fix : FSDP 优化器顺序、温度除零、调度器静默死亡、ref模型设备
- executor: use_orig_params 硬编码 True,FSDP 不替换 Parameter 对象 - strategy: DPO/GRPO ref 模型创建后移到 device - sample: TemperatureStrategy clamp 1e-8,engine 验证改为 >0 - scheduler: 异常不 re-raise 避免 daemon 静默死亡,stop() 发回调给 waiting 任务
This commit is contained in:
@@ -218,7 +218,6 @@ class FSDPExecutor(BaseExecutor):
|
||||
sync_module_states: bool = False,
|
||||
forward_prefetch: bool = False,
|
||||
limit_all_gathers: bool = True,
|
||||
use_orig_params: bool = False,
|
||||
ignored_states=None,
|
||||
device_mesh=None,
|
||||
):
|
||||
@@ -237,7 +236,7 @@ class FSDPExecutor(BaseExecutor):
|
||||
sync_module_states=sync_module_states,
|
||||
forward_prefetch=forward_prefetch,
|
||||
limit_all_gathers=limit_all_gathers,
|
||||
use_orig_params=use_orig_params,
|
||||
use_orig_params=True,
|
||||
ignored_states=ignored_states,
|
||||
device_mesh=device_mesh,
|
||||
).items()
|
||||
|
||||
Reference in New Issue
Block a user