Upload 7 files
Browse files- config.json +116 -0
- preprocessor_config.json +10 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2057 -0
- training_args.bin +3 -0
- vocab.json +1 -0
config.json
ADDED
|
@@ -0,0 +1,116 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/wav2vec2-large-xlsr-53",
|
| 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": 0.0,
|
| 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_channel_length": 10,
|
| 64 |
+
"mask_channel_min_space": 1,
|
| 65 |
+
"mask_channel_other": 0.0,
|
| 66 |
+
"mask_channel_prob": 0.0,
|
| 67 |
+
"mask_channel_selection": "static",
|
| 68 |
+
"mask_feature_length": 10,
|
| 69 |
+
"mask_feature_min_masks": 0,
|
| 70 |
+
"mask_feature_prob": 0.0,
|
| 71 |
+
"mask_time_length": 10,
|
| 72 |
+
"mask_time_min_masks": 2,
|
| 73 |
+
"mask_time_min_space": 1,
|
| 74 |
+
"mask_time_other": 0.0,
|
| 75 |
+
"mask_time_prob": 0.05,
|
| 76 |
+
"mask_time_selection": "static",
|
| 77 |
+
"model_type": "wav2vec2",
|
| 78 |
+
"num_adapter_layers": 3,
|
| 79 |
+
"num_attention_heads": 16,
|
| 80 |
+
"num_codevector_groups": 2,
|
| 81 |
+
"num_codevectors_per_group": 320,
|
| 82 |
+
"num_conv_pos_embedding_groups": 16,
|
| 83 |
+
"num_conv_pos_embeddings": 128,
|
| 84 |
+
"num_feat_extract_layers": 7,
|
| 85 |
+
"num_hidden_layers": 24,
|
| 86 |
+
"num_negatives": 100,
|
| 87 |
+
"output_hidden_size": 1024,
|
| 88 |
+
"pad_token_id": 26,
|
| 89 |
+
"proj_codevector_dim": 768,
|
| 90 |
+
"tdnn_dilation": [
|
| 91 |
+
1,
|
| 92 |
+
2,
|
| 93 |
+
3,
|
| 94 |
+
1,
|
| 95 |
+
1
|
| 96 |
+
],
|
| 97 |
+
"tdnn_dim": [
|
| 98 |
+
512,
|
| 99 |
+
512,
|
| 100 |
+
512,
|
| 101 |
+
512,
|
| 102 |
+
1500
|
| 103 |
+
],
|
| 104 |
+
"tdnn_kernel": [
|
| 105 |
+
5,
|
| 106 |
+
3,
|
| 107 |
+
3,
|
| 108 |
+
1,
|
| 109 |
+
1
|
| 110 |
+
],
|
| 111 |
+
"torch_dtype": "float32",
|
| 112 |
+
"transformers_version": "4.41.1",
|
| 113 |
+
"use_weighted_layer_sum": false,
|
| 114 |
+
"vocab_size": 29,
|
| 115 |
+
"xvector_output_dim": 512
|
| 116 |
+
}
|
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": 0.0,
|
| 7 |
+
"processor_class": "Wav2Vec2Processor",
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000
|
| 10 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc6d8be337d23ba212f987b1383487c3e673290d8ec4a9a819f5aab5b08628c5
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac2ea29afe0c11f282dd386c9043289e8b142c9be1c58f45cdd17f545ae4647f
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2057 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.923943661971831,
|
| 3 |
+
"best_model_checkpoint": "./ssw-finetune/checkpoint-1150",
|
| 4 |
+
"epoch": 115.0,
|
| 5 |
+
"eval_steps": 25,
|
| 6 |
+
"global_step": 1150,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.5,
|
| 13 |
+
"grad_norm": 5.561491012573242,
|
| 14 |
+
"learning_rate": 2.9999999999999997e-06,
|
| 15 |
+
"loss": 7.7799,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"grad_norm": 4.18166971206665,
|
| 21 |
+
"learning_rate": 6.749999999999999e-06,
|
| 22 |
+
"loss": 7.4713,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.5,
|
| 27 |
+
"grad_norm": 6.806884288787842,
|
| 28 |
+
"learning_rate": 1.05e-05,
|
| 29 |
+
"loss": 7.696,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.0,
|
| 34 |
+
"grad_norm": NaN,
|
| 35 |
+
"learning_rate": 1.3499999999999998e-05,
|
| 36 |
+
"loss": 7.9462,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 2.5,
|
| 41 |
+
"grad_norm": NaN,
|
| 42 |
+
"learning_rate": 1.6499999999999998e-05,
|
| 43 |
+
"loss": 8.0165,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.5,
|
| 48 |
+
"eval_loss": 7.467132091522217,
|
| 49 |
+
"eval_runtime": 1.054,
|
| 50 |
+
"eval_samples_per_second": 22.77,
|
| 51 |
+
"eval_steps_per_second": 0.949,
|
| 52 |
+
"eval_wer": 1.0,
|
| 53 |
+
"step": 25
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 3.0,
|
| 57 |
+
"grad_norm": 2.7979044914245605,
|
| 58 |
+
"learning_rate": 2.025e-05,
|
| 59 |
+
"loss": 6.5204,
|
| 60 |
+
"step": 30
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 3.5,
|
| 64 |
+
"grad_norm": 9.716986656188965,
|
| 65 |
+
"learning_rate": 2.3999999999999997e-05,
|
| 66 |
+
"loss": 7.6715,
|
| 67 |
+
"step": 35
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 4.0,
|
| 71 |
+
"grad_norm": 3.3519299030303955,
|
| 72 |
+
"learning_rate": 2.7749999999999997e-05,
|
| 73 |
+
"loss": 7.0161,
|
| 74 |
+
"step": 40
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.5,
|
| 78 |
+
"grad_norm": 5.6749138832092285,
|
| 79 |
+
"learning_rate": 3.149999999999999e-05,
|
| 80 |
+
"loss": 8.0617,
|
| 81 |
+
"step": 45
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 5.0,
|
| 85 |
+
"grad_norm": 8.150848388671875,
|
| 86 |
+
"learning_rate": 3.5249999999999996e-05,
|
| 87 |
+
"loss": 6.3142,
|
| 88 |
+
"step": 50
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 5.0,
|
| 92 |
+
"eval_loss": 6.626723766326904,
|
| 93 |
+
"eval_runtime": 1.071,
|
| 94 |
+
"eval_samples_per_second": 22.409,
|
| 95 |
+
"eval_steps_per_second": 0.934,
|
| 96 |
+
"eval_wer": 1.0,
|
| 97 |
+
"step": 50
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 5.5,
|
| 101 |
+
"grad_norm": 6.339476585388184,
|
| 102 |
+
"learning_rate": 3.9e-05,
|
| 103 |
+
"loss": 6.2643,
|
| 104 |
+
"step": 55
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 6.0,
|
| 108 |
+
"grad_norm": 13.012835502624512,
|
| 109 |
+
"learning_rate": 4.2749999999999996e-05,
|
| 110 |
+
"loss": 7.1655,
|
| 111 |
+
"step": 60
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 6.5,
|
| 115 |
+
"grad_norm": 11.24893569946289,
|
| 116 |
+
"learning_rate": 4.65e-05,
|
| 117 |
+
"loss": 5.8178,
|
| 118 |
+
"step": 65
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 7.0,
|
| 122 |
+
"grad_norm": 24.677473068237305,
|
| 123 |
+
"learning_rate": 5.025e-05,
|
| 124 |
+
"loss": 5.5684,
|
| 125 |
+
"step": 70
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 7.5,
|
| 129 |
+
"grad_norm": 8.214367866516113,
|
| 130 |
+
"learning_rate": 5.399999999999999e-05,
|
| 131 |
+
"loss": 4.3185,
|
| 132 |
+
"step": 75
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 7.5,
|
| 136 |
+
"eval_loss": 3.72790789604187,
|
| 137 |
+
"eval_runtime": 1.0391,
|
| 138 |
+
"eval_samples_per_second": 23.097,
|
| 139 |
+
"eval_steps_per_second": 0.962,
|
| 140 |
+
"eval_wer": 1.0,
|
| 141 |
+
"step": 75
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 8.0,
|
| 145 |
+
"grad_norm": 8.643641471862793,
|
| 146 |
+
"learning_rate": 5.7749999999999994e-05,
|
| 147 |
+
"loss": 4.1807,
|
| 148 |
+
"step": 80
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 8.5,
|
| 152 |
+
"grad_norm": 10.54008674621582,
|
| 153 |
+
"learning_rate": 6.149999999999999e-05,
|
| 154 |
+
"loss": 3.7552,
|
| 155 |
+
"step": 85
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 9.0,
|
| 159 |
+
"grad_norm": 3.332289934158325,
|
| 160 |
+
"learning_rate": 6.525e-05,
|
| 161 |
+
"loss": 3.7053,
|
| 162 |
+
"step": 90
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 9.5,
|
| 166 |
+
"grad_norm": 4.925398349761963,
|
| 167 |
+
"learning_rate": 6.9e-05,
|
| 168 |
+
"loss": 3.3661,
|
| 169 |
+
"step": 95
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.0,
|
| 173 |
+
"grad_norm": 5.291933536529541,
|
| 174 |
+
"learning_rate": 7.274999999999999e-05,
|
| 175 |
+
"loss": 3.1777,
|
| 176 |
+
"step": 100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 10.0,
|
| 180 |
+
"eval_loss": 3.050647735595703,
|
| 181 |
+
"eval_runtime": 1.0273,
|
| 182 |
+
"eval_samples_per_second": 23.362,
|
| 183 |
+
"eval_steps_per_second": 0.973,
|
| 184 |
+
"eval_wer": 1.0,
|
| 185 |
+
"step": 100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 10.5,
|
| 189 |
+
"grad_norm": 1.6660319566726685,
|
| 190 |
+
"learning_rate": 7.649999999999999e-05,
|
| 191 |
+
"loss": 3.0435,
|
| 192 |
+
"step": 105
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 11.0,
|
| 196 |
+
"grad_norm": 0.681082546710968,
|
| 197 |
+
"learning_rate": 8.025e-05,
|
| 198 |
+
"loss": 3.254,
|
| 199 |
+
"step": 110
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 11.5,
|
| 203 |
+
"grad_norm": 2.713016986846924,
|
| 204 |
+
"learning_rate": 8.4e-05,
|
| 205 |
+
"loss": 2.972,
|
| 206 |
+
"step": 115
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 12.0,
|
| 210 |
+
"grad_norm": 7.211615085601807,
|
| 211 |
+
"learning_rate": 8.774999999999999e-05,
|
| 212 |
+
"loss": 3.1145,
|
| 213 |
+
"step": 120
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 12.5,
|
| 217 |
+
"grad_norm": 3.3372182846069336,
|
| 218 |
+
"learning_rate": 9.149999999999999e-05,
|
| 219 |
+
"loss": 3.0587,
|
| 220 |
+
"step": 125
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 12.5,
|
| 224 |
+
"eval_loss": 2.936924695968628,
|
| 225 |
+
"eval_runtime": 1.0424,
|
| 226 |
+
"eval_samples_per_second": 23.023,
|
| 227 |
+
"eval_steps_per_second": 0.959,
|
| 228 |
+
"eval_wer": 1.0,
|
| 229 |
+
"step": 125
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 13.0,
|
| 233 |
+
"grad_norm": 1.895374059677124,
|
| 234 |
+
"learning_rate": 9.525e-05,
|
| 235 |
+
"loss": 2.9096,
|
| 236 |
+
"step": 130
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 13.5,
|
| 240 |
+
"grad_norm": 8.356375694274902,
|
| 241 |
+
"learning_rate": 9.9e-05,
|
| 242 |
+
"loss": 3.3159,
|
| 243 |
+
"step": 135
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 14.0,
|
| 247 |
+
"grad_norm": 1.6825320720672607,
|
| 248 |
+
"learning_rate": 0.00010275,
|
| 249 |
+
"loss": 2.9022,
|
| 250 |
+
"step": 140
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 14.5,
|
| 254 |
+
"grad_norm": 0.7314967513084412,
|
| 255 |
+
"learning_rate": 0.00010649999999999999,
|
| 256 |
+
"loss": 2.9058,
|
| 257 |
+
"step": 145
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 15.0,
|
| 261 |
+
"grad_norm": 3.183772563934326,
|
| 262 |
+
"learning_rate": 0.00011024999999999998,
|
| 263 |
+
"loss": 3.0633,
|
| 264 |
+
"step": 150
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 15.0,
|
| 268 |
+
"eval_loss": 2.9296257495880127,
|
| 269 |
+
"eval_runtime": 1.0347,
|
| 270 |
+
"eval_samples_per_second": 23.194,
|
| 271 |
+
"eval_steps_per_second": 0.966,
|
| 272 |
+
"eval_wer": 1.0,
|
| 273 |
+
"step": 150
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 15.5,
|
| 277 |
+
"grad_norm": 1.4891362190246582,
|
| 278 |
+
"learning_rate": 0.00011399999999999999,
|
| 279 |
+
"loss": 2.9901,
|
| 280 |
+
"step": 155
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 16.0,
|
| 284 |
+
"grad_norm": 5.58284854888916,
|
| 285 |
+
"learning_rate": 0.00011774999999999999,
|
| 286 |
+
"loss": 2.9861,
|
| 287 |
+
"step": 160
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 16.5,
|
| 291 |
+
"grad_norm": 1.3804000616073608,
|
| 292 |
+
"learning_rate": 0.0001215,
|
| 293 |
+
"loss": 2.9584,
|
| 294 |
+
"step": 165
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 17.0,
|
| 298 |
+
"grad_norm": 1.6562563180923462,
|
| 299 |
+
"learning_rate": 0.00012524999999999998,
|
| 300 |
+
"loss": 3.0194,
|
| 301 |
+
"step": 170
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 17.5,
|
| 305 |
+
"grad_norm": 0.653541088104248,
|
| 306 |
+
"learning_rate": 0.000129,
|
| 307 |
+
"loss": 2.9639,
|
| 308 |
+
"step": 175
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 17.5,
|
| 312 |
+
"eval_loss": 2.926556348800659,
|
| 313 |
+
"eval_runtime": 1.0442,
|
| 314 |
+
"eval_samples_per_second": 22.985,
|
| 315 |
+
"eval_steps_per_second": 0.958,
|
| 316 |
+
"eval_wer": 1.0,
|
| 317 |
+
"step": 175
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 18.0,
|
| 321 |
+
"grad_norm": 9.456038475036621,
|
| 322 |
+
"learning_rate": 0.00013275,
|
| 323 |
+
"loss": 2.8944,
|
| 324 |
+
"step": 180
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 18.5,
|
| 328 |
+
"grad_norm": 0.3759576082229614,
|
| 329 |
+
"learning_rate": 0.00013649999999999998,
|
| 330 |
+
"loss": 2.9149,
|
| 331 |
+
"step": 185
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 19.0,
|
| 335 |
+
"grad_norm": 3.0567305088043213,
|
| 336 |
+
"learning_rate": 0.00014025,
|
| 337 |
+
"loss": 3.0321,
|
| 338 |
+
"step": 190
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 19.5,
|
| 342 |
+
"grad_norm": 8.436885833740234,
|
| 343 |
+
"learning_rate": 0.00014399999999999998,
|
| 344 |
+
"loss": 2.9683,
|
| 345 |
+
"step": 195
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 20.0,
|
| 349 |
+
"grad_norm": 1.9778860807418823,
|
| 350 |
+
"learning_rate": 0.00014774999999999999,
|
| 351 |
+
"loss": 2.9576,
|
| 352 |
+
"step": 200
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 20.0,
|
| 356 |
+
"eval_loss": 2.9644908905029297,
|
| 357 |
+
"eval_runtime": 1.0268,
|
| 358 |
+
"eval_samples_per_second": 23.374,
|
| 359 |
+
"eval_steps_per_second": 0.974,
|
| 360 |
+
"eval_wer": 1.0,
|
| 361 |
+
"step": 200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 20.5,
|
| 365 |
+
"grad_norm": 0.6856608390808105,
|
| 366 |
+
"learning_rate": 0.00014976923076923077,
|
| 367 |
+
"loss": 2.9374,
|
| 368 |
+
"step": 205
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 21.0,
|
| 372 |
+
"grad_norm": 1.157402515411377,
|
| 373 |
+
"learning_rate": 0.00014919230769230767,
|
| 374 |
+
"loss": 2.875,
|
| 375 |
+
"step": 210
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 21.5,
|
| 379 |
+
"grad_norm": 0.42920613288879395,
|
| 380 |
+
"learning_rate": 0.0001486153846153846,
|
| 381 |
+
"loss": 2.9796,
|
| 382 |
+
"step": 215
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 22.0,
|
| 386 |
+
"grad_norm": 4.603660583496094,
|
| 387 |
+
"learning_rate": 0.00014803846153846152,
|
| 388 |
+
"loss": 2.9233,
|
| 389 |
+
"step": 220
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 22.5,
|
| 393 |
+
"grad_norm": 1.3661619424819946,
|
| 394 |
+
"learning_rate": 0.00014746153846153845,
|
| 395 |
+
"loss": 2.8708,
|
| 396 |
+
"step": 225
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 22.5,
|
| 400 |
+
"eval_loss": 2.9085776805877686,
|
| 401 |
+
"eval_runtime": 1.0387,
|
| 402 |
+
"eval_samples_per_second": 23.106,
|
| 403 |
+
"eval_steps_per_second": 0.963,
|
| 404 |
+
"eval_wer": 1.0,
|
| 405 |
+
"step": 225
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 23.0,
|
| 409 |
+
"grad_norm": 0.7445681691169739,
|
| 410 |
+
"learning_rate": 0.00014688461538461537,
|
| 411 |
+
"loss": 2.933,
|
| 412 |
+
"step": 230
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 23.5,
|
| 416 |
+
"grad_norm": 1.2040903568267822,
|
| 417 |
+
"learning_rate": 0.0001463076923076923,
|
| 418 |
+
"loss": 2.9217,
|
| 419 |
+
"step": 235
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 24.0,
|
| 423 |
+
"grad_norm": 4.538419246673584,
|
| 424 |
+
"learning_rate": 0.00014573076923076923,
|
| 425 |
+
"loss": 2.9043,
|
| 426 |
+
"step": 240
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 24.5,
|
| 430 |
+
"grad_norm": 0.36169031262397766,
|
| 431 |
+
"learning_rate": 0.00014515384615384615,
|
| 432 |
+
"loss": 2.8554,
|
| 433 |
+
"step": 245
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 25.0,
|
| 437 |
+
"grad_norm": 1.2133870124816895,
|
| 438 |
+
"learning_rate": 0.00014457692307692305,
|
| 439 |
+
"loss": 2.943,
|
| 440 |
+
"step": 250
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 25.0,
|
| 444 |
+
"eval_loss": 2.900446653366089,
|
| 445 |
+
"eval_runtime": 1.0279,
|
| 446 |
+
"eval_samples_per_second": 23.348,
|
| 447 |
+
"eval_steps_per_second": 0.973,
|
| 448 |
+
"eval_wer": 1.0,
|
| 449 |
+
"step": 250
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 25.5,
|
| 453 |
+
"grad_norm": 1.1455128192901611,
|
| 454 |
+
"learning_rate": 0.00014399999999999998,
|
| 455 |
+
"loss": 2.8775,
|
| 456 |
+
"step": 255
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 26.0,
|
| 460 |
+
"grad_norm": 3.7162177562713623,
|
| 461 |
+
"learning_rate": 0.0001434230769230769,
|
| 462 |
+
"loss": 2.9401,
|
| 463 |
+
"step": 260
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 26.5,
|
| 467 |
+
"grad_norm": 4.095553398132324,
|
| 468 |
+
"learning_rate": 0.00014284615384615383,
|
| 469 |
+
"loss": 2.9053,
|
| 470 |
+
"step": 265
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 27.0,
|
| 474 |
+
"grad_norm": 2.0302634239196777,
|
| 475 |
+
"learning_rate": 0.00014226923076923075,
|
| 476 |
+
"loss": 2.975,
|
| 477 |
+
"step": 270
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 27.5,
|
| 481 |
+
"grad_norm": 3.123234510421753,
|
| 482 |
+
"learning_rate": 0.00014169230769230768,
|
| 483 |
+
"loss": 2.9225,
|
| 484 |
+
"step": 275
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 27.5,
|
| 488 |
+
"eval_loss": 2.9469966888427734,
|
| 489 |
+
"eval_runtime": 1.023,
|
| 490 |
+
"eval_samples_per_second": 23.46,
|
| 491 |
+
"eval_steps_per_second": 0.978,
|
| 492 |
+
"eval_wer": 1.0,
|
| 493 |
+
"step": 275
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 28.0,
|
| 497 |
+
"grad_norm": 0.886202871799469,
|
| 498 |
+
"learning_rate": 0.0001411153846153846,
|
| 499 |
+
"loss": 2.8783,
|
| 500 |
+
"step": 280
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 28.5,
|
| 504 |
+
"grad_norm": 0.48980531096458435,
|
| 505 |
+
"learning_rate": 0.00014053846153846153,
|
| 506 |
+
"loss": 2.8977,
|
| 507 |
+
"step": 285
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 29.0,
|
| 511 |
+
"grad_norm": 2.4499869346618652,
|
| 512 |
+
"learning_rate": 0.00013996153846153843,
|
| 513 |
+
"loss": 2.9178,
|
| 514 |
+
"step": 290
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 29.5,
|
| 518 |
+
"grad_norm": 3.5155863761901855,
|
| 519 |
+
"learning_rate": 0.00013938461538461536,
|
| 520 |
+
"loss": 2.8955,
|
| 521 |
+
"step": 295
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 30.0,
|
| 525 |
+
"grad_norm": 3.8240697383880615,
|
| 526 |
+
"learning_rate": 0.00013880769230769228,
|
| 527 |
+
"loss": 2.9897,
|
| 528 |
+
"step": 300
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 30.0,
|
| 532 |
+
"eval_loss": 2.9530646800994873,
|
| 533 |
+
"eval_runtime": 1.0334,
|
| 534 |
+
"eval_samples_per_second": 23.224,
|
| 535 |
+
"eval_steps_per_second": 0.968,
|
| 536 |
+
"eval_wer": 1.0,
|
| 537 |
+
"step": 300
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 30.5,
|
| 541 |
+
"grad_norm": 1.4881560802459717,
|
| 542 |
+
"learning_rate": 0.0001382307692307692,
|
| 543 |
+
"loss": 2.8732,
|
| 544 |
+
"step": 305
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 31.0,
|
| 548 |
+
"grad_norm": 5.950206756591797,
|
| 549 |
+
"learning_rate": 0.00013765384615384613,
|
| 550 |
+
"loss": 2.9688,
|
| 551 |
+
"step": 310
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 31.5,
|
| 555 |
+
"grad_norm": 0.8825148940086365,
|
| 556 |
+
"learning_rate": 0.00013707692307692306,
|
| 557 |
+
"loss": 2.869,
|
| 558 |
+
"step": 315
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 32.0,
|
| 562 |
+
"grad_norm": 1.6368755102157593,
|
| 563 |
+
"learning_rate": 0.00013649999999999998,
|
| 564 |
+
"loss": 2.8843,
|
| 565 |
+
"step": 320
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 32.5,
|
| 569 |
+
"grad_norm": 1.556404709815979,
|
| 570 |
+
"learning_rate": 0.0001359230769230769,
|
| 571 |
+
"loss": 2.8514,
|
| 572 |
+
"step": 325
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 32.5,
|
| 576 |
+
"eval_loss": 2.911478042602539,
|
| 577 |
+
"eval_runtime": 1.021,
|
| 578 |
+
"eval_samples_per_second": 23.505,
|
| 579 |
+
"eval_steps_per_second": 0.979,
|
| 580 |
+
"eval_wer": 1.0,
|
| 581 |
+
"step": 325
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 33.0,
|
| 585 |
+
"grad_norm": 0.6802976131439209,
|
| 586 |
+
"learning_rate": 0.00013534615384615384,
|
| 587 |
+
"loss": 2.8542,
|
| 588 |
+
"step": 330
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 33.5,
|
| 592 |
+
"grad_norm": 0.5035978555679321,
|
| 593 |
+
"learning_rate": 0.00013476923076923076,
|
| 594 |
+
"loss": 2.9064,
|
| 595 |
+
"step": 335
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 34.0,
|
| 599 |
+
"grad_norm": 1.6443456411361694,
|
| 600 |
+
"learning_rate": 0.0001341923076923077,
|
| 601 |
+
"loss": 2.8498,
|
| 602 |
+
"step": 340
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 34.5,
|
| 606 |
+
"grad_norm": 0.6262179017066956,
|
| 607 |
+
"learning_rate": 0.0001336153846153846,
|
| 608 |
+
"loss": 2.8368,
|
| 609 |
+
"step": 345
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 35.0,
|
| 613 |
+
"grad_norm": 0.8266497850418091,
|
| 614 |
+
"learning_rate": 0.00013303846153846154,
|
| 615 |
+
"loss": 2.8681,
|
| 616 |
+
"step": 350
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 35.0,
|
| 620 |
+
"eval_loss": 2.9094789028167725,
|
| 621 |
+
"eval_runtime": 1.0369,
|
| 622 |
+
"eval_samples_per_second": 23.145,
|
| 623 |
+
"eval_steps_per_second": 0.964,
|
| 624 |
+
"eval_wer": 1.0,
|
| 625 |
+
"step": 350
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 35.5,
|
| 629 |
+
"grad_norm": 0.33677324652671814,
|
| 630 |
+
"learning_rate": 0.00013246153846153846,
|
| 631 |
+
"loss": 2.8163,
|
| 632 |
+
"step": 355
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 36.0,
|
| 636 |
+
"grad_norm": 0.6221341490745544,
|
| 637 |
+
"learning_rate": 0.0001318846153846154,
|
| 638 |
+
"loss": 2.8746,
|
| 639 |
+
"step": 360
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 36.5,
|
| 643 |
+
"grad_norm": 0.5015878677368164,
|
| 644 |
+
"learning_rate": 0.00013130769230769232,
|
| 645 |
+
"loss": 2.8477,
|
| 646 |
+
"step": 365
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 37.0,
|
| 650 |
+
"grad_norm": 0.6005992889404297,
|
| 651 |
+
"learning_rate": 0.00013073076923076921,
|
| 652 |
+
"loss": 2.838,
|
| 653 |
+
"step": 370
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 37.5,
|
| 657 |
+
"grad_norm": 0.4997330605983734,
|
| 658 |
+
"learning_rate": 0.00013015384615384614,
|
| 659 |
+
"loss": 2.8431,
|
| 660 |
+
"step": 375
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 37.5,
|
| 664 |
+
"eval_loss": 2.90104603767395,
|
| 665 |
+
"eval_runtime": 1.017,
|
| 666 |
+
"eval_samples_per_second": 23.599,
|
| 667 |
+
"eval_steps_per_second": 0.983,
|
| 668 |
+
"eval_wer": 1.0,
|
| 669 |
+
"step": 375
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 38.0,
|
| 673 |
+
"grad_norm": 1.342210292816162,
|
| 674 |
+
"learning_rate": 0.00012957692307692307,
|
| 675 |
+
"loss": 2.8672,
|
| 676 |
+
"step": 380
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 38.5,
|
| 680 |
+
"grad_norm": 1.2935914993286133,
|
| 681 |
+
"learning_rate": 0.000129,
|
| 682 |
+
"loss": 2.848,
|
| 683 |
+
"step": 385
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 39.0,
|
| 687 |
+
"grad_norm": 0.41487249732017517,
|
| 688 |
+
"learning_rate": 0.00012842307692307692,
|
| 689 |
+
"loss": 2.8244,
|
| 690 |
+
"step": 390
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 39.5,
|
| 694 |
+
"grad_norm": 1.1988450288772583,
|
| 695 |
+
"learning_rate": 0.00012784615384615384,
|
| 696 |
+
"loss": 2.8328,
|
| 697 |
+
"step": 395
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 40.0,
|
| 701 |
+
"grad_norm": 2.2671468257904053,
|
| 702 |
+
"learning_rate": 0.00012726923076923077,
|
| 703 |
+
"loss": 2.8843,
|
| 704 |
+
"step": 400
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 40.0,
|
| 708 |
+
"eval_loss": 2.9156665802001953,
|
| 709 |
+
"eval_runtime": 1.0579,
|
| 710 |
+
"eval_samples_per_second": 22.686,
|
| 711 |
+
"eval_steps_per_second": 0.945,
|
| 712 |
+
"eval_wer": 1.0,
|
| 713 |
+
"step": 400
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 40.5,
|
| 717 |
+
"grad_norm": 1.003772497177124,
|
| 718 |
+
"learning_rate": 0.0001266923076923077,
|
| 719 |
+
"loss": 2.8312,
|
| 720 |
+
"step": 405
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 41.0,
|
| 724 |
+
"grad_norm": 1.2402571439743042,
|
| 725 |
+
"learning_rate": 0.00012611538461538462,
|
| 726 |
+
"loss": 2.8291,
|
| 727 |
+
"step": 410
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 41.5,
|
| 731 |
+
"grad_norm": 0.29388442635536194,
|
| 732 |
+
"learning_rate": 0.00012553846153846152,
|
| 733 |
+
"loss": 2.8275,
|
| 734 |
+
"step": 415
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 42.0,
|
| 738 |
+
"grad_norm": 0.9477460980415344,
|
| 739 |
+
"learning_rate": 0.00012496153846153844,
|
| 740 |
+
"loss": 2.8384,
|
| 741 |
+
"step": 420
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 42.5,
|
| 745 |
+
"grad_norm": 1.4519686698913574,
|
| 746 |
+
"learning_rate": 0.00012438461538461537,
|
| 747 |
+
"loss": 2.9357,
|
| 748 |
+
"step": 425
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 42.5,
|
| 752 |
+
"eval_loss": 2.902658462524414,
|
| 753 |
+
"eval_runtime": 1.0363,
|
| 754 |
+
"eval_samples_per_second": 23.158,
|
| 755 |
+
"eval_steps_per_second": 0.965,
|
| 756 |
+
"eval_wer": 1.0,
|
| 757 |
+
"step": 425
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 43.0,
|
| 761 |
+
"grad_norm": 0.4391646087169647,
|
| 762 |
+
"learning_rate": 0.0001238076923076923,
|
| 763 |
+
"loss": 2.8395,
|
| 764 |
+
"step": 430
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 43.5,
|
| 768 |
+
"grad_norm": 2.1784377098083496,
|
| 769 |
+
"learning_rate": 0.00012323076923076922,
|
| 770 |
+
"loss": 2.8599,
|
| 771 |
+
"step": 435
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 44.0,
|
| 775 |
+
"grad_norm": 0.9729048609733582,
|
| 776 |
+
"learning_rate": 0.00012265384615384615,
|
| 777 |
+
"loss": 2.8489,
|
| 778 |
+
"step": 440
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 44.5,
|
| 782 |
+
"grad_norm": 0.5243009328842163,
|
| 783 |
+
"learning_rate": 0.00012207692307692307,
|
| 784 |
+
"loss": 2.83,
|
| 785 |
+
"step": 445
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 45.0,
|
| 789 |
+
"grad_norm": 0.7081323862075806,
|
| 790 |
+
"learning_rate": 0.0001215,
|
| 791 |
+
"loss": 2.8236,
|
| 792 |
+
"step": 450
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 45.0,
|
| 796 |
+
"eval_loss": 2.901521682739258,
|
| 797 |
+
"eval_runtime": 1.0318,
|
| 798 |
+
"eval_samples_per_second": 23.261,
|
| 799 |
+
"eval_steps_per_second": 0.969,
|
| 800 |
+
"eval_wer": 1.0,
|
| 801 |
+
"step": 450
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 45.5,
|
| 805 |
+
"grad_norm": 0.3105088770389557,
|
| 806 |
+
"learning_rate": 0.00012092307692307691,
|
| 807 |
+
"loss": 2.8189,
|
| 808 |
+
"step": 455
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 46.0,
|
| 812 |
+
"grad_norm": 0.6120209097862244,
|
| 813 |
+
"learning_rate": 0.00012034615384615384,
|
| 814 |
+
"loss": 2.8075,
|
| 815 |
+
"step": 460
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 46.5,
|
| 819 |
+
"grad_norm": 0.996507465839386,
|
| 820 |
+
"learning_rate": 0.00011976923076923076,
|
| 821 |
+
"loss": 2.8318,
|
| 822 |
+
"step": 465
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 47.0,
|
| 826 |
+
"grad_norm": 7.280458927154541,
|
| 827 |
+
"learning_rate": 0.00011919230769230767,
|
| 828 |
+
"loss": 2.871,
|
| 829 |
+
"step": 470
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 47.5,
|
| 833 |
+
"grad_norm": 0.8332684636116028,
|
| 834 |
+
"learning_rate": 0.0001186153846153846,
|
| 835 |
+
"loss": 2.8376,
|
| 836 |
+
"step": 475
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 47.5,
|
| 840 |
+
"eval_loss": 2.900068998336792,
|
| 841 |
+
"eval_runtime": 1.0322,
|
| 842 |
+
"eval_samples_per_second": 23.251,
|
| 843 |
+
"eval_steps_per_second": 0.969,
|
| 844 |
+
"eval_wer": 1.0,
|
| 845 |
+
"step": 475
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 48.0,
|
| 849 |
+
"grad_norm": 0.6555355191230774,
|
| 850 |
+
"learning_rate": 0.00011803846153846153,
|
| 851 |
+
"loss": 2.7954,
|
| 852 |
+
"step": 480
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 48.5,
|
| 856 |
+
"grad_norm": 1.127866268157959,
|
| 857 |
+
"learning_rate": 0.00011746153846153845,
|
| 858 |
+
"loss": 2.8494,
|
| 859 |
+
"step": 485
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 49.0,
|
| 863 |
+
"grad_norm": 0.7961714863777161,
|
| 864 |
+
"learning_rate": 0.00011688461538461538,
|
| 865 |
+
"loss": 2.8446,
|
| 866 |
+
"step": 490
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 49.5,
|
| 870 |
+
"grad_norm": 1.9832100868225098,
|
| 871 |
+
"learning_rate": 0.00011630769230769229,
|
| 872 |
+
"loss": 2.8353,
|
| 873 |
+
"step": 495
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 50.0,
|
| 877 |
+
"grad_norm": 0.9229313731193542,
|
| 878 |
+
"learning_rate": 0.00011573076923076922,
|
| 879 |
+
"loss": 2.8148,
|
| 880 |
+
"step": 500
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 50.0,
|
| 884 |
+
"eval_loss": 2.8878333568573,
|
| 885 |
+
"eval_runtime": 1.0279,
|
| 886 |
+
"eval_samples_per_second": 23.349,
|
| 887 |
+
"eval_steps_per_second": 0.973,
|
| 888 |
+
"eval_wer": 1.0,
|
| 889 |
+
"step": 500
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 50.5,
|
| 893 |
+
"grad_norm": 2.113555669784546,
|
| 894 |
+
"learning_rate": 0.00011515384615384614,
|
| 895 |
+
"loss": 2.816,
|
| 896 |
+
"step": 505
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 51.0,
|
| 900 |
+
"grad_norm": 2.10042667388916,
|
| 901 |
+
"learning_rate": 0.00011457692307692307,
|
| 902 |
+
"loss": 2.8544,
|
| 903 |
+
"step": 510
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 51.5,
|
| 907 |
+
"grad_norm": 0.48272839188575745,
|
| 908 |
+
"learning_rate": 0.00011399999999999999,
|
| 909 |
+
"loss": 2.8207,
|
| 910 |
+
"step": 515
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 52.0,
|
| 914 |
+
"grad_norm": 0.9009172320365906,
|
| 915 |
+
"learning_rate": 0.00011342307692307692,
|
| 916 |
+
"loss": 2.8008,
|
| 917 |
+
"step": 520
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 52.5,
|
| 921 |
+
"grad_norm": 1.0341640710830688,
|
| 922 |
+
"learning_rate": 0.00011284615384615384,
|
| 923 |
+
"loss": 2.8057,
|
| 924 |
+
"step": 525
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 52.5,
|
| 928 |
+
"eval_loss": 2.8624706268310547,
|
| 929 |
+
"eval_runtime": 1.037,
|
| 930 |
+
"eval_samples_per_second": 23.144,
|
| 931 |
+
"eval_steps_per_second": 0.964,
|
| 932 |
+
"eval_wer": 1.0,
|
| 933 |
+
"step": 525
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 53.0,
|
| 937 |
+
"grad_norm": 1.3395497798919678,
|
| 938 |
+
"learning_rate": 0.00011226923076923077,
|
| 939 |
+
"loss": 2.7866,
|
| 940 |
+
"step": 530
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 53.5,
|
| 944 |
+
"grad_norm": 0.3619355261325836,
|
| 945 |
+
"learning_rate": 0.00011169230769230768,
|
| 946 |
+
"loss": 2.7779,
|
| 947 |
+
"step": 535
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 54.0,
|
| 951 |
+
"grad_norm": 1.4029289484024048,
|
| 952 |
+
"learning_rate": 0.0001111153846153846,
|
| 953 |
+
"loss": 2.789,
|
| 954 |
+
"step": 540
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 54.5,
|
| 958 |
+
"grad_norm": 0.29736635088920593,
|
| 959 |
+
"learning_rate": 0.00011053846153846152,
|
| 960 |
+
"loss": 2.7452,
|
| 961 |
+
"step": 545
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 55.0,
|
| 965 |
+
"grad_norm": 1.7570823431015015,
|
| 966 |
+
"learning_rate": 0.00010996153846153845,
|
| 967 |
+
"loss": 2.7268,
|
| 968 |
+
"step": 550
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 55.0,
|
| 972 |
+
"eval_loss": 2.819674253463745,
|
| 973 |
+
"eval_runtime": 1.0343,
|
| 974 |
+
"eval_samples_per_second": 23.205,
|
| 975 |
+
"eval_steps_per_second": 0.967,
|
| 976 |
+
"eval_wer": 1.0,
|
| 977 |
+
"step": 550
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 55.5,
|
| 981 |
+
"grad_norm": 0.3762887418270111,
|
| 982 |
+
"learning_rate": 0.00010938461538461537,
|
| 983 |
+
"loss": 2.7224,
|
| 984 |
+
"step": 555
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 56.0,
|
| 988 |
+
"grad_norm": 1.0835281610488892,
|
| 989 |
+
"learning_rate": 0.0001088076923076923,
|
| 990 |
+
"loss": 2.7022,
|
| 991 |
+
"step": 560
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 56.5,
|
| 995 |
+
"grad_norm": 1.721433401107788,
|
| 996 |
+
"learning_rate": 0.00010823076923076922,
|
| 997 |
+
"loss": 2.6927,
|
| 998 |
+
"step": 565
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 57.0,
|
| 1002 |
+
"grad_norm": 2.9872403144836426,
|
| 1003 |
+
"learning_rate": 0.00010765384615384615,
|
| 1004 |
+
"loss": 2.7924,
|
| 1005 |
+
"step": 570
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 57.5,
|
| 1009 |
+
"grad_norm": 0.5493649840354919,
|
| 1010 |
+
"learning_rate": 0.00010707692307692306,
|
| 1011 |
+
"loss": 2.6252,
|
| 1012 |
+
"step": 575
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 57.5,
|
| 1016 |
+
"eval_loss": 2.807591676712036,
|
| 1017 |
+
"eval_runtime": 1.0323,
|
| 1018 |
+
"eval_samples_per_second": 23.25,
|
| 1019 |
+
"eval_steps_per_second": 0.969,
|
| 1020 |
+
"eval_wer": 1.0,
|
| 1021 |
+
"step": 575
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 58.0,
|
| 1025 |
+
"grad_norm": 1.2353851795196533,
|
| 1026 |
+
"learning_rate": 0.00010649999999999999,
|
| 1027 |
+
"loss": 2.6458,
|
| 1028 |
+
"step": 580
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 58.5,
|
| 1032 |
+
"grad_norm": 0.7240511775016785,
|
| 1033 |
+
"learning_rate": 0.00010592307692307691,
|
| 1034 |
+
"loss": 2.5911,
|
| 1035 |
+
"step": 585
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 59.0,
|
| 1039 |
+
"grad_norm": 0.9982340335845947,
|
| 1040 |
+
"learning_rate": 0.00010534615384615384,
|
| 1041 |
+
"loss": 2.6489,
|
| 1042 |
+
"step": 590
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 59.5,
|
| 1046 |
+
"grad_norm": 0.6784680485725403,
|
| 1047 |
+
"learning_rate": 0.00010476923076923076,
|
| 1048 |
+
"loss": 2.5169,
|
| 1049 |
+
"step": 595
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 60.0,
|
| 1053 |
+
"grad_norm": 1.9756778478622437,
|
| 1054 |
+
"learning_rate": 0.00010419230769230769,
|
| 1055 |
+
"loss": 2.5511,
|
| 1056 |
+
"step": 600
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 60.0,
|
| 1060 |
+
"eval_loss": 2.615316152572632,
|
| 1061 |
+
"eval_runtime": 1.0274,
|
| 1062 |
+
"eval_samples_per_second": 23.361,
|
| 1063 |
+
"eval_steps_per_second": 0.973,
|
| 1064 |
+
"eval_wer": 1.0056338028169014,
|
| 1065 |
+
"step": 600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 60.5,
|
| 1069 |
+
"grad_norm": 1.3284317255020142,
|
| 1070 |
+
"learning_rate": 0.00010361538461538462,
|
| 1071 |
+
"loss": 2.4731,
|
| 1072 |
+
"step": 605
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 61.0,
|
| 1076 |
+
"grad_norm": 1.3110464811325073,
|
| 1077 |
+
"learning_rate": 0.00010303846153846154,
|
| 1078 |
+
"loss": 2.4817,
|
| 1079 |
+
"step": 610
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 61.5,
|
| 1083 |
+
"grad_norm": 1.003812551498413,
|
| 1084 |
+
"learning_rate": 0.00010246153846153844,
|
| 1085 |
+
"loss": 2.3945,
|
| 1086 |
+
"step": 615
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 62.0,
|
| 1090 |
+
"grad_norm": 1.148573398590088,
|
| 1091 |
+
"learning_rate": 0.00010188461538461537,
|
| 1092 |
+
"loss": 2.399,
|
| 1093 |
+
"step": 620
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 62.5,
|
| 1097 |
+
"grad_norm": 0.5585479736328125,
|
| 1098 |
+
"learning_rate": 0.00010130769230769229,
|
| 1099 |
+
"loss": 2.323,
|
| 1100 |
+
"step": 625
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 62.5,
|
| 1104 |
+
"eval_loss": 2.4444546699523926,
|
| 1105 |
+
"eval_runtime": 1.0272,
|
| 1106 |
+
"eval_samples_per_second": 23.365,
|
| 1107 |
+
"eval_steps_per_second": 0.974,
|
| 1108 |
+
"eval_wer": 1.0169014084507042,
|
| 1109 |
+
"step": 625
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 63.0,
|
| 1113 |
+
"grad_norm": 2.2142958641052246,
|
| 1114 |
+
"learning_rate": 0.00010073076923076922,
|
| 1115 |
+
"loss": 2.2927,
|
| 1116 |
+
"step": 630
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 63.5,
|
| 1120 |
+
"grad_norm": 1.0168890953063965,
|
| 1121 |
+
"learning_rate": 0.00010015384615384614,
|
| 1122 |
+
"loss": 2.2108,
|
| 1123 |
+
"step": 635
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 64.0,
|
| 1127 |
+
"grad_norm": 1.312639832496643,
|
| 1128 |
+
"learning_rate": 9.957692307692307e-05,
|
| 1129 |
+
"loss": 2.1866,
|
| 1130 |
+
"step": 640
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 64.5,
|
| 1134 |
+
"grad_norm": 0.5699294209480286,
|
| 1135 |
+
"learning_rate": 9.9e-05,
|
| 1136 |
+
"loss": 2.1114,
|
| 1137 |
+
"step": 645
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 65.0,
|
| 1141 |
+
"grad_norm": 1.4273818731307983,
|
| 1142 |
+
"learning_rate": 9.842307692307692e-05,
|
| 1143 |
+
"loss": 2.1119,
|
| 1144 |
+
"step": 650
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 65.0,
|
| 1148 |
+
"eval_loss": 2.2476181983947754,
|
| 1149 |
+
"eval_runtime": 1.0519,
|
| 1150 |
+
"eval_samples_per_second": 22.815,
|
| 1151 |
+
"eval_steps_per_second": 0.951,
|
| 1152 |
+
"eval_wer": 1.1183098591549296,
|
| 1153 |
+
"step": 650
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 65.5,
|
| 1157 |
+
"grad_norm": 0.5214980244636536,
|
| 1158 |
+
"learning_rate": 9.784615384615383e-05,
|
| 1159 |
+
"loss": 2.0414,
|
| 1160 |
+
"step": 655
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 66.0,
|
| 1164 |
+
"grad_norm": 2.480297803878784,
|
| 1165 |
+
"learning_rate": 9.726923076923076e-05,
|
| 1166 |
+
"loss": 2.0609,
|
| 1167 |
+
"step": 660
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 66.5,
|
| 1171 |
+
"grad_norm": 3.5270726680755615,
|
| 1172 |
+
"learning_rate": 9.669230769230768e-05,
|
| 1173 |
+
"loss": 1.9963,
|
| 1174 |
+
"step": 665
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 67.0,
|
| 1178 |
+
"grad_norm": 14.827882766723633,
|
| 1179 |
+
"learning_rate": 9.611538461538461e-05,
|
| 1180 |
+
"loss": 1.9333,
|
| 1181 |
+
"step": 670
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 67.5,
|
| 1185 |
+
"grad_norm": 1.1005451679229736,
|
| 1186 |
+
"learning_rate": 9.553846153846153e-05,
|
| 1187 |
+
"loss": 1.8514,
|
| 1188 |
+
"step": 675
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 67.5,
|
| 1192 |
+
"eval_loss": 2.173093318939209,
|
| 1193 |
+
"eval_runtime": 1.033,
|
| 1194 |
+
"eval_samples_per_second": 23.233,
|
| 1195 |
+
"eval_steps_per_second": 0.968,
|
| 1196 |
+
"eval_wer": 1.095774647887324,
|
| 1197 |
+
"step": 675
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 68.0,
|
| 1201 |
+
"grad_norm": 1.5897767543792725,
|
| 1202 |
+
"learning_rate": 9.496153846153846e-05,
|
| 1203 |
+
"loss": 1.9986,
|
| 1204 |
+
"step": 680
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 68.5,
|
| 1208 |
+
"grad_norm": 0.8863438963890076,
|
| 1209 |
+
"learning_rate": 9.438461538461539e-05,
|
| 1210 |
+
"loss": 1.8067,
|
| 1211 |
+
"step": 685
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 69.0,
|
| 1215 |
+
"grad_norm": 1.305874228477478,
|
| 1216 |
+
"learning_rate": 9.380769230769231e-05,
|
| 1217 |
+
"loss": 1.7975,
|
| 1218 |
+
"step": 690
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 69.5,
|
| 1222 |
+
"grad_norm": 0.6541560292243958,
|
| 1223 |
+
"learning_rate": 9.323076923076921e-05,
|
| 1224 |
+
"loss": 1.7655,
|
| 1225 |
+
"step": 695
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 70.0,
|
| 1229 |
+
"grad_norm": 1.056104063987732,
|
| 1230 |
+
"learning_rate": 9.265384615384614e-05,
|
| 1231 |
+
"loss": 1.7094,
|
| 1232 |
+
"step": 700
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 70.0,
|
| 1236 |
+
"eval_loss": 2.0642001628875732,
|
| 1237 |
+
"eval_runtime": 1.0377,
|
| 1238 |
+
"eval_samples_per_second": 23.129,
|
| 1239 |
+
"eval_steps_per_second": 0.964,
|
| 1240 |
+
"eval_wer": 1.0309859154929577,
|
| 1241 |
+
"step": 700
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 70.5,
|
| 1245 |
+
"grad_norm": 0.5228053331375122,
|
| 1246 |
+
"learning_rate": 9.207692307692306e-05,
|
| 1247 |
+
"loss": 1.6764,
|
| 1248 |
+
"step": 705
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 71.0,
|
| 1252 |
+
"grad_norm": 6.9655256271362305,
|
| 1253 |
+
"learning_rate": 9.149999999999999e-05,
|
| 1254 |
+
"loss": 1.7414,
|
| 1255 |
+
"step": 710
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 71.5,
|
| 1259 |
+
"grad_norm": 0.6360809206962585,
|
| 1260 |
+
"learning_rate": 9.092307692307691e-05,
|
| 1261 |
+
"loss": 1.6232,
|
| 1262 |
+
"step": 715
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 72.0,
|
| 1266 |
+
"grad_norm": 1.2141180038452148,
|
| 1267 |
+
"learning_rate": 9.034615384615384e-05,
|
| 1268 |
+
"loss": 1.6497,
|
| 1269 |
+
"step": 720
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 72.5,
|
| 1273 |
+
"grad_norm": 0.874902606010437,
|
| 1274 |
+
"learning_rate": 8.976923076923077e-05,
|
| 1275 |
+
"loss": 1.6069,
|
| 1276 |
+
"step": 725
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 72.5,
|
| 1280 |
+
"eval_loss": 2.0792412757873535,
|
| 1281 |
+
"eval_runtime": 1.0243,
|
| 1282 |
+
"eval_samples_per_second": 23.431,
|
| 1283 |
+
"eval_steps_per_second": 0.976,
|
| 1284 |
+
"eval_wer": 1.0788732394366196,
|
| 1285 |
+
"step": 725
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 73.0,
|
| 1289 |
+
"grad_norm": 0.9335172176361084,
|
| 1290 |
+
"learning_rate": 8.919230769230769e-05,
|
| 1291 |
+
"loss": 1.4947,
|
| 1292 |
+
"step": 730
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 73.5,
|
| 1296 |
+
"grad_norm": 1.299177885055542,
|
| 1297 |
+
"learning_rate": 8.861538461538462e-05,
|
| 1298 |
+
"loss": 1.5304,
|
| 1299 |
+
"step": 735
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 74.0,
|
| 1303 |
+
"grad_norm": 1.6317135095596313,
|
| 1304 |
+
"learning_rate": 8.803846153846153e-05,
|
| 1305 |
+
"loss": 1.5218,
|
| 1306 |
+
"step": 740
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 74.5,
|
| 1310 |
+
"grad_norm": 0.8083561062812805,
|
| 1311 |
+
"learning_rate": 8.746153846153845e-05,
|
| 1312 |
+
"loss": 1.5259,
|
| 1313 |
+
"step": 745
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 75.0,
|
| 1317 |
+
"grad_norm": 1.805677890777588,
|
| 1318 |
+
"learning_rate": 8.688461538461538e-05,
|
| 1319 |
+
"loss": 1.4663,
|
| 1320 |
+
"step": 750
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 75.0,
|
| 1324 |
+
"eval_loss": 2.0323963165283203,
|
| 1325 |
+
"eval_runtime": 1.0407,
|
| 1326 |
+
"eval_samples_per_second": 23.062,
|
| 1327 |
+
"eval_steps_per_second": 0.961,
|
| 1328 |
+
"eval_wer": 1.036619718309859,
|
| 1329 |
+
"step": 750
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 75.5,
|
| 1333 |
+
"grad_norm": 0.8463692665100098,
|
| 1334 |
+
"learning_rate": 8.63076923076923e-05,
|
| 1335 |
+
"loss": 1.4244,
|
| 1336 |
+
"step": 755
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 76.0,
|
| 1340 |
+
"grad_norm": 2.091686248779297,
|
| 1341 |
+
"learning_rate": 8.573076923076923e-05,
|
| 1342 |
+
"loss": 1.3791,
|
| 1343 |
+
"step": 760
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 76.5,
|
| 1347 |
+
"grad_norm": 0.7040625810623169,
|
| 1348 |
+
"learning_rate": 8.515384615384614e-05,
|
| 1349 |
+
"loss": 1.3495,
|
| 1350 |
+
"step": 765
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 77.0,
|
| 1354 |
+
"grad_norm": 1.7725024223327637,
|
| 1355 |
+
"learning_rate": 8.457692307692307e-05,
|
| 1356 |
+
"loss": 1.3497,
|
| 1357 |
+
"step": 770
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 77.5,
|
| 1361 |
+
"grad_norm": 0.808942437171936,
|
| 1362 |
+
"learning_rate": 8.4e-05,
|
| 1363 |
+
"loss": 1.288,
|
| 1364 |
+
"step": 775
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 77.5,
|
| 1368 |
+
"eval_loss": 2.0642640590667725,
|
| 1369 |
+
"eval_runtime": 1.0443,
|
| 1370 |
+
"eval_samples_per_second": 22.982,
|
| 1371 |
+
"eval_steps_per_second": 0.958,
|
| 1372 |
+
"eval_wer": 1.0929577464788733,
|
| 1373 |
+
"step": 775
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 78.0,
|
| 1377 |
+
"grad_norm": 3.843997001647949,
|
| 1378 |
+
"learning_rate": 8.342307692307691e-05,
|
| 1379 |
+
"loss": 1.2597,
|
| 1380 |
+
"step": 780
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 78.5,
|
| 1384 |
+
"grad_norm": 0.9082187414169312,
|
| 1385 |
+
"learning_rate": 8.284615384615383e-05,
|
| 1386 |
+
"loss": 1.2702,
|
| 1387 |
+
"step": 785
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 79.0,
|
| 1391 |
+
"grad_norm": 1.4159339666366577,
|
| 1392 |
+
"learning_rate": 8.226923076923076e-05,
|
| 1393 |
+
"loss": 1.2833,
|
| 1394 |
+
"step": 790
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 79.5,
|
| 1398 |
+
"grad_norm": 1.0848701000213623,
|
| 1399 |
+
"learning_rate": 8.169230769230768e-05,
|
| 1400 |
+
"loss": 1.2117,
|
| 1401 |
+
"step": 795
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 80.0,
|
| 1405 |
+
"grad_norm": 2.275663137435913,
|
| 1406 |
+
"learning_rate": 8.111538461538461e-05,
|
| 1407 |
+
"loss": 1.262,
|
| 1408 |
+
"step": 800
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 80.0,
|
| 1412 |
+
"eval_loss": 2.084003210067749,
|
| 1413 |
+
"eval_runtime": 1.0408,
|
| 1414 |
+
"eval_samples_per_second": 23.059,
|
| 1415 |
+
"eval_steps_per_second": 0.961,
|
| 1416 |
+
"eval_wer": 1.076056338028169,
|
| 1417 |
+
"step": 800
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 80.5,
|
| 1421 |
+
"grad_norm": 0.9842613339424133,
|
| 1422 |
+
"learning_rate": 8.053846153846154e-05,
|
| 1423 |
+
"loss": 1.2799,
|
| 1424 |
+
"step": 805
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 81.0,
|
| 1428 |
+
"grad_norm": 20.336593627929688,
|
| 1429 |
+
"learning_rate": 7.996153846153846e-05,
|
| 1430 |
+
"loss": 1.2903,
|
| 1431 |
+
"step": 810
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 81.5,
|
| 1435 |
+
"grad_norm": 0.8291641473770142,
|
| 1436 |
+
"learning_rate": 7.938461538461539e-05,
|
| 1437 |
+
"loss": 1.1215,
|
| 1438 |
+
"step": 815
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 82.0,
|
| 1442 |
+
"grad_norm": 1.6971830129623413,
|
| 1443 |
+
"learning_rate": 7.88076923076923e-05,
|
| 1444 |
+
"loss": 1.1435,
|
| 1445 |
+
"step": 820
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 82.5,
|
| 1449 |
+
"grad_norm": 0.69861900806427,
|
| 1450 |
+
"learning_rate": 7.823076923076923e-05,
|
| 1451 |
+
"loss": 1.043,
|
| 1452 |
+
"step": 825
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 82.5,
|
| 1456 |
+
"eval_loss": 2.149214506149292,
|
| 1457 |
+
"eval_runtime": 1.0296,
|
| 1458 |
+
"eval_samples_per_second": 23.311,
|
| 1459 |
+
"eval_steps_per_second": 0.971,
|
| 1460 |
+
"eval_wer": 1.0901408450704226,
|
| 1461 |
+
"step": 825
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 83.0,
|
| 1465 |
+
"grad_norm": 1.7208884954452515,
|
| 1466 |
+
"learning_rate": 7.776923076923076e-05,
|
| 1467 |
+
"loss": 1.203,
|
| 1468 |
+
"step": 830
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 83.5,
|
| 1472 |
+
"grad_norm": 0.8559800982475281,
|
| 1473 |
+
"learning_rate": 7.719230769230768e-05,
|
| 1474 |
+
"loss": 1.0825,
|
| 1475 |
+
"step": 835
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 84.0,
|
| 1479 |
+
"grad_norm": 1.6605381965637207,
|
| 1480 |
+
"learning_rate": 7.661538461538461e-05,
|
| 1481 |
+
"loss": 1.1121,
|
| 1482 |
+
"step": 840
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 84.5,
|
| 1486 |
+
"grad_norm": 1.077573537826538,
|
| 1487 |
+
"learning_rate": 7.603846153846154e-05,
|
| 1488 |
+
"loss": 1.0145,
|
| 1489 |
+
"step": 845
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 85.0,
|
| 1493 |
+
"grad_norm": 2.7091293334960938,
|
| 1494 |
+
"learning_rate": 7.546153846153846e-05,
|
| 1495 |
+
"loss": 1.0501,
|
| 1496 |
+
"step": 850
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 85.0,
|
| 1500 |
+
"eval_loss": 2.177476644515991,
|
| 1501 |
+
"eval_runtime": 1.0552,
|
| 1502 |
+
"eval_samples_per_second": 22.744,
|
| 1503 |
+
"eval_steps_per_second": 0.948,
|
| 1504 |
+
"eval_wer": 1.0591549295774647,
|
| 1505 |
+
"step": 850
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 85.5,
|
| 1509 |
+
"grad_norm": 1.3562541007995605,
|
| 1510 |
+
"learning_rate": 7.488461538461539e-05,
|
| 1511 |
+
"loss": 1.1098,
|
| 1512 |
+
"step": 855
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 86.0,
|
| 1516 |
+
"grad_norm": 2.6526386737823486,
|
| 1517 |
+
"learning_rate": 7.43076923076923e-05,
|
| 1518 |
+
"loss": 0.8642,
|
| 1519 |
+
"step": 860
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 86.5,
|
| 1523 |
+
"grad_norm": 1.1710244417190552,
|
| 1524 |
+
"learning_rate": 7.373076923076922e-05,
|
| 1525 |
+
"loss": 0.9004,
|
| 1526 |
+
"step": 865
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 87.0,
|
| 1530 |
+
"grad_norm": 2.9008164405822754,
|
| 1531 |
+
"learning_rate": 7.315384615384615e-05,
|
| 1532 |
+
"loss": 1.037,
|
| 1533 |
+
"step": 870
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 87.5,
|
| 1537 |
+
"grad_norm": 0.6306678056716919,
|
| 1538 |
+
"learning_rate": 7.257692307692308e-05,
|
| 1539 |
+
"loss": 0.9726,
|
| 1540 |
+
"step": 875
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 87.5,
|
| 1544 |
+
"eval_loss": 2.176731586456299,
|
| 1545 |
+
"eval_runtime": 1.0313,
|
| 1546 |
+
"eval_samples_per_second": 23.271,
|
| 1547 |
+
"eval_steps_per_second": 0.97,
|
| 1548 |
+
"eval_wer": 1.028169014084507,
|
| 1549 |
+
"step": 875
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 88.0,
|
| 1553 |
+
"grad_norm": 1.6984366178512573,
|
| 1554 |
+
"learning_rate": 7.199999999999999e-05,
|
| 1555 |
+
"loss": 1.1201,
|
| 1556 |
+
"step": 880
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 88.5,
|
| 1560 |
+
"grad_norm": 0.803970992565155,
|
| 1561 |
+
"learning_rate": 7.142307692307691e-05,
|
| 1562 |
+
"loss": 0.908,
|
| 1563 |
+
"step": 885
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 89.0,
|
| 1567 |
+
"grad_norm": 2.103391408920288,
|
| 1568 |
+
"learning_rate": 7.084615384615384e-05,
|
| 1569 |
+
"loss": 0.8684,
|
| 1570 |
+
"step": 890
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 89.5,
|
| 1574 |
+
"grad_norm": 0.9575273990631104,
|
| 1575 |
+
"learning_rate": 7.026923076923077e-05,
|
| 1576 |
+
"loss": 0.9791,
|
| 1577 |
+
"step": 895
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 90.0,
|
| 1581 |
+
"grad_norm": 3.000880479812622,
|
| 1582 |
+
"learning_rate": 6.969230769230768e-05,
|
| 1583 |
+
"loss": 0.8079,
|
| 1584 |
+
"step": 900
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 90.0,
|
| 1588 |
+
"eval_loss": 2.1965668201446533,
|
| 1589 |
+
"eval_runtime": 1.0433,
|
| 1590 |
+
"eval_samples_per_second": 23.003,
|
| 1591 |
+
"eval_steps_per_second": 0.958,
|
| 1592 |
+
"eval_wer": 0.9943661971830986,
|
| 1593 |
+
"step": 900
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 90.5,
|
| 1597 |
+
"grad_norm": 0.6576473712921143,
|
| 1598 |
+
"learning_rate": 6.91153846153846e-05,
|
| 1599 |
+
"loss": 0.846,
|
| 1600 |
+
"step": 905
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 91.0,
|
| 1604 |
+
"grad_norm": 2.2526416778564453,
|
| 1605 |
+
"learning_rate": 6.853846153846153e-05,
|
| 1606 |
+
"loss": 0.8868,
|
| 1607 |
+
"step": 910
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 91.5,
|
| 1611 |
+
"grad_norm": 0.5678216814994812,
|
| 1612 |
+
"learning_rate": 6.796153846153845e-05,
|
| 1613 |
+
"loss": 0.8925,
|
| 1614 |
+
"step": 915
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 92.0,
|
| 1618 |
+
"grad_norm": 2.549266815185547,
|
| 1619 |
+
"learning_rate": 6.738461538461538e-05,
|
| 1620 |
+
"loss": 1.0163,
|
| 1621 |
+
"step": 920
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 92.5,
|
| 1625 |
+
"grad_norm": 0.7736966013908386,
|
| 1626 |
+
"learning_rate": 6.68076923076923e-05,
|
| 1627 |
+
"loss": 0.7198,
|
| 1628 |
+
"step": 925
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 92.5,
|
| 1632 |
+
"eval_loss": 2.2433066368103027,
|
| 1633 |
+
"eval_runtime": 1.0523,
|
| 1634 |
+
"eval_samples_per_second": 22.808,
|
| 1635 |
+
"eval_steps_per_second": 0.95,
|
| 1636 |
+
"eval_wer": 1.0028169014084507,
|
| 1637 |
+
"step": 925
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 93.0,
|
| 1641 |
+
"grad_norm": 3.742175817489624,
|
| 1642 |
+
"learning_rate": 6.623076923076923e-05,
|
| 1643 |
+
"loss": 1.011,
|
| 1644 |
+
"step": 930
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 93.5,
|
| 1648 |
+
"grad_norm": 0.748150110244751,
|
| 1649 |
+
"learning_rate": 6.565384615384616e-05,
|
| 1650 |
+
"loss": 0.7659,
|
| 1651 |
+
"step": 935
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 94.0,
|
| 1655 |
+
"grad_norm": 2.121845006942749,
|
| 1656 |
+
"learning_rate": 6.507692307692307e-05,
|
| 1657 |
+
"loss": 0.7862,
|
| 1658 |
+
"step": 940
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 94.5,
|
| 1662 |
+
"grad_norm": 0.7966519594192505,
|
| 1663 |
+
"learning_rate": 6.45e-05,
|
| 1664 |
+
"loss": 0.8271,
|
| 1665 |
+
"step": 945
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 95.0,
|
| 1669 |
+
"grad_norm": 1.6206731796264648,
|
| 1670 |
+
"learning_rate": 6.392307692307692e-05,
|
| 1671 |
+
"loss": 0.6312,
|
| 1672 |
+
"step": 950
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 95.0,
|
| 1676 |
+
"eval_loss": 2.309884786605835,
|
| 1677 |
+
"eval_runtime": 1.062,
|
| 1678 |
+
"eval_samples_per_second": 22.599,
|
| 1679 |
+
"eval_steps_per_second": 0.942,
|
| 1680 |
+
"eval_wer": 0.9971830985915493,
|
| 1681 |
+
"step": 950
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 95.5,
|
| 1685 |
+
"grad_norm": 3.240893602371216,
|
| 1686 |
+
"learning_rate": 6.334615384615385e-05,
|
| 1687 |
+
"loss": 0.723,
|
| 1688 |
+
"step": 955
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 96.0,
|
| 1692 |
+
"grad_norm": 1.4926756620407104,
|
| 1693 |
+
"learning_rate": 6.276923076923076e-05,
|
| 1694 |
+
"loss": 0.7344,
|
| 1695 |
+
"step": 960
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 96.5,
|
| 1699 |
+
"grad_norm": 0.8542086482048035,
|
| 1700 |
+
"learning_rate": 6.219230769230769e-05,
|
| 1701 |
+
"loss": 0.7649,
|
| 1702 |
+
"step": 965
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 97.0,
|
| 1706 |
+
"grad_norm": 2.2014851570129395,
|
| 1707 |
+
"learning_rate": 6.161538461538461e-05,
|
| 1708 |
+
"loss": 0.6969,
|
| 1709 |
+
"step": 970
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 97.5,
|
| 1713 |
+
"grad_norm": 0.6612327694892883,
|
| 1714 |
+
"learning_rate": 6.103846153846154e-05,
|
| 1715 |
+
"loss": 0.6336,
|
| 1716 |
+
"step": 975
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 97.5,
|
| 1720 |
+
"eval_loss": 2.3546626567840576,
|
| 1721 |
+
"eval_runtime": 1.0484,
|
| 1722 |
+
"eval_samples_per_second": 22.893,
|
| 1723 |
+
"eval_steps_per_second": 0.954,
|
| 1724 |
+
"eval_wer": 0.9971830985915493,
|
| 1725 |
+
"step": 975
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 98.0,
|
| 1729 |
+
"grad_norm": 2.117011547088623,
|
| 1730 |
+
"learning_rate": 6.0461538461538456e-05,
|
| 1731 |
+
"loss": 0.7537,
|
| 1732 |
+
"step": 980
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 98.5,
|
| 1736 |
+
"grad_norm": 8.142460823059082,
|
| 1737 |
+
"learning_rate": 5.988461538461538e-05,
|
| 1738 |
+
"loss": 0.6593,
|
| 1739 |
+
"step": 985
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 99.0,
|
| 1743 |
+
"grad_norm": 2.6468851566314697,
|
| 1744 |
+
"learning_rate": 5.93076923076923e-05,
|
| 1745 |
+
"loss": 0.8069,
|
| 1746 |
+
"step": 990
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 99.5,
|
| 1750 |
+
"grad_norm": 1.392821192741394,
|
| 1751 |
+
"learning_rate": 5.8730769230769226e-05,
|
| 1752 |
+
"loss": 0.746,
|
| 1753 |
+
"step": 995
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 100.0,
|
| 1757 |
+
"grad_norm": 2.0805888175964355,
|
| 1758 |
+
"learning_rate": 5.8153846153846145e-05,
|
| 1759 |
+
"loss": 0.9073,
|
| 1760 |
+
"step": 1000
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 100.0,
|
| 1764 |
+
"eval_loss": 2.350856304168701,
|
| 1765 |
+
"eval_runtime": 1.0707,
|
| 1766 |
+
"eval_samples_per_second": 22.414,
|
| 1767 |
+
"eval_steps_per_second": 0.934,
|
| 1768 |
+
"eval_wer": 0.9943661971830986,
|
| 1769 |
+
"step": 1000
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 100.5,
|
| 1773 |
+
"grad_norm": 18.686534881591797,
|
| 1774 |
+
"learning_rate": 5.757692307692307e-05,
|
| 1775 |
+
"loss": 0.7907,
|
| 1776 |
+
"step": 1005
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 101.0,
|
| 1780 |
+
"grad_norm": 1.7688676118850708,
|
| 1781 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 1782 |
+
"loss": 0.5693,
|
| 1783 |
+
"step": 1010
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 101.5,
|
| 1787 |
+
"grad_norm": 0.9006216526031494,
|
| 1788 |
+
"learning_rate": 5.642307692307692e-05,
|
| 1789 |
+
"loss": 0.6408,
|
| 1790 |
+
"step": 1015
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 102.0,
|
| 1794 |
+
"grad_norm": 2.382704496383667,
|
| 1795 |
+
"learning_rate": 5.584615384615384e-05,
|
| 1796 |
+
"loss": 0.7203,
|
| 1797 |
+
"step": 1020
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 102.5,
|
| 1801 |
+
"grad_norm": 0.8852857351303101,
|
| 1802 |
+
"learning_rate": 5.526923076923076e-05,
|
| 1803 |
+
"loss": 0.6431,
|
| 1804 |
+
"step": 1025
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 102.5,
|
| 1808 |
+
"eval_loss": 2.4202942848205566,
|
| 1809 |
+
"eval_runtime": 1.0529,
|
| 1810 |
+
"eval_samples_per_second": 22.794,
|
| 1811 |
+
"eval_steps_per_second": 0.95,
|
| 1812 |
+
"eval_wer": 1.0056338028169014,
|
| 1813 |
+
"step": 1025
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 103.0,
|
| 1817 |
+
"grad_norm": 3.3610403537750244,
|
| 1818 |
+
"learning_rate": 5.4692307692307686e-05,
|
| 1819 |
+
"loss": 0.6476,
|
| 1820 |
+
"step": 1030
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 103.5,
|
| 1824 |
+
"grad_norm": 0.8738270401954651,
|
| 1825 |
+
"learning_rate": 5.411538461538461e-05,
|
| 1826 |
+
"loss": 0.5492,
|
| 1827 |
+
"step": 1035
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 104.0,
|
| 1831 |
+
"grad_norm": 2.4251339435577393,
|
| 1832 |
+
"learning_rate": 5.353846153846153e-05,
|
| 1833 |
+
"loss": 0.6005,
|
| 1834 |
+
"step": 1040
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 104.5,
|
| 1838 |
+
"grad_norm": 0.7935536503791809,
|
| 1839 |
+
"learning_rate": 5.2961538461538456e-05,
|
| 1840 |
+
"loss": 0.5855,
|
| 1841 |
+
"step": 1045
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 105.0,
|
| 1845 |
+
"grad_norm": 2.805385112762451,
|
| 1846 |
+
"learning_rate": 5.238461538461538e-05,
|
| 1847 |
+
"loss": 0.62,
|
| 1848 |
+
"step": 1050
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 105.0,
|
| 1852 |
+
"eval_loss": 2.3933348655700684,
|
| 1853 |
+
"eval_runtime": 1.0674,
|
| 1854 |
+
"eval_samples_per_second": 22.485,
|
| 1855 |
+
"eval_steps_per_second": 0.937,
|
| 1856 |
+
"eval_wer": 0.9746478873239437,
|
| 1857 |
+
"step": 1050
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 105.5,
|
| 1861 |
+
"grad_norm": 1.2249245643615723,
|
| 1862 |
+
"learning_rate": 5.180769230769231e-05,
|
| 1863 |
+
"loss": 0.652,
|
| 1864 |
+
"step": 1055
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 106.0,
|
| 1868 |
+
"grad_norm": 1.2247533798217773,
|
| 1869 |
+
"learning_rate": 5.123076923076922e-05,
|
| 1870 |
+
"loss": 0.6108,
|
| 1871 |
+
"step": 1060
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 106.5,
|
| 1875 |
+
"grad_norm": 0.8812918663024902,
|
| 1876 |
+
"learning_rate": 5.0653846153846146e-05,
|
| 1877 |
+
"loss": 0.6453,
|
| 1878 |
+
"step": 1065
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 107.0,
|
| 1882 |
+
"grad_norm": 2.7638535499572754,
|
| 1883 |
+
"learning_rate": 5.007692307692307e-05,
|
| 1884 |
+
"loss": 0.568,
|
| 1885 |
+
"step": 1070
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 107.5,
|
| 1889 |
+
"grad_norm": 1.3182368278503418,
|
| 1890 |
+
"learning_rate": 4.95e-05,
|
| 1891 |
+
"loss": 0.708,
|
| 1892 |
+
"step": 1075
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 107.5,
|
| 1896 |
+
"eval_loss": 2.4381346702575684,
|
| 1897 |
+
"eval_runtime": 1.061,
|
| 1898 |
+
"eval_samples_per_second": 22.619,
|
| 1899 |
+
"eval_steps_per_second": 0.942,
|
| 1900 |
+
"eval_wer": 0.9690140845070423,
|
| 1901 |
+
"step": 1075
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 108.0,
|
| 1905 |
+
"grad_norm": 2.4760406017303467,
|
| 1906 |
+
"learning_rate": 4.8923076923076916e-05,
|
| 1907 |
+
"loss": 0.6171,
|
| 1908 |
+
"step": 1080
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 108.5,
|
| 1912 |
+
"grad_norm": 0.5409008264541626,
|
| 1913 |
+
"learning_rate": 4.834615384615384e-05,
|
| 1914 |
+
"loss": 0.5542,
|
| 1915 |
+
"step": 1085
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 109.0,
|
| 1919 |
+
"grad_norm": 1.675410509109497,
|
| 1920 |
+
"learning_rate": 4.776923076923077e-05,
|
| 1921 |
+
"loss": 0.6491,
|
| 1922 |
+
"step": 1090
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 109.5,
|
| 1926 |
+
"grad_norm": 0.8941754698753357,
|
| 1927 |
+
"learning_rate": 4.719230769230769e-05,
|
| 1928 |
+
"loss": 0.7266,
|
| 1929 |
+
"step": 1095
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 110.0,
|
| 1933 |
+
"grad_norm": 1.9851211309432983,
|
| 1934 |
+
"learning_rate": 4.6615384615384605e-05,
|
| 1935 |
+
"loss": 0.6729,
|
| 1936 |
+
"step": 1100
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 110.0,
|
| 1940 |
+
"eval_loss": 2.474308967590332,
|
| 1941 |
+
"eval_runtime": 1.0636,
|
| 1942 |
+
"eval_samples_per_second": 22.564,
|
| 1943 |
+
"eval_steps_per_second": 0.94,
|
| 1944 |
+
"eval_wer": 1.0,
|
| 1945 |
+
"step": 1100
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 110.5,
|
| 1949 |
+
"grad_norm": 0.677306592464447,
|
| 1950 |
+
"learning_rate": 4.603846153846153e-05,
|
| 1951 |
+
"loss": 0.7625,
|
| 1952 |
+
"step": 1105
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 111.0,
|
| 1956 |
+
"grad_norm": 2.572356700897217,
|
| 1957 |
+
"learning_rate": 4.546153846153846e-05,
|
| 1958 |
+
"loss": 0.5146,
|
| 1959 |
+
"step": 1110
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 111.5,
|
| 1963 |
+
"grad_norm": 1.2789101600646973,
|
| 1964 |
+
"learning_rate": 4.488461538461538e-05,
|
| 1965 |
+
"loss": 0.5504,
|
| 1966 |
+
"step": 1115
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 112.0,
|
| 1970 |
+
"grad_norm": 2.3920390605926514,
|
| 1971 |
+
"learning_rate": 4.430769230769231e-05,
|
| 1972 |
+
"loss": 0.4821,
|
| 1973 |
+
"step": 1120
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 112.5,
|
| 1977 |
+
"grad_norm": 1.219436764717102,
|
| 1978 |
+
"learning_rate": 4.373076923076923e-05,
|
| 1979 |
+
"loss": 0.5779,
|
| 1980 |
+
"step": 1125
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 112.5,
|
| 1984 |
+
"eval_loss": 2.492933988571167,
|
| 1985 |
+
"eval_runtime": 1.0274,
|
| 1986 |
+
"eval_samples_per_second": 23.36,
|
| 1987 |
+
"eval_steps_per_second": 0.973,
|
| 1988 |
+
"eval_wer": 0.9549295774647887,
|
| 1989 |
+
"step": 1125
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 113.0,
|
| 1993 |
+
"grad_norm": 3.558155059814453,
|
| 1994 |
+
"learning_rate": 4.315384615384615e-05,
|
| 1995 |
+
"loss": 0.4743,
|
| 1996 |
+
"step": 1130
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 113.5,
|
| 2000 |
+
"grad_norm": 0.9398171901702881,
|
| 2001 |
+
"learning_rate": 4.257692307692307e-05,
|
| 2002 |
+
"loss": 0.493,
|
| 2003 |
+
"step": 1135
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 114.0,
|
| 2007 |
+
"grad_norm": 4.514529705047607,
|
| 2008 |
+
"learning_rate": 4.2e-05,
|
| 2009 |
+
"loss": 0.4341,
|
| 2010 |
+
"step": 1140
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 114.5,
|
| 2014 |
+
"grad_norm": 1.015120029449463,
|
| 2015 |
+
"learning_rate": 4.142307692307692e-05,
|
| 2016 |
+
"loss": 0.5069,
|
| 2017 |
+
"step": 1145
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 115.0,
|
| 2021 |
+
"grad_norm": 2.043063163757324,
|
| 2022 |
+
"learning_rate": 4.084615384615384e-05,
|
| 2023 |
+
"loss": 0.6303,
|
| 2024 |
+
"step": 1150
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 115.0,
|
| 2028 |
+
"eval_loss": 2.5056331157684326,
|
| 2029 |
+
"eval_runtime": 1.0408,
|
| 2030 |
+
"eval_samples_per_second": 23.06,
|
| 2031 |
+
"eval_steps_per_second": 0.961,
|
| 2032 |
+
"eval_wer": 0.923943661971831,
|
| 2033 |
+
"step": 1150
|
| 2034 |
+
}
|
| 2035 |
+
],
|
| 2036 |
+
"logging_steps": 5,
|
| 2037 |
+
"max_steps": 1500,
|
| 2038 |
+
"num_input_tokens_seen": 0,
|
| 2039 |
+
"num_train_epochs": 150,
|
| 2040 |
+
"save_steps": 25,
|
| 2041 |
+
"stateful_callbacks": {
|
| 2042 |
+
"TrainerControl": {
|
| 2043 |
+
"args": {
|
| 2044 |
+
"should_epoch_stop": false,
|
| 2045 |
+
"should_evaluate": false,
|
| 2046 |
+
"should_log": false,
|
| 2047 |
+
"should_save": true,
|
| 2048 |
+
"should_training_stop": false
|
| 2049 |
+
},
|
| 2050 |
+
"attributes": {}
|
| 2051 |
+
}
|
| 2052 |
+
},
|
| 2053 |
+
"total_flos": 4.771505223996499e+18,
|
| 2054 |
+
"train_batch_size": 16,
|
| 2055 |
+
"trial_name": null,
|
| 2056 |
+
"trial_params": null
|
| 2057 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d684a542fe2c2e932faa66172a7cc1f48ba500827ae1b3d29c9b64338152a672
|
| 3 |
+
size 5112
|
vocab.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"i": 0, "3": 1, "d": 2, "k": 3, "y": 5, "x": 6, "4": 7, "t": 8, "o": 9, "s": 10, "e": 11, "b": 12, "u": 13, "2": 14, "h": 15, "c": 16, "m": 17, "1": 18, "n": 19, "\u00f1": 20, "l": 21, "'": 22, "a": 23, "r": 24, "|": 4, "[UNK]": 25, "[PAD]": 26}
|