Commit
•
aff1fdf
1
Parent(s):
7ebe59a
Upload 11 files
Browse files- config.json +108 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +9 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +871 -0
- training_args.bin +3 -0
- training_args/byteorder +1 -0
- training_args/data.pkl +3 -0
- training_args/version +1 -0
config.json
ADDED
|
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/wav2vec2-xls-r-300m",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_attn_dim": null,
|
| 5 |
+
"adapter_kernel_size": 3,
|
| 6 |
+
"adapter_stride": 2,
|
| 7 |
+
"add_adapter": false,
|
| 8 |
+
"apply_spec_augment": true,
|
| 9 |
+
"architectures": [
|
| 10 |
+
"Wav2Vec2ForCTC"
|
| 11 |
+
],
|
| 12 |
+
"attention_dropout": 0.1,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"classifier_proj_size": 256,
|
| 15 |
+
"codevector_dim": 768,
|
| 16 |
+
"contrastive_logits_temperature": 0.1,
|
| 17 |
+
"conv_bias": true,
|
| 18 |
+
"conv_dim": [
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512,
|
| 25 |
+
512
|
| 26 |
+
],
|
| 27 |
+
"conv_kernel": [
|
| 28 |
+
10,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
3,
|
| 33 |
+
2,
|
| 34 |
+
2
|
| 35 |
+
],
|
| 36 |
+
"conv_stride": [
|
| 37 |
+
5,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2,
|
| 43 |
+
2
|
| 44 |
+
],
|
| 45 |
+
"ctc_loss_reduction": "mean",
|
| 46 |
+
"ctc_zero_infinity": false,
|
| 47 |
+
"diversity_loss_weight": 0.1,
|
| 48 |
+
"do_stable_layer_norm": true,
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_dropout": 0.0,
|
| 52 |
+
"feat_extract_norm": "layer",
|
| 53 |
+
"feat_proj_dropout": 0.1,
|
| 54 |
+
"feat_quantizer_dropout": 0.0,
|
| 55 |
+
"final_dropout": 0.0,
|
| 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": 0.1,
|
| 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": 0.075,
|
| 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": 47,
|
| 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": "4.38.2",
|
| 105 |
+
"use_weighted_layer_sum": false,
|
| 106 |
+
"vocab_size": 50,
|
| 107 |
+
"xvector_output_dim": 512
|
| 108 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b3bcf8ad360c7621f1d04757a1122d14e00786f44dee9885e6f7f88efdac3be
|
| 3 |
+
size 1262012480
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6dcbb67b228f3a9d540c30c4ec4908e52066fd5e05932d65dea1049e4394beb4
|
| 3 |
+
size 2490569782
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"return_attention_mask": false,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8fcbb4a702f96dfc82b6c0b385989dc162a6a585a73d12a769238bccc10368dd
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a9bbbb6333d7f5a82d7204046ded86a435d7caf39a8db15924f97da626cd069
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,871 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.32769803615398474,
|
| 3 |
+
"best_model_checkpoint": "./wav2vec2-base-hy/checkpoint-36331",
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 38650,
|
| 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": 1.0217593908309937,
|
| 14 |
+
"learning_rate": 9.928440366972478e-05,
|
| 15 |
+
"loss": 4.8189,
|
| 16 |
+
"step": 773
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_cer": 0.9914830021613891,
|
| 21 |
+
"eval_loss": 3.259536027908325,
|
| 22 |
+
"eval_runtime": 538.6244,
|
| 23 |
+
"eval_samples_per_second": 7.948,
|
| 24 |
+
"eval_steps_per_second": 0.995,
|
| 25 |
+
"eval_wer": 1.0,
|
| 26 |
+
"step": 773
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.0,
|
| 30 |
+
"grad_norm": 6.301283359527588,
|
| 31 |
+
"learning_rate": 9.725819134993447e-05,
|
| 32 |
+
"loss": 1.7898,
|
| 33 |
+
"step": 1546
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 2.0,
|
| 37 |
+
"eval_cer": 0.1494773295194324,
|
| 38 |
+
"eval_loss": 0.5445137023925781,
|
| 39 |
+
"eval_runtime": 538.4983,
|
| 40 |
+
"eval_samples_per_second": 7.95,
|
| 41 |
+
"eval_steps_per_second": 0.995,
|
| 42 |
+
"eval_wer": 0.7093567953876644,
|
| 43 |
+
"step": 1546
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 3.0,
|
| 47 |
+
"grad_norm": 3.109081983566284,
|
| 48 |
+
"learning_rate": 9.523197903014417e-05,
|
| 49 |
+
"loss": 0.6005,
|
| 50 |
+
"step": 2319
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.0,
|
| 54 |
+
"eval_cer": 0.11405947435660509,
|
| 55 |
+
"eval_loss": 0.3848719298839569,
|
| 56 |
+
"eval_runtime": 538.0999,
|
| 57 |
+
"eval_samples_per_second": 7.956,
|
| 58 |
+
"eval_steps_per_second": 0.996,
|
| 59 |
+
"eval_wer": 0.56861449762777,
|
| 60 |
+
"step": 2319
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 4.0,
|
| 64 |
+
"grad_norm": 10.241255760192871,
|
| 65 |
+
"learning_rate": 9.320576671035387e-05,
|
| 66 |
+
"loss": 0.4408,
|
| 67 |
+
"step": 3092
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 4.0,
|
| 71 |
+
"eval_cer": 0.09999839301927574,
|
| 72 |
+
"eval_loss": 0.33175399899482727,
|
| 73 |
+
"eval_runtime": 538.012,
|
| 74 |
+
"eval_samples_per_second": 7.957,
|
| 75 |
+
"eval_steps_per_second": 0.996,
|
| 76 |
+
"eval_wer": 0.5128821091826317,
|
| 77 |
+
"step": 3092
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.0,
|
| 81 |
+
"grad_norm": 3.3970553874969482,
|
| 82 |
+
"learning_rate": 9.117955439056357e-05,
|
| 83 |
+
"loss": 0.3726,
|
| 84 |
+
"step": 3865
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 5.0,
|
| 88 |
+
"eval_cer": 0.09280313682637377,
|
| 89 |
+
"eval_loss": 0.31485071778297424,
|
| 90 |
+
"eval_runtime": 538.2419,
|
| 91 |
+
"eval_samples_per_second": 7.954,
|
| 92 |
+
"eval_steps_per_second": 0.996,
|
| 93 |
+
"eval_wer": 0.4864572698336436,
|
| 94 |
+
"step": 3865
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 6.0,
|
| 98 |
+
"grad_norm": 3.6496012210845947,
|
| 99 |
+
"learning_rate": 8.915334207077327e-05,
|
| 100 |
+
"loss": 0.324,
|
| 101 |
+
"step": 4638
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 6.0,
|
| 105 |
+
"eval_cer": 0.08554360140450115,
|
| 106 |
+
"eval_loss": 0.2828996181488037,
|
| 107 |
+
"eval_runtime": 538.7731,
|
| 108 |
+
"eval_samples_per_second": 7.946,
|
| 109 |
+
"eval_steps_per_second": 0.995,
|
| 110 |
+
"eval_wer": 0.4492823253858627,
|
| 111 |
+
"step": 4638
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 7.0,
|
| 115 |
+
"grad_norm": 1.054186463356018,
|
| 116 |
+
"learning_rate": 8.712712975098296e-05,
|
| 117 |
+
"loss": 0.2885,
|
| 118 |
+
"step": 5411
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 7.0,
|
| 122 |
+
"eval_cer": 0.08390448106574962,
|
| 123 |
+
"eval_loss": 0.2892753779888153,
|
| 124 |
+
"eval_runtime": 538.8258,
|
| 125 |
+
"eval_samples_per_second": 7.945,
|
| 126 |
+
"eval_steps_per_second": 0.995,
|
| 127 |
+
"eval_wer": 0.4395832082157228,
|
| 128 |
+
"step": 5411
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"grad_norm": 4.084922790527344,
|
| 133 |
+
"learning_rate": 8.510091743119266e-05,
|
| 134 |
+
"loss": 0.256,
|
| 135 |
+
"step": 6184
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 8.0,
|
| 139 |
+
"eval_cer": 0.0834987184328724,
|
| 140 |
+
"eval_loss": 0.2856770157814026,
|
| 141 |
+
"eval_runtime": 538.8477,
|
| 142 |
+
"eval_samples_per_second": 7.945,
|
| 143 |
+
"eval_steps_per_second": 0.995,
|
| 144 |
+
"eval_wer": 0.4325566032070146,
|
| 145 |
+
"step": 6184
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 9.0,
|
| 149 |
+
"grad_norm": 2.03792405128479,
|
| 150 |
+
"learning_rate": 8.307470511140236e-05,
|
| 151 |
+
"loss": 0.2425,
|
| 152 |
+
"step": 6957
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 9.0,
|
| 156 |
+
"eval_cer": 0.08157034156375294,
|
| 157 |
+
"eval_loss": 0.2794096767902374,
|
| 158 |
+
"eval_runtime": 538.7335,
|
| 159 |
+
"eval_samples_per_second": 7.946,
|
| 160 |
+
"eval_steps_per_second": 0.995,
|
| 161 |
+
"eval_wer": 0.4256501111044382,
|
| 162 |
+
"step": 6957
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 10.0,
|
| 166 |
+
"grad_norm": 4.848245143890381,
|
| 167 |
+
"learning_rate": 8.104849279161207e-05,
|
| 168 |
+
"loss": 0.2239,
|
| 169 |
+
"step": 7730
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.0,
|
| 173 |
+
"eval_cer": 0.07841262444056983,
|
| 174 |
+
"eval_loss": 0.2803117334842682,
|
| 175 |
+
"eval_runtime": 539.9556,
|
| 176 |
+
"eval_samples_per_second": 7.928,
|
| 177 |
+
"eval_steps_per_second": 0.993,
|
| 178 |
+
"eval_wer": 0.4096750945889136,
|
| 179 |
+
"step": 7730
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 11.0,
|
| 183 |
+
"grad_norm": 5.475008964538574,
|
| 184 |
+
"learning_rate": 7.902228047182177e-05,
|
| 185 |
+
"loss": 0.202,
|
| 186 |
+
"step": 8503
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 11.0,
|
| 190 |
+
"eval_cer": 0.07942100484504688,
|
| 191 |
+
"eval_loss": 0.2747463583946228,
|
| 192 |
+
"eval_runtime": 540.606,
|
| 193 |
+
"eval_samples_per_second": 7.919,
|
| 194 |
+
"eval_steps_per_second": 0.991,
|
| 195 |
+
"eval_wer": 0.4147198366464477,
|
| 196 |
+
"step": 8503
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 12.0,
|
| 200 |
+
"grad_norm": 2.023427724838257,
|
| 201 |
+
"learning_rate": 7.699606815203146e-05,
|
| 202 |
+
"loss": 0.1912,
|
| 203 |
+
"step": 9276
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 12.0,
|
| 207 |
+
"eval_cer": 0.07573700153466659,
|
| 208 |
+
"eval_loss": 0.2822825014591217,
|
| 209 |
+
"eval_runtime": 538.7132,
|
| 210 |
+
"eval_samples_per_second": 7.947,
|
| 211 |
+
"eval_steps_per_second": 0.995,
|
| 212 |
+
"eval_wer": 0.3974536063900066,
|
| 213 |
+
"step": 9276
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 13.0,
|
| 217 |
+
"grad_norm": 4.661593437194824,
|
| 218 |
+
"learning_rate": 7.496985583224116e-05,
|
| 219 |
+
"loss": 0.1755,
|
| 220 |
+
"step": 10049
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 13.0,
|
| 224 |
+
"eval_cer": 0.07552005913689065,
|
| 225 |
+
"eval_loss": 0.2788126468658447,
|
| 226 |
+
"eval_runtime": 539.2919,
|
| 227 |
+
"eval_samples_per_second": 7.938,
|
| 228 |
+
"eval_steps_per_second": 0.994,
|
| 229 |
+
"eval_wer": 0.39339979580805956,
|
| 230 |
+
"step": 10049
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 14.0,
|
| 234 |
+
"grad_norm": 2.8817713260650635,
|
| 235 |
+
"learning_rate": 7.294364351245086e-05,
|
| 236 |
+
"loss": 0.1669,
|
| 237 |
+
"step": 10822
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 14.0,
|
| 241 |
+
"eval_cer": 0.0756767397575066,
|
| 242 |
+
"eval_loss": 0.3079923391342163,
|
| 243 |
+
"eval_runtime": 538.3008,
|
| 244 |
+
"eval_samples_per_second": 7.953,
|
| 245 |
+
"eval_steps_per_second": 0.996,
|
| 246 |
+
"eval_wer": 0.39333973935499367,
|
| 247 |
+
"step": 10822
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 15.0,
|
| 251 |
+
"grad_norm": 4.330118656158447,
|
| 252 |
+
"learning_rate": 7.091743119266056e-05,
|
| 253 |
+
"loss": 0.1622,
|
| 254 |
+
"step": 11595
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 15.0,
|
| 258 |
+
"eval_cer": 0.07294085507444338,
|
| 259 |
+
"eval_loss": 0.2876891791820526,
|
| 260 |
+
"eval_runtime": 539.8362,
|
| 261 |
+
"eval_samples_per_second": 7.93,
|
| 262 |
+
"eval_steps_per_second": 0.993,
|
| 263 |
+
"eval_wer": 0.3854122875502973,
|
| 264 |
+
"step": 11595
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 16.0,
|
| 268 |
+
"grad_norm": 0.5675534009933472,
|
| 269 |
+
"learning_rate": 6.889121887287026e-05,
|
| 270 |
+
"loss": 0.1485,
|
| 271 |
+
"step": 12368
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 16.0,
|
| 275 |
+
"eval_cer": 0.07495359843158682,
|
| 276 |
+
"eval_loss": 0.30401086807250977,
|
| 277 |
+
"eval_runtime": 540.0181,
|
| 278 |
+
"eval_samples_per_second": 7.928,
|
| 279 |
+
"eval_steps_per_second": 0.993,
|
| 280 |
+
"eval_wer": 0.38766440454026785,
|
| 281 |
+
"step": 12368
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.0,
|
| 285 |
+
"grad_norm": 1.4718132019042969,
|
| 286 |
+
"learning_rate": 6.686500655307996e-05,
|
| 287 |
+
"loss": 0.1426,
|
| 288 |
+
"step": 13141
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 17.0,
|
| 292 |
+
"eval_cer": 0.07407779393686173,
|
| 293 |
+
"eval_loss": 0.2863430678844452,
|
| 294 |
+
"eval_runtime": 539.387,
|
| 295 |
+
"eval_samples_per_second": 7.937,
|
| 296 |
+
"eval_steps_per_second": 0.994,
|
| 297 |
+
"eval_wer": 0.38862530778932197,
|
| 298 |
+
"step": 13141
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 18.0,
|
| 302 |
+
"grad_norm": 4.17996883392334,
|
| 303 |
+
"learning_rate": 6.483879423328964e-05,
|
| 304 |
+
"loss": 0.137,
|
| 305 |
+
"step": 13914
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 18.0,
|
| 309 |
+
"eval_cer": 0.07082365797022265,
|
| 310 |
+
"eval_loss": 0.29217201471328735,
|
| 311 |
+
"eval_runtime": 540.1379,
|
| 312 |
+
"eval_samples_per_second": 7.926,
|
| 313 |
+
"eval_steps_per_second": 0.992,
|
| 314 |
+
"eval_wer": 0.37601345264548675,
|
| 315 |
+
"step": 13914
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 19.0,
|
| 319 |
+
"grad_norm": 1.8616777658462524,
|
| 320 |
+
"learning_rate": 6.281258191349934e-05,
|
| 321 |
+
"loss": 0.1324,
|
| 322 |
+
"step": 14687
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 19.0,
|
| 326 |
+
"eval_cer": 0.07016077842146283,
|
| 327 |
+
"eval_loss": 0.28696727752685547,
|
| 328 |
+
"eval_runtime": 539.124,
|
| 329 |
+
"eval_samples_per_second": 7.941,
|
| 330 |
+
"eval_steps_per_second": 0.994,
|
| 331 |
+
"eval_wer": 0.3717794727043421,
|
| 332 |
+
"step": 14687
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 20.0,
|
| 336 |
+
"grad_norm": 2.518717050552368,
|
| 337 |
+
"learning_rate": 6.078636959370905e-05,
|
| 338 |
+
"loss": 0.1247,
|
| 339 |
+
"step": 15460
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 20.0,
|
| 343 |
+
"eval_cer": 0.06962645733064432,
|
| 344 |
+
"eval_loss": 0.2821752727031708,
|
| 345 |
+
"eval_runtime": 540.8162,
|
| 346 |
+
"eval_samples_per_second": 7.916,
|
| 347 |
+
"eval_steps_per_second": 0.991,
|
| 348 |
+
"eval_wer": 0.3689868476367786,
|
| 349 |
+
"step": 15460
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 21.0,
|
| 353 |
+
"grad_norm": 3.7167491912841797,
|
| 354 |
+
"learning_rate": 5.876015727391875e-05,
|
| 355 |
+
"loss": 0.1226,
|
| 356 |
+
"step": 16233
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 21.0,
|
| 360 |
+
"eval_cer": 0.06858191985987128,
|
| 361 |
+
"eval_loss": 0.28035247325897217,
|
| 362 |
+
"eval_runtime": 540.4383,
|
| 363 |
+
"eval_samples_per_second": 7.921,
|
| 364 |
+
"eval_steps_per_second": 0.992,
|
| 365 |
+
"eval_wer": 0.3664945048345445,
|
| 366 |
+
"step": 16233
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 22.0,
|
| 370 |
+
"grad_norm": 1.6708425283432007,
|
| 371 |
+
"learning_rate": 5.673394495412845e-05,
|
| 372 |
+
"loss": 0.1162,
|
| 373 |
+
"step": 17006
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 22.0,
|
| 377 |
+
"eval_cer": 0.07046208730726275,
|
| 378 |
+
"eval_loss": 0.3041548728942871,
|
| 379 |
+
"eval_runtime": 539.3197,
|
| 380 |
+
"eval_samples_per_second": 7.938,
|
| 381 |
+
"eval_steps_per_second": 0.994,
|
| 382 |
+
"eval_wer": 0.37355113806978557,
|
| 383 |
+
"step": 17006
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 23.0,
|
| 387 |
+
"grad_norm": 0.9431765675544739,
|
| 388 |
+
"learning_rate": 5.4707732634338135e-05,
|
| 389 |
+
"loss": 0.1116,
|
| 390 |
+
"step": 17779
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 23.0,
|
| 394 |
+
"eval_cer": 0.06816410487156206,
|
| 395 |
+
"eval_loss": 0.293849378824234,
|
| 396 |
+
"eval_runtime": 540.8252,
|
| 397 |
+
"eval_samples_per_second": 7.916,
|
| 398 |
+
"eval_steps_per_second": 0.991,
|
| 399 |
+
"eval_wer": 0.3625608071587292,
|
| 400 |
+
"step": 17779
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 24.0,
|
| 404 |
+
"grad_norm": 1.915635347366333,
|
| 405 |
+
"learning_rate": 5.268152031454784e-05,
|
| 406 |
+
"loss": 0.1068,
|
| 407 |
+
"step": 18552
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 24.0,
|
| 411 |
+
"eval_cer": 0.06715170701527436,
|
| 412 |
+
"eval_loss": 0.29729682207107544,
|
| 413 |
+
"eval_runtime": 540.0658,
|
| 414 |
+
"eval_samples_per_second": 7.927,
|
| 415 |
+
"eval_steps_per_second": 0.992,
|
| 416 |
+
"eval_wer": 0.3608191700198186,
|
| 417 |
+
"step": 18552
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 25.0,
|
| 421 |
+
"grad_norm": 1.3635971546173096,
|
| 422 |
+
"learning_rate": 5.065530799475754e-05,
|
| 423 |
+
"loss": 0.1014,
|
| 424 |
+
"step": 19325
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 25.0,
|
| 428 |
+
"eval_cer": 0.06892340326377785,
|
| 429 |
+
"eval_loss": 0.31454575061798096,
|
| 430 |
+
"eval_runtime": 538.5303,
|
| 431 |
+
"eval_samples_per_second": 7.949,
|
| 432 |
+
"eval_steps_per_second": 0.995,
|
| 433 |
+
"eval_wer": 0.36304125878325627,
|
| 434 |
+
"step": 19325
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 26.0,
|
| 438 |
+
"grad_norm": 2.388526201248169,
|
| 439 |
+
"learning_rate": 4.8629095674967235e-05,
|
| 440 |
+
"loss": 0.0996,
|
| 441 |
+
"step": 20098
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 26.0,
|
| 445 |
+
"eval_cer": 0.06636830391219457,
|
| 446 |
+
"eval_loss": 0.30341199040412903,
|
| 447 |
+
"eval_runtime": 543.2486,
|
| 448 |
+
"eval_samples_per_second": 7.88,
|
| 449 |
+
"eval_steps_per_second": 0.987,
|
| 450 |
+
"eval_wer": 0.35634496426641044,
|
| 451 |
+
"step": 20098
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 27.0,
|
| 455 |
+
"grad_norm": 7.999663829803467,
|
| 456 |
+
"learning_rate": 4.6602883355176935e-05,
|
| 457 |
+
"loss": 0.096,
|
| 458 |
+
"step": 20871
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 27.0,
|
| 462 |
+
"eval_cer": 0.06646070530383988,
|
| 463 |
+
"eval_loss": 0.2997344732284546,
|
| 464 |
+
"eval_runtime": 541.4586,
|
| 465 |
+
"eval_samples_per_second": 7.906,
|
| 466 |
+
"eval_steps_per_second": 0.99,
|
| 467 |
+
"eval_wer": 0.35541408924388923,
|
| 468 |
+
"step": 20871
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 28.0,
|
| 472 |
+
"grad_norm": 3.0532593727111816,
|
| 473 |
+
"learning_rate": 4.4576671035386635e-05,
|
| 474 |
+
"loss": 0.0936,
|
| 475 |
+
"step": 21644
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 28.0,
|
| 479 |
+
"eval_cer": 0.0656773022007601,
|
| 480 |
+
"eval_loss": 0.3077986538410187,
|
| 481 |
+
"eval_runtime": 541.1099,
|
| 482 |
+
"eval_samples_per_second": 7.912,
|
| 483 |
+
"eval_steps_per_second": 0.991,
|
| 484 |
+
"eval_wer": 0.35099993994354695,
|
| 485 |
+
"step": 21644
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 29.0,
|
| 489 |
+
"grad_norm": 1.7395150661468506,
|
| 490 |
+
"learning_rate": 4.255045871559633e-05,
|
| 491 |
+
"loss": 0.091,
|
| 492 |
+
"step": 22417
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 29.0,
|
| 496 |
+
"eval_cer": 0.06572952907429876,
|
| 497 |
+
"eval_loss": 0.299217164516449,
|
| 498 |
+
"eval_runtime": 544.6033,
|
| 499 |
+
"eval_samples_per_second": 7.861,
|
| 500 |
+
"eval_steps_per_second": 0.984,
|
| 501 |
+
"eval_wer": 0.34874782295357637,
|
| 502 |
+
"step": 22417
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 30.0,
|
| 506 |
+
"grad_norm": 2.331186532974243,
|
| 507 |
+
"learning_rate": 4.0524246395806035e-05,
|
| 508 |
+
"loss": 0.0914,
|
| 509 |
+
"step": 23190
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 30.0,
|
| 513 |
+
"eval_cer": 0.06598664599018135,
|
| 514 |
+
"eval_loss": 0.31265002489089966,
|
| 515 |
+
"eval_runtime": 539.5662,
|
| 516 |
+
"eval_samples_per_second": 7.934,
|
| 517 |
+
"eval_steps_per_second": 0.993,
|
| 518 |
+
"eval_wer": 0.3516305327007387,
|
| 519 |
+
"step": 23190
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 31.0,
|
| 523 |
+
"grad_norm": 0.5025594234466553,
|
| 524 |
+
"learning_rate": 3.849803407601573e-05,
|
| 525 |
+
"loss": 0.0849,
|
| 526 |
+
"step": 23963
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 31.0,
|
| 530 |
+
"eval_cer": 0.06473721847706437,
|
| 531 |
+
"eval_loss": 0.3014875054359436,
|
| 532 |
+
"eval_runtime": 540.1447,
|
| 533 |
+
"eval_samples_per_second": 7.926,
|
| 534 |
+
"eval_steps_per_second": 0.992,
|
| 535 |
+
"eval_wer": 0.34751666566572575,
|
| 536 |
+
"step": 23963
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 32.0,
|
| 540 |
+
"grad_norm": 3.843916416168213,
|
| 541 |
+
"learning_rate": 3.647182175622543e-05,
|
| 542 |
+
"loss": 0.0819,
|
| 543 |
+
"step": 24736
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 32.0,
|
| 547 |
+
"eval_cer": 0.06535992350771752,
|
| 548 |
+
"eval_loss": 0.3202614486217499,
|
| 549 |
+
"eval_runtime": 539.8225,
|
| 550 |
+
"eval_samples_per_second": 7.93,
|
| 551 |
+
"eval_steps_per_second": 0.993,
|
| 552 |
+
"eval_wer": 0.348927992312774,
|
| 553 |
+
"step": 24736
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 33.0,
|
| 557 |
+
"grad_norm": 3.8599517345428467,
|
| 558 |
+
"learning_rate": 3.444560943643513e-05,
|
| 559 |
+
"loss": 0.0806,
|
| 560 |
+
"step": 25509
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 33.0,
|
| 564 |
+
"eval_cer": 0.06497023068208296,
|
| 565 |
+
"eval_loss": 0.3200363516807556,
|
| 566 |
+
"eval_runtime": 539.477,
|
| 567 |
+
"eval_samples_per_second": 7.935,
|
| 568 |
+
"eval_steps_per_second": 0.994,
|
| 569 |
+
"eval_wer": 0.3485676535943787,
|
| 570 |
+
"step": 25509
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 34.0,
|
| 574 |
+
"grad_norm": 3.0424208641052246,
|
| 575 |
+
"learning_rate": 3.241939711664482e-05,
|
| 576 |
+
"loss": 0.08,
|
| 577 |
+
"step": 26282
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 34.0,
|
| 581 |
+
"eval_cer": 0.06427119406702717,
|
| 582 |
+
"eval_loss": 0.3235025107860565,
|
| 583 |
+
"eval_runtime": 540.33,
|
| 584 |
+
"eval_samples_per_second": 7.923,
|
| 585 |
+
"eval_steps_per_second": 0.992,
|
| 586 |
+
"eval_wer": 0.3427722058735211,
|
| 587 |
+
"step": 26282
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 35.0,
|
| 591 |
+
"grad_norm": 1.9409205913543701,
|
| 592 |
+
"learning_rate": 3.0393184796854524e-05,
|
| 593 |
+
"loss": 0.0762,
|
| 594 |
+
"step": 27055
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 35.0,
|
| 598 |
+
"eval_cer": 0.06393774556674192,
|
| 599 |
+
"eval_loss": 0.31750521063804626,
|
| 600 |
+
"eval_runtime": 540.1387,
|
| 601 |
+
"eval_samples_per_second": 7.926,
|
| 602 |
+
"eval_steps_per_second": 0.992,
|
| 603 |
+
"eval_wer": 0.34442375833283284,
|
| 604 |
+
"step": 27055
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 36.0,
|
| 608 |
+
"grad_norm": 2.8859474658966064,
|
| 609 |
+
"learning_rate": 2.8366972477064224e-05,
|
| 610 |
+
"loss": 0.0753,
|
| 611 |
+
"step": 27828
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 36.0,
|
| 615 |
+
"eval_cer": 0.06355207019291803,
|
| 616 |
+
"eval_loss": 0.3156121075153351,
|
| 617 |
+
"eval_runtime": 541.5718,
|
| 618 |
+
"eval_samples_per_second": 7.905,
|
| 619 |
+
"eval_steps_per_second": 0.99,
|
| 620 |
+
"eval_wer": 0.34202150021019756,
|
| 621 |
+
"step": 27828
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 37.0,
|
| 625 |
+
"grad_norm": 1.7962828874588013,
|
| 626 |
+
"learning_rate": 2.634076015727392e-05,
|
| 627 |
+
"loss": 0.0709,
|
| 628 |
+
"step": 28601
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 37.0,
|
| 632 |
+
"eval_cer": 0.06434752565142982,
|
| 633 |
+
"eval_loss": 0.31625601649284363,
|
| 634 |
+
"eval_runtime": 540.5673,
|
| 635 |
+
"eval_samples_per_second": 7.919,
|
| 636 |
+
"eval_steps_per_second": 0.992,
|
| 637 |
+
"eval_wer": 0.3438832502552399,
|
| 638 |
+
"step": 28601
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 38.0,
|
| 642 |
+
"grad_norm": 0.47632133960723877,
|
| 643 |
+
"learning_rate": 2.4314547837483617e-05,
|
| 644 |
+
"loss": 0.0709,
|
| 645 |
+
"step": 29374
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 38.0,
|
| 649 |
+
"eval_cer": 0.06292534771045422,
|
| 650 |
+
"eval_loss": 0.3177834451198578,
|
| 651 |
+
"eval_runtime": 542.3771,
|
| 652 |
+
"eval_samples_per_second": 7.893,
|
| 653 |
+
"eval_steps_per_second": 0.988,
|
| 654 |
+
"eval_wer": 0.33988949612635877,
|
| 655 |
+
"step": 29374
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 39.0,
|
| 659 |
+
"grad_norm": 1.1476659774780273,
|
| 660 |
+
"learning_rate": 2.2288335517693317e-05,
|
| 661 |
+
"loss": 0.0674,
|
| 662 |
+
"step": 30147
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 39.0,
|
| 666 |
+
"eval_cer": 0.06325879621073945,
|
| 667 |
+
"eval_loss": 0.31504514813423157,
|
| 668 |
+
"eval_runtime": 539.8002,
|
| 669 |
+
"eval_samples_per_second": 7.931,
|
| 670 |
+
"eval_steps_per_second": 0.993,
|
| 671 |
+
"eval_wer": 0.34009969371208937,
|
| 672 |
+
"step": 30147
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 40.0,
|
| 676 |
+
"grad_norm": 3.619795560836792,
|
| 677 |
+
"learning_rate": 2.0262123197903017e-05,
|
| 678 |
+
"loss": 0.0679,
|
| 679 |
+
"step": 30920
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 40.0,
|
| 683 |
+
"eval_cer": 0.06262002137284363,
|
| 684 |
+
"eval_loss": 0.3333224356174469,
|
| 685 |
+
"eval_runtime": 540.4897,
|
| 686 |
+
"eval_samples_per_second": 7.921,
|
| 687 |
+
"eval_steps_per_second": 0.992,
|
| 688 |
+
"eval_wer": 0.3378776049486517,
|
| 689 |
+
"step": 30920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 41.0,
|
| 693 |
+
"grad_norm": 5.51519775390625,
|
| 694 |
+
"learning_rate": 1.8235910878112714e-05,
|
| 695 |
+
"loss": 0.0636,
|
| 696 |
+
"step": 31693
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 41.0,
|
| 700 |
+
"eval_cer": 0.06255975959568365,
|
| 701 |
+
"eval_loss": 0.32877811789512634,
|
| 702 |
+
"eval_runtime": 541.1158,
|
| 703 |
+
"eval_samples_per_second": 7.911,
|
| 704 |
+
"eval_steps_per_second": 0.991,
|
| 705 |
+
"eval_wer": 0.3371869557383941,
|
| 706 |
+
"step": 31693
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 42.0,
|
| 710 |
+
"grad_norm": 3.61895751953125,
|
| 711 |
+
"learning_rate": 1.620969855832241e-05,
|
| 712 |
+
"loss": 0.0614,
|
| 713 |
+
"step": 32466
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 42.0,
|
| 717 |
+
"eval_cer": 0.062358887005150375,
|
| 718 |
+
"eval_loss": 0.3240591883659363,
|
| 719 |
+
"eval_runtime": 540.0535,
|
| 720 |
+
"eval_samples_per_second": 7.927,
|
| 721 |
+
"eval_steps_per_second": 0.992,
|
| 722 |
+
"eval_wer": 0.33619602426280704,
|
| 723 |
+
"step": 32466
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 43.0,
|
| 727 |
+
"grad_norm": 1.1857914924621582,
|
| 728 |
+
"learning_rate": 1.4183486238532112e-05,
|
| 729 |
+
"loss": 0.0629,
|
| 730 |
+
"step": 33239
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 43.0,
|
| 734 |
+
"eval_cer": 0.06161967587198792,
|
| 735 |
+
"eval_loss": 0.3232579231262207,
|
| 736 |
+
"eval_runtime": 540.6476,
|
| 737 |
+
"eval_samples_per_second": 7.918,
|
| 738 |
+
"eval_steps_per_second": 0.991,
|
| 739 |
+
"eval_wer": 0.3335235121013753,
|
| 740 |
+
"step": 33239
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 44.0,
|
| 744 |
+
"grad_norm": 0.7397142052650452,
|
| 745 |
+
"learning_rate": 1.2157273918741809e-05,
|
| 746 |
+
"loss": 0.061,
|
| 747 |
+
"step": 34012
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 44.0,
|
| 751 |
+
"eval_cer": 0.06244325349317435,
|
| 752 |
+
"eval_loss": 0.32336390018463135,
|
| 753 |
+
"eval_runtime": 539.148,
|
| 754 |
+
"eval_samples_per_second": 7.94,
|
| 755 |
+
"eval_steps_per_second": 0.994,
|
| 756 |
+
"eval_wer": 0.3349648669749565,
|
| 757 |
+
"step": 34012
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 45.0,
|
| 761 |
+
"grad_norm": 2.209028482437134,
|
| 762 |
+
"learning_rate": 1.0131061598951509e-05,
|
| 763 |
+
"loss": 0.0604,
|
| 764 |
+
"step": 34785
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 45.0,
|
| 768 |
+
"eval_cer": 0.061691990004579894,
|
| 769 |
+
"eval_loss": 0.32257187366485596,
|
| 770 |
+
"eval_runtime": 539.0839,
|
| 771 |
+
"eval_samples_per_second": 7.941,
|
| 772 |
+
"eval_steps_per_second": 0.994,
|
| 773 |
+
"eval_wer": 0.33421416131163295,
|
| 774 |
+
"step": 34785
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 46.0,
|
| 778 |
+
"grad_norm": 0.40143582224845886,
|
| 779 |
+
"learning_rate": 8.104849279161205e-06,
|
| 780 |
+
"loss": 0.0582,
|
| 781 |
+
"step": 35558
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 46.0,
|
| 785 |
+
"eval_cer": 0.06054299878672955,
|
| 786 |
+
"eval_loss": 0.326095849275589,
|
| 787 |
+
"eval_runtime": 538.3445,
|
| 788 |
+
"eval_samples_per_second": 7.952,
|
| 789 |
+
"eval_steps_per_second": 0.996,
|
| 790 |
+
"eval_wer": 0.3291994474806318,
|
| 791 |
+
"step": 35558
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 47.0,
|
| 795 |
+
"grad_norm": 2.2391343116760254,
|
| 796 |
+
"learning_rate": 6.078636959370904e-06,
|
| 797 |
+
"loss": 0.0575,
|
| 798 |
+
"step": 36331
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 47.0,
|
| 802 |
+
"eval_cer": 0.060346143648006945,
|
| 803 |
+
"eval_loss": 0.31861406564712524,
|
| 804 |
+
"eval_runtime": 541.1135,
|
| 805 |
+
"eval_samples_per_second": 7.911,
|
| 806 |
+
"eval_steps_per_second": 0.991,
|
| 807 |
+
"eval_wer": 0.32769803615398474,
|
| 808 |
+
"step": 36331
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 48.0,
|
| 812 |
+
"grad_norm": 2.6581954956054688,
|
| 813 |
+
"learning_rate": 4.052424639580603e-06,
|
| 814 |
+
"loss": 0.0553,
|
| 815 |
+
"step": 37104
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 48.0,
|
| 819 |
+
"eval_cer": 0.060695661955534844,
|
| 820 |
+
"eval_loss": 0.32371774315834045,
|
| 821 |
+
"eval_runtime": 539.7027,
|
| 822 |
+
"eval_samples_per_second": 7.932,
|
| 823 |
+
"eval_steps_per_second": 0.993,
|
| 824 |
+
"eval_wer": 0.3286889676295718,
|
| 825 |
+
"step": 37104
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 49.0,
|
| 829 |
+
"grad_norm": 1.2842614650726318,
|
| 830 |
+
"learning_rate": 2.0262123197903013e-06,
|
| 831 |
+
"loss": 0.0554,
|
| 832 |
+
"step": 37877
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 49.0,
|
| 836 |
+
"eval_cer": 0.060611295467510866,
|
| 837 |
+
"eval_loss": 0.3244434893131256,
|
| 838 |
+
"eval_runtime": 539.5278,
|
| 839 |
+
"eval_samples_per_second": 7.935,
|
| 840 |
+
"eval_steps_per_second": 0.993,
|
| 841 |
+
"eval_wer": 0.3289291934418353,
|
| 842 |
+
"step": 37877
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 50.0,
|
| 846 |
+
"grad_norm": 1.3330661058425903,
|
| 847 |
+
"learning_rate": 0.0,
|
| 848 |
+
"loss": 0.0543,
|
| 849 |
+
"step": 38650
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 50.0,
|
| 853 |
+
"eval_cer": 0.0604787195577589,
|
| 854 |
+
"eval_loss": 0.32520928978919983,
|
| 855 |
+
"eval_runtime": 540.1631,
|
| 856 |
+
"eval_samples_per_second": 7.925,
|
| 857 |
+
"eval_steps_per_second": 0.992,
|
| 858 |
+
"eval_wer": 0.32829860068464356,
|
| 859 |
+
"step": 38650
|
| 860 |
+
}
|
| 861 |
+
],
|
| 862 |
+
"logging_steps": 500,
|
| 863 |
+
"max_steps": 38650,
|
| 864 |
+
"num_input_tokens_seen": 0,
|
| 865 |
+
"num_train_epochs": 50,
|
| 866 |
+
"save_steps": 500,
|
| 867 |
+
"total_flos": 4.930233517238343e+19,
|
| 868 |
+
"train_batch_size": 8,
|
| 869 |
+
"trial_name": null,
|
| 870 |
+
"trial_params": null
|
| 871 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c6ab60db645952785375eda3a692c5162e6a70391f370abfe5ca9d3fe0738f4
|
| 3 |
+
size 4856
|
training_args/byteorder
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
little
|
training_args/data.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59968a613ed68c67977829690cdcbb5e8a7b48fc351624ed25a32ed9e8293564
|
| 3 |
+
size 3988
|
training_args/version
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
3
|