Llama-3.2-11B-Vision-Instruct-FP8-KV

Llama-3.2-11B-Vision-Instruct-FP8-KV

• Introduction

  This model was created by applying Quark with calibration samples from Pile dataset.

• Quantization Stragegy

  • Quantized Layers: All linear layers in MllamaForCausalLM excluding "lm_head"
  • Weight: FP8 symmetric per-tensor
  • Activation: FP8 symmetric per-tensor
  • KV Cache: FP8 symmetric per-tensor
  • Note: The Llama-3.2-11B-Vision-Instruct consists of two parts: the language model (MllamaForCausalLM) and the vision model (MllamaVisionModel). Here, we only quantize the MllamaForCausalLM.

• Quick Start

1. Download and install Quark
2. Run the quantization script in the example folder using the following command line:

export MODEL_DIR = [local model checkpoint folder] or meta-llama/Llama-3.2-11B-Vision-Instruct 
# single GPU
python3 quantize_quark.py \ 
        --model_dir $MODEL_DIR \
        --output_dir Llama-3.2-11B-Vision-Instruct-FP8-KV \                           
        --quant_scheme w_fp8_a_fp8 \
        --kv_cache_dtype fp8 \
        --num_calib_data 128 \
        --model_export quark_safetensors \
        --no_weight_matrix_merge \
        --custom_mode fp8
# If model size is too large for single GPU, please use multi GPU instead.
python3 quantize_quark.py \ 
        --model_dir $MODEL_DIR \
        --output_dir Llama-3.2-11B-Vision-Instruct-FP8-KV \                           
        --quant_scheme w_fp8_a_fp8 \
        --kv_cache_dtype fp8 \
        --num_calib_data 128 \
        --model_export quark_safetensors \
        --no_weight_matrix_merge \
        --multi_gpu \
        --custom_mode fp8

Deployment

Quark has its own export format and allows FP8 quantized models to be efficiently deployed using the vLLM backend(vLLM-compatible).

Evaluation

Quark currently uses perplexity(PPL) as the evaluation metric for accuracy loss before and after quantization.The specific PPL algorithm can be referenced in the quantize_quark.py. The quantization evaluation results are conducted in pseudo-quantization mode, which may slightly differ from the actual quantized inference accuracy. These results are provided for reference only.

Evaluation scores

<table> <tr> <td><strong>Benchmark</strong> </td> <td><strong>Llama-3.2-11B-Vision-Instruct </strong> </td> <td><strong>Llama-3.2-11B-Vision-Instruct-FP8-KV(this model)</strong> </td> </tr> <tr> <td>Perplexity-wikitext2 </td> <td>7.2285 </td> <td>7.2799 </td> </tr> </table>

License

Modifications copyright(c) 2024 Advanced Micro Devices,Inc. All rights reserved.