Hugging Face's logo

egerber1
/
sap_predictions_model

Text Classification
Transformers
Safetensors
xlm-roberta
Generated from Trainer
Model card Files
xet
Community
egerber1 commited on
Commit
073d409
·
verified ·
1 Parent(s): 4519796

End of training

Browse files
Files changed (2) hide show
  1. README.md +17 -82
  2. model.safetensors +1 -1
README.md CHANGED
@@ -19,9 +19,9 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on an unknown dataset.
21
  It achieves the following results on the evaluation set:
22
- - Loss: 1.7320
23
- - Accuracy: 0.5827
24
- - F1: 0.5204
25
 
26
  ## Model description
27
 
@@ -40,93 +40,28 @@ More information needed
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
- - learning_rate: 3e-05
44
- - train_batch_size: 128
45
- - eval_batch_size: 128
46
  - seed: 42
47
  - gradient_accumulation_steps: 2
48
- - total_train_batch_size: 256
49
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
  - lr_scheduler_type: linear
51
  - lr_scheduler_warmup_ratio: 0.1
52
- - num_epochs: 25
53
 
54
  ### Training results
55
 
56
- | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
57
- |:-------------:|:-------:|:-----:|:---------------:|:--------:|:------:|
58
- | 8.6779 | 0.3433 | 200 | 8.6561 | 0.0085 | 0.0004 |
59
- | 8.5189 | 0.6867 | 400 | 8.4598 | 0.0106 | 0.0002 |
60
- | 8.1366 | 1.0292 | 600 | 8.1444 | 0.0156 | 0.0023 |
61
- | 7.8642 | 1.3725 | 800 | 7.7846 | 0.0298 | 0.0040 |
62
- | 7.4619 | 1.7159 | 1000 | 7.3664 | 0.0557 | 0.0120 |
63
- | 7.0007 | 2.0584 | 1200 | 6.9190 | 0.0875 | 0.0271 |
64
- | 6.5573 | 2.4017 | 1400 | 6.4318 | 0.1219 | 0.0450 |
65
- | 6.113 | 2.7451 | 1600 | 5.9448 | 0.1626 | 0.0697 |
66
- | 5.6278 | 3.0876 | 1800 | 5.5131 | 0.1849 | 0.0852 |
67
- | 5.205 | 3.4309 | 2000 | 5.1058 | 0.2187 | 0.1143 |
68
- | 4.8582 | 3.7742 | 2200 | 4.7616 | 0.2429 | 0.1346 |
69
- | 4.5232 | 4.1167 | 2400 | 4.4627 | 0.2623 | 0.1532 |
70
- | 4.2714 | 4.4601 | 2600 | 4.2170 | 0.2752 | 0.1656 |
71
- | 4.0822 | 4.8034 | 2800 | 3.9888 | 0.2923 | 0.1855 |
72
- | 3.7851 | 5.1459 | 3000 | 3.7960 | 0.3083 | 0.2026 |
73
- | 3.6781 | 5.4893 | 3200 | 3.6184 | 0.3237 | 0.2192 |
74
- | 3.4707 | 5.8326 | 3400 | 3.4664 | 0.3451 | 0.2400 |
75
- | 3.2795 | 6.1751 | 3600 | 3.3434 | 0.3534 | 0.2476 |
76
- | 3.195 | 6.5185 | 3800 | 3.2099 | 0.3706 | 0.2685 |
77
- | 3.0989 | 6.8618 | 4000 | 3.1027 | 0.3754 | 0.2753 |
78
- | 2.9292 | 7.2043 | 4200 | 2.9945 | 0.3914 | 0.2943 |
79
- | 2.8712 | 7.5476 | 4400 | 2.9152 | 0.3951 | 0.2973 |
80
- | 2.8199 | 7.8910 | 4600 | 2.8185 | 0.4115 | 0.3158 |
81
- | 2.6806 | 8.2335 | 4800 | 2.7421 | 0.4202 | 0.3290 |
82
- | 2.5815 | 8.5768 | 5000 | 2.6696 | 0.4307 | 0.3380 |
83
- | 2.5966 | 8.9202 | 5200 | 2.6012 | 0.4329 | 0.3429 |
84
- | 2.4303 | 9.2627 | 5400 | 2.5515 | 0.4391 | 0.3504 |
85
- | 2.4154 | 9.6060 | 5600 | 2.4824 | 0.4506 | 0.3641 |
86
- | 2.3508 | 9.9494 | 5800 | 2.4303 | 0.4563 | 0.3701 |
87
- | 2.2627 | 10.2918 | 6000 | 2.3746 | 0.4668 | 0.3839 |
88
- | 2.2356 | 10.6352 | 6200 | 2.3306 | 0.4731 | 0.3903 |
89
- | 2.1811 | 10.9785 | 6400 | 2.2785 | 0.4838 | 0.4038 |
90
- | 2.1012 | 11.3210 | 6600 | 2.2387 | 0.4906 | 0.4118 |
91
- | 2.0846 | 11.6644 | 6800 | 2.2147 | 0.4855 | 0.4061 |
92
- | 2.0169 | 12.0069 | 7000 | 2.1568 | 0.5024 | 0.4276 |
93
- | 1.9768 | 12.3502 | 7200 | 2.1300 | 0.5096 | 0.4365 |
94
- | 1.9513 | 12.6936 | 7400 | 2.0956 | 0.5117 | 0.4408 |
95
- | 1.9021 | 13.0361 | 7600 | 2.0663 | 0.5167 | 0.4457 |
96
- | 1.8442 | 13.3794 | 7800 | 2.0414 | 0.5155 | 0.4438 |
97
- | 1.8643 | 13.7227 | 8000 | 2.0045 | 0.5253 | 0.4581 |
98
- | 1.7843 | 14.0652 | 8200 | 1.9779 | 0.5327 | 0.4676 |
99
- | 1.7562 | 14.4086 | 8400 | 1.9544 | 0.5372 | 0.4722 |
100
- | 1.7709 | 14.7519 | 8600 | 1.9323 | 0.5376 | 0.4735 |
101
- | 1.6962 | 15.0944 | 8800 | 1.9107 | 0.5441 | 0.4827 |
102
- | 1.6768 | 15.4378 | 9000 | 1.8918 | 0.5449 | 0.4843 |
103
- | 1.6797 | 15.7811 | 9200 | 1.8661 | 0.5523 | 0.4928 |
104
- | 1.6334 | 16.1236 | 9400 | 1.8505 | 0.5526 | 0.4947 |
105
- | 1.5961 | 16.4670 | 9600 | 1.8361 | 0.5532 | 0.4951 |
106
- | 1.5864 | 16.8103 | 9800 | 1.8088 | 0.5611 | 0.5064 |
107
- | 1.536 | 17.1528 | 10000 | 1.7966 | 0.5628 | 0.5065 |
108
- | 1.5433 | 17.4961 | 10200 | 1.7893 | 0.5667 | 0.5114 |
109
- | 1.5713 | 17.8395 | 10400 | 1.7664 | 0.5689 | 0.5145 |
110
- | 1.4823 | 18.1820 | 10600 | 1.7529 | 0.5754 | 0.5240 |
111
- | 1.5039 | 18.5253 | 10800 | 1.7361 | 0.5771 | 0.5255 |
112
- | 1.4963 | 18.8687 | 11000 | 1.7273 | 0.5811 | 0.5303 |
113
- | 1.4484 | 19.2112 | 11200 | 1.7186 | 0.5832 | 0.5327 |
114
- | 1.4299 | 19.5545 | 11400 | 1.7115 | 0.5823 | 0.5321 |
115
- | 1.4561 | 19.8979 | 11600 | 1.6955 | 0.5895 | 0.5418 |
116
- | 1.402 | 20.2403 | 11800 | 1.6911 | 0.5879 | 0.5388 |
117
- | 1.3908 | 20.5837 | 12000 | 1.6811 | 0.5913 | 0.5440 |
118
- | 1.3983 | 20.9270 | 12200 | 1.6687 | 0.5923 | 0.5458 |
119
- | 1.3577 | 21.2695 | 12400 | 1.6634 | 0.5944 | 0.5480 |
120
- | 1.3851 | 21.6129 | 12600 | 1.6626 | 0.5932 | 0.5474 |
121
- | 1.4019 | 21.9562 | 12800 | 1.6511 | 0.5989 | 0.5544 |
122
- | 1.3533 | 22.2987 | 13000 | 1.6502 | 0.5963 | 0.5506 |
123
- | 1.3468 | 22.6421 | 13200 | 1.6421 | 0.5989 | 0.5540 |
124
- | 1.3518 | 22.9854 | 13400 | 1.6371 | 0.6006 | 0.5552 |
125
- | 1.3326 | 23.3279 | 13600 | 1.6348 | 0.6013 | 0.5566 |
126
- | 1.3298 | 23.6712 | 13800 | 1.6316 | 0.6024 | 0.5581 |
127
- | 1.3309 | 24.0137 | 14000 | 1.6289 | 0.6038 | 0.5591 |
128
- | 1.3277 | 24.3571 | 14200 | 1.6295 | 0.6024 | 0.5581 |
129
- | 1.3267 | 24.7004 | 14400 | 1.6276 | 0.6028 | 0.5586 |
130
 
131
 
132
  ### Framework versions
 
19
 
20
  This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on an unknown dataset.
21
  It achieves the following results on the evaluation set:
22
+ - Loss: 6.3177
23
+ - Accuracy: 0.1599
24
+ - F1: 0.0713
25
 
26
  ## Model description
27
 
 
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
+ - learning_rate: 2e-05
44
+ - train_batch_size: 64
45
+ - eval_batch_size: 64
46
  - seed: 42
47
  - gradient_accumulation_steps: 2
48
+ - total_train_batch_size: 128
49
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
  - lr_scheduler_type: linear
51
  - lr_scheduler_warmup_ratio: 0.1
52
+ - num_epochs: 5
53
 
54
  ### Training results
55
 
56
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
57
+ |:-------------:|:------:|:----:|:---------------:|:--------:|:------:|
58
+ | 8.7072 | 0.6425 | 1000 | 8.5615 | 0.0156 | 0.0018 |
59
+ | 7.9463 | 1.2846 | 2000 | 7.8865 | 0.0445 | 0.0110 |
60
+ | 7.3576 | 1.9271 | 3000 | 7.2356 | 0.1019 | 0.0376 |
61
+ | 6.8566 | 2.5692 | 4000 | 6.7092 | 0.1424 | 0.0591 |
62
+ | 6.3983 | 3.2114 | 5000 | 6.3177 | 0.1599 | 0.0713 |
63
+ | 6.1392 | 3.8538 | 6000 | 6.0647 | 0.1756 | 0.0821 |
64
+ | 6.0378 | 4.4960 | 7000 | 5.9330 | 0.1819 | 0.0866 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
65
 
66
 
67
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ec8e9f6a01de6aad898d2ac9ad62183fc1063a141af256f77ec3498924680105
3
  size 1170366104
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17228c677f0def1e6cbca04b0db4ff02a17811c4e66be4b4f5d111aa35c98b13
3
  size 1170366104