Update README.md

4256015 verified 3 months ago

3.08 kB

license: other
license_name: modified-mit
license_link: LICENSE
base_model:
  - moonshotai/Kimi-K2-Thinking

Model Overview

Model Architecture: Kimi-K2-Thinking
- Input: Text
- Output: Text
Supported Hardware Microarchitecture: AMD MI350/MI355
ROCm: 7.0
Operating System(s): Linux
Inference Engine: vLLM
Model Optimizer: AMD-Quark
- Weight quantization: OCP MXFP4, Static
- Activation quantization: OCP MXFP4, Dynamic
Calibration Dataset: Pile

This model was built with Kimi-K2-Thinking model by applying AMD-Quark for MXFP4 quantization.

Model Quantization

The model was quantized from moonshotai/Kimi-K2-Thinking using AMD-Quark. The weights and activations are quantized to MXFP4.

Quantization scripts:

cd Quark/examples/torch/language_modeling/llm_ptq/
exclude_layers="*self_attn* *mlp.gate *lm_head *mlp.gate_proj *mlp.up_proj *mlp.down_proj *shared_experts.*"

python3 quantize_quark.py \
    --model_dir moonshotai/Kimi-K2-Thinking \
    --quant_scheme mxfp4 \
    --exclude_layers $exclude_layers \
    --num_calib_data 128 \
    --output_dir amd/Kimi-K2-Thinking-MXFP4 \
    --model_export hf_format \
    --device cpu

Deployment

Use with vLLM

This model can be deployed efficiently using the vLLM backend.

Evaluation

The model was evaluated on GSM8K benchmarks.

Accuracy

Benchmark	Kimi-K2-Thinking	Kimi-K2-Thinking-MXFP4(this model)	Recovery
GSM8K	94.16	93.48	99.28%

Reproduction

The GSM8K results were obtained using the lm-evaluation-harness framework, based on the Docker image rocm/vllm-dev:base, with vLLM and lm-eval compiled and installed from source inside the container.

Launching server

export VLLM_ATTENTION_BACKEND="TRITON_MLA"
export VLLM_ROCM_USE_AITER=1
export VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS=0

vllm serve amd/Kimi-K2-Thinking-MXFP4 \
  --tensor-parallel-size 8 \
  --enable-auto-tool-choice \
  --tool-call-parser kimi_k2 \
  --reasoning-parser kimi_k2 \
  --trust-remote-code

Evaluating model in a new terminal

lm_eval \
  --model local-completions \
  --model_args "model=amd/Kimi-K2-Thinking-MXFP4,base_url=http://0.0.0.0:8000/v1/completions,tokenized_requests=False,tokenizer_backend=None,num_concurrent=32" \
  --tasks gsm8k \
  --num_fewshot 5 \
  --batch_size 1

amd
/

Kimi-K2-Thinking-MXFP4