Bug Fix: Correct model name when using a local model (#75)

* fix naming

* fix unexpected change
This commit is contained in:
Zhiyuan He
2025-08-28 16:05:33 +08:00
committed by GitHub
parent d00cc3e4aa
commit 6bd01959cd
+4 -1
View File
@@ -278,7 +278,10 @@ class AgentLightningTrainer(RayPPOTrainer):
self.config.agentlightning.port,
self.config.actor_rollout_ref.rollout.n,
train_information={
"model": self.config.actor_rollout_ref.model.path,
# Note (Zhiyuan): To avoid further patch into vllm async server, using the same sentence to get the naming here.
# However, it is possible that verl updates the naming and causes incompatibility.
# Reference: https://github.com/volcengine/verl/blob/5b5e09d9cc20625e436d01f69d9cc739ff681c54/verl/workers/rollout/vllm_rollout/vllm_async_server.py#L217
"model": "/".join(self.config.actor_rollout_ref.model.path.split("/")[-2:]),
"temperature": self.config.actor_rollout_ref.rollout.temperature,
},
tokenizer=self.tokenizer,