Add convert_finetuning_dataset to CLI

[KuuCi]

This PR allows users to call llmfoundry convert_finetuning_dataset {ARGS} while maintaining correctness with existing convert_finetuning_dataset script. The motivation is for DLE where we want to make the CLI much more intuitive in the docker images

[KuuCi]

test-data-txt-cEry3n runs:
llmfoundry convert_finetuning_dataset \ --dataset "Muennighoff/P3" \ --splits train,validation \ --preprocessor llmfoundry.data.finetuning.tasks:p3_preprocessing_function \ --out_root data_folder

test-data-txt-ccNcwV runs:
python convert_finetuning_dataset.py \ --dataset "Muennighoff/P3" \ --splits train validation \ --preprocessor llmfoundry.data.finetuning.tasks:p3_preprocessing_function \ --out_root data_folder

[KuuCi]

ft-cli-JmJNkX runs:
cd llm-foundry/scripts/data_prep mkdir shakespeare && cd shakespeare curl -O https://ocw.mit.edu/ans7870/6/6.006/s08/lecturenotes/files/t8.shakespeare.txt cd .. llmfoundry convert_finetuning_dataset \ --dataset "Muennighoff/P3" \ --splits train,validation \ --preprocessor "llmfoundry.data.finetuning.tasks:p3_preprocessing_function" \ --out-root "data_folder"

ft-orig-5VBmib runs:
cd llm-foundry/scripts/data_prep mkdir shakespeare && cd shakespeare curl -O https://ocw.mit.edu/ans7870/6/6.006/s08/lecturenotes/files/t8.shakespeare.txt cd .. python convert_finetuning_dataset.py \ --dataset "Muennighoff/P3" \ --splits train validation \ --preprocessor "llmfoundry.data.finetuning.tasks:p3_preprocessing_function" \ --out_root "data_folder"

[KuuCi]

mcli logs ft-orig-8sBz3c runs:
python convert_finetuning_dataset.py \ --dataset "Muennighoff/P3" \ --splits train validation \ --preprocessor "llmfoundry.data.finetuning.tasks:p3_preprocessing_function" \ --out_root "data_folder"

mcli logs ft-cli-KIX0U6 runs:
llmfoundry data_prep convert_finetuning_dataset \ --dataset "Muennighoff/P3" \ --splits train,validation \ --preprocessor "llmfoundry.data.finetuning.tasks:p3_preprocessing_function" \ --out-root "data_folder"