Hugging Face's logo

jraramhoej
/
whisper-small-lt-sr-v2

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
whisper
whisper-event
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
2
jraramhoej commited on
Commit
4381de3
·
1 Parent(s): 36a641a

Training in progress, step 2000

Browse files
Files changed (3) hide show
  1. fine-tune-whisper-streaming.ipynb +125 -2
  2. pytorch_model.bin +1 -1
  3. runs/Dec15_14-18-11_129-213-18-196/events.out.tfevents.1671113929.129-213-18-196.4128231.0 +2 -2
fine-tune-whisper-streaming.ipynb CHANGED
@@ -876,8 +876,8 @@
876
  "\n",
877
  " <div>\n",
878
  " \n",
879
- " <progress value='1001' max='2000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
880
- " [1001/2000 1:45:14 < 1:45:14, 0.16 it/s, Epoch 49.01/9223372036854775807]\n",
881
  " </div>\n",
882
  " <table border=\"1\" class=\"dataframe\">\n",
883
  " <thead>\n",
@@ -885,9 +885,16 @@
885
  " <th>Step</th>\n",
886
  " <th>Training Loss</th>\n",
887
  " <th>Validation Loss</th>\n",
 
888
  " </tr>\n",
889
  " </thead>\n",
890
  " <tbody>\n",
 
 
 
 
 
 
891
  " </tbody>\n",
892
  "</table><p>"
893
  ],
@@ -1004,6 +1011,122 @@
1004
  " Num examples: Unknown\n",
1005
  " Batch size = 8\n",
1006
  "Reading metadata...: 677it [00:00, 4738.28it/s]\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1007
  "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"
1008
  ]
1009
  }
 
876
  "\n",
877
  " <div>\n",
878
  " \n",
879
+ " <progress value='2001' max='2000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
880
+ " [2000/2000 3:36:26, Epoch 99.01/9223372036854775807]\n",
881
  " </div>\n",
882
  " <table border=\"1\" class=\"dataframe\">\n",
883
  " <thead>\n",
 
885
  " <th>Step</th>\n",
886
  " <th>Training Loss</th>\n",
887
  " <th>Validation Loss</th>\n",
888
+ " <th>Wer</th>\n",
889
  " </tr>\n",
890
  " </thead>\n",
891
  " <tbody>\n",
892
+ " <tr>\n",
893
+ " <td>1000</td>\n",
894
+ " <td>0.000300</td>\n",
895
+ " <td>0.603500</td>\n",
896
+ " <td>26.790773</td>\n",
897
+ " </tr>\n",
898
  " </tbody>\n",
899
  "</table><p>"
900
  ],
 
1011
  " Num examples: Unknown\n",
1012
  " Batch size = 8\n",
1013
  "Reading metadata...: 677it [00:00, 4738.28it/s]\n",
1014
+ "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",
1015
+ "Saving model checkpoint to ./checkpoint-1000\n",
1016
+ "Configuration saved in ./checkpoint-1000/config.json\n",
1017
+ "Model weights saved in ./checkpoint-1000/pytorch_model.bin\n",
1018
+ "Feature extractor saved in ./checkpoint-1000/preprocessor_config.json\n",
1019
+ "tokenizer config file saved in ./checkpoint-1000/tokenizer_config.json\n",
1020
+ "Special tokens file saved in ./checkpoint-1000/special_tokens_map.json\n",
1021
+ "added tokens file saved in ./checkpoint-1000/added_tokens.json\n",
1022
+ "Feature extractor saved in ./preprocessor_config.json\n",
1023
+ "tokenizer config file saved in ./tokenizer_config.json\n",
1024
+ "Special tokens file saved in ./special_tokens_map.json\n",
1025
+ "added tokens file saved in ./added_tokens.json\n",
1026
+ "Reading metadata...: 1045it [00:00, 34758.23it/s]\n",
1027
+ "Reading metadata...: 623it [00:00, 30640.85it/s]\n",
1028
+ "Reading metadata...: 1045it [00:00, 8336.63it/s]\n",
1029
+ "Reading metadata...: 623it [00:00, 28144.84it/s]\n",
1030
+ "Reading metadata...: 1045it [00:00, 39837.56it/s]\n",
1031
+ "Reading metadata...: 623it [00:00, 13953.02it/s]\n",
1032
+ "Reading metadata...: 1045it [00:00, 4443.95it/s]\n",
1033
+ "Reading metadata...: 623it [00:00, 2417.91it/s]\n",
1034
+ "Reading metadata...: 1045it [00:00, 44040.11it/s]\n",
1035
+ "Reading metadata...: 623it [00:00, 32592.25it/s]\n",
1036
+ "Reading metadata...: 1045it [00:00, 36922.93it/s]\n",
1037
+ "Reading metadata...: 623it [00:00, 28641.20it/s]\n",
1038
+ "Reading metadata...: 1045it [00:00, 43164.02it/s]\n",
1039
+ "Reading metadata...: 623it [00:00, 19506.79it/s]\n",
1040
+ "Reading metadata...: 1045it [00:00, 20683.53it/s]\n",
1041
+ "Reading metadata...: 623it [00:00, 18646.81it/s]\n",
1042
+ "Reading metadata...: 1045it [00:00, 12424.52it/s]\n",
1043
+ "Reading metadata...: 623it [00:00, 25188.47it/s]\n",
1044
+ "Reading metadata...: 1045it [00:00, 36641.43it/s]\n",
1045
+ "Reading metadata...: 623it [00:00, 33033.95it/s]\n",
1046
+ "Reading metadata...: 1045it [00:00, 40875.20it/s]\n",
1047
+ "Reading metadata...: 623it [00:00, 28758.46it/s]\n",
1048
+ "Reading metadata...: 1045it [00:00, 42691.05it/s]\n",
1049
+ "Reading metadata...: 623it [00:00, 14031.61it/s]\n",
1050
+ "Reading metadata...: 1045it [00:00, 4653.45it/s]\n",
1051
+ "Reading metadata...: 623it [00:00, 2405.97it/s]\n",
1052
+ "Reading metadata...: 1045it [00:00, 9632.14it/s]\n",
1053
+ "Reading metadata...: 623it [00:00, 4707.13it/s]\n",
1054
+ "Reading metadata...: 1045it [00:00, 33352.97it/s]\n",
1055
+ "Reading metadata...: 623it [00:00, 28545.46it/s]\n",
1056
+ "Reading metadata...: 1045it [00:00, 39528.22it/s]\n",
1057
+ "Reading metadata...: 623it [00:00, 21992.24it/s]\n",
1058
+ "Reading metadata...: 1045it [00:00, 8611.72it/s]\n",
1059
+ "Reading metadata...: 623it [00:00, 5601.53it/s]\n",
1060
+ "Reading metadata...: 1045it [00:00, 36653.38it/s]\n",
1061
+ "Reading metadata...: 623it [00:00, 29095.65it/s]\n",
1062
+ "Reading metadata...: 1045it [00:00, 22574.76it/s]\n",
1063
+ "Reading metadata...: 623it [00:00, 29994.74it/s]\n",
1064
+ "Reading metadata...: 1045it [00:00, 40011.75it/s]\n",
1065
+ "Reading metadata...: 623it [00:00, 29296.93it/s]\n",
1066
+ "Reading metadata...: 1045it [00:00, 37057.48it/s]\n",
1067
+ "Reading metadata...: 623it [00:00, 21988.53it/s]\n",
1068
+ "Reading metadata...: 1045it [00:00, 41209.16it/s]\n",
1069
+ "Reading metadata...: 623it [00:00, 17949.25it/s]\n",
1070
+ "Reading metadata...: 1045it [00:00, 33757.56it/s]\n",
1071
+ "Reading metadata...: 623it [00:00, 21740.64it/s]\n",
1072
+ "Reading metadata...: 1045it [00:00, 25435.96it/s]\n",
1073
+ "Reading metadata...: 623it [00:00, 17388.35it/s]\n",
1074
+ "Reading metadata...: 1045it [00:00, 41428.45it/s]\n",
1075
+ "Reading metadata...: 623it [00:00, 29616.02it/s]\n",
1076
+ "Reading metadata...: 1045it [00:00, 43414.57it/s]\n",
1077
+ "Reading metadata...: 623it [00:00, 33342.07it/s]\n",
1078
+ "Reading metadata...: 1045it [00:00, 43046.18it/s]\n",
1079
+ "Reading metadata...: 623it [00:00, 32253.92it/s]\n",
1080
+ "Reading metadata...: 1045it [00:00, 6045.31it/s]\n",
1081
+ "Reading metadata...: 623it [00:00, 3404.43it/s]\n",
1082
+ "Reading metadata...: 1045it [00:00, 36673.01it/s]\n",
1083
+ "Reading metadata...: 623it [00:00, 28305.51it/s]\n",
1084
+ "Reading metadata...: 1045it [00:00, 40369.22it/s]\n",
1085
+ "Reading metadata...: 623it [00:00, 28357.11it/s]\n",
1086
+ "Reading metadata...: 1045it [00:00, 40457.91it/s]\n",
1087
+ "Reading metadata...: 623it [00:00, 27286.93it/s]\n",
1088
+ "Reading metadata...: 1045it [00:00, 33758.34it/s]\n",
1089
+ "Reading metadata...: 623it [00:00, 17969.24it/s]\n",
1090
+ "Reading metadata...: 1045it [00:00, 38347.21it/s]\n",
1091
+ "Reading metadata...: 623it [00:00, 13549.52it/s]\n",
1092
+ "Reading metadata...: 1045it [00:00, 36556.78it/s]\n",
1093
+ "Reading metadata...: 623it [00:00, 24387.99it/s]\n",
1094
+ "Reading metadata...: 1045it [00:00, 38171.88it/s]\n",
1095
+ "Reading metadata...: 623it [00:00, 18286.26it/s]\n",
1096
+ "Reading metadata...: 1045it [00:00, 14985.75it/s]\n",
1097
+ "Reading metadata...: 623it [00:00, 14077.95it/s]\n",
1098
+ "Reading metadata...: 1045it [00:00, 5754.51it/s]\n",
1099
+ "Reading metadata...: 623it [00:00, 5569.51it/s]\n",
1100
+ "Reading metadata...: 1045it [00:00, 13035.98it/s]\n",
1101
+ "Reading metadata...: 623it [00:00, 11232.79it/s]\n",
1102
+ "Reading metadata...: 1045it [00:00, 11289.33it/s]\n",
1103
+ "Reading metadata...: 623it [00:00, 19555.25it/s]\n",
1104
+ "Reading metadata...: 1045it [00:00, 9592.82it/s]\n",
1105
+ "Reading metadata...: 623it [00:00, 19744.54it/s]\n",
1106
+ "Reading metadata...: 1045it [00:00, 32887.25it/s]\n",
1107
+ "Reading metadata...: 623it [00:00, 22827.59it/s]\n",
1108
+ "Reading metadata...: 1045it [00:00, 4995.65it/s]\n",
1109
+ "Reading metadata...: 623it [00:00, 3313.49it/s]\n",
1110
+ "Reading metadata...: 1045it [00:00, 3087.78it/s]\n",
1111
+ "Reading metadata...: 623it [00:00, 2727.80it/s]\n",
1112
+ "Reading metadata...: 1045it [00:01, 665.59it/s]\n",
1113
+ "Reading metadata...: 623it [00:00, 1464.44it/s]\n",
1114
+ "Reading metadata...: 1045it [00:05, 190.45it/s]\n",
1115
+ "Reading metadata...: 623it [00:01, 542.05it/s]\n",
1116
+ "Reading metadata...: 1045it [00:02, 473.84it/s]\n",
1117
+ "Reading metadata...: 623it [00:00, 1116.31it/s]\n",
1118
+ "Reading metadata...: 1045it [00:00, 9172.01it/s]\n",
1119
+ "Reading metadata...: 623it [00:00, 9288.57it/s]\n",
1120
+ "Reading metadata...: 1045it [00:00, 7151.01it/s]\n",
1121
+ "Reading metadata...: 623it [00:00, 4058.46it/s]\n",
1122
+ "Reading metadata...: 1045it [00:00, 22239.72it/s]\n",
1123
+ "Reading metadata...: 623it [00:00, 17949.37it/s]\n",
1124
+ "Reading metadata...: 1045it [00:00, 24927.19it/s]\n",
1125
+ "Reading metadata...: 623it [00:00, 20978.93it/s]\n",
1126
+ "***** Running Evaluation *****\n",
1127
+ " Num examples: Unknown\n",
1128
+ " Batch size = 8\n",
1129
+ "Reading metadata...: 677it [00:00, 6344.98it/s]\n",
1130
  "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"
1131
  ]
1132
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0c102a56223c9d4c0073bf307faa42881b324cda17921f66be674baeaac81536
3
  size 967102601
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e6ff01a3ce8803bfef56cf5cca68c92577d50423a3d1edf2a66a1338a93cd7d
3
  size 967102601
runs/Dec15_14-18-11_129-213-18-196/events.out.tfevents.1671113929.129-213-18-196.4128231.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:79e3c278a91be84dc37fe3af83dfcdc461de029979a79195988c9f7a1dbaf16e
3
- size 10865
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c42467d082604e0422febf3b84ff084d42c13e1839cdcd4976014a09b2a401e8
3
+ size 17463