!torchrun train.py --config configs/demo.yaml may perform faster

[YunyaGaoTree]

Hello authors,

Thanks so much for sharing these codes. The codes are very useful to fine-tune SAM for downstream works : )

I reduced datasize, adapted the codes and run them in Google Colab with one A100 GPU. I found using torchrun is faster and more convenient compared to the original commands as shows below. https://pytorch.org/docs/stable/elastic/run.html

!torchrun train.py --config configs/demo.yaml

CUDA_VISIBLE_DEVICES=0,1,2,3 python -m torch.distributed.launch --nnodes 1 --nproc_per_node 4 loadddptrain.py --config configs/demo.yaml Besides loadddptrain.py seems wrong in the GitHub.

Thanks so much again. Kind regards.

[tianrun-chen]

Thanks for your comments!

[langge52]

Hello YunyaGaoTree

Thank you very much for your contribution. I would like to know the content of your loadddptrain.py file? I am currently unable to achieve single machine multi card training.

Thank you again!

[ericzw]

where is the loadddptrain.py?

[langge52]

Thank you very much for your reply on the solution. We have also reproduced the relevant code ourselves and completed distributed training. Thank you again for your reply.

------------------ 原始邮件 ------------------ 发件人: "tianrun-chen/SAM-Adapter-PyTorch" @.>; 发送时间: 2023年12月8日(星期五) 中午12:06 @.>; @.@.>; 主题: Re: [tianrun-chen/SAM-Adapter-PyTorch] !torchrun train.py --config configs/demo.yaml may perform faster (Issue #39)

where is the loadddptrain.py?

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

[ericzw]

Thank you very much for your reply on the solution. We have also reproduced the relevant code ourselves and completed distributed training. Thank you again for your reply. … ------------------ 原始邮件 ------------------ 发件人: "tianrun-chen/SAM-Adapter-PyTorch" @.>; 发送时间: 2023年12月8日(星期五) 中午12:06 @.>; @.@.>; 主题: Re: [tianrun-chen/SAM-Adapter-PyTorch] !torchrun train.py --config configs/demo.yaml may perform faster (Issue #39) where is the loadddptrain.py? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

thanks, but it seems that you do not update the train.py?

[YunyaGaoTree]

Thank you very much for your reply on the solution. We have also reproduced the relevant code ourselves and completed distributed training. Thank you again for your reply. … ------------------ 原始邮件 ------------------ 发件人: "tianrun-chen/SAM-Adapter-PyTorch" @.>; 发送时间: 2023年12月8日(星期五) 中午12:06 @.>; @.@.>; 主题: Re: [tianrun-chen/SAM-Adapter-PyTorch] !torchrun train.py --config configs/demo.yaml may perform faster (Issue #39) where is the loadddptrain.py? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

sorry for the late reply, so glad to hear that you solved the problem!

[demoyu123]

Hello authors,

Thanks so much for sharing these codes. The codes are very useful to fine-tune SAM for downstream works : )

I reduced datasize, adapted the codes and run them in Google Colab with one A100 GPU. I found using torchrun is faster and more convenient compared to the original commands as shows below. https://pytorch.org/docs/stable/elastic/run.html

!torchrun train.py --config configs/demo.yaml

CUDA_VISIBLE_DEVICES=0,1,2,3 python -m torch.distributed.launch --nnodes 1 --nproc_per_node 4 loadddptrain.py --config configs/demo.yaml Besides loadddptrain.py seems wrong in the GitHub.

Thanks so much again. Kind regards.

Would I like to ask if you have a loadddptrain.py？

[parisaalipour]

Hello, Thank you for sharing these codes! I noticed that the loadddptrain.py file seems to be missing, and I’m unable to run SAM2Adapter without it. Could you kindly provide this file?