#	--dataset_name="NbAiLab/NPSC" \
#	--dataset_config_name="48K_mp3" \
#	--text_column_name="text" \
	

python run_speech_recognition_ctc_bnb.py \
	--dataset_name="mozilla-foundation/common_voice_7_0" \
	--dataset_config_name="sv-SE" \
	--text_column_name="sentence" \
	--model_name_or_path="KBLab/wav2vec2-large-voxrex" \
        --output_dir="./" \
        --overwrite_output_dir \
        --hub_model_id="NbAiLab/xls-npsc-oh" \
        --num_train_epochs="5" \
        --per_device_train_batch_size="32" \
        --per_device_eval_batch_size="32" \
        --gradient_accumulation_steps="4" \
        --learning_rate="5e-5" \
        --warmup_steps="1000" \
        --evaluation_strategy="steps" \
	--length_column_name="input_length" \
        --chars_to_ignore , ? . ! \- \; \: \" " % ‘ " � — ’ … – \
        --save_steps="1000" \
        --eval_steps="1000" \
        --logging_steps="1000" \
        --layerdrop="0.0" \
        --activation_dropout="0.1" \
        --save_total_limit="3" \
        --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" \
        --gradient_checkpointing \
        --use_auth_token \
        --fp16 \
        --group_by_length \
        --do_train --do_eval \
        --max_eval_samples="100"  \
        --push_to_hub