End of training
Browse files- all_results.json +8 -0
- preprocessor_config.json +28 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +58 -0
- train_results.json +8 -0
- trainer_state.json +743 -0
- vocab.txt +0 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"total_flos": 2.0309886965211648e+18,
|
| 4 |
+
"train_loss": 1.927858536594598,
|
| 5 |
+
"train_runtime": 42079.8744,
|
| 6 |
+
"train_samples_per_second": 183.867,
|
| 7 |
+
"train_steps_per_second": 0.92
|
| 8 |
+
}
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"crop_size": {
|
| 3 |
+
"height": 224,
|
| 4 |
+
"width": 224
|
| 5 |
+
},
|
| 6 |
+
"do_center_crop": false,
|
| 7 |
+
"do_convert_rgb": true,
|
| 8 |
+
"do_normalize": true,
|
| 9 |
+
"do_rescale": true,
|
| 10 |
+
"do_resize": true,
|
| 11 |
+
"image_mean": [
|
| 12 |
+
0.48145466,
|
| 13 |
+
0.4578275,
|
| 14 |
+
0.40821073
|
| 15 |
+
],
|
| 16 |
+
"image_processor_type": "ChineseCLIPImageProcessor",
|
| 17 |
+
"image_std": [
|
| 18 |
+
0.26862954,
|
| 19 |
+
0.26130258,
|
| 20 |
+
0.27577711
|
| 21 |
+
],
|
| 22 |
+
"resample": 3,
|
| 23 |
+
"rescale_factor": 0.00392156862745098,
|
| 24 |
+
"size": {
|
| 25 |
+
"height": 224,
|
| 26 |
+
"width": 224
|
| 27 |
+
}
|
| 28 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_basic_tokenize": true,
|
| 47 |
+
"do_lower_case": true,
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "[MASK]",
|
| 50 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 51 |
+
"never_split": null,
|
| 52 |
+
"pad_token": "[PAD]",
|
| 53 |
+
"sep_token": "[SEP]",
|
| 54 |
+
"strip_accents": null,
|
| 55 |
+
"tokenize_chinese_chars": true,
|
| 56 |
+
"tokenizer_class": "BertTokenizer",
|
| 57 |
+
"unk_token": "[UNK]"
|
| 58 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"total_flos": 2.0309886965211648e+18,
|
| 4 |
+
"train_loss": 1.927858536594598,
|
| 5 |
+
"train_runtime": 42079.8744,
|
| 6 |
+
"train_samples_per_second": 183.867,
|
| 7 |
+
"train_steps_per_second": 0.92
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,743 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 100.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 38700,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.0,
|
| 14 |
+
"grad_norm": 0.4617389738559723,
|
| 15 |
+
"learning_rate": 9.900516795865633e-06,
|
| 16 |
+
"loss": 2.3944,
|
| 17 |
+
"step": 387
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 2.0,
|
| 21 |
+
"grad_norm": 1.3490279912948608,
|
| 22 |
+
"learning_rate": 9.800775193798451e-06,
|
| 23 |
+
"loss": 2.3064,
|
| 24 |
+
"step": 774
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 3.0,
|
| 28 |
+
"grad_norm": 1.569773554801941,
|
| 29 |
+
"learning_rate": 9.70077519379845e-06,
|
| 30 |
+
"loss": 2.2698,
|
| 31 |
+
"step": 1161
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 4.0,
|
| 35 |
+
"grad_norm": 1.8885618448257446,
|
| 36 |
+
"learning_rate": 9.600775193798451e-06,
|
| 37 |
+
"loss": 2.2307,
|
| 38 |
+
"step": 1548
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 5.0,
|
| 42 |
+
"grad_norm": 4.19589900970459,
|
| 43 |
+
"learning_rate": 9.50077519379845e-06,
|
| 44 |
+
"loss": 2.1967,
|
| 45 |
+
"step": 1935
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 6.0,
|
| 49 |
+
"grad_norm": 2.8683922290802,
|
| 50 |
+
"learning_rate": 9.40077519379845e-06,
|
| 51 |
+
"loss": 2.1706,
|
| 52 |
+
"step": 2322
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 7.0,
|
| 56 |
+
"grad_norm": 2.789121389389038,
|
| 57 |
+
"learning_rate": 9.301033591731267e-06,
|
| 58 |
+
"loss": 2.1468,
|
| 59 |
+
"step": 2709
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 8.0,
|
| 63 |
+
"grad_norm": 2.4587924480438232,
|
| 64 |
+
"learning_rate": 9.201033591731266e-06,
|
| 65 |
+
"loss": 2.1146,
|
| 66 |
+
"step": 3096
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 9.0,
|
| 70 |
+
"grad_norm": 3.5958080291748047,
|
| 71 |
+
"learning_rate": 9.101033591731266e-06,
|
| 72 |
+
"loss": 2.1352,
|
| 73 |
+
"step": 3483
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 10.0,
|
| 77 |
+
"grad_norm": 3.489946126937866,
|
| 78 |
+
"learning_rate": 9.001033591731267e-06,
|
| 79 |
+
"loss": 2.0821,
|
| 80 |
+
"step": 3870
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 11.0,
|
| 84 |
+
"grad_norm": 4.598156929016113,
|
| 85 |
+
"learning_rate": 8.901033591731268e-06,
|
| 86 |
+
"loss": 2.0615,
|
| 87 |
+
"step": 4257
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 12.0,
|
| 91 |
+
"grad_norm": 3.008528709411621,
|
| 92 |
+
"learning_rate": 8.801291989664084e-06,
|
| 93 |
+
"loss": 2.0678,
|
| 94 |
+
"step": 4644
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 13.0,
|
| 98 |
+
"grad_norm": 3.5459799766540527,
|
| 99 |
+
"learning_rate": 8.701291989664083e-06,
|
| 100 |
+
"loss": 2.0456,
|
| 101 |
+
"step": 5031
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 14.0,
|
| 105 |
+
"grad_norm": 3.375678300857544,
|
| 106 |
+
"learning_rate": 8.601291989664084e-06,
|
| 107 |
+
"loss": 2.0267,
|
| 108 |
+
"step": 5418
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 15.0,
|
| 112 |
+
"grad_norm": 3.702049970626831,
|
| 113 |
+
"learning_rate": 8.501291989664083e-06,
|
| 114 |
+
"loss": 2.0277,
|
| 115 |
+
"step": 5805
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 16.0,
|
| 119 |
+
"grad_norm": 3.2726526260375977,
|
| 120 |
+
"learning_rate": 8.401291989664084e-06,
|
| 121 |
+
"loss": 2.0316,
|
| 122 |
+
"step": 6192
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 17.0,
|
| 126 |
+
"grad_norm": 4.813005447387695,
|
| 127 |
+
"learning_rate": 8.301291989664083e-06,
|
| 128 |
+
"loss": 2.0194,
|
| 129 |
+
"step": 6579
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 18.0,
|
| 133 |
+
"grad_norm": 4.172696590423584,
|
| 134 |
+
"learning_rate": 8.201550387596899e-06,
|
| 135 |
+
"loss": 1.9955,
|
| 136 |
+
"step": 6966
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 19.0,
|
| 140 |
+
"grad_norm": 4.579292297363281,
|
| 141 |
+
"learning_rate": 8.1015503875969e-06,
|
| 142 |
+
"loss": 2.0042,
|
| 143 |
+
"step": 7353
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 20.0,
|
| 147 |
+
"grad_norm": 2.4596266746520996,
|
| 148 |
+
"learning_rate": 8.001550387596901e-06,
|
| 149 |
+
"loss": 1.9876,
|
| 150 |
+
"step": 7740
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 21.0,
|
| 154 |
+
"grad_norm": 3.477017641067505,
|
| 155 |
+
"learning_rate": 7.9015503875969e-06,
|
| 156 |
+
"loss": 1.9997,
|
| 157 |
+
"step": 8127
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 22.0,
|
| 161 |
+
"grad_norm": 3.163322925567627,
|
| 162 |
+
"learning_rate": 7.8015503875969e-06,
|
| 163 |
+
"loss": 1.9954,
|
| 164 |
+
"step": 8514
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 23.0,
|
| 168 |
+
"grad_norm": 3.398844003677368,
|
| 169 |
+
"learning_rate": 7.701808785529717e-06,
|
| 170 |
+
"loss": 1.9795,
|
| 171 |
+
"step": 8901
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 24.0,
|
| 175 |
+
"grad_norm": 2.4849860668182373,
|
| 176 |
+
"learning_rate": 7.601808785529716e-06,
|
| 177 |
+
"loss": 1.9721,
|
| 178 |
+
"step": 9288
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 25.0,
|
| 182 |
+
"grad_norm": 3.936396598815918,
|
| 183 |
+
"learning_rate": 7.5018087855297165e-06,
|
| 184 |
+
"loss": 1.9711,
|
| 185 |
+
"step": 9675
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 26.0,
|
| 189 |
+
"grad_norm": 3.5928125381469727,
|
| 190 |
+
"learning_rate": 7.401808785529716e-06,
|
| 191 |
+
"loss": 1.9699,
|
| 192 |
+
"step": 10062
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 27.0,
|
| 196 |
+
"grad_norm": 4.320117473602295,
|
| 197 |
+
"learning_rate": 7.301808785529716e-06,
|
| 198 |
+
"loss": 1.9653,
|
| 199 |
+
"step": 10449
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 28.0,
|
| 203 |
+
"grad_norm": 3.197213649749756,
|
| 204 |
+
"learning_rate": 7.2020671834625335e-06,
|
| 205 |
+
"loss": 1.9485,
|
| 206 |
+
"step": 10836
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 29.0,
|
| 210 |
+
"grad_norm": 5.386129379272461,
|
| 211 |
+
"learning_rate": 7.102067183462534e-06,
|
| 212 |
+
"loss": 1.9496,
|
| 213 |
+
"step": 11223
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 30.0,
|
| 217 |
+
"grad_norm": 3.7523598670959473,
|
| 218 |
+
"learning_rate": 7.002067183462533e-06,
|
| 219 |
+
"loss": 1.9362,
|
| 220 |
+
"step": 11610
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 31.0,
|
| 224 |
+
"grad_norm": 2.684497356414795,
|
| 225 |
+
"learning_rate": 6.902067183462533e-06,
|
| 226 |
+
"loss": 1.9541,
|
| 227 |
+
"step": 11997
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 32.0,
|
| 231 |
+
"grad_norm": 3.1632211208343506,
|
| 232 |
+
"learning_rate": 6.802067183462533e-06,
|
| 233 |
+
"loss": 1.9393,
|
| 234 |
+
"step": 12384
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 33.0,
|
| 238 |
+
"grad_norm": 3.8801021575927734,
|
| 239 |
+
"learning_rate": 6.702067183462533e-06,
|
| 240 |
+
"loss": 1.9434,
|
| 241 |
+
"step": 12771
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 34.0,
|
| 245 |
+
"grad_norm": 5.696396350860596,
|
| 246 |
+
"learning_rate": 6.602325581395349e-06,
|
| 247 |
+
"loss": 1.9381,
|
| 248 |
+
"step": 13158
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 35.0,
|
| 252 |
+
"grad_norm": 4.289793968200684,
|
| 253 |
+
"learning_rate": 6.502325581395349e-06,
|
| 254 |
+
"loss": 1.907,
|
| 255 |
+
"step": 13545
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 36.0,
|
| 259 |
+
"grad_norm": 2.529639720916748,
|
| 260 |
+
"learning_rate": 6.402325581395349e-06,
|
| 261 |
+
"loss": 1.9195,
|
| 262 |
+
"step": 13932
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 37.0,
|
| 266 |
+
"grad_norm": 3.81178879737854,
|
| 267 |
+
"learning_rate": 6.30232558139535e-06,
|
| 268 |
+
"loss": 1.9165,
|
| 269 |
+
"step": 14319
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 38.0,
|
| 273 |
+
"grad_norm": 2.9904661178588867,
|
| 274 |
+
"learning_rate": 6.2023255813953495e-06,
|
| 275 |
+
"loss": 1.9256,
|
| 276 |
+
"step": 14706
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 39.0,
|
| 280 |
+
"grad_norm": 3.6421236991882324,
|
| 281 |
+
"learning_rate": 6.10232558139535e-06,
|
| 282 |
+
"loss": 1.9113,
|
| 283 |
+
"step": 15093
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 40.0,
|
| 287 |
+
"grad_norm": 2.2317612171173096,
|
| 288 |
+
"learning_rate": 6.002583979328166e-06,
|
| 289 |
+
"loss": 1.9097,
|
| 290 |
+
"step": 15480
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 41.0,
|
| 294 |
+
"grad_norm": 3.1146578788757324,
|
| 295 |
+
"learning_rate": 5.902583979328166e-06,
|
| 296 |
+
"loss": 1.9052,
|
| 297 |
+
"step": 15867
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 42.0,
|
| 301 |
+
"grad_norm": 2.7279844284057617,
|
| 302 |
+
"learning_rate": 5.802583979328166e-06,
|
| 303 |
+
"loss": 1.8954,
|
| 304 |
+
"step": 16254
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 43.0,
|
| 308 |
+
"grad_norm": 4.926011085510254,
|
| 309 |
+
"learning_rate": 5.702583979328166e-06,
|
| 310 |
+
"loss": 1.8875,
|
| 311 |
+
"step": 16641
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 44.0,
|
| 315 |
+
"grad_norm": 3.347382068634033,
|
| 316 |
+
"learning_rate": 5.602583979328166e-06,
|
| 317 |
+
"loss": 1.8773,
|
| 318 |
+
"step": 17028
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 45.0,
|
| 322 |
+
"grad_norm": 4.180445671081543,
|
| 323 |
+
"learning_rate": 5.502583979328165e-06,
|
| 324 |
+
"loss": 1.9004,
|
| 325 |
+
"step": 17415
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 46.0,
|
| 329 |
+
"grad_norm": 4.547725677490234,
|
| 330 |
+
"learning_rate": 5.402842377260983e-06,
|
| 331 |
+
"loss": 1.8963,
|
| 332 |
+
"step": 17802
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 47.0,
|
| 336 |
+
"grad_norm": 3.7642316818237305,
|
| 337 |
+
"learning_rate": 5.302842377260983e-06,
|
| 338 |
+
"loss": 1.8963,
|
| 339 |
+
"step": 18189
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 48.0,
|
| 343 |
+
"grad_norm": 3.2858059406280518,
|
| 344 |
+
"learning_rate": 5.202842377260983e-06,
|
| 345 |
+
"loss": 1.8892,
|
| 346 |
+
"step": 18576
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 49.0,
|
| 350 |
+
"grad_norm": 2.402951240539551,
|
| 351 |
+
"learning_rate": 5.102842377260982e-06,
|
| 352 |
+
"loss": 1.8924,
|
| 353 |
+
"step": 18963
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 50.0,
|
| 357 |
+
"grad_norm": 2.7793326377868652,
|
| 358 |
+
"learning_rate": 5.002842377260982e-06,
|
| 359 |
+
"loss": 1.893,
|
| 360 |
+
"step": 19350
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 51.0,
|
| 364 |
+
"grad_norm": 3.6147751808166504,
|
| 365 |
+
"learning_rate": 4.9028423772609825e-06,
|
| 366 |
+
"loss": 1.8781,
|
| 367 |
+
"step": 19737
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 52.0,
|
| 371 |
+
"grad_norm": 4.644423484802246,
|
| 372 |
+
"learning_rate": 4.803100775193798e-06,
|
| 373 |
+
"loss": 1.8664,
|
| 374 |
+
"step": 20124
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 53.0,
|
| 378 |
+
"grad_norm": 3.9448044300079346,
|
| 379 |
+
"learning_rate": 4.703100775193799e-06,
|
| 380 |
+
"loss": 1.8898,
|
| 381 |
+
"step": 20511
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 54.0,
|
| 385 |
+
"grad_norm": 3.2738654613494873,
|
| 386 |
+
"learning_rate": 4.6031007751937986e-06,
|
| 387 |
+
"loss": 1.876,
|
| 388 |
+
"step": 20898
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 55.0,
|
| 392 |
+
"grad_norm": 2.4791011810302734,
|
| 393 |
+
"learning_rate": 4.503100775193799e-06,
|
| 394 |
+
"loss": 1.864,
|
| 395 |
+
"step": 21285
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 56.0,
|
| 399 |
+
"grad_norm": 2.873185873031616,
|
| 400 |
+
"learning_rate": 4.403100775193799e-06,
|
| 401 |
+
"loss": 1.8607,
|
| 402 |
+
"step": 21672
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 57.0,
|
| 406 |
+
"grad_norm": 3.4380970001220703,
|
| 407 |
+
"learning_rate": 4.303100775193799e-06,
|
| 408 |
+
"loss": 1.8788,
|
| 409 |
+
"step": 22059
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 58.0,
|
| 413 |
+
"grad_norm": 6.063451766967773,
|
| 414 |
+
"learning_rate": 4.203100775193798e-06,
|
| 415 |
+
"loss": 1.8555,
|
| 416 |
+
"step": 22446
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 59.0,
|
| 420 |
+
"grad_norm": 3.49165415763855,
|
| 421 |
+
"learning_rate": 4.103100775193799e-06,
|
| 422 |
+
"loss": 1.8728,
|
| 423 |
+
"step": 22833
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 60.0,
|
| 427 |
+
"grad_norm": 3.0042760372161865,
|
| 428 |
+
"learning_rate": 4.003100775193799e-06,
|
| 429 |
+
"loss": 1.8626,
|
| 430 |
+
"step": 23220
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 61.0,
|
| 434 |
+
"grad_norm": 5.144619464874268,
|
| 435 |
+
"learning_rate": 3.9031007751937985e-06,
|
| 436 |
+
"loss": 1.8748,
|
| 437 |
+
"step": 23607
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 62.0,
|
| 441 |
+
"grad_norm": 3.293827533721924,
|
| 442 |
+
"learning_rate": 3.8031007751937986e-06,
|
| 443 |
+
"loss": 1.8672,
|
| 444 |
+
"step": 23994
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 63.0,
|
| 448 |
+
"grad_norm": 3.602724313735962,
|
| 449 |
+
"learning_rate": 3.7033591731266153e-06,
|
| 450 |
+
"loss": 1.8595,
|
| 451 |
+
"step": 24381
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 64.0,
|
| 455 |
+
"grad_norm": 4.562984466552734,
|
| 456 |
+
"learning_rate": 3.6033591731266154e-06,
|
| 457 |
+
"loss": 1.8689,
|
| 458 |
+
"step": 24768
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 65.0,
|
| 462 |
+
"grad_norm": 1.9634078741073608,
|
| 463 |
+
"learning_rate": 3.5036175710594318e-06,
|
| 464 |
+
"loss": 1.8655,
|
| 465 |
+
"step": 25155
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 66.0,
|
| 469 |
+
"grad_norm": 2.359875440597534,
|
| 470 |
+
"learning_rate": 3.4036175710594314e-06,
|
| 471 |
+
"loss": 1.853,
|
| 472 |
+
"step": 25542
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 67.0,
|
| 476 |
+
"grad_norm": 2.7939252853393555,
|
| 477 |
+
"learning_rate": 3.303617571059432e-06,
|
| 478 |
+
"loss": 1.8539,
|
| 479 |
+
"step": 25929
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 68.0,
|
| 483 |
+
"grad_norm": 3.8576886653900146,
|
| 484 |
+
"learning_rate": 3.203617571059432e-06,
|
| 485 |
+
"loss": 1.8579,
|
| 486 |
+
"step": 26316
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 69.0,
|
| 490 |
+
"grad_norm": 3.667564630508423,
|
| 491 |
+
"learning_rate": 3.1036175710594318e-06,
|
| 492 |
+
"loss": 1.8474,
|
| 493 |
+
"step": 26703
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 70.0,
|
| 497 |
+
"grad_norm": 2.811600923538208,
|
| 498 |
+
"learning_rate": 3.003875968992248e-06,
|
| 499 |
+
"loss": 1.8442,
|
| 500 |
+
"step": 27090
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 71.0,
|
| 504 |
+
"grad_norm": 4.078073978424072,
|
| 505 |
+
"learning_rate": 2.9038759689922486e-06,
|
| 506 |
+
"loss": 1.8308,
|
| 507 |
+
"step": 27477
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 72.0,
|
| 511 |
+
"grad_norm": 2.925967216491699,
|
| 512 |
+
"learning_rate": 2.8038759689922483e-06,
|
| 513 |
+
"loss": 1.852,
|
| 514 |
+
"step": 27864
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 73.0,
|
| 518 |
+
"grad_norm": 3.747209072113037,
|
| 519 |
+
"learning_rate": 2.7038759689922484e-06,
|
| 520 |
+
"loss": 1.8525,
|
| 521 |
+
"step": 28251
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 74.0,
|
| 525 |
+
"grad_norm": 3.608250141143799,
|
| 526 |
+
"learning_rate": 2.603875968992248e-06,
|
| 527 |
+
"loss": 1.8525,
|
| 528 |
+
"step": 28638
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 75.0,
|
| 532 |
+
"grad_norm": 3.8558218479156494,
|
| 533 |
+
"learning_rate": 2.503875968992248e-06,
|
| 534 |
+
"loss": 1.854,
|
| 535 |
+
"step": 29025
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 76.0,
|
| 539 |
+
"grad_norm": 5.439558982849121,
|
| 540 |
+
"learning_rate": 2.4041343669250645e-06,
|
| 541 |
+
"loss": 1.8398,
|
| 542 |
+
"step": 29412
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 77.0,
|
| 546 |
+
"grad_norm": 3.1695291996002197,
|
| 547 |
+
"learning_rate": 2.304134366925065e-06,
|
| 548 |
+
"loss": 1.8546,
|
| 549 |
+
"step": 29799
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 78.0,
|
| 553 |
+
"grad_norm": 5.20058012008667,
|
| 554 |
+
"learning_rate": 2.2041343669250647e-06,
|
| 555 |
+
"loss": 1.8436,
|
| 556 |
+
"step": 30186
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 79.0,
|
| 560 |
+
"grad_norm": 3.13725209236145,
|
| 561 |
+
"learning_rate": 2.104134366925065e-06,
|
| 562 |
+
"loss": 1.8588,
|
| 563 |
+
"step": 30573
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 80.0,
|
| 567 |
+
"grad_norm": 3.2893519401550293,
|
| 568 |
+
"learning_rate": 2.004134366925065e-06,
|
| 569 |
+
"loss": 1.8483,
|
| 570 |
+
"step": 30960
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 81.0,
|
| 574 |
+
"grad_norm": 3.6456849575042725,
|
| 575 |
+
"learning_rate": 1.9041343669250649e-06,
|
| 576 |
+
"loss": 1.8335,
|
| 577 |
+
"step": 31347
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 82.0,
|
| 581 |
+
"grad_norm": 2.9292547702789307,
|
| 582 |
+
"learning_rate": 1.8041343669250648e-06,
|
| 583 |
+
"loss": 1.8392,
|
| 584 |
+
"step": 31734
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 83.0,
|
| 588 |
+
"grad_norm": 1.7664443254470825,
|
| 589 |
+
"learning_rate": 1.7041343669250649e-06,
|
| 590 |
+
"loss": 1.8599,
|
| 591 |
+
"step": 32121
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 84.0,
|
| 595 |
+
"grad_norm": 2.2330174446105957,
|
| 596 |
+
"learning_rate": 1.6041343669250648e-06,
|
| 597 |
+
"loss": 1.8354,
|
| 598 |
+
"step": 32508
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 85.0,
|
| 602 |
+
"grad_norm": 2.521230936050415,
|
| 603 |
+
"learning_rate": 1.5043927648578813e-06,
|
| 604 |
+
"loss": 1.8449,
|
| 605 |
+
"step": 32895
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 86.0,
|
| 609 |
+
"grad_norm": 3.481978178024292,
|
| 610 |
+
"learning_rate": 1.4043927648578812e-06,
|
| 611 |
+
"loss": 1.8249,
|
| 612 |
+
"step": 33282
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 87.0,
|
| 616 |
+
"grad_norm": 2.547130823135376,
|
| 617 |
+
"learning_rate": 1.3043927648578813e-06,
|
| 618 |
+
"loss": 1.8433,
|
| 619 |
+
"step": 33669
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 88.0,
|
| 623 |
+
"grad_norm": 3.2912960052490234,
|
| 624 |
+
"learning_rate": 1.2043927648578812e-06,
|
| 625 |
+
"loss": 1.8207,
|
| 626 |
+
"step": 34056
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 89.0,
|
| 630 |
+
"grad_norm": 5.078946113586426,
|
| 631 |
+
"learning_rate": 1.104392764857881e-06,
|
| 632 |
+
"loss": 1.8373,
|
| 633 |
+
"step": 34443
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 90.0,
|
| 637 |
+
"grad_norm": 3.1195156574249268,
|
| 638 |
+
"learning_rate": 1.0043927648578812e-06,
|
| 639 |
+
"loss": 1.8433,
|
| 640 |
+
"step": 34830
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 91.0,
|
| 644 |
+
"grad_norm": 3.4918460845947266,
|
| 645 |
+
"learning_rate": 9.043927648578812e-07,
|
| 646 |
+
"loss": 1.8358,
|
| 647 |
+
"step": 35217
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 92.0,
|
| 651 |
+
"grad_norm": 3.529872179031372,
|
| 652 |
+
"learning_rate": 8.043927648578811e-07,
|
| 653 |
+
"loss": 1.8395,
|
| 654 |
+
"step": 35604
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 93.0,
|
| 658 |
+
"grad_norm": 1.9922144412994385,
|
| 659 |
+
"learning_rate": 7.043927648578812e-07,
|
| 660 |
+
"loss": 1.8397,
|
| 661 |
+
"step": 35991
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 94.0,
|
| 665 |
+
"grad_norm": 2.5432543754577637,
|
| 666 |
+
"learning_rate": 6.043927648578812e-07,
|
| 667 |
+
"loss": 1.8323,
|
| 668 |
+
"step": 36378
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 95.0,
|
| 672 |
+
"grad_norm": 2.306816816329956,
|
| 673 |
+
"learning_rate": 5.043927648578812e-07,
|
| 674 |
+
"loss": 1.85,
|
| 675 |
+
"step": 36765
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 96.0,
|
| 679 |
+
"grad_norm": 3.3451128005981445,
|
| 680 |
+
"learning_rate": 4.0465116279069773e-07,
|
| 681 |
+
"loss": 1.8461,
|
| 682 |
+
"step": 37152
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 97.0,
|
| 686 |
+
"grad_norm": 2.5839929580688477,
|
| 687 |
+
"learning_rate": 3.046511627906977e-07,
|
| 688 |
+
"loss": 1.8522,
|
| 689 |
+
"step": 37539
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 98.0,
|
| 693 |
+
"grad_norm": 3.5228729248046875,
|
| 694 |
+
"learning_rate": 2.0465116279069766e-07,
|
| 695 |
+
"loss": 1.8449,
|
| 696 |
+
"step": 37926
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 99.0,
|
| 700 |
+
"grad_norm": 2.9826176166534424,
|
| 701 |
+
"learning_rate": 1.0465116279069768e-07,
|
| 702 |
+
"loss": 1.8333,
|
| 703 |
+
"step": 38313
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 100.0,
|
| 707 |
+
"grad_norm": 3.2512013912200928,
|
| 708 |
+
"learning_rate": 4.651162790697675e-09,
|
| 709 |
+
"loss": 1.8459,
|
| 710 |
+
"step": 38700
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 100.0,
|
| 714 |
+
"step": 38700,
|
| 715 |
+
"total_flos": 2.0309886965211648e+18,
|
| 716 |
+
"train_loss": 1.927858536594598,
|
| 717 |
+
"train_runtime": 42079.8744,
|
| 718 |
+
"train_samples_per_second": 183.867,
|
| 719 |
+
"train_steps_per_second": 0.92
|
| 720 |
+
}
|
| 721 |
+
],
|
| 722 |
+
"logging_steps": 387,
|
| 723 |
+
"max_steps": 38700,
|
| 724 |
+
"num_input_tokens_seen": 0,
|
| 725 |
+
"num_train_epochs": 100,
|
| 726 |
+
"save_steps": 6800,
|
| 727 |
+
"stateful_callbacks": {
|
| 728 |
+
"TrainerControl": {
|
| 729 |
+
"args": {
|
| 730 |
+
"should_epoch_stop": false,
|
| 731 |
+
"should_evaluate": false,
|
| 732 |
+
"should_log": false,
|
| 733 |
+
"should_save": true,
|
| 734 |
+
"should_training_stop": true
|
| 735 |
+
},
|
| 736 |
+
"attributes": {}
|
| 737 |
+
}
|
| 738 |
+
},
|
| 739 |
+
"total_flos": 2.0309886965211648e+18,
|
| 740 |
+
"train_batch_size": 50,
|
| 741 |
+
"trial_name": null,
|
| 742 |
+
"trial_params": null
|
| 743 |
+
}
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|