Commit
65fe514
·
verified ·
1 Parent(s): f40fdea

Add pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue

Browse files

Upload fine-tuned rerankers for BioASQ 14B

Co-authored-by: André Ribeiro <andrepedro2004@hotmail.com>
Co-authored-by: Rúben Garrido <rubengarrido@ua.pt>

pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": null,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 768,
14
+ "id2label": {
15
+ "0": "LABEL_0"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 3072,
19
+ "is_decoder": false,
20
+ "label2id": {
21
+ "LABEL_0": 0
22
+ },
23
+ "layer_norm_eps": 1e-12,
24
+ "max_position_embeddings": 512,
25
+ "model_type": "bert",
26
+ "num_attention_heads": 12,
27
+ "num_hidden_layers": 12,
28
+ "pad_token_id": 0,
29
+ "position_embedding_type": "absolute",
30
+ "tie_word_embeddings": true,
31
+ "transformers_version": "5.2.0",
32
+ "type_vocab_size": 2,
33
+ "use_cache": false,
34
+ "vocab_size": 30522
35
+ }
pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f441fdd4f715885374f60cd50838d0fe29b0d4471af505b0070b2215d4648ab5
3
+ size 218989610
pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/ranx_results.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "pritamdeka-S-PubMedBert-MS-MARCO",
3
+ "val_files": [
4
+ "../../data/val_data/13B3_golden.json",
5
+ "../../data/val_data/13B1_golden.json",
6
+ "../../data/val_data/13B2_golden.json",
7
+ "../../data/val_data/13B4_golden.json"
8
+ ],
9
+ "total": {
10
+ "ndcg@5": 0.7109487692574652,
11
+ "mrr": 0.8277499787793906,
12
+ "recall@10": 0.4153044743717044,
13
+ "recall@100": 0.9404008699092695,
14
+ "map@10": 0.3563486104744832,
15
+ "map-bioasq@10": 0.5839477258495991
16
+ },
17
+ "13B3_golden.json": {
18
+ "ndcg@5": 0.7338792695416825,
19
+ "mrr": 0.8422417451829216,
20
+ "recall@10": 0.4340862345875427,
21
+ "recall@100": 0.9515166035341741,
22
+ "map@10": 0.3738720932849196,
23
+ "map-bioasq@10": 0.5926828898225958
24
+ },
25
+ "13B1_golden.json": {
26
+ "ndcg@5": 0.6323908123831076,
27
+ "mrr": 0.7293464052287582,
28
+ "recall@10": 0.4659199748585041,
29
+ "recall@100": 0.9605964285948642,
30
+ "map@10": 0.3775489539858533,
31
+ "map-bioasq@10": 0.518151053013798
32
+ },
33
+ "13B2_golden.json": {
34
+ "ndcg@5": 0.8062057525365898,
35
+ "mrr": 0.9070588235294119,
36
+ "recall@10": 0.5003984745773042,
37
+ "recall@100": 0.9437935968329072,
38
+ "map@10": 0.46314647110849727,
39
+ "map-bioasq@10": 0.7019981992797119
40
+ },
41
+ "13B4_golden.json": {
42
+ "ndcg@5": 0.6713192425684811,
43
+ "mrr": 0.8323529411764706,
44
+ "recall@10": 0.2608132134634668,
45
+ "recall@100": 0.9056968506751328,
46
+ "map@10": 0.21082692351866253,
47
+ "map-bioasq@10": 0.5229587612822908
48
+ }
49
+ }
pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "cls_token": "[CLS]",
4
+ "do_basic_tokenize": true,
5
+ "do_lower_case": true,
6
+ "is_local": false,
7
+ "mask_token": "[MASK]",
8
+ "model_max_length": 1000000000000000019884624838656,
9
+ "never_split": null,
10
+ "pad_token": "[PAD]",
11
+ "sep_token": "[SEP]",
12
+ "strip_accents": null,
13
+ "tokenize_chinese_chars": true,
14
+ "tokenizer_class": "BertTokenizer",
15
+ "unk_token": "[UNK]"
16
+ }
pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/trainer_state.json ADDED
@@ -0,0 +1,673 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 2.0,
6
+ "eval_steps": 500,
7
+ "global_step": 4452,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.022461814914645103,
14
+ "grad_norm": 16.75,
15
+ "learning_rate": 1.977987421383648e-05,
16
+ "loss": 0.3261573791503906,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.044923629829290206,
21
+ "grad_norm": 11.4375,
22
+ "learning_rate": 1.955525606469003e-05,
23
+ "loss": 0.36861457824707033,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.0673854447439353,
28
+ "grad_norm": 8.1875,
29
+ "learning_rate": 1.9330637915543576e-05,
30
+ "loss": 0.30939388275146484,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.08984725965858041,
35
+ "grad_norm": 7.75,
36
+ "learning_rate": 1.9106019766397127e-05,
37
+ "loss": 0.3264142608642578,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.11230907457322552,
42
+ "grad_norm": 4.375,
43
+ "learning_rate": 1.8881401617250677e-05,
44
+ "loss": 0.3411271667480469,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.1347708894878706,
49
+ "grad_norm": 12.625,
50
+ "learning_rate": 1.8656783468104224e-05,
51
+ "loss": 0.32702728271484377,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.15723270440251572,
56
+ "grad_norm": 11.5625,
57
+ "learning_rate": 1.843216531895777e-05,
58
+ "loss": 0.3568133926391602,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.17969451931716082,
63
+ "grad_norm": 13.375,
64
+ "learning_rate": 1.8207547169811322e-05,
65
+ "loss": 0.3119574165344238,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.20215633423180593,
70
+ "grad_norm": 6.46875,
71
+ "learning_rate": 1.7982929020664873e-05,
72
+ "loss": 0.3093097114562988,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.22461814914645103,
77
+ "grad_norm": 9.6875,
78
+ "learning_rate": 1.775831087151842e-05,
79
+ "loss": 0.3411777114868164,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.24707996406109614,
84
+ "grad_norm": 13.0625,
85
+ "learning_rate": 1.753369272237197e-05,
86
+ "loss": 0.33546173095703125,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 0.2695417789757412,
91
+ "grad_norm": 12.125,
92
+ "learning_rate": 1.7309074573225517e-05,
93
+ "loss": 0.3275848388671875,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 0.2920035938903863,
98
+ "grad_norm": 11.5,
99
+ "learning_rate": 1.7084456424079068e-05,
100
+ "loss": 0.33126487731933596,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 0.31446540880503143,
105
+ "grad_norm": 10.5,
106
+ "learning_rate": 1.6859838274932615e-05,
107
+ "loss": 0.32779808044433595,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 0.33692722371967654,
112
+ "grad_norm": 12.25,
113
+ "learning_rate": 1.6635220125786166e-05,
114
+ "loss": 0.32774574279785157,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 0.35938903863432164,
119
+ "grad_norm": 8.5625,
120
+ "learning_rate": 1.6410601976639713e-05,
121
+ "loss": 0.33253490447998046,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 0.38185085354896675,
126
+ "grad_norm": 13.6875,
127
+ "learning_rate": 1.6185983827493263e-05,
128
+ "loss": 0.32540748596191404,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 0.40431266846361186,
133
+ "grad_norm": 14.25,
134
+ "learning_rate": 1.596136567834681e-05,
135
+ "loss": 0.2918030548095703,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 0.42677448337825696,
140
+ "grad_norm": 11.8125,
141
+ "learning_rate": 1.573674752920036e-05,
142
+ "loss": 0.3029415893554688,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 0.44923629829290207,
147
+ "grad_norm": 12.5,
148
+ "learning_rate": 1.5512129380053908e-05,
149
+ "loss": 0.3138488578796387,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 0.4716981132075472,
154
+ "grad_norm": 10.4375,
155
+ "learning_rate": 1.528751123090746e-05,
156
+ "loss": 0.33728736877441406,
157
+ "step": 1050
158
+ },
159
+ {
160
+ "epoch": 0.4941599281221923,
161
+ "grad_norm": 10.5625,
162
+ "learning_rate": 1.5062893081761007e-05,
163
+ "loss": 0.314833984375,
164
+ "step": 1100
165
+ },
166
+ {
167
+ "epoch": 0.5166217430368374,
168
+ "grad_norm": 13.1875,
169
+ "learning_rate": 1.4838274932614556e-05,
170
+ "loss": 0.3515630340576172,
171
+ "step": 1150
172
+ },
173
+ {
174
+ "epoch": 0.5390835579514824,
175
+ "grad_norm": 13.875,
176
+ "learning_rate": 1.4613656783468107e-05,
177
+ "loss": 0.2913247299194336,
178
+ "step": 1200
179
+ },
180
+ {
181
+ "epoch": 0.5615453728661276,
182
+ "grad_norm": 10.6875,
183
+ "learning_rate": 1.4389038634321656e-05,
184
+ "loss": 0.3084243392944336,
185
+ "step": 1250
186
+ },
187
+ {
188
+ "epoch": 0.5840071877807727,
189
+ "grad_norm": 16.25,
190
+ "learning_rate": 1.4164420485175203e-05,
191
+ "loss": 0.3340093994140625,
192
+ "step": 1300
193
+ },
194
+ {
195
+ "epoch": 0.6064690026954178,
196
+ "grad_norm": 13.6875,
197
+ "learning_rate": 1.3939802336028752e-05,
198
+ "loss": 0.28631263732910156,
199
+ "step": 1350
200
+ },
201
+ {
202
+ "epoch": 0.6289308176100629,
203
+ "grad_norm": 11.5625,
204
+ "learning_rate": 1.3715184186882302e-05,
205
+ "loss": 0.32017528533935546,
206
+ "step": 1400
207
+ },
208
+ {
209
+ "epoch": 0.651392632524708,
210
+ "grad_norm": 12.375,
211
+ "learning_rate": 1.3490566037735851e-05,
212
+ "loss": 0.34149707794189454,
213
+ "step": 1450
214
+ },
215
+ {
216
+ "epoch": 0.6738544474393531,
217
+ "grad_norm": 12.0625,
218
+ "learning_rate": 1.3265947888589398e-05,
219
+ "loss": 0.3179684257507324,
220
+ "step": 1500
221
+ },
222
+ {
223
+ "epoch": 0.6963162623539982,
224
+ "grad_norm": 8.9375,
225
+ "learning_rate": 1.3041329739442947e-05,
226
+ "loss": 0.3083969497680664,
227
+ "step": 1550
228
+ },
229
+ {
230
+ "epoch": 0.7187780772686433,
231
+ "grad_norm": 11.5,
232
+ "learning_rate": 1.2816711590296497e-05,
233
+ "loss": 0.3271527481079102,
234
+ "step": 1600
235
+ },
236
+ {
237
+ "epoch": 0.7412398921832885,
238
+ "grad_norm": 8.625,
239
+ "learning_rate": 1.2592093441150046e-05,
240
+ "loss": 0.33942337036132814,
241
+ "step": 1650
242
+ },
243
+ {
244
+ "epoch": 0.7637017070979335,
245
+ "grad_norm": 12.375,
246
+ "learning_rate": 1.2367475292003595e-05,
247
+ "loss": 0.32219131469726564,
248
+ "step": 1700
249
+ },
250
+ {
251
+ "epoch": 0.7861635220125787,
252
+ "grad_norm": 13.6875,
253
+ "learning_rate": 1.2142857142857142e-05,
254
+ "loss": 0.3429621505737305,
255
+ "step": 1750
256
+ },
257
+ {
258
+ "epoch": 0.8086253369272237,
259
+ "grad_norm": 10.9375,
260
+ "learning_rate": 1.1918238993710693e-05,
261
+ "loss": 0.29558956146240234,
262
+ "step": 1800
263
+ },
264
+ {
265
+ "epoch": 0.8310871518418689,
266
+ "grad_norm": 8.875,
267
+ "learning_rate": 1.1693620844564242e-05,
268
+ "loss": 0.28521089553833007,
269
+ "step": 1850
270
+ },
271
+ {
272
+ "epoch": 0.8535489667565139,
273
+ "grad_norm": 14.6875,
274
+ "learning_rate": 1.146900269541779e-05,
275
+ "loss": 0.2883760452270508,
276
+ "step": 1900
277
+ },
278
+ {
279
+ "epoch": 0.876010781671159,
280
+ "grad_norm": 11.1875,
281
+ "learning_rate": 1.124438454627134e-05,
282
+ "loss": 0.2875115394592285,
283
+ "step": 1950
284
+ },
285
+ {
286
+ "epoch": 0.8984725965858041,
287
+ "grad_norm": 15.125,
288
+ "learning_rate": 1.101976639712489e-05,
289
+ "loss": 0.30944171905517576,
290
+ "step": 2000
291
+ },
292
+ {
293
+ "epoch": 0.9209344115004492,
294
+ "grad_norm": 12.9375,
295
+ "learning_rate": 1.0795148247978437e-05,
296
+ "loss": 0.2973375129699707,
297
+ "step": 2050
298
+ },
299
+ {
300
+ "epoch": 0.9433962264150944,
301
+ "grad_norm": 17.0,
302
+ "learning_rate": 1.0570530098831986e-05,
303
+ "loss": 0.3215830993652344,
304
+ "step": 2100
305
+ },
306
+ {
307
+ "epoch": 0.9658580413297394,
308
+ "grad_norm": 13.3125,
309
+ "learning_rate": 1.0345911949685536e-05,
310
+ "loss": 0.346932487487793,
311
+ "step": 2150
312
+ },
313
+ {
314
+ "epoch": 0.9883198562443846,
315
+ "grad_norm": 13.125,
316
+ "learning_rate": 1.0121293800539085e-05,
317
+ "loss": 0.3510477066040039,
318
+ "step": 2200
319
+ },
320
+ {
321
+ "epoch": 1.0,
322
+ "eval_loss": 0.41021817922592163,
323
+ "eval_runtime": 1.5984,
324
+ "eval_samples_per_second": 211.468,
325
+ "eval_steps_per_second": 6.882,
326
+ "step": 2226
327
+ },
328
+ {
329
+ "epoch": 1.0107816711590296,
330
+ "grad_norm": 13.4375,
331
+ "learning_rate": 9.896675651392634e-06,
332
+ "loss": 0.2997927474975586,
333
+ "step": 2250
334
+ },
335
+ {
336
+ "epoch": 1.0332434860736748,
337
+ "grad_norm": 12.9375,
338
+ "learning_rate": 9.672057502246183e-06,
339
+ "loss": 0.2939089012145996,
340
+ "step": 2300
341
+ },
342
+ {
343
+ "epoch": 1.05570530098832,
344
+ "grad_norm": 14.5,
345
+ "learning_rate": 9.447439353099732e-06,
346
+ "loss": 0.3152497863769531,
347
+ "step": 2350
348
+ },
349
+ {
350
+ "epoch": 1.0781671159029649,
351
+ "grad_norm": 18.75,
352
+ "learning_rate": 9.22282120395328e-06,
353
+ "loss": 0.3313028335571289,
354
+ "step": 2400
355
+ },
356
+ {
357
+ "epoch": 1.10062893081761,
358
+ "grad_norm": 13.8125,
359
+ "learning_rate": 8.99820305480683e-06,
360
+ "loss": 0.31833568572998044,
361
+ "step": 2450
362
+ },
363
+ {
364
+ "epoch": 1.1230907457322552,
365
+ "grad_norm": 10.125,
366
+ "learning_rate": 8.773584905660378e-06,
367
+ "loss": 0.30160762786865236,
368
+ "step": 2500
369
+ },
370
+ {
371
+ "epoch": 1.1455525606469004,
372
+ "grad_norm": 11.125,
373
+ "learning_rate": 8.548966756513927e-06,
374
+ "loss": 0.3140912628173828,
375
+ "step": 2550
376
+ },
377
+ {
378
+ "epoch": 1.1680143755615453,
379
+ "grad_norm": 13.1875,
380
+ "learning_rate": 8.324348607367476e-06,
381
+ "loss": 0.3247405242919922,
382
+ "step": 2600
383
+ },
384
+ {
385
+ "epoch": 1.1904761904761905,
386
+ "grad_norm": 13.25,
387
+ "learning_rate": 8.099730458221025e-06,
388
+ "loss": 0.3353820037841797,
389
+ "step": 2650
390
+ },
391
+ {
392
+ "epoch": 1.2129380053908356,
393
+ "grad_norm": 12.5,
394
+ "learning_rate": 7.875112309074573e-06,
395
+ "loss": 0.3192539405822754,
396
+ "step": 2700
397
+ },
398
+ {
399
+ "epoch": 1.2353998203054806,
400
+ "grad_norm": 7.21875,
401
+ "learning_rate": 7.650494159928122e-06,
402
+ "loss": 0.2817390060424805,
403
+ "step": 2750
404
+ },
405
+ {
406
+ "epoch": 1.2578616352201257,
407
+ "grad_norm": 9.6875,
408
+ "learning_rate": 7.425876010781672e-06,
409
+ "loss": 0.29697341918945314,
410
+ "step": 2800
411
+ },
412
+ {
413
+ "epoch": 1.280323450134771,
414
+ "grad_norm": 14.1875,
415
+ "learning_rate": 7.20125786163522e-06,
416
+ "loss": 0.28406223297119143,
417
+ "step": 2850
418
+ },
419
+ {
420
+ "epoch": 1.302785265049416,
421
+ "grad_norm": 8.9375,
422
+ "learning_rate": 6.97663971248877e-06,
423
+ "loss": 0.3698112487792969,
424
+ "step": 2900
425
+ },
426
+ {
427
+ "epoch": 1.3252470799640612,
428
+ "grad_norm": 13.25,
429
+ "learning_rate": 6.752021563342318e-06,
430
+ "loss": 0.3122553825378418,
431
+ "step": 2950
432
+ },
433
+ {
434
+ "epoch": 1.3477088948787062,
435
+ "grad_norm": 13.375,
436
+ "learning_rate": 6.527403414195867e-06,
437
+ "loss": 0.3351892852783203,
438
+ "step": 3000
439
+ },
440
+ {
441
+ "epoch": 1.3701707097933513,
442
+ "grad_norm": 11.125,
443
+ "learning_rate": 6.302785265049417e-06,
444
+ "loss": 0.298923397064209,
445
+ "step": 3050
446
+ },
447
+ {
448
+ "epoch": 1.3926325247079965,
449
+ "grad_norm": 11.5,
450
+ "learning_rate": 6.078167115902965e-06,
451
+ "loss": 0.33939796447753906,
452
+ "step": 3100
453
+ },
454
+ {
455
+ "epoch": 1.4150943396226414,
456
+ "grad_norm": 12.1875,
457
+ "learning_rate": 5.853548966756515e-06,
458
+ "loss": 0.29243343353271484,
459
+ "step": 3150
460
+ },
461
+ {
462
+ "epoch": 1.4375561545372866,
463
+ "grad_norm": 11.125,
464
+ "learning_rate": 5.6289308176100635e-06,
465
+ "loss": 0.31228084564208985,
466
+ "step": 3200
467
+ },
468
+ {
469
+ "epoch": 1.4600179694519317,
470
+ "grad_norm": 13.75,
471
+ "learning_rate": 5.404312668463612e-06,
472
+ "loss": 0.3152684020996094,
473
+ "step": 3250
474
+ },
475
+ {
476
+ "epoch": 1.482479784366577,
477
+ "grad_norm": 13.375,
478
+ "learning_rate": 5.179694519317161e-06,
479
+ "loss": 0.2882042694091797,
480
+ "step": 3300
481
+ },
482
+ {
483
+ "epoch": 1.504941599281222,
484
+ "grad_norm": 8.4375,
485
+ "learning_rate": 4.95507637017071e-06,
486
+ "loss": 0.3170111465454102,
487
+ "step": 3350
488
+ },
489
+ {
490
+ "epoch": 1.527403414195867,
491
+ "grad_norm": 12.5,
492
+ "learning_rate": 4.730458221024259e-06,
493
+ "loss": 0.2965633201599121,
494
+ "step": 3400
495
+ },
496
+ {
497
+ "epoch": 1.5498652291105122,
498
+ "grad_norm": 17.125,
499
+ "learning_rate": 4.5058400718778085e-06,
500
+ "loss": 0.2729132080078125,
501
+ "step": 3450
502
+ },
503
+ {
504
+ "epoch": 1.5723270440251573,
505
+ "grad_norm": 10.25,
506
+ "learning_rate": 4.281221922731357e-06,
507
+ "loss": 0.30944385528564455,
508
+ "step": 3500
509
+ },
510
+ {
511
+ "epoch": 1.5947888589398023,
512
+ "grad_norm": 10.25,
513
+ "learning_rate": 4.056603773584906e-06,
514
+ "loss": 0.2877353286743164,
515
+ "step": 3550
516
+ },
517
+ {
518
+ "epoch": 1.6172506738544474,
519
+ "grad_norm": 17.625,
520
+ "learning_rate": 3.831985624438455e-06,
521
+ "loss": 0.3187435340881348,
522
+ "step": 3600
523
+ },
524
+ {
525
+ "epoch": 1.6397124887690926,
526
+ "grad_norm": 12.5625,
527
+ "learning_rate": 3.607367475292004e-06,
528
+ "loss": 0.3127108573913574,
529
+ "step": 3650
530
+ },
531
+ {
532
+ "epoch": 1.6621743036837375,
533
+ "grad_norm": 14.9375,
534
+ "learning_rate": 3.382749326145553e-06,
535
+ "loss": 0.3153807067871094,
536
+ "step": 3700
537
+ },
538
+ {
539
+ "epoch": 1.684636118598383,
540
+ "grad_norm": 11.75,
541
+ "learning_rate": 3.158131176999102e-06,
542
+ "loss": 0.34931446075439454,
543
+ "step": 3750
544
+ },
545
+ {
546
+ "epoch": 1.7070979335130279,
547
+ "grad_norm": 13.3125,
548
+ "learning_rate": 2.9335130278526508e-06,
549
+ "loss": 0.31531978607177735,
550
+ "step": 3800
551
+ },
552
+ {
553
+ "epoch": 1.7295597484276728,
554
+ "grad_norm": 6.6875,
555
+ "learning_rate": 2.7088948787061996e-06,
556
+ "loss": 0.31528482437133787,
557
+ "step": 3850
558
+ },
559
+ {
560
+ "epoch": 1.7520215633423182,
561
+ "grad_norm": 9.75,
562
+ "learning_rate": 2.4842767295597484e-06,
563
+ "loss": 0.3385495758056641,
564
+ "step": 3900
565
+ },
566
+ {
567
+ "epoch": 1.7744833782569631,
568
+ "grad_norm": 13.75,
569
+ "learning_rate": 2.2596585804132977e-06,
570
+ "loss": 0.3233124542236328,
571
+ "step": 3950
572
+ },
573
+ {
574
+ "epoch": 1.7969451931716083,
575
+ "grad_norm": 14.5,
576
+ "learning_rate": 2.0350404312668465e-06,
577
+ "loss": 0.2921269798278809,
578
+ "step": 4000
579
+ },
580
+ {
581
+ "epoch": 1.8194070080862534,
582
+ "grad_norm": 9.25,
583
+ "learning_rate": 1.8104222821203954e-06,
584
+ "loss": 0.33798755645751954,
585
+ "step": 4050
586
+ },
587
+ {
588
+ "epoch": 1.8418688230008984,
589
+ "grad_norm": 12.4375,
590
+ "learning_rate": 1.5858041329739444e-06,
591
+ "loss": 0.3124137115478516,
592
+ "step": 4100
593
+ },
594
+ {
595
+ "epoch": 1.8643306379155435,
596
+ "grad_norm": 13.4375,
597
+ "learning_rate": 1.3611859838274934e-06,
598
+ "loss": 0.28721511840820313,
599
+ "step": 4150
600
+ },
601
+ {
602
+ "epoch": 1.8867924528301887,
603
+ "grad_norm": 16.25,
604
+ "learning_rate": 1.1365678346810423e-06,
605
+ "loss": 0.3201523971557617,
606
+ "step": 4200
607
+ },
608
+ {
609
+ "epoch": 1.9092542677448336,
610
+ "grad_norm": 4.375,
611
+ "learning_rate": 9.119496855345912e-07,
612
+ "loss": 0.286939697265625,
613
+ "step": 4250
614
+ },
615
+ {
616
+ "epoch": 1.931716082659479,
617
+ "grad_norm": 12.25,
618
+ "learning_rate": 6.873315363881403e-07,
619
+ "loss": 0.32165679931640623,
620
+ "step": 4300
621
+ },
622
+ {
623
+ "epoch": 1.954177897574124,
624
+ "grad_norm": 13.125,
625
+ "learning_rate": 4.627133872416892e-07,
626
+ "loss": 0.28595523834228515,
627
+ "step": 4350
628
+ },
629
+ {
630
+ "epoch": 1.9766397124887691,
631
+ "grad_norm": 14.625,
632
+ "learning_rate": 2.3809523809523811e-07,
633
+ "loss": 0.27817085266113284,
634
+ "step": 4400
635
+ },
636
+ {
637
+ "epoch": 1.9991015274034143,
638
+ "grad_norm": 11.5625,
639
+ "learning_rate": 1.3477088948787064e-08,
640
+ "loss": 0.3066493225097656,
641
+ "step": 4450
642
+ },
643
+ {
644
+ "epoch": 2.0,
645
+ "eval_loss": 0.40981459617614746,
646
+ "eval_runtime": 1.535,
647
+ "eval_samples_per_second": 220.202,
648
+ "eval_steps_per_second": 7.166,
649
+ "step": 4452
650
+ }
651
+ ],
652
+ "logging_steps": 50,
653
+ "max_steps": 4452,
654
+ "num_input_tokens_seen": 0,
655
+ "num_train_epochs": 2,
656
+ "save_steps": 500,
657
+ "stateful_callbacks": {
658
+ "TrainerControl": {
659
+ "args": {
660
+ "should_epoch_stop": false,
661
+ "should_evaluate": false,
662
+ "should_log": false,
663
+ "should_save": true,
664
+ "should_training_stop": true
665
+ },
666
+ "attributes": {}
667
+ }
668
+ },
669
+ "total_flos": 0.0,
670
+ "train_batch_size": 16,
671
+ "trial_name": null,
672
+ "trial_params": null
673
+ }
pritamdeka-S-PubMedBert-MS-MARCO-E2-S1-Mpairwise-FullDataTrue/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30c5e511d396f13071cc2ab051f988477f1460d67d9df8cae36980064b579364
3
+ size 5329