mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
24 lines
526 B
Bash
24 lines
526 B
Bash
#! /usr/bin/env bash
|
|
|
|
cat >./config.yml <<EOF
|
|
cuda_graph_config:
|
|
enable_padding: true
|
|
max_batch_size: 512
|
|
enable_attention_dp: true
|
|
kv_cache_config:
|
|
dtype: fp8
|
|
free_gpu_memory_fraction: 0.8
|
|
stream_interval: 10
|
|
moe_config:
|
|
backend: DEEPGEMM
|
|
EOF
|
|
|
|
trtllm-serve \
|
|
deepseek-ai/DeepSeek-R1 \
|
|
--host localhost --port 8000 \
|
|
--trust_remote_code \
|
|
--max_batch_size 1024 --max_num_tokens 8192 \
|
|
--tp_size 8 --ep_size 8 --pp_size 1 \
|
|
--config ./config.yml \
|
|
--reasoning_parser deepseek-r1
|