Regular lora target module cannot use imdeploy

modelscope / ms-swift

Use PEFT or Full-parameter to finetune 400+ LLMs or 100+ MLLMs. (LLM: Qwen2.5, Llama3.2, GLM4, Internlm2.5, Yi1.5, Mistral, Baichuan2, DeepSeek, Gemma2, ...; MLLM: Qwen2-VL, Qwen2-Audio, Llama3.2-Vision, Llava, InternVL2, MiniCPM-V-2.6, GLM4v, Xcomposer2.5, Yi-VL, DeepSeek-VL, Phi3.5-Vision, ...)

https://swift.readthedocs.io/zh-cn/latest/Instruction/index.html

Apache License 2.0

4.15k stars 368 forks source link

Regular lora target module cannot use imdeploy #2007

Open orzgugu opened 2 months ago

orzgugu commented 2 months ago

Hello, the checkpoint of internvl2-8b obtained after training with Swift 2.25 and later versions defaults to the regular lora target module, and the current regular lora target module cannot use imdeploy. Is there any solution? Looking forward to your reply, it is very important to me!

orzgugu commented 2 months ago

Or is there a way to force the lora target module to look like a non-regular expression? I really hope to hear from you

tastelikefeet commented 1 month ago

What is the error stack, and the lora settings you are using now please?