Abdul Waheed

Results 22 comments of Abdul Waheed

Does dataset class still have some problems?

Hi @noshad-vida, can you try running this on Kaggle using this [Notebook](https://www.kaggle.com/eabdul/casa-dialogue-act-classifier?scriptVersionId=52851658) ?

Does dataset class still have some problems?

@noshad-vida this should not happen, may be because you're using a different python environment on Kaggle ....try without installing datasets now the first cell should be `!pip install wandb pytorch-lightning...

Does dataset class still have some problems?

Hi @noshad-vida , Hugginface datasets package has compatibility issues with pytorch and tokenizer/transforemrs due to which I use pandas to read csv datafiles. Try without installing datasets package because its...

Restart from checkpoint not implemented

Hi @glicerico, Apologies for the late response, been busy with conference deadlines and exams due to which I could not review the PR and some of your comments. I Will...

Multi-gpu training still has issues

It's still not working. @pacman100 can you share your working notebook if possible?

ModuleNotFoundError with Multi-node training using SLURM

For me, multi-GPU on a single node works fine. I get that error when I try to train on multiple nodes where not all the nodes have access to the...

ModuleNotFoundError with Multi-node training using SLURM

Please help @jeffra @ShadenSmith @samyam @molly-smith @arashashari @arashb

Automatic adjustment of ZeRO's optimizer state partitioning with a new world size is not currently supported.

Any update on this? @cdj0311

Int8 training example error with multi-gpu - flan-t5

Any update on this? @djaym7 did you manage to fix it?

Example Finetune_opt_bnb_peft.ipynb runs in colab notebook but fails on my system with seemingly the same requirements installed: "expected scalar type Half but found Float"

Did anyone manage to fix it?