wyclike

Results 9 comments of wyclike

could it finetune in qlora or quant 4/8?

补充一个问题：尝试用quant 4 量化 cogAgent的时候，会报一个维度不匹配的错误： File "/root/miniconda3/envs/CogVLM/lib/python3.10/site-packages/sat/model/finetune/lora2.py", line 97, in __init__ self.original.weight.data.copy_(original_obj.weight.data.detach().clone()) RuntimeError: The size of tensor a (4096) must match the size of tensor b (2048) at non-singleton dimension 1

max_depth argument in retrieval_lm/run_short_form.py

Hi, did you find out if removing it affects the end result?

The supplementary materials and complete code

What a great work!Could u send me your experomental codes? I have greate interest in this work and hope to follow it . My email is : [email protected]

failed (exitcode: -11) local_rank: 5 (pid: 11514) of binary: /home/jovyan/data-ws-enr/zconda/envs/swift_ft/bin/python

> 请问你解决这个问题了吗？我现在用qwen-vl也是lora时会训练很久之后遇到这个childfailederror，而且最后给的也是traceback : Signal 11 (SIGSEGV) received by PID xxx的错误，为了避免是显存的问题，特意调成了2b的模型，而且开了bf16，帧率和pixel，token数量都做了压缩，确认显存是足够的 我单机操作也遇到这个问题了

question about evaluator.py

同求

vllm多卡推理遇到的问题

同样的问题

vllm多卡推理遇到的问题

试一下卸载flashattention

About model ckpt

I apologize for my previous doubts. I have replicated your results, and they are indeed solid.

单机多卡微调 Signal 11 (SIGSEGV)

> [@luhan1999](https://github.com/luhan1999) Hi, have you solved this issue, I met the same problem on the same GPU NVIDIA A800-SXM4-80GB hey,i got same problem in same machine as u