Ekansh Verma

Request to release embeddings for 1M words

1

If possible, can you please release embeddings for the top 1M words? I believe this is a good tradeoff between the two released embeddings and will be helpful in using...

Set gradients to zero for integrated gradients ?

1

I am trying to calculate integrated gradients for a specific model and multiple inputs in a loop. Do we need to set the model parameter and input gradients to zero...

Error encountered while Data Preparation

6

First of all, Thank you for releasing the code and creating detailed instructions for data preparation and model training. While using the code repository, on executing the command 'java -Xmx4g...

Pass device in launch_concordia_challenge_evaluation.py

2

If we are using local model, we need to pass device to utilise the gpu for inference. However, in `launch_concordia_challenge_evaluation.py ` ``` # Language Model setup model = utils.language_model_setup( api_type=args.api_type,...