Commit
•
a2c8228
1
Parent(s):
26c078e
update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- iva_mt_wslot
|
7 |
+
metrics:
|
8 |
+
- bleu
|
9 |
+
model-index:
|
10 |
+
- name: iva_mt_wslot-m2m100_418M-zh-en
|
11 |
+
results:
|
12 |
+
- task:
|
13 |
+
name: Sequence-to-sequence Language Modeling
|
14 |
+
type: text2text-generation
|
15 |
+
dataset:
|
16 |
+
name: iva_mt_wslot
|
17 |
+
type: iva_mt_wslot
|
18 |
+
config: en-zh
|
19 |
+
split: validation
|
20 |
+
args: en-zh
|
21 |
+
metrics:
|
22 |
+
- name: Bleu
|
23 |
+
type: bleu
|
24 |
+
value: 67.9385
|
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 |
+
# iva_mt_wslot-m2m100_418M-zh-en
|
31 |
+
|
32 |
+
This model is a fine-tuned version of [facebook/m2m100_418M](https://huggingface.co/facebook/m2m100_418M) on the iva_mt_wslot dataset.
|
33 |
+
It achieves the following results on the evaluation set:
|
34 |
+
- Loss: 0.0139
|
35 |
+
- Bleu: 67.9385
|
36 |
+
- Gen Len: 18.9988
|
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-05
|
56 |
+
- train_batch_size: 4
|
57 |
+
- eval_batch_size: 4
|
58 |
+
- seed: 42
|
59 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
60 |
+
- lr_scheduler_type: linear
|
61 |
+
- num_epochs: 10
|
62 |
+
- mixed_precision_training: Native AMP
|
63 |
+
|
64 |
+
### Training results
|
65 |
+
|
66 |
+
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|
67 |
+
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|
|
68 |
+
| 0.016 | 1.0 | 2437 | 0.0138 | 62.077 | 18.6077 |
|
69 |
+
| 0.0114 | 2.0 | 4874 | 0.0126 | 64.3834 | 18.9019 |
|
70 |
+
| 0.0084 | 3.0 | 7311 | 0.0123 | 66.0012 | 18.9206 |
|
71 |
+
| 0.0067 | 4.0 | 9748 | 0.0123 | 66.7838 | 19.0018 |
|
72 |
+
| 0.005 | 5.0 | 12185 | 0.0124 | 66.9053 | 18.9527 |
|
73 |
+
| 0.0039 | 6.0 | 14622 | 0.0128 | 67.5252 | 18.9918 |
|
74 |
+
| 0.003 | 7.0 | 17059 | 0.0131 | 67.3664 | 18.9609 |
|
75 |
+
| 0.0025 | 8.0 | 19496 | 0.0135 | 67.792 | 19.0198 |
|
76 |
+
| 0.0019 | 9.0 | 21933 | 0.0137 | 67.7256 | 18.9591 |
|
77 |
+
| 0.0015 | 10.0 | 24370 | 0.0139 | 67.9385 | 18.9988 |
|
78 |
+
|
79 |
+
|
80 |
+
### Framework versions
|
81 |
+
|
82 |
+
- Transformers 4.28.1
|
83 |
+
- Pytorch 2.1.0+cu121
|
84 |
+
- Datasets 2.15.0
|
85 |
+
- Tokenizers 0.13.3
|