training / flax /evaluation_scripts /run_distilled_16_2.sh
GalaktischeGurke's picture
Saving train state of step 5000
4008bf9 verified
raw
history blame
1.1 kB
#!/usr/bin/env bash
python run_eval.py \
--model_name_or_path "sanchit-gandhi/large-16-2-ts-28k-wer-10" \
--subfolder "checkpoint-10000" \
--dataset_name "librispeech_asr+librispeech_asr+common_voice_13_0+voxpopuli+ami-ihm+ami-sdm+peoples_speech-clean+tedlium+switchboard-data+gigaspeech-l+spgispeech+chime4+google/fleurs" \
--dataset_config_name "all+all+en+en+ihm+sdm+clean+release3+all+l+L+1-channel+en_us" \
--dataset_split_name "validation.clean+validation.other+validation+validation+validation+validation+validation+validation+validation+validation+validation+validation+validation" \
--text_column_name "text+text+text+text+text+text+text+text+text+text+text+text+transcription" \
--cache_dir "/home/sanchitgandhi/.cache" \
--dataset_cache_dir "/home/sanchitgandhi/.cache" \
--output_dir "./" \
--wandb_dir "/home/sanchitgandhi/.cache" \
--wandb_project "distil-whisper-eval" \
--wandb_name "large-16-2-eval" \
--per_device_eval_batch_size 64 \
--dtype "bfloat16" \
--dataloader_num_workers 0 \
--report_to "wandb" \
--streaming \
--predict_with_generate