Support limit thinking lengths (#4070)

Co-authored-by: K11OntheBoat <“ruianmaidanglao@163.com”>
2025-10-05 08:37:06 +08:00 · 2025-09-17 12:40:08 +08:00
parent b41988f4bc
commit 7f9a9b37f3
8 changed files with 184 additions and 26 deletions
--- a/fastdeploy/config.py
+++ b/fastdeploy/config.py
@@ -130,6 +130,7 @@ class ModelConfig:
        self.quantization = None
        self.pad_token_id: int = -1
        self.eos_tokens_lens: int = 2
+        self.think_end_id = None
        self.lm_head_fp32: bool = False
        self.model_format = "auto"
        self.partial_rotary_factor: float = 1.0