marinone94
/

whisper-medium-nordic

Automatic Speech Recognition

Generated from Trainer

hf-asr-leaderboard

Inference Endpoints

Model card Files Files and versions Community

marinone94 commited on Dec 17, 2022

Commit

67e34d0

•

1 Parent(s): 688bd6b

define num steps

Files changed (1) hide show

run.sh +1 -1

run.sh CHANGED Viewed

@@ -9,7 +9,7 @@ python $1run_speech_recognition_seq2seq_streaming.py \
 	--language_eval="sv" \
 	--eval_split_name="test" \
 	--model_index_name="Whisper Medium Nordic" \
-	--num_train_epochs="1" \
 	--output_dir="./" \
 	--per_device_train_batch_size="32" \
 	--per_device_eval_batch_size="16" \

 	--language_eval="sv" \
 	--eval_split_name="test" \
 	--model_index_name="Whisper Medium Nordic" \
+	--max_steps="10000" \
 	--output_dir="./" \
 	--per_device_train_batch_size="32" \
 	--per_device_eval_batch_size="16" \