Parth Chokhra

Results 16 comments of Parth Chokhra

Hey @CaptainStabs @akovia HOW DID YOU SOLVE THIS ISSUE ?

Hey @shivam07a , How to get threshold for each table with this? Any idea? Also how you know how to interpret result variable

Hey @antoine77340 , nothing is saved in bumpy files in my case. Can you please help with this?

> Hey, @martiansideofthemoon I trained a model on a custom dataset. It completed the training but then threw this error. > > ``` > 100%|█████████████████████████████| 1042301/1042301 [00:00 100%|███████████████████████████████| 456318/456318 [00:00...

Well, I wanted to retrain on my dataset since I had a specific type of question that I wanted to generate. ``` model_class = GPT2LMHeadModel if "gpt2" in args.model_checkpoint else...

``` checkpoint_mymodel_4714.pth model_training_args.bin config.json special_tokens.txt events.out.tfevents.1589258707.2ab40f8c7b8c vocab.json merges.txt ``` This is a list of files created by the model after training and throwing the above the error. All the files...

Well, I am currently using a single Kaggle GPU with 16gb capacity. So are u suggesting to --nproc_per_node =1 ? @martiansideofthemoon How could I increase the accumulation? Tesla P100-PCIE-16GB to...

Well, thanks for your input. It seems to work but it started to show this new error ``` 100%|██████████████████████████████| 1042301/1042301 [00:01

Hey @juandados , I was also facing sometime random duplication any workaround?