_M.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q5_K_M.gguf) | Q5_K_M | 0.305 GB | large, very low quality loss - recommended |
| [pythia-410m-sentiment-first-ft-Q6_K.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q6_K.gguf) | Q6_K | 0.335 GB | very large, extremely low quality loss |
| [pythia-410m-sentiment-first-ft-Q8_0.gguf](https://huggingface.co/tensorblock/pythia-410m-sentiment-first-ft-GGUF/blob/main/pythia-410m-sentiment-first-ft-Q8_0.gguf) | Q8_0 | 0.433 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/pythia-410m-sentiment-first-ft-GGUF --include "pythia-410m-sentiment-first-ft-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/pythia-410m-sentiment-first-ft-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/stablelm-2-12b-GGUF | tensorblock | 2025-04-21T00:24:16Z | 150 | 0 | null | [
"gguf",
"causal-lm",
"TensorBlock",
"GGUF",
"en",
"de",
"es",
"fr",
"it",
"nl",
"pt",
"dataset:tiiuae/falcon-refinedweb",
"dataset:togethercomputer/RedPajama-Data-1T",
"dataset:uonlp/CulturaX",
"dataset:CarperAI/pilev2-dev",
"dataset:bigcode/starcoderdata",
"dataset:DataProvenanceInitiative/Commercially-Verified-Licenses",
"base_model:stabilityai/stablelm-2-12b",
"base_model:quantized:stabilityai/stablelm-2-12b",
"license:other",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T21:10:37Z | ---
language:
- en
- de
- es
- fr
- it
- nl
- pt
license: other
tags:
- causal-lm
- TensorBlock
- GGUF
datasets:
- tiiuae/falcon-refinedweb
- togethercomputer/RedPajama-Data-1T
- uonlp/CulturaX
- CarperAI/pilev2-dev
- bigcode/starcoderdata
- DataProvenanceInitiative/Commercially-Verified-Licenses
base_model: stabilityai/stablelm-2-12b
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## stabilityai/stablelm-2-12b - GGUF
This repo contains GGUF format model files for [stabilityai/stablelm-2-12b](https://huggingface.co/stabilityai/stablelm-2-12b).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [stablelm-2-12b-Q2_K.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q2_K.gguf) | Q2_K | 4.699 GB | smallest, significant quality loss - not recommended for most purposes |
| [stablelm-2-12b-Q3_K_S.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q3_K_S.gguf) | Q3_K_S | 5.424 GB | very small, high quality loss |
| [stablelm-2-12b-Q3_K_M.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q3_K_M.gguf) | Q3_K_M | 5.994 GB | very small, high quality loss |
| [stablelm-2-12b-Q3_K_L.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q3_K_L.gguf) | Q3_K_L | 6.492 GB | small, substantial quality loss |
| [stablelm-2-12b-Q4_0.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q4_0.gguf) | Q4_0 | 6.969 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [stablelm-2-12b-Q4_K_S.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q4_K_S.gguf) | Q4_K_S | 7.016 GB | small, greater quality loss |
| [stablelm-2-12b-Q4_K_M.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q4_K_M.gguf) | Q4_K_M | 7.368 GB | medium, balanced quality - recommended |
| [stablelm-2-12b-Q5_0.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q5_0.gguf) | Q5_0 | 8.422 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [stablelm-2-12b-Q5_K_S.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q5_K_S.gguf) | Q5_K_S | 8.422 GB | large, low quality loss - recommended |
| [stablelm-2-12b-Q5_K_M.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q5_K_M.gguf) | Q5_K_M | 8.628 GB | large, very low quality loss - recommended |
| [stablelm-2-12b-Q6_K.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q6_K.gguf) | Q6_K | 9.967 GB | very large, extremely low quality loss |
| [stablelm-2-12b-Q8_0.gguf](https://huggingface.co/tensorblock/stablelm-2-12b-GGUF/blob/main/stablelm-2-12b-Q8_0.gguf) | Q8_0 | 12.908 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/stablelm-2-12b-GGUF --include "stablelm-2-12b-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/stablelm-2-12b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/nous-0-GGUF | tensorblock | 2025-04-21T00:24:14Z | 13 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"base_model:kalytm/nous-0",
"base_model:quantized:kalytm/nous-0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T20:09:17Z | ---
library_name: transformers
tags:
- TensorBlock
- GGUF
base_model: kalytm/nous-0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## kalytm/nous-0 - GGUF
This repo contains GGUF format model files for [kalytm/nous-0](https://huggingface.co/kalytm/nous-0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}<|endoftext|>
<|user|>
{prompt}<|endoftext|>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [nous-0-Q2_K.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q2_K.gguf) | Q2_K | 0.694 GB | smallest, significant quality loss - not recommended for most purposes |
| [nous-0-Q3_K_S.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q3_K_S.gguf) | Q3_K_S | 0.792 GB | very small, high quality loss |
| [nous-0-Q3_K_M.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q3_K_M.gguf) | Q3_K_M | 0.858 GB | very small, high quality loss |
| [nous-0-Q3_K_L.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q3_K_L.gguf) | Q3_K_L | 0.915 GB | small, substantial quality loss |
| [nous-0-Q4_0.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q4_0.gguf) | Q4_0 | 0.983 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [nous-0-Q4_K_S.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q4_K_S.gguf) | Q4_K_S | 0.989 GB | small, greater quality loss |
| [nous-0-Q4_K_M.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q4_K_M.gguf) | Q4_K_M | 1.031 GB | medium, balanced quality - recommended |
| [nous-0-Q5_0.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q5_0.gguf) | Q5_0 | 1.163 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [nous-0-Q5_K_S.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q5_K_S.gguf) | Q5_K_S | 1.163 GB | large, low quality loss - recommended |
| [nous-0-Q5_K_M.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q5_K_M.gguf) | Q5_K_M | 1.188 GB | large, very low quality loss - recommended |
| [nous-0-Q6_K.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q6_K.gguf) | Q6_K | 1.354 GB | very large, extremely low quality loss |
| [nous-0-Q8_0.gguf](https://huggingface.co/tensorblock/nous-0-GGUF/blob/main/nous-0-Q8_0.gguf) | Q8_0 | 1.752 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/nous-0-GGUF --include "nous-0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/nous-0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Llama-3-13B-Instruct-v0.1-GGUF | tensorblock | 2025-04-21T00:24:05Z | 140 | 0 | transformers | [
"transformers",
"gguf",
"mergekit",
"merge",
"facebook",
"meta",
"pytorch",
"llama",
"llama-3",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"base_model:MaziyarPanahi/Llama-3-13B-Instruct-v0.1",
"base_model:quantized:MaziyarPanahi/Llama-3-13B-Instruct-v0.1",
"license:other",
"region:us",
"conversational"
] | text-generation | 2024-12-13T19:05:54Z | ---
base_model: MaziyarPanahi/Llama-3-13B-Instruct-v0.1
library_name: transformers
tags:
- mergekit
- merge
- facebook
- meta
- pytorch
- llama
- llama-3
- TensorBlock
- GGUF
language:
- en
pipeline_tag: text-generation
license: other
license_name: llama3
license_link: LICENSE
inference: false
model_creator: MaziyarPanahi
model_name: Llama-3-13B-Instruct-v0.1
quantized_by: MaziyarPanahi
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## MaziyarPanahi/Llama-3-13B-Instruct-v0.1 - GGUF
This repo contains GGUF format model files for [MaziyarPanahi/Llama-3-13B-Instruct-v0.1](https://huggingface.co/MaziyarPanahi/Llama-3-13B-Instruct-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
{prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Llama-3-13B-Instruct-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q2_K.gguf) | Q2_K | 5.105 GB | smallest, significant quality loss - not recommended for most purposes |
| [Llama-3-13B-Instruct-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q3_K_S.gguf) | Q3_K_S | 5.914 GB | very small, high quality loss |
| [Llama-3-13B-Instruct-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q3_K_M.gguf) | Q3_K_M | 6.530 GB | very small, high quality loss |
| [Llama-3-13B-Instruct-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q3_K_L.gguf) | Q3_K_L | 7.065 GB | small, substantial quality loss |
| [Llama-3-13B-Instruct-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q4_0.gguf) | Q4_0 | 7.606 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Llama-3-13B-Instruct-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q4_K_S.gguf) | Q4_K_S | 7.660 GB | small, greater quality loss |
| [Llama-3-13B-Instruct-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q4_K_M.gguf) | Q4_K_M | 8.061 GB | medium, balanced quality - recommended |
| [Llama-3-13B-Instruct-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q5_0.gguf) | Q5_0 | 9.199 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Llama-3-13B-Instruct-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q5_K_S.gguf) | Q5_K_S | 9.199 GB | large, low quality loss - recommended |
| [Llama-3-13B-Instruct-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q5_K_M.gguf) | Q5_K_M | 9.433 GB | large, very low quality loss - recommended |
| [Llama-3-13B-Instruct-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q6_K.gguf) | Q6_K | 10.891 GB | very large, extremely low quality loss |
| [Llama-3-13B-Instruct-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/Llama-3-13B-Instruct-v0.1-GGUF/blob/main/Llama-3-13B-Instruct-v0.1-Q8_0.gguf) | Q8_0 | 14.103 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Llama-3-13B-Instruct-v0.1-GGUF --include "Llama-3-13B-Instruct-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Llama-3-13B-Instruct-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/llama2-13b-lima-sft-dpo-GGUF | tensorblock | 2025-04-21T00:24:04Z | 37 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:etri-xainlp/llama2-13b-lima-sft-dpo",
"base_model:quantized:etri-xainlp/llama2-13b-lima-sft-dpo",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T18:18:50Z | ---
license: apache-2.0
base_model: etri-xainlp/llama2-13b-lima-sft-dpo
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## etri-xainlp/llama2-13b-lima-sft-dpo - GGUF
This repo contains GGUF format model files for [etri-xainlp/llama2-13b-lima-sft-dpo](https://huggingface.co/etri-xainlp/llama2-13b-lima-sft-dpo).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [llama2-13b-lima-sft-dpo-Q2_K.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q2_K.gguf) | Q2_K | 4.854 GB | smallest, significant quality loss - not recommended for most purposes |
| [llama2-13b-lima-sft-dpo-Q3_K_S.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q3_K_S.gguf) | Q3_K_S | 5.659 GB | very small, high quality loss |
| [llama2-13b-lima-sft-dpo-Q3_K_M.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q3_K_M.gguf) | Q3_K_M | 6.338 GB | very small, high quality loss |
| [llama2-13b-lima-sft-dpo-Q3_K_L.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q3_K_L.gguf) | Q3_K_L | 6.930 GB | small, substantial quality loss |
| [llama2-13b-lima-sft-dpo-Q4_0.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q4_0.gguf) | Q4_0 | 7.366 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [llama2-13b-lima-sft-dpo-Q4_K_S.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q4_K_S.gguf) | Q4_K_S | 7.423 GB | small, greater quality loss |
| [llama2-13b-lima-sft-dpo-Q4_K_M.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q4_K_M.gguf) | Q4_K_M | 7.866 GB | medium, balanced quality - recommended |
| [llama2-13b-lima-sft-dpo-Q5_0.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q5_0.gguf) | Q5_0 | 8.972 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [llama2-13b-lima-sft-dpo-Q5_K_S.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q5_K_S.gguf) | Q5_K_S | 8.972 GB | large, low quality loss - recommended |
| [llama2-13b-lima-sft-dpo-Q5_K_M.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q5_K_M.gguf) | Q5_K_M | 9.230 GB | large, very low quality loss - recommended |
| [llama2-13b-lima-sft-dpo-Q6_K.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q6_K.gguf) | Q6_K | 10.679 GB | very large, extremely low quality loss |
| [llama2-13b-lima-sft-dpo-Q8_0.gguf](https://huggingface.co/tensorblock/llama2-13b-lima-sft-dpo-GGUF/blob/main/llama2-13b-lima-sft-dpo-Q8_0.gguf) | Q8_0 | 13.831 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/llama2-13b-lima-sft-dpo-GGUF --include "llama2-13b-lima-sft-dpo-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/llama2-13b-lima-sft-dpo-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/prometheus-bgb-8x7b-v2.0-GGUF | tensorblock | 2025-04-21T00:24:02Z | 97 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"base_model:prometheus-eval/prometheus-bgb-8x7b-v2.0",
"base_model:quantized:prometheus-eval/prometheus-bgb-8x7b-v2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T17:09:39Z | ---
library_name: transformers
tags:
- TensorBlock
- GGUF
base_model: prometheus-eval/prometheus-bgb-8x7b-v2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## prometheus-eval/prometheus-bgb-8x7b-v2.0 - GGUF
This repo contains GGUF format model files for [prometheus-eval/prometheus-bgb-8x7b-v2.0](https://huggingface.co/prometheus-eval/prometheus-bgb-8x7b-v2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>[INST] {prompt} [/INST]
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [prometheus-bgb-8x7b-v2.0-Q2_K.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q2_K.gguf) | Q2_K | 17.311 GB | smallest, significant quality loss - not recommended for most purposes |
| [prometheus-bgb-8x7b-v2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q3_K_S.gguf) | Q3_K_S | 20.433 GB | very small, high quality loss |
| [prometheus-bgb-8x7b-v2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q3_K_M.gguf) | Q3_K_M | 22.546 GB | very small, high quality loss |
| [prometheus-bgb-8x7b-v2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q3_K_L.gguf) | Q3_K_L | 24.170 GB | small, substantial quality loss |
| [prometheus-bgb-8x7b-v2.0-Q4_0.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q4_0.gguf) | Q4_0 | 26.444 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [prometheus-bgb-8x7b-v2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q4_K_S.gguf) | Q4_K_S | 26.746 GB | small, greater quality loss |
| [prometheus-bgb-8x7b-v2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q4_K_M.gguf) | Q4_K_M | 28.448 GB | medium, balanced quality - recommended |
| [prometheus-bgb-8x7b-v2.0-Q5_0.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q5_0.gguf) | Q5_0 | 32.231 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [prometheus-bgb-8x7b-v2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q5_K_S.gguf) | Q5_K_S | 32.231 GB | large, low quality loss - recommended |
| [prometheus-bgb-8x7b-v2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q5_K_M.gguf) | Q5_K_M | 33.230 GB | large, very low quality loss - recommended |
| [prometheus-bgb-8x7b-v2.0-Q6_K.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q6_K.gguf) | Q6_K | 38.381 GB | very large, extremely low quality loss |
| [prometheus-bgb-8x7b-v2.0-Q8_0.gguf](https://huggingface.co/tensorblock/prometheus-bgb-8x7b-v2.0-GGUF/blob/main/prometheus-bgb-8x7b-v2.0-Q8_0.gguf) | Q8_0 | 49.626 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/prometheus-bgb-8x7b-v2.0-GGUF --include "prometheus-bgb-8x7b-v2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/prometheus-bgb-8x7b-v2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/mCoT-GGUF | tensorblock | 2025-04-21T00:23:59Z | 30 | 0 | null | [
"gguf",
"text-generation",
"TensorBlock",
"GGUF",
"sw",
"bn",
"te",
"th",
"ja",
"zh",
"ru",
"es",
"fr",
"de",
"en",
"base_model:laihuiyuan/mCoT",
"base_model:quantized:laihuiyuan/mCoT",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-13T14:49:50Z | ---
license: apache-2.0
language:
- sw
- bn
- te
- th
- ja
- zh
- ru
- es
- fr
- de
- en
tags:
- text-generation
- TensorBlock
- GGUF
base_model: laihuiyuan/mCoT
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## laihuiyuan/mCoT - GGUF
This repo contains GGUF format model files for [laihuiyuan/mCoT](https://huggingface.co/laihuiyuan/mCoT).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [mCoT-Q2_K.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [mCoT-Q3_K_S.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [mCoT-Q3_K_M.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [mCoT-Q3_K_L.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [mCoT-Q4_0.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [mCoT-Q4_K_S.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [mCoT-Q4_K_M.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [mCoT-Q5_0.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [mCoT-Q5_K_S.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [mCoT-Q5_K_M.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [mCoT-Q6_K.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [mCoT-Q8_0.gguf](https://huggingface.co/tensorblock/mCoT-GGUF/blob/main/mCoT-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/mCoT-GGUF --include "mCoT-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/mCoT-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/GritLM-8x7B-KTO-GGUF | tensorblock | 2025-04-21T00:23:54Z | 119 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"dataset:GritLM/tulu2",
"base_model:GritLM/GritLM-8x7B-KTO",
"base_model:quantized:GritLM/GritLM-8x7B-KTO",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T14:12:47Z | ---
pipeline_tag: text-generation
inference: true
license: apache-2.0
datasets:
- GritLM/tulu2
tags:
- TensorBlock
- GGUF
base_model: GritLM/GritLM-8x7B-KTO
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## GritLM/GritLM-8x7B-KTO - GGUF
This repo contains GGUF format model files for [GritLM/GritLM-8x7B-KTO](https://huggingface.co/GritLM/GritLM-8x7B-KTO).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s><|user|>
{prompt}
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [GritLM-8x7B-KTO-Q2_K.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q2_K.gguf) | Q2_K | 17.311 GB | smallest, significant quality loss - not recommended for most purposes |
| [GritLM-8x7B-KTO-Q3_K_S.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q3_K_S.gguf) | Q3_K_S | 20.433 GB | very small, high quality loss |
| [GritLM-8x7B-KTO-Q3_K_M.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q3_K_M.gguf) | Q3_K_M | 22.546 GB | very small, high quality loss |
| [GritLM-8x7B-KTO-Q3_K_L.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q3_K_L.gguf) | Q3_K_L | 24.170 GB | small, substantial quality loss |
| [GritLM-8x7B-KTO-Q4_0.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q4_0.gguf) | Q4_0 | 26.444 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [GritLM-8x7B-KTO-Q4_K_S.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q4_K_S.gguf) | Q4_K_S | 26.746 GB | small, greater quality loss |
| [GritLM-8x7B-KTO-Q4_K_M.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q4_K_M.gguf) | Q4_K_M | 28.448 GB | medium, balanced quality - recommended |
| [GritLM-8x7B-KTO-Q5_0.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q5_0.gguf) | Q5_0 | 32.231 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [GritLM-8x7B-KTO-Q5_K_S.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q5_K_S.gguf) | Q5_K_S | 32.231 GB | large, low quality loss - recommended |
| [GritLM-8x7B-KTO-Q5_K_M.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q5_K_M.gguf) | Q5_K_M | 33.230 GB | large, very low quality loss - recommended |
| [GritLM-8x7B-KTO-Q6_K.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q6_K.gguf) | Q6_K | 38.381 GB | very large, extremely low quality loss |
| [GritLM-8x7B-KTO-Q8_0.gguf](https://huggingface.co/tensorblock/GritLM-8x7B-KTO-GGUF/blob/main/GritLM-8x7B-KTO-Q8_0.gguf) | Q8_0 | 49.626 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/GritLM-8x7B-KTO-GGUF --include "GritLM-8x7B-KTO-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/GritLM-8x7B-KTO-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Leia-Swallow-13b-GGUF | tensorblock | 2025-04-21T00:23:53Z | 24 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"ja",
"base_model:leia-llm/Leia-Swallow-13b",
"base_model:quantized:leia-llm/Leia-Swallow-13b",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T13:34:25Z | ---
license: apache-2.0
language:
- ja
tags:
- TensorBlock
- GGUF
base_model: leia-llm/Leia-Swallow-13b
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## leia-llm/Leia-Swallow-13b - GGUF
This repo contains GGUF format model files for [leia-llm/Leia-Swallow-13b](https://huggingface.co/leia-llm/Leia-Swallow-13b).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Leia-Swallow-13b-Q2_K.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q2_K.gguf) | Q2_K | 4.920 GB | smallest, significant quality loss - not recommended for most purposes |
| [Leia-Swallow-13b-Q3_K_S.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q3_K_S.gguf) | Q3_K_S | 5.731 GB | very small, high quality loss |
| [Leia-Swallow-13b-Q3_K_M.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q3_K_M.gguf) | Q3_K_M | 6.410 GB | very small, high quality loss |
| [Leia-Swallow-13b-Q3_K_L.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q3_K_L.gguf) | Q3_K_L | 7.001 GB | small, substantial quality loss |
| [Leia-Swallow-13b-Q4_0.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q4_0.gguf) | Q4_0 | 7.445 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Leia-Swallow-13b-Q4_K_S.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q4_K_S.gguf) | Q4_K_S | 7.503 GB | small, greater quality loss |
| [Leia-Swallow-13b-Q4_K_M.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q4_K_M.gguf) | Q4_K_M | 7.945 GB | medium, balanced quality - recommended |
| [Leia-Swallow-13b-Q5_0.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q5_0.gguf) | Q5_0 | 9.059 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Leia-Swallow-13b-Q5_K_S.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q5_K_S.gguf) | Q5_K_S | 9.059 GB | large, low quality loss - recommended |
| [Leia-Swallow-13b-Q5_K_M.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q5_K_M.gguf) | Q5_K_M | 9.316 GB | large, very low quality loss - recommended |
| [Leia-Swallow-13b-Q6_K.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q6_K.gguf) | Q6_K | 10.773 GB | very large, extremely low quality loss |
| [Leia-Swallow-13b-Q8_0.gguf](https://huggingface.co/tensorblock/Leia-Swallow-13b-GGUF/blob/main/Leia-Swallow-13b-Q8_0.gguf) | Q8_0 | 13.953 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Leia-Swallow-13b-GGUF --include "Leia-Swallow-13b-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Leia-Swallow-13b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF | tensorblock | 2025-04-21T00:23:48Z | 1 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T13:26:42Z | ---
library_name: transformers
tags:
- TensorBlock
- GGUF
base_model: Goekdeniz-Guelmez/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Goekdeniz-Guelmez/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta - GGUF
This repo contains GGUF format model files for [Goekdeniz-Guelmez/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta](https://huggingface.co/Goekdeniz-Guelmez/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q2_K.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q2_K.gguf) | Q2_K | 0.105 GB | smallest, significant quality loss - not recommended for most purposes |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q3_K_S.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q3_K_S.gguf) | Q3_K_S | 0.121 GB | very small, high quality loss |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q3_K_M.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q3_K_M.gguf) | Q3_K_M | 0.129 GB | very small, high quality loss |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q3_K_L.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q3_K_L.gguf) | Q3_K_L | 0.136 GB | small, substantial quality loss |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q4_0.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q4_0.gguf) | Q4_0 | 0.149 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q4_K_S.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q4_K_S.gguf) | Q4_K_S | 0.149 GB | small, greater quality loss |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q4_K_M.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q4_K_M.gguf) | Q4_K_M | 0.157 GB | medium, balanced quality - recommended |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q5_0.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q5_0.gguf) | Q5_0 | 0.175 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q5_K_S.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q5_K_S.gguf) | Q5_K_S | 0.175 GB | large, low quality loss - recommended |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q5_K_M.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q5_K_M.gguf) | Q5_K_M | 0.179 GB | large, very low quality loss - recommended |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q6_K.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q6_K.gguf) | Q6_K | 0.204 GB | very large, extremely low quality loss |
| [TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q8_0.gguf](https://huggingface.co/tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF/blob/main/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q8_0.gguf) | Q8_0 | 0.263 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF --include "TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/TinyJ.O.S.I.E.-8x220M-Chat-Checkpoint-30000-steps-Beta-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/code_gpt2-GGUF | tensorblock | 2025-04-21T00:23:44Z | 86 | 0 | null | [
"gguf",
"gpt2",
"dpo",
"code",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:HuggingFaceH4/ultrachat_200k",
"dataset:mlabonne/CodeLlama-2-20k",
"dataset:Intel/orca_dpo_pairs",
"dataset:Sharathhebbar24/Evol-Instruct-Code-80k-v1",
"dataset:Sharathhebbar24/sql-create-context",
"base_model:Sharathhebbar24/code_gpt2",
"base_model:quantized:Sharathhebbar24/code_gpt2",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-13T13:20:52Z | ---
language:
- en
license: apache-2.0
tags:
- gpt2
- dpo
- code
- TensorBlock
- GGUF
datasets:
- HuggingFaceH4/ultrachat_200k
- mlabonne/CodeLlama-2-20k
- Intel/orca_dpo_pairs
- Sharathhebbar24/Evol-Instruct-Code-80k-v1
- Sharathhebbar24/sql-create-context
pipeline_tag: text-generation
base_model: Sharathhebbar24/code_gpt2
model-index:
- name: code_gpt2
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 23.29
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Sharathhebbar24/code_gpt2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 30.99
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Sharathhebbar24/code_gpt2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 25.03
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Sharathhebbar24/code_gpt2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 40.6
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Sharathhebbar24/code_gpt2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 49.25
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Sharathhebbar24/code_gpt2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 0.0
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Sharathhebbar24/code_gpt2
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Sharathhebbar24/code_gpt2 - GGUF
This repo contains GGUF format model files for [Sharathhebbar24/code_gpt2](https://huggingface.co/Sharathhebbar24/code_gpt2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [code_gpt2-Q2_K.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q2_K.gguf) | Q2_K | 0.081 GB | smallest, significant quality loss - not recommended for most purposes |
| [code_gpt2-Q3_K_S.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q3_K_S.gguf) | Q3_K_S | 0.090 GB | very small, high quality loss |
| [code_gpt2-Q3_K_M.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q3_K_M.gguf) | Q3_K_M | 0.098 GB | very small, high quality loss |
| [code_gpt2-Q3_K_L.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q3_K_L.gguf) | Q3_K_L | 0.102 GB | small, substantial quality loss |
| [code_gpt2-Q4_0.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q4_0.gguf) | Q4_0 | 0.107 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [code_gpt2-Q4_K_S.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q4_K_S.gguf) | Q4_K_S | 0.107 GB | small, greater quality loss |
| [code_gpt2-Q4_K_M.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q4_K_M.gguf) | Q4_K_M | 0.113 GB | medium, balanced quality - recommended |
| [code_gpt2-Q5_0.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q5_0.gguf) | Q5_0 | 0.122 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [code_gpt2-Q5_K_S.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q5_K_S.gguf) | Q5_K_S | 0.122 GB | large, low quality loss - recommended |
| [code_gpt2-Q5_K_M.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q5_K_M.gguf) | Q5_K_M | 0.127 GB | large, very low quality loss - recommended |
| [code_gpt2-Q6_K.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q6_K.gguf) | Q6_K | 0.138 GB | very large, extremely low quality loss |
| [code_gpt2-Q8_0.gguf](https://huggingface.co/tensorblock/code_gpt2-GGUF/blob/main/code_gpt2-Q8_0.gguf) | Q8_0 | 0.178 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/code_gpt2-GGUF --include "code_gpt2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/code_gpt2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF | tensorblock | 2025-04-21T00:23:37Z | 37 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"base_model:chihoonlee10/T3Q-ko-solar-dpo-v5.0",
"base_model:quantized:chihoonlee10/T3Q-ko-solar-dpo-v5.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-generation | 2024-12-13T10:54:27Z | ---
library_name: transformers
license: apache-2.0
pipeline_tag: text-generation
tags:
- TensorBlock
- GGUF
base_model: chihoonlee10/T3Q-ko-solar-dpo-v5.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## chihoonlee10/T3Q-ko-solar-dpo-v5.0 - GGUF
This repo contains GGUF format model files for [chihoonlee10/T3Q-ko-solar-dpo-v5.0](https://huggingface.co/chihoonlee10/T3Q-ko-solar-dpo-v5.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [T3Q-ko-solar-dpo-v5.0-Q2_K.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [T3Q-ko-solar-dpo-v5.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [T3Q-ko-solar-dpo-v5.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [T3Q-ko-solar-dpo-v5.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [T3Q-ko-solar-dpo-v5.0-Q4_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [T3Q-ko-solar-dpo-v5.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [T3Q-ko-solar-dpo-v5.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [T3Q-ko-solar-dpo-v5.0-Q5_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [T3Q-ko-solar-dpo-v5.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [T3Q-ko-solar-dpo-v5.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [T3Q-ko-solar-dpo-v5.0-Q6_K.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [T3Q-ko-solar-dpo-v5.0-Q8_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF/blob/main/T3Q-ko-solar-dpo-v5.0-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF --include "T3Q-ko-solar-dpo-v5.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/T3Q-ko-solar-dpo-v5.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Algae-550M-GGUF | tensorblock | 2025-04-21T00:23:35Z | 13 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"en",
"base_model:PhelixZhen/Algae-550M",
"base_model:quantized:PhelixZhen/Algae-550M",
"license:mit",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T10:46:56Z | ---
license: mit
language: en
base_model: PhelixZhen/Algae-550M
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## PhelixZhen/Algae-550M - GGUF
This repo contains GGUF format model files for [PhelixZhen/Algae-550M](https://huggingface.co/PhelixZhen/Algae-550M).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Algae-550M-Q2_K.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q2_K.gguf) | Q2_K | 0.239 GB | smallest, significant quality loss - not recommended for most purposes |
| [Algae-550M-Q3_K_S.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q3_K_S.gguf) | Q3_K_S | 0.267 GB | very small, high quality loss |
| [Algae-550M-Q3_K_M.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q3_K_M.gguf) | Q3_K_M | 0.296 GB | very small, high quality loss |
| [Algae-550M-Q3_K_L.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q3_K_L.gguf) | Q3_K_L | 0.322 GB | small, substantial quality loss |
| [Algae-550M-Q4_0.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q4_0.gguf) | Q4_0 | 0.332 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Algae-550M-Q4_K_S.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q4_K_S.gguf) | Q4_K_S | 0.335 GB | small, greater quality loss |
| [Algae-550M-Q4_K_M.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q4_K_M.gguf) | Q4_K_M | 0.355 GB | medium, balanced quality - recommended |
| [Algae-550M-Q5_0.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q5_0.gguf) | Q5_0 | 0.393 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Algae-550M-Q5_K_S.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q5_K_S.gguf) | Q5_K_S | 0.393 GB | large, low quality loss - recommended |
| [Algae-550M-Q5_K_M.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q5_K_M.gguf) | Q5_K_M | 0.405 GB | large, very low quality loss - recommended |
| [Algae-550M-Q6_K.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q6_K.gguf) | Q6_K | 0.458 GB | very large, extremely low quality loss |
| [Algae-550M-Q8_0.gguf](https://huggingface.co/tensorblock/Algae-550M-GGUF/blob/main/Algae-550M-Q8_0.gguf) | Q8_0 | 0.593 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Algae-550M-GGUF --include "Algae-550M-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Algae-550M-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF | tensorblock | 2025-04-21T00:23:28Z | 90 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"ja",
"base_model:tokyotech-llm/Swallow-MS-7b-instruct-v0.1",
"base_model:quantized:tokyotech-llm/Swallow-MS-7b-instruct-v0.1",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T09:03:08Z | ---
language:
- en
- ja
library_name: transformers
pipeline_tag: text-generation
model_type: mistral
license: apache-2.0
base_model: tokyotech-llm/Swallow-MS-7b-instruct-v0.1
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## tokyotech-llm/Swallow-MS-7b-instruct-v0.1 - GGUF
This repo contains GGUF format model files for [tokyotech-llm/Swallow-MS-7b-instruct-v0.1](https://huggingface.co/tokyotech-llm/Swallow-MS-7b-instruct-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>[INST] <<SYS>>
{system_prompt}
<</SYS>>
{prompt} [/INST]
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Swallow-MS-7b-instruct-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q2_K.gguf) | Q2_K | 2.770 GB | smallest, significant quality loss - not recommended for most purposes |
| [Swallow-MS-7b-instruct-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q3_K_S.gguf) | Q3_K_S | 3.220 GB | very small, high quality loss |
| [Swallow-MS-7b-instruct-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q3_K_M.gguf) | Q3_K_M | 3.575 GB | very small, high quality loss |
| [Swallow-MS-7b-instruct-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q3_K_L.gguf) | Q3_K_L | 3.878 GB | small, substantial quality loss |
| [Swallow-MS-7b-instruct-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q4_0.gguf) | Q4_0 | 4.170 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Swallow-MS-7b-instruct-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q4_K_S.gguf) | Q4_K_S | 4.202 GB | small, greater quality loss |
| [Swallow-MS-7b-instruct-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q4_K_M.gguf) | Q4_K_M | 4.430 GB | medium, balanced quality - recommended |
| [Swallow-MS-7b-instruct-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q5_0.gguf) | Q5_0 | 5.065 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Swallow-MS-7b-instruct-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q5_K_S.gguf) | Q5_K_S | 5.065 GB | large, low quality loss - recommended |
| [Swallow-MS-7b-instruct-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q5_K_M.gguf) | Q5_K_M | 5.198 GB | large, very low quality loss - recommended |
| [Swallow-MS-7b-instruct-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q6_K.gguf) | Q6_K | 6.015 GB | very large, extremely low quality loss |
| [Swallow-MS-7b-instruct-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF/blob/main/Swallow-MS-7b-instruct-v0.1-Q8_0.gguf) | Q8_0 | 7.790 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF --include "Swallow-MS-7b-instruct-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Swallow-MS-7b-instruct-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Myrrh_solar_10.7b_2.0-GGUF | tensorblock | 2025-04-21T00:23:26Z | 23 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"ko",
"base_model:MoaData/Myrrh_solar_10.7b_2.0",
"base_model:quantized:MoaData/Myrrh_solar_10.7b_2.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T07:15:11Z | ---
license: apache-2.0
language:
- ko
tags:
- TensorBlock
- GGUF
base_model: MoaData/Myrrh_solar_10.7b_2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## MoaData/Myrrh_solar_10.7b_2.0 - GGUF
This repo contains GGUF format model files for [MoaData/Myrrh_solar_10.7b_2.0](https://huggingface.co/MoaData/Myrrh_solar_10.7b_2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Myrrh_solar_10.7b_2.0-Q2_K.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [Myrrh_solar_10.7b_2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [Myrrh_solar_10.7b_2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [Myrrh_solar_10.7b_2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [Myrrh_solar_10.7b_2.0-Q4_0.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Myrrh_solar_10.7b_2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [Myrrh_solar_10.7b_2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [Myrrh_solar_10.7b_2.0-Q5_0.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Myrrh_solar_10.7b_2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [Myrrh_solar_10.7b_2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [Myrrh_solar_10.7b_2.0-Q6_K.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [Myrrh_solar_10.7b_2.0-Q8_0.gguf](https://huggingface.co/tensorblock/Myrrh_solar_10.7b_2.0-GGUF/blob/main/Myrrh_solar_10.7b_2.0-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Myrrh_solar_10.7b_2.0-GGUF --include "Myrrh_solar_10.7b_2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Myrrh_solar_10.7b_2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Zamfir-7B-slerp-GGUF | tensorblock | 2025-04-21T00:23:25Z | 26 | 0 | null | [
"gguf",
"merge",
"mergekit",
"lazymergekit",
"gordicaleksa/YugoGPT",
"HuggingFaceH4/zephyr-7b-beta",
"TensorBlock",
"GGUF",
"base_model:Stopwolf/Zamfir-7B-slerp",
"base_model:quantized:Stopwolf/Zamfir-7B-slerp",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T06:53:26Z | ---
license: apache-2.0
tags:
- merge
- mergekit
- lazymergekit
- gordicaleksa/YugoGPT
- HuggingFaceH4/zephyr-7b-beta
- TensorBlock
- GGUF
base_model: Stopwolf/Zamfir-7B-slerp
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Stopwolf/Zamfir-7B-slerp - GGUF
This repo contains GGUF format model files for [Stopwolf/Zamfir-7B-slerp](https://huggingface.co/Stopwolf/Zamfir-7B-slerp).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Zamfir-7B-slerp-Q2_K.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Zamfir-7B-slerp-Q3_K_S.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Zamfir-7B-slerp-Q3_K_M.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Zamfir-7B-slerp-Q3_K_L.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Zamfir-7B-slerp-Q4_0.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Zamfir-7B-slerp-Q4_K_S.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Zamfir-7B-slerp-Q4_K_M.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Zamfir-7B-slerp-Q5_0.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Zamfir-7B-slerp-Q5_K_S.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Zamfir-7B-slerp-Q5_K_M.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Zamfir-7B-slerp-Q6_K.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Zamfir-7B-slerp-Q8_0.gguf](https://huggingface.co/tensorblock/Zamfir-7B-slerp-GGUF/blob/main/Zamfir-7B-slerp-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Zamfir-7B-slerp-GGUF --include "Zamfir-7B-slerp-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Zamfir-7B-slerp-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Confluence-Renegade-7B-GGUF | tensorblock | 2025-04-21T00:23:23Z | 26 | 0 | transformers | [
"transformers",
"gguf",
"mergekit",
"merge",
"TensorBlock",
"GGUF",
"base_model:Nekochu/Confluence-Renegade-7B",
"base_model:quantized:Nekochu/Confluence-Renegade-7B",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-13T06:15:23Z | ---
license: apache-2.0
base_model: Nekochu/Confluence-Renegade-7B
library_name: transformers
tags:
- mergekit
- merge
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## Nekochu/Confluence-Renegade-7B - GGUF
This repo contains GGUF format model files for [Nekochu/Confluence-Renegade-7B](https://huggingface.co/Nekochu/Confluence-Renegade-7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Confluence-Renegade-7B-Q2_K.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [Confluence-Renegade-7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [Confluence-Renegade-7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [Confluence-Renegade-7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [Confluence-Renegade-7B-Q4_0.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Confluence-Renegade-7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [Confluence-Renegade-7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [Confluence-Renegade-7B-Q5_0.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Confluence-Renegade-7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [Confluence-Renegade-7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [Confluence-Renegade-7B-Q6_K.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [Confluence-Renegade-7B-Q8_0.gguf](https://huggingface.co/tensorblock/Confluence-Renegade-7B-GGUF/blob/main/Confluence-Renegade-7B-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Confluence-Renegade-7B-GGUF --include "Confluence-Renegade-7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Confluence-Renegade-7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Meltemi-7B-Instruct-v1-GGUF | tensorblock | 2025-04-21T00:23:22Z | 27 | 0 | null | [
"gguf",
"finetuned",
"TensorBlock",
"GGUF",
"text-generation",
"el",
"en",
"base_model:ilsp/Meltemi-7B-Instruct-v1",
"base_model:quantized:ilsp/Meltemi-7B-Instruct-v1",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T05:32:09Z | ---
license: apache-2.0
language:
- el
- en
tags:
- finetuned
- TensorBlock
- GGUF
inference: true
pipeline_tag: text-generation
base_model: ilsp/Meltemi-7B-Instruct-v1
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## ilsp/Meltemi-7B-Instruct-v1 - GGUF
This repo contains GGUF format model files for [ilsp/Meltemi-7B-Instruct-v1](https://huggingface.co/ilsp/Meltemi-7B-Instruct-v1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}</s>
<|user|>
{prompt}</s>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Meltemi-7B-Instruct-v1-Q2_K.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q2_K.gguf) | Q2_K | 2.859 GB | smallest, significant quality loss - not recommended for most purposes |
| [Meltemi-7B-Instruct-v1-Q3_K_S.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q3_K_S.gguf) | Q3_K_S | 3.317 GB | very small, high quality loss |
| [Meltemi-7B-Instruct-v1-Q3_K_M.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q3_K_M.gguf) | Q3_K_M | 3.671 GB | very small, high quality loss |
| [Meltemi-7B-Instruct-v1-Q3_K_L.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q3_K_L.gguf) | Q3_K_L | 3.974 GB | small, substantial quality loss |
| [Meltemi-7B-Instruct-v1-Q4_0.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q4_0.gguf) | Q4_0 | 4.277 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Meltemi-7B-Instruct-v1-Q4_K_S.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q4_K_S.gguf) | Q4_K_S | 4.308 GB | small, greater quality loss |
| [Meltemi-7B-Instruct-v1-Q4_K_M.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q4_K_M.gguf) | Q4_K_M | 4.537 GB | medium, balanced quality - recommended |
| [Meltemi-7B-Instruct-v1-Q5_0.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q5_0.gguf) | Q5_0 | 5.181 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Meltemi-7B-Instruct-v1-Q5_K_S.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q5_K_S.gguf) | Q5_K_S | 5.181 GB | large, low quality loss - recommended |
| [Meltemi-7B-Instruct-v1-Q5_K_M.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q5_K_M.gguf) | Q5_K_M | 5.315 GB | large, very low quality loss - recommended |
| [Meltemi-7B-Instruct-v1-Q6_K.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q6_K.gguf) | Q6_K | 6.141 GB | very large, extremely low quality loss |
| [Meltemi-7B-Instruct-v1-Q8_0.gguf](https://huggingface.co/tensorblock/Meltemi-7B-Instruct-v1-GGUF/blob/main/Meltemi-7B-Instruct-v1-Q8_0.gguf) | Q8_0 | 7.954 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Meltemi-7B-Instruct-v1-GGUF --include "Meltemi-7B-Instruct-v1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Meltemi-7B-Instruct-v1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Qwen1.5-0.4B-Chat-GGUF | tensorblock | 2025-04-21T00:23:19Z | 28 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"base_model:nathanrchn/Qwen1.5-0.4B-Chat",
"base_model:quantized:nathanrchn/Qwen1.5-0.4B-Chat",
"license:other",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T05:17:37Z | ---
license: other
license_name: tongyi-qianwen-research
license_link: https://huggingface.co/Qwen/Qwen1.5-0.5B-Chat/blob/main/LICENSE
library_name: transformers
tags:
- TensorBlock
- GGUF
base_model: nathanrchn/Qwen1.5-0.4B-Chat
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## nathanrchn/Qwen1.5-0.4B-Chat - GGUF
This repo contains GGUF format model files for [nathanrchn/Qwen1.5-0.4B-Chat](https://huggingface.co/nathanrchn/Qwen1.5-0.4B-Chat).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Qwen1.5-0.4B-Chat-Q2_K.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q2_K.gguf) | Q2_K | 0.228 GB | smallest, significant quality loss - not recommended for most purposes |
| [Qwen1.5-0.4B-Chat-Q3_K_S.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q3_K_S.gguf) | Q3_K_S | 0.244 GB | very small, high quality loss |
| [Qwen1.5-0.4B-Chat-Q3_K_M.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q3_K_M.gguf) | Q3_K_M | 0.258 GB | very small, high quality loss |
| [Qwen1.5-0.4B-Chat-Q3_K_L.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q3_K_L.gguf) | Q3_K_L | 0.270 GB | small, substantial quality loss |
| [Qwen1.5-0.4B-Chat-Q4_0.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q4_0.gguf) | Q4_0 | 0.278 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Qwen1.5-0.4B-Chat-Q4_K_S.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q4_K_S.gguf) | Q4_K_S | 0.280 GB | small, greater quality loss |
| [Qwen1.5-0.4B-Chat-Q4_K_M.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q4_K_M.gguf) | Q4_K_M | 0.289 GB | medium, balanced quality - recommended |
| [Qwen1.5-0.4B-Chat-Q5_0.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q5_0.gguf) | Q5_0 | 0.311 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Qwen1.5-0.4B-Chat-Q5_K_S.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q5_K_S.gguf) | Q5_K_S | 0.311 GB | large, low quality loss - recommended |
| [Qwen1.5-0.4B-Chat-Q5_K_M.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q5_K_M.gguf) | Q5_K_M | 0.316 GB | large, very low quality loss - recommended |
| [Qwen1.5-0.4B-Chat-Q6_K.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q6_K.gguf) | Q6_K | 0.345 GB | very large, extremely low quality loss |
| [Qwen1.5-0.4B-Chat-Q8_0.gguf](https://huggingface.co/tensorblock/Qwen1.5-0.4B-Chat-GGUF/blob/main/Qwen1.5-0.4B-Chat-Q8_0.gguf) | Q8_0 | 0.445 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Qwen1.5-0.4B-Chat-GGUF --include "Qwen1.5-0.4B-Chat-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Qwen1.5-0.4B-Chat-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/nous-6-GGUF | tensorblock | 2025-04-21T00:22:59Z | 12 | 0 | transformers | [
"transformers",
"gguf",
"TensorBlock",
"GGUF",
"base_model:kalytm/nous-6",
"base_model:quantized:kalytm/nous-6",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T03:51:01Z | ---
library_name: transformers
tags:
- TensorBlock
- GGUF
base_model: kalytm/nous-6
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## kalytm/nous-6 - GGUF
This repo contains GGUF format model files for [kalytm/nous-6](https://huggingface.co/kalytm/nous-6).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|system|>
{system_prompt}<|endoftext|>
<|user|>
{prompt}<|endoftext|>
<|assistant|>
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [nous-6-Q2_K.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q2_K.gguf) | Q2_K | 0.694 GB | smallest, significant quality loss - not recommended for most purposes |
| [nous-6-Q3_K_S.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q3_K_S.gguf) | Q3_K_S | 0.792 GB | very small, high quality loss |
| [nous-6-Q3_K_M.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q3_K_M.gguf) | Q3_K_M | 0.858 GB | very small, high quality loss |
| [nous-6-Q3_K_L.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q3_K_L.gguf) | Q3_K_L | 0.915 GB | small, substantial quality loss |
| [nous-6-Q4_0.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q4_0.gguf) | Q4_0 | 0.983 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [nous-6-Q4_K_S.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q4_K_S.gguf) | Q4_K_S | 0.989 GB | small, greater quality loss |
| [nous-6-Q4_K_M.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q4_K_M.gguf) | Q4_K_M | 1.031 GB | medium, balanced quality - recommended |
| [nous-6-Q5_0.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q5_0.gguf) | Q5_0 | 1.163 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [nous-6-Q5_K_S.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q5_K_S.gguf) | Q5_K_S | 1.163 GB | large, low quality loss - recommended |
| [nous-6-Q5_K_M.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q5_K_M.gguf) | Q5_K_M | 1.188 GB | large, very low quality loss - recommended |
| [nous-6-Q6_K.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q6_K.gguf) | Q6_K | 1.354 GB | very large, extremely low quality loss |
| [nous-6-Q8_0.gguf](https://huggingface.co/tensorblock/nous-6-GGUF/blob/main/nous-6-Q8_0.gguf) | Q8_0 | 1.752 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/nous-6-GGUF --include "nous-6-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/nous-6-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/google-gemma-7b-it-dequantized-GGUF | tensorblock | 2025-04-21T00:22:49Z | 32 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"base_model:justinxzhao/google-gemma-7b-it-dequantized",
"base_model:quantized:justinxzhao/google-gemma-7b-it-dequantized",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T02:54:00Z | ---
base_model: justinxzhao/google-gemma-7b-it-dequantized
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## justinxzhao/google-gemma-7b-it-dequantized - GGUF
This repo contains GGUF format model files for [justinxzhao/google-gemma-7b-it-dequantized](https://huggingface.co/justinxzhao/google-gemma-7b-it-dequantized).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<bos><start_of_turn>user
{prompt}<end_of_turn>
<start_of_turn>model
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [google-gemma-7b-it-dequantized-Q2_K.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q2_K.gguf) | Q2_K | 3.481 GB | smallest, significant quality loss - not recommended for most purposes |
| [google-gemma-7b-it-dequantized-Q3_K_S.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q3_K_S.gguf) | Q3_K_S | 3.982 GB | very small, high quality loss |
| [google-gemma-7b-it-dequantized-Q3_K_M.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q3_K_M.gguf) | Q3_K_M | 4.369 GB | very small, high quality loss |
| [google-gemma-7b-it-dequantized-Q3_K_L.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q3_K_L.gguf) | Q3_K_L | 4.709 GB | small, substantial quality loss |
| [google-gemma-7b-it-dequantized-Q4_0.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q4_0.gguf) | Q4_0 | 5.012 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [google-gemma-7b-it-dequantized-Q4_K_S.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q4_K_S.gguf) | Q4_K_S | 5.046 GB | small, greater quality loss |
| [google-gemma-7b-it-dequantized-Q4_K_M.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q4_K_M.gguf) | Q4_K_M | 5.330 GB | medium, balanced quality - recommended |
| [google-gemma-7b-it-dequantized-Q5_0.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q5_0.gguf) | Q5_0 | 5.981 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [google-gemma-7b-it-dequantized-Q5_K_S.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q5_K_S.gguf) | Q5_K_S | 5.981 GB | large, low quality loss - recommended |
| [google-gemma-7b-it-dequantized-Q5_K_M.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q5_K_M.gguf) | Q5_K_M | 6.145 GB | large, very low quality loss - recommended |
| [google-gemma-7b-it-dequantized-Q6_K.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q6_K.gguf) | Q6_K | 7.010 GB | very large, extremely low quality loss |
| [google-gemma-7b-it-dequantized-Q8_0.gguf](https://huggingface.co/tensorblock/google-gemma-7b-it-dequantized-GGUF/blob/main/google-gemma-7b-it-dequantized-Q8_0.gguf) | Q8_0 | 9.078 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/google-gemma-7b-it-dequantized-GGUF --include "google-gemma-7b-it-dequantized-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/google-gemma-7b-it-dequantized-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF | tensorblock | 2025-04-21T00:22:47Z | 26 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"ko",
"en",
"base_model:KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0",
"base_model:quantized:KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-13T02:39:01Z | ---
license: cc-by-nc-4.0
language:
- ko
- en
tags:
- TensorBlock
- GGUF
base_model: KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0 - GGUF
This repo contains GGUF format model files for [KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0](https://huggingface.co/KBNIT/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q2_K.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q2_K.gguf) | Q2_K | 4.046 GB | smallest, significant quality loss - not recommended for most purposes |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q3_K_S.gguf) | Q3_K_S | 4.711 GB | very small, high quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q3_K_M.gguf) | Q3_K_M | 5.242 GB | very small, high quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q3_K_L.gguf) | Q3_K_L | 5.697 GB | small, substantial quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q4_0.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q4_0.gguf) | Q4_0 | 6.123 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q4_K_S.gguf) | Q4_K_S | 6.169 GB | small, greater quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q4_K_M.gguf) | Q4_K_M | 6.513 GB | medium, balanced quality - recommended |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q5_0.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q5_0.gguf) | Q5_0 | 7.453 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q5_K_S.gguf) | Q5_K_S | 7.453 GB | large, low quality loss - recommended |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q5_K_M.gguf) | Q5_K_M | 7.653 GB | large, very low quality loss - recommended |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q6_K.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q6_K.gguf) | Q6_K | 8.866 GB | very large, extremely low quality loss |
| [KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q8_0.gguf](https://huggingface.co/tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF/blob/main/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q8_0.gguf) | Q8_0 | 11.482 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF --include "KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/KoSOLAR-10.7B-QLoRA-NEFTune-kolon-v2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/gemma-ko-7b-it-v0.41-GGUF | tensorblock | 2025-04-21T00:22:45Z | 37 | 0 | transformers | [
"transformers",
"gguf",
"korean",
"gemma",
"pytorch",
"TensorBlock",
"GGUF",
"text-generation",
"ko",
"en",
"base_model:lemon-mint/gemma-ko-7b-it-v0.41",
"base_model:quantized:lemon-mint/gemma-ko-7b-it-v0.41",
"license:other",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T01:50:11Z | ---
library_name: transformers
license: other
license_name: gemma-terms-of-use
license_link: https://ai.google.dev/gemma/terms
language:
- ko
- en
tags:
- korean
- gemma
- pytorch
- TensorBlock
- GGUF
pipeline_tag: text-generation
base_model: lemon-mint/gemma-ko-7b-it-v0.41
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## lemon-mint/gemma-ko-7b-it-v0.41 - GGUF
This repo contains GGUF format model files for [lemon-mint/gemma-ko-7b-it-v0.41](https://huggingface.co/lemon-mint/gemma-ko-7b-it-v0.41).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<bos><start_of_turn>user
{prompt}<end_of_turn>
<start_of_turn>model
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [gemma-ko-7b-it-v0.41-Q2_K.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q2_K.gguf) | Q2_K | 3.481 GB | smallest, significant quality loss - not recommended for most purposes |
| [gemma-ko-7b-it-v0.41-Q3_K_S.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q3_K_S.gguf) | Q3_K_S | 3.982 GB | very small, high quality loss |
| [gemma-ko-7b-it-v0.41-Q3_K_M.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q3_K_M.gguf) | Q3_K_M | 4.369 GB | very small, high quality loss |
| [gemma-ko-7b-it-v0.41-Q3_K_L.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q3_K_L.gguf) | Q3_K_L | 4.709 GB | small, substantial quality loss |
| [gemma-ko-7b-it-v0.41-Q4_0.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q4_0.gguf) | Q4_0 | 5.012 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [gemma-ko-7b-it-v0.41-Q4_K_S.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q4_K_S.gguf) | Q4_K_S | 5.046 GB | small, greater quality loss |
| [gemma-ko-7b-it-v0.41-Q4_K_M.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q4_K_M.gguf) | Q4_K_M | 5.330 GB | medium, balanced quality - recommended |
| [gemma-ko-7b-it-v0.41-Q5_0.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q5_0.gguf) | Q5_0 | 5.981 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [gemma-ko-7b-it-v0.41-Q5_K_S.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q5_K_S.gguf) | Q5_K_S | 5.981 GB | large, low quality loss - recommended |
| [gemma-ko-7b-it-v0.41-Q5_K_M.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q5_K_M.gguf) | Q5_K_M | 6.145 GB | large, very low quality loss - recommended |
| [gemma-ko-7b-it-v0.41-Q6_K.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q6_K.gguf) | Q6_K | 7.010 GB | very large, extremely low quality loss |
| [gemma-ko-7b-it-v0.41-Q8_0.gguf](https://huggingface.co/tensorblock/gemma-ko-7b-it-v0.41-GGUF/blob/main/gemma-ko-7b-it-v0.41-Q8_0.gguf) | Q8_0 | 9.078 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/gemma-ko-7b-it-v0.41-GGUF --include "gemma-ko-7b-it-v0.41-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/gemma-ko-7b-it-v0.41-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/T3Q-ko-solar-sft-v2.0-GGUF | tensorblock | 2025-04-21T00:22:42Z | 28 | 0 | null | [
"gguf",
"T3Q-ko-solar-sft-v2.0",
"nlpai-lab/kullm-v2",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:nlpai-lab/kullm-v2",
"base_model:chlee10/T3Q-ko-solar-sft-v2.0",
"base_model:quantized:chlee10/T3Q-ko-solar-sft-v2.0",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-13T01:12:17Z | ---
pipeline_tag: text-generation
license: apache-2.0
language:
- en
tags:
- T3Q-ko-solar-sft-v2.0
- nlpai-lab/kullm-v2
- TensorBlock
- GGUF
base_model: chlee10/T3Q-ko-solar-sft-v2.0
datasets:
- nlpai-lab/kullm-v2
model-index:
- name: T3Q-ko-solar-sft-v2.0
results: []
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## chlee10/T3Q-ko-solar-sft-v2.0 - GGUF
This repo contains GGUF format model files for [chlee10/T3Q-ko-solar-sft-v2.0](https://huggingface.co/chlee10/T3Q-ko-solar-sft-v2.0).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### System:
{system_prompt}
### User:
{prompt}
### Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [T3Q-ko-solar-sft-v2.0-Q2_K.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [T3Q-ko-solar-sft-v2.0-Q3_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [T3Q-ko-solar-sft-v2.0-Q3_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [T3Q-ko-solar-sft-v2.0-Q3_K_L.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [T3Q-ko-solar-sft-v2.0-Q4_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [T3Q-ko-solar-sft-v2.0-Q4_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [T3Q-ko-solar-sft-v2.0-Q4_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [T3Q-ko-solar-sft-v2.0-Q5_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [T3Q-ko-solar-sft-v2.0-Q5_K_S.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [T3Q-ko-solar-sft-v2.0-Q5_K_M.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [T3Q-ko-solar-sft-v2.0-Q6_K.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [T3Q-ko-solar-sft-v2.0-Q8_0.gguf](https://huggingface.co/tensorblock/T3Q-ko-solar-sft-v2.0-GGUF/blob/main/T3Q-ko-solar-sft-v2.0-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/T3Q-ko-solar-sft-v2.0-GGUF --include "T3Q-ko-solar-sft-v2.0-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/T3Q-ko-solar-sft-v2.0-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/SambaLingo-Arabic-Base-70B-GGUF | tensorblock | 2025-04-21T00:22:35Z | 26 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"ar",
"en",
"dataset:uonlp/CulturaX",
"base_model:sambanovasystems/SambaLingo-Arabic-Base-70B",
"base_model:quantized:sambanovasystems/SambaLingo-Arabic-Base-70B",
"license:llama2",
"endpoints_compatible",
"region:us"
] | null | 2024-12-12T23:34:41Z | ---
license: llama2
datasets:
- uonlp/CulturaX
language:
- ar
- en
metrics:
- chrf
- accuracy
- bleu
tags:
- TensorBlock
- GGUF
base_model: sambanovasystems/SambaLingo-Arabic-Base-70B
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## sambanovasystems/SambaLingo-Arabic-Base-70B - GGUF
This repo contains GGUF format model files for [sambanovasystems/SambaLingo-Arabic-Base-70B](https://huggingface.co/sambanovasystems/SambaLingo-Arabic-Base-70B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [SambaLingo-Arabic-Base-70B-Q2_K.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q2_K.gguf) | Q2_K | 25.702 GB | smallest, significant quality loss - not recommended for most purposes |
| [SambaLingo-Arabic-Base-70B-Q3_K_S.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q3_K_S.gguf) | Q3_K_S | 30.180 GB | very small, high quality loss |
| [SambaLingo-Arabic-Base-70B-Q3_K_M.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q3_K_M.gguf) | Q3_K_M | 33.535 GB | very small, high quality loss |
| [SambaLingo-Arabic-Base-70B-Q3_K_L.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q3_K_L.gguf) | Q3_K_L | 36.408 GB | small, substantial quality loss |
| [SambaLingo-Arabic-Base-70B-Q4_0.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q4_0.gguf) | Q4_0 | 39.160 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [SambaLingo-Arabic-Base-70B-Q4_K_S.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q4_K_S.gguf) | Q4_K_S | 39.538 GB | small, greater quality loss |
| [SambaLingo-Arabic-Base-70B-Q4_K_M.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q4_K_M.gguf) | Q4_K_M | 41.711 GB | medium, balanced quality - recommended |
| [SambaLingo-Arabic-Base-70B-Q5_0.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q5_0.gguf) | Q5_0 | 47.775 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [SambaLingo-Arabic-Base-70B-Q5_K_S.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q5_K_S.gguf) | Q5_K_S | 47.775 GB | large, low quality loss - recommended |
| [SambaLingo-Arabic-Base-70B-Q5_K_M.gguf](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q5_K_M.gguf) | Q5_K_M | 49.068 GB | large, very low quality loss - recommended |
| [SambaLingo-Arabic-Base-70B-Q6_K](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q6_K) | Q6_K | 56.929 GB | very large, extremely low quality loss |
| [SambaLingo-Arabic-Base-70B-Q8_0](https://huggingface.co/tensorblock/SambaLingo-Arabic-Base-70B-GGUF/blob/main/SambaLingo-Arabic-Base-70B-Q8_0) | Q8_0 | 73.734 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/SambaLingo-Arabic-Base-70B-GGUF --include "SambaLingo-Arabic-Base-70B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/SambaLingo-Arabic-Base-70B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF | tensorblock | 2025-04-21T00:22:34Z | 36 | 0 | transformers | [
"transformers",
"gguf",
"mergekit",
"merge",
"TensorBlock",
"GGUF",
"base_model:saishf/Fimbulvetr-Kuro-Lotus-10.7B",
"base_model:quantized:saishf/Fimbulvetr-Kuro-Lotus-10.7B",
"license:cc-by-nc-4.0",
"model-index",
"endpoints_compatible",
"region:us"
] | null | 2024-12-12T22:49:07Z | ---
license: cc-by-nc-4.0
library_name: transformers
tags:
- mergekit
- merge
- TensorBlock
- GGUF
base_model: saishf/Fimbulvetr-Kuro-Lotus-10.7B
model-index:
- name: Fimbulvetr-Kuro-Lotus-10.7B
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 69.54
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 87.87
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 66.99
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 60.95
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 84.14
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 66.87
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saishf/Fimbulvetr-Kuro-Lotus-10.7B
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## saishf/Fimbulvetr-Kuro-Lotus-10.7B - GGUF
This repo contains GGUF format model files for [saishf/Fimbulvetr-Kuro-Lotus-10.7B](https://huggingface.co/saishf/Fimbulvetr-Kuro-Lotus-10.7B).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q2_K.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q3_K_S.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q3_K_M.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q3_K_L.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q4_0.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q4_K_S.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q4_K_M.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q5_0.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q5_K_S.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q5_K_M.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q6_K.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [Fimbulvetr-Kuro-Lotus-10.7B-Q8_0.gguf](https://huggingface.co/tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF/blob/main/Fimbulvetr-Kuro-Lotus-10.7B-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF --include "Fimbulvetr-Kuro-Lotus-10.7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Fimbulvetr-Kuro-Lotus-10.7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Reyna-Mini-1.8B-v0.2-GGUF | tensorblock | 2025-04-21T00:22:31Z | 16 | 0 | transformers | [
"transformers",
"gguf",
"chatml",
"finetune",
"gpt4",
"synthetic data",
"custom_code",
"qwen2",
"TensorBlock",
"GGUF",
"dataset:Locutusque/Hercules-v3.0",
"base_model:aloobun/Reyna-Mini-1.8B-v0.2",
"base_model:quantized:aloobun/Reyna-Mini-1.8B-v0.2",
"license:other",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | null | 2024-12-12T22:05:20Z | ---
license: other
library_name: transformers
tags:
- chatml
- finetune
- gpt4
- synthetic data
- custom_code
- qwen2
- TensorBlock
- GGUF
datasets:
- Locutusque/Hercules-v3.0
license_name: tongyi-qianwen-research
license_link: https://huggingface.co/Qwen/Qwen1.5-1.8B-Chat/raw/main/LICENSE
base_model: aloobun/Reyna-Mini-1.8B-v0.2
model-index:
- name: Reyna-Mini-1.8B-v0.2
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 36.6
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=aloobun/Reyna-Mini-1.8B-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 60.19
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=aloobun/Reyna-Mini-1.8B-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 44.75
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=aloobun/Reyna-Mini-1.8B-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 41.24
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=aloobun/Reyna-Mini-1.8B-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 61.56
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=aloobun/Reyna-Mini-1.8B-v0.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 31.31
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=aloobun/Reyna-Mini-1.8B-v0.2
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## aloobun/Reyna-Mini-1.8B-v0.2 - GGUF
This repo contains GGUF format model files for [aloobun/Reyna-Mini-1.8B-v0.2](https://huggingface.co/aloobun/Reyna-Mini-1.8B-v0.2).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Reyna-Mini-1.8B-v0.2-Q2_K.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q2_K.gguf) | Q2_K | 0.847 GB | smallest, significant quality loss - not recommended for most purposes |
| [Reyna-Mini-1.8B-v0.2-Q3_K_S.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q3_K_S.gguf) | Q3_K_S | 0.954 GB | very small, high quality loss |
| [Reyna-Mini-1.8B-v0.2-Q3_K_M.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q3_K_M.gguf) | Q3_K_M | 1.016 GB | very small, high quality loss |
| [Reyna-Mini-1.8B-v0.2-Q3_K_L.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q3_K_L.gguf) | Q3_K_L | 1.056 GB | small, substantial quality loss |
| [Reyna-Mini-1.8B-v0.2-Q4_0.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q4_0.gguf) | Q4_0 | 1.120 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Reyna-Mini-1.8B-v0.2-Q4_K_S.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q4_K_S.gguf) | Q4_K_S | 1.158 GB | small, greater quality loss |
| [Reyna-Mini-1.8B-v0.2-Q4_K_M.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q4_K_M.gguf) | Q4_K_M | 1.218 GB | medium, balanced quality - recommended |
| [Reyna-Mini-1.8B-v0.2-Q5_0.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q5_0.gguf) | Q5_0 | 1.311 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Reyna-Mini-1.8B-v0.2-Q5_K_S.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q5_K_S.gguf) | Q5_K_S | 1.328 GB | large, low quality loss - recommended |
| [Reyna-Mini-1.8B-v0.2-Q5_K_M.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q5_K_M.gguf) | Q5_K_M | 1.377 GB | large, very low quality loss - recommended |
| [Reyna-Mini-1.8B-v0.2-Q6_K.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q6_K.gguf) | Q6_K | 1.579 GB | very large, extremely low quality loss |
| [Reyna-Mini-1.8B-v0.2-Q8_0.gguf](https://huggingface.co/tensorblock/Reyna-Mini-1.8B-v0.2-GGUF/blob/main/Reyna-Mini-1.8B-v0.2-Q8_0.gguf) | Q8_0 | 1.958 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Reyna-Mini-1.8B-v0.2-GGUF --include "Reyna-Mini-1.8B-v0.2-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Reyna-Mini-1.8B-v0.2-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/FuseChat-7B-VaRM-GGUF | tensorblock | 2025-04-21T00:22:23Z | 131 | 0 | transformers | [
"transformers",
"gguf",
"mistral",
"mixtral",
"solar",
"model-fusion",
"fusechat",
"TensorBlock",
"GGUF",
"text-generation",
"en",
"dataset:FuseAI/FuseChat-Mixture",
"base_model:FuseAI/FuseChat-7B-VaRM",
"base_model:quantized:FuseAI/FuseChat-7B-VaRM",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-12T20:51:34Z | ---
license: apache-2.0
language:
- en
base_model: FuseAI/FuseChat-7B-VaRM
datasets:
- FuseAI/FuseChat-Mixture
pipeline_tag: text-generation
tags:
- mistral
- mixtral
- solar
- model-fusion
- fusechat
- TensorBlock
- GGUF
library_name: transformers
model-index:
- name: FuseChat-7B-VaRM
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: MT-Bench
type: unknown
metrics:
- type: unknown
value: 8.22
name: score
source:
url: https://huggingface.co/spaces/lmsys/mt-bench
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 62.88
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=FuseAI/FuseChat-7B-VaRM
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 84.25
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=FuseAI/FuseChat-7B-VaRM
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 63.71
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=FuseAI/FuseChat-7B-VaRM
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 45.67
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=FuseAI/FuseChat-7B-VaRM
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 79.16
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=FuseAI/FuseChat-7B-VaRM
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 63.46
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=FuseAI/FuseChat-7B-VaRM
name: Open LLM Leaderboard
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## FuseAI/FuseChat-7B-VaRM - GGUF
This repo contains GGUF format model files for [FuseAI/FuseChat-7B-VaRM](https://huggingface.co/FuseAI/FuseChat-7B-VaRM).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>GPT4 Correct System: {system_prompt}<|end_of_turn|>GPT4 Correct User: {prompt}<|end_of_turn|>GPT4 Correct Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [FuseChat-7B-VaRM-Q2_K.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [FuseChat-7B-VaRM-Q3_K_S.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [FuseChat-7B-VaRM-Q3_K_M.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [FuseChat-7B-VaRM-Q3_K_L.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [FuseChat-7B-VaRM-Q4_0.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [FuseChat-7B-VaRM-Q4_K_S.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [FuseChat-7B-VaRM-Q4_K_M.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [FuseChat-7B-VaRM-Q5_0.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [FuseChat-7B-VaRM-Q5_K_S.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [FuseChat-7B-VaRM-Q5_K_M.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [FuseChat-7B-VaRM-Q6_K.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [FuseChat-7B-VaRM-Q8_0.gguf](https://huggingface.co/tensorblock/FuseChat-7B-VaRM-GGUF/blob/main/FuseChat-7B-VaRM-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/FuseChat-7B-VaRM-GGUF --include "FuseChat-7B-VaRM-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/FuseChat-7B-VaRM-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/SambaLingo-Thai-Base-GGUF | tensorblock | 2025-04-21T00:22:17Z | 33 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"th",
"en",
"dataset:uonlp/CulturaX",
"base_model:sambanovasystems/SambaLingo-Thai-Base",
"base_model:quantized:sambanovasystems/SambaLingo-Thai-Base",
"license:llama2",
"endpoints_compatible",
"region:us"
] | null | 2024-12-12T20:09:36Z | ---
license: llama2
datasets:
- uonlp/CulturaX
language:
- th
- en
metrics:
- chrf
- accuracy
- bleu
base_model: sambanovasystems/SambaLingo-Thai-Base
tags:
- TensorBlock
- GGUF
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## sambanovasystems/SambaLingo-Thai-Base - GGUF
This repo contains GGUF format model files for [sambanovasystems/SambaLingo-Thai-Base](https://huggingface.co/sambanovasystems/SambaLingo-Thai-Base).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [SambaLingo-Thai-Base-Q2_K.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q2_K.gguf) | Q2_K | 2.653 GB | smallest, significant quality loss - not recommended for most purposes |
| [SambaLingo-Thai-Base-Q3_K_S.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q3_K_S.gguf) | Q3_K_S | 3.079 GB | very small, high quality loss |
| [SambaLingo-Thai-Base-Q3_K_M.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q3_K_M.gguf) | Q3_K_M | 3.429 GB | very small, high quality loss |
| [SambaLingo-Thai-Base-Q3_K_L.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q3_K_L.gguf) | Q3_K_L | 3.728 GB | small, substantial quality loss |
| [SambaLingo-Thai-Base-Q4_0.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q4_0.gguf) | Q4_0 | 3.970 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [SambaLingo-Thai-Base-Q4_K_S.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q4_K_S.gguf) | Q4_K_S | 4.001 GB | small, greater quality loss |
| [SambaLingo-Thai-Base-Q4_K_M.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q4_K_M.gguf) | Q4_K_M | 4.225 GB | medium, balanced quality - recommended |
| [SambaLingo-Thai-Base-Q5_0.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q5_0.gguf) | Q5_0 | 4.809 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [SambaLingo-Thai-Base-Q5_K_S.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q5_K_S.gguf) | Q5_K_S | 4.809 GB | large, low quality loss - recommended |
| [SambaLingo-Thai-Base-Q5_K_M.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q5_K_M.gguf) | Q5_K_M | 4.941 GB | large, very low quality loss - recommended |
| [SambaLingo-Thai-Base-Q6_K.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q6_K.gguf) | Q6_K | 5.700 GB | very large, extremely low quality loss |
| [SambaLingo-Thai-Base-Q8_0.gguf](https://huggingface.co/tensorblock/SambaLingo-Thai-Base-GGUF/blob/main/SambaLingo-Thai-Base-Q8_0.gguf) | Q8_0 | 7.383 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/SambaLingo-Thai-Base-GGUF --include "SambaLingo-Thai-Base-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/SambaLingo-Thai-Base-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/openchat_3.5-GGUF | tensorblock | 2025-04-21T00:22:11Z | 48 | 0 | transformers | [
"transformers",
"gguf",
"openchat",
"mistral",
"C-RLFT",
"TensorBlock",
"GGUF",
"text-generation",
"dataset:openchat/openchat_sharegpt4_dataset",
"dataset:imone/OpenOrca_FLAN",
"dataset:LDJnr/LessWrong-Amplify-Instruct",
"dataset:LDJnr/Pure-Dove",
"dataset:LDJnr/Verified-Camel",
"dataset:tiedong/goat",
"dataset:glaiveai/glaive-code-assistant",
"dataset:meta-math/MetaMathQA",
"dataset:OpenAssistant/oasst_top1_2023-08-25",
"dataset:TIGER-Lab/MathInstruct",
"base_model:openchat/openchat_3.5",
"base_model:quantized:openchat/openchat_3.5",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | text-generation | 2024-12-12T19:05:01Z | ---
license: apache-2.0
tags:
- openchat
- mistral
- C-RLFT
- TensorBlock
- GGUF
datasets:
- openchat/openchat_sharegpt4_dataset
- imone/OpenOrca_FLAN
- LDJnr/LessWrong-Amplify-Instruct
- LDJnr/Pure-Dove
- LDJnr/Verified-Camel
- tiedong/goat
- glaiveai/glaive-code-assistant
- meta-math/MetaMathQA
- OpenAssistant/oasst_top1_2023-08-25
- TIGER-Lab/MathInstruct
library_name: transformers
pipeline_tag: text-generation
base_model: openchat/openchat_3.5
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## openchat/openchat_3.5 - GGUF
This repo contains GGUF format model files for [openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<s>GPT4 Correct System: {system_prompt}<|end_of_turn|>GPT4 Correct User: {prompt}<|end_of_turn|>GPT4 Correct Assistant:
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [openchat_3.5-Q2_K.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q2_K.gguf) | Q2_K | 2.719 GB | smallest, significant quality loss - not recommended for most purposes |
| [openchat_3.5-Q3_K_S.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q3_K_S.gguf) | Q3_K_S | 3.165 GB | very small, high quality loss |
| [openchat_3.5-Q3_K_M.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q3_K_M.gguf) | Q3_K_M | 3.519 GB | very small, high quality loss |
| [openchat_3.5-Q3_K_L.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q3_K_L.gguf) | Q3_K_L | 3.822 GB | small, substantial quality loss |
| [openchat_3.5-Q4_0.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q4_0.gguf) | Q4_0 | 4.109 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [openchat_3.5-Q4_K_S.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q4_K_S.gguf) | Q4_K_S | 4.140 GB | small, greater quality loss |
| [openchat_3.5-Q4_K_M.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q4_K_M.gguf) | Q4_K_M | 4.368 GB | medium, balanced quality - recommended |
| [openchat_3.5-Q5_0.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q5_0.gguf) | Q5_0 | 4.998 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [openchat_3.5-Q5_K_S.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q5_K_S.gguf) | Q5_K_S | 4.998 GB | large, low quality loss - recommended |
| [openchat_3.5-Q5_K_M.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q5_K_M.gguf) | Q5_K_M | 5.131 GB | large, very low quality loss - recommended |
| [openchat_3.5-Q6_K.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q6_K.gguf) | Q6_K | 5.942 GB | very large, extremely low quality loss |
| [openchat_3.5-Q8_0.gguf](https://huggingface.co/tensorblock/openchat_3.5-GGUF/blob/main/openchat_3.5-Q8_0.gguf) | Q8_0 | 7.696 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/openchat_3.5-GGUF --include "openchat_3.5-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/openchat_3.5-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/aanaphi2-v0.1-GGUF | tensorblock | 2025-04-21T00:21:59Z | 19 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"text-generation",
"base_model:mobiuslabsgmbh/aanaphi2-v0.1",
"base_model:quantized:mobiuslabsgmbh/aanaphi2-v0.1",
"license:mit",
"region:us",
"conversational"
] | text-generation | 2024-12-12T17:05:44Z | ---
license: mit
train: false
inference: false
pipeline_tag: text-generation
tags:
- TensorBlock
- GGUF
base_model: mobiuslabsgmbh/aanaphi2-v0.1
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## mobiuslabsgmbh/aanaphi2-v0.1 - GGUF
This repo contains GGUF format model files for [mobiuslabsgmbh/aanaphi2-v0.1](https://huggingface.co/mobiuslabsgmbh/aanaphi2-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
### Human: {prompt}
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [aanaphi2-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q2_K.gguf) | Q2_K | 1.110 GB | smallest, significant quality loss - not recommended for most purposes |
| [aanaphi2-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q3_K_S.gguf) | Q3_K_S | 1.251 GB | very small, high quality loss |
| [aanaphi2-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q3_K_M.gguf) | Q3_K_M | 1.426 GB | very small, high quality loss |
| [aanaphi2-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q3_K_L.gguf) | Q3_K_L | 1.575 GB | small, substantial quality loss |
| [aanaphi2-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q4_0.gguf) | Q4_0 | 1.602 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [aanaphi2-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q4_K_S.gguf) | Q4_K_S | 1.619 GB | small, greater quality loss |
| [aanaphi2-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q4_K_M.gguf) | Q4_K_M | 1.738 GB | medium, balanced quality - recommended |
| [aanaphi2-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q5_0.gguf) | Q5_0 | 1.933 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [aanaphi2-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q5_K_S.gguf) | Q5_K_S | 1.933 GB | large, low quality loss - recommended |
| [aanaphi2-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q5_K_M.gguf) | Q5_K_M | 2.003 GB | large, very low quality loss - recommended |
| [aanaphi2-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q6_K.gguf) | Q6_K | 2.285 GB | very large, extremely low quality loss |
| [aanaphi2-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/aanaphi2-v0.1-GGUF/blob/main/aanaphi2-v0.1-Q8_0.gguf) | Q8_0 | 2.958 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/aanaphi2-v0.1-GGUF --include "aanaphi2-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/aanaphi2-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/KS-SOLAR-10.7B-v0.1-GGUF | tensorblock | 2025-04-21T00:21:55Z | 26 | 0 | null | [
"gguf",
"TensorBlock",
"GGUF",
"en",
"dataset:kyujinpy/Open-platypus-Commercial",
"license:cc-by-4.0",
"endpoints_compatible",
"region:us"
] | null | 2024-12-12T16:29:22Z | ---
license: cc-by-4.0
datasets:
- kyujinpy/Open-platypus-Commercial
language:
- en
tags:
- TensorBlock
- GGUF
base_model: knlp/KS-SOLAR-10.7B-v0.1
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## knlp/KS-SOLAR-10.7B-v0.1 - GGUF
This repo contains GGUF format model files for [knlp/KS-SOLAR-10.7B-v0.1](https://huggingface.co/knlp/KS-SOLAR-10.7B-v0.1).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [KS-SOLAR-10.7B-v0.1-Q2_K.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q2_K.gguf) | Q2_K | 4.003 GB | smallest, significant quality loss - not recommended for most purposes |
| [KS-SOLAR-10.7B-v0.1-Q3_K_S.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q3_K_S.gguf) | Q3_K_S | 4.665 GB | very small, high quality loss |
| [KS-SOLAR-10.7B-v0.1-Q3_K_M.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q3_K_M.gguf) | Q3_K_M | 5.196 GB | very small, high quality loss |
| [KS-SOLAR-10.7B-v0.1-Q3_K_L.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q3_K_L.gguf) | Q3_K_L | 5.651 GB | small, substantial quality loss |
| [KS-SOLAR-10.7B-v0.1-Q4_0.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q4_0.gguf) | Q4_0 | 6.072 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [KS-SOLAR-10.7B-v0.1-Q4_K_S.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q4_K_S.gguf) | Q4_K_S | 6.119 GB | small, greater quality loss |
| [KS-SOLAR-10.7B-v0.1-Q4_K_M.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q4_K_M.gguf) | Q4_K_M | 6.462 GB | medium, balanced quality - recommended |
| [KS-SOLAR-10.7B-v0.1-Q5_0.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q5_0.gguf) | Q5_0 | 7.397 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [KS-SOLAR-10.7B-v0.1-Q5_K_S.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q5_K_S.gguf) | Q5_K_S | 7.397 GB | large, low quality loss - recommended |
| [KS-SOLAR-10.7B-v0.1-Q5_K_M.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q5_K_M.gguf) | Q5_K_M | 7.598 GB | large, very low quality loss - recommended |
| [KS-SOLAR-10.7B-v0.1-Q6_K.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q6_K.gguf) | Q6_K | 8.805 GB | very large, extremely low quality loss |
| [KS-SOLAR-10.7B-v0.1-Q8_0.gguf](https://huggingface.co/tensorblock/KS-SOLAR-10.7B-v0.1-GGUF/blob/main/KS-SOLAR-10.7B-v0.1-Q8_0.gguf) | Q8_0 | 11.404 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/KS-SOLAR-10.7B-v0.1-GGUF --include "KS-SOLAR-10.7B-v0.1-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/KS-SOLAR-10.7B-v0.1-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
tensorblock/Sailor-1.8B-Chat-GGUF | tensorblock | 2025-04-21T00:21:47Z | 19 | 0 | null | [
"gguf",
"multilingual",
"sea",
"sailor",
"sft",
"chat",
"instruction",
"TensorBlock",
"GGUF",
"en",
"zh",
"id",
"th",
"vi",
"ms",
"lo",
"dataset:CohereForAI/aya_dataset",
"dataset:CohereForAI/aya_collection",
"dataset:Open-Orca/OpenOrca",
"base_model:sail/Sailor-1.8B-Chat",
"base_model:quantized:sail/Sailor-1.8B-Chat",
"license:apache-2.0",
"region:us",
"conversational"
] | null | 2024-12-12T15:25:54Z | ---
language:
- en
- zh
- id
- th
- vi
- ms
- lo
datasets:
- CohereForAI/aya_dataset
- CohereForAI/aya_collection
- Open-Orca/OpenOrca
tags:
- multilingual
- sea
- sailor
- sft
- chat
- instruction
- TensorBlock
- GGUF
widget:
- text: ε¦δ½εΆδ½η€ι±ΌοΌ
example_title: Chinese
- text: How to bake fish?
example_title: English
- text: Bagaimana cara memanggang ikan?
example_title: Malay
- text: ΰΈ§ΰΈ΄ΰΈΰΈ΅ΰΈ’ΰΉΰΈ²ΰΈΰΈΰΈ₯ΰΈ²?
example_title: Thai
- text: Bagaimana membuat bakaran ikan?
example_title: Indonesian
- text: LΓ m thαΊΏ nΓ o Δα» nΖ°α»ng cΓ‘?
example_title: Vietnamese
license: apache-2.0
base_model: sail/Sailor-1.8B-Chat
inference: false
---
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/jC7kdl8.jpeg" alt="TensorBlock" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;">
Feedback and support: TensorBlock's <a href="https://x.com/tensorblock_aoi">Twitter/X</a>, <a href="https://t.me/TensorBlock">Telegram Group</a> and <a href="https://x.com/tensorblock_aoi">Discord server</a>
</p>
</div>
</div>
## sail/Sailor-1.8B-Chat - GGUF
This repo contains GGUF format model files for [sail/Sailor-1.8B-Chat](https://huggingface.co/sail/Sailor-1.8B-Chat).
The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4242](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
## Our projects
<table border="1" cellspacing="0" cellpadding="10">
<tr>
<th style="font-size: 25px;">Awesome MCP Servers</th>
<th style="font-size: 25px;">TensorBlock Studio</th>
</tr>
<tr>
<th><img src="https://imgur.com/2Xov7B7.jpeg" alt="Project A" width="450"/></th>
<th><img src="https://imgur.com/pJcmF5u.jpeg" alt="Project B" width="450"/></th>
</tr>
<tr>
<th>A comprehensive collection of Model Context Protocol (MCP) servers.</th>
<th>A lightweight, open, and extensible multi-LLM interaction studio.</th>
</tr>
<tr>
<th>
<a href="https://github.com/TensorBlock/awesome-mcp-servers" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
<th>
<a href="https://github.com/TensorBlock/TensorBlock-Studio" target="_blank" style="
display: inline-block;
padding: 8px 16px;
background-color: #FF7F50;
color: white;
text-decoration: none;
border-radius: 6px;
font-weight: bold;
font-family: sans-serif;
">π See what we built π</a>
</th>
</tr>
</table>
## Prompt template
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>answer
```
## Model file specification
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Sailor-1.8B-Chat-Q2_K.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q2_K.gguf) | Q2_K | 0.847 GB | smallest, significant quality loss - not recommended for most purposes |
| [Sailor-1.8B-Chat-Q3_K_S.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q3_K_S.gguf) | Q3_K_S | 0.954 GB | very small, high quality loss |
| [Sailor-1.8B-Chat-Q3_K_M.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q3_K_M.gguf) | Q3_K_M | 1.016 GB | very small, high quality loss |
| [Sailor-1.8B-Chat-Q3_K_L.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q3_K_L.gguf) | Q3_K_L | 1.056 GB | small, substantial quality loss |
| [Sailor-1.8B-Chat-Q4_0.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q4_0.gguf) | Q4_0 | 1.120 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Sailor-1.8B-Chat-Q4_K_S.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q4_K_S.gguf) | Q4_K_S | 1.158 GB | small, greater quality loss |
| [Sailor-1.8B-Chat-Q4_K_M.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q4_K_M.gguf) | Q4_K_M | 1.218 GB | medium, balanced quality - recommended |
| [Sailor-1.8B-Chat-Q5_0.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q5_0.gguf) | Q5_0 | 1.311 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Sailor-1.8B-Chat-Q5_K_S.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q5_K_S.gguf) | Q5_K_S | 1.328 GB | large, low quality loss - recommended |
| [Sailor-1.8B-Chat-Q5_K_M.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q5_K_M.gguf) | Q5_K_M | 1.377 GB | large, very low quality loss - recommended |
| [Sailor-1.8B-Chat-Q6_K.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q6_K.gguf) | Q6_K | 1.579 GB | very large, extremely low quality loss |
| [Sailor-1.8B-Chat-Q8_0.gguf](https://huggingface.co/tensorblock/Sailor-1.8B-Chat-GGUF/blob/main/Sailor-1.8B-Chat-Q8_0.gguf) | Q8_0 | 1.958 GB | very large, extremely low quality loss - not recommended |
## Downloading instruction
### Command line
Firstly, install Huggingface Client
```shell
pip install -U "huggingface_hub[cli]"
```
Then, downoad the individual model file the a local directory
```shell
huggingface-cli download tensorblock/Sailor-1.8B-Chat-GGUF --include "Sailor-1.8B-Chat-Q2_K.gguf" --local-dir MY_LOCAL_DIR
```
If you wanna download multiple model files with a pattern (e.g., `*Q4_K*gguf`), you can try:
```shell
huggingface-cli download tensorblock/Sailor-1.8B-Chat-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
|
Subsets and Splits