fix

2025-06-10 08:52:58 +08:00
parent a5a4909e6a
commit b98e1ca305
10 changed files with 39 additions and 26 deletions
--- a/nanovllm/config.py
+++ b/nanovllm/config.py
@@ -5,7 +5,7 @@ from transformers import AutoConfig
@dataclass
 class Config:
    model: str = ''
-    max_num_batched_tokens: int = 16384
+    max_num_batched_tokens: int = 32768
    max_num_seqs: int = 512
    max_model_len: int = 4096
    gpu_memory_utilization: float = 0.95