Training in progress, step 20400, checkpoint
Browse files- last-checkpoint/config.json +109 -0
- last-checkpoint/model.safetensors +3 -0
- last-checkpoint/optimizer.pt +3 -0
- last-checkpoint/preprocessor_config.json +9 -0
- last-checkpoint/rng_state.pth +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/trainer_state.json +837 -0
- last-checkpoint/training_args.bin +3 -0
last-checkpoint/config.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 0.0,
|
| 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 |
+
"gradient_checkpointing": false,
|
| 57 |
+
"hidden_act": "gelu",
|
| 58 |
+
"hidden_dropout": 0.0,
|
| 59 |
+
"hidden_size": 1024,
|
| 60 |
+
"initializer_range": 0.02,
|
| 61 |
+
"intermediate_size": 4096,
|
| 62 |
+
"layer_norm_eps": 1e-05,
|
| 63 |
+
"layerdrop": 0.0,
|
| 64 |
+
"mask_feature_length": 10,
|
| 65 |
+
"mask_feature_min_masks": 0,
|
| 66 |
+
"mask_feature_prob": 0.0,
|
| 67 |
+
"mask_time_length": 10,
|
| 68 |
+
"mask_time_min_masks": 2,
|
| 69 |
+
"mask_time_prob": 0.05,
|
| 70 |
+
"model_type": "wav2vec2",
|
| 71 |
+
"num_adapter_layers": 3,
|
| 72 |
+
"num_attention_heads": 16,
|
| 73 |
+
"num_codevector_groups": 2,
|
| 74 |
+
"num_codevectors_per_group": 320,
|
| 75 |
+
"num_conv_pos_embedding_groups": 16,
|
| 76 |
+
"num_conv_pos_embeddings": 128,
|
| 77 |
+
"num_feat_extract_layers": 7,
|
| 78 |
+
"num_hidden_layers": 24,
|
| 79 |
+
"num_negatives": 100,
|
| 80 |
+
"output_hidden_size": 1024,
|
| 81 |
+
"pad_token_id": 77,
|
| 82 |
+
"proj_codevector_dim": 768,
|
| 83 |
+
"tdnn_dilation": [
|
| 84 |
+
1,
|
| 85 |
+
2,
|
| 86 |
+
3,
|
| 87 |
+
1,
|
| 88 |
+
1
|
| 89 |
+
],
|
| 90 |
+
"tdnn_dim": [
|
| 91 |
+
512,
|
| 92 |
+
512,
|
| 93 |
+
512,
|
| 94 |
+
512,
|
| 95 |
+
1500
|
| 96 |
+
],
|
| 97 |
+
"tdnn_kernel": [
|
| 98 |
+
5,
|
| 99 |
+
3,
|
| 100 |
+
3,
|
| 101 |
+
1,
|
| 102 |
+
1
|
| 103 |
+
],
|
| 104 |
+
"torch_dtype": "float32",
|
| 105 |
+
"transformers_version": "4.39.3",
|
| 106 |
+
"use_weighted_layer_sum": false,
|
| 107 |
+
"vocab_size": 80,
|
| 108 |
+
"xvector_output_dim": 512
|
| 109 |
+
}
|
last-checkpoint/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01af806bfb43537f473716c6921c56c3f1b6cff5b24f430ec50b63d6432ff62c
|
| 3 |
+
size 1262135480
|
last-checkpoint/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f54746b0feae2e036d16930d4e673363c77f9f9aafe635895c1098194f55d6ef
|
| 3 |
+
size 2490815798
|
last-checkpoint/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": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
last-checkpoint/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7ec026adbf9bcbd8a783c266184ac6a8dfb47b9e7d5a7687187770c89b84faf
|
| 3 |
+
size 14308
|
last-checkpoint/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05c969971a5565a3fd0897b347349de9a9a330022bc9b3755b121094b8004da8
|
| 3 |
+
size 1064
|
last-checkpoint/trainer_state.json
ADDED
|
@@ -0,0 +1,837 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 13.83989145183175,
|
| 5 |
+
"eval_steps": 400,
|
| 6 |
+
"global_step": 20400,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.27,
|
| 13 |
+
"grad_norm": 4.026116371154785,
|
| 14 |
+
"learning_rate": 0.00023999999999999998,
|
| 15 |
+
"loss": 5.8396,
|
| 16 |
+
"step": 400
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.27,
|
| 20 |
+
"eval_loss": 0.9751555919647217,
|
| 21 |
+
"eval_runtime": 211.3976,
|
| 22 |
+
"eval_samples_per_second": 6.618,
|
| 23 |
+
"eval_steps_per_second": 0.828,
|
| 24 |
+
"eval_wer": 0.8373902326297951,
|
| 25 |
+
"step": 400
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.54,
|
| 29 |
+
"grad_norm": 1.9866678714752197,
|
| 30 |
+
"learning_rate": 0.00029794144556267153,
|
| 31 |
+
"loss": 0.6967,
|
| 32 |
+
"step": 800
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.54,
|
| 36 |
+
"eval_loss": 0.38119566440582275,
|
| 37 |
+
"eval_runtime": 210.7469,
|
| 38 |
+
"eval_samples_per_second": 6.638,
|
| 39 |
+
"eval_steps_per_second": 0.83,
|
| 40 |
+
"eval_wer": 0.5935140964412263,
|
| 41 |
+
"step": 800
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.81,
|
| 45 |
+
"grad_norm": 1.540819525718689,
|
| 46 |
+
"learning_rate": 0.00029519670631290027,
|
| 47 |
+
"loss": 0.4806,
|
| 48 |
+
"step": 1200
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.81,
|
| 52 |
+
"eval_loss": 0.3368071913719177,
|
| 53 |
+
"eval_runtime": 211.9206,
|
| 54 |
+
"eval_samples_per_second": 6.602,
|
| 55 |
+
"eval_steps_per_second": 0.826,
|
| 56 |
+
"eval_wer": 0.4757356339547065,
|
| 57 |
+
"step": 1200
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.09,
|
| 61 |
+
"grad_norm": 0.9303820729255676,
|
| 62 |
+
"learning_rate": 0.000292451967063129,
|
| 63 |
+
"loss": 0.3996,
|
| 64 |
+
"step": 1600
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 1.09,
|
| 68 |
+
"eval_loss": 0.199427992105484,
|
| 69 |
+
"eval_runtime": 210.2196,
|
| 70 |
+
"eval_samples_per_second": 6.655,
|
| 71 |
+
"eval_steps_per_second": 0.832,
|
| 72 |
+
"eval_wer": 0.3142813125866584,
|
| 73 |
+
"step": 1600
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.36,
|
| 77 |
+
"grad_norm": 0.5186350345611572,
|
| 78 |
+
"learning_rate": 0.0002897072278133577,
|
| 79 |
+
"loss": 0.3497,
|
| 80 |
+
"step": 2000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.36,
|
| 84 |
+
"eval_loss": 0.16836047172546387,
|
| 85 |
+
"eval_runtime": 210.345,
|
| 86 |
+
"eval_samples_per_second": 6.651,
|
| 87 |
+
"eval_steps_per_second": 0.832,
|
| 88 |
+
"eval_wer": 0.2564319827453397,
|
| 89 |
+
"step": 2000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 1.63,
|
| 93 |
+
"grad_norm": 0.9710769057273865,
|
| 94 |
+
"learning_rate": 0.00028696248856358643,
|
| 95 |
+
"loss": 0.3372,
|
| 96 |
+
"step": 2400
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.63,
|
| 100 |
+
"eval_loss": 0.15862207114696503,
|
| 101 |
+
"eval_runtime": 212.5997,
|
| 102 |
+
"eval_samples_per_second": 6.58,
|
| 103 |
+
"eval_steps_per_second": 0.823,
|
| 104 |
+
"eval_wer": 0.2418733631181636,
|
| 105 |
+
"step": 2400
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.9,
|
| 109 |
+
"grad_norm": 0.8559458255767822,
|
| 110 |
+
"learning_rate": 0.00028421774931381517,
|
| 111 |
+
"loss": 0.312,
|
| 112 |
+
"step": 2800
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.9,
|
| 116 |
+
"eval_loss": 0.1412586122751236,
|
| 117 |
+
"eval_runtime": 211.389,
|
| 118 |
+
"eval_samples_per_second": 6.618,
|
| 119 |
+
"eval_steps_per_second": 0.828,
|
| 120 |
+
"eval_wer": 0.2226929594823602,
|
| 121 |
+
"step": 2800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.17,
|
| 125 |
+
"grad_norm": 0.6652762293815613,
|
| 126 |
+
"learning_rate": 0.0002814730100640439,
|
| 127 |
+
"loss": 0.2797,
|
| 128 |
+
"step": 3200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.17,
|
| 132 |
+
"eval_loss": 0.1465640813112259,
|
| 133 |
+
"eval_runtime": 210.5595,
|
| 134 |
+
"eval_samples_per_second": 6.644,
|
| 135 |
+
"eval_steps_per_second": 0.831,
|
| 136 |
+
"eval_wer": 0.22731474349098751,
|
| 137 |
+
"step": 3200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.44,
|
| 141 |
+
"grad_norm": 0.361288458108902,
|
| 142 |
+
"learning_rate": 0.0002787282708142726,
|
| 143 |
+
"loss": 0.2554,
|
| 144 |
+
"step": 3600
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.44,
|
| 148 |
+
"eval_loss": 0.1542745977640152,
|
| 149 |
+
"eval_runtime": 211.2213,
|
| 150 |
+
"eval_samples_per_second": 6.623,
|
| 151 |
+
"eval_steps_per_second": 0.829,
|
| 152 |
+
"eval_wer": 0.2365583115082422,
|
| 153 |
+
"step": 3600
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.71,
|
| 157 |
+
"grad_norm": 0.5574541091918945,
|
| 158 |
+
"learning_rate": 0.00027598353156450134,
|
| 159 |
+
"loss": 0.2613,
|
| 160 |
+
"step": 4000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.71,
|
| 164 |
+
"eval_loss": 0.14501063525676727,
|
| 165 |
+
"eval_runtime": 212.1205,
|
| 166 |
+
"eval_samples_per_second": 6.595,
|
| 167 |
+
"eval_steps_per_second": 0.825,
|
| 168 |
+
"eval_wer": 0.23255276536743183,
|
| 169 |
+
"step": 4000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 2.99,
|
| 173 |
+
"grad_norm": 0.4116342067718506,
|
| 174 |
+
"learning_rate": 0.0002732387923147301,
|
| 175 |
+
"loss": 0.2399,
|
| 176 |
+
"step": 4400
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 2.99,
|
| 180 |
+
"eval_loss": 0.12376150488853455,
|
| 181 |
+
"eval_runtime": 212.9821,
|
| 182 |
+
"eval_samples_per_second": 6.569,
|
| 183 |
+
"eval_steps_per_second": 0.822,
|
| 184 |
+
"eval_wer": 0.20297334771221692,
|
| 185 |
+
"step": 4400
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 3.26,
|
| 189 |
+
"grad_norm": 0.834938645362854,
|
| 190 |
+
"learning_rate": 0.0002704940530649588,
|
| 191 |
+
"loss": 0.2125,
|
| 192 |
+
"step": 4800
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 3.26,
|
| 196 |
+
"eval_loss": 0.09888482838869095,
|
| 197 |
+
"eval_runtime": 213.2706,
|
| 198 |
+
"eval_samples_per_second": 6.56,
|
| 199 |
+
"eval_steps_per_second": 0.821,
|
| 200 |
+
"eval_wer": 0.16099214296718534,
|
| 201 |
+
"step": 4800
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 3.53,
|
| 205 |
+
"grad_norm": 0.4636085331439972,
|
| 206 |
+
"learning_rate": 0.0002677493138151875,
|
| 207 |
+
"loss": 0.2144,
|
| 208 |
+
"step": 5200
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 3.53,
|
| 212 |
+
"eval_loss": 0.09842105209827423,
|
| 213 |
+
"eval_runtime": 210.8983,
|
| 214 |
+
"eval_samples_per_second": 6.634,
|
| 215 |
+
"eval_steps_per_second": 0.83,
|
| 216 |
+
"eval_wer": 0.1612232321676167,
|
| 217 |
+
"step": 5200
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 3.8,
|
| 221 |
+
"grad_norm": 0.4843950569629669,
|
| 222 |
+
"learning_rate": 0.0002650045745654163,
|
| 223 |
+
"loss": 0.212,
|
| 224 |
+
"step": 5600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 3.8,
|
| 228 |
+
"eval_loss": 0.08756741881370544,
|
| 229 |
+
"eval_runtime": 210.9696,
|
| 230 |
+
"eval_samples_per_second": 6.631,
|
| 231 |
+
"eval_steps_per_second": 0.83,
|
| 232 |
+
"eval_wer": 0.1507471884147281,
|
| 233 |
+
"step": 5600
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.07,
|
| 237 |
+
"grad_norm": 0.847865641117096,
|
| 238 |
+
"learning_rate": 0.000262259835315645,
|
| 239 |
+
"loss": 0.1964,
|
| 240 |
+
"step": 6000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.07,
|
| 244 |
+
"eval_loss": 0.1017051413655281,
|
| 245 |
+
"eval_runtime": 211.736,
|
| 246 |
+
"eval_samples_per_second": 6.607,
|
| 247 |
+
"eval_steps_per_second": 0.827,
|
| 248 |
+
"eval_wer": 0.17531967339393006,
|
| 249 |
+
"step": 6000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 4.34,
|
| 253 |
+
"grad_norm": 0.35252293944358826,
|
| 254 |
+
"learning_rate": 0.0002595150960658737,
|
| 255 |
+
"loss": 0.1814,
|
| 256 |
+
"step": 6400
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 4.34,
|
| 260 |
+
"eval_loss": 0.0966610386967659,
|
| 261 |
+
"eval_runtime": 211.5792,
|
| 262 |
+
"eval_samples_per_second": 6.612,
|
| 263 |
+
"eval_steps_per_second": 0.827,
|
| 264 |
+
"eval_wer": 0.1653828377753813,
|
| 265 |
+
"step": 6400
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 4.61,
|
| 269 |
+
"grad_norm": 0.38534751534461975,
|
| 270 |
+
"learning_rate": 0.00025677035681610246,
|
| 271 |
+
"loss": 0.1772,
|
| 272 |
+
"step": 6800
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 4.61,
|
| 276 |
+
"eval_loss": 0.0955633744597435,
|
| 277 |
+
"eval_runtime": 211.2723,
|
| 278 |
+
"eval_samples_per_second": 6.622,
|
| 279 |
+
"eval_steps_per_second": 0.828,
|
| 280 |
+
"eval_wer": 0.16314897550454474,
|
| 281 |
+
"step": 6800
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 4.88,
|
| 285 |
+
"grad_norm": 0.3151361346244812,
|
| 286 |
+
"learning_rate": 0.0002540256175663312,
|
| 287 |
+
"loss": 0.1748,
|
| 288 |
+
"step": 7200
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 4.88,
|
| 292 |
+
"eval_loss": 0.08700825273990631,
|
| 293 |
+
"eval_runtime": 213.02,
|
| 294 |
+
"eval_samples_per_second": 6.567,
|
| 295 |
+
"eval_steps_per_second": 0.822,
|
| 296 |
+
"eval_wer": 0.14828223694346018,
|
| 297 |
+
"step": 7200
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 5.16,
|
| 301 |
+
"grad_norm": 1.1274446249008179,
|
| 302 |
+
"learning_rate": 0.0002512808783165599,
|
| 303 |
+
"loss": 0.1706,
|
| 304 |
+
"step": 7600
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.16,
|
| 308 |
+
"eval_loss": 0.07706322520971298,
|
| 309 |
+
"eval_runtime": 212.6781,
|
| 310 |
+
"eval_samples_per_second": 6.578,
|
| 311 |
+
"eval_steps_per_second": 0.823,
|
| 312 |
+
"eval_wer": 0.1306424279771992,
|
| 313 |
+
"step": 7600
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.43,
|
| 317 |
+
"grad_norm": 1.3828215599060059,
|
| 318 |
+
"learning_rate": 0.0002485361390667886,
|
| 319 |
+
"loss": 0.1545,
|
| 320 |
+
"step": 8000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.43,
|
| 324 |
+
"eval_loss": 0.06534561514854431,
|
| 325 |
+
"eval_runtime": 211.3365,
|
| 326 |
+
"eval_samples_per_second": 6.62,
|
| 327 |
+
"eval_steps_per_second": 0.828,
|
| 328 |
+
"eval_wer": 0.1198582652904021,
|
| 329 |
+
"step": 8000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 5.7,
|
| 333 |
+
"grad_norm": 0.4036734402179718,
|
| 334 |
+
"learning_rate": 0.00024579139981701736,
|
| 335 |
+
"loss": 0.1627,
|
| 336 |
+
"step": 8400
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 5.7,
|
| 340 |
+
"eval_loss": 0.060022782534360886,
|
| 341 |
+
"eval_runtime": 211.4007,
|
| 342 |
+
"eval_samples_per_second": 6.618,
|
| 343 |
+
"eval_steps_per_second": 0.828,
|
| 344 |
+
"eval_wer": 0.11030657833923894,
|
| 345 |
+
"step": 8400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.97,
|
| 349 |
+
"grad_norm": 0.3543277084827423,
|
| 350 |
+
"learning_rate": 0.0002430466605672461,
|
| 351 |
+
"loss": 0.1541,
|
| 352 |
+
"step": 8800
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.97,
|
| 356 |
+
"eval_loss": 0.058936990797519684,
|
| 357 |
+
"eval_runtime": 212.2764,
|
| 358 |
+
"eval_samples_per_second": 6.59,
|
| 359 |
+
"eval_steps_per_second": 0.824,
|
| 360 |
+
"eval_wer": 0.10676321059929132,
|
| 361 |
+
"step": 8800
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 6.24,
|
| 365 |
+
"grad_norm": 0.23857054114341736,
|
| 366 |
+
"learning_rate": 0.00024030192131747481,
|
| 367 |
+
"loss": 0.1382,
|
| 368 |
+
"step": 9200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.24,
|
| 372 |
+
"eval_loss": 0.07100619375705719,
|
| 373 |
+
"eval_runtime": 213.3512,
|
| 374 |
+
"eval_samples_per_second": 6.557,
|
| 375 |
+
"eval_steps_per_second": 0.82,
|
| 376 |
+
"eval_wer": 0.12309351409644123,
|
| 377 |
+
"step": 9200
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.51,
|
| 381 |
+
"grad_norm": 0.6902241706848145,
|
| 382 |
+
"learning_rate": 0.00023755718206770355,
|
| 383 |
+
"loss": 0.1397,
|
| 384 |
+
"step": 9600
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.51,
|
| 388 |
+
"eval_loss": 0.06513579189777374,
|
| 389 |
+
"eval_runtime": 211.6334,
|
| 390 |
+
"eval_samples_per_second": 6.61,
|
| 391 |
+
"eval_steps_per_second": 0.827,
|
| 392 |
+
"eval_wer": 0.12478816823293791,
|
| 393 |
+
"step": 9600
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 6.78,
|
| 397 |
+
"grad_norm": 0.3040298819541931,
|
| 398 |
+
"learning_rate": 0.00023481244281793227,
|
| 399 |
+
"loss": 0.1345,
|
| 400 |
+
"step": 10000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 6.78,
|
| 404 |
+
"eval_loss": 0.06700535863637924,
|
| 405 |
+
"eval_runtime": 211.9668,
|
| 406 |
+
"eval_samples_per_second": 6.6,
|
| 407 |
+
"eval_steps_per_second": 0.826,
|
| 408 |
+
"eval_wer": 0.11939608688953936,
|
| 409 |
+
"step": 10000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 7.06,
|
| 413 |
+
"grad_norm": 1.0794196128845215,
|
| 414 |
+
"learning_rate": 0.000232067703568161,
|
| 415 |
+
"loss": 0.1281,
|
| 416 |
+
"step": 10400
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.06,
|
| 420 |
+
"eval_loss": 0.054103270173072815,
|
| 421 |
+
"eval_runtime": 209.427,
|
| 422 |
+
"eval_samples_per_second": 6.68,
|
| 423 |
+
"eval_steps_per_second": 0.836,
|
| 424 |
+
"eval_wer": 0.10060083192112156,
|
| 425 |
+
"step": 10400
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 7.33,
|
| 429 |
+
"grad_norm": 0.8765202164649963,
|
| 430 |
+
"learning_rate": 0.00022932296431838972,
|
| 431 |
+
"loss": 0.1315,
|
| 432 |
+
"step": 10800
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.33,
|
| 436 |
+
"eval_loss": 0.05593947321176529,
|
| 437 |
+
"eval_runtime": 208.8917,
|
| 438 |
+
"eval_samples_per_second": 6.697,
|
| 439 |
+
"eval_steps_per_second": 0.838,
|
| 440 |
+
"eval_wer": 0.10622400246495148,
|
| 441 |
+
"step": 10800
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 7.6,
|
| 445 |
+
"grad_norm": 0.5989521741867065,
|
| 446 |
+
"learning_rate": 0.00022657822506861846,
|
| 447 |
+
"loss": 0.1234,
|
| 448 |
+
"step": 11200
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 7.6,
|
| 452 |
+
"eval_loss": 0.05276945233345032,
|
| 453 |
+
"eval_runtime": 209.5765,
|
| 454 |
+
"eval_samples_per_second": 6.675,
|
| 455 |
+
"eval_steps_per_second": 0.835,
|
| 456 |
+
"eval_wer": 0.09698043444769681,
|
| 457 |
+
"step": 11200
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.87,
|
| 461 |
+
"grad_norm": 0.3679282069206238,
|
| 462 |
+
"learning_rate": 0.00022383348581884717,
|
| 463 |
+
"loss": 0.1248,
|
| 464 |
+
"step": 11600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.87,
|
| 468 |
+
"eval_loss": 0.044818080961704254,
|
| 469 |
+
"eval_runtime": 209.7397,
|
| 470 |
+
"eval_samples_per_second": 6.67,
|
| 471 |
+
"eval_steps_per_second": 0.834,
|
| 472 |
+
"eval_wer": 0.0865043906948082,
|
| 473 |
+
"step": 11600
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 8.14,
|
| 477 |
+
"grad_norm": 0.3467447757720947,
|
| 478 |
+
"learning_rate": 0.00022108874656907594,
|
| 479 |
+
"loss": 0.115,
|
| 480 |
+
"step": 12000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 8.14,
|
| 484 |
+
"eval_loss": 0.0545777752995491,
|
| 485 |
+
"eval_runtime": 209.6701,
|
| 486 |
+
"eval_samples_per_second": 6.672,
|
| 487 |
+
"eval_steps_per_second": 0.835,
|
| 488 |
+
"eval_wer": 0.0993683561854876,
|
| 489 |
+
"step": 12000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 8.41,
|
| 493 |
+
"grad_norm": 1.4704556465148926,
|
| 494 |
+
"learning_rate": 0.00021834400731930465,
|
| 495 |
+
"loss": 0.1143,
|
| 496 |
+
"step": 12400
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 8.41,
|
| 500 |
+
"eval_loss": 0.05952217057347298,
|
| 501 |
+
"eval_runtime": 210.7235,
|
| 502 |
+
"eval_samples_per_second": 6.639,
|
| 503 |
+
"eval_steps_per_second": 0.83,
|
| 504 |
+
"eval_wer": 0.10861192420274225,
|
| 505 |
+
"step": 12400
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 8.68,
|
| 509 |
+
"grad_norm": 0.2509690821170807,
|
| 510 |
+
"learning_rate": 0.0002155992680695334,
|
| 511 |
+
"loss": 0.1169,
|
| 512 |
+
"step": 12800
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 8.68,
|
| 516 |
+
"eval_loss": 0.04853290319442749,
|
| 517 |
+
"eval_runtime": 209.7854,
|
| 518 |
+
"eval_samples_per_second": 6.669,
|
| 519 |
+
"eval_steps_per_second": 0.834,
|
| 520 |
+
"eval_wer": 0.08742874749653366,
|
| 521 |
+
"step": 12800
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 8.96,
|
| 525 |
+
"grad_norm": 0.2134234458208084,
|
| 526 |
+
"learning_rate": 0.0002128545288197621,
|
| 527 |
+
"loss": 0.1165,
|
| 528 |
+
"step": 13200
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 8.96,
|
| 532 |
+
"eval_loss": 0.05237515643239021,
|
| 533 |
+
"eval_runtime": 210.2596,
|
| 534 |
+
"eval_samples_per_second": 6.654,
|
| 535 |
+
"eval_steps_per_second": 0.832,
|
| 536 |
+
"eval_wer": 0.09767370204899091,
|
| 537 |
+
"step": 13200
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 9.23,
|
| 541 |
+
"grad_norm": 0.4314160943031311,
|
| 542 |
+
"learning_rate": 0.00021010978956999084,
|
| 543 |
+
"loss": 0.1035,
|
| 544 |
+
"step": 13600
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 9.23,
|
| 548 |
+
"eval_loss": 0.044478464871644974,
|
| 549 |
+
"eval_runtime": 210.5273,
|
| 550 |
+
"eval_samples_per_second": 6.645,
|
| 551 |
+
"eval_steps_per_second": 0.831,
|
| 552 |
+
"eval_wer": 0.0837313202896318,
|
| 553 |
+
"step": 13600
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.5,
|
| 557 |
+
"grad_norm": 0.4133451581001282,
|
| 558 |
+
"learning_rate": 0.00020736505032021955,
|
| 559 |
+
"loss": 0.1017,
|
| 560 |
+
"step": 14000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 9.5,
|
| 564 |
+
"eval_loss": 0.04130551964044571,
|
| 565 |
+
"eval_runtime": 210.2347,
|
| 566 |
+
"eval_samples_per_second": 6.654,
|
| 567 |
+
"eval_steps_per_second": 0.832,
|
| 568 |
+
"eval_wer": 0.0791865660144816,
|
| 569 |
+
"step": 14000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 9.77,
|
| 573 |
+
"grad_norm": 0.29300200939178467,
|
| 574 |
+
"learning_rate": 0.0002046203110704483,
|
| 575 |
+
"loss": 0.109,
|
| 576 |
+
"step": 14400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 9.77,
|
| 580 |
+
"eval_loss": 0.04202074185013771,
|
| 581 |
+
"eval_runtime": 211.2523,
|
| 582 |
+
"eval_samples_per_second": 6.622,
|
| 583 |
+
"eval_steps_per_second": 0.828,
|
| 584 |
+
"eval_wer": 0.08326914188876906,
|
| 585 |
+
"step": 14400
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 10.04,
|
| 589 |
+
"grad_norm": 0.2547191083431244,
|
| 590 |
+
"learning_rate": 0.000201875571820677,
|
| 591 |
+
"loss": 0.1018,
|
| 592 |
+
"step": 14800
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 10.04,
|
| 596 |
+
"eval_loss": 0.04540720209479332,
|
| 597 |
+
"eval_runtime": 210.6476,
|
| 598 |
+
"eval_samples_per_second": 6.641,
|
| 599 |
+
"eval_steps_per_second": 0.831,
|
| 600 |
+
"eval_wer": 0.0823447850870436,
|
| 601 |
+
"step": 14800
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 10.31,
|
| 605 |
+
"grad_norm": 0.2799816131591797,
|
| 606 |
+
"learning_rate": 0.00019913083257090574,
|
| 607 |
+
"loss": 0.0929,
|
| 608 |
+
"step": 15200
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 10.31,
|
| 612 |
+
"eval_loss": 0.042868729680776596,
|
| 613 |
+
"eval_runtime": 209.1492,
|
| 614 |
+
"eval_samples_per_second": 6.689,
|
| 615 |
+
"eval_steps_per_second": 0.837,
|
| 616 |
+
"eval_wer": 0.0785703281466646,
|
| 617 |
+
"step": 15200
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 10.58,
|
| 621 |
+
"grad_norm": 0.18747101724147797,
|
| 622 |
+
"learning_rate": 0.00019638609332113446,
|
| 623 |
+
"loss": 0.0956,
|
| 624 |
+
"step": 15600
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.58,
|
| 628 |
+
"eval_loss": 0.04033521190285683,
|
| 629 |
+
"eval_runtime": 209.4853,
|
| 630 |
+
"eval_samples_per_second": 6.678,
|
| 631 |
+
"eval_steps_per_second": 0.835,
|
| 632 |
+
"eval_wer": 0.07718379294407642,
|
| 633 |
+
"step": 15600
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 10.85,
|
| 637 |
+
"grad_norm": 0.3078967034816742,
|
| 638 |
+
"learning_rate": 0.0001936413540713632,
|
| 639 |
+
"loss": 0.0986,
|
| 640 |
+
"step": 16000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 10.85,
|
| 644 |
+
"eval_loss": 0.046831514686346054,
|
| 645 |
+
"eval_runtime": 209.7149,
|
| 646 |
+
"eval_samples_per_second": 6.671,
|
| 647 |
+
"eval_steps_per_second": 0.834,
|
| 648 |
+
"eval_wer": 0.09058696656909566,
|
| 649 |
+
"step": 16000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 11.13,
|
| 653 |
+
"grad_norm": 0.10461205244064331,
|
| 654 |
+
"learning_rate": 0.0001908966148215919,
|
| 655 |
+
"loss": 0.0941,
|
| 656 |
+
"step": 16400
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 11.13,
|
| 660 |
+
"eval_loss": 0.03615270182490349,
|
| 661 |
+
"eval_runtime": 209.392,
|
| 662 |
+
"eval_samples_per_second": 6.681,
|
| 663 |
+
"eval_steps_per_second": 0.836,
|
| 664 |
+
"eval_wer": 0.06940378986288707,
|
| 665 |
+
"step": 16400
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 11.4,
|
| 669 |
+
"grad_norm": 0.5313912630081177,
|
| 670 |
+
"learning_rate": 0.00018815187557182068,
|
| 671 |
+
"loss": 0.0845,
|
| 672 |
+
"step": 16800
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 11.4,
|
| 676 |
+
"eval_loss": 0.038695286959409714,
|
| 677 |
+
"eval_runtime": 210.5021,
|
| 678 |
+
"eval_samples_per_second": 6.646,
|
| 679 |
+
"eval_steps_per_second": 0.831,
|
| 680 |
+
"eval_wer": 0.0701740871976583,
|
| 681 |
+
"step": 16800
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 11.67,
|
| 685 |
+
"grad_norm": 0.5434714555740356,
|
| 686 |
+
"learning_rate": 0.0001854071363220494,
|
| 687 |
+
"loss": 0.0955,
|
| 688 |
+
"step": 17200
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 11.67,
|
| 692 |
+
"eval_loss": 0.03512905538082123,
|
| 693 |
+
"eval_runtime": 210.6242,
|
| 694 |
+
"eval_samples_per_second": 6.642,
|
| 695 |
+
"eval_steps_per_second": 0.831,
|
| 696 |
+
"eval_wer": 0.06270220305037745,
|
| 697 |
+
"step": 17200
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 11.94,
|
| 701 |
+
"grad_norm": 0.6010161638259888,
|
| 702 |
+
"learning_rate": 0.00018266239707227813,
|
| 703 |
+
"loss": 0.089,
|
| 704 |
+
"step": 17600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 11.94,
|
| 708 |
+
"eval_loss": 0.03609244525432587,
|
| 709 |
+
"eval_runtime": 210.1814,
|
| 710 |
+
"eval_samples_per_second": 6.656,
|
| 711 |
+
"eval_steps_per_second": 0.833,
|
| 712 |
+
"eval_wer": 0.06747804652595903,
|
| 713 |
+
"step": 17600
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 12.21,
|
| 717 |
+
"grad_norm": 0.2907910943031311,
|
| 718 |
+
"learning_rate": 0.00017991765782250684,
|
| 719 |
+
"loss": 0.0806,
|
| 720 |
+
"step": 18000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 12.21,
|
| 724 |
+
"eval_loss": 0.03811544552445412,
|
| 725 |
+
"eval_runtime": 210.0932,
|
| 726 |
+
"eval_samples_per_second": 6.659,
|
| 727 |
+
"eval_steps_per_second": 0.833,
|
| 728 |
+
"eval_wer": 0.06847943306116161,
|
| 729 |
+
"step": 18000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 12.48,
|
| 733 |
+
"grad_norm": 0.25918644666671753,
|
| 734 |
+
"learning_rate": 0.00017717291857273558,
|
| 735 |
+
"loss": 0.0803,
|
| 736 |
+
"step": 18400
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 12.48,
|
| 740 |
+
"eval_loss": 0.0369856134057045,
|
| 741 |
+
"eval_runtime": 210.0481,
|
| 742 |
+
"eval_samples_per_second": 6.66,
|
| 743 |
+
"eval_steps_per_second": 0.833,
|
| 744 |
+
"eval_wer": 0.06747804652595903,
|
| 745 |
+
"step": 18400
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 12.75,
|
| 749 |
+
"grad_norm": 0.3781072199344635,
|
| 750 |
+
"learning_rate": 0.0001744281793229643,
|
| 751 |
+
"loss": 0.0839,
|
| 752 |
+
"step": 18800
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 12.75,
|
| 756 |
+
"eval_loss": 0.0333174392580986,
|
| 757 |
+
"eval_runtime": 210.517,
|
| 758 |
+
"eval_samples_per_second": 6.646,
|
| 759 |
+
"eval_steps_per_second": 0.831,
|
| 760 |
+
"eval_wer": 0.06193190571560622,
|
| 761 |
+
"step": 18800
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 13.03,
|
| 765 |
+
"grad_norm": 0.42045047879219055,
|
| 766 |
+
"learning_rate": 0.00017168344007319303,
|
| 767 |
+
"loss": 0.0834,
|
| 768 |
+
"step": 19200
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 13.03,
|
| 772 |
+
"eval_loss": 0.033445805311203,
|
| 773 |
+
"eval_runtime": 210.6516,
|
| 774 |
+
"eval_samples_per_second": 6.641,
|
| 775 |
+
"eval_steps_per_second": 0.831,
|
| 776 |
+
"eval_wer": 0.057695270374364505,
|
| 777 |
+
"step": 19200
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 13.3,
|
| 781 |
+
"grad_norm": 0.4078648090362549,
|
| 782 |
+
"learning_rate": 0.00016893870082342174,
|
| 783 |
+
"loss": 0.0779,
|
| 784 |
+
"step": 19600
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 13.3,
|
| 788 |
+
"eval_loss": 0.03577824681997299,
|
| 789 |
+
"eval_runtime": 210.5433,
|
| 790 |
+
"eval_samples_per_second": 6.645,
|
| 791 |
+
"eval_steps_per_second": 0.831,
|
| 792 |
+
"eval_wer": 0.06208596518256047,
|
| 793 |
+
"step": 19600
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 13.57,
|
| 797 |
+
"grad_norm": 0.8884561657905579,
|
| 798 |
+
"learning_rate": 0.00016619396157365048,
|
| 799 |
+
"loss": 0.0773,
|
| 800 |
+
"step": 20000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 13.57,
|
| 804 |
+
"eval_loss": 0.032973628491163254,
|
| 805 |
+
"eval_runtime": 210.0131,
|
| 806 |
+
"eval_samples_per_second": 6.661,
|
| 807 |
+
"eval_steps_per_second": 0.833,
|
| 808 |
+
"eval_wer": 0.05646279463873055,
|
| 809 |
+
"step": 20000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 13.84,
|
| 813 |
+
"grad_norm": 0.5232133269309998,
|
| 814 |
+
"learning_rate": 0.0001634492223238792,
|
| 815 |
+
"loss": 0.0717,
|
| 816 |
+
"step": 20400
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 13.84,
|
| 820 |
+
"eval_loss": 0.03495289012789726,
|
| 821 |
+
"eval_runtime": 210.8453,
|
| 822 |
+
"eval_samples_per_second": 6.635,
|
| 823 |
+
"eval_steps_per_second": 0.83,
|
| 824 |
+
"eval_wer": 0.0625481435834232,
|
| 825 |
+
"step": 20400
|
| 826 |
+
}
|
| 827 |
+
],
|
| 828 |
+
"logging_steps": 400,
|
| 829 |
+
"max_steps": 44220,
|
| 830 |
+
"num_input_tokens_seen": 0,
|
| 831 |
+
"num_train_epochs": 30,
|
| 832 |
+
"save_steps": 400,
|
| 833 |
+
"total_flos": 6.2228407342062576e+19,
|
| 834 |
+
"train_batch_size": 8,
|
| 835 |
+
"trial_name": null,
|
| 836 |
+
"trial_params": null
|
| 837 |
+
}
|
last-checkpoint/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6cdc5b20075b1fce2a136d9f42ed3f0b21546cf27b594236f7beba8f280a4aa
|
| 3 |
+
size 5048
|