File size: 1,194 Bytes
7d30caa 0a1c33d a90a966 0a1c33d fbe783d a90a966 0a1c33d fbe783d 0a1c33d 3870354 0a1c33d 14d27e0 0a1c33d |
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 39 |
python join_datasets_asr_ctc.py \
--dataset_name="mozilla-foundation/common_voice_9_0,marinone94/nst_sv" \
--model_name_or_path="KBLab/wav2vec2-large-voxrex" \
--dataset_config_name="sv-SE,distant_channel" \
--train_split_name="train+validation,train" \
--eval_split_name="test,None" \
--output_dir="./" \
--overwrite_output_dir \
--num_train_epochs="10" \
--per_device_train_batch_size="32" \
--per_device_eval_batch_size="32" \
--gradient_accumulation_steps="4" \
--learning_rate="7.5e-4" \
--max_train_samples="10000" \
--max_eval_samples="100" \
--warmup_ratio="0.0" \
--length_column_name="input_length" \
--evaluation_strategy="steps" \
--save_strategy="steps" \
--eval_steps="50" \
--save_steps="50" \
--text_column_name="sentence" \
--logging_steps="20" \
--layerdrop="0.0" \
--activation_dropout="0.1" \
--save_total_limit="2" \
--freeze_feature_encoder \
--feat_proj_dropout="0.0" \
--mask_time_prob="0.75" \
--mask_time_length="10" \
--mask_feature_prob="0.25" \
--mask_feature_length="64" \
--dataset_seed="42" \
--gradient_checkpointing \
--use_auth_token \
--preprocessing_only \
--group_by_length \
--do_train --do_eval \
--push_to_hub |