Commit
•
e3f67a1
1
Parent(s):
64e0ff0
Upload 13 files
Browse files- README.md +106 -0
- all_results.json +39 -0
- config.json +46 -0
- eval_results.json +34 -0
- merges.txt +0 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +15 -0
- train_results.json +8 -0
- trainer_state.json +1259 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model: roberta-base
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- accuracy
|
| 8 |
+
- recall
|
| 9 |
+
- f1
|
| 10 |
+
model-index:
|
| 11 |
+
- name: train
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
+
|
| 18 |
+
# train
|
| 19 |
+
|
| 20 |
+
This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
|
| 21 |
+
It achieves the following results on the evaluation set:
|
| 22 |
+
- Loss: 0.6648
|
| 23 |
+
- Accuracy: 0.7617
|
| 24 |
+
- B Acc: 0.6394
|
| 25 |
+
- Prec: 0.7595
|
| 26 |
+
- Recall: 0.7617
|
| 27 |
+
- F1: 0.7602
|
| 28 |
+
- Prec Joy: 0.7315
|
| 29 |
+
- Recall Joy: 0.7793
|
| 30 |
+
- F1 Joy: 0.7547
|
| 31 |
+
- Prec Anger: 0.6467
|
| 32 |
+
- Recall Anger: 0.6507
|
| 33 |
+
- F1 Anger: 0.6487
|
| 34 |
+
- Prec Disgust: 0.4710
|
| 35 |
+
- Recall Disgust: 0.45
|
| 36 |
+
- F1 Disgust: 0.4603
|
| 37 |
+
- Prec Fear: 0.6963
|
| 38 |
+
- Recall Fear: 0.6409
|
| 39 |
+
- F1 Fear: 0.6675
|
| 40 |
+
- Prec Neutral: 0.8457
|
| 41 |
+
- Recall Neutral: 0.8490
|
| 42 |
+
- F1 Neutral: 0.8474
|
| 43 |
+
- Prec Sadness: 0.7094
|
| 44 |
+
- Recall Sadness: 0.6738
|
| 45 |
+
- F1 Sadness: 0.6911
|
| 46 |
+
- Prec Surprise: 0.5228
|
| 47 |
+
- Recall Surprise: 0.4323
|
| 48 |
+
- F1 Surprise: 0.4732
|
| 49 |
+
|
| 50 |
+
## Model description
|
| 51 |
+
|
| 52 |
+
More information needed
|
| 53 |
+
|
| 54 |
+
## Intended uses & limitations
|
| 55 |
+
|
| 56 |
+
More information needed
|
| 57 |
+
|
| 58 |
+
## Training and evaluation data
|
| 59 |
+
|
| 60 |
+
More information needed
|
| 61 |
+
|
| 62 |
+
## Training procedure
|
| 63 |
+
|
| 64 |
+
### Training hyperparameters
|
| 65 |
+
|
| 66 |
+
The following hyperparameters were used during training:
|
| 67 |
+
- learning_rate: 5e-05
|
| 68 |
+
- train_batch_size: 128
|
| 69 |
+
- eval_batch_size: 128
|
| 70 |
+
- seed: 42
|
| 71 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 72 |
+
- lr_scheduler_type: linear
|
| 73 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 74 |
+
- num_epochs: 3.0
|
| 75 |
+
|
| 76 |
+
### Training results
|
| 77 |
+
|
| 78 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | B Acc | Prec | Recall | F1 | Prec Joy | Recall Joy | F1 Joy | Prec Anger | Recall Anger | F1 Anger | Prec Disgust | Recall Disgust | F1 Disgust | Prec Fear | Recall Fear | F1 Fear | Prec Neutral | Recall Neutral | F1 Neutral | Prec Sadness | Recall Sadness | F1 Sadness | Prec Surprise | Recall Surprise | F1 Surprise |
|
| 79 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:------:|:------:|:------:|:--------:|:----------:|:------:|:----------:|:------------:|:--------:|:------------:|:--------------:|:----------:|:---------:|:-----------:|:-------:|:------------:|:--------------:|:----------:|:------------:|:--------------:|:----------:|:-------------:|:---------------:|:-----------:|
|
| 80 |
+
| 0.9538 | 0.15 | 232 | 0.8701 | 0.6961 | 0.4790 | 0.6837 | 0.6961 | 0.6837 | 0.7401 | 0.6381 | 0.6853 | 0.4622 | 0.5391 | 0.4977 | 0.25 | 0.0018 | 0.0035 | 0.5527 | 0.4292 | 0.4832 | 0.7965 | 0.8618 | 0.8279 | 0.5281 | 0.6431 | 0.5800 | 0.3562 | 0.2398 | 0.2866 |
|
| 81 |
+
| 0.7952 | 0.3 | 464 | 0.8010 | 0.7168 | 0.5242 | 0.7098 | 0.7168 | 0.7025 | 0.8084 | 0.5948 | 0.6853 | 0.5732 | 0.4710 | 0.5171 | 0.4713 | 0.2643 | 0.3387 | 0.6156 | 0.5263 | 0.5675 | 0.7405 | 0.9250 | 0.8226 | 0.6858 | 0.5676 | 0.6211 | 0.4448 | 0.3204 | 0.3725 |
|
| 82 |
+
| 0.7528 | 0.45 | 696 | 0.7560 | 0.7261 | 0.5878 | 0.7309 | 0.7261 | 0.7256 | 0.6969 | 0.7646 | 0.7292 | 0.5550 | 0.5534 | 0.5542 | 0.3409 | 0.4821 | 0.3994 | 0.7225 | 0.4842 | 0.5798 | 0.8476 | 0.8159 | 0.8314 | 0.6118 | 0.7027 | 0.6541 | 0.4957 | 0.3118 | 0.3828 |
|
| 83 |
+
| 0.7334 | 0.6 | 928 | 0.7310 | 0.7370 | 0.5868 | 0.7345 | 0.7370 | 0.7283 | 0.7170 | 0.7458 | 0.7311 | 0.7129 | 0.4116 | 0.5219 | 0.3727 | 0.5696 | 0.4506 | 0.6671 | 0.5626 | 0.6104 | 0.7898 | 0.8859 | 0.8351 | 0.7318 | 0.5844 | 0.6499 | 0.5252 | 0.3473 | 0.4181 |
|
| 84 |
+
| 0.7216 | 0.75 | 1160 | 0.7043 | 0.7448 | 0.6009 | 0.7403 | 0.7448 | 0.7389 | 0.7767 | 0.6826 | 0.7266 | 0.6159 | 0.5386 | 0.5746 | 0.5302 | 0.4393 | 0.4805 | 0.8023 | 0.5602 | 0.6598 | 0.7854 | 0.8926 | 0.8356 | 0.7005 | 0.632 | 0.6645 | 0.4815 | 0.4613 | 0.4712 |
|
| 85 |
+
| 0.7259 | 0.9 | 1392 | 0.6962 | 0.7475 | 0.6082 | 0.7433 | 0.7475 | 0.7412 | 0.7355 | 0.7586 | 0.7469 | 0.6758 | 0.4504 | 0.5405 | 0.3908 | 0.5589 | 0.4600 | 0.6939 | 0.6070 | 0.6475 | 0.8122 | 0.8744 | 0.8421 | 0.6830 | 0.6676 | 0.6752 | 0.5494 | 0.3409 | 0.4207 |
|
| 86 |
+
| 0.6362 | 1.05 | 1624 | 0.6771 | 0.7526 | 0.6055 | 0.7472 | 0.7526 | 0.7484 | 0.7392 | 0.7483 | 0.7437 | 0.5873 | 0.6191 | 0.6028 | 0.5302 | 0.3768 | 0.4405 | 0.7388 | 0.5789 | 0.6492 | 0.8213 | 0.8670 | 0.8435 | 0.7090 | 0.6507 | 0.6786 | 0.5301 | 0.3978 | 0.4545 |
|
| 87 |
+
| 0.621 | 1.2 | 1856 | 0.6779 | 0.7528 | 0.6120 | 0.7494 | 0.7528 | 0.7487 | 0.7107 | 0.7828 | 0.7450 | 0.6508 | 0.5913 | 0.6196 | 0.4980 | 0.4518 | 0.4738 | 0.7963 | 0.5532 | 0.6529 | 0.8165 | 0.8590 | 0.8372 | 0.7499 | 0.6236 | 0.6809 | 0.5078 | 0.4226 | 0.4613 |
|
| 88 |
+
| 0.6241 | 1.35 | 2088 | 0.6849 | 0.7513 | 0.6367 | 0.7526 | 0.7513 | 0.7514 | 0.7429 | 0.7592 | 0.7510 | 0.5795 | 0.6531 | 0.6141 | 0.4372 | 0.4661 | 0.4512 | 0.6462 | 0.6515 | 0.6488 | 0.8492 | 0.8372 | 0.8432 | 0.6887 | 0.6609 | 0.6745 | 0.5271 | 0.4290 | 0.4730 |
|
| 89 |
+
| 0.6188 | 1.5 | 2320 | 0.6713 | 0.7579 | 0.6159 | 0.7539 | 0.7579 | 0.7534 | 0.7071 | 0.7971 | 0.7494 | 0.6343 | 0.6267 | 0.6305 | 0.5877 | 0.3768 | 0.4592 | 0.7247 | 0.6281 | 0.6729 | 0.8361 | 0.8496 | 0.8428 | 0.6943 | 0.6693 | 0.6816 | 0.5919 | 0.3634 | 0.4504 |
|
| 90 |
+
| 0.6182 | 1.65 | 2552 | 0.6608 | 0.7601 | 0.6199 | 0.7567 | 0.7601 | 0.7566 | 0.7143 | 0.7891 | 0.7498 | 0.6163 | 0.6358 | 0.6259 | 0.5607 | 0.3875 | 0.4583 | 0.7591 | 0.6082 | 0.6753 | 0.8375 | 0.8578 | 0.8475 | 0.7324 | 0.6436 | 0.6851 | 0.5381 | 0.4172 | 0.4700 |
|
| 91 |
+
| 0.6392 | 1.8 | 2784 | 0.6542 | 0.7624 | 0.6261 | 0.7593 | 0.7624 | 0.7596 | 0.7513 | 0.7584 | 0.7548 | 0.5970 | 0.6708 | 0.6318 | 0.5711 | 0.3875 | 0.4617 | 0.7482 | 0.6152 | 0.6752 | 0.8379 | 0.8635 | 0.8505 | 0.7076 | 0.668 | 0.6872 | 0.5132 | 0.4194 | 0.4615 |
|
| 92 |
+
| 0.6158 | 1.95 | 3016 | 0.6456 | 0.7649 | 0.6279 | 0.7599 | 0.7649 | 0.7614 | 0.7490 | 0.7548 | 0.7519 | 0.6402 | 0.6378 | 0.6390 | 0.5314 | 0.4232 | 0.4712 | 0.7569 | 0.6117 | 0.6766 | 0.8310 | 0.8753 | 0.8526 | 0.7199 | 0.6627 | 0.6901 | 0.5063 | 0.4301 | 0.4651 |
|
| 93 |
+
| 0.554 | 2.1 | 3248 | 0.6742 | 0.7584 | 0.6346 | 0.7555 | 0.7584 | 0.7564 | 0.7293 | 0.7732 | 0.7506 | 0.6433 | 0.6430 | 0.6432 | 0.5031 | 0.4393 | 0.4690 | 0.7292 | 0.6363 | 0.6796 | 0.8347 | 0.8496 | 0.8421 | 0.7163 | 0.6587 | 0.6863 | 0.5049 | 0.4419 | 0.4713 |
|
| 94 |
+
| 0.5537 | 2.25 | 3480 | 0.6708 | 0.7633 | 0.6283 | 0.7604 | 0.7633 | 0.7605 | 0.7263 | 0.7801 | 0.7523 | 0.6304 | 0.6612 | 0.6455 | 0.5806 | 0.3732 | 0.4543 | 0.7486 | 0.6094 | 0.6718 | 0.8442 | 0.8528 | 0.8485 | 0.6982 | 0.692 | 0.6951 | 0.5356 | 0.4290 | 0.4764 |
|
| 95 |
+
| 0.5375 | 2.4 | 3712 | 0.6712 | 0.7606 | 0.6402 | 0.7592 | 0.7606 | 0.7595 | 0.7373 | 0.7709 | 0.7537 | 0.6245 | 0.6608 | 0.6421 | 0.4827 | 0.4482 | 0.4648 | 0.7319 | 0.6257 | 0.6747 | 0.8454 | 0.8474 | 0.8464 | 0.7006 | 0.6769 | 0.6885 | 0.5204 | 0.4516 | 0.4836 |
|
| 96 |
+
| 0.5175 | 2.55 | 3944 | 0.6625 | 0.7625 | 0.6369 | 0.7600 | 0.7625 | 0.7604 | 0.7422 | 0.7642 | 0.7530 | 0.6335 | 0.6526 | 0.6429 | 0.4481 | 0.4929 | 0.4694 | 0.7482 | 0.6187 | 0.6773 | 0.8374 | 0.8604 | 0.8488 | 0.7252 | 0.6684 | 0.6957 | 0.5321 | 0.4011 | 0.4574 |
|
| 97 |
+
| 0.5182 | 2.7 | 4176 | 0.6621 | 0.7631 | 0.6404 | 0.7602 | 0.7631 | 0.7612 | 0.7343 | 0.7766 | 0.7549 | 0.6491 | 0.6392 | 0.6441 | 0.4739 | 0.4536 | 0.4635 | 0.6784 | 0.6538 | 0.6659 | 0.8444 | 0.8529 | 0.8486 | 0.7109 | 0.684 | 0.6972 | 0.5458 | 0.4226 | 0.4764 |
|
| 98 |
+
| 0.5148 | 2.85 | 4408 | 0.6638 | 0.7637 | 0.6383 | 0.7598 | 0.7637 | 0.7612 | 0.7394 | 0.7741 | 0.7563 | 0.6741 | 0.6205 | 0.6462 | 0.5 | 0.4375 | 0.4667 | 0.6813 | 0.6550 | 0.6679 | 0.8400 | 0.8572 | 0.8485 | 0.6922 | 0.6916 | 0.6919 | 0.5296 | 0.4323 | 0.4760 |
|
| 99 |
+
|
| 100 |
+
|
| 101 |
+
### Framework versions
|
| 102 |
+
|
| 103 |
+
- Transformers 4.31.0
|
| 104 |
+
- Pytorch 2.0.1+cu117
|
| 105 |
+
- Datasets 2.14.2
|
| 106 |
+
- Tokenizers 0.13.3
|
all_results.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.7617346477344669,
|
| 4 |
+
"eval_b_acc": 0.6394282040393414,
|
| 5 |
+
"eval_f1": 0.7601905030141421,
|
| 6 |
+
"eval_f1_anger": 0.6486744685932649,
|
| 7 |
+
"eval_f1_disgust": 0.46027397260273967,
|
| 8 |
+
"eval_f1_fear": 0.6674786845310596,
|
| 9 |
+
"eval_f1_joy": 0.7546596434359805,
|
| 10 |
+
"eval_f1_neutral": 0.8473718729504358,
|
| 11 |
+
"eval_f1_sadness": 0.6911328926373376,
|
| 12 |
+
"eval_f1_surprise": 0.47321954090641555,
|
| 13 |
+
"eval_loss": 0.6647528409957886,
|
| 14 |
+
"eval_prec": 0.7595317468114701,
|
| 15 |
+
"eval_prec_anger": 0.6466666666666666,
|
| 16 |
+
"eval_prec_disgust": 0.47102803738317756,
|
| 17 |
+
"eval_prec_fear": 0.6963151207115629,
|
| 18 |
+
"eval_prec_joy": 0.7315396700706991,
|
| 19 |
+
"eval_prec_neutral": 0.8457078735739667,
|
| 20 |
+
"eval_prec_sadness": 0.7094057089377632,
|
| 21 |
+
"eval_prec_surprise": 0.5227568270481144,
|
| 22 |
+
"eval_recall": 0.7617346477344669,
|
| 23 |
+
"eval_recall_anger": 0.6506947771921419,
|
| 24 |
+
"eval_recall_disgust": 0.45,
|
| 25 |
+
"eval_recall_fear": 0.6409356725146199,
|
| 26 |
+
"eval_recall_joy": 0.7792887029288703,
|
| 27 |
+
"eval_recall_neutral": 0.8490424333458505,
|
| 28 |
+
"eval_recall_sadness": 0.6737777777777778,
|
| 29 |
+
"eval_recall_surprise": 0.432258064516129,
|
| 30 |
+
"eval_runtime": 14.5728,
|
| 31 |
+
"eval_samples": 22114,
|
| 32 |
+
"eval_samples_per_second": 1517.489,
|
| 33 |
+
"eval_steps_per_second": 11.871,
|
| 34 |
+
"train_loss": 0.6570020180521673,
|
| 35 |
+
"train_runtime": 1377.6596,
|
| 36 |
+
"train_samples": 197800,
|
| 37 |
+
"train_samples_per_second": 430.73,
|
| 38 |
+
"train_steps_per_second": 3.367
|
| 39 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "roberta-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "anger",
|
| 15 |
+
"1": "disgust",
|
| 16 |
+
"2": "fear",
|
| 17 |
+
"3": "joy",
|
| 18 |
+
"4": "neutral",
|
| 19 |
+
"5": "sadness",
|
| 20 |
+
"6": "surprise"
|
| 21 |
+
},
|
| 22 |
+
"initializer_range": 0.02,
|
| 23 |
+
"intermediate_size": 3072,
|
| 24 |
+
"label2id": {
|
| 25 |
+
"anger": 0,
|
| 26 |
+
"disgust": 1,
|
| 27 |
+
"fear": 2,
|
| 28 |
+
"joy": 3,
|
| 29 |
+
"neutral": 4,
|
| 30 |
+
"sadness": 5,
|
| 31 |
+
"surprise": 6
|
| 32 |
+
},
|
| 33 |
+
"layer_norm_eps": 1e-05,
|
| 34 |
+
"max_position_embeddings": 514,
|
| 35 |
+
"model_type": "roberta",
|
| 36 |
+
"num_attention_heads": 12,
|
| 37 |
+
"num_hidden_layers": 12,
|
| 38 |
+
"pad_token_id": 1,
|
| 39 |
+
"position_embedding_type": "absolute",
|
| 40 |
+
"problem_type": "single_label_classification",
|
| 41 |
+
"torch_dtype": "float32",
|
| 42 |
+
"transformers_version": "4.31.0",
|
| 43 |
+
"type_vocab_size": 1,
|
| 44 |
+
"use_cache": true,
|
| 45 |
+
"vocab_size": 50265
|
| 46 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.7617346477344669,
|
| 4 |
+
"eval_b_acc": 0.6394282040393414,
|
| 5 |
+
"eval_f1": 0.7601905030141421,
|
| 6 |
+
"eval_f1_anger": 0.6486744685932649,
|
| 7 |
+
"eval_f1_disgust": 0.46027397260273967,
|
| 8 |
+
"eval_f1_fear": 0.6674786845310596,
|
| 9 |
+
"eval_f1_joy": 0.7546596434359805,
|
| 10 |
+
"eval_f1_neutral": 0.8473718729504358,
|
| 11 |
+
"eval_f1_sadness": 0.6911328926373376,
|
| 12 |
+
"eval_f1_surprise": 0.47321954090641555,
|
| 13 |
+
"eval_loss": 0.6647528409957886,
|
| 14 |
+
"eval_prec": 0.7595317468114701,
|
| 15 |
+
"eval_prec_anger": 0.6466666666666666,
|
| 16 |
+
"eval_prec_disgust": 0.47102803738317756,
|
| 17 |
+
"eval_prec_fear": 0.6963151207115629,
|
| 18 |
+
"eval_prec_joy": 0.7315396700706991,
|
| 19 |
+
"eval_prec_neutral": 0.8457078735739667,
|
| 20 |
+
"eval_prec_sadness": 0.7094057089377632,
|
| 21 |
+
"eval_prec_surprise": 0.5227568270481144,
|
| 22 |
+
"eval_recall": 0.7617346477344669,
|
| 23 |
+
"eval_recall_anger": 0.6506947771921419,
|
| 24 |
+
"eval_recall_disgust": 0.45,
|
| 25 |
+
"eval_recall_fear": 0.6409356725146199,
|
| 26 |
+
"eval_recall_joy": 0.7792887029288703,
|
| 27 |
+
"eval_recall_neutral": 0.8490424333458505,
|
| 28 |
+
"eval_recall_sadness": 0.6737777777777778,
|
| 29 |
+
"eval_recall_surprise": 0.432258064516129,
|
| 30 |
+
"eval_runtime": 14.5728,
|
| 31 |
+
"eval_samples": 22114,
|
| 32 |
+
"eval_samples_per_second": 1517.489,
|
| 33 |
+
"eval_steps_per_second": 11.871
|
| 34 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52ff68740dbd51618e05d8e926e72bd32831e0cc00fa4d8693a1c4b6ccd39cc0
|
| 3 |
+
size 498673009
|
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 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": "<s>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"cls_token": "<s>",
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"mask_token": "<mask>",
|
| 9 |
+
"model_max_length": 512,
|
| 10 |
+
"pad_token": "<pad>",
|
| 11 |
+
"sep_token": "</s>",
|
| 12 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 13 |
+
"trim_offsets": true,
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.6570020180521673,
|
| 4 |
+
"train_runtime": 1377.6596,
|
| 5 |
+
"train_samples": 197800,
|
| 6 |
+
"train_samples_per_second": 430.73,
|
| 7 |
+
"train_steps_per_second": 3.367
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1259 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 4638,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 5.064655172413793e-06,
|
| 13 |
+
"loss": 1.827,
|
| 14 |
+
"step": 47
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.06,
|
| 18 |
+
"learning_rate": 1.0129310344827586e-05,
|
| 19 |
+
"loss": 1.288,
|
| 20 |
+
"step": 94
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"learning_rate": 1.5086206896551724e-05,
|
| 25 |
+
"loss": 1.0612,
|
| 26 |
+
"step": 141
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.12,
|
| 30 |
+
"learning_rate": 2.0150862068965517e-05,
|
| 31 |
+
"loss": 0.9538,
|
| 32 |
+
"step": 188
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.15,
|
| 36 |
+
"eval_accuracy": 0.6960748846884327,
|
| 37 |
+
"eval_b_acc": 0.4789797300032158,
|
| 38 |
+
"eval_f1": 0.683716357682452,
|
| 39 |
+
"eval_f1_anger": 0.497677504976775,
|
| 40 |
+
"eval_f1_disgust": 0.0035460992907801418,
|
| 41 |
+
"eval_f1_fear": 0.4832126398946675,
|
| 42 |
+
"eval_f1_joy": 0.685316256600382,
|
| 43 |
+
"eval_f1_neutral": 0.827884745457005,
|
| 44 |
+
"eval_f1_sadness": 0.5799599198396794,
|
| 45 |
+
"eval_f1_surprise": 0.2866323907455013,
|
| 46 |
+
"eval_loss": 0.8701004981994629,
|
| 47 |
+
"eval_prec": 0.6836877535141815,
|
| 48 |
+
"eval_prec_anger": 0.462202136400986,
|
| 49 |
+
"eval_prec_disgust": 0.25,
|
| 50 |
+
"eval_prec_fear": 0.552710843373494,
|
| 51 |
+
"eval_prec_joy": 0.7401116233923805,
|
| 52 |
+
"eval_prec_neutral": 0.796529284164859,
|
| 53 |
+
"eval_prec_sadness": 0.5281021897810219,
|
| 54 |
+
"eval_prec_surprise": 0.3562300319488818,
|
| 55 |
+
"eval_recall": 0.6960748846884327,
|
| 56 |
+
"eval_recall_anger": 0.5390512697652132,
|
| 57 |
+
"eval_recall_disgust": 0.0017857142857142857,
|
| 58 |
+
"eval_recall_fear": 0.42923976608187137,
|
| 59 |
+
"eval_recall_joy": 0.6380753138075314,
|
| 60 |
+
"eval_recall_neutral": 0.86180998873451,
|
| 61 |
+
"eval_recall_sadness": 0.6431111111111111,
|
| 62 |
+
"eval_recall_surprise": 0.23978494623655913,
|
| 63 |
+
"eval_runtime": 16.0297,
|
| 64 |
+
"eval_samples_per_second": 1379.566,
|
| 65 |
+
"eval_steps_per_second": 10.792,
|
| 66 |
+
"step": 232
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.15,
|
| 70 |
+
"learning_rate": 2.521551724137931e-05,
|
| 71 |
+
"loss": 0.8837,
|
| 72 |
+
"step": 235
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18,
|
| 76 |
+
"learning_rate": 3.0280172413793107e-05,
|
| 77 |
+
"loss": 0.881,
|
| 78 |
+
"step": 282
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.21,
|
| 82 |
+
"learning_rate": 3.53448275862069e-05,
|
| 83 |
+
"loss": 0.8462,
|
| 84 |
+
"step": 329
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.24,
|
| 88 |
+
"learning_rate": 4.0409482758620696e-05,
|
| 89 |
+
"loss": 0.8242,
|
| 90 |
+
"step": 376
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.27,
|
| 94 |
+
"learning_rate": 4.5474137931034485e-05,
|
| 95 |
+
"loss": 0.7952,
|
| 96 |
+
"step": 423
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.3,
|
| 100 |
+
"eval_accuracy": 0.7168309668083567,
|
| 101 |
+
"eval_b_acc": 0.5241940962877409,
|
| 102 |
+
"eval_f1": 0.7025228198992769,
|
| 103 |
+
"eval_f1_anger": 0.5170962651236192,
|
| 104 |
+
"eval_f1_disgust": 0.3386727688787185,
|
| 105 |
+
"eval_f1_fear": 0.5674653215636821,
|
| 106 |
+
"eval_f1_joy": 0.6853079426298663,
|
| 107 |
+
"eval_f1_neutral": 0.8225570814375757,
|
| 108 |
+
"eval_f1_sadness": 0.6211089494163424,
|
| 109 |
+
"eval_f1_surprise": 0.37250000000000005,
|
| 110 |
+
"eval_loss": 0.8009697198867798,
|
| 111 |
+
"eval_prec": 0.7097552640353452,
|
| 112 |
+
"eval_prec_anger": 0.5731778425655977,
|
| 113 |
+
"eval_prec_disgust": 0.4713375796178344,
|
| 114 |
+
"eval_prec_fear": 0.615595075239398,
|
| 115 |
+
"eval_prec_joy": 0.8083593972135342,
|
| 116 |
+
"eval_prec_neutral": 0.7405486659150695,
|
| 117 |
+
"eval_prec_sadness": 0.6858216970998926,
|
| 118 |
+
"eval_prec_surprise": 0.44477611940298506,
|
| 119 |
+
"eval_recall": 0.7168309668083567,
|
| 120 |
+
"eval_recall_anger": 0.4710110206037374,
|
| 121 |
+
"eval_recall_disgust": 0.2642857142857143,
|
| 122 |
+
"eval_recall_fear": 0.5263157894736842,
|
| 123 |
+
"eval_recall_joy": 0.5947698744769875,
|
| 124 |
+
"eval_recall_neutral": 0.924990612091626,
|
| 125 |
+
"eval_recall_sadness": 0.5675555555555556,
|
| 126 |
+
"eval_recall_surprise": 0.3204301075268817,
|
| 127 |
+
"eval_runtime": 15.1427,
|
| 128 |
+
"eval_samples_per_second": 1460.377,
|
| 129 |
+
"eval_steps_per_second": 11.425,
|
| 130 |
+
"step": 464
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.3,
|
| 134 |
+
"learning_rate": 4.9940105414470535e-05,
|
| 135 |
+
"loss": 0.7998,
|
| 136 |
+
"step": 470
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.33,
|
| 140 |
+
"learning_rate": 4.938907522759943e-05,
|
| 141 |
+
"loss": 0.7996,
|
| 142 |
+
"step": 517
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.36,
|
| 146 |
+
"learning_rate": 4.882606612362243e-05,
|
| 147 |
+
"loss": 0.776,
|
| 148 |
+
"step": 564
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.4,
|
| 152 |
+
"learning_rate": 4.826305701964542e-05,
|
| 153 |
+
"loss": 0.8018,
|
| 154 |
+
"step": 611
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.43,
|
| 158 |
+
"learning_rate": 4.770004791566843e-05,
|
| 159 |
+
"loss": 0.7528,
|
| 160 |
+
"step": 658
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.45,
|
| 164 |
+
"eval_accuracy": 0.7260558921949896,
|
| 165 |
+
"eval_b_acc": 0.5878316350938373,
|
| 166 |
+
"eval_f1": 0.7255951803151626,
|
| 167 |
+
"eval_f1_anger": 0.5542226487523992,
|
| 168 |
+
"eval_f1_disgust": 0.3994082840236686,
|
| 169 |
+
"eval_f1_fear": 0.5798319327731093,
|
| 170 |
+
"eval_f1_joy": 0.7291770573566084,
|
| 171 |
+
"eval_f1_neutral": 0.8314359514015114,
|
| 172 |
+
"eval_f1_sadness": 0.6541166735622672,
|
| 173 |
+
"eval_f1_surprise": 0.38283828382838286,
|
| 174 |
+
"eval_loss": 0.7560049891471863,
|
| 175 |
+
"eval_prec": 0.7309376250219672,
|
| 176 |
+
"eval_prec_anger": 0.5550216242191254,
|
| 177 |
+
"eval_prec_disgust": 0.3409090909090909,
|
| 178 |
+
"eval_prec_fear": 0.7225130890052356,
|
| 179 |
+
"eval_prec_joy": 0.6968541468064824,
|
| 180 |
+
"eval_prec_neutral": 0.847571679344646,
|
| 181 |
+
"eval_prec_sadness": 0.6118421052631579,
|
| 182 |
+
"eval_prec_surprise": 0.49572649572649574,
|
| 183 |
+
"eval_recall": 0.7260558921949896,
|
| 184 |
+
"eval_recall_anger": 0.5534259702922856,
|
| 185 |
+
"eval_recall_disgust": 0.48214285714285715,
|
| 186 |
+
"eval_recall_fear": 0.4842105263157895,
|
| 187 |
+
"eval_recall_joy": 0.7646443514644351,
|
| 188 |
+
"eval_recall_neutral": 0.8159031167855801,
|
| 189 |
+
"eval_recall_sadness": 0.7026666666666667,
|
| 190 |
+
"eval_recall_surprise": 0.3118279569892473,
|
| 191 |
+
"eval_runtime": 14.7708,
|
| 192 |
+
"eval_samples_per_second": 1497.143,
|
| 193 |
+
"eval_steps_per_second": 11.712,
|
| 194 |
+
"step": 696
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.46,
|
| 198 |
+
"learning_rate": 4.713703881169142e-05,
|
| 199 |
+
"loss": 0.7459,
|
| 200 |
+
"step": 705
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.49,
|
| 204 |
+
"learning_rate": 4.657402970771442e-05,
|
| 205 |
+
"loss": 0.7599,
|
| 206 |
+
"step": 752
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.52,
|
| 210 |
+
"learning_rate": 4.601102060373742e-05,
|
| 211 |
+
"loss": 0.7648,
|
| 212 |
+
"step": 799
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.55,
|
| 216 |
+
"learning_rate": 4.544801149976042e-05,
|
| 217 |
+
"loss": 0.7436,
|
| 218 |
+
"step": 846
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.58,
|
| 222 |
+
"learning_rate": 4.488500239578342e-05,
|
| 223 |
+
"loss": 0.7334,
|
| 224 |
+
"step": 893
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.6,
|
| 228 |
+
"eval_accuracy": 0.7370444062584788,
|
| 229 |
+
"eval_b_acc": 0.5867600905079025,
|
| 230 |
+
"eval_f1": 0.7282740654207969,
|
| 231 |
+
"eval_f1_anger": 0.5218712029161603,
|
| 232 |
+
"eval_f1_disgust": 0.4505649717514124,
|
| 233 |
+
"eval_f1_fear": 0.6104060913705585,
|
| 234 |
+
"eval_f1_joy": 0.7311320754716981,
|
| 235 |
+
"eval_f1_neutral": 0.8351327433628318,
|
| 236 |
+
"eval_f1_sadness": 0.6498640968618731,
|
| 237 |
+
"eval_f1_surprise": 0.4181229773462783,
|
| 238 |
+
"eval_loss": 0.7309762835502625,
|
| 239 |
+
"eval_prec": 0.7344879537976949,
|
| 240 |
+
"eval_prec_anger": 0.7128630705394191,
|
| 241 |
+
"eval_prec_disgust": 0.3726635514018692,
|
| 242 |
+
"eval_prec_fear": 0.6671289875173371,
|
| 243 |
+
"eval_prec_joy": 0.7170152855993563,
|
| 244 |
+
"eval_prec_neutral": 0.7898393036491463,
|
| 245 |
+
"eval_prec_sadness": 0.7317751808569839,
|
| 246 |
+
"eval_prec_surprise": 0.5252032520325203,
|
| 247 |
+
"eval_recall": 0.7370444062584788,
|
| 248 |
+
"eval_recall_anger": 0.411595591758505,
|
| 249 |
+
"eval_recall_disgust": 0.5696428571428571,
|
| 250 |
+
"eval_recall_fear": 0.5625730994152047,
|
| 251 |
+
"eval_recall_joy": 0.74581589958159,
|
| 252 |
+
"eval_recall_neutral": 0.8859369132557267,
|
| 253 |
+
"eval_recall_sadness": 0.5844444444444444,
|
| 254 |
+
"eval_recall_surprise": 0.3473118279569892,
|
| 255 |
+
"eval_runtime": 14.5647,
|
| 256 |
+
"eval_samples_per_second": 1518.332,
|
| 257 |
+
"eval_steps_per_second": 11.878,
|
| 258 |
+
"step": 928
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.61,
|
| 262 |
+
"learning_rate": 4.432199329180642e-05,
|
| 263 |
+
"loss": 0.7493,
|
| 264 |
+
"step": 940
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.64,
|
| 268 |
+
"learning_rate": 4.375898418782942e-05,
|
| 269 |
+
"loss": 0.7194,
|
| 270 |
+
"step": 987
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.67,
|
| 274 |
+
"learning_rate": 4.319597508385242e-05,
|
| 275 |
+
"loss": 0.7158,
|
| 276 |
+
"step": 1034
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7,
|
| 280 |
+
"learning_rate": 4.263296597987542e-05,
|
| 281 |
+
"loss": 0.719,
|
| 282 |
+
"step": 1081
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.73,
|
| 286 |
+
"learning_rate": 4.206995687589842e-05,
|
| 287 |
+
"loss": 0.7216,
|
| 288 |
+
"step": 1128
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.75,
|
| 292 |
+
"eval_accuracy": 0.7448222845256399,
|
| 293 |
+
"eval_b_acc": 0.6009457685058381,
|
| 294 |
+
"eval_f1": 0.7388819166581033,
|
| 295 |
+
"eval_f1_anger": 0.5746421267893661,
|
| 296 |
+
"eval_f1_disgust": 0.48046875,
|
| 297 |
+
"eval_f1_fear": 0.6597796143250689,
|
| 298 |
+
"eval_f1_joy": 0.7266451397394499,
|
| 299 |
+
"eval_f1_neutral": 0.8355743035416118,
|
| 300 |
+
"eval_f1_sadness": 0.6644859813084112,
|
| 301 |
+
"eval_f1_surprise": 0.4711696869851729,
|
| 302 |
+
"eval_loss": 0.7043296694755554,
|
| 303 |
+
"eval_prec": 0.7402960337790491,
|
| 304 |
+
"eval_prec_anger": 0.6158904109589041,
|
| 305 |
+
"eval_prec_disgust": 0.5301724137931034,
|
| 306 |
+
"eval_prec_fear": 0.8023450586264657,
|
| 307 |
+
"eval_prec_joy": 0.7767198286122352,
|
| 308 |
+
"eval_prec_neutral": 0.7853956715678176,
|
| 309 |
+
"eval_prec_sadness": 0.7004926108374384,
|
| 310 |
+
"eval_prec_surprise": 0.48148148148148145,
|
| 311 |
+
"eval_recall": 0.7448222845256399,
|
| 312 |
+
"eval_recall_anger": 0.5385721130809775,
|
| 313 |
+
"eval_recall_disgust": 0.4392857142857143,
|
| 314 |
+
"eval_recall_fear": 0.560233918128655,
|
| 315 |
+
"eval_recall_joy": 0.6826359832635983,
|
| 316 |
+
"eval_recall_neutral": 0.8926023282012767,
|
| 317 |
+
"eval_recall_sadness": 0.632,
|
| 318 |
+
"eval_recall_surprise": 0.4612903225806452,
|
| 319 |
+
"eval_runtime": 14.6801,
|
| 320 |
+
"eval_samples_per_second": 1506.393,
|
| 321 |
+
"eval_steps_per_second": 11.785,
|
| 322 |
+
"step": 1160
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.76,
|
| 326 |
+
"learning_rate": 4.150694777192142e-05,
|
| 327 |
+
"loss": 0.7134,
|
| 328 |
+
"step": 1175
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.79,
|
| 332 |
+
"learning_rate": 4.094393866794442e-05,
|
| 333 |
+
"loss": 0.7029,
|
| 334 |
+
"step": 1222
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.82,
|
| 338 |
+
"learning_rate": 4.0380929563967415e-05,
|
| 339 |
+
"loss": 0.6945,
|
| 340 |
+
"step": 1269
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.85,
|
| 344 |
+
"learning_rate": 3.981792045999042e-05,
|
| 345 |
+
"loss": 0.7033,
|
| 346 |
+
"step": 1316
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.88,
|
| 350 |
+
"learning_rate": 3.9254911356013415e-05,
|
| 351 |
+
"loss": 0.7259,
|
| 352 |
+
"step": 1363
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9,
|
| 356 |
+
"eval_accuracy": 0.7475354978746496,
|
| 357 |
+
"eval_b_acc": 0.6082480515560814,
|
| 358 |
+
"eval_f1": 0.741168197686153,
|
| 359 |
+
"eval_f1_anger": 0.5405405405405406,
|
| 360 |
+
"eval_f1_disgust": 0.4599559147685525,
|
| 361 |
+
"eval_f1_fear": 0.6475358702432937,
|
| 362 |
+
"eval_f1_joy": 0.7468589083419156,
|
| 363 |
+
"eval_f1_neutral": 0.8421338155515371,
|
| 364 |
+
"eval_f1_sadness": 0.6752079118903124,
|
| 365 |
+
"eval_f1_surprise": 0.42070338420703385,
|
| 366 |
+
"eval_loss": 0.6962281465530396,
|
| 367 |
+
"eval_prec": 0.7432901970674257,
|
| 368 |
+
"eval_prec_anger": 0.6757728253055356,
|
| 369 |
+
"eval_prec_disgust": 0.39076154806491886,
|
| 370 |
+
"eval_prec_fear": 0.6938502673796791,
|
| 371 |
+
"eval_prec_joy": 0.7354969574036511,
|
| 372 |
+
"eval_prec_neutral": 0.8121730031391698,
|
| 373 |
+
"eval_prec_sadness": 0.683037744429286,
|
| 374 |
+
"eval_prec_surprise": 0.5493934142114385,
|
| 375 |
+
"eval_recall": 0.7475354978746496,
|
| 376 |
+
"eval_recall_anger": 0.4504072831816004,
|
| 377 |
+
"eval_recall_disgust": 0.5589285714285714,
|
| 378 |
+
"eval_recall_fear": 0.6070175438596491,
|
| 379 |
+
"eval_recall_joy": 0.7585774058577406,
|
| 380 |
+
"eval_recall_neutral": 0.8743897859556891,
|
| 381 |
+
"eval_recall_sadness": 0.6675555555555556,
|
| 382 |
+
"eval_recall_surprise": 0.34086021505376346,
|
| 383 |
+
"eval_runtime": 14.6912,
|
| 384 |
+
"eval_samples_per_second": 1505.258,
|
| 385 |
+
"eval_steps_per_second": 11.776,
|
| 386 |
+
"step": 1392
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.91,
|
| 390 |
+
"learning_rate": 3.8691902252036415e-05,
|
| 391 |
+
"loss": 0.7143,
|
| 392 |
+
"step": 1410
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.94,
|
| 396 |
+
"learning_rate": 3.812889314805942e-05,
|
| 397 |
+
"loss": 0.6903,
|
| 398 |
+
"step": 1457
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.97,
|
| 402 |
+
"learning_rate": 3.7565884044082415e-05,
|
| 403 |
+
"loss": 0.6812,
|
| 404 |
+
"step": 1504
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.0,
|
| 408 |
+
"learning_rate": 3.7002874940105415e-05,
|
| 409 |
+
"loss": 0.7182,
|
| 410 |
+
"step": 1551
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.03,
|
| 414 |
+
"learning_rate": 3.6439865836128415e-05,
|
| 415 |
+
"loss": 0.6362,
|
| 416 |
+
"step": 1598
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.05,
|
| 420 |
+
"eval_accuracy": 0.7526453830152844,
|
| 421 |
+
"eval_b_acc": 0.6055170494206373,
|
| 422 |
+
"eval_f1": 0.7483739984911967,
|
| 423 |
+
"eval_f1_anger": 0.602752507581059,
|
| 424 |
+
"eval_f1_disgust": 0.4405010438413361,
|
| 425 |
+
"eval_f1_fear": 0.6491803278688525,
|
| 426 |
+
"eval_f1_joy": 0.7437363551304709,
|
| 427 |
+
"eval_f1_neutral": 0.8435330654000731,
|
| 428 |
+
"eval_f1_sadness": 0.6785631517960602,
|
| 429 |
+
"eval_f1_surprise": 0.45454545454545453,
|
| 430 |
+
"eval_loss": 0.6771188974380493,
|
| 431 |
+
"eval_prec": 0.7472415916170276,
|
| 432 |
+
"eval_prec_anger": 0.5872727272727273,
|
| 433 |
+
"eval_prec_disgust": 0.5301507537688442,
|
| 434 |
+
"eval_prec_fear": 0.7388059701492538,
|
| 435 |
+
"eval_prec_joy": 0.739202314527795,
|
| 436 |
+
"eval_prec_neutral": 0.8213269299181786,
|
| 437 |
+
"eval_prec_sadness": 0.7089588377723971,
|
| 438 |
+
"eval_prec_surprise": 0.5300859598853869,
|
| 439 |
+
"eval_recall": 0.7526453830152844,
|
| 440 |
+
"eval_recall_anger": 0.6190704360325826,
|
| 441 |
+
"eval_recall_disgust": 0.3767857142857143,
|
| 442 |
+
"eval_recall_fear": 0.5789473684210527,
|
| 443 |
+
"eval_recall_joy": 0.748326359832636,
|
| 444 |
+
"eval_recall_neutral": 0.8669733383402178,
|
| 445 |
+
"eval_recall_sadness": 0.6506666666666666,
|
| 446 |
+
"eval_recall_surprise": 0.3978494623655914,
|
| 447 |
+
"eval_runtime": 14.8683,
|
| 448 |
+
"eval_samples_per_second": 1487.33,
|
| 449 |
+
"eval_steps_per_second": 11.636,
|
| 450 |
+
"step": 1624
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.06,
|
| 454 |
+
"learning_rate": 3.5876856732151415e-05,
|
| 455 |
+
"loss": 0.6432,
|
| 456 |
+
"step": 1645
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 1.09,
|
| 460 |
+
"learning_rate": 3.5313847628174415e-05,
|
| 461 |
+
"loss": 0.652,
|
| 462 |
+
"step": 1692
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.12,
|
| 466 |
+
"learning_rate": 3.4750838524197415e-05,
|
| 467 |
+
"loss": 0.6474,
|
| 468 |
+
"step": 1739
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.16,
|
| 472 |
+
"learning_rate": 3.4187829420220415e-05,
|
| 473 |
+
"loss": 0.6364,
|
| 474 |
+
"step": 1786
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.19,
|
| 478 |
+
"learning_rate": 3.362482031624341e-05,
|
| 479 |
+
"loss": 0.621,
|
| 480 |
+
"step": 1833
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.2,
|
| 484 |
+
"eval_accuracy": 0.7528262639052185,
|
| 485 |
+
"eval_b_acc": 0.6120366345893286,
|
| 486 |
+
"eval_f1": 0.7487165396015127,
|
| 487 |
+
"eval_f1_anger": 0.6196334421290485,
|
| 488 |
+
"eval_f1_disgust": 0.47378277153558057,
|
| 489 |
+
"eval_f1_fear": 0.6528640441683919,
|
| 490 |
+
"eval_f1_joy": 0.7450472872075661,
|
| 491 |
+
"eval_f1_neutral": 0.8372220697227559,
|
| 492 |
+
"eval_f1_sadness": 0.6809026935209901,
|
| 493 |
+
"eval_f1_surprise": 0.4612676056338028,
|
| 494 |
+
"eval_loss": 0.67792809009552,
|
| 495 |
+
"eval_prec": 0.749407175516463,
|
| 496 |
+
"eval_prec_anger": 0.6508438818565401,
|
| 497 |
+
"eval_prec_disgust": 0.49803149606299213,
|
| 498 |
+
"eval_prec_fear": 0.7962962962962963,
|
| 499 |
+
"eval_prec_joy": 0.7107312440645774,
|
| 500 |
+
"eval_prec_neutral": 0.8165268606103873,
|
| 501 |
+
"eval_prec_sadness": 0.7498663816141101,
|
| 502 |
+
"eval_prec_surprise": 0.5077519379844961,
|
| 503 |
+
"eval_recall": 0.7528262639052185,
|
| 504 |
+
"eval_recall_anger": 0.5912793483469094,
|
| 505 |
+
"eval_recall_disgust": 0.4517857142857143,
|
| 506 |
+
"eval_recall_fear": 0.5532163742690058,
|
| 507 |
+
"eval_recall_joy": 0.7828451882845189,
|
| 508 |
+
"eval_recall_neutral": 0.8589936162223056,
|
| 509 |
+
"eval_recall_sadness": 0.6235555555555555,
|
| 510 |
+
"eval_recall_surprise": 0.42258064516129035,
|
| 511 |
+
"eval_runtime": 14.7115,
|
| 512 |
+
"eval_samples_per_second": 1503.179,
|
| 513 |
+
"eval_steps_per_second": 11.76,
|
| 514 |
+
"step": 1856
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.22,
|
| 518 |
+
"learning_rate": 3.3061811212266415e-05,
|
| 519 |
+
"loss": 0.6195,
|
| 520 |
+
"step": 1880
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.25,
|
| 524 |
+
"learning_rate": 3.249880210828941e-05,
|
| 525 |
+
"loss": 0.6332,
|
| 526 |
+
"step": 1927
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.28,
|
| 530 |
+
"learning_rate": 3.1935793004312415e-05,
|
| 531 |
+
"loss": 0.6251,
|
| 532 |
+
"step": 1974
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.31,
|
| 536 |
+
"learning_rate": 3.1372783900335415e-05,
|
| 537 |
+
"loss": 0.63,
|
| 538 |
+
"step": 2021
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.34,
|
| 542 |
+
"learning_rate": 3.080977479635841e-05,
|
| 543 |
+
"loss": 0.6241,
|
| 544 |
+
"step": 2068
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.35,
|
| 548 |
+
"eval_accuracy": 0.7512887763407796,
|
| 549 |
+
"eval_b_acc": 0.6367091163081916,
|
| 550 |
+
"eval_f1": 0.7514440246825209,
|
| 551 |
+
"eval_f1_anger": 0.6141022752872269,
|
| 552 |
+
"eval_f1_disgust": 0.4511668107173725,
|
| 553 |
+
"eval_f1_fear": 0.6488060570762958,
|
| 554 |
+
"eval_f1_joy": 0.7509570615623384,
|
| 555 |
+
"eval_f1_neutral": 0.843150231634679,
|
| 556 |
+
"eval_f1_sadness": 0.6745293717396236,
|
| 557 |
+
"eval_f1_surprise": 0.4730290456431535,
|
| 558 |
+
"eval_loss": 0.6849319338798523,
|
| 559 |
+
"eval_prec": 0.7525985800729426,
|
| 560 |
+
"eval_prec_anger": 0.5795068027210885,
|
| 561 |
+
"eval_prec_disgust": 0.4371859296482412,
|
| 562 |
+
"eval_prec_fear": 0.6461716937354989,
|
| 563 |
+
"eval_prec_joy": 0.7428863868986694,
|
| 564 |
+
"eval_prec_neutral": 0.849171586364502,
|
| 565 |
+
"eval_prec_sadness": 0.6887447892542844,
|
| 566 |
+
"eval_prec_surprise": 0.5270805812417437,
|
| 567 |
+
"eval_recall": 0.7512887763407796,
|
| 568 |
+
"eval_recall_anger": 0.6530905606133206,
|
| 569 |
+
"eval_recall_disgust": 0.4660714285714286,
|
| 570 |
+
"eval_recall_fear": 0.6514619883040935,
|
| 571 |
+
"eval_recall_joy": 0.7592050209205021,
|
| 572 |
+
"eval_recall_neutral": 0.8372136687945926,
|
| 573 |
+
"eval_recall_sadness": 0.6608888888888889,
|
| 574 |
+
"eval_recall_surprise": 0.4290322580645161,
|
| 575 |
+
"eval_runtime": 14.591,
|
| 576 |
+
"eval_samples_per_second": 1515.587,
|
| 577 |
+
"eval_steps_per_second": 11.857,
|
| 578 |
+
"step": 2088
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.37,
|
| 582 |
+
"learning_rate": 3.024676569238141e-05,
|
| 583 |
+
"loss": 0.5997,
|
| 584 |
+
"step": 2115
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.4,
|
| 588 |
+
"learning_rate": 2.968375658840441e-05,
|
| 589 |
+
"loss": 0.6348,
|
| 590 |
+
"step": 2162
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.43,
|
| 594 |
+
"learning_rate": 2.9120747484427408e-05,
|
| 595 |
+
"loss": 0.6269,
|
| 596 |
+
"step": 2209
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.46,
|
| 600 |
+
"learning_rate": 2.855773838045041e-05,
|
| 601 |
+
"loss": 0.6411,
|
| 602 |
+
"step": 2256
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.49,
|
| 606 |
+
"learning_rate": 2.7994729276473408e-05,
|
| 607 |
+
"loss": 0.6188,
|
| 608 |
+
"step": 2303
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.5,
|
| 612 |
+
"eval_accuracy": 0.7578909288233698,
|
| 613 |
+
"eval_b_acc": 0.6158634091154939,
|
| 614 |
+
"eval_f1": 0.7533929534836733,
|
| 615 |
+
"eval_f1_anger": 0.6305133767172812,
|
| 616 |
+
"eval_f1_disgust": 0.45919477693144717,
|
| 617 |
+
"eval_f1_fear": 0.6729323308270676,
|
| 618 |
+
"eval_f1_joy": 0.7494099134539732,
|
| 619 |
+
"eval_f1_neutral": 0.84280126653008,
|
| 620 |
+
"eval_f1_sadness": 0.681602172437203,
|
| 621 |
+
"eval_f1_surprise": 0.4503664223850767,
|
| 622 |
+
"eval_loss": 0.6713131070137024,
|
| 623 |
+
"eval_prec": 0.7538941653702528,
|
| 624 |
+
"eval_prec_anger": 0.6343355965082444,
|
| 625 |
+
"eval_prec_disgust": 0.5877437325905293,
|
| 626 |
+
"eval_prec_fear": 0.7246963562753036,
|
| 627 |
+
"eval_prec_joy": 0.7071269487750557,
|
| 628 |
+
"eval_prec_neutral": 0.8361049519586105,
|
| 629 |
+
"eval_prec_sadness": 0.69432918395574,
|
| 630 |
+
"eval_prec_surprise": 0.5919439579684763,
|
| 631 |
+
"eval_recall": 0.7578909288233698,
|
| 632 |
+
"eval_recall_anger": 0.6267369429803545,
|
| 633 |
+
"eval_recall_disgust": 0.3767857142857143,
|
| 634 |
+
"eval_recall_fear": 0.6280701754385964,
|
| 635 |
+
"eval_recall_joy": 0.797071129707113,
|
| 636 |
+
"eval_recall_neutral": 0.8496057078482914,
|
| 637 |
+
"eval_recall_sadness": 0.6693333333333333,
|
| 638 |
+
"eval_recall_surprise": 0.3634408602150538,
|
| 639 |
+
"eval_runtime": 14.588,
|
| 640 |
+
"eval_samples_per_second": 1515.9,
|
| 641 |
+
"eval_steps_per_second": 11.859,
|
| 642 |
+
"step": 2320
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.52,
|
| 646 |
+
"learning_rate": 2.7431720172496405e-05,
|
| 647 |
+
"loss": 0.622,
|
| 648 |
+
"step": 2350
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.55,
|
| 652 |
+
"learning_rate": 2.6868711068519408e-05,
|
| 653 |
+
"loss": 0.6278,
|
| 654 |
+
"step": 2397
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.58,
|
| 658 |
+
"learning_rate": 2.6305701964542405e-05,
|
| 659 |
+
"loss": 0.6227,
|
| 660 |
+
"step": 2444
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.61,
|
| 664 |
+
"learning_rate": 2.5742692860565405e-05,
|
| 665 |
+
"loss": 0.6337,
|
| 666 |
+
"step": 2491
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 1.64,
|
| 670 |
+
"learning_rate": 2.5179683756588408e-05,
|
| 671 |
+
"loss": 0.6182,
|
| 672 |
+
"step": 2538
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.65,
|
| 676 |
+
"eval_accuracy": 0.760106719725061,
|
| 677 |
+
"eval_b_acc": 0.6198832054514857,
|
| 678 |
+
"eval_f1": 0.7565733944354605,
|
| 679 |
+
"eval_f1_anger": 0.625943396226415,
|
| 680 |
+
"eval_f1_disgust": 0.4582893347412883,
|
| 681 |
+
"eval_f1_fear": 0.6753246753246753,
|
| 682 |
+
"eval_f1_joy": 0.7498260610277309,
|
| 683 |
+
"eval_f1_neutral": 0.8475095074668398,
|
| 684 |
+
"eval_f1_sadness": 0.6851194700733381,
|
| 685 |
+
"eval_f1_surprise": 0.4700181708055723,
|
| 686 |
+
"eval_loss": 0.6608244180679321,
|
| 687 |
+
"eval_prec": 0.7566647637236428,
|
| 688 |
+
"eval_prec_anger": 0.6163492800743149,
|
| 689 |
+
"eval_prec_disgust": 0.5607235142118863,
|
| 690 |
+
"eval_prec_fear": 0.7591240875912408,
|
| 691 |
+
"eval_prec_joy": 0.7142586631319826,
|
| 692 |
+
"eval_prec_neutral": 0.8374885426214482,
|
| 693 |
+
"eval_prec_sadness": 0.7324228629236217,
|
| 694 |
+
"eval_prec_surprise": 0.5381414701803051,
|
| 695 |
+
"eval_recall": 0.760106719725061,
|
| 696 |
+
"eval_recall_anger": 0.6358409199808337,
|
| 697 |
+
"eval_recall_disgust": 0.3875,
|
| 698 |
+
"eval_recall_fear": 0.6081871345029239,
|
| 699 |
+
"eval_recall_joy": 0.7891213389121339,
|
| 700 |
+
"eval_recall_neutral": 0.8577731881336839,
|
| 701 |
+
"eval_recall_sadness": 0.6435555555555555,
|
| 702 |
+
"eval_recall_surprise": 0.4172043010752688,
|
| 703 |
+
"eval_runtime": 14.5162,
|
| 704 |
+
"eval_samples_per_second": 1523.406,
|
| 705 |
+
"eval_steps_per_second": 11.918,
|
| 706 |
+
"step": 2552
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.67,
|
| 710 |
+
"learning_rate": 2.4616674652611405e-05,
|
| 711 |
+
"loss": 0.6078,
|
| 712 |
+
"step": 2585
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.7,
|
| 716 |
+
"learning_rate": 2.4053665548634405e-05,
|
| 717 |
+
"loss": 0.6167,
|
| 718 |
+
"step": 2632
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.73,
|
| 722 |
+
"learning_rate": 2.3490656444657405e-05,
|
| 723 |
+
"loss": 0.6329,
|
| 724 |
+
"step": 2679
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.76,
|
| 728 |
+
"learning_rate": 2.29396262577863e-05,
|
| 729 |
+
"loss": 0.6153,
|
| 730 |
+
"step": 2726
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.79,
|
| 734 |
+
"learning_rate": 2.2376617153809296e-05,
|
| 735 |
+
"loss": 0.6392,
|
| 736 |
+
"step": 2773
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 1.8,
|
| 740 |
+
"eval_accuracy": 0.7623677308492358,
|
| 741 |
+
"eval_b_acc": 0.626106698297278,
|
| 742 |
+
"eval_f1": 0.7595959808260688,
|
| 743 |
+
"eval_f1_anger": 0.6317689530685919,
|
| 744 |
+
"eval_f1_disgust": 0.46170212765957447,
|
| 745 |
+
"eval_f1_fear": 0.6752246469833119,
|
| 746 |
+
"eval_f1_joy": 0.7548152004164499,
|
| 747 |
+
"eval_f1_neutral": 0.8505247584261871,
|
| 748 |
+
"eval_f1_sadness": 0.6872427983539094,
|
| 749 |
+
"eval_f1_surprise": 0.4615384615384616,
|
| 750 |
+
"eval_loss": 0.6541842818260193,
|
| 751 |
+
"eval_prec": 0.7593266897219236,
|
| 752 |
+
"eval_prec_anger": 0.5970149253731343,
|
| 753 |
+
"eval_prec_disgust": 0.5710526315789474,
|
| 754 |
+
"eval_prec_fear": 0.748221906116643,
|
| 755 |
+
"eval_prec_joy": 0.7512953367875648,
|
| 756 |
+
"eval_prec_neutral": 0.8379338617108499,
|
| 757 |
+
"eval_prec_sadness": 0.7076271186440678,
|
| 758 |
+
"eval_prec_surprise": 0.5131578947368421,
|
| 759 |
+
"eval_recall": 0.7623677308492358,
|
| 760 |
+
"eval_recall_anger": 0.6708193579300431,
|
| 761 |
+
"eval_recall_disgust": 0.3875,
|
| 762 |
+
"eval_recall_fear": 0.6152046783625731,
|
| 763 |
+
"eval_recall_joy": 0.7583682008368201,
|
| 764 |
+
"eval_recall_neutral": 0.8634998122418325,
|
| 765 |
+
"eval_recall_sadness": 0.668,
|
| 766 |
+
"eval_recall_surprise": 0.41935483870967744,
|
| 767 |
+
"eval_runtime": 14.6661,
|
| 768 |
+
"eval_samples_per_second": 1507.827,
|
| 769 |
+
"eval_steps_per_second": 11.796,
|
| 770 |
+
"step": 2784
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.82,
|
| 774 |
+
"learning_rate": 2.1813608049832296e-05,
|
| 775 |
+
"loss": 0.6221,
|
| 776 |
+
"step": 2820
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.85,
|
| 780 |
+
"learning_rate": 2.1250598945855296e-05,
|
| 781 |
+
"loss": 0.6109,
|
| 782 |
+
"step": 2867
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.88,
|
| 786 |
+
"learning_rate": 2.0687589841878296e-05,
|
| 787 |
+
"loss": 0.6029,
|
| 788 |
+
"step": 2914
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.92,
|
| 792 |
+
"learning_rate": 2.0124580737901292e-05,
|
| 793 |
+
"loss": 0.6243,
|
| 794 |
+
"step": 2961
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.95,
|
| 798 |
+
"learning_rate": 1.9561571633924296e-05,
|
| 799 |
+
"loss": 0.6158,
|
| 800 |
+
"step": 3008
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.95,
|
| 804 |
+
"eval_accuracy": 0.764854843085828,
|
| 805 |
+
"eval_b_acc": 0.6279403178124943,
|
| 806 |
+
"eval_f1": 0.761380516026678,
|
| 807 |
+
"eval_f1_anger": 0.6389822371579453,
|
| 808 |
+
"eval_f1_disgust": 0.4711729622266402,
|
| 809 |
+
"eval_f1_fear": 0.6765847347994824,
|
| 810 |
+
"eval_f1_joy": 0.751901635927894,
|
| 811 |
+
"eval_f1_neutral": 0.8525969275786394,
|
| 812 |
+
"eval_f1_sadness": 0.6901180282342051,
|
| 813 |
+
"eval_f1_surprise": 0.46511627906976744,
|
| 814 |
+
"eval_loss": 0.6456097960472107,
|
| 815 |
+
"eval_prec": 0.7598739530313577,
|
| 816 |
+
"eval_prec_anger": 0.6402116402116402,
|
| 817 |
+
"eval_prec_disgust": 0.531390134529148,
|
| 818 |
+
"eval_prec_fear": 0.7568740955137482,
|
| 819 |
+
"eval_prec_joy": 0.7490139090720366,
|
| 820 |
+
"eval_prec_neutral": 0.8310160427807487,
|
| 821 |
+
"eval_prec_sadness": 0.7199420569773056,
|
| 822 |
+
"eval_prec_surprise": 0.5063291139240507,
|
| 823 |
+
"eval_recall": 0.764854843085828,
|
| 824 |
+
"eval_recall_anger": 0.6377575467177767,
|
| 825 |
+
"eval_recall_disgust": 0.4232142857142857,
|
| 826 |
+
"eval_recall_fear": 0.6116959064327485,
|
| 827 |
+
"eval_recall_joy": 0.7548117154811715,
|
| 828 |
+
"eval_recall_neutral": 0.8753285767930905,
|
| 829 |
+
"eval_recall_sadness": 0.6626666666666666,
|
| 830 |
+
"eval_recall_surprise": 0.43010752688172044,
|
| 831 |
+
"eval_runtime": 14.622,
|
| 832 |
+
"eval_samples_per_second": 1512.381,
|
| 833 |
+
"eval_steps_per_second": 11.832,
|
| 834 |
+
"step": 3016
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 1.98,
|
| 838 |
+
"learning_rate": 1.8998562529947296e-05,
|
| 839 |
+
"loss": 0.5901,
|
| 840 |
+
"step": 3055
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.01,
|
| 844 |
+
"learning_rate": 1.8435553425970292e-05,
|
| 845 |
+
"loss": 0.5923,
|
| 846 |
+
"step": 3102
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.04,
|
| 850 |
+
"learning_rate": 1.7872544321993292e-05,
|
| 851 |
+
"loss": 0.5239,
|
| 852 |
+
"step": 3149
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.07,
|
| 856 |
+
"learning_rate": 1.7309535218016292e-05,
|
| 857 |
+
"loss": 0.5329,
|
| 858 |
+
"step": 3196
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.1,
|
| 862 |
+
"learning_rate": 1.6758505031145187e-05,
|
| 863 |
+
"loss": 0.554,
|
| 864 |
+
"step": 3243
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.1,
|
| 868 |
+
"eval_accuracy": 0.7583883512706883,
|
| 869 |
+
"eval_b_acc": 0.634571558597101,
|
| 870 |
+
"eval_f1": 0.7563700086517053,
|
| 871 |
+
"eval_f1_anger": 0.643182362808531,
|
| 872 |
+
"eval_f1_disgust": 0.4690181124880839,
|
| 873 |
+
"eval_f1_fear": 0.6795752654590881,
|
| 874 |
+
"eval_f1_joy": 0.7506092607636069,
|
| 875 |
+
"eval_f1_neutral": 0.842095468502838,
|
| 876 |
+
"eval_f1_sadness": 0.6862699699004399,
|
| 877 |
+
"eval_f1_surprise": 0.47133027522935783,
|
| 878 |
+
"eval_loss": 0.6742041707038879,
|
| 879 |
+
"eval_prec": 0.7554687165264222,
|
| 880 |
+
"eval_prec_anger": 0.6433365292425696,
|
| 881 |
+
"eval_prec_disgust": 0.5030674846625767,
|
| 882 |
+
"eval_prec_fear": 0.7292225201072386,
|
| 883 |
+
"eval_prec_joy": 0.7292817679558011,
|
| 884 |
+
"eval_prec_neutral": 0.8347168419110865,
|
| 885 |
+
"eval_prec_sadness": 0.7162880618656355,
|
| 886 |
+
"eval_prec_surprise": 0.504914004914005,
|
| 887 |
+
"eval_recall": 0.7583883512706883,
|
| 888 |
+
"eval_recall_anger": 0.6430282702443699,
|
| 889 |
+
"eval_recall_disgust": 0.4392857142857143,
|
| 890 |
+
"eval_recall_fear": 0.6362573099415205,
|
| 891 |
+
"eval_recall_joy": 0.7732217573221757,
|
| 892 |
+
"eval_recall_neutral": 0.8496057078482914,
|
| 893 |
+
"eval_recall_sadness": 0.6586666666666666,
|
| 894 |
+
"eval_recall_surprise": 0.44193548387096776,
|
| 895 |
+
"eval_runtime": 14.8261,
|
| 896 |
+
"eval_samples_per_second": 1491.559,
|
| 897 |
+
"eval_steps_per_second": 11.669,
|
| 898 |
+
"step": 3248
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.13,
|
| 902 |
+
"learning_rate": 1.6195495927168187e-05,
|
| 903 |
+
"loss": 0.5568,
|
| 904 |
+
"step": 3290
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 2.16,
|
| 908 |
+
"learning_rate": 1.5632486823191184e-05,
|
| 909 |
+
"loss": 0.5375,
|
| 910 |
+
"step": 3337
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 2.19,
|
| 914 |
+
"learning_rate": 1.5069477719214184e-05,
|
| 915 |
+
"loss": 0.5198,
|
| 916 |
+
"step": 3384
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.22,
|
| 920 |
+
"learning_rate": 1.4506468615237184e-05,
|
| 921 |
+
"loss": 0.5255,
|
| 922 |
+
"step": 3431
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 2.25,
|
| 926 |
+
"learning_rate": 1.3943459511260182e-05,
|
| 927 |
+
"loss": 0.5537,
|
| 928 |
+
"step": 3478
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 2.25,
|
| 932 |
+
"eval_accuracy": 0.7632721352989057,
|
| 933 |
+
"eval_b_acc": 0.6282518018397625,
|
| 934 |
+
"eval_f1": 0.7604673635172584,
|
| 935 |
+
"eval_f1_anger": 0.6454630495790459,
|
| 936 |
+
"eval_f1_disgust": 0.45434782608695656,
|
| 937 |
+
"eval_f1_fear": 0.6718246292714378,
|
| 938 |
+
"eval_f1_joy": 0.7522695178535405,
|
| 939 |
+
"eval_f1_neutral": 0.8484961703717542,
|
| 940 |
+
"eval_f1_sadness": 0.6950892857142857,
|
| 941 |
+
"eval_f1_surprise": 0.47641791044776116,
|
| 942 |
+
"eval_loss": 0.670791745185852,
|
| 943 |
+
"eval_prec": 0.7603587760472442,
|
| 944 |
+
"eval_prec_anger": 0.6304248515303792,
|
| 945 |
+
"eval_prec_disgust": 0.5805555555555556,
|
| 946 |
+
"eval_prec_fear": 0.7485632183908046,
|
| 947 |
+
"eval_prec_joy": 0.726334242306194,
|
| 948 |
+
"eval_prec_neutral": 0.8442379182156133,
|
| 949 |
+
"eval_prec_sadness": 0.6982062780269058,
|
| 950 |
+
"eval_prec_surprise": 0.5355704697986577,
|
| 951 |
+
"eval_recall": 0.7632721352989057,
|
| 952 |
+
"eval_recall_anger": 0.6612362242453282,
|
| 953 |
+
"eval_recall_disgust": 0.3732142857142857,
|
| 954 |
+
"eval_recall_fear": 0.6093567251461989,
|
| 955 |
+
"eval_recall_joy": 0.7801255230125523,
|
| 956 |
+
"eval_recall_neutral": 0.8527975966954563,
|
| 957 |
+
"eval_recall_sadness": 0.692,
|
| 958 |
+
"eval_recall_surprise": 0.4290322580645161,
|
| 959 |
+
"eval_runtime": 14.5517,
|
| 960 |
+
"eval_samples_per_second": 1519.68,
|
| 961 |
+
"eval_steps_per_second": 11.889,
|
| 962 |
+
"step": 3480
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.28,
|
| 966 |
+
"learning_rate": 1.3380450407283182e-05,
|
| 967 |
+
"loss": 0.5305,
|
| 968 |
+
"step": 3525
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.31,
|
| 972 |
+
"learning_rate": 1.2817441303306182e-05,
|
| 973 |
+
"loss": 0.5313,
|
| 974 |
+
"step": 3572
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.34,
|
| 978 |
+
"learning_rate": 1.2254432199329182e-05,
|
| 979 |
+
"loss": 0.5182,
|
| 980 |
+
"step": 3619
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.37,
|
| 984 |
+
"learning_rate": 1.169142309535218e-05,
|
| 985 |
+
"loss": 0.5375,
|
| 986 |
+
"step": 3666
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.4,
|
| 990 |
+
"eval_accuracy": 0.7605589219498959,
|
| 991 |
+
"eval_b_acc": 0.6402110359246711,
|
| 992 |
+
"eval_f1": 0.7594675823716318,
|
| 993 |
+
"eval_f1_anger": 0.6421420256111757,
|
| 994 |
+
"eval_f1_disgust": 0.4648148148148148,
|
| 995 |
+
"eval_f1_fear": 0.6746532156368222,
|
| 996 |
+
"eval_f1_joy": 0.7537328697075065,
|
| 997 |
+
"eval_f1_neutral": 0.8463990997749438,
|
| 998 |
+
"eval_f1_sadness": 0.6885171790235082,
|
| 999 |
+
"eval_f1_surprise": 0.4835924006908463,
|
| 1000 |
+
"eval_loss": 0.671157717704773,
|
| 1001 |
+
"eval_prec": 0.7592351302120464,
|
| 1002 |
+
"eval_prec_anger": 0.6245471014492754,
|
| 1003 |
+
"eval_prec_disgust": 0.4826923076923077,
|
| 1004 |
+
"eval_prec_fear": 0.7318741450068399,
|
| 1005 |
+
"eval_prec_joy": 0.7372949179671868,
|
| 1006 |
+
"eval_prec_neutral": 0.8454477332334207,
|
| 1007 |
+
"eval_prec_sadness": 0.7005519779208832,
|
| 1008 |
+
"eval_prec_surprise": 0.5204460966542751,
|
| 1009 |
+
"eval_recall": 0.7605589219498959,
|
| 1010 |
+
"eval_recall_anger": 0.6607570675610924,
|
| 1011 |
+
"eval_recall_disgust": 0.44821428571428573,
|
| 1012 |
+
"eval_recall_fear": 0.6257309941520468,
|
| 1013 |
+
"eval_recall_joy": 0.7709205020920502,
|
| 1014 |
+
"eval_recall_neutral": 0.847352609838528,
|
| 1015 |
+
"eval_recall_sadness": 0.6768888888888889,
|
| 1016 |
+
"eval_recall_surprise": 0.45161290322580644,
|
| 1017 |
+
"eval_runtime": 14.5753,
|
| 1018 |
+
"eval_samples_per_second": 1517.225,
|
| 1019 |
+
"eval_steps_per_second": 11.869,
|
| 1020 |
+
"step": 3712
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 2.4,
|
| 1024 |
+
"learning_rate": 1.112841399137518e-05,
|
| 1025 |
+
"loss": 0.5266,
|
| 1026 |
+
"step": 3713
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.43,
|
| 1030 |
+
"learning_rate": 1.056540488739818e-05,
|
| 1031 |
+
"loss": 0.5276,
|
| 1032 |
+
"step": 3760
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.46,
|
| 1036 |
+
"learning_rate": 1.000239578342118e-05,
|
| 1037 |
+
"loss": 0.5432,
|
| 1038 |
+
"step": 3807
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.49,
|
| 1042 |
+
"learning_rate": 9.439386679444178e-06,
|
| 1043 |
+
"loss": 0.5262,
|
| 1044 |
+
"step": 3854
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 2.52,
|
| 1048 |
+
"learning_rate": 8.876377575467178e-06,
|
| 1049 |
+
"loss": 0.5175,
|
| 1050 |
+
"step": 3901
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 2.55,
|
| 1054 |
+
"eval_accuracy": 0.7625033915166862,
|
| 1055 |
+
"eval_b_acc": 0.6369042077487537,
|
| 1056 |
+
"eval_f1": 0.7603749093215698,
|
| 1057 |
+
"eval_f1_anger": 0.6429077177248053,
|
| 1058 |
+
"eval_f1_disgust": 0.46938775510204084,
|
| 1059 |
+
"eval_f1_fear": 0.677336747759283,
|
| 1060 |
+
"eval_f1_joy": 0.7530406101834674,
|
| 1061 |
+
"eval_f1_neutral": 0.8487682904241526,
|
| 1062 |
+
"eval_f1_sadness": 0.6956521739130433,
|
| 1063 |
+
"eval_f1_surprise": 0.457388105456775,
|
| 1064 |
+
"eval_loss": 0.66249018907547,
|
| 1065 |
+
"eval_prec": 0.7600295878449174,
|
| 1066 |
+
"eval_prec_anger": 0.6334883720930232,
|
| 1067 |
+
"eval_prec_disgust": 0.44805194805194803,
|
| 1068 |
+
"eval_prec_fear": 0.7482319660537482,
|
| 1069 |
+
"eval_prec_joy": 0.7421779764323446,
|
| 1070 |
+
"eval_prec_neutral": 0.8374451754385965,
|
| 1071 |
+
"eval_prec_sadness": 0.7251687560270009,
|
| 1072 |
+
"eval_prec_surprise": 0.5320970042796006,
|
| 1073 |
+
"eval_recall": 0.7625033915166862,
|
| 1074 |
+
"eval_recall_anger": 0.6526114039290848,
|
| 1075 |
+
"eval_recall_disgust": 0.4928571428571429,
|
| 1076 |
+
"eval_recall_fear": 0.6187134502923977,
|
| 1077 |
+
"eval_recall_joy": 0.7642259414225941,
|
| 1078 |
+
"eval_recall_neutral": 0.8604018024784078,
|
| 1079 |
+
"eval_recall_sadness": 0.6684444444444444,
|
| 1080 |
+
"eval_recall_surprise": 0.4010752688172043,
|
| 1081 |
+
"eval_runtime": 14.6386,
|
| 1082 |
+
"eval_samples_per_second": 1510.666,
|
| 1083 |
+
"eval_steps_per_second": 11.818,
|
| 1084 |
+
"step": 3944
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.55,
|
| 1088 |
+
"learning_rate": 8.313368471490178e-06,
|
| 1089 |
+
"loss": 0.5337,
|
| 1090 |
+
"step": 3948
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.58,
|
| 1094 |
+
"learning_rate": 7.750359367513177e-06,
|
| 1095 |
+
"loss": 0.5148,
|
| 1096 |
+
"step": 3995
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 2.61,
|
| 1100 |
+
"learning_rate": 7.1873502635361775e-06,
|
| 1101 |
+
"loss": 0.5234,
|
| 1102 |
+
"step": 4042
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.64,
|
| 1106 |
+
"learning_rate": 6.624341159559177e-06,
|
| 1107 |
+
"loss": 0.5286,
|
| 1108 |
+
"step": 4089
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.68,
|
| 1112 |
+
"learning_rate": 6.061332055582176e-06,
|
| 1113 |
+
"loss": 0.5182,
|
| 1114 |
+
"step": 4136
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 2.7,
|
| 1118 |
+
"eval_accuracy": 0.7631364746314552,
|
| 1119 |
+
"eval_b_acc": 0.6403726819328496,
|
| 1120 |
+
"eval_f1": 0.7611758530200116,
|
| 1121 |
+
"eval_f1_anger": 0.6441332689521971,
|
| 1122 |
+
"eval_f1_disgust": 0.4635036496350365,
|
| 1123 |
+
"eval_f1_fear": 0.6658725431804646,
|
| 1124 |
+
"eval_f1_joy": 0.7548551093035079,
|
| 1125 |
+
"eval_f1_neutral": 0.848629209284947,
|
| 1126 |
+
"eval_f1_sadness": 0.6971687429218574,
|
| 1127 |
+
"eval_f1_surprise": 0.4763636363636364,
|
| 1128 |
+
"eval_loss": 0.6621036529541016,
|
| 1129 |
+
"eval_prec": 0.7602393679680954,
|
| 1130 |
+
"eval_prec_anger": 0.6491484184914842,
|
| 1131 |
+
"eval_prec_disgust": 0.47388059701492535,
|
| 1132 |
+
"eval_prec_fear": 0.6783980582524272,
|
| 1133 |
+
"eval_prec_joy": 0.734322453016815,
|
| 1134 |
+
"eval_prec_neutral": 0.844409331722279,
|
| 1135 |
+
"eval_prec_sadness": 0.7108545034642032,
|
| 1136 |
+
"eval_prec_surprise": 0.5458333333333333,
|
| 1137 |
+
"eval_recall": 0.7631364746314552,
|
| 1138 |
+
"eval_recall_anger": 0.639195016770484,
|
| 1139 |
+
"eval_recall_disgust": 0.45357142857142857,
|
| 1140 |
+
"eval_recall_fear": 0.6538011695906433,
|
| 1141 |
+
"eval_recall_joy": 0.7765690376569038,
|
| 1142 |
+
"eval_recall_neutral": 0.8528914757791964,
|
| 1143 |
+
"eval_recall_sadness": 0.684,
|
| 1144 |
+
"eval_recall_surprise": 0.42258064516129035,
|
| 1145 |
+
"eval_runtime": 14.5758,
|
| 1146 |
+
"eval_samples_per_second": 1517.169,
|
| 1147 |
+
"eval_steps_per_second": 11.869,
|
| 1148 |
+
"step": 4176
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.71,
|
| 1152 |
+
"learning_rate": 5.498322951605176e-06,
|
| 1153 |
+
"loss": 0.5301,
|
| 1154 |
+
"step": 4183
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.74,
|
| 1158 |
+
"learning_rate": 4.935313847628174e-06,
|
| 1159 |
+
"loss": 0.5402,
|
| 1160 |
+
"step": 4230
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.77,
|
| 1164 |
+
"learning_rate": 4.372304743651174e-06,
|
| 1165 |
+
"loss": 0.5167,
|
| 1166 |
+
"step": 4277
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.8,
|
| 1170 |
+
"learning_rate": 3.8092956396741735e-06,
|
| 1171 |
+
"loss": 0.5004,
|
| 1172 |
+
"step": 4324
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.83,
|
| 1176 |
+
"learning_rate": 3.246286535697173e-06,
|
| 1177 |
+
"loss": 0.5148,
|
| 1178 |
+
"step": 4371
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.85,
|
| 1182 |
+
"eval_accuracy": 0.7637243375237406,
|
| 1183 |
+
"eval_b_acc": 0.6382943967754277,
|
| 1184 |
+
"eval_f1": 0.7612453985473553,
|
| 1185 |
+
"eval_f1_anger": 0.6462075848303392,
|
| 1186 |
+
"eval_f1_disgust": 0.4666666666666667,
|
| 1187 |
+
"eval_f1_fear": 0.6678592725104352,
|
| 1188 |
+
"eval_f1_joy": 0.7563368765331154,
|
| 1189 |
+
"eval_f1_neutral": 0.8485270885605427,
|
| 1190 |
+
"eval_f1_sadness": 0.6918630502445532,
|
| 1191 |
+
"eval_f1_surprise": 0.4760213143872114,
|
| 1192 |
+
"eval_loss": 0.6637689471244812,
|
| 1193 |
+
"eval_prec": 0.7597715932015663,
|
| 1194 |
+
"eval_prec_anger": 0.6741280583029672,
|
| 1195 |
+
"eval_prec_disgust": 0.5,
|
| 1196 |
+
"eval_prec_fear": 0.681265206812652,
|
| 1197 |
+
"eval_prec_joy": 0.7394084732214229,
|
| 1198 |
+
"eval_prec_neutral": 0.8400183992640294,
|
| 1199 |
+
"eval_prec_sadness": 0.6921708185053381,
|
| 1200 |
+
"eval_prec_surprise": 0.5296442687747036,
|
| 1201 |
+
"eval_recall": 0.7637243375237406,
|
| 1202 |
+
"eval_recall_anger": 0.6205079060852899,
|
| 1203 |
+
"eval_recall_disgust": 0.4375,
|
| 1204 |
+
"eval_recall_fear": 0.6549707602339181,
|
| 1205 |
+
"eval_recall_joy": 0.7740585774058577,
|
| 1206 |
+
"eval_recall_neutral": 0.857209913631243,
|
| 1207 |
+
"eval_recall_sadness": 0.6915555555555556,
|
| 1208 |
+
"eval_recall_surprise": 0.432258064516129,
|
| 1209 |
+
"eval_runtime": 14.5644,
|
| 1210 |
+
"eval_samples_per_second": 1518.363,
|
| 1211 |
+
"eval_steps_per_second": 11.878,
|
| 1212 |
+
"step": 4408
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 2.86,
|
| 1216 |
+
"learning_rate": 2.6832774317201726e-06,
|
| 1217 |
+
"loss": 0.5231,
|
| 1218 |
+
"step": 4418
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 2.89,
|
| 1222 |
+
"learning_rate": 2.120268327743172e-06,
|
| 1223 |
+
"loss": 0.5223,
|
| 1224 |
+
"step": 4465
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.92,
|
| 1228 |
+
"learning_rate": 1.5572592237661715e-06,
|
| 1229 |
+
"loss": 0.525,
|
| 1230 |
+
"step": 4512
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 2.95,
|
| 1234 |
+
"learning_rate": 9.942501197891712e-07,
|
| 1235 |
+
"loss": 0.5219,
|
| 1236 |
+
"step": 4559
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 2.98,
|
| 1240 |
+
"learning_rate": 4.312410158121706e-07,
|
| 1241 |
+
"loss": 0.516,
|
| 1242 |
+
"step": 4606
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.0,
|
| 1246 |
+
"step": 4638,
|
| 1247 |
+
"total_flos": 3.903427734912e+16,
|
| 1248 |
+
"train_loss": 0.6570020180521673,
|
| 1249 |
+
"train_runtime": 1377.6596,
|
| 1250 |
+
"train_samples_per_second": 430.73,
|
| 1251 |
+
"train_steps_per_second": 3.367
|
| 1252 |
+
}
|
| 1253 |
+
],
|
| 1254 |
+
"max_steps": 4638,
|
| 1255 |
+
"num_train_epochs": 3,
|
| 1256 |
+
"total_flos": 3.903427734912e+16,
|
| 1257 |
+
"trial_name": null,
|
| 1258 |
+
"trial_params": null
|
| 1259 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8b6d654a0fd9d25c51cd952b5043595530c71ec715e7e1fb0c43d08734c2b34
|
| 3 |
+
size 4027
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|