zero3保存的模型无法加载

jianzhnie / LLamaTuner

Easy and Efficient Finetuning LLMs. (Supported LLama, LLama2, LLama3, Qwen, Baichuan, GLM , Falcon) 大模型高效量化训练+部署.

https://jianzhnie.github.io/llmtech/

Apache License 2.0

547 stars 60 forks source link

Open charliedream1 opened 10 months ago

charliedream1 commented 10 months ago

zero3保存出来的pytorch.bin只有几兆，推理时加载存在问题

jianzhnie commented 10 months ago

是使用的 lora 训练的吗

charliedream1 commented 10 months ago

Lora和full都用了，都不行，只存出几兆的东西

---原始邮件--- 发件人: @.> 发送时间: 2023年8月24日(周四) 下午3:32 收件人: @.>; 抄送: "Optimus @.**@.>; 主题: Re: [jianzhnie/Efficient-Tuning-LLMs] zero3保存的模型无法加载 (Issue #90)

是使用的 lora 训练的吗

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>