Commit
•
4704864
1
Parent(s):
a95f140
Upload TFXLNetForSequenceClassification
Browse files- README.md +73 -0
- config.json +49 -0
- tf_model.h5 +3 -0
README.md
ADDED
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
tags:
|
4 |
+
- generated_from_keras_callback
|
5 |
+
model-index:
|
6 |
+
- name: Regression_xlnet_aug_CustomLoss
|
7 |
+
results: []
|
8 |
+
---
|
9 |
+
|
10 |
+
<!-- This model card has been generated automatically according to the information Keras had access to. You should
|
11 |
+
probably proofread and complete it, then remove this comment. -->
|
12 |
+
|
13 |
+
# Regression_xlnet_aug_CustomLoss
|
14 |
+
|
15 |
+
This model is a fine-tuned version of [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on an unknown dataset.
|
16 |
+
It achieves the following results on the evaluation set:
|
17 |
+
- Train Loss: 0.3652
|
18 |
+
- Train Mae: 0.5686
|
19 |
+
- Train Mse: 0.4521
|
20 |
+
- Train R2-score: 0.7022
|
21 |
+
- Validation Loss: 0.3538
|
22 |
+
- Validation Mae: 0.5478
|
23 |
+
- Validation Mse: 0.4335
|
24 |
+
- Validation R2-score: 0.7272
|
25 |
+
- Epoch: 14
|
26 |
+
|
27 |
+
## Model description
|
28 |
+
|
29 |
+
More information needed
|
30 |
+
|
31 |
+
## Intended uses & limitations
|
32 |
+
|
33 |
+
More information needed
|
34 |
+
|
35 |
+
## Training and evaluation data
|
36 |
+
|
37 |
+
More information needed
|
38 |
+
|
39 |
+
## Training procedure
|
40 |
+
|
41 |
+
### Training hyperparameters
|
42 |
+
|
43 |
+
The following hyperparameters were used during training:
|
44 |
+
- optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': 1e-04, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
|
45 |
+
- training_precision: float32
|
46 |
+
|
47 |
+
### Training results
|
48 |
+
|
49 |
+
| Train Loss | Train Mae | Train Mse | Train R2-score | Validation Loss | Validation Mae | Validation Mse | Validation R2-score | Epoch |
|
50 |
+
|:----------:|:---------:|:---------:|:--------------:|:---------------:|:--------------:|:--------------:|:-------------------:|:-----:|
|
51 |
+
| 0.3976 | 0.5843 | 0.4994 | 0.6170 | 0.3698 | 0.6232 | 0.4576 | 0.5391 | 0 |
|
52 |
+
| 0.3938 | 0.5830 | 0.4901 | 0.5860 | 0.4330 | 0.6806 | 0.5537 | 0.2841 | 1 |
|
53 |
+
| 0.3822 | 0.5739 | 0.4691 | -2.6800 | 0.3660 | 0.5198 | 0.4614 | 0.7579 | 2 |
|
54 |
+
| 0.3716 | 0.5723 | 0.4607 | -0.3420 | 0.3541 | 0.5461 | 0.4343 | 0.7295 | 3 |
|
55 |
+
| 0.3852 | 0.5794 | 0.4793 | 0.6654 | 0.3524 | 0.5654 | 0.4288 | 0.6982 | 4 |
|
56 |
+
| 0.3826 | 0.5782 | 0.4787 | 0.5114 | 0.3584 | 0.5991 | 0.4374 | 0.6172 | 5 |
|
57 |
+
| 0.3817 | 0.5746 | 0.4754 | 0.6431 | 0.3569 | 0.5945 | 0.4348 | 0.6304 | 6 |
|
58 |
+
| 0.3705 | 0.5689 | 0.4541 | 0.6679 | 0.3621 | 0.5210 | 0.4528 | 0.7540 | 7 |
|
59 |
+
| 0.3719 | 0.5654 | 0.4567 | 0.6807 | 0.3544 | 0.5446 | 0.4350 | 0.7315 | 8 |
|
60 |
+
| 0.3752 | 0.5698 | 0.4638 | 0.6324 | 0.3571 | 0.5343 | 0.4413 | 0.7433 | 9 |
|
61 |
+
| 0.3773 | 0.5714 | 0.4655 | 0.7523 | 0.3534 | 0.5787 | 0.4294 | 0.6701 | 10 |
|
62 |
+
| 0.3651 | 0.5615 | 0.4490 | 0.6223 | 0.3766 | 0.5372 | 0.4846 | 0.7600 | 11 |
|
63 |
+
| 0.3706 | 0.5649 | 0.4529 | 0.5998 | 0.3525 | 0.5683 | 0.4286 | 0.6924 | 12 |
|
64 |
+
| 0.3779 | 0.5687 | 0.4619 | 0.6900 | 0.3532 | 0.5771 | 0.4291 | 0.6737 | 13 |
|
65 |
+
| 0.3652 | 0.5686 | 0.4521 | 0.7022 | 0.3538 | 0.5478 | 0.4335 | 0.7272 | 14 |
|
66 |
+
|
67 |
+
|
68 |
+
### Framework versions
|
69 |
+
|
70 |
+
- Transformers 4.28.1
|
71 |
+
- TensorFlow 2.12.0
|
72 |
+
- Datasets 2.11.0
|
73 |
+
- Tokenizers 0.13.3
|
config.json
ADDED
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "xlnet-base-cased",
|
3 |
+
"architectures": [
|
4 |
+
"XLNetForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attn_type": "bi",
|
7 |
+
"bi_data": false,
|
8 |
+
"bos_token_id": 1,
|
9 |
+
"clamp_len": -1,
|
10 |
+
"d_head": 64,
|
11 |
+
"d_inner": 3072,
|
12 |
+
"d_model": 768,
|
13 |
+
"dropout": 0.1,
|
14 |
+
"end_n_top": 5,
|
15 |
+
"eos_token_id": 2,
|
16 |
+
"ff_activation": "gelu",
|
17 |
+
"id2label": {
|
18 |
+
"0": "LABEL_0"
|
19 |
+
},
|
20 |
+
"initializer_range": 0.02,
|
21 |
+
"label2id": {
|
22 |
+
"LABEL_0": 0
|
23 |
+
},
|
24 |
+
"layer_norm_eps": 1e-12,
|
25 |
+
"mem_len": null,
|
26 |
+
"model_type": "xlnet",
|
27 |
+
"n_head": 12,
|
28 |
+
"n_layer": 12,
|
29 |
+
"pad_token_id": 5,
|
30 |
+
"problem_type": "regression",
|
31 |
+
"reuse_len": null,
|
32 |
+
"same_length": false,
|
33 |
+
"start_n_top": 5,
|
34 |
+
"summary_activation": "tanh",
|
35 |
+
"summary_last_dropout": 0.1,
|
36 |
+
"summary_type": "last",
|
37 |
+
"summary_use_proj": true,
|
38 |
+
"task_specific_params": {
|
39 |
+
"text-generation": {
|
40 |
+
"do_sample": true,
|
41 |
+
"max_length": 250
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"transformers_version": "4.28.1",
|
45 |
+
"untie_r": true,
|
46 |
+
"use_mems_eval": true,
|
47 |
+
"use_mems_train": false,
|
48 |
+
"vocab_size": 32000
|
49 |
+
}
|
tf_model.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7c7195d7f0b091346512a582bb130c9865eb2d4f2b3c02e3f30afcc5648a4e69
|
3 |
+
size 469445640
|