add tokenizer

added_tokens.json

@@ -1 +1 @@
-{"<s>": 27, "</s>": 28}
+{"<s>": 26, "</s>": 27}

runs/{May09_23-39-31_184ce95389f2 → xlsr-vlbg-22-attnDrp0.5}/events.out.tfevents.1652139584.184ce95389f2.17100.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61d3746dbdf638e01f993fc4b8d3d9bb3d0f3ead95f7f0d674a2e76e06e440eb
-size 11738
+oid sha256:f87a0879aefcf3b235e42d7b40d63a092d2cb9e8a6ee483a2ac86999babcc19d
+size 12092

vocab.json

@@ -1 +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}
+{"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, "z": 23, "|": 0, "[UNK]": 24, "[PAD]": 25}

xlsr-vlbg-22-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": 25,
+  "proj_codevector_dim": 768,
+  "torch_dtype": "float32",
+  "transformers_version": "4.11.3",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 28
+}

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b05912608ce9feaa3579c8fb6ab2c38bb7aa827e325366acab8cb763661df3b4
+size 2490288529

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41bef192788d19ece4ad73676665b6a55eff75069a8ee6e6b2d4fda1ad83e8da
+size 1262038449

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2398ba45045bc92d88875757e4a606c93e6695aa024858e25f2c47b052b1f39b
+size 14567

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aea1b65af4b6e176c6413ca69a6348ca2d2f1b302a05b3300a961ad4fdf21f04
+size 559

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05a6351238f909334158af1b6fbf11c210cbdf090fc78ba41efb68eb8d22a53e
+size 623

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

@@ -0,0 +1,256 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 114.27586206896552,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 7.14,
+      "learning_rate": 5.82e-05,
+      "loss": 5.8515,
+      "step": 100
+    },
+    {
+      "epoch": 7.14,
+      "eval_loss": 3.176227331161499,
+      "eval_runtime": 20.423,
+      "eval_samples_per_second": 5.386,
+      "eval_steps_per_second": 0.686,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 0.0001182,
+      "loss": 3.0387,
+      "step": 200
+    },
+    {
+      "epoch": 14.28,
+      "eval_loss": 2.77799129486084,
+      "eval_runtime": 5.5207,
+      "eval_samples_per_second": 19.925,
+      "eval_steps_per_second": 2.536,
+      "eval_wer": 1.0,
+      "step": 200
+    },
+    {
+      "epoch": 21.41,
+      "learning_rate": 0.00017819999999999997,
+      "loss": 2.3167,
+      "step": 300
+    },
+    {
+      "epoch": 21.41,
+      "eval_loss": 1.1196272373199463,
+      "eval_runtime": 5.5506,
+      "eval_samples_per_second": 19.818,
+      "eval_steps_per_second": 2.522,
+      "eval_wer": 0.8452579034941764,
+      "step": 300
+    },
+    {
+      "epoch": 28.55,
+      "learning_rate": 0.0002382,
+      "loss": 0.571,
+      "step": 400
+    },
+    {
+      "epoch": 28.55,
+      "eval_loss": 0.2766098380088806,
+      "eval_runtime": 5.4507,
+      "eval_samples_per_second": 20.181,
+      "eval_steps_per_second": 2.568,
+      "eval_wer": 0.30948419301164726,
+      "step": 400
+    },
+    {
+      "epoch": 35.69,
+      "learning_rate": 0.0002982,
+      "loss": 0.21,
+      "step": 500
+    },
+    {
+      "epoch": 35.69,
+      "eval_loss": 0.26628053188323975,
+      "eval_runtime": 5.4244,
+      "eval_samples_per_second": 20.279,
+      "eval_steps_per_second": 2.581,
+      "eval_wer": 0.2978369384359401,
+      "step": 500
+    },
+    {
+      "epoch": 42.83,
+      "learning_rate": 0.00027378378378378374,
+      "loss": 0.1352,
+      "step": 600
+    },
+    {
+      "epoch": 42.83,
+      "eval_loss": 0.1913282722234726,
+      "eval_runtime": 5.4513,
+      "eval_samples_per_second": 20.179,
+      "eval_steps_per_second": 2.568,
+      "eval_wer": 0.23460898502495842,
+      "step": 600
+    },
+    {
+      "epoch": 49.97,
+      "learning_rate": 0.00024675675675675675,
+      "loss": 0.0937,
+      "step": 700
+    },
+    {
+      "epoch": 49.97,
+      "eval_loss": 0.17986251413822174,
+      "eval_runtime": 5.4633,
+      "eval_samples_per_second": 20.134,
+      "eval_steps_per_second": 2.563,
+      "eval_wer": 0.194675540765391,
+      "step": 700
+    },
+    {
+      "epoch": 57.14,
+      "learning_rate": 0.0002197297297297297,
+      "loss": 0.0546,
+      "step": 800
+    },
+    {
+      "epoch": 57.14,
+      "eval_loss": 0.17105376720428467,
+      "eval_runtime": 5.5914,
+      "eval_samples_per_second": 19.673,
+      "eval_steps_per_second": 2.504,
+      "eval_wer": 0.15973377703826955,
+      "step": 800
+    },
+    {
+      "epoch": 64.28,
+      "learning_rate": 0.0001927027027027027,
+      "loss": 0.0462,
+      "step": 900
+    },
+    {
+      "epoch": 64.28,
+      "eval_loss": 0.17690478265285492,
+      "eval_runtime": 5.5906,
+      "eval_samples_per_second": 19.676,
+      "eval_steps_per_second": 2.504,
+      "eval_wer": 0.15973377703826955,
+      "step": 900
+    },
+    {
+      "epoch": 71.41,
+      "learning_rate": 0.00016567567567567568,
+      "loss": 0.0336,
+      "step": 1000
+    },
+    {
+      "epoch": 71.41,
+      "eval_loss": 0.19472338259220123,
+      "eval_runtime": 5.751,
+      "eval_samples_per_second": 19.127,
+      "eval_steps_per_second": 2.434,
+      "eval_wer": 0.1447587354409318,
+      "step": 1000
+    },
+    {
+      "epoch": 78.55,
+      "learning_rate": 0.00013864864864864863,
+      "loss": 0.0253,
+      "step": 1100
+    },
+    {
+      "epoch": 78.55,
+      "eval_loss": 0.17213842272758484,
+      "eval_runtime": 5.7965,
+      "eval_samples_per_second": 18.977,
+      "eval_steps_per_second": 2.415,
+      "eval_wer": 0.11148086522462562,
+      "step": 1100
+    },
+    {
+      "epoch": 85.69,
+      "learning_rate": 0.0001116216216216216,
+      "loss": 0.0203,
+      "step": 1200
+    },
+    {
+      "epoch": 85.69,
+      "eval_loss": 0.2061169594526291,
+      "eval_runtime": 5.6059,
+      "eval_samples_per_second": 19.622,
+      "eval_steps_per_second": 2.497,
+      "eval_wer": 0.1447587354409318,
+      "step": 1200
+    },
+    {
+      "epoch": 92.83,
+      "learning_rate": 8.459459459459459e-05,
+      "loss": 0.0157,
+      "step": 1300
+    },
+    {
+      "epoch": 92.83,
+      "eval_loss": 0.17522919178009033,
+      "eval_runtime": 5.6866,
+      "eval_samples_per_second": 19.344,
+      "eval_steps_per_second": 2.462,
+      "eval_wer": 0.1264559068219634,
+      "step": 1300
+    },
+    {
+      "epoch": 99.97,
+      "learning_rate": 5.756756756756756e-05,
+      "loss": 0.0134,
+      "step": 1400
+    },
+    {
+      "epoch": 99.97,
+      "eval_loss": 0.19270485639572144,
+      "eval_runtime": 5.7092,
+      "eval_samples_per_second": 19.267,
+      "eval_steps_per_second": 2.452,
+      "eval_wer": 0.1314475873544093,
+      "step": 1400
+    },
+    {
+      "epoch": 107.14,
+      "learning_rate": 3.054054054054054e-05,
+      "loss": 0.0107,
+      "step": 1500
+    },
+    {
+      "epoch": 107.14,
+      "eval_loss": 0.1782664954662323,
+      "eval_runtime": 5.6232,
+      "eval_samples_per_second": 19.562,
+      "eval_steps_per_second": 2.49,
+      "eval_wer": 0.13643926788685523,
+      "step": 1500
+    },
+    {
+      "epoch": 114.28,
+      "learning_rate": 3.5135135135135134e-06,
+      "loss": 0.0131,
+      "step": 1600
+    },
+    {
+      "epoch": 114.28,
+      "eval_loss": 0.18077482283115387,
+      "eval_runtime": 5.5344,
+      "eval_samples_per_second": 19.876,
+      "eval_steps_per_second": 2.53,
+      "eval_wer": 0.1314475873544093,
+      "step": 1600
+    }
+  ],
+  "max_steps": 1610,
+  "num_train_epochs": 115,
+  "total_flos": 4.549145358095393e+18,
+  "trial_name": null,
+  "trial_params": null
+}

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:121c3f44095d4bb2b67ecc969a8fe41aa92c7b7c8d65ecd7245126f2cd323d6a
+size 2799