File size: 2,278 Bytes
0c1a6df
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
59ab818
0c1a6df
 
 
 
 
 
 
 
 
59ab818
 
 
 
 
0c1a6df
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a80318f
 
0c1a6df
 
 
a80318f
 
0c1a6df
 
 
 
 
59ab818
 
 
 
 
0c1a6df
 
 
 
a80318f
0c1a6df
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
---
license: apache-2.0
base_model: google-t5/t5-small
tags:
- summarization
- generated_from_trainer
datasets:
- scientific_papers
metrics:
- rouge
model-index:
- name: t5-small-finetuned-scientific-articles
  results:
  - task:
      name: Sequence-to-sequence Language Modeling
      type: text2text-generation
    dataset:
      name: scientific_papers
      type: scientific_papers
      config: pubmed
      split: train
      args: pubmed
    metrics:
    - name: Rouge1
      type: rouge
      value: 7.8805
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# t5-small-finetuned-scientific-articles

This model is a fine-tuned version of [google-t5/t5-small](https://huggingface.co/google-t5/t5-small) on the scientific_papers dataset.
It achieves the following results on the evaluation set:
- Loss: 3.2246
- Rouge1: 7.8805
- Rouge2: 2.622
- Rougel: 6.7327
- Rougelsum: 7.3172

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5.6e-05
- train_batch_size: 9
- eval_batch_size: 9
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
| 7.3084        | 1.0   | 56   | 5.4563          | 6.7162 | 2.0525 | 5.6729 | 6.2076    |
| 4.5048        | 2.0   | 112  | 3.8325          | 7.2382 | 2.4034 | 6.1187 | 6.674     |
| 3.7194        | 3.0   | 168  | 3.4523          | 7.7064 | 2.2556 | 6.4334 | 7.1049    |
| 3.5218        | 4.0   | 224  | 3.3173          | 8.1033 | 2.6122 | 6.8396 | 7.4375    |
| 3.4221        | 5.0   | 280  | 3.2246          | 7.8805 | 2.622  | 6.7327 | 7.3172    |


### Framework versions

- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1