Upload 14 files
Browse files- all_results.json +8 -0
- config.json +108 -0
- optimizer.pt +3 -0
- preprocessor_config.json +10 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +6 -0
- tokenizer_config.json +14 -0
- train_results.json +8 -0
- trainer_state.json +666 -0
- training_args.bin +3 -0
- vocab.json +46 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 11.39,
|
| 3 |
+
"train_loss": 218.4774083984375,
|
| 4 |
+
"train_runtime": 9202.4528,
|
| 5 |
+
"train_samples": 7024,
|
| 6 |
+
"train_samples_per_second": 8.693,
|
| 7 |
+
"train_steps_per_second": 1.087
|
| 8 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "lgris/bp_400h_xlsr2_300M",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"adapter_kernel_size": 3,
|
| 5 |
+
"adapter_stride": 2,
|
| 6 |
+
"add_adapter": false,
|
| 7 |
+
"apply_spec_augment": true,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Wav2Vec2ForCTC"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.1,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"classifier_proj_size": 256,
|
| 14 |
+
"codevector_dim": 768,
|
| 15 |
+
"contrastive_logits_temperature": 0.1,
|
| 16 |
+
"conv_bias": true,
|
| 17 |
+
"conv_dim": [
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512
|
| 25 |
+
],
|
| 26 |
+
"conv_kernel": [
|
| 27 |
+
10,
|
| 28 |
+
3,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
2,
|
| 33 |
+
2
|
| 34 |
+
],
|
| 35 |
+
"conv_stride": [
|
| 36 |
+
5,
|
| 37 |
+
2,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2
|
| 43 |
+
],
|
| 44 |
+
"ctc_loss_reduction": "sum",
|
| 45 |
+
"ctc_zero_infinity": false,
|
| 46 |
+
"diversity_loss_weight": 0.1,
|
| 47 |
+
"do_stable_layer_norm": true,
|
| 48 |
+
"eos_token_id": 2,
|
| 49 |
+
"feat_extract_activation": "gelu",
|
| 50 |
+
"feat_extract_dropout": 0.0,
|
| 51 |
+
"feat_extract_norm": "layer",
|
| 52 |
+
"feat_proj_dropout": 0.1,
|
| 53 |
+
"feat_quantizer_dropout": 0.0,
|
| 54 |
+
"final_dropout": 0.1,
|
| 55 |
+
"gradient_checkpointing": false,
|
| 56 |
+
"hidden_act": "gelu",
|
| 57 |
+
"hidden_dropout": 0.1,
|
| 58 |
+
"hidden_size": 1024,
|
| 59 |
+
"initializer_range": 0.02,
|
| 60 |
+
"intermediate_size": 4096,
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "wav2vec2",
|
| 70 |
+
"num_adapter_layers": 3,
|
| 71 |
+
"num_attention_heads": 16,
|
| 72 |
+
"num_codevector_groups": 2,
|
| 73 |
+
"num_codevectors_per_group": 320,
|
| 74 |
+
"num_conv_pos_embedding_groups": 16,
|
| 75 |
+
"num_conv_pos_embeddings": 128,
|
| 76 |
+
"num_feat_extract_layers": 7,
|
| 77 |
+
"num_hidden_layers": 24,
|
| 78 |
+
"num_negatives": 100,
|
| 79 |
+
"output_hidden_size": 1024,
|
| 80 |
+
"pad_token_id": 0,
|
| 81 |
+
"proj_codevector_dim": 768,
|
| 82 |
+
"tdnn_dilation": [
|
| 83 |
+
1,
|
| 84 |
+
2,
|
| 85 |
+
3,
|
| 86 |
+
1,
|
| 87 |
+
1
|
| 88 |
+
],
|
| 89 |
+
"tdnn_dim": [
|
| 90 |
+
512,
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
1500
|
| 95 |
+
],
|
| 96 |
+
"tdnn_kernel": [
|
| 97 |
+
5,
|
| 98 |
+
3,
|
| 99 |
+
3,
|
| 100 |
+
1,
|
| 101 |
+
1
|
| 102 |
+
],
|
| 103 |
+
"torch_dtype": "float32",
|
| 104 |
+
"transformers_version": "4.29.2",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 44,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cd629b244ea9bd81dbab89b7a3dbbfff1cf52ee6e8f5de5d5ea8055402862ee
|
| 3 |
+
size 2490517301
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0,
|
| 7 |
+
"processor_class": "Wav2Vec2Processor",
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000
|
| 10 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bbab804e00abceaf5f77c8b91673929aab2d54752d6759906e65c5c8ab4fcad
|
| 3 |
+
size 1262079089
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99d283846fc8d61b2b8a3296e54a50406c2ebaea763f9e97993614fd0f7acdcc
|
| 3 |
+
size 14567
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33dc576dbca460909cd313e403b166b8f027764ab0301b0689155b7d90783bf1
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:edfb59c94cc0f2c5014ae5820494437671f2c32d0fd2dab3bece1a23c7e0e7d7
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"pad_token": "<pad>",
|
| 5 |
+
"unk_token": "<unk>"
|
| 6 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"do_lower_case": false,
|
| 5 |
+
"eos_token": "</s>",
|
| 6 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"pad_token": "<pad>",
|
| 8 |
+
"processor_class": "Wav2Vec2Processor",
|
| 9 |
+
"replace_word_delimiter_char": " ",
|
| 10 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 11 |
+
"tokenizer_file": null,
|
| 12 |
+
"unk_token": "<unk>",
|
| 13 |
+
"word_delimiter_token": "|"
|
| 14 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 11.39,
|
| 3 |
+
"train_loss": 218.4774083984375,
|
| 4 |
+
"train_runtime": 9202.4528,
|
| 5 |
+
"train_samples": 7024,
|
| 6 |
+
"train_samples_per_second": 8.693,
|
| 7 |
+
"train_steps_per_second": 1.087
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,666 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 11.389521640091116,
|
| 5 |
+
"global_step": 10000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.11,
|
| 12 |
+
"learning_rate": 5e-05,
|
| 13 |
+
"loss": 467.0219,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.23,
|
| 18 |
+
"learning_rate": 5e-05,
|
| 19 |
+
"loss": 446.5955,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.34,
|
| 24 |
+
"learning_rate": 5e-05,
|
| 25 |
+
"loss": 393.423,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.46,
|
| 30 |
+
"learning_rate": 5e-05,
|
| 31 |
+
"loss": 360.4873,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.57,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 369.4461,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.68,
|
| 42 |
+
"learning_rate": 5e-05,
|
| 43 |
+
"loss": 364.8889,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.8,
|
| 48 |
+
"learning_rate": 5e-05,
|
| 49 |
+
"loss": 362.8324,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.91,
|
| 54 |
+
"learning_rate": 5e-05,
|
| 55 |
+
"loss": 345.4677,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.03,
|
| 60 |
+
"learning_rate": 5e-05,
|
| 61 |
+
"loss": 377.1422,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.14,
|
| 66 |
+
"learning_rate": 5e-05,
|
| 67 |
+
"loss": 319.0323,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.25,
|
| 72 |
+
"learning_rate": 5e-05,
|
| 73 |
+
"loss": 298.1518,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.37,
|
| 78 |
+
"learning_rate": 5e-05,
|
| 79 |
+
"loss": 335.3308,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.48,
|
| 84 |
+
"learning_rate": 5e-05,
|
| 85 |
+
"loss": 290.9471,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.59,
|
| 90 |
+
"learning_rate": 5e-05,
|
| 91 |
+
"loss": 310.6187,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.71,
|
| 96 |
+
"learning_rate": 5e-05,
|
| 97 |
+
"loss": 333.6563,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.82,
|
| 102 |
+
"learning_rate": 5e-05,
|
| 103 |
+
"loss": 287.3142,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.94,
|
| 108 |
+
"learning_rate": 5e-05,
|
| 109 |
+
"loss": 304.1946,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.05,
|
| 114 |
+
"learning_rate": 5e-05,
|
| 115 |
+
"loss": 295.4456,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.16,
|
| 120 |
+
"learning_rate": 5e-05,
|
| 121 |
+
"loss": 288.6632,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 2.28,
|
| 126 |
+
"learning_rate": 5e-05,
|
| 127 |
+
"loss": 268.8452,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.28,
|
| 132 |
+
"eval_cer": 0.1040043410507569,
|
| 133 |
+
"eval_loss": 262.4808654785156,
|
| 134 |
+
"eval_runtime": 146.2892,
|
| 135 |
+
"eval_samples_per_second": 12.783,
|
| 136 |
+
"eval_steps_per_second": 1.6,
|
| 137 |
+
"eval_wer": 0.22893966022261278,
|
| 138 |
+
"step": 2000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.39,
|
| 142 |
+
"learning_rate": 5e-05,
|
| 143 |
+
"loss": 279.8674,
|
| 144 |
+
"step": 2100
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.51,
|
| 148 |
+
"learning_rate": 5e-05,
|
| 149 |
+
"loss": 288.3673,
|
| 150 |
+
"step": 2200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 2.62,
|
| 154 |
+
"learning_rate": 5e-05,
|
| 155 |
+
"loss": 266.2094,
|
| 156 |
+
"step": 2300
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.73,
|
| 160 |
+
"learning_rate": 5e-05,
|
| 161 |
+
"loss": 266.3993,
|
| 162 |
+
"step": 2400
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 2.85,
|
| 166 |
+
"learning_rate": 5e-05,
|
| 167 |
+
"loss": 265.0983,
|
| 168 |
+
"step": 2500
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 2.96,
|
| 172 |
+
"learning_rate": 5e-05,
|
| 173 |
+
"loss": 280.5026,
|
| 174 |
+
"step": 2600
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 3.08,
|
| 178 |
+
"learning_rate": 5e-05,
|
| 179 |
+
"loss": 253.5466,
|
| 180 |
+
"step": 2700
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 3.19,
|
| 184 |
+
"learning_rate": 5e-05,
|
| 185 |
+
"loss": 259.7672,
|
| 186 |
+
"step": 2800
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.3,
|
| 190 |
+
"learning_rate": 5e-05,
|
| 191 |
+
"loss": 241.1983,
|
| 192 |
+
"step": 2900
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 3.42,
|
| 196 |
+
"learning_rate": 5e-05,
|
| 197 |
+
"loss": 234.3184,
|
| 198 |
+
"step": 3000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.53,
|
| 202 |
+
"learning_rate": 5e-05,
|
| 203 |
+
"loss": 241.2133,
|
| 204 |
+
"step": 3100
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.64,
|
| 208 |
+
"learning_rate": 5e-05,
|
| 209 |
+
"loss": 264.1351,
|
| 210 |
+
"step": 3200
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 3.76,
|
| 214 |
+
"learning_rate": 5e-05,
|
| 215 |
+
"loss": 247.2135,
|
| 216 |
+
"step": 3300
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 3.87,
|
| 220 |
+
"learning_rate": 5e-05,
|
| 221 |
+
"loss": 227.8616,
|
| 222 |
+
"step": 3400
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 3.99,
|
| 226 |
+
"learning_rate": 5e-05,
|
| 227 |
+
"loss": 236.7091,
|
| 228 |
+
"step": 3500
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.1,
|
| 232 |
+
"learning_rate": 5e-05,
|
| 233 |
+
"loss": 234.9673,
|
| 234 |
+
"step": 3600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 5e-05,
|
| 239 |
+
"loss": 211.579,
|
| 240 |
+
"step": 3700
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.33,
|
| 244 |
+
"learning_rate": 5e-05,
|
| 245 |
+
"loss": 226.7139,
|
| 246 |
+
"step": 3800
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.44,
|
| 250 |
+
"learning_rate": 5e-05,
|
| 251 |
+
"loss": 239.6923,
|
| 252 |
+
"step": 3900
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.56,
|
| 256 |
+
"learning_rate": 5e-05,
|
| 257 |
+
"loss": 212.4134,
|
| 258 |
+
"step": 4000
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.56,
|
| 262 |
+
"eval_cer": 0.09499526936776491,
|
| 263 |
+
"eval_loss": 273.2842712402344,
|
| 264 |
+
"eval_runtime": 145.12,
|
| 265 |
+
"eval_samples_per_second": 12.886,
|
| 266 |
+
"eval_steps_per_second": 1.612,
|
| 267 |
+
"eval_wer": 0.20917789494239405,
|
| 268 |
+
"step": 4000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.67,
|
| 272 |
+
"learning_rate": 5e-05,
|
| 273 |
+
"loss": 222.3978,
|
| 274 |
+
"step": 4100
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 4.78,
|
| 278 |
+
"learning_rate": 5e-05,
|
| 279 |
+
"loss": 220.115,
|
| 280 |
+
"step": 4200
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 4.9,
|
| 284 |
+
"learning_rate": 5e-05,
|
| 285 |
+
"loss": 214.0393,
|
| 286 |
+
"step": 4300
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.01,
|
| 290 |
+
"learning_rate": 5e-05,
|
| 291 |
+
"loss": 217.6124,
|
| 292 |
+
"step": 4400
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.13,
|
| 296 |
+
"learning_rate": 5e-05,
|
| 297 |
+
"loss": 207.3773,
|
| 298 |
+
"step": 4500
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.24,
|
| 302 |
+
"learning_rate": 5e-05,
|
| 303 |
+
"loss": 210.2402,
|
| 304 |
+
"step": 4600
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.35,
|
| 308 |
+
"learning_rate": 5e-05,
|
| 309 |
+
"loss": 183.2794,
|
| 310 |
+
"step": 4700
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.47,
|
| 314 |
+
"learning_rate": 5e-05,
|
| 315 |
+
"loss": 212.013,
|
| 316 |
+
"step": 4800
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.58,
|
| 320 |
+
"learning_rate": 5e-05,
|
| 321 |
+
"loss": 201.9256,
|
| 322 |
+
"step": 4900
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.69,
|
| 326 |
+
"learning_rate": 5e-05,
|
| 327 |
+
"loss": 204.919,
|
| 328 |
+
"step": 5000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 5.81,
|
| 332 |
+
"learning_rate": 5e-05,
|
| 333 |
+
"loss": 200.8947,
|
| 334 |
+
"step": 5100
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 5.92,
|
| 338 |
+
"learning_rate": 5e-05,
|
| 339 |
+
"loss": 205.1183,
|
| 340 |
+
"step": 5200
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 6.04,
|
| 344 |
+
"learning_rate": 5e-05,
|
| 345 |
+
"loss": 183.53,
|
| 346 |
+
"step": 5300
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 6.15,
|
| 350 |
+
"learning_rate": 5e-05,
|
| 351 |
+
"loss": 176.2721,
|
| 352 |
+
"step": 5400
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.26,
|
| 356 |
+
"learning_rate": 5e-05,
|
| 357 |
+
"loss": 182.444,
|
| 358 |
+
"step": 5500
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 6.38,
|
| 362 |
+
"learning_rate": 5e-05,
|
| 363 |
+
"loss": 179.9177,
|
| 364 |
+
"step": 5600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 6.49,
|
| 368 |
+
"learning_rate": 5e-05,
|
| 369 |
+
"loss": 184.7364,
|
| 370 |
+
"step": 5700
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.61,
|
| 374 |
+
"learning_rate": 5e-05,
|
| 375 |
+
"loss": 170.2941,
|
| 376 |
+
"step": 5800
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 6.72,
|
| 380 |
+
"learning_rate": 5e-05,
|
| 381 |
+
"loss": 184.6765,
|
| 382 |
+
"step": 5900
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 6.83,
|
| 386 |
+
"learning_rate": 5e-05,
|
| 387 |
+
"loss": 200.3965,
|
| 388 |
+
"step": 6000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 6.83,
|
| 392 |
+
"eval_cer": 0.09548224621549421,
|
| 393 |
+
"eval_loss": 279.6430358886719,
|
| 394 |
+
"eval_runtime": 147.4008,
|
| 395 |
+
"eval_samples_per_second": 12.687,
|
| 396 |
+
"eval_steps_per_second": 1.588,
|
| 397 |
+
"eval_wer": 0.2134348760007811,
|
| 398 |
+
"step": 6000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.95,
|
| 402 |
+
"learning_rate": 5e-05,
|
| 403 |
+
"loss": 203.8473,
|
| 404 |
+
"step": 6100
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 7.06,
|
| 408 |
+
"learning_rate": 5e-05,
|
| 409 |
+
"loss": 180.4022,
|
| 410 |
+
"step": 6200
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 7.18,
|
| 414 |
+
"learning_rate": 5e-05,
|
| 415 |
+
"loss": 167.2582,
|
| 416 |
+
"step": 6300
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.29,
|
| 420 |
+
"learning_rate": 5e-05,
|
| 421 |
+
"loss": 171.7429,
|
| 422 |
+
"step": 6400
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.4,
|
| 426 |
+
"learning_rate": 5e-05,
|
| 427 |
+
"loss": 161.9806,
|
| 428 |
+
"step": 6500
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 7.52,
|
| 432 |
+
"learning_rate": 5e-05,
|
| 433 |
+
"loss": 181.7546,
|
| 434 |
+
"step": 6600
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 7.63,
|
| 438 |
+
"learning_rate": 5e-05,
|
| 439 |
+
"loss": 166.6461,
|
| 440 |
+
"step": 6700
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 7.74,
|
| 444 |
+
"learning_rate": 5e-05,
|
| 445 |
+
"loss": 156.8281,
|
| 446 |
+
"step": 6800
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 7.86,
|
| 450 |
+
"learning_rate": 5e-05,
|
| 451 |
+
"loss": 162.8639,
|
| 452 |
+
"step": 6900
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 7.97,
|
| 456 |
+
"learning_rate": 5e-05,
|
| 457 |
+
"loss": 177.6029,
|
| 458 |
+
"step": 7000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 8.09,
|
| 462 |
+
"learning_rate": 5e-05,
|
| 463 |
+
"loss": 162.6876,
|
| 464 |
+
"step": 7100
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 8.2,
|
| 468 |
+
"learning_rate": 5e-05,
|
| 469 |
+
"loss": 154.6953,
|
| 470 |
+
"step": 7200
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 8.31,
|
| 474 |
+
"learning_rate": 5e-05,
|
| 475 |
+
"loss": 157.7532,
|
| 476 |
+
"step": 7300
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 8.43,
|
| 480 |
+
"learning_rate": 5e-05,
|
| 481 |
+
"loss": 155.4144,
|
| 482 |
+
"step": 7400
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 8.54,
|
| 486 |
+
"learning_rate": 5e-05,
|
| 487 |
+
"loss": 160.9859,
|
| 488 |
+
"step": 7500
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 8.66,
|
| 492 |
+
"learning_rate": 5e-05,
|
| 493 |
+
"loss": 157.5087,
|
| 494 |
+
"step": 7600
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 8.77,
|
| 498 |
+
"learning_rate": 5e-05,
|
| 499 |
+
"loss": 158.3729,
|
| 500 |
+
"step": 7700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 8.88,
|
| 504 |
+
"learning_rate": 5e-05,
|
| 505 |
+
"loss": 161.7107,
|
| 506 |
+
"step": 7800
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.0,
|
| 510 |
+
"learning_rate": 5e-05,
|
| 511 |
+
"loss": 158.3218,
|
| 512 |
+
"step": 7900
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.11,
|
| 516 |
+
"learning_rate": 5e-05,
|
| 517 |
+
"loss": 130.7524,
|
| 518 |
+
"step": 8000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.11,
|
| 522 |
+
"eval_cer": 0.09586487088156723,
|
| 523 |
+
"eval_loss": 320.47552490234375,
|
| 524 |
+
"eval_runtime": 144.7814,
|
| 525 |
+
"eval_samples_per_second": 12.916,
|
| 526 |
+
"eval_steps_per_second": 1.616,
|
| 527 |
+
"eval_wer": 0.20667838312829526,
|
| 528 |
+
"step": 8000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 9.23,
|
| 532 |
+
"learning_rate": 5e-05,
|
| 533 |
+
"loss": 142.4579,
|
| 534 |
+
"step": 8100
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 9.34,
|
| 538 |
+
"learning_rate": 5e-05,
|
| 539 |
+
"loss": 147.8841,
|
| 540 |
+
"step": 8200
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 9.45,
|
| 544 |
+
"learning_rate": 5e-05,
|
| 545 |
+
"loss": 158.0805,
|
| 546 |
+
"step": 8300
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 9.57,
|
| 550 |
+
"learning_rate": 5e-05,
|
| 551 |
+
"loss": 145.2056,
|
| 552 |
+
"step": 8400
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 9.68,
|
| 556 |
+
"learning_rate": 5e-05,
|
| 557 |
+
"loss": 144.9573,
|
| 558 |
+
"step": 8500
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 9.79,
|
| 562 |
+
"learning_rate": 5e-05,
|
| 563 |
+
"loss": 148.1379,
|
| 564 |
+
"step": 8600
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.91,
|
| 568 |
+
"learning_rate": 5e-05,
|
| 569 |
+
"loss": 148.8498,
|
| 570 |
+
"step": 8700
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.02,
|
| 574 |
+
"learning_rate": 5e-05,
|
| 575 |
+
"loss": 160.1041,
|
| 576 |
+
"step": 8800
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.14,
|
| 580 |
+
"learning_rate": 5e-05,
|
| 581 |
+
"loss": 141.2527,
|
| 582 |
+
"step": 8900
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.25,
|
| 586 |
+
"learning_rate": 5e-05,
|
| 587 |
+
"loss": 127.7468,
|
| 588 |
+
"step": 9000
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.36,
|
| 592 |
+
"learning_rate": 5e-05,
|
| 593 |
+
"loss": 136.7523,
|
| 594 |
+
"step": 9100
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.48,
|
| 598 |
+
"learning_rate": 5e-05,
|
| 599 |
+
"loss": 134.6625,
|
| 600 |
+
"step": 9200
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.59,
|
| 604 |
+
"learning_rate": 5e-05,
|
| 605 |
+
"loss": 138.4917,
|
| 606 |
+
"step": 9300
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 10.71,
|
| 610 |
+
"learning_rate": 5e-05,
|
| 611 |
+
"loss": 135.3308,
|
| 612 |
+
"step": 9400
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 10.82,
|
| 616 |
+
"learning_rate": 5e-05,
|
| 617 |
+
"loss": 143.3156,
|
| 618 |
+
"step": 9500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 10.93,
|
| 622 |
+
"learning_rate": 5e-05,
|
| 623 |
+
"loss": 130.6054,
|
| 624 |
+
"step": 9600
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 11.05,
|
| 628 |
+
"learning_rate": 5e-05,
|
| 629 |
+
"loss": 138.9151,
|
| 630 |
+
"step": 9700
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 11.16,
|
| 634 |
+
"learning_rate": 5e-05,
|
| 635 |
+
"loss": 125.203,
|
| 636 |
+
"step": 9800
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 11.28,
|
| 640 |
+
"learning_rate": 5e-05,
|
| 641 |
+
"loss": 131.7745,
|
| 642 |
+
"step": 9900
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 11.39,
|
| 646 |
+
"learning_rate": 5e-05,
|
| 647 |
+
"loss": 119.4062,
|
| 648 |
+
"step": 10000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 11.39,
|
| 652 |
+
"eval_cer": 0.09683186776491541,
|
| 653 |
+
"eval_loss": 327.7448425292969,
|
| 654 |
+
"eval_runtime": 145.6399,
|
| 655 |
+
"eval_samples_per_second": 12.84,
|
| 656 |
+
"eval_steps_per_second": 1.607,
|
| 657 |
+
"eval_wer": 0.20937316930287053,
|
| 658 |
+
"step": 10000
|
| 659 |
+
}
|
| 660 |
+
],
|
| 661 |
+
"max_steps": 10000,
|
| 662 |
+
"num_train_epochs": 12,
|
| 663 |
+
"total_flos": 1.2933525176475093e+19,
|
| 664 |
+
"trial_name": null,
|
| 665 |
+
"trial_params": null
|
| 666 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8df666f0a9af7a95cf6bd16018c9b643dc7cb2a0395205e4cfa917f8781d6759
|
| 3 |
+
size 3887
|
vocab.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"-": 35,
|
| 3 |
+
"</s>": 2,
|
| 4 |
+
"<pad>": 0,
|
| 5 |
+
"<s>": 1,
|
| 6 |
+
"<unk>": 3,
|
| 7 |
+
"a": 5,
|
| 8 |
+
"b": 24,
|
| 9 |
+
"c": 16,
|
| 10 |
+
"d": 13,
|
| 11 |
+
"e": 6,
|
| 12 |
+
"f": 22,
|
| 13 |
+
"g": 21,
|
| 14 |
+
"h": 23,
|
| 15 |
+
"i": 10,
|
| 16 |
+
"j": 30,
|
| 17 |
+
"k": 41,
|
| 18 |
+
"l": 17,
|
| 19 |
+
"m": 12,
|
| 20 |
+
"n": 11,
|
| 21 |
+
"o": 7,
|
| 22 |
+
"p": 18,
|
| 23 |
+
"q": 20,
|
| 24 |
+
"r": 9,
|
| 25 |
+
"s": 8,
|
| 26 |
+
"t": 14,
|
| 27 |
+
"u": 15,
|
| 28 |
+
"v": 19,
|
| 29 |
+
"w": 43,
|
| 30 |
+
"x": 33,
|
| 31 |
+
"y": 42,
|
| 32 |
+
"z": 29,
|
| 33 |
+
"|": 4,
|
| 34 |
+
"à": 38,
|
| 35 |
+
"á": 28,
|
| 36 |
+
"â": 40,
|
| 37 |
+
"ã": 25,
|
| 38 |
+
"ç": 27,
|
| 39 |
+
"é": 26,
|
| 40 |
+
"ê": 32,
|
| 41 |
+
"í": 31,
|
| 42 |
+
"ó": 34,
|
| 43 |
+
"ô": 39,
|
| 44 |
+
"õ": 36,
|
| 45 |
+
"ú": 37
|
| 46 |
+
}
|