update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- summarization
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- rouge
|
8 |
+
model-index:
|
9 |
+
- name: mt5-base-wikinewssum-all-languages
|
10 |
+
results: []
|
11 |
+
---
|
12 |
+
|
13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
+
should probably proofread and complete it, then remove this comment. -->
|
15 |
+
|
16 |
+
# mt5-base-wikinewssum-all-languages
|
17 |
+
|
18 |
+
This model is a fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) on an unknown dataset.
|
19 |
+
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 2.2454
|
21 |
+
- Rouge1: 8.3826
|
22 |
+
- Rouge2: 3.5524
|
23 |
+
- Rougel: 6.8656
|
24 |
+
- Rougelsum: 7.8362
|
25 |
+
|
26 |
+
## Model description
|
27 |
+
|
28 |
+
More information needed
|
29 |
+
|
30 |
+
## Intended uses & limitations
|
31 |
+
|
32 |
+
More information needed
|
33 |
+
|
34 |
+
## Training and evaluation data
|
35 |
+
|
36 |
+
More information needed
|
37 |
+
|
38 |
+
## Training procedure
|
39 |
+
|
40 |
+
### Training hyperparameters
|
41 |
+
|
42 |
+
The following hyperparameters were used during training:
|
43 |
+
- learning_rate: 5.6e-05
|
44 |
+
- train_batch_size: 4
|
45 |
+
- eval_batch_size: 4
|
46 |
+
- seed: 42
|
47 |
+
- gradient_accumulation_steps: 2
|
48 |
+
- total_train_batch_size: 8
|
49 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
50 |
+
- lr_scheduler_type: linear
|
51 |
+
- num_epochs: 8
|
52 |
+
|
53 |
+
### Training results
|
54 |
+
|
55 |
+
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|
56 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:---------:|
|
57 |
+
| No log | 1.0 | 3467 | 2.4034 | 8.0363 | 3.2484 | 6.5409 | 7.477 |
|
58 |
+
| No log | 2.0 | 6934 | 2.3276 | 8.1054 | 3.2905 | 6.5765 | 7.5687 |
|
59 |
+
| No log | 3.0 | 10401 | 2.2976 | 8.169 | 3.4272 | 6.6597 | 7.6435 |
|
60 |
+
| No log | 4.0 | 13868 | 2.2795 | 8.2941 | 3.5353 | 6.7881 | 7.7664 |
|
61 |
+
| 2.8057 | 5.0 | 17335 | 2.2621 | 8.3302 | 3.5599 | 6.8238 | 7.7928 |
|
62 |
+
| 2.8057 | 6.0 | 20802 | 2.2547 | 8.3818 | 3.5886 | 6.8672 | 7.844 |
|
63 |
+
| 2.8057 | 7.0 | 24269 | 2.2472 | 8.3809 | 3.5696 | 6.8575 | 7.8327 |
|
64 |
+
| 2.8057 | 8.0 | 27736 | 2.2454 | 8.3826 | 3.5524 | 6.8656 | 7.8362 |
|
65 |
+
|
66 |
+
|
67 |
+
### Framework versions
|
68 |
+
|
69 |
+
- Transformers 4.13.0
|
70 |
+
- Pytorch 1.10.1
|
71 |
+
- Datasets 1.16.1
|
72 |
+
- Tokenizers 0.10.3
|