AlexSu1108

Results 5 comments of AlexSu1108

update: Although the release note of 0.6.9 mentioned the support of Azure OpenAI GPT4o, the same error was shown again.

> same problem here. any updates? I think it's caused by the Azure Dall-e 3 tool No update. Other function tools also meet the same problem. I think the problem...

I met the same problem. And I found azure openai gpt4 will also reproduce this error. Here are some info: - Model name: gpt-4o - Model version: 2024-05-13 - Region:...

mark一下,我最近也在跑dpo训练的实验 给我的感觉是dpo训练的时候,z3 offload不会起作用,并不会offload到cpu,不确定是不是真的是这样

这里我也想麻烦问下,我是了dpo full跟dpo lora,给我的感觉显存占用没什么区别?是我的yaml脚本不对吗? dpo_lora.yaml `### model model_name_or_path: /opt/ml/pretrain_model trust_remote_code: true ### method stage: dpo do_train: true finetuning_type: lora lora_rank: 32 lora_dropout: 0.05 lora_target: all pref_beta: 0.1 pref_loss: sigmoid # choices:...