Update from

checkpoint-2400/config.json

@@ -0,0 +1,84 @@
+{
+  "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
+  "activation_dropout": 0.0,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.05,
+  "bos_token_id": 1,
+  "codevector_dim": 768,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": true,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "mean",
+  "ctc_zero_infinity": false,
+  "diversity_loss_weight": 0.1,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "layer",
+  "feat_proj_dropout": 0.0,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.0,
+  "gradient_checkpointing": true,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.05,
+  "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_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,
+  "pad_token_id": 37,
+  "proj_codevector_dim": 768,
+  "transformers_version": "4.4.0",
+  "vocab_size": 38
+}

checkpoint-2400/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
+}

checkpoint-2400/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c57ed53272b57d91e7b12b2d76cce9d9168e2a2d8cb54cbbb8bdf78c9ba44f5
+size 1262079473

checkpoint-2400/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d2eb2798cdbca9c3a4d8db40b57e982dc36dd6ec9d325cdd869685b0ed68d30
+size 623

checkpoint-2400/trainer_state.json

@@ -0,0 +1,100 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 30.0,
+  "global_step": 2400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 5.0,
+      "learning_rate": 8e-05,
+      "loss": 9.4971,
+      "step": 400
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 3.1448328495025635,
+      "eval_runtime": 45.2941,
+      "eval_samples_per_second": 14.13,
+      "eval_wer": 1.0,
+      "step": 400
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 8.421052631578948e-05,
+      "loss": 2.4507,
+      "step": 800
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 1.3176685571670532,
+      "eval_runtime": 45.3981,
+      "eval_samples_per_second": 14.098,
+      "eval_wer": 1.029591836734694,
+      "step": 800
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 6.31578947368421e-05,
+      "loss": 0.6394,
+      "step": 1200
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 1.0083409547805786,
+      "eval_runtime": 45.4018,
+      "eval_samples_per_second": 14.096,
+      "eval_wer": 0.7188775510204082,
+      "step": 1200
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 0.3395,
+      "step": 1600
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 1.0249711275100708,
+      "eval_runtime": 45.9714,
+      "eval_samples_per_second": 13.922,
+      "eval_wer": 0.6642857142857143,
+      "step": 1600
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 0.2434,
+      "step": 2000
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 1.1575790643692017,
+      "eval_runtime": 46.2674,
+      "eval_samples_per_second": 13.833,
+      "eval_wer": 0.6357142857142857,
+      "step": 2000
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 0.0,
+      "loss": 0.2048,
+      "step": 2400
+    },
+    {
+      "epoch": 30.0,
+      "eval_loss": 1.2169420719146729,
+      "eval_runtime": 49.8509,
+      "eval_samples_per_second": 12.838,
+      "eval_wer": 0.625,
+      "step": 2400
+    }
+  ],
+  "max_steps": 2400,
+  "num_train_epochs": 30,
+  "total_flos": 7.003249081477079e+18,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2400/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c89d8a6fb4c2849f998a76427cb73f01c7599f4d172a0e01b3e96a7b39fee63b
+size 2287

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
+}

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": "|"}

vocab.json

@@ -0,0 +1 @@
+{"f": 0, "o": 1, "’": 2, "é": 3, "t": 4, "î": 5, "p": 6, "x": 7, "s": 8, "h": 9, "è": 10, "â": 11, "u": 12, "v": 13, "y": 14, "e": 15, "r": 16, "c": 17, "ë": 18, "b": 19, "'": 20, "m": 21, "k": 22, "z": 23, "g": 25, "œ": 26, "ô": 27, "w": 28, "n": 29, "i": 30, "l": 31, "j": 32, "d": 33, "a": 34, "q": 35, "|": 24, "[UNK]": 36, "[PAD]": 37}