Training in progress, epoch 25, checkpoint
Browse files- checkpoint-6450/config.json +82 -0
- checkpoint-6450/model.safetensors +3 -0
- checkpoint-6450/optimizer.pt +3 -0
- checkpoint-6450/preprocessor_config.json +11 -0
- checkpoint-6450/rng_state.pth +3 -0
- checkpoint-6450/scheduler.pt +3 -0
- checkpoint-6450/trainer_state.json +467 -0
- checkpoint-6450/training_args.bin +3 -0
checkpoint-6450/config.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/w2v-bert-2.0",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_act": "relu",
|
| 5 |
+
"adapter_kernel_size": 3,
|
| 6 |
+
"adapter_stride": 2,
|
| 7 |
+
"add_adapter": true,
|
| 8 |
+
"apply_spec_augment": false,
|
| 9 |
+
"architectures": [
|
| 10 |
+
"Wav2Vec2BertForCTC"
|
| 11 |
+
],
|
| 12 |
+
"attention_dropout": 0.0,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"classifier_proj_size": 768,
|
| 15 |
+
"codevector_dim": 768,
|
| 16 |
+
"conformer_conv_dropout": 0.1,
|
| 17 |
+
"contrastive_logits_temperature": 0.1,
|
| 18 |
+
"conv_depthwise_kernel_size": 31,
|
| 19 |
+
"ctc_loss_reduction": "mean",
|
| 20 |
+
"ctc_zero_infinity": true,
|
| 21 |
+
"diversity_loss_weight": 0.1,
|
| 22 |
+
"eos_token_id": 2,
|
| 23 |
+
"feat_proj_dropout": 0.0,
|
| 24 |
+
"feat_quantizer_dropout": 0.0,
|
| 25 |
+
"feature_projection_input_dim": 160,
|
| 26 |
+
"final_dropout": 0.1,
|
| 27 |
+
"hidden_act": "swish",
|
| 28 |
+
"hidden_dropout": 0.0,
|
| 29 |
+
"hidden_size": 1024,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 4096,
|
| 32 |
+
"layer_norm_eps": 1e-05,
|
| 33 |
+
"layerdrop": 0.0,
|
| 34 |
+
"left_max_position_embeddings": 64,
|
| 35 |
+
"mask_feature_length": 10,
|
| 36 |
+
"mask_feature_min_masks": 0,
|
| 37 |
+
"mask_feature_prob": 0.0,
|
| 38 |
+
"mask_time_length": 10,
|
| 39 |
+
"mask_time_min_masks": 2,
|
| 40 |
+
"mask_time_prob": 0.0,
|
| 41 |
+
"max_source_positions": 5000,
|
| 42 |
+
"model_type": "wav2vec2-bert",
|
| 43 |
+
"num_adapter_layers": 1,
|
| 44 |
+
"num_attention_heads": 16,
|
| 45 |
+
"num_codevector_groups": 2,
|
| 46 |
+
"num_codevectors_per_group": 320,
|
| 47 |
+
"num_hidden_layers": 24,
|
| 48 |
+
"num_negatives": 100,
|
| 49 |
+
"output_hidden_size": 1024,
|
| 50 |
+
"pad_token_id": 43,
|
| 51 |
+
"position_embeddings_type": "relative_key",
|
| 52 |
+
"proj_codevector_dim": 768,
|
| 53 |
+
"right_max_position_embeddings": 8,
|
| 54 |
+
"rotary_embedding_base": 10000,
|
| 55 |
+
"tdnn_dilation": [
|
| 56 |
+
1,
|
| 57 |
+
2,
|
| 58 |
+
3,
|
| 59 |
+
1,
|
| 60 |
+
1
|
| 61 |
+
],
|
| 62 |
+
"tdnn_dim": [
|
| 63 |
+
512,
|
| 64 |
+
512,
|
| 65 |
+
512,
|
| 66 |
+
512,
|
| 67 |
+
1500
|
| 68 |
+
],
|
| 69 |
+
"tdnn_kernel": [
|
| 70 |
+
5,
|
| 71 |
+
3,
|
| 72 |
+
3,
|
| 73 |
+
1,
|
| 74 |
+
1
|
| 75 |
+
],
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.46.0",
|
| 78 |
+
"use_intermediate_ffn_before_adapter": false,
|
| 79 |
+
"use_weighted_layer_sum": false,
|
| 80 |
+
"vocab_size": 46,
|
| 81 |
+
"xvector_output_dim": 512
|
| 82 |
+
}
|
checkpoint-6450/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:651d4ab1a939a60328167420b4198d49fd46c063364f7ae81851515cf978b6a8
|
| 3 |
+
size 2423003160
|
checkpoint-6450/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:197c39417deff8903d763a98bde904cb4f9a2d8dd17df2c9c195a8fe1804738a
|
| 3 |
+
size 4846467050
|
checkpoint-6450/preprocessor_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"feature_extractor_type": "SeamlessM4TFeatureExtractor",
|
| 3 |
+
"feature_size": 80,
|
| 4 |
+
"num_mel_bins": 80,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 1,
|
| 7 |
+
"processor_class": "Wav2Vec2BertProcessor",
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000,
|
| 10 |
+
"stride": 2
|
| 11 |
+
}
|
checkpoint-6450/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a1796469fe1cae90b24cab33d0b4f5d25b0e07d3c4278049286a455d4d9f4e9
|
| 3 |
+
size 14244
|
checkpoint-6450/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f52fdbf431f333341f47e5b047d4255f77560a279f8a2b38abcb68b467c4eaa
|
| 3 |
+
size 1064
|
checkpoint-6450/trainer_state.json
ADDED
|
@@ -0,0 +1,467 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4217069089170452,
|
| 3 |
+
"best_model_checkpoint": "w2v-bert-2.0-CV_Fleurs-lg-5hrs-v4/checkpoint-3354",
|
| 4 |
+
"epoch": 25.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 6450,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"grad_norm": 11.246596336364746,
|
| 14 |
+
"learning_rate": 5.1200000000000004e-05,
|
| 15 |
+
"loss": 1.9959,
|
| 16 |
+
"step": 258
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_cer": 0.13379445701894616,
|
| 21 |
+
"eval_loss": 0.5475470423698425,
|
| 22 |
+
"eval_runtime": 29.6601,
|
| 23 |
+
"eval_samples_per_second": 27.849,
|
| 24 |
+
"eval_steps_per_second": 6.979,
|
| 25 |
+
"eval_wer": 0.6078173559646187,
|
| 26 |
+
"step": 258
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.0,
|
| 30 |
+
"grad_norm": 1.1338785886764526,
|
| 31 |
+
"learning_rate": 9.994466403162056e-05,
|
| 32 |
+
"loss": 0.5447,
|
| 33 |
+
"step": 516
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 2.0,
|
| 37 |
+
"eval_cer": 0.1309612371199291,
|
| 38 |
+
"eval_loss": 0.5649892687797546,
|
| 39 |
+
"eval_runtime": 29.1014,
|
| 40 |
+
"eval_samples_per_second": 28.384,
|
| 41 |
+
"eval_steps_per_second": 7.113,
|
| 42 |
+
"eval_wer": 0.5978962467128854,
|
| 43 |
+
"step": 516
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 3.0,
|
| 47 |
+
"grad_norm": 5.585188388824463,
|
| 48 |
+
"learning_rate": 9.892490118577076e-05,
|
| 49 |
+
"loss": 0.4574,
|
| 50 |
+
"step": 774
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.0,
|
| 54 |
+
"eval_cer": 0.11218917678342487,
|
| 55 |
+
"eval_loss": 0.45441317558288574,
|
| 56 |
+
"eval_runtime": 29.5398,
|
| 57 |
+
"eval_samples_per_second": 27.962,
|
| 58 |
+
"eval_steps_per_second": 7.007,
|
| 59 |
+
"eval_wer": 0.5168539325842697,
|
| 60 |
+
"step": 774
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 4.0,
|
| 64 |
+
"grad_norm": 3.5034706592559814,
|
| 65 |
+
"learning_rate": 9.790513833992096e-05,
|
| 66 |
+
"loss": 0.3712,
|
| 67 |
+
"step": 1032
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 4.0,
|
| 71 |
+
"eval_cer": 0.10859621076623562,
|
| 72 |
+
"eval_loss": 0.4465671181678772,
|
| 73 |
+
"eval_runtime": 29.2068,
|
| 74 |
+
"eval_samples_per_second": 28.281,
|
| 75 |
+
"eval_steps_per_second": 7.087,
|
| 76 |
+
"eval_wer": 0.5049007889074827,
|
| 77 |
+
"step": 1032
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.0,
|
| 81 |
+
"grad_norm": 17.74321174621582,
|
| 82 |
+
"learning_rate": 9.688932806324111e-05,
|
| 83 |
+
"loss": 0.3187,
|
| 84 |
+
"step": 1290
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 5.0,
|
| 88 |
+
"eval_cer": 0.10036562781937036,
|
| 89 |
+
"eval_loss": 0.44379866123199463,
|
| 90 |
+
"eval_runtime": 29.231,
|
| 91 |
+
"eval_samples_per_second": 28.258,
|
| 92 |
+
"eval_steps_per_second": 7.082,
|
| 93 |
+
"eval_wer": 0.4769304327038011,
|
| 94 |
+
"step": 1290
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 6.0,
|
| 98 |
+
"grad_norm": 10.85033893585205,
|
| 99 |
+
"learning_rate": 9.586956521739131e-05,
|
| 100 |
+
"loss": 0.276,
|
| 101 |
+
"step": 1548
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 6.0,
|
| 105 |
+
"eval_cer": 0.1056996786907042,
|
| 106 |
+
"eval_loss": 0.4188626706600189,
|
| 107 |
+
"eval_runtime": 29.2483,
|
| 108 |
+
"eval_samples_per_second": 28.241,
|
| 109 |
+
"eval_steps_per_second": 7.077,
|
| 110 |
+
"eval_wer": 0.4794405928759264,
|
| 111 |
+
"step": 1548
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 7.0,
|
| 115 |
+
"grad_norm": 1.633737564086914,
|
| 116 |
+
"learning_rate": 9.484980237154151e-05,
|
| 117 |
+
"loss": 0.235,
|
| 118 |
+
"step": 1806
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 7.0,
|
| 122 |
+
"eval_cer": 0.1004447680400133,
|
| 123 |
+
"eval_loss": 0.4695667326450348,
|
| 124 |
+
"eval_runtime": 29.6156,
|
| 125 |
+
"eval_samples_per_second": 27.891,
|
| 126 |
+
"eval_steps_per_second": 6.99,
|
| 127 |
+
"eval_wer": 0.4796796557494621,
|
| 128 |
+
"step": 1806
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"grad_norm": 7.645702838897705,
|
| 133 |
+
"learning_rate": 9.383003952569171e-05,
|
| 134 |
+
"loss": 0.2086,
|
| 135 |
+
"step": 2064
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 8.0,
|
| 139 |
+
"eval_cer": 0.09632947656658067,
|
| 140 |
+
"eval_loss": 0.4218059182167053,
|
| 141 |
+
"eval_runtime": 29.7338,
|
| 142 |
+
"eval_samples_per_second": 27.78,
|
| 143 |
+
"eval_steps_per_second": 6.962,
|
| 144 |
+
"eval_wer": 0.4603155629930672,
|
| 145 |
+
"step": 2064
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 9.0,
|
| 149 |
+
"grad_norm": 5.886650562286377,
|
| 150 |
+
"learning_rate": 9.281027667984191e-05,
|
| 151 |
+
"loss": 0.1883,
|
| 152 |
+
"step": 2322
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 9.0,
|
| 156 |
+
"eval_cer": 0.09757989205273904,
|
| 157 |
+
"eval_loss": 0.4250072240829468,
|
| 158 |
+
"eval_runtime": 29.3849,
|
| 159 |
+
"eval_samples_per_second": 28.11,
|
| 160 |
+
"eval_steps_per_second": 7.044,
|
| 161 |
+
"eval_wer": 0.45206789385608415,
|
| 162 |
+
"step": 2322
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 10.0,
|
| 166 |
+
"grad_norm": 1.1357783079147339,
|
| 167 |
+
"learning_rate": 9.179051383399209e-05,
|
| 168 |
+
"loss": 0.1671,
|
| 169 |
+
"step": 2580
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.0,
|
| 173 |
+
"eval_cer": 0.0972633311701673,
|
| 174 |
+
"eval_loss": 0.45072105526924133,
|
| 175 |
+
"eval_runtime": 29.0789,
|
| 176 |
+
"eval_samples_per_second": 28.405,
|
| 177 |
+
"eval_steps_per_second": 7.119,
|
| 178 |
+
"eval_wer": 0.458642122878317,
|
| 179 |
+
"step": 2580
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 11.0,
|
| 183 |
+
"grad_norm": 0.13285361230373383,
|
| 184 |
+
"learning_rate": 9.077075098814229e-05,
|
| 185 |
+
"loss": 0.1466,
|
| 186 |
+
"step": 2838
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 11.0,
|
| 190 |
+
"eval_cer": 0.09696259833172415,
|
| 191 |
+
"eval_loss": 0.4737579822540283,
|
| 192 |
+
"eval_runtime": 29.2438,
|
| 193 |
+
"eval_samples_per_second": 28.245,
|
| 194 |
+
"eval_steps_per_second": 7.078,
|
| 195 |
+
"eval_wer": 0.4634233803490318,
|
| 196 |
+
"step": 2838
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 12.0,
|
| 200 |
+
"grad_norm": 18.049510955810547,
|
| 201 |
+
"learning_rate": 8.975098814229249e-05,
|
| 202 |
+
"loss": 0.1313,
|
| 203 |
+
"step": 3096
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 12.0,
|
| 207 |
+
"eval_cer": 0.09357539688820653,
|
| 208 |
+
"eval_loss": 0.46005183458328247,
|
| 209 |
+
"eval_runtime": 29.8062,
|
| 210 |
+
"eval_samples_per_second": 27.712,
|
| 211 |
+
"eval_steps_per_second": 6.945,
|
| 212 |
+
"eval_wer": 0.4410710016734401,
|
| 213 |
+
"step": 3096
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 13.0,
|
| 217 |
+
"grad_norm": 1.1974396705627441,
|
| 218 |
+
"learning_rate": 8.873517786561266e-05,
|
| 219 |
+
"loss": 0.1181,
|
| 220 |
+
"step": 3354
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 13.0,
|
| 224 |
+
"eval_cer": 0.08908023235568781,
|
| 225 |
+
"eval_loss": 0.47361627221107483,
|
| 226 |
+
"eval_runtime": 29.1477,
|
| 227 |
+
"eval_samples_per_second": 28.338,
|
| 228 |
+
"eval_steps_per_second": 7.102,
|
| 229 |
+
"eval_wer": 0.4217069089170452,
|
| 230 |
+
"step": 3354
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 14.0,
|
| 234 |
+
"grad_norm": 1.3642749786376953,
|
| 235 |
+
"learning_rate": 8.771541501976284e-05,
|
| 236 |
+
"loss": 0.1026,
|
| 237 |
+
"step": 3612
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 14.0,
|
| 241 |
+
"eval_cer": 0.09121701831304706,
|
| 242 |
+
"eval_loss": 0.4470472037792206,
|
| 243 |
+
"eval_runtime": 29.2097,
|
| 244 |
+
"eval_samples_per_second": 28.278,
|
| 245 |
+
"eval_steps_per_second": 7.087,
|
| 246 |
+
"eval_wer": 0.42780301219220657,
|
| 247 |
+
"step": 3612
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 15.0,
|
| 251 |
+
"grad_norm": 50.79623031616211,
|
| 252 |
+
"learning_rate": 8.669565217391304e-05,
|
| 253 |
+
"loss": 0.092,
|
| 254 |
+
"step": 3870
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 15.0,
|
| 258 |
+
"eval_cer": 0.09560138653666567,
|
| 259 |
+
"eval_loss": 0.4730050265789032,
|
| 260 |
+
"eval_runtime": 29.5306,
|
| 261 |
+
"eval_samples_per_second": 27.971,
|
| 262 |
+
"eval_steps_per_second": 7.01,
|
| 263 |
+
"eval_wer": 0.4619890031078174,
|
| 264 |
+
"step": 3870
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 16.0,
|
| 268 |
+
"grad_norm": 0.18782581388950348,
|
| 269 |
+
"learning_rate": 8.567588932806324e-05,
|
| 270 |
+
"loss": 0.083,
|
| 271 |
+
"step": 4128
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 16.0,
|
| 275 |
+
"eval_cer": 0.09370202124123522,
|
| 276 |
+
"eval_loss": 0.533858597278595,
|
| 277 |
+
"eval_runtime": 30.1467,
|
| 278 |
+
"eval_samples_per_second": 27.399,
|
| 279 |
+
"eval_steps_per_second": 6.866,
|
| 280 |
+
"eval_wer": 0.44824288787951233,
|
| 281 |
+
"step": 4128
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.0,
|
| 285 |
+
"grad_norm": 0.19636400043964386,
|
| 286 |
+
"learning_rate": 8.465612648221344e-05,
|
| 287 |
+
"loss": 0.0744,
|
| 288 |
+
"step": 4386
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 17.0,
|
| 292 |
+
"eval_cer": 0.09436679909463587,
|
| 293 |
+
"eval_loss": 0.485524982213974,
|
| 294 |
+
"eval_runtime": 30.0117,
|
| 295 |
+
"eval_samples_per_second": 27.523,
|
| 296 |
+
"eval_steps_per_second": 6.897,
|
| 297 |
+
"eval_wer": 0.4508725794884055,
|
| 298 |
+
"step": 4386
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 18.0,
|
| 302 |
+
"grad_norm": 3.5025405883789062,
|
| 303 |
+
"learning_rate": 8.363636363636364e-05,
|
| 304 |
+
"loss": 0.0697,
|
| 305 |
+
"step": 4644
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 18.0,
|
| 309 |
+
"eval_cer": 0.09033064784184619,
|
| 310 |
+
"eval_loss": 0.5221356153488159,
|
| 311 |
+
"eval_runtime": 29.9898,
|
| 312 |
+
"eval_samples_per_second": 27.543,
|
| 313 |
+
"eval_steps_per_second": 6.902,
|
| 314 |
+
"eval_wer": 0.43748505857040404,
|
| 315 |
+
"step": 4644
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 19.0,
|
| 319 |
+
"grad_norm": 1.851936936378479,
|
| 320 |
+
"learning_rate": 8.261660079051384e-05,
|
| 321 |
+
"loss": 0.0602,
|
| 322 |
+
"step": 4902
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 19.0,
|
| 326 |
+
"eval_cer": 0.08944427737064531,
|
| 327 |
+
"eval_loss": 0.51482093334198,
|
| 328 |
+
"eval_runtime": 29.3263,
|
| 329 |
+
"eval_samples_per_second": 28.166,
|
| 330 |
+
"eval_steps_per_second": 7.059,
|
| 331 |
+
"eval_wer": 0.42708582357159935,
|
| 332 |
+
"step": 4902
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 20.0,
|
| 336 |
+
"grad_norm": 0.04486146196722984,
|
| 337 |
+
"learning_rate": 8.159683794466404e-05,
|
| 338 |
+
"loss": 0.056,
|
| 339 |
+
"step": 5160
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 20.0,
|
| 343 |
+
"eval_cer": 0.08983997847385998,
|
| 344 |
+
"eval_loss": 0.5517579317092896,
|
| 345 |
+
"eval_runtime": 29.8422,
|
| 346 |
+
"eval_samples_per_second": 27.679,
|
| 347 |
+
"eval_steps_per_second": 6.936,
|
| 348 |
+
"eval_wer": 0.4312694238584748,
|
| 349 |
+
"step": 5160
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 21.0,
|
| 353 |
+
"grad_norm": 0.09858173131942749,
|
| 354 |
+
"learning_rate": 8.057707509881424e-05,
|
| 355 |
+
"loss": 0.05,
|
| 356 |
+
"step": 5418
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 21.0,
|
| 360 |
+
"eval_cer": 0.09120119026891847,
|
| 361 |
+
"eval_loss": 0.5373868346214294,
|
| 362 |
+
"eval_runtime": 29.6954,
|
| 363 |
+
"eval_samples_per_second": 27.816,
|
| 364 |
+
"eval_steps_per_second": 6.971,
|
| 365 |
+
"eval_wer": 0.431030360984939,
|
| 366 |
+
"step": 5418
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 22.0,
|
| 370 |
+
"grad_norm": 0.7042592167854309,
|
| 371 |
+
"learning_rate": 7.955731225296444e-05,
|
| 372 |
+
"loss": 0.0464,
|
| 373 |
+
"step": 5676
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 22.0,
|
| 377 |
+
"eval_cer": 0.08990329065037433,
|
| 378 |
+
"eval_loss": 0.5166798233985901,
|
| 379 |
+
"eval_runtime": 29.6186,
|
| 380 |
+
"eval_samples_per_second": 27.888,
|
| 381 |
+
"eval_steps_per_second": 6.989,
|
| 382 |
+
"eval_wer": 0.42648816638776,
|
| 383 |
+
"step": 5676
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 23.0,
|
| 387 |
+
"grad_norm": 8.724380493164062,
|
| 388 |
+
"learning_rate": 7.853754940711463e-05,
|
| 389 |
+
"loss": 0.0453,
|
| 390 |
+
"step": 5934
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 23.0,
|
| 394 |
+
"eval_cer": 0.08813054970797259,
|
| 395 |
+
"eval_loss": 0.5781718492507935,
|
| 396 |
+
"eval_runtime": 29.9397,
|
| 397 |
+
"eval_samples_per_second": 27.589,
|
| 398 |
+
"eval_steps_per_second": 6.914,
|
| 399 |
+
"eval_wer": 0.4226631604111881,
|
| 400 |
+
"step": 5934
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 24.0,
|
| 404 |
+
"grad_norm": 0.4805084764957428,
|
| 405 |
+
"learning_rate": 7.751778656126482e-05,
|
| 406 |
+
"loss": 0.0412,
|
| 407 |
+
"step": 6192
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 24.0,
|
| 411 |
+
"eval_cer": 0.09294227512306304,
|
| 412 |
+
"eval_loss": 0.5274640321731567,
|
| 413 |
+
"eval_runtime": 29.0914,
|
| 414 |
+
"eval_samples_per_second": 28.393,
|
| 415 |
+
"eval_steps_per_second": 7.116,
|
| 416 |
+
"eval_wer": 0.43533349270858235,
|
| 417 |
+
"step": 6192
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 25.0,
|
| 421 |
+
"grad_norm": 0.1090458482503891,
|
| 422 |
+
"learning_rate": 7.649802371541502e-05,
|
| 423 |
+
"loss": 0.0369,
|
| 424 |
+
"step": 6450
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 25.0,
|
| 428 |
+
"eval_cer": 0.09186596812231912,
|
| 429 |
+
"eval_loss": 0.6112269759178162,
|
| 430 |
+
"eval_runtime": 29.3026,
|
| 431 |
+
"eval_samples_per_second": 28.189,
|
| 432 |
+
"eval_steps_per_second": 7.064,
|
| 433 |
+
"eval_wer": 0.4233803490317954,
|
| 434 |
+
"step": 6450
|
| 435 |
+
}
|
| 436 |
+
],
|
| 437 |
+
"logging_steps": 500,
|
| 438 |
+
"max_steps": 25800,
|
| 439 |
+
"num_input_tokens_seen": 0,
|
| 440 |
+
"num_train_epochs": 100,
|
| 441 |
+
"save_steps": 500,
|
| 442 |
+
"stateful_callbacks": {
|
| 443 |
+
"EarlyStoppingCallback": {
|
| 444 |
+
"args": {
|
| 445 |
+
"early_stopping_patience": 20,
|
| 446 |
+
"early_stopping_threshold": 0.001
|
| 447 |
+
},
|
| 448 |
+
"attributes": {
|
| 449 |
+
"early_stopping_patience_counter": 12
|
| 450 |
+
}
|
| 451 |
+
},
|
| 452 |
+
"TrainerControl": {
|
| 453 |
+
"args": {
|
| 454 |
+
"should_epoch_stop": false,
|
| 455 |
+
"should_evaluate": false,
|
| 456 |
+
"should_log": false,
|
| 457 |
+
"should_save": true,
|
| 458 |
+
"should_training_stop": false
|
| 459 |
+
},
|
| 460 |
+
"attributes": {}
|
| 461 |
+
}
|
| 462 |
+
},
|
| 463 |
+
"total_flos": 1.3654467547432352e+19,
|
| 464 |
+
"train_batch_size": 8,
|
| 465 |
+
"trial_name": null,
|
| 466 |
+
"trial_params": null
|
| 467 |
+
}
|
checkpoint-6450/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:badc6886eb2995443a897620ab2c36decc3b66898e976799a420ad7fb69a195c
|
| 3 |
+
size 5304
|