Commit
•
42f3566
1
Parent(s):
a5b29e5
update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,101 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- esnli
|
7 |
+
metrics:
|
8 |
+
- accuracy
|
9 |
+
- f1
|
10 |
+
- rouge
|
11 |
+
- bleu
|
12 |
+
model-index:
|
13 |
+
- name: t5-small-e-snli-generation-label_and_explanation-selected-b48
|
14 |
+
results:
|
15 |
+
- task:
|
16 |
+
name: Sequence-to-sequence Language Modeling
|
17 |
+
type: text2text-generation
|
18 |
+
dataset:
|
19 |
+
name: esnli
|
20 |
+
type: esnli
|
21 |
+
config: plain_text
|
22 |
+
split: validation
|
23 |
+
args: plain_text
|
24 |
+
metrics:
|
25 |
+
- name: Accuracy
|
26 |
+
type: accuracy
|
27 |
+
value: 0.8657793131477342
|
28 |
+
- name: F1
|
29 |
+
type: f1
|
30 |
+
value: 0.8658628497423001
|
31 |
+
- name: Rouge1
|
32 |
+
type: rouge
|
33 |
+
value: 0.6049779979620054
|
34 |
+
- name: Bleu
|
35 |
+
type: bleu
|
36 |
+
value: 0.4039391893498565
|
37 |
+
---
|
38 |
+
|
39 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
40 |
+
should probably proofread and complete it, then remove this comment. -->
|
41 |
+
|
42 |
+
# t5-small-e-snli-generation-label_and_explanation-selected-b48
|
43 |
+
|
44 |
+
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the esnli dataset.
|
45 |
+
It achieves the following results on the evaluation set:
|
46 |
+
- Loss: 1.9091
|
47 |
+
- Accuracy: 0.8658
|
48 |
+
- F1: 0.8659
|
49 |
+
- Bertscore F1: 0.9337
|
50 |
+
- Rouge1: 0.6050
|
51 |
+
- Rouge2: 0.3983
|
52 |
+
- Rougel: 0.5492
|
53 |
+
- Rougelsum: 0.5513
|
54 |
+
- Bleu: 0.4039
|
55 |
+
|
56 |
+
## Model description
|
57 |
+
|
58 |
+
More information needed
|
59 |
+
|
60 |
+
## Intended uses & limitations
|
61 |
+
|
62 |
+
More information needed
|
63 |
+
|
64 |
+
## Training and evaluation data
|
65 |
+
|
66 |
+
More information needed
|
67 |
+
|
68 |
+
## Training procedure
|
69 |
+
|
70 |
+
### Training hyperparameters
|
71 |
+
|
72 |
+
The following hyperparameters were used during training:
|
73 |
+
- learning_rate: 0.001
|
74 |
+
- train_batch_size: 48
|
75 |
+
- eval_batch_size: 48
|
76 |
+
- seed: 42
|
77 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
78 |
+
- lr_scheduler_type: linear
|
79 |
+
- lr_scheduler_warmup_ratio: 0.05
|
80 |
+
- num_epochs: 10
|
81 |
+
|
82 |
+
### Training results
|
83 |
+
|
84 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Bertscore F1 | Rouge1 | Rouge2 | Rougel | Rougelsum | Bleu |
|
85 |
+
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:------:|:------------:|:------:|:------:|:------:|:---------:|:------:|
|
86 |
+
| 1.7285 | 0.17 | 2000 | 1.9945 | 0.7799 | 0.7792 | 0.9249 | 0.5631 | 0.3517 | 0.5091 | 0.5116 | 0.3617 |
|
87 |
+
| 1.3318 | 0.35 | 4000 | 1.9494 | 0.7980 | 0.7971 | 0.9295 | 0.5766 | 0.3656 | 0.5218 | 0.5234 | 0.3785 |
|
88 |
+
| 1.2662 | 0.52 | 6000 | 1.8983 | 0.8322 | 0.8331 | 0.9289 | 0.5769 | 0.3656 | 0.5205 | 0.5225 | 0.3727 |
|
89 |
+
| 1.2285 | 0.7 | 8000 | 1.9078 | 0.8391 | 0.8396 | 0.9313 | 0.5833 | 0.3734 | 0.5304 | 0.5321 | 0.3884 |
|
90 |
+
| 1.1973 | 0.87 | 10000 | 1.9246 | 0.8485 | 0.8470 | 0.9303 | 0.5888 | 0.3782 | 0.5322 | 0.5339 | 0.3868 |
|
91 |
+
| 1.1715 | 1.05 | 12000 | 1.9262 | 0.8561 | 0.8565 | 0.9331 | 0.6020 | 0.3950 | 0.5464 | 0.5479 | 0.4039 |
|
92 |
+
| 1.1368 | 1.22 | 14000 | 1.9155 | 0.8621 | 0.8612 | 0.9313 | 0.6027 | 0.3918 | 0.5442 | 0.5463 | 0.3889 |
|
93 |
+
| 1.1281 | 1.4 | 16000 | 1.9091 | 0.8658 | 0.8659 | 0.9337 | 0.6050 | 0.3983 | 0.5492 | 0.5513 | 0.4039 |
|
94 |
+
|
95 |
+
|
96 |
+
### Framework versions
|
97 |
+
|
98 |
+
- Transformers 4.27.4
|
99 |
+
- Pytorch 2.0.0+cu117
|
100 |
+
- Datasets 2.11.0
|
101 |
+
- Tokenizers 0.13.2
|