From da31ce7e1a27a698c60030692c1980881dc7346f Mon Sep 17 00:00:00 2001 From: biondizzle Date: Sat, 16 May 2026 19:23:41 +0000 Subject: [PATCH] allow for cuda graphs again --- docker-compose.yml | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/docker-compose.yml b/docker-compose.yml index 6dded3f7..fd7b48cf 100644 --- a/docker-compose.yml +++ b/docker-compose.yml @@ -15,9 +15,10 @@ services: - --trust-remote-code - --enable-expert-parallel - --tensor-parallel-size=8 - - --enforce-eager - #- --compilation-config + #- --enforce-eager + - --compilation-config #- '{"cudagraph_mode": "FULL_DECODE_ONLY", "custom_ops": ["all"], "cudagraph_capture_sizes": [1, 2, 4, 8], "max_cudagraph_capture_size": 8}' + - '{"cudagraph_mode":"FULL_AND_PIECEWISE", "custom_ops":["all"]}' - --tokenizer-mode=deepseek_v4 - --host=0.0.0.0 - --port=8000