C0uchP0tat0 commited on
Commit
d2c772a
1 Parent(s): 07263fc

End of training

Browse files
README.md CHANGED
@@ -12,9 +12,9 @@ should probably proofread and complete it, then remove this comment. -->
12
 
13
  # my_rugpt3medium_finetune
14
 
15
- This model is a fine-tuned version of [ai-forever/rugpt3medium_based_on_gpt2](https://huggingface.co/ai-forever/rugpt3medium_based_on_gpt2) on the None dataset.
16
  It achieves the following results on the evaluation set:
17
- - Loss: 2.9269
18
 
19
  ## Model description
20
 
@@ -42,28 +42,88 @@ The following hyperparameters were used during training:
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: cosine
44
  - lr_scheduler_warmup_steps: 1000
45
- - num_epochs: 25
46
  - mixed_precision_training: Native AMP
47
 
48
  ### Training results
49
 
50
  | Training Loss | Epoch | Step | Validation Loss |
51
  |:-------------:|:-----:|:----:|:---------------:|
52
- | 3.601 | 1.6 | 25 | 3.6157 |
53
- | 3.601 | 3.19 | 50 | 3.6010 |
54
- | 3.5542 | 4.79 | 75 | 3.5621 |
55
- | 3.5309 | 6.38 | 100 | 3.5117 |
56
- | 3.496 | 7.98 | 125 | 3.4615 |
57
- | 3.446 | 9.57 | 150 | 3.4173 |
58
- | 3.34 | 11.17 | 175 | 3.3699 |
59
- | 3.3581 | 12.77 | 200 | 3.3214 |
60
- | 3.3136 | 14.36 | 225 | 3.2743 |
61
- | 3.214 | 15.96 | 250 | 3.2227 |
62
- | 3.2098 | 17.55 | 275 | 3.1738 |
63
- | 3.1348 | 19.15 | 300 | 3.1153 |
64
- | 3.0931 | 20.74 | 325 | 3.0561 |
65
- | 3.0383 | 22.34 | 350 | 2.9922 |
66
- | 2.9739 | 23.94 | 375 | 2.9269 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
67
 
68
 
69
  ### Framework versions
 
12
 
13
  # my_rugpt3medium_finetune
14
 
15
+ This model is a fine-tuned version of [ai-forever/rugpt3medium_based_on_gpt2](https://huggingface.co/ai-forever/rugpt3medium_based_on_gpt2) on an unknown dataset.
16
  It achieves the following results on the evaluation set:
17
+ - Loss: 0.9955
18
 
19
  ## Model description
20
 
 
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: cosine
44
  - lr_scheduler_warmup_steps: 1000
45
+ - num_epochs: 35
46
  - mixed_precision_training: Native AMP
47
 
48
  ### Training results
49
 
50
  | Training Loss | Epoch | Step | Validation Loss |
51
  |:-------------:|:-----:|:----:|:---------------:|
52
+ | 3.5373 | 0.46 | 25 | 3.4828 |
53
+ | 3.5265 | 0.93 | 50 | 3.4708 |
54
+ | 3.478 | 1.39 | 75 | 3.4398 |
55
+ | 3.4851 | 1.85 | 100 | 3.3995 |
56
+ | 3.4407 | 2.31 | 125 | 3.3609 |
57
+ | 3.3731 | 2.78 | 150 | 3.3241 |
58
+ | 3.3584 | 3.24 | 175 | 3.2886 |
59
+ | 3.3267 | 3.7 | 200 | 3.2540 |
60
+ | 3.3043 | 4.17 | 225 | 3.2200 |
61
+ | 3.229 | 4.63 | 250 | 3.1853 |
62
+ | 3.2618 | 5.09 | 275 | 3.1508 |
63
+ | 3.1823 | 5.56 | 300 | 3.1164 |
64
+ | 3.172 | 6.02 | 325 | 3.0779 |
65
+ | 3.1354 | 6.48 | 350 | 3.0395 |
66
+ | 3.0899 | 6.94 | 375 | 2.9987 |
67
+ | 3.0741 | 7.41 | 400 | 2.9577 |
68
+ | 3.009 | 7.87 | 425 | 2.9140 |
69
+ | 2.9598 | 8.33 | 450 | 2.8737 |
70
+ | 2.9187 | 8.8 | 475 | 2.8294 |
71
+ | 2.9378 | 9.26 | 500 | 2.7842 |
72
+ | 2.8396 | 9.72 | 525 | 2.7374 |
73
+ | 2.8608 | 10.19 | 550 | 2.6889 |
74
+ | 2.7296 | 10.65 | 575 | 2.6405 |
75
+ | 2.7452 | 11.11 | 600 | 2.5926 |
76
+ | 2.6882 | 11.57 | 625 | 2.5389 |
77
+ | 2.6463 | 12.04 | 650 | 2.4893 |
78
+ | 2.572 | 12.5 | 675 | 2.4356 |
79
+ | 2.5384 | 12.96 | 700 | 2.3788 |
80
+ | 2.5246 | 13.43 | 725 | 2.3296 |
81
+ | 2.4055 | 13.89 | 750 | 2.2747 |
82
+ | 2.3759 | 14.35 | 775 | 2.2155 |
83
+ | 2.3351 | 14.81 | 800 | 2.1606 |
84
+ | 2.286 | 15.28 | 825 | 2.1061 |
85
+ | 2.2694 | 15.74 | 850 | 2.0504 |
86
+ | 2.1745 | 16.2 | 875 | 1.9967 |
87
+ | 2.1053 | 16.67 | 900 | 1.9411 |
88
+ | 2.1184 | 17.13 | 925 | 1.8878 |
89
+ | 2.0107 | 17.59 | 950 | 1.8362 |
90
+ | 2.027 | 18.06 | 975 | 1.7854 |
91
+ | 1.9153 | 18.52 | 1000 | 1.7304 |
92
+ | 1.9267 | 18.98 | 1025 | 1.6854 |
93
+ | 1.8131 | 19.44 | 1050 | 1.6331 |
94
+ | 1.8405 | 19.91 | 1075 | 1.5839 |
95
+ | 1.7294 | 20.37 | 1100 | 1.5370 |
96
+ | 1.7154 | 20.83 | 1125 | 1.4971 |
97
+ | 1.6573 | 21.3 | 1150 | 1.4476 |
98
+ | 1.6391 | 21.76 | 1175 | 1.4130 |
99
+ | 1.5497 | 22.22 | 1200 | 1.3727 |
100
+ | 1.5194 | 22.69 | 1225 | 1.3378 |
101
+ | 1.535 | 23.15 | 1250 | 1.3000 |
102
+ | 1.4514 | 23.61 | 1275 | 1.2714 |
103
+ | 1.4711 | 24.07 | 1300 | 1.2388 |
104
+ | 1.4105 | 24.54 | 1325 | 1.2136 |
105
+ | 1.4202 | 25.0 | 1350 | 1.1890 |
106
+ | 1.3351 | 25.46 | 1375 | 1.1679 |
107
+ | 1.3575 | 25.93 | 1400 | 1.1440 |
108
+ | 1.2882 | 26.39 | 1425 | 1.1202 |
109
+ | 1.3378 | 26.85 | 1450 | 1.1074 |
110
+ | 1.3094 | 27.31 | 1475 | 1.0864 |
111
+ | 1.2793 | 27.78 | 1500 | 1.0743 |
112
+ | 1.2377 | 28.24 | 1525 | 1.0626 |
113
+ | 1.2693 | 28.7 | 1550 | 1.0468 |
114
+ | 1.2157 | 29.17 | 1575 | 1.0368 |
115
+ | 1.2007 | 29.63 | 1600 | 1.0263 |
116
+ | 1.2376 | 30.09 | 1625 | 1.0221 |
117
+ | 1.2216 | 30.56 | 1650 | 1.0136 |
118
+ | 1.1923 | 31.02 | 1675 | 1.0102 |
119
+ | 1.2143 | 31.48 | 1700 | 1.0039 |
120
+ | 1.1764 | 31.94 | 1725 | 1.0014 |
121
+ | 1.1654 | 32.41 | 1750 | 0.9990 |
122
+ | 1.2031 | 32.87 | 1775 | 0.9976 |
123
+ | 1.1952 | 33.33 | 1800 | 0.9965 |
124
+ | 1.1852 | 33.8 | 1825 | 0.9961 |
125
+ | 1.1737 | 34.26 | 1850 | 0.9959 |
126
+ | 1.1609 | 34.72 | 1875 | 0.9955 |
127
 
128
 
129
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:720095fc544006d19dd75064e06645c36f75a32cf3956b55faab42c04a85a284
3
  size 1423517184
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2cc3c3c6dc63dea0e60a4f481edcbe973fad68350d039223ce490a333f92225a
3
  size 1423517184
runs/Dec28_09-55-46_cd8a8367193a/events.out.tfevents.1703757350.cd8a8367193a.207.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:792683a36556972dec13aca2a22237e80d0244162719553d1c37c6ebb6e3d61b
3
- size 21636
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28a91ff56dcfd391f3205ff84df870f3a887ce4b07558a2e26fb976d7765d70a
3
+ size 36970