Hugging Face's logo

MaryaAI
/
opus-mt-ar-en-finetunedTanzil-v6-ar-to-en

Text2Text Generation
Transformers
TensorFlow
marian
generated_from_keras_callback
Inference Endpoints
Model card Files Community
MaryaAI commited on
Commit
26dbf35
•
1 Parent(s): c05fd7d

add model

Browse files
Files changed (2) hide show
  1. README.md +23 -19
  2. tf_model.h5 +1 -1
README.md CHANGED
@@ -14,14 +14,14 @@ probably proofread and complete it, then remove this comment. -->
14
 
15
  This model is a fine-tuned version of [Helsinki-NLP/opus-mt-ar-en](https://huggingface.co/Helsinki-NLP/opus-mt-ar-en) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
- - Train Loss: 0.8405
18
- - Validation Loss: 0.9594
19
- - Train Rouge1: 58.4345
20
- - Train Rouge2: 34.4502
21
- - Train Rougel: 54.6634
22
- - Train Rougelsum: 54.9469
23
- - Train Gen Len: 96.98
24
- - Epoch: 10
25
 
26
  ## Model description
27
 
@@ -47,17 +47,21 @@ The following hyperparameters were used during training:
47
 
48
  | Train Loss | Validation Loss | Train Rouge1 | Train Rouge2 | Train Rougel | Train Rougelsum | Train Gen Len | Epoch |
49
  |:----------:|:---------------:|:------------:|:------------:|:------------:|:---------------:|:-------------:|:-----:|
50
- | 1.7341 | 1.4426 | 55.0505 | 31.6295 | 51.3925 | 51.8558 | 76.44 | 0 |
51
- | 1.4329 | 1.2987 | 56.7672 | 33.8519 | 53.2796 | 53.6077 | 75.905 | 1 |
52
- | 1.2849 | 1.2132 | 59.1926 | 36.8577 | 55.8615 | 56.2321 | 94.415 | 2 |
53
- | 1.1854 | 1.1560 | 59.5654 | 36.0992 | 56.2438 | 56.6259 | 109.725 | 3 |
54
- | 1.1092 | 1.1101 | 59.2882 | 36.3247 | 56.1751 | 56.6098 | 110.7 | 4 |
55
- | 1.0478 | 1.0738 | 59.3702 | 35.9485 | 56.2157 | 56.6418 | 100.835 | 5 |
56
- | 0.9957 | 1.0464 | 59.3582 | 35.5831 | 55.8989 | 56.3337 | 103.575 | 6 |
57
- | 0.9503 | 1.0190 | 59.2327 | 35.7054 | 55.6769 | 56.0504 | 97.95 | 7 |
58
- | 0.9099 | 0.9959 | 58.9992 | 34.9965 | 55.5582 | 55.8486 | 97.05 | 8 |
59
- | 0.8735 | 0.9776 | 58.9665 | 35.1024 | 55.6433 | 55.8640 | 100.195 | 9 |
60
- | 0.8405 | 0.9594 | 58.4345 | 34.4502 | 54.6634 | 54.9469 | 96.98 | 10 |
 
 
 
 
61
 
62
 
63
  ### Framework versions
 
14
 
15
  This model is a fine-tuned version of [Helsinki-NLP/opus-mt-ar-en](https://huggingface.co/Helsinki-NLP/opus-mt-ar-en) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
+ - Train Loss: 0.6863
18
+ - Validation Loss: 0.8789
19
+ - Train Rouge1: 56.4999
20
+ - Train Rouge2: 33.3152
21
+ - Train Rougel: 52.7922
22
+ - Train Rougelsum: 53.4004
23
+ - Train Gen Len: 65.66
24
+ - Epoch: 14
25
 
26
  ## Model description
27
 
 
47
 
48
  | Train Loss | Validation Loss | Train Rouge1 | Train Rouge2 | Train Rougel | Train Rougelsum | Train Gen Len | Epoch |
49
  |:----------:|:---------------:|:------------:|:------------:|:------------:|:---------------:|:-------------:|:-----:|
50
+ | 1.2958 | 1.2191 | 55.4125 | 33.4250 | 52.1444 | 52.4427 | 109.29 | 0 |
51
+ | 1.1928 | 1.1607 | 57.1190 | 35.0261 | 53.9195 | 54.3028 | 103.37 | 1 |
52
+ | 1.1152 | 1.1133 | 56.7848 | 34.9767 | 53.6885 | 54.0708 | 81.705 | 2 |
53
+ | 1.0531 | 1.0795 | 56.5101 | 35.1214 | 53.2131 | 53.4955 | 104.285 | 3 |
54
+ | 0.9999 | 1.0493 | 57.7130 | 35.7997 | 54.7519 | 55.0195 | 98.01 | 4 |
55
+ | 0.9544 | 1.0233 | 57.4034 | 35.3118 | 54.2593 | 54.6092 | 56.94 | 5 |
56
+ | 0.9135 | 1.0009 | 56.7008 | 35.3027 | 53.6876 | 53.8885 | 94.855 | 6 |
57
+ | 0.8770 | 0.9793 | 56.6660 | 34.4890 | 53.1461 | 53.5412 | 93.33 | 7 |
58
+ | 0.8434 | 0.9624 | 55.3198 | 33.4441 | 52.3099 | 52.7027 | 94.63 | 8 |
59
+ | 0.8135 | 0.9468 | 55.2723 | 33.4791 | 51.9936 | 52.2751 | 69.765 | 9 |
60
+ | 0.7842 | 0.9335 | 55.9241 | 33.1444 | 52.5591 | 52.8279 | 63.895 | 10 |
61
+ | 0.7572 | 0.9170 | 57.7109 | 34.8199 | 54.0893 | 54.6004 | 75.57 | 11 |
62
+ | 0.7322 | 0.9009 | 55.7067 | 33.2829 | 52.2642 | 52.5643 | 96.515 | 12 |
63
+ | 0.7084 | 0.8912 | 55.2352 | 32.1104 | 51.6885 | 52.1647 | 91.595 | 13 |
64
+ | 0.6863 | 0.8789 | 56.4999 | 33.3152 | 52.7922 | 53.4004 | 65.66 | 14 |
65
 
66
 
67
  ### Framework versions
tf_model.h5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5c19e1e49298b8af0992b7c32715333deddba68431ee4ef674ec1fdc03bc111e
3
  size 307854560
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:292f18d4a336ba706b061aa97a48e2187f7143943466a7d56f038e90289ea84f
3
  size 307854560