Hugging Face's logo

Heather-Driver
/
distilbert-classn-LinearAlg-finetuned-span-width-2

Transformers
TensorBoard
Safetensors
distilbert
Generated from Trainer
Model card Files Metrics Community
Heather-Driver commited on
Commit
ebfae22
·
verified ·
1 Parent(s): 1007091

Classification Training

Browse files
Files changed (1) hide show
  1. README.md +33 -23
README.md CHANGED
@@ -21,11 +21,11 @@ should probably proofread and complete it, then remove this comment. -->
21
 
22
  This model is a fine-tuned version of [dslim/distilbert-NER](https://huggingface.co/dslim/distilbert-NER) on the None dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 1.0960
25
- - Accuracy: 0.7619
26
- - F1: 0.7595
27
- - Precision: 0.7766
28
- - Recall: 0.7619
29
 
30
  ## Model description
31
 
@@ -45,34 +45,44 @@ More information needed
45
 
46
  The following hyperparameters were used during training:
47
  - learning_rate: 5.314885705504048e-06
48
- - train_batch_size: 4
49
- - eval_batch_size: 4
50
  - seed: 42
51
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
52
  - lr_scheduler_type: linear
53
  - lr_scheduler_warmup_steps: 100
54
- - num_epochs: 15
55
  - mixed_precision_training: Native AMP
56
 
57
  ### Training results
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
60
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
61
- | 2.5453 | 1.0 | 73 | 2.5010 | 0.1270 | 0.0686 | 0.0848 | 0.1270 |
62
- | 2.4282 | 2.0 | 146 | 2.3230 | 0.1825 | 0.1798 | 0.2204 | 0.1825 |
63
- | 2.2774 | 3.0 | 219 | 2.1517 | 0.3968 | 0.3888 | 0.4138 | 0.3968 |
64
- | 2.1333 | 4.0 | 292 | 1.9495 | 0.5079 | 0.5086 | 0.5547 | 0.5079 |
65
- | 1.7985 | 5.0 | 365 | 1.7535 | 0.6111 | 0.6047 | 0.6447 | 0.6111 |
66
- | 1.734 | 6.0 | 438 | 1.6085 | 0.6508 | 0.6453 | 0.6606 | 0.6508 |
67
- | 1.5049 | 7.0 | 511 | 1.4771 | 0.6984 | 0.6938 | 0.7330 | 0.6984 |
68
- | 1.3924 | 8.0 | 584 | 1.3954 | 0.7222 | 0.7190 | 0.7298 | 0.7222 |
69
- | 1.2377 | 9.0 | 657 | 1.3074 | 0.7460 | 0.7421 | 0.7662 | 0.7460 |
70
- | 1.1501 | 10.0 | 730 | 1.2365 | 0.7381 | 0.7332 | 0.7406 | 0.7381 |
71
- | 0.9999 | 11.0 | 803 | 1.1857 | 0.7460 | 0.7417 | 0.7597 | 0.7460 |
72
- | 0.9838 | 12.0 | 876 | 1.1483 | 0.7540 | 0.7471 | 0.7600 | 0.7540 |
73
- | 0.9232 | 13.0 | 949 | 1.1151 | 0.7540 | 0.7485 | 0.7623 | 0.7540 |
74
- | 0.8965 | 14.0 | 1022 | 1.1026 | 0.7540 | 0.7485 | 0.7623 | 0.7540 |
75
- | 0.8475 | 15.0 | 1095 | 1.0960 | 0.7619 | 0.7595 | 0.7766 | 0.7619 |
 
 
 
 
 
 
 
 
 
 
76
 
77
 
78
  ### Framework versions
 
21
 
22
  This model is a fine-tuned version of [dslim/distilbert-NER](https://huggingface.co/dslim/distilbert-NER) on the None dataset.
23
  It achieves the following results on the evaluation set:
24
+ - Loss: 1.6148
25
+ - Accuracy: 0.6429
26
+ - F1: 0.6377
27
+ - Precision: 0.6600
28
+ - Recall: 0.6429
29
 
30
  ## Model description
31
 
 
45
 
46
  The following hyperparameters were used during training:
47
  - learning_rate: 5.314885705504048e-06
48
+ - train_batch_size: 32
49
+ - eval_batch_size: 32
50
  - seed: 42
51
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
52
  - lr_scheduler_type: linear
53
  - lr_scheduler_warmup_steps: 100
54
+ - num_epochs: 25
55
  - mixed_precision_training: Native AMP
56
 
57
  ### Training results
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
60
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
61
+ | No log | 1.0 | 10 | 2.4233 | 0.1429 | 0.0916 | 0.0802 | 0.1429 |
62
+ | No log | 2.0 | 20 | 2.4147 | 0.1429 | 0.0920 | 0.0804 | 0.1429 |
63
+ | No log | 3.0 | 30 | 2.4004 | 0.1667 | 0.1208 | 0.1178 | 0.1667 |
64
+ | No log | 4.0 | 40 | 2.3840 | 0.1587 | 0.1192 | 0.1064 | 0.1587 |
65
+ | No log | 5.0 | 50 | 2.3674 | 0.1667 | 0.1293 | 0.1176 | 0.1667 |
66
+ | No log | 6.0 | 60 | 2.3469 | 0.1905 | 0.1467 | 0.1312 | 0.1905 |
67
+ | No log | 7.0 | 70 | 2.3135 | 0.1984 | 0.1750 | 0.2072 | 0.1984 |
68
+ | No log | 8.0 | 80 | 2.2775 | 0.2143 | 0.1964 | 0.2092 | 0.2143 |
69
+ | No log | 9.0 | 90 | 2.2407 | 0.2302 | 0.2121 | 0.2260 | 0.2302 |
70
+ | No log | 10.0 | 100 | 2.1807 | 0.3016 | 0.2911 | 0.3070 | 0.3016 |
71
+ | No log | 11.0 | 110 | 2.1111 | 0.3413 | 0.3332 | 0.3604 | 0.3413 |
72
+ | No log | 12.0 | 120 | 2.0411 | 0.4048 | 0.3799 | 0.3880 | 0.4048 |
73
+ | No log | 13.0 | 130 | 1.9745 | 0.4444 | 0.4421 | 0.4658 | 0.4444 |
74
+ | No log | 14.0 | 140 | 1.9188 | 0.5238 | 0.5207 | 0.5477 | 0.5238 |
75
+ | No log | 15.0 | 150 | 1.8678 | 0.5635 | 0.5585 | 0.5787 | 0.5635 |
76
+ | No log | 16.0 | 160 | 1.8253 | 0.5794 | 0.5685 | 0.5826 | 0.5794 |
77
+ | No log | 17.0 | 170 | 1.7861 | 0.5952 | 0.5879 | 0.6104 | 0.5952 |
78
+ | No log | 18.0 | 180 | 1.7471 | 0.6032 | 0.5896 | 0.6028 | 0.6032 |
79
+ | No log | 19.0 | 190 | 1.7105 | 0.6190 | 0.6132 | 0.6383 | 0.6190 |
80
+ | No log | 20.0 | 200 | 1.6786 | 0.6270 | 0.6221 | 0.6484 | 0.6270 |
81
+ | No log | 21.0 | 210 | 1.6556 | 0.6508 | 0.6428 | 0.6627 | 0.6508 |
82
+ | No log | 22.0 | 220 | 1.6385 | 0.6349 | 0.6302 | 0.6532 | 0.6349 |
83
+ | No log | 23.0 | 230 | 1.6256 | 0.6349 | 0.6298 | 0.6546 | 0.6349 |
84
+ | No log | 24.0 | 240 | 1.6173 | 0.6429 | 0.6377 | 0.6600 | 0.6429 |
85
+ | No log | 25.0 | 250 | 1.6148 | 0.6429 | 0.6377 | 0.6600 | 0.6429 |
86
 
87
 
88
  ### Framework versions