First commit

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": 47,
+  "transformers_version": "4.5.0.dev0",
+  "vocab_size": 48
+}

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 sha256:30dd761aa957f2bcb00e94b04f5edf09ac416b6577ac222c67020567e8e6f09d
+size 1262130647

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 +1,170 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 29.528428093645484,
+  "global_step": 4400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 2.4877,
+      "step": 400
+    },
+    {
+      "epoch": 2.68,
+      "eval_loss": 1.0936583280563354,
+      "eval_runtime": 158.101,
+      "eval_samples_per_second": 13.2,
+      "eval_wer": 0.8853237689026755,
+      "step": 400
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.0002773299748110831,
+      "loss": 0.7501,
+      "step": 800
+    },
+    {
+      "epoch": 5.37,
+      "eval_loss": 0.8788074851036072,
+      "eval_runtime": 158.2433,
+      "eval_samples_per_second": 13.189,
+      "eval_wer": 0.768030244280729,
+      "step": 800
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 0.00024710327455919394,
+      "loss": 0.4114,
+      "step": 1200
+    },
+    {
+      "epoch": 8.05,
+      "eval_loss": 0.8623701333999634,
+      "eval_runtime": 160.5429,
+      "eval_samples_per_second": 13.0,
+      "eval_wer": 0.702016285381931,
+      "step": 1200
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 0.00021687657430730477,
+      "loss": 0.2676,
+      "step": 1600
+    },
+    {
+      "epoch": 10.74,
+      "eval_loss": 0.8885830640792847,
+      "eval_runtime": 159.9177,
+      "eval_samples_per_second": 13.05,
+      "eval_wer": 0.6879604497867391,
+      "step": 1600
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 0.0001866498740554156,
+      "loss": 0.194,
+      "step": 2000
+    },
+    {
+      "epoch": 13.42,
+      "eval_loss": 0.9405023455619812,
+      "eval_runtime": 154.7945,
+      "eval_samples_per_second": 13.482,
+      "eval_wer": 0.6577161690577743,
+      "step": 2000
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 0.00015642317380352644,
+      "loss": 0.1496,
+      "step": 2400
+    },
+    {
+      "epoch": 16.11,
+      "eval_loss": 1.0035845041275024,
+      "eval_runtime": 153.8389,
+      "eval_samples_per_second": 13.566,
+      "eval_wer": 0.6553896859247771,
+      "step": 2400
+    },
+    {
+      "epoch": 18.79,
+      "learning_rate": 0.00012619647355163727,
+      "loss": 0.1227,
+      "step": 2800
+    },
+    {
+      "epoch": 18.79,
+      "eval_loss": 0.9778856635093689,
+      "eval_runtime": 153.4023,
+      "eval_samples_per_second": 13.605,
+      "eval_wer": 0.6457929430011632,
+      "step": 2800
+    },
+    {
+      "epoch": 21.47,
+      "learning_rate": 9.59697732997481e-05,
+      "loss": 0.0982,
+      "step": 3200
+    },
+    {
+      "epoch": 21.47,
+      "eval_loss": 1.0781689882278442,
+      "eval_runtime": 154.2908,
+      "eval_samples_per_second": 13.526,
+      "eval_wer": 0.6339666537417604,
+      "step": 3200
+    },
+    {
+      "epoch": 24.16,
+      "learning_rate": 6.574307304785894e-05,
+      "loss": 0.0852,
+      "step": 3600
+    },
+    {
+      "epoch": 24.16,
+      "eval_loss": 1.066897988319397,
+      "eval_runtime": 154.9275,
+      "eval_samples_per_second": 13.471,
+      "eval_wer": 0.6234974796432726,
+      "step": 3600
+    },
+    {
+      "epoch": 26.84,
+      "learning_rate": 3.551637279596977e-05,
+      "loss": 0.0715,
+      "step": 4000
+    },
+    {
+      "epoch": 26.84,
+      "eval_loss": 1.0754976272583008,
+      "eval_runtime": 156.8919,
+      "eval_samples_per_second": 13.302,
+      "eval_wer": 0.6209771229158588,
+      "step": 4000
+    },
+    {
+      "epoch": 29.53,
+      "learning_rate": 5.289672544080604e-06,
+      "loss": 0.0678,
+      "step": 4400
+    },
+    {
+      "epoch": 29.53,
+      "eval_loss": 1.0621362924575806,
+      "eval_runtime": 157.2718,
+      "eval_samples_per_second": 13.27,
+      "eval_wer": 0.6191353237689027,
+      "step": 4400
+    }
+  ],
+  "max_steps": 4470,
+  "num_train_epochs": 30,
+  "total_flos": 1.2337798222367355e+19,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40246ed3ec9b8172ded7d6279be610159fca8d75d6448bcbab83840ad4318c6c
+size 2287

vocab.json

@@ -0,0 +1 @@
+{"à": 0, "ù": 1, "«": 2, "x": 3, "’": 5, "c": 6, "w": 7, "b": 8, "(": 9, "é": 10, "j": 11, "r": 12, "h": 13, "p": 14, "z": 15, "l": 16, "d": 17, "'": 18, "n": 19, "/": 20, "è": 21, "ê": 22, "u": 23, ")": 24, "i": 25, "…": 26, "o": 27, "½": 28, "ʼ": 29, "m": 30, "a": 31, "t": 32, "»": 33, "e": 34, "ü": 35, "û": 36, "y": 37, "v": 38, "k": 39, "â": 40, "f": 41, "q": 42, "g": 43, "ñ": 44, "s": 45, "|": 4, "[UNK]": 46, "[PAD]": 47}