Hugging Face's logo

sourabsb
/
english_kumaoni-mbart_large

PEFT
Safetensors
Generated from Trainer
Model card Files
xet
Community
sourabsb commited on
Commit
e7af2f1
·
verified ·
1 Parent(s): de56ebe

Pushing best model after training.

Browse files
Files changed (1) hide show
  1. README.md +19 -11
README.md CHANGED
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  This model is a fine-tuned version of [facebook/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 4.2736
19
 
20
  ## Model description
21
 
@@ -40,20 +40,28 @@ The following hyperparameters were used during training:
40
  - seed: 42
41
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
42
  - lr_scheduler_type: linear
43
- - num_epochs: 7
44
  - mixed_precision_training: Native AMP
45
 
46
  ### Training results
47
 
48
- | Training Loss | Epoch | Step | Validation Loss |
49
- |:-------------:|:-----:|:----:|:---------------:|
50
- | 5.1787 | 1.0 | 1132 | 4.9447 |
51
- | 4.8369 | 2.0 | 2264 | 4.6666 |
52
- | 4.6559 | 3.0 | 3396 | 4.5080 |
53
- | 4.5016 | 4.0 | 4528 | 4.4247 |
54
- | 4.4583 | 5.0 | 5660 | 4.3958 |
55
- | 4.4426 | 6.0 | 6792 | 4.2922 |
56
- | 4.3882 | 7.0 | 7924 | 4.2736 |
 
 
 
 
 
 
 
 
57
 
58
 
59
  ### Framework versions
 
15
 
16
  This model is a fine-tuned version of [facebook/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 3.8143
19
 
20
  ## Model description
21
 
 
40
  - seed: 42
41
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
42
  - lr_scheduler_type: linear
43
+ - num_epochs: 15
44
  - mixed_precision_training: Native AMP
45
 
46
  ### Training results
47
 
48
+ | Training Loss | Epoch | Step | Validation Loss |
49
+ |:-------------:|:-----:|:-----:|:---------------:|
50
+ | 5.1837 | 1.0 | 1133 | 5.0117 |
51
+ | 4.78 | 2.0 | 2266 | 4.6802 |
52
+ | 4.5691 | 3.0 | 3399 | 4.4787 |
53
+ | 4.3752 | 4.0 | 4532 | 4.3266 |
54
+ | 4.2656 | 5.0 | 5665 | 4.2151 |
55
+ | 4.1741 | 6.0 | 6798 | 4.1338 |
56
+ | 4.0978 | 7.0 | 7931 | 4.0474 |
57
+ | 3.9998 | 8.0 | 9064 | 3.9952 |
58
+ | 3.9682 | 9.0 | 10197 | 3.9317 |
59
+ | 3.9121 | 10.0 | 11330 | 3.9082 |
60
+ | 3.8589 | 11.0 | 12463 | 3.8674 |
61
+ | 3.8396 | 12.0 | 13596 | 3.8455 |
62
+ | 3.8058 | 13.0 | 14729 | 3.8335 |
63
+ | 3.7847 | 14.0 | 15862 | 3.8232 |
64
+ | 3.7661 | 15.0 | 16995 | 3.8143 |
65
 
66
 
67
  ### Framework versions