Downloading: 0%| | 0.00/9.79k [00:00 to the vocabulary Adding to the vocabulary Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained. loading configuration file https://huggingface.co/facebook/wav2vec2-xls-r-300m/resolve/main/config.json from cache at /home/sagrilaft/.cache/huggingface/transformers/dabc27df63e37bd2a7a221c7774e35f36a280fbdf917cf54cadfc7df8c786f6f.a3e4c3c967d9985881e0ae550a5f6f668f897db5ab2e0802f9b97973b15970e6 Model config Wav2Vec2Config { "_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": [ "Wav2Vec2ForPreTraining" ], "attention_dropout": 0.1, "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": "sum", "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": 0.1, "feat_quantizer_dropout": 0.0, "final_dropout": 0.0, "gradient_checkpointing": false, "hidden_act": "gelu", "hidden_dropout": 0.1, "hidden_size": 1024, "initializer_range": 0.02, "intermediate_size": 4096, "layer_norm_eps": 1e-05, "layerdrop": 0.1, "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": 0.075, "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": 0, "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": "4.16.0.dev0", "use_weighted_layer_sum": false, "vocab_size": 32, "xvector_output_dim": 512 } loading feature extractor configuration file https://huggingface.co/facebook/wav2vec2-xls-r-300m/resolve/main/preprocessor_config.json from cache at /home/sagrilaft/.cache/huggingface/transformers/6fb028b95b394059e7d3b367bbca2382b576c66aebe896f04d2cd34e1b575f5b.d4484dc1c81456a2461485e7168b04347a7b9a4e3b1ef3aba723323b33e12326 Feature extractor Wav2Vec2FeatureExtractor { "do_normalize": true, "feature_extractor_type": "Wav2Vec2FeatureExtractor", "feature_size": 1, "padding_side": "right", "padding_value": 0, "return_attention_mask": true, "sampling_rate": 16000 } loading weights file https://huggingface.co/facebook/wav2vec2-xls-r-300m/resolve/main/pytorch_model.bin from cache at /home/sagrilaft/.cache/huggingface/transformers/1e6a6507f3b689035cd4b247e2a37c154e27f39143f31357a49b4e38baeccc36.1edb32803799e27ed554eb7dd935f6745b1a0b17b0ea256442fe24db6eb546cd Some weights of the model checkpoint at facebook/wav2vec2-xls-r-300m were not used when initializing Wav2Vec2ForCTC: ['project_q.bias', 'project_q.weight', 'quantizer.weight_proj.weight', 'project_hid.bias', 'project_hid.weight', 'quantizer.codevectors', 'quantizer.weight_proj.bias'] - 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). - 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). Some weights of Wav2Vec2ForCTC were not initialized from the model checkpoint at facebook/wav2vec2-xls-r-300m and are newly initialized: ['lm_head.bias', 'lm_head.weight'] You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference. preprocess datasets: 0ex [00:00, ?ex/s] preprocess datasets: 3ex [00:00, 29.00ex/s] preprocess datasets: 11ex [00:00, 58.57ex/s] preprocess datasets: 20ex [00:00, 71.44ex/s] preprocess datasets: 28ex [00:00, 73.26ex/s] preprocess datasets: 36ex [00:00, 74.53ex/s] preprocess datasets: 44ex [00:00, 75.67ex/s] preprocess datasets: 52ex [00:00, 74.21ex/s] preprocess datasets: 60ex [00:00, 72.30ex/s] preprocess datasets: 68ex [00:00, 74.29ex/s] preprocess datasets: 76ex [00:01, 72.58ex/s] preprocess datasets: 85ex [00:01, 74.21ex/s] preprocess datasets: 93ex [00:01, 72.98ex/s] preprocess datasets: 101ex [00:01, 74.67ex/s] preprocess datasets: 109ex [00:01, 75.60ex/s] preprocess datasets: 118ex [00:01, 78.71ex/s] preprocess datasets: 128ex [00:01, 81.43ex/s] preprocess datasets: 137ex [00:01, 80.04ex/s] preprocess datasets: 146ex [00:01, 79.61ex/s] preprocess datasets: 155ex [00:02, 81.56ex/s] preprocess datasets: 164ex [00:02, 83.30ex/s] preprocess datasets: 173ex [00:02, 77.46ex/s] preprocess datasets: 182ex [00:02, 80.58ex/s] preprocess datasets: 192ex [00:02, 84.87ex/s] preprocess datasets: 201ex [00:02, 82.85ex/s] preprocess datasets: 210ex [00:02, 81.85ex/s] preprocess datasets: 219ex [00:02, 83.03ex/s] preprocess datasets: 228ex [00:02, 84.59ex/s] preprocess datasets: 238ex [00:03, 86.50ex/s] preprocess datasets: 247ex [00:03, 86.72ex/s] preprocess datasets: 257ex [00:03, 89.71ex/s] preprocess datasets: 267ex [00:03, 91.83ex/s] preprocess datasets: 277ex [00:03, 87.54ex/s] preprocess datasets: 286ex [00:03, 86.42ex/s] preprocess datasets: 295ex [00:03, 85.19ex/s] preprocess datasets: 304ex [00:03, 85.70ex/s] preprocess datasets: 313ex [00:03, 83.44ex/s] preprocess datasets: 322ex [00:04, 82.27ex/s] preprocess datasets: 331ex [00:04, 83.74ex/s] preprocess datasets: 340ex [00:04, 85.11ex/s] preprocess datasets: 350ex [00:04, 87.55ex/s] preprocess datasets: 359ex [00:04, 79.55ex/s] preprocess datasets: 368ex [00:04, 81.08ex/s] preprocess datasets: 379ex [00:04, 88.77ex/s] preprocess datasets: 389ex [00:04, 85.07ex/s] preprocess datasets: 398ex [00:04, 83.37ex/s] preprocess datasets: 407ex [00:05, 83.14ex/s] preprocess datasets: 417ex [00:05, 86.75ex/s] preprocess datasets: 427ex [00:05, 87.85ex/s] preprocess datasets: 437ex [00:05, 90.32ex/s] preprocess datasets: 447ex [00:05, 88.86ex/s] preprocess datasets: 457ex [00:05, 91.26ex/s] preprocess datasets: 467ex [00:05, 93.58ex/s] preprocess datasets: 477ex [00:05, 93.19ex/s] preprocess datasets: 487ex [00:05, 88.87ex/s] preprocess datasets: 496ex [00:06, 86.09ex/s] preprocess datasets: 506ex [00:06, 87.60ex/s] preprocess datasets: 516ex [00:06, 89.99ex/s] preprocess datasets: 526ex [00:06, 90.10ex/s] preprocess datasets: 537ex [00:06, 94.69ex/s] preprocess datasets: 547ex [00:06, 89.48ex/s] preprocess datasets: 557ex [00:06, 87.14ex/s] preprocess datasets: 567ex [00:06, 89.24ex/s] preprocess datasets: 576ex [00:06, 88.63ex/s] preprocess datasets: 586ex [00:07, 89.78ex/s] preprocess datasets: 596ex [00:07, 84.93ex/s] preprocess datasets: 605ex [00:07, 81.82ex/s] preprocess datasets: 614ex [00:07, 82.78ex/s] preprocess datasets: 623ex [00:07, 83.91ex/s] preprocess datasets: 632ex [00:07, 85.07ex/s] preprocess datasets: 643ex [00:07, 91.34ex/s] preprocess datasets: 656ex [00:07, 100.68ex/s] preprocess datasets: 667ex [00:07, 93.63ex/s] preprocess datasets: 677ex [00:08, 91.95ex/s] preprocess datasets: 689ex [00:08, 98.85ex/s] preprocess datasets: 700ex [00:08, 92.04ex/s] preprocess datasets: 710ex [00:08, 90.41ex/s] preprocess datasets: 720ex [00:08, 86.97ex/s] preprocess datasets: 729ex [00:08, 85.06ex/s] preprocess datasets: 739ex [00:08, 87.01ex/s] preprocess datasets: 749ex [00:08, 89.40ex/s] preprocess datasets: 758ex [00:08, 87.40ex/s] preprocess datasets: 767ex [00:09, 85.09ex/s] preprocess datasets: 776ex [00:09, 83.74ex/s] preprocess datasets: 786ex [00:09, 85.74ex/s] preprocess datasets: 799ex [00:09, 95.03ex/s] preprocess datasets: 811ex [00:09, 100.75ex/s] preprocess datasets: 822ex [00:09, 87.09ex/s] preprocess datasets: 832ex [00:09, 83.73ex/s] preprocess datasets: 841ex [00:09, 82.61ex/s] preprocess datasets: 850ex [00:10, 84.03ex/s] preprocess datasets: 859ex [00:10, 81.30ex/s] preprocess datasets: 868ex [00:10, 83.59ex/s] preprocess datasets: 878ex [00:10, 85.72ex/s] preprocess datasets: 889ex [00:10, 90.10ex/s] preprocess datasets: 900ex [00:10, 93.52ex/s] preprocess datasets: 910ex [00:10, 90.09ex/s] preprocess datasets: 920ex [00:10, 83.38ex/s] preprocess datasets: 929ex [00:10, 82.47ex/s] preprocess datasets: 938ex [00:11, 81.67ex/s] preprocess datasets: 948ex [00:11, 83.70ex/s] preprocess datasets: 957ex [00:11, 82.21ex/s] preprocess datasets: 967ex [00:11, 84.06ex/s] preprocess datasets: 976ex [00:11, 80.47ex/s] preprocess datasets: 985ex [00:11, 81.07ex/s] preprocess datasets: 994ex [00:11, 83.14ex/s] preprocess datasets: 1003ex [00:12, 24.82ex/s] preprocess datasets: 1011ex [00:12, 30.41ex/s] preprocess datasets: 1019ex [00:12, 36.65ex/s] preprocess datasets: 1027ex [00:13, 42.62ex/s] preprocess datasets: 1039ex [00:13, 55.75ex/s] preprocess datasets: 1050ex [00:13, 65.64ex/s] preprocess datasets: 1060ex [00:13, 72.35ex/s] preprocess datasets: 1070ex [00:13, 77.76ex/s] preprocess datasets: 1080ex [00:13, 82.65ex/s] preprocess datasets: 1090ex [00:13, 85.68ex/s] preprocess datasets: 1100ex [00:13, 84.89ex/s] preprocess datasets: 1110ex [00:13, 87.13ex/s] preprocess datasets: 1120ex [00:13, 88.55ex/s] preprocess datasets: 1130ex [00:14, 84.09ex/s] preprocess datasets: 1140ex [00:14, 86.99ex/s] preprocess datasets: 1150ex [00:14, 87.38ex/s] preprocess datasets: 1162ex [00:14, 94.59ex/s] preprocess datasets: 1175ex [00:14, 100.52ex/s] preprocess datasets: 1186ex [00:14, 94.68ex/s] preprocess datasets: 1196ex [00:14, 95.97ex/s] preprocess datasets: 1207ex [00:14, 99.66ex/s] preprocess datasets: 1218ex [00:15, 99.27ex/s] preprocess datasets: 1228ex [00:15, 96.92ex/s] preprocess datasets: 1238ex [00:15, 95.96ex/s] preprocess datasets: 1248ex [00:15, 96.38ex/s] preprocess datasets: 1258ex [00:15, 91.72ex/s] preprocess datasets: 1268ex [00:15, 90.02ex/s] preprocess datasets: 1278ex [00:15, 91.57ex/s] preprocess datasets: 1288ex [00:15, 92.82ex/s] preprocess datasets: 1298ex [00:15, 94.80ex/s] preprocess datasets: 1310ex [00:15, 98.76ex/s] preprocess datasets: 1320ex [00:16, 91.01ex/s] preprocess datasets: 1330ex [00:16, 89.87ex/s] preprocess datasets: 1340ex [00:16, 90.94ex/s] preprocess datasets: 1350ex [00:16, 83.15ex/s] preprocess datasets: 1359ex [00:16, 82.34ex/s] preprocess datasets: 1369ex [00:16, 86.38ex/s] preprocess datasets: 1378ex [00:16, 86.56ex/s] preprocess datasets: 1387ex [00:16, 85.94ex/s] preprocess datasets: 1398ex [00:17, 91.90ex/s] preprocess datasets: 1411ex [00:17, 101.18ex/s] preprocess datasets: 1422ex [00:17, 98.08ex/s] preprocess datasets: 1434ex [00:17, 101.93ex/s] preprocess datasets: 1445ex [00:17, 101.20ex/s] preprocess datasets: 1456ex [00:17, 98.95ex/s] preprocess datasets: 1467ex [00:17, 100.01ex/s] preprocess datasets: 1478ex [00:17, 98.29ex/s] preprocess datasets: 1489ex [00:17, 100.07ex/s] preprocess datasets: 1501ex [00:18, 104.28ex/s] preprocess datasets: 1512ex [00:18, 105.21ex/s] preprocess datasets: 1523ex [00:18, 101.23ex/s] preprocess datasets: 1534ex [00:18, 99.39ex/s] preprocess datasets: 1546ex [00:18, 103.36ex/s] preprocess datasets: 1557ex [00:18, 100.75ex/s] preprocess datasets: 1568ex [00:18, 95.12ex/s] preprocess datasets: 1578ex [00:18, 91.76ex/s] preprocess datasets: 1588ex [00:18, 90.24ex/s] preprocess datasets: 1598ex [00:19, 89.75ex/s] preprocess datasets: 1608ex [00:19, 92.06ex/s] preprocess datasets: 1618ex [00:19, 94.10ex/s] preprocess datasets: 1628ex [00:19, 88.50ex/s] preprocess datasets: 1639ex [00:19, 93.44ex/s] preprocess datasets: 1650ex [00:19, 94.57ex/s] preprocess datasets: 1660ex [00:19, 90.99ex/s] preprocess datasets: 1670ex [00:19, 90.10ex/s] preprocess datasets: 1680ex [00:19, 91.95ex/s] preprocess datasets: 1690ex [00:20, 87.13ex/s] preprocess datasets: 1700ex [00:20, 88.98ex/s] preprocess datasets: 1711ex [00:20, 94.53ex/s] preprocess datasets: 1721ex [00:20, 93.21ex/s] preprocess datasets: 1731ex [00:20, 94.10ex/s] preprocess datasets: 1742ex [00:20, 95.75ex/s] preprocess datasets: 1752ex [00:20, 92.93ex/s] preprocess datasets: 1764ex [00:20, 97.52ex/s] preprocess datasets: 1774ex [00:20, 97.73ex/s] preprocess datasets: 1785ex [00:21, 99.83ex/s] preprocess datasets: 1796ex [00:21, 101.74ex/s] preprocess datasets: 1808ex [00:21, 106.86ex/s] preprocess datasets: 1822ex [00:21, 115.16ex/s] preprocess datasets: 1836ex [00:21, 119.85ex/s] preprocess datasets: 1849ex [00:21, 121.77ex/s] preprocess datasets: 1863ex [00:21, 126.58ex/s] preprocess datasets: 1877ex [00:21, 127.93ex/s] preprocess datasets: 1890ex [00:21, 124.27ex/s] preprocess datasets: 1903ex [00:21, 123.77ex/s] preprocess datasets: 1917ex [00:22, 126.90ex/s] preprocess datasets: 1931ex [00:22, 129.09ex/s] preprocess datasets: 1944ex [00:22, 125.27ex/s] preprocess datasets: 1957ex [00:22, 124.26ex/s] preprocess datasets: 1971ex [00:22, 128.07ex/s] preprocess datasets: 1984ex [00:22, 123.20ex/s] preprocess datasets: 1999ex [00:22, 129.49ex/s] preprocess datasets: 2013ex [00:23, 45.09ex/s] preprocess datasets: 2024ex [00:23, 52.79ex/s] preprocess datasets: 2038ex [00:23, 64.80ex/s] preprocess datasets: 2053ex [00:23, 78.69ex/s] preprocess datasets: 2066ex [00:23, 87.92ex/s] preprocess datasets: 2079ex [00:24, 95.58ex/s] preprocess datasets: 2093ex [00:24, 103.79ex/s] preprocess datasets: 2106ex [00:24, 107.62ex/s] preprocess datasets: 2119ex [00:24, 110.49ex/s] preprocess datasets: 2132ex [00:24, 112.85ex/s] preprocess datasets: 2144ex [00:24, 114.39ex/s] preprocess datasets: 2156ex [00:24, 115.88ex/s] preprocess datasets: 2168ex [00:24, 115.56ex/s] preprocess datasets: 2180ex [00:24, 113.61ex/s] preprocess datasets: 2196ex [00:24, 125.22ex/s] preprocess datasets: 2213ex [00:25, 135.94ex/s] preprocess datasets: 2228ex [00:25, 138.49ex/s] preprocess datasets: 2242ex [00:25, 127.18ex/s] preprocess datasets: 2255ex [00:25, 119.44ex/s] preprocess datasets: 2268ex [00:25, 120.72ex/s] preprocess datasets: 2282ex [00:25, 123.88ex/s] preprocess datasets: 2295ex [00:25, 120.85ex/s] preprocess datasets: 2309ex [00:25, 126.12ex/s] preprocess datasets: 2322ex [00:25, 121.05ex/s] preprocess datasets: 2335ex [00:26, 119.89ex/s] preprocess datasets: 2348ex [00:26, 118.36ex/s] preprocess datasets: 2360ex [00:26, 118.57ex/s] preprocess datasets: 2372ex [00:26, 118.10ex/s] preprocess datasets: 2384ex [00:26, 114.96ex/s] preprocess datasets: 2396ex [00:26, 112.79ex/s] preprocess datasets: 2408ex [00:26, 113.06ex/s] preprocess datasets: 2422ex [00:26, 117.76ex/s] preprocess datasets: 2438ex [00:26, 127.95ex/s] preprocess datasets: 2451ex [00:27, 128.53ex/s] preprocess datasets: 2466ex [00:27, 131.24ex/s] preprocess datasets: 2480ex [00:27, 130.72ex/s] preprocess datasets: 2494ex [00:27, 129.32ex/s] preprocess datasets: 2507ex [00:27, 126.36ex/s] preprocess datasets: 2520ex [00:27, 122.84ex/s] preprocess datasets: 2533ex [00:27, 122.67ex/s] preprocess datasets: 2546ex [00:27, 123.67ex/s] preprocess datasets: 2559ex [00:27, 124.92ex/s] preprocess datasets: 2572ex [00:28, 123.78ex/s] preprocess datasets: 2585ex [00:28, 115.47ex/s] preprocess datasets: 2597ex [00:28, 115.60ex/s] preprocess datasets: 2609ex [00:28, 112.62ex/s] preprocess datasets: 2621ex [00:28, 112.19ex/s] preprocess datasets: 2633ex [00:28, 111.91ex/s] preprocess datasets: 2646ex [00:28, 116.08ex/s] preprocess datasets: 2660ex [00:28, 120.48ex/s] preprocess datasets: 2673ex [00:28, 117.74ex/s] preprocess datasets: 2688ex [00:29, 126.35ex/s] preprocess datasets: 2702ex [00:29, 129.58ex/s] preprocess datasets: 2717ex [00:29, 133.46ex/s] preprocess datasets: 2732ex [00:29, 138.22ex/s] preprocess datasets: 2746ex [00:29, 138.50ex/s] preprocess datasets: 2760ex [00:29, 136.69ex/s] preprocess datasets: 2776ex [00:29, 140.96ex/s] preprocess datasets: 2791ex [00:29, 139.52ex/s] preprocess datasets: 2805ex [00:29, 128.21ex/s] preprocess datasets: 2818ex [00:29, 124.94ex/s] preprocess datasets: 2831ex [00:30, 122.41ex/s] preprocess datasets: 2844ex [00:30, 120.10ex/s] preprocess datasets: 2857ex [00:30, 120.78ex/s] preprocess datasets: 2873ex [00:30, 129.65ex/s] preprocess datasets: 2888ex [00:30, 135.36ex/s] preprocess datasets: 2902ex [00:30, 126.96ex/s] preprocess datasets: 2915ex [00:30, 122.03ex/s] preprocess datasets: 2928ex [00:30, 118.41ex/s] preprocess datasets: 2940ex [00:30, 116.78ex/s] preprocess datasets: 2953ex [00:31, 117.24ex/s] preprocess datasets: 2965ex [00:31, 116.43ex/s] preprocess datasets: 2977ex [00:31, 114.94ex/s] preprocess datasets: 2989ex [00:31, 111.58ex/s] preprocess datasets: 3001ex [00:32, 48.05ex/s] preprocess datasets: 3012ex [00:32, 56.69ex/s] preprocess datasets: 3025ex [00:32, 68.34ex/s] preprocess datasets: 3037ex [00:32, 78.27ex/s] preprocess datasets: 3057ex [00:32, 102.93ex/s] preprocess datasets: 3074ex [00:32, 118.21ex/s] preprocess datasets: 3091ex [00:32, 130.95ex/s] preprocess datasets: 3106ex [00:32, 126.47ex/s] preprocess datasets: 3120ex [00:32, 113.72ex/s] preprocess datasets: 3133ex [00:33, 108.54ex/s] preprocess datasets: 3145ex [00:33, 107.88ex/s] preprocess datasets: 3157ex [00:33, 101.29ex/s] preprocess datasets: 3168ex [00:33, 97.74ex/s] preprocess datasets: 3179ex [00:33, 94.22ex/s] preprocess datasets: 3189ex [00:33, 94.95ex/s] preprocess datasets: 3199ex [00:33, 92.44ex/s] preprocess datasets: 3209ex [00:33, 93.58ex/s] preprocess datasets: 3219ex [00:34, 91.17ex/s] preprocess datasets: 3230ex [00:34, 92.65ex/s] preprocess datasets: 3240ex [00:34, 91.37ex/s] preprocess datasets: 3250ex [00:34, 89.73ex/s] preprocess datasets: 3259ex [00:34, 87.80ex/s] preprocess datasets: 3268ex [00:34, 87.14ex/s] preprocess datasets: 3277ex [00:34, 86.84ex/s] preprocess datasets: 3287ex [00:34, 88.86ex/s] preprocess datasets: 3298ex [00:34, 91.27ex/s] preprocess datasets: 3310ex [00:34, 98.31ex/s] preprocess datasets: 3322ex [00:35, 101.67ex/s] preprocess datasets: 3334ex [00:35, 102.95ex/s] preprocess datasets: 3345ex [00:35, 97.85ex/s] preprocess datasets: 3356ex [00:35, 99.95ex/s] preprocess datasets: 3367ex [00:35, 98.33ex/s] preprocess datasets: 3377ex [00:35, 98.01ex/s] preprocess datasets: 3387ex [00:35, 97.96ex/s] preprocess datasets: 3397ex [00:35, 94.37ex/s] preprocess datasets: 3407ex [00:35, 95.20ex/s] preprocess datasets: 3417ex [00:36, 95.93ex/s] preprocess datasets: 3428ex [00:36, 97.89ex/s] preprocess datasets: 3438ex [00:36, 98.39ex/s] preprocess datasets: 3448ex [00:36, 95.84ex/s] preprocess datasets: 3459ex [00:36, 99.20ex/s] preprocess datasets: 3469ex [00:36, 94.05ex/s] preprocess datasets: 3479ex [00:36, 89.50ex/s] preprocess datasets: 3489ex [00:36, 85.54ex/s] preprocess datasets: 3498ex [00:37, 82.21ex/s] preprocess datasets: 3508ex [00:37, 85.31ex/s] preprocess datasets: 3519ex [00:37, 90.89ex/s] preprocess datasets: 3529ex [00:37, 92.56ex/s] preprocess datasets: 3539ex [00:37, 91.82ex/s] preprocess datasets: 3549ex [00:37, 93.11ex/s] preprocess datasets: 3560ex [00:37, 96.09ex/s] preprocess datasets: 3571ex [00:37, 97.62ex/s] preprocess datasets: 3581ex [00:37, 96.36ex/s] preprocess datasets: 3591ex [00:37, 95.59ex/s] preprocess datasets: 3601ex [00:38, 93.96ex/s] preprocess datasets: 3611ex [00:38, 90.80ex/s] preprocess datasets: 3621ex [00:38, 87.57ex/s] preprocess datasets: 3630ex [00:38, 85.10ex/s] preprocess datasets: 3640ex [00:38, 87.75ex/s] preprocess datasets: 3653ex [00:38, 96.00ex/s] preprocess datasets: 3664ex [00:38, 98.84ex/s] preprocess datasets: 3674ex [00:38, 97.69ex/s] preprocess datasets: 3684ex [00:38, 97.01ex/s] preprocess datasets: 3694ex [00:39, 93.31ex/s] preprocess datasets: 3704ex [00:39, 94.64ex/s] preprocess datasets: 3715ex [00:39, 98.60ex/s] preprocess datasets: 3725ex [00:39, 98.22ex/s] preprocess datasets: 3735ex [00:39, 96.63ex/s] preprocess datasets: 3746ex [00:39, 99.60ex/s] preprocess datasets: 3756ex [00:39, 99.34ex/s] preprocess datasets: 3766ex [00:39, 91.86ex/s] preprocess datasets: 3776ex [00:39, 90.54ex/s] preprocess datasets: 3786ex [00:40, 92.53ex/s] preprocess datasets: 3796ex [00:40, 94.60ex/s] preprocess datasets: 3807ex [00:40, 98.28ex/s] preprocess datasets: 3820ex [00:40, 105.99ex/s] preprocess datasets: 3831ex [00:40, 103.79ex/s] preprocess datasets: 3842ex [00:40, 101.12ex/s] preprocess datasets: 3853ex [00:40, 96.37ex/s] preprocess datasets: 3863ex [00:40, 93.72ex/s] preprocess datasets: 3874ex [00:40, 98.04ex/s] preprocess datasets: 3884ex [00:41, 92.46ex/s] preprocess datasets: 3894ex [00:41, 90.59ex/s] preprocess datasets: 3904ex [00:41, 86.94ex/s] preprocess datasets: 3915ex [00:41, 92.70ex/s] preprocess datasets: 3925ex [00:41, 90.00ex/s] preprocess datasets: 3936ex [00:41, 95.18ex/s] preprocess datasets: 3946ex [00:41, 95.87ex/s] preprocess datasets: 3956ex [00:41, 95.08ex/s] preprocess datasets: 3966ex [00:41, 95.62ex/s] preprocess datasets: 3976ex [00:42, 92.43ex/s] preprocess datasets: 3986ex [00:42, 89.73ex/s] preprocess datasets: 3996ex [00:42, 91.77ex/s] preprocess datasets: 4006ex [00:43, 30.61ex/s] preprocess datasets: 4017ex [00:43, 39.54ex/s] preprocess datasets: 4027ex [00:43, 47.80ex/s] preprocess datasets: 4036ex [00:43, 53.34ex/s] preprocess datasets: 4045ex [00:43, 58.27ex/s] preprocess datasets: 4055ex [00:43, 65.13ex/s] preprocess datasets: 4064ex [00:43, 69.51ex/s] preprocess datasets: 4074ex [00:43, 74.58ex/s] preprocess datasets: 4083ex [00:43, 75.37ex/s] preprocess datasets: 4092ex [00:44, 75.40ex/s] preprocess datasets: 4101ex [00:44, 78.31ex/s] preprocess datasets: 4110ex [00:44, 80.65ex/s] preprocess datasets: 4120ex [00:44, 84.39ex/s] preprocess datasets: 4131ex [00:44, 89.75ex/s] preprocess datasets: 4141ex [00:44, 85.15ex/s] preprocess datasets: 4151ex [00:44, 86.54ex/s] preprocess datasets: 4160ex [00:44, 84.63ex/s] preprocess datasets: 4169ex [00:45, 81.90ex/s] preprocess datasets: 4178ex [00:45, 81.32ex/s] preprocess datasets: 4187ex [00:45, 77.91ex/s] preprocess datasets: 4196ex [00:45, 79.13ex/s] preprocess datasets: 4204ex [00:45, 78.89ex/s] preprocess datasets: 4214ex [00:45, 83.90ex/s] preprocess datasets: 4223ex [00:45, 83.27ex/s] preprocess datasets: 4233ex [00:45, 86.73ex/s] preprocess datasets: 4244ex [00:45, 91.08ex/s] preprocess datasets: 4254ex [00:45, 92.93ex/s] preprocess datasets: 4264ex [00:46, 90.92ex/s] preprocess datasets: 4275ex [00:46, 95.87ex/s] preprocess datasets: 4285ex [00:46, 95.66ex/s] preprocess datasets: 4295ex [00:46, 93.41ex/s] preprocess datasets: 4307ex [00:46, 99.87ex/s] preprocess datasets: 4318ex [00:46, 102.34ex/s] preprocess datasets: 4329ex [00:46, 97.20ex/s] preprocess datasets: 4339ex [00:46, 88.69ex/s] preprocess datasets: 4349ex [00:47, 86.67ex/s] preprocess datasets: 4358ex [00:47, 85.78ex/s] preprocess datasets: 4367ex [00:47, 83.36ex/s] preprocess datasets: 4376ex [00:47, 79.00ex/s] preprocess datasets: 4385ex [00:47, 80.62ex/s] preprocess datasets: 4394ex [00:47, 81.41ex/s] preprocess datasets: 4404ex [00:47, 84.71ex/s] preprocess datasets: 4413ex [00:47, 84.23ex/s] preprocess datasets: 4422ex [00:47, 82.34ex/s] preprocess datasets: 4431ex [00:48, 84.20ex/s] preprocess datasets: 4440ex [00:48, 85.52ex/s] preprocess datasets: 4449ex [00:48, 86.27ex/s] preprocess datasets: 4458ex [00:48, 84.53ex/s] preprocess datasets: 4468ex [00:48, 87.45ex/s] preprocess datasets: 4478ex [00:48, 89.97ex/s] preprocess datasets: 4488ex [00:48, 90.33ex/s] preprocess datasets: 4498ex [00:48, 86.68ex/s] preprocess datasets: 4507ex [00:48, 86.03ex/s] preprocess datasets: 4516ex [00:48, 86.81ex/s] preprocess datasets: 4525ex [00:49, 86.56ex/s] preprocess datasets: 4534ex [00:49, 83.93ex/s] preprocess datasets: 4543ex [00:49, 84.16ex/s] preprocess datasets: 4554ex [00:49, 88.81ex/s] preprocess datasets: 4563ex [00:49, 87.78ex/s] preprocess datasets: 4573ex [00:49, 88.00ex/s] preprocess datasets: 4582ex [00:49, 83.89ex/s] preprocess datasets: 4591ex [00:49, 83.84ex/s] preprocess datasets: 4600ex [00:49, 82.54ex/s] preprocess datasets: 4611ex [00:50, 89.05ex/s] preprocess datasets: 4621ex [00:50, 91.89ex/s] preprocess datasets: 4631ex [00:50, 89.77ex/s] preprocess datasets: 4641ex [00:50, 83.32ex/s] preprocess datasets: 4650ex [00:50, 84.06ex/s] preprocess datasets: 4659ex [00:50, 83.85ex/s] preprocess datasets: 4670ex [00:50, 90.34ex/s] preprocess datasets: 4680ex [00:50, 87.08ex/s] preprocess datasets: 4689ex [00:51, 80.93ex/s] preprocess datasets: 4698ex [00:51, 77.58ex/s] preprocess datasets: 4707ex [00:51, 78.15ex/s] preprocess datasets: 4715ex [00:51, 78.47ex/s] preprocess datasets: 4725ex [00:51, 83.13ex/s] preprocess datasets: 4734ex [00:51, 81.61ex/s] preprocess datasets: 4744ex [00:51, 85.59ex/s] preprocess datasets: 4753ex [00:51, 81.66ex/s] preprocess datasets: 4763ex [00:51, 86.49ex/s] preprocess datasets: 4773ex [00:51, 89.65ex/s] preprocess datasets: 4784ex [00:52, 92.91ex/s] preprocess datasets: 4794ex [00:52, 88.37ex/s] preprocess datasets: 4803ex [00:52, 86.35ex/s] preprocess datasets: 4812ex [00:52, 87.13ex/s] preprocess datasets: 4821ex [00:52, 85.08ex/s] preprocess datasets: 4831ex [00:52, 87.45ex/s] preprocess datasets: 4840ex [00:52, 83.80ex/s] preprocess datasets: 4852ex [00:52, 93.42ex/s] preprocess datasets: 4867ex [00:52, 107.65ex/s] preprocess datasets: 4880ex [00:53, 110.61ex/s] preprocess datasets: 4892ex [00:53, 113.16ex/s] preprocess datasets: 4906ex [00:53, 120.00ex/s] preprocess datasets: 4922ex [00:53, 130.06ex/s] preprocess datasets: 4939ex [00:53, 140.42ex/s] preprocess datasets: 4954ex [00:53, 138.28ex/s] preprocess datasets: 4969ex [00:53, 139.61ex/s] preprocess datasets: 4983ex [00:53, 137.19ex/s] preprocess datasets: 4997ex [00:53, 136.41ex/s] preprocess datasets: 5011ex [00:54, 43.62ex/s] preprocess datasets: 5022ex [00:54, 51.17ex/s] preprocess datasets: 5033ex [00:54, 59.46ex/s] preprocess datasets: 5046ex [00:55, 70.98ex/s] preprocess datasets: 5061ex [00:55, 86.11ex/s] preprocess datasets: 5075ex [00:55, 95.90ex/s] preprocess datasets: 5088ex [00:55, 99.85ex/s] preprocess datasets: 5101ex [00:55, 105.57ex/s] preprocess datasets: 5114ex [00:55, 106.37ex/s] preprocess datasets: 5126ex [00:55, 108.24ex/s] preprocess datasets: 5138ex [00:55, 109.78ex/s] preprocess datasets: 5151ex [00:55, 113.69ex/s] preprocess datasets: 5163ex [00:56, 111.74ex/s] preprocess datasets: 5175ex [00:56, 113.06ex/s] preprocess datasets: 5189ex [00:56, 119.15ex/s] preprocess datasets: 5202ex [00:56, 118.43ex/s] preprocess datasets: 5216ex [00:56, 122.96ex/s] preprocess datasets: 5229ex [00:56, 119.10ex/s] preprocess datasets: 5241ex [00:56, 119.10ex/s] preprocess datasets: 5255ex [00:56, 123.44ex/s] preprocess datasets: 5268ex [00:56, 118.11ex/s] preprocess datasets: 5280ex [00:57, 113.61ex/s] preprocess datasets: 5292ex [00:57, 114.51ex/s] preprocess datasets: 5304ex [00:57, 112.20ex/s] preprocess datasets: 5316ex [00:57, 111.42ex/s] preprocess datasets: 5328ex [00:57, 106.56ex/s] preprocess datasets: 5344ex [00:57, 120.64ex/s] preprocess datasets: 5359ex [00:57, 127.34ex/s] preprocess datasets: 5372ex [00:57, 125.33ex/s] preprocess datasets: 5385ex [00:57, 124.46ex/s] preprocess datasets: 5399ex [00:58, 127.32ex/s] preprocess datasets: 5412ex [00:58, 121.90ex/s] preprocess datasets: 5425ex [00:58, 118.99ex/s] preprocess datasets: 5437ex [00:58, 119.05ex/s] preprocess datasets: 5450ex [00:58, 119.10ex/s] preprocess datasets: 5465ex [00:58, 126.12ex/s] preprocess datasets: 5478ex [00:58, 117.45ex/s] preprocess datasets: 5490ex [00:58, 112.71ex/s] preprocess datasets: 5502ex [00:58, 114.00ex/s] preprocess datasets: 5516ex [00:59, 118.69ex/s] preprocess datasets: 5531ex [00:59, 126.34ex/s] preprocess datasets: 5544ex [00:59, 124.51ex/s] preprocess datasets: 5557ex [00:59, 124.82ex/s] preprocess datasets: 5570ex [00:59, 121.27ex/s] preprocess datasets: 5583ex [00:59, 116.83ex/s] preprocess datasets: 5595ex [00:59, 117.69ex/s] preprocess datasets: 5607ex [00:59, 110.47ex/s] preprocess datasets: 5621ex [00:59, 116.64ex/s] preprocess datasets: 5633ex [01:00, 114.80ex/s] preprocess datasets: 5645ex [01:00, 115.14ex/s] preprocess datasets: 5657ex [01:00, 112.38ex/s] preprocess datasets: 5669ex [01:00, 112.81ex/s] preprocess datasets: 5681ex [01:00, 109.24ex/s] preprocess datasets: 5695ex [01:00, 117.10ex/s] preprocess datasets: 5705ex [01:00, 94.09ex/s] preprocess datasets: 0ex [00:00, ?ex/s] preprocess datasets: 8ex [00:00, 77.71ex/s] preprocess datasets: 19ex [00:00, 92.78ex/s] preprocess datasets: 31ex [00:00, 103.82ex/s] preprocess datasets: 42ex [00:00, 95.92ex/s] preprocess datasets: 52ex [00:00, 91.68ex/s] preprocess datasets: 62ex [00:00, 88.57ex/s] preprocess datasets: 71ex [00:00, 87.32ex/s] preprocess datasets: 80ex [00:00, 87.07ex/s] preprocess datasets: 89ex [00:01, 84.91ex/s] preprocess datasets: 98ex [00:01, 85.23ex/s] preprocess datasets: 107ex [00:01, 81.78ex/s] preprocess datasets: 116ex [00:01, 83.49ex/s] preprocess datasets: 126ex [00:01, 87.29ex/s] preprocess datasets: 135ex [00:01, 86.41ex/s] preprocess datasets: 144ex [00:01, 86.79ex/s] preprocess datasets: 153ex [00:01, 87.55ex/s] preprocess datasets: 163ex [00:01, 90.33ex/s] preprocess datasets: 173ex [00:01, 89.99ex/s] preprocess datasets: 183ex [00:02, 86.86ex/s] preprocess datasets: 194ex [00:02, 91.46ex/s] preprocess datasets: 204ex [00:02, 88.28ex/s] preprocess datasets: 213ex [00:02, 86.51ex/s] preprocess datasets: 222ex [00:02, 84.34ex/s] preprocess datasets: 232ex [00:02, 86.84ex/s] preprocess datasets: 242ex [00:02, 88.49ex/s] preprocess datasets: 254ex [00:02, 95.58ex/s] preprocess datasets: 264ex [00:02, 93.21ex/s] preprocess datasets: 274ex [00:03, 92.83ex/s] preprocess datasets: 287ex [00:03, 103.11ex/s] preprocess datasets: 298ex [00:03, 95.93ex/s] preprocess datasets: 308ex [00:03, 93.05ex/s] preprocess datasets: 318ex [00:03, 88.64ex/s] preprocess datasets: 330ex [00:03, 93.92ex/s] preprocess datasets: 340ex [00:03, 93.91ex/s] preprocess datasets: 350ex [00:03, 92.64ex/s] preprocess datasets: 360ex [00:03, 91.97ex/s] preprocess datasets: 370ex [00:04, 93.46ex/s] preprocess datasets: 380ex [00:04, 92.49ex/s] preprocess datasets: 390ex [00:04, 92.43ex/s] preprocess datasets: 402ex [00:04, 99.33ex/s] preprocess datasets: 413ex [00:04, 100.96ex/s] preprocess datasets: 424ex [00:04, 95.21ex/s] preprocess datasets: 434ex [00:04, 92.81ex/s] preprocess datasets: 444ex [00:04, 91.75ex/s] preprocess datasets: 454ex [00:04, 93.24ex/s] preprocess datasets: 464ex [00:05, 93.83ex/s] preprocess datasets: 474ex [00:05, 93.75ex/s] preprocess datasets: 484ex [00:05, 95.08ex/s] preprocess datasets: 495ex [00:05, 97.26ex/s] preprocess datasets: 507ex [00:05, 103.70ex/s] preprocess datasets: 518ex [00:05, 103.68ex/s] preprocess datasets: 529ex [00:05, 100.81ex/s] preprocess datasets: 540ex [00:05, 102.17ex/s] preprocess datasets: 551ex [00:05, 102.70ex/s] preprocess datasets: 562ex [00:06, 103.74ex/s] preprocess datasets: 573ex [00:06, 102.21ex/s] preprocess datasets: 584ex [00:06, 97.33ex/s] preprocess datasets: 597ex [00:06, 105.50ex/s] preprocess datasets: 609ex [00:06, 109.12ex/s] preprocess datasets: 621ex [00:06, 105.87ex/s] preprocess datasets: 632ex [00:06, 97.88ex/s] preprocess datasets: 642ex [00:06, 94.65ex/s] preprocess datasets: 654ex [00:06, 99.70ex/s] preprocess datasets: 667ex [00:07, 106.84ex/s] preprocess datasets: 678ex [00:07, 102.63ex/s] preprocess datasets: 689ex [00:07, 102.64ex/s] preprocess datasets: 700ex [00:07, 95.70ex/s] preprocess datasets: 711ex [00:07, 97.74ex/s] preprocess datasets: 721ex [00:07, 94.26ex/s] preprocess datasets: 732ex [00:07, 97.69ex/s] preprocess datasets: 742ex [00:07, 94.22ex/s] preprocess datasets: 752ex [00:07, 89.85ex/s] preprocess datasets: 762ex [00:08, 89.74ex/s] preprocess datasets: 772ex [00:08, 88.47ex/s] preprocess datasets: 781ex [00:08, 86.26ex/s] preprocess datasets: 790ex [00:08, 83.02ex/s] preprocess datasets: 799ex [00:08, 81.06ex/s] preprocess datasets: 808ex [00:08, 82.41ex/s] preprocess datasets: 817ex [00:08, 84.07ex/s] preprocess datasets: 828ex [00:08, 90.15ex/s] preprocess datasets: 839ex [00:08, 94.93ex/s] preprocess datasets: 849ex [00:09, 94.50ex/s] preprocess datasets: 859ex [00:09, 90.89ex/s] preprocess datasets: 871ex [00:09, 98.59ex/s] preprocess datasets: 881ex [00:09, 97.89ex/s] preprocess datasets: 892ex [00:09, 99.19ex/s] preprocess datasets: 902ex [00:09, 96.39ex/s] preprocess datasets: 912ex [00:09, 94.40ex/s] preprocess datasets: 923ex [00:09, 96.95ex/s] preprocess datasets: 934ex [00:09, 99.43ex/s] preprocess datasets: 947ex [00:10, 106.00ex/s] preprocess datasets: 958ex [00:10, 102.33ex/s] preprocess datasets: 970ex [00:10, 105.87ex/s] preprocess datasets: 982ex [00:10, 108.34ex/s] preprocess datasets: 993ex [00:10, 101.62ex/s] preprocess datasets: 1004ex [00:11, 33.77ex/s] preprocess datasets: 1014ex [00:11, 40.96ex/s] preprocess datasets: 1025ex [00:11, 50.42ex/s] preprocess datasets: 1037ex [00:11, 61.81ex/s] preprocess datasets: 1048ex [00:11, 70.91ex/s] preprocess datasets: 1060ex [00:11, 80.90ex/s] preprocess datasets: 1071ex [00:11, 86.03ex/s] preprocess datasets: 1082ex [00:12, 88.28ex/s] preprocess datasets: 1093ex [00:12, 84.92ex/s] preprocess datasets: 1103ex [00:12, 86.49ex/s] preprocess datasets: 1113ex [00:12, 89.80ex/s] preprocess datasets: 1123ex [00:12, 88.72ex/s] preprocess datasets: 1134ex [00:12, 93.44ex/s] preprocess datasets: 1145ex [00:12, 96.53ex/s] preprocess datasets: 1155ex [00:12, 94.36ex/s] preprocess datasets: 1165ex [00:12, 94.66ex/s] preprocess datasets: 1175ex [00:13, 94.33ex/s] preprocess datasets: 1186ex [00:13, 96.67ex/s] preprocess datasets: 1197ex [00:13, 99.86ex/s] preprocess datasets: 1208ex [00:13, 92.83ex/s] preprocess datasets: 1218ex [00:13, 89.98ex/s] preprocess datasets: 1228ex [00:13, 88.04ex/s] preprocess datasets: 1238ex [00:13, 87.35ex/s] preprocess datasets: 1247ex [00:13, 84.82ex/s] preprocess datasets: 1256ex [00:14, 83.20ex/s] preprocess datasets: 1267ex [00:14, 88.90ex/s] preprocess datasets: 1276ex [00:14, 84.79ex/s] preprocess datasets: 1286ex [00:14, 87.69ex/s] preprocess datasets: 1295ex [00:14, 82.15ex/s] preprocess datasets: 1304ex [00:14, 81.14ex/s] preprocess datasets: 1313ex [00:14, 77.13ex/s] preprocess datasets: 1323ex [00:14, 82.11ex/s] preprocess datasets: 1333ex [00:14, 86.44ex/s] preprocess datasets: 1342ex [00:15, 87.36ex/s] preprocess datasets: 1351ex [00:15, 84.75ex/s] preprocess datasets: 1360ex [00:15, 81.41ex/s] preprocess datasets: 1369ex [00:15, 82.77ex/s] preprocess datasets: 1378ex [00:15, 82.62ex/s] preprocess datasets: 1388ex [00:15, 85.29ex/s] preprocess datasets: 1397ex [00:15, 85.52ex/s] preprocess datasets: 1407ex [00:15, 89.03ex/s] preprocess datasets: 1417ex [00:15, 91.52ex/s] preprocess datasets: 1427ex [00:16, 91.88ex/s] preprocess datasets: 1440ex [00:16, 99.95ex/s] preprocess datasets: 1452ex [00:16, 102.29ex/s] preprocess datasets: 1463ex [00:16, 95.60ex/s] preprocess datasets: 1473ex [00:16, 96.07ex/s] preprocess datasets: 1483ex [00:16, 96.21ex/s] preprocess datasets: 1493ex [00:16, 92.69ex/s] preprocess datasets: 1503ex [00:16, 91.71ex/s] preprocess datasets: 1513ex [00:16, 91.90ex/s] preprocess datasets: 1523ex [00:17, 91.12ex/s] preprocess datasets: 1533ex [00:17, 93.42ex/s] preprocess datasets: 1543ex [00:17, 94.79ex/s] preprocess datasets: 1553ex [00:17, 94.75ex/s] preprocess datasets: 1563ex [00:17, 89.95ex/s] preprocess datasets: 1573ex [00:17, 86.70ex/s] preprocess datasets: 1582ex [00:17, 85.62ex/s] preprocess datasets: 1591ex [00:17, 84.11ex/s] preprocess datasets: 1600ex [00:17, 84.29ex/s] preprocess datasets: 1609ex [00:18, 80.60ex/s] preprocess datasets: 1618ex [00:18, 80.68ex/s] preprocess datasets: 1627ex [00:18, 81.69ex/s] preprocess datasets: 1636ex [00:18, 80.89ex/s] preprocess datasets: 1647ex [00:18, 86.51ex/s] preprocess datasets: 1656ex [00:18, 84.46ex/s] preprocess datasets: 1666ex [00:18, 88.33ex/s] preprocess datasets: 1676ex [00:18, 88.63ex/s] preprocess datasets: 1687ex [00:18, 93.16ex/s] preprocess datasets: 1697ex [00:19, 94.22ex/s] preprocess datasets: 1707ex [00:19, 91.42ex/s] preprocess datasets: 1717ex [00:19, 90.53ex/s] preprocess datasets: 1727ex [00:19, 91.24ex/s] preprocess datasets: 1737ex [00:19, 87.50ex/s] preprocess datasets: 1747ex [00:19, 89.51ex/s] preprocess datasets: 1756ex [00:19, 86.60ex/s] preprocess datasets: 1766ex [00:19, 88.84ex/s] preprocess datasets: 1777ex [00:19, 92.88ex/s] preprocess datasets: 1787ex [00:20, 86.17ex/s] preprocess datasets: 1796ex [00:20, 82.25ex/s] preprocess datasets: 1805ex [00:20, 80.99ex/s] preprocess datasets: 1814ex [00:20, 78.12ex/s] preprocess datasets: 1823ex [00:20, 79.55ex/s] preprocess datasets: 1832ex [00:20, 80.64ex/s] preprocess datasets: 1842ex [00:20, 85.23ex/s] preprocess datasets: 1851ex [00:20, 84.40ex/s] preprocess datasets: 1861ex [00:20, 87.43ex/s] preprocess datasets: 1870ex [00:21, 83.76ex/s] preprocess datasets: 1879ex [00:21, 84.55ex/s] preprocess datasets: 1888ex [00:21, 85.51ex/s] preprocess datasets: 1898ex [00:21, 88.57ex/s] preprocess datasets: 1907ex [00:21, 87.80ex/s] preprocess datasets: 1916ex [00:21, 87.25ex/s] preprocess datasets: 1926ex [00:21, 90.34ex/s] preprocess datasets: 1936ex [00:21, 87.23ex/s] preprocess datasets: 1946ex [00:21, 89.53ex/s] preprocess datasets: 1955ex [00:22, 88.17ex/s] preprocess datasets: 1964ex [00:22, 87.57ex/s] preprocess datasets: 1973ex [00:22, 86.65ex/s] preprocess datasets: 1982ex [00:22, 86.71ex/s] preprocess datasets: 1992ex [00:22, 89.83ex/s] preprocess datasets: 2001ex [00:23, 26.54ex/s] preprocess datasets: 2010ex [00:23, 33.16ex/s] preprocess datasets: 2018ex [00:23, 39.15ex/s] preprocess datasets: 2027ex [00:23, 47.21ex/s] preprocess datasets: 2035ex [00:23, 53.11ex/s] preprocess datasets: 2044ex [00:23, 60.06ex/s] preprocess datasets: 2053ex [00:23, 65.53ex/s] preprocess datasets: 2062ex [00:24, 70.56ex/s] preprocess datasets: 2071ex [00:24, 72.94ex/s] preprocess datasets: 2080ex [00:24, 76.17ex/s] preprocess datasets: 2090ex [00:24, 79.21ex/s] preprocess datasets: 2100ex [00:24, 82.32ex/s] preprocess datasets: 2110ex [00:24, 86.42ex/s] preprocess datasets: 2119ex [00:24, 86.30ex/s] preprocess datasets: 2129ex [00:24, 90.01ex/s] preprocess datasets: 2140ex [00:24, 94.08ex/s] preprocess datasets: 2150ex [00:25, 88.56ex/s] preprocess datasets: 2159ex [00:25, 88.34ex/s] preprocess datasets: 2168ex [00:25, 88.13ex/s] preprocess datasets: 2178ex [00:25, 91.00ex/s] preprocess datasets: 2188ex [00:25, 88.01ex/s] preprocess datasets: 2197ex [00:25, 87.28ex/s] preprocess datasets: 2206ex [00:25, 83.32ex/s] preprocess datasets: 2215ex [00:25, 84.99ex/s] preprocess datasets: 2225ex [00:25, 87.52ex/s] preprocess datasets: 2234ex [00:26, 86.81ex/s] preprocess datasets: 2243ex [00:26, 87.35ex/s] preprocess datasets: 2252ex [00:26, 84.82ex/s] preprocess datasets: 2262ex [00:26, 88.96ex/s] preprocess datasets: 2274ex [00:26, 97.03ex/s] preprocess datasets: 2288ex [00:26, 108.25ex/s] preprocess datasets: 2301ex [00:26, 113.83ex/s] preprocess datasets: 2314ex [00:26, 115.40ex/s] preprocess datasets: 2326ex [00:26, 108.30ex/s] preprocess datasets: 2339ex [00:27, 110.97ex/s] preprocess datasets: 2351ex [00:27, 110.24ex/s] preprocess datasets: 2365ex [00:27, 116.86ex/s] preprocess datasets: 2377ex [00:27, 117.60ex/s] preprocess datasets: 2390ex [00:27, 120.27ex/s] preprocess datasets: 2406ex [00:27, 130.36ex/s] preprocess datasets: 2420ex [00:27, 132.10ex/s] preprocess datasets: 2434ex [00:27, 129.56ex/s] preprocess datasets: 2447ex [00:27, 126.98ex/s] preprocess datasets: 2460ex [00:27, 120.45ex/s] preprocess datasets: 2473ex [00:28, 119.84ex/s] preprocess datasets: 2486ex [00:28, 117.62ex/s] preprocess datasets: 2498ex [00:28, 114.30ex/s] preprocess datasets: 2510ex [00:28, 115.34ex/s] preprocess datasets: 2525ex [00:28, 123.85ex/s] preprocess datasets: 2539ex [00:28, 126.50ex/s] preprocess datasets: 2553ex [00:28, 129.77ex/s] preprocess datasets: 2571ex [00:28, 140.69ex/s] preprocess datasets: 2586ex [00:28, 141.61ex/s] preprocess datasets: 2601ex [00:29, 143.63ex/s] preprocess datasets: 2609ex [00:29, 89.62ex/s] 0%| | 0/6 [00:00 to the vocabulary Adding to the vocabulary /home/sagrilaft/Project/audio/xls-r-et/./ is already a clone of https://huggingface.co/shpotes/xls-r-et. Make sure you pull the latest changes with `repo.git_pull()`. Using amp half precision backend The following columns in the training set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length. ***** Running training ***** Num examples = 5705 Num Epochs = 500 Instantaneous batch size per device = 80 Total train batch size (w. parallel, distributed & accumulation) = 160 Gradient Accumulation steps = 2 Total optimization steps = 18000 Automatic Weights & Biases logging enabled, to disable set os.environ["WANDB_DISABLED"] = "true" wandb: Currently logged in as: shpotes (use `wandb login --relogin` to force relogin) wandb: Tracking run with wandb version 0.12.9 wandb: Syncing run cosine+drop_proj+low_specaugment-300M wandb: View project at https://wandb.ai/shpotes/xls-r-estonian wandb: View run at https://wandb.ai/shpotes/xls-r-estonian/runs/1xdiy2kf wandb: Run data is saved locally in /home/sagrilaft/Project/audio/xls-r-et/wandb/run-20220126_105847-1xdiy2kf wandb: Run `wandb offline` to turn off syncing. 0% 0/18000 [00:00