Add files using upload-large-folder tool
Browse files- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/special_tokens_map.json +15 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/tokenizer.json +0 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/tokenizer_config.json +56 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/trainer_state.json +396 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/special_tokens_map.json +15 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/tokenizer.json +0 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/tokenizer_config.json +56 -0
- uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/trainer_state.json +216 -0
- uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/special_tokens_map.json +15 -0
- uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/tokenizer.json +0 -0
- uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/tokenizer_config.json +56 -0
- uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/trainer_state.json +456 -0
- uit_vsmec-visobert-s44-20260829T184728Z-863c2de1/trainer/checkpoint-2784/trainer_state.json +996 -0
- vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/special_tokens_map.json +15 -0
- vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/tokenizer.json +0 -0
- vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/tokenizer_config.json +56 -0
- vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/trainer_state.json +834 -0
- vigoemotions-visobert-s42-20260829T190105Z-5eeb966a/trainer/checkpoint-5687/special_tokens_map.json +15 -0
- vigoemotions-visobert-s42-20260829T190105Z-5eeb966a/trainer/checkpoint-5687/tokenizer.json +0 -0
- vigoemotions-visobert-s42-20260829T190105Z-5eeb966a/trainer/checkpoint-5687/tokenizer_config.json +56 -0
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"15001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"sp_model_kwargs": {},
|
| 54 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 55 |
+
"unk_token": "<unk>"
|
| 56 |
+
}
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-1044/trainer_state.json
ADDED
|
@@ -0,0 +1,396 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 522,
|
| 3 |
+
"best_metric": 0.6127241623305312,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1044,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1440922190201729,
|
| 14 |
+
"grad_norm": 38.22502899169922,
|
| 15 |
+
"learning_rate": 1.3793103448275862e-06,
|
| 16 |
+
"loss": 8.4291,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2881844380403458,
|
| 21 |
+
"grad_norm": 35.24979019165039,
|
| 22 |
+
"learning_rate": 2.8160919540229887e-06,
|
| 23 |
+
"loss": 7.6722,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4322766570605187,
|
| 28 |
+
"grad_norm": 36.06541061401367,
|
| 29 |
+
"learning_rate": 4.252873563218391e-06,
|
| 30 |
+
"loss": 6.8824,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5763688760806917,
|
| 35 |
+
"grad_norm": 29.078420639038086,
|
| 36 |
+
"learning_rate": 5.689655172413794e-06,
|
| 37 |
+
"loss": 6.3125,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7204610951008645,
|
| 42 |
+
"grad_norm": 42.46487045288086,
|
| 43 |
+
"learning_rate": 7.126436781609196e-06,
|
| 44 |
+
"loss": 5.985,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8645533141210374,
|
| 49 |
+
"grad_norm": 29.571460723876953,
|
| 50 |
+
"learning_rate": 8.563218390804599e-06,
|
| 51 |
+
"loss": 5.4702,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.607871720116618,
|
| 57 |
+
"eval_loss": 1.0908740758895874,
|
| 58 |
+
"eval_macro_f1": 0.49109327504848377,
|
| 59 |
+
"eval_runtime": 1.14,
|
| 60 |
+
"eval_samples_per_second": 601.751,
|
| 61 |
+
"eval_steps_per_second": 37.719,
|
| 62 |
+
"eval_weighted_f1": 0.5877912480877648,
|
| 63 |
+
"step": 174
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.005763688760807,
|
| 67 |
+
"grad_norm": 36.474754333496094,
|
| 68 |
+
"learning_rate": 1e-05,
|
| 69 |
+
"loss": 4.6922,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1498559077809798,
|
| 74 |
+
"grad_norm": 30.144718170166016,
|
| 75 |
+
"learning_rate": 1.1436781609195405e-05,
|
| 76 |
+
"loss": 4.084,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2939481268011528,
|
| 81 |
+
"grad_norm": 33.71086120605469,
|
| 82 |
+
"learning_rate": 1.2873563218390805e-05,
|
| 83 |
+
"loss": 3.7871,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4380403458213258,
|
| 88 |
+
"grad_norm": 42.127891540527344,
|
| 89 |
+
"learning_rate": 1.4310344827586209e-05,
|
| 90 |
+
"loss": 3.6227,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5821325648414986,
|
| 95 |
+
"grad_norm": 36.8306884765625,
|
| 96 |
+
"learning_rate": 1.574712643678161e-05,
|
| 97 |
+
"loss": 3.7176,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7262247838616713,
|
| 102 |
+
"grad_norm": 34.83902359008789,
|
| 103 |
+
"learning_rate": 1.7183908045977015e-05,
|
| 104 |
+
"loss": 3.62,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.8703170028818443,
|
| 109 |
+
"grad_norm": 35.82661819458008,
|
| 110 |
+
"learning_rate": 1.8620689655172415e-05,
|
| 111 |
+
"loss": 3.4783,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.6559766763848397,
|
| 117 |
+
"eval_loss": 0.9278486371040344,
|
| 118 |
+
"eval_macro_f1": 0.5995889137222778,
|
| 119 |
+
"eval_runtime": 1.1519,
|
| 120 |
+
"eval_samples_per_second": 595.556,
|
| 121 |
+
"eval_steps_per_second": 37.331,
|
| 122 |
+
"eval_weighted_f1": 0.6522077306052753,
|
| 123 |
+
"step": 348
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.011527377521614,
|
| 127 |
+
"grad_norm": 36.957427978515625,
|
| 128 |
+
"learning_rate": 1.999361430395913e-05,
|
| 129 |
+
"loss": 3.2858,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.155619596541787,
|
| 134 |
+
"grad_norm": 37.81496047973633,
|
| 135 |
+
"learning_rate": 1.983397190293742e-05,
|
| 136 |
+
"loss": 2.2024,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2997118155619596,
|
| 141 |
+
"grad_norm": 40.775150299072266,
|
| 142 |
+
"learning_rate": 1.967432950191571e-05,
|
| 143 |
+
"loss": 2.2638,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.443804034582133,
|
| 148 |
+
"grad_norm": 33.59572982788086,
|
| 149 |
+
"learning_rate": 1.9514687100894e-05,
|
| 150 |
+
"loss": 2.1861,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5878962536023056,
|
| 155 |
+
"grad_norm": 40.19615936279297,
|
| 156 |
+
"learning_rate": 1.9355044699872286e-05,
|
| 157 |
+
"loss": 2.2974,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.7319884726224783,
|
| 162 |
+
"grad_norm": 44.34323501586914,
|
| 163 |
+
"learning_rate": 1.9195402298850576e-05,
|
| 164 |
+
"loss": 2.3014,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8760806916426516,
|
| 169 |
+
"grad_norm": 34.9240608215332,
|
| 170 |
+
"learning_rate": 1.9035759897828865e-05,
|
| 171 |
+
"loss": 2.3619,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"eval_accuracy": 0.6720116618075802,
|
| 177 |
+
"eval_loss": 0.9691091775894165,
|
| 178 |
+
"eval_macro_f1": 0.6127241623305312,
|
| 179 |
+
"eval_runtime": 1.1146,
|
| 180 |
+
"eval_samples_per_second": 615.464,
|
| 181 |
+
"eval_steps_per_second": 38.579,
|
| 182 |
+
"eval_weighted_f1": 0.6652856405133254,
|
| 183 |
+
"step": 522
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0172910662824206,
|
| 187 |
+
"grad_norm": 27.3934268951416,
|
| 188 |
+
"learning_rate": 1.8876117496807155e-05,
|
| 189 |
+
"loss": 2.1046,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.161383285302594,
|
| 194 |
+
"grad_norm": 28.96881866455078,
|
| 195 |
+
"learning_rate": 1.871647509578544e-05,
|
| 196 |
+
"loss": 1.0097,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.3054755043227666,
|
| 201 |
+
"grad_norm": 36.40007781982422,
|
| 202 |
+
"learning_rate": 1.855683269476373e-05,
|
| 203 |
+
"loss": 1.0124,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.4495677233429394,
|
| 208 |
+
"grad_norm": 30.516788482666016,
|
| 209 |
+
"learning_rate": 1.839719029374202e-05,
|
| 210 |
+
"loss": 0.9584,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5936599423631126,
|
| 215 |
+
"grad_norm": 28.956357955932617,
|
| 216 |
+
"learning_rate": 1.823754789272031e-05,
|
| 217 |
+
"loss": 1.0372,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.7377521613832854,
|
| 222 |
+
"grad_norm": 28.7249698638916,
|
| 223 |
+
"learning_rate": 1.8077905491698596e-05,
|
| 224 |
+
"loss": 1.0111,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.881844380403458,
|
| 229 |
+
"grad_norm": 44.42074966430664,
|
| 230 |
+
"learning_rate": 1.7918263090676886e-05,
|
| 231 |
+
"loss": 1.0208,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"eval_accuracy": 0.6370262390670554,
|
| 237 |
+
"eval_loss": 1.2697323560714722,
|
| 238 |
+
"eval_macro_f1": 0.5980103547642622,
|
| 239 |
+
"eval_runtime": 1.026,
|
| 240 |
+
"eval_samples_per_second": 668.591,
|
| 241 |
+
"eval_steps_per_second": 41.909,
|
| 242 |
+
"eval_weighted_f1": 0.6371660432993673,
|
| 243 |
+
"step": 696
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.023054755043228,
|
| 247 |
+
"grad_norm": 27.69520378112793,
|
| 248 |
+
"learning_rate": 1.7758620689655175e-05,
|
| 249 |
+
"loss": 0.7987,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.1671469740634,
|
| 254 |
+
"grad_norm": 30.78780174255371,
|
| 255 |
+
"learning_rate": 1.759897828863346e-05,
|
| 256 |
+
"loss": 0.4189,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.311239193083574,
|
| 261 |
+
"grad_norm": 15.431654930114746,
|
| 262 |
+
"learning_rate": 1.743933588761175e-05,
|
| 263 |
+
"loss": 0.4017,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.455331412103746,
|
| 268 |
+
"grad_norm": 28.855989456176758,
|
| 269 |
+
"learning_rate": 1.7279693486590037e-05,
|
| 270 |
+
"loss": 0.3422,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.599423631123919,
|
| 275 |
+
"grad_norm": 22.899911880493164,
|
| 276 |
+
"learning_rate": 1.7120051085568327e-05,
|
| 277 |
+
"loss": 0.3003,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.743515850144092,
|
| 282 |
+
"grad_norm": 38.18413543701172,
|
| 283 |
+
"learning_rate": 1.6960408684546617e-05,
|
| 284 |
+
"loss": 0.398,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.887608069164266,
|
| 289 |
+
"grad_norm": 32.55165481567383,
|
| 290 |
+
"learning_rate": 1.6800766283524906e-05,
|
| 291 |
+
"loss": 0.323,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"eval_accuracy": 0.6588921282798834,
|
| 297 |
+
"eval_loss": 1.578009009361267,
|
| 298 |
+
"eval_macro_f1": 0.6073350146881163,
|
| 299 |
+
"eval_runtime": 1.077,
|
| 300 |
+
"eval_samples_per_second": 636.957,
|
| 301 |
+
"eval_steps_per_second": 39.926,
|
| 302 |
+
"eval_weighted_f1": 0.6542667347294129,
|
| 303 |
+
"step": 870
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.028818443804035,
|
| 307 |
+
"grad_norm": 26.64670181274414,
|
| 308 |
+
"learning_rate": 1.6641123882503192e-05,
|
| 309 |
+
"loss": 0.313,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.172910662824208,
|
| 314 |
+
"grad_norm": 8.356332778930664,
|
| 315 |
+
"learning_rate": 1.6481481481481482e-05,
|
| 316 |
+
"loss": 0.17,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.31700288184438,
|
| 321 |
+
"grad_norm": 11.126396179199219,
|
| 322 |
+
"learning_rate": 1.632183908045977e-05,
|
| 323 |
+
"loss": 0.1258,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.461095100864553,
|
| 328 |
+
"grad_norm": 3.7577595710754395,
|
| 329 |
+
"learning_rate": 1.616219667943806e-05,
|
| 330 |
+
"loss": 0.1172,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.605187319884726,
|
| 335 |
+
"grad_norm": 4.574268341064453,
|
| 336 |
+
"learning_rate": 1.6002554278416347e-05,
|
| 337 |
+
"loss": 0.1375,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.749279538904899,
|
| 342 |
+
"grad_norm": 20.50524139404297,
|
| 343 |
+
"learning_rate": 1.5842911877394637e-05,
|
| 344 |
+
"loss": 0.1414,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.893371757925072,
|
| 349 |
+
"grad_norm": 10.564369201660156,
|
| 350 |
+
"learning_rate": 1.5683269476372927e-05,
|
| 351 |
+
"loss": 0.174,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"eval_accuracy": 0.6530612244897959,
|
| 357 |
+
"eval_loss": 1.9646137952804565,
|
| 358 |
+
"eval_macro_f1": 0.5909335920986452,
|
| 359 |
+
"eval_runtime": 1.2818,
|
| 360 |
+
"eval_samples_per_second": 535.189,
|
| 361 |
+
"eval_steps_per_second": 33.547,
|
| 362 |
+
"eval_weighted_f1": 0.6511196399070723,
|
| 363 |
+
"step": 1044
|
| 364 |
+
}
|
| 365 |
+
],
|
| 366 |
+
"logging_steps": 25,
|
| 367 |
+
"max_steps": 3480,
|
| 368 |
+
"num_input_tokens_seen": 0,
|
| 369 |
+
"num_train_epochs": 20,
|
| 370 |
+
"save_steps": 500,
|
| 371 |
+
"stateful_callbacks": {
|
| 372 |
+
"EarlyStoppingCallback": {
|
| 373 |
+
"args": {
|
| 374 |
+
"early_stopping_patience": 3,
|
| 375 |
+
"early_stopping_threshold": 0.0
|
| 376 |
+
},
|
| 377 |
+
"attributes": {
|
| 378 |
+
"early_stopping_patience_counter": 3
|
| 379 |
+
}
|
| 380 |
+
},
|
| 381 |
+
"TrainerControl": {
|
| 382 |
+
"args": {
|
| 383 |
+
"should_epoch_stop": false,
|
| 384 |
+
"should_evaluate": false,
|
| 385 |
+
"should_log": false,
|
| 386 |
+
"should_save": true,
|
| 387 |
+
"should_training_stop": true
|
| 388 |
+
},
|
| 389 |
+
"attributes": {}
|
| 390 |
+
}
|
| 391 |
+
},
|
| 392 |
+
"total_flos": 0.0,
|
| 393 |
+
"train_batch_size": 8,
|
| 394 |
+
"trial_name": null,
|
| 395 |
+
"trial_params": null
|
| 396 |
+
}
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"15001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"sp_model_kwargs": {},
|
| 54 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 55 |
+
"unk_token": "<unk>"
|
| 56 |
+
}
|
uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522/trainer_state.json
ADDED
|
@@ -0,0 +1,216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 522,
|
| 3 |
+
"best_metric": 0.6127241623305312,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsmec-visobert-s42-20260829T183400Z-804458b3/trainer/checkpoint-522",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 522,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1440922190201729,
|
| 14 |
+
"grad_norm": 38.22502899169922,
|
| 15 |
+
"learning_rate": 1.3793103448275862e-06,
|
| 16 |
+
"loss": 8.4291,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2881844380403458,
|
| 21 |
+
"grad_norm": 35.24979019165039,
|
| 22 |
+
"learning_rate": 2.8160919540229887e-06,
|
| 23 |
+
"loss": 7.6722,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4322766570605187,
|
| 28 |
+
"grad_norm": 36.06541061401367,
|
| 29 |
+
"learning_rate": 4.252873563218391e-06,
|
| 30 |
+
"loss": 6.8824,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5763688760806917,
|
| 35 |
+
"grad_norm": 29.078420639038086,
|
| 36 |
+
"learning_rate": 5.689655172413794e-06,
|
| 37 |
+
"loss": 6.3125,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7204610951008645,
|
| 42 |
+
"grad_norm": 42.46487045288086,
|
| 43 |
+
"learning_rate": 7.126436781609196e-06,
|
| 44 |
+
"loss": 5.985,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8645533141210374,
|
| 49 |
+
"grad_norm": 29.571460723876953,
|
| 50 |
+
"learning_rate": 8.563218390804599e-06,
|
| 51 |
+
"loss": 5.4702,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.607871720116618,
|
| 57 |
+
"eval_loss": 1.0908740758895874,
|
| 58 |
+
"eval_macro_f1": 0.49109327504848377,
|
| 59 |
+
"eval_runtime": 1.14,
|
| 60 |
+
"eval_samples_per_second": 601.751,
|
| 61 |
+
"eval_steps_per_second": 37.719,
|
| 62 |
+
"eval_weighted_f1": 0.5877912480877648,
|
| 63 |
+
"step": 174
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.005763688760807,
|
| 67 |
+
"grad_norm": 36.474754333496094,
|
| 68 |
+
"learning_rate": 1e-05,
|
| 69 |
+
"loss": 4.6922,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1498559077809798,
|
| 74 |
+
"grad_norm": 30.144718170166016,
|
| 75 |
+
"learning_rate": 1.1436781609195405e-05,
|
| 76 |
+
"loss": 4.084,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2939481268011528,
|
| 81 |
+
"grad_norm": 33.71086120605469,
|
| 82 |
+
"learning_rate": 1.2873563218390805e-05,
|
| 83 |
+
"loss": 3.7871,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4380403458213258,
|
| 88 |
+
"grad_norm": 42.127891540527344,
|
| 89 |
+
"learning_rate": 1.4310344827586209e-05,
|
| 90 |
+
"loss": 3.6227,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5821325648414986,
|
| 95 |
+
"grad_norm": 36.8306884765625,
|
| 96 |
+
"learning_rate": 1.574712643678161e-05,
|
| 97 |
+
"loss": 3.7176,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7262247838616713,
|
| 102 |
+
"grad_norm": 34.83902359008789,
|
| 103 |
+
"learning_rate": 1.7183908045977015e-05,
|
| 104 |
+
"loss": 3.62,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.8703170028818443,
|
| 109 |
+
"grad_norm": 35.82661819458008,
|
| 110 |
+
"learning_rate": 1.8620689655172415e-05,
|
| 111 |
+
"loss": 3.4783,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.6559766763848397,
|
| 117 |
+
"eval_loss": 0.9278486371040344,
|
| 118 |
+
"eval_macro_f1": 0.5995889137222778,
|
| 119 |
+
"eval_runtime": 1.1519,
|
| 120 |
+
"eval_samples_per_second": 595.556,
|
| 121 |
+
"eval_steps_per_second": 37.331,
|
| 122 |
+
"eval_weighted_f1": 0.6522077306052753,
|
| 123 |
+
"step": 348
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.011527377521614,
|
| 127 |
+
"grad_norm": 36.957427978515625,
|
| 128 |
+
"learning_rate": 1.999361430395913e-05,
|
| 129 |
+
"loss": 3.2858,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.155619596541787,
|
| 134 |
+
"grad_norm": 37.81496047973633,
|
| 135 |
+
"learning_rate": 1.983397190293742e-05,
|
| 136 |
+
"loss": 2.2024,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2997118155619596,
|
| 141 |
+
"grad_norm": 40.775150299072266,
|
| 142 |
+
"learning_rate": 1.967432950191571e-05,
|
| 143 |
+
"loss": 2.2638,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.443804034582133,
|
| 148 |
+
"grad_norm": 33.59572982788086,
|
| 149 |
+
"learning_rate": 1.9514687100894e-05,
|
| 150 |
+
"loss": 2.1861,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5878962536023056,
|
| 155 |
+
"grad_norm": 40.19615936279297,
|
| 156 |
+
"learning_rate": 1.9355044699872286e-05,
|
| 157 |
+
"loss": 2.2974,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.7319884726224783,
|
| 162 |
+
"grad_norm": 44.34323501586914,
|
| 163 |
+
"learning_rate": 1.9195402298850576e-05,
|
| 164 |
+
"loss": 2.3014,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8760806916426516,
|
| 169 |
+
"grad_norm": 34.9240608215332,
|
| 170 |
+
"learning_rate": 1.9035759897828865e-05,
|
| 171 |
+
"loss": 2.3619,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"eval_accuracy": 0.6720116618075802,
|
| 177 |
+
"eval_loss": 0.9691091775894165,
|
| 178 |
+
"eval_macro_f1": 0.6127241623305312,
|
| 179 |
+
"eval_runtime": 1.1146,
|
| 180 |
+
"eval_samples_per_second": 615.464,
|
| 181 |
+
"eval_steps_per_second": 38.579,
|
| 182 |
+
"eval_weighted_f1": 0.6652856405133254,
|
| 183 |
+
"step": 522
|
| 184 |
+
}
|
| 185 |
+
],
|
| 186 |
+
"logging_steps": 25,
|
| 187 |
+
"max_steps": 3480,
|
| 188 |
+
"num_input_tokens_seen": 0,
|
| 189 |
+
"num_train_epochs": 20,
|
| 190 |
+
"save_steps": 500,
|
| 191 |
+
"stateful_callbacks": {
|
| 192 |
+
"EarlyStoppingCallback": {
|
| 193 |
+
"args": {
|
| 194 |
+
"early_stopping_patience": 3,
|
| 195 |
+
"early_stopping_threshold": 0.0
|
| 196 |
+
},
|
| 197 |
+
"attributes": {
|
| 198 |
+
"early_stopping_patience_counter": 0
|
| 199 |
+
}
|
| 200 |
+
},
|
| 201 |
+
"TrainerControl": {
|
| 202 |
+
"args": {
|
| 203 |
+
"should_epoch_stop": false,
|
| 204 |
+
"should_evaluate": false,
|
| 205 |
+
"should_log": false,
|
| 206 |
+
"should_save": true,
|
| 207 |
+
"should_training_stop": false
|
| 208 |
+
},
|
| 209 |
+
"attributes": {}
|
| 210 |
+
}
|
| 211 |
+
},
|
| 212 |
+
"total_flos": 0.0,
|
| 213 |
+
"train_batch_size": 8,
|
| 214 |
+
"trial_name": null,
|
| 215 |
+
"trial_params": null
|
| 216 |
+
}
|
uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"15001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"sp_model_kwargs": {},
|
| 54 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 55 |
+
"unk_token": "<unk>"
|
| 56 |
+
}
|
uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218/trainer_state.json
ADDED
|
@@ -0,0 +1,456 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1218,
|
| 3 |
+
"best_metric": 0.6293701407785159,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsmec-visobert-s43-20260829T183854Z-c0d1ca74/trainer/checkpoint-1218",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1218,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1440922190201729,
|
| 14 |
+
"grad_norm": 37.81015396118164,
|
| 15 |
+
"learning_rate": 1.3793103448275862e-06,
|
| 16 |
+
"loss": 7.8747,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2881844380403458,
|
| 21 |
+
"grad_norm": 36.96730422973633,
|
| 22 |
+
"learning_rate": 2.8160919540229887e-06,
|
| 23 |
+
"loss": 7.4611,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4322766570605187,
|
| 28 |
+
"grad_norm": 36.00823974609375,
|
| 29 |
+
"learning_rate": 4.252873563218391e-06,
|
| 30 |
+
"loss": 7.0067,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5763688760806917,
|
| 35 |
+
"grad_norm": 32.074058532714844,
|
| 36 |
+
"learning_rate": 5.689655172413794e-06,
|
| 37 |
+
"loss": 6.4648,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7204610951008645,
|
| 42 |
+
"grad_norm": 31.639789581298828,
|
| 43 |
+
"learning_rate": 7.126436781609196e-06,
|
| 44 |
+
"loss": 5.9512,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8645533141210374,
|
| 49 |
+
"grad_norm": 35.208580017089844,
|
| 50 |
+
"learning_rate": 8.563218390804599e-06,
|
| 51 |
+
"loss": 5.1826,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.5816326530612245,
|
| 57 |
+
"eval_loss": 1.0798285007476807,
|
| 58 |
+
"eval_macro_f1": 0.4484534425183768,
|
| 59 |
+
"eval_runtime": 1.0789,
|
| 60 |
+
"eval_samples_per_second": 635.842,
|
| 61 |
+
"eval_steps_per_second": 39.856,
|
| 62 |
+
"eval_weighted_f1": 0.5564766125357652,
|
| 63 |
+
"step": 174
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.005763688760807,
|
| 67 |
+
"grad_norm": 30.983509063720703,
|
| 68 |
+
"learning_rate": 1e-05,
|
| 69 |
+
"loss": 4.6204,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1498559077809798,
|
| 74 |
+
"grad_norm": 31.968477249145508,
|
| 75 |
+
"learning_rate": 1.1436781609195405e-05,
|
| 76 |
+
"loss": 4.2192,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2939481268011528,
|
| 81 |
+
"grad_norm": 33.485923767089844,
|
| 82 |
+
"learning_rate": 1.2873563218390805e-05,
|
| 83 |
+
"loss": 3.8234,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4380403458213258,
|
| 88 |
+
"grad_norm": 33.73719787597656,
|
| 89 |
+
"learning_rate": 1.4310344827586209e-05,
|
| 90 |
+
"loss": 3.7014,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5821325648414986,
|
| 95 |
+
"grad_norm": 38.27572250366211,
|
| 96 |
+
"learning_rate": 1.574712643678161e-05,
|
| 97 |
+
"loss": 3.6003,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7262247838616713,
|
| 102 |
+
"grad_norm": 32.76981735229492,
|
| 103 |
+
"learning_rate": 1.7183908045977015e-05,
|
| 104 |
+
"loss": 3.5286,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.8703170028818443,
|
| 109 |
+
"grad_norm": 35.05574417114258,
|
| 110 |
+
"learning_rate": 1.8620689655172415e-05,
|
| 111 |
+
"loss": 3.5884,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.6574344023323615,
|
| 117 |
+
"eval_loss": 0.925201416015625,
|
| 118 |
+
"eval_macro_f1": 0.5544747298968196,
|
| 119 |
+
"eval_runtime": 1.1164,
|
| 120 |
+
"eval_samples_per_second": 614.468,
|
| 121 |
+
"eval_steps_per_second": 38.516,
|
| 122 |
+
"eval_weighted_f1": 0.6443525684514556,
|
| 123 |
+
"step": 348
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.011527377521614,
|
| 127 |
+
"grad_norm": 32.71429443359375,
|
| 128 |
+
"learning_rate": 1.999361430395913e-05,
|
| 129 |
+
"loss": 3.3028,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.155619596541787,
|
| 134 |
+
"grad_norm": 24.289871215820312,
|
| 135 |
+
"learning_rate": 1.983397190293742e-05,
|
| 136 |
+
"loss": 2.1476,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2997118155619596,
|
| 141 |
+
"grad_norm": 32.85432815551758,
|
| 142 |
+
"learning_rate": 1.967432950191571e-05,
|
| 143 |
+
"loss": 2.3055,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.443804034582133,
|
| 148 |
+
"grad_norm": 29.986528396606445,
|
| 149 |
+
"learning_rate": 1.9514687100894e-05,
|
| 150 |
+
"loss": 2.2233,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5878962536023056,
|
| 155 |
+
"grad_norm": 39.08875274658203,
|
| 156 |
+
"learning_rate": 1.9355044699872286e-05,
|
| 157 |
+
"loss": 2.5057,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.7319884726224783,
|
| 162 |
+
"grad_norm": 30.903278350830078,
|
| 163 |
+
"learning_rate": 1.9195402298850576e-05,
|
| 164 |
+
"loss": 2.3106,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8760806916426516,
|
| 169 |
+
"grad_norm": 43.84492111206055,
|
| 170 |
+
"learning_rate": 1.9035759897828865e-05,
|
| 171 |
+
"loss": 2.3182,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"eval_accuracy": 0.6501457725947521,
|
| 177 |
+
"eval_loss": 0.9864262342453003,
|
| 178 |
+
"eval_macro_f1": 0.5999468077687296,
|
| 179 |
+
"eval_runtime": 1.2267,
|
| 180 |
+
"eval_samples_per_second": 559.218,
|
| 181 |
+
"eval_steps_per_second": 35.053,
|
| 182 |
+
"eval_weighted_f1": 0.6491266975171134,
|
| 183 |
+
"step": 522
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0172910662824206,
|
| 187 |
+
"grad_norm": 33.274166107177734,
|
| 188 |
+
"learning_rate": 1.8876117496807155e-05,
|
| 189 |
+
"loss": 2.0816,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.161383285302594,
|
| 194 |
+
"grad_norm": 25.669147491455078,
|
| 195 |
+
"learning_rate": 1.871647509578544e-05,
|
| 196 |
+
"loss": 1.185,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.3054755043227666,
|
| 201 |
+
"grad_norm": 19.854557037353516,
|
| 202 |
+
"learning_rate": 1.855683269476373e-05,
|
| 203 |
+
"loss": 1.0029,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.4495677233429394,
|
| 208 |
+
"grad_norm": 25.799800872802734,
|
| 209 |
+
"learning_rate": 1.839719029374202e-05,
|
| 210 |
+
"loss": 1.0524,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5936599423631126,
|
| 215 |
+
"grad_norm": 33.403865814208984,
|
| 216 |
+
"learning_rate": 1.823754789272031e-05,
|
| 217 |
+
"loss": 0.8661,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.7377521613832854,
|
| 222 |
+
"grad_norm": 28.626461029052734,
|
| 223 |
+
"learning_rate": 1.8077905491698596e-05,
|
| 224 |
+
"loss": 1.2073,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.881844380403458,
|
| 229 |
+
"grad_norm": 42.90084457397461,
|
| 230 |
+
"learning_rate": 1.7918263090676886e-05,
|
| 231 |
+
"loss": 0.8702,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"eval_accuracy": 0.6574344023323615,
|
| 237 |
+
"eval_loss": 1.2197622060775757,
|
| 238 |
+
"eval_macro_f1": 0.5987751645847678,
|
| 239 |
+
"eval_runtime": 1.1058,
|
| 240 |
+
"eval_samples_per_second": 620.363,
|
| 241 |
+
"eval_steps_per_second": 38.886,
|
| 242 |
+
"eval_weighted_f1": 0.6526032923630904,
|
| 243 |
+
"step": 696
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.023054755043228,
|
| 247 |
+
"grad_norm": 21.690303802490234,
|
| 248 |
+
"learning_rate": 1.7758620689655175e-05,
|
| 249 |
+
"loss": 0.8188,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.1671469740634,
|
| 254 |
+
"grad_norm": 7.896020889282227,
|
| 255 |
+
"learning_rate": 1.759897828863346e-05,
|
| 256 |
+
"loss": 0.3972,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.311239193083574,
|
| 261 |
+
"grad_norm": 26.9348087310791,
|
| 262 |
+
"learning_rate": 1.743933588761175e-05,
|
| 263 |
+
"loss": 0.4757,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.455331412103746,
|
| 268 |
+
"grad_norm": 6.688484191894531,
|
| 269 |
+
"learning_rate": 1.7279693486590037e-05,
|
| 270 |
+
"loss": 0.3661,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.599423631123919,
|
| 275 |
+
"grad_norm": 29.710420608520508,
|
| 276 |
+
"learning_rate": 1.7120051085568327e-05,
|
| 277 |
+
"loss": 0.3892,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.743515850144092,
|
| 282 |
+
"grad_norm": 50.58494186401367,
|
| 283 |
+
"learning_rate": 1.6960408684546617e-05,
|
| 284 |
+
"loss": 0.4685,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.887608069164266,
|
| 289 |
+
"grad_norm": 16.130414962768555,
|
| 290 |
+
"learning_rate": 1.6800766283524906e-05,
|
| 291 |
+
"loss": 0.4359,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"eval_accuracy": 0.641399416909621,
|
| 297 |
+
"eval_loss": 1.4822856187820435,
|
| 298 |
+
"eval_macro_f1": 0.5858791071664031,
|
| 299 |
+
"eval_runtime": 1.276,
|
| 300 |
+
"eval_samples_per_second": 537.603,
|
| 301 |
+
"eval_steps_per_second": 33.698,
|
| 302 |
+
"eval_weighted_f1": 0.6389097283655994,
|
| 303 |
+
"step": 870
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.028818443804035,
|
| 307 |
+
"grad_norm": 24.854923248291016,
|
| 308 |
+
"learning_rate": 1.6641123882503192e-05,
|
| 309 |
+
"loss": 0.3586,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.172910662824208,
|
| 314 |
+
"grad_norm": 3.022062301635742,
|
| 315 |
+
"learning_rate": 1.6481481481481482e-05,
|
| 316 |
+
"loss": 0.1867,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.31700288184438,
|
| 321 |
+
"grad_norm": 10.752510070800781,
|
| 322 |
+
"learning_rate": 1.632183908045977e-05,
|
| 323 |
+
"loss": 0.1332,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.461095100864553,
|
| 328 |
+
"grad_norm": 36.71868896484375,
|
| 329 |
+
"learning_rate": 1.616219667943806e-05,
|
| 330 |
+
"loss": 0.1336,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.605187319884726,
|
| 335 |
+
"grad_norm": 2.1257846355438232,
|
| 336 |
+
"learning_rate": 1.6002554278416347e-05,
|
| 337 |
+
"loss": 0.1929,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.749279538904899,
|
| 342 |
+
"grad_norm": 66.12401580810547,
|
| 343 |
+
"learning_rate": 1.5842911877394637e-05,
|
| 344 |
+
"loss": 0.1425,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.893371757925072,
|
| 349 |
+
"grad_norm": 3.548274040222168,
|
| 350 |
+
"learning_rate": 1.5683269476372927e-05,
|
| 351 |
+
"loss": 0.1632,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"eval_accuracy": 0.6516034985422741,
|
| 357 |
+
"eval_loss": 1.8195829391479492,
|
| 358 |
+
"eval_macro_f1": 0.612226437302956,
|
| 359 |
+
"eval_runtime": 1.1185,
|
| 360 |
+
"eval_samples_per_second": 613.342,
|
| 361 |
+
"eval_steps_per_second": 38.446,
|
| 362 |
+
"eval_weighted_f1": 0.6523423198710622,
|
| 363 |
+
"step": 1044
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.034582132564841,
|
| 367 |
+
"grad_norm": 29.878677368164062,
|
| 368 |
+
"learning_rate": 1.5523627075351216e-05,
|
| 369 |
+
"loss": 0.2029,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.178674351585014,
|
| 374 |
+
"grad_norm": 14.024438858032227,
|
| 375 |
+
"learning_rate": 1.5363984674329502e-05,
|
| 376 |
+
"loss": 0.0803,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.322766570605188,
|
| 381 |
+
"grad_norm": 37.29269790649414,
|
| 382 |
+
"learning_rate": 1.5204342273307792e-05,
|
| 383 |
+
"loss": 0.0752,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.46685878962536,
|
| 388 |
+
"grad_norm": 18.938222885131836,
|
| 389 |
+
"learning_rate": 1.504469987228608e-05,
|
| 390 |
+
"loss": 0.0853,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.610951008645533,
|
| 395 |
+
"grad_norm": 19.586759567260742,
|
| 396 |
+
"learning_rate": 1.4885057471264368e-05,
|
| 397 |
+
"loss": 0.0575,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.755043227665706,
|
| 402 |
+
"grad_norm": 3.7584567070007324,
|
| 403 |
+
"learning_rate": 1.4725415070242657e-05,
|
| 404 |
+
"loss": 0.063,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.899135446685879,
|
| 409 |
+
"grad_norm": 37.0729866027832,
|
| 410 |
+
"learning_rate": 1.4565772669220945e-05,
|
| 411 |
+
"loss": 0.1913,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 7.0,
|
| 416 |
+
"eval_accuracy": 0.6720116618075802,
|
| 417 |
+
"eval_loss": 2.132646322250366,
|
| 418 |
+
"eval_macro_f1": 0.6293701407785159,
|
| 419 |
+
"eval_runtime": 1.1621,
|
| 420 |
+
"eval_samples_per_second": 590.317,
|
| 421 |
+
"eval_steps_per_second": 37.002,
|
| 422 |
+
"eval_weighted_f1": 0.6703315130630516,
|
| 423 |
+
"step": 1218
|
| 424 |
+
}
|
| 425 |
+
],
|
| 426 |
+
"logging_steps": 25,
|
| 427 |
+
"max_steps": 3480,
|
| 428 |
+
"num_input_tokens_seen": 0,
|
| 429 |
+
"num_train_epochs": 20,
|
| 430 |
+
"save_steps": 500,
|
| 431 |
+
"stateful_callbacks": {
|
| 432 |
+
"EarlyStoppingCallback": {
|
| 433 |
+
"args": {
|
| 434 |
+
"early_stopping_patience": 3,
|
| 435 |
+
"early_stopping_threshold": 0.0
|
| 436 |
+
},
|
| 437 |
+
"attributes": {
|
| 438 |
+
"early_stopping_patience_counter": 0
|
| 439 |
+
}
|
| 440 |
+
},
|
| 441 |
+
"TrainerControl": {
|
| 442 |
+
"args": {
|
| 443 |
+
"should_epoch_stop": false,
|
| 444 |
+
"should_evaluate": false,
|
| 445 |
+
"should_log": false,
|
| 446 |
+
"should_save": true,
|
| 447 |
+
"should_training_stop": false
|
| 448 |
+
},
|
| 449 |
+
"attributes": {}
|
| 450 |
+
}
|
| 451 |
+
},
|
| 452 |
+
"total_flos": 0.0,
|
| 453 |
+
"train_batch_size": 8,
|
| 454 |
+
"trial_name": null,
|
| 455 |
+
"trial_params": null
|
| 456 |
+
}
|
uit_vsmec-visobert-s44-20260829T184728Z-863c2de1/trainer/checkpoint-2784/trainer_state.json
ADDED
|
@@ -0,0 +1,996 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2262,
|
| 3 |
+
"best_metric": 0.637598474535291,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsmec-visobert-s44-20260829T184728Z-863c2de1/trainer/checkpoint-2262",
|
| 5 |
+
"epoch": 16.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2784,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1440922190201729,
|
| 14 |
+
"grad_norm": 40.45027160644531,
|
| 15 |
+
"learning_rate": 1.3793103448275862e-06,
|
| 16 |
+
"loss": 8.0136,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2881844380403458,
|
| 21 |
+
"grad_norm": 32.17760467529297,
|
| 22 |
+
"learning_rate": 2.8160919540229887e-06,
|
| 23 |
+
"loss": 7.3676,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4322766570605187,
|
| 28 |
+
"grad_norm": 31.06029510498047,
|
| 29 |
+
"learning_rate": 4.252873563218391e-06,
|
| 30 |
+
"loss": 6.8516,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5763688760806917,
|
| 35 |
+
"grad_norm": 28.91082191467285,
|
| 36 |
+
"learning_rate": 5.689655172413794e-06,
|
| 37 |
+
"loss": 6.3662,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7204610951008645,
|
| 42 |
+
"grad_norm": 38.7941780090332,
|
| 43 |
+
"learning_rate": 7.126436781609196e-06,
|
| 44 |
+
"loss": 5.7246,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8645533141210374,
|
| 49 |
+
"grad_norm": 30.64776039123535,
|
| 50 |
+
"learning_rate": 8.563218390804599e-06,
|
| 51 |
+
"loss": 5.2698,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.5962099125364432,
|
| 57 |
+
"eval_loss": 1.0838309526443481,
|
| 58 |
+
"eval_macro_f1": 0.4405885745011722,
|
| 59 |
+
"eval_runtime": 1.3247,
|
| 60 |
+
"eval_samples_per_second": 517.852,
|
| 61 |
+
"eval_steps_per_second": 32.46,
|
| 62 |
+
"eval_weighted_f1": 0.5691508225252657,
|
| 63 |
+
"step": 174
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.005763688760807,
|
| 67 |
+
"grad_norm": 35.94602966308594,
|
| 68 |
+
"learning_rate": 1e-05,
|
| 69 |
+
"loss": 4.4774,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1498559077809798,
|
| 74 |
+
"grad_norm": 43.05365753173828,
|
| 75 |
+
"learning_rate": 1.1436781609195405e-05,
|
| 76 |
+
"loss": 3.9836,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2939481268011528,
|
| 81 |
+
"grad_norm": 30.314823150634766,
|
| 82 |
+
"learning_rate": 1.2873563218390805e-05,
|
| 83 |
+
"loss": 3.7859,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4380403458213258,
|
| 88 |
+
"grad_norm": 33.88893127441406,
|
| 89 |
+
"learning_rate": 1.4310344827586209e-05,
|
| 90 |
+
"loss": 3.6843,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5821325648414986,
|
| 95 |
+
"grad_norm": 32.667911529541016,
|
| 96 |
+
"learning_rate": 1.574712643678161e-05,
|
| 97 |
+
"loss": 3.7409,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7262247838616713,
|
| 102 |
+
"grad_norm": 40.5688591003418,
|
| 103 |
+
"learning_rate": 1.7183908045977015e-05,
|
| 104 |
+
"loss": 3.5667,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.8703170028818443,
|
| 109 |
+
"grad_norm": 35.93356704711914,
|
| 110 |
+
"learning_rate": 1.8620689655172415e-05,
|
| 111 |
+
"loss": 3.5757,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.6326530612244898,
|
| 117 |
+
"eval_loss": 0.956699013710022,
|
| 118 |
+
"eval_macro_f1": 0.5783879033678574,
|
| 119 |
+
"eval_runtime": 1.1307,
|
| 120 |
+
"eval_samples_per_second": 606.68,
|
| 121 |
+
"eval_steps_per_second": 38.028,
|
| 122 |
+
"eval_weighted_f1": 0.6316722186823505,
|
| 123 |
+
"step": 348
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.011527377521614,
|
| 127 |
+
"grad_norm": 23.907941818237305,
|
| 128 |
+
"learning_rate": 1.999361430395913e-05,
|
| 129 |
+
"loss": 3.4776,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.155619596541787,
|
| 134 |
+
"grad_norm": 29.50314712524414,
|
| 135 |
+
"learning_rate": 1.983397190293742e-05,
|
| 136 |
+
"loss": 2.3515,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2997118155619596,
|
| 141 |
+
"grad_norm": 30.915699005126953,
|
| 142 |
+
"learning_rate": 1.967432950191571e-05,
|
| 143 |
+
"loss": 2.2055,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.443804034582133,
|
| 148 |
+
"grad_norm": 41.33200454711914,
|
| 149 |
+
"learning_rate": 1.9514687100894e-05,
|
| 150 |
+
"loss": 2.2972,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5878962536023056,
|
| 155 |
+
"grad_norm": 26.23257064819336,
|
| 156 |
+
"learning_rate": 1.9355044699872286e-05,
|
| 157 |
+
"loss": 2.0362,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.7319884726224783,
|
| 162 |
+
"grad_norm": 45.42691421508789,
|
| 163 |
+
"learning_rate": 1.9195402298850576e-05,
|
| 164 |
+
"loss": 2.5396,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8760806916426516,
|
| 169 |
+
"grad_norm": 39.118629455566406,
|
| 170 |
+
"learning_rate": 1.9035759897828865e-05,
|
| 171 |
+
"loss": 2.1574,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"eval_accuracy": 0.6618075801749271,
|
| 177 |
+
"eval_loss": 1.0008624792099,
|
| 178 |
+
"eval_macro_f1": 0.6084706214511898,
|
| 179 |
+
"eval_runtime": 1.313,
|
| 180 |
+
"eval_samples_per_second": 522.452,
|
| 181 |
+
"eval_steps_per_second": 32.748,
|
| 182 |
+
"eval_weighted_f1": 0.6534163014910526,
|
| 183 |
+
"step": 522
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0172910662824206,
|
| 187 |
+
"grad_norm": 21.546226501464844,
|
| 188 |
+
"learning_rate": 1.8876117496807155e-05,
|
| 189 |
+
"loss": 1.9429,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.161383285302594,
|
| 194 |
+
"grad_norm": 32.71604537963867,
|
| 195 |
+
"learning_rate": 1.871647509578544e-05,
|
| 196 |
+
"loss": 1.1974,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.3054755043227666,
|
| 201 |
+
"grad_norm": 29.053213119506836,
|
| 202 |
+
"learning_rate": 1.855683269476373e-05,
|
| 203 |
+
"loss": 1.0577,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.4495677233429394,
|
| 208 |
+
"grad_norm": 28.296419143676758,
|
| 209 |
+
"learning_rate": 1.839719029374202e-05,
|
| 210 |
+
"loss": 0.9895,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5936599423631126,
|
| 215 |
+
"grad_norm": 33.610748291015625,
|
| 216 |
+
"learning_rate": 1.823754789272031e-05,
|
| 217 |
+
"loss": 1.0258,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.7377521613832854,
|
| 222 |
+
"grad_norm": 47.01686477661133,
|
| 223 |
+
"learning_rate": 1.8077905491698596e-05,
|
| 224 |
+
"loss": 0.9368,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.881844380403458,
|
| 229 |
+
"grad_norm": 30.013246536254883,
|
| 230 |
+
"learning_rate": 1.7918263090676886e-05,
|
| 231 |
+
"loss": 1.0416,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"eval_accuracy": 0.6647230320699709,
|
| 237 |
+
"eval_loss": 1.19782292842865,
|
| 238 |
+
"eval_macro_f1": 0.6094987634437494,
|
| 239 |
+
"eval_runtime": 1.0721,
|
| 240 |
+
"eval_samples_per_second": 639.839,
|
| 241 |
+
"eval_steps_per_second": 40.107,
|
| 242 |
+
"eval_weighted_f1": 0.6611207551300284,
|
| 243 |
+
"step": 696
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.023054755043228,
|
| 247 |
+
"grad_norm": 30.42293357849121,
|
| 248 |
+
"learning_rate": 1.7758620689655175e-05,
|
| 249 |
+
"loss": 0.9359,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.1671469740634,
|
| 254 |
+
"grad_norm": 25.70184326171875,
|
| 255 |
+
"learning_rate": 1.759897828863346e-05,
|
| 256 |
+
"loss": 0.3643,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.311239193083574,
|
| 261 |
+
"grad_norm": 15.36362075805664,
|
| 262 |
+
"learning_rate": 1.743933588761175e-05,
|
| 263 |
+
"loss": 0.3725,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.455331412103746,
|
| 268 |
+
"grad_norm": 6.803077220916748,
|
| 269 |
+
"learning_rate": 1.7279693486590037e-05,
|
| 270 |
+
"loss": 0.3143,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.599423631123919,
|
| 275 |
+
"grad_norm": 30.110551834106445,
|
| 276 |
+
"learning_rate": 1.7120051085568327e-05,
|
| 277 |
+
"loss": 0.3664,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.743515850144092,
|
| 282 |
+
"grad_norm": 40.79981231689453,
|
| 283 |
+
"learning_rate": 1.6960408684546617e-05,
|
| 284 |
+
"loss": 0.4078,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.887608069164266,
|
| 289 |
+
"grad_norm": 23.093769073486328,
|
| 290 |
+
"learning_rate": 1.6800766283524906e-05,
|
| 291 |
+
"loss": 0.443,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"eval_accuracy": 0.6559766763848397,
|
| 297 |
+
"eval_loss": 1.53959059715271,
|
| 298 |
+
"eval_macro_f1": 0.6025615219039925,
|
| 299 |
+
"eval_runtime": 1.0331,
|
| 300 |
+
"eval_samples_per_second": 664.012,
|
| 301 |
+
"eval_steps_per_second": 41.622,
|
| 302 |
+
"eval_weighted_f1": 0.6504290426860866,
|
| 303 |
+
"step": 870
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.028818443804035,
|
| 307 |
+
"grad_norm": 9.220955848693848,
|
| 308 |
+
"learning_rate": 1.6641123882503192e-05,
|
| 309 |
+
"loss": 0.3286,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.172910662824208,
|
| 314 |
+
"grad_norm": 1.4264723062515259,
|
| 315 |
+
"learning_rate": 1.6481481481481482e-05,
|
| 316 |
+
"loss": 0.1547,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.31700288184438,
|
| 321 |
+
"grad_norm": 6.127079963684082,
|
| 322 |
+
"learning_rate": 1.632183908045977e-05,
|
| 323 |
+
"loss": 0.1457,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.461095100864553,
|
| 328 |
+
"grad_norm": 16.20818328857422,
|
| 329 |
+
"learning_rate": 1.616219667943806e-05,
|
| 330 |
+
"loss": 0.1483,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.605187319884726,
|
| 335 |
+
"grad_norm": 1.372633695602417,
|
| 336 |
+
"learning_rate": 1.6002554278416347e-05,
|
| 337 |
+
"loss": 0.16,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.749279538904899,
|
| 342 |
+
"grad_norm": 13.250823020935059,
|
| 343 |
+
"learning_rate": 1.5842911877394637e-05,
|
| 344 |
+
"loss": 0.1373,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.893371757925072,
|
| 349 |
+
"grad_norm": 11.020379066467285,
|
| 350 |
+
"learning_rate": 1.5683269476372927e-05,
|
| 351 |
+
"loss": 0.2921,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"eval_accuracy": 0.6603498542274052,
|
| 357 |
+
"eval_loss": 1.8537579774856567,
|
| 358 |
+
"eval_macro_f1": 0.6209504120577762,
|
| 359 |
+
"eval_runtime": 1.102,
|
| 360 |
+
"eval_samples_per_second": 622.483,
|
| 361 |
+
"eval_steps_per_second": 39.019,
|
| 362 |
+
"eval_weighted_f1": 0.6605366930199218,
|
| 363 |
+
"step": 1044
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.034582132564841,
|
| 367 |
+
"grad_norm": 1.9109667539596558,
|
| 368 |
+
"learning_rate": 1.5523627075351216e-05,
|
| 369 |
+
"loss": 0.1709,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.178674351585014,
|
| 374 |
+
"grad_norm": 3.536971092224121,
|
| 375 |
+
"learning_rate": 1.5363984674329502e-05,
|
| 376 |
+
"loss": 0.0659,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.322766570605188,
|
| 381 |
+
"grad_norm": 6.847039699554443,
|
| 382 |
+
"learning_rate": 1.5204342273307792e-05,
|
| 383 |
+
"loss": 0.0814,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.46685878962536,
|
| 388 |
+
"grad_norm": 36.487308502197266,
|
| 389 |
+
"learning_rate": 1.504469987228608e-05,
|
| 390 |
+
"loss": 0.042,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.610951008645533,
|
| 395 |
+
"grad_norm": 15.233339309692383,
|
| 396 |
+
"learning_rate": 1.4885057471264368e-05,
|
| 397 |
+
"loss": 0.0763,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.755043227665706,
|
| 402 |
+
"grad_norm": 0.16093553602695465,
|
| 403 |
+
"learning_rate": 1.4725415070242657e-05,
|
| 404 |
+
"loss": 0.0551,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.899135446685879,
|
| 409 |
+
"grad_norm": 20.91053581237793,
|
| 410 |
+
"learning_rate": 1.4565772669220945e-05,
|
| 411 |
+
"loss": 0.0668,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 7.0,
|
| 416 |
+
"eval_accuracy": 0.6705539358600583,
|
| 417 |
+
"eval_loss": 2.1843667030334473,
|
| 418 |
+
"eval_macro_f1": 0.6169649727681762,
|
| 419 |
+
"eval_runtime": 1.0985,
|
| 420 |
+
"eval_samples_per_second": 624.509,
|
| 421 |
+
"eval_steps_per_second": 39.146,
|
| 422 |
+
"eval_weighted_f1": 0.6651173598130018,
|
| 423 |
+
"step": 1218
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.040345821325649,
|
| 427 |
+
"grad_norm": 0.4691203832626343,
|
| 428 |
+
"learning_rate": 1.4406130268199235e-05,
|
| 429 |
+
"loss": 0.065,
|
| 430 |
+
"step": 1225
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.184438040345821,
|
| 434 |
+
"grad_norm": 1.504001259803772,
|
| 435 |
+
"learning_rate": 1.4246487867177523e-05,
|
| 436 |
+
"loss": 0.013,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.328530259365994,
|
| 441 |
+
"grad_norm": 1.9196845293045044,
|
| 442 |
+
"learning_rate": 1.4086845466155812e-05,
|
| 443 |
+
"loss": 0.0851,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.472622478386167,
|
| 448 |
+
"grad_norm": 56.49351119995117,
|
| 449 |
+
"learning_rate": 1.39272030651341e-05,
|
| 450 |
+
"loss": 0.0712,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.61671469740634,
|
| 455 |
+
"grad_norm": 0.2787981927394867,
|
| 456 |
+
"learning_rate": 1.376756066411239e-05,
|
| 457 |
+
"loss": 0.0296,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.760806916426513,
|
| 462 |
+
"grad_norm": 12.182669639587402,
|
| 463 |
+
"learning_rate": 1.3607918263090678e-05,
|
| 464 |
+
"loss": 0.0163,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.904899135446686,
|
| 469 |
+
"grad_norm": 0.42971697449684143,
|
| 470 |
+
"learning_rate": 1.3448275862068967e-05,
|
| 471 |
+
"loss": 0.0616,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.0,
|
| 476 |
+
"eval_accuracy": 0.6661807580174927,
|
| 477 |
+
"eval_loss": 2.3372700214385986,
|
| 478 |
+
"eval_macro_f1": 0.6166848665531429,
|
| 479 |
+
"eval_runtime": 1.1236,
|
| 480 |
+
"eval_samples_per_second": 610.517,
|
| 481 |
+
"eval_steps_per_second": 38.269,
|
| 482 |
+
"eval_weighted_f1": 0.6659107567577827,
|
| 483 |
+
"step": 1392
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 8.046109510086456,
|
| 487 |
+
"grad_norm": 6.766666889190674,
|
| 488 |
+
"learning_rate": 1.3288633461047255e-05,
|
| 489 |
+
"loss": 0.0209,
|
| 490 |
+
"step": 1400
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.190201729106628,
|
| 494 |
+
"grad_norm": 0.30378708243370056,
|
| 495 |
+
"learning_rate": 1.3128991060025545e-05,
|
| 496 |
+
"loss": 0.0119,
|
| 497 |
+
"step": 1425
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.3342939481268,
|
| 501 |
+
"grad_norm": 0.11394540965557098,
|
| 502 |
+
"learning_rate": 1.2969348659003833e-05,
|
| 503 |
+
"loss": 0.0141,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.478386167146974,
|
| 508 |
+
"grad_norm": 0.9965255260467529,
|
| 509 |
+
"learning_rate": 1.2809706257982123e-05,
|
| 510 |
+
"loss": 0.0155,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.622478386167147,
|
| 515 |
+
"grad_norm": 12.061532020568848,
|
| 516 |
+
"learning_rate": 1.265006385696041e-05,
|
| 517 |
+
"loss": 0.0408,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.76657060518732,
|
| 522 |
+
"grad_norm": 0.26100677251815796,
|
| 523 |
+
"learning_rate": 1.24904214559387e-05,
|
| 524 |
+
"loss": 0.0074,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.910662824207492,
|
| 529 |
+
"grad_norm": 0.2961542308330536,
|
| 530 |
+
"learning_rate": 1.2330779054916986e-05,
|
| 531 |
+
"loss": 0.0125,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 9.0,
|
| 536 |
+
"eval_accuracy": 0.6676384839650146,
|
| 537 |
+
"eval_loss": 2.4361181259155273,
|
| 538 |
+
"eval_macro_f1": 0.6306161152930151,
|
| 539 |
+
"eval_runtime": 1.1276,
|
| 540 |
+
"eval_samples_per_second": 608.395,
|
| 541 |
+
"eval_steps_per_second": 38.136,
|
| 542 |
+
"eval_weighted_f1": 0.6664173681447428,
|
| 543 |
+
"step": 1566
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.051873198847263,
|
| 547 |
+
"grad_norm": 0.03754047676920891,
|
| 548 |
+
"learning_rate": 1.2171136653895274e-05,
|
| 549 |
+
"loss": 0.0148,
|
| 550 |
+
"step": 1575
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.195965417867436,
|
| 554 |
+
"grad_norm": 19.575042724609375,
|
| 555 |
+
"learning_rate": 1.2011494252873564e-05,
|
| 556 |
+
"loss": 0.0038,
|
| 557 |
+
"step": 1600
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.340057636887607,
|
| 561 |
+
"grad_norm": 0.04034111648797989,
|
| 562 |
+
"learning_rate": 1.1851851851851852e-05,
|
| 563 |
+
"loss": 0.0057,
|
| 564 |
+
"step": 1625
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.48414985590778,
|
| 568 |
+
"grad_norm": 0.13508199155330658,
|
| 569 |
+
"learning_rate": 1.1692209450830141e-05,
|
| 570 |
+
"loss": 0.0352,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.628242074927954,
|
| 575 |
+
"grad_norm": 0.021814337000250816,
|
| 576 |
+
"learning_rate": 1.1532567049808429e-05,
|
| 577 |
+
"loss": 0.015,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.772334293948127,
|
| 582 |
+
"grad_norm": 0.026802001520991325,
|
| 583 |
+
"learning_rate": 1.1372924648786719e-05,
|
| 584 |
+
"loss": 0.0061,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.9164265129683,
|
| 589 |
+
"grad_norm": 1.7753314971923828,
|
| 590 |
+
"learning_rate": 1.1213282247765007e-05,
|
| 591 |
+
"loss": 0.0037,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 10.0,
|
| 596 |
+
"eval_accuracy": 0.6822157434402333,
|
| 597 |
+
"eval_loss": 2.56136155128479,
|
| 598 |
+
"eval_macro_f1": 0.6352333829256098,
|
| 599 |
+
"eval_runtime": 1.0515,
|
| 600 |
+
"eval_samples_per_second": 652.389,
|
| 601 |
+
"eval_steps_per_second": 40.893,
|
| 602 |
+
"eval_weighted_f1": 0.6801640503920754,
|
| 603 |
+
"step": 1740
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.05763688760807,
|
| 607 |
+
"grad_norm": 0.09404151141643524,
|
| 608 |
+
"learning_rate": 1.1053639846743296e-05,
|
| 609 |
+
"loss": 0.02,
|
| 610 |
+
"step": 1750
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 10.201729106628243,
|
| 614 |
+
"grad_norm": 8.37583065032959,
|
| 615 |
+
"learning_rate": 1.0893997445721584e-05,
|
| 616 |
+
"loss": 0.0532,
|
| 617 |
+
"step": 1775
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 10.345821325648416,
|
| 621 |
+
"grad_norm": 0.22101309895515442,
|
| 622 |
+
"learning_rate": 1.0734355044699874e-05,
|
| 623 |
+
"loss": 0.0032,
|
| 624 |
+
"step": 1800
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.489913544668587,
|
| 628 |
+
"grad_norm": 0.024156736209988594,
|
| 629 |
+
"learning_rate": 1.0574712643678162e-05,
|
| 630 |
+
"loss": 0.0547,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.63400576368876,
|
| 635 |
+
"grad_norm": 0.0766325294971466,
|
| 636 |
+
"learning_rate": 1.0415070242656451e-05,
|
| 637 |
+
"loss": 0.0009,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.778097982708934,
|
| 642 |
+
"grad_norm": 0.0264886487275362,
|
| 643 |
+
"learning_rate": 1.025542784163474e-05,
|
| 644 |
+
"loss": 0.0007,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.922190201729107,
|
| 649 |
+
"grad_norm": 0.9045348763465881,
|
| 650 |
+
"learning_rate": 1.0095785440613029e-05,
|
| 651 |
+
"loss": 0.006,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 11.0,
|
| 656 |
+
"eval_accuracy": 0.6690962099125365,
|
| 657 |
+
"eval_loss": 2.6275720596313477,
|
| 658 |
+
"eval_macro_f1": 0.6223726448600777,
|
| 659 |
+
"eval_runtime": 1.0381,
|
| 660 |
+
"eval_samples_per_second": 660.834,
|
| 661 |
+
"eval_steps_per_second": 41.423,
|
| 662 |
+
"eval_weighted_f1": 0.6684647909703425,
|
| 663 |
+
"step": 1914
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 11.063400576368876,
|
| 667 |
+
"grad_norm": 0.15793076157569885,
|
| 668 |
+
"learning_rate": 9.936143039591317e-06,
|
| 669 |
+
"loss": 0.002,
|
| 670 |
+
"step": 1925
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 11.207492795389049,
|
| 674 |
+
"grad_norm": 0.006426218431442976,
|
| 675 |
+
"learning_rate": 9.776500638569605e-06,
|
| 676 |
+
"loss": 0.0013,
|
| 677 |
+
"step": 1950
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 11.351585014409222,
|
| 681 |
+
"grad_norm": 0.3995853364467621,
|
| 682 |
+
"learning_rate": 9.616858237547894e-06,
|
| 683 |
+
"loss": 0.0058,
|
| 684 |
+
"step": 1975
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 11.495677233429396,
|
| 688 |
+
"grad_norm": 0.023134293034672737,
|
| 689 |
+
"learning_rate": 9.457215836526182e-06,
|
| 690 |
+
"loss": 0.003,
|
| 691 |
+
"step": 2000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.639769452449567,
|
| 695 |
+
"grad_norm": 0.08001644909381866,
|
| 696 |
+
"learning_rate": 9.297573435504472e-06,
|
| 697 |
+
"loss": 0.0007,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.78386167146974,
|
| 702 |
+
"grad_norm": 0.023221734911203384,
|
| 703 |
+
"learning_rate": 9.13793103448276e-06,
|
| 704 |
+
"loss": 0.0091,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.927953890489913,
|
| 709 |
+
"grad_norm": 0.10720467567443848,
|
| 710 |
+
"learning_rate": 8.978288633461047e-06,
|
| 711 |
+
"loss": 0.0008,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 12.0,
|
| 716 |
+
"eval_accuracy": 0.6778425655976676,
|
| 717 |
+
"eval_loss": 2.6376121044158936,
|
| 718 |
+
"eval_macro_f1": 0.6293068527924854,
|
| 719 |
+
"eval_runtime": 1.0899,
|
| 720 |
+
"eval_samples_per_second": 629.392,
|
| 721 |
+
"eval_steps_per_second": 39.452,
|
| 722 |
+
"eval_weighted_f1": 0.6754674257965746,
|
| 723 |
+
"step": 2088
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.069164265129682,
|
| 727 |
+
"grad_norm": 0.026265041902661324,
|
| 728 |
+
"learning_rate": 8.818646232439337e-06,
|
| 729 |
+
"loss": 0.0672,
|
| 730 |
+
"step": 2100
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 12.213256484149856,
|
| 734 |
+
"grad_norm": 0.0390600711107254,
|
| 735 |
+
"learning_rate": 8.659003831417625e-06,
|
| 736 |
+
"loss": 0.0006,
|
| 737 |
+
"step": 2125
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 12.357348703170029,
|
| 741 |
+
"grad_norm": 0.09783770143985748,
|
| 742 |
+
"learning_rate": 8.499361430395915e-06,
|
| 743 |
+
"loss": 0.0218,
|
| 744 |
+
"step": 2150
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 12.501440922190202,
|
| 748 |
+
"grad_norm": 0.023970406502485275,
|
| 749 |
+
"learning_rate": 8.339719029374202e-06,
|
| 750 |
+
"loss": 0.0347,
|
| 751 |
+
"step": 2175
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.645533141210375,
|
| 755 |
+
"grad_norm": 0.08880403637886047,
|
| 756 |
+
"learning_rate": 8.180076628352492e-06,
|
| 757 |
+
"loss": 0.0015,
|
| 758 |
+
"step": 2200
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.789625360230547,
|
| 762 |
+
"grad_norm": 0.01197072584182024,
|
| 763 |
+
"learning_rate": 8.02043422733078e-06,
|
| 764 |
+
"loss": 0.001,
|
| 765 |
+
"step": 2225
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.93371757925072,
|
| 769 |
+
"grad_norm": 0.08891931176185608,
|
| 770 |
+
"learning_rate": 7.86079182630907e-06,
|
| 771 |
+
"loss": 0.0006,
|
| 772 |
+
"step": 2250
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 13.0,
|
| 776 |
+
"eval_accuracy": 0.6807580174927114,
|
| 777 |
+
"eval_loss": 2.6285221576690674,
|
| 778 |
+
"eval_macro_f1": 0.637598474535291,
|
| 779 |
+
"eval_runtime": 1.0562,
|
| 780 |
+
"eval_samples_per_second": 649.524,
|
| 781 |
+
"eval_steps_per_second": 40.714,
|
| 782 |
+
"eval_weighted_f1": 0.679078398096636,
|
| 783 |
+
"step": 2262
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 13.07492795389049,
|
| 787 |
+
"grad_norm": 0.012412037700414658,
|
| 788 |
+
"learning_rate": 7.701149425287356e-06,
|
| 789 |
+
"loss": 0.0104,
|
| 790 |
+
"step": 2275
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 13.219020172910662,
|
| 794 |
+
"grad_norm": 0.019839955493807793,
|
| 795 |
+
"learning_rate": 7.541507024265645e-06,
|
| 796 |
+
"loss": 0.0326,
|
| 797 |
+
"step": 2300
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 13.363112391930835,
|
| 801 |
+
"grad_norm": 0.019731566309928894,
|
| 802 |
+
"learning_rate": 7.381864623243934e-06,
|
| 803 |
+
"loss": 0.0004,
|
| 804 |
+
"step": 2325
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 13.507204610951009,
|
| 808 |
+
"grad_norm": 0.12854623794555664,
|
| 809 |
+
"learning_rate": 7.222222222222223e-06,
|
| 810 |
+
"loss": 0.001,
|
| 811 |
+
"step": 2350
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 13.651296829971182,
|
| 815 |
+
"grad_norm": 0.012279336340725422,
|
| 816 |
+
"learning_rate": 7.062579821200512e-06,
|
| 817 |
+
"loss": 0.0003,
|
| 818 |
+
"step": 2375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 13.795389048991355,
|
| 822 |
+
"grad_norm": 0.021951761096715927,
|
| 823 |
+
"learning_rate": 6.9029374201788e-06,
|
| 824 |
+
"loss": 0.0004,
|
| 825 |
+
"step": 2400
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 13.939481268011527,
|
| 829 |
+
"grad_norm": 0.25726187229156494,
|
| 830 |
+
"learning_rate": 6.743295019157089e-06,
|
| 831 |
+
"loss": 0.0035,
|
| 832 |
+
"step": 2425
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 14.0,
|
| 836 |
+
"eval_accuracy": 0.6822157434402333,
|
| 837 |
+
"eval_loss": 2.656416893005371,
|
| 838 |
+
"eval_macro_f1": 0.629416377925499,
|
| 839 |
+
"eval_runtime": 1.2906,
|
| 840 |
+
"eval_samples_per_second": 531.537,
|
| 841 |
+
"eval_steps_per_second": 33.318,
|
| 842 |
+
"eval_weighted_f1": 0.6788806203037029,
|
| 843 |
+
"step": 2436
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 14.080691642651297,
|
| 847 |
+
"grad_norm": 0.03417178988456726,
|
| 848 |
+
"learning_rate": 6.583652618135378e-06,
|
| 849 |
+
"loss": 0.0022,
|
| 850 |
+
"step": 2450
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 14.22478386167147,
|
| 854 |
+
"grad_norm": 0.016545239835977554,
|
| 855 |
+
"learning_rate": 6.424010217113667e-06,
|
| 856 |
+
"loss": 0.0027,
|
| 857 |
+
"step": 2475
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 14.368876080691642,
|
| 861 |
+
"grad_norm": 0.05657158046960831,
|
| 862 |
+
"learning_rate": 6.264367816091954e-06,
|
| 863 |
+
"loss": 0.0303,
|
| 864 |
+
"step": 2500
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 14.512968299711815,
|
| 868 |
+
"grad_norm": 0.011330910958349705,
|
| 869 |
+
"learning_rate": 6.1047254150702425e-06,
|
| 870 |
+
"loss": 0.0005,
|
| 871 |
+
"step": 2525
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 14.657060518731988,
|
| 875 |
+
"grad_norm": 0.01636470854282379,
|
| 876 |
+
"learning_rate": 5.945083014048531e-06,
|
| 877 |
+
"loss": 0.0003,
|
| 878 |
+
"step": 2550
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 14.801152737752162,
|
| 882 |
+
"grad_norm": 0.019321203231811523,
|
| 883 |
+
"learning_rate": 5.78544061302682e-06,
|
| 884 |
+
"loss": 0.0025,
|
| 885 |
+
"step": 2575
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 14.945244956772335,
|
| 889 |
+
"grad_norm": 0.011153306812047958,
|
| 890 |
+
"learning_rate": 5.625798212005109e-06,
|
| 891 |
+
"loss": 0.0004,
|
| 892 |
+
"step": 2600
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 15.0,
|
| 896 |
+
"eval_accuracy": 0.6749271137026239,
|
| 897 |
+
"eval_loss": 2.7105910778045654,
|
| 898 |
+
"eval_macro_f1": 0.6254400564663937,
|
| 899 |
+
"eval_runtime": 1.3102,
|
| 900 |
+
"eval_samples_per_second": 523.595,
|
| 901 |
+
"eval_steps_per_second": 32.82,
|
| 902 |
+
"eval_weighted_f1": 0.67065544336299,
|
| 903 |
+
"step": 2610
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 15.086455331412104,
|
| 907 |
+
"grad_norm": 0.19627001881599426,
|
| 908 |
+
"learning_rate": 5.4661558109833975e-06,
|
| 909 |
+
"loss": 0.0004,
|
| 910 |
+
"step": 2625
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 15.230547550432277,
|
| 914 |
+
"grad_norm": 0.01699591428041458,
|
| 915 |
+
"learning_rate": 5.306513409961686e-06,
|
| 916 |
+
"loss": 0.0002,
|
| 917 |
+
"step": 2650
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 15.37463976945245,
|
| 921 |
+
"grad_norm": 0.14286349713802338,
|
| 922 |
+
"learning_rate": 5.146871008939975e-06,
|
| 923 |
+
"loss": 0.0002,
|
| 924 |
+
"step": 2675
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 15.518731988472622,
|
| 928 |
+
"grad_norm": 0.007836956530809402,
|
| 929 |
+
"learning_rate": 4.987228607918264e-06,
|
| 930 |
+
"loss": 0.0003,
|
| 931 |
+
"step": 2700
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 15.662824207492795,
|
| 935 |
+
"grad_norm": 0.033788181841373444,
|
| 936 |
+
"learning_rate": 4.8275862068965525e-06,
|
| 937 |
+
"loss": 0.006,
|
| 938 |
+
"step": 2725
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 15.806916426512968,
|
| 942 |
+
"grad_norm": 0.23998883366584778,
|
| 943 |
+
"learning_rate": 4.66794380587484e-06,
|
| 944 |
+
"loss": 0.0211,
|
| 945 |
+
"step": 2750
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 15.951008645533141,
|
| 949 |
+
"grad_norm": 0.03011343628168106,
|
| 950 |
+
"learning_rate": 4.508301404853129e-06,
|
| 951 |
+
"loss": 0.0002,
|
| 952 |
+
"step": 2775
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 16.0,
|
| 956 |
+
"eval_accuracy": 0.6807580174927114,
|
| 957 |
+
"eval_loss": 2.690096855163574,
|
| 958 |
+
"eval_macro_f1": 0.6342184294336438,
|
| 959 |
+
"eval_runtime": 1.1197,
|
| 960 |
+
"eval_samples_per_second": 612.642,
|
| 961 |
+
"eval_steps_per_second": 38.402,
|
| 962 |
+
"eval_weighted_f1": 0.6788607166953128,
|
| 963 |
+
"step": 2784
|
| 964 |
+
}
|
| 965 |
+
],
|
| 966 |
+
"logging_steps": 25,
|
| 967 |
+
"max_steps": 3480,
|
| 968 |
+
"num_input_tokens_seen": 0,
|
| 969 |
+
"num_train_epochs": 20,
|
| 970 |
+
"save_steps": 500,
|
| 971 |
+
"stateful_callbacks": {
|
| 972 |
+
"EarlyStoppingCallback": {
|
| 973 |
+
"args": {
|
| 974 |
+
"early_stopping_patience": 3,
|
| 975 |
+
"early_stopping_threshold": 0.0
|
| 976 |
+
},
|
| 977 |
+
"attributes": {
|
| 978 |
+
"early_stopping_patience_counter": 3
|
| 979 |
+
}
|
| 980 |
+
},
|
| 981 |
+
"TrainerControl": {
|
| 982 |
+
"args": {
|
| 983 |
+
"should_epoch_stop": false,
|
| 984 |
+
"should_evaluate": false,
|
| 985 |
+
"should_log": false,
|
| 986 |
+
"should_save": true,
|
| 987 |
+
"should_training_stop": true
|
| 988 |
+
},
|
| 989 |
+
"attributes": {}
|
| 990 |
+
}
|
| 991 |
+
},
|
| 992 |
+
"total_flos": 0.0,
|
| 993 |
+
"train_batch_size": 8,
|
| 994 |
+
"trial_name": null,
|
| 995 |
+
"trial_params": null
|
| 996 |
+
}
|
vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"15001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"sp_model_kwargs": {},
|
| 54 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 55 |
+
"unk_token": "<unk>"
|
| 56 |
+
}
|
vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-2282/trainer_state.json
ADDED
|
@@ -0,0 +1,834 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1793,
|
| 3 |
+
"best_metric": 0.5450922531222987,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vifactcheck-visobert-s44-20260829T182248Z-59e46264/trainer/checkpoint-1793",
|
| 5 |
+
"epoch": 14.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2282,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.15408320493066255,
|
| 14 |
+
"grad_norm": 22.301509857177734,
|
| 15 |
+
"learning_rate": 1.47239263803681e-06,
|
| 16 |
+
"loss": 4.7401,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.3081664098613251,
|
| 21 |
+
"grad_norm": 20.221092224121094,
|
| 22 |
+
"learning_rate": 3.0061349693251535e-06,
|
| 23 |
+
"loss": 4.5514,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4622496147919877,
|
| 28 |
+
"grad_norm": 29.94936752319336,
|
| 29 |
+
"learning_rate": 4.539877300613497e-06,
|
| 30 |
+
"loss": 4.4596,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.6163328197226502,
|
| 35 |
+
"grad_norm": 26.135087966918945,
|
| 36 |
+
"learning_rate": 6.073619631901841e-06,
|
| 37 |
+
"loss": 4.3972,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7704160246533128,
|
| 42 |
+
"grad_norm": 18.76947784423828,
|
| 43 |
+
"learning_rate": 7.6073619631901856e-06,
|
| 44 |
+
"loss": 4.3924,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.9244992295839753,
|
| 49 |
+
"grad_norm": 30.230634689331055,
|
| 50 |
+
"learning_rate": 9.14110429447853e-06,
|
| 51 |
+
"loss": 4.3625,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.40988372093023256,
|
| 57 |
+
"eval_loss": 1.0818296670913696,
|
| 58 |
+
"eval_macro_f1": 0.3922304631718285,
|
| 59 |
+
"eval_runtime": 1.3308,
|
| 60 |
+
"eval_samples_per_second": 516.964,
|
| 61 |
+
"eval_steps_per_second": 32.31,
|
| 62 |
+
"eval_weighted_f1": 0.39419148185628594,
|
| 63 |
+
"step": 163
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.073959938366718,
|
| 67 |
+
"grad_norm": 25.277162551879883,
|
| 68 |
+
"learning_rate": 1.0674846625766873e-05,
|
| 69 |
+
"loss": 4.2113,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.2280431432973806,
|
| 74 |
+
"grad_norm": 28.43084144592285,
|
| 75 |
+
"learning_rate": 1.2208588957055216e-05,
|
| 76 |
+
"loss": 4.348,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.3821263482280433,
|
| 81 |
+
"grad_norm": 23.55897331237793,
|
| 82 |
+
"learning_rate": 1.374233128834356e-05,
|
| 83 |
+
"loss": 4.2152,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.5362095531587057,
|
| 88 |
+
"grad_norm": 51.33898162841797,
|
| 89 |
+
"learning_rate": 1.5276073619631903e-05,
|
| 90 |
+
"loss": 4.3203,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.6902927580893683,
|
| 95 |
+
"grad_norm": 19.284208297729492,
|
| 96 |
+
"learning_rate": 1.6809815950920248e-05,
|
| 97 |
+
"loss": 4.2034,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.8443759630200307,
|
| 102 |
+
"grad_norm": 20.314197540283203,
|
| 103 |
+
"learning_rate": 1.834355828220859e-05,
|
| 104 |
+
"loss": 4.1263,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.9984591679506933,
|
| 109 |
+
"grad_norm": 28.18060302734375,
|
| 110 |
+
"learning_rate": 1.9877300613496935e-05,
|
| 111 |
+
"loss": 4.2162,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.40406976744186046,
|
| 117 |
+
"eval_loss": 1.098442554473877,
|
| 118 |
+
"eval_macro_f1": 0.35844521871291385,
|
| 119 |
+
"eval_runtime": 1.2482,
|
| 120 |
+
"eval_samples_per_second": 551.176,
|
| 121 |
+
"eval_steps_per_second": 34.448,
|
| 122 |
+
"eval_weighted_f1": 0.35546987758431847,
|
| 123 |
+
"step": 326
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.147919876733436,
|
| 127 |
+
"grad_norm": 56.9528694152832,
|
| 128 |
+
"learning_rate": 1.9843217450579417e-05,
|
| 129 |
+
"loss": 3.8295,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.3020030816640986,
|
| 134 |
+
"grad_norm": 19.001956939697266,
|
| 135 |
+
"learning_rate": 1.967280163599182e-05,
|
| 136 |
+
"loss": 3.7898,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.4560862865947612,
|
| 141 |
+
"grad_norm": 47.619361877441406,
|
| 142 |
+
"learning_rate": 1.9502385821404228e-05,
|
| 143 |
+
"loss": 3.6991,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.610169491525424,
|
| 148 |
+
"grad_norm": 28.823945999145508,
|
| 149 |
+
"learning_rate": 1.9331970006816635e-05,
|
| 150 |
+
"loss": 3.9033,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.7642526964560865,
|
| 155 |
+
"grad_norm": 57.12553405761719,
|
| 156 |
+
"learning_rate": 1.916155419222904e-05,
|
| 157 |
+
"loss": 3.5999,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.9183359013867487,
|
| 162 |
+
"grad_norm": 18.48838233947754,
|
| 163 |
+
"learning_rate": 1.8991138377641446e-05,
|
| 164 |
+
"loss": 3.6755,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 3.0,
|
| 169 |
+
"eval_accuracy": 0.5218023255813954,
|
| 170 |
+
"eval_loss": 0.9917183518409729,
|
| 171 |
+
"eval_macro_f1": 0.5199300680686414,
|
| 172 |
+
"eval_runtime": 1.3239,
|
| 173 |
+
"eval_samples_per_second": 519.691,
|
| 174 |
+
"eval_steps_per_second": 32.481,
|
| 175 |
+
"eval_weighted_f1": 0.5202113165109993,
|
| 176 |
+
"step": 489
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.0677966101694913,
|
| 180 |
+
"grad_norm": 47.14376449584961,
|
| 181 |
+
"learning_rate": 1.8820722563053853e-05,
|
| 182 |
+
"loss": 3.2505,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.221879815100154,
|
| 187 |
+
"grad_norm": 23.565784454345703,
|
| 188 |
+
"learning_rate": 1.8650306748466257e-05,
|
| 189 |
+
"loss": 2.9527,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.3759630200308166,
|
| 194 |
+
"grad_norm": 30.684024810791016,
|
| 195 |
+
"learning_rate": 1.8479890933878664e-05,
|
| 196 |
+
"loss": 2.9458,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.5300462249614792,
|
| 201 |
+
"grad_norm": 34.59135055541992,
|
| 202 |
+
"learning_rate": 1.8309475119291072e-05,
|
| 203 |
+
"loss": 2.831,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.684129429892142,
|
| 208 |
+
"grad_norm": 38.83842849731445,
|
| 209 |
+
"learning_rate": 1.8139059304703476e-05,
|
| 210 |
+
"loss": 3.1647,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.8382126348228045,
|
| 215 |
+
"grad_norm": 45.43645477294922,
|
| 216 |
+
"learning_rate": 1.7968643490115886e-05,
|
| 217 |
+
"loss": 2.8642,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.992295839753467,
|
| 222 |
+
"grad_norm": 52.25239181518555,
|
| 223 |
+
"learning_rate": 1.779822767552829e-05,
|
| 224 |
+
"loss": 2.9152,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 4.0,
|
| 229 |
+
"eval_accuracy": 0.5261627906976745,
|
| 230 |
+
"eval_loss": 1.098737359046936,
|
| 231 |
+
"eval_macro_f1": 0.5298340462675778,
|
| 232 |
+
"eval_runtime": 1.1524,
|
| 233 |
+
"eval_samples_per_second": 596.992,
|
| 234 |
+
"eval_steps_per_second": 37.312,
|
| 235 |
+
"eval_weighted_f1": 0.5295335051142966,
|
| 236 |
+
"step": 652
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 4.141756548536209,
|
| 240 |
+
"grad_norm": 30.164339065551758,
|
| 241 |
+
"learning_rate": 1.7627811860940697e-05,
|
| 242 |
+
"loss": 2.1327,
|
| 243 |
+
"step": 675
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.295839753466872,
|
| 247 |
+
"grad_norm": 32.94883728027344,
|
| 248 |
+
"learning_rate": 1.7457396046353105e-05,
|
| 249 |
+
"loss": 1.938,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.449922958397535,
|
| 254 |
+
"grad_norm": 68.37862396240234,
|
| 255 |
+
"learning_rate": 1.728698023176551e-05,
|
| 256 |
+
"loss": 1.8231,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.604006163328197,
|
| 261 |
+
"grad_norm": 60.21028518676758,
|
| 262 |
+
"learning_rate": 1.7116564417177916e-05,
|
| 263 |
+
"loss": 2.2355,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.75808936825886,
|
| 268 |
+
"grad_norm": 41.61284255981445,
|
| 269 |
+
"learning_rate": 1.6946148602590323e-05,
|
| 270 |
+
"loss": 2.0929,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.9121725731895225,
|
| 275 |
+
"grad_norm": 74.66960144042969,
|
| 276 |
+
"learning_rate": 1.6775732788002727e-05,
|
| 277 |
+
"loss": 1.9251,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 5.0,
|
| 282 |
+
"eval_accuracy": 0.5450581395348837,
|
| 283 |
+
"eval_loss": 1.2722961902618408,
|
| 284 |
+
"eval_macro_f1": 0.5250357317817511,
|
| 285 |
+
"eval_runtime": 1.1685,
|
| 286 |
+
"eval_samples_per_second": 588.78,
|
| 287 |
+
"eval_steps_per_second": 36.799,
|
| 288 |
+
"eval_weighted_f1": 0.5257288199302057,
|
| 289 |
+
"step": 815
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 5.061633281972265,
|
| 293 |
+
"grad_norm": 37.04746627807617,
|
| 294 |
+
"learning_rate": 1.6605316973415134e-05,
|
| 295 |
+
"loss": 1.742,
|
| 296 |
+
"step": 825
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 5.215716486902927,
|
| 300 |
+
"grad_norm": 36.91507339477539,
|
| 301 |
+
"learning_rate": 1.643490115882754e-05,
|
| 302 |
+
"loss": 1.0258,
|
| 303 |
+
"step": 850
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.36979969183359,
|
| 307 |
+
"grad_norm": 59.55405807495117,
|
| 308 |
+
"learning_rate": 1.6264485344239945e-05,
|
| 309 |
+
"loss": 1.0554,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.523882896764253,
|
| 314 |
+
"grad_norm": 70.8587417602539,
|
| 315 |
+
"learning_rate": 1.6094069529652356e-05,
|
| 316 |
+
"loss": 1.2274,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.677966101694915,
|
| 321 |
+
"grad_norm": 97.20299530029297,
|
| 322 |
+
"learning_rate": 1.592365371506476e-05,
|
| 323 |
+
"loss": 1.1048,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.832049306625578,
|
| 328 |
+
"grad_norm": 97.20658111572266,
|
| 329 |
+
"learning_rate": 1.5753237900477167e-05,
|
| 330 |
+
"loss": 1.252,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.9861325115562405,
|
| 335 |
+
"grad_norm": 57.26484298706055,
|
| 336 |
+
"learning_rate": 1.5582822085889574e-05,
|
| 337 |
+
"loss": 1.4221,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.0,
|
| 342 |
+
"eval_accuracy": 0.5087209302325582,
|
| 343 |
+
"eval_loss": 1.5389418601989746,
|
| 344 |
+
"eval_macro_f1": 0.4930922526757539,
|
| 345 |
+
"eval_runtime": 1.1384,
|
| 346 |
+
"eval_samples_per_second": 604.379,
|
| 347 |
+
"eval_steps_per_second": 37.774,
|
| 348 |
+
"eval_weighted_f1": 0.4923670176443556,
|
| 349 |
+
"step": 978
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 6.135593220338983,
|
| 353 |
+
"grad_norm": 22.37522315979004,
|
| 354 |
+
"learning_rate": 1.5412406271301978e-05,
|
| 355 |
+
"loss": 0.6636,
|
| 356 |
+
"step": 1000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.289676425269645,
|
| 360 |
+
"grad_norm": 91.99688720703125,
|
| 361 |
+
"learning_rate": 1.5241990456714383e-05,
|
| 362 |
+
"loss": 1.0649,
|
| 363 |
+
"step": 1025
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.443759630200308,
|
| 367 |
+
"grad_norm": 45.110321044921875,
|
| 368 |
+
"learning_rate": 1.5071574642126792e-05,
|
| 369 |
+
"loss": 0.642,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.597842835130971,
|
| 374 |
+
"grad_norm": 70.322509765625,
|
| 375 |
+
"learning_rate": 1.4901158827539198e-05,
|
| 376 |
+
"loss": 0.6997,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.751926040061633,
|
| 381 |
+
"grad_norm": 98.82887268066406,
|
| 382 |
+
"learning_rate": 1.4730743012951603e-05,
|
| 383 |
+
"loss": 0.633,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.906009244992296,
|
| 388 |
+
"grad_norm": 20.150808334350586,
|
| 389 |
+
"learning_rate": 1.456032719836401e-05,
|
| 390 |
+
"loss": 0.7364,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 7.0,
|
| 395 |
+
"eval_accuracy": 0.5319767441860465,
|
| 396 |
+
"eval_loss": 1.8516525030136108,
|
| 397 |
+
"eval_macro_f1": 0.534303020732136,
|
| 398 |
+
"eval_runtime": 1.1908,
|
| 399 |
+
"eval_samples_per_second": 577.744,
|
| 400 |
+
"eval_steps_per_second": 36.109,
|
| 401 |
+
"eval_weighted_f1": 0.5336971385418576,
|
| 402 |
+
"step": 1141
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.055469953775039,
|
| 406 |
+
"grad_norm": 14.736871719360352,
|
| 407 |
+
"learning_rate": 1.4389911383776416e-05,
|
| 408 |
+
"loss": 0.5445,
|
| 409 |
+
"step": 1150
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 7.209553158705701,
|
| 413 |
+
"grad_norm": 9.65927791595459,
|
| 414 |
+
"learning_rate": 1.4219495569188822e-05,
|
| 415 |
+
"loss": 0.3656,
|
| 416 |
+
"step": 1175
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.363636363636363,
|
| 420 |
+
"grad_norm": 107.58206176757812,
|
| 421 |
+
"learning_rate": 1.4049079754601229e-05,
|
| 422 |
+
"loss": 0.4211,
|
| 423 |
+
"step": 1200
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.517719568567026,
|
| 427 |
+
"grad_norm": 13.92328929901123,
|
| 428 |
+
"learning_rate": 1.3878663940013635e-05,
|
| 429 |
+
"loss": 0.3471,
|
| 430 |
+
"step": 1225
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.671802773497689,
|
| 434 |
+
"grad_norm": 62.75358581542969,
|
| 435 |
+
"learning_rate": 1.370824812542604e-05,
|
| 436 |
+
"loss": 0.4898,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.825885978428351,
|
| 441 |
+
"grad_norm": 33.541160583496094,
|
| 442 |
+
"learning_rate": 1.3537832310838447e-05,
|
| 443 |
+
"loss": 0.3111,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.979969183359014,
|
| 448 |
+
"grad_norm": 25.23708724975586,
|
| 449 |
+
"learning_rate": 1.3367416496250853e-05,
|
| 450 |
+
"loss": 0.4277,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 8.0,
|
| 455 |
+
"eval_accuracy": 0.5276162790697675,
|
| 456 |
+
"eval_loss": 2.4985873699188232,
|
| 457 |
+
"eval_macro_f1": 0.5265026929933844,
|
| 458 |
+
"eval_runtime": 1.1206,
|
| 459 |
+
"eval_samples_per_second": 613.958,
|
| 460 |
+
"eval_steps_per_second": 38.372,
|
| 461 |
+
"eval_weighted_f1": 0.5248343285583679,
|
| 462 |
+
"step": 1304
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 8.129429892141756,
|
| 466 |
+
"grad_norm": 43.25033950805664,
|
| 467 |
+
"learning_rate": 1.3197000681663258e-05,
|
| 468 |
+
"loss": 0.142,
|
| 469 |
+
"step": 1325
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 8.283513097072419,
|
| 473 |
+
"grad_norm": 67.72445678710938,
|
| 474 |
+
"learning_rate": 1.3026584867075667e-05,
|
| 475 |
+
"loss": 0.2648,
|
| 476 |
+
"step": 1350
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 8.437596302003081,
|
| 480 |
+
"grad_norm": 44.03306579589844,
|
| 481 |
+
"learning_rate": 1.2856169052488073e-05,
|
| 482 |
+
"loss": 0.3789,
|
| 483 |
+
"step": 1375
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 8.591679506933744,
|
| 487 |
+
"grad_norm": 63.89084243774414,
|
| 488 |
+
"learning_rate": 1.2685753237900478e-05,
|
| 489 |
+
"loss": 0.3502,
|
| 490 |
+
"step": 1400
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.745762711864407,
|
| 494 |
+
"grad_norm": 36.28083419799805,
|
| 495 |
+
"learning_rate": 1.2515337423312886e-05,
|
| 496 |
+
"loss": 0.1828,
|
| 497 |
+
"step": 1425
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.89984591679507,
|
| 501 |
+
"grad_norm": 92.56626892089844,
|
| 502 |
+
"learning_rate": 1.2344921608725291e-05,
|
| 503 |
+
"loss": 0.3767,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 9.0,
|
| 508 |
+
"eval_accuracy": 0.5334302325581395,
|
| 509 |
+
"eval_loss": 2.8457882404327393,
|
| 510 |
+
"eval_macro_f1": 0.5357835177016838,
|
| 511 |
+
"eval_runtime": 1.1735,
|
| 512 |
+
"eval_samples_per_second": 586.302,
|
| 513 |
+
"eval_steps_per_second": 36.644,
|
| 514 |
+
"eval_weighted_f1": 0.5347919934359653,
|
| 515 |
+
"step": 1467
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 9.049306625577811,
|
| 519 |
+
"grad_norm": 21.610509872436523,
|
| 520 |
+
"learning_rate": 1.2174505794137697e-05,
|
| 521 |
+
"loss": 0.1969,
|
| 522 |
+
"step": 1475
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 9.203389830508474,
|
| 526 |
+
"grad_norm": 9.576517105102539,
|
| 527 |
+
"learning_rate": 1.2004089979550104e-05,
|
| 528 |
+
"loss": 0.2576,
|
| 529 |
+
"step": 1500
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 9.357473035439137,
|
| 533 |
+
"grad_norm": 65.83451080322266,
|
| 534 |
+
"learning_rate": 1.183367416496251e-05,
|
| 535 |
+
"loss": 0.1715,
|
| 536 |
+
"step": 1525
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.5115562403698,
|
| 540 |
+
"grad_norm": 72.24199676513672,
|
| 541 |
+
"learning_rate": 1.1663258350374915e-05,
|
| 542 |
+
"loss": 0.1148,
|
| 543 |
+
"step": 1550
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.665639445300462,
|
| 547 |
+
"grad_norm": 100.49017333984375,
|
| 548 |
+
"learning_rate": 1.1492842535787322e-05,
|
| 549 |
+
"loss": 0.2852,
|
| 550 |
+
"step": 1575
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.819722650231125,
|
| 554 |
+
"grad_norm": 87.1134262084961,
|
| 555 |
+
"learning_rate": 1.1322426721199728e-05,
|
| 556 |
+
"loss": 0.2295,
|
| 557 |
+
"step": 1600
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.973805855161787,
|
| 561 |
+
"grad_norm": 13.292170524597168,
|
| 562 |
+
"learning_rate": 1.1152010906612133e-05,
|
| 563 |
+
"loss": 0.1435,
|
| 564 |
+
"step": 1625
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.0,
|
| 568 |
+
"eval_accuracy": 0.5232558139534884,
|
| 569 |
+
"eval_loss": 3.1596717834472656,
|
| 570 |
+
"eval_macro_f1": 0.526687720808056,
|
| 571 |
+
"eval_runtime": 1.1787,
|
| 572 |
+
"eval_samples_per_second": 583.7,
|
| 573 |
+
"eval_steps_per_second": 36.481,
|
| 574 |
+
"eval_weighted_f1": 0.5259946851032898,
|
| 575 |
+
"step": 1630
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 10.12326656394453,
|
| 579 |
+
"grad_norm": 46.83879470825195,
|
| 580 |
+
"learning_rate": 1.0981595092024542e-05,
|
| 581 |
+
"loss": 0.0794,
|
| 582 |
+
"step": 1650
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.277349768875192,
|
| 586 |
+
"grad_norm": 32.3950080871582,
|
| 587 |
+
"learning_rate": 1.0811179277436948e-05,
|
| 588 |
+
"loss": 0.1614,
|
| 589 |
+
"step": 1675
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 10.431432973805855,
|
| 593 |
+
"grad_norm": 12.704472541809082,
|
| 594 |
+
"learning_rate": 1.0640763462849353e-05,
|
| 595 |
+
"loss": 0.2561,
|
| 596 |
+
"step": 1700
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 10.585516178736517,
|
| 600 |
+
"grad_norm": 45.62063217163086,
|
| 601 |
+
"learning_rate": 1.047034764826176e-05,
|
| 602 |
+
"loss": 0.0774,
|
| 603 |
+
"step": 1725
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.73959938366718,
|
| 607 |
+
"grad_norm": 19.634353637695312,
|
| 608 |
+
"learning_rate": 1.0299931833674166e-05,
|
| 609 |
+
"loss": 0.2027,
|
| 610 |
+
"step": 1750
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 10.893682588597843,
|
| 614 |
+
"grad_norm": 128.722412109375,
|
| 615 |
+
"learning_rate": 1.0129516019086572e-05,
|
| 616 |
+
"loss": 0.2195,
|
| 617 |
+
"step": 1775
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 11.0,
|
| 621 |
+
"eval_accuracy": 0.5436046511627907,
|
| 622 |
+
"eval_loss": 3.533062696456909,
|
| 623 |
+
"eval_macro_f1": 0.5450922531222987,
|
| 624 |
+
"eval_runtime": 1.1678,
|
| 625 |
+
"eval_samples_per_second": 589.132,
|
| 626 |
+
"eval_steps_per_second": 36.821,
|
| 627 |
+
"eval_weighted_f1": 0.5432349818765609,
|
| 628 |
+
"step": 1793
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 11.043143297380585,
|
| 632 |
+
"grad_norm": 134.18212890625,
|
| 633 |
+
"learning_rate": 9.959100204498979e-06,
|
| 634 |
+
"loss": 0.2283,
|
| 635 |
+
"step": 1800
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 11.197226502311247,
|
| 639 |
+
"grad_norm": 59.07541275024414,
|
| 640 |
+
"learning_rate": 9.788684389911385e-06,
|
| 641 |
+
"loss": 0.1032,
|
| 642 |
+
"step": 1825
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 11.35130970724191,
|
| 646 |
+
"grad_norm": 0.024077802896499634,
|
| 647 |
+
"learning_rate": 9.61826857532379e-06,
|
| 648 |
+
"loss": 0.1312,
|
| 649 |
+
"step": 1850
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 11.505392912172573,
|
| 653 |
+
"grad_norm": 0.34076356887817383,
|
| 654 |
+
"learning_rate": 9.447852760736197e-06,
|
| 655 |
+
"loss": 0.1567,
|
| 656 |
+
"step": 1875
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 11.659476117103235,
|
| 660 |
+
"grad_norm": 0.02346443198621273,
|
| 661 |
+
"learning_rate": 9.277436946148603e-06,
|
| 662 |
+
"loss": 0.0971,
|
| 663 |
+
"step": 1900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 11.813559322033898,
|
| 667 |
+
"grad_norm": 208.256103515625,
|
| 668 |
+
"learning_rate": 9.107021131561008e-06,
|
| 669 |
+
"loss": 0.1709,
|
| 670 |
+
"step": 1925
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 11.96764252696456,
|
| 674 |
+
"grad_norm": 5.23779296875,
|
| 675 |
+
"learning_rate": 8.936605316973416e-06,
|
| 676 |
+
"loss": 0.1708,
|
| 677 |
+
"step": 1950
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 12.0,
|
| 681 |
+
"eval_accuracy": 0.5348837209302325,
|
| 682 |
+
"eval_loss": 3.8550264835357666,
|
| 683 |
+
"eval_macro_f1": 0.5334088907577671,
|
| 684 |
+
"eval_runtime": 1.1302,
|
| 685 |
+
"eval_samples_per_second": 608.743,
|
| 686 |
+
"eval_steps_per_second": 38.046,
|
| 687 |
+
"eval_weighted_f1": 0.53179255314788,
|
| 688 |
+
"step": 1956
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 12.117103235747303,
|
| 692 |
+
"grad_norm": 5.9993510246276855,
|
| 693 |
+
"learning_rate": 8.766189502385823e-06,
|
| 694 |
+
"loss": 0.034,
|
| 695 |
+
"step": 1975
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 12.271186440677965,
|
| 699 |
+
"grad_norm": 130.4667510986328,
|
| 700 |
+
"learning_rate": 8.595773687798228e-06,
|
| 701 |
+
"loss": 0.0849,
|
| 702 |
+
"step": 2000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 12.425269645608628,
|
| 706 |
+
"grad_norm": 8.46444034576416,
|
| 707 |
+
"learning_rate": 8.425357873210634e-06,
|
| 708 |
+
"loss": 0.1614,
|
| 709 |
+
"step": 2025
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 12.57935285053929,
|
| 713 |
+
"grad_norm": 96.05816650390625,
|
| 714 |
+
"learning_rate": 8.254942058623041e-06,
|
| 715 |
+
"loss": 0.1958,
|
| 716 |
+
"step": 2050
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 12.733436055469953,
|
| 720 |
+
"grad_norm": 4.017282009124756,
|
| 721 |
+
"learning_rate": 8.084526244035447e-06,
|
| 722 |
+
"loss": 0.0253,
|
| 723 |
+
"step": 2075
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.887519260400616,
|
| 727 |
+
"grad_norm": 0.8230677247047424,
|
| 728 |
+
"learning_rate": 7.914110429447854e-06,
|
| 729 |
+
"loss": 0.0727,
|
| 730 |
+
"step": 2100
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 13.0,
|
| 734 |
+
"eval_accuracy": 0.5043604651162791,
|
| 735 |
+
"eval_loss": 4.332130432128906,
|
| 736 |
+
"eval_macro_f1": 0.5062903117139593,
|
| 737 |
+
"eval_runtime": 1.1656,
|
| 738 |
+
"eval_samples_per_second": 590.229,
|
| 739 |
+
"eval_steps_per_second": 36.889,
|
| 740 |
+
"eval_weighted_f1": 0.5043046593195004,
|
| 741 |
+
"step": 2119
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 13.03697996918336,
|
| 745 |
+
"grad_norm": 1.5189831256866455,
|
| 746 |
+
"learning_rate": 7.74369461486026e-06,
|
| 747 |
+
"loss": 0.1613,
|
| 748 |
+
"step": 2125
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 13.19106317411402,
|
| 752 |
+
"grad_norm": 0.16000854969024658,
|
| 753 |
+
"learning_rate": 7.573278800272665e-06,
|
| 754 |
+
"loss": 0.0316,
|
| 755 |
+
"step": 2150
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 13.345146379044683,
|
| 759 |
+
"grad_norm": 4.500791072845459,
|
| 760 |
+
"learning_rate": 7.402862985685072e-06,
|
| 761 |
+
"loss": 0.0303,
|
| 762 |
+
"step": 2175
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 13.499229583975346,
|
| 766 |
+
"grad_norm": 93.62580871582031,
|
| 767 |
+
"learning_rate": 7.232447171097479e-06,
|
| 768 |
+
"loss": 0.0184,
|
| 769 |
+
"step": 2200
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 13.653312788906009,
|
| 773 |
+
"grad_norm": 55.62511444091797,
|
| 774 |
+
"learning_rate": 7.062031356509885e-06,
|
| 775 |
+
"loss": 0.0313,
|
| 776 |
+
"step": 2225
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 13.807395993836671,
|
| 780 |
+
"grad_norm": 113.67203521728516,
|
| 781 |
+
"learning_rate": 6.891615541922291e-06,
|
| 782 |
+
"loss": 0.1363,
|
| 783 |
+
"step": 2250
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 13.961479198767334,
|
| 787 |
+
"grad_norm": 0.3834678530693054,
|
| 788 |
+
"learning_rate": 6.721199727334697e-06,
|
| 789 |
+
"loss": 0.0097,
|
| 790 |
+
"step": 2275
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 14.0,
|
| 794 |
+
"eval_accuracy": 0.5232558139534884,
|
| 795 |
+
"eval_loss": 4.539718151092529,
|
| 796 |
+
"eval_macro_f1": 0.5265144884763863,
|
| 797 |
+
"eval_runtime": 1.2498,
|
| 798 |
+
"eval_samples_per_second": 550.491,
|
| 799 |
+
"eval_steps_per_second": 34.406,
|
| 800 |
+
"eval_weighted_f1": 0.5266126588243202,
|
| 801 |
+
"step": 2282
|
| 802 |
+
}
|
| 803 |
+
],
|
| 804 |
+
"logging_steps": 25,
|
| 805 |
+
"max_steps": 3260,
|
| 806 |
+
"num_input_tokens_seen": 0,
|
| 807 |
+
"num_train_epochs": 20,
|
| 808 |
+
"save_steps": 500,
|
| 809 |
+
"stateful_callbacks": {
|
| 810 |
+
"EarlyStoppingCallback": {
|
| 811 |
+
"args": {
|
| 812 |
+
"early_stopping_patience": 3,
|
| 813 |
+
"early_stopping_threshold": 0.0
|
| 814 |
+
},
|
| 815 |
+
"attributes": {
|
| 816 |
+
"early_stopping_patience_counter": 3
|
| 817 |
+
}
|
| 818 |
+
},
|
| 819 |
+
"TrainerControl": {
|
| 820 |
+
"args": {
|
| 821 |
+
"should_epoch_stop": false,
|
| 822 |
+
"should_evaluate": false,
|
| 823 |
+
"should_log": false,
|
| 824 |
+
"should_save": true,
|
| 825 |
+
"should_training_stop": true
|
| 826 |
+
},
|
| 827 |
+
"attributes": {}
|
| 828 |
+
}
|
| 829 |
+
},
|
| 830 |
+
"total_flos": 0.0,
|
| 831 |
+
"train_batch_size": 8,
|
| 832 |
+
"trial_name": null,
|
| 833 |
+
"trial_params": null
|
| 834 |
+
}
|
vigoemotions-visobert-s42-20260829T190105Z-5eeb966a/trainer/checkpoint-5687/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
vigoemotions-visobert-s42-20260829T190105Z-5eeb966a/trainer/checkpoint-5687/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vigoemotions-visobert-s42-20260829T190105Z-5eeb966a/trainer/checkpoint-5687/tokenizer_config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"15001": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"sp_model_kwargs": {},
|
| 54 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 55 |
+
"unk_token": "<unk>"
|
| 56 |
+
}
|