Hugging Face's logo

YagiASAFAS
/
PoliBERT-MY

Safetensors
bert
politics
Model card Files
xet
Community
YagiASAFAS commited on
Commit
6b0894a
·
verified ·
1 Parent(s): c9cd63d

Training in progress, epoch 1

Browse files
Files changed (17) hide show
  1. model.safetensors +1 -1
  2. run-4/checkpoint-2400/config.json +125 -0
  3. run-4/checkpoint-2400/model.safetensors +3 -0
  4. run-4/checkpoint-2400/optimizer.pt +3 -0
  5. run-4/checkpoint-2400/rng_state.pth +3 -0
  6. run-4/checkpoint-2400/scaler.pt +3 -0
  7. run-4/checkpoint-2400/scheduler.pt +3 -0
  8. run-4/checkpoint-2400/trainer_state.json +351 -0
  9. run-4/checkpoint-2400/training_args.bin +3 -0
  10. run-4/checkpoint-2995/config.json +125 -0
  11. run-4/checkpoint-2995/model.safetensors +3 -0
  12. run-4/checkpoint-2995/optimizer.pt +3 -0
  13. run-4/checkpoint-2995/rng_state.pth +3 -0
  14. run-4/checkpoint-2995/scaler.pt +3 -0
  15. run-4/checkpoint-2995/scheduler.pt +3 -0
  16. run-4/checkpoint-2995/trainer_state.json +420 -0
  17. run-4/checkpoint-2995/training_args.bin +3 -0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:978b28c3324ee9817d6f8d6fb6dbab5ad38e5350ed244b9ec8d04b0920357224
3
  size 438100144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89c05c9e831c4f2efb61107f2e0d6a7ddd2fe127c70f43a826f0ce72e6aec65e
3
  size 438100144
run-4/checkpoint-2400/config.json ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "18": "LABEL_18",
31
+ "19": "LABEL_19",
32
+ "20": "LABEL_20",
33
+ "21": "LABEL_21",
34
+ "22": "LABEL_22",
35
+ "23": "LABEL_23",
36
+ "24": "LABEL_24",
37
+ "25": "LABEL_25",
38
+ "26": "LABEL_26",
39
+ "27": "LABEL_27",
40
+ "28": "LABEL_28",
41
+ "29": "LABEL_29",
42
+ "30": "LABEL_30",
43
+ "31": "LABEL_31",
44
+ "32": "LABEL_32",
45
+ "33": "LABEL_33",
46
+ "34": "LABEL_34",
47
+ "35": "LABEL_35",
48
+ "36": "LABEL_36",
49
+ "37": "LABEL_37",
50
+ "38": "LABEL_38",
51
+ "39": "LABEL_39",
52
+ "40": "LABEL_40",
53
+ "41": "LABEL_41",
54
+ "42": "LABEL_42",
55
+ "43": "LABEL_43",
56
+ "44": "LABEL_44",
57
+ "45": "LABEL_45",
58
+ "46": "LABEL_46",
59
+ "47": "LABEL_47"
60
+ },
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 3072,
63
+ "label2id": {
64
+ "LABEL_0": 0,
65
+ "LABEL_1": 1,
66
+ "LABEL_10": 10,
67
+ "LABEL_11": 11,
68
+ "LABEL_12": 12,
69
+ "LABEL_13": 13,
70
+ "LABEL_14": 14,
71
+ "LABEL_15": 15,
72
+ "LABEL_16": 16,
73
+ "LABEL_17": 17,
74
+ "LABEL_18": 18,
75
+ "LABEL_19": 19,
76
+ "LABEL_2": 2,
77
+ "LABEL_20": 20,
78
+ "LABEL_21": 21,
79
+ "LABEL_22": 22,
80
+ "LABEL_23": 23,
81
+ "LABEL_24": 24,
82
+ "LABEL_25": 25,
83
+ "LABEL_26": 26,
84
+ "LABEL_27": 27,
85
+ "LABEL_28": 28,
86
+ "LABEL_29": 29,
87
+ "LABEL_3": 3,
88
+ "LABEL_30": 30,
89
+ "LABEL_31": 31,
90
+ "LABEL_32": 32,
91
+ "LABEL_33": 33,
92
+ "LABEL_34": 34,
93
+ "LABEL_35": 35,
94
+ "LABEL_36": 36,
95
+ "LABEL_37": 37,
96
+ "LABEL_38": 38,
97
+ "LABEL_39": 39,
98
+ "LABEL_4": 4,
99
+ "LABEL_40": 40,
100
+ "LABEL_41": 41,
101
+ "LABEL_42": 42,
102
+ "LABEL_43": 43,
103
+ "LABEL_44": 44,
104
+ "LABEL_45": 45,
105
+ "LABEL_46": 46,
106
+ "LABEL_47": 47,
107
+ "LABEL_5": 5,
108
+ "LABEL_6": 6,
109
+ "LABEL_7": 7,
110
+ "LABEL_8": 8,
111
+ "LABEL_9": 9
112
+ },
113
+ "layer_norm_eps": 1e-12,
114
+ "max_position_embeddings": 512,
115
+ "model_type": "bert",
116
+ "num_attention_heads": 12,
117
+ "num_hidden_layers": 12,
118
+ "pad_token_id": 0,
119
+ "position_embedding_type": "absolute",
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.50.3",
122
+ "type_vocab_size": 2,
123
+ "use_cache": true,
124
+ "vocab_size": 30522
125
+ }
run-4/checkpoint-2400/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43f0980f28535ee7d74d6b164742a5a055c40378bc2807048e67b35056095853
3
+ size 438100144
run-4/checkpoint-2400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48c2e9279cdea73cd8ab6e54aabc622fc8493996e9d3fa5e7b62eb0993cbeaab
3
+ size 876321402
run-4/checkpoint-2400/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ee8bdec2336829d8b665872a25fb80e042d5fa43fe24ad4b60b39cb823c692f
3
+ size 14244
run-4/checkpoint-2400/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:900f26e6b66c9eae5e85a376345ba03c270e502825dc9e8a4ca0b1cd6fc50bd2
3
+ size 988
run-4/checkpoint-2400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a3068bb9a2cdf6d4c873156e5b80042e43fdbee4d14e76328f40b2969217f8c
3
+ size 1064
run-4/checkpoint-2400/trainer_state.json ADDED
@@ -0,0 +1,351 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2400,
3
+ "best_metric": 0.9399009152601657,
4
+ "best_model_checkpoint": "./results/run-4/checkpoint-2400",
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2400,
8
+ "is_hyper_param_search": true,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.16673614005835766,
14
+ "grad_norm": 1.7758623361587524,
15
+ "learning_rate": 1e-05,
16
+ "loss": 1.3361,
17
+ "step": 100
18
+ },
19
+ {
20
+ "epoch": 0.3334722801167153,
21
+ "grad_norm": 0.6607375144958496,
22
+ "learning_rate": 2e-05,
23
+ "loss": 0.5099,
24
+ "step": 200
25
+ },
26
+ {
27
+ "epoch": 0.5002084201750729,
28
+ "grad_norm": 0.5690075159072876,
29
+ "learning_rate": 3e-05,
30
+ "loss": 0.4223,
31
+ "step": 300
32
+ },
33
+ {
34
+ "epoch": 0.6669445602334306,
35
+ "grad_norm": 0.848582923412323,
36
+ "learning_rate": 4e-05,
37
+ "loss": 0.3466,
38
+ "step": 400
39
+ },
40
+ {
41
+ "epoch": 0.8336807002917882,
42
+ "grad_norm": 0.7310414910316467,
43
+ "learning_rate": 5e-05,
44
+ "loss": 0.3032,
45
+ "step": 500
46
+ },
47
+ {
48
+ "epoch": 1.0,
49
+ "grad_norm": 0.5762994289398193,
50
+ "learning_rate": 4.7995991983967934e-05,
51
+ "loss": 0.2757,
52
+ "step": 600
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "eval_administration_accuracy": 0.9163018553262455,
57
+ "eval_administration_f1": 0.8799341161157713,
58
+ "eval_corruption_accuracy": 0.9427767354596623,
59
+ "eval_corruption_f1": 0.9297320702956003,
60
+ "eval_democracy_accuracy": 0.9418386491557224,
61
+ "eval_democracy_f1": 0.9228010908644549,
62
+ "eval_development_accuracy": 0.8867000208463623,
63
+ "eval_development_f1": 0.8490864044420574,
64
+ "eval_economy_accuracy": 0.9122368146758391,
65
+ "eval_economy_f1": 0.8967581427500766,
66
+ "eval_education_accuracy": 0.9583072753804461,
67
+ "eval_education_f1": 0.9515050567699045,
68
+ "eval_environment_accuracy": 0.9754012924744632,
69
+ "eval_environment_f1": 0.971384333060466,
70
+ "eval_instability_accuracy": 0.924119241192412,
71
+ "eval_instability_f1": 0.8935408759758452,
72
+ "eval_leadership_accuracy": 0.7805920366895976,
73
+ "eval_leadership_f1": 0.7320671383079426,
74
+ "eval_loss": 0.26127132773399353,
75
+ "eval_overall_accuracy": 0.9182127718713086,
76
+ "eval_overall_f1": 0.8968188861118868,
77
+ "eval_race_accuracy": 0.9360016677089847,
78
+ "eval_race_f1": 0.9209694401775296,
79
+ "eval_religion_accuracy": 0.9441317490097978,
80
+ "eval_religion_f1": 0.9328887566718251,
81
+ "eval_runtime": 10.8983,
82
+ "eval_safety_accuracy": 0.9001459245361685,
83
+ "eval_safety_f1": 0.8811592079111662,
84
+ "eval_samples_per_second": 880.318,
85
+ "eval_steps_per_second": 55.054,
86
+ "step": 600
87
+ },
88
+ {
89
+ "epoch": 1.1667361400583576,
90
+ "grad_norm": 0.6828027367591858,
91
+ "learning_rate": 4.599198396793588e-05,
92
+ "loss": 0.2396,
93
+ "step": 700
94
+ },
95
+ {
96
+ "epoch": 1.3334722801167154,
97
+ "grad_norm": 0.67964106798172,
98
+ "learning_rate": 4.398797595190381e-05,
99
+ "loss": 0.2312,
100
+ "step": 800
101
+ },
102
+ {
103
+ "epoch": 1.500208420175073,
104
+ "grad_norm": 0.7818251252174377,
105
+ "learning_rate": 4.198396793587174e-05,
106
+ "loss": 0.2237,
107
+ "step": 900
108
+ },
109
+ {
110
+ "epoch": 1.6669445602334307,
111
+ "grad_norm": 0.8807923197746277,
112
+ "learning_rate": 3.997995991983968e-05,
113
+ "loss": 0.2085,
114
+ "step": 1000
115
+ },
116
+ {
117
+ "epoch": 1.8336807002917883,
118
+ "grad_norm": 0.8586667776107788,
119
+ "learning_rate": 3.797595190380762e-05,
120
+ "loss": 0.2007,
121
+ "step": 1100
122
+ },
123
+ {
124
+ "epoch": 2.0,
125
+ "grad_norm": 0.9014652371406555,
126
+ "learning_rate": 3.5971943887775555e-05,
127
+ "loss": 0.2,
128
+ "step": 1200
129
+ },
130
+ {
131
+ "epoch": 2.0,
132
+ "eval_administration_accuracy": 0.9375651448822181,
133
+ "eval_administration_f1": 0.9290316070134363,
134
+ "eval_corruption_accuracy": 0.9589326662497394,
135
+ "eval_corruption_f1": 0.9550373529738804,
136
+ "eval_democracy_accuracy": 0.9521575984990619,
137
+ "eval_democracy_f1": 0.9433460958983528,
138
+ "eval_development_accuracy": 0.9138002918490723,
139
+ "eval_development_f1": 0.8974478772752896,
140
+ "eval_economy_accuracy": 0.9307900771315405,
141
+ "eval_economy_f1": 0.9241148503263762,
142
+ "eval_education_accuracy": 0.9664373566812591,
143
+ "eval_education_f1": 0.9649731388208568,
144
+ "eval_environment_accuracy": 0.9822805920366896,
145
+ "eval_environment_f1": 0.9806789699363689,
146
+ "eval_instability_accuracy": 0.9471544715447154,
147
+ "eval_instability_f1": 0.9400721946847931,
148
+ "eval_leadership_accuracy": 0.8124869710235564,
149
+ "eval_leadership_f1": 0.7871605097392734,
150
+ "eval_loss": 0.20422659814357758,
151
+ "eval_overall_accuracy": 0.9359582377875061,
152
+ "eval_overall_f1": 0.9285199435647921,
153
+ "eval_race_accuracy": 0.9536168438607463,
154
+ "eval_race_f1": 0.950384319474048,
155
+ "eval_religion_accuracy": 0.9576818845111528,
156
+ "eval_religion_f1": 0.956169902275164,
157
+ "eval_runtime": 11.0011,
158
+ "eval_safety_accuracy": 0.9185949551803211,
159
+ "eval_safety_f1": 0.913822504359663,
160
+ "eval_samples_per_second": 872.094,
161
+ "eval_steps_per_second": 54.54,
162
+ "step": 1200
163
+ },
164
+ {
165
+ "epoch": 2.1667361400583576,
166
+ "grad_norm": 0.7211724519729614,
167
+ "learning_rate": 3.3967935871743486e-05,
168
+ "loss": 0.1565,
169
+ "step": 1300
170
+ },
171
+ {
172
+ "epoch": 2.333472280116715,
173
+ "grad_norm": 0.8795008063316345,
174
+ "learning_rate": 3.1963927855711424e-05,
175
+ "loss": 0.1491,
176
+ "step": 1400
177
+ },
178
+ {
179
+ "epoch": 2.500208420175073,
180
+ "grad_norm": 0.9686936140060425,
181
+ "learning_rate": 2.9959919839679363e-05,
182
+ "loss": 0.1514,
183
+ "step": 1500
184
+ },
185
+ {
186
+ "epoch": 2.6669445602334307,
187
+ "grad_norm": 0.8608136773109436,
188
+ "learning_rate": 2.7955911823647297e-05,
189
+ "loss": 0.1495,
190
+ "step": 1600
191
+ },
192
+ {
193
+ "epoch": 2.8336807002917883,
194
+ "grad_norm": 0.9045342803001404,
195
+ "learning_rate": 2.595190380761523e-05,
196
+ "loss": 0.1433,
197
+ "step": 1700
198
+ },
199
+ {
200
+ "epoch": 3.0,
201
+ "grad_norm": 0.7755804061889648,
202
+ "learning_rate": 2.3947895791583166e-05,
203
+ "loss": 0.1425,
204
+ "step": 1800
205
+ },
206
+ {
207
+ "epoch": 3.0,
208
+ "eval_administration_accuracy": 0.9407963310402335,
209
+ "eval_administration_f1": 0.9350680574385688,
210
+ "eval_corruption_accuracy": 0.9607046070460704,
211
+ "eval_corruption_f1": 0.9599629345245179,
212
+ "eval_democracy_accuracy": 0.9569522618303106,
213
+ "eval_democracy_f1": 0.9539836532509403,
214
+ "eval_development_accuracy": 0.9252657911194496,
215
+ "eval_development_f1": 0.918888388565598,
216
+ "eval_economy_accuracy": 0.9362101313320825,
217
+ "eval_economy_f1": 0.9331727962045289,
218
+ "eval_education_accuracy": 0.9673754429851991,
219
+ "eval_education_f1": 0.9653887107414594,
220
+ "eval_environment_accuracy": 0.9832186783406296,
221
+ "eval_environment_f1": 0.9819944959159485,
222
+ "eval_instability_accuracy": 0.9492391077756931,
223
+ "eval_instability_f1": 0.9467721148205744,
224
+ "eval_leadership_accuracy": 0.8369814467375443,
225
+ "eval_leadership_f1": 0.8295655888282574,
226
+ "eval_loss": 0.189470112323761,
227
+ "eval_overall_accuracy": 0.94157806962685,
228
+ "eval_overall_f1": 0.9384264348969675,
229
+ "eval_race_accuracy": 0.957473420888055,
230
+ "eval_race_f1": 0.9558244178036731,
231
+ "eval_religion_accuracy": 0.960287679799875,
232
+ "eval_religion_f1": 0.9595273358731586,
233
+ "eval_runtime": 11.03,
234
+ "eval_safety_accuracy": 0.9244319366270586,
235
+ "eval_safety_f1": 0.9209687247963856,
236
+ "eval_samples_per_second": 869.81,
237
+ "eval_steps_per_second": 54.397,
238
+ "step": 1800
239
+ },
240
+ {
241
+ "epoch": 3.1667361400583576,
242
+ "grad_norm": 0.737460196018219,
243
+ "learning_rate": 2.1943887775551105e-05,
244
+ "loss": 0.1122,
245
+ "step": 1900
246
+ },
247
+ {
248
+ "epoch": 3.333472280116715,
249
+ "grad_norm": 0.7762249708175659,
250
+ "learning_rate": 1.993987975951904e-05,
251
+ "loss": 0.1122,
252
+ "step": 2000
253
+ },
254
+ {
255
+ "epoch": 3.500208420175073,
256
+ "grad_norm": 1.0788854360580444,
257
+ "learning_rate": 1.7935871743486977e-05,
258
+ "loss": 0.1083,
259
+ "step": 2100
260
+ },
261
+ {
262
+ "epoch": 3.6669445602334307,
263
+ "grad_norm": 0.7059413194656372,
264
+ "learning_rate": 1.593186372745491e-05,
265
+ "loss": 0.1081,
266
+ "step": 2200
267
+ },
268
+ {
269
+ "epoch": 3.8336807002917883,
270
+ "grad_norm": 0.9918854236602783,
271
+ "learning_rate": 1.3927855711422847e-05,
272
+ "loss": 0.1086,
273
+ "step": 2300
274
+ },
275
+ {
276
+ "epoch": 4.0,
277
+ "grad_norm": 0.9893817901611328,
278
+ "learning_rate": 1.1923847695390781e-05,
279
+ "loss": 0.1024,
280
+ "step": 2400
281
+ },
282
+ {
283
+ "epoch": 4.0,
284
+ "eval_administration_accuracy": 0.9428809672712112,
285
+ "eval_administration_f1": 0.938180470700318,
286
+ "eval_corruption_accuracy": 0.9623723160308526,
287
+ "eval_corruption_f1": 0.9612272557957268,
288
+ "eval_democracy_accuracy": 0.9570564936418595,
289
+ "eval_democracy_f1": 0.9540522922098595,
290
+ "eval_development_accuracy": 0.9252657911194496,
291
+ "eval_development_f1": 0.9205294368258055,
292
+ "eval_economy_accuracy": 0.9373566812591203,
293
+ "eval_economy_f1": 0.9360975905522984,
294
+ "eval_education_accuracy": 0.9690431519699813,
295
+ "eval_education_f1": 0.9675757412855236,
296
+ "eval_environment_accuracy": 0.9822805920366896,
297
+ "eval_environment_f1": 0.9813159598680739,
298
+ "eval_instability_accuracy": 0.9490306441525954,
299
+ "eval_instability_f1": 0.9457765995039474,
300
+ "eval_leadership_accuracy": 0.8419845736918907,
301
+ "eval_leadership_f1": 0.8366095676531852,
302
+ "eval_loss": 0.1909668743610382,
303
+ "eval_overall_accuracy": 0.9425769578208602,
304
+ "eval_overall_f1": 0.9399009152601657,
305
+ "eval_race_accuracy": 0.958411507191995,
306
+ "eval_race_f1": 0.9572260766271818,
307
+ "eval_religion_accuracy": 0.9600792161767772,
308
+ "eval_religion_f1": 0.9597895636198337,
309
+ "eval_runtime": 11.044,
310
+ "eval_safety_accuracy": 0.9251615593079008,
311
+ "eval_safety_f1": 0.9204304284802344,
312
+ "eval_samples_per_second": 868.704,
313
+ "eval_steps_per_second": 54.328,
314
+ "step": 2400
315
+ }
316
+ ],
317
+ "logging_steps": 100,
318
+ "max_steps": 2995,
319
+ "num_input_tokens_seen": 0,
320
+ "num_train_epochs": 5,
321
+ "save_steps": 500,
322
+ "stateful_callbacks": {
323
+ "EarlyStoppingCallback": {
324
+ "args": {
325
+ "early_stopping_patience": 2,
326
+ "early_stopping_threshold": 0.0
327
+ },
328
+ "attributes": {
329
+ "early_stopping_patience_counter": 0
330
+ }
331
+ },
332
+ "TrainerControl": {
333
+ "args": {
334
+ "should_epoch_stop": false,
335
+ "should_evaluate": false,
336
+ "should_log": false,
337
+ "should_save": true,
338
+ "should_training_stop": false
339
+ },
340
+ "attributes": {}
341
+ }
342
+ },
343
+ "total_flos": 4.040106904151654e+16,
344
+ "train_batch_size": 16,
345
+ "trial_name": null,
346
+ "trial_params": {
347
+ "gradient_accumulation_steps": 4,
348
+ "learning_rate": 5e-05,
349
+ "num_train_epochs": 5
350
+ }
351
+ }
run-4/checkpoint-2400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d23c4b505069b659bf843cf83d2a0a253f99487d90fb9c81ceef3ac1152d69eb
3
+ size 5368
run-4/checkpoint-2995/config.json ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "18": "LABEL_18",
31
+ "19": "LABEL_19",
32
+ "20": "LABEL_20",
33
+ "21": "LABEL_21",
34
+ "22": "LABEL_22",
35
+ "23": "LABEL_23",
36
+ "24": "LABEL_24",
37
+ "25": "LABEL_25",
38
+ "26": "LABEL_26",
39
+ "27": "LABEL_27",
40
+ "28": "LABEL_28",
41
+ "29": "LABEL_29",
42
+ "30": "LABEL_30",
43
+ "31": "LABEL_31",
44
+ "32": "LABEL_32",
45
+ "33": "LABEL_33",
46
+ "34": "LABEL_34",
47
+ "35": "LABEL_35",
48
+ "36": "LABEL_36",
49
+ "37": "LABEL_37",
50
+ "38": "LABEL_38",
51
+ "39": "LABEL_39",
52
+ "40": "LABEL_40",
53
+ "41": "LABEL_41",
54
+ "42": "LABEL_42",
55
+ "43": "LABEL_43",
56
+ "44": "LABEL_44",
57
+ "45": "LABEL_45",
58
+ "46": "LABEL_46",
59
+ "47": "LABEL_47"
60
+ },
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 3072,
63
+ "label2id": {
64
+ "LABEL_0": 0,
65
+ "LABEL_1": 1,
66
+ "LABEL_10": 10,
67
+ "LABEL_11": 11,
68
+ "LABEL_12": 12,
69
+ "LABEL_13": 13,
70
+ "LABEL_14": 14,
71
+ "LABEL_15": 15,
72
+ "LABEL_16": 16,
73
+ "LABEL_17": 17,
74
+ "LABEL_18": 18,
75
+ "LABEL_19": 19,
76
+ "LABEL_2": 2,
77
+ "LABEL_20": 20,
78
+ "LABEL_21": 21,
79
+ "LABEL_22": 22,
80
+ "LABEL_23": 23,
81
+ "LABEL_24": 24,
82
+ "LABEL_25": 25,
83
+ "LABEL_26": 26,
84
+ "LABEL_27": 27,
85
+ "LABEL_28": 28,
86
+ "LABEL_29": 29,
87
+ "LABEL_3": 3,
88
+ "LABEL_30": 30,
89
+ "LABEL_31": 31,
90
+ "LABEL_32": 32,
91
+ "LABEL_33": 33,
92
+ "LABEL_34": 34,
93
+ "LABEL_35": 35,
94
+ "LABEL_36": 36,
95
+ "LABEL_37": 37,
96
+ "LABEL_38": 38,
97
+ "LABEL_39": 39,
98
+ "LABEL_4": 4,
99
+ "LABEL_40": 40,
100
+ "LABEL_41": 41,
101
+ "LABEL_42": 42,
102
+ "LABEL_43": 43,
103
+ "LABEL_44": 44,
104
+ "LABEL_45": 45,
105
+ "LABEL_46": 46,
106
+ "LABEL_47": 47,
107
+ "LABEL_5": 5,
108
+ "LABEL_6": 6,
109
+ "LABEL_7": 7,
110
+ "LABEL_8": 8,
111
+ "LABEL_9": 9
112
+ },
113
+ "layer_norm_eps": 1e-12,
114
+ "max_position_embeddings": 512,
115
+ "model_type": "bert",
116
+ "num_attention_heads": 12,
117
+ "num_hidden_layers": 12,
118
+ "pad_token_id": 0,
119
+ "position_embedding_type": "absolute",
120
+ "torch_dtype": "float32",
121
+ "transformers_version": "4.50.3",
122
+ "type_vocab_size": 2,
123
+ "use_cache": true,
124
+ "vocab_size": 30522
125
+ }
run-4/checkpoint-2995/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6aac157c69c5a342355cd3b65b06e09ca017e043bae6a4955140b66139c3d53d
3
+ size 438100144
run-4/checkpoint-2995/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:615282c0c67b2bc4887e68cafaa732f0dbe33240d16aceaf32fccd27f3eff433
3
+ size 876321402
run-4/checkpoint-2995/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab97b90684981c7dc315117d733b9e57684d6b194320677f6c90432a975bebd3
3
+ size 14244
run-4/checkpoint-2995/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7aad8d28ae06cf3603ef16a53be9b74295b0750a2bc8bb9a319dd832fcad75f6
3
+ size 988
run-4/checkpoint-2995/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a32f59012b14dddd7c98640dfb1b798800a2b23b7b3cd7400a943d07283b331a
3
+ size 1064
run-4/checkpoint-2995/trainer_state.json ADDED
@@ -0,0 +1,420 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2995,
3
+ "best_metric": 0.9403059159534481,
4
+ "best_model_checkpoint": "./results/run-4/checkpoint-2995",
5
+ "epoch": 4.992080033347228,
6
+ "eval_steps": 500,
7
+ "global_step": 2995,
8
+ "is_hyper_param_search": true,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.16673614005835766,
14
+ "grad_norm": 1.7758623361587524,
15
+ "learning_rate": 1e-05,
16
+ "loss": 1.3361,
17
+ "step": 100
18
+ },
19
+ {
20
+ "epoch": 0.3334722801167153,
21
+ "grad_norm": 0.6607375144958496,
22
+ "learning_rate": 2e-05,
23
+ "loss": 0.5099,
24
+ "step": 200
25
+ },
26
+ {
27
+ "epoch": 0.5002084201750729,
28
+ "grad_norm": 0.5690075159072876,
29
+ "learning_rate": 3e-05,
30
+ "loss": 0.4223,
31
+ "step": 300
32
+ },
33
+ {
34
+ "epoch": 0.6669445602334306,
35
+ "grad_norm": 0.848582923412323,
36
+ "learning_rate": 4e-05,
37
+ "loss": 0.3466,
38
+ "step": 400
39
+ },
40
+ {
41
+ "epoch": 0.8336807002917882,
42
+ "grad_norm": 0.7310414910316467,
43
+ "learning_rate": 5e-05,
44
+ "loss": 0.3032,
45
+ "step": 500
46
+ },
47
+ {
48
+ "epoch": 1.0,
49
+ "grad_norm": 0.5762994289398193,
50
+ "learning_rate": 4.7995991983967934e-05,
51
+ "loss": 0.2757,
52
+ "step": 600
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "eval_administration_accuracy": 0.9163018553262455,
57
+ "eval_administration_f1": 0.8799341161157713,
58
+ "eval_corruption_accuracy": 0.9427767354596623,
59
+ "eval_corruption_f1": 0.9297320702956003,
60
+ "eval_democracy_accuracy": 0.9418386491557224,
61
+ "eval_democracy_f1": 0.9228010908644549,
62
+ "eval_development_accuracy": 0.8867000208463623,
63
+ "eval_development_f1": 0.8490864044420574,
64
+ "eval_economy_accuracy": 0.9122368146758391,
65
+ "eval_economy_f1": 0.8967581427500766,
66
+ "eval_education_accuracy": 0.9583072753804461,
67
+ "eval_education_f1": 0.9515050567699045,
68
+ "eval_environment_accuracy": 0.9754012924744632,
69
+ "eval_environment_f1": 0.971384333060466,
70
+ "eval_instability_accuracy": 0.924119241192412,
71
+ "eval_instability_f1": 0.8935408759758452,
72
+ "eval_leadership_accuracy": 0.7805920366895976,
73
+ "eval_leadership_f1": 0.7320671383079426,
74
+ "eval_loss": 0.26127132773399353,
75
+ "eval_overall_accuracy": 0.9182127718713086,
76
+ "eval_overall_f1": 0.8968188861118868,
77
+ "eval_race_accuracy": 0.9360016677089847,
78
+ "eval_race_f1": 0.9209694401775296,
79
+ "eval_religion_accuracy": 0.9441317490097978,
80
+ "eval_religion_f1": 0.9328887566718251,
81
+ "eval_runtime": 10.8983,
82
+ "eval_safety_accuracy": 0.9001459245361685,
83
+ "eval_safety_f1": 0.8811592079111662,
84
+ "eval_samples_per_second": 880.318,
85
+ "eval_steps_per_second": 55.054,
86
+ "step": 600
87
+ },
88
+ {
89
+ "epoch": 1.1667361400583576,
90
+ "grad_norm": 0.6828027367591858,
91
+ "learning_rate": 4.599198396793588e-05,
92
+ "loss": 0.2396,
93
+ "step": 700
94
+ },
95
+ {
96
+ "epoch": 1.3334722801167154,
97
+ "grad_norm": 0.67964106798172,
98
+ "learning_rate": 4.398797595190381e-05,
99
+ "loss": 0.2312,
100
+ "step": 800
101
+ },
102
+ {
103
+ "epoch": 1.500208420175073,
104
+ "grad_norm": 0.7818251252174377,
105
+ "learning_rate": 4.198396793587174e-05,
106
+ "loss": 0.2237,
107
+ "step": 900
108
+ },
109
+ {
110
+ "epoch": 1.6669445602334307,
111
+ "grad_norm": 0.8807923197746277,
112
+ "learning_rate": 3.997995991983968e-05,
113
+ "loss": 0.2085,
114
+ "step": 1000
115
+ },
116
+ {
117
+ "epoch": 1.8336807002917883,
118
+ "grad_norm": 0.8586667776107788,
119
+ "learning_rate": 3.797595190380762e-05,
120
+ "loss": 0.2007,
121
+ "step": 1100
122
+ },
123
+ {
124
+ "epoch": 2.0,
125
+ "grad_norm": 0.9014652371406555,
126
+ "learning_rate": 3.5971943887775555e-05,
127
+ "loss": 0.2,
128
+ "step": 1200
129
+ },
130
+ {
131
+ "epoch": 2.0,
132
+ "eval_administration_accuracy": 0.9375651448822181,
133
+ "eval_administration_f1": 0.9290316070134363,
134
+ "eval_corruption_accuracy": 0.9589326662497394,
135
+ "eval_corruption_f1": 0.9550373529738804,
136
+ "eval_democracy_accuracy": 0.9521575984990619,
137
+ "eval_democracy_f1": 0.9433460958983528,
138
+ "eval_development_accuracy": 0.9138002918490723,
139
+ "eval_development_f1": 0.8974478772752896,
140
+ "eval_economy_accuracy": 0.9307900771315405,
141
+ "eval_economy_f1": 0.9241148503263762,
142
+ "eval_education_accuracy": 0.9664373566812591,
143
+ "eval_education_f1": 0.9649731388208568,
144
+ "eval_environment_accuracy": 0.9822805920366896,
145
+ "eval_environment_f1": 0.9806789699363689,
146
+ "eval_instability_accuracy": 0.9471544715447154,
147
+ "eval_instability_f1": 0.9400721946847931,
148
+ "eval_leadership_accuracy": 0.8124869710235564,
149
+ "eval_leadership_f1": 0.7871605097392734,
150
+ "eval_loss": 0.20422659814357758,
151
+ "eval_overall_accuracy": 0.9359582377875061,
152
+ "eval_overall_f1": 0.9285199435647921,
153
+ "eval_race_accuracy": 0.9536168438607463,
154
+ "eval_race_f1": 0.950384319474048,
155
+ "eval_religion_accuracy": 0.9576818845111528,
156
+ "eval_religion_f1": 0.956169902275164,
157
+ "eval_runtime": 11.0011,
158
+ "eval_safety_accuracy": 0.9185949551803211,
159
+ "eval_safety_f1": 0.913822504359663,
160
+ "eval_samples_per_second": 872.094,
161
+ "eval_steps_per_second": 54.54,
162
+ "step": 1200
163
+ },
164
+ {
165
+ "epoch": 2.1667361400583576,
166
+ "grad_norm": 0.7211724519729614,
167
+ "learning_rate": 3.3967935871743486e-05,
168
+ "loss": 0.1565,
169
+ "step": 1300
170
+ },
171
+ {
172
+ "epoch": 2.333472280116715,
173
+ "grad_norm": 0.8795008063316345,
174
+ "learning_rate": 3.1963927855711424e-05,
175
+ "loss": 0.1491,
176
+ "step": 1400
177
+ },
178
+ {
179
+ "epoch": 2.500208420175073,
180
+ "grad_norm": 0.9686936140060425,
181
+ "learning_rate": 2.9959919839679363e-05,
182
+ "loss": 0.1514,
183
+ "step": 1500
184
+ },
185
+ {
186
+ "epoch": 2.6669445602334307,
187
+ "grad_norm": 0.8608136773109436,
188
+ "learning_rate": 2.7955911823647297e-05,
189
+ "loss": 0.1495,
190
+ "step": 1600
191
+ },
192
+ {
193
+ "epoch": 2.8336807002917883,
194
+ "grad_norm": 0.9045342803001404,
195
+ "learning_rate": 2.595190380761523e-05,
196
+ "loss": 0.1433,
197
+ "step": 1700
198
+ },
199
+ {
200
+ "epoch": 3.0,
201
+ "grad_norm": 0.7755804061889648,
202
+ "learning_rate": 2.3947895791583166e-05,
203
+ "loss": 0.1425,
204
+ "step": 1800
205
+ },
206
+ {
207
+ "epoch": 3.0,
208
+ "eval_administration_accuracy": 0.9407963310402335,
209
+ "eval_administration_f1": 0.9350680574385688,
210
+ "eval_corruption_accuracy": 0.9607046070460704,
211
+ "eval_corruption_f1": 0.9599629345245179,
212
+ "eval_democracy_accuracy": 0.9569522618303106,
213
+ "eval_democracy_f1": 0.9539836532509403,
214
+ "eval_development_accuracy": 0.9252657911194496,
215
+ "eval_development_f1": 0.918888388565598,
216
+ "eval_economy_accuracy": 0.9362101313320825,
217
+ "eval_economy_f1": 0.9331727962045289,
218
+ "eval_education_accuracy": 0.9673754429851991,
219
+ "eval_education_f1": 0.9653887107414594,
220
+ "eval_environment_accuracy": 0.9832186783406296,
221
+ "eval_environment_f1": 0.9819944959159485,
222
+ "eval_instability_accuracy": 0.9492391077756931,
223
+ "eval_instability_f1": 0.9467721148205744,
224
+ "eval_leadership_accuracy": 0.8369814467375443,
225
+ "eval_leadership_f1": 0.8295655888282574,
226
+ "eval_loss": 0.189470112323761,
227
+ "eval_overall_accuracy": 0.94157806962685,
228
+ "eval_overall_f1": 0.9384264348969675,
229
+ "eval_race_accuracy": 0.957473420888055,
230
+ "eval_race_f1": 0.9558244178036731,
231
+ "eval_religion_accuracy": 0.960287679799875,
232
+ "eval_religion_f1": 0.9595273358731586,
233
+ "eval_runtime": 11.03,
234
+ "eval_safety_accuracy": 0.9244319366270586,
235
+ "eval_safety_f1": 0.9209687247963856,
236
+ "eval_samples_per_second": 869.81,
237
+ "eval_steps_per_second": 54.397,
238
+ "step": 1800
239
+ },
240
+ {
241
+ "epoch": 3.1667361400583576,
242
+ "grad_norm": 0.737460196018219,
243
+ "learning_rate": 2.1943887775551105e-05,
244
+ "loss": 0.1122,
245
+ "step": 1900
246
+ },
247
+ {
248
+ "epoch": 3.333472280116715,
249
+ "grad_norm": 0.7762249708175659,
250
+ "learning_rate": 1.993987975951904e-05,
251
+ "loss": 0.1122,
252
+ "step": 2000
253
+ },
254
+ {
255
+ "epoch": 3.500208420175073,
256
+ "grad_norm": 1.0788854360580444,
257
+ "learning_rate": 1.7935871743486977e-05,
258
+ "loss": 0.1083,
259
+ "step": 2100
260
+ },
261
+ {
262
+ "epoch": 3.6669445602334307,
263
+ "grad_norm": 0.7059413194656372,
264
+ "learning_rate": 1.593186372745491e-05,
265
+ "loss": 0.1081,
266
+ "step": 2200
267
+ },
268
+ {
269
+ "epoch": 3.8336807002917883,
270
+ "grad_norm": 0.9918854236602783,
271
+ "learning_rate": 1.3927855711422847e-05,
272
+ "loss": 0.1086,
273
+ "step": 2300
274
+ },
275
+ {
276
+ "epoch": 4.0,
277
+ "grad_norm": 0.9893817901611328,
278
+ "learning_rate": 1.1923847695390781e-05,
279
+ "loss": 0.1024,
280
+ "step": 2400
281
+ },
282
+ {
283
+ "epoch": 4.0,
284
+ "eval_administration_accuracy": 0.9428809672712112,
285
+ "eval_administration_f1": 0.938180470700318,
286
+ "eval_corruption_accuracy": 0.9623723160308526,
287
+ "eval_corruption_f1": 0.9612272557957268,
288
+ "eval_democracy_accuracy": 0.9570564936418595,
289
+ "eval_democracy_f1": 0.9540522922098595,
290
+ "eval_development_accuracy": 0.9252657911194496,
291
+ "eval_development_f1": 0.9205294368258055,
292
+ "eval_economy_accuracy": 0.9373566812591203,
293
+ "eval_economy_f1": 0.9360975905522984,
294
+ "eval_education_accuracy": 0.9690431519699813,
295
+ "eval_education_f1": 0.9675757412855236,
296
+ "eval_environment_accuracy": 0.9822805920366896,
297
+ "eval_environment_f1": 0.9813159598680739,
298
+ "eval_instability_accuracy": 0.9490306441525954,
299
+ "eval_instability_f1": 0.9457765995039474,
300
+ "eval_leadership_accuracy": 0.8419845736918907,
301
+ "eval_leadership_f1": 0.8366095676531852,
302
+ "eval_loss": 0.1909668743610382,
303
+ "eval_overall_accuracy": 0.9425769578208602,
304
+ "eval_overall_f1": 0.9399009152601657,
305
+ "eval_race_accuracy": 0.958411507191995,
306
+ "eval_race_f1": 0.9572260766271818,
307
+ "eval_religion_accuracy": 0.9600792161767772,
308
+ "eval_religion_f1": 0.9597895636198337,
309
+ "eval_runtime": 11.044,
310
+ "eval_safety_accuracy": 0.9251615593079008,
311
+ "eval_safety_f1": 0.9204304284802344,
312
+ "eval_samples_per_second": 868.704,
313
+ "eval_steps_per_second": 54.328,
314
+ "step": 2400
315
+ },
316
+ {
317
+ "epoch": 4.166736140058358,
318
+ "grad_norm": 0.612309992313385,
319
+ "learning_rate": 9.919839679358718e-06,
320
+ "loss": 0.0861,
321
+ "step": 2500
322
+ },
323
+ {
324
+ "epoch": 4.333472280116715,
325
+ "grad_norm": 0.6645572185516357,
326
+ "learning_rate": 7.915831663326654e-06,
327
+ "loss": 0.0907,
328
+ "step": 2600
329
+ },
330
+ {
331
+ "epoch": 4.500208420175073,
332
+ "grad_norm": 0.4853862524032593,
333
+ "learning_rate": 5.911823647294589e-06,
334
+ "loss": 0.089,
335
+ "step": 2700
336
+ },
337
+ {
338
+ "epoch": 4.66694456023343,
339
+ "grad_norm": 0.9452826380729675,
340
+ "learning_rate": 3.907815631262525e-06,
341
+ "loss": 0.0854,
342
+ "step": 2800
343
+ },
344
+ {
345
+ "epoch": 4.833680700291788,
346
+ "grad_norm": 0.5481389760971069,
347
+ "learning_rate": 1.9038076152304612e-06,
348
+ "loss": 0.0831,
349
+ "step": 2900
350
+ },
351
+ {
352
+ "epoch": 4.992080033347228,
353
+ "eval_administration_accuracy": 0.9416301855326246,
354
+ "eval_administration_f1": 0.9373957175884656,
355
+ "eval_corruption_accuracy": 0.962997706900146,
356
+ "eval_corruption_f1": 0.9615877547001083,
357
+ "eval_democracy_accuracy": 0.958411507191995,
358
+ "eval_democracy_f1": 0.956671714795931,
359
+ "eval_development_accuracy": 0.9251615593079008,
360
+ "eval_development_f1": 0.9203586042336891,
361
+ "eval_economy_accuracy": 0.9391286220554513,
362
+ "eval_economy_f1": 0.9362693111073657,
363
+ "eval_education_accuracy": 0.9680008338544924,
364
+ "eval_education_f1": 0.9665263839816942,
365
+ "eval_environment_accuracy": 0.9825932874713362,
366
+ "eval_environment_f1": 0.9812821621070215,
367
+ "eval_instability_accuracy": 0.9483010214717532,
368
+ "eval_instability_f1": 0.9453908405296636,
369
+ "eval_leadership_accuracy": 0.8457369189076506,
370
+ "eval_leadership_f1": 0.8409596087566152,
371
+ "eval_loss": 0.19213667511940002,
372
+ "eval_overall_accuracy": 0.9429330831769857,
373
+ "eval_overall_f1": 0.9403059159534481,
374
+ "eval_race_accuracy": 0.9570564936418595,
375
+ "eval_race_f1": 0.9560341943790861,
376
+ "eval_religion_accuracy": 0.9609130706691682,
377
+ "eval_religion_f1": 0.9603128429572957,
378
+ "eval_runtime": 10.9978,
379
+ "eval_safety_accuracy": 0.9252657911194496,
380
+ "eval_safety_f1": 0.9208818563044412,
381
+ "eval_samples_per_second": 872.356,
382
+ "eval_steps_per_second": 54.556,
383
+ "step": 2995
384
+ }
385
+ ],
386
+ "logging_steps": 100,
387
+ "max_steps": 2995,
388
+ "num_input_tokens_seen": 0,
389
+ "num_train_epochs": 5,
390
+ "save_steps": 500,
391
+ "stateful_callbacks": {
392
+ "EarlyStoppingCallback": {
393
+ "args": {
394
+ "early_stopping_patience": 2,
395
+ "early_stopping_threshold": 0.0
396
+ },
397
+ "attributes": {
398
+ "early_stopping_patience_counter": 0
399
+ }
400
+ },
401
+ "TrainerControl": {
402
+ "args": {
403
+ "should_epoch_stop": false,
404
+ "should_evaluate": false,
405
+ "should_log": false,
406
+ "should_save": true,
407
+ "should_training_stop": true
408
+ },
409
+ "attributes": {}
410
+ }
411
+ },
412
+ "total_flos": 4.882410021873254e+16,
413
+ "train_batch_size": 16,
414
+ "trial_name": null,
415
+ "trial_params": {
416
+ "gradient_accumulation_steps": 4,
417
+ "learning_rate": 5e-05,
418
+ "num_train_epochs": 5
419
+ }
420
+ }
run-4/checkpoint-2995/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d23c4b505069b659bf843cf83d2a0a253f99487d90fb9c81ceef3ac1152d69eb
3
+ size 5368