PontifexMaximus commited on
Commit
5a647e2
1 Parent(s): e71fa31

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +105 -0
README.md ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - id_panl_bppt
7
+ metrics:
8
+ - bleu
9
+ model-index:
10
+ - name: opus-mt-id-en-finetuned-id-to-en
11
+ results:
12
+ - task:
13
+ name: Sequence-to-sequence Language Modeling
14
+ type: text2text-generation
15
+ dataset:
16
+ name: id_panl_bppt
17
+ type: id_panl_bppt
18
+ config: id_panl_bppt
19
+ split: train
20
+ args: id_panl_bppt
21
+ metrics:
22
+ - name: Bleu
23
+ type: bleu
24
+ value: 30.557
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
+ # opus-mt-id-en-finetuned-id-to-en
31
+
32
+ This model is a fine-tuned version of [Helsinki-NLP/opus-mt-id-en](https://huggingface.co/Helsinki-NLP/opus-mt-id-en) on the id_panl_bppt dataset.
33
+ It achieves the following results on the evaluation set:
34
+ - Loss: 1.6469
35
+ - Bleu: 30.557
36
+ - Gen Len: 29.8247
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 2e-06
56
+ - train_batch_size: 32
57
+ - eval_batch_size: 32
58
+ - seed: 42
59
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
+ - lr_scheduler_type: linear
61
+ - num_epochs: 30
62
+ - mixed_precision_training: Native AMP
63
+
64
+ ### Training results
65
+
66
+ | Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
67
+ |:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|
68
+ | 2.5737 | 1.0 | 751 | 2.2222 | 24.4223 | 30.3344 |
69
+ | 2.3756 | 2.0 | 1502 | 2.1264 | 25.419 | 30.3147 |
70
+ | 2.3146 | 3.0 | 2253 | 2.0588 | 26.0995 | 30.1959 |
71
+ | 2.2411 | 4.0 | 3004 | 2.0072 | 26.5944 | 30.0763 |
72
+ | 2.1927 | 5.0 | 3755 | 1.9657 | 27.0422 | 30.0773 |
73
+ | 2.1554 | 6.0 | 4506 | 1.9284 | 27.4151 | 30.0715 |
74
+ | 2.1105 | 7.0 | 5257 | 1.8980 | 27.6645 | 29.9426 |
75
+ | 2.0841 | 8.0 | 6008 | 1.8680 | 28.023 | 29.9797 |
76
+ | 2.0491 | 9.0 | 6759 | 1.8438 | 28.2456 | 29.9342 |
77
+ | 2.0265 | 10.0 | 7510 | 1.8218 | 28.5378 | 29.8968 |
78
+ | 2.0065 | 11.0 | 8261 | 1.8012 | 28.7599 | 29.8907 |
79
+ | 1.9764 | 12.0 | 9012 | 1.7835 | 28.9369 | 29.8796 |
80
+ | 1.969 | 13.0 | 9763 | 1.7663 | 29.1565 | 29.8671 |
81
+ | 1.9474 | 14.0 | 10514 | 1.7506 | 29.3313 | 29.893 |
82
+ | 1.9397 | 15.0 | 11265 | 1.7378 | 29.4567 | 29.8512 |
83
+ | 1.9217 | 16.0 | 12016 | 1.7239 | 29.6245 | 29.8361 |
84
+ | 1.9174 | 17.0 | 12767 | 1.7127 | 29.7464 | 29.8398 |
85
+ | 1.9021 | 18.0 | 13518 | 1.7030 | 29.9035 | 29.8621 |
86
+ | 1.89 | 19.0 | 14269 | 1.6934 | 29.9669 | 29.8225 |
87
+ | 1.878 | 20.0 | 15020 | 1.6847 | 30.0961 | 29.8398 |
88
+ | 1.8671 | 21.0 | 15771 | 1.6774 | 30.1878 | 29.839 |
89
+ | 1.8634 | 22.0 | 16522 | 1.6717 | 30.2341 | 29.8134 |
90
+ | 1.8536 | 23.0 | 17273 | 1.6653 | 30.3356 | 29.816 |
91
+ | 1.8533 | 24.0 | 18024 | 1.6602 | 30.3548 | 29.8251 |
92
+ | 1.8476 | 25.0 | 18775 | 1.6560 | 30.4323 | 29.8315 |
93
+ | 1.8362 | 26.0 | 19526 | 1.6528 | 30.4682 | 29.8277 |
94
+ | 1.8463 | 27.0 | 20277 | 1.6501 | 30.5002 | 29.8236 |
95
+ | 1.8369 | 28.0 | 21028 | 1.6484 | 30.5236 | 29.8257 |
96
+ | 1.8313 | 29.0 | 21779 | 1.6472 | 30.55 | 29.8259 |
97
+ | 1.8332 | 30.0 | 22530 | 1.6469 | 30.557 | 29.8247 |
98
+
99
+
100
+ ### Framework versions
101
+
102
+ - Transformers 4.21.1
103
+ - Pytorch 1.12.0
104
+ - Datasets 2.4.0
105
+ - Tokenizers 0.12.1