Options:
-i, --input_data_path TEXT Path to dataset in .csv format
-t, --table_name TEXT Table name related to the supplied dataset
-e, --eval_data_path TEXT Path to eval dataset in .csv format
-m, --model_name TEXT Model name to use for inference
-s, --sample_qna_path TEXT Path to sample QnA in .csv format
-k, --kwargs TEXT Additional arguments
--help Show this message and exit.
- `python eval/inference.py run-eval -i <input_data.csv> -t "eval_test" -e <eval/input_context.csv> -s <sample_qa.csv>`
Currently, simply use `cosine similarity`, other metrics pending - e.g. capturing syntax error might be helpful
Options: -i, --input_data_path TEXT Path to dataset in .csv format -t, --table_name TEXT Table name related to the supplied dataset -e, --eval_data_path TEXT Path to eval dataset in .csv format -m, --model_name TEXT Model name to use for inference -s, --sample_qna_path TEXT Path to sample QnA in .csv format -k, --kwargs TEXT Additional arguments --help Show this message and exit.