File size: 7,033 Bytes

---
license: mit
tags:
- deepseek
- fp8
- vllm
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
library_name: transformers
---

# DeepSeek-R1-Distill-Qwen-1.5B-FP8-dynamic

## Model Overview
- **Model Architecture:** Qwen2ForCausalLM
  - **Input:** Text
  - **Output:** Text
- **Model Optimizations:**
  - **Weight quantization:** FP8
  - **Activation quantization:** FP8
- **Release Date:** 2/5/2025
- **Version:** 1.0
- **Model Developers:** Neural Magic

Quantized version of [DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B).


### Model Optimizations

This model was obtained by quantizing the weights and activations of [DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) to FP8 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 and activations of the linear operators within transformers blocks are quantized.
Weights are quantized using a symmetric per-channel scheme, whereas quantizations are quantized using a symmetric per-token scheme.
[LLM Compressor](https://github.com/vllm-project/llm-compressor) is used for quantization.


## Use with vLLM

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

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

number_gpus = 1
model_name = "neuralmagic/DeepSeek-R1-Distill-Qwen-1.5B-dynamic"

tokenizer = AutoTokenizer.from_pretrained(model_name)
sampling_params = SamplingParams(temperature=0.6, max_tokens=256, stop_token_ids=[tokenizer.eos_token_id])
llm = LLM(model=model_name, tensor_parallel_size=number_gpus, trust_remote_code=True)

messages_list = [
    [{"role": "user", "content": "Who are you? Please respond in pirate speak!"}],
]

prompt_token_ids = [tokenizer.apply_chat_template(messages, add_generation_prompt=True) for messages in messages_list]

outputs = llm.generate(prompt_token_ids=prompt_token_ids, sampling_params=sampling_params)

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

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

## Creation

This model was created with [llm-compressor](https://github.com/vllm-project/llm-compressor) by running the code snippet below. 


```python
from transformers import AutoModelForCausalLM, AutoTokenizer
from llmcompressor.modifiers.quantization import QuantizationModifier
from llmcompressor.transformers import oneshot
import os

# Load model
model_stub = "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B"
model_name = model_stub.split("/")[-1]

model = AutoModelForCausalLM.from_pretrained(
    model_stub,
    torch_dtype="auto",
)

tokenizer = AutoTokenizer.from_pretrained(model_stub)

# Configure the quantization algorithm and scheme
recipe = QuantizationModifier(
    targets="Linear",
    scheme="FP8_DYNAMIC",
    ignore=["lm_head"],
)

# Apply quantization
oneshot(
    model=model,
    recipe=recipe,
)

# Save to disk in compressed-tensors format
save_path = model_name + "-FP8-dynamic
model.save_pretrained(save_path)
tokenizer.save_pretrained(save_path)
print(f"Model and tokenizer saved to: {save_path}")
```

## Evaluation

The model was evaluated on OpenLLM Leaderboard [V1](https://huggingface.co/spaces/open-llm-leaderboard-old/open_llm_leaderboard) and [V2](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/), using the following commands:

OpenLLM Leaderboard V1:
```
lm_eval \
  --model vllm \
  --model_args pretrained="neuralmagic/DeepSeek-R1-Distill-Qwen-1.5B-FP8-dynamic",dtype=auto,max_model_len=4096,tensor_parallel_size=1,enable_chunked_prefill=True \
  --tasks openllm \
  --write_out \
  --batch_size auto \
  --output_path output_dir \
  --show_config
```

OpenLLM Leaderboard V2:
```
lm_eval \
  --model vllm \
  --model_args pretrained="neuralmagic/DeepSeek-R1-Distill-Qwen-1.5B-FP8-dynamic",dtype=auto,max_model_len=4096,tensor_parallel_size=1,enable_chunked_prefill=True \
  --apply_chat_template \
  --fewshot_as_multiturn \
  --tasks leaderboard \
  --write_out \
  --batch_size auto \
  --output_path output_dir \
  --show_config
```

### Accuracy

<table>
  <thead>
    <tr>
      <th>Category</th>
      <th>Metric</th>
      <th>deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B</th>
      <th>neuralmagic/DeepSeek-R1-Distill-Qwen-1.5B-FP8-dynamic</th>
      <th>Recovery</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td rowspan="7"><b>OpenLLM V1</b></td>
      <td>ARC-Challenge (Acc-Norm, 25-shot)</td>
      <td>37.02</td>
      <td>37.71</td>
      <td>101.4%</td>
    </tr>
    <tr>
      <td>GSM8K (Strict-Match, 5-shot)</td>
      <td>69.98</td>
      <td>68.99</td>
      <td>98.6%</td>
    </tr>
    <tr>
      <td>HellaSwag (Acc-Norm, 10-shot)</td>
      <td>43.86</td>
      <td>43.61</td>
      <td>99.4%</td>
    </tr>
    <tr>
      <td>MMLU (Acc, 5-shot)</td>
      <td>37.38</td>
      <td>37.22</td>
      <td>99.6%</td>
    </tr>
    <tr>
      <td>TruthfulQA (MC2, 0-shot)</td>
      <td>45.21</td>
      <td>44.77</td>
      <td>99.0%</td>
    </tr>
    <tr>
      <td>Winogrande (Acc, 5-shot)</td>
      <td>54.30</td>
      <td>54.62</td>
      <td>100.6%</td>
    </tr>
    <tr>
      <td><b>Average Score</b></td>
      <td><b>47.99</b></td>
      <td><b>47.82</b></td>
      <td><b>99.7%</b></td>
    </tr>
    <tr>
      <td rowspan="7"><b>OpenLLM V2</b></td>
      <td>IFEval (Inst Level Strict Acc, 0-shot)</td>
      <td>34.37</td>
      <td>34.91</td>
      <td>101.6%</td>
    </tr>
    <tr>
      <td>BBH (Acc-Norm, 3-shot)</td>
      <td>34.44</td>
      <td>34.40</td>
      <td>99.9%</td>
    </tr>
    <tr>
      <td>Math-Hard (Exact-Match, 4-shot)</td>
      <td>0.00</td>
      <td>0.00</td>
      <td>---</td>
    </tr>
    <tr>
      <td>GPQA (Acc-Norm, 0-shot)</td>
      <td>24.67</td>
      <td>25.16</td>
      <td>102.0%</td>
    </tr>
    <tr>
      <td>MUSR (Acc-Norm, 0-shot)</td>
      <td>35.82</td>
      <td>36.61</td>
      <td>102.2%</td>
    </tr>
    <tr>
      <td>MMLU-Pro (Acc, 5-shot)</td>
      <td>11.80</td>
      <td>11.69</td>
      <td>99.1%</td>
    </tr>
    <tr>
      <td><b>Average Score</b></td>
      <td><b>23.52</b></td>
      <td><b>23.79</b></td>
      <td><b>101.2%</b></td>
    </tr>
    <tr>
      <td rowspan="4"><b>Coding</b></td>
      <td>HumanEval (pass@1)</td>
      <td>37.90</td>
      <td>36.40</td>
      <td><b>96.0%</b></td>
    </tr>
    <tr>
      <td>HumanEval (pass@10)</td>
      <td>61.30</td>
      <td>61.30</td>
      <td>100.0%</td>
    </tr>
    <tr>
      <td>HumanEval+ (pass@10)</td>
      <td>33.00</td>
      <td>32.60</td>
      <td>98.8%</td>
    </tr>
    <tr>
      <td>HumanEval+ (pass@10)</td>
      <td>55.90</td>
      <td>56.30</td>
      <td>100.7%</td>
    </tr>
  </tbody>
</table>