Hugging Face's logo

xshubhamx
/
InLegalBERT

Text Classification
Transformers
TensorBoard
Safetensors
bert
Generated from Trainer
Model card Files
xet
Metrics Community
xshubhamx commited on
Commit
cae5f9e
·
verified ·
1 Parent(s): e47107a

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. training_checkpoints/checkpoint-9002/config.json +63 -0
  2. training_checkpoints/checkpoint-9002/model.safetensors +3 -0
  3. training_checkpoints/checkpoint-9002/optimizer.pt +3 -0
  4. training_checkpoints/checkpoint-9002/rng_state.pth +3 -0
  5. training_checkpoints/checkpoint-9002/scheduler.pt +3 -0
  6. training_checkpoints/checkpoint-9002/trainer_state.json +455 -0
  7. training_checkpoints/checkpoint-9002/training_args.bin +3 -0
training_checkpoints/checkpoint-9002/config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "law-ai/InLegalBERT",
3
+ "architectures": [
4
+ "BertForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_ids": 0,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "Issue",
15
+ "1": "Court Discourse",
16
+ "2": "Conclusion",
17
+ "3": "Precedent Analysis",
18
+ "4": "Section Analysis",
19
+ "5": "Argument by Petitioner",
20
+ "6": "Fact",
21
+ "7": "Argument by Respondent",
22
+ "8": "Ratio",
23
+ "9": "Appellant",
24
+ "10": "Respondent",
25
+ "11": "Argument by Appellant",
26
+ "12": "Petitioner",
27
+ "13": "Judge",
28
+ "14": "Argument by Defendant"
29
+ },
30
+ "initializer_range": 0.02,
31
+ "intermediate_size": 3072,
32
+ "label2id": {
33
+ "Appellant": 9,
34
+ "Argument by Appellant": 11,
35
+ "Argument by Defendant": 14,
36
+ "Argument by Petitioner": 5,
37
+ "Argument by Respondent": 7,
38
+ "Conclusion": 2,
39
+ "Court Discourse": 1,
40
+ "Fact": 6,
41
+ "Issue": 0,
42
+ "Judge": 13,
43
+ "Petitioner": 12,
44
+ "Precedent Analysis": 3,
45
+ "Ratio": 8,
46
+ "Respondent": 10,
47
+ "Section Analysis": 4
48
+ },
49
+ "layer_norm_eps": 1e-12,
50
+ "max_position_embeddings": 512,
51
+ "model_type": "bert",
52
+ "num_attention_heads": 12,
53
+ "num_hidden_layers": 12,
54
+ "output_past": true,
55
+ "pad_token_id": 0,
56
+ "position_embedding_type": "absolute",
57
+ "problem_type": "single_label_classification",
58
+ "torch_dtype": "float32",
59
+ "transformers_version": "4.38.2",
60
+ "type_vocab_size": 2,
61
+ "use_cache": true,
62
+ "vocab_size": 30522
63
+ }
training_checkpoints/checkpoint-9002/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:098b35dceb782b55013a517bfc6d004ff56bc166b467247f2452e618935794c9
3
+ size 437998636
training_checkpoints/checkpoint-9002/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0f920b2b2ea120facb6739106559ed0f13c1c755e2ab531a919f756c6a2cdb3
3
+ size 876118266
training_checkpoints/checkpoint-9002/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:665de590dae01c642ec85fbd3e83ddd5fcaedc475031ed6788afb25c0b5cd310
3
+ size 14244
training_checkpoints/checkpoint-9002/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78bd0c494fea2f6752fe12805c48143182753af2a1fa6aac546cfd20a82fa8ec
3
+ size 1064
training_checkpoints/checkpoint-9002/trainer_state.json ADDED
@@ -0,0 +1,455 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7985673319510695,
3
+ "best_model_checkpoint": "InLegalBERT/checkpoint-3215",
4
+ "epoch": 14.0,
5
+ "eval_steps": 500,
6
+ "global_step": 9002,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.78,
13
+ "grad_norm": 6.225058078765869,
14
+ "learning_rate": 4.742871954380509e-05,
15
+ "loss": 1.065,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 1.0,
20
+ "eval_accuracy": 0.7993803253292022,
21
+ "eval_f1_macro": 0.6028764374506596,
22
+ "eval_f1_micro": 0.7993803253292022,
23
+ "eval_f1_weighted": 0.780434475079005,
24
+ "eval_loss": 0.6395189166069031,
25
+ "eval_macro_fpr": 0.01853296480715729,
26
+ "eval_macro_sensitivity": 0.6307714529979154,
27
+ "eval_macro_specificity": 0.9847175193097543,
28
+ "eval_precision": 0.7817573111259606,
29
+ "eval_precision_macro": 0.6194286814679265,
30
+ "eval_recall": 0.7993803253292022,
31
+ "eval_recall_macro": 0.6307714529979154,
32
+ "eval_runtime": 29.4696,
33
+ "eval_samples_per_second": 43.808,
34
+ "eval_steps_per_second": 5.497,
35
+ "eval_weighted_fpr": 0.017610661589719183,
36
+ "eval_weighted_sensitivity": 0.7993803253292022,
37
+ "eval_weighted_specificity": 0.9713824643171124,
38
+ "step": 643
39
+ },
40
+ {
41
+ "epoch": 1.56,
42
+ "grad_norm": 8.756747245788574,
43
+ "learning_rate": 4.483670295489891e-05,
44
+ "loss": 0.5866,
45
+ "step": 1000
46
+ },
47
+ {
48
+ "epoch": 2.0,
49
+ "eval_accuracy": 0.8187451587916343,
50
+ "eval_f1_macro": 0.7276084501581802,
51
+ "eval_f1_micro": 0.8187451587916342,
52
+ "eval_f1_weighted": 0.8152450677099363,
53
+ "eval_loss": 0.6906521916389465,
54
+ "eval_macro_fpr": 0.01610571025604924,
55
+ "eval_macro_sensitivity": 0.7365748487318147,
56
+ "eval_macro_specificity": 0.9863504064994671,
57
+ "eval_precision": 0.8198905458065566,
58
+ "eval_precision_macro": 0.7284980957351811,
59
+ "eval_recall": 0.8187451587916343,
60
+ "eval_recall_macro": 0.7365748487318147,
61
+ "eval_runtime": 57.1075,
62
+ "eval_samples_per_second": 22.606,
63
+ "eval_steps_per_second": 2.837,
64
+ "eval_weighted_fpr": 0.015566790846194785,
65
+ "eval_weighted_sensitivity": 0.8187451587916343,
66
+ "eval_weighted_specificity": 0.976510938700373,
67
+ "step": 1286
68
+ },
69
+ {
70
+ "epoch": 2.33,
71
+ "grad_norm": 8.241662979125977,
72
+ "learning_rate": 4.224468636599274e-05,
73
+ "loss": 0.4622,
74
+ "step": 1500
75
+ },
76
+ {
77
+ "epoch": 3.0,
78
+ "eval_accuracy": 0.8179705654531371,
79
+ "eval_f1_macro": 0.7282676953256336,
80
+ "eval_f1_micro": 0.817970565453137,
81
+ "eval_f1_weighted": 0.8150036178041801,
82
+ "eval_loss": 0.8055915236473083,
83
+ "eval_macro_fpr": 0.016168738933500626,
84
+ "eval_macro_sensitivity": 0.737597935006265,
85
+ "eval_macro_specificity": 0.9862931306285473,
86
+ "eval_precision": 0.8137371356291269,
87
+ "eval_precision_macro": 0.7226669118270935,
88
+ "eval_recall": 0.8179705654531371,
89
+ "eval_recall_macro": 0.737597935006265,
90
+ "eval_runtime": 56.5581,
91
+ "eval_samples_per_second": 22.826,
92
+ "eval_steps_per_second": 2.864,
93
+ "eval_weighted_fpr": 0.01564684732671949,
94
+ "eval_weighted_sensitivity": 0.8179705654531371,
95
+ "eval_weighted_specificity": 0.9764263939750704,
96
+ "step": 1929
97
+ },
98
+ {
99
+ "epoch": 3.11,
100
+ "grad_norm": 5.718050003051758,
101
+ "learning_rate": 3.9652669777086575e-05,
102
+ "loss": 0.346,
103
+ "step": 2000
104
+ },
105
+ {
106
+ "epoch": 3.89,
107
+ "grad_norm": 28.698440551757812,
108
+ "learning_rate": 3.7060653188180404e-05,
109
+ "loss": 0.2398,
110
+ "step": 2500
111
+ },
112
+ {
113
+ "epoch": 4.0,
114
+ "eval_accuracy": 0.8171959721146398,
115
+ "eval_f1_macro": 0.7407336443567221,
116
+ "eval_f1_micro": 0.8171959721146398,
117
+ "eval_f1_weighted": 0.8160506897887181,
118
+ "eval_loss": 0.9310164451599121,
119
+ "eval_macro_fpr": 0.016134598971936556,
120
+ "eval_macro_sensitivity": 0.7424877811517157,
121
+ "eval_macro_specificity": 0.9862275325034797,
122
+ "eval_precision": 0.8234764041889034,
123
+ "eval_precision_macro": 0.7660803499273104,
124
+ "eval_recall": 0.8171959721146398,
125
+ "eval_recall_macro": 0.7424877811517157,
126
+ "eval_runtime": 58.71,
127
+ "eval_samples_per_second": 21.989,
128
+ "eval_steps_per_second": 2.759,
129
+ "eval_weighted_fpr": 0.015727042516326804,
130
+ "eval_weighted_sensitivity": 0.8171959721146398,
131
+ "eval_weighted_specificity": 0.9762170154375561,
132
+ "step": 2572
133
+ },
134
+ {
135
+ "epoch": 4.67,
136
+ "grad_norm": 8.432526588439941,
137
+ "learning_rate": 3.447382063245205e-05,
138
+ "loss": 0.1611,
139
+ "step": 3000
140
+ },
141
+ {
142
+ "epoch": 5.0,
143
+ "eval_accuracy": 0.8303640588690937,
144
+ "eval_f1_macro": 0.7985673319510695,
145
+ "eval_f1_micro": 0.8303640588690937,
146
+ "eval_f1_weighted": 0.8304130023375266,
147
+ "eval_loss": 1.0763046741485596,
148
+ "eval_macro_fpr": 0.014791024663918677,
149
+ "eval_macro_sensitivity": 0.7918299512118986,
150
+ "eval_macro_specificity": 0.9872504344639083,
151
+ "eval_precision": 0.8362631231068205,
152
+ "eval_precision_macro": 0.8174407838611729,
153
+ "eval_recall": 0.8303640588690937,
154
+ "eval_recall_macro": 0.7918299512118986,
155
+ "eval_runtime": 71.8128,
156
+ "eval_samples_per_second": 17.977,
157
+ "eval_steps_per_second": 2.256,
158
+ "eval_weighted_fpr": 0.014382347146516056,
159
+ "eval_weighted_sensitivity": 0.8303640588690937,
160
+ "eval_weighted_specificity": 0.978392458089531,
161
+ "step": 3215
162
+ },
163
+ {
164
+ "epoch": 5.44,
165
+ "grad_norm": 0.10359270125627518,
166
+ "learning_rate": 3.1881804043545884e-05,
167
+ "loss": 0.1055,
168
+ "step": 3500
169
+ },
170
+ {
171
+ "epoch": 6.0,
172
+ "eval_accuracy": 0.82571649883811,
173
+ "eval_f1_macro": 0.786305894104014,
174
+ "eval_f1_micro": 0.82571649883811,
175
+ "eval_f1_weighted": 0.824551084474796,
176
+ "eval_loss": 1.1377127170562744,
177
+ "eval_macro_fpr": 0.015367631824352218,
178
+ "eval_macro_sensitivity": 0.7810139173900718,
179
+ "eval_macro_specificity": 0.9868830526127108,
180
+ "eval_precision": 0.8275394319817495,
181
+ "eval_precision_macro": 0.8039087862954798,
182
+ "eval_recall": 0.82571649883811,
183
+ "eval_recall_macro": 0.7810139173900718,
184
+ "eval_runtime": 64.6694,
185
+ "eval_samples_per_second": 19.963,
186
+ "eval_steps_per_second": 2.505,
187
+ "eval_weighted_fpr": 0.01485246550927454,
188
+ "eval_weighted_sensitivity": 0.82571649883811,
189
+ "eval_weighted_specificity": 0.9775292903525532,
190
+ "step": 3858
191
+ },
192
+ {
193
+ "epoch": 6.22,
194
+ "grad_norm": 0.07212593406438828,
195
+ "learning_rate": 2.928978745463971e-05,
196
+ "loss": 0.1152,
197
+ "step": 4000
198
+ },
199
+ {
200
+ "epoch": 7.0,
201
+ "grad_norm": 0.15883693099021912,
202
+ "learning_rate": 2.669777086573354e-05,
203
+ "loss": 0.0463,
204
+ "step": 4500
205
+ },
206
+ {
207
+ "epoch": 7.0,
208
+ "eval_accuracy": 0.8071262587141751,
209
+ "eval_f1_macro": 0.7661147949067431,
210
+ "eval_f1_micro": 0.8071262587141751,
211
+ "eval_f1_weighted": 0.8078396785965507,
212
+ "eval_loss": 1.321502447128296,
213
+ "eval_macro_fpr": 0.017229971151571283,
214
+ "eval_macro_sensitivity": 0.7688718615633127,
215
+ "eval_macro_specificity": 0.9855509290260365,
216
+ "eval_precision": 0.8111033305830805,
217
+ "eval_precision_macro": 0.7691608912363075,
218
+ "eval_recall": 0.8071262587141751,
219
+ "eval_recall_macro": 0.7688718615633127,
220
+ "eval_runtime": 77.5552,
221
+ "eval_samples_per_second": 16.646,
222
+ "eval_steps_per_second": 2.089,
223
+ "eval_weighted_fpr": 0.016782368403316036,
224
+ "eval_weighted_sensitivity": 0.8071262587141751,
225
+ "eval_weighted_specificity": 0.9761376766763701,
226
+ "step": 4501
227
+ },
228
+ {
229
+ "epoch": 7.78,
230
+ "grad_norm": 0.003739361884072423,
231
+ "learning_rate": 2.4105754276827372e-05,
232
+ "loss": 0.031,
233
+ "step": 5000
234
+ },
235
+ {
236
+ "epoch": 8.0,
237
+ "eval_accuracy": 0.820294345468629,
238
+ "eval_f1_macro": 0.7689822136833361,
239
+ "eval_f1_micro": 0.8202943454686291,
240
+ "eval_f1_weighted": 0.8174949305492484,
241
+ "eval_loss": 1.3483381271362305,
242
+ "eval_macro_fpr": 0.01612405504801943,
243
+ "eval_macro_sensitivity": 0.7726825622003671,
244
+ "eval_macro_specificity": 0.9863605879945143,
245
+ "eval_precision": 0.8169871783387602,
246
+ "eval_precision_macro": 0.7773067113000881,
247
+ "eval_recall": 0.820294345468629,
248
+ "eval_recall_macro": 0.7726825622003671,
249
+ "eval_runtime": 57.45,
250
+ "eval_samples_per_second": 22.472,
251
+ "eval_steps_per_second": 2.82,
252
+ "eval_weighted_fpr": 0.015407092575375215,
253
+ "eval_weighted_sensitivity": 0.820294345468629,
254
+ "eval_weighted_specificity": 0.9751144744490858,
255
+ "step": 5144
256
+ },
257
+ {
258
+ "epoch": 8.55,
259
+ "grad_norm": 0.008842483162879944,
260
+ "learning_rate": 2.151892172109902e-05,
261
+ "loss": 0.0202,
262
+ "step": 5500
263
+ },
264
+ {
265
+ "epoch": 9.0,
266
+ "eval_accuracy": 0.8280402788536019,
267
+ "eval_f1_macro": 0.7753479744325771,
268
+ "eval_f1_micro": 0.8280402788536019,
269
+ "eval_f1_weighted": 0.825561089776041,
270
+ "eval_loss": 1.3729532957077026,
271
+ "eval_macro_fpr": 0.015177813835714501,
272
+ "eval_macro_sensitivity": 0.7802568330601105,
273
+ "eval_macro_specificity": 0.9870641550691818,
274
+ "eval_precision": 0.8263269633953413,
275
+ "eval_precision_macro": 0.7817620978620624,
276
+ "eval_recall": 0.8280402788536019,
277
+ "eval_recall_macro": 0.7802568330601105,
278
+ "eval_runtime": 69.0336,
279
+ "eval_samples_per_second": 18.701,
280
+ "eval_steps_per_second": 2.347,
281
+ "eval_weighted_fpr": 0.014616802739004477,
282
+ "eval_weighted_sensitivity": 0.8280402788536019,
283
+ "eval_weighted_specificity": 0.977922047184124,
284
+ "step": 5787
285
+ },
286
+ {
287
+ "epoch": 9.33,
288
+ "grad_norm": 0.0021986050996929407,
289
+ "learning_rate": 1.8926905132192844e-05,
290
+ "loss": 0.0133,
291
+ "step": 6000
292
+ },
293
+ {
294
+ "epoch": 10.0,
295
+ "eval_accuracy": 0.8164213787761425,
296
+ "eval_f1_macro": 0.7655491262772139,
297
+ "eval_f1_micro": 0.8164213787761425,
298
+ "eval_f1_weighted": 0.8134862058117613,
299
+ "eval_loss": 1.5407140254974365,
300
+ "eval_macro_fpr": 0.016532295658760295,
301
+ "eval_macro_sensitivity": 0.7779131623925402,
302
+ "eval_macro_specificity": 0.9861038702342115,
303
+ "eval_precision": 0.8162862069364494,
304
+ "eval_precision_macro": 0.7687513177071161,
305
+ "eval_recall": 0.8164213787761425,
306
+ "eval_recall_macro": 0.7779131623925402,
307
+ "eval_runtime": 56.3906,
308
+ "eval_samples_per_second": 22.894,
309
+ "eval_steps_per_second": 2.873,
310
+ "eval_weighted_fpr": 0.01580737677582872,
311
+ "eval_weighted_sensitivity": 0.8164213787761425,
312
+ "eval_weighted_specificity": 0.9751366747370328,
313
+ "step": 6430
314
+ },
315
+ {
316
+ "epoch": 10.11,
317
+ "grad_norm": 0.006160721182823181,
318
+ "learning_rate": 1.6334888543286677e-05,
319
+ "loss": 0.0059,
320
+ "step": 6500
321
+ },
322
+ {
323
+ "epoch": 10.89,
324
+ "grad_norm": 0.04344305023550987,
325
+ "learning_rate": 1.374287195438051e-05,
326
+ "loss": 0.0051,
327
+ "step": 7000
328
+ },
329
+ {
330
+ "epoch": 11.0,
331
+ "eval_accuracy": 0.8226181254841208,
332
+ "eval_f1_macro": 0.7743760234813677,
333
+ "eval_f1_micro": 0.8226181254841208,
334
+ "eval_f1_weighted": 0.8234395708523029,
335
+ "eval_loss": 1.5235263109207153,
336
+ "eval_macro_fpr": 0.015631611736392083,
337
+ "eval_macro_sensitivity": 0.7679556612821589,
338
+ "eval_macro_specificity": 0.9866336477591318,
339
+ "eval_precision": 0.8265113395379454,
340
+ "eval_precision_macro": 0.7900275628638693,
341
+ "eval_recall": 0.8226181254841208,
342
+ "eval_recall_macro": 0.7679556612821589,
343
+ "eval_runtime": 61.2541,
344
+ "eval_samples_per_second": 21.076,
345
+ "eval_steps_per_second": 2.645,
346
+ "eval_weighted_fpr": 0.01516857653838511,
347
+ "eval_weighted_sensitivity": 0.8226181254841208,
348
+ "eval_weighted_specificity": 0.9768865909028567,
349
+ "step": 7073
350
+ },
351
+ {
352
+ "epoch": 11.66,
353
+ "grad_norm": 0.004823015071451664,
354
+ "learning_rate": 1.115085536547434e-05,
355
+ "loss": 0.0027,
356
+ "step": 7500
357
+ },
358
+ {
359
+ "epoch": 12.0,
360
+ "eval_accuracy": 0.8264910921766073,
361
+ "eval_f1_macro": 0.7775395030897111,
362
+ "eval_f1_micro": 0.8264910921766073,
363
+ "eval_f1_weighted": 0.8245039474797854,
364
+ "eval_loss": 1.5643447637557983,
365
+ "eval_macro_fpr": 0.015383831928275008,
366
+ "eval_macro_sensitivity": 0.7840707270984979,
367
+ "eval_macro_specificity": 0.9869113657916206,
368
+ "eval_precision": 0.8259137679191498,
369
+ "eval_precision_macro": 0.780456051036287,
370
+ "eval_recall": 0.8264910921766073,
371
+ "eval_recall_macro": 0.7840707270984979,
372
+ "eval_runtime": 57.1114,
373
+ "eval_samples_per_second": 22.605,
374
+ "eval_steps_per_second": 2.837,
375
+ "eval_weighted_fpr": 0.014773776546629732,
376
+ "eval_weighted_sensitivity": 0.8264910921766073,
377
+ "eval_weighted_specificity": 0.9771793946977024,
378
+ "step": 7716
379
+ },
380
+ {
381
+ "epoch": 12.44,
382
+ "grad_norm": 0.0019840807653963566,
383
+ "learning_rate": 8.564022809745984e-06,
384
+ "loss": 0.002,
385
+ "step": 8000
386
+ },
387
+ {
388
+ "epoch": 13.0,
389
+ "eval_accuracy": 0.8280402788536019,
390
+ "eval_f1_macro": 0.7859866551548631,
391
+ "eval_f1_micro": 0.8280402788536019,
392
+ "eval_f1_weighted": 0.8262088681957308,
393
+ "eval_loss": 1.551638126373291,
394
+ "eval_macro_fpr": 0.015177283746053602,
395
+ "eval_macro_sensitivity": 0.7902099894008585,
396
+ "eval_macro_specificity": 0.9870602401116764,
397
+ "eval_precision": 0.8272528468694974,
398
+ "eval_precision_macro": 0.7881702744112012,
399
+ "eval_recall": 0.8280402788536019,
400
+ "eval_recall_macro": 0.7902099894008585,
401
+ "eval_runtime": 70.3586,
402
+ "eval_samples_per_second": 18.349,
403
+ "eval_steps_per_second": 2.302,
404
+ "eval_weighted_fpr": 0.014616802739004477,
405
+ "eval_weighted_sensitivity": 0.8280402788536019,
406
+ "eval_weighted_specificity": 0.9778633228215452,
407
+ "step": 8359
408
+ },
409
+ {
410
+ "epoch": 13.22,
411
+ "grad_norm": 0.002753123641014099,
412
+ "learning_rate": 5.9720062208398135e-06,
413
+ "loss": 0.0014,
414
+ "step": 8500
415
+ },
416
+ {
417
+ "epoch": 14.0,
418
+ "grad_norm": 0.0008325249655172229,
419
+ "learning_rate": 3.3851736651114567e-06,
420
+ "loss": 0.0015,
421
+ "step": 9000
422
+ },
423
+ {
424
+ "epoch": 14.0,
425
+ "eval_accuracy": 0.8272656855151046,
426
+ "eval_f1_macro": 0.7942950839039143,
427
+ "eval_f1_micro": 0.8272656855151046,
428
+ "eval_f1_weighted": 0.8259391886185965,
429
+ "eval_loss": 1.5834864377975464,
430
+ "eval_macro_fpr": 0.01525435506530012,
431
+ "eval_macro_sensitivity": 0.8022311610261105,
432
+ "eval_macro_specificity": 0.9869687722325965,
433
+ "eval_precision": 0.8268053163162928,
434
+ "eval_precision_macro": 0.7942523809700115,
435
+ "eval_recall": 0.8272656855151046,
436
+ "eval_recall_macro": 0.8022311610261105,
437
+ "eval_runtime": 67.4962,
438
+ "eval_samples_per_second": 19.127,
439
+ "eval_steps_per_second": 2.4,
440
+ "eval_weighted_fpr": 0.014695222405271829,
441
+ "eval_weighted_sensitivity": 0.8272656855151046,
442
+ "eval_weighted_specificity": 0.9772658979738418,
443
+ "step": 9002
444
+ }
445
+ ],
446
+ "logging_steps": 500,
447
+ "max_steps": 9645,
448
+ "num_input_tokens_seen": 0,
449
+ "num_train_epochs": 15,
450
+ "save_steps": 500,
451
+ "total_flos": 1.892831352227021e+16,
452
+ "train_batch_size": 8,
453
+ "trial_name": null,
454
+ "trial_params": null
455
+ }
training_checkpoints/checkpoint-9002/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d67f433b74e0c5015495d7470ff35e34ab20bb7abc2e1887dfc6d557fdf2970
3
+ size 4920