Hugging Face's logo

vitouphy
/
wav2vec2-xls-r-300m-khmer

Automatic Speech Recognition
Transformers
PyTorch
Safetensors
Khmer
wav2vec2
openslr
robust-speech-event
Generated from Trainer
hf-asr-leaderboard
Eval Results
Inference Endpoints
Model card Files Community
2
vitouphy commited on
Commit
2a395ac
•
1 Parent(s): e256a57

update readme for 2 step trainings

Browse files
Files changed (7) hide show
  1. README.md +14 -3
  2. README_inital_step.md +76 -0
  3. config.json +1 -1
  4. inference.ipynb +22 -48
  5. pytorch_model.bin +1 -1
  6. train_kh.ipynb +184 -550
  7. training_args.bin +1 -1
README.md CHANGED
@@ -20,8 +20,8 @@ should probably proofread and complete it, then remove this comment. -->
20
 
21
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the openslr dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.4638
24
- - Wer: 0.4944
25
 
26
  ## Model description
27
 
@@ -48,7 +48,7 @@ The following hyperparameters were used during training:
48
  - total_train_batch_size: 32
49
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
  - lr_scheduler_type: linear
51
- - lr_scheduler_warmup_steps: 1000
52
  - num_epochs: 50
53
  - mixed_precision_training: Native AMP
54
 
@@ -66,6 +66,17 @@ The following hyperparameters were used during training:
66
  | 1.4696 | 39.5 | 3200 | 0.5002 | 0.5130 |
67
  | 1.4175 | 44.44 | 3600 | 0.4752 | 0.5021 |
68
  | 1.3943 | 49.38 | 4000 | 0.4638 | 0.4944 |
 
 
 
 
 
 
 
 
 
 
 
69
 
70
 
71
  ### Framework versions
 
20
 
21
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the openslr dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.3142
24
+ - Wer: 0.3512
25
 
26
  ## Model description
27
 
 
48
  - total_train_batch_size: 32
49
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
  - lr_scheduler_type: linear
51
+ - lr_scheduler_warmup_steps: 100
52
  - num_epochs: 50
53
  - mixed_precision_training: Native AMP
54
 
 
66
  | 1.4696 | 39.5 | 3200 | 0.5002 | 0.5130 |
67
  | 1.4175 | 44.44 | 3600 | 0.4752 | 0.5021 |
68
  | 1.3943 | 49.38 | 4000 | 0.4638 | 0.4944 |
69
+ | Pause and Resume | | | | |
70
+ | 1.3829 | 4.93 | 400 | 0.4290 | 0.4796 |
71
+ | 1.3156 | 9.87 | 800 | 0.3856 | 0.4474 |
72
+ | 1.2396 | 14.81 | 1200 | 0.3600 | 0.4307 |
73
+ | 1.1444 | 19.75 | 1600 | 0.3423 | 0.4179 |
74
+ | 1.0979 | 24.69 | 2000 | 0.3370 | 0.3884 |
75
+ | 1.0714 | 29.62 | 2400 | 0.3237 | 0.3710 |
76
+ | 1.0442 | 34.56 | 2800 | 0.3336 | 0.3683 |
77
+ | 1.0492 | 39.5 | 3200 | 0.3166 | 0.3527 |
78
+ | 1.0284 | 44.44 | 3600 | 0.3178 | 0.3566 |
79
+ | 1.0302 | 49.38 | 4000 | 0.3142 | 0.3512 |
80
 
81
 
82
  ### Framework versions
README_inital_step.md ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - km
4
+ license: apache-2.0
5
+ tags:
6
+ - automatic-speech-recognition
7
+ - openslr
8
+ - robust-speech-event
9
+ - km
10
+ - generated_from_trainer
11
+ model-index:
12
+ - name: ''
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ #
20
+
21
+ This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the openslr dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 0.4638
24
+ - Wer: 0.4944
25
+
26
+ ## Model description
27
+
28
+ More information needed
29
+
30
+ ## Intended uses & limitations
31
+
32
+ More information needed
33
+
34
+ ## Training and evaluation data
35
+
36
+ More information needed
37
+
38
+ ## Training procedure
39
+
40
+ ### Training hyperparameters
41
+
42
+ The following hyperparameters were used during training:
43
+ - learning_rate: 5e-05
44
+ - train_batch_size: 8
45
+ - eval_batch_size: 8
46
+ - seed: 42
47
+ - gradient_accumulation_steps: 4
48
+ - total_train_batch_size: 32
49
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
+ - lr_scheduler_type: linear
51
+ - lr_scheduler_warmup_steps: 1000
52
+ - num_epochs: 50
53
+ - mixed_precision_training: Native AMP
54
+
55
+ ### Training results
56
+
57
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
58
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
59
+ | 5.2049 | 4.93 | 400 | 4.5570 | 1.0 |
60
+ | 3.569 | 9.87 | 800 | 3.5415 | 1.0 |
61
+ | 3.483 | 14.81 | 1200 | 3.3956 | 1.0 |
62
+ | 2.1906 | 19.75 | 1600 | 1.1732 | 0.7897 |
63
+ | 1.7968 | 24.69 | 2000 | 0.7634 | 0.6678 |
64
+ | 1.615 | 29.62 | 2400 | 0.6182 | 0.5922 |
65
+ | 1.52 | 34.56 | 2800 | 0.5473 | 0.5479 |
66
+ | 1.4696 | 39.5 | 3200 | 0.5002 | 0.5130 |
67
+ | 1.4175 | 44.44 | 3600 | 0.4752 | 0.5021 |
68
+ | 1.3943 | 49.38 | 4000 | 0.4638 | 0.4944 |
69
+
70
+
71
+ ### Framework versions
72
+
73
+ - Transformers 4.17.0.dev0
74
+ - Pytorch 1.10.2+cu102
75
+ - Datasets 1.18.2.dev0
76
+ - Tokenizers 0.11.0
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "facebook/wav2vec2-xls-r-300m",
3
  "activation_dropout": 0.0,
4
  "adapter_kernel_size": 3,
5
  "adapter_stride": 2,
 
1
  {
2
+ "_name_or_path": "checkpoint-4000",
3
  "activation_dropout": 0.0,
4
  "adapter_kernel_size": 3,
5
  "adapter_stride": 2,
inference.ipynb CHANGED
@@ -3,7 +3,7 @@
3
  {
4
  "cell_type": "code",
5
  "execution_count": 1,
6
- "id": "438927ca",
7
  "metadata": {},
8
  "outputs": [],
9
  "source": [
@@ -16,46 +16,20 @@
16
  {
17
  "cell_type": "code",
18
  "execution_count": 5,
19
- "id": "27a57965",
20
  "metadata": {},
21
  "outputs": [],
22
  "source": [
23
- "model = AutoModelForCTC.from_pretrained(\".\").to('cuda')\n",
24
- "processor = Wav2Vec2Processor.from_pretrained(\".\")"
25
  ]
26
  },
27
  {
28
  "cell_type": "code",
29
- "execution_count": 3,
30
- "id": "1d4324df",
31
- "metadata": {
32
- "collapsed": true,
33
- "jupyter": {
34
- "outputs_hidden": true
35
- }
36
- },
37
- "outputs": [
38
- {
39
- "ename": "JSONDecodeError",
40
- "evalue": "Expecting value: line 1 column 1 (char 0)",
41
- "output_type": "error",
42
- "traceback": [
43
- "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
44
- "\u001b[0;31mJSONDecodeError\u001b[0m Traceback (most recent call last)",
45
- "Input \u001b[0;32mIn [3]\u001b[0m, in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 1\u001b[0m model \u001b[38;5;241m=\u001b[39m AutoModelForCTC\u001b[38;5;241m.\u001b[39mfrom_pretrained(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mvitouphy/xls-r-300m-km\u001b[39m\u001b[38;5;124m\"\u001b[39m)\u001b[38;5;241m.\u001b[39mto(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcuda\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[0;32m----> 2\u001b[0m processor \u001b[38;5;241m=\u001b[39m \u001b[43mWav2Vec2Processor\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mvitouphy/xls-r-300m-km\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n",
46
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/models/wav2vec2/processing_wav2vec2.py:117\u001b[0m, in \u001b[0;36mWav2Vec2Processor.from_pretrained\u001b[0;34m(cls, pretrained_model_name_or_path, **kwargs)\u001b[0m\n\u001b[1;32m 112\u001b[0m \u001b[38;5;66;03m# load generic `AutoTokenizer`\u001b[39;00m\n\u001b[1;32m 113\u001b[0m \u001b[38;5;66;03m# need fallback here for backward compatibility in case processor is\u001b[39;00m\n\u001b[1;32m 114\u001b[0m \u001b[38;5;66;03m# loaded from just a tokenizer file that does not have a `tokenizer_class` attribute\u001b[39;00m\n\u001b[1;32m 115\u001b[0m \u001b[38;5;66;03m# behavior should be deprecated in major future release\u001b[39;00m\n\u001b[1;32m 116\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m--> 117\u001b[0m tokenizer \u001b[38;5;241m=\u001b[39m \u001b[43mAutoTokenizer\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43mpretrained_model_name_or_path\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 118\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m:\n\u001b[1;32m 119\u001b[0m warnings\u001b[38;5;241m.\u001b[39mwarn(\n\u001b[1;32m 120\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mLoading a tokenizer inside \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m from a config that does not\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 121\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m include a `tokenizer_class` attribute is deprecated and will be \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 125\u001b[0m \u001b[38;5;167;01mFutureWarning\u001b[39;00m,\n\u001b[1;32m 126\u001b[0m )\n",
47
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/models/auto/tokenization_auto.py:514\u001b[0m, in \u001b[0;36mAutoTokenizer.from_pretrained\u001b[0;34m(cls, pretrained_model_name_or_path, *inputs, **kwargs)\u001b[0m\n\u001b[1;32m 510\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m tokenizer_class \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m 511\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\n\u001b[1;32m 512\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mTokenizer class \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mtokenizer_class_candidate\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m does not exist or is not currently imported.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 513\u001b[0m )\n\u001b[0;32m--> 514\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mtokenizer_class\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[43mpretrained_model_name_or_path\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 516\u001b[0m \u001b[38;5;66;03m# Otherwise we have to be creative.\u001b[39;00m\n\u001b[1;32m 517\u001b[0m \u001b[38;5;66;03m# if model is an encoder decoder, the encoder tokenizer class is used by default\u001b[39;00m\n\u001b[1;32m 518\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(config, EncoderDecoderConfig):\n",
48
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/tokenization_utils_base.py:1773\u001b[0m, in \u001b[0;36mPreTrainedTokenizerBase.from_pretrained\u001b[0;34m(cls, pretrained_model_name_or_path, *init_inputs, **kwargs)\u001b[0m\n\u001b[1;32m 1770\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m 1771\u001b[0m logger\u001b[38;5;241m.\u001b[39minfo(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mloading file \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mfile_path\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m from cache at \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mresolved_vocab_files[file_id]\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m-> 1773\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_from_pretrained\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 1774\u001b[0m \u001b[43m \u001b[49m\u001b[43mresolved_vocab_files\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1775\u001b[0m \u001b[43m \u001b[49m\u001b[43mpretrained_model_name_or_path\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1776\u001b[0m \u001b[43m \u001b[49m\u001b[43minit_configuration\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1777\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minit_inputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1778\u001b[0m \u001b[43m \u001b[49m\u001b[43muse_auth_token\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43muse_auth_token\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1779\u001b[0m \u001b[43m \u001b[49m\u001b[43mcache_dir\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcache_dir\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1780\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1781\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
49
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/tokenization_utils_base.py:1908\u001b[0m, in \u001b[0;36mPreTrainedTokenizerBase._from_pretrained\u001b[0;34m(cls, resolved_vocab_files, pretrained_model_name_or_path, init_configuration, use_auth_token, cache_dir, *init_inputs, **kwargs)\u001b[0m\n\u001b[1;32m 1906\u001b[0m \u001b[38;5;66;03m# Instantiate tokenizer.\u001b[39;00m\n\u001b[1;32m 1907\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m-> 1908\u001b[0m tokenizer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minit_inputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minit_kwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1909\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m:\n\u001b[1;32m 1910\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m(\n\u001b[1;32m 1911\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mUnable to load vocabulary from file. \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 1912\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mPlease check that the provided vocabulary is accessible and not corrupted.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 1913\u001b[0m )\n",
50
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/models/wav2vec2/tokenization_wav2vec2.py:142\u001b[0m, in \u001b[0;36mWav2Vec2CTCTokenizer.__init__\u001b[0;34m(self, vocab_file, bos_token, eos_token, unk_token, pad_token, word_delimiter_token, do_lower_case, **kwargs)\u001b[0m\n\u001b[1;32m 139\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdo_lower_case \u001b[38;5;241m=\u001b[39m do_lower_case\n\u001b[1;32m 141\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[38;5;28mopen\u001b[39m(vocab_file, encoding\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mutf-8\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;28;01mas\u001b[39;00m vocab_handle:\n\u001b[0;32m--> 142\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder \u001b[38;5;241m=\u001b[39m \u001b[43mjson\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload\u001b[49m\u001b[43m(\u001b[49m\u001b[43mvocab_handle\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 143\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdecoder \u001b[38;5;241m=\u001b[39m {v: k \u001b[38;5;28;01mfor\u001b[39;00m k, v \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder\u001b[38;5;241m.\u001b[39mitems()}\n\u001b[1;32m 145\u001b[0m \u001b[38;5;66;03m# make sure that tokens made of several\u001b[39;00m\n\u001b[1;32m 146\u001b[0m \u001b[38;5;66;03m# characters are not split at tokenization\u001b[39;00m\n",
51
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/__init__.py:293\u001b[0m, in \u001b[0;36mload\u001b[0;34m(fp, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)\u001b[0m\n\u001b[1;32m 274\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload\u001b[39m(fp, \u001b[38;5;241m*\u001b[39m, \u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, object_hook\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, parse_float\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m 275\u001b[0m parse_int\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, parse_constant\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, object_pairs_hook\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkw):\n\u001b[1;32m 276\u001b[0m \u001b[38;5;124;03m\"\"\"Deserialize ``fp`` (a ``.read()``-supporting file-like object containing\u001b[39;00m\n\u001b[1;32m 277\u001b[0m \u001b[38;5;124;03m a JSON document) to a Python object.\u001b[39;00m\n\u001b[1;32m 278\u001b[0m \n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 291\u001b[0m \u001b[38;5;124;03m kwarg; otherwise ``JSONDecoder`` is used.\u001b[39;00m\n\u001b[1;32m 292\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n\u001b[0;32m--> 293\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mloads\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfp\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mread\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 294\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mcls\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mobject_hook\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mobject_hook\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 295\u001b[0m \u001b[43m \u001b[49m\u001b[43mparse_float\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparse_float\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mparse_int\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparse_int\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 296\u001b[0m \u001b[43m \u001b[49m\u001b[43mparse_constant\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparse_constant\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mobject_pairs_hook\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mobject_pairs_hook\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkw\u001b[49m\u001b[43m)\u001b[49m\n",
52
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/__init__.py:357\u001b[0m, in \u001b[0;36mloads\u001b[0;34m(s, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)\u001b[0m\n\u001b[1;32m 352\u001b[0m \u001b[38;5;28;01mdel\u001b[39;00m kw[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mencoding\u001b[39m\u001b[38;5;124m'\u001b[39m]\n\u001b[1;32m 354\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (\u001b[38;5;28mcls\u001b[39m \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m object_hook \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m\n\u001b[1;32m 355\u001b[0m parse_int \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m parse_float \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m\n\u001b[1;32m 356\u001b[0m parse_constant \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m object_pairs_hook \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m kw):\n\u001b[0;32m--> 357\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43m_default_decoder\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdecode\u001b[49m\u001b[43m(\u001b[49m\u001b[43ms\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 358\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mcls\u001b[39m \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m 359\u001b[0m \u001b[38;5;28mcls\u001b[39m \u001b[38;5;241m=\u001b[39m JSONDecoder\n",
53
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/decoder.py:337\u001b[0m, in \u001b[0;36mJSONDecoder.decode\u001b[0;34m(self, s, _w)\u001b[0m\n\u001b[1;32m 332\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdecode\u001b[39m(\u001b[38;5;28mself\u001b[39m, s, _w\u001b[38;5;241m=\u001b[39mWHITESPACE\u001b[38;5;241m.\u001b[39mmatch):\n\u001b[1;32m 333\u001b[0m \u001b[38;5;124;03m\"\"\"Return the Python representation of ``s`` (a ``str`` instance\u001b[39;00m\n\u001b[1;32m 334\u001b[0m \u001b[38;5;124;03m containing a JSON document).\u001b[39;00m\n\u001b[1;32m 335\u001b[0m \n\u001b[1;32m 336\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n\u001b[0;32m--> 337\u001b[0m obj, end \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mraw_decode\u001b[49m\u001b[43m(\u001b[49m\u001b[43ms\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43midx\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m_w\u001b[49m\u001b[43m(\u001b[49m\u001b[43ms\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mend\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 338\u001b[0m end \u001b[38;5;241m=\u001b[39m _w(s, end)\u001b[38;5;241m.\u001b[39mend()\n\u001b[1;32m 339\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m end \u001b[38;5;241m!=\u001b[39m \u001b[38;5;28mlen\u001b[39m(s):\n",
54
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/decoder.py:355\u001b[0m, in \u001b[0;36mJSONDecoder.raw_decode\u001b[0;34m(self, s, idx)\u001b[0m\n\u001b[1;32m 353\u001b[0m obj, end \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscan_once(s, idx)\n\u001b[1;32m 354\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mStopIteration\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m err:\n\u001b[0;32m--> 355\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m JSONDecodeError(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mExpecting value\u001b[39m\u001b[38;5;124m\"\u001b[39m, s, err\u001b[38;5;241m.\u001b[39mvalue) \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;28mNone\u001b[39m\n\u001b[1;32m 356\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m obj, end\n",
55
- "\u001b[0;31mJSONDecodeError\u001b[0m: Expecting value: line 1 column 1 (char 0)"
56
- ]
57
- }
58
- ],
59
  "source": [
60
  "model = AutoModelForCTC.from_pretrained(\"vitouphy/xls-r-300m-km\").to('cuda')\n",
61
  "processor = Wav2Vec2Processor.from_pretrained(\"vitouphy/xls-r-300m-km\")"
@@ -64,7 +38,7 @@
64
  {
65
  "cell_type": "code",
66
  "execution_count": 8,
67
- "id": "3d61ff3b",
68
  "metadata": {},
69
  "outputs": [
70
  {
@@ -83,7 +57,7 @@
83
  {
84
  "cell_type": "code",
85
  "execution_count": 9,
86
- "id": "a03f3af4",
87
  "metadata": {},
88
  "outputs": [],
89
  "source": [
@@ -95,7 +69,7 @@
95
  {
96
  "cell_type": "code",
97
  "execution_count": 10,
98
- "id": "9c88048b",
99
  "metadata": {},
100
  "outputs": [],
101
  "source": [
@@ -105,7 +79,7 @@
105
  {
106
  "cell_type": "code",
107
  "execution_count": 11,
108
- "id": "f3bfc930",
109
  "metadata": {},
110
  "outputs": [
111
  {
@@ -130,7 +104,7 @@
130
  {
131
  "cell_type": "code",
132
  "execution_count": 12,
133
- "id": "122a898b",
134
  "metadata": {},
135
  "outputs": [],
136
  "source": [
@@ -149,7 +123,7 @@
149
  {
150
  "cell_type": "code",
151
  "execution_count": 13,
152
- "id": "153e7f45",
153
  "metadata": {},
154
  "outputs": [
155
  {
@@ -173,8 +147,8 @@
173
  },
174
  {
175
  "cell_type": "code",
176
- "execution_count": 17,
177
- "id": "8947d307",
178
  "metadata": {},
179
  "outputs": [],
180
  "source": [
@@ -183,8 +157,8 @@
183
  },
184
  {
185
  "cell_type": "code",
186
- "execution_count": 18,
187
- "id": "3d6b46ca",
188
  "metadata": {},
189
  "outputs": [
190
  {
@@ -203,8 +177,8 @@
203
  },
204
  {
205
  "cell_type": "code",
206
- "execution_count": 19,
207
- "id": "d1550ddc",
208
  "metadata": {},
209
  "outputs": [
210
  {
@@ -232,7 +206,7 @@
232
  {
233
  "cell_type": "code",
234
  "execution_count": null,
235
- "id": "5bbf1c82",
236
  "metadata": {},
237
  "outputs": [],
238
  "source": []
@@ -240,7 +214,7 @@
240
  {
241
  "cell_type": "code",
242
  "execution_count": null,
243
- "id": "71b6f502",
244
  "metadata": {},
245
  "outputs": [],
246
  "source": []
 
3
  {
4
  "cell_type": "code",
5
  "execution_count": 1,
6
+ "id": "310fea8f",
7
  "metadata": {},
8
  "outputs": [],
9
  "source": [
 
16
  {
17
  "cell_type": "code",
18
  "execution_count": 5,
19
+ "id": "555c8316",
20
  "metadata": {},
21
  "outputs": [],
22
  "source": [
23
+ "# model = AutoModelForCTC.from_pretrained(\".\").to('cuda')\n",
24
+ "# processor = Wav2Vec2Processor.from_pretrained(\".\")"
25
  ]
26
  },
27
  {
28
  "cell_type": "code",
29
+ "execution_count": 20,
30
+ "id": "24cc91e8",
31
+ "metadata": {},
32
+ "outputs": [],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  "source": [
34
  "model = AutoModelForCTC.from_pretrained(\"vitouphy/xls-r-300m-km\").to('cuda')\n",
35
  "processor = Wav2Vec2Processor.from_pretrained(\"vitouphy/xls-r-300m-km\")"
 
38
  {
39
  "cell_type": "code",
40
  "execution_count": 8,
41
+ "id": "69d79b00",
42
  "metadata": {},
43
  "outputs": [
44
  {
 
57
  {
58
  "cell_type": "code",
59
  "execution_count": 9,
60
+ "id": "9c9a59b3",
61
  "metadata": {},
62
  "outputs": [],
63
  "source": [
 
69
  {
70
  "cell_type": "code",
71
  "execution_count": 10,
72
+ "id": "868afb48",
73
  "metadata": {},
74
  "outputs": [],
75
  "source": [
 
79
  {
80
  "cell_type": "code",
81
  "execution_count": 11,
82
+ "id": "f93e7f2a",
83
  "metadata": {},
84
  "outputs": [
85
  {
 
104
  {
105
  "cell_type": "code",
106
  "execution_count": 12,
107
+ "id": "c97bf6c8",
108
  "metadata": {},
109
  "outputs": [],
110
  "source": [
 
123
  {
124
  "cell_type": "code",
125
  "execution_count": 13,
126
+ "id": "8e6b77e3",
127
  "metadata": {},
128
  "outputs": [
129
  {
 
147
  },
148
  {
149
  "cell_type": "code",
150
+ "execution_count": 21,
151
+ "id": "53b5be56",
152
  "metadata": {},
153
  "outputs": [],
154
  "source": [
 
157
  },
158
  {
159
  "cell_type": "code",
160
+ "execution_count": 22,
161
+ "id": "15dda9d3",
162
  "metadata": {},
163
  "outputs": [
164
  {
 
177
  },
178
  {
179
  "cell_type": "code",
180
+ "execution_count": 23,
181
+ "id": "bc40d9dc",
182
  "metadata": {},
183
  "outputs": [
184
  {
 
206
  {
207
  "cell_type": "code",
208
  "execution_count": null,
209
+ "id": "f755f572",
210
  "metadata": {},
211
  "outputs": [],
212
  "source": []
 
214
  {
215
  "cell_type": "code",
216
  "execution_count": null,
217
+ "id": "16aa56dc",
218
  "metadata": {},
219
  "outputs": [],
220
  "source": []
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d0ba956e9839aa07016519fdfa526ee53a8b6a22d5f2e4b17268e9859f128846
3
  size 1262231153
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b254f3cb4af80138f33d06456c61d7e3730f18b51646fbede34871daeafcc7e
3
  size 1262231153
train_kh.ipynb CHANGED
@@ -3,7 +3,7 @@
3
  {
4
  "cell_type": "code",
5
  "execution_count": 1,
6
- "id": "1bf32ef8",
7
  "metadata": {},
8
  "outputs": [],
9
  "source": [
@@ -16,7 +16,7 @@
16
  {
17
  "cell_type": "code",
18
  "execution_count": null,
19
- "id": "d2deec6c",
20
  "metadata": {
21
  "collapsed": true,
22
  "jupyter": {
@@ -19167,7 +19167,7 @@
19167
  },
19168
  {
19169
  "cell_type": "markdown",
19170
- "id": "6fe38e7a",
19171
  "metadata": {},
19172
  "source": [
19173
  "### Load KH Data"
@@ -19176,7 +19176,7 @@
19176
  {
19177
  "cell_type": "code",
19178
  "execution_count": 4,
19179
- "id": "b75f1fec",
19180
  "metadata": {},
19181
  "outputs": [],
19182
  "source": [
@@ -19197,7 +19197,7 @@
19197
  {
19198
  "cell_type": "code",
19199
  "execution_count": 5,
19200
- "id": "433fe749",
19201
  "metadata": {},
19202
  "outputs": [
19203
  {
@@ -19307,7 +19307,7 @@
19307
  {
19308
  "cell_type": "code",
19309
  "execution_count": 6,
19310
- "id": "c6d633ad",
19311
  "metadata": {},
19312
  "outputs": [],
19313
  "source": [
@@ -19321,7 +19321,7 @@
19321
  },
19322
  {
19323
  "cell_type": "markdown",
19324
- "id": "acb914d0",
19325
  "metadata": {},
19326
  "source": [
19327
  "### Clean Up the Text"
@@ -19330,7 +19330,7 @@
19330
  {
19331
  "cell_type": "code",
19332
  "execution_count": 6,
19333
- "id": "bc3a017b",
19334
  "metadata": {},
19335
  "outputs": [],
19336
  "source": [
@@ -19346,7 +19346,7 @@
19346
  {
19347
  "cell_type": "code",
19348
  "execution_count": 7,
19349
- "id": "4a7b6a10",
19350
  "metadata": {
19351
  "collapsed": true,
19352
  "jupyter": {
@@ -19402,7 +19402,7 @@
19402
  {
19403
  "cell_type": "code",
19404
  "execution_count": 7,
19405
- "id": "7f511e3f",
19406
  "metadata": {},
19407
  "outputs": [
19408
  {
@@ -19423,7 +19423,7 @@
19423
  },
19424
  {
19425
  "cell_type": "markdown",
19426
- "id": "205a6e23",
19427
  "metadata": {},
19428
  "source": [
19429
  "### Build Character"
@@ -19432,7 +19432,7 @@
19432
  {
19433
  "cell_type": "code",
19434
  "execution_count": 8,
19435
- "id": "48a97fac",
19436
  "metadata": {},
19437
  "outputs": [
19438
  {
@@ -19480,7 +19480,7 @@
19480
  {
19481
  "cell_type": "code",
19482
  "execution_count": 9,
19483
- "id": "9b4ac5f7",
19484
  "metadata": {},
19485
  "outputs": [],
19486
  "source": [
@@ -19491,7 +19491,7 @@
19491
  {
19492
  "cell_type": "code",
19493
  "execution_count": 10,
19494
- "id": "a9a07875",
19495
  "metadata": {},
19496
  "outputs": [
19497
  {
@@ -19509,7 +19509,7 @@
19509
  {
19510
  "cell_type": "code",
19511
  "execution_count": 11,
19512
- "id": "8a3d39d8",
19513
  "metadata": {},
19514
  "outputs": [
19515
  {
@@ -19536,7 +19536,7 @@
19536
  {
19537
  "cell_type": "code",
19538
  "execution_count": 12,
19539
- "id": "934a4070",
19540
  "metadata": {},
19541
  "outputs": [
19542
  {
@@ -19554,7 +19554,7 @@
19554
  {
19555
  "cell_type": "code",
19556
  "execution_count": 13,
19557
- "id": "7f42a2b4",
19558
  "metadata": {},
19559
  "outputs": [],
19560
  "source": [
@@ -19565,7 +19565,7 @@
19565
  },
19566
  {
19567
  "cell_type": "markdown",
19568
- "id": "9a504bc4",
19569
  "metadata": {},
19570
  "source": [
19571
  "# Tokenizer"
@@ -19574,7 +19574,7 @@
19574
  {
19575
  "cell_type": "code",
19576
  "execution_count": 14,
19577
- "id": "0cec90b4",
19578
  "metadata": {},
19579
  "outputs": [],
19580
  "source": [
@@ -19585,8 +19585,8 @@
19585
  },
19586
  {
19587
  "cell_type": "code",
19588
- "execution_count": 62,
19589
- "id": "dc9e79da",
19590
  "metadata": {},
19591
  "outputs": [
19592
  {
@@ -19598,25 +19598,9 @@
19598
  "loading file ./tokenizer_config.json\n",
19599
  "loading file ./added_tokens.json\n",
19600
  "loading file ./special_tokens_map.json\n",
19601
- "loading file None\n"
19602
- ]
19603
- },
19604
- {
19605
- "ename": "JSONDecodeError",
19606
- "evalue": "Expecting value: line 1 column 1 (char 0)",
19607
- "output_type": "error",
19608
- "traceback": [
19609
- "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
19610
- "\u001b[0;31mJSONDecodeError\u001b[0m Traceback (most recent call last)",
19611
- "Input \u001b[0;32mIn [62]\u001b[0m, in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0m tokenizer \u001b[38;5;241m=\u001b[39m \u001b[43mWav2Vec2CTCTokenizer\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfrom_pretrained\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m./\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43munk_token\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m[UNK]\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mpad_token\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m[PAD]\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mword_delimiter_token\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43m|\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m \u001b[38;5;66;03m# './' load vocab.json in the current directory\u001b[39;00m\n\u001b[1;32m 2\u001b[0m feature_extractor \u001b[38;5;241m=\u001b[39m Wav2Vec2FeatureExtractor(feature_size\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m, sampling_rate\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m16000\u001b[39m, padding_value\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0.0\u001b[39m, do_normalize\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m, return_attention_mask\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m) \n\u001b[1;32m 3\u001b[0m processor \u001b[38;5;241m=\u001b[39m Wav2Vec2Processor(feature_extractor\u001b[38;5;241m=\u001b[39mfeature_extractor, tokenizer\u001b[38;5;241m=\u001b[39mtokenizer)\n",
19612
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/tokenization_utils_base.py:1773\u001b[0m, in \u001b[0;36mPreTrainedTokenizerBase.from_pretrained\u001b[0;34m(cls, pretrained_model_name_or_path, *init_inputs, **kwargs)\u001b[0m\n\u001b[1;32m 1770\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m 1771\u001b[0m logger\u001b[38;5;241m.\u001b[39minfo(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mloading file \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mfile_path\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m from cache at \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mresolved_vocab_files[file_id]\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m-> 1773\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_from_pretrained\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m 1774\u001b[0m \u001b[43m \u001b[49m\u001b[43mresolved_vocab_files\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1775\u001b[0m \u001b[43m \u001b[49m\u001b[43mpretrained_model_name_or_path\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1776\u001b[0m \u001b[43m \u001b[49m\u001b[43minit_configuration\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1777\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minit_inputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1778\u001b[0m \u001b[43m \u001b[49m\u001b[43muse_auth_token\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43muse_auth_token\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1779\u001b[0m \u001b[43m \u001b[49m\u001b[43mcache_dir\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcache_dir\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1780\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 1781\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
19613
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/tokenization_utils_base.py:1908\u001b[0m, in \u001b[0;36mPreTrainedTokenizerBase._from_pretrained\u001b[0;34m(cls, resolved_vocab_files, pretrained_model_name_or_path, init_configuration, use_auth_token, cache_dir, *init_inputs, **kwargs)\u001b[0m\n\u001b[1;32m 1906\u001b[0m \u001b[38;5;66;03m# Instantiate tokenizer.\u001b[39;00m\n\u001b[1;32m 1907\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m-> 1908\u001b[0m tokenizer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minit_inputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43minit_kwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1909\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m:\n\u001b[1;32m 1910\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mOSError\u001b[39;00m(\n\u001b[1;32m 1911\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mUnable to load vocabulary from file. \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 1912\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mPlease check that the provided vocabulary is accessible and not corrupted.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 1913\u001b[0m )\n",
19614
- "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/transformers/models/wav2vec2/tokenization_wav2vec2.py:142\u001b[0m, in \u001b[0;36mWav2Vec2CTCTokenizer.__init__\u001b[0;34m(self, vocab_file, bos_token, eos_token, unk_token, pad_token, word_delimiter_token, do_lower_case, **kwargs)\u001b[0m\n\u001b[1;32m 139\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdo_lower_case \u001b[38;5;241m=\u001b[39m do_lower_case\n\u001b[1;32m 141\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[38;5;28mopen\u001b[39m(vocab_file, encoding\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mutf-8\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;28;01mas\u001b[39;00m vocab_handle:\n\u001b[0;32m--> 142\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder \u001b[38;5;241m=\u001b[39m \u001b[43mjson\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload\u001b[49m\u001b[43m(\u001b[49m\u001b[43mvocab_handle\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 143\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdecoder \u001b[38;5;241m=\u001b[39m {v: k \u001b[38;5;28;01mfor\u001b[39;00m k, v \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder\u001b[38;5;241m.\u001b[39mitems()}\n\u001b[1;32m 145\u001b[0m \u001b[38;5;66;03m# make sure that tokens made of several\u001b[39;00m\n\u001b[1;32m 146\u001b[0m \u001b[38;5;66;03m# characters are not split at tokenization\u001b[39;00m\n",
19615
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/__init__.py:293\u001b[0m, in \u001b[0;36mload\u001b[0;34m(fp, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)\u001b[0m\n\u001b[1;32m 274\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mload\u001b[39m(fp, \u001b[38;5;241m*\u001b[39m, \u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, object_hook\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, parse_float\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m 275\u001b[0m parse_int\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, parse_constant\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, object_pairs_hook\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkw):\n\u001b[1;32m 276\u001b[0m \u001b[38;5;124;03m\"\"\"Deserialize ``fp`` (a ``.read()``-supporting file-like object containing\u001b[39;00m\n\u001b[1;32m 277\u001b[0m \u001b[38;5;124;03m a JSON document) to a Python object.\u001b[39;00m\n\u001b[1;32m 278\u001b[0m \n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 291\u001b[0m \u001b[38;5;124;03m kwarg; otherwise ``JSONDecoder`` is used.\u001b[39;00m\n\u001b[1;32m 292\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n\u001b[0;32m--> 293\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mloads\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfp\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mread\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 294\u001b[0m \u001b[43m \u001b[49m\u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mcls\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mobject_hook\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mobject_hook\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 295\u001b[0m \u001b[43m \u001b[49m\u001b[43mparse_float\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparse_float\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mparse_int\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparse_int\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 296\u001b[0m \u001b[43m \u001b[49m\u001b[43mparse_constant\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparse_constant\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mobject_pairs_hook\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mobject_pairs_hook\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkw\u001b[49m\u001b[43m)\u001b[49m\n",
19616
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/__init__.py:357\u001b[0m, in \u001b[0;36mloads\u001b[0;34m(s, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)\u001b[0m\n\u001b[1;32m 352\u001b[0m \u001b[38;5;28;01mdel\u001b[39;00m kw[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mencoding\u001b[39m\u001b[38;5;124m'\u001b[39m]\n\u001b[1;32m 354\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (\u001b[38;5;28mcls\u001b[39m \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m object_hook \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m\n\u001b[1;32m 355\u001b[0m parse_int \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m parse_float \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m\n\u001b[1;32m 356\u001b[0m parse_constant \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m object_pairs_hook \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m kw):\n\u001b[0;32m--> 357\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43m_default_decoder\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdecode\u001b[49m\u001b[43m(\u001b[49m\u001b[43ms\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 358\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mcls\u001b[39m \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m 359\u001b[0m \u001b[38;5;28mcls\u001b[39m \u001b[38;5;241m=\u001b[39m JSONDecoder\n",
19617
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/decoder.py:337\u001b[0m, in \u001b[0;36mJSONDecoder.decode\u001b[0;34m(self, s, _w)\u001b[0m\n\u001b[1;32m 332\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdecode\u001b[39m(\u001b[38;5;28mself\u001b[39m, s, _w\u001b[38;5;241m=\u001b[39mWHITESPACE\u001b[38;5;241m.\u001b[39mmatch):\n\u001b[1;32m 333\u001b[0m \u001b[38;5;124;03m\"\"\"Return the Python representation of ``s`` (a ``str`` instance\u001b[39;00m\n\u001b[1;32m 334\u001b[0m \u001b[38;5;124;03m containing a JSON document).\u001b[39;00m\n\u001b[1;32m 335\u001b[0m \n\u001b[1;32m 336\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n\u001b[0;32m--> 337\u001b[0m obj, end \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mraw_decode\u001b[49m\u001b[43m(\u001b[49m\u001b[43ms\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43midx\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m_w\u001b[49m\u001b[43m(\u001b[49m\u001b[43ms\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mend\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 338\u001b[0m end \u001b[38;5;241m=\u001b[39m _w(s, end)\u001b[38;5;241m.\u001b[39mend()\n\u001b[1;32m 339\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m end \u001b[38;5;241m!=\u001b[39m \u001b[38;5;28mlen\u001b[39m(s):\n",
19618
- "File \u001b[0;32m/opt/conda/lib/python3.8/json/decoder.py:355\u001b[0m, in \u001b[0;36mJSONDecoder.raw_decode\u001b[0;34m(self, s, idx)\u001b[0m\n\u001b[1;32m 353\u001b[0m obj, end \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscan_once(s, idx)\n\u001b[1;32m 354\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mStopIteration\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m err:\n\u001b[0;32m--> 355\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m JSONDecodeError(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mExpecting value\u001b[39m\u001b[38;5;124m\"\u001b[39m, s, err\u001b[38;5;241m.\u001b[39mvalue) \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;28mNone\u001b[39m\n\u001b[1;32m 356\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m obj, end\n",
19619
- "\u001b[0;31mJSONDecodeError\u001b[0m: Expecting value: line 1 column 1 (char 0)"
19620
  ]
19621
  }
19622
  ],
@@ -19629,7 +19613,7 @@
19629
  {
19630
  "cell_type": "code",
19631
  "execution_count": 26,
19632
- "id": "61738038",
19633
  "metadata": {},
19634
  "outputs": [],
19635
  "source": [
@@ -19646,7 +19630,7 @@
19646
  {
19647
  "cell_type": "code",
19648
  "execution_count": 27,
19649
- "id": "4b72b9b8",
19650
  "metadata": {},
19651
  "outputs": [
19652
  {
@@ -19686,7 +19670,7 @@
19686
  {
19687
  "cell_type": "code",
19688
  "execution_count": 17,
19689
- "id": "ccb4a36c",
19690
  "metadata": {},
19691
  "outputs": [],
19692
  "source": [
@@ -19697,7 +19681,7 @@
19697
  {
19698
  "cell_type": "code",
19699
  "execution_count": 18,
19700
- "id": "cf9d1391",
19701
  "metadata": {},
19702
  "outputs": [
19703
  {
@@ -19722,7 +19706,7 @@
19722
  {
19723
  "cell_type": "code",
19724
  "execution_count": 19,
19725
- "id": "57ea4c6f",
19726
  "metadata": {},
19727
  "outputs": [
19728
  {
@@ -19769,7 +19753,7 @@
19769
  {
19770
  "cell_type": "code",
19771
  "execution_count": 20,
19772
- "id": "7cb9fd2a",
19773
  "metadata": {},
19774
  "outputs": [],
19775
  "source": [
@@ -19791,7 +19775,7 @@
19791
  {
19792
  "cell_type": "code",
19793
  "execution_count": 22,
19794
- "id": "42f2952f",
19795
  "metadata": {},
19796
  "outputs": [],
19797
  "source": [
@@ -19802,7 +19786,7 @@
19802
  {
19803
  "cell_type": "code",
19804
  "execution_count": 41,
19805
- "id": "fe093630",
19806
  "metadata": {},
19807
  "outputs": [],
19808
  "source": [
@@ -19814,7 +19798,7 @@
19814
  {
19815
  "cell_type": "code",
19816
  "execution_count": 25,
19817
- "id": "a6efe782",
19818
  "metadata": {},
19819
  "outputs": [],
19820
  "source": [
@@ -19874,7 +19858,7 @@
19874
  {
19875
  "cell_type": "code",
19876
  "execution_count": 26,
19877
- "id": "e82a3663",
19878
  "metadata": {},
19879
  "outputs": [],
19880
  "source": [
@@ -19884,7 +19868,7 @@
19884
  {
19885
  "cell_type": "code",
19886
  "execution_count": 27,
19887
- "id": "1df03ab8",
19888
  "metadata": {},
19889
  "outputs": [],
19890
  "source": [
@@ -19894,8 +19878,8 @@
19894
  },
19895
  {
19896
  "cell_type": "code",
19897
- "execution_count": 44,
19898
- "id": "8304f047",
19899
  "metadata": {},
19900
  "outputs": [],
19901
  "source": [
@@ -19908,9 +19892,9 @@
19908
  " pred_str = tokenizer.batch_decode(pred_ids)\n",
19909
  " label_str = tokenizer.batch_decode(pred.label_ids, group_tokens=False)\n",
19910
  "\n",
19911
- " print(\"pred : \", pred_ids[0])\n",
19912
- " print(\"label: \", pred.label_ids[0])\n",
19913
- " print(\"-----------------\")\n",
19914
  " \n",
19915
  " wer = wer_metric.compute(predictions=pred_str, references=label_str)\n",
19916
  "\n",
@@ -19919,8 +19903,8 @@
19919
  },
19920
  {
19921
  "cell_type": "code",
19922
- "execution_count": 45,
19923
- "id": "f92c9b4d",
19924
  "metadata": {
19925
  "collapsed": true,
19926
  "jupyter": {
@@ -19932,15 +19916,16 @@
19932
  "name": "stderr",
19933
  "output_type": "stream",
19934
  "text": [
19935
- "loading configuration file https://huggingface.co/facebook/wav2vec2-xls-r-300m/resolve/main/config.json from cache at /workspace/.cache/huggingface/transformers/dabc27df63e37bd2a7a221c7774e35f36a280fbdf917cf54cadfc7df8c786f6f.a3e4c3c967d9985881e0ae550a5f6f668f897db5ab2e0802f9b97973b15970e6\n",
19936
  "Model config Wav2Vec2Config {\n",
 
19937
  " \"activation_dropout\": 0.0,\n",
19938
  " \"adapter_kernel_size\": 3,\n",
19939
  " \"adapter_stride\": 2,\n",
19940
  " \"add_adapter\": false,\n",
19941
  " \"apply_spec_augment\": true,\n",
19942
  " \"architectures\": [\n",
19943
- " \"Wav2Vec2ForPreTraining\"\n",
19944
  " ],\n",
19945
  " \"attention_dropout\": 0.1,\n",
19946
  " \"bos_token_id\": 1,\n",
@@ -20041,12 +20026,11 @@
20041
  " \"xvector_output_dim\": 512\n",
20042
  "}\n",
20043
  "\n",
20044
- "loading weights file https://huggingface.co/facebook/wav2vec2-xls-r-300m/resolve/main/pytorch_model.bin from cache at /workspace/.cache/huggingface/transformers/1e6a6507f3b689035cd4b247e2a37c154e27f39143f31357a49b4e38baeccc36.1edb32803799e27ed554eb7dd935f6745b1a0b17b0ea256442fe24db6eb546cd\n",
20045
- "Some weights of the model checkpoint at facebook/wav2vec2-xls-r-300m were not used when initializing Wav2Vec2ForCTC: ['project_q.bias', 'project_hid.weight', 'project_hid.bias', 'quantizer.weight_proj.bias', 'quantizer.weight_proj.weight', 'project_q.weight', 'quantizer.codevectors']\n",
20046
- "- This IS expected if you are initializing Wav2Vec2ForCTC from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
20047
- "- This IS NOT expected if you are initializing Wav2Vec2ForCTC from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
20048
- "Some weights of Wav2Vec2ForCTC were not initialized from the model checkpoint at facebook/wav2vec2-xls-r-300m and are newly initialized: ['lm_head.weight', 'lm_head.bias']\n",
20049
- "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
20050
  ]
20051
  }
20052
  ],
@@ -20054,7 +20038,8 @@
20054
  "from transformers import Wav2Vec2ForCTC\n",
20055
  "\n",
20056
  "model = Wav2Vec2ForCTC.from_pretrained(\n",
20057
- " \"facebook/wav2vec2-xls-r-300m\", \n",
 
20058
  " attention_dropout=0.1,\n",
20059
  " layerdrop=0.0,\n",
20060
  " feat_proj_dropout=0.0,\n",
@@ -20070,8 +20055,8 @@
20070
  },
20071
  {
20072
  "cell_type": "code",
20073
- "execution_count": 46,
20074
- "id": "7f2dd147",
20075
  "metadata": {},
20076
  "outputs": [],
20077
  "source": [
@@ -20080,8 +20065,8 @@
20080
  },
20081
  {
20082
  "cell_type": "code",
20083
- "execution_count": 47,
20084
- "id": "3d27466c",
20085
  "metadata": {},
20086
  "outputs": [
20087
  {
@@ -20109,7 +20094,7 @@
20109
  " eval_steps=400,\n",
20110
  " logging_steps=100,\n",
20111
  " learning_rate=5e-5,\n",
20112
- " warmup_steps=1000,\n",
20113
  " save_total_limit=3,\n",
20114
  " load_best_model_at_end=True\n",
20115
  ")"
@@ -20117,8 +20102,8 @@
20117
  },
20118
  {
20119
  "cell_type": "code",
20120
- "execution_count": 48,
20121
- "id": "014ac4c9",
20122
  "metadata": {},
20123
  "outputs": [
20124
  {
@@ -20145,14 +20130,9 @@
20145
  },
20146
  {
20147
  "cell_type": "code",
20148
- "execution_count": 49,
20149
- "id": "e6cb809a",
20150
- "metadata": {
20151
- "collapsed": true,
20152
- "jupyter": {
20153
- "outputs_hidden": true
20154
- }
20155
- },
20156
  "outputs": [
20157
  {
20158
  "name": "stderr",
@@ -20177,7 +20157,7 @@
20177
  " <div>\n",
20178
  " \n",
20179
  " <progress value='4050' max='4050' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
20180
- " [4050/4050 2:16:04, Epoch 49/50]\n",
20181
  " </div>\n",
20182
  " <table border=\"1\" class=\"dataframe\">\n",
20183
  " <thead>\n",
@@ -20191,63 +20171,63 @@
20191
  " <tbody>\n",
20192
  " <tr>\n",
20193
  " <td>400</td>\n",
20194
- " <td>5.204900</td>\n",
20195
- " <td>4.556981</td>\n",
20196
- " <td>1.000000</td>\n",
20197
  " </tr>\n",
20198
  " <tr>\n",
20199
  " <td>800</td>\n",
20200
- " <td>3.569000</td>\n",
20201
- " <td>3.541533</td>\n",
20202
- " <td>1.000000</td>\n",
20203
  " </tr>\n",
20204
  " <tr>\n",
20205
  " <td>1200</td>\n",
20206
- " <td>3.483000</td>\n",
20207
- " <td>3.395552</td>\n",
20208
- " <td>1.000000</td>\n",
20209
  " </tr>\n",
20210
  " <tr>\n",
20211
  " <td>1600</td>\n",
20212
- " <td>2.190600</td>\n",
20213
- " <td>1.173165</td>\n",
20214
- " <td>0.789678</td>\n",
20215
  " </tr>\n",
20216
  " <tr>\n",
20217
  " <td>2000</td>\n",
20218
- " <td>1.796800</td>\n",
20219
- " <td>0.763436</td>\n",
20220
- " <td>0.667831</td>\n",
20221
  " </tr>\n",
20222
  " <tr>\n",
20223
  " <td>2400</td>\n",
20224
- " <td>1.615000</td>\n",
20225
- " <td>0.618224</td>\n",
20226
- " <td>0.592161</td>\n",
20227
  " </tr>\n",
20228
  " <tr>\n",
20229
  " <td>2800</td>\n",
20230
- " <td>1.520000</td>\n",
20231
- " <td>0.547277</td>\n",
20232
- " <td>0.547924</td>\n",
20233
  " </tr>\n",
20234
  " <tr>\n",
20235
  " <td>3200</td>\n",
20236
- " <td>1.469600</td>\n",
20237
- " <td>0.500246</td>\n",
20238
- " <td>0.513000</td>\n",
20239
  " </tr>\n",
20240
  " <tr>\n",
20241
  " <td>3600</td>\n",
20242
- " <td>1.417500</td>\n",
20243
- " <td>0.475214</td>\n",
20244
- " <td>0.502134</td>\n",
20245
  " </tr>\n",
20246
  " <tr>\n",
20247
  " <td>4000</td>\n",
20248
- " <td>1.394300</td>\n",
20249
- " <td>0.463765</td>\n",
20250
- " <td>0.494373</td>\n",
20251
  " </tr>\n",
20252
  " </tbody>\n",
20253
  "</table><p>"
@@ -20266,196 +20246,35 @@
20266
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20267
  "***** Running Evaluation *****\n",
20268
  " Num examples = 291\n",
20269
- " Batch size = 8\n"
20270
- ]
20271
- },
20272
- {
20273
- "name": "stdout",
20274
- "output_type": "stream",
20275
- "text": [
20276
- "pred : [72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20277
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20278
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20279
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20280
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20281
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20282
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20283
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20284
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20285
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20286
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20287
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20288
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20289
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20290
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 0 0 0 0 0 0 0\n",
20291
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20292
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20293
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20294
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20295
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20296
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20297
- " 0 0 0 0 0 0 0]\n",
20298
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20299
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20300
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20301
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20302
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20303
- " 72 72 72 72 72 72 72]\n",
20304
- "-----------------\n"
20305
- ]
20306
- },
20307
- {
20308
- "name": "stderr",
20309
- "output_type": "stream",
20310
- "text": [
20311
  "Saving model checkpoint to ./checkpoint-400\n",
20312
  "Configuration saved in ./checkpoint-400/config.json\n",
20313
  "Model weights saved in ./checkpoint-400/pytorch_model.bin\n",
20314
  "Configuration saved in ./checkpoint-400/preprocessor_config.json\n",
 
 
20315
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20316
  "***** Running Evaluation *****\n",
20317
  " Num examples = 291\n",
20318
- " Batch size = 8\n"
20319
- ]
20320
- },
20321
- {
20322
- "name": "stdout",
20323
- "output_type": "stream",
20324
- "text": [
20325
- "pred : [72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20326
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20327
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20328
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20329
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20330
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20331
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20332
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20333
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20334
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20335
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20336
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20337
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20338
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20339
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 0 0 0 0 0 0 0\n",
20340
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20341
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20342
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20343
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20344
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20345
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20346
- " 0 0 0 0 0 0 0]\n",
20347
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20348
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20349
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20350
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20351
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20352
- " 72 72 72 72 72 72 72]\n",
20353
- "-----------------\n"
20354
- ]
20355
- },
20356
- {
20357
- "name": "stderr",
20358
- "output_type": "stream",
20359
- "text": [
20360
  "Saving model checkpoint to ./checkpoint-800\n",
20361
  "Configuration saved in ./checkpoint-800/config.json\n",
20362
  "Model weights saved in ./checkpoint-800/pytorch_model.bin\n",
20363
  "Configuration saved in ./checkpoint-800/preprocessor_config.json\n",
 
20364
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20365
  "***** Running Evaluation *****\n",
20366
  " Num examples = 291\n",
20367
- " Batch size = 8\n"
20368
- ]
20369
- },
20370
- {
20371
- "name": "stdout",
20372
- "output_type": "stream",
20373
- "text": [
20374
- "pred : [ 1 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20375
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20376
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20377
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20378
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20379
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20380
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20381
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20382
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20383
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20384
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20385
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20386
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20387
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20388
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 0 0 0 0 0 0 0\n",
20389
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20390
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20391
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20392
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20393
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20394
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20395
- " 0 0 0 0 0 0 0]\n",
20396
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20397
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20398
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20399
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20400
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20401
- " 72 72 72 72 72 72 72]\n",
20402
- "-----------------\n"
20403
- ]
20404
- },
20405
- {
20406
- "name": "stderr",
20407
- "output_type": "stream",
20408
- "text": [
20409
  "Saving model checkpoint to ./checkpoint-1200\n",
20410
  "Configuration saved in ./checkpoint-1200/config.json\n",
20411
  "Model weights saved in ./checkpoint-1200/pytorch_model.bin\n",
20412
  "Configuration saved in ./checkpoint-1200/preprocessor_config.json\n",
20413
- "Deleting older checkpoint [checkpoint-500] due to args.save_total_limit\n",
20414
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20415
  "***** Running Evaluation *****\n",
20416
  " Num examples = 291\n",
20417
- " Batch size = 8\n"
20418
- ]
20419
- },
20420
- {
20421
- "name": "stdout",
20422
- "output_type": "stream",
20423
- "text": [
20424
- "pred : [30 45 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20425
- " 72 72 72 72 72 72 72 72 72 11 43 72 72 72 72 6 26 0 72 25 72 72 72 72\n",
20426
- " 72 72 72 18 49 72 72 72 72 72 0 72 20 58 72 72 0 0 72 72 72 23 54 72\n",
20427
- " 72 72 0 72 11 55 72 72 28 0 0 72 72 21 70 70 27 43 72 72 72 72 72 72\n",
20428
- " 0 0 0 33 72 72 72 72 26 26 72 72 11 48 72 72 72 21 21 64 0 72 72 30\n",
20429
- " 72 72 72 72 72 72 59 72 72 72 23 54 72 72 72 27 27 72 72 72 72 72 1 72\n",
20430
- " 72 0 0 72 72 72 72 72 3 70 27 27 50 72 72 72 5 0 0 72 30 30 44 72\n",
20431
- " 72 5 5 70 72 31 31 43 72 72 72 72 72 72 72 72 27 27 72 72 72 72 72 25\n",
20432
- " 72 0 0 72 26 72 72 72 72 72 72 72 1 0 72 72 18 58 72 0 0 0 33 72\n",
20433
- " 72 72 72 72 72 72 26 26 0 0 72 72 25 25 49 72 72 72 72 72 72 72 72 26\n",
20434
- " 0 0 72 20 58 72 72 72 72 0 0 21 25 72 70 70 70 72 11 72 72 72 72 72\n",
20435
- " 72 59 72 72 72 72 72 29 72 72 72 72 72 70 70 16 0 0 30 30 72 72 72 25\n",
20436
- " 70 70 72 27 48 72 72 72 5 5 0 33 72 72 20 70 70 70 70 11 55 72 72 72\n",
20437
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20438
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20439
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20440
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20441
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20442
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20443
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20444
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20445
- " 0 0 0 0 0 0 0]\n",
20446
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20447
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20448
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20449
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20450
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20451
- " 72 72 72 72 72 72 72]\n",
20452
- "-----------------\n"
20453
- ]
20454
- },
20455
- {
20456
- "name": "stderr",
20457
- "output_type": "stream",
20458
- "text": [
20459
  "Saving model checkpoint to ./checkpoint-1600\n",
20460
  "Configuration saved in ./checkpoint-1600/config.json\n",
20461
  "Model weights saved in ./checkpoint-1600/pytorch_model.bin\n",
@@ -20464,48 +20283,7 @@
20464
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20465
  "***** Running Evaluation *****\n",
20466
  " Num examples = 291\n",
20467
- " Batch size = 8\n"
20468
- ]
20469
- },
20470
- {
20471
- "name": "stdout",
20472
- "output_type": "stream",
20473
- "text": [
20474
- "pred : [30 63 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20475
- " 72 72 72 45 72 72 72 72 72 11 43 72 72 72 72 6 72 0 72 25 72 72 72 72\n",
20476
- " 72 72 72 16 72 72 72 72 72 72 0 72 20 58 72 72 0 0 72 72 72 23 54 72\n",
20477
- " 72 27 0 72 11 55 72 72 28 0 0 72 72 21 70 72 27 51 72 72 72 72 72 72\n",
20478
- " 0 0 0 33 70 72 72 72 26 0 0 72 72 11 72 72 72 21 21 64 0 72 72 30\n",
20479
- " 30 72 72 72 72 72 59 72 72 72 23 54 72 72 72 27 27 72 72 72 72 1 72 72\n",
20480
- " 72 72 72 72 72 72 72 72 3 70 72 27 50 72 72 72 5 0 0 72 30 30 44 72\n",
20481
- " 72 5 70 70 70 31 31 43 72 72 72 72 72 72 72 72 27 27 44 72 72 72 72 25\n",
20482
- " 72 0 0 72 26 72 72 72 72 72 72 1 1 0 72 72 18 58 72 0 0 0 33 70\n",
20483
- " 70 72 72 72 72 72 26 72 0 0 72 72 72 25 50 72 72 72 72 72 72 72 26 26\n",
20484
- " 0 0 72 20 58 72 72 72 72 0 0 72 25 72 72 70 70 72 11 72 72 72 72 72\n",
20485
- " 72 59 72 0 0 72 72 29 29 16 72 72 72 70 16 16 0 0 72 30 72 72 72 25\n",
20486
- " 70 70 72 27 72 72 72 72 5 5 0 33 72 72 15 70 70 70 72 11 55 72 72 72\n",
20487
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20488
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20489
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20490
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20491
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20492
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20493
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20494
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20495
- " 0 0 0 0 0 0 0]\n",
20496
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20497
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20498
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20499
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20500
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20501
- " 72 72 72 72 72 72 72]\n",
20502
- "-----------------\n"
20503
- ]
20504
- },
20505
- {
20506
- "name": "stderr",
20507
- "output_type": "stream",
20508
- "text": [
20509
  "Saving model checkpoint to ./checkpoint-2000\n",
20510
  "Configuration saved in ./checkpoint-2000/config.json\n",
20511
  "Model weights saved in ./checkpoint-2000/pytorch_model.bin\n",
@@ -20514,48 +20292,7 @@
20514
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20515
  "***** Running Evaluation *****\n",
20516
  " Num examples = 291\n",
20517
- " Batch size = 8\n"
20518
- ]
20519
- },
20520
- {
20521
- "name": "stdout",
20522
- "output_type": "stream",
20523
- "text": [
20524
- "pred : [30 63 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20525
- " 72 72 45 72 72 72 72 72 72 11 43 72 72 72 72 6 6 0 72 25 62 72 72 72\n",
20526
- " 72 72 72 16 49 72 72 72 72 72 0 72 20 58 72 72 0 0 72 72 23 54 72 72\n",
20527
- " 72 28 0 11 55 72 72 72 28 0 0 72 72 21 70 70 27 51 72 72 72 72 72 72\n",
20528
- " 0 0 0 33 70 72 72 72 26 0 0 72 11 72 72 72 72 21 21 64 0 72 72 30\n",
20529
- " 72 72 72 72 72 72 59 72 72 72 23 54 72 72 72 27 27 72 72 72 72 1 72 72\n",
20530
- " 0 0 72 72 72 72 72 72 3 70 72 27 52 72 72 72 5 0 0 72 30 30 44 72\n",
20531
- " 72 5 70 70 72 31 43 72 72 72 72 72 72 72 72 72 27 44 44 72 72 72 25 25\n",
20532
- " 72 0 0 72 26 26 72 72 72 72 72 1 1 0 72 18 58 72 72 0 0 72 33 70\n",
20533
- " 72 72 72 72 72 72 26 72 0 0 72 72 72 25 49 72 72 72 72 72 72 72 26 26\n",
20534
- " 0 0 72 20 58 72 72 72 72 72 0 72 25 72 70 70 72 11 48 72 72 72 72 72\n",
20535
- " 59 59 72 0 0 72 72 29 16 16 72 72 70 70 16 72 0 0 30 30 72 72 72 25\n",
20536
- " 70 70 72 27 72 72 72 72 5 72 0 33 72 72 15 70 70 72 11 55 72 72 72 72\n",
20537
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20538
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20539
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20540
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20541
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20542
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20543
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20544
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20545
- " 0 0 0 0 0 0 0]\n",
20546
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20547
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20548
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20549
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20550
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20551
- " 72 72 72 72 72 72 72]\n",
20552
- "-----------------\n"
20553
- ]
20554
- },
20555
- {
20556
- "name": "stderr",
20557
- "output_type": "stream",
20558
- "text": [
20559
  "Saving model checkpoint to ./checkpoint-2400\n",
20560
  "Configuration saved in ./checkpoint-2400/config.json\n",
20561
  "Model weights saved in ./checkpoint-2400/pytorch_model.bin\n",
@@ -20564,48 +20301,7 @@
20564
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20565
  "***** Running Evaluation *****\n",
20566
  " Num examples = 291\n",
20567
- " Batch size = 8\n"
20568
- ]
20569
- },
20570
- {
20571
- "name": "stdout",
20572
- "output_type": "stream",
20573
- "text": [
20574
- "pred : [30 63 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20575
- " 72 72 72 72 72 72 72 72 72 11 54 72 72 72 72 6 72 0 72 25 62 49 72 72\n",
20576
- " 72 72 72 16 49 72 72 72 72 72 0 72 20 58 72 72 0 0 72 72 23 54 72 72\n",
20577
- " 72 28 0 72 11 55 72 72 28 0 0 72 72 21 70 70 27 51 72 72 72 72 72 72\n",
20578
- " 0 0 0 33 70 72 72 72 26 0 0 72 11 48 72 72 72 21 21 64 0 72 72 30\n",
20579
- " 72 72 72 72 72 25 59 72 72 72 23 54 72 72 72 27 27 72 72 72 72 72 72 72\n",
20580
- " 0 0 72 72 72 72 72 72 3 70 72 27 52 72 72 72 5 0 0 72 30 30 44 72\n",
20581
- " 72 5 70 70 72 31 43 72 72 72 72 72 72 72 72 72 27 44 72 72 72 72 25 25\n",
20582
- " 72 0 0 72 26 72 72 72 72 72 72 1 1 0 72 18 58 72 72 0 0 0 33 70\n",
20583
- " 72 72 72 72 72 72 26 72 0 0 72 72 72 25 50 72 72 72 72 72 72 72 26 26\n",
20584
- " 0 0 72 20 58 72 72 72 72 72 0 72 25 72 70 70 70 72 11 48 72 72 72 72\n",
20585
- " 72 59 72 0 0 72 72 29 16 72 72 72 70 70 16 72 0 0 30 30 72 72 72 25\n",
20586
- " 70 72 72 27 72 72 72 72 5 72 0 33 72 72 15 70 70 72 11 55 72 72 72 72\n",
20587
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20588
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20589
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20590
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20591
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20592
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20593
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20594
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20595
- " 0 0 0 0 0 0 0]\n",
20596
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20597
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20598
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20599
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20600
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20601
- " 72 72 72 72 72 72 72]\n",
20602
- "-----------------\n"
20603
- ]
20604
- },
20605
- {
20606
- "name": "stderr",
20607
- "output_type": "stream",
20608
- "text": [
20609
  "Saving model checkpoint to ./checkpoint-2800\n",
20610
  "Configuration saved in ./checkpoint-2800/config.json\n",
20611
  "Model weights saved in ./checkpoint-2800/pytorch_model.bin\n",
@@ -20614,48 +20310,7 @@
20614
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20615
  "***** Running Evaluation *****\n",
20616
  " Num examples = 291\n",
20617
- " Batch size = 8\n"
20618
- ]
20619
- },
20620
- {
20621
- "name": "stdout",
20622
- "output_type": "stream",
20623
- "text": [
20624
- "pred : [30 63 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20625
- " 72 72 72 72 72 72 72 72 72 11 43 72 72 72 72 6 72 0 72 25 62 62 49 72\n",
20626
- " 72 72 72 16 49 72 72 72 72 0 0 72 20 58 72 72 0 0 72 72 23 54 18 72\n",
20627
- " 72 28 0 72 11 55 72 28 28 0 0 72 72 21 70 72 27 51 72 72 72 72 72 72\n",
20628
- " 0 0 0 33 70 72 72 72 26 0 0 72 11 48 72 72 72 21 64 64 0 72 72 30\n",
20629
- " 72 72 72 72 72 25 72 72 72 72 23 54 72 72 72 27 27 72 72 72 72 1 72 72\n",
20630
- " 0 0 72 72 72 72 72 72 3 70 27 52 72 72 72 72 5 0 0 72 30 30 44 72\n",
20631
- " 72 5 70 70 72 31 43 72 72 72 72 72 72 72 72 72 27 44 72 72 72 72 25 72\n",
20632
- " 72 0 0 72 26 26 72 72 72 72 72 1 72 0 72 18 58 72 72 0 0 0 33 70\n",
20633
- " 72 72 72 72 72 72 26 72 0 0 72 72 72 25 50 72 72 72 72 72 72 72 26 72\n",
20634
- " 0 0 72 20 58 72 72 72 72 0 0 72 25 72 72 70 70 72 11 48 72 72 72 72\n",
20635
- " 59 72 72 0 0 72 72 29 16 72 72 72 70 70 16 72 0 0 72 30 72 72 72 25\n",
20636
- " 70 72 27 27 72 72 72 72 5 72 0 33 72 72 15 70 70 72 11 55 72 72 72 72\n",
20637
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20638
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20639
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20640
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20641
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20642
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20643
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20644
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20645
- " 0 0 0 0 0 0 0]\n",
20646
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20647
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20648
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20649
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20650
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20651
- " 72 72 72 72 72 72 72]\n",
20652
- "-----------------\n"
20653
- ]
20654
- },
20655
- {
20656
- "name": "stderr",
20657
- "output_type": "stream",
20658
- "text": [
20659
  "Saving model checkpoint to ./checkpoint-3200\n",
20660
  "Configuration saved in ./checkpoint-3200/config.json\n",
20661
  "Model weights saved in ./checkpoint-3200/pytorch_model.bin\n",
@@ -20664,48 +20319,7 @@
20664
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20665
  "***** Running Evaluation *****\n",
20666
  " Num examples = 291\n",
20667
- " Batch size = 8\n"
20668
- ]
20669
- },
20670
- {
20671
- "name": "stdout",
20672
- "output_type": "stream",
20673
- "text": [
20674
- "pred : [30 63 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20675
- " 72 72 72 72 72 72 72 72 72 11 44 72 72 72 72 6 72 0 72 25 62 49 49 72\n",
20676
- " 72 72 72 16 49 72 72 72 72 72 0 72 20 58 72 72 0 0 72 72 23 54 72 72\n",
20677
- " 72 28 0 72 11 55 72 72 28 0 0 72 72 21 70 70 27 51 72 72 72 72 72 72\n",
20678
- " 0 0 0 42 70 72 72 72 26 0 0 72 11 48 72 72 72 21 64 64 0 72 72 30\n",
20679
- " 72 72 72 72 72 25 72 72 72 72 23 54 72 72 72 27 27 72 72 72 72 72 72 72\n",
20680
- " 0 0 72 72 72 72 72 72 3 70 72 27 52 72 72 72 5 0 0 72 30 30 44 72\n",
20681
- " 72 5 70 70 72 31 43 72 72 72 72 72 72 72 72 72 27 44 72 72 72 72 25 72\n",
20682
- " 72 0 0 72 72 26 72 72 72 72 72 1 72 0 72 18 58 72 72 0 0 0 33 70\n",
20683
- " 72 72 72 72 72 72 26 72 0 0 72 72 72 25 50 72 72 72 72 72 72 72 26 26\n",
20684
- " 0 0 72 20 58 72 72 72 72 72 0 72 25 72 72 70 70 72 11 48 72 72 72 72\n",
20685
- " 72 59 72 72 0 72 72 29 16 72 72 72 72 70 16 72 0 0 72 30 72 72 72 25\n",
20686
- " 70 72 72 27 72 72 72 72 5 72 0 33 72 72 15 70 70 72 11 55 72 72 72 72\n",
20687
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20688
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20689
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20690
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20691
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20692
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20693
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20694
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20695
- " 0 0 0 0 0 0 0]\n",
20696
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20697
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20698
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20699
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20700
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20701
- " 72 72 72 72 72 72 72]\n",
20702
- "-----------------\n"
20703
- ]
20704
- },
20705
- {
20706
- "name": "stderr",
20707
- "output_type": "stream",
20708
- "text": [
20709
  "Saving model checkpoint to ./checkpoint-3600\n",
20710
  "Configuration saved in ./checkpoint-3600/config.json\n",
20711
  "Model weights saved in ./checkpoint-3600/pytorch_model.bin\n",
@@ -20714,48 +20328,7 @@
20714
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20715
  "***** Running Evaluation *****\n",
20716
  " Num examples = 291\n",
20717
- " Batch size = 8\n"
20718
- ]
20719
- },
20720
- {
20721
- "name": "stdout",
20722
- "output_type": "stream",
20723
- "text": [
20724
- "pred : [30 63 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20725
- " 72 72 72 72 72 72 72 72 72 11 43 72 72 72 72 6 72 0 72 25 62 62 49 72\n",
20726
- " 72 72 72 16 49 72 72 72 72 72 0 72 20 58 72 72 0 0 72 72 23 54 18 72\n",
20727
- " 72 28 0 0 11 55 72 72 28 0 0 72 72 21 70 70 27 51 72 72 72 72 72 72\n",
20728
- " 0 0 0 42 70 72 72 72 26 0 0 72 11 48 72 72 72 21 21 64 0 72 72 30\n",
20729
- " 30 72 72 72 72 25 72 72 72 72 23 54 72 72 72 27 27 72 72 72 72 72 72 72\n",
20730
- " 0 0 72 72 72 72 72 72 3 70 72 27 52 72 72 72 5 0 0 72 30 30 44 72\n",
20731
- " 72 5 70 70 72 31 43 72 72 72 72 72 72 72 72 72 27 46 72 72 72 72 25 72\n",
20732
- " 72 0 0 72 72 26 72 72 72 72 72 1 72 0 72 18 58 72 72 0 0 0 33 70\n",
20733
- " 72 72 72 72 72 72 26 72 0 0 72 72 72 25 50 72 72 72 72 72 72 72 26 26\n",
20734
- " 0 0 72 20 58 72 72 72 72 72 0 72 25 72 70 70 70 72 11 48 72 72 72 72\n",
20735
- " 72 59 72 72 0 72 72 29 16 72 72 72 70 70 16 72 0 0 72 30 72 72 72 25\n",
20736
- " 72 72 72 27 72 72 72 72 5 72 0 33 72 72 15 70 70 72 72 12 55 72 72 72\n",
20737
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20738
- " 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 16 16 0 0 0 0 0 0 0\n",
20739
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20740
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20741
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20742
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20743
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20744
- " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
20745
- " 0 0 0 0 0 0 0]\n",
20746
- "label: [30 63 45 0 11 43 6 64 0 25 62 49 16 49 0 20 58 0 23 54 28 0 11 55\n",
20747
- " 28 0 21 70 27 51 0 42 70 26 0 13 48 21 0 30 25 70 24 43 27 61 0 3\n",
20748
- " 70 27 52 5 0 30 5 70 31 43 27 46 25 0 26 1 0 18 58 0 42 70 26 0\n",
20749
- " 25 62 49 26 0 20 58 0 25 70 11 48 59 0 29 16 70 11 0 30 59 27 57 5\n",
20750
- " 33 15 70 11 55 16 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72 72\n",
20751
- " 72 72 72 72 72 72 72]\n",
20752
- "-----------------\n"
20753
- ]
20754
- },
20755
- {
20756
- "name": "stderr",
20757
- "output_type": "stream",
20758
- "text": [
20759
  "Saving model checkpoint to ./checkpoint-4000\n",
20760
  "Configuration saved in ./checkpoint-4000/config.json\n",
20761
  "Model weights saved in ./checkpoint-4000/pytorch_model.bin\n",
@@ -20766,16 +20339,16 @@
20766
  "Training completed. Do not forget to share your model on huggingface.co/models =)\n",
20767
  "\n",
20768
  "\n",
20769
- "Loading best model from ./checkpoint-4000 (score: 0.46376487612724304).\n"
20770
  ]
20771
  },
20772
  {
20773
  "data": {
20774
  "text/plain": [
20775
- "TrainOutput(global_step=4050, training_loss=2.89372775796019, metrics={'train_runtime': 8168.6927, 'train_samples_per_second': 16.006, 'train_steps_per_second': 0.496, 'total_flos': 1.9735608328149316e+19, 'train_loss': 2.89372775796019, 'epoch': 49.99})"
20776
  ]
20777
  },
20778
- "execution_count": 49,
20779
  "metadata": {},
20780
  "output_type": "execute_result"
20781
  }
@@ -20787,7 +20360,7 @@
20787
  {
20788
  "cell_type": "code",
20789
  "execution_count": 57,
20790
- "id": "57c2527b",
20791
  "metadata": {},
20792
  "outputs": [
20793
  {
@@ -20806,8 +20379,8 @@
20806
  },
20807
  {
20808
  "cell_type": "code",
20809
- "execution_count": 53,
20810
- "id": "0211e267",
20811
  "metadata": {},
20812
  "outputs": [],
20813
  "source": [
@@ -20823,8 +20396,8 @@
20823
  },
20824
  {
20825
  "cell_type": "code",
20826
- "execution_count": 54,
20827
- "id": "62f6fd3e",
20828
  "metadata": {},
20829
  "outputs": [
20830
  {
@@ -20842,8 +20415,8 @@
20842
  },
20843
  {
20844
  "cell_type": "code",
20845
- "execution_count": 60,
20846
- "id": "b050fb9f",
20847
  "metadata": {},
20848
  "outputs": [
20849
  {
@@ -20858,12 +20431,12 @@
20858
  {
20859
  "data": {
20860
  "application/vnd.jupyter.widget-view+json": {
20861
- "model_id": "331db7acce774ee3b699aa82a0451092",
20862
  "version_major": 2,
20863
  "version_minor": 0
20864
  },
20865
  "text/plain": [
20866
- "Download file pytorch_model.bin: 0%| | 3.47k/1.18G [00:00<?, ?B/s]"
20867
  ]
20868
  },
20869
  "metadata": {},
@@ -20872,7 +20445,35 @@
20872
  {
20873
  "data": {
20874
  "application/vnd.jupyter.widget-view+json": {
20875
- "model_id": "db90465291c64e9f82988698d2473234",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20876
  "version_major": 2,
20877
  "version_minor": 0
20878
  },
@@ -20890,6 +20491,39 @@
20890
  "Configuration saved in vitouphy/xls-r-300m-km/config.json\n",
20891
  "Model weights saved in vitouphy/xls-r-300m-km/pytorch_model.bin\n"
20892
  ]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20893
  }
20894
  ],
20895
  "source": [
@@ -20898,8 +20532,8 @@
20898
  },
20899
  {
20900
  "cell_type": "code",
20901
- "execution_count": 61,
20902
- "id": "9d7cb173",
20903
  "metadata": {},
20904
  "outputs": [
20905
  {
@@ -20920,7 +20554,7 @@
20920
  {
20921
  "cell_type": "code",
20922
  "execution_count": null,
20923
- "id": "8dc01ad4",
20924
  "metadata": {},
20925
  "outputs": [],
20926
  "source": []
 
3
  {
4
  "cell_type": "code",
5
  "execution_count": 1,
6
+ "id": "bff05704",
7
  "metadata": {},
8
  "outputs": [],
9
  "source": [
 
16
  {
17
  "cell_type": "code",
18
  "execution_count": null,
19
+ "id": "9637cdfd",
20
  "metadata": {
21
  "collapsed": true,
22
  "jupyter": {
 
19167
  },
19168
  {
19169
  "cell_type": "markdown",
19170
+ "id": "b11b1d53",
19171
  "metadata": {},
19172
  "source": [
19173
  "### Load KH Data"
 
19176
  {
19177
  "cell_type": "code",
19178
  "execution_count": 4,
19179
+ "id": "f35b6d68",
19180
  "metadata": {},
19181
  "outputs": [],
19182
  "source": [
 
19197
  {
19198
  "cell_type": "code",
19199
  "execution_count": 5,
19200
+ "id": "a0b561cb",
19201
  "metadata": {},
19202
  "outputs": [
19203
  {
 
19307
  {
19308
  "cell_type": "code",
19309
  "execution_count": 6,
19310
+ "id": "c8ae4532",
19311
  "metadata": {},
19312
  "outputs": [],
19313
  "source": [
 
19321
  },
19322
  {
19323
  "cell_type": "markdown",
19324
+ "id": "4649ca2b",
19325
  "metadata": {},
19326
  "source": [
19327
  "### Clean Up the Text"
 
19330
  {
19331
  "cell_type": "code",
19332
  "execution_count": 6,
19333
+ "id": "363283a2",
19334
  "metadata": {},
19335
  "outputs": [],
19336
  "source": [
 
19346
  {
19347
  "cell_type": "code",
19348
  "execution_count": 7,
19349
+ "id": "51f70aa8",
19350
  "metadata": {
19351
  "collapsed": true,
19352
  "jupyter": {
 
19402
  {
19403
  "cell_type": "code",
19404
  "execution_count": 7,
19405
+ "id": "fbc089d7",
19406
  "metadata": {},
19407
  "outputs": [
19408
  {
 
19423
  },
19424
  {
19425
  "cell_type": "markdown",
19426
+ "id": "af02801f",
19427
  "metadata": {},
19428
  "source": [
19429
  "### Build Character"
 
19432
  {
19433
  "cell_type": "code",
19434
  "execution_count": 8,
19435
+ "id": "a9e58b43",
19436
  "metadata": {},
19437
  "outputs": [
19438
  {
 
19480
  {
19481
  "cell_type": "code",
19482
  "execution_count": 9,
19483
+ "id": "4480543c",
19484
  "metadata": {},
19485
  "outputs": [],
19486
  "source": [
 
19491
  {
19492
  "cell_type": "code",
19493
  "execution_count": 10,
19494
+ "id": "99857f4d",
19495
  "metadata": {},
19496
  "outputs": [
19497
  {
 
19509
  {
19510
  "cell_type": "code",
19511
  "execution_count": 11,
19512
+ "id": "bec53215",
19513
  "metadata": {},
19514
  "outputs": [
19515
  {
 
19536
  {
19537
  "cell_type": "code",
19538
  "execution_count": 12,
19539
+ "id": "cf58f8a4",
19540
  "metadata": {},
19541
  "outputs": [
19542
  {
 
19554
  {
19555
  "cell_type": "code",
19556
  "execution_count": 13,
19557
+ "id": "0c621a15",
19558
  "metadata": {},
19559
  "outputs": [],
19560
  "source": [
 
19565
  },
19566
  {
19567
  "cell_type": "markdown",
19568
+ "id": "bb8b5aa3",
19569
  "metadata": {},
19570
  "source": [
19571
  "# Tokenizer"
 
19574
  {
19575
  "cell_type": "code",
19576
  "execution_count": 14,
19577
+ "id": "dc1c1984",
19578
  "metadata": {},
19579
  "outputs": [],
19580
  "source": [
 
19585
  },
19586
  {
19587
  "cell_type": "code",
19588
+ "execution_count": 63,
19589
+ "id": "6324377d",
19590
  "metadata": {},
19591
  "outputs": [
19592
  {
 
19598
  "loading file ./tokenizer_config.json\n",
19599
  "loading file ./added_tokens.json\n",
19600
  "loading file ./special_tokens_map.json\n",
19601
+ "loading file None\n",
19602
+ "Adding <s> to the vocabulary\n",
19603
+ "Adding </s> to the vocabulary\n"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19604
  ]
19605
  }
19606
  ],
 
19613
  {
19614
  "cell_type": "code",
19615
  "execution_count": 26,
19616
+ "id": "f971580d",
19617
  "metadata": {},
19618
  "outputs": [],
19619
  "source": [
 
19630
  {
19631
  "cell_type": "code",
19632
  "execution_count": 27,
19633
+ "id": "d0368c7a",
19634
  "metadata": {},
19635
  "outputs": [
19636
  {
 
19670
  {
19671
  "cell_type": "code",
19672
  "execution_count": 17,
19673
+ "id": "62e9d0c6",
19674
  "metadata": {},
19675
  "outputs": [],
19676
  "source": [
 
19681
  {
19682
  "cell_type": "code",
19683
  "execution_count": 18,
19684
+ "id": "f642a861",
19685
  "metadata": {},
19686
  "outputs": [
19687
  {
 
19706
  {
19707
  "cell_type": "code",
19708
  "execution_count": 19,
19709
+ "id": "0c756a07",
19710
  "metadata": {},
19711
  "outputs": [
19712
  {
 
19753
  {
19754
  "cell_type": "code",
19755
  "execution_count": 20,
19756
+ "id": "d2a5374c",
19757
  "metadata": {},
19758
  "outputs": [],
19759
  "source": [
 
19775
  {
19776
  "cell_type": "code",
19777
  "execution_count": 22,
19778
+ "id": "9c3697ba",
19779
  "metadata": {},
19780
  "outputs": [],
19781
  "source": [
 
19786
  {
19787
  "cell_type": "code",
19788
  "execution_count": 41,
19789
+ "id": "d5bd0662",
19790
  "metadata": {},
19791
  "outputs": [],
19792
  "source": [
 
19798
  {
19799
  "cell_type": "code",
19800
  "execution_count": 25,
19801
+ "id": "639dd5a7",
19802
  "metadata": {},
19803
  "outputs": [],
19804
  "source": [
 
19858
  {
19859
  "cell_type": "code",
19860
  "execution_count": 26,
19861
+ "id": "c4fe1643",
19862
  "metadata": {},
19863
  "outputs": [],
19864
  "source": [
 
19868
  {
19869
  "cell_type": "code",
19870
  "execution_count": 27,
19871
+ "id": "9fb388e3",
19872
  "metadata": {},
19873
  "outputs": [],
19874
  "source": [
 
19878
  },
19879
  {
19880
  "cell_type": "code",
19881
+ "execution_count": 64,
19882
+ "id": "96611455",
19883
  "metadata": {},
19884
  "outputs": [],
19885
  "source": [
 
19892
  " pred_str = tokenizer.batch_decode(pred_ids)\n",
19893
  " label_str = tokenizer.batch_decode(pred.label_ids, group_tokens=False)\n",
19894
  "\n",
19895
+ "# print(\"pred : \", pred_ids[0])\n",
19896
+ "# print(\"label: \", pred.label_ids[0])\n",
19897
+ "# print(\"-----------------\")\n",
19898
  " \n",
19899
  " wer = wer_metric.compute(predictions=pred_str, references=label_str)\n",
19900
  "\n",
 
19903
  },
19904
  {
19905
  "cell_type": "code",
19906
+ "execution_count": 66,
19907
+ "id": "bb429520",
19908
  "metadata": {
19909
  "collapsed": true,
19910
  "jupyter": {
 
19916
  "name": "stderr",
19917
  "output_type": "stream",
19918
  "text": [
19919
+ "loading configuration file checkpoint-4000/config.json\n",
19920
  "Model config Wav2Vec2Config {\n",
19921
+ " \"_name_or_path\": \"facebook/wav2vec2-xls-r-300m\",\n",
19922
  " \"activation_dropout\": 0.0,\n",
19923
  " \"adapter_kernel_size\": 3,\n",
19924
  " \"adapter_stride\": 2,\n",
19925
  " \"add_adapter\": false,\n",
19926
  " \"apply_spec_augment\": true,\n",
19927
  " \"architectures\": [\n",
19928
+ " \"Wav2Vec2ForCTC\"\n",
19929
  " ],\n",
19930
  " \"attention_dropout\": 0.1,\n",
19931
  " \"bos_token_id\": 1,\n",
 
20026
  " \"xvector_output_dim\": 512\n",
20027
  "}\n",
20028
  "\n",
20029
+ "loading weights file checkpoint-4000/pytorch_model.bin\n",
20030
+ "All model checkpoint weights were used when initializing Wav2Vec2ForCTC.\n",
20031
+ "\n",
20032
+ "All the weights of Wav2Vec2ForCTC were initialized from the model checkpoint at checkpoint-4000.\n",
20033
+ "If your task is similar to the task the model of the checkpoint was trained on, you can already use Wav2Vec2ForCTC for predictions without further training.\n"
 
20034
  ]
20035
  }
20036
  ],
 
20038
  "from transformers import Wav2Vec2ForCTC\n",
20039
  "\n",
20040
  "model = Wav2Vec2ForCTC.from_pretrained(\n",
20041
+ "# \"facebook/wav2vec2-xls-r-300m\", \n",
20042
+ " \"checkpoint-4000\",\n",
20043
  " attention_dropout=0.1,\n",
20044
  " layerdrop=0.0,\n",
20045
  " feat_proj_dropout=0.0,\n",
 
20055
  },
20056
  {
20057
  "cell_type": "code",
20058
+ "execution_count": 68,
20059
+ "id": "ffcd9012",
20060
  "metadata": {},
20061
  "outputs": [],
20062
  "source": [
 
20065
  },
20066
  {
20067
  "cell_type": "code",
20068
+ "execution_count": 69,
20069
+ "id": "b07418cf",
20070
  "metadata": {},
20071
  "outputs": [
20072
  {
 
20094
  " eval_steps=400,\n",
20095
  " logging_steps=100,\n",
20096
  " learning_rate=5e-5,\n",
20097
+ " warmup_steps=100,\n",
20098
  " save_total_limit=3,\n",
20099
  " load_best_model_at_end=True\n",
20100
  ")"
 
20102
  },
20103
  {
20104
  "cell_type": "code",
20105
+ "execution_count": 70,
20106
+ "id": "7776cd7d",
20107
  "metadata": {},
20108
  "outputs": [
20109
  {
 
20130
  },
20131
  {
20132
  "cell_type": "code",
20133
+ "execution_count": 71,
20134
+ "id": "ac33ed4c",
20135
+ "metadata": {},
 
 
 
 
 
20136
  "outputs": [
20137
  {
20138
  "name": "stderr",
 
20157
  " <div>\n",
20158
  " \n",
20159
  " <progress value='4050' max='4050' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
20160
+ " [4050/4050 2:16:09, Epoch 49/50]\n",
20161
  " </div>\n",
20162
  " <table border=\"1\" class=\"dataframe\">\n",
20163
  " <thead>\n",
 
20171
  " <tbody>\n",
20172
  " <tr>\n",
20173
  " <td>400</td>\n",
20174
+ " <td>1.382900</td>\n",
20175
+ " <td>0.429020</td>\n",
20176
+ " <td>0.479627</td>\n",
20177
  " </tr>\n",
20178
  " <tr>\n",
20179
  " <td>800</td>\n",
20180
+ " <td>1.315600</td>\n",
20181
+ " <td>0.385632</td>\n",
20182
+ " <td>0.447419</td>\n",
20183
  " </tr>\n",
20184
  " <tr>\n",
20185
  " <td>1200</td>\n",
20186
+ " <td>1.239600</td>\n",
20187
+ " <td>0.359977</td>\n",
20188
+ " <td>0.430733</td>\n",
20189
  " </tr>\n",
20190
  " <tr>\n",
20191
  " <td>1600</td>\n",
20192
+ " <td>1.144400</td>\n",
20193
+ " <td>0.342276</td>\n",
20194
+ " <td>0.417928</td>\n",
20195
  " </tr>\n",
20196
  " <tr>\n",
20197
  " <td>2000</td>\n",
20198
+ " <td>1.097900</td>\n",
20199
+ " <td>0.337029</td>\n",
20200
+ " <td>0.388436</td>\n",
20201
  " </tr>\n",
20202
  " <tr>\n",
20203
  " <td>2400</td>\n",
20204
+ " <td>1.071400</td>\n",
20205
+ " <td>0.323725</td>\n",
20206
+ " <td>0.370974</td>\n",
20207
  " </tr>\n",
20208
  " <tr>\n",
20209
  " <td>2800</td>\n",
20210
+ " <td>1.044200</td>\n",
20211
+ " <td>0.333624</td>\n",
20212
+ " <td>0.368258</td>\n",
20213
  " </tr>\n",
20214
  " <tr>\n",
20215
  " <td>3200</td>\n",
20216
+ " <td>1.049200</td>\n",
20217
+ " <td>0.316629</td>\n",
20218
+ " <td>0.352736</td>\n",
20219
  " </tr>\n",
20220
  " <tr>\n",
20221
  " <td>3600</td>\n",
20222
+ " <td>1.028400</td>\n",
20223
+ " <td>0.317763</td>\n",
20224
+ " <td>0.356616</td>\n",
20225
  " </tr>\n",
20226
  " <tr>\n",
20227
  " <td>4000</td>\n",
20228
+ " <td>1.030200</td>\n",
20229
+ " <td>0.314151</td>\n",
20230
+ " <td>0.351184</td>\n",
20231
  " </tr>\n",
20232
  " </tbody>\n",
20233
  "</table><p>"
 
20246
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20247
  "***** Running Evaluation *****\n",
20248
  " Num examples = 291\n",
20249
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20250
  "Saving model checkpoint to ./checkpoint-400\n",
20251
  "Configuration saved in ./checkpoint-400/config.json\n",
20252
  "Model weights saved in ./checkpoint-400/pytorch_model.bin\n",
20253
  "Configuration saved in ./checkpoint-400/preprocessor_config.json\n",
20254
+ "Deleting older checkpoint [checkpoint-3200] due to args.save_total_limit\n",
20255
+ "Deleting older checkpoint [checkpoint-3600] due to args.save_total_limit\n",
20256
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20257
  "***** Running Evaluation *****\n",
20258
  " Num examples = 291\n",
20259
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20260
  "Saving model checkpoint to ./checkpoint-800\n",
20261
  "Configuration saved in ./checkpoint-800/config.json\n",
20262
  "Model weights saved in ./checkpoint-800/pytorch_model.bin\n",
20263
  "Configuration saved in ./checkpoint-800/preprocessor_config.json\n",
20264
+ "Deleting older checkpoint [checkpoint-4000-prev-best] due to args.save_total_limit\n",
20265
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20266
  "***** Running Evaluation *****\n",
20267
  " Num examples = 291\n",
20268
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20269
  "Saving model checkpoint to ./checkpoint-1200\n",
20270
  "Configuration saved in ./checkpoint-1200/config.json\n",
20271
  "Model weights saved in ./checkpoint-1200/pytorch_model.bin\n",
20272
  "Configuration saved in ./checkpoint-1200/preprocessor_config.json\n",
20273
+ "Deleting older checkpoint [checkpoint-4000] due to args.save_total_limit\n",
20274
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20275
  "***** Running Evaluation *****\n",
20276
  " Num examples = 291\n",
20277
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20278
  "Saving model checkpoint to ./checkpoint-1600\n",
20279
  "Configuration saved in ./checkpoint-1600/config.json\n",
20280
  "Model weights saved in ./checkpoint-1600/pytorch_model.bin\n",
 
20283
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20284
  "***** Running Evaluation *****\n",
20285
  " Num examples = 291\n",
20286
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20287
  "Saving model checkpoint to ./checkpoint-2000\n",
20288
  "Configuration saved in ./checkpoint-2000/config.json\n",
20289
  "Model weights saved in ./checkpoint-2000/pytorch_model.bin\n",
 
20292
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20293
  "***** Running Evaluation *****\n",
20294
  " Num examples = 291\n",
20295
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20296
  "Saving model checkpoint to ./checkpoint-2400\n",
20297
  "Configuration saved in ./checkpoint-2400/config.json\n",
20298
  "Model weights saved in ./checkpoint-2400/pytorch_model.bin\n",
 
20301
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20302
  "***** Running Evaluation *****\n",
20303
  " Num examples = 291\n",
20304
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20305
  "Saving model checkpoint to ./checkpoint-2800\n",
20306
  "Configuration saved in ./checkpoint-2800/config.json\n",
20307
  "Model weights saved in ./checkpoint-2800/pytorch_model.bin\n",
 
20310
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20311
  "***** Running Evaluation *****\n",
20312
  " Num examples = 291\n",
20313
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20314
  "Saving model checkpoint to ./checkpoint-3200\n",
20315
  "Configuration saved in ./checkpoint-3200/config.json\n",
20316
  "Model weights saved in ./checkpoint-3200/pytorch_model.bin\n",
 
20319
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20320
  "***** Running Evaluation *****\n",
20321
  " Num examples = 291\n",
20322
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20323
  "Saving model checkpoint to ./checkpoint-3600\n",
20324
  "Configuration saved in ./checkpoint-3600/config.json\n",
20325
  "Model weights saved in ./checkpoint-3600/pytorch_model.bin\n",
 
20328
  "The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
20329
  "***** Running Evaluation *****\n",
20330
  " Num examples = 291\n",
20331
+ " Batch size = 8\n",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20332
  "Saving model checkpoint to ./checkpoint-4000\n",
20333
  "Configuration saved in ./checkpoint-4000/config.json\n",
20334
  "Model weights saved in ./checkpoint-4000/pytorch_model.bin\n",
 
20339
  "Training completed. Do not forget to share your model on huggingface.co/models =)\n",
20340
  "\n",
20341
  "\n",
20342
+ "Loading best model from ./checkpoint-4000 (score: 0.3141506016254425).\n"
20343
  ]
20344
  },
20345
  {
20346
  "data": {
20347
  "text/plain": [
20348
+ "TrainOutput(global_step=4050, training_loss=1.1567209813624253, metrics={'train_runtime': 8173.6251, 'train_samples_per_second': 15.997, 'train_steps_per_second': 0.495, 'total_flos': 1.9735608328149316e+19, 'train_loss': 1.1567209813624253, 'epoch': 49.99})"
20349
  ]
20350
  },
20351
+ "execution_count": 71,
20352
  "metadata": {},
20353
  "output_type": "execute_result"
20354
  }
 
20360
  {
20361
  "cell_type": "code",
20362
  "execution_count": 57,
20363
+ "id": "19b3350f",
20364
  "metadata": {},
20365
  "outputs": [
20366
  {
 
20379
  },
20380
  {
20381
  "cell_type": "code",
20382
+ "execution_count": 72,
20383
+ "id": "724e14ef",
20384
  "metadata": {},
20385
  "outputs": [],
20386
  "source": [
 
20396
  },
20397
  {
20398
  "cell_type": "code",
20399
+ "execution_count": 73,
20400
+ "id": "75b87f11",
20401
  "metadata": {},
20402
  "outputs": [
20403
  {
 
20415
  },
20416
  {
20417
  "cell_type": "code",
20418
+ "execution_count": 74,
20419
+ "id": "9e4a2ec9",
20420
  "metadata": {},
20421
  "outputs": [
20422
  {
 
20431
  {
20432
  "data": {
20433
  "application/vnd.jupyter.widget-view+json": {
20434
+ "model_id": "ae4aa0641113454c801089fa2dbd6777",
20435
  "version_major": 2,
20436
  "version_minor": 0
20437
  },
20438
  "text/plain": [
20439
+ "Download file pytorch_model.bin: 0%| | 2.83k/1.18G [00:00<?, ?B/s]"
20440
  ]
20441
  },
20442
  "metadata": {},
 
20445
  {
20446
  "data": {
20447
  "application/vnd.jupyter.widget-view+json": {
20448
+ "model_id": "9a3129d18855473ba7da0f290f26419b",
20449
+ "version_major": 2,
20450
+ "version_minor": 0
20451
+ },
20452
+ "text/plain": [
20453
+ "Download file training_args.bin: 63%|######2 | 1.84k/2.92k [00:00<?, ?B/s]"
20454
+ ]
20455
+ },
20456
+ "metadata": {},
20457
+ "output_type": "display_data"
20458
+ },
20459
+ {
20460
+ "data": {
20461
+ "application/vnd.jupyter.widget-view+json": {
20462
+ "model_id": "dabccfa9f14045919cf70a905afb5506",
20463
+ "version_major": 2,
20464
+ "version_minor": 0
20465
+ },
20466
+ "text/plain": [
20467
+ "Clean file training_args.bin: 34%|###4 | 1.00k/2.92k [00:00<?, ?B/s]"
20468
+ ]
20469
+ },
20470
+ "metadata": {},
20471
+ "output_type": "display_data"
20472
+ },
20473
+ {
20474
+ "data": {
20475
+ "application/vnd.jupyter.widget-view+json": {
20476
+ "model_id": "ee7e633b1e784625b2d3695176f6c0f2",
20477
  "version_major": 2,
20478
  "version_minor": 0
20479
  },
 
20491
  "Configuration saved in vitouphy/xls-r-300m-km/config.json\n",
20492
  "Model weights saved in vitouphy/xls-r-300m-km/pytorch_model.bin\n"
20493
  ]
20494
+ },
20495
+ {
20496
+ "data": {
20497
+ "application/vnd.jupyter.widget-view+json": {
20498
+ "model_id": "9738e4743ca3470f863dfd4d85f6e411",
20499
+ "version_major": 2,
20500
+ "version_minor": 0
20501
+ },
20502
+ "text/plain": [
20503
+ "Upload file pytorch_model.bin: 0%| | 3.39k/1.18G [00:00<?, ?B/s]"
20504
+ ]
20505
+ },
20506
+ "metadata": {},
20507
+ "output_type": "display_data"
20508
+ },
20509
+ {
20510
+ "name": "stderr",
20511
+ "output_type": "stream",
20512
+ "text": [
20513
+ "To https://huggingface.co/vitouphy/xls-r-300m-km\n",
20514
+ " 6f203d5..74be6ec main -> main\n",
20515
+ "\n"
20516
+ ]
20517
+ },
20518
+ {
20519
+ "data": {
20520
+ "text/plain": [
20521
+ "'https://huggingface.co/vitouphy/xls-r-300m-km/commit/74be6ece8cca85ef00972b1f3f88460217d0acf5'"
20522
+ ]
20523
+ },
20524
+ "execution_count": 74,
20525
+ "metadata": {},
20526
+ "output_type": "execute_result"
20527
  }
20528
  ],
20529
  "source": [
 
20532
  },
20533
  {
20534
  "cell_type": "code",
20535
+ "execution_count": 75,
20536
+ "id": "8c70b0b9",
20537
  "metadata": {},
20538
  "outputs": [
20539
  {
 
20554
  {
20555
  "cell_type": "code",
20556
  "execution_count": null,
20557
+ "id": "96cd8308",
20558
  "metadata": {},
20559
  "outputs": [],
20560
  "source": []
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c93624168d989aabdf33bd32b0dbfa8b32857515b2a2f04190df98bd15ae4e61
3
  size 2991
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59beaed2af6d1171b371d53a9d0077d3ab22cd9f2392cf839539bb2e9f36d978
3
  size 2991