initial commit

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": 0.05,
+  "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": 0.05,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.05,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.05,
+  "hidden_size": 1280,
+  "initializer_range": 0.02,
+  "intermediate_size": 5120,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.05,
+  "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": 0,
+  "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": "4.30.2",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 65,
+  "xvector_output_dim": 512
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d9746dbc5d904270046128ec6bee4d31608f5481650db878035f9b51cd9ea5e
+size 7684809701

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0,
+  "processor_class": "Wav2Vec2Processor",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa353898ebeca60a27788649409f7d7dc04876a2bec30c266f563e74f9553d6a
+size 3859308813

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:017692ef1e0125f2380f568c62e6b18c41b04406e304160fd5a613234b395dc4
+size 14639

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5693dec8f7c54d61012e9efa68332ae2395902c7c278078845182b48b7a60b05
+size 627

trainer_state.json

@@ -0,0 +1,88 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 16.107382550335572,
+  "global_step": 1200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 2.68,
+      "eval_loss": 0.2483854740858078,
+      "eval_runtime": 123.9729,
+      "eval_samples_per_second": 17.762,
+      "eval_steps_per_second": 2.226,
+      "eval_wer": 0.6468182250546942,
+      "step": 200
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.00027,
+      "loss": 0.1823,
+      "step": 400
+    },
+    {
+      "epoch": 5.37,
+      "eval_loss": 0.3248055875301361,
+      "eval_runtime": 124.5509,
+      "eval_samples_per_second": 17.68,
+      "eval_steps_per_second": 2.216,
+      "eval_wer": 0.6482450299629031,
+      "step": 400
+    },
+    {
+      "epoch": 8.05,
+      "eval_loss": 0.1808338314294815,
+      "eval_runtime": 124.8994,
+      "eval_samples_per_second": 17.63,
+      "eval_steps_per_second": 2.21,
+      "eval_wer": 0.5620660135070865,
+      "step": 600
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 0.00014755102040816325,
+      "loss": 0.024,
+      "step": 800
+    },
+    {
+      "epoch": 10.74,
+      "eval_loss": 0.17792771756649017,
+      "eval_runtime": 124.4855,
+      "eval_samples_per_second": 17.689,
+      "eval_steps_per_second": 2.217,
+      "eval_wer": 0.557500237800818,
+      "step": 800
+    },
+    {
+      "epoch": 13.42,
+      "eval_loss": 0.1329699456691742,
+      "eval_runtime": 125.075,
+      "eval_samples_per_second": 17.605,
+      "eval_steps_per_second": 2.207,
+      "eval_wer": 0.5046133358698754,
+      "step": 1000
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 2.510204081632653e-05,
+      "loss": -0.1501,
+      "step": 1200
+    },
+    {
+      "epoch": 16.11,
+      "eval_loss": 0.09411804378032684,
+      "eval_runtime": 125.3304,
+      "eval_samples_per_second": 17.57,
+      "eval_steps_per_second": 2.202,
+      "eval_wer": 0.47132122134500143,
+      "step": 1200
+    }
+  ],
+  "max_steps": 1480,
+  "num_train_epochs": 20,
+  "total_flos": 4.7101086681748005e+19,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c197bd059b2b3feaa99249b9cac168a0d4a0634decd06f0d882c3791330d0f55
+size 3899

vocab.json

@@ -0,0 +1 @@
+{"\u0651": 1, "\u0780": 2, "\u0781": 3, "\u0782": 4, "\u0783": 5, "\u0784": 6, "\u0785": 7, "\u0786": 8, "\u0787": 9, "\u0788": 10, "\u0789": 11, "\u078a": 12, "\u078b": 13, "\u078c": 14, "\u078d": 15, "\u078e": 16, "\u078f": 17, "\u0790": 18, "\u0791": 19, "\u0792": 20, "\u0793": 21, "\u0794": 22, "\u0795": 23, "\u0796": 24, "\u0797": 25, "\u0798": 26, "\u0799": 27, "\u079a": 28, "\u079b": 29, "\u079d": 30, "\u079e": 31, "\u079f": 32, "\u07a0": 33, "\u07a1": 34, "\u07a2": 35, "\u07a3": 36, "\u07a4": 37, "\u07a5": 38, "\u07a6": 39, "\u07a7": 40, "\u07a8": 41, "\u07a9": 42, "\u07aa": 43, "\u07ab": 44, "\u07ac": 45, "\u07ad": 46, "\u07ae": 47, "\u07af": 48, "\u07b0": 49, "\ufdf2": 50, "\ufdfa": 51, "|": 0, "[UNK]": 52, "[PAD]": 53}