---
tags:
- fp8
- vllm
---

# Phi-3-mini-128k-instruct-FP8

## Model Overview
* <h3 style="display: inline;">Model Architecture:</h3> Based on and identical to the Phi-3-mini-128k-instruct architecture
* <h3 style="display: inline;">Model Optimizations:</h3> Weights and activations quantized to FP8
* <h3 style="display: inline;">Release Date:</h3> June 29, 2024
* <h3 style="display: inline;">Model Developers:</h3> Neural Magic

Phi-3-mini-128k-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 ~50%.
Part of the [FP8 LLMs for vLLM collection](https://huggingface.co/collections/neuralmagic/fp8-llms-for-vllm-666742ed2b78b7ac8df13127).


## Usage and Creation
Produced using [AutoFP8 with calibration samples from ultrachat](https://github.com/neuralmagic/AutoFP8/blob/147fa4d9e1a90ef8a93f96fc7d9c33056ddc017a/example_dataset.py).

```python
from datasets import load_dataset
from transformers import AutoTokenizer

from auto_fp8 import AutoFP8ForCausalLM, BaseQuantizeConfig

pretrained_model_dir = "microsoft/Phi-3-mini-128k-instruct"
quantized_model_dir = "Phi-3-mini-128k-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")

model = AutoFP8ForCausalLM.from_pretrained(
    pretrained_model_dir, quantize_config=quantize_config
)
model.quantize(examples)
model.save_quantized(quantized_model_dir)
```

Evaluated through a modified version of vLLM with the following script:

```
#!/bin/bash

# Example usage:
# CUDA_VISIBLE_DEVICES=0 ./eval_openllm.sh "neuralmagic/Phi-3-mini-128k-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
```


## Evaluation

Evaluated on the Open LLM Leaderboard evaluations through vLLM.

### Open LLM Leaderboard evaluation scores
|                      | Phi-3-mini-128k-instruct-FP8       | neuralmagic/Phi-3-mini-128k-instruct-FP8<br>(this model) |
| :------------------: | :----------------------: | :------------------------------------------------: |
| arc-c<br>25-shot     | 63.65                    | 64.24                                              |
| hellaswag<br>10-shot | 79.76                    | 79.79                                              |
| mmlu<br>5-shot       | 68.10                    | 67.93                                              |
| truthfulqa<br>0-shot | 53.97                    | 53.50                                              |
| winogrande<br>5-shot | 73.72                    | 74.11                                              |
| gsm8k<br>5-shot      | 75.59                    | 74.37                                              |
| **Average<br>Accuracy**  | **69.13**                    |              **68.99**                                      |
| **Recovery**             | **100%**                     |              **99.80%**                                     |