update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- scientific_papers
|
7 |
+
metrics:
|
8 |
+
- rouge
|
9 |
+
model-index:
|
10 |
+
- name: t5-small-science-papers
|
11 |
+
results:
|
12 |
+
- task:
|
13 |
+
name: Sequence-to-sequence Language Modeling
|
14 |
+
type: text2text-generation
|
15 |
+
dataset:
|
16 |
+
name: scientific_papers
|
17 |
+
type: scientific_papers
|
18 |
+
config: arxiv
|
19 |
+
split: train
|
20 |
+
args: arxiv
|
21 |
+
metrics:
|
22 |
+
- name: Rouge1
|
23 |
+
type: rouge
|
24 |
+
value: 12.3568
|
25 |
+
---
|
26 |
+
|
27 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
28 |
+
should probably proofread and complete it, then remove this comment. -->
|
29 |
+
|
30 |
+
# t5-small-science-papers
|
31 |
+
|
32 |
+
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the scientific_papers dataset.
|
33 |
+
It achieves the following results on the evaluation set:
|
34 |
+
- Loss: 3.6405
|
35 |
+
- Rouge1: 12.3568
|
36 |
+
- Rouge2: 2.4449
|
37 |
+
- Rougel: 10.2371
|
38 |
+
- Rougelsum: 11.4209
|
39 |
+
- Gen Len: 19.0
|
40 |
+
|
41 |
+
## Model description
|
42 |
+
|
43 |
+
More information needed
|
44 |
+
|
45 |
+
## Intended uses & limitations
|
46 |
+
|
47 |
+
More information needed
|
48 |
+
|
49 |
+
## Training and evaluation data
|
50 |
+
|
51 |
+
More information needed
|
52 |
+
|
53 |
+
## Training procedure
|
54 |
+
|
55 |
+
### Training hyperparameters
|
56 |
+
|
57 |
+
The following hyperparameters were used during training:
|
58 |
+
- learning_rate: 2e-05
|
59 |
+
- train_batch_size: 16
|
60 |
+
- eval_batch_size: 16
|
61 |
+
- seed: 42
|
62 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
63 |
+
- lr_scheduler_type: linear
|
64 |
+
- num_epochs: 5
|
65 |
+
- mixed_precision_training: Native AMP
|
66 |
+
|
67 |
+
### Training results
|
68 |
+
|
69 |
+
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
70 |
+
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:------:|:-------:|:---------:|:-------:|
|
71 |
+
| 4.4735 | 1.0 | 12690 | 4.3727 | 9.9604 | 1.7641 | 8.6213 | 9.2779 | 19.0 |
|
72 |
+
| 4.0104 | 2.0 | 25380 | 3.9384 | 11.4001 | 2.1474 | 9.6516 | 10.6602 | 19.0 |
|
73 |
+
| 3.8237 | 3.0 | 38070 | 3.7580 | 11.1806 | 2.1229 | 9.3881 | 10.3853 | 19.0 |
|
74 |
+
| 3.7382 | 4.0 | 50760 | 3.6738 | 11.9298 | 2.3222 | 9.9077 | 11.045 | 19.0 |
|
75 |
+
| 3.6994 | 5.0 | 63450 | 3.6405 | 12.3568 | 2.4449 | 10.2371 | 11.4209 | 19.0 |
|
76 |
+
|
77 |
+
|
78 |
+
### Framework versions
|
79 |
+
|
80 |
+
- Transformers 4.24.0
|
81 |
+
- Pytorch 1.12.1+cu113
|
82 |
+
- Datasets 2.6.1
|
83 |
+
- Tokenizers 0.13.1
|