Commit
•
4829a0a
1
Parent(s):
90fefbc
Training complete
Browse files- README.md +87 -0
- generation_config.json +7 -0
README.md
ADDED
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: kravchenko/uk-mt5-base
|
3 |
+
tags:
|
4 |
+
- summarization
|
5 |
+
- generated_from_trainer
|
6 |
+
datasets:
|
7 |
+
- xlsum
|
8 |
+
metrics:
|
9 |
+
- rouge
|
10 |
+
model-index:
|
11 |
+
- name: uk-mt5-base-xlsum-v2
|
12 |
+
results:
|
13 |
+
- task:
|
14 |
+
name: Sequence-to-sequence Language Modeling
|
15 |
+
type: text2text-generation
|
16 |
+
dataset:
|
17 |
+
name: xlsum
|
18 |
+
type: xlsum
|
19 |
+
config: ukrainian
|
20 |
+
split: validation
|
21 |
+
args: ukrainian
|
22 |
+
metrics:
|
23 |
+
- name: Rouge1
|
24 |
+
type: rouge
|
25 |
+
value: 4.4311
|
26 |
+
---
|
27 |
+
|
28 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
29 |
+
should probably proofread and complete it, then remove this comment. -->
|
30 |
+
|
31 |
+
# uk-mt5-base-xlsum-v2
|
32 |
+
|
33 |
+
This model is a fine-tuned version of [kravchenko/uk-mt5-base](https://huggingface.co/kravchenko/uk-mt5-base) on the xlsum dataset.
|
34 |
+
It achieves the following results on the evaluation set:
|
35 |
+
- Loss: 2.0401
|
36 |
+
- Rouge1: 4.4311
|
37 |
+
- Rouge2: 0.8944
|
38 |
+
- Rougel: 4.4294
|
39 |
+
- Rougelsum: 4.4527
|
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: 5.6e-05
|
59 |
+
- train_batch_size: 4
|
60 |
+
- eval_batch_size: 4
|
61 |
+
- seed: 42
|
62 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
63 |
+
- lr_scheduler_type: linear
|
64 |
+
- num_epochs: 10
|
65 |
+
|
66 |
+
### Training results
|
67 |
+
|
68 |
+
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|
69 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:---------:|
|
70 |
+
| 3.2519 | 1.0 | 2000 | 2.0993 | 4.1141 | 0.5944 | 4.1014 | 4.11 |
|
71 |
+
| 2.5587 | 2.0 | 4000 | 2.0428 | 4.5015 | 0.6167 | 4.4863 | 4.518 |
|
72 |
+
| 2.3299 | 3.0 | 6000 | 2.0175 | 4.4642 | 1.0833 | 4.4528 | 4.5167 |
|
73 |
+
| 2.1543 | 4.0 | 8000 | 2.0183 | 4.3294 | 0.9444 | 4.3408 | 4.3611 |
|
74 |
+
| 2.0276 | 5.0 | 10000 | 2.0039 | 4.6694 | 0.9444 | 4.6264 | 4.6527 |
|
75 |
+
| 1.9119 | 6.0 | 12000 | 2.0139 | 4.9447 | 1.0675 | 4.8908 | 4.9633 |
|
76 |
+
| 1.8305 | 7.0 | 14000 | 2.0134 | 4.9385 | 1.1595 | 4.8774 | 4.9294 |
|
77 |
+
| 1.7669 | 8.0 | 16000 | 2.0253 | 4.2697 | 0.9667 | 4.2524 | 4.3167 |
|
78 |
+
| 1.7141 | 9.0 | 18000 | 2.0354 | 4.4527 | 0.9 | 4.448 | 4.4941 |
|
79 |
+
| 1.681 | 10.0 | 20000 | 2.0401 | 4.4311 | 0.8944 | 4.4294 | 4.4527 |
|
80 |
+
|
81 |
+
|
82 |
+
### Framework versions
|
83 |
+
|
84 |
+
- Transformers 4.34.1
|
85 |
+
- Pytorch 2.0.1+cu118
|
86 |
+
- Datasets 2.14.5
|
87 |
+
- Tokenizers 0.14.1
|
generation_config.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"decoder_start_token_id": 0,
|
3 |
+
"eos_token_id": 1,
|
4 |
+
"pad_token_id": 0,
|
5 |
+
"transformers_version": "4.34.1",
|
6 |
+
"use_cache": false
|
7 |
+
}
|