File size: 1,183 Bytes
1fb32b9 1d53e66 1fb32b9 1d53e66 1fb32b9 1d53e66 1fb32b9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 |
export WANDB_PROJECT="xls-r-300m-cv_8-fr"
python run_speech_recognition_ctc.py \
--activation_dropout="0.1" \
--dataset_name="mozilla-foundation/common_voice_8_0" \
--dataset_config_name="fr" \
--eval_steps="500" \
--evaluation_strategy="steps" \
--feat_proj_dropout="0.0" \
--freeze_feature_encoder \
--fp16 \
--gradient_accumulation_steps="8" \
--gradient_checkpointing \
--group_by_length \
--layerdrop="0.0" \
--learning_rate="7.5e-5" \
--length_column_name="input_length" \
--load_best_model_at_end \
--logging_steps="100" \
--mask_feature_length="64" \
--mask_feature_prob="0.25" \
--mask_time_length="10" \
--mask_time_prob="0.75" \
--model_name_or_path="./checkpoint-17000" \
--num_train_epochs="7.0" \
--output_dir="./" \
--overwrite_output_dir \
--per_device_train_batch_size="16" \
--per_device_eval_batch_size="16" \
--preprocessing_num_workers="4" \
--push_to_hub \
--report_to="wandb" \
--save_steps="500" \
--save_total_limit="3" \
--text_column_name="sentence" \
--use_auth_token \
--warmup_steps="2000" \
--do_train --do_eval
|