Training in progress, epoch 3, checkpoint

checkpoint-2187/config.json

@@ -0,0 +1,108 @@
+{
+  "_name_or_path": "facebook/mms-1b-all",
+  "activation_dropout": 0.05,
+  "adapter_attn_dim": 16,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 1024,
+  "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": 0.05,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1280,
+  "initializer_range": 0.02,
+  "intermediate_size": 5120,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.0,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
+  "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": 48,
+  "num_negatives": 100,
+  "output_hidden_size": 1280,
+  "pad_token_id": 92,
+  "proj_codevector_dim": 1024,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.34.1",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 95,
+  "xvector_output_dim": 512
+}

checkpoint-2187/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b75197a7945988e498f8263f3819a59b170e33ab52043eae433d5b763bd9ce2
+size 18424186

checkpoint-2187/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
+}

checkpoint-2187/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d80f8747310baf27c89e59fd339c535237f2f31238a3e41a0937112739780b3e
+size 3859463050

checkpoint-2187/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf0e8e4344b6c88a49561eca2e9290c3733a932e5ca69913ab8352e77753bf54
+size 14308

checkpoint-2187/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb5b4da30becc9902ebfa0f77ce0641de395dd146f92c926bfc3b0f02895b3ef
+size 1064

checkpoint-2187/trainer_state.json

@@ -0,0 +1,181 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2187,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00048,
+      "loss": 7.0366,
+      "step": 100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0004933240611961058,
+      "loss": 0.4467,
+      "step": 200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00048643949930458967,
+      "loss": 0.3946,
+      "step": 300
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0004794853963838665,
+      "loss": 0.399,
+      "step": 400
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0004726008344923505,
+      "loss": 0.3696,
+      "step": 500
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.3043254017829895,
+      "eval_runtime": 65.1553,
+      "eval_samples_per_second": 11.664,
+      "eval_steps_per_second": 1.458,
+      "eval_wer": 0.265993265993266,
+      "step": 500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0004656467315716273,
+      "loss": 0.3677,
+      "step": 600
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00045869262865090403,
+      "loss": 0.3534,
+      "step": 700
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00045173852573018083,
+      "loss": 0.34,
+      "step": 800
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0004447844228094576,
+      "loss": 0.3428,
+      "step": 900
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0004378998609179416,
+      "loss": 0.3533,
+      "step": 1000
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 0.27337372303009033,
+      "eval_runtime": 65.7997,
+      "eval_samples_per_second": 11.55,
+      "eval_steps_per_second": 1.444,
+      "eval_wer": 0.23931623931623933,
+      "step": 1000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00043094575799721834,
+      "loss": 0.336,
+      "step": 1100
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00042399165507649514,
+      "loss": 0.3453,
+      "step": 1200
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00041703755215577194,
+      "loss": 0.342,
+      "step": 1300
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0004100834492350487,
+      "loss": 0.3232,
+      "step": 1400
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0004031293463143255,
+      "loss": 0.3419,
+      "step": 1500
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 0.2687598764896393,
+      "eval_runtime": 66.7716,
+      "eval_samples_per_second": 11.382,
+      "eval_steps_per_second": 1.423,
+      "eval_wer": 0.23776223776223776,
+      "step": 1500
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0003961752433936022,
+      "loss": 0.3045,
+      "step": 1600
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00038922114047287897,
+      "loss": 0.327,
+      "step": 1700
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0003822670375521558,
+      "loss": 0.3162,
+      "step": 1800
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00037531293463143256,
+      "loss": 0.3111,
+      "step": 1900
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0003684283727399166,
+      "loss": 0.3399,
+      "step": 2000
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 0.2612040042877197,
+      "eval_runtime": 66.305,
+      "eval_samples_per_second": 11.462,
+      "eval_steps_per_second": 1.433,
+      "eval_wer": 0.22895622895622897,
+      "step": 2000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00036147426981919333,
+      "loss": 0.3113,
+      "step": 2100
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 7290,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 1.5151017238545934e+19,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2187/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d3aa848da266dd9fcefac1319c4009241b1015d8306bc1c8253b19be17b56d9
+size 4536