Hugging Face's logo

nananatsu
/
opus-mt-en-zh-finetuned-audio-product

Text2Text Generation
Transformers
ONNX
Safetensors
marian
Generated from Trainer
Inference Endpoints
Model card Files Community
nananatsu commited on
Commit
22998b9
·
verified ·
1 Parent(s): 0855c2f

Model save

Browse files
Files changed (3) hide show
  1. README.md +27 -27
  2. model.safetensors +1 -1
  3. training_args.bin +1 -1
README.md CHANGED
@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-zh](https://huggingface.co/Helsinki-NLP/opus-mt-en-zh) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 0.0999
22
- - Bleu: 58.9781
23
 
24
  ## Model description
25
 
@@ -39,12 +39,12 @@ More information needed
39
 
40
  The following hyperparameters were used during training:
41
  - learning_rate: 5e-05
42
- - train_batch_size: 16
43
- - eval_batch_size: 16
44
  - seed: 42
45
- - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
46
- - lr_scheduler_type: linear
47
- - lr_scheduler_warmup_steps: 500
48
  - num_epochs: 20
49
  - mixed_precision_training: Native AMP
50
 
@@ -52,26 +52,26 @@ The following hyperparameters were used during training:
52
 
53
  | Training Loss | Epoch | Step | Validation Loss | Bleu |
54
  |:-------------:|:-----:|:----:|:---------------:|:-------:|
55
- | 3.5089 | 1.0 | 108 | 0.2055 | 29.8305 |
56
- | 0.2267 | 2.0 | 216 | 0.1524 | 38.9089 |
57
- | 0.1471 | 3.0 | 324 | 0.1248 | 47.2948 |
58
- | 0.1093 | 4.0 | 432 | 0.1093 | 51.7524 |
59
- | 0.0782 | 5.0 | 540 | 0.1025 | 53.4004 |
60
- | 0.0522 | 6.0 | 648 | 0.0992 | 54.9161 |
61
- | 0.0395 | 7.0 | 756 | 0.0972 | 56.7778 |
62
- | 0.0299 | 8.0 | 864 | 0.0960 | 57.2394 |
63
- | 0.0227 | 9.0 | 972 | 0.0965 | 57.8044 |
64
- | 0.0163 | 10.0 | 1080 | 0.0965 | 59.7562 |
65
- | 0.0148 | 11.0 | 1188 | 0.0970 | 58.9913 |
66
- | 0.0113 | 12.0 | 1296 | 0.0968 | 59.3438 |
67
- | 0.0085 | 13.0 | 1404 | 0.0983 | 58.0797 |
68
- | 0.0071 | 14.0 | 1512 | 0.0982 | 58.7092 |
69
- | 0.0065 | 15.0 | 1620 | 0.0994 | 58.2769 |
70
- | 0.0063 | 16.0 | 1728 | 0.0992 | 59.2400 |
71
- | 0.0053 | 17.0 | 1836 | 0.0992 | 59.1475 |
72
- | 0.0056 | 18.0 | 1944 | 0.0995 | 59.6588 |
73
- | 0.0044 | 19.0 | 2052 | 0.0998 | 59.3419 |
74
- | 0.0044 | 20.0 | 2160 | 0.0999 | 58.9781 |
75
 
76
 
77
  ### Framework versions
 
18
 
19
  This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-zh](https://huggingface.co/Helsinki-NLP/opus-mt-en-zh) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.0734
22
+ - Bleu: 54.8194
23
 
24
  ## Model description
25
 
 
39
 
40
  The following hyperparameters were used during training:
41
  - learning_rate: 5e-05
42
+ - train_batch_size: 64
43
+ - eval_batch_size: 64
44
  - seed: 42
45
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
46
+ - lr_scheduler_type: cosine
47
+ - lr_scheduler_warmup_steps: 100
48
  - num_epochs: 20
49
  - mixed_precision_training: Native AMP
50
 
 
52
 
53
  | Training Loss | Epoch | Step | Validation Loss | Bleu |
54
  |:-------------:|:-----:|:----:|:---------------:|:-------:|
55
+ | 3.7778 | 1.0 | 54 | 0.1557 | 28.3526 |
56
+ | 0.1726 | 2.0 | 108 | 0.1068 | 39.8488 |
57
+ | 0.1128 | 3.0 | 162 | 0.0895 | 46.0427 |
58
+ | 0.0777 | 4.0 | 216 | 0.0813 | 47.7747 |
59
+ | 0.0634 | 5.0 | 270 | 0.0768 | 48.3524 |
60
+ | 0.0498 | 6.0 | 324 | 0.0751 | 49.6404 |
61
+ | 0.0392 | 7.0 | 378 | 0.0729 | 51.0547 |
62
+ | 0.0324 | 8.0 | 432 | 0.0728 | 52.2582 |
63
+ | 0.0274 | 9.0 | 486 | 0.0730 | 53.4640 |
64
+ | 0.0238 | 10.0 | 540 | 0.0728 | 53.5532 |
65
+ | 0.02 | 11.0 | 594 | 0.0732 | 53.6670 |
66
+ | 0.0173 | 12.0 | 648 | 0.0731 | 54.2053 |
67
+ | 0.0144 | 13.0 | 702 | 0.0730 | 54.1739 |
68
+ | 0.0133 | 14.0 | 756 | 0.0730 | 54.0315 |
69
+ | 0.0121 | 15.0 | 810 | 0.0731 | 54.0839 |
70
+ | 0.012 | 16.0 | 864 | 0.0733 | 54.9269 |
71
+ | 0.0113 | 17.0 | 918 | 0.0734 | 54.2207 |
72
+ | 0.0113 | 18.0 | 972 | 0.0735 | 54.6988 |
73
+ | 0.0108 | 19.0 | 1026 | 0.0734 | 54.7575 |
74
+ | 0.0108 | 20.0 | 1080 | 0.0734 | 54.8194 |
75
 
76
 
77
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0b52e8d4e0d1add3ae7b84e815a9dac4b11580a1f825dca5afc31220bb775e41
3
  size 309965092
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4cd6a5ec93c9be420fba42c881cac4cfbd6c4617e3dc3033e347bd239fe58a5
3
  size 309965092
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7b0ac191ff134e53ded51676452edb7b656c1984043a48a1090f3d3e66f6edcf
3
  size 5496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28c715ccf566f1de094ea30d760a5ba49916fee950cccfd41056063c6f5b4f41
3
  size 5496