search

DRAGNLabs / 301r_retnet

2 stars 1 forks source link

Non-standardized form for parser arguments. #27

Closed KimballNJardine closed 4 months ago

KimballNJardine commented 5 months ago

This will probably be made irrelevant by the YAML branch. However, there are two different styles for parser args. See train_model.py, and tokenize_data.py for example. Examples: train_model.py: parser.add_argument("--dataset-subset", type=str, default="wikitext-2-v1", help="Subset/config to use for Hugging Face dataset.") tokenize_data.py for example: parser.add_argument( "--tokenized_data_name", type=str, required=True, help="Name of tokenized data.")

nprisbrey commented 5 months ago

@KimballNJardine, has this now been resolved after #29?

nprisbrey commented 4 months ago

Since all parser arguments have been dropped as of f91cca2, I'm considering this resolved.