File size: 4,285 Bytes
31138f4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
---
tags:
- int8
- vllm
- llm-compressor
language:
- en
pipeline_tag: text-generation
license: apache-2.0
base_model:
- Qwen/Qwen2.5-3B
---

# Qwen2.5-3B-quantized.w8a16

## Model Overview
- **Model Architecture:** Qwen2
  - **Input:** Text
  - **Output:** Text
- **Model Optimizations:**
  - **Weight quantization:** INT8
- **Intended Use Cases:** Similarly to [Qwen2.5-3B](https://huggingface.co/Qwen/Qwen2.5-3B), this is a base language model.
- **Out-of-scope:** Use in any manner that violates applicable laws or regulations (including trade compliance laws).
- **Release Date:** 10/09/2024
- **Version:** 1.0
- **Model Developers:** Neural Magic

Quantized version of [Qwen2.5-3B](https://huggingface.co/Qwen/Qwen2.5-3B).
It achieves an OpenLLMv1 score of 63.8, compared to 63.6 for [Qwen2.5-3B](https://huggingface.co/Qwen/Qwen2.5-3B).

### Model Optimizations

This model was obtained by quantizing the weights of [Qwen2.5-3B](https://huggingface.co/Qwen/Qwen2.5-3B) to INT8 data type.
This optimization reduces the number of bits per parameter from 16 to 8, reducing the disk size and GPU memory requirements by approximately 50%.

Only the weights of the linear operators within transformers blocks are quantized.
Symmetric per-channel quantization is applied, in which a linear scaling per output dimension maps the INT8 and floating point representations of the quantized weights.
The [GPTQ](https://arxiv.org/abs/2210.17323) algorithm is applied for quantization, as implemented in the [llm-compressor](https://github.com/vllm-project/llm-compressor) library.


## Deployment

This model can be deployed efficiently using the [vLLM](https://docs.vllm.ai/en/latest/) backend, as shown in the example below.

```python
from vllm import LLM, SamplingParams
from transformers import AutoTokenizer

model_id = "neuralmagic/Qwen2.5-3B-quantized.w8a16"
number_gpus = 1
max_model_len = 8192

sampling_params = SamplingParams(temperature=0.7, top_p=0.8, max_tokens=256)

tokenizer = AutoTokenizer.from_pretrained(model_id)

prompt = "Give me a short introduction to large language model."

llm = LLM(model=model_id, tensor_parallel_size=number_gpus, max_model_len=max_model_len)

outputs = llm.generate(prompt, sampling_params)

generated_text = outputs[0].outputs[0].text
print(generated_text)
```

vLLM aslo supports OpenAI-compatible serving. See the [documentation](https://docs.vllm.ai/en/latest/) for more details.



## Evaluation

The model was evaluated on the OpenLLMv1 benchmark, composed of MMLU, ARC-Challenge, GSM-8K, Hellaswag, Winogrande and TruthfulQA.
Evaluation was conducted using [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) and the [vLLM](https://docs.vllm.ai/en/stable/) engine.

### Accuracy

<table>
  <tr>
   <td><strong>Category</strong>
   </td>
   <td><strong>Benchmark</strong>
   </td>
   <td><strong>Qwen2.5-3B</strong>
   </td>
   <td><strong>Qwen2.5-3B-quantized.w8a16<br>(this model)</strong>
   </td>
   <td><strong>Recovery</strong>
   </td>
  </tr>
  <tr>
   <td rowspan="8" ><strong>OpenLLM v1</strong>
   </td>
  </tr>
  <tr>
   <td>MMLU (5-shot)
   </td>
   <td>65.68
   </td>
   <td>65.65
   </td>
   <td>100.0%
   </td>
  </tr>
  <tr>
   <td>ARC Challenge (25-shot)
   </td>
   <td>53.58
   </td>
   <td>53.07
   </td>
   <td>99.0%
   </td>
  </tr>
  <tr>
   <td>GSM-8k (5-shot, strict-match)
   </td>
   <td>68.23
   </td>
   <td>70.05
   </td>
   <td>102.7%
   </td>
  </tr>
  <tr>
   <td>Hellaswag (10-shot)
   </td>
   <td>51.83
   </td>
   <td>51.78
   </td>
   <td>99.9%
   </td>
  </tr>
  <tr>
   <td>Winogrande (5-shot)
   </td>
   <td>70.64
   </td>
   <td>70.56
   </td>
   <td>99.9%
   </td>
  </tr>
  <tr>
  <td>TruthfulQA (0-shot, mc2)
   </td>
   <td>49.93
   </td>
   <td>48.88
   </td>
   <td>99.9%
   </td>
  </tr>
  <tr>
   <td><strong>Average</strong>
   </td>
   <td><strong>63.59</strong>
   </td>
   <td><strong>63.78</strong>
   </td>
   <td><strong>100.3%</strong>
   </td>
  </tr>
</table>

### Reproduction

The results were obtained using the following command:

```
lm_eval \
  --model vllm \
  --model_args pretrained="neuralmagic/Qwen2.5-3B-quantized.w8a16",dtype=auto,max_model_len=4096,add_bos_token=True,tensor_parallel_size=1 \
  --tasks openllm \
  --batch_size auto
```