add tokenizer

added_tokens.json

@@ -0,0 +1 @@
+{"<s>": 27, "</s>": 28}

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}

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": "|", "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "./", "tokenizer_class": "Wav2Vec2CTCTokenizer"}

vocab.json

@@ -0,0 +1 @@
+{"a": 1, "b": 2, "c": 3, "d": 4, "e": 5, "f": 6, "g": 7, "h": 8, "i": 9, "j": 10, "k": 11, "l": 12, "m": 13, "n": 14, "o": 15, "p": 16, "r": 17, "s": 18, "t": 19, "u": 20, "v": 21, "w": 22, "x": 23, "z": 24, "|": 0, "[UNK]": 25, "[PAD]": 26}

xlsr-vlbg-22DL-attnDrp0.5/config.json

@@ -0,0 +1,79 @@
+{
+  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
+  "activation_dropout": 0.0,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.5,
+  "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": "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": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.0,
+  "mask_feature_length": 10,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_prob": 0.05,
+  "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": 26,
+  "proj_codevector_dim": 768,
+  "torch_dtype": "float32",
+  "transformers_version": "4.11.3",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 29
+}

xlsr-vlbg-22DL-attnDrp0.5/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47edb237e1251089dac43affae9caf272841511885a1192eeeb7e41094781feb
+size 2490296721

xlsr-vlbg-22DL-attnDrp0.5/preprocessor_config.json

@@ -0,0 +1,9 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

xlsr-vlbg-22DL-attnDrp0.5/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:446f01593c67a73c6be324669a04604e096d13a61f4ad0037ac39957eeccc9b4
+size 1262042545

xlsr-vlbg-22DL-attnDrp0.5/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:274c31383ff8c3ef7bf3c340a89a6941fe11463c9b33d2608fbdef0bea84ed7d
+size 14567

xlsr-vlbg-22DL-attnDrp0.5/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c42decc74b17e4851fcc51766fe3ddc893e4d6f043335e1c4a2764999ef5968
+size 559

xlsr-vlbg-22DL-attnDrp0.5/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:767d1694e9ef2cc1caba77a48749c3f99ece3803691fca634506d9ae870315b2
+size 623

xlsr-vlbg-22DL-attnDrp0.5/trainer_state.json

@@ -0,0 +1,256 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 57.14035087719298,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.82e-05,
+      "loss": 10.4898,
+      "step": 100
+    },
+    {
+      "epoch": 3.56,
+      "eval_loss": 3.874396562576294,
+      "eval_runtime": 10.7449,
+      "eval_samples_per_second": 21.126,
+      "eval_steps_per_second": 2.699,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 0.0001182,
+      "loss": 3.4178,
+      "step": 200
+    },
+    {
+      "epoch": 7.14,
+      "eval_loss": 3.0382564067840576,
+      "eval_runtime": 10.7699,
+      "eval_samples_per_second": 21.077,
+      "eval_steps_per_second": 2.693,
+      "eval_wer": 1.0,
+      "step": 200
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 0.00017819999999999997,
+      "loss": 2.9644,
+      "step": 300
+    },
+    {
+      "epoch": 10.7,
+      "eval_loss": 2.7548253536224365,
+      "eval_runtime": 10.8099,
+      "eval_samples_per_second": 20.999,
+      "eval_steps_per_second": 2.683,
+      "eval_wer": 1.0,
+      "step": 300
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 0.0002382,
+      "loss": 2.0967,
+      "step": 400
+    },
+    {
+      "epoch": 14.28,
+      "eval_loss": 1.3797626495361328,
+      "eval_runtime": 10.7742,
+      "eval_samples_per_second": 21.069,
+      "eval_steps_per_second": 2.692,
+      "eval_wer": 0.7384916748285995,
+      "step": 400
+    },
+    {
+      "epoch": 17.84,
+      "learning_rate": 0.0002982,
+      "loss": 0.9283,
+      "step": 500
+    },
+    {
+      "epoch": 17.84,
+      "eval_loss": 0.8976249098777771,
+      "eval_runtime": 10.7567,
+      "eval_samples_per_second": 21.103,
+      "eval_steps_per_second": 2.696,
+      "eval_wer": 0.5494613124387855,
+      "step": 500
+    },
+    {
+      "epoch": 21.42,
+      "learning_rate": 0.00027533898305084746,
+      "loss": 0.5498,
+      "step": 600
+    },
+    {
+      "epoch": 21.42,
+      "eval_loss": 0.912022054195404,
+      "eval_runtime": 10.7884,
+      "eval_samples_per_second": 21.041,
+      "eval_steps_per_second": 2.688,
+      "eval_wer": 0.47110675808031344,
+      "step": 600
+    },
+    {
+      "epoch": 24.98,
+      "learning_rate": 0.00024991525423728815,
+      "loss": 0.3561,
+      "step": 700
+    },
+    {
+      "epoch": 24.98,
+      "eval_loss": 0.9388850331306458,
+      "eval_runtime": 10.7935,
+      "eval_samples_per_second": 21.031,
+      "eval_steps_per_second": 2.687,
+      "eval_wer": 0.4505386875612145,
+      "step": 700
+    },
+    {
+      "epoch": 28.56,
+      "learning_rate": 0.0002244915254237288,
+      "loss": 0.2544,
+      "step": 800
+    },
+    {
+      "epoch": 28.56,
+      "eval_loss": 0.8813793659210205,
+      "eval_runtime": 10.7967,
+      "eval_samples_per_second": 21.025,
+      "eval_steps_per_second": 2.686,
+      "eval_wer": 0.40842311459353575,
+      "step": 800
+    },
+    {
+      "epoch": 32.14,
+      "learning_rate": 0.0001990677966101695,
+      "loss": 0.1859,
+      "step": 900
+    },
+    {
+      "epoch": 32.14,
+      "eval_loss": 1.0429902076721191,
+      "eval_runtime": 10.7927,
+      "eval_samples_per_second": 21.033,
+      "eval_steps_per_second": 2.687,
+      "eval_wer": 0.37904015670910873,
+      "step": 900
+    },
+    {
+      "epoch": 35.7,
+      "learning_rate": 0.00017364406779661016,
+      "loss": 0.1517,
+      "step": 1000
+    },
+    {
+      "epoch": 35.7,
+      "eval_loss": 0.9847303032875061,
+      "eval_runtime": 10.7475,
+      "eval_samples_per_second": 21.121,
+      "eval_steps_per_second": 2.698,
+      "eval_wer": 0.39764936336924583,
+      "step": 1000
+    },
+    {
+      "epoch": 39.28,
+      "learning_rate": 0.00014822033898305082,
+      "loss": 0.107,
+      "step": 1100
+    },
+    {
+      "epoch": 39.28,
+      "eval_loss": 0.9529415965080261,
+      "eval_runtime": 10.7532,
+      "eval_samples_per_second": 21.11,
+      "eval_steps_per_second": 2.697,
+      "eval_wer": 0.3702252693437806,
+      "step": 1100
+    },
+    {
+      "epoch": 42.84,
+      "learning_rate": 0.0001227966101694915,
+      "loss": 0.0863,
+      "step": 1200
+    },
+    {
+      "epoch": 42.84,
+      "eval_loss": 1.104305624961853,
+      "eval_runtime": 10.7454,
+      "eval_samples_per_second": 21.125,
+      "eval_steps_per_second": 2.699,
+      "eval_wer": 0.3476983349657199,
+      "step": 1200
+    },
+    {
+      "epoch": 46.42,
+      "learning_rate": 9.73728813559322e-05,
+      "loss": 0.0686,
+      "step": 1300
+    },
+    {
+      "epoch": 46.42,
+      "eval_loss": 1.1166020631790161,
+      "eval_runtime": 10.8153,
+      "eval_samples_per_second": 20.989,
+      "eval_steps_per_second": 2.681,
+      "eval_wer": 0.3731635651322233,
+      "step": 1300
+    },
+    {
+      "epoch": 49.98,
+      "learning_rate": 7.194915254237287e-05,
+      "loss": 0.0508,
+      "step": 1400
+    },
+    {
+      "epoch": 49.98,
+      "eval_loss": 1.0854098796844482,
+      "eval_runtime": 10.8097,
+      "eval_samples_per_second": 21.0,
+      "eval_steps_per_second": 2.683,
+      "eval_wer": 0.356513222331048,
+      "step": 1400
+    },
+    {
+      "epoch": 53.56,
+      "learning_rate": 4.652542372881355e-05,
+      "loss": 0.0425,
+      "step": 1500
+    },
+    {
+      "epoch": 53.56,
+      "eval_loss": 1.0830148458480835,
+      "eval_runtime": 10.7781,
+      "eval_samples_per_second": 21.061,
+      "eval_steps_per_second": 2.691,
+      "eval_wer": 0.3614103819784525,
+      "step": 1500
+    },
+    {
+      "epoch": 57.14,
+      "learning_rate": 2.1101694915254236e-05,
+      "loss": 0.0362,
+      "step": 1600
+    },
+    {
+      "epoch": 57.14,
+      "eval_loss": 1.0556385517120361,
+      "eval_runtime": 10.7827,
+      "eval_samples_per_second": 21.052,
+      "eval_steps_per_second": 2.689,
+      "eval_wer": 0.3574926542605289,
+      "step": 1600
+    }
+  ],
+  "max_steps": 1680,
+  "num_train_epochs": 60,
+  "total_flos": 4.130905589155786e+18,
+  "trial_name": null,
+  "trial_params": null
+}

xlsr-vlbg-22DL-attnDrp0.5/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a37d4c2d38f022542bfbd6b6927ef9b65bc24966937168dbb488be3cc250a40c
+size 2799