diff --git a/composte-hermes.yml b/composte-hermes.yml index 2454755..f8426d6 100644 --- a/composte-hermes.yml +++ b/composte-hermes.yml @@ -26,7 +26,7 @@ services: - --quantization=awq - --tensor-parallel-size=2 - --gpu-memory-utilization=0.86 - - --max-model-len=65536 + - --max-model-len=60000 - --hf-overrides={"rope_scaling":{"rope_type":"yarn","factor":2.0,"original_max_position_embeddings":32768}} - --max-num-seqs=4 - --served-model-name=qwen3-14b-awq