nm-research's picture
Create README.md
2eefe8c verified
|
raw
history blame
4.8 kB
---
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen2.5-0.5B/blob/main/LICENSE
language:
- en
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-0.5B
tags:
- chat
- neuralmagic
- llmcompressor
---
# Qwen2.5-0.5B-quantized.w8a8
## Model Overview
- **Model Architecture:** Qwen2
- **Input:** Text
- **Output:** Text
- **Model Optimizations:**
- **Activation quantization:** INT8
- **Weight quantization:** INT8
- **Intended Use Cases:** Intended for commercial and research use multiple languages. Similarly to [Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B), this models is intended for assistant-like chat.
- **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 [Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B).
It achieves an average score of 43.93 on the [OpenLLM](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) benchmark (version 1), whereas the unquantized model achieves 44.03.
### Model Optimizations
This model was obtained by quantizing the weights of [Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) to INT8 data type.
This optimization reduces the number of bits used to represent weights and activations from 16 to 8, reducing GPU memory requirements (by approximately 50%) and increasing matrix-multiply compute throughput (by approximately 2x).
Weight quantization also reduces disk size requirements by approximately 50%.
Only weights and activations of the linear operators within transformers blocks are quantized.
Weights are quantized with a symmetric static per-channel scheme, where a fixed linear scaling factor is applied between INT8 and floating point representations for each output channel dimension.
Activations are quantized with a symmetric dynamic per-token scheme, computing a linear scaling factor at runtime for each token between INT8 and floating point representations.
## 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/Qwen2.5-0.5B-quantized.w8a8"
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 [OpenLLM](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) leaderboard tasks (version 1) with the [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness/tree/383bbd54bc621086e05aa1b030d8d4d5635b25e6) (commit 383bbd54bc621086e05aa1b030d8d4d5635b25e6) and the [vLLM](https://docs.vllm.ai/en/stable/) engine, using the following command:
```
lm_eval \
--model vllm \
--model_args pretrained="neuralmagic/Qwen2.5-0.5B-quantized.w8a8",dtype=auto,gpu_memory_utilization=0.9,add_bos_token=True,max_model_len=4096,enable_chunk_prefill=True,tensor_parallel_size=1 \
--tasks openllm \
--batch_size auto
```
### Accuracy
#### Open LLM Leaderboard evaluation scores
<table>
<tr>
<td><strong>Benchmark</strong>
</td>
<td><strong>Qwen2.5-0.5B</strong>
</td>
<td><strong>Qwen2.5-0.5B-quantized.w8a8 (this model)</strong>
</td>
<td><strong>Recovery</strong>
</td>
</tr>
<tr>
<td>MMLU (5-shot)
</td>
<td>47.57
</td>
<td>47.35
</td>
<td>99.5%
</td>
</tr>
<tr>
<td>ARC Challenge (25-shot)
</td>
<td>34.90
</td>
<td>34.47
</td>
<td>98.8%
</td>
</tr>
<tr>
<td>GSM-8K (5-shot, strict-match)
</td>
<td>34.19
</td>
<td>34.19
</td>
<td>100.0%
</td>
</tr>
<tr>
<td>Hellaswag (10-shot)
</td>
<td>51.83
</td>
<td>51.63
</td>
<td>99.6%
</td>
</tr>
<tr>
<td>Winogrande (5-shot)
</td>
<td>55.80
</td>
<td>55.64
</td>
<td>99.7%
</td>
</tr>
<tr>
<td>TruthfulQA (0-shot, mc2)
</td>
<td>39.90
</td>
<td>40.32
</td>
<td>101.1%
</td>
</tr>
<tr>
<td><strong>Average</strong>
</td>
<td><strong>44.03</strong>
</td>
<td><strong>43.93</strong>
</td>
<td><strong>99.8%</strong>
</td>
</tr>
</table>