DataAnalysist comments

Results 15 comments of


                                            DataAnalysist

使用KTO进行多机训练过程中再进行验证，报错RuntimeError: still have inglight params [{id：388, "status":"AVALIBLE"}]

我也碰到相似的问题。如下所示： Invalidate trace cache @ step 0 and module 0: cache has only 0 modules Invalidate trace cache @ step 9: expected module 11, but got module 18 RuntimeErrorstill have...

使用KTO进行多机训练过程中再进行验证，报错RuntimeError: still have inglight params [{id：388, "status":"AVALIBLE"}]

> 使用 ZeRO3 时候请关闭验证集相关参数 > > ```yaml > val_size: 0.1 > per_device_eval_batch_size: 1 > evaluation_strategy: steps > eval_steps: 500 > ``` zero3关闭后已经解决

[NPU]目前华为昇腾+llamaFactory多卡训练和推理

请问支持Atlas 300I Duo推理卡吗

模型训练，出现deepspeed_init() got an unexpected keyword argument 'resume_from_checkpoint'

> 请问是怎么解决的 transformer版本调整为4.28.1就可以

cannot import binding error

i have same issue，solved by： export LD_PRELOAD=/your path/hnswlib.xxx.so: