Training in progress, epoch 26, checkpoint
Browse files- checkpoint-13338/config.json +82 -0
- checkpoint-13338/model.safetensors +3 -0
- checkpoint-13338/optimizer.pt +3 -0
- checkpoint-13338/preprocessor_config.json +11 -0
- checkpoint-13338/rng_state.pth +3 -0
- checkpoint-13338/scheduler.pt +3 -0
- checkpoint-13338/trainer_state.json +484 -0
- checkpoint-13338/training_args.bin +3 -0
checkpoint-13338/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": 45,
|
| 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": 48,
|
| 81 |
+
"xvector_output_dim": 512
|
| 82 |
+
}
|
checkpoint-13338/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:742bbd160a92d02474b4bbe90a29e2caa5ebd0a5ec6b42ad9a725e8807ea2ac1
|
| 3 |
+
size 2423011360
|
checkpoint-13338/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0b5aee09e9da4208dd8eb505200289b75f69a088cb2f0da981341a6bd84cb79
|
| 3 |
+
size 4846483562
|
checkpoint-13338/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-13338/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ece7901419c3763c1113df5095e1e388239513290a3afdfd2a0785f731b21fb3
|
| 3 |
+
size 14244
|
checkpoint-13338/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:527a10dc0042d6368b90bc1fec48b2c2ec248f9dabfe47c0322d92a3991f7b56
|
| 3 |
+
size 1064
|
checkpoint-13338/trainer_state.json
ADDED
|
@@ -0,0 +1,484 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.377719340186469,
|
| 3 |
+
"best_model_checkpoint": "w2v-bert-2.0-CV_Fleurs-lg-10hrs-v4/checkpoint-11286",
|
| 4 |
+
"epoch": 26.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 13338,
|
| 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": 30.792112350463867,
|
| 14 |
+
"learning_rate": 9.998031496062993e-05,
|
| 15 |
+
"loss": 1.4106,
|
| 16 |
+
"step": 513
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_cer": 0.1264185884550246,
|
| 21 |
+
"eval_loss": 0.5293365716934204,
|
| 22 |
+
"eval_runtime": 31.0205,
|
| 23 |
+
"eval_samples_per_second": 26.628,
|
| 24 |
+
"eval_steps_per_second": 6.673,
|
| 25 |
+
"eval_wer": 0.5665790102797036,
|
| 26 |
+
"step": 513
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.0,
|
| 30 |
+
"grad_norm": 10.16055679321289,
|
| 31 |
+
"learning_rate": 9.897047244094488e-05,
|
| 32 |
+
"loss": 0.4775,
|
| 33 |
+
"step": 1026
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 2.0,
|
| 37 |
+
"eval_cer": 0.11182513176846737,
|
| 38 |
+
"eval_loss": 0.4882170259952545,
|
| 39 |
+
"eval_runtime": 30.4398,
|
| 40 |
+
"eval_samples_per_second": 27.135,
|
| 41 |
+
"eval_steps_per_second": 6.8,
|
| 42 |
+
"eval_wer": 0.5251016017212526,
|
| 43 |
+
"step": 1026
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 3.0,
|
| 47 |
+
"grad_norm": 24.60751724243164,
|
| 48 |
+
"learning_rate": 9.796062992125985e-05,
|
| 49 |
+
"loss": 0.3696,
|
| 50 |
+
"step": 1539
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.0,
|
| 54 |
+
"eval_cer": 0.09981164627486981,
|
| 55 |
+
"eval_loss": 0.397329181432724,
|
| 56 |
+
"eval_runtime": 30.5019,
|
| 57 |
+
"eval_samples_per_second": 27.08,
|
| 58 |
+
"eval_steps_per_second": 6.786,
|
| 59 |
+
"eval_wer": 0.4564905570164953,
|
| 60 |
+
"step": 1539
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 4.0,
|
| 64 |
+
"grad_norm": 8.973501205444336,
|
| 65 |
+
"learning_rate": 9.695078740157481e-05,
|
| 66 |
+
"loss": 0.3007,
|
| 67 |
+
"step": 2052
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 4.0,
|
| 71 |
+
"eval_cer": 0.09322717991737761,
|
| 72 |
+
"eval_loss": 0.3895888030529022,
|
| 73 |
+
"eval_runtime": 29.8209,
|
| 74 |
+
"eval_samples_per_second": 27.699,
|
| 75 |
+
"eval_steps_per_second": 6.941,
|
| 76 |
+
"eval_wer": 0.4311498924217069,
|
| 77 |
+
"step": 2052
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.0,
|
| 81 |
+
"grad_norm": 3.461097240447998,
|
| 82 |
+
"learning_rate": 9.594094488188978e-05,
|
| 83 |
+
"loss": 0.2564,
|
| 84 |
+
"step": 2565
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 5.0,
|
| 88 |
+
"eval_cer": 0.09659855331676664,
|
| 89 |
+
"eval_loss": 0.3743041455745697,
|
| 90 |
+
"eval_runtime": 29.9644,
|
| 91 |
+
"eval_samples_per_second": 27.566,
|
| 92 |
+
"eval_steps_per_second": 6.908,
|
| 93 |
+
"eval_wer": 0.4372459956968683,
|
| 94 |
+
"step": 2565
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 6.0,
|
| 98 |
+
"grad_norm": 9.60593032836914,
|
| 99 |
+
"learning_rate": 9.493110236220472e-05,
|
| 100 |
+
"loss": 0.2273,
|
| 101 |
+
"step": 3078
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 6.0,
|
| 105 |
+
"eval_cer": 0.0879880973108153,
|
| 106 |
+
"eval_loss": 0.3549309968948364,
|
| 107 |
+
"eval_runtime": 30.2119,
|
| 108 |
+
"eval_samples_per_second": 27.34,
|
| 109 |
+
"eval_steps_per_second": 6.852,
|
| 110 |
+
"eval_wer": 0.41836002868754485,
|
| 111 |
+
"step": 3078
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 7.0,
|
| 115 |
+
"grad_norm": 2.7023348808288574,
|
| 116 |
+
"learning_rate": 9.392125984251969e-05,
|
| 117 |
+
"loss": 0.203,
|
| 118 |
+
"step": 3591
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 7.0,
|
| 122 |
+
"eval_cer": 0.08468003608794061,
|
| 123 |
+
"eval_loss": 0.3835783302783966,
|
| 124 |
+
"eval_runtime": 29.6315,
|
| 125 |
+
"eval_samples_per_second": 27.876,
|
| 126 |
+
"eval_steps_per_second": 6.986,
|
| 127 |
+
"eval_wer": 0.4000717188620607,
|
| 128 |
+
"step": 3591
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"grad_norm": 2.196669340133667,
|
| 133 |
+
"learning_rate": 9.291141732283465e-05,
|
| 134 |
+
"loss": 0.1775,
|
| 135 |
+
"step": 4104
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 8.0,
|
| 139 |
+
"eval_cer": 0.08550309438262714,
|
| 140 |
+
"eval_loss": 0.3693934977054596,
|
| 141 |
+
"eval_runtime": 30.2095,
|
| 142 |
+
"eval_samples_per_second": 27.342,
|
| 143 |
+
"eval_steps_per_second": 6.852,
|
| 144 |
+
"eval_wer": 0.4132201769065264,
|
| 145 |
+
"step": 4104
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 9.0,
|
| 149 |
+
"grad_norm": 3.045036792755127,
|
| 150 |
+
"learning_rate": 9.190354330708662e-05,
|
| 151 |
+
"loss": 0.159,
|
| 152 |
+
"step": 4617
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 9.0,
|
| 156 |
+
"eval_cer": 0.0889694360467877,
|
| 157 |
+
"eval_loss": 0.3487217128276825,
|
| 158 |
+
"eval_runtime": 29.9136,
|
| 159 |
+
"eval_samples_per_second": 27.613,
|
| 160 |
+
"eval_steps_per_second": 6.92,
|
| 161 |
+
"eval_wer": 0.4136983026535979,
|
| 162 |
+
"step": 4617
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 10.0,
|
| 166 |
+
"grad_norm": 3.3439998626708984,
|
| 167 |
+
"learning_rate": 9.089370078740158e-05,
|
| 168 |
+
"loss": 0.1444,
|
| 169 |
+
"step": 5130
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.0,
|
| 173 |
+
"eval_cer": 0.08915937257633075,
|
| 174 |
+
"eval_loss": 0.3705226480960846,
|
| 175 |
+
"eval_runtime": 30.2455,
|
| 176 |
+
"eval_samples_per_second": 27.31,
|
| 177 |
+
"eval_steps_per_second": 6.844,
|
| 178 |
+
"eval_wer": 0.42768348075543866,
|
| 179 |
+
"step": 5130
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 11.0,
|
| 183 |
+
"grad_norm": 3.339082956314087,
|
| 184 |
+
"learning_rate": 8.988385826771653e-05,
|
| 185 |
+
"loss": 0.1262,
|
| 186 |
+
"step": 5643
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 11.0,
|
| 190 |
+
"eval_cer": 0.08284398296902452,
|
| 191 |
+
"eval_loss": 0.3619905114173889,
|
| 192 |
+
"eval_runtime": 30.7416,
|
| 193 |
+
"eval_samples_per_second": 26.869,
|
| 194 |
+
"eval_steps_per_second": 6.734,
|
| 195 |
+
"eval_wer": 0.3981592158737748,
|
| 196 |
+
"step": 5643
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 12.0,
|
| 200 |
+
"grad_norm": 0.1660577654838562,
|
| 201 |
+
"learning_rate": 8.88740157480315e-05,
|
| 202 |
+
"loss": 0.1139,
|
| 203 |
+
"step": 6156
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 12.0,
|
| 207 |
+
"eval_cer": 0.08422102280821159,
|
| 208 |
+
"eval_loss": 0.3665634095668793,
|
| 209 |
+
"eval_runtime": 30.3996,
|
| 210 |
+
"eval_samples_per_second": 27.171,
|
| 211 |
+
"eval_steps_per_second": 6.809,
|
| 212 |
+
"eval_wer": 0.40545063351661487,
|
| 213 |
+
"step": 6156
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 13.0,
|
| 217 |
+
"grad_norm": 0.19767990708351135,
|
| 218 |
+
"learning_rate": 8.786417322834646e-05,
|
| 219 |
+
"loss": 0.1015,
|
| 220 |
+
"step": 6669
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 13.0,
|
| 224 |
+
"eval_cer": 0.08355624495481094,
|
| 225 |
+
"eval_loss": 0.3656393885612488,
|
| 226 |
+
"eval_runtime": 30.0647,
|
| 227 |
+
"eval_samples_per_second": 27.474,
|
| 228 |
+
"eval_steps_per_second": 6.885,
|
| 229 |
+
"eval_wer": 0.40114750179297154,
|
| 230 |
+
"step": 6669
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 14.0,
|
| 234 |
+
"grad_norm": 2.6208112239837646,
|
| 235 |
+
"learning_rate": 8.685433070866143e-05,
|
| 236 |
+
"loss": 0.0914,
|
| 237 |
+
"step": 7182
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 14.0,
|
| 241 |
+
"eval_cer": 0.08339796451352506,
|
| 242 |
+
"eval_loss": 0.3572923541069031,
|
| 243 |
+
"eval_runtime": 30.3516,
|
| 244 |
+
"eval_samples_per_second": 27.214,
|
| 245 |
+
"eval_steps_per_second": 6.82,
|
| 246 |
+
"eval_wer": 0.39278030121922064,
|
| 247 |
+
"step": 7182
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 15.0,
|
| 251 |
+
"grad_norm": 3.129692316055298,
|
| 252 |
+
"learning_rate": 8.584448818897639e-05,
|
| 253 |
+
"loss": 0.0836,
|
| 254 |
+
"step": 7695
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 15.0,
|
| 258 |
+
"eval_cer": 0.07895028411339211,
|
| 259 |
+
"eval_loss": 0.379753053188324,
|
| 260 |
+
"eval_runtime": 30.1224,
|
| 261 |
+
"eval_samples_per_second": 27.421,
|
| 262 |
+
"eval_steps_per_second": 6.872,
|
| 263 |
+
"eval_wer": 0.3795123117379871,
|
| 264 |
+
"step": 7695
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 16.0,
|
| 268 |
+
"grad_norm": 4.029957294464111,
|
| 269 |
+
"learning_rate": 8.483464566929134e-05,
|
| 270 |
+
"loss": 0.0729,
|
| 271 |
+
"step": 8208
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 16.0,
|
| 275 |
+
"eval_cer": 0.08219503315975245,
|
| 276 |
+
"eval_loss": 0.4288513660430908,
|
| 277 |
+
"eval_runtime": 30.1728,
|
| 278 |
+
"eval_samples_per_second": 27.376,
|
| 279 |
+
"eval_steps_per_second": 6.86,
|
| 280 |
+
"eval_wer": 0.3895529524264882,
|
| 281 |
+
"step": 8208
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.0,
|
| 285 |
+
"grad_norm": 0.5638870596885681,
|
| 286 |
+
"learning_rate": 8.38248031496063e-05,
|
| 287 |
+
"loss": 0.0652,
|
| 288 |
+
"step": 8721
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 17.0,
|
| 292 |
+
"eval_cer": 0.087386631633929,
|
| 293 |
+
"eval_loss": 0.4778367280960083,
|
| 294 |
+
"eval_runtime": 30.6189,
|
| 295 |
+
"eval_samples_per_second": 26.977,
|
| 296 |
+
"eval_steps_per_second": 6.761,
|
| 297 |
+
"eval_wer": 0.4157303370786517,
|
| 298 |
+
"step": 8721
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 18.0,
|
| 302 |
+
"grad_norm": 0.010239621624350548,
|
| 303 |
+
"learning_rate": 8.281496062992127e-05,
|
| 304 |
+
"loss": 0.0599,
|
| 305 |
+
"step": 9234
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 18.0,
|
| 309 |
+
"eval_cer": 0.07918770477532092,
|
| 310 |
+
"eval_loss": 0.4570063352584839,
|
| 311 |
+
"eval_runtime": 30.2821,
|
| 312 |
+
"eval_samples_per_second": 27.277,
|
| 313 |
+
"eval_steps_per_second": 6.836,
|
| 314 |
+
"eval_wer": 0.37963184317475496,
|
| 315 |
+
"step": 9234
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 19.0,
|
| 319 |
+
"grad_norm": 3.2985517978668213,
|
| 320 |
+
"learning_rate": 8.180511811023623e-05,
|
| 321 |
+
"loss": 0.0504,
|
| 322 |
+
"step": 9747
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 19.0,
|
| 326 |
+
"eval_cer": 0.08564554677978442,
|
| 327 |
+
"eval_loss": 0.44442683458328247,
|
| 328 |
+
"eval_runtime": 30.5951,
|
| 329 |
+
"eval_samples_per_second": 26.998,
|
| 330 |
+
"eval_steps_per_second": 6.766,
|
| 331 |
+
"eval_wer": 0.4125029882859192,
|
| 332 |
+
"step": 9747
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 20.0,
|
| 336 |
+
"grad_norm": 0.014229847118258476,
|
| 337 |
+
"learning_rate": 8.079724409448819e-05,
|
| 338 |
+
"loss": 0.0498,
|
| 339 |
+
"step": 10260
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 20.0,
|
| 343 |
+
"eval_cer": 0.08422102280821159,
|
| 344 |
+
"eval_loss": 0.4611632823944092,
|
| 345 |
+
"eval_runtime": 30.4634,
|
| 346 |
+
"eval_samples_per_second": 27.115,
|
| 347 |
+
"eval_steps_per_second": 6.795,
|
| 348 |
+
"eval_wer": 0.3944537413339708,
|
| 349 |
+
"step": 10260
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 21.0,
|
| 353 |
+
"grad_norm": 0.029411843046545982,
|
| 354 |
+
"learning_rate": 7.978740157480316e-05,
|
| 355 |
+
"loss": 0.0434,
|
| 356 |
+
"step": 10773
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 21.0,
|
| 360 |
+
"eval_cer": 0.08431599107298311,
|
| 361 |
+
"eval_loss": 0.48810598254203796,
|
| 362 |
+
"eval_runtime": 30.3298,
|
| 363 |
+
"eval_samples_per_second": 27.234,
|
| 364 |
+
"eval_steps_per_second": 6.825,
|
| 365 |
+
"eval_wer": 0.40019125029882857,
|
| 366 |
+
"step": 10773
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 22.0,
|
| 370 |
+
"grad_norm": 0.19160379469394684,
|
| 371 |
+
"learning_rate": 7.877755905511811e-05,
|
| 372 |
+
"loss": 0.0386,
|
| 373 |
+
"step": 11286
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 22.0,
|
| 377 |
+
"eval_cer": 0.08035898004083636,
|
| 378 |
+
"eval_loss": 0.509917676448822,
|
| 379 |
+
"eval_runtime": 31.249,
|
| 380 |
+
"eval_samples_per_second": 26.433,
|
| 381 |
+
"eval_steps_per_second": 6.624,
|
| 382 |
+
"eval_wer": 0.377719340186469,
|
| 383 |
+
"step": 11286
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 23.0,
|
| 387 |
+
"grad_norm": 0.18706804513931274,
|
| 388 |
+
"learning_rate": 7.776771653543308e-05,
|
| 389 |
+
"loss": 0.0381,
|
| 390 |
+
"step": 11799
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 23.0,
|
| 394 |
+
"eval_cer": 0.08241662577755267,
|
| 395 |
+
"eval_loss": 0.49042966961860657,
|
| 396 |
+
"eval_runtime": 30.3145,
|
| 397 |
+
"eval_samples_per_second": 27.248,
|
| 398 |
+
"eval_steps_per_second": 6.828,
|
| 399 |
+
"eval_wer": 0.38656466650729143,
|
| 400 |
+
"step": 11799
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 24.0,
|
| 404 |
+
"grad_norm": 0.048221975564956665,
|
| 405 |
+
"learning_rate": 7.675984251968504e-05,
|
| 406 |
+
"loss": 0.0344,
|
| 407 |
+
"step": 12312
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 24.0,
|
| 411 |
+
"eval_cer": 0.08344544864591083,
|
| 412 |
+
"eval_loss": 0.4621882736682892,
|
| 413 |
+
"eval_runtime": 30.4774,
|
| 414 |
+
"eval_samples_per_second": 27.102,
|
| 415 |
+
"eval_steps_per_second": 6.792,
|
| 416 |
+
"eval_wer": 0.4028209419077217,
|
| 417 |
+
"step": 12312
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 25.0,
|
| 421 |
+
"grad_norm": 1.851014494895935,
|
| 422 |
+
"learning_rate": 7.575e-05,
|
| 423 |
+
"loss": 0.0302,
|
| 424 |
+
"step": 12825
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 25.0,
|
| 428 |
+
"eval_cer": 0.08198926858608081,
|
| 429 |
+
"eval_loss": 0.4985555112361908,
|
| 430 |
+
"eval_runtime": 30.5557,
|
| 431 |
+
"eval_samples_per_second": 27.033,
|
| 432 |
+
"eval_steps_per_second": 6.775,
|
| 433 |
+
"eval_wer": 0.39182404972507767,
|
| 434 |
+
"step": 12825
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 26.0,
|
| 438 |
+
"grad_norm": 0.03737611323595047,
|
| 439 |
+
"learning_rate": 7.474015748031497e-05,
|
| 440 |
+
"loss": 0.0268,
|
| 441 |
+
"step": 13338
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 26.0,
|
| 445 |
+
"eval_cer": 0.08118203833552289,
|
| 446 |
+
"eval_loss": 0.5162001252174377,
|
| 447 |
+
"eval_runtime": 30.3567,
|
| 448 |
+
"eval_samples_per_second": 27.21,
|
| 449 |
+
"eval_steps_per_second": 6.819,
|
| 450 |
+
"eval_wer": 0.3954099928281138,
|
| 451 |
+
"step": 13338
|
| 452 |
+
}
|
| 453 |
+
],
|
| 454 |
+
"logging_steps": 500,
|
| 455 |
+
"max_steps": 51300,
|
| 456 |
+
"num_input_tokens_seen": 0,
|
| 457 |
+
"num_train_epochs": 100,
|
| 458 |
+
"save_steps": 500,
|
| 459 |
+
"stateful_callbacks": {
|
| 460 |
+
"EarlyStoppingCallback": {
|
| 461 |
+
"args": {
|
| 462 |
+
"early_stopping_patience": 20,
|
| 463 |
+
"early_stopping_threshold": 0.001
|
| 464 |
+
},
|
| 465 |
+
"attributes": {
|
| 466 |
+
"early_stopping_patience_counter": 4
|
| 467 |
+
}
|
| 468 |
+
},
|
| 469 |
+
"TrainerControl": {
|
| 470 |
+
"args": {
|
| 471 |
+
"should_epoch_stop": false,
|
| 472 |
+
"should_evaluate": false,
|
| 473 |
+
"should_log": false,
|
| 474 |
+
"should_save": true,
|
| 475 |
+
"should_training_stop": false
|
| 476 |
+
},
|
| 477 |
+
"attributes": {}
|
| 478 |
+
}
|
| 479 |
+
},
|
| 480 |
+
"total_flos": 2.8373469830456582e+19,
|
| 481 |
+
"train_batch_size": 8,
|
| 482 |
+
"trial_name": null,
|
| 483 |
+
"trial_params": null
|
| 484 |
+
}
|
checkpoint-13338/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b71a1c487e8560b2d69fe2832962ae0eb6d2d0e7304ab8707b68c92ffe58d66
|
| 3 |
+
size 5304
|