Upload odia_training_script.ipynb

odia_training_script.ipynb

@@ -0,0 +1,1449 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# HuggingFace challenge - Debugger notebook\n",
+    "Run this notebook to verify your libraries versions, check GPU config and run a quick training"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {
+    "id": "T2utsYSKszvv"
+   },
+   "outputs": [],
+   "source": [
+    "import platform\n",
+    "import multiprocessing\n",
+    "\n",
+    "import torch\n",
+    "import transformers\n",
+    "import datasets\n",
+    "\n",
+    "import soundfile"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Print main infos"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "5P6I-W9ts-kR",
+    "outputId": "939bd550-1486-46a6-8371-e82ada0f448c"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Platform: Linux-5.11.0-37-generic-x86_64-with-glibc2.10\n",
+      "CPU cores: 60\n",
+      "Python version: 3.8.8\n",
+      "PyTorch version: 1.10.1+cu102\n",
+      "GPU is visible: True\n",
+      "Transformers version: 4.16.0.dev0\n",
+      "Datasets version: 1.17.1.dev0\n",
+      "soundfile version: 0.10.3\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(f\"Platform: {platform.platform()}\")\n",
+    "print(f\"CPU cores: {multiprocessing.cpu_count()}\")\n",
+    "\n",
+    "print(f\"Python version: {platform.python_version()}\")\n",
+    "\n",
+    "print(f\"PyTorch version: {torch.__version__}\")\n",
+    "print(f\"GPU is visible: {torch.cuda.is_available()}\")\n",
+    "\n",
+    "print(f\"Transformers version: {transformers.__version__}\")\n",
+    "print(f\"Datasets version: {datasets.__version__}\")\n",
+    "\n",
+    "print(f\"soundfile version: {soundfile.__version__}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Check your GPU informations (if any)\n",
+    "If you launched an AI Training job with GPU resources, they should be listed below (Tesla V100s 32GB).\n",
+    "Driver and CUDA version "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "YT7fRnKctggU",
+    "outputId": "f355a3e0-20da-489f-bd1f-5e508e792a68"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Fri Jan 21 03:07:38 2022       \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| NVIDIA-SMI 470.57.02    Driver Version: 470.57.02    CUDA Version: 11.4     |\n",
+      "|-------------------------------+----------------------+----------------------+\n",
+      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
+      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
+      "|                               |                      |               MIG M. |\n",
+      "|===============================+======================+======================|\n",
+      "|   0  Tesla V100S-PCI...  Off  | 00000000:00:06.0 Off |                    0 |\n",
+      "| N/A   35C    P0    26W / 250W |      4MiB / 32510MiB |      0%      Default |\n",
+      "|                               |                      |                  N/A |\n",
+      "+-------------------------------+----------------------+----------------------+\n",
+      "                                                                               \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| Processes:                                                                  |\n",
+      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
+      "|        ID   ID                                                   Usage      |\n",
+      "|=============================================================================|\n",
+      "|  No running processes found                                                 |\n",
+      "+-----------------------------------------------------------------------------+\n"
+     ]
+    }
+   ],
+   "source": [
+    "!nvidia-smi"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "1f72bffe678b4bdca366b35305baaab5",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "VBox(children=(HTML(value='<center>\\n<img src=https://huggingface.co/front/assets/huggingface_logo-noborder.sv…"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "from huggingface_hub import notebook_login\n",
+    "\n",
+    "notebook_login()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "id": "TorMtpwPv6RQ"
+   },
+   "source": [
+    "## Quick training run with a dummy model and data\n",
+    "more information on https://github.com/huggingface/transformers/tree/master/examples/pytorch/speech-recognition"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "fevoJD15u4Ss",
+    "outputId": "5861d34e-745b-45ee-e780-ed363043e655"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "--2022-01-21 03:07:52--  https://raw.githubusercontent.com/huggingface/transformers/master/examples/pytorch/speech-recognition/run_speech_recognition_ctc.py\n",
+      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n",
+      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n",
+      "HTTP request sent, awaiting response... 200 OK\n",
+      "Length: 30348 (30K) [text/plain]\n",
+      "Saving to: ‘run_speech_recognition_ctc.py’\n",
+      "\n",
+      "run_speech_recognit 100%[===================>]  29.64K  --.-KB/s    in 0.001s  \n",
+      "\n",
+      "2022-01-21 03:07:52 (21.5 MB/s) - ‘run_speech_recognition_ctc.py’ saved [30348/30348]\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "!wget -O run_speech_recognition_ctc.py https://raw.githubusercontent.com/huggingface/transformers/master/examples/pytorch/speech-recognition/run_speech_recognition_ctc.py"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 46,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "Mz4bubhxxsad",
+    "outputId": "23398525-cc19-43c2-9fec-497e06214f29"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "01/21/2022 06:29:10 - WARNING - __main__ - Process rank: -1, device: cuda:0, n_gpu: 1distributed training: False, 16-bits training: True\n",
+      "01/21/2022 06:29:10 - 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=500,\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=2,\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=7.5e-05,\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=./wav2vec2-large-xls-r-300m-odia/runs/Jan21_06-29-10_job-8be8b741-e32e-4579-bbec-1e00d9824b4f,\n",
+      "logging_first_step=False,\n",
+      "logging_nan_inf_filter=True,\n",
+      "logging_steps=100,\n",
+      "logging_strategy=IntervalStrategy.STEPS,\n",
+      "lr_scheduler_type=SchedulerType.LINEAR,\n",
+      "max_grad_norm=1.0,\n",
+      "max_steps=-1,\n",
+      "metric_for_best_model=None,\n",
+      "mp_parameters=,\n",
+      "no_cuda=False,\n",
+      "num_train_epochs=120.0,\n",
+      "optim=OptimizerNames.ADAMW_HF,\n",
+      "output_dir=./wav2vec2-large-xls-r-300m-odia,\n",
+      "overwrite_output_dir=True,\n",
+      "past_index=-1,\n",
+      "per_device_eval_batch_size=16,\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=./wav2vec2-large-xls-r-300m-odia,\n",
+      "save_on_each_node=False,\n",
+      "save_steps=500,\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",
+      "01/21/2022 06:29:12 - WARNING - datasets.builder - Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/or/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n",
+      "01/21/2022 06:29:15 - WARNING - datasets.builder - Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/or/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n",
+      "remove special characters from datasets: 100%|█| 537/537 [00:00<00:00, 5280.48ex\n",
+      "remove special characters from datasets: 100%|█| 112/112 [00:00<00:00, 6377.61ex\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\": \"4.16.0.dev0\",\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, 44.57ba/s]\n",
+      "100%|████████████████████████████████████████████| 1/1 [00:00<00:00, 236.54ba/s]\n",
+      "Didn't find file ./wav2vec2-large-xls-r-300m-odia/tokenizer.json. We won't load it.\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/vocab.json\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/tokenizer_config.json\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/added_tokens.json\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/special_tokens_map.json\n",
+      "loading file None\n",
+      "Adding <s> to the vocabulary\n",
+      "Adding </s> to the vocabulary\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\": \"4.16.0.dev0\",\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",
+      "Some weights of the model checkpoint at facebook/wav2vec2-xls-r-300m were not used when initializing Wav2Vec2ForCTC: ['project_q.weight', 'project_hid.weight', 'project_q.bias', 'project_hid.bias', 'quantizer.weight_proj.weight', 'quantizer.codevectors', 'quantizer.weight_proj.bias']\n",
+      "- 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",
+      "- 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",
+      "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",
+      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n",
+      "preprocess datasets: 100%|████████████████████| 537/537 [00:05<00:00, 97.46ex/s]\n",
+      "preprocess datasets: 100%|███████████████████| 112/112 [00:01<00:00, 107.18ex/s]\n",
+      "100%|████████████████████████████████████████████| 1/1 [00:00<00:00, 739.08ba/s]\n",
+      "100%|███████████████████████████████████████████| 1/1 [00:00<00:00, 1258.42ba/s]\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "tokenizer config file saved in ./wav2vec2-large-xls-r-300m-odia/tokenizer_config.json\n",
+      "Special tokens file saved in ./wav2vec2-large-xls-r-300m-odia/special_tokens_map.json\n",
+      "added tokens file saved in ./wav2vec2-large-xls-r-300m-odia/added_tokens.json\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/config.json\n",
+      "loading feature extractor configuration file ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "loading configuration file ./wav2vec2-large-xls-r-300m-odia/config.json\n",
+      "Model config Wav2Vec2Config {\n",
+      "  \"_name_or_path\": \"./wav2vec2-large-xls-r-300m-odia\",\n",
+      "  \"activation_dropout\": 0.1,\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\": 0.0,\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\": \"mean\",\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\": 0.0,\n",
+      "  \"feat_quantizer_dropout\": 0.0,\n",
+      "  \"final_dropout\": 0.0,\n",
+      "  \"hidden_act\": \"gelu\",\n",
+      "  \"hidden_dropout\": 0.0,\n",
+      "  \"hidden_size\": 1024,\n",
+      "  \"initializer_range\": 0.02,\n",
+      "  \"intermediate_size\": 4096,\n",
+      "  \"layer_norm_eps\": 1e-05,\n",
+      "  \"layerdrop\": 0.0,\n",
+      "  \"mask_feature_length\": 64,\n",
+      "  \"mask_feature_min_masks\": 0,\n",
+      "  \"mask_feature_prob\": 0.25,\n",
+      "  \"mask_time_length\": 10,\n",
+      "  \"mask_time_min_masks\": 2,\n",
+      "  \"mask_time_prob\": 0.75,\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\": 62,\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\": \"4.16.0.dev0\",\n",
+      "  \"use_weighted_layer_sum\": false,\n",
+      "  \"vocab_size\": 64,\n",
+      "  \"xvector_output_dim\": 512\n",
+      "}\n",
+      "\n",
+      "loading feature extractor configuration file ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\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",
+      "Didn't find file ./wav2vec2-large-xls-r-300m-odia/tokenizer.json. We won't load it.\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/vocab.json\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/tokenizer_config.json\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/added_tokens.json\n",
+      "loading file ./wav2vec2-large-xls-r-300m-odia/special_tokens_map.json\n",
+      "loading file None\n",
+      "Adding <s> to the vocabulary\n",
+      "Adding </s> to the vocabulary\n",
+      "/workspace/oriya_training/./wav2vec2-large-xls-r-300m-odia is already a clone of https://huggingface.co/infinitejoy/wav2vec2-large-xls-r-300m-odia. Make sure you pull the latest changes with `repo.git_pull()`.\n",
+      "01/21/2022 06:29:36 - WARNING - huggingface_hub.repository - /workspace/oriya_training/./wav2vec2-large-xls-r-300m-odia is already a clone of https://huggingface.co/infinitejoy/wav2vec2-large-xls-r-300m-odia. Make sure you pull the latest changes with `repo.git_pull()`.\n",
+      "Using amp half precision backend\n",
+      "The following columns in the training set  don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
+      "/opt/conda/lib/python3.8/site-packages/transformers/optimization.py:306: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use thePyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
+      "  warnings.warn(\n",
+      "***** Running training *****\n",
+      "  Num examples = 537\n",
+      "  Num Epochs = 120\n",
+      "  Instantaneous batch size per device = 16\n",
+      "  Total train batch size (w. parallel, distributed & accumulation) = 32\n",
+      "  Gradient Accumulation steps = 2\n",
+      "  Total optimization steps = 2040\n",
+      "{'loss': 16.0345, 'learning_rate': 1.455e-05, 'epoch': 5.88}                    \n",
+      "{'loss': 6.2463, 'learning_rate': 2.955e-05, 'epoch': 11.76}                    \n",
+      "{'loss': 4.2277, 'learning_rate': 4.454999999999999e-05, 'epoch': 17.65}        \n",
+      "{'loss': 3.4915, 'learning_rate': 5.955e-05, 'epoch': 23.53}                    \n",
+      "{'loss': 3.3231, 'learning_rate': 7.455e-05, 'epoch': 29.41}                    \n",
+      " 25%|█████████▊                              | 500/2040 [15:40<47:23,  1.85s/it]The following columns in the evaluation set  don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
+      "***** Running Evaluation *****\n",
+      "  Num examples = 112\n",
+      "  Batch size = 16\n",
+      "\n",
+      "  0%|                                                     | 0/7 [00:00<?, ?it/s]\u001b[A\n",
+      " 29%|████████████▊                                | 2/7 [00:00<00:02,  2.32it/s]\u001b[A\n",
+      " 43%|███████████████████▎                         | 3/7 [00:01<00:02,  1.94it/s]\u001b[A\n",
+      " 57%|█████████████████████████▋                   | 4/7 [00:02<00:01,  1.67it/s]\u001b[A\n",
+      " 71%|████████████████████████████████▏            | 5/7 [00:02<00:01,  1.69it/s]\u001b[A\n",
+      " 86%|██████████████████████████████████████▌      | 6/7 [00:03<00:00,  1.56it/s]\u001b[A\n",
+      "                                                                                \u001b[A\n",
+      "\u001b[A{'eval_loss': 3.352241277694702, 'eval_wer': 0.998972250770812, 'eval_runtime': 5.0475, 'eval_samples_per_second': 22.189, 'eval_steps_per_second': 1.387, 'epoch': 29.41}\n",
+      " 25%|█████████▊                              | 500/2040 [15:45<47:23,  1.85s/it]\n",
+      "100%|█████████████████████████████████████████████| 7/7 [00:04<00:00,  1.60it/s]\u001b[A\n",
+      "                                                                                \u001b[ASaving model checkpoint to ./wav2vec2-large-xls-r-300m-odia/checkpoint-500\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-500/config.json\n",
+      "Model weights saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-500/pytorch_model.bin\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-500/preprocessor_config.json\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "{'loss': 3.2264, 'learning_rate': 7.027597402597401e-05, 'epoch': 35.29}        \n",
+      "{'loss': 3.1652, 'learning_rate': 6.540584415584416e-05, 'epoch': 41.18}        \n",
+      "{'loss': 3.019, 'learning_rate': 6.0535714285714285e-05, 'epoch': 47.06}        \n",
+      "{'loss': 2.6429, 'learning_rate': 5.566558441558441e-05, 'epoch': 52.94}        \n",
+      "{'loss': 2.1146, 'learning_rate': 5.0795454545454536e-05, 'epoch': 58.82}       \n",
+      " 49%|███████████████████                    | 1000/2040 [32:52<33:58,  1.96s/it]The following columns in the evaluation set  don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
+      "***** Running Evaluation *****\n",
+      "  Num examples = 112\n",
+      "  Batch size = 16\n",
+      "\n",
+      "  0%|                                                     | 0/7 [00:00<?, ?it/s]\u001b[A\n",
+      " 29%|████████████▊                                | 2/7 [00:00<00:02,  2.36it/s]\u001b[A\n",
+      " 43%|███████████████████▎                         | 3/7 [00:01<00:02,  1.95it/s]\u001b[A\n",
+      " 57%|█████████████████████████▋                   | 4/7 [00:02<00:01,  1.67it/s]\u001b[A\n",
+      " 71%|████████████████████████████████▏            | 5/7 [00:02<00:01,  1.70it/s]\u001b[A\n",
+      " 86%|██████████████████████████████████████▌      | 6/7 [00:03<00:00,  1.55it/s]\u001b[A\n",
+      "                                                                                \u001b[A\n",
+      "\u001b[A{'eval_loss': 1.3367875814437866, 'eval_wer': 0.9383350462487153, 'eval_runtime': 5.0633, 'eval_samples_per_second': 22.12, 'eval_steps_per_second': 1.382, 'epoch': 58.82}\n",
+      " 49%|███████████████████                    | 1000/2040 [32:57<33:58,  1.96s/it]\n",
+      "100%|█████████████████████████████████████████████| 7/7 [00:04<00:00,  1.60it/s]\u001b[A\n",
+      "                                                                                \u001b[ASaving model checkpoint to ./wav2vec2-large-xls-r-300m-odia/checkpoint-1000\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-1000/config.json\n",
+      "Model weights saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-1000/pytorch_model.bin\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-1000/preprocessor_config.json\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "{'loss': 1.7769, 'learning_rate': 4.592532467532467e-05, 'epoch': 64.71}        \n",
+      "{'loss': 1.5937, 'learning_rate': 4.10551948051948e-05, 'epoch': 70.59}         \n",
+      "{'loss': 1.4694, 'learning_rate': 3.6185064935064934e-05, 'epoch': 76.47}       \n",
+      "{'loss': 1.3781, 'learning_rate': 3.131493506493506e-05, 'epoch': 82.35}        \n",
+      "{'loss': 1.3134, 'learning_rate': 2.6444805194805193e-05, 'epoch': 88.24}       \n",
+      " 74%|████████████████████████████▋          | 1500/2040 [50:05<15:32,  1.73s/it]The following columns in the evaluation set  don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
+      "***** Running Evaluation *****\n",
+      "  Num examples = 112\n",
+      "  Batch size = 16\n",
+      "\n",
+      "  0%|                                                     | 0/7 [00:00<?, ?it/s]\u001b[A\n",
+      " 29%|████████████▊                                | 2/7 [00:00<00:02,  2.39it/s]\u001b[A\n",
+      " 43%|███████████████████▎                         | 3/7 [00:01<00:02,  1.98it/s]\u001b[A\n",
+      " 57%|█████████████████████████▋                   | 4/7 [00:02<00:01,  1.68it/s]\u001b[A\n",
+      " 71%|████████████████████████████████▏            | 5/7 [00:02<00:01,  1.72it/s]\u001b[A\n",
+      " 86%|██████████████████████████████████████▌      | 6/7 [00:03<00:00,  1.52it/s]\u001b[A\n",
+      "                                                                                \u001b[A\n",
+      "\u001b[A{'eval_loss': 0.7372016310691833, 'eval_wer': 0.9578622816032888, 'eval_runtime': 5.0913, 'eval_samples_per_second': 21.998, 'eval_steps_per_second': 1.375, 'epoch': 88.24}\n",
+      " 74%|████████████████████████████▋          | 1500/2040 [50:10<15:32,  1.73s/it]\n",
+      "100%|█████████████████████████████████████████████| 7/7 [00:04<00:00,  1.57it/s]\u001b[A\n",
+      "                                                                                \u001b[ASaving model checkpoint to ./wav2vec2-large-xls-r-300m-odia/checkpoint-1500\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-1500/config.json\n",
+      "Model weights saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-1500/pytorch_model.bin\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-1500/preprocessor_config.json\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "{'loss': 1.271, 'learning_rate': 2.162337662337662e-05, 'epoch': 94.12}         \n",
+      "{'loss': 1.2273, 'learning_rate': 1.6753246753246752e-05, 'epoch': 100.0}       \n",
+      "{'loss': 1.2121, 'learning_rate': 1.1883116883116881e-05, 'epoch': 105.88}      \n",
+      "{'loss': 1.169, 'learning_rate': 7.012987012987012e-06, 'epoch': 111.76}        \n",
+      "{'loss': 1.1506, 'learning_rate': 2.1428571428571427e-06, 'epoch': 117.65}      \n",
+      " 98%|████████████████████████████████████▎| 2000/2040 [1:07:17<01:13,  1.85s/it]The following columns in the evaluation set  don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
+      "***** Running Evaluation *****\n",
+      "  Num examples = 112\n",
+      "  Batch size = 16\n",
+      "\n",
+      "  0%|                                                     | 0/7 [00:00<?, ?it/s]\u001b[A\n",
+      " 29%|████████████▊                                | 2/7 [00:00<00:02,  2.36it/s]\u001b[A\n",
+      " 43%|███████████████████▎                         | 3/7 [00:01<00:02,  1.96it/s]\u001b[A\n",
+      " 57%|█████████████████████████▋                   | 4/7 [00:02<00:01,  1.67it/s]\u001b[A\n",
+      " 71%|████████████████████████████████▏            | 5/7 [00:02<00:01,  1.69it/s]\u001b[A\n",
+      " 86%|██████████████████████████████████████▌      | 6/7 [00:03<00:00,  1.54it/s]\u001b[A\n",
+      "                                                                                \u001b[A\n",
+      "\u001b[A{'eval_loss': 0.6582115292549133, 'eval_wer': 0.9681397738951696, 'eval_runtime': 5.0953, 'eval_samples_per_second': 21.981, 'eval_steps_per_second': 1.374, 'epoch': 117.65}\n",
+      " 98%|████████████████████████████████████▎| 2000/2040 [1:07:22<01:13,  1.85s/it]\n",
+      "100%|█████████████████████████████████████████████| 7/7 [00:04<00:00,  1.57it/s]\u001b[A\n",
+      "                                                                                \u001b[ASaving model checkpoint to ./wav2vec2-large-xls-r-300m-odia/checkpoint-2000\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-2000/config.json\n",
+      "Model weights saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-2000/pytorch_model.bin\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/checkpoint-2000/preprocessor_config.json\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "Deleting older checkpoint [wav2vec2-large-xls-r-300m-odia/checkpoint-500] due to args.save_total_limit\n",
+      "100%|█████████████████████████████████████| 2040/2040 [1:10:04<00:00,  1.69s/it]\n",
+      "\n",
+      "Training completed. Do not forget to share your model on huggingface.co/models =)\n",
+      "\n",
+      "\n",
+      "{'train_runtime': 4204.6915, 'train_samples_per_second': 15.326, 'train_steps_per_second': 0.485, 'train_loss': 3.015083034365785, 'epoch': 120.0}\n",
+      "100%|█████████████████████████████████████| 2040/2040 [1:10:04<00:00,  2.06s/it]\n",
+      "Saving model checkpoint to ./wav2vec2-large-xls-r-300m-odia\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/config.json\n",
+      "Model weights saved in ./wav2vec2-large-xls-r-300m-odia/pytorch_model.bin\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "***** train metrics *****\n",
+      "  epoch                    =      120.0\n",
+      "  train_loss               =     3.0151\n",
+      "  train_runtime            = 1:10:04.69\n",
+      "  train_samples            =        537\n",
+      "  train_samples_per_second =     15.326\n",
+      "  train_steps_per_second   =      0.485\n",
+      "01/21/2022 07:39:46 - INFO - __main__ - *** Evaluate ***\n",
+      "The following columns in the evaluation set  don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.\n",
+      "***** Running Evaluation *****\n",
+      "  Num examples = 112\n",
+      "  Batch size = 16\n",
+      "100%|█████████████████████████████████████████████| 7/7 [00:04<00:00,  1.66it/s]\n",
+      "***** eval metrics *****\n",
+      "  epoch                   =      120.0\n",
+      "  eval_loss               =      0.658\n",
+      "  eval_runtime            = 0:00:05.06\n",
+      "  eval_samples            =        112\n",
+      "  eval_samples_per_second =     22.115\n",
+      "  eval_steps_per_second   =      1.382\n",
+      "  eval_wer                =     0.9712\n",
+      "Saving model checkpoint to ./wav2vec2-large-xls-r-300m-odia\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/config.json\n",
+      "Model weights saved in ./wav2vec2-large-xls-r-300m-odia/pytorch_model.bin\n",
+      "Configuration saved in ./wav2vec2-large-xls-r-300m-odia/preprocessor_config.json\n",
+      "Upload file pytorch_model.bin:  98%|██████▊| 1.15G/1.18G [00:41<00:01, 28.4MB/s]To https://huggingface.co/infinitejoy/wav2vec2-large-xls-r-300m-odia\n",
+      "   543c1c1..e870548  main -> main\n",
+      "\n",
+      "01/21/2022 07:41:54 - WARNING - huggingface_hub.repository - To https://huggingface.co/infinitejoy/wav2vec2-large-xls-r-300m-odia\n",
+      "   543c1c1..e870548  main -> main\n",
+      "\n",
+      "Upload file pytorch_model.bin: 100%|███████| 1.18G/1.18G [00:42<00:00, 29.8MB/s]\n",
+      "Dropping the following result as it does not have all the necessary fields:\n",
+      "{'dataset': {'name': 'MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - OR', 'type': 'common_voice', 'args': 'Config: or, Training split: train+validation, Eval split: test'}}\n",
+      "To https://huggingface.co/infinitejoy/wav2vec2-large-xls-r-300m-odia\n",
+      "   e870548..eb50e66  main -> main\n",
+      "\n",
+      "01/21/2022 07:42:00 - WARNING - huggingface_hub.repository - To https://huggingface.co/infinitejoy/wav2vec2-large-xls-r-300m-odia\n",
+      "   e870548..eb50e66  main -> main\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "!python run_speech_recognition_ctc.py \\\n",
+    "\t--dataset_name=\"mozilla-foundation/common_voice_7_0\" \\\n",
+    "\t--model_name_or_path=\"facebook/wav2vec2-xls-r-300m\" \\\n",
+    "\t--dataset_config_name=\"or\" \\\n",
+    "\t--output_dir=\"./wav2vec2-large-xls-r-300m-odia\" \\\n",
+    "\t--overwrite_output_dir \\\n",
+    "\t--num_train_epochs=\"120\" \\\n",
+    "\t--per_device_train_batch_size=\"16\" \\\n",
+    "\t--per_device_eval_batch_size=\"16\" \\\n",
+    "\t--gradient_accumulation_steps=\"2\" \\\n",
+    "\t--learning_rate=\"7.5e-5\" \\\n",
+    "\t--warmup_steps=\"500\" \\\n",
+    "\t--length_column_name=\"input_length\" \\\n",
+    "\t--evaluation_strategy=\"steps\" \\\n",
+    "\t--text_column_name=\"sentence\" \\\n",
+    "\t--chars_to_ignore , ? . ! \\- \\; \\: \\\" “ % ‘ ” � — \\’ … \\– \\' \\’ \\– \\\n",
+    "\t--save_steps=\"500\" \\\n",
+    "\t--eval_steps=\"500\" \\\n",
+    "\t--logging_steps=\"100\" \\\n",
+    "\t--layerdrop=\"0.0\" \\\n",
+    "\t--activation_dropout=\"0.1\" \\\n",
+    "\t--save_total_limit=\"3\" \\\n",
+    "\t--freeze_feature_encoder \\\n",
+    "\t--feat_proj_dropout=\"0.0\" \\\n",
+    "\t--mask_time_prob=\"0.75\" \\\n",
+    "\t--mask_time_length=\"10\" \\\n",
+    "\t--mask_feature_prob=\"0.25\" \\\n",
+    "\t--mask_feature_length=\"64\" \\\n",
+    "\t--gradient_checkpointing \\\n",
+    "\t--use_auth_token \\\n",
+    "\t--fp16 \\\n",
+    "\t--group_by_length \\\n",
+    "\t--do_train --do_eval \\\n",
+    "    --push_to_hub"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "df = pd.DataFrame([\n",
+    "    {}\n",
+    "])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# !zip -r wav2vec2-large-xls-r-300m-odia.zip wav2vec2-large-xls-r-300m-odia/\n",
+    "# !rm wav2vec2-large-xls-r-300m-odia.zip"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {
+    "collapsed": true,
+    "jupyter": {
+     "outputs_hidden": true
+    }
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Filesystem      Size  Used Avail Use% Mounted on\n",
+      "overlay         3.5T  557G  2.8T  17% /\n",
+      "tmpfs            64M     0   64M   0% /dev\n",
+      "tmpfs            87G     0   87G   0% /sys/fs/cgroup\n",
+      "tmpfs            87G     0   87G   0% /dev/shm\n",
+      "/dev/md0        3.5T  557G  2.8T  17% /etc/group\n",
+      "tmpfs            87G   12K   87G   1% /proc/driver/nvidia\n",
+      "/dev/vda1        49G  6.6G   42G  14% /usr/bin/nvidia-smi\n",
+      "udev             87G     0   87G   0% /dev/nvidia0\n",
+      "tmpfs            87G     0   87G   0% /proc/acpi\n",
+      "tmpfs            87G     0   87G   0% /proc/scsi\n",
+      "tmpfs            87G     0   87G   0% /sys/firmware\n"
+     ]
+    }
+   ],
+   "source": [
+    "!df -h"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/or/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n",
+      "Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/or/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n"
+     ]
+    }
+   ],
+   "source": [
+    "from datasets import load_dataset, load_metric, Audio\n",
+    "\n",
+    "common_voice_train = load_dataset(\"mozilla-foundation/common_voice_7_0\", \"or\", use_auth_token=True, split=\"train+validation\")\n",
+    "common_voice_test = load_dataset(\"mozilla-foundation/common_voice_7_0\", \"or\", use_auth_token=True, split=\"test\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 54,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "2013.75"
+      ]
+     },
+     "execution_count": 54,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "len(common_voice_train) * 120 / 32"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "common_voice_train = common_voice_train.remove_columns([\"accent\", \"age\", \"client_id\", \"down_votes\", \"gender\", \"locale\", \"segment\", \"up_votes\"])\n",
+    "common_voice_test = common_voice_test.remove_columns([\"accent\", \"age\", \"client_id\", \"down_votes\", \"gender\", \"locale\", \"segment\", \"up_votes\"])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import ClassLabel\n",
+    "import random\n",
+    "import pandas as pd\n",
+    "from IPython.display import display, HTML\n",
+    "\n",
+    "def show_random_elements(dataset, num_examples=10):\n",
+    "    assert num_examples <= len(dataset), \"Can't pick more elements than there are in the dataset.\"\n",
+    "    picks = []\n",
+    "    for _ in range(num_examples):\n",
+    "        pick = random.randint(0, len(dataset)-1)\n",
+    "        while pick in picks:\n",
+    "            pick = random.randint(0, len(dataset)-1)\n",
+    "        picks.append(pick)\n",
+    "    \n",
+    "    df = pd.DataFrame(dataset[picks])\n",
+    "    display(HTML(df.to_html()))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>sentence</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>ସେ କଥା ଯାଉ, ଆମ୍ଭମାନଙ୍କୁ ଆଉ କଥା ଲେଖିବାକୁ ହେବ ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>ଯାହା ଦରମା ଗଣ୍ଡାକ ପାଉଥିଲେ, ପେଟ ପିଠିକୁ ନିଅଣ୍ଟ, ବିଧବା ଲାଗି ସାଇତି ଯିବେ କଣ?</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>ଯେ ଯେଡ଼େ ହୁସିଆର ହେବ, ଆପଦ ବିପଦ କାହାରିକୁ ଛାଡ଼ିନାହିଁ ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>ମୁଁ ପୂଜା ସାରି ସେମାନଙ୍କୁ କିଛି ଭୋଗ ଦେଇ ଘରେ ଛାଡ଼ିଆସିଲି, ବାକି ଭୋଗକୁ ବାନ୍ଧିଲି ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>ବାସୁ ଦୁଇ ଟଙ୍କାର ନଡ଼ା କିଣି ବାଡ଼ିରେ ଗଦେଇଅଛି, ଶରଣ ଦେବାରୁ ଛପରବନ୍ଦି ହୋଇପାରି ନାହିଁ ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>ଦେଖି ଦେଖି ମନରେ କଲା, ଆଜି ଏ କଣ ହେଉଛି ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>ଶାଶୁ ମୁହଁକୁ ଚାହିଁ ଗାଳି ଦିଏ ନାହିଁ; ଓଢ଼ଣା ପଡ଼ିଥାଏ, ପଛ କରି ବରବର କରି ବକିଯାଏ ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>ଆଜି ମହାପ୍ରସାଦ ଉଠା ପରା ।</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>\"\"\"ଯାହାର ବାହା ସେ ଖେଳୁଛି ପଶା ଧାଇଁ ବୁଲୁଛନ୍ତି ସାଇ ପଡିଶା ।\"\"\"</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>ଅଶୀ ବର୍ଷର ପୁରୁଷ ବି ବିଭା ହୋଇ ପାରେ ।</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "show_random_elements(common_voice_train.remove_columns([\"path\", \"audio\"]), num_examples=10)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 36,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import re\n",
+    "chars_to_remove_regex = '[\\,\\?\\.\\!\\-\\;\\:\\\"\\“\\%\\‘\\”\\�\\'\\’\\–]'\n",
+    "\n",
+    "def remove_special_characters(batch):\n",
+    "    batch[\"sentence\"] = re.sub(chars_to_remove_regex, '', batch[\"sentence\"]).lower()\n",
+    "    return batch"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 37,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "a9df324b393840628b6a038aa00aa697",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/537 [00:00<?, ?ex/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "35fb5a05a9484b5892a1b41be076c55c",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/112 [00:00<?, ?ex/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "common_voice_train = common_voice_train.map(remove_special_characters)\n",
+    "common_voice_test = common_voice_test.map(remove_special_characters)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 38,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def replace_hatted_characters(batch):\n",
+    "    batch[\"sentence\"] = re.sub('[â]', 'a', batch[\"sentence\"])\n",
+    "    batch[\"sentence\"] = re.sub('[î]', 'i', batch[\"sentence\"])\n",
+    "    batch[\"sentence\"] = re.sub('[ô]', 'o', batch[\"sentence\"])\n",
+    "    batch[\"sentence\"] = re.sub('[û]', 'u', batch[\"sentence\"])\n",
+    "    return batch"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 39,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "e57ed45f40c440dc8df26f140b226c0d",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/537 [00:00<?, ?ex/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "863c62f8741f4efcb130ff1f44f3e0e4",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/112 [00:00<?, ?ex/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "common_voice_train = common_voice_train.map(replace_hatted_characters)\n",
+    "common_voice_test = common_voice_test.map(replace_hatted_characters)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 40,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def extract_all_chars(batch):\n",
+    "  all_text = \" \".join(batch[\"sentence\"])\n",
+    "  vocab = list(set(all_text))\n",
+    "  return {\"vocab\": [vocab], \"all_text\": [all_text]}"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 41,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "a8ab1aeb1bb240ca821b5558280495f3",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/1 [00:00<?, ?ba/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "f3d8bf90831a4e5d8a3feb7f30cf5966",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/1 [00:00<?, ?ba/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "vocab_train = common_voice_train.map(extract_all_chars, batched=True, batch_size=-1, keep_in_memory=True, remove_columns=common_voice_train.column_names)\n",
+    "vocab_test = common_voice_test.map(extract_all_chars, batched=True, batch_size=-1, keep_in_memory=True, remove_columns=common_voice_test.column_names)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 42,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "vocab_list = list(set(vocab_train[\"vocab\"][0]) | set(vocab_test[\"vocab\"][0]))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 43,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{' ': 0,\n",
+       " '|': 1,\n",
+       " '।': 2,\n",
+       " 'ଁ': 3,\n",
+       " 'ଂ': 4,\n",
+       " 'ଃ': 5,\n",
+       " 'ଅ': 6,\n",
+       " 'ଆ': 7,\n",
+       " 'ଇ': 8,\n",
+       " 'ଈ': 9,\n",
+       " 'ଉ': 10,\n",
+       " 'ଊ': 11,\n",
+       " 'ଏ': 12,\n",
+       " 'ଓ': 13,\n",
+       " 'କ': 14,\n",
+       " 'ଖ': 15,\n",
+       " 'ଗ': 16,\n",
+       " 'ଘ': 17,\n",
+       " 'ଙ': 18,\n",
+       " 'ଚ': 19,\n",
+       " 'ଛ': 20,\n",
+       " 'ଜ': 21,\n",
+       " 'ଝ': 22,\n",
+       " 'ଞ': 23,\n",
+       " 'ଟ': 24,\n",
+       " 'ଠ': 25,\n",
+       " 'ଡ': 26,\n",
+       " 'ଢ': 27,\n",
+       " 'ଣ': 28,\n",
+       " 'ତ': 29,\n",
+       " 'ଥ': 30,\n",
+       " 'ଦ': 31,\n",
+       " 'ଧ': 32,\n",
+       " 'ନ': 33,\n",
+       " 'ପ': 34,\n",
+       " 'ଫ': 35,\n",
+       " 'ବ': 36,\n",
+       " 'ଭ': 37,\n",
+       " 'ମ': 38,\n",
+       " 'ଯ': 39,\n",
+       " 'ର': 40,\n",
+       " 'ଲ': 41,\n",
+       " 'ଳ': 42,\n",
+       " 'ଵ': 43,\n",
+       " 'ଶ': 44,\n",
+       " 'ଷ': 45,\n",
+       " 'ସ': 46,\n",
+       " 'ହ': 47,\n",
+       " '଼': 48,\n",
+       " 'ା': 49,\n",
+       " 'ି': 50,\n",
+       " 'ୀ': 51,\n",
+       " 'ୁ': 52,\n",
+       " 'ୂ': 53,\n",
+       " 'ୃ': 54,\n",
+       " 'େ': 55,\n",
+       " 'ୈ': 56,\n",
+       " 'ୋ': 57,\n",
+       " 'ୌ': 58,\n",
+       " '୍': 59,\n",
+       " 'ୟ': 60,\n",
+       " 'ୱ': 61}"
+      ]
+     },
+     "execution_count": 43,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "vocab_dict = {v: k for k, v in enumerate(sorted(vocab_list))}\n",
+    "vocab_dict"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 48,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "--2022-01-21 08:33:50--  https://raw.githubusercontent.com/huggingface/transformers/master/examples/research_projects/robust-speech-event/eval.py\n",
+      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.109.133, 185.199.110.133, 185.199.111.133, ...\n",
+      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.109.133|:443... connected.\n",
+      "HTTP request sent, awaiting response... 200 OK\n",
+      "Length: 4419 (4.3K) [text/plain]\n",
+      "Saving to: ‘eval.py’\n",
+      "\n",
+      "eval.py             100%[===================>]   4.32K  --.-KB/s    in 0s      \n",
+      "\n",
+      "2022-01-21 08:33:50 (14.9 MB/s) - ‘eval.py’ saved [4419/4419]\n",
+      "\n",
+      "total 1232676\n",
+      "-rw-r--r-- 1 ovh ovh        686 Jan 21 06:29 vocab.json\n",
+      "-rw-r--r-- 1 ovh ovh        290 Jan 21 06:29 tokenizer_config.json\n",
+      "-rw-r--r-- 1 ovh ovh        502 Jan 21 06:29 special_tokens_map.json\n",
+      "-rw-r--r-- 1 ovh ovh         23 Jan 21 06:29 added_tokens.json\n",
+      "drwxr-xr-x 2 ovh ovh       4096 Jan 21 07:02 checkpoint-1000\n",
+      "drwxr-xr-x 2 ovh ovh       4096 Jan 21 07:19 checkpoint-1500\n",
+      "drwxr-xr-x 2 ovh ovh       4096 Jan 21 07:37 checkpoint-2000\n",
+      "-rw-r--r-- 1 ovh ovh       3953 Jan 21 07:39 trainer_state.json\n",
+      "-rw-r--r-- 1 ovh ovh        194 Jan 21 07:39 train_results.json\n",
+      "-rw-r--r-- 1 ovh ovh        222 Jan 21 07:39 eval_results.json\n",
+      "-rw-r--r-- 1 ovh ovh       2033 Jan 21 07:39 config.json\n",
+      "-rw-r--r-- 1 ovh ovh        394 Jan 21 07:39 all_results.json\n",
+      "-rw-r--r-- 1 ovh ovh 1262186097 Jan 21 07:39 pytorch_model.bin\n",
+      "-rw-r--r-- 1 ovh ovh       3055 Jan 21 07:39 training_args.bin\n",
+      "-rw-r--r-- 1 ovh ovh        212 Jan 21 07:39 preprocessor_config.json\n",
+      "-rw-r--r-- 1 ovh ovh       1825 Jan 21 07:41 README.md\n",
+      "-rw-r--r-- 1 ovh ovh       4419 Jan 21 08:33 eval.py\n"
+     ]
+    }
+   ],
+   "source": [
+    "!wget -O eval.py https://raw.githubusercontent.com/huggingface/transformers/master/examples/research_projects/robust-speech-event/eval.py\n",
+    "!cp eval.py wav2vec2-large-xls-r-300m-odia\n",
+    "!ls -ltr wav2vec2-large-xls-r-300m-odia"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 50,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/or/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n",
+      "100%|███████████████████████████████████████████| 10/10 [00:06<00:00,  1.55ex/s]\n",
+      "Downloading: 5.61kB [00:00, 2.23MB/s]                                           \n",
+      "WER: 1.0921052631578947\n",
+      "CER: 2.5547945205479454\n",
+      "100%|████████████████████████████████████████| 10/10 [00:00<00:00, 13001.56ex/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "!cd wav2vec2-large-xls-r-300m-odia; python eval.py --model_id ./ --dataset mozilla-foundation/common_voice_7_0 --config or --split test --log_outputs"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "accelerator": "GPU",
+  "colab": {
+   "authorship_tag": "ABX9TyM3OaMlm9YQtKpl28c8gBBd",
+   "include_colab_link": true,
+   "name": "DebugOVHTransformers.ipynb",
+   "provenance": []
+  },
+  "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": 4
+}