Hugging Face's logo

tsilva
/
clinical-field-mapper-classification

Text Classification
Transformers
ONNX
Safetensors
gpt2
healthcare
column-normalization
distilgpt2
Eval Results (legacy)
Model card Files
xet
Community
tsilva commited on
Commit
014aeb1
·
verified ·
1 Parent(s): 092cde2

Push classification fine-tuned model

Browse files
Files changed (4) hide show
  1. README.md +56 -56
  2. config.json +22 -24
  3. model.safetensors +2 -2
  4. training_args.bin +1 -1
README.md CHANGED
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model is a fine-tuned version of [distilbert/distilgpt2](https://huggingface.co/distilbert/distilgpt2) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 0.6943
20
 
21
  ## Model description
22
 
@@ -35,9 +35,9 @@ More information needed
35
  ### Training hyperparameters
36
 
37
  The following hyperparameters were used during training:
38
- - learning_rate: 0.0005
39
- - train_batch_size: 512
40
- - eval_batch_size: 512
41
  - seed: 42
42
  - distributed_type: multi-GPU
43
  - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
@@ -48,58 +48,58 @@ The following hyperparameters were used during training:
48
 
49
  ### Training results
50
 
51
- | Training Loss | Epoch | Step | Validation Loss |
52
- |:-------------:|:-----:|:-----:|:---------------:|
53
- | 4.9577 | 1.0 | 486 | 3.5547 |
54
- | 3.1944 | 2.0 | 972 | 2.6211 |
55
- | 2.5386 | 3.0 | 1458 | 2.1523 |
56
- | 2.1691 | 4.0 | 1944 | 1.8564 |
57
- | 1.9271 | 5.0 | 2430 | 1.6533 |
58
- | 1.753 | 6.0 | 2916 | 1.5059 |
59
- | 1.6207 | 7.0 | 3402 | 1.3906 |
60
- | 1.5174 | 8.0 | 3888 | 1.2998 |
61
- | 1.433 | 9.0 | 4374 | 1.2236 |
62
- | 1.3652 | 10.0 | 4860 | 1.1621 |
63
- | 1.3062 | 11.0 | 5346 | 1.1113 |
64
- | 1.2586 | 12.0 | 5832 | 1.0674 |
65
- | 1.2175 | 13.0 | 6318 | 1.0283 |
66
- | 1.1805 | 14.0 | 6804 | 0.9956 |
67
- | 1.148 | 15.0 | 7290 | 0.9668 |
68
- | 1.1207 | 16.0 | 7776 | 0.9399 |
69
- | 1.0959 | 17.0 | 8262 | 0.9165 |
70
- | 1.0718 | 18.0 | 8748 | 0.8979 |
71
- | 1.0529 | 19.0 | 9234 | 0.8813 |
72
- | 1.036 | 20.0 | 9720 | 0.8652 |
73
- | 1.0196 | 21.0 | 10206 | 0.8511 |
74
- | 1.0057 | 22.0 | 10692 | 0.8398 |
75
- | 0.9922 | 23.0 | 11178 | 0.8252 |
76
- | 0.9802 | 24.0 | 11664 | 0.8135 |
77
- | 0.9686 | 25.0 | 12150 | 0.8022 |
78
- | 0.9621 | 26.0 | 12636 | 0.7974 |
79
- | 0.9494 | 27.0 | 13122 | 0.7852 |
80
- | 0.9415 | 28.0 | 13608 | 0.7788 |
81
- | 0.9358 | 29.0 | 14094 | 0.7700 |
82
- | 0.9276 | 30.0 | 14580 | 0.7642 |
83
- | 0.9188 | 31.0 | 15066 | 0.7622 |
84
- | 0.9128 | 32.0 | 15552 | 0.7510 |
85
- | 0.9076 | 33.0 | 16038 | 0.7471 |
86
- | 0.9022 | 34.0 | 16524 | 0.7397 |
87
- | 0.8975 | 35.0 | 17010 | 0.7397 |
88
- | 0.8938 | 36.0 | 17496 | 0.7344 |
89
- | 0.8866 | 37.0 | 17982 | 0.7271 |
90
- | 0.8842 | 38.0 | 18468 | 0.7197 |
91
- | 0.88 | 39.0 | 18954 | 0.7202 |
92
- | 0.8775 | 40.0 | 19440 | 0.7183 |
93
- | 0.8742 | 41.0 | 19926 | 0.7144 |
94
- | 0.8704 | 42.0 | 20412 | 0.7090 |
95
- | 0.8691 | 43.0 | 20898 | 0.7085 |
96
- | 0.8644 | 44.0 | 21384 | 0.7036 |
97
- | 0.8628 | 45.0 | 21870 | 0.7017 |
98
- | 0.8567 | 46.0 | 22356 | 0.7036 |
99
- | 0.8579 | 47.0 | 22842 | 0.6987 |
100
- | 0.8546 | 48.0 | 23328 | 0.6973 |
101
- | 0.8555 | 49.0 | 23814 | 0.6934 |
102
- | 0.8509 | 50.0 | 24300 | 0.6943 |
103
 
104
 
105
  ### Framework versions
 
16
 
17
  This model is a fine-tuned version of [distilbert/distilgpt2](https://huggingface.co/distilbert/distilgpt2) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.5728
20
 
21
  ## Model description
22
 
 
35
  ### Training hyperparameters
36
 
37
  The following hyperparameters were used during training:
38
+ - learning_rate: 0.005
39
+ - train_batch_size: 3072
40
+ - eval_batch_size: 1024
41
  - seed: 42
42
  - distributed_type: multi-GPU
43
  - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 
48
 
49
  ### Training results
50
 
51
+ | Training Loss | Epoch | Step | Validation Loss |
52
+ |:-------------:|:-----:|:----:|:---------------:|
53
+ | 8.921 | 1.0 | 9 | 7.9531 |
54
+ | 6.1949 | 2.0 | 18 | 5.2070 |
55
+ | 4.6541 | 3.0 | 27 | 4.0664 |
56
+ | 3.6725 | 4.0 | 36 | 3.2637 |
57
+ | 2.9716 | 5.0 | 45 | 2.6895 |
58
+ | 2.4881 | 6.0 | 54 | 2.3086 |
59
+ | 2.1437 | 7.0 | 63 | 2.0430 |
60
+ | 1.8888 | 8.0 | 72 | 1.8438 |
61
+ | 1.7087 | 9.0 | 81 | 1.6875 |
62
+ | 1.5463 | 10.0 | 90 | 1.5508 |
63
+ | 1.4124 | 11.0 | 99 | 1.4395 |
64
+ | 1.3009 | 12.0 | 108 | 1.3467 |
65
+ | 1.2058 | 13.0 | 117 | 1.2627 |
66
+ | 1.1296 | 14.0 | 126 | 1.2031 |
67
+ | 1.0643 | 15.0 | 135 | 1.1465 |
68
+ | 0.9962 | 16.0 | 144 | 1.0898 |
69
+ | 0.9387 | 17.0 | 153 | 1.0449 |
70
+ | 0.8919 | 18.0 | 162 | 1.0049 |
71
+ | 0.8522 | 19.0 | 171 | 0.9648 |
72
+ | 0.8161 | 20.0 | 180 | 0.9346 |
73
+ | 0.7829 | 21.0 | 189 | 0.8999 |
74
+ | 0.7489 | 22.0 | 198 | 0.8784 |
75
+ | 0.7249 | 23.0 | 207 | 0.8516 |
76
+ | 0.6945 | 24.0 | 216 | 0.8311 |
77
+ | 0.6763 | 25.0 | 225 | 0.8076 |
78
+ | 0.6529 | 26.0 | 234 | 0.7915 |
79
+ | 0.6309 | 27.0 | 243 | 0.7793 |
80
+ | 0.6121 | 28.0 | 252 | 0.7617 |
81
+ | 0.6009 | 29.0 | 261 | 0.7485 |
82
+ | 0.5841 | 30.0 | 270 | 0.7314 |
83
+ | 0.5598 | 31.0 | 279 | 0.7197 |
84
+ | 0.5529 | 32.0 | 288 | 0.7085 |
85
+ | 0.5378 | 33.0 | 297 | 0.6997 |
86
+ | 0.522 | 34.0 | 306 | 0.6846 |
87
+ | 0.5097 | 35.0 | 315 | 0.6650 |
88
+ | 0.5017 | 36.0 | 324 | 0.6602 |
89
+ | 0.4889 | 37.0 | 333 | 0.6567 |
90
+ | 0.4795 | 38.0 | 342 | 0.6426 |
91
+ | 0.4682 | 39.0 | 351 | 0.6396 |
92
+ | 0.4646 | 40.0 | 360 | 0.6323 |
93
+ | 0.4526 | 41.0 | 369 | 0.6226 |
94
+ | 0.4474 | 42.0 | 378 | 0.6133 |
95
+ | 0.4387 | 43.0 | 387 | 0.6040 |
96
+ | 0.432 | 44.0 | 396 | 0.6064 |
97
+ | 0.4258 | 45.0 | 405 | 0.6011 |
98
+ | 0.4194 | 46.0 | 414 | 0.5938 |
99
+ | 0.4113 | 47.0 | 423 | 0.5854 |
100
+ | 0.4076 | 48.0 | 432 | 0.5850 |
101
+ | 0.402 | 49.0 | 441 | 0.5703 |
102
+ | 0.3934 | 50.0 | 450 | 0.5728 |
103
 
104
 
105
  ### Framework versions
config.json CHANGED
@@ -335,18 +335,17 @@
335
  "323": "tumor_stage_n",
336
  "324": "tumor_stage_t",
337
  "325": "ultrasound_doppler_grade",
338
- "326": "unknown_field",
339
- "327": "urine_albumin_creatinine_ratio",
340
- "328": "visit_end_date",
341
- "329": "visit_id",
342
- "330": "visit_start_date",
343
- "331": "visit_type",
344
- "332": "wbc_count",
345
- "333": "weight",
346
- "334": "worsening_heart_failure_episode_order",
347
- "335": "worsening_heart_failure_event_type",
348
- "336": "worsening_heart_failure_flag",
349
- "337": "worsening_heart_failure_start_date"
350
  },
351
  "initializer_range": 0.02,
352
  "label2id": {
@@ -676,18 +675,17 @@
676
  "tumor_stage_n": 323,
677
  "tumor_stage_t": 324,
678
  "ultrasound_doppler_grade": 325,
679
- "unknown_field": 326,
680
- "urine_albumin_creatinine_ratio": 327,
681
- "visit_end_date": 328,
682
- "visit_id": 329,
683
- "visit_start_date": 330,
684
- "visit_type": 331,
685
- "wbc_count": 332,
686
- "weight": 333,
687
- "worsening_heart_failure_episode_order": 334,
688
- "worsening_heart_failure_event_type": 335,
689
- "worsening_heart_failure_flag": 336,
690
- "worsening_heart_failure_start_date": 337
691
  },
692
  "layer_norm_epsilon": 1e-05,
693
  "model_type": "gpt2",
 
335
  "323": "tumor_stage_n",
336
  "324": "tumor_stage_t",
337
  "325": "ultrasound_doppler_grade",
338
+ "326": "urine_albumin_creatinine_ratio",
339
+ "327": "visit_end_date",
340
+ "328": "visit_id",
341
+ "329": "visit_start_date",
342
+ "330": "visit_type",
343
+ "331": "wbc_count",
344
+ "332": "weight",
345
+ "333": "worsening_heart_failure_episode_order",
346
+ "334": "worsening_heart_failure_event_type",
347
+ "335": "worsening_heart_failure_flag",
348
+ "336": "worsening_heart_failure_start_date"
 
349
  },
350
  "initializer_range": 0.02,
351
  "label2id": {
 
675
  "tumor_stage_n": 323,
676
  "tumor_stage_t": 324,
677
  "ultrasound_doppler_grade": 325,
678
+ "urine_albumin_creatinine_ratio": 326,
679
+ "visit_end_date": 327,
680
+ "visit_id": 328,
681
+ "visit_start_date": 329,
682
+ "visit_type": 330,
683
+ "wbc_count": 331,
684
+ "weight": 332,
685
+ "worsening_heart_failure_episode_order": 333,
686
+ "worsening_heart_failure_event_type": 334,
687
+ "worsening_heart_failure_flag": 335,
688
+ "worsening_heart_failure_start_date": 336
 
689
  },
690
  "layer_norm_epsilon": 1e-05,
691
  "model_type": "gpt2",
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:db1daabaa4834518963e72874d49c0f55e880948963fbdf5f47a38fb01bd6d9b
3
- size 164353504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3da2a50385230c5fbc8a06c831e6d9923d39a4e7d1a1622f21c93d585cfb33d1
3
+ size 164351968
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:633ad78c5a738c48af0a15fcca5bb413c8061991604c00897f41351ae8c70650
3
  size 7416
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20f6401f4171ba69319ed6ea406c608618cc9b7cff9c69647950762dfe6d6707
3
  size 7416