Training in progress, step 200

.gitignore

@@ -0,0 +1 @@
+checkpoint-*/

.ipynb_checkpoints/Untitled-checkpoint.ipynb

@@ -0,0 +1,6 @@
+{
+ "cells": [],
+ "metadata": {},
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

Untitled.ipynb

@@ -0,0 +1,420 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "85541f92",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "02/07/2022 11:34:12 - WARNING - __main__ - Process rank: -1, device: cuda:0, n_gpu: 1distributed training: False, 16-bits training: True\n",
+      "02/07/2022 11:34:12 - INFO - __main__ - Training/evaluation parameters TrainingArguments(\n",
+      "_n_gpu=1,\n",
+      "adafactor=False,\n",
+      "adam_beta1=0.9,\n",
+      "adam_beta2=0.999,\n",
+      "adam_epsilon=1e-08,\n",
+      "bf16=False,\n",
+      "bf16_full_eval=False,\n",
+      "dataloader_drop_last=False,\n",
+      "dataloader_num_workers=0,\n",
+      "dataloader_pin_memory=True,\n",
+      "ddp_bucket_cap_mb=None,\n",
+      "ddp_find_unused_parameters=None,\n",
+      "debug=[],\n",
+      "deepspeed=None,\n",
+      "disable_tqdm=False,\n",
+      "do_eval=True,\n",
+      "do_predict=False,\n",
+      "do_train=True,\n",
+      "eval_accumulation_steps=None,\n",
+      "eval_steps=400,\n",
+      "evaluation_strategy=IntervalStrategy.STEPS,\n",
+      "fp16=True,\n",
+      "fp16_backend=auto,\n",
+      "fp16_full_eval=False,\n",
+      "fp16_opt_level=O1,\n",
+      "gradient_accumulation_steps=1,\n",
+      "gradient_checkpointing=True,\n",
+      "greater_is_better=None,\n",
+      "group_by_length=True,\n",
+      "half_precision_backend=auto,\n",
+      "hub_model_id=None,\n",
+      "hub_strategy=HubStrategy.EVERY_SAVE,\n",
+      "hub_token=<HUB_TOKEN>,\n",
+      "ignore_data_skip=False,\n",
+      "label_names=None,\n",
+      "label_smoothing_factor=0.0,\n",
+      "learning_rate=0.0003,\n",
+      "length_column_name=input_length,\n",
+      "load_best_model_at_end=False,\n",
+      "local_rank=-1,\n",
+      "log_level=-1,\n",
+      "log_level_replica=-1,\n",
+      "log_on_each_node=True,\n",
+      "logging_dir=./runs/Feb07_11-34-12_job-7e123c9a-c8eb-4ec4-9153-164c740ace86,\n",
+      "logging_first_step=False,\n",
+      "logging_nan_inf_filter=True,\n",
+      "logging_steps=500,\n",
+      "logging_strategy=IntervalStrategy.STEPS,\n",
+      "lr_scheduler_type=SchedulerType.LINEAR,\n",
+      "max_grad_norm=1.0,\n",
+      "max_steps=8000,\n",
+      "metric_for_best_model=None,\n",
+      "mp_parameters=,\n",
+      "no_cuda=False,\n",
+      "num_train_epochs=3.0,\n",
+      "optim=OptimizerNames.ADAMW_HF,\n",
+      "output_dir=./,\n",
+      "overwrite_output_dir=True,\n",
+      "past_index=-1,\n",
+      "per_device_eval_batch_size=8,\n",
+      "per_device_train_batch_size=16,\n",
+      "prediction_loss_only=False,\n",
+      "push_to_hub=True,\n",
+      "push_to_hub_model_id=None,\n",
+      "push_to_hub_organization=None,\n",
+      "push_to_hub_token=<PUSH_TO_HUB_TOKEN>,\n",
+      "remove_unused_columns=True,\n",
+      "report_to=[],\n",
+      "resume_from_checkpoint=None,\n",
+      "run_name=./,\n",
+      "save_on_each_node=False,\n",
+      "save_steps=200,\n",
+      "save_strategy=IntervalStrategy.STEPS,\n",
+      "save_total_limit=3,\n",
+      "seed=42,\n",
+      "sharded_ddp=[],\n",
+      "skip_memory_metrics=True,\n",
+      "tf32=None,\n",
+      "tpu_metrics_debug=False,\n",
+      "tpu_num_cores=None,\n",
+      "use_legacy_prediction_loop=False,\n",
+      "warmup_ratio=0.0,\n",
+      "warmup_steps=500,\n",
+      "weight_decay=0.0,\n",
+      "xpu_backend=None,\n",
+      ")\n",
+      "02/07/2022 11:34:14 - WARNING - datasets.builder - Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/hi/7.0.0/fe20cac47c166e25b1f096ab661832e3da7cf298ed4a91dcaa1343ad972d175b)\n",
+      "02/07/2022 11:34:17 - WARNING - datasets.builder - Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/hi/7.0.0/fe20cac47c166e25b1f096ab661832e3da7cf298ed4a91dcaa1343ad972d175b)\n",
+      "02/07/2022 11:34:17 - WARNING - datasets.arrow_dataset - Loading cached processed dataset at /workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/hi/7.0.0/fe20cac47c166e25b1f096ab661832e3da7cf298ed4a91dcaa1343ad972d175b/cache-56c362c60f5a4e8d.arrow\n",
+      "02/07/2022 11:34:17 - WARNING - datasets.arrow_dataset - Loading cached processed dataset at /workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/hi/7.0.0/fe20cac47c166e25b1f096ab661832e3da7cf298ed4a91dcaa1343ad972d175b/cache-c2794d9326d1e793.arrow\n",
+      "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",
+      "Model config Wav2Vec2Config {\n",
+      "  \"_name_or_path\": \"facebook/wav2vec2-xls-r-300m\",\n",
+      "  \"activation_dropout\": 0.0,\n",
+      "  \"adapter_kernel_size\": 3,\n",
+      "  \"adapter_stride\": 2,\n",
+      "  \"add_adapter\": false,\n",
+      "  \"apply_spec_augment\": true,\n",
+      "  \"architectures\": [\n",
+      "    \"Wav2Vec2ForPreTraining\"\n",
+      "  ],\n",
+      "  \"attention_dropout\": 0.1,\n",
+      "  \"bos_token_id\": 1,\n",
+      "  \"classifier_proj_size\": 256,\n",
+      "  \"codevector_dim\": 768,\n",
+      "  \"contrastive_logits_temperature\": 0.1,\n",
+      "  \"conv_bias\": true,\n",
+      "  \"conv_dim\": [\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512\n",
+      "  ],\n",
+      "  \"conv_kernel\": [\n",
+      "    10,\n",
+      "    3,\n",
+      "    3,\n",
+      "    3,\n",
+      "    3,\n",
+      "    2,\n",
+      "    2\n",
+      "  ],\n",
+      "  \"conv_stride\": [\n",
+      "    5,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2\n",
+      "  ],\n",
+      "  \"ctc_loss_reduction\": \"sum\",\n",
+      "  \"ctc_zero_infinity\": false,\n",
+      "  \"diversity_loss_weight\": 0.1,\n",
+      "  \"do_stable_layer_norm\": true,\n",
+      "  \"eos_token_id\": 2,\n",
+      "  \"feat_extract_activation\": \"gelu\",\n",
+      "  \"feat_extract_dropout\": 0.0,\n",
+      "  \"feat_extract_norm\": \"layer\",\n",
+      "  \"feat_proj_dropout\": 0.1,\n",
+      "  \"feat_quantizer_dropout\": 0.0,\n",
+      "  \"final_dropout\": 0.0,\n",
+      "  \"gradient_checkpointing\": false,\n",
+      "  \"hidden_act\": \"gelu\",\n",
+      "  \"hidden_dropout\": 0.1,\n",
+      "  \"hidden_size\": 1024,\n",
+      "  \"initializer_range\": 0.02,\n",
+      "  \"intermediate_size\": 4096,\n",
+      "  \"layer_norm_eps\": 1e-05,\n",
+      "  \"layerdrop\": 0.1,\n",
+      "  \"mask_feature_length\": 10,\n",
+      "  \"mask_feature_min_masks\": 0,\n",
+      "  \"mask_feature_prob\": 0.0,\n",
+      "  \"mask_time_length\": 10,\n",
+      "  \"mask_time_min_masks\": 2,\n",
+      "  \"mask_time_prob\": 0.075,\n",
+      "  \"model_type\": \"wav2vec2\",\n",
+      "  \"num_adapter_layers\": 3,\n",
+      "  \"num_attention_heads\": 16,\n",
+      "  \"num_codevector_groups\": 2,\n",
+      "  \"num_codevectors_per_group\": 320,\n",
+      "  \"num_conv_pos_embedding_groups\": 16,\n",
+      "  \"num_conv_pos_embeddings\": 128,\n",
+      "  \"num_feat_extract_layers\": 7,\n",
+      "  \"num_hidden_layers\": 24,\n",
+      "  \"num_negatives\": 100,\n",
+      "  \"output_hidden_size\": 1024,\n",
+      "  \"pad_token_id\": 0,\n",
+      "  \"proj_codevector_dim\": 768,\n",
+      "  \"tdnn_dilation\": [\n",
+      "    1,\n",
+      "    2,\n",
+      "    3,\n",
+      "    1,\n",
+      "    1\n",
+      "  ],\n",
+      "  \"tdnn_dim\": [\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    1500\n",
+      "  ],\n",
+      "  \"tdnn_kernel\": [\n",
+      "    5,\n",
+      "    3,\n",
+      "    3,\n",
+      "    1,\n",
+      "    1\n",
+      "  ],\n",
+      "  \"torch_dtype\": \"float32\",\n",
+      "  \"transformers_version\": \"4.16.2\",\n",
+      "  \"use_weighted_layer_sum\": false,\n",
+      "  \"vocab_size\": 32,\n",
+      "  \"xvector_output_dim\": 512\n",
+      "}\n",
+      "\n",
+      "100%|█████████████████████████████████████████████| 1/1 [00:00<00:00,  3.97ba/s]\n",
+      "100%|█████████████████████████████████████████████| 1/1 [00:00<00:00, 15.14ba/s]\n",
+      "{'j', 'y', 'x', 'l', 'm', 'ज़', 'क', 'ऐ', 'ऊ', 'त', 'ञ', 'p', 'n', 'u', 'ी', 'ऋ', 'ठ', 'छ', 'ा', 'क़', '&', 'c', 'ण', 'ढ़', 'w', 'अ', 'r', ' ', 'ष', 'ट', \"'\", 'ग़', 'f', 'k', 't', 'ृ', 'v', 'भ', 'g', 'स', 'ऑ', 'े', 'झ', 'z', 'ो', 'इ', '|', '।', 'ु', 'ड़', 'ए', 'h', 'ब', 'ध', 'ग', 'ः', 'i', 'श', 'औ', 'र', 'e', 'य', 'ड', 'प', 'ि', 's', 'म', 'b', 'ख', '़', 'ल', 'ई', 'उ', 'द', 'ज', 'ढ', 'ओ', 'ॉ', 'ं', 'च', 'a', 'न', 'ै', 'घ', 'थ', 'o', 'फ', 'ँ', 'आ', 'd', 'ौ', 'ॅ', 'व', 'ू', '्', 'ह'} 96 <class 'set'> {' ': 0, '&': 1, \"'\": 2, 'a': 3, 'b': 4, 'c': 5, 'd': 6, 'e': 7, 'f': 8, 'g': 9, 'h': 10, 'i': 11, 'j': 12, 'k': 13, 'l': 14, 'm': 15, 'n': 16, 'o': 17, 'p': 18, 'r': 19, 's': 20, 't': 21, 'u': 22, 'v': 23, 'w': 24, 'x': 25, 'y': 26, 'z': 27, '|': 28, 'ँ': 29, 'ं': 30, 'ः': 31, 'अ': 32, 'आ': 33, 'इ': 34, 'ई': 35, 'उ': 36, 'ऊ': 37, 'ऋ': 38, 'ए': 39, 'ऐ': 40, 'ऑ': 41, 'ओ': 42, 'औ': 43, 'क': 44, 'ख': 45, 'ग': 46, 'घ': 47, 'च': 48, 'छ': 49, 'ज': 50, 'झ': 51, 'ञ': 52, 'ट': 53, 'ठ': 54, 'ड': 55, 'ढ': 56, 'ण': 57, 'त': 58, 'थ': 59, 'द': 60, 'ध': 61, 'न': 62, 'प': 63, 'फ': 64, 'ब': 65, 'भ': 66, 'म': 67, 'य': 68, 'र': 69, 'ल': 70, 'व': 71, 'श': 72, 'ष': 73, 'स': 74, 'ह': 75, '़': 76, 'ा': 77, 'ि': 78, 'ी': 79, 'ु': 80, 'ू': 81, 'ृ': 82, 'ॅ': 83, 'े': 84, 'ै': 85, 'ॉ': 86, 'ो': 87, 'ौ': 88, '्': 89, 'क़': 90, 'ग़': 91, 'ज़': 92, 'ड़': 93, 'ढ़': 94, '।': 95}\n",
+      "{'&': 1, \"'\": 2, 'a': 3, 'b': 4, 'c': 5, 'd': 6, 'e': 7, 'f': 8, 'g': 9, 'h': 10, 'i': 11, 'j': 12, 'k': 13, 'l': 14, 'm': 15, 'n': 16, 'o': 17, 'p': 18, 'r': 19, 's': 20, 't': 21, 'u': 22, 'v': 23, 'w': 24, 'x': 25, 'y': 26, 'z': 27, '|': 28, 'ँ': 29, 'ं': 30, 'ः': 31, 'अ': 32, 'आ': 33, 'इ': 34, 'ई': 35, 'उ': 36, 'ऊ': 37, 'ऋ': 38, 'ए': 39, 'ऐ': 40, 'ऑ': 41, 'ओ': 42, 'औ': 43, 'क': 44, 'ख': 45, 'ग': 46, 'घ': 47, 'च': 48, 'छ': 49, 'ज': 50, 'झ': 51, 'ञ': 52, 'ट': 53, 'ठ': 54, 'ड': 55, 'ढ': 56, 'ण': 57, 'त': 58, 'थ': 59, 'द': 60, 'ध': 61, 'न': 62, 'प': 63, 'फ': 64, 'ब': 65, 'भ': 66, 'म': 67, 'य': 68, 'र': 69, 'ल': 70, 'व': 71, 'श': 72, 'ष': 73, 'स': 74, 'ह': 75, '़': 76, 'ा': 77, 'ि': 78, 'ी': 79, 'ु': 80, 'ू': 81, 'ृ': 82, 'ॅ': 83, 'े': 84, 'ै': 85, 'ॉ': 86, 'ो': 87, 'ौ': 88, '्': 89, 'क़': 90, 'ग़': 91, 'ज़': 92, 'ड़': 93, 'ढ़': 94, '।': 95, '$': 0, '[UNK]': 96, '[PAD]': 97}\n",
+      "Didn't find file ./tokenizer_config.json. We won't load it.\n",
+      "Didn't find file ./added_tokens.json. We won't load it.\n",
+      "Didn't find file ./special_tokens_map.json. We won't load it.\n",
+      "Didn't find file ./tokenizer.json. We won't load it.\n",
+      "loading file ./vocab.json\n",
+      "loading file None\n",
+      "loading file None\n",
+      "loading file None\n",
+      "loading file None\n",
+      "file ./config.json not found\n",
+      "Adding <s> to the vocabulary\n",
+      "Adding </s> to the vocabulary\n",
+      "Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.\n",
+      "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",
+      "Model config Wav2Vec2Config {\n",
+      "  \"_name_or_path\": \"facebook/wav2vec2-xls-r-300m\",\n",
+      "  \"activation_dropout\": 0.0,\n",
+      "  \"adapter_kernel_size\": 3,\n",
+      "  \"adapter_stride\": 2,\n",
+      "  \"add_adapter\": false,\n",
+      "  \"apply_spec_augment\": true,\n",
+      "  \"architectures\": [\n",
+      "    \"Wav2Vec2ForPreTraining\"\n",
+      "  ],\n",
+      "  \"attention_dropout\": 0.1,\n",
+      "  \"bos_token_id\": 1,\n",
+      "  \"classifier_proj_size\": 256,\n",
+      "  \"codevector_dim\": 768,\n",
+      "  \"contrastive_logits_temperature\": 0.1,\n",
+      "  \"conv_bias\": true,\n",
+      "  \"conv_dim\": [\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512\n",
+      "  ],\n",
+      "  \"conv_kernel\": [\n",
+      "    10,\n",
+      "    3,\n",
+      "    3,\n",
+      "    3,\n",
+      "    3,\n",
+      "    2,\n",
+      "    2\n",
+      "  ],\n",
+      "  \"conv_stride\": [\n",
+      "    5,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2,\n",
+      "    2\n",
+      "  ],\n",
+      "  \"ctc_loss_reduction\": \"sum\",\n",
+      "  \"ctc_zero_infinity\": false,\n",
+      "  \"diversity_loss_weight\": 0.1,\n",
+      "  \"do_stable_layer_norm\": true,\n",
+      "  \"eos_token_id\": 2,\n",
+      "  \"feat_extract_activation\": \"gelu\",\n",
+      "  \"feat_extract_dropout\": 0.0,\n",
+      "  \"feat_extract_norm\": \"layer\",\n",
+      "  \"feat_proj_dropout\": 0.1,\n",
+      "  \"feat_quantizer_dropout\": 0.0,\n",
+      "  \"final_dropout\": 0.0,\n",
+      "  \"gradient_checkpointing\": false,\n",
+      "  \"hidden_act\": \"gelu\",\n",
+      "  \"hidden_dropout\": 0.1,\n",
+      "  \"hidden_size\": 1024,\n",
+      "  \"initializer_range\": 0.02,\n",
+      "  \"intermediate_size\": 4096,\n",
+      "  \"layer_norm_eps\": 1e-05,\n",
+      "  \"layerdrop\": 0.1,\n",
+      "  \"mask_feature_length\": 10,\n",
+      "  \"mask_feature_min_masks\": 0,\n",
+      "  \"mask_feature_prob\": 0.0,\n",
+      "  \"mask_time_length\": 10,\n",
+      "  \"mask_time_min_masks\": 2,\n",
+      "  \"mask_time_prob\": 0.075,\n",
+      "  \"model_type\": \"wav2vec2\",\n",
+      "  \"num_adapter_layers\": 3,\n",
+      "  \"num_attention_heads\": 16,\n",
+      "  \"num_codevector_groups\": 2,\n",
+      "  \"num_codevectors_per_group\": 320,\n",
+      "  \"num_conv_pos_embedding_groups\": 16,\n",
+      "  \"num_conv_pos_embeddings\": 128,\n",
+      "  \"num_feat_extract_layers\": 7,\n",
+      "  \"num_hidden_layers\": 24,\n",
+      "  \"num_negatives\": 100,\n",
+      "  \"output_hidden_size\": 1024,\n",
+      "  \"pad_token_id\": 0,\n",
+      "  \"proj_codevector_dim\": 768,\n",
+      "  \"tdnn_dilation\": [\n",
+      "    1,\n",
+      "    2,\n",
+      "    3,\n",
+      "    1,\n",
+      "    1\n",
+      "  ],\n",
+      "  \"tdnn_dim\": [\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    512,\n",
+      "    1500\n",
+      "  ],\n",
+      "  \"tdnn_kernel\": [\n",
+      "    5,\n",
+      "    3,\n",
+      "    3,\n",
+      "    1,\n",
+      "    1\n",
+      "  ],\n",
+      "  \"torch_dtype\": \"float32\",\n",
+      "  \"transformers_version\": \"4.16.2\",\n",
+      "  \"use_weighted_layer_sum\": false,\n",
+      "  \"vocab_size\": 32,\n",
+      "  \"xvector_output_dim\": 512\n",
+      "}\n",
+      "\n",
+      "loading feature extractor configuration file https://huggingface.co/facebook/wav2vec2-xls-r-300m/resolve/main/preprocessor_config.json from cache at /workspace/.cache/huggingface/transformers/6fb028b95b394059e7d3b367bbca2382b576c66aebe896f04d2cd34e1b575f5b.d4484dc1c81456a2461485e7168b04347a7b9a4e3b1ef3aba723323b33e12326\n",
+      "Feature extractor Wav2Vec2FeatureExtractor {\n",
+      "  \"do_normalize\": true,\n",
+      "  \"feature_extractor_type\": \"Wav2Vec2FeatureExtractor\",\n",
+      "  \"feature_size\": 1,\n",
+      "  \"padding_side\": \"right\",\n",
+      "  \"padding_value\": 0,\n",
+      "  \"return_attention_mask\": true,\n",
+      "  \"sampling_rate\": 16000\n",
+      "}\n",
+      "\n",
+      "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",
+      "^C\n",
+      "Traceback (most recent call last):\n",
+      "  File \"run_speech_recognition_ctc.py\", line 765, in <module>\n",
+      "    main()\n",
+      "  File \"run_speech_recognition_ctc.py\", line 572, in main\n",
+      "    model = AutoModelForCTC.from_pretrained(\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/auto/auto_factory.py\", line 447, in from_pretrained\n",
+      "    return model_class.from_pretrained(pretrained_model_name_or_path, *model_args, config=config, **kwargs)\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/modeling_utils.py\", line 1489, in from_pretrained\n",
+      "    model = cls(config, *model_args, **model_kwargs)\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/wav2vec2/modeling_wav2vec2.py\", line 1680, in __init__\n",
+      "    self.wav2vec2 = Wav2Vec2Model(config)\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/wav2vec2/modeling_wav2vec2.py\", line 1234, in __init__\n",
+      "    self.encoder = Wav2Vec2EncoderStableLayerNorm(config)\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/wav2vec2/modeling_wav2vec2.py\", line 859, in __init__\n",
+      "    [Wav2Vec2EncoderLayerStableLayerNorm(config) for _ in range(config.num_hidden_layers)]\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/wav2vec2/modeling_wav2vec2.py\", line 859, in <listcomp>\n",
+      "    [Wav2Vec2EncoderLayerStableLayerNorm(config) for _ in range(config.num_hidden_layers)]\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/wav2vec2/modeling_wav2vec2.py\", line 737, in __init__\n",
+      "    self.attention = Wav2Vec2Attention(\n",
+      "  File \"/workspace/.local/lib/python3.8/site-packages/transformers/models/wav2vec2/modeling_wav2vec2.py\", line 559, in __init__\n",
+      "    self.q_proj = nn.Linear(embed_dim, embed_dim, bias=bias)\n",
+      "  File \"/opt/conda/lib/python3.8/site-packages/torch/nn/modules/linear.py\", line 90, in __init__\n",
+      "    self.reset_parameters()\n",
+      "  File \"/opt/conda/lib/python3.8/site-packages/torch/nn/modules/linear.py\", line 96, in reset_parameters\n",
+      "    init.kaiming_uniform_(self.weight, a=math.sqrt(5))\n",
+      "  File \"/opt/conda/lib/python3.8/site-packages/torch/nn/init.py\", line 395, in kaiming_uniform_\n",
+      "    return tensor.uniform_(-bound, bound)\n",
+      "KeyboardInterrupt\n"
+     ]
+    }
+   ],
+   "source": [
+    "!. run.sh"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5c2d3236",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.8.8"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

added_tokens.json

@@ -0,0 +1 @@
+{"<s>": 98, "</s>": 99}

config.json

@@ -0,0 +1,107 @@
+{
+  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
+  "activation_dropout": 0.0,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 768,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": true,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "mean",
+  "ctc_zero_infinity": false,
+  "diversity_loss_weight": 0.1,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "layer",
+  "feat_proj_dropout": 0.0,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.0,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
+  "num_attention_heads": 16,
+  "num_codevector_groups": 2,
+  "num_codevectors_per_group": 320,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 24,
+  "num_negatives": 100,
+  "output_hidden_size": 1024,
+  "pad_token_id": 97,
+  "proj_codevector_dim": 768,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.16.2",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 100,
+  "xvector_output_dim": 512
+}

preprocessor_config.json

@@ -0,0 +1,9 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0,
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b94df9afab6a29dde619f3dae5b5c27e9f1d4aba16828352b07ee3a2b43a008
+size 1262333681

run.sh

@@ -0,0 +1,27 @@
+python run_speech_recognition_ctc.py  \
+--dataset_name="mozilla-foundation/common_voice_7_0" \
+--model_name_or_path="facebook/wav2vec2-xls-r-300m" \
+--dataset_config_name="hi" \
+--output_dir="./" \
+--overwrite_output_dir \
+--max_steps="8000" \
+--per_device_train_batch_size="16" \
+--learning_rate="3e-4" \
+ --warmup_steps="500" \
+ --save_steps="200" \
+ --eval_steps="400" \
+--save_total_limit="3" \
+--evaluation_strategy="steps" \
+--text_column_name="sentence" \
+--length_column_name="input_length"  \
+--layerdrop="0.0"  \
+--freeze_feature_encoder \
+--gradient_checkpointing  \
+--fp16 \
+--group_by_length \
+--push_to_hub \
+--use_auth_token \
+--do_train \
+--do_eval \
+--chars_to_ignore , ? . ! - \; \: \" “ % ‘ ” � \
+

run_speech_recognition_ctc.py

@@ -0,0 +1,765 @@
+#!/usr/bin/env python
+# coding=utf-8
+# Copyright 2021 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http:a
+# //www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+
+""" Fine-tuning a 🤗 Transformers CTC model for automatic speech recognition"""
+
+import functools
+import json
+import logging
+import os
+import re
+import sys
+import warnings
+from dataclasses import dataclass, field
+from typing import Dict, List, Optional, Union
+
+import datasets
+import numpy as np
+import torch
+from datasets import DatasetDict, load_dataset, load_metric
+
+import transformers
+from transformers import (
+    AutoConfig,
+    AutoFeatureExtractor,
+    AutoModelForCTC,
+    AutoProcessor,
+    AutoTokenizer,
+    HfArgumentParser,
+    Trainer,
+    TrainingArguments,
+    Wav2Vec2Processor,
+    set_seed,
+)
+from transformers.trainer_utils import get_last_checkpoint, is_main_process
+from transformers.utils import check_min_version
+from transformers.utils.versions import require_version
+
+
+# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
+check_min_version("4.16.0.dev0")
+
+require_version("datasets>=1.13.3", "To fix: pip install -r examples/pytorch/text-classification/requirements.txt")
+
+
+logger = logging.getLogger(__name__)
+
+
+def list_field(default=None, metadata=None):
+    return field(default_factory=lambda: default, metadata=metadata)
+
+
+@dataclass
+class ModelArguments:
+    """
+    Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.
+    """
+
+    model_name_or_path: str = field(
+        metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"}
+    )
+    tokenizer_name_or_path: Optional[str] = field(
+        default=None,
+        metadata={"help": "Path to pretrained tokenizer or tokenizer identifier from huggingface.co/models"},
+    )
+    cache_dir: Optional[str] = field(
+        default=None,
+        metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
+    )
+    freeze_feature_encoder: bool = field(
+        default=True, metadata={"help": "Whether to freeze the feature encoder layers of the model."}
+    )
+    attention_dropout: float = field(
+        default=0.0, metadata={"help": "The dropout ratio for the attention probabilities."}
+    )
+    activation_dropout: float = field(
+        default=0.0, metadata={"help": "The dropout ratio for activations inside the fully connected layer."}
+    )
+    feat_proj_dropout: float = field(default=0.0, metadata={"help": "The dropout ratio for the projected features."})
+    hidden_dropout: float = field(
+        default=0.0,
+        metadata={
+            "help": "The dropout probability for all fully connected layers in the embeddings, encoder, and pooler."
+        },
+    )
+    final_dropout: float = field(
+        default=0.0,
+        metadata={"help": "The dropout probability for the final projection layer."},
+    )
+    mask_time_prob: float = field(
+        default=0.05,
+        metadata={
+            "help": "Probability of each feature vector along the time axis to be chosen as the start of the vector"
+            "span to be masked. Approximately ``mask_time_prob * sequence_length // mask_time_length`` feature"
+            "vectors will be masked along the time axis."
+        },
+    )
+    mask_time_length: int = field(
+        default=10,
+        metadata={"help": "Length of vector span to mask along the time axis."},
+    )
+    mask_feature_prob: float = field(
+        default=0.0,
+        metadata={
+            "help": "Probability of each feature vector along the feature axis to be chosen as the start of the vector"
+            "span to be masked. Approximately ``mask_feature_prob * sequence_length // mask_feature_length`` feature bins will be masked along the time axis."
+        },
+    )
+    mask_feature_length: int = field(
+        default=10,
+        metadata={"help": "Length of vector span to mask along the feature axis."},
+    )
+    layerdrop: float = field(default=0.0, metadata={"help": "The LayerDrop probability."})
+    ctc_loss_reduction: Optional[str] = field(
+        default="mean", metadata={"help": "The way the ctc loss should be reduced. Should be one of 'mean' or 'sum'."}
+    )
+
+
+@dataclass
+class DataTrainingArguments:
+    """
+    Arguments pertaining to what data we are going to input our model for training and eval.
+
+    Using `HfArgumentParser` we can turn this class
+    into argparse arguments to be able to specify them on
+    the command line.
+    """
+
+    dataset_name: str = field(
+        metadata={"help": "The configuration name of the dataset to use (via the datasets library)."}
+    )
+    dataset_config_name: str = field(
+        default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."}
+    )
+    train_split_name: str = field(
+        default="train+validation",
+        metadata={
+            "help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
+        },
+    )
+    eval_split_name: str = field(
+        default="test",
+        metadata={
+            "help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
+        },
+    )
+    audio_column_name: str = field(
+        default="audio",
+        metadata={"help": "The name of the dataset column containing the audio data. Defaults to 'audio'"},
+    )
+    text_column_name: str = field(
+        default="text",
+        metadata={"help": "The name of the dataset column containing the text data. Defaults to 'text'"},
+    )
+    overwrite_cache: bool = field(
+        default=False, metadata={"help": "Overwrite the cached preprocessed datasets or not."}
+    )
+    preprocessing_num_workers: Optional[int] = field(
+        default=None,
+        metadata={"help": "The number of processes to use for the preprocessing."},
+    )
+    max_train_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of training examples to this "
+            "value if set."
+        },
+    )
+    max_eval_samples: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "For debugging purposes or quicker training, truncate the number of validation examples to this "
+            "value if set."
+        },
+    )
+    chars_to_ignore: Optional[List[str]] = list_field(
+        default=None,
+        metadata={"help": "A list of characters to remove from the transcripts."},
+    )
+    eval_metrics: List[str] = list_field(
+        default=["wer"],
+        metadata={"help": "A list of metrics the model should be evaluated on. E.g. `'wer cer'`"},
+    )
+    max_duration_in_seconds: float = field(
+        default=20.0,
+        metadata={
+            "help": "Filter audio files that are longer than `max_duration_in_seconds` seconds to 'max_duration_in_seconds`"
+        },
+    )
+    min_duration_in_seconds: float = field(
+        default=0.0, metadata={"help": "Filter audio files that are shorter than `min_duration_in_seconds` seconds"}
+    )
+    preprocessing_only: bool = field(
+        default=False,
+        metadata={
+            "help": "Whether to only do data preprocessing and skip training. "
+            "This is especially useful when data preprocessing errors out in distributed training due to timeout. "
+            "In this case, one should run the preprocessing in a non-distributed setup with `preprocessing_only=True` "
+            "so that the cached datasets can consequently be loaded in distributed training"
+        },
+    )
+    use_auth_token: bool = field(
+        default=False,
+        metadata={
+            "help": "If :obj:`True`, will use the token generated when running"
+            ":obj:`transformers-cli login` as HTTP bearer authorization for remote files."
+        },
+    )
+    unk_token: str = field(
+        default="[UNK]",
+        metadata={"help": "The unk token for the tokenizer"},
+    )
+    pad_token: str = field(
+        default="[PAD]",
+        metadata={"help": "The padding token for the tokenizer"},
+    )
+    word_delimiter_token: str = field(
+        default="$",
+        metadata={"help": "The word delimiter token for the tokenizer"},
+    )
+    phoneme_language: Optional[str] = field(
+        default=None,
+        metadata={
+            "help": "The target language that should be used be"
+            " passed to the tokenizer for tokenization. Note that"
+            " this is only relevant if the model classifies the"
+            " input audio to a sequence of phoneme sequences."
+        },
+    )
+
+
+@dataclass
+class DataCollatorCTCWithPadding:
+    """
+    Data collator that will dynamically pad the inputs received.
+    Args:
+        processor (:class:`~transformers.AutoProcessor`)
+            The processor used for proccessing the data.
+        padding (:obj:`bool`, :obj:`str` or :class:`~transformers.tokenization_utils_base.PaddingStrategy`, `optional`, defaults to :obj:`True`):
+            Select a strategy to pad the returned sequences (according to the model's padding side and padding index)
+            among:
+            * :obj:`True` or :obj:`'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
+              sequence if provided).
+            * :obj:`'max_length'`: Pad to a maximum length specified with the argument :obj:`max_length` or to the
+              maximum acceptable input length for the model if that argument is not provided.
+            * :obj:`False` or :obj:`'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of
+              different lengths).
+        max_length (:obj:`int`, `optional`):
+            Maximum length of the ``input_values`` of the returned list and optionally padding length (see above).
+        max_length_labels (:obj:`int`, `optional`):
+            Maximum length of the ``labels`` returned list and optionally padding length (see above).
+        pad_to_multiple_of (:obj:`int`, `optional`):
+            If set will pad the sequence to a multiple of the provided value.
+            This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >=
+            7.5 (Volta).
+    """
+
+    processor: AutoProcessor
+    padding: Union[bool, str] = "longest"
+    pad_to_multiple_of: Optional[int] = None
+    pad_to_multiple_of_labels: Optional[int] = None
+
+    def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:
+        # split inputs and labels since they have to be of different lenghts and need
+        # different padding methods
+        input_features = [{"input_values": feature["input_values"]} for feature in features]
+        label_features = [{"input_ids": feature["labels"]} for feature in features]
+
+        batch = self.processor.pad(
+            input_features,
+            padding=self.padding,
+            pad_to_multiple_of=self.pad_to_multiple_of,
+            return_tensors="pt",
+        )
+
+        with self.processor.as_target_processor():
+            labels_batch = self.processor.pad(
+                label_features,
+                padding=self.padding,
+                pad_to_multiple_of=self.pad_to_multiple_of_labels,
+                return_tensors="pt",
+            )
+
+        # replace padding with -100 to ignore loss correctly
+        labels = labels_batch["input_ids"].masked_fill(labels_batch.attention_mask.ne(1), -100)
+
+        batch["labels"] = labels
+
+        return batch
+
+
+def create_vocabulary_from_data(
+    datasets: DatasetDict,
+    word_delimiter_token: Optional[str] = None,
+    unk_token: Optional[str] = None,
+    pad_token: Optional[str] = None,
+):
+    # Given training and test labels create vocabulary
+    def extract_all_chars(batch):
+        all_text = " ".join(batch["target_text"])
+        vocab = list(set(all_text))
+        return {"vocab": [vocab], "all_text": [all_text]}
+
+    vocabs = datasets.map(
+        extract_all_chars,
+        batched=True,
+        batch_size=-1,
+        keep_in_memory=True,
+        remove_columns=datasets["train"].column_names,
+    )
+
+    # take union of all unique characters in each dataset
+    vocab_set = functools.reduce(
+        lambda vocab_1, vocab_2: set(vocab_1["vocab"][0]) | set(vocab_2["vocab"][0]), vocabs.values()
+    )
+    
+    vocab_dict = {v: k for k, v in enumerate(sorted(list(set(vocab_set))))}
+    print(vocab_set,len(vocab_set),type(vocab_set),vocab_dict)
+    # replace white space with delimiter token
+    if word_delimiter_token is not None:
+        vocab_dict[word_delimiter_token] = vocab_dict[" "]
+        del vocab_dict[" "]
+
+    # add unk and pad token
+    if unk_token is not None:
+        vocab_dict[unk_token] = len(vocab_dict)
+
+    if pad_token is not None:
+        vocab_dict[pad_token] = len(vocab_dict)
+        
+    print(vocab_dict)
+    return vocab_dict
+
+
+def main():
+    # See all possible arguments in src/transformers/training_args.py
+    # or by passing the --help flag to this script.
+    # We now keep distinct sets of args, for a cleaner separation of concerns.
+
+    parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TrainingArguments))
+    if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
+        # If we pass only one argument to the script and it's the path to a json file,
+        # let's parse it to get our arguments.
+        model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
+    else:
+        model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+
+    # Detecting last checkpoint.
+    last_checkpoint = None
+    if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
+        last_checkpoint = get_last_checkpoint(training_args.output_dir)
+        if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
+            raise ValueError(
+                f"Output directory ({training_args.output_dir}) already exists and is not empty. "
+                "Use --overwrite_output_dir to overcome."
+            )
+        elif last_checkpoint is not None:
+            logger.info(
+                f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
+                "the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
+            )
+
+    # Setup logging
+    logging.basicConfig(
+        format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
+        datefmt="%m/%d/%Y %H:%M:%S",
+        handlers=[logging.StreamHandler(sys.stdout)],
+    )
+    logger.setLevel(logging.INFO if is_main_process(training_args.local_rank) else logging.WARN)
+
+    # Log on each process the small summary:
+    logger.warning(
+        f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
+        f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
+    )
+    # Set the verbosity to info of the Transformers logger (on main process only):
+    if is_main_process(training_args.local_rank):
+        transformers.utils.logging.set_verbosity_info()
+    logger.info("Training/evaluation parameters %s", training_args)
+
+    # Set seed before initializing model.
+    set_seed(training_args.seed)
+
+    # 1. First, let's load the dataset
+    raw_datasets = DatasetDict()
+
+    if training_args.do_train:
+        raw_datasets["train"] = load_dataset(
+            data_args.dataset_name,
+            data_args.dataset_config_name,
+            split=data_args.train_split_name,
+            use_auth_token=data_args.use_auth_token,
+        )
+
+        if data_args.audio_column_name not in raw_datasets["train"].column_names:
+            raise ValueError(
+                f"--audio_column_name '{data_args.audio_column_name}' not found in dataset '{data_args.dataset_name}'. "
+                "Make sure to set `--audio_column_name` to the correct audio column - one of "
+                f"{', '.join(raw_datasets['train'].column_names)}."
+            )
+
+        if data_args.text_column_name not in raw_datasets["train"].column_names:
+            raise ValueError(
+                f"--text_column_name {data_args.text_column_name} not found in dataset '{data_args.dataset_name}'. "
+                "Make sure to set `--text_column_name` to the correct text column - one of "
+                f"{', '.join(raw_datasets['train'].column_names)}."
+            )
+
+        if data_args.max_train_samples is not None:
+            raw_datasets["train"] = raw_datasets["train"].select(range(data_args.max_train_samples))
+
+    if training_args.do_eval:
+        raw_datasets["eval"] = load_dataset(
+            data_args.dataset_name,
+            data_args.dataset_config_name,
+            split=data_args.eval_split_name,
+            use_auth_token=data_args.use_auth_token,
+        )
+
+        if data_args.max_eval_samples is not None:
+            raw_datasets["eval"] = raw_datasets["eval"].select(range(data_args.max_eval_samples))
+
+    # 2. We remove some special characters from the datasets
+    # that make training complicated and do not help in transcribing the speech
+    # E.g. characters, such as `,` and `.` do not really have an acoustic characteristic
+    # that could be easily picked up by the model
+#     chars_to_ignore_regex = (
+#         f'[{"".join(data_args.chars_to_ignore)}]' if data_args.chars_to_ignore is not None else None
+#     )
+    chars_to_ignore_regex = '[,?.!\-\;\:"“%‘”�—’…–]'
+    text_column_name = data_args.text_column_name
+
+    
+    def replace_text(text):
+        text=text.replace('„', r'"')
+        text=text.replace('“', r'"')
+        text=text.replace('”', r'"')
+        text=text.replace('–', r'-')
+        text=text.replace('—', r' - ')
+        text=text.replace('´', r"'")
+        text=text.replace('‘', r"'")
+        text=text.replace('‚', r"'")
+        text=text.replace('’', r"'")
+        text=text.replace("''", r'"')
+        text=text.replace('´´', r'"')
+
+        token_sequences_to_ignore = ["\n\n", "\n", "   ", "  "]
+        for t in token_sequences_to_ignore:
+            text = " ".join(text.split(t))
+        return text
+    
+    def remove_special_characters(batch):
+        text=batch[text_column_name]
+        text=replace_text(text)
+    
+        
+        if chars_to_ignore_regex is not None:
+            target_text = re.sub(chars_to_ignore_regex, "", text).lower() + " "
+        else:
+            target_text = text.lower() + " "
+            
+        batch["target_text"]=target_text
+        return batch
+
+    with training_args.main_process_first(desc="dataset map special characters removal"):
+        raw_datasets = raw_datasets.map(
+            remove_special_characters,
+            remove_columns=[text_column_name],
+            desc="remove special characters from datasets",
+        )
+
+    # save special tokens for tokenizer
+    word_delimiter_token = data_args.word_delimiter_token
+    unk_token = data_args.unk_token
+    pad_token = data_args.pad_token
+
+    # 3. Next, let's load the config as we might need it to create
+    # the tokenizer
+    # load config
+    config = AutoConfig.from_pretrained(
+        model_args.model_name_or_path, cache_dir=model_args.cache_dir, use_auth_token=data_args.use_auth_token
+    )
+
+    # 4. Next, if no tokenizer file is defined,
+    # we create the vocabulary of the model by extracting all unique characters from
+    # the training and evaluation datasets
+    # We need to make sure that only first rank saves vocabulary
+    # make sure all processes wait until vocab is created
+    tokenizer_name_or_path = model_args.tokenizer_name_or_path
+    tokenizer_kwargs = {}
+    if tokenizer_name_or_path is None:
+        # save vocab in training output dir
+        tokenizer_name_or_path = training_args.output_dir
+
+        vocab_file = os.path.join(tokenizer_name_or_path, "vocab.json")
+
+        with training_args.main_process_first():
+            if training_args.overwrite_output_dir and os.path.isfile(vocab_file):
+                os.remove(vocab_file)
+
+        with training_args.main_process_first(desc="dataset map vocabulary creation"):
+            if not os.path.isfile(vocab_file):
+                os.makedirs(tokenizer_name_or_path, exist_ok=True)
+                vocab_dict = create_vocabulary_from_data(
+                    raw_datasets,
+                    word_delimiter_token=word_delimiter_token,
+                    unk_token=unk_token,
+                    pad_token=pad_token,
+                )
+
+                # save vocab dict to be loaded into tokenizer
+                with open(vocab_file, "w") as file:
+                    json.dump(vocab_dict, file)
+
+        # if tokenizer has just been created
+        # it is defined by `tokenizer_class` if present in config else by `model_type`
+        tokenizer_kwargs = {
+            "config": config if config.tokenizer_class is not None else None,
+            "tokenizer_type": config.model_type if config.tokenizer_class is None else None,
+            "unk_token": unk_token,
+            "pad_token": pad_token,
+            "word_delimiter_token": word_delimiter_token,
+        }
+
+    # 5. Now we can instantiate the feature extractor, tokenizer and model
+    # Note for distributed training, the .from_pretrained methods guarantee that only
+    # one local process can concurrently download model & vocab.
+
+    # load feature_extractor and tokenizer
+    tokenizer = AutoTokenizer.from_pretrained(
+        tokenizer_name_or_path,
+        use_auth_token=data_args.use_auth_token,
+        **tokenizer_kwargs,
+    )
+    feature_extractor = AutoFeatureExtractor.from_pretrained(
+        model_args.model_name_or_path, cache_dir=model_args.cache_dir, use_auth_token=data_args.use_auth_token
+    )
+
+    # adapt config
+    config.update(
+        {
+            "feat_proj_dropout": model_args.feat_proj_dropout,
+            "attention_dropout": model_args.attention_dropout,
+            "hidden_dropout": model_args.hidden_dropout,
+            "final_dropout": model_args.final_dropout,
+            "mask_time_prob": model_args.mask_time_prob,
+            "mask_time_length": model_args.mask_time_length,
+            "mask_feature_prob": model_args.mask_feature_prob,
+            "mask_feature_length": model_args.mask_feature_length,
+            "gradient_checkpointing": training_args.gradient_checkpointing,
+            "layerdrop": model_args.layerdrop,
+            "ctc_loss_reduction": model_args.ctc_loss_reduction,
+            "pad_token_id": tokenizer.pad_token_id,
+            "vocab_size": len(tokenizer),
+            "activation_dropout": model_args.activation_dropout,
+        }
+    )
+
+    # create model
+    model = AutoModelForCTC.from_pretrained(
+        model_args.model_name_or_path,
+        cache_dir=model_args.cache_dir,
+        config=config,
+        use_auth_token=data_args.use_auth_token,
+    )
+
+    # freeze encoder
+    if model_args.freeze_feature_encoder:
+        model.freeze_feature_encoder()
+
+    # 6. Now we preprocess the datasets including loading the audio, resampling and normalization
+    # Thankfully, `datasets` takes care of automatically loading and resampling the audio,
+    # so that we just need to set the correct target sampling rate and normalize the input
+    # via the `feature_extractor`
+
+    # make sure that dataset decodes audio with correct sampling rate
+    dataset_sampling_rate = next(iter(raw_datasets.values())).features[data_args.audio_column_name].sampling_rate
+    if dataset_sampling_rate != feature_extractor.sampling_rate:
+        raw_datasets = raw_datasets.cast_column(
+            data_args.audio_column_name, datasets.features.Audio(sampling_rate=feature_extractor.sampling_rate)
+        )
+
+    # derive max & min input length for sample rate & max duration
+    max_input_length = data_args.max_duration_in_seconds * feature_extractor.sampling_rate
+    min_input_length = data_args.min_duration_in_seconds * feature_extractor.sampling_rate
+    audio_column_name = data_args.audio_column_name
+    num_workers = data_args.preprocessing_num_workers
+
+    # `phoneme_language` is only relevant if the model is fine-tuned on phoneme classification
+    phoneme_language = data_args.phoneme_language
+
+    # Preprocessing the datasets.
+    # We need to read the audio files as arrays and tokenize the targets.
+    def prepare_dataset(batch):
+        # load audio
+        sample = batch[audio_column_name]
+
+        inputs = feature_extractor(sample["array"], sampling_rate=sample["sampling_rate"])
+        batch["input_values"] = inputs.input_values[0]
+        batch["input_length"] = len(batch["input_values"])
+
+        # encode targets
+        additional_kwargs = {}
+        if phoneme_language is not None:
+            additional_kwargs["phonemizer_lang"] = phoneme_language
+
+        batch["labels"] = tokenizer(batch["target_text"], **additional_kwargs).input_ids
+        return batch
+
+    with training_args.main_process_first(desc="dataset map preprocessing"):
+        vectorized_datasets = raw_datasets.map(
+            prepare_dataset,
+            remove_columns=next(iter(raw_datasets.values())).column_names,
+            num_proc=num_workers,
+            desc="preprocess datasets",
+        )
+
+        def is_audio_in_length_range(length):
+            return length > min_input_length and length < max_input_length
+
+        # filter data that is shorter than min_input_length
+        vectorized_datasets = vectorized_datasets.filter(
+            is_audio_in_length_range,
+            num_proc=num_workers,
+            input_columns=["input_length"],
+        )
+
+    # 7. Next, we can prepare the training.
+    # Let's use word error rate (WER) as our evaluation metric,
+    # instantiate a data collator and the trainer
+
+    # Define evaluation metrics during training, *i.e.* word error rate, character error rate
+    eval_metrics = {metric: load_metric(metric) for metric in data_args.eval_metrics}
+
+    # for large datasets it is advised to run the preprocessing on a
+    # single machine first with ``args.preprocessing_only`` since there will mostly likely
+    # be a timeout when running the script in distributed mode.
+    # In a second step ``args.preprocessing_only`` can then be set to `False` to load the
+    # cached dataset
+    if data_args.preprocessing_only:
+        logger.info(f"Data preprocessing finished. Files cached at {vectorized_datasets.cache_files}")
+        return
+
+    def compute_metrics(pred):
+        pred_logits = pred.predictions
+        pred_ids = np.argmax(pred_logits, axis=-1)
+
+        pred.label_ids[pred.label_ids == -100] = tokenizer.pad_token_id
+
+        pred_str = tokenizer.batch_decode(pred_ids)
+        # we do not want to group tokens when computing the metrics
+        label_str = tokenizer.batch_decode(pred.label_ids, group_tokens=False)
+
+        metrics = {k: v.compute(predictions=pred_str, references=label_str) for k, v in eval_metrics.items()}
+
+        return metrics
+
+    # Now save everything to be able to create a single processor later
+    if is_main_process(training_args.local_rank):
+        # save feature extractor, tokenizer and config
+        feature_extractor.save_pretrained(training_args.output_dir)
+        tokenizer.save_pretrained(training_args.output_dir)
+        config.save_pretrained(training_args.output_dir)
+
+    try:
+        processor = AutoProcessor.from_pretrained(training_args.output_dir)
+    except (OSError, KeyError):
+        warnings.warn(
+            "Loading a processor from a feature extractor config that does not"
+            " include a `processor_class` attribute is deprecated and will be removed in v5. Please add the following "
+            " attribute to your `preprocessor_config.json` file to suppress this warning: "
+            " `'processor_class': 'Wav2Vec2Processor'`",
+            FutureWarning,
+        )
+        processor = Wav2Vec2Processor.from_pretrained(training_args.output_dir)
+
+    # Instantiate custom data collator
+    data_collator = DataCollatorCTCWithPadding(processor=processor)
+
+    # Initialize Trainer
+    trainer = Trainer(
+        model=model,
+        data_collator=data_collator,
+        args=training_args,
+        compute_metrics=compute_metrics,
+        train_dataset=vectorized_datasets["train"] if training_args.do_train else None,
+        eval_dataset=vectorized_datasets["eval"] if training_args.do_eval else None,
+        tokenizer=feature_extractor,
+    )
+
+    # 8. Finally, we can start training
+
+    # Training
+    if training_args.do_train:
+
+        # use last checkpoint if exist
+        if last_checkpoint is not None:
+            checkpoint = last_checkpoint
+        elif os.path.isdir(model_args.model_name_or_path):
+            checkpoint = model_args.model_name_or_path
+        else:
+            checkpoint = None
+
+        train_result = trainer.train(resume_from_checkpoint=checkpoint)
+        trainer.save_model()
+
+        metrics = train_result.metrics
+        max_train_samples = (
+            data_args.max_train_samples
+            if data_args.max_train_samples is not None
+            else len(vectorized_datasets["train"])
+        )
+        metrics["train_samples"] = min(max_train_samples, len(vectorized_datasets["train"]))
+
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
+
+    # Evaluation
+    results = {}
+    if training_args.do_eval:
+        logger.info("*** Evaluate ***")
+        metrics = trainer.evaluate()
+        max_eval_samples = (
+            data_args.max_eval_samples if data_args.max_eval_samples is not None else len(vectorized_datasets["eval"])
+        )
+        metrics["eval_samples"] = min(max_eval_samples, len(vectorized_datasets["eval"]))
+
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
+
+    # Write model card and (optionally) push to hub
+    config_name = data_args.dataset_config_name if data_args.dataset_config_name is not None else "na"
+    kwargs = {
+        "finetuned_from": model_args.model_name_or_path,
+        "tasks": "speech-recognition",
+        "tags": ["automatic-speech-recognition", data_args.dataset_name,"robust-speech-event"],
+        "dataset_args": f"Config: {config_name}, Training split: {data_args.train_split_name}, Eval split: {data_args.eval_split_name}",
+        "dataset": f"{data_args.dataset_name.upper()} - {config_name.upper()}",
+    }
+    if "common_voice" in data_args.dataset_name:
+        kwargs["language"] = config_name
+
+    if training_args.push_to_hub:
+        trainer.push_to_hub(**kwargs)
+    else:
+        trainer.create_model_card(**kwargs)
+
+    return results
+
+
+if __name__ == "__main__":
+    main()

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "$", "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "./", "tokenizer_class": "Wav2Vec2CTCTokenizer"}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b14a8f9f8277105473cbf033c9a27c340ba9432e25e5aaa2aef3d10c81d85e9
+size 2991

vocab.json

@@ -0,0 +1 @@
+{"&": 1, "'": 2, "a": 3, "b": 4, "c": 5, "d": 6, "e": 7, "f": 8, "g": 9, "h": 10, "i": 11, "j": 12, "k": 13, "l": 14, "m": 15, "n": 16, "o": 17, "p": 18, "r": 19, "s": 20, "t": 21, "u": 22, "v": 23, "w": 24, "x": 25, "y": 26, "z": 27, "|": 28, "ँ": 29, "ं": 30, "ः": 31, "अ": 32, "आ": 33, "इ": 34, "ई": 35, "उ": 36, "ऊ": 37, "ऋ": 38, "ए": 39, "ऐ": 40, "ऑ": 41, "ओ": 42, "औ": 43, "क": 44, "ख": 45, "ग": 46, "घ": 47, "च": 48, "छ": 49, "ज": 50, "झ": 51, "ञ": 52, "ट": 53, "ठ": 54, "ड": 55, "ढ": 56, "ण": 57, "त": 58, "थ": 59, "द": 60, "ध": 61, "न": 62, "प": 63, "फ": 64, "ब": 65, "भ": 66, "म": 67, "य": 68, "र": 69, "ल": 70, "व": 71, "श": 72, "ष": 73, "स": 74, "ह": 75, "़": 76, "ा": 77, "ि": 78, "ी": 79, "ु": 80, "ू": 81, "ृ": 82, "ॅ": 83, "े": 84, "ै": 85, "ॉ": 86, "ो": 87, "ौ": 88, "्": 89, "क़": 90, "ग़": 91, "ज़": 92, "ड़": 93, "ढ़": 94, "।": 95, "$": 0, "[UNK]": 96, "[PAD]": 97}