File size: 6,065 Bytes
b38cf0b f53748a b38cf0b 14b7574 3d039b6 14b7574 019cb18 14b7574 b38cf0b 14b7574 3d039b6 14b7574 b38cf0b 14b7574 b38cf0b 14b7574 b38cf0b 14b7574 019cb18 14b7574 019cb18 14b7574 019cb18 14b7574 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 |
---
tags:
- fp8
- vllm
license: apache-2.0
---
# Qwen2-7B-Instruct-FP8
## Model Overview
- **Model Architecture:** Qwen2
- **Input:** Text
- **Output:** Text
- **Model Optimizations:**
- **Weight quantization:** FP8
- **Activation quantization:** FP8
- **Intended Use Cases:** Intended for commercial and research use in English. Similarly to [Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct), 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). Use in languages other than English.
- **Release Date:** 6/14/2024
- **Version:** 1.0
- **License(s):** [apache-2.0](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/apache-2.0.md)
- **Model Developers:** Neural Magic
Quantized version of [Qwen2-7B-Instruct](https://huggingface.co/Qwen/Qwen2-7B-Instruct).
It achieves an average score of 69.44 on the [OpenLLM](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) benchmark (version 1), whereas the unquantized model achieves 69.55.
### Model Optimizations
This model was obtained by quantizing the weights and activations of [Qwen2-7B-Instruct](https://huggingface.co/Qwen/Qwen2-7B-Instruct) to FP8 data type, ready for inference with vLLM >= 0.5.0.
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. Symmetric per-tensor quantization is applied, in which a single linear scaling maps the FP8 representations of the quantized weights and activations.
[AutoFP8](https://github.com/neuralmagic/AutoFP8) is used for quantization with 512 sequences of UltraChat.
## Deployment
### 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 vllm import LLM, SamplingParams
from transformers import AutoTokenizer
model_id = "neuralmagic/Qwen2-7B-Instruct-FP8"
sampling_params = SamplingParams(temperature=0.6, top_p=0.9, max_tokens=256)
tokenizer = AutoTokenizer.from_pretrained(model_id)
messages = [
{"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"},
{"role": "user", "content": "Who are you?"},
]
prompts = tokenizer.apply_chat_template(messages, tokenize=False)
llm = LLM(model=model_id)
outputs = llm.generate(prompts, 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.
## Creation
This model was created by applying [AutoFP8 with calibration samples from ultrachat](https://github.com/neuralmagic/AutoFP8/blob/147fa4d9e1a90ef8a93f96fc7d9c33056ddc017a/example_dataset.py), as presented in the code snipet below.
Although AutoFP8 was used for this particular model, Neural Magic is transitioning to using [llm-compressor](https://github.com/vllm-project/llm-compressor) which supports several quantization schemes and models not supported by AutoFP8.
```python
from datasets import load_dataset
from transformers import AutoTokenizer
from auto_fp8 import AutoFP8ForCausalLM, BaseQuantizeConfig
pretrained_model_dir = "Qwen/Qwen2-7B-Instruct"
quantized_model_dir = "Qwen2-7B-Instruct-FP8"
tokenizer = AutoTokenizer.from_pretrained(pretrained_model_dir, use_fast=True, model_max_length=4096)
tokenizer.pad_token = tokenizer.eos_token
ds = load_dataset("mgoin/ultrachat_2k", split="train_sft").select(range(512))
examples = [tokenizer.apply_chat_template(batch["messages"], tokenize=False) for batch in ds]
examples = tokenizer(examples, padding=True, truncation=True, return_tensors="pt").to("cuda")
quantize_config = BaseQuantizeConfig(
quant_method="fp8",
activation_scheme="static"
ignore_patterns=["re:.*lm_head"],
)
model = AutoFP8ForCausalLM.from_pretrained(
pretrained_model_dir, quantize_config=quantize_config
)
model.quantize(examples)
model.save_quantized(quantized_model_dir)
```
## 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/Qwen2-7B-Instruct-FP8",dtype=auto,gpu_memory_utilization=0.4,add_bos_token=True,max_model_len=4096 \
--tasks openllm \
--batch_size auto
```
### Accuracy
#### Open LLM Leaderboard evaluation scores
<table>
<tr>
<td><strong>Benchmark</strong>
</td>
<td><strong>Qwen2-7B-Instruct</strong>
</td>
<td><strong>Qwen2-7B-Instruct-FP8(this model)</strong>
</td>
<td><strong>Recovery</strong>
</td>
</tr>
<tr>
<td>MMLU (5-shot)
</td>
<td>70.82
</td>
<td>70.27
</td>
<td>99.22%
</td>
</tr>
<tr>
<td>ARC Challenge (25-shot)
</td>
<td>62.37
</td>
<td>62.03
</td>
<td>99.45%
</td>
</tr>
<tr>
<td>GSM-8K (5-shot, strict-match)
</td>
<td>68.84
</td>
<td>69.83
</td>
<td>101.4%
</td>
</tr>
<tr>
<td>Hellaswag (10-shot)
</td>
<td>81.77
</td>
<td>81.46
</td>
<td>99.62%
</td>
</tr>
<tr>
<td>Winogrande (5-shot)
</td>
<td>76.16
</td>
<td>76.72
</td>
<td>100.7%
</td>
</tr>
<tr>
<td>TruthfulQA (0-shot)
</td>
<td>57.36
</td>
<td>56.34
</td>
<td>98.22%
</td>
</tr>
<tr>
<td><strong>Average</strong>
</td>
<td><strong>69.55</strong>
</td>
<td><strong>69.44</strong>
</td>
<td><strong>99.84%</strong>
</td>
</tr>
</table> |