End of training

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:396841b33d5a2242a9ca2553ea39793fc38d384f03f83a250cecc4d932cd84ff
+oid sha256:9b4ef4ef05959005bb560e6cacb65731fd31dabb4026e692b83e8b87b04300de
 size 377670039

runs/{May10_11-00-49_726f2b3fea59 → wav2vec-vlbg-22D-attnDrp0.5}/events.out.tfevents.1652180603.726f2b3fea59.13393.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f502a5b55e9f7fb603cdb4594b1ff0b76487e8d8caefaed5781a4e89a605c2b
-size 11731
+oid sha256:618ef3980b013a8817f383a53f424f30bb170f102b609bac45566cdbf9eb7b5c
+size 12085

wav2vec-vlbg-22D-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": 25,
+  "proj_codevector_dim": 256,
+  "torch_dtype": "float32",
+  "transformers_version": "4.11.3",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 32
+}

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fcd1922295a68da3b54699edcd7a10321a5dead69e9d3e3bfc0a239a5fa08fb
+size 721685265

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:396841b33d5a2242a9ca2553ea39793fc38d384f03f83a250cecc4d932cd84ff
+size 377670039

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05f8b8ccd2d70a71bee4e42e042cf7150d88e3aea94fbe945dc084d182956226
+size 14503

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7acedd6e65f60c847328ab1edfe16736a602ffd876ea00a5175226b958f0587a
+size 559

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:501b8c1db8a1e604b70a87a72165753d4acb6f83afdd825a61f20fea16cbbb7c
+size 623

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

@@ -0,0 +1,256 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 69.55319148936171,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.34,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 15.3922,
+      "step": 100
+    },
+    {
+      "epoch": 4.34,
+      "eval_loss": 3.9500844478607178,
+      "eval_runtime": 148.4293,
+      "eval_samples_per_second": 1.112,
+      "eval_steps_per_second": 0.141,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 0.0001164,
+      "loss": 3.1463,
+      "step": 200
+    },
+    {
+      "epoch": 8.68,
+      "eval_loss": 3.1115620136260986,
+      "eval_runtime": 4.6891,
+      "eval_samples_per_second": 35.188,
+      "eval_steps_per_second": 4.478,
+      "eval_wer": 1.0,
+      "step": 200
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 0.00017639999999999998,
+      "loss": 2.6716,
+      "step": 300
+    },
+    {
+      "epoch": 13.04,
+      "eval_loss": 1.9330189228057861,
+      "eval_runtime": 4.6705,
+      "eval_samples_per_second": 35.328,
+      "eval_steps_per_second": 4.496,
+      "eval_wer": 0.9731743666169895,
+      "step": 300
+    },
+    {
+      "epoch": 17.38,
+      "learning_rate": 0.0002364,
+      "loss": 1.7477,
+      "step": 400
+    },
+    {
+      "epoch": 17.38,
+      "eval_loss": 1.2021902799606323,
+      "eval_runtime": 4.7103,
+      "eval_samples_per_second": 35.03,
+      "eval_steps_per_second": 4.458,
+      "eval_wer": 0.6870342771982116,
+      "step": 400
+    },
+    {
+      "epoch": 21.72,
+      "learning_rate": 0.0002964,
+      "loss": 1.0917,
+      "step": 500
+    },
+    {
+      "epoch": 21.72,
+      "eval_loss": 1.1227840185165405,
+      "eval_runtime": 4.6735,
+      "eval_samples_per_second": 35.305,
+      "eval_steps_per_second": 4.493,
+      "eval_wer": 0.503725782414307,
+      "step": 500
+    },
+    {
+      "epoch": 26.09,
+      "learning_rate": 0.00027459459459459456,
+      "loss": 0.8226,
+      "step": 600
+    },
+    {
+      "epoch": 26.09,
+      "eval_loss": 0.9606490731239319,
+      "eval_runtime": 4.6745,
+      "eval_samples_per_second": 35.298,
+      "eval_steps_per_second": 4.492,
+      "eval_wer": 0.3740685543964232,
+      "step": 600
+    },
+    {
+      "epoch": 30.43,
+      "learning_rate": 0.00024756756756756757,
+      "loss": 0.6129,
+      "step": 700
+    },
+    {
+      "epoch": 30.43,
+      "eval_loss": 1.1125797033309937,
+      "eval_runtime": 4.7099,
+      "eval_samples_per_second": 35.033,
+      "eval_steps_per_second": 4.459,
+      "eval_wer": 0.43517138599105815,
+      "step": 700
+    },
+    {
+      "epoch": 34.77,
+      "learning_rate": 0.00022054054054054052,
+      "loss": 0.4598,
+      "step": 800
+    },
+    {
+      "epoch": 34.77,
+      "eval_loss": 0.9925816655158997,
+      "eval_runtime": 4.7294,
+      "eval_samples_per_second": 34.888,
+      "eval_steps_per_second": 4.44,
+      "eval_wer": 0.27719821162444114,
+      "step": 800
+    },
+    {
+      "epoch": 39.13,
+      "learning_rate": 0.00019351351351351348,
+      "loss": 0.3705,
+      "step": 900
+    },
+    {
+      "epoch": 39.13,
+      "eval_loss": 1.2002172470092773,
+      "eval_runtime": 4.7834,
+      "eval_samples_per_second": 34.495,
+      "eval_steps_per_second": 4.39,
+      "eval_wer": 0.2578241430700447,
+      "step": 900
+    },
+    {
+      "epoch": 43.47,
+      "learning_rate": 0.00016648648648648646,
+      "loss": 0.2999,
+      "step": 1000
+    },
+    {
+      "epoch": 43.47,
+      "eval_loss": 1.2655400037765503,
+      "eval_runtime": 4.7236,
+      "eval_samples_per_second": 34.931,
+      "eval_steps_per_second": 4.446,
+      "eval_wer": 0.26229508196721313,
+      "step": 1000
+    },
+    {
+      "epoch": 47.81,
+      "learning_rate": 0.00013945945945945945,
+      "loss": 0.2338,
+      "step": 1100
+    },
+    {
+      "epoch": 47.81,
+      "eval_loss": 1.2947732210159302,
+      "eval_runtime": 4.7062,
+      "eval_samples_per_second": 35.06,
+      "eval_steps_per_second": 4.462,
+      "eval_wer": 0.21013412816691504,
+      "step": 1100
+    },
+    {
+      "epoch": 52.17,
+      "learning_rate": 0.00011243243243243243,
+      "loss": 0.1817,
+      "step": 1200
+    },
+    {
+      "epoch": 52.17,
+      "eval_loss": 1.337550163269043,
+      "eval_runtime": 4.7558,
+      "eval_samples_per_second": 34.694,
+      "eval_steps_per_second": 4.416,
+      "eval_wer": 0.20268256333830104,
+      "step": 1200
+    },
+    {
+      "epoch": 56.51,
+      "learning_rate": 8.540540540540539e-05,
+      "loss": 0.1545,
+      "step": 1300
+    },
+    {
+      "epoch": 56.51,
+      "eval_loss": 1.2345869541168213,
+      "eval_runtime": 4.7487,
+      "eval_samples_per_second": 34.746,
+      "eval_steps_per_second": 4.422,
+      "eval_wer": 0.18628912071535023,
+      "step": 1300
+    },
+    {
+      "epoch": 60.85,
+      "learning_rate": 5.8378378378378376e-05,
+      "loss": 0.1285,
+      "step": 1400
+    },
+    {
+      "epoch": 60.85,
+      "eval_loss": 1.3973815441131592,
+      "eval_runtime": 4.6899,
+      "eval_samples_per_second": 35.182,
+      "eval_steps_per_second": 4.478,
+      "eval_wer": 0.19374068554396423,
+      "step": 1400
+    },
+    {
+      "epoch": 65.21,
+      "learning_rate": 3.135135135135135e-05,
+      "loss": 0.1024,
+      "step": 1500
+    },
+    {
+      "epoch": 65.21,
+      "eval_loss": 1.4698290824890137,
+      "eval_runtime": 4.7387,
+      "eval_samples_per_second": 34.82,
+      "eval_steps_per_second": 4.432,
+      "eval_wer": 0.18032786885245902,
+      "step": 1500
+    },
+    {
+      "epoch": 69.55,
+      "learning_rate": 4.324324324324324e-06,
+      "loss": 0.0777,
+      "step": 1600
+    },
+    {
+      "epoch": 69.55,
+      "eval_loss": 1.4972994327545166,
+      "eval_runtime": 4.6929,
+      "eval_samples_per_second": 35.16,
+      "eval_steps_per_second": 4.475,
+      "eval_wer": 0.17734724292101342,
+      "step": 1600
+    }
+  ],
+  "max_steps": 1610,
+  "num_train_epochs": 70,
+  "total_flos": 1.271106313435803e+18,
+  "trial_name": null,
+  "trial_params": null
+}

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80db73272fd4ef1ac29e278a443e3fe82686deb5df9f79bb4001958b3556ad8b
+size 2799