Hugging Face's logo

YagiASAFAS
/
MsIssuesBERT

Text Classification
Transformers
Safetensors
bert
Generated from Trainer
text-embeddings-inference
Model card Files
xet
Community
YagiASAFAS commited on
Commit
4b17557
·
verified ·
1 Parent(s): 2bfe524

Training in progress, epoch 1

Browse files
Files changed (25) hide show
  1. model.safetensors +1 -1
  2. run-4/checkpoint-3000/config.json +65 -0
  3. run-4/checkpoint-3000/model.safetensors +3 -0
  4. run-4/checkpoint-3000/optimizer.pt +3 -0
  5. run-4/checkpoint-3000/rng_state.pth +3 -0
  6. run-4/checkpoint-3000/scaler.pt +3 -0
  7. run-4/checkpoint-3000/scheduler.pt +3 -0
  8. run-4/checkpoint-3000/trainer_state.json +326 -0
  9. run-4/checkpoint-3000/training_args.bin +3 -0
  10. run-4/checkpoint-4000/config.json +65 -0
  11. run-4/checkpoint-4000/model.safetensors +3 -0
  12. run-4/checkpoint-4000/optimizer.pt +3 -0
  13. run-4/checkpoint-4000/rng_state.pth +3 -0
  14. run-4/checkpoint-4000/scaler.pt +3 -0
  15. run-4/checkpoint-4000/scheduler.pt +3 -0
  16. run-4/checkpoint-4000/trainer_state.json +418 -0
  17. run-4/checkpoint-4000/training_args.bin +3 -0
  18. run-4/checkpoint-5000/config.json +65 -0
  19. run-4/checkpoint-5000/model.safetensors +3 -0
  20. run-4/checkpoint-5000/optimizer.pt +3 -0
  21. run-4/checkpoint-5000/rng_state.pth +3 -0
  22. run-4/checkpoint-5000/scaler.pt +3 -0
  23. run-4/checkpoint-5000/scheduler.pt +3 -0
  24. run-4/checkpoint-5000/trainer_state.json +510 -0
  25. run-4/checkpoint-5000/training_args.bin +3 -0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b153083e3247a53ddfa45f8ace862dd036680a0af16004e032aae994694c3ccb
3
  size 438007864
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ae22f9512206808c3dc8ef2632f32d5372ead88bd2843333db89b743ec3d9ca
3
  size 438007864
run-4/checkpoint-3000/config.json ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "LABEL_0",
13
+ "1": "LABEL_1",
14
+ "2": "LABEL_2",
15
+ "3": "LABEL_3",
16
+ "4": "LABEL_4",
17
+ "5": "LABEL_5",
18
+ "6": "LABEL_6",
19
+ "7": "LABEL_7",
20
+ "8": "LABEL_8",
21
+ "9": "LABEL_9",
22
+ "10": "LABEL_10",
23
+ "11": "LABEL_11",
24
+ "12": "LABEL_12",
25
+ "13": "LABEL_13",
26
+ "14": "LABEL_14",
27
+ "15": "LABEL_15",
28
+ "16": "LABEL_16",
29
+ "17": "LABEL_17"
30
+ },
31
+ "initializer_range": 0.02,
32
+ "intermediate_size": 3072,
33
+ "label2id": {
34
+ "LABEL_0": 0,
35
+ "LABEL_1": 1,
36
+ "LABEL_10": 10,
37
+ "LABEL_11": 11,
38
+ "LABEL_12": 12,
39
+ "LABEL_13": 13,
40
+ "LABEL_14": 14,
41
+ "LABEL_15": 15,
42
+ "LABEL_16": 16,
43
+ "LABEL_17": 17,
44
+ "LABEL_2": 2,
45
+ "LABEL_3": 3,
46
+ "LABEL_4": 4,
47
+ "LABEL_5": 5,
48
+ "LABEL_6": 6,
49
+ "LABEL_7": 7,
50
+ "LABEL_8": 8,
51
+ "LABEL_9": 9
52
+ },
53
+ "layer_norm_eps": 1e-12,
54
+ "max_position_embeddings": 512,
55
+ "model_type": "bert",
56
+ "num_attention_heads": 12,
57
+ "num_hidden_layers": 12,
58
+ "pad_token_id": 0,
59
+ "position_embedding_type": "absolute",
60
+ "torch_dtype": "float32",
61
+ "transformers_version": "4.55.4",
62
+ "type_vocab_size": 2,
63
+ "use_cache": true,
64
+ "vocab_size": 30522
65
+ }
run-4/checkpoint-3000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6096e7e6471c8f02bd355e3bd5860464fb1d8ebdf365b9b68c387fed8907c58e
3
+ size 438007864
run-4/checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bdd88f9d862cdd0f511c7d1b157b6202e663ff88a28f5a0b69bba88bd1cd85b
3
+ size 876140043
run-4/checkpoint-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a555b0edf62e8f522d8cd4624c35c5a1c5442d645e06f9fdcb5b5a2274c2e88d
3
+ size 14645
run-4/checkpoint-3000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c82336e9109f719d709251f8b29ee8fb4036c133e374069c706e8cacde5d6c0
3
+ size 1383
run-4/checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0561a411304954bd1af84ccf2f7c9b36ae975db19db6b7e5d786e003f991e6fe
3
+ size 1465
run-4/checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,326 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 3000,
3
+ "best_metric": 0.7988061056554884,
4
+ "best_model_checkpoint": "./results/run-4/checkpoint-3000",
5
+ "epoch": 3.0,
6
+ "eval_steps": 500,
7
+ "global_step": 3000,
8
+ "is_hyper_param_search": true,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1,
14
+ "grad_norm": 3.002380609512329,
15
+ "learning_rate": 4.572943108695439e-06,
16
+ "loss": 0.296,
17
+ "step": 100
18
+ },
19
+ {
20
+ "epoch": 0.2,
21
+ "grad_norm": 0.9367781281471252,
22
+ "learning_rate": 9.192077561923155e-06,
23
+ "loss": 1.0274,
24
+ "step": 200
25
+ },
26
+ {
27
+ "epoch": 0.3,
28
+ "grad_norm": 0.7274576425552368,
29
+ "learning_rate": 1.3811212015150872e-05,
30
+ "loss": 0.0,
31
+ "step": 300
32
+ },
33
+ {
34
+ "epoch": 0.4,
35
+ "grad_norm": 44.297061920166016,
36
+ "learning_rate": 1.8430346468378586e-05,
37
+ "loss": 0.0,
38
+ "step": 400
39
+ },
40
+ {
41
+ "epoch": 0.5,
42
+ "grad_norm": 1.0151411294937134,
43
+ "learning_rate": 2.3049480921606306e-05,
44
+ "loss": 0.0,
45
+ "step": 500
46
+ },
47
+ {
48
+ "epoch": 0.6,
49
+ "grad_norm": 1.208032488822937,
50
+ "learning_rate": 2.766861537483402e-05,
51
+ "loss": 0.0,
52
+ "step": 600
53
+ },
54
+ {
55
+ "epoch": 0.7,
56
+ "grad_norm": 6.984821319580078,
57
+ "learning_rate": 3.2287749828061733e-05,
58
+ "loss": 0.0342,
59
+ "step": 700
60
+ },
61
+ {
62
+ "epoch": 0.8,
63
+ "grad_norm": 1.5860588550567627,
64
+ "learning_rate": 3.690688428128945e-05,
65
+ "loss": 0.0488,
66
+ "step": 800
67
+ },
68
+ {
69
+ "epoch": 0.9,
70
+ "grad_norm": 0.7841507196426392,
71
+ "learning_rate": 4.152601873451717e-05,
72
+ "loss": 0.0314,
73
+ "step": 900
74
+ },
75
+ {
76
+ "epoch": 1.0,
77
+ "grad_norm": 0.9344207644462585,
78
+ "learning_rate": 4.414230747586468e-05,
79
+ "loss": 0.0356,
80
+ "step": 1000
81
+ },
82
+ {
83
+ "epoch": 1.0,
84
+ "eval_economic_inequality_accuracy": 0.7783375314861462,
85
+ "eval_economic_inequality_f1": 0.6829252350399464,
86
+ "eval_economic_policy_benefits_accuracy": 0.8215488215488216,
87
+ "eval_economic_policy_benefits_f1": 0.752833443442648,
88
+ "eval_ethnic_boundaries_accuracy": 0.9493464052287581,
89
+ "eval_ethnic_boundaries_f1": 0.9305506350938352,
90
+ "eval_language_policy_accuracy": 0.6764705882352942,
91
+ "eval_language_policy_f1": 0.555672268907563,
92
+ "eval_loss": NaN,
93
+ "eval_mother_tongue_education_accuracy": 0.7777777777777778,
94
+ "eval_mother_tongue_education_f1": 0.6805555555555556,
95
+ "eval_overall_accuracy": 0.8090908256773385,
96
+ "eval_overall_f1": 0.7331392486819855,
97
+ "eval_religion_ethnic_identity_accuracy": 0.851063829787234,
98
+ "eval_religion_ethnic_identity_f1": 0.7962983540523647,
99
+ "eval_runtime": 4.4716,
100
+ "eval_samples_per_second": 894.535,
101
+ "eval_steps_per_second": 55.908,
102
+ "step": 1000
103
+ },
104
+ {
105
+ "epoch": 1.1,
106
+ "grad_norm": 0.8767189383506775,
107
+ "learning_rate": 4.303902560943467e-05,
108
+ "loss": 0.3117,
109
+ "step": 1100
110
+ },
111
+ {
112
+ "epoch": 1.2,
113
+ "grad_norm": 32.66496658325195,
114
+ "learning_rate": 4.193574374300466e-05,
115
+ "loss": 0.0098,
116
+ "step": 1200
117
+ },
118
+ {
119
+ "epoch": 1.3,
120
+ "grad_norm": 6.170604705810547,
121
+ "learning_rate": 4.083246187657465e-05,
122
+ "loss": 0.0,
123
+ "step": 1300
124
+ },
125
+ {
126
+ "epoch": 1.4,
127
+ "grad_norm": 0.40170523524284363,
128
+ "learning_rate": 3.972918001014464e-05,
129
+ "loss": 0.2259,
130
+ "step": 1400
131
+ },
132
+ {
133
+ "epoch": 1.5,
134
+ "grad_norm": 3.7784886360168457,
135
+ "learning_rate": 3.862589814371463e-05,
136
+ "loss": 0.0,
137
+ "step": 1500
138
+ },
139
+ {
140
+ "epoch": 1.6,
141
+ "grad_norm": 1.0622837543487549,
142
+ "learning_rate": 3.752261627728462e-05,
143
+ "loss": 0.0,
144
+ "step": 1600
145
+ },
146
+ {
147
+ "epoch": 1.7,
148
+ "grad_norm": 0.3284582793712616,
149
+ "learning_rate": 3.641933441085462e-05,
150
+ "loss": 0.0012,
151
+ "step": 1700
152
+ },
153
+ {
154
+ "epoch": 1.8,
155
+ "grad_norm": 3.2630698680877686,
156
+ "learning_rate": 3.53160525444246e-05,
157
+ "loss": 0.0045,
158
+ "step": 1800
159
+ },
160
+ {
161
+ "epoch": 1.9,
162
+ "grad_norm": 0.4251757860183716,
163
+ "learning_rate": 3.421277067799459e-05,
164
+ "loss": 0.1053,
165
+ "step": 1900
166
+ },
167
+ {
168
+ "epoch": 2.0,
169
+ "grad_norm": 2.5576839447021484,
170
+ "learning_rate": 3.310948881156459e-05,
171
+ "loss": 0.0826,
172
+ "step": 2000
173
+ },
174
+ {
175
+ "epoch": 2.0,
176
+ "eval_economic_inequality_accuracy": 0.7984886649874056,
177
+ "eval_economic_inequality_f1": 0.792653055614301,
178
+ "eval_economic_policy_benefits_accuracy": 0.8215488215488216,
179
+ "eval_economic_policy_benefits_f1": 0.8093298577169545,
180
+ "eval_ethnic_boundaries_accuracy": 0.9444444444444444,
181
+ "eval_ethnic_boundaries_f1": 0.9370640143333165,
182
+ "eval_language_policy_accuracy": 0.5882352941176471,
183
+ "eval_language_policy_f1": 0.5458823529411765,
184
+ "eval_loss": NaN,
185
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
186
+ "eval_mother_tongue_education_f1": 0.837037037037037,
187
+ "eval_overall_accuracy": 0.8148002453034774,
188
+ "eval_overall_f1": 0.7921517057928192,
189
+ "eval_religion_ethnic_identity_accuracy": 0.8471953578336557,
190
+ "eval_religion_ethnic_identity_f1": 0.8309439171141297,
191
+ "eval_runtime": 4.6859,
192
+ "eval_samples_per_second": 853.616,
193
+ "eval_steps_per_second": 53.351,
194
+ "step": 2000
195
+ },
196
+ {
197
+ "epoch": 2.1,
198
+ "grad_norm": 210.1742706298828,
199
+ "learning_rate": 3.200620694513457e-05,
200
+ "loss": 1.2274,
201
+ "step": 2100
202
+ },
203
+ {
204
+ "epoch": 2.2,
205
+ "grad_norm": 3.0127012729644775,
206
+ "learning_rate": 3.090292507870457e-05,
207
+ "loss": 0.0,
208
+ "step": 2200
209
+ },
210
+ {
211
+ "epoch": 2.3,
212
+ "grad_norm": 0.38292792439460754,
213
+ "learning_rate": 2.9799643212274557e-05,
214
+ "loss": 0.0351,
215
+ "step": 2300
216
+ },
217
+ {
218
+ "epoch": 2.4,
219
+ "grad_norm": 2.8801395893096924,
220
+ "learning_rate": 2.8696361345844544e-05,
221
+ "loss": 0.0026,
222
+ "step": 2400
223
+ },
224
+ {
225
+ "epoch": 2.5,
226
+ "grad_norm": 1.5863370895385742,
227
+ "learning_rate": 2.7593079479414537e-05,
228
+ "loss": 0.0,
229
+ "step": 2500
230
+ },
231
+ {
232
+ "epoch": 2.6,
233
+ "grad_norm": 0.23432491719722748,
234
+ "learning_rate": 2.6489797612984527e-05,
235
+ "loss": 0.1192,
236
+ "step": 2600
237
+ },
238
+ {
239
+ "epoch": 2.7,
240
+ "grad_norm": 1.943647027015686,
241
+ "learning_rate": 2.5386515746554517e-05,
242
+ "loss": 0.0195,
243
+ "step": 2700
244
+ },
245
+ {
246
+ "epoch": 2.8,
247
+ "grad_norm": 0.5006923079490662,
248
+ "learning_rate": 2.4283233880124507e-05,
249
+ "loss": 0.0,
250
+ "step": 2800
251
+ },
252
+ {
253
+ "epoch": 2.9,
254
+ "grad_norm": 0.26087385416030884,
255
+ "learning_rate": 2.31799520136945e-05,
256
+ "loss": 0.0,
257
+ "step": 2900
258
+ },
259
+ {
260
+ "epoch": 3.0,
261
+ "grad_norm": 1.207767367362976,
262
+ "learning_rate": 2.207667014726449e-05,
263
+ "loss": 0.1565,
264
+ "step": 3000
265
+ },
266
+ {
267
+ "epoch": 3.0,
268
+ "eval_economic_inequality_accuracy": 0.8148614609571788,
269
+ "eval_economic_inequality_f1": 0.7997421102149133,
270
+ "eval_economic_policy_benefits_accuracy": 0.8518518518518519,
271
+ "eval_economic_policy_benefits_f1": 0.8175032474456659,
272
+ "eval_ethnic_boundaries_accuracy": 0.9477124183006536,
273
+ "eval_ethnic_boundaries_f1": 0.9334997758629278,
274
+ "eval_language_policy_accuracy": 0.6764705882352942,
275
+ "eval_language_policy_f1": 0.5869281045751634,
276
+ "eval_loss": NaN,
277
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
278
+ "eval_mother_tongue_education_f1": 0.837037037037037,
279
+ "eval_overall_accuracy": 0.8371853490189908,
280
+ "eval_overall_f1": 0.7988061056554884,
281
+ "eval_religion_ethnic_identity_accuracy": 0.8433268858800773,
282
+ "eval_religion_ethnic_identity_f1": 0.8181263587972226,
283
+ "eval_runtime": 4.4894,
284
+ "eval_samples_per_second": 890.983,
285
+ "eval_steps_per_second": 55.686,
286
+ "step": 3000
287
+ }
288
+ ],
289
+ "logging_steps": 100,
290
+ "max_steps": 5000,
291
+ "num_input_tokens_seen": 0,
292
+ "num_train_epochs": 5,
293
+ "save_steps": 500,
294
+ "stateful_callbacks": {
295
+ "EarlyStoppingCallback": {
296
+ "args": {
297
+ "early_stopping_patience": 2,
298
+ "early_stopping_threshold": 0.0
299
+ },
300
+ "attributes": {
301
+ "early_stopping_patience_counter": 0
302
+ }
303
+ },
304
+ "TrainerControl": {
305
+ "args": {
306
+ "should_epoch_stop": false,
307
+ "should_evaluate": false,
308
+ "should_log": false,
309
+ "should_save": true,
310
+ "should_training_stop": false
311
+ },
312
+ "attributes": {}
313
+ }
314
+ },
315
+ "total_flos": 1.2631144955904e+16,
316
+ "train_batch_size": 16,
317
+ "trial_name": null,
318
+ "trial_params": {
319
+ "gradient_accumulation_steps": 1,
320
+ "learning_rate": 4.452845612911518e-05,
321
+ "num_train_epochs": 5,
322
+ "per_device_train_batch_size": 16,
323
+ "warmup_steps": 964,
324
+ "weight_decay": 0.08427803271887896
325
+ }
326
+ }
run-4/checkpoint-3000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4ad559f091ae7757869f5224f652bb49e6e3989ffd03fbb0eefdb9252513c40
3
+ size 5777
run-4/checkpoint-4000/config.json ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "LABEL_0",
13
+ "1": "LABEL_1",
14
+ "2": "LABEL_2",
15
+ "3": "LABEL_3",
16
+ "4": "LABEL_4",
17
+ "5": "LABEL_5",
18
+ "6": "LABEL_6",
19
+ "7": "LABEL_7",
20
+ "8": "LABEL_8",
21
+ "9": "LABEL_9",
22
+ "10": "LABEL_10",
23
+ "11": "LABEL_11",
24
+ "12": "LABEL_12",
25
+ "13": "LABEL_13",
26
+ "14": "LABEL_14",
27
+ "15": "LABEL_15",
28
+ "16": "LABEL_16",
29
+ "17": "LABEL_17"
30
+ },
31
+ "initializer_range": 0.02,
32
+ "intermediate_size": 3072,
33
+ "label2id": {
34
+ "LABEL_0": 0,
35
+ "LABEL_1": 1,
36
+ "LABEL_10": 10,
37
+ "LABEL_11": 11,
38
+ "LABEL_12": 12,
39
+ "LABEL_13": 13,
40
+ "LABEL_14": 14,
41
+ "LABEL_15": 15,
42
+ "LABEL_16": 16,
43
+ "LABEL_17": 17,
44
+ "LABEL_2": 2,
45
+ "LABEL_3": 3,
46
+ "LABEL_4": 4,
47
+ "LABEL_5": 5,
48
+ "LABEL_6": 6,
49
+ "LABEL_7": 7,
50
+ "LABEL_8": 8,
51
+ "LABEL_9": 9
52
+ },
53
+ "layer_norm_eps": 1e-12,
54
+ "max_position_embeddings": 512,
55
+ "model_type": "bert",
56
+ "num_attention_heads": 12,
57
+ "num_hidden_layers": 12,
58
+ "pad_token_id": 0,
59
+ "position_embedding_type": "absolute",
60
+ "torch_dtype": "float32",
61
+ "transformers_version": "4.55.4",
62
+ "type_vocab_size": 2,
63
+ "use_cache": true,
64
+ "vocab_size": 30522
65
+ }
run-4/checkpoint-4000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3bac87b5c062ea0d27a770b51e3950a2933b0dfd9d9c203e259ad2567c57d19
3
+ size 438007864
run-4/checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d40344be585ee37464135b83c24195198e2456208a008bb7521f165aa19a8e17
3
+ size 876140043
run-4/checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fb97b06d0acaf299ecfc2800cce464cd782890febfa378a20e287092543287a
3
+ size 14645
run-4/checkpoint-4000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63e069d76e5e48d538cbcdcb40c54890cef37c6fcb59cede3f543ecafdb98bed
3
+ size 1383
run-4/checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f53b64b82dadbe92a95863b0ae8097107860248f84ec3966d66f68e06a65c5a8
3
+ size 1465
run-4/checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,418 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 4000,
3
+ "best_metric": 0.8135900008491362,
4
+ "best_model_checkpoint": "./results/run-4/checkpoint-4000",
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 4000,
8
+ "is_hyper_param_search": true,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1,
14
+ "grad_norm": 3.002380609512329,
15
+ "learning_rate": 4.572943108695439e-06,
16
+ "loss": 0.296,
17
+ "step": 100
18
+ },
19
+ {
20
+ "epoch": 0.2,
21
+ "grad_norm": 0.9367781281471252,
22
+ "learning_rate": 9.192077561923155e-06,
23
+ "loss": 1.0274,
24
+ "step": 200
25
+ },
26
+ {
27
+ "epoch": 0.3,
28
+ "grad_norm": 0.7274576425552368,
29
+ "learning_rate": 1.3811212015150872e-05,
30
+ "loss": 0.0,
31
+ "step": 300
32
+ },
33
+ {
34
+ "epoch": 0.4,
35
+ "grad_norm": 44.297061920166016,
36
+ "learning_rate": 1.8430346468378586e-05,
37
+ "loss": 0.0,
38
+ "step": 400
39
+ },
40
+ {
41
+ "epoch": 0.5,
42
+ "grad_norm": 1.0151411294937134,
43
+ "learning_rate": 2.3049480921606306e-05,
44
+ "loss": 0.0,
45
+ "step": 500
46
+ },
47
+ {
48
+ "epoch": 0.6,
49
+ "grad_norm": 1.208032488822937,
50
+ "learning_rate": 2.766861537483402e-05,
51
+ "loss": 0.0,
52
+ "step": 600
53
+ },
54
+ {
55
+ "epoch": 0.7,
56
+ "grad_norm": 6.984821319580078,
57
+ "learning_rate": 3.2287749828061733e-05,
58
+ "loss": 0.0342,
59
+ "step": 700
60
+ },
61
+ {
62
+ "epoch": 0.8,
63
+ "grad_norm": 1.5860588550567627,
64
+ "learning_rate": 3.690688428128945e-05,
65
+ "loss": 0.0488,
66
+ "step": 800
67
+ },
68
+ {
69
+ "epoch": 0.9,
70
+ "grad_norm": 0.7841507196426392,
71
+ "learning_rate": 4.152601873451717e-05,
72
+ "loss": 0.0314,
73
+ "step": 900
74
+ },
75
+ {
76
+ "epoch": 1.0,
77
+ "grad_norm": 0.9344207644462585,
78
+ "learning_rate": 4.414230747586468e-05,
79
+ "loss": 0.0356,
80
+ "step": 1000
81
+ },
82
+ {
83
+ "epoch": 1.0,
84
+ "eval_economic_inequality_accuracy": 0.7783375314861462,
85
+ "eval_economic_inequality_f1": 0.6829252350399464,
86
+ "eval_economic_policy_benefits_accuracy": 0.8215488215488216,
87
+ "eval_economic_policy_benefits_f1": 0.752833443442648,
88
+ "eval_ethnic_boundaries_accuracy": 0.9493464052287581,
89
+ "eval_ethnic_boundaries_f1": 0.9305506350938352,
90
+ "eval_language_policy_accuracy": 0.6764705882352942,
91
+ "eval_language_policy_f1": 0.555672268907563,
92
+ "eval_loss": NaN,
93
+ "eval_mother_tongue_education_accuracy": 0.7777777777777778,
94
+ "eval_mother_tongue_education_f1": 0.6805555555555556,
95
+ "eval_overall_accuracy": 0.8090908256773385,
96
+ "eval_overall_f1": 0.7331392486819855,
97
+ "eval_religion_ethnic_identity_accuracy": 0.851063829787234,
98
+ "eval_religion_ethnic_identity_f1": 0.7962983540523647,
99
+ "eval_runtime": 4.4716,
100
+ "eval_samples_per_second": 894.535,
101
+ "eval_steps_per_second": 55.908,
102
+ "step": 1000
103
+ },
104
+ {
105
+ "epoch": 1.1,
106
+ "grad_norm": 0.8767189383506775,
107
+ "learning_rate": 4.303902560943467e-05,
108
+ "loss": 0.3117,
109
+ "step": 1100
110
+ },
111
+ {
112
+ "epoch": 1.2,
113
+ "grad_norm": 32.66496658325195,
114
+ "learning_rate": 4.193574374300466e-05,
115
+ "loss": 0.0098,
116
+ "step": 1200
117
+ },
118
+ {
119
+ "epoch": 1.3,
120
+ "grad_norm": 6.170604705810547,
121
+ "learning_rate": 4.083246187657465e-05,
122
+ "loss": 0.0,
123
+ "step": 1300
124
+ },
125
+ {
126
+ "epoch": 1.4,
127
+ "grad_norm": 0.40170523524284363,
128
+ "learning_rate": 3.972918001014464e-05,
129
+ "loss": 0.2259,
130
+ "step": 1400
131
+ },
132
+ {
133
+ "epoch": 1.5,
134
+ "grad_norm": 3.7784886360168457,
135
+ "learning_rate": 3.862589814371463e-05,
136
+ "loss": 0.0,
137
+ "step": 1500
138
+ },
139
+ {
140
+ "epoch": 1.6,
141
+ "grad_norm": 1.0622837543487549,
142
+ "learning_rate": 3.752261627728462e-05,
143
+ "loss": 0.0,
144
+ "step": 1600
145
+ },
146
+ {
147
+ "epoch": 1.7,
148
+ "grad_norm": 0.3284582793712616,
149
+ "learning_rate": 3.641933441085462e-05,
150
+ "loss": 0.0012,
151
+ "step": 1700
152
+ },
153
+ {
154
+ "epoch": 1.8,
155
+ "grad_norm": 3.2630698680877686,
156
+ "learning_rate": 3.53160525444246e-05,
157
+ "loss": 0.0045,
158
+ "step": 1800
159
+ },
160
+ {
161
+ "epoch": 1.9,
162
+ "grad_norm": 0.4251757860183716,
163
+ "learning_rate": 3.421277067799459e-05,
164
+ "loss": 0.1053,
165
+ "step": 1900
166
+ },
167
+ {
168
+ "epoch": 2.0,
169
+ "grad_norm": 2.5576839447021484,
170
+ "learning_rate": 3.310948881156459e-05,
171
+ "loss": 0.0826,
172
+ "step": 2000
173
+ },
174
+ {
175
+ "epoch": 2.0,
176
+ "eval_economic_inequality_accuracy": 0.7984886649874056,
177
+ "eval_economic_inequality_f1": 0.792653055614301,
178
+ "eval_economic_policy_benefits_accuracy": 0.8215488215488216,
179
+ "eval_economic_policy_benefits_f1": 0.8093298577169545,
180
+ "eval_ethnic_boundaries_accuracy": 0.9444444444444444,
181
+ "eval_ethnic_boundaries_f1": 0.9370640143333165,
182
+ "eval_language_policy_accuracy": 0.5882352941176471,
183
+ "eval_language_policy_f1": 0.5458823529411765,
184
+ "eval_loss": NaN,
185
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
186
+ "eval_mother_tongue_education_f1": 0.837037037037037,
187
+ "eval_overall_accuracy": 0.8148002453034774,
188
+ "eval_overall_f1": 0.7921517057928192,
189
+ "eval_religion_ethnic_identity_accuracy": 0.8471953578336557,
190
+ "eval_religion_ethnic_identity_f1": 0.8309439171141297,
191
+ "eval_runtime": 4.6859,
192
+ "eval_samples_per_second": 853.616,
193
+ "eval_steps_per_second": 53.351,
194
+ "step": 2000
195
+ },
196
+ {
197
+ "epoch": 2.1,
198
+ "grad_norm": 210.1742706298828,
199
+ "learning_rate": 3.200620694513457e-05,
200
+ "loss": 1.2274,
201
+ "step": 2100
202
+ },
203
+ {
204
+ "epoch": 2.2,
205
+ "grad_norm": 3.0127012729644775,
206
+ "learning_rate": 3.090292507870457e-05,
207
+ "loss": 0.0,
208
+ "step": 2200
209
+ },
210
+ {
211
+ "epoch": 2.3,
212
+ "grad_norm": 0.38292792439460754,
213
+ "learning_rate": 2.9799643212274557e-05,
214
+ "loss": 0.0351,
215
+ "step": 2300
216
+ },
217
+ {
218
+ "epoch": 2.4,
219
+ "grad_norm": 2.8801395893096924,
220
+ "learning_rate": 2.8696361345844544e-05,
221
+ "loss": 0.0026,
222
+ "step": 2400
223
+ },
224
+ {
225
+ "epoch": 2.5,
226
+ "grad_norm": 1.5863370895385742,
227
+ "learning_rate": 2.7593079479414537e-05,
228
+ "loss": 0.0,
229
+ "step": 2500
230
+ },
231
+ {
232
+ "epoch": 2.6,
233
+ "grad_norm": 0.23432491719722748,
234
+ "learning_rate": 2.6489797612984527e-05,
235
+ "loss": 0.1192,
236
+ "step": 2600
237
+ },
238
+ {
239
+ "epoch": 2.7,
240
+ "grad_norm": 1.943647027015686,
241
+ "learning_rate": 2.5386515746554517e-05,
242
+ "loss": 0.0195,
243
+ "step": 2700
244
+ },
245
+ {
246
+ "epoch": 2.8,
247
+ "grad_norm": 0.5006923079490662,
248
+ "learning_rate": 2.4283233880124507e-05,
249
+ "loss": 0.0,
250
+ "step": 2800
251
+ },
252
+ {
253
+ "epoch": 2.9,
254
+ "grad_norm": 0.26087385416030884,
255
+ "learning_rate": 2.31799520136945e-05,
256
+ "loss": 0.0,
257
+ "step": 2900
258
+ },
259
+ {
260
+ "epoch": 3.0,
261
+ "grad_norm": 1.207767367362976,
262
+ "learning_rate": 2.207667014726449e-05,
263
+ "loss": 0.1565,
264
+ "step": 3000
265
+ },
266
+ {
267
+ "epoch": 3.0,
268
+ "eval_economic_inequality_accuracy": 0.8148614609571788,
269
+ "eval_economic_inequality_f1": 0.7997421102149133,
270
+ "eval_economic_policy_benefits_accuracy": 0.8518518518518519,
271
+ "eval_economic_policy_benefits_f1": 0.8175032474456659,
272
+ "eval_ethnic_boundaries_accuracy": 0.9477124183006536,
273
+ "eval_ethnic_boundaries_f1": 0.9334997758629278,
274
+ "eval_language_policy_accuracy": 0.6764705882352942,
275
+ "eval_language_policy_f1": 0.5869281045751634,
276
+ "eval_loss": NaN,
277
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
278
+ "eval_mother_tongue_education_f1": 0.837037037037037,
279
+ "eval_overall_accuracy": 0.8371853490189908,
280
+ "eval_overall_f1": 0.7988061056554884,
281
+ "eval_religion_ethnic_identity_accuracy": 0.8433268858800773,
282
+ "eval_religion_ethnic_identity_f1": 0.8181263587972226,
283
+ "eval_runtime": 4.4894,
284
+ "eval_samples_per_second": 890.983,
285
+ "eval_steps_per_second": 55.686,
286
+ "step": 3000
287
+ },
288
+ {
289
+ "epoch": 3.1,
290
+ "grad_norm": 0.10033319890499115,
291
+ "learning_rate": 2.097338828083448e-05,
292
+ "loss": 0.0013,
293
+ "step": 3100
294
+ },
295
+ {
296
+ "epoch": 3.2,
297
+ "grad_norm": 0.0292133130133152,
298
+ "learning_rate": 1.987010641440447e-05,
299
+ "loss": 0.1935,
300
+ "step": 3200
301
+ },
302
+ {
303
+ "epoch": 3.3,
304
+ "grad_norm": 0.9771420955657959,
305
+ "learning_rate": 1.876682454797446e-05,
306
+ "loss": 0.0026,
307
+ "step": 3300
308
+ },
309
+ {
310
+ "epoch": 3.4,
311
+ "grad_norm": 0.13967016339302063,
312
+ "learning_rate": 1.7663542681544454e-05,
313
+ "loss": 0.0042,
314
+ "step": 3400
315
+ },
316
+ {
317
+ "epoch": 3.5,
318
+ "grad_norm": 44.07572937011719,
319
+ "learning_rate": 1.6560260815114444e-05,
320
+ "loss": 0.0002,
321
+ "step": 3500
322
+ },
323
+ {
324
+ "epoch": 3.6,
325
+ "grad_norm": 4.053230285644531,
326
+ "learning_rate": 1.545697894868443e-05,
327
+ "loss": 0.0109,
328
+ "step": 3600
329
+ },
330
+ {
331
+ "epoch": 3.7,
332
+ "grad_norm": 2.6551458835601807,
333
+ "learning_rate": 1.4353697082254424e-05,
334
+ "loss": 0.0001,
335
+ "step": 3700
336
+ },
337
+ {
338
+ "epoch": 3.8,
339
+ "grad_norm": 0.14367616176605225,
340
+ "learning_rate": 1.3250415215824414e-05,
341
+ "loss": 0.0001,
342
+ "step": 3800
343
+ },
344
+ {
345
+ "epoch": 3.9,
346
+ "grad_norm": 0.05590776354074478,
347
+ "learning_rate": 1.2147133349394404e-05,
348
+ "loss": 0.0086,
349
+ "step": 3900
350
+ },
351
+ {
352
+ "epoch": 4.0,
353
+ "grad_norm": 0.6181541681289673,
354
+ "learning_rate": 1.1043851482964394e-05,
355
+ "loss": 0.0789,
356
+ "step": 4000
357
+ },
358
+ {
359
+ "epoch": 4.0,
360
+ "eval_economic_inequality_accuracy": 0.8123425692695214,
361
+ "eval_economic_inequality_f1": 0.8054350386595589,
362
+ "eval_economic_policy_benefits_accuracy": 0.8518518518518519,
363
+ "eval_economic_policy_benefits_f1": 0.8269803894803894,
364
+ "eval_ethnic_boundaries_accuracy": 0.9395424836601307,
365
+ "eval_ethnic_boundaries_f1": 0.9411192800317346,
366
+ "eval_language_policy_accuracy": 0.7058823529411765,
367
+ "eval_language_policy_f1": 0.6335511982570806,
368
+ "eval_loss": NaN,
369
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
370
+ "eval_mother_tongue_education_f1": 0.837037037037037,
371
+ "eval_overall_accuracy": 0.8390163480974149,
372
+ "eval_overall_f1": 0.8135900008491362,
373
+ "eval_religion_ethnic_identity_accuracy": 0.8355899419729207,
374
+ "eval_religion_ethnic_identity_f1": 0.8374170616290167,
375
+ "eval_runtime": 5.6175,
376
+ "eval_samples_per_second": 712.064,
377
+ "eval_steps_per_second": 44.504,
378
+ "step": 4000
379
+ }
380
+ ],
381
+ "logging_steps": 100,
382
+ "max_steps": 5000,
383
+ "num_input_tokens_seen": 0,
384
+ "num_train_epochs": 5,
385
+ "save_steps": 500,
386
+ "stateful_callbacks": {
387
+ "EarlyStoppingCallback": {
388
+ "args": {
389
+ "early_stopping_patience": 2,
390
+ "early_stopping_threshold": 0.0
391
+ },
392
+ "attributes": {
393
+ "early_stopping_patience_counter": 0
394
+ }
395
+ },
396
+ "TrainerControl": {
397
+ "args": {
398
+ "should_epoch_stop": false,
399
+ "should_evaluate": false,
400
+ "should_log": false,
401
+ "should_save": true,
402
+ "should_training_stop": false
403
+ },
404
+ "attributes": {}
405
+ }
406
+ },
407
+ "total_flos": 1.6841526607872e+16,
408
+ "train_batch_size": 16,
409
+ "trial_name": null,
410
+ "trial_params": {
411
+ "gradient_accumulation_steps": 1,
412
+ "learning_rate": 4.452845612911518e-05,
413
+ "num_train_epochs": 5,
414
+ "per_device_train_batch_size": 16,
415
+ "warmup_steps": 964,
416
+ "weight_decay": 0.08427803271887896
417
+ }
418
+ }
run-4/checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4ad559f091ae7757869f5224f652bb49e6e3989ffd03fbb0eefdb9252513c40
3
+ size 5777
run-4/checkpoint-5000/config.json ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "LABEL_0",
13
+ "1": "LABEL_1",
14
+ "2": "LABEL_2",
15
+ "3": "LABEL_3",
16
+ "4": "LABEL_4",
17
+ "5": "LABEL_5",
18
+ "6": "LABEL_6",
19
+ "7": "LABEL_7",
20
+ "8": "LABEL_8",
21
+ "9": "LABEL_9",
22
+ "10": "LABEL_10",
23
+ "11": "LABEL_11",
24
+ "12": "LABEL_12",
25
+ "13": "LABEL_13",
26
+ "14": "LABEL_14",
27
+ "15": "LABEL_15",
28
+ "16": "LABEL_16",
29
+ "17": "LABEL_17"
30
+ },
31
+ "initializer_range": 0.02,
32
+ "intermediate_size": 3072,
33
+ "label2id": {
34
+ "LABEL_0": 0,
35
+ "LABEL_1": 1,
36
+ "LABEL_10": 10,
37
+ "LABEL_11": 11,
38
+ "LABEL_12": 12,
39
+ "LABEL_13": 13,
40
+ "LABEL_14": 14,
41
+ "LABEL_15": 15,
42
+ "LABEL_16": 16,
43
+ "LABEL_17": 17,
44
+ "LABEL_2": 2,
45
+ "LABEL_3": 3,
46
+ "LABEL_4": 4,
47
+ "LABEL_5": 5,
48
+ "LABEL_6": 6,
49
+ "LABEL_7": 7,
50
+ "LABEL_8": 8,
51
+ "LABEL_9": 9
52
+ },
53
+ "layer_norm_eps": 1e-12,
54
+ "max_position_embeddings": 512,
55
+ "model_type": "bert",
56
+ "num_attention_heads": 12,
57
+ "num_hidden_layers": 12,
58
+ "pad_token_id": 0,
59
+ "position_embedding_type": "absolute",
60
+ "torch_dtype": "float32",
61
+ "transformers_version": "4.55.4",
62
+ "type_vocab_size": 2,
63
+ "use_cache": true,
64
+ "vocab_size": 30522
65
+ }
run-4/checkpoint-5000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93aefbf4224c7c4a25fabf43ef101a9e27c0037edb24c5ca01420badb9cd7c80
3
+ size 438007864
run-4/checkpoint-5000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a26ec9be0dd102951d52449d0333ddbe6f8b9a61cf96251075d5fb22c2ad2b8e
3
+ size 876140043
run-4/checkpoint-5000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c375d96697a678bb14ca85ae4c2d3936cefbaf222cc9914537051fc32cab6df1
3
+ size 14645
run-4/checkpoint-5000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80adc64877f6fd242e1e9aa8a3c041631d448006055749fbb933806f057fc3e9
3
+ size 1383
run-4/checkpoint-5000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:326856e02406ea0bae41a46c778ec9615cf54e6c5335b2cd4ef58326598ec8b2
3
+ size 1465
run-4/checkpoint-5000/trainer_state.json ADDED
@@ -0,0 +1,510 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 5000,
3
+ "best_metric": 0.8217308919703874,
4
+ "best_model_checkpoint": "./results/run-4/checkpoint-5000",
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 5000,
8
+ "is_hyper_param_search": true,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1,
14
+ "grad_norm": 3.002380609512329,
15
+ "learning_rate": 4.572943108695439e-06,
16
+ "loss": 0.296,
17
+ "step": 100
18
+ },
19
+ {
20
+ "epoch": 0.2,
21
+ "grad_norm": 0.9367781281471252,
22
+ "learning_rate": 9.192077561923155e-06,
23
+ "loss": 1.0274,
24
+ "step": 200
25
+ },
26
+ {
27
+ "epoch": 0.3,
28
+ "grad_norm": 0.7274576425552368,
29
+ "learning_rate": 1.3811212015150872e-05,
30
+ "loss": 0.0,
31
+ "step": 300
32
+ },
33
+ {
34
+ "epoch": 0.4,
35
+ "grad_norm": 44.297061920166016,
36
+ "learning_rate": 1.8430346468378586e-05,
37
+ "loss": 0.0,
38
+ "step": 400
39
+ },
40
+ {
41
+ "epoch": 0.5,
42
+ "grad_norm": 1.0151411294937134,
43
+ "learning_rate": 2.3049480921606306e-05,
44
+ "loss": 0.0,
45
+ "step": 500
46
+ },
47
+ {
48
+ "epoch": 0.6,
49
+ "grad_norm": 1.208032488822937,
50
+ "learning_rate": 2.766861537483402e-05,
51
+ "loss": 0.0,
52
+ "step": 600
53
+ },
54
+ {
55
+ "epoch": 0.7,
56
+ "grad_norm": 6.984821319580078,
57
+ "learning_rate": 3.2287749828061733e-05,
58
+ "loss": 0.0342,
59
+ "step": 700
60
+ },
61
+ {
62
+ "epoch": 0.8,
63
+ "grad_norm": 1.5860588550567627,
64
+ "learning_rate": 3.690688428128945e-05,
65
+ "loss": 0.0488,
66
+ "step": 800
67
+ },
68
+ {
69
+ "epoch": 0.9,
70
+ "grad_norm": 0.7841507196426392,
71
+ "learning_rate": 4.152601873451717e-05,
72
+ "loss": 0.0314,
73
+ "step": 900
74
+ },
75
+ {
76
+ "epoch": 1.0,
77
+ "grad_norm": 0.9344207644462585,
78
+ "learning_rate": 4.414230747586468e-05,
79
+ "loss": 0.0356,
80
+ "step": 1000
81
+ },
82
+ {
83
+ "epoch": 1.0,
84
+ "eval_economic_inequality_accuracy": 0.7783375314861462,
85
+ "eval_economic_inequality_f1": 0.6829252350399464,
86
+ "eval_economic_policy_benefits_accuracy": 0.8215488215488216,
87
+ "eval_economic_policy_benefits_f1": 0.752833443442648,
88
+ "eval_ethnic_boundaries_accuracy": 0.9493464052287581,
89
+ "eval_ethnic_boundaries_f1": 0.9305506350938352,
90
+ "eval_language_policy_accuracy": 0.6764705882352942,
91
+ "eval_language_policy_f1": 0.555672268907563,
92
+ "eval_loss": NaN,
93
+ "eval_mother_tongue_education_accuracy": 0.7777777777777778,
94
+ "eval_mother_tongue_education_f1": 0.6805555555555556,
95
+ "eval_overall_accuracy": 0.8090908256773385,
96
+ "eval_overall_f1": 0.7331392486819855,
97
+ "eval_religion_ethnic_identity_accuracy": 0.851063829787234,
98
+ "eval_religion_ethnic_identity_f1": 0.7962983540523647,
99
+ "eval_runtime": 4.4716,
100
+ "eval_samples_per_second": 894.535,
101
+ "eval_steps_per_second": 55.908,
102
+ "step": 1000
103
+ },
104
+ {
105
+ "epoch": 1.1,
106
+ "grad_norm": 0.8767189383506775,
107
+ "learning_rate": 4.303902560943467e-05,
108
+ "loss": 0.3117,
109
+ "step": 1100
110
+ },
111
+ {
112
+ "epoch": 1.2,
113
+ "grad_norm": 32.66496658325195,
114
+ "learning_rate": 4.193574374300466e-05,
115
+ "loss": 0.0098,
116
+ "step": 1200
117
+ },
118
+ {
119
+ "epoch": 1.3,
120
+ "grad_norm": 6.170604705810547,
121
+ "learning_rate": 4.083246187657465e-05,
122
+ "loss": 0.0,
123
+ "step": 1300
124
+ },
125
+ {
126
+ "epoch": 1.4,
127
+ "grad_norm": 0.40170523524284363,
128
+ "learning_rate": 3.972918001014464e-05,
129
+ "loss": 0.2259,
130
+ "step": 1400
131
+ },
132
+ {
133
+ "epoch": 1.5,
134
+ "grad_norm": 3.7784886360168457,
135
+ "learning_rate": 3.862589814371463e-05,
136
+ "loss": 0.0,
137
+ "step": 1500
138
+ },
139
+ {
140
+ "epoch": 1.6,
141
+ "grad_norm": 1.0622837543487549,
142
+ "learning_rate": 3.752261627728462e-05,
143
+ "loss": 0.0,
144
+ "step": 1600
145
+ },
146
+ {
147
+ "epoch": 1.7,
148
+ "grad_norm": 0.3284582793712616,
149
+ "learning_rate": 3.641933441085462e-05,
150
+ "loss": 0.0012,
151
+ "step": 1700
152
+ },
153
+ {
154
+ "epoch": 1.8,
155
+ "grad_norm": 3.2630698680877686,
156
+ "learning_rate": 3.53160525444246e-05,
157
+ "loss": 0.0045,
158
+ "step": 1800
159
+ },
160
+ {
161
+ "epoch": 1.9,
162
+ "grad_norm": 0.4251757860183716,
163
+ "learning_rate": 3.421277067799459e-05,
164
+ "loss": 0.1053,
165
+ "step": 1900
166
+ },
167
+ {
168
+ "epoch": 2.0,
169
+ "grad_norm": 2.5576839447021484,
170
+ "learning_rate": 3.310948881156459e-05,
171
+ "loss": 0.0826,
172
+ "step": 2000
173
+ },
174
+ {
175
+ "epoch": 2.0,
176
+ "eval_economic_inequality_accuracy": 0.7984886649874056,
177
+ "eval_economic_inequality_f1": 0.792653055614301,
178
+ "eval_economic_policy_benefits_accuracy": 0.8215488215488216,
179
+ "eval_economic_policy_benefits_f1": 0.8093298577169545,
180
+ "eval_ethnic_boundaries_accuracy": 0.9444444444444444,
181
+ "eval_ethnic_boundaries_f1": 0.9370640143333165,
182
+ "eval_language_policy_accuracy": 0.5882352941176471,
183
+ "eval_language_policy_f1": 0.5458823529411765,
184
+ "eval_loss": NaN,
185
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
186
+ "eval_mother_tongue_education_f1": 0.837037037037037,
187
+ "eval_overall_accuracy": 0.8148002453034774,
188
+ "eval_overall_f1": 0.7921517057928192,
189
+ "eval_religion_ethnic_identity_accuracy": 0.8471953578336557,
190
+ "eval_religion_ethnic_identity_f1": 0.8309439171141297,
191
+ "eval_runtime": 4.6859,
192
+ "eval_samples_per_second": 853.616,
193
+ "eval_steps_per_second": 53.351,
194
+ "step": 2000
195
+ },
196
+ {
197
+ "epoch": 2.1,
198
+ "grad_norm": 210.1742706298828,
199
+ "learning_rate": 3.200620694513457e-05,
200
+ "loss": 1.2274,
201
+ "step": 2100
202
+ },
203
+ {
204
+ "epoch": 2.2,
205
+ "grad_norm": 3.0127012729644775,
206
+ "learning_rate": 3.090292507870457e-05,
207
+ "loss": 0.0,
208
+ "step": 2200
209
+ },
210
+ {
211
+ "epoch": 2.3,
212
+ "grad_norm": 0.38292792439460754,
213
+ "learning_rate": 2.9799643212274557e-05,
214
+ "loss": 0.0351,
215
+ "step": 2300
216
+ },
217
+ {
218
+ "epoch": 2.4,
219
+ "grad_norm": 2.8801395893096924,
220
+ "learning_rate": 2.8696361345844544e-05,
221
+ "loss": 0.0026,
222
+ "step": 2400
223
+ },
224
+ {
225
+ "epoch": 2.5,
226
+ "grad_norm": 1.5863370895385742,
227
+ "learning_rate": 2.7593079479414537e-05,
228
+ "loss": 0.0,
229
+ "step": 2500
230
+ },
231
+ {
232
+ "epoch": 2.6,
233
+ "grad_norm": 0.23432491719722748,
234
+ "learning_rate": 2.6489797612984527e-05,
235
+ "loss": 0.1192,
236
+ "step": 2600
237
+ },
238
+ {
239
+ "epoch": 2.7,
240
+ "grad_norm": 1.943647027015686,
241
+ "learning_rate": 2.5386515746554517e-05,
242
+ "loss": 0.0195,
243
+ "step": 2700
244
+ },
245
+ {
246
+ "epoch": 2.8,
247
+ "grad_norm": 0.5006923079490662,
248
+ "learning_rate": 2.4283233880124507e-05,
249
+ "loss": 0.0,
250
+ "step": 2800
251
+ },
252
+ {
253
+ "epoch": 2.9,
254
+ "grad_norm": 0.26087385416030884,
255
+ "learning_rate": 2.31799520136945e-05,
256
+ "loss": 0.0,
257
+ "step": 2900
258
+ },
259
+ {
260
+ "epoch": 3.0,
261
+ "grad_norm": 1.207767367362976,
262
+ "learning_rate": 2.207667014726449e-05,
263
+ "loss": 0.1565,
264
+ "step": 3000
265
+ },
266
+ {
267
+ "epoch": 3.0,
268
+ "eval_economic_inequality_accuracy": 0.8148614609571788,
269
+ "eval_economic_inequality_f1": 0.7997421102149133,
270
+ "eval_economic_policy_benefits_accuracy": 0.8518518518518519,
271
+ "eval_economic_policy_benefits_f1": 0.8175032474456659,
272
+ "eval_ethnic_boundaries_accuracy": 0.9477124183006536,
273
+ "eval_ethnic_boundaries_f1": 0.9334997758629278,
274
+ "eval_language_policy_accuracy": 0.6764705882352942,
275
+ "eval_language_policy_f1": 0.5869281045751634,
276
+ "eval_loss": NaN,
277
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
278
+ "eval_mother_tongue_education_f1": 0.837037037037037,
279
+ "eval_overall_accuracy": 0.8371853490189908,
280
+ "eval_overall_f1": 0.7988061056554884,
281
+ "eval_religion_ethnic_identity_accuracy": 0.8433268858800773,
282
+ "eval_religion_ethnic_identity_f1": 0.8181263587972226,
283
+ "eval_runtime": 4.4894,
284
+ "eval_samples_per_second": 890.983,
285
+ "eval_steps_per_second": 55.686,
286
+ "step": 3000
287
+ },
288
+ {
289
+ "epoch": 3.1,
290
+ "grad_norm": 0.10033319890499115,
291
+ "learning_rate": 2.097338828083448e-05,
292
+ "loss": 0.0013,
293
+ "step": 3100
294
+ },
295
+ {
296
+ "epoch": 3.2,
297
+ "grad_norm": 0.0292133130133152,
298
+ "learning_rate": 1.987010641440447e-05,
299
+ "loss": 0.1935,
300
+ "step": 3200
301
+ },
302
+ {
303
+ "epoch": 3.3,
304
+ "grad_norm": 0.9771420955657959,
305
+ "learning_rate": 1.876682454797446e-05,
306
+ "loss": 0.0026,
307
+ "step": 3300
308
+ },
309
+ {
310
+ "epoch": 3.4,
311
+ "grad_norm": 0.13967016339302063,
312
+ "learning_rate": 1.7663542681544454e-05,
313
+ "loss": 0.0042,
314
+ "step": 3400
315
+ },
316
+ {
317
+ "epoch": 3.5,
318
+ "grad_norm": 44.07572937011719,
319
+ "learning_rate": 1.6560260815114444e-05,
320
+ "loss": 0.0002,
321
+ "step": 3500
322
+ },
323
+ {
324
+ "epoch": 3.6,
325
+ "grad_norm": 4.053230285644531,
326
+ "learning_rate": 1.545697894868443e-05,
327
+ "loss": 0.0109,
328
+ "step": 3600
329
+ },
330
+ {
331
+ "epoch": 3.7,
332
+ "grad_norm": 2.6551458835601807,
333
+ "learning_rate": 1.4353697082254424e-05,
334
+ "loss": 0.0001,
335
+ "step": 3700
336
+ },
337
+ {
338
+ "epoch": 3.8,
339
+ "grad_norm": 0.14367616176605225,
340
+ "learning_rate": 1.3250415215824414e-05,
341
+ "loss": 0.0001,
342
+ "step": 3800
343
+ },
344
+ {
345
+ "epoch": 3.9,
346
+ "grad_norm": 0.05590776354074478,
347
+ "learning_rate": 1.2147133349394404e-05,
348
+ "loss": 0.0086,
349
+ "step": 3900
350
+ },
351
+ {
352
+ "epoch": 4.0,
353
+ "grad_norm": 0.6181541681289673,
354
+ "learning_rate": 1.1043851482964394e-05,
355
+ "loss": 0.0789,
356
+ "step": 4000
357
+ },
358
+ {
359
+ "epoch": 4.0,
360
+ "eval_economic_inequality_accuracy": 0.8123425692695214,
361
+ "eval_economic_inequality_f1": 0.8054350386595589,
362
+ "eval_economic_policy_benefits_accuracy": 0.8518518518518519,
363
+ "eval_economic_policy_benefits_f1": 0.8269803894803894,
364
+ "eval_ethnic_boundaries_accuracy": 0.9395424836601307,
365
+ "eval_ethnic_boundaries_f1": 0.9411192800317346,
366
+ "eval_language_policy_accuracy": 0.7058823529411765,
367
+ "eval_language_policy_f1": 0.6335511982570806,
368
+ "eval_loss": NaN,
369
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
370
+ "eval_mother_tongue_education_f1": 0.837037037037037,
371
+ "eval_overall_accuracy": 0.8390163480974149,
372
+ "eval_overall_f1": 0.8135900008491362,
373
+ "eval_religion_ethnic_identity_accuracy": 0.8355899419729207,
374
+ "eval_religion_ethnic_identity_f1": 0.8374170616290167,
375
+ "eval_runtime": 5.6175,
376
+ "eval_samples_per_second": 712.064,
377
+ "eval_steps_per_second": 44.504,
378
+ "step": 4000
379
+ },
380
+ {
381
+ "epoch": 4.1,
382
+ "grad_norm": 0.10239959508180618,
383
+ "learning_rate": 9.940569616534386e-06,
384
+ "loss": 0.0,
385
+ "step": 4100
386
+ },
387
+ {
388
+ "epoch": 4.2,
389
+ "grad_norm": 1.4125248193740845,
390
+ "learning_rate": 8.837287750104376e-06,
391
+ "loss": 0.0,
392
+ "step": 4200
393
+ },
394
+ {
395
+ "epoch": 4.3,
396
+ "grad_norm": 0.03549821674823761,
397
+ "learning_rate": 7.734005883674367e-06,
398
+ "loss": 0.0001,
399
+ "step": 4300
400
+ },
401
+ {
402
+ "epoch": 4.4,
403
+ "grad_norm": 0.016542043536901474,
404
+ "learning_rate": 6.630724017244356e-06,
405
+ "loss": 0.0001,
406
+ "step": 4400
407
+ },
408
+ {
409
+ "epoch": 4.5,
410
+ "grad_norm": 0.016069957986474037,
411
+ "learning_rate": 5.527442150814347e-06,
412
+ "loss": 0.0029,
413
+ "step": 4500
414
+ },
415
+ {
416
+ "epoch": 4.6,
417
+ "grad_norm": 1.0463519096374512,
418
+ "learning_rate": 4.424160284384338e-06,
419
+ "loss": 0.0,
420
+ "step": 4600
421
+ },
422
+ {
423
+ "epoch": 4.7,
424
+ "grad_norm": 0.226824089884758,
425
+ "learning_rate": 3.3208784179543285e-06,
426
+ "loss": 0.3585,
427
+ "step": 4700
428
+ },
429
+ {
430
+ "epoch": 4.8,
431
+ "grad_norm": 0.06588530540466309,
432
+ "learning_rate": 2.217596551524319e-06,
433
+ "loss": 0.0178,
434
+ "step": 4800
435
+ },
436
+ {
437
+ "epoch": 4.9,
438
+ "grad_norm": 0.018818065524101257,
439
+ "learning_rate": 1.1143146850943096e-06,
440
+ "loss": 0.0,
441
+ "step": 4900
442
+ },
443
+ {
444
+ "epoch": 5.0,
445
+ "grad_norm": 0.020508505403995514,
446
+ "learning_rate": 1.1032818664300094e-08,
447
+ "loss": 0.0087,
448
+ "step": 5000
449
+ },
450
+ {
451
+ "epoch": 5.0,
452
+ "eval_economic_inequality_accuracy": 0.8047858942065491,
453
+ "eval_economic_inequality_f1": 0.8017329203783851,
454
+ "eval_economic_policy_benefits_accuracy": 0.835016835016835,
455
+ "eval_economic_policy_benefits_f1": 0.8163396640810648,
456
+ "eval_ethnic_boundaries_accuracy": 0.9362745098039216,
457
+ "eval_ethnic_boundaries_f1": 0.9334180476356357,
458
+ "eval_language_policy_accuracy": 0.7352941176470589,
459
+ "eval_language_policy_f1": 0.6910099889012209,
460
+ "eval_loss": NaN,
461
+ "eval_mother_tongue_education_accuracy": 0.8888888888888888,
462
+ "eval_mother_tongue_education_f1": 0.837037037037037,
463
+ "eval_overall_accuracy": 0.8428544638801426,
464
+ "eval_overall_f1": 0.8217308919703874,
465
+ "eval_religion_ethnic_identity_accuracy": 0.8568665377176016,
466
+ "eval_religion_ethnic_identity_f1": 0.8508476937889805,
467
+ "eval_runtime": 4.4951,
468
+ "eval_samples_per_second": 889.857,
469
+ "eval_steps_per_second": 55.616,
470
+ "step": 5000
471
+ }
472
+ ],
473
+ "logging_steps": 100,
474
+ "max_steps": 5000,
475
+ "num_input_tokens_seen": 0,
476
+ "num_train_epochs": 5,
477
+ "save_steps": 500,
478
+ "stateful_callbacks": {
479
+ "EarlyStoppingCallback": {
480
+ "args": {
481
+ "early_stopping_patience": 2,
482
+ "early_stopping_threshold": 0.0
483
+ },
484
+ "attributes": {
485
+ "early_stopping_patience_counter": 0
486
+ }
487
+ },
488
+ "TrainerControl": {
489
+ "args": {
490
+ "should_epoch_stop": false,
491
+ "should_evaluate": false,
492
+ "should_log": false,
493
+ "should_save": true,
494
+ "should_training_stop": true
495
+ },
496
+ "attributes": {}
497
+ }
498
+ },
499
+ "total_flos": 2.105190825984e+16,
500
+ "train_batch_size": 16,
501
+ "trial_name": null,
502
+ "trial_params": {
503
+ "gradient_accumulation_steps": 1,
504
+ "learning_rate": 4.452845612911518e-05,
505
+ "num_train_epochs": 5,
506
+ "per_device_train_batch_size": 16,
507
+ "warmup_steps": 964,
508
+ "weight_decay": 0.08427803271887896
509
+ }
510
+ }
run-4/checkpoint-5000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4ad559f091ae7757869f5224f652bb49e6e3989ffd03fbb0eefdb9252513c40
3
+ size 5777