HuiGao-NV

Results 17 comments of


                                            HuiGao-NV

fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP

/bot run

fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP

/bot run

fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP

/bot run --disable-fail-fast

fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP

/bot run --stage-list="H100_PCIe-5,H100_PCIe-7"

fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP

/bot reuse-pipeline

fix: Set correct draft_token_nums to dummy requests for torch compilation with MTP

/bot reuse-pipeline

feat: Run PyExecutor's inference flow to estimate max_num_tokens for kv_cache_manager

/bot run --disable-fail-fast

feat: Run PyExecutor's inference flow to estimate max_num_tokens for kv_cache_manager

/bot run --stage-list="A100X-1,A100X-2,A30-7,B200_PCIe-1"

feat: Run PyExecutor's inference flow to estimate max_num_tokens for kv_cache_manager

/bot run --stage-list="A100X-1,A100X-2,A30-7,B200_PCIe-1" --disable-fail-fast

feat: Run PyExecutor's inference flow to estimate max_num_tokens for kv_cache_manager

/bot run --disable-fail-fast --multi-gpu-test

1
2
›