Kevincp560
commited on
Commit
•
d236099
1
Parent(s):
f456a9c
update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- pub_med_summarization_dataset
|
7 |
+
metrics:
|
8 |
+
- rouge
|
9 |
+
model-index:
|
10 |
+
- name: bigbird-pegasus-large-bigpatent-finetuned-pubMed
|
11 |
+
results:
|
12 |
+
- task:
|
13 |
+
name: Sequence-to-sequence Language Modeling
|
14 |
+
type: text2text-generation
|
15 |
+
dataset:
|
16 |
+
name: pub_med_summarization_dataset
|
17 |
+
type: pub_med_summarization_dataset
|
18 |
+
args: document
|
19 |
+
metrics:
|
20 |
+
- name: Rouge1
|
21 |
+
type: rouge
|
22 |
+
value: 45.0851
|
23 |
+
---
|
24 |
+
|
25 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
26 |
+
should probably proofread and complete it, then remove this comment. -->
|
27 |
+
|
28 |
+
# bigbird-pegasus-large-bigpatent-finetuned-pubMed
|
29 |
+
|
30 |
+
This model is a fine-tuned version of [google/bigbird-pegasus-large-bigpatent](https://huggingface.co/google/bigbird-pegasus-large-bigpatent) on the pub_med_summarization_dataset dataset.
|
31 |
+
It achieves the following results on the evaluation set:
|
32 |
+
- Loss: 1.5403
|
33 |
+
- Rouge1: 45.0851
|
34 |
+
- Rouge2: 19.5488
|
35 |
+
- Rougel: 27.391
|
36 |
+
- Rougelsum: 41.112
|
37 |
+
- Gen Len: 231.608
|
38 |
+
|
39 |
+
## Model description
|
40 |
+
|
41 |
+
More information needed
|
42 |
+
|
43 |
+
## Intended uses & limitations
|
44 |
+
|
45 |
+
More information needed
|
46 |
+
|
47 |
+
## Training and evaluation data
|
48 |
+
|
49 |
+
More information needed
|
50 |
+
|
51 |
+
## Training procedure
|
52 |
+
|
53 |
+
### Training hyperparameters
|
54 |
+
|
55 |
+
The following hyperparameters were used during training:
|
56 |
+
- learning_rate: 2e-05
|
57 |
+
- train_batch_size: 4
|
58 |
+
- eval_batch_size: 4
|
59 |
+
- seed: 42
|
60 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
61 |
+
- lr_scheduler_type: linear
|
62 |
+
- num_epochs: 5
|
63 |
+
- mixed_precision_training: Native AMP
|
64 |
+
|
65 |
+
### Training results
|
66 |
+
|
67 |
+
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
68 |
+
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
|
69 |
+
| 2.1198 | 1.0 | 500 | 1.6285 | 43.0579 | 18.1792 | 26.421 | 39.0769 | 214.924 |
|
70 |
+
| 1.6939 | 2.0 | 1000 | 1.5696 | 44.0679 | 18.9331 | 26.84 | 40.0684 | 222.814 |
|
71 |
+
| 1.6195 | 3.0 | 1500 | 1.5506 | 44.7352 | 19.3532 | 27.2418 | 40.7454 | 229.396 |
|
72 |
+
| 1.5798 | 4.0 | 2000 | 1.5403 | 45.0415 | 19.5019 | 27.2969 | 40.951 | 231.044 |
|
73 |
+
| 1.5592 | 5.0 | 2500 | 1.5403 | 45.0851 | 19.5488 | 27.391 | 41.112 | 231.608 |
|
74 |
+
|
75 |
+
|
76 |
+
### Framework versions
|
77 |
+
|
78 |
+
- Transformers 4.17.0
|
79 |
+
- Pytorch 1.9.1
|
80 |
+
- Datasets 1.18.4
|
81 |
+
- Tokenizers 0.11.6
|