End of training

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c73935dcf07a080d4d90af3a2c076507a844fc6e6140ec67e0e27e56afcbdb08
+oid sha256:a6a59099d2ea765b74f76375ee49bd21a4517d4007ffefed6854f5393875170b
 size 377670039

runs/{May10_14-51-33_726f2b3fea59 → wav2vec-vlbg-22L-attnDrp0.5}/events.out.tfevents.1652194309.726f2b3fea59.21337.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86ebd3bce387987b530ec83c941e5b359fdf60d0110dffef509faa295be39367
-size 11731
+oid sha256:37aa19857a6c4fdfd53d4eb84a257dcd286c7872dce7e80e333068c695c5887c
+size 12085

wav2vec-vlbg-22L-attnDrp0.5/config.json

@@ -0,0 +1,80 @@
+{
+  "_name_or_path": "facebook/wav2vec2-base-960h",
+  "activation_dropout": 0.1,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.5,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 256,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": false,
+  "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": false,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "group",
+  "feat_proj_dropout": 0.0,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "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": 12,
+  "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": 12,
+  "num_negatives": 100,
+  "pad_token_id": 26,
+  "proj_codevector_dim": 256,
+  "torch_dtype": "float32",
+  "transformers_version": "4.11.3",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 32
+}

wav2vec-vlbg-22L-attnDrp0.5/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2739feed6b031dd2da439ce81d04874cef37990024b6311e7f56c3d740a875f0
+size 721685265

wav2vec-vlbg-22L-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
+}

wav2vec-vlbg-22L-attnDrp0.5/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c73935dcf07a080d4d90af3a2c076507a844fc6e6140ec67e0e27e56afcbdb08
+size 377670039

wav2vec-vlbg-22L-attnDrp0.5/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4445c83c8ce2be83638144c7dd1c808c2cc69b49776aba61e5a1434cbf52ca60
+size 14567

wav2vec-vlbg-22L-attnDrp0.5/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6bb1026cd68e31b7a4d1d768ed5fffbec2ac4fa768ed45351858ef85b1a53bc
+size 559

wav2vec-vlbg-22L-attnDrp0.5/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c83890e8f1db351034070297b46eacce5e4bea5a780c39cbae5d6ac077a2d146
+size 623

wav2vec-vlbg-22L-attnDrp0.5/trainer_state.json

@@ -0,0 +1,256 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 84.2051282051282,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 5.26,
+      "learning_rate": 5.82e-05,
+      "loss": 5.4816,
+      "step": 100
+    },
+    {
+      "epoch": 5.26,
+      "eval_loss": 3.476776123046875,
+      "eval_runtime": 5.1924,
+      "eval_samples_per_second": 33.125,
+      "eval_steps_per_second": 4.237,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 0.0001182,
+      "loss": 3.1242,
+      "step": 200
+    },
+    {
+      "epoch": 10.51,
+      "eval_loss": 2.9408938884735107,
+      "eval_runtime": 5.301,
+      "eval_samples_per_second": 32.447,
+      "eval_steps_per_second": 4.15,
+      "eval_wer": 1.0,
+      "step": 200
+    },
+    {
+      "epoch": 15.77,
+      "learning_rate": 0.00017819999999999997,
+      "loss": 2.5963,
+      "step": 300
+    },
+    {
+      "epoch": 15.77,
+      "eval_loss": 1.7765510082244873,
+      "eval_runtime": 5.3164,
+      "eval_samples_per_second": 32.353,
+      "eval_steps_per_second": 4.138,
+      "eval_wer": 0.961864406779661,
+      "step": 300
+    },
+    {
+      "epoch": 21.05,
+      "learning_rate": 0.0002382,
+      "loss": 1.4721,
+      "step": 400
+    },
+    {
+      "epoch": 21.05,
+      "eval_loss": 1.0526211261749268,
+      "eval_runtime": 5.2743,
+      "eval_samples_per_second": 32.611,
+      "eval_steps_per_second": 4.171,
+      "eval_wer": 0.7097457627118644,
+      "step": 400
+    },
+    {
+      "epoch": 26.31,
+      "learning_rate": 0.0002982,
+      "loss": 0.8627,
+      "step": 500
+    },
+    {
+      "epoch": 26.31,
+      "eval_loss": 0.8283404111862183,
+      "eval_runtime": 5.3117,
+      "eval_samples_per_second": 32.381,
+      "eval_steps_per_second": 4.142,
+      "eval_wer": 0.5497881355932204,
+      "step": 500
+    },
+    {
+      "epoch": 31.56,
+      "learning_rate": 0.0002741704035874439,
+      "loss": 0.709,
+      "step": 600
+    },
+    {
+      "epoch": 31.56,
+      "eval_loss": 0.844669759273529,
+      "eval_runtime": 5.282,
+      "eval_samples_per_second": 32.563,
+      "eval_steps_per_second": 4.165,
+      "eval_wer": 0.4925847457627119,
+      "step": 600
+    },
+    {
+      "epoch": 36.82,
+      "learning_rate": 0.0002472645739910314,
+      "loss": 0.5004,
+      "step": 700
+    },
+    {
+      "epoch": 36.82,
+      "eval_loss": 0.8127376437187195,
+      "eval_runtime": 5.3345,
+      "eval_samples_per_second": 32.243,
+      "eval_steps_per_second": 4.124,
+      "eval_wer": 0.4396186440677966,
+      "step": 700
+    },
+    {
+      "epoch": 42.1,
+      "learning_rate": 0.00022035874439461881,
+      "loss": 0.3627,
+      "step": 800
+    },
+    {
+      "epoch": 42.1,
+      "eval_loss": 0.8336098194122314,
+      "eval_runtime": 5.2871,
+      "eval_samples_per_second": 32.532,
+      "eval_steps_per_second": 4.161,
+      "eval_wer": 0.4046610169491525,
+      "step": 800
+    },
+    {
+      "epoch": 47.36,
+      "learning_rate": 0.00019345291479820625,
+      "loss": 0.2899,
+      "step": 900
+    },
+    {
+      "epoch": 47.36,
+      "eval_loss": 0.8964722156524658,
+      "eval_runtime": 5.3008,
+      "eval_samples_per_second": 32.448,
+      "eval_steps_per_second": 4.15,
+      "eval_wer": 0.3654661016949153,
+      "step": 900
+    },
+    {
+      "epoch": 52.62,
+      "learning_rate": 0.0001665470852017937,
+      "loss": 0.2198,
+      "step": 1000
+    },
+    {
+      "epoch": 52.62,
+      "eval_loss": 0.959164023399353,
+      "eval_runtime": 5.2971,
+      "eval_samples_per_second": 32.471,
+      "eval_steps_per_second": 4.153,
+      "eval_wer": 0.3495762711864407,
+      "step": 1000
+    },
+    {
+      "epoch": 57.87,
+      "learning_rate": 0.00013964125560538115,
+      "loss": 0.1775,
+      "step": 1100
+    },
+    {
+      "epoch": 57.87,
+      "eval_loss": 0.9055131673812866,
+      "eval_runtime": 5.3332,
+      "eval_samples_per_second": 32.251,
+      "eval_steps_per_second": 4.125,
+      "eval_wer": 0.3516949152542373,
+      "step": 1100
+    },
+    {
+      "epoch": 63.15,
+      "learning_rate": 0.0001127354260089686,
+      "loss": 0.1494,
+      "step": 1200
+    },
+    {
+      "epoch": 63.15,
+      "eval_loss": 1.008966088294983,
+      "eval_runtime": 5.3032,
+      "eval_samples_per_second": 32.433,
+      "eval_steps_per_second": 4.148,
+      "eval_wer": 0.3336864406779661,
+      "step": 1200
+    },
+    {
+      "epoch": 68.41,
+      "learning_rate": 8.582959641255606e-05,
+      "loss": 0.1224,
+      "step": 1300
+    },
+    {
+      "epoch": 68.41,
+      "eval_loss": 0.9990301728248596,
+      "eval_runtime": 5.2827,
+      "eval_samples_per_second": 32.559,
+      "eval_steps_per_second": 4.165,
+      "eval_wer": 0.3305084745762712,
+      "step": 1300
+    },
+    {
+      "epoch": 73.67,
+      "learning_rate": 5.892376681614349e-05,
+      "loss": 0.102,
+      "step": 1400
+    },
+    {
+      "epoch": 73.67,
+      "eval_loss": 1.068620204925537,
+      "eval_runtime": 5.2912,
+      "eval_samples_per_second": 32.507,
+      "eval_steps_per_second": 4.158,
+      "eval_wer": 0.3072033898305085,
+      "step": 1400
+    },
+    {
+      "epoch": 78.92,
+      "learning_rate": 3.2017937219730936e-05,
+      "loss": 0.0787,
+      "step": 1500
+    },
+    {
+      "epoch": 78.92,
+      "eval_loss": 1.097746729850769,
+      "eval_runtime": 5.2813,
+      "eval_samples_per_second": 32.568,
+      "eval_steps_per_second": 4.166,
+      "eval_wer": 0.3019067796610169,
+      "step": 1500
+    },
+    {
+      "epoch": 84.21,
+      "learning_rate": 5.112107623318384e-06,
+      "loss": 0.0719,
+      "step": 1600
+    },
+    {
+      "epoch": 84.21,
+      "eval_loss": 1.0722393989562988,
+      "eval_runtime": 5.2609,
+      "eval_samples_per_second": 32.694,
+      "eval_steps_per_second": 4.182,
+      "eval_wer": 0.2997881355932203,
+      "step": 1600
+    }
+  ],
+  "max_steps": 1615,
+  "num_train_epochs": 85,
+  "total_flos": 1.363504277312058e+18,
+  "trial_name": null,
+  "trial_params": null
+}

wav2vec-vlbg-22L-attnDrp0.5/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f36f9ccd48f3f01b0e7bcec3703095b0f71f41e3eb46969b06b47ac3a2c12dc4
+size 2799