Create README.md

README.md

@@ -0,0 +1,233 @@
+---
+tags:
+- fp8
+- vllm
+license: other
+license_name: bigcode-openrail-m
+license_link: https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement
+---
+
+# starcoder2-3b-FP8
+
+## Model Overview
+- **Model Architecture:** starcoder2-3b
+  - **Input:** Text
+  - **Output:** Text
+- **Model Optimizations:**
+  - **Weight quantization:** FP8
+  - **Activation quantization:** FP8
+- **Intended Use Cases:** Intended for commercial and research use in English.
+- **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:** 8/1/2024
+- **Version:** 1.0
+- **License(s):** [bigcode-openrail-m](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement)
+- **Model Developers:** Neural Magic
+
+Quantized version of [starcoder2-3b](https://huggingface.co/bigcode/starcoder2-3b).
+<!-- It achieves an average score of 73.19 on the [OpenLLM](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) benchmark (version 1), whereas the unquantized model achieves 73.48. -->
+It achieves an average score of 35.53 on the [HumanEval+](https://github.com/openai/human-eval?tab=readme-ov-file) benchmark, whereas the unquantized model achieves 35.35.
+
+### Model Optimizations
+
+This model was obtained by quantizing the weights and activations of [starcoder2-3b](https://huggingface.co/bigcode/starcoder2-3b) to FP8 data type, ready for inference with vLLM >= 0.5.2.
+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/starcoder2-3b-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, add_generation_prompt=True)
+
+llm = LLM(model=model_id, trust_remote_code=True, max_model_len=4096)
+
+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 [LLM Compressor with calibration samples from UltraChat](https://github.com/vllm-project/llm-compressor/blob/sa/big_model_support/examples/big_model_offloading/big_model_w8a8_calibrate.py), as presented in the code snipet below.
+A slight modification to the code was made due to the parameters of the model. Running the below code will throw an index error, and simply replacing the erroneous line with ```max_quant_shape = param.shape[0]``` resolves the issue. 
+
+```python
+import torch
+from datasets import load_dataset
+from transformers import AutoTokenizer
+
+from llmcompressor.transformers import SparseAutoModelForCausalLM, oneshot
+from llmcompressor.transformers.compression.helpers import (
+    calculate_offload_device_map,
+    custom_offload_device_map,
+)
+
+recipe = """
+quant_stage:
+    quant_modifiers:
+        QuantizationModifier:
+            ignore: ["lm_head"]
+            config_groups:
+                group_0:
+                    weights:
+                        num_bits: 8
+                        type: float
+                        strategy: tensor
+                        dynamic: false
+                        symmetric: true
+                    input_activations:
+                        num_bits: 8
+                        type: float
+                        strategy: tensor
+                        dynamic: false
+                        symmetric: true
+                    targets: ["Linear"]
+"""
+
+model_stub = "bigcode/starcoder2-3b"
+model_name = model_stub.split("/")[-1]
+
+device_map = calculate_offload_device_map(
+    model_stub, reserve_for_hessians=False, num_gpus=8, torch_dtype=torch.float16
+)
+
+model = SparseAutoModelForCausalLM.from_pretrained(
+    model_stub, torch_dtype=torch.float16, device_map=device_map
+)
+tokenizer = AutoTokenizer.from_pretrained(model_stub)
+
+output_dir = f"./{model_name}-FP8"
+
+DATASET_ID = "HuggingFaceH4/ultrachat_200k"
+DATASET_SPLIT = "train_sft"
+NUM_CALIBRATION_SAMPLES = 512
+MAX_SEQUENCE_LENGTH = 4096
+
+ds = load_dataset(DATASET_ID, split=DATASET_SPLIT)
+ds = ds.shuffle(seed=42).select(range(NUM_CALIBRATION_SAMPLES))
+
+def preprocess(example):
+    return {
+        "text": " ".join([msg["content"] for msg in example["messages"]])
+    }
+
+ds = ds.map(preprocess)
+
+def tokenize(sample):
+    return tokenizer(
+        sample["text"],
+        padding=False,
+        max_length=MAX_SEQUENCE_LENGTH,
+        truncation=True,
+        add_special_tokens=False,
+    )
+
+ds = ds.map(tokenize, remove_columns=ds.column_names)
+
+oneshot(
+    model=model,
+    output_dir=output_dir,
+    dataset=ds,
+    recipe=recipe,
+    max_seq_length=MAX_SEQUENCE_LENGTH,
+    num_calibration_samples=NUM_CALIBRATION_SAMPLES,
+    save_compressed=True,
+)
+```
+
+## Evaluation
+
+The model was evaluated on the [HumanEval+](https://github.com/openai/human-eval?tab=readme-ov-file) benchmark with the [Neural Magic fork](https://github.com/neuralmagic/evalplus) of the [EvalPlus implementation of HumanEval+](https://github.com/evalplus/evalplus) and the [vLLM](https://docs.vllm.ai/en/stable/) engine, using the following command:
+```
+python codegen/generate.py --model neuralmagic/starcoder2-3b-FP8 --temperature 0.2 --n_samples 50 --resume --root ~ --dataset humaneval
+python evalplus/sanitize.py ~/humaneval/neuralmagic--starcoder2-3b-FP8_vllm_temp_0.2
+evalplus.evaluate --dataset humaneval --samples ~/humaneval/neuralmagic--starcoder2-3b-FP8_vllm_temp_0.2-sanitized
+```
+
+### Accuracy
+
+#### HumanEval+ evaluation scores
+<table>
+  <tr>
+   <td><strong>Benchmark</strong>
+   </td>
+   <td><strong>starcoder2-3b</strong>
+   </td>
+   <td><strong>starcoder2-3b-FP8(this model)</strong>
+   </td>
+   <td><strong>Recovery</strong>
+   </td>
+  </tr>
+  <tr>
+   <td>base pass@1
+   </td>
+   <td>30.7
+   </td>
+   <td>30.8
+   </td>
+   <td>100.3%
+   </td>
+  </tr>
+  <tr>
+   <td>base pass@10
+   </td>
+   <td>44.9
+   </td>
+   <td>45.4
+   </td>
+   <td>101.1%
+   </td>
+  </tr>
+  <tr>
+   <td>base+extra pass@1
+   </td>
+   <td>26.6
+   </td>
+   <td>26.5
+   </td>
+   <td>99.62%
+   </td>
+  </tr>
+  <tr>
+   <td>base+extra pass@10
+   </td>
+   <td>39.2
+   </td>
+   <td>39.4
+   </td>
+   <td>100.5%
+   </td>
+  </tr>
+  <tr>
+   <td><strong>Average</strong>
+   </td>
+   <td><strong>35.35</strong>
+   </td>
+   <td><strong>35.53</strong>
+   </td>
+   <td><strong>100.3%</strong>
+   </td>
+  </tr>
+</table>