marciovbarbosa
commited on
Commit
•
81485df
1
Parent(s):
796e369
update model card README.md
Browse files
README.md
CHANGED
@@ -19,7 +19,7 @@ model-index:
|
|
19 |
metrics:
|
20 |
- name: Bleu
|
21 |
type: bleu
|
22 |
-
value:
|
23 |
---
|
24 |
|
25 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -29,9 +29,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
29 |
|
30 |
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the wmt16 dataset.
|
31 |
It achieves the following results on the evaluation set:
|
32 |
-
- Loss: 1.
|
33 |
-
- Bleu:
|
34 |
-
- Gen Len: 17.
|
35 |
|
36 |
## Model description
|
37 |
|
@@ -56,62 +56,22 @@ The following hyperparameters were used during training:
|
|
56 |
- seed: 42
|
57 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
58 |
- lr_scheduler_type: linear
|
59 |
-
- num_epochs:
|
60 |
|
61 |
### Training results
|
62 |
|
63 |
-
| Training Loss | Epoch | Step
|
64 |
-
|
65 |
-
| No log | 1.0 | 272
|
66 |
-
| 2.
|
67 |
-
| 2.
|
68 |
-
| 2.
|
69 |
-
| 2.
|
70 |
-
| 2.
|
71 |
-
| 2.
|
72 |
-
| 2.
|
73 |
-
| 2.
|
74 |
-
| 2.
|
75 |
-
| 2.1501 | 11.0 | 2992 | 1.8738 | 10.3086 | 17.3258 |
|
76 |
-
| 2.1025 | 12.0 | 3264 | 1.8677 | 10.3851 | 17.3181 |
|
77 |
-
| 2.0638 | 13.0 | 3536 | 1.8623 | 10.489 | 17.3014 |
|
78 |
-
| 2.0638 | 14.0 | 3808 | 1.8574 | 10.4969 | 17.3204 |
|
79 |
-
| 2.034 | 15.0 | 4080 | 1.8528 | 10.7067 | 17.3178 |
|
80 |
-
| 2.034 | 16.0 | 4352 | 1.8493 | 10.6867 | 17.3408 |
|
81 |
-
| 1.9852 | 17.0 | 4624 | 1.8473 | 10.8333 | 17.3198 |
|
82 |
-
| 1.9852 | 18.0 | 4896 | 1.8429 | 10.8907 | 17.3001 |
|
83 |
-
| 1.9646 | 19.0 | 5168 | 1.8405 | 10.9049 | 17.3154 |
|
84 |
-
| 1.9646 | 20.0 | 5440 | 1.8385 | 10.9549 | 17.3124 |
|
85 |
-
| 1.9264 | 21.0 | 5712 | 1.8361 | 11.0046 | 17.3068 |
|
86 |
-
| 1.9264 | 22.0 | 5984 | 1.8338 | 11.1415 | 17.2954 |
|
87 |
-
| 1.9161 | 23.0 | 6256 | 1.8333 | 11.1041 | 17.2938 |
|
88 |
-
| 1.882 | 24.0 | 6528 | 1.8323 | 11.0801 | 17.2651 |
|
89 |
-
| 1.882 | 25.0 | 6800 | 1.8309 | 11.157 | 17.2921 |
|
90 |
-
| 1.8751 | 26.0 | 7072 | 1.8290 | 11.1713 | 17.2951 |
|
91 |
-
| 1.8751 | 27.0 | 7344 | 1.8279 | 11.2006 | 17.2861 |
|
92 |
-
| 1.8425 | 28.0 | 7616 | 1.8267 | 11.1761 | 17.2658 |
|
93 |
-
| 1.8425 | 29.0 | 7888 | 1.8278 | 11.148 | 17.2841 |
|
94 |
-
| 1.8306 | 30.0 | 8160 | 1.8261 | 11.1765 | 17.2748 |
|
95 |
-
| 1.8306 | 31.0 | 8432 | 1.8255 | 11.2723 | 17.2454 |
|
96 |
-
| 1.8229 | 32.0 | 8704 | 1.8247 | 11.2715 | 17.2621 |
|
97 |
-
| 1.8229 | 33.0 | 8976 | 1.8231 | 11.2896 | 17.2698 |
|
98 |
-
| 1.7975 | 34.0 | 9248 | 1.8245 | 11.322 | 17.2491 |
|
99 |
-
| 1.7919 | 35.0 | 9520 | 1.8238 | 11.3854 | 17.2711 |
|
100 |
-
| 1.7919 | 36.0 | 9792 | 1.8237 | 11.3304 | 17.2634 |
|
101 |
-
| 1.7781 | 37.0 | 10064 | 1.8225 | 11.3184 | 17.2644 |
|
102 |
-
| 1.7781 | 38.0 | 10336 | 1.8230 | 11.3382 | 17.2651 |
|
103 |
-
| 1.7819 | 39.0 | 10608 | 1.8228 | 11.3656 | 17.2658 |
|
104 |
-
| 1.7819 | 40.0 | 10880 | 1.8221 | 11.3934 | 17.2544 |
|
105 |
-
| 1.7592 | 41.0 | 11152 | 1.8223 | 11.3625 | 17.2421 |
|
106 |
-
| 1.7592 | 42.0 | 11424 | 1.8221 | 11.4068 | 17.2511 |
|
107 |
-
| 1.7529 | 43.0 | 11696 | 1.8224 | 11.4199 | 17.2541 |
|
108 |
-
| 1.7529 | 44.0 | 11968 | 1.8224 | 11.4051 | 17.2561 |
|
109 |
-
| 1.7482 | 45.0 | 12240 | 1.8223 | 11.4195 | 17.2504 |
|
110 |
-
| 1.7461 | 46.0 | 12512 | 1.8220 | 11.3873 | 17.2497 |
|
111 |
-
| 1.7461 | 47.0 | 12784 | 1.8220 | 11.4214 | 17.2431 |
|
112 |
-
| 1.739 | 48.0 | 13056 | 1.8218 | 11.3972 | 17.2441 |
|
113 |
-
| 1.739 | 49.0 | 13328 | 1.8219 | 11.3952 | 17.2457 |
|
114 |
-
| 1.7362 | 50.0 | 13600 | 1.8219 | 11.3921 | 17.2471 |
|
115 |
|
116 |
|
117 |
### Framework versions
|
19 |
metrics:
|
20 |
- name: Bleu
|
21 |
type: bleu
|
22 |
+
value: 9.2166
|
23 |
---
|
24 |
|
25 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
29 |
|
30 |
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the wmt16 dataset.
|
31 |
It achieves the following results on the evaluation set:
|
32 |
+
- Loss: 1.9417
|
33 |
+
- Bleu: 9.2166
|
34 |
+
- Gen Len: 17.3404
|
35 |
|
36 |
## Model description
|
37 |
|
56 |
- seed: 42
|
57 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
58 |
- lr_scheduler_type: linear
|
59 |
+
- num_epochs: 10
|
60 |
|
61 |
### Training results
|
62 |
|
63 |
+
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|
64 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|:-------:|
|
65 |
+
| No log | 1.0 | 272 | 2.1660 | 3.8515 | 17.6289 |
|
66 |
+
| 2.6678 | 2.0 | 544 | 2.0656 | 6.4422 | 17.4842 |
|
67 |
+
| 2.6678 | 3.0 | 816 | 2.0203 | 7.4348 | 17.3741 |
|
68 |
+
| 2.4316 | 4.0 | 1088 | 1.9926 | 8.0914 | 17.3658 |
|
69 |
+
| 2.4316 | 5.0 | 1360 | 1.9739 | 8.6535 | 17.3461 |
|
70 |
+
| 2.3307 | 6.0 | 1632 | 1.9603 | 8.8757 | 17.3768 |
|
71 |
+
| 2.3307 | 7.0 | 1904 | 1.9509 | 9.0744 | 17.3511 |
|
72 |
+
| 2.2945 | 8.0 | 2176 | 1.9466 | 9.1111 | 17.3418 |
|
73 |
+
| 2.2945 | 9.0 | 2448 | 1.9427 | 9.1969 | 17.3351 |
|
74 |
+
| 2.2666 | 10.0 | 2720 | 1.9417 | 9.2166 | 17.3404 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
75 |
|
76 |
|
77 |
### Framework versions
|