export CUDA_VISIBLE_DEVICES=0,1,2,3
export PYTHONPATH=/home/qichu_qle/zhiwei/dsv4/atom_m3_eagle${PYTHONPATH:+:${PYTHONPATH}}
export SGLANG_PLUGINS=atom_sglang
export SGLANG_EXTERNAL_MODEL_PACKAGE=atom.plugin.sglang.models
export SGLANG_EXTERNAL_MM_PROCESSOR_PACKAGE=atom.plugin.sglang.models
export AITER_QUICK_REDUCE_QUANTIZATION=INT4
export SGLANG_USE_AITER=1
export SGLANG_AITER_FP8_PREFILL_ATTN=0
export ATOM_FORCE_ATTN_TRITON=1

python3 -m sglang.launch_server \
  --model-path /workspace/shared/data/amd_int/models/MiniMax-M3-MXFP8 \
  --host 127.0.0.1 \
  --port 8000 \
  --trust-remote-code \
  --tensor-parallel-size 4 \
  --attention-backend aiter \
  --mem-fraction-static 0.8 \
  --page-size 128 \
  --context-length 32768 \
  --max-running-requests 128 \
  --disable-radix-cache \
  --model-loader-extra-config '{"online_quant_config": {"global_quant_config": "ptpc_fp8", "exclude_layer": ["lm_head", "model.embed_tokens", "vision_tower", "multi_modal_projector", "patch_merge_mlp", "*block_sparse_moe"]}}' \
  --speculative-algorithm EAGLE3 \
  --speculative-draft-model-path Inferact/MiniMax-M3-EAGLE3 \
  --speculative-num-steps 3 \
  --speculative-eagle-topk 1 \
  2>&1 | tee minimax-m3-mxfp8-eagle3-sglang-server.log