Create README.md

README.md

@@ -0,0 +1,211 @@
+---
+tags:
+- FP8
+- vllm
+- audio
+license: apache-2.0
+license_link: https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/apache-2.0.md
+language:
+  - en
+base_model: openai/whisper-medium
+library_name: transformers
+---
+
+# whisper-medium-FP8-Dynamic
+
+## Model Overview
+- **Model Architecture:** whisper-medium
+  - **Input:** Audio-Text
+  - **Output:** Text
+- **Model Optimizations:**
+  - **Weight quantization:** FP8
+  - **Activation quantization:** FP8
+- **Release Date:** 04/16/2025
+- **Version:** 1.0
+- **Model Developers:** Neural Magic
+
+Quantized version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium).
+
+### Model Optimizations
+
+This model was obtained by quantizing the weights of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) to FP8 data type, ready for inference with vLLM >= 0.5.2.
+
+## 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.assets.audio import AudioAsset
+from vllm import LLM, SamplingParams
+
+# prepare model
+llm = LLM(
+    model="neuralmagic/whisper-medium-FP8-Dynamic",
+    max_model_len=448,
+    max_num_seqs=400,
+    limit_mm_per_prompt={"audio": 1},
+)
+
+# prepare inputs
+inputs = {  # Test explicit encoder/decoder prompt
+    "encoder_prompt": {
+        "prompt": "",
+        "multi_modal_data": {
+            "audio": AudioAsset("winning_call").audio_and_sample_rate,
+        },
+    },
+    "decoder_prompt": "<|startoftranscript|>",
+}
+
+# generate response
+print("========== SAMPLE GENERATION ==============")
+outputs = llm.generate(inputs, SamplingParams(temperature=0.0, max_tokens=64))
+print(f"PROMPT  : {outputs[0].prompt}")
+print(f"RESPONSE: {outputs[0].outputs[0].text}")
+print("==========================================")
+```
+
+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. 
+
+<details>
+  <summary>Model Creation Code</summary>
+
+```bash
+python quantize.py \
+    --model_path openai/whisper-medium \
+    --quant_path output_dir/whisper-medium-FP8-Dynamic
+```
+
+
+```python
+import argparse
+import torch
+import os
+from datasets import load_dataset
+from transformers import WhisperProcessor
+from llmcompressor import oneshot
+from llmcompressor.modifiers.quantization import QuantizationModifier
+from llmcompressor.transformers.tracing import TraceableWhisperForConditionalGeneration
+from compressed_tensors.quantization import QuantizationType
+
+# --- Args ---
+parser = argparse.ArgumentParser()
+parser.add_argument('--model_path', type=str, required=True)
+parser.add_argument('--quant_path', type=str, required=True)
+parser.add_argument('--observer', type=str, default="minmax")
+args = parser.parse_args()
+
+# --- Load Model ---
+model = TraceableWhisperForConditionalGeneration.from_pretrained(
+    args.model_path,
+    device_map="auto",
+    torch_dtype="auto",
+)
+model.config.forced_decoder_ids = None
+processor = WhisperProcessor.from_pretrained(args.model_path)
+
+# --- Recipe (FP8 Dynamic) ---
+recipe = [
+    QuantizationModifier(
+        targets="Linear",
+        scheme="FP8_DYNAMIC",
+        sequential_targets=["WhisperEncoderLayer", "WhisperDecoderLayer"],
+        ignore=["re:.*lm_head"],
+    )
+]
+
+# --- Run oneshot ---
+oneshot(
+    model=model,
+    recipe=recipe,
+    trust_remote_code_model=True,
+)
+
+# --- Save ---
+os.makedirs(args.quant_path, exist_ok=True)
+model.save_pretrained(args.quant_path, save_compressed=True)
+processor.save_pretrained(args.quant_path)
+
+
+```
+</details>
+
+## Evaluation
+
+The model was evaluated on [LibriSpeech](https://huggingface.co/datasets/lmms-lab/librispeech) and [Fleurs](https://huggingface.co/datasets/lmms-lab/fleurs) datasets using [lmms-eval](https://github.com/EvolvingLMMs-Lab/lmms-eval), via the following commands:
+
+<details>
+<summary>Evaluation Commands</summary>
+  
+Librispeech:
+```
+lmms-eval \
+    --model=whisper_vllm \
+    --model_args="pretrained=neuralmagic-ent/whisper-medium-FP8-Dynamic" \
+    --batch_size 64 \
+    --output_path <output_file_path> \
+    --tasks librispeech
+```
+
+Fleurs:
+```
+lmms-eval \
+    --model=whisper_vllm \
+    --model_args="pretrained=neuralmagic-ent/whisper-medium-FP8-Dynamic" \
+    --batch_size 64 \
+    --output_path <output_file_path> \
+    --tasks fleurs
+```
+</details>
+
+<table>
+  <thead>
+    <tr>
+      <th>Benchmark</th>
+      <th>Split</th>
+      <th>BF16</th>
+      <th>w8a8</th>
+      <th>Recovery (%)</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td rowspan="2"><b>LibriSpeech (WER)</b></td>
+      <td>test-clean</td>
+      <td>2.8269</td>
+      <td>2.8155</td>
+      <td>100.40%</td>
+    </tr>
+    <tr>
+      <td>test-other</td>
+      <td>6.4445</td>
+      <td>6.4124</td>
+      <td>100.50%</td>
+    </tr>
+    <tr>
+      <td rowspan="3"><b>Fleurs (X→en, BLEU)</b></td>
+      <td>cmn_hans_cn</td>
+      <td></td>
+      <td></td>
+      <td></td>
+    </tr>
+    <tr>
+      <td>en</td>
+      <td></td>
+      <td></td>
+      <td></td>
+    </tr>
+    <tr>
+      <td>yue_hant_hk</td>
+      <td></td>
+      <td></td>
+      <td></td>
+    </tr>
+  </tbody>
+</table>