jraramhoej commited on
Commit
0a68295
·
1 Parent(s): f23ba24

Training in progress, step 2000

Browse files
fine-tune-whisper-streaming.ipynb CHANGED
@@ -875,8 +875,8 @@
875
  "\n",
876
  " <div>\n",
877
  " \n",
878
- " <progress value='1001' max='5000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
879
- " [1001/5000 1:00:56 < 4:03:57, 0.27 it/s, Epoch 6.00/9223372036854775807]\n",
880
  " </div>\n",
881
  " <table border=\"1\" class=\"dataframe\">\n",
882
  " <thead>\n",
@@ -884,9 +884,16 @@
884
  " <th>Step</th>\n",
885
  " <th>Training Loss</th>\n",
886
  " <th>Validation Loss</th>\n",
 
887
  " </tr>\n",
888
  " </thead>\n",
889
  " <tbody>\n",
 
 
 
 
 
 
890
  " </tbody>\n",
891
  "</table><p>"
892
  ],
@@ -911,6 +918,28 @@
911
  " Num examples: Unknown\n",
912
  " Batch size = 8\n",
913
  "Reading metadata...: 3749it [00:00, 12140.15it/s]\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
914
  "The following columns in the evaluation set don't have a corresponding argument in `WhisperForConditionalGeneration.forward` and have been ignored: input_length. If input_length are not expected by `WhisperForConditionalGeneration.forward`, you can safely ignore this message.\n"
915
  ]
916
  }
@@ -949,8 +978,8 @@
949
  "kwargs = {\n",
950
  " \"dataset_tags\": \"mozilla-foundation/common_voice_11_0\",\n",
951
  " \"dataset\": \"Common Voice 11.0\", # a 'pretty' name for the training dataset\n",
952
- " \"language\": \"es\",\n",
953
- " \"model_name\": \"Whisper Small Es - Sanchit Gandhi\", # a 'pretty' name for your model\n",
954
  " \"finetuned_from\": \"openai/whisper-small\",\n",
955
  " \"tasks\": \"automatic-speech-recognition\",\n",
956
  " \"tags\": \"whisper-event\",\n",
 
875
  "\n",
876
  " <div>\n",
877
  " \n",
878
+ " <progress value='2001' max='5000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
879
+ " [2001/5000 2:23:52 < 3:35:51, 0.23 it/s, Epoch 12.01/9223372036854775807]\n",
880
  " </div>\n",
881
  " <table border=\"1\" class=\"dataframe\">\n",
882
  " <thead>\n",
 
884
  " <th>Step</th>\n",
885
  " <th>Training Loss</th>\n",
886
  " <th>Validation Loss</th>\n",
887
+ " <th>Wer</th>\n",
888
  " </tr>\n",
889
  " </thead>\n",
890
  " <tbody>\n",
891
+ " <tr>\n",
892
+ " <td>1000</td>\n",
893
+ " <td>0.023700</td>\n",
894
+ " <td>0.474486</td>\n",
895
+ " <td>37.983881</td>\n",
896
+ " </tr>\n",
897
  " </tbody>\n",
898
  "</table><p>"
899
  ],
 
918
  " Num examples: Unknown\n",
919
  " Batch size = 8\n",
920
  "Reading metadata...: 3749it [00:00, 12140.15it/s]\n",
921
+ "The following columns in the evaluation set don't have a corresponding argument in `WhisperForConditionalGeneration.forward` and have been ignored: input_length. If input_length are not expected by `WhisperForConditionalGeneration.forward`, you can safely ignore this message.\n",
922
+ "Saving model checkpoint to ./checkpoint-1000\n",
923
+ "Configuration saved in ./checkpoint-1000/config.json\n",
924
+ "Model weights saved in ./checkpoint-1000/pytorch_model.bin\n",
925
+ "Feature extractor saved in ./checkpoint-1000/preprocessor_config.json\n",
926
+ "tokenizer config file saved in ./checkpoint-1000/tokenizer_config.json\n",
927
+ "Special tokens file saved in ./checkpoint-1000/special_tokens_map.json\n",
928
+ "added tokens file saved in ./checkpoint-1000/added_tokens.json\n",
929
+ "Feature extractor saved in ./preprocessor_config.json\n",
930
+ "tokenizer config file saved in ./tokenizer_config.json\n",
931
+ "Special tokens file saved in ./special_tokens_map.json\n",
932
+ "added tokens file saved in ./added_tokens.json\n",
933
+ "Reading metadata...: 5194it [00:00, 29994.50it/s]\n",
934
+ "Reading metadata...: 5194it [00:00, 6191.43it/s]\n",
935
+ "Reading metadata...: 5194it [00:00, 30728.37it/s]\n",
936
+ "Reading metadata...: 5194it [00:00, 24715.43it/s]\n",
937
+ "Reading metadata...: 5194it [00:00, 28694.77it/s]\n",
938
+ "Reading metadata...: 5194it [00:00, 30271.03it/s]\n",
939
+ "***** Running Evaluation *****\n",
940
+ " Num examples: Unknown\n",
941
+ " Batch size = 8\n",
942
+ "Reading metadata...: 3749it [00:00, 31235.80it/s]\n",
943
  "The following columns in the evaluation set don't have a corresponding argument in `WhisperForConditionalGeneration.forward` and have been ignored: input_length. If input_length are not expected by `WhisperForConditionalGeneration.forward`, you can safely ignore this message.\n"
944
  ]
945
  }
 
978
  "kwargs = {\n",
979
  " \"dataset_tags\": \"mozilla-foundation/common_voice_11_0\",\n",
980
  " \"dataset\": \"Common Voice 11.0\", # a 'pretty' name for the training dataset\n",
981
+ " \"language\": \"lt\",\n",
982
+ " \"model_name\": \"Whisper Small Lt and Sr\", # a 'pretty' name for your model\n",
983
  " \"finetuned_from\": \"openai/whisper-small\",\n",
984
  " \"tasks\": \"automatic-speech-recognition\",\n",
985
  " \"tags\": \"whisper-event\",\n",
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8ee06303a7fd2e39f1d80b96a96479db82a437369d8c3dc21a61d87f463bc16
3
  size 967102601
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74518845ad11c903cfe2f2d172e21e95d419ae7ec46c4505a27ed98d81746aa0
3
  size 967102601
runs/Dec14_09-49-59_129-146-98-164/events.out.tfevents.1671011416.129-146-98-164.123311.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fca38d91d82dde2b2698413e8423a6cd530616dfa1d87519695e04daaa628ee9
3
- size 10870
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd39fe404f5570f2032a2819e9886b44b2c3ddeece590b74103a283f5638f5b8
3
+ size 17468