Hugging Face's logo

gayanin
/
pubmed-abs-sub-03

Text2Text Generation
Transformers
PyTorch
bart
Generated from Trainer
Model card Files Community
gayanin commited on
Commit
2121e08
·
1 Parent(s): 2af18c8

End of training

Browse files
Files changed (1) hide show
  1. README.md +19 -33
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/bart-base](https://huggingface.co/facebook/bart-base) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 0.1916
19
 
20
  ## Model description
21
 
@@ -35,8 +35,8 @@ More information needed
35
 
36
  The following hyperparameters were used during training:
37
  - learning_rate: 5e-05
38
- - train_batch_size: 16
39
- - eval_batch_size: 16
40
  - seed: 42
41
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
42
  - lr_scheduler_type: linear
@@ -45,36 +45,22 @@ The following hyperparameters were used during training:
45
 
46
  ### Training results
47
 
48
- | Training Loss | Epoch | Step | Validation Loss |
49
- |:-------------:|:-----:|:-----:|:---------------:|
50
- | 0.5622 | 0.11 | 500 | 0.4754 |
51
- | 0.4051 | 0.21 | 1000 | 0.3731 |
52
- | 0.3978 | 0.32 | 1500 | 0.3332 |
53
- | 0.3669 | 0.43 | 2000 | 0.3047 |
54
- | 0.3059 | 0.54 | 2500 | 0.2824 |
55
- | 0.294 | 0.64 | 3000 | 0.2632 |
56
- | 0.2774 | 0.75 | 3500 | 0.2579 |
57
- | 0.3134 | 0.86 | 4000 | 0.2408 |
58
- | 0.2865 | 0.96 | 4500 | 0.2345 |
59
- | 0.2103 | 1.07 | 5000 | 0.2320 |
60
- | 0.1974 | 1.18 | 5500 | 0.2289 |
61
- | 0.2026 | 1.28 | 6000 | 0.2303 |
62
- | 0.1931 | 1.39 | 6500 | 0.2155 |
63
- | 0.2027 | 1.5 | 7000 | 0.2167 |
64
- | 0.1977 | 1.61 | 7500 | 0.2120 |
65
- | 0.1836 | 1.71 | 8000 | 0.2100 |
66
- | 0.2204 | 1.82 | 8500 | 0.2043 |
67
- | 0.1731 | 1.93 | 9000 | 0.2009 |
68
- | 0.1479 | 2.03 | 9500 | 0.2004 |
69
- | 0.1277 | 2.14 | 10000 | 0.2040 |
70
- | 0.136 | 2.25 | 10500 | 0.2022 |
71
- | 0.1577 | 2.35 | 11000 | 0.1986 |
72
- | 0.1305 | 2.46 | 11500 | 0.1963 |
73
- | 0.159 | 2.57 | 12000 | 0.1936 |
74
- | 0.1603 | 2.68 | 12500 | 0.1939 |
75
- | 0.1195 | 2.78 | 13000 | 0.1926 |
76
- | 0.1203 | 2.89 | 13500 | 0.1922 |
77
- | 0.1437 | 3.0 | 14000 | 0.1916 |
78
 
79
 
80
  ### Framework versions
 
15
 
16
  This model is a fine-tuned version of [facebook/bart-base](https://huggingface.co/facebook/bart-base) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 0.1963
19
 
20
  ## Model description
21
 
 
35
 
36
  The following hyperparameters were used during training:
37
  - learning_rate: 5e-05
38
+ - train_batch_size: 32
39
+ - eval_batch_size: 32
40
  - seed: 42
41
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
42
  - lr_scheduler_type: linear
 
45
 
46
  ### Training results
47
 
48
+ | Training Loss | Epoch | Step | Validation Loss |
49
+ |:-------------:|:-----:|:----:|:---------------:|
50
+ | 0.4453 | 0.21 | 500 | 0.3989 |
51
+ | 0.3803 | 0.43 | 1000 | 0.3145 |
52
+ | 0.3093 | 0.64 | 1500 | 0.2789 |
53
+ | 0.3014 | 0.86 | 2000 | 0.2489 |
54
+ | 0.251 | 1.07 | 2500 | 0.2404 |
55
+ | 0.2304 | 1.28 | 3000 | 0.2312 |
56
+ | 0.2248 | 1.5 | 3500 | 0.2185 |
57
+ | 0.1944 | 1.71 | 4000 | 0.2134 |
58
+ | 0.1961 | 1.93 | 4500 | 0.2066 |
59
+ | 0.17 | 2.14 | 5000 | 0.2062 |
60
+ | 0.1606 | 2.35 | 5500 | 0.2021 |
61
+ | 0.1678 | 2.57 | 6000 | 0.1987 |
62
+ | 0.1535 | 2.78 | 6500 | 0.1978 |
63
+ | 0.1714 | 3.0 | 7000 | 0.1963 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
64
 
65
 
66
  ### Framework versions