Initial commit from mrm8488

config.json

@@ -0,0 +1,76 @@
+{
+  "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
+  "activation_dropout": 0.0,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 1,
+  "conv_bias": true,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "mean",
+  "ctc_zero_infinity": false,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "layer",
+  "feat_proj_dropout": 0.0,
+  "final_dropout": 0.0,
+  "gradient_checkpointing": true,
+  "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_channel_length": 10,
+  "mask_channel_min_space": 1,
+  "mask_channel_other": 0.0,
+  "mask_channel_prob": 0.0,
+  "mask_channel_selection": "static",
+  "mask_feature_length": 10,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_space": 1,
+  "mask_time_other": 0.0,
+  "mask_time_prob": 0.05,
+  "mask_time_selection": "static",
+  "model_type": "wav2vec2",
+  "num_attention_heads": 16,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 24,
+  "pad_token_id": 77,
+  "transformers_version": "4.5.0.dev0",
+  "vocab_size": 78
+}

preprocessor_config.json

@@ -0,0 +1,8 @@
+{
+  "do_normalize": true,
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9964d894b2e9c738e253a5cad8fe874ca415784d33db5eb657f567093fc73607
+size 1262253655

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|"}

trainer_state.json

@@ -0,0 +1,268 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 28.458498023715414,
+  "global_step": 7200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 6.005,
+      "step": 400
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 3.0256412029266357,
+      "eval_runtime": 103.2564,
+      "eval_samples_per_second": 7.302,
+      "eval_wer": 1.0,
+      "step": 400
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0002873060648801128,
+      "loss": 2.0338,
+      "step": 800
+    },
+    {
+      "epoch": 3.16,
+      "eval_loss": 0.6479684710502625,
+      "eval_runtime": 104.5189,
+      "eval_samples_per_second": 7.214,
+      "eval_wer": 0.5284067231559939,
+      "step": 800
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.0002703808180535966,
+      "loss": 0.399,
+      "step": 1200
+    },
+    {
+      "epoch": 4.74,
+      "eval_loss": 0.49810782074928284,
+      "eval_runtime": 104.6121,
+      "eval_samples_per_second": 7.208,
+      "eval_wer": 0.4114460341714127,
+      "step": 1200
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00025345557122708037,
+      "loss": 0.2602,
+      "step": 1600
+    },
+    {
+      "epoch": 6.32,
+      "eval_loss": 0.49113723635673523,
+      "eval_runtime": 105.9638,
+      "eval_samples_per_second": 7.116,
+      "eval_wer": 0.38324767328795667,
+      "step": 1600
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 0.00023653032440056415,
+      "loss": 0.1991,
+      "step": 2000
+    },
+    {
+      "epoch": 7.91,
+      "eval_loss": 0.5186901092529297,
+      "eval_runtime": 104.8564,
+      "eval_samples_per_second": 7.191,
+      "eval_wer": 0.3838033060147243,
+      "step": 2000
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 0.00021960507757404793,
+      "loss": 0.159,
+      "step": 2400
+    },
+    {
+      "epoch": 9.49,
+      "eval_loss": 0.5159746408462524,
+      "eval_runtime": 100.948,
+      "eval_samples_per_second": 7.469,
+      "eval_wer": 0.3706070287539936,
+      "step": 2400
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 0.0002026798307475317,
+      "loss": 0.1378,
+      "step": 2800
+    },
+    {
+      "epoch": 11.07,
+      "eval_loss": 0.5382620692253113,
+      "eval_runtime": 100.0461,
+      "eval_samples_per_second": 7.537,
+      "eval_wer": 0.36741214057507987,
+      "step": 2800
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 0.00018575458392101551,
+      "loss": 0.116,
+      "step": 3200
+    },
+    {
+      "epoch": 12.65,
+      "eval_loss": 0.5114040970802307,
+      "eval_runtime": 101.876,
+      "eval_samples_per_second": 7.401,
+      "eval_wer": 0.3560216696763439,
+      "step": 3200
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 0.0001688293370944993,
+      "loss": 0.1022,
+      "step": 3600
+    },
+    {
+      "epoch": 14.23,
+      "eval_loss": 0.5726804733276367,
+      "eval_runtime": 101.0522,
+      "eval_samples_per_second": 7.461,
+      "eval_wer": 0.36352271148770665,
+      "step": 3600
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 0.00015190409026798307,
+      "loss": 0.0896,
+      "step": 4000
+    },
+    {
+      "epoch": 15.81,
+      "eval_loss": 0.5884941220283508,
+      "eval_runtime": 101.7591,
+      "eval_samples_per_second": 7.41,
+      "eval_wer": 0.357271843311571,
+      "step": 4000
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 0.00013497884344146685,
+      "loss": 0.0794,
+      "step": 4400
+    },
+    {
+      "epoch": 17.39,
+      "eval_loss": 0.5713954567909241,
+      "eval_runtime": 102.4951,
+      "eval_samples_per_second": 7.356,
+      "eval_wer": 0.3504653424086679,
+      "step": 4400
+    },
+    {
+      "epoch": 18.97,
+      "learning_rate": 0.00011805359661495063,
+      "loss": 0.0719,
+      "step": 4800
+    },
+    {
+      "epoch": 18.97,
+      "eval_loss": 0.5819716453552246,
+      "eval_runtime": 100.7538,
+      "eval_samples_per_second": 7.484,
+      "eval_wer": 0.34879844422836503,
+      "step": 4800
+    },
+    {
+      "epoch": 20.55,
+      "learning_rate": 0.0001011283497884344,
+      "loss": 0.0602,
+      "step": 5200
+    },
+    {
+      "epoch": 20.55,
+      "eval_loss": 0.6186490654945374,
+      "eval_runtime": 101.0143,
+      "eval_samples_per_second": 7.464,
+      "eval_wer": 0.3453257396860675,
+      "step": 5200
+    },
+    {
+      "epoch": 22.13,
+      "learning_rate": 8.420310296191819e-05,
+      "loss": 0.0604,
+      "step": 5600
+    },
+    {
+      "epoch": 22.13,
+      "eval_loss": 0.6135568618774414,
+      "eval_runtime": 101.9403,
+      "eval_samples_per_second": 7.396,
+      "eval_wer": 0.3394915960550076,
+      "step": 5600
+    },
+    {
+      "epoch": 23.72,
+      "learning_rate": 6.727785613540198e-05,
+      "loss": 0.0532,
+      "step": 6000
+    },
+    {
+      "epoch": 23.72,
+      "eval_loss": 0.594706654548645,
+      "eval_runtime": 100.6719,
+      "eval_samples_per_second": 7.49,
+      "eval_wer": 0.34324211696068896,
+      "step": 6000
+    },
+    {
+      "epoch": 25.3,
+      "learning_rate": 5.0352609308885744e-05,
+      "loss": 0.0457,
+      "step": 6400
+    },
+    {
+      "epoch": 25.3,
+      "eval_loss": 0.6127145886421204,
+      "eval_runtime": 101.5213,
+      "eval_samples_per_second": 7.427,
+      "eval_wer": 0.3442144742325323,
+      "step": 6400
+    },
+    {
+      "epoch": 26.88,
+      "learning_rate": 3.342736248236953e-05,
+      "loss": 0.0478,
+      "step": 6800
+    },
+    {
+      "epoch": 26.88,
+      "eval_loss": 0.6120516061782837,
+      "eval_runtime": 102.6766,
+      "eval_samples_per_second": 7.343,
+      "eval_wer": 0.336157799694402,
+      "step": 6800
+    },
+    {
+      "epoch": 28.46,
+      "learning_rate": 1.650211565585331e-05,
+      "loss": 0.0406,
+      "step": 7200
+    },
+    {
+      "epoch": 28.46,
+      "eval_loss": 0.6265799403190613,
+      "eval_runtime": 104.4584,
+      "eval_samples_per_second": 7.218,
+      "eval_wer": 0.33282400333379636,
+      "step": 7200
+    }
+  ],
+  "max_steps": 7590,
+  "num_train_epochs": 30,
+  "total_flos": 4.099942107044385e+19,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9474bdaf87aa787d1020bec02dafa9b781ca4c63af73ca8033ddec02fd423a4d
+size 2287

vocab.json

@@ -0,0 +1 @@
+{"o": 0, "r": 1, "ü": 2, "n": 3, "ə": 4, "ã": 5, "ô": 6, "e": 7, "x": 8, "q": 9, "ō": 10, "ê": 11, "–": 12, "d": 13, "毵": 14, "л": 15, "ş": 16, "í": 17, "â": 18, "т": 19, "=": 20, "¿": 21, "s": 22, "h": 24, "p": 25, "ы": 26, "¡": 27, "g": 28, "é": 29, "ñ": 30, "j": 31, "к": 32, "е": 33, "y": 34, "ú": 35, "w": 36, "ø": 37, "ʻ": 38, "б": 39, "ö": 40, "о": 41, "i": 42, "ä": 43, "ë": 44, "р": 45, "а": 46, "ő": 47, "v": 48, "и": 49, "ө": 50, "m": 51, "ò": 52, "’": 53, "´": 54, "'": 55, "н": 56, "t": 57, "š": 58, "ð": 59, "á": 60, "c": 61, "—": 62, "ү": 63, "п": 64, "°": 65, "k": 66, "ū": 67, "z": 68, "a": 69, "ш": 70, "l": 71, "b": 72, "ó": 73, "f": 74, "u": 75, "|": 23, "[UNK]": 76, "[PAD]": 77}