jagoldz commited on
Commit
edc3791
1 Parent(s): c7f964f

Upload model files

Browse files
config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "deepset/gelectra-large",
3
+ "architectures": [
4
+ "ElectraForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "embedding_size": 1024,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 1024,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4096,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "electra",
17
+ "num_attention_heads": 16,
18
+ "num_hidden_layers": 24,
19
+ "pad_token_id": 0,
20
+ "position_embedding_type": "absolute",
21
+ "problem_type": "single_label_classification",
22
+ "summary_activation": "gelu",
23
+ "summary_last_dropout": 0.1,
24
+ "summary_type": "first",
25
+ "summary_use_proj": true,
26
+ "torch_dtype": "float32",
27
+ "transformers_version": "4.32.0",
28
+ "type_vocab_size": 2,
29
+ "use_cache": true,
30
+ "vocab_size": 31102
31
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d613e7b17cd14664663f88952a2b1641c0123b3a200c1755ed9bc7d07305cb01
3
+ size 2686231352
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b44e96a636082a086fb2929ef6c872199c14ece574bbbbc968dda4d81cc4840
3
+ size 1343087025
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9fb0acb3ffc9440a147beeb0054bd2dbd2e49b42e8a7e972376d9399f27f397
3
+ size 14575
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50bfc4db7dd4a88e8db8c4692609b79e38de4f41d75bcaa19ed03152bdd08664
3
+ size 627
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "cls_token": "[CLS]",
4
+ "do_basic_tokenize": true,
5
+ "do_lower_case": false,
6
+ "mask_token": "[MASK]",
7
+ "max_len": 512,
8
+ "model_max_length": 512,
9
+ "never_split": null,
10
+ "pad_token": "[PAD]",
11
+ "sep_token": "[SEP]",
12
+ "strip_accents": false,
13
+ "tokenize_chinese_chars": true,
14
+ "tokenizer_class": "ElectraTokenizer",
15
+ "unk_token": "[UNK]"
16
+ }
trainer_state.json ADDED
@@ -0,0 +1,1465 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "eval_steps": 500,
6
+ "global_step": 2010,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "learning_rate": 9.950248756218906e-06,
14
+ "loss": 0.6611,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.05,
19
+ "learning_rate": 9.900497512437812e-06,
20
+ "loss": 0.6622,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.07,
25
+ "learning_rate": 9.850746268656717e-06,
26
+ "loss": 0.6558,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.1,
31
+ "learning_rate": 9.800995024875622e-06,
32
+ "loss": 0.6535,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.12,
37
+ "learning_rate": 9.751243781094527e-06,
38
+ "loss": 0.6359,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.15,
43
+ "learning_rate": 9.701492537313434e-06,
44
+ "loss": 0.6274,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.17,
49
+ "learning_rate": 9.651741293532339e-06,
50
+ "loss": 0.6193,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.2,
55
+ "learning_rate": 9.601990049751244e-06,
56
+ "loss": 0.5715,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.22,
61
+ "learning_rate": 9.552238805970149e-06,
62
+ "loss": 0.5749,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.25,
67
+ "learning_rate": 9.502487562189056e-06,
68
+ "loss": 0.536,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.27,
73
+ "learning_rate": 9.45273631840796e-06,
74
+ "loss": 0.5248,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.3,
79
+ "learning_rate": 9.402985074626867e-06,
80
+ "loss": 0.525,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.32,
85
+ "learning_rate": 9.353233830845772e-06,
86
+ "loss": 0.5185,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.35,
91
+ "learning_rate": 9.303482587064677e-06,
92
+ "loss": 0.4787,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.37,
97
+ "learning_rate": 9.253731343283582e-06,
98
+ "loss": 0.5257,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.4,
103
+ "learning_rate": 9.203980099502489e-06,
104
+ "loss": 0.5418,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.42,
109
+ "learning_rate": 9.154228855721394e-06,
110
+ "loss": 0.4907,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.45,
115
+ "learning_rate": 9.104477611940299e-06,
116
+ "loss": 0.4937,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.47,
121
+ "learning_rate": 9.054726368159204e-06,
122
+ "loss": 0.5307,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.5,
127
+ "learning_rate": 9.00497512437811e-06,
128
+ "loss": 0.5081,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.52,
133
+ "learning_rate": 8.955223880597016e-06,
134
+ "loss": 0.4422,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.55,
139
+ "learning_rate": 8.905472636815922e-06,
140
+ "loss": 0.4795,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.57,
145
+ "learning_rate": 8.855721393034826e-06,
146
+ "loss": 0.5052,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.6,
151
+ "learning_rate": 8.805970149253732e-06,
152
+ "loss": 0.4952,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.62,
157
+ "learning_rate": 8.756218905472637e-06,
158
+ "loss": 0.4394,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.65,
163
+ "learning_rate": 8.706467661691544e-06,
164
+ "loss": 0.5088,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.67,
169
+ "learning_rate": 8.656716417910447e-06,
170
+ "loss": 0.4878,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.7,
175
+ "learning_rate": 8.606965174129354e-06,
176
+ "loss": 0.4851,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.72,
181
+ "learning_rate": 8.557213930348259e-06,
182
+ "loss": 0.4383,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.75,
187
+ "learning_rate": 8.507462686567165e-06,
188
+ "loss": 0.4377,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.77,
193
+ "learning_rate": 8.45771144278607e-06,
194
+ "loss": 0.4818,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.8,
199
+ "learning_rate": 8.407960199004975e-06,
200
+ "loss": 0.4399,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.82,
205
+ "learning_rate": 8.35820895522388e-06,
206
+ "loss": 0.5072,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.85,
211
+ "learning_rate": 8.308457711442787e-06,
212
+ "loss": 0.4585,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.87,
217
+ "learning_rate": 8.258706467661692e-06,
218
+ "loss": 0.4994,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.9,
223
+ "learning_rate": 8.208955223880599e-06,
224
+ "loss": 0.4836,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.92,
229
+ "learning_rate": 8.159203980099502e-06,
230
+ "loss": 0.4744,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.95,
235
+ "learning_rate": 8.109452736318409e-06,
236
+ "loss": 0.4244,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.97,
241
+ "learning_rate": 8.059701492537314e-06,
242
+ "loss": 0.4591,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 1.0,
247
+ "learning_rate": 8.00995024875622e-06,
248
+ "loss": 0.4019,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 1.0,
253
+ "eval_DeTox:accuracy": 0.7757009345794392,
254
+ "eval_DeTox:f1-macro": 0.7292154841128502,
255
+ "eval_DeTox:precision": 0.6744186046511628,
256
+ "eval_DeTox:recall": 0.5686274509803921,
257
+ "eval_HASOC2019:accuracy": 0.7575757575757576,
258
+ "eval_HASOC2019:f1-macro": 0.7380952380952381,
259
+ "eval_HASOC2019:precision": 0.6153846153846154,
260
+ "eval_HASOC2019:recall": 0.7272727272727273,
261
+ "eval_HASOC2020:accuracy": 0.6511627906976745,
262
+ "eval_HASOC2020:f1-macro": 0.5274725274725275,
263
+ "eval_HASOC2020:precision": 0.3333333333333333,
264
+ "eval_HASOC2020:recall": 0.25,
265
+ "eval_MHC:accuracy": NaN,
266
+ "eval_MHC:f1-macro": NaN,
267
+ "eval_MHC:precision": 0.0,
268
+ "eval_MHC:recall": 0.0,
269
+ "eval_RP:accuracy": 0.8848684210526315,
270
+ "eval_RP:f1-macro": 0.8664105364922722,
271
+ "eval_RP:precision": 0.9397590361445783,
272
+ "eval_RP:recall": 0.7222222222222222,
273
+ "eval_Round1:accuracy": NaN,
274
+ "eval_Round1:f1-macro": NaN,
275
+ "eval_Round1:precision": 0.0,
276
+ "eval_Round1:recall": 0.0,
277
+ "eval_Round2:accuracy": NaN,
278
+ "eval_Round2:f1-macro": NaN,
279
+ "eval_Round2:precision": 0.0,
280
+ "eval_Round2:recall": 0.0,
281
+ "eval_Round3:accuracy": NaN,
282
+ "eval_Round3:f1-macro": NaN,
283
+ "eval_Round3:precision": 0.0,
284
+ "eval_Round3:recall": 0.0,
285
+ "eval_Round4:accuracy": NaN,
286
+ "eval_Round4:f1-macro": NaN,
287
+ "eval_Round4:precision": 0.0,
288
+ "eval_Round4:recall": 0.0,
289
+ "eval_accuracy": 0.7868085106382978,
290
+ "eval_f1-macro": 0.7744064408987412,
291
+ "eval_loss": 0.46475154161453247,
292
+ "eval_precision": 0.7572928821470245,
293
+ "eval_recall": 0.6889596602972399,
294
+ "eval_runtime": 9.743,
295
+ "eval_samples_per_second": 241.2,
296
+ "eval_steps_per_second": 30.176,
297
+ "step": 402
298
+ },
299
+ {
300
+ "epoch": 1.02,
301
+ "learning_rate": 7.960199004975125e-06,
302
+ "loss": 0.4151,
303
+ "step": 410
304
+ },
305
+ {
306
+ "epoch": 1.04,
307
+ "learning_rate": 7.91044776119403e-06,
308
+ "loss": 0.4307,
309
+ "step": 420
310
+ },
311
+ {
312
+ "epoch": 1.07,
313
+ "learning_rate": 7.860696517412935e-06,
314
+ "loss": 0.4405,
315
+ "step": 430
316
+ },
317
+ {
318
+ "epoch": 1.09,
319
+ "learning_rate": 7.810945273631842e-06,
320
+ "loss": 0.4275,
321
+ "step": 440
322
+ },
323
+ {
324
+ "epoch": 1.12,
325
+ "learning_rate": 7.761194029850747e-06,
326
+ "loss": 0.4203,
327
+ "step": 450
328
+ },
329
+ {
330
+ "epoch": 1.14,
331
+ "learning_rate": 7.711442786069654e-06,
332
+ "loss": 0.4934,
333
+ "step": 460
334
+ },
335
+ {
336
+ "epoch": 1.17,
337
+ "learning_rate": 7.661691542288557e-06,
338
+ "loss": 0.4166,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 1.19,
343
+ "learning_rate": 7.611940298507463e-06,
344
+ "loss": 0.3872,
345
+ "step": 480
346
+ },
347
+ {
348
+ "epoch": 1.22,
349
+ "learning_rate": 7.5621890547263685e-06,
350
+ "loss": 0.362,
351
+ "step": 490
352
+ },
353
+ {
354
+ "epoch": 1.24,
355
+ "learning_rate": 7.512437810945274e-06,
356
+ "loss": 0.3795,
357
+ "step": 500
358
+ },
359
+ {
360
+ "epoch": 1.27,
361
+ "learning_rate": 7.46268656716418e-06,
362
+ "loss": 0.3736,
363
+ "step": 510
364
+ },
365
+ {
366
+ "epoch": 1.29,
367
+ "learning_rate": 7.412935323383084e-06,
368
+ "loss": 0.3603,
369
+ "step": 520
370
+ },
371
+ {
372
+ "epoch": 1.32,
373
+ "learning_rate": 7.36318407960199e-06,
374
+ "loss": 0.3526,
375
+ "step": 530
376
+ },
377
+ {
378
+ "epoch": 1.34,
379
+ "learning_rate": 7.313432835820896e-06,
380
+ "loss": 0.3054,
381
+ "step": 540
382
+ },
383
+ {
384
+ "epoch": 1.37,
385
+ "learning_rate": 7.263681592039802e-06,
386
+ "loss": 0.3601,
387
+ "step": 550
388
+ },
389
+ {
390
+ "epoch": 1.39,
391
+ "learning_rate": 7.213930348258708e-06,
392
+ "loss": 0.4352,
393
+ "step": 560
394
+ },
395
+ {
396
+ "epoch": 1.42,
397
+ "learning_rate": 7.164179104477612e-06,
398
+ "loss": 0.3177,
399
+ "step": 570
400
+ },
401
+ {
402
+ "epoch": 1.44,
403
+ "learning_rate": 7.114427860696518e-06,
404
+ "loss": 0.3465,
405
+ "step": 580
406
+ },
407
+ {
408
+ "epoch": 1.47,
409
+ "learning_rate": 7.064676616915423e-06,
410
+ "loss": 0.3679,
411
+ "step": 590
412
+ },
413
+ {
414
+ "epoch": 1.49,
415
+ "learning_rate": 7.014925373134329e-06,
416
+ "loss": 0.3933,
417
+ "step": 600
418
+ },
419
+ {
420
+ "epoch": 1.52,
421
+ "learning_rate": 6.965174129353235e-06,
422
+ "loss": 0.3282,
423
+ "step": 610
424
+ },
425
+ {
426
+ "epoch": 1.54,
427
+ "learning_rate": 6.915422885572139e-06,
428
+ "loss": 0.361,
429
+ "step": 620
430
+ },
431
+ {
432
+ "epoch": 1.57,
433
+ "learning_rate": 6.865671641791045e-06,
434
+ "loss": 0.3475,
435
+ "step": 630
436
+ },
437
+ {
438
+ "epoch": 1.59,
439
+ "learning_rate": 6.815920398009951e-06,
440
+ "loss": 0.3196,
441
+ "step": 640
442
+ },
443
+ {
444
+ "epoch": 1.62,
445
+ "learning_rate": 6.766169154228857e-06,
446
+ "loss": 0.2777,
447
+ "step": 650
448
+ },
449
+ {
450
+ "epoch": 1.64,
451
+ "learning_rate": 6.7164179104477625e-06,
452
+ "loss": 0.2905,
453
+ "step": 660
454
+ },
455
+ {
456
+ "epoch": 1.67,
457
+ "learning_rate": 6.666666666666667e-06,
458
+ "loss": 0.3959,
459
+ "step": 670
460
+ },
461
+ {
462
+ "epoch": 1.69,
463
+ "learning_rate": 6.6169154228855725e-06,
464
+ "loss": 0.3215,
465
+ "step": 680
466
+ },
467
+ {
468
+ "epoch": 1.72,
469
+ "learning_rate": 6.567164179104478e-06,
470
+ "loss": 0.3283,
471
+ "step": 690
472
+ },
473
+ {
474
+ "epoch": 1.74,
475
+ "learning_rate": 6.517412935323384e-06,
476
+ "loss": 0.2946,
477
+ "step": 700
478
+ },
479
+ {
480
+ "epoch": 1.77,
481
+ "learning_rate": 6.46766169154229e-06,
482
+ "loss": 0.3383,
483
+ "step": 710
484
+ },
485
+ {
486
+ "epoch": 1.79,
487
+ "learning_rate": 6.417910447761194e-06,
488
+ "loss": 0.279,
489
+ "step": 720
490
+ },
491
+ {
492
+ "epoch": 1.82,
493
+ "learning_rate": 6.3681592039801e-06,
494
+ "loss": 0.3557,
495
+ "step": 730
496
+ },
497
+ {
498
+ "epoch": 1.84,
499
+ "learning_rate": 6.318407960199006e-06,
500
+ "loss": 0.352,
501
+ "step": 740
502
+ },
503
+ {
504
+ "epoch": 1.87,
505
+ "learning_rate": 6.2686567164179116e-06,
506
+ "loss": 0.2888,
507
+ "step": 750
508
+ },
509
+ {
510
+ "epoch": 1.89,
511
+ "learning_rate": 6.218905472636816e-06,
512
+ "loss": 0.3714,
513
+ "step": 760
514
+ },
515
+ {
516
+ "epoch": 1.92,
517
+ "learning_rate": 6.1691542288557215e-06,
518
+ "loss": 0.2962,
519
+ "step": 770
520
+ },
521
+ {
522
+ "epoch": 1.94,
523
+ "learning_rate": 6.119402985074627e-06,
524
+ "loss": 0.3303,
525
+ "step": 780
526
+ },
527
+ {
528
+ "epoch": 1.97,
529
+ "learning_rate": 6.069651741293533e-06,
530
+ "loss": 0.311,
531
+ "step": 790
532
+ },
533
+ {
534
+ "epoch": 1.99,
535
+ "learning_rate": 6.019900497512439e-06,
536
+ "loss": 0.2763,
537
+ "step": 800
538
+ },
539
+ {
540
+ "epoch": 2.0,
541
+ "eval_DeTox:accuracy": 0.778816199376947,
542
+ "eval_DeTox:f1-macro": 0.7338028662531974,
543
+ "eval_DeTox:precision": 0.6781609195402298,
544
+ "eval_DeTox:recall": 0.5784313725490197,
545
+ "eval_HASOC2019:accuracy": 0.7878787878787878,
546
+ "eval_HASOC2019:f1-macro": 0.7664307381193125,
547
+ "eval_HASOC2019:precision": 0.6666666666666666,
548
+ "eval_HASOC2019:recall": 0.7272727272727273,
549
+ "eval_HASOC2020:accuracy": 0.6511627906976745,
550
+ "eval_HASOC2020:f1-macro": 0.5274725274725275,
551
+ "eval_HASOC2020:precision": 0.3333333333333333,
552
+ "eval_HASOC2020:recall": 0.25,
553
+ "eval_MHC:accuracy": NaN,
554
+ "eval_MHC:f1-macro": NaN,
555
+ "eval_MHC:precision": 0.0,
556
+ "eval_MHC:recall": 0.0,
557
+ "eval_RP:accuracy": 0.8618421052631579,
558
+ "eval_RP:f1-macro": 0.8342679127725857,
559
+ "eval_RP:precision": 0.9583333333333334,
560
+ "eval_RP:recall": 0.6388888888888888,
561
+ "eval_Round1:accuracy": NaN,
562
+ "eval_Round1:f1-macro": NaN,
563
+ "eval_Round1:precision": 0.0,
564
+ "eval_Round1:recall": 0.0,
565
+ "eval_Round2:accuracy": NaN,
566
+ "eval_Round2:f1-macro": NaN,
567
+ "eval_Round2:precision": 0.0,
568
+ "eval_Round2:recall": 0.0,
569
+ "eval_Round3:accuracy": NaN,
570
+ "eval_Round3:f1-macro": NaN,
571
+ "eval_Round3:precision": 0.0,
572
+ "eval_Round3:recall": 0.0,
573
+ "eval_Round4:accuracy": NaN,
574
+ "eval_Round4:f1-macro": NaN,
575
+ "eval_Round4:precision": 0.0,
576
+ "eval_Round4:recall": 0.0,
577
+ "eval_accuracy": 0.7919148936170213,
578
+ "eval_f1-macro": 0.7729607214046385,
579
+ "eval_loss": 0.5181771516799927,
580
+ "eval_precision": 0.8106995884773662,
581
+ "eval_recall": 0.6273885350318471,
582
+ "eval_runtime": 9.762,
583
+ "eval_samples_per_second": 240.729,
584
+ "eval_steps_per_second": 30.117,
585
+ "step": 804
586
+ },
587
+ {
588
+ "epoch": 2.01,
589
+ "learning_rate": 5.970149253731343e-06,
590
+ "loss": 0.2548,
591
+ "step": 810
592
+ },
593
+ {
594
+ "epoch": 2.04,
595
+ "learning_rate": 5.920398009950249e-06,
596
+ "loss": 0.3048,
597
+ "step": 820
598
+ },
599
+ {
600
+ "epoch": 2.06,
601
+ "learning_rate": 5.870646766169155e-06,
602
+ "loss": 0.2669,
603
+ "step": 830
604
+ },
605
+ {
606
+ "epoch": 2.09,
607
+ "learning_rate": 5.820895522388061e-06,
608
+ "loss": 0.2954,
609
+ "step": 840
610
+ },
611
+ {
612
+ "epoch": 2.11,
613
+ "learning_rate": 5.771144278606966e-06,
614
+ "loss": 0.2801,
615
+ "step": 850
616
+ },
617
+ {
618
+ "epoch": 2.14,
619
+ "learning_rate": 5.721393034825871e-06,
620
+ "loss": 0.4024,
621
+ "step": 860
622
+ },
623
+ {
624
+ "epoch": 2.16,
625
+ "learning_rate": 5.671641791044776e-06,
626
+ "loss": 0.2838,
627
+ "step": 870
628
+ },
629
+ {
630
+ "epoch": 2.19,
631
+ "learning_rate": 5.621890547263682e-06,
632
+ "loss": 0.2733,
633
+ "step": 880
634
+ },
635
+ {
636
+ "epoch": 2.21,
637
+ "learning_rate": 5.572139303482588e-06,
638
+ "loss": 0.2462,
639
+ "step": 890
640
+ },
641
+ {
642
+ "epoch": 2.24,
643
+ "learning_rate": 5.522388059701493e-06,
644
+ "loss": 0.225,
645
+ "step": 900
646
+ },
647
+ {
648
+ "epoch": 2.26,
649
+ "learning_rate": 5.472636815920398e-06,
650
+ "loss": 0.2447,
651
+ "step": 910
652
+ },
653
+ {
654
+ "epoch": 2.29,
655
+ "learning_rate": 5.422885572139304e-06,
656
+ "loss": 0.2392,
657
+ "step": 920
658
+ },
659
+ {
660
+ "epoch": 2.31,
661
+ "learning_rate": 5.37313432835821e-06,
662
+ "loss": 0.2926,
663
+ "step": 930
664
+ },
665
+ {
666
+ "epoch": 2.34,
667
+ "learning_rate": 5.323383084577115e-06,
668
+ "loss": 0.1636,
669
+ "step": 940
670
+ },
671
+ {
672
+ "epoch": 2.36,
673
+ "learning_rate": 5.2736318407960205e-06,
674
+ "loss": 0.2151,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 2.39,
679
+ "learning_rate": 5.2238805970149255e-06,
680
+ "loss": 0.26,
681
+ "step": 960
682
+ },
683
+ {
684
+ "epoch": 2.41,
685
+ "learning_rate": 5.174129353233831e-06,
686
+ "loss": 0.2943,
687
+ "step": 970
688
+ },
689
+ {
690
+ "epoch": 2.44,
691
+ "learning_rate": 5.124378109452737e-06,
692
+ "loss": 0.2122,
693
+ "step": 980
694
+ },
695
+ {
696
+ "epoch": 2.46,
697
+ "learning_rate": 5.074626865671642e-06,
698
+ "loss": 0.2709,
699
+ "step": 990
700
+ },
701
+ {
702
+ "epoch": 2.49,
703
+ "learning_rate": 5.024875621890548e-06,
704
+ "loss": 0.2713,
705
+ "step": 1000
706
+ },
707
+ {
708
+ "epoch": 2.51,
709
+ "learning_rate": 4.975124378109453e-06,
710
+ "loss": 0.2157,
711
+ "step": 1010
712
+ },
713
+ {
714
+ "epoch": 2.54,
715
+ "learning_rate": 4.925373134328359e-06,
716
+ "loss": 0.2441,
717
+ "step": 1020
718
+ },
719
+ {
720
+ "epoch": 2.56,
721
+ "learning_rate": 4.875621890547264e-06,
722
+ "loss": 0.2299,
723
+ "step": 1030
724
+ },
725
+ {
726
+ "epoch": 2.59,
727
+ "learning_rate": 4.8258706467661695e-06,
728
+ "loss": 0.2002,
729
+ "step": 1040
730
+ },
731
+ {
732
+ "epoch": 2.61,
733
+ "learning_rate": 4.7761194029850745e-06,
734
+ "loss": 0.1875,
735
+ "step": 1050
736
+ },
737
+ {
738
+ "epoch": 2.64,
739
+ "learning_rate": 4.72636815920398e-06,
740
+ "loss": 0.2272,
741
+ "step": 1060
742
+ },
743
+ {
744
+ "epoch": 2.66,
745
+ "learning_rate": 4.676616915422886e-06,
746
+ "loss": 0.2605,
747
+ "step": 1070
748
+ },
749
+ {
750
+ "epoch": 2.69,
751
+ "learning_rate": 4.626865671641791e-06,
752
+ "loss": 0.2487,
753
+ "step": 1080
754
+ },
755
+ {
756
+ "epoch": 2.71,
757
+ "learning_rate": 4.577114427860697e-06,
758
+ "loss": 0.2037,
759
+ "step": 1090
760
+ },
761
+ {
762
+ "epoch": 2.74,
763
+ "learning_rate": 4.527363184079602e-06,
764
+ "loss": 0.2093,
765
+ "step": 1100
766
+ },
767
+ {
768
+ "epoch": 2.76,
769
+ "learning_rate": 4.477611940298508e-06,
770
+ "loss": 0.1975,
771
+ "step": 1110
772
+ },
773
+ {
774
+ "epoch": 2.79,
775
+ "learning_rate": 4.427860696517413e-06,
776
+ "loss": 0.1983,
777
+ "step": 1120
778
+ },
779
+ {
780
+ "epoch": 2.81,
781
+ "learning_rate": 4.378109452736319e-06,
782
+ "loss": 0.2215,
783
+ "step": 1130
784
+ },
785
+ {
786
+ "epoch": 2.84,
787
+ "learning_rate": 4.3283582089552236e-06,
788
+ "loss": 0.2411,
789
+ "step": 1140
790
+ },
791
+ {
792
+ "epoch": 2.86,
793
+ "learning_rate": 4.278606965174129e-06,
794
+ "loss": 0.2356,
795
+ "step": 1150
796
+ },
797
+ {
798
+ "epoch": 2.89,
799
+ "learning_rate": 4.228855721393035e-06,
800
+ "loss": 0.2352,
801
+ "step": 1160
802
+ },
803
+ {
804
+ "epoch": 2.91,
805
+ "learning_rate": 4.17910447761194e-06,
806
+ "loss": 0.1669,
807
+ "step": 1170
808
+ },
809
+ {
810
+ "epoch": 2.94,
811
+ "learning_rate": 4.129353233830846e-06,
812
+ "loss": 0.2586,
813
+ "step": 1180
814
+ },
815
+ {
816
+ "epoch": 2.96,
817
+ "learning_rate": 4.079601990049751e-06,
818
+ "loss": 0.2224,
819
+ "step": 1190
820
+ },
821
+ {
822
+ "epoch": 2.99,
823
+ "learning_rate": 4.029850746268657e-06,
824
+ "loss": 0.2033,
825
+ "step": 1200
826
+ },
827
+ {
828
+ "epoch": 3.0,
829
+ "eval_DeTox:accuracy": 0.8037383177570093,
830
+ "eval_DeTox:f1-macro": 0.774259083551934,
831
+ "eval_DeTox:precision": 0.6893203883495146,
832
+ "eval_DeTox:recall": 0.696078431372549,
833
+ "eval_HASOC2019:accuracy": 0.7575757575757576,
834
+ "eval_HASOC2019:f1-macro": 0.7518796992481203,
835
+ "eval_HASOC2019:precision": 0.5882352941176471,
836
+ "eval_HASOC2019:recall": 0.9090909090909091,
837
+ "eval_HASOC2020:accuracy": 0.7209302325581395,
838
+ "eval_HASOC2020:f1-macro": 0.6692307692307692,
839
+ "eval_HASOC2020:precision": 0.5,
840
+ "eval_HASOC2020:recall": 0.5833333333333334,
841
+ "eval_MHC:accuracy": NaN,
842
+ "eval_MHC:f1-macro": NaN,
843
+ "eval_MHC:precision": 0.0,
844
+ "eval_MHC:recall": 0.0,
845
+ "eval_RP:accuracy": 0.8782894736842105,
846
+ "eval_RP:f1-macro": 0.8595667644671952,
847
+ "eval_RP:precision": 0.9176470588235294,
848
+ "eval_RP:recall": 0.7222222222222222,
849
+ "eval_Round1:accuracy": NaN,
850
+ "eval_Round1:f1-macro": NaN,
851
+ "eval_Round1:precision": 0.0,
852
+ "eval_Round1:recall": 0.0,
853
+ "eval_Round2:accuracy": NaN,
854
+ "eval_Round2:f1-macro": NaN,
855
+ "eval_Round2:precision": 0.0,
856
+ "eval_Round2:recall": 0.0,
857
+ "eval_Round3:accuracy": NaN,
858
+ "eval_Round3:f1-macro": NaN,
859
+ "eval_Round3:precision": 0.0,
860
+ "eval_Round3:recall": 0.0,
861
+ "eval_Round4:accuracy": NaN,
862
+ "eval_Round4:f1-macro": NaN,
863
+ "eval_Round4:precision": 0.0,
864
+ "eval_Round4:recall": 0.0,
865
+ "eval_accuracy": 0.8246808510638298,
866
+ "eval_f1-macro": 0.8160155062328975,
867
+ "eval_loss": 0.4982839524745941,
868
+ "eval_precision": 0.7951002227171492,
869
+ "eval_recall": 0.7579617834394905,
870
+ "eval_runtime": 9.7615,
871
+ "eval_samples_per_second": 240.741,
872
+ "eval_steps_per_second": 30.118,
873
+ "step": 1206
874
+ },
875
+ {
876
+ "epoch": 3.01,
877
+ "learning_rate": 3.980099502487563e-06,
878
+ "loss": 0.15,
879
+ "step": 1210
880
+ },
881
+ {
882
+ "epoch": 3.03,
883
+ "learning_rate": 3.930348258706468e-06,
884
+ "loss": 0.2823,
885
+ "step": 1220
886
+ },
887
+ {
888
+ "epoch": 3.06,
889
+ "learning_rate": 3.8805970149253735e-06,
890
+ "loss": 0.2278,
891
+ "step": 1230
892
+ },
893
+ {
894
+ "epoch": 3.08,
895
+ "learning_rate": 3.8308457711442784e-06,
896
+ "loss": 0.1614,
897
+ "step": 1240
898
+ },
899
+ {
900
+ "epoch": 3.11,
901
+ "learning_rate": 3.7810945273631843e-06,
902
+ "loss": 0.1793,
903
+ "step": 1250
904
+ },
905
+ {
906
+ "epoch": 3.13,
907
+ "learning_rate": 3.73134328358209e-06,
908
+ "loss": 0.3277,
909
+ "step": 1260
910
+ },
911
+ {
912
+ "epoch": 3.16,
913
+ "learning_rate": 3.681592039800995e-06,
914
+ "loss": 0.185,
915
+ "step": 1270
916
+ },
917
+ {
918
+ "epoch": 3.18,
919
+ "learning_rate": 3.631840796019901e-06,
920
+ "loss": 0.1947,
921
+ "step": 1280
922
+ },
923
+ {
924
+ "epoch": 3.21,
925
+ "learning_rate": 3.582089552238806e-06,
926
+ "loss": 0.1548,
927
+ "step": 1290
928
+ },
929
+ {
930
+ "epoch": 3.23,
931
+ "learning_rate": 3.5323383084577117e-06,
932
+ "loss": 0.1412,
933
+ "step": 1300
934
+ },
935
+ {
936
+ "epoch": 3.26,
937
+ "learning_rate": 3.4825870646766175e-06,
938
+ "loss": 0.1608,
939
+ "step": 1310
940
+ },
941
+ {
942
+ "epoch": 3.28,
943
+ "learning_rate": 3.4328358208955225e-06,
944
+ "loss": 0.1994,
945
+ "step": 1320
946
+ },
947
+ {
948
+ "epoch": 3.31,
949
+ "learning_rate": 3.3830845771144283e-06,
950
+ "loss": 0.2287,
951
+ "step": 1330
952
+ },
953
+ {
954
+ "epoch": 3.33,
955
+ "learning_rate": 3.3333333333333333e-06,
956
+ "loss": 0.2052,
957
+ "step": 1340
958
+ },
959
+ {
960
+ "epoch": 3.36,
961
+ "learning_rate": 3.283582089552239e-06,
962
+ "loss": 0.1324,
963
+ "step": 1350
964
+ },
965
+ {
966
+ "epoch": 3.38,
967
+ "learning_rate": 3.233830845771145e-06,
968
+ "loss": 0.1727,
969
+ "step": 1360
970
+ },
971
+ {
972
+ "epoch": 3.41,
973
+ "learning_rate": 3.18407960199005e-06,
974
+ "loss": 0.2653,
975
+ "step": 1370
976
+ },
977
+ {
978
+ "epoch": 3.43,
979
+ "learning_rate": 3.1343283582089558e-06,
980
+ "loss": 0.1253,
981
+ "step": 1380
982
+ },
983
+ {
984
+ "epoch": 3.46,
985
+ "learning_rate": 3.0845771144278608e-06,
986
+ "loss": 0.1939,
987
+ "step": 1390
988
+ },
989
+ {
990
+ "epoch": 3.48,
991
+ "learning_rate": 3.0348258706467666e-06,
992
+ "loss": 0.1671,
993
+ "step": 1400
994
+ },
995
+ {
996
+ "epoch": 3.51,
997
+ "learning_rate": 2.9850746268656716e-06,
998
+ "loss": 0.1425,
999
+ "step": 1410
1000
+ },
1001
+ {
1002
+ "epoch": 3.53,
1003
+ "learning_rate": 2.9353233830845774e-06,
1004
+ "loss": 0.2261,
1005
+ "step": 1420
1006
+ },
1007
+ {
1008
+ "epoch": 3.56,
1009
+ "learning_rate": 2.885572139303483e-06,
1010
+ "loss": 0.152,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 3.58,
1015
+ "learning_rate": 2.835820895522388e-06,
1016
+ "loss": 0.15,
1017
+ "step": 1440
1018
+ },
1019
+ {
1020
+ "epoch": 3.61,
1021
+ "learning_rate": 2.786069651741294e-06,
1022
+ "loss": 0.1616,
1023
+ "step": 1450
1024
+ },
1025
+ {
1026
+ "epoch": 3.63,
1027
+ "learning_rate": 2.736318407960199e-06,
1028
+ "loss": 0.147,
1029
+ "step": 1460
1030
+ },
1031
+ {
1032
+ "epoch": 3.66,
1033
+ "learning_rate": 2.686567164179105e-06,
1034
+ "loss": 0.2288,
1035
+ "step": 1470
1036
+ },
1037
+ {
1038
+ "epoch": 3.68,
1039
+ "learning_rate": 2.6368159203980102e-06,
1040
+ "loss": 0.1681,
1041
+ "step": 1480
1042
+ },
1043
+ {
1044
+ "epoch": 3.71,
1045
+ "learning_rate": 2.5870646766169156e-06,
1046
+ "loss": 0.1668,
1047
+ "step": 1490
1048
+ },
1049
+ {
1050
+ "epoch": 3.73,
1051
+ "learning_rate": 2.537313432835821e-06,
1052
+ "loss": 0.159,
1053
+ "step": 1500
1054
+ },
1055
+ {
1056
+ "epoch": 3.76,
1057
+ "learning_rate": 2.4875621890547264e-06,
1058
+ "loss": 0.1464,
1059
+ "step": 1510
1060
+ },
1061
+ {
1062
+ "epoch": 3.78,
1063
+ "learning_rate": 2.437810945273632e-06,
1064
+ "loss": 0.1449,
1065
+ "step": 1520
1066
+ },
1067
+ {
1068
+ "epoch": 3.81,
1069
+ "learning_rate": 2.3880597014925373e-06,
1070
+ "loss": 0.1459,
1071
+ "step": 1530
1072
+ },
1073
+ {
1074
+ "epoch": 3.83,
1075
+ "learning_rate": 2.338308457711443e-06,
1076
+ "loss": 0.2006,
1077
+ "step": 1540
1078
+ },
1079
+ {
1080
+ "epoch": 3.86,
1081
+ "learning_rate": 2.2885572139303485e-06,
1082
+ "loss": 0.1621,
1083
+ "step": 1550
1084
+ },
1085
+ {
1086
+ "epoch": 3.88,
1087
+ "learning_rate": 2.238805970149254e-06,
1088
+ "loss": 0.2158,
1089
+ "step": 1560
1090
+ },
1091
+ {
1092
+ "epoch": 3.91,
1093
+ "learning_rate": 2.1890547263681593e-06,
1094
+ "loss": 0.1351,
1095
+ "step": 1570
1096
+ },
1097
+ {
1098
+ "epoch": 3.93,
1099
+ "learning_rate": 2.1393034825870647e-06,
1100
+ "loss": 0.1395,
1101
+ "step": 1580
1102
+ },
1103
+ {
1104
+ "epoch": 3.96,
1105
+ "learning_rate": 2.08955223880597e-06,
1106
+ "loss": 0.1569,
1107
+ "step": 1590
1108
+ },
1109
+ {
1110
+ "epoch": 3.98,
1111
+ "learning_rate": 2.0398009950248755e-06,
1112
+ "loss": 0.1868,
1113
+ "step": 1600
1114
+ },
1115
+ {
1116
+ "epoch": 4.0,
1117
+ "eval_DeTox:accuracy": 0.8130841121495327,
1118
+ "eval_DeTox:f1-macro": 0.7855615925892937,
1119
+ "eval_DeTox:precision": 0.7019230769230769,
1120
+ "eval_DeTox:recall": 0.7156862745098039,
1121
+ "eval_HASOC2019:accuracy": 0.7575757575757576,
1122
+ "eval_HASOC2019:f1-macro": 0.7518796992481203,
1123
+ "eval_HASOC2019:precision": 0.5882352941176471,
1124
+ "eval_HASOC2019:recall": 0.9090909090909091,
1125
+ "eval_HASOC2020:accuracy": 0.7441860465116279,
1126
+ "eval_HASOC2020:f1-macro": 0.7030759573132455,
1127
+ "eval_HASOC2020:precision": 0.5333333333333333,
1128
+ "eval_HASOC2020:recall": 0.6666666666666666,
1129
+ "eval_MHC:accuracy": NaN,
1130
+ "eval_MHC:f1-macro": NaN,
1131
+ "eval_MHC:precision": 0.0,
1132
+ "eval_MHC:recall": 0.0,
1133
+ "eval_RP:accuracy": 0.881578947368421,
1134
+ "eval_RP:f1-macro": 0.8651884700665189,
1135
+ "eval_RP:precision": 0.9,
1136
+ "eval_RP:recall": 0.75,
1137
+ "eval_Round1:accuracy": NaN,
1138
+ "eval_Round1:f1-macro": NaN,
1139
+ "eval_Round1:precision": 0.0,
1140
+ "eval_Round1:recall": 0.0,
1141
+ "eval_Round2:accuracy": NaN,
1142
+ "eval_Round2:f1-macro": NaN,
1143
+ "eval_Round2:precision": 0.0,
1144
+ "eval_Round2:recall": 0.0,
1145
+ "eval_Round3:accuracy": NaN,
1146
+ "eval_Round3:f1-macro": NaN,
1147
+ "eval_Round3:precision": 0.0,
1148
+ "eval_Round3:recall": 0.0,
1149
+ "eval_Round4:accuracy": NaN,
1150
+ "eval_Round4:f1-macro": NaN,
1151
+ "eval_Round4:precision": 0.0,
1152
+ "eval_Round4:recall": 0.0,
1153
+ "eval_accuracy": 0.828936170212766,
1154
+ "eval_f1-macro": 0.8213584698436758,
1155
+ "eval_loss": 0.5340345501899719,
1156
+ "eval_precision": 0.7922077922077922,
1157
+ "eval_recall": 0.7770700636942676,
1158
+ "eval_runtime": 10.0181,
1159
+ "eval_samples_per_second": 234.576,
1160
+ "eval_steps_per_second": 29.347,
1161
+ "step": 1608
1162
+ },
1163
+ {
1164
+ "epoch": 4.0,
1165
+ "learning_rate": 1.9900497512437813e-06,
1166
+ "loss": 0.0959,
1167
+ "step": 1610
1168
+ },
1169
+ {
1170
+ "epoch": 4.03,
1171
+ "learning_rate": 1.9402985074626867e-06,
1172
+ "loss": 0.2036,
1173
+ "step": 1620
1174
+ },
1175
+ {
1176
+ "epoch": 4.05,
1177
+ "learning_rate": 1.8905472636815921e-06,
1178
+ "loss": 0.1506,
1179
+ "step": 1630
1180
+ },
1181
+ {
1182
+ "epoch": 4.08,
1183
+ "learning_rate": 1.8407960199004975e-06,
1184
+ "loss": 0.1104,
1185
+ "step": 1640
1186
+ },
1187
+ {
1188
+ "epoch": 4.1,
1189
+ "learning_rate": 1.791044776119403e-06,
1190
+ "loss": 0.1229,
1191
+ "step": 1650
1192
+ },
1193
+ {
1194
+ "epoch": 4.13,
1195
+ "learning_rate": 1.7412935323383088e-06,
1196
+ "loss": 0.1676,
1197
+ "step": 1660
1198
+ },
1199
+ {
1200
+ "epoch": 4.15,
1201
+ "learning_rate": 1.6915422885572142e-06,
1202
+ "loss": 0.1602,
1203
+ "step": 1670
1204
+ },
1205
+ {
1206
+ "epoch": 4.18,
1207
+ "learning_rate": 1.6417910447761196e-06,
1208
+ "loss": 0.1397,
1209
+ "step": 1680
1210
+ },
1211
+ {
1212
+ "epoch": 4.2,
1213
+ "learning_rate": 1.592039800995025e-06,
1214
+ "loss": 0.1205,
1215
+ "step": 1690
1216
+ },
1217
+ {
1218
+ "epoch": 4.23,
1219
+ "learning_rate": 1.5422885572139304e-06,
1220
+ "loss": 0.1497,
1221
+ "step": 1700
1222
+ },
1223
+ {
1224
+ "epoch": 4.25,
1225
+ "learning_rate": 1.4925373134328358e-06,
1226
+ "loss": 0.085,
1227
+ "step": 1710
1228
+ },
1229
+ {
1230
+ "epoch": 4.28,
1231
+ "learning_rate": 1.4427860696517414e-06,
1232
+ "loss": 0.1142,
1233
+ "step": 1720
1234
+ },
1235
+ {
1236
+ "epoch": 4.3,
1237
+ "learning_rate": 1.393034825870647e-06,
1238
+ "loss": 0.1601,
1239
+ "step": 1730
1240
+ },
1241
+ {
1242
+ "epoch": 4.33,
1243
+ "learning_rate": 1.3432835820895524e-06,
1244
+ "loss": 0.1047,
1245
+ "step": 1740
1246
+ },
1247
+ {
1248
+ "epoch": 4.35,
1249
+ "learning_rate": 1.2935323383084578e-06,
1250
+ "loss": 0.0812,
1251
+ "step": 1750
1252
+ },
1253
+ {
1254
+ "epoch": 4.38,
1255
+ "learning_rate": 1.2437810945273632e-06,
1256
+ "loss": 0.0614,
1257
+ "step": 1760
1258
+ },
1259
+ {
1260
+ "epoch": 4.4,
1261
+ "learning_rate": 1.1940298507462686e-06,
1262
+ "loss": 0.169,
1263
+ "step": 1770
1264
+ },
1265
+ {
1266
+ "epoch": 4.43,
1267
+ "learning_rate": 1.1442786069651742e-06,
1268
+ "loss": 0.0794,
1269
+ "step": 1780
1270
+ },
1271
+ {
1272
+ "epoch": 4.45,
1273
+ "learning_rate": 1.0945273631840796e-06,
1274
+ "loss": 0.1185,
1275
+ "step": 1790
1276
+ },
1277
+ {
1278
+ "epoch": 4.48,
1279
+ "learning_rate": 1.044776119402985e-06,
1280
+ "loss": 0.1428,
1281
+ "step": 1800
1282
+ },
1283
+ {
1284
+ "epoch": 4.5,
1285
+ "learning_rate": 9.950248756218907e-07,
1286
+ "loss": 0.1351,
1287
+ "step": 1810
1288
+ },
1289
+ {
1290
+ "epoch": 4.53,
1291
+ "learning_rate": 9.452736318407961e-07,
1292
+ "loss": 0.1341,
1293
+ "step": 1820
1294
+ },
1295
+ {
1296
+ "epoch": 4.55,
1297
+ "learning_rate": 8.955223880597015e-07,
1298
+ "loss": 0.0735,
1299
+ "step": 1830
1300
+ },
1301
+ {
1302
+ "epoch": 4.58,
1303
+ "learning_rate": 8.457711442786071e-07,
1304
+ "loss": 0.1164,
1305
+ "step": 1840
1306
+ },
1307
+ {
1308
+ "epoch": 4.6,
1309
+ "learning_rate": 7.960199004975125e-07,
1310
+ "loss": 0.0471,
1311
+ "step": 1850
1312
+ },
1313
+ {
1314
+ "epoch": 4.63,
1315
+ "learning_rate": 7.462686567164179e-07,
1316
+ "loss": 0.1176,
1317
+ "step": 1860
1318
+ },
1319
+ {
1320
+ "epoch": 4.65,
1321
+ "learning_rate": 6.965174129353235e-07,
1322
+ "loss": 0.1874,
1323
+ "step": 1870
1324
+ },
1325
+ {
1326
+ "epoch": 4.68,
1327
+ "learning_rate": 6.467661691542289e-07,
1328
+ "loss": 0.159,
1329
+ "step": 1880
1330
+ },
1331
+ {
1332
+ "epoch": 4.7,
1333
+ "learning_rate": 5.970149253731343e-07,
1334
+ "loss": 0.0975,
1335
+ "step": 1890
1336
+ },
1337
+ {
1338
+ "epoch": 4.73,
1339
+ "learning_rate": 5.472636815920398e-07,
1340
+ "loss": 0.0947,
1341
+ "step": 1900
1342
+ },
1343
+ {
1344
+ "epoch": 4.75,
1345
+ "learning_rate": 4.975124378109453e-07,
1346
+ "loss": 0.0872,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 4.78,
1351
+ "learning_rate": 4.4776119402985074e-07,
1352
+ "loss": 0.0993,
1353
+ "step": 1920
1354
+ },
1355
+ {
1356
+ "epoch": 4.8,
1357
+ "learning_rate": 3.9800995024875624e-07,
1358
+ "loss": 0.1159,
1359
+ "step": 1930
1360
+ },
1361
+ {
1362
+ "epoch": 4.83,
1363
+ "learning_rate": 3.4825870646766175e-07,
1364
+ "loss": 0.1865,
1365
+ "step": 1940
1366
+ },
1367
+ {
1368
+ "epoch": 4.85,
1369
+ "learning_rate": 2.9850746268656716e-07,
1370
+ "loss": 0.1229,
1371
+ "step": 1950
1372
+ },
1373
+ {
1374
+ "epoch": 4.88,
1375
+ "learning_rate": 2.4875621890547267e-07,
1376
+ "loss": 0.1395,
1377
+ "step": 1960
1378
+ },
1379
+ {
1380
+ "epoch": 4.9,
1381
+ "learning_rate": 1.9900497512437812e-07,
1382
+ "loss": 0.1113,
1383
+ "step": 1970
1384
+ },
1385
+ {
1386
+ "epoch": 4.93,
1387
+ "learning_rate": 1.4925373134328358e-07,
1388
+ "loss": 0.1167,
1389
+ "step": 1980
1390
+ },
1391
+ {
1392
+ "epoch": 4.95,
1393
+ "learning_rate": 9.950248756218906e-08,
1394
+ "loss": 0.093,
1395
+ "step": 1990
1396
+ },
1397
+ {
1398
+ "epoch": 4.98,
1399
+ "learning_rate": 4.975124378109453e-08,
1400
+ "loss": 0.1402,
1401
+ "step": 2000
1402
+ },
1403
+ {
1404
+ "epoch": 5.0,
1405
+ "learning_rate": 0.0,
1406
+ "loss": 0.0734,
1407
+ "step": 2010
1408
+ },
1409
+ {
1410
+ "epoch": 5.0,
1411
+ "eval_DeTox:accuracy": 0.8193146417445483,
1412
+ "eval_DeTox:f1-macro": 0.7985762506491259,
1413
+ "eval_DeTox:precision": 0.6896551724137931,
1414
+ "eval_DeTox:recall": 0.7843137254901961,
1415
+ "eval_HASOC2019:accuracy": 0.7272727272727273,
1416
+ "eval_HASOC2019:f1-macro": 0.723205964585275,
1417
+ "eval_HASOC2019:precision": 0.5555555555555556,
1418
+ "eval_HASOC2019:recall": 0.9090909090909091,
1419
+ "eval_HASOC2020:accuracy": 0.6976744186046512,
1420
+ "eval_HASOC2020:f1-macro": 0.661826981246219,
1421
+ "eval_HASOC2020:precision": 0.47058823529411764,
1422
+ "eval_HASOC2020:recall": 0.6666666666666666,
1423
+ "eval_MHC:accuracy": NaN,
1424
+ "eval_MHC:f1-macro": NaN,
1425
+ "eval_MHC:precision": 0.0,
1426
+ "eval_MHC:recall": 0.0,
1427
+ "eval_RP:accuracy": 0.881578947368421,
1428
+ "eval_RP:f1-macro": 0.8672102504368084,
1429
+ "eval_RP:precision": 0.875,
1430
+ "eval_RP:recall": 0.7777777777777778,
1431
+ "eval_Round1:accuracy": NaN,
1432
+ "eval_Round1:f1-macro": NaN,
1433
+ "eval_Round1:precision": 0.0,
1434
+ "eval_Round1:recall": 0.0,
1435
+ "eval_Round2:accuracy": NaN,
1436
+ "eval_Round2:f1-macro": NaN,
1437
+ "eval_Round2:precision": 0.0,
1438
+ "eval_Round2:recall": 0.0,
1439
+ "eval_Round3:accuracy": NaN,
1440
+ "eval_Round3:f1-macro": NaN,
1441
+ "eval_Round3:precision": 0.0,
1442
+ "eval_Round3:recall": 0.0,
1443
+ "eval_Round4:accuracy": NaN,
1444
+ "eval_Round4:f1-macro": NaN,
1445
+ "eval_Round4:precision": 0.0,
1446
+ "eval_Round4:recall": 0.0,
1447
+ "eval_accuracy": 0.8336170212765958,
1448
+ "eval_f1-macro": 0.8283671540104293,
1449
+ "eval_loss": 0.6073299050331116,
1450
+ "eval_precision": 0.7763289869608826,
1451
+ "eval_recall": 0.821656050955414,
1452
+ "eval_runtime": 9.7194,
1453
+ "eval_samples_per_second": 241.784,
1454
+ "eval_steps_per_second": 30.249,
1455
+ "step": 2010
1456
+ }
1457
+ ],
1458
+ "logging_steps": 10,
1459
+ "max_steps": 2010,
1460
+ "num_train_epochs": 5,
1461
+ "save_steps": 500,
1462
+ "total_flos": 9308565479568180.0,
1463
+ "trial_name": null,
1464
+ "trial_params": null
1465
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8837d57b537787bd5a353447f8234e4d6725e03f559cf5139cb6033716f8e5a
3
+ size 4027
vocab.txt ADDED
The diff for this file is too large to render. See raw diff