Update README.md

README.md

@@ -7,18 +7,66 @@ tags:
 # Qwen2-7B-Instruct-FP8
 
 ## Model Overview
-* <h3 style="display: inline;">Model Architecture:</h3> Based on and identical to the Qwen2-7B-Instruct architecture
-* <h3 style="display: inline;">Model Optimizations:</h3> Weights and activations quantized to FP8
-* <h3 style="display: inline;">Release Date:</h3> June 14, 2024
-* <h3 style="display: inline;">Model Developers:</h3> Neural Magic
+- **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
+- **Model Developers:** Neural Magic
 
-Qwen2-7B-Instruct quantized to FP8 weights and activations using per-tensor quantization through the [AutoFP8 repository](https://github.com/neuralmagic/AutoFP8), ready for inference with vLLM >= 0.5.0. 
-Calibrated with 512 UltraChat samples to achieve 100% performance recovery on the Open LLM Benchmark evaluations. 
-Reduces space on disk by ~45%.
-Part of the [FP8 LLMs for vLLM collection](https://huggingface.co/collections/neuralmagic/fp8-llms-for-vllm-666742ed2b78b7ac8df13127).
+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.
 
-## Usage and Creation
-Produced using [AutoFP8 with calibration samples from ultrachat](https://github.com/neuralmagic/AutoFP8/blob/147fa4d9e1a90ef8a93f96fc7d9c33056ddc017a/example_dataset.py).
+### 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-channel quantization is applied, in which a linear scaling per output dimension 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
@@ -36,72 +84,113 @@ 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")
+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)
 ```
 
-Evaluated through vLLM with the following script:
+## 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:
 ```
-#!/bin/bash
-
-# Example usage:
-# CUDA_VISIBLE_DEVICES=0 ./eval_openllm.sh "neuralmagic/Qwen2-7B-Instruct-FP8" "tensor_parallel_size=1,max_model_len=4096,add_bos_token=True,gpu_memory_utilization=0.7"
-
-export MODEL_DIR=${1}
-export MODEL_ARGS=${2}
-
-declare -A tasks_fewshot=(
-    ["arc_challenge"]=25
-    ["winogrande"]=5
-    ["truthfulqa_mc2"]=0
-    ["hellaswag"]=10
-    ["mmlu"]=5
-    ["gsm8k"]=5
-)
-
-declare -A batch_sizes=(
-    ["arc_challenge"]="auto"
-    ["winogrande"]="auto"
-    ["truthfulqa_mc2"]="auto"
-    ["hellaswag"]="auto"
-    ["mmlu"]=1
-    ["gsm8k"]="auto"
-)
-
-for TASK in "${!tasks_fewshot[@]}"; do
-    NUM_FEWSHOT=${tasks_fewshot[$TASK]}
-    BATCH_SIZE=${batch_sizes[$TASK]}
-    lm_eval --model vllm \
-        --model_args pretrained=$MODEL_DIR,$MODEL_ARGS \
-        --tasks ${TASK} \
-        --num_fewshot ${NUM_FEWSHOT} \
-        --write_out \
-        --show_config \
-        --device cuda \
-        --batch_size ${BATCH_SIZE} \
-        --output_path="results/${TASK}"
-done
+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
 ```
 
-## Evaluation
-
-Evaluated on the Open LLM Leaderboard evaluations through vLLM.
-
-### Open LLM Leaderboard evaluation scores
-|                      | Qwen2-7B-Instruct       | Qwen2-7B-Instruct-FP8<br>(this model) |
-| :------------------: | :----------------------: | :------------------------------------------------: |
-| arc-c<br>25-shot     | 62.37                    | 62.03                                              |
-| hellaswag<br>10-shot | 81.77                    | 81.46                                              |
-| mmlu<br>5-shot       | 70.82                    | 70.27                                              |
-| truthfulqa<br>0-shot | 57.36                    | 56.34                                              |
-| winogrande<br>5-shot | 76.16                    | 76.72                                              |
-| gsm8k<br>5-shot      | 68.84                    | 69.83                                              |
-| **Average<br>Accuracy**  | **69.55**                    |              **69.44**                                     |
-| **Recovery**             | **100%**                     |              **99.84%**                                     |
-
+### 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>