Training in progress, step 1000

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a27130315b5147db7b9315597e9857ff6aea1ac62aa8f8a1e9b4b3f10ad5fa9
+oid sha256:d209be948a097f78946d50db10de9282b601215dc1a3b48b1770e9ef2a71a643
 size 967102601

run.sh

@@ -226,3 +226,155 @@ python run_speech_recognition_seq2seq_streaming.py \
 --streaming \
 --use_auth_token \
 --push_to_hub
+python run_speech_recognition_seq2seq_streaming.py \
+--model_name_or_path="openai/whisper-small" \
+-- dataset_name = "google/fleurs"
+--dataset_config_name="cmn_hans_cn" \
+--language="chinese" \
+--train_split_name="train+validation" \
+--eval_split_name="test" \
+--model_index_name="Whisper Small Chinese Mandarin Base Dataset" \
+--max_steps="5000" \
+--output_dir="./" \
+--per_device_train_batch_size="64" \
+--per_device_eval_batch_size="32" \
+--logging_steps="25" \
+--learning_rate="1e-5" \
+--warmup_steps="500" \
+--evaluation_strategy="steps" \
+--eval_steps="1000" \
+--save_strategy="steps" \
+--save_steps="1000" \
+--generation_max_length="225" \
+--length_column_name="input_length" \
+--max_duration_in_seconds="30" \
+--text_column_name="sentence" \
+--freeze_feature_encoder="False" \
+--report_to="tensorboard" \
+--metric_for_best_model="wer" \
+--greater_is_better="False" \
+--load_best_model_at_end \
+--gradient_checkpointing \
+--fp16 \
+--overwrite_output_dir \
+--do_train \
+--do_eval \
+--predict_with_generate \
+--do_normalize_eval \
+--streaming \
+--use_auth_token \
+--push_to_hub
+python run_speech_recognition_seq2seq_streaming.py \
+--model_name_or_path="openai/whisper-small" \
+-- dataset_name = "google/fleurs"
+--dataset_config_name="cmn_hans_cn" \
+--language="chinese" \
+--train_split_name="train+validation" \
+--eval_split_name="test" \
+--model_index_name="Whisper Small Chinese Mandarin Base Dataset" \
+--max_steps="5000" \
+--output_dir="./" \
+--per_device_train_batch_size="64" \
+--per_device_eval_batch_size="32" \
+--logging_steps="25" \
+--learning_rate="1e-5" \
+--warmup_steps="500" \
+--evaluation_strategy="steps" \
+--eval_steps="1000" \
+--save_strategy="steps" \
+--save_steps="1000" \
+--generation_max_length="225" \
+--length_column_name="input_length" \
+--max_duration_in_seconds="30" \
+--text_column_name="sentence" \
+--freeze_feature_encoder="False" \
+--report_to="tensorboard" \
+--metric_for_best_model="wer" \
+--greater_is_better="False" \
+--load_best_model_at_end \
+--gradient_checkpointing \
+--fp16 \
+--overwrite_output_dir \
+--do_train \
+--do_eval \
+--predict_with_generate \
+--do_normalize_eval \
+--streaming \
+--use_auth_token \
+--push_to_hub
+python run_speech_recognition_seq2seq_streaming.py \
+--model_name_or_path="Jingmiao/whisper-small-chinese_base" \
+--dataset_name="mozilla-foundation/common_voice_11_" \
+--dataset_config_name="zh-TW" \
+--language="chinese" \
+--train_split_name="train+validation" \
+--eval_split_name="test" \
+--model_index_name="Whisper Small TW on Chinese base" \
+--max_steps="5000" \
+--output_dir="./" \
+--per_device_train_batch_size="64" \
+--per_device_eval_batch_size="32" \
+--logging_steps="25" \
+--learning_rate="1e-5" \
+--warmup_steps="500" \
+--evaluation_strategy="steps" \
+--eval_steps="1000" \
+--save_strategy="steps" \
+--save_steps="1000" \
+--generation_max_length="225" \
+--length_column_name="input_length" \
+--max_duration_in_seconds="30" \
+--text_column_name="sentence" \
+--freeze_feature_encoder="False" \
+--report_to="tensorboard" \
+--metric_for_best_model="wer" \
+--greater_is_better="False" \
+--load_best_model_at_end \
+--gradient_checkpointing \
+--fp16 \
+--overwrite_output_dir \
+--do_train \
+--do_eval \
+--predict_with_generate \
+--do_normalize_eval \
+--streaming \
+--use_auth_token \
+--push_to_hub
+python run_speech_recognition_seq2seq_streaming.py \
+--model_name_or_path="Jingmiao/whisper-small-chinese_base" \
+--dataset_name="mozilla-foundation/common_voice_11_0" \
+--dataset_config_name="zh-TW" \
+--language="chinese" \
+--train_split_name="train+validation" \
+--eval_split_name="test" \
+--model_index_name="Whisper Small TW on Chinese base" \
+--max_steps="5000" \
+--output_dir="./" \
+--per_device_train_batch_size="64" \
+--per_device_eval_batch_size="32" \
+--logging_steps="25" \
+--learning_rate="1e-5" \
+--warmup_steps="500" \
+--evaluation_strategy="steps" \
+--eval_steps="1000" \
+--save_strategy="steps" \
+--save_steps="1000" \
+--generation_max_length="225" \
+--length_column_name="input_length" \
+--max_duration_in_seconds="30" \
+--text_column_name="sentence" \
+--freeze_feature_encoder="False" \
+--report_to="tensorboard" \
+--metric_for_best_model="wer" \
+--greater_is_better="False" \
+--load_best_model_at_end \
+--gradient_checkpointing \
+--fp16 \
+--overwrite_output_dir \
+--do_train \
+--do_eval \
+--predict_with_generate \
+--do_normalize_eval \
+--streaming \
+--use_auth_token \
+--push_to_hub

runs/Dec19_00-14-40_129-213-131-109/1671408888.219753/events.out.tfevents.1671408888.129-213-131-109.84528.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f433a234dc2508d82126b0731dd5cc8f1560762e9d440118665891c86387fc2
+size 5871

runs/Dec19_00-14-40_129-213-131-109/events.out.tfevents.1671408888.129-213-131-109.84528.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a84617fdbd2048bc0117d1c482462c70606a74aee8007acdde8362e5cb73839
+size 4268

runs/Dec19_00-15-22_129-213-131-109/1671408930.5727222/events.out.tfevents.1671408930.129-213-131-109.84922.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c14780a7b4c8a3b50424218afcb84663c3152dfd8d2dc249b2fb0948c3bd2ed6
+size 5871

runs/Dec19_00-15-22_129-213-131-109/events.out.tfevents.1671408930.129-213-131-109.84922.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5377c4b8ee88d4102c4b38f75c1fdf984a6904b6283488ee12a1df4b948f51fe
+size 10851

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:04e3593c44117435b70a5e928d1be360435aa01916b3b5063b037e743189a76f
+oid sha256:66eea850f06cf43eb6ea191a013b50870ed87ef30fcfb5aad2c92fbf844595bd
 size 3579