Transformers
PyTorch
Indonesian
bert
Iftitahu commited on
Commit
5a12e3f
·
verified ·
1 Parent(s): 2745c39

Upload 11 files

Browse files
config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "result/unsup/simcse-indobert-content-FakeCLSTrain-precise",
3
+ "_num_labels": 5,
4
+ "architectures": [
5
+ "BertForCL"
6
+ ],
7
+ "attention_probs_dropout_prob": 0.1,
8
+ "classifier_dropout": null,
9
+ "directionality": "bidi",
10
+ "gradient_checkpointing": false,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 768,
14
+ "id2label": {
15
+ "0": "LABEL_0",
16
+ "1": "LABEL_1",
17
+ "2": "LABEL_2",
18
+ "3": "LABEL_3",
19
+ "4": "LABEL_4"
20
+ },
21
+ "initializer_range": 0.02,
22
+ "intermediate_size": 3072,
23
+ "label2id": {
24
+ "LABEL_0": 0,
25
+ "LABEL_1": 1,
26
+ "LABEL_2": 2,
27
+ "LABEL_3": 3,
28
+ "LABEL_4": 4
29
+ },
30
+ "layer_norm_eps": 1e-12,
31
+ "max_position_embeddings": 512,
32
+ "model_type": "bert",
33
+ "num_attention_heads": 12,
34
+ "num_hidden_layers": 12,
35
+ "output_past": true,
36
+ "pad_token_id": 0,
37
+ "pooler_fc_size": 768,
38
+ "pooler_num_attention_heads": 12,
39
+ "pooler_num_fc_layers": 3,
40
+ "pooler_size_per_head": 128,
41
+ "pooler_type": "first_token_transform",
42
+ "position_embedding_type": "absolute",
43
+ "torch_dtype": "float32",
44
+ "transformers_version": "4.2.1",
45
+ "type_vocab_size": 2,
46
+ "use_cache": true,
47
+ "vocab_size": 30521
48
+ }
eval_results.txt ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ epoch = 3.0
2
+ eval_CR = 72.4
3
+ eval_FakeCLSDev = 0.75
4
+ eval_FakeCLSTest = 0.6
5
+ eval_FakeCLSTrain = 0.5866261398176292
6
+ eval_FakePairs = 0.009186111367833594
7
+ eval_FakePairsNLI = 61.31
8
+ eval_MPQA = 77.71
9
+ eval_MR = 62.08
10
+ eval_MRPC = 70.66
11
+ eval_SST2 = 68.92
12
+ eval_SUBJ = 81.17
13
+ eval_TREC = 62.38
14
+ eval_avg_sts = 0.6025418380141683
15
+ eval_avg_transfer = 70.76
16
+ eval_sickr_spearman = 0.5790834682182003
17
+ eval_stsb_indo = 0.2856567184812426
18
+ eval_stsb_spearman = 0.6260002078101363
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60bafe8fc2f177201b419b92d517ab4fb3f2046c0e52b750aecfdee192dd3b8d
3
+ size 875968954
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:661da7ffb02996db36fc801db16e8c61dac9acc40359203c8e2e4c3a4dab3cc2
3
+ size 438011626
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6a8f190d2b477700dfe75d0e6a8f67d67b0672fe16cf3d5b42ddbba0f514fa6
3
+ size 1064
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "clean_up_tokenization_spaces": true, "do_basic_tokenize": true, "model_max_length": 1000000000000000019884624838656, "never_split": null, "tokenizer_class": "BertTokenizer", "special_tokens_map_file": "/home/ifti001/.cache/huggingface/transformers/7cc9f436cefa78ab595ac4e5b35da41475c4703c7dfa70ca087711ea68d89a2f.7da70648c6cb9951e284c9685f9ba7ae083dd59ed1d6d84bdfc0584a4ea94b6d", "name_or_path": "LazarusNLP/simcse-indobert-base"}
train_results.txt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ epoch = 3.0
2
+ train_runtime = 51381.4965
3
+ train_samples_per_second = 0.015
trainer_state.json ADDED
@@ -0,0 +1,805 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.5866261398176292,
3
+ "best_model_checkpoint": "result/unsup/simcse-indobert-content-FakeCLSTrain-precise",
4
+ "epoch": 3.0,
5
+ "global_step": 756,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "eval_CR": 73.16,
13
+ "eval_FakeCLSDev": 0.9856897538637664,
14
+ "eval_FakeCLSTest": 0.5,
15
+ "eval_FakeCLSTrain": 0.5635838150289018,
16
+ "eval_FakePairs": -0.0605484557984162,
17
+ "eval_FakePairsNLI": 62.09,
18
+ "eval_MPQA": 77.41,
19
+ "eval_MR": 61.65,
20
+ "eval_MRPC": 69.68,
21
+ "eval_SST2": 69.95,
22
+ "eval_SUBJ": 82.09,
23
+ "eval_TREC": 65.04,
24
+ "eval_avg_sts": 0.594301042267096,
25
+ "eval_avg_transfer": 71.28285714285714,
26
+ "eval_sickr_spearman": 0.5739718062787506,
27
+ "eval_stsb_indo": 0.3209488720594688,
28
+ "eval_stsb_spearman": 0.6146302782554416,
29
+ "step": 20
30
+ },
31
+ {
32
+ "epoch": 0.16,
33
+ "eval_CR": 72.46,
34
+ "eval_FakeCLSDev": 0.9851598173515982,
35
+ "eval_FakeCLSTest": 0.44155844155844154,
36
+ "eval_FakeCLSTrain": 0.5580736543909348,
37
+ "eval_FakePairs": -0.06458236398195215,
38
+ "eval_FakePairsNLI": 62.56,
39
+ "eval_MPQA": 77.78,
40
+ "eval_MR": 61.94,
41
+ "eval_MRPC": 69.92,
42
+ "eval_SST2": 69.72,
43
+ "eval_SUBJ": 81.95,
44
+ "eval_TREC": 63.56,
45
+ "eval_avg_sts": 0.5942119231714487,
46
+ "eval_avg_transfer": 71.04714285714286,
47
+ "eval_sickr_spearman": 0.572479335863764,
48
+ "eval_stsb_indo": 0.33555082512700035,
49
+ "eval_stsb_spearman": 0.6159445104791332,
50
+ "step": 40
51
+ },
52
+ {
53
+ "epoch": 0.24,
54
+ "eval_CR": 72.2,
55
+ "eval_FakeCLSDev": 0.9883973894126178,
56
+ "eval_FakeCLSTest": 0.49557522123893805,
57
+ "eval_FakeCLSTrain": 0.5488126649076517,
58
+ "eval_FakePairs": -0.06162683895274021,
59
+ "eval_FakePairsNLI": 62.25,
60
+ "eval_MPQA": 77.85,
61
+ "eval_MR": 61.31,
62
+ "eval_MRPC": 70.17,
63
+ "eval_SST2": 69.95,
64
+ "eval_SUBJ": 82.79,
65
+ "eval_TREC": 63.96,
66
+ "eval_avg_sts": 0.5899751766827189,
67
+ "eval_avg_transfer": 71.17571428571428,
68
+ "eval_sickr_spearman": 0.5702007884426414,
69
+ "eval_stsb_indo": 0.31928216322520336,
70
+ "eval_stsb_spearman": 0.6097495649227964,
71
+ "step": 60
72
+ },
73
+ {
74
+ "epoch": 0.32,
75
+ "eval_CR": 72.32,
76
+ "eval_FakeCLSDev": 0.9828571428571429,
77
+ "eval_FakeCLSTest": 0.4358974358974359,
78
+ "eval_FakeCLSTrain": 0.505720823798627,
79
+ "eval_FakePairs": -0.057193528168424795,
80
+ "eval_FakePairsNLI": 60.53,
81
+ "eval_MPQA": 77.43,
82
+ "eval_MR": 62.04,
83
+ "eval_MRPC": 69.38,
84
+ "eval_SST2": 69.72,
85
+ "eval_SUBJ": 82.67,
86
+ "eval_TREC": 63.44,
87
+ "eval_avg_sts": 0.6008571910373384,
88
+ "eval_avg_transfer": 70.99999999999999,
89
+ "eval_sickr_spearman": 0.5721567109554042,
90
+ "eval_stsb_indo": 0.3125770393791657,
91
+ "eval_stsb_spearman": 0.6295576711192724,
92
+ "step": 80
93
+ },
94
+ {
95
+ "epoch": 0.4,
96
+ "eval_CR": 71.95,
97
+ "eval_FakeCLSDev": 1.0,
98
+ "eval_FakeCLSTest": 0.4166666666666667,
99
+ "eval_FakeCLSTrain": 0.5508474576271186,
100
+ "eval_FakePairs": -0.05351908362129137,
101
+ "eval_FakePairsNLI": 60.53,
102
+ "eval_MPQA": 77.06,
103
+ "eval_MR": 62.45,
104
+ "eval_MRPC": 70.22,
105
+ "eval_SST2": 69.04,
106
+ "eval_SUBJ": 82.31,
107
+ "eval_TREC": 63.44,
108
+ "eval_avg_sts": 0.5941874894842695,
109
+ "eval_avg_transfer": 70.92428571428572,
110
+ "eval_sickr_spearman": 0.5704407998565277,
111
+ "eval_stsb_indo": 0.30895400373418236,
112
+ "eval_stsb_spearman": 0.6179341791120112,
113
+ "step": 100
114
+ },
115
+ {
116
+ "epoch": 0.48,
117
+ "eval_CR": 72.01,
118
+ "eval_FakeCLSDev": 1.0,
119
+ "eval_FakeCLSTest": 0.47058823529411764,
120
+ "eval_FakeCLSTrain": 0.5430809399477807,
121
+ "eval_FakePairs": -0.030993140962777686,
122
+ "eval_FakePairsNLI": 60.53,
123
+ "eval_MPQA": 76.94,
124
+ "eval_MR": 61.46,
125
+ "eval_MRPC": 69.68,
126
+ "eval_SST2": 67.32,
127
+ "eval_SUBJ": 81.42,
128
+ "eval_TREC": 64.38,
129
+ "eval_avg_sts": 0.5907439625277868,
130
+ "eval_avg_transfer": 70.45857142857143,
131
+ "eval_sickr_spearman": 0.5674561472131634,
132
+ "eval_stsb_indo": 0.2897446195810389,
133
+ "eval_stsb_spearman": 0.6140317778424101,
134
+ "step": 120
135
+ },
136
+ {
137
+ "epoch": 0.56,
138
+ "eval_CR": 72.67,
139
+ "eval_FakeCLSDev": 0.9917948717948718,
140
+ "eval_FakeCLSTest": 0.5957446808510638,
141
+ "eval_FakeCLSTrain": 0.5555555555555556,
142
+ "eval_FakePairs": -0.04577079838059695,
143
+ "eval_FakePairsNLI": 63.18,
144
+ "eval_MPQA": 77.25,
145
+ "eval_MR": 61.59,
146
+ "eval_MRPC": 69.87,
147
+ "eval_SST2": 69.95,
148
+ "eval_SUBJ": 81.79,
149
+ "eval_TREC": 62.4,
150
+ "eval_avg_sts": 0.601502215900739,
151
+ "eval_avg_transfer": 70.78857142857143,
152
+ "eval_sickr_spearman": 0.5684590846422926,
153
+ "eval_stsb_indo": 0.3097638766411074,
154
+ "eval_stsb_spearman": 0.6345453471591856,
155
+ "step": 140
156
+ },
157
+ {
158
+ "epoch": 0.63,
159
+ "eval_CR": 72.47,
160
+ "eval_FakeCLSDev": 0.9908735332464146,
161
+ "eval_FakeCLSTest": 0.5632911392405063,
162
+ "eval_FakeCLSTrain": 0.556786703601108,
163
+ "eval_FakePairs": -0.05599533972914216,
164
+ "eval_FakePairsNLI": 60.69,
165
+ "eval_MPQA": 77.78,
166
+ "eval_MR": 61.95,
167
+ "eval_MRPC": 69.87,
168
+ "eval_SST2": 70.3,
169
+ "eval_SUBJ": 81.82,
170
+ "eval_TREC": 61.57,
171
+ "eval_avg_sts": 0.5932109567201365,
172
+ "eval_avg_transfer": 70.82285714285715,
173
+ "eval_sickr_spearman": 0.5648520449864932,
174
+ "eval_stsb_indo": 0.31916108185452513,
175
+ "eval_stsb_spearman": 0.6215698684537796,
176
+ "step": 160
177
+ },
178
+ {
179
+ "epoch": 0.71,
180
+ "eval_CR": 72.56,
181
+ "eval_FakeCLSDev": 0.9912587412587412,
182
+ "eval_FakeCLSTest": 0.5294117647058824,
183
+ "eval_FakeCLSTrain": 0.5282555282555282,
184
+ "eval_FakePairs": -0.06366374574055107,
185
+ "eval_FakePairsNLI": 63.96,
186
+ "eval_MPQA": 77.65,
187
+ "eval_MR": 62.88,
188
+ "eval_MRPC": 69.85,
189
+ "eval_SST2": 68.69,
190
+ "eval_SUBJ": 82.11,
191
+ "eval_TREC": 64.78,
192
+ "eval_avg_sts": 0.5965775256163806,
193
+ "eval_avg_transfer": 71.21714285714287,
194
+ "eval_sickr_spearman": 0.5614600886029337,
195
+ "eval_stsb_indo": 0.331965762222026,
196
+ "eval_stsb_spearman": 0.6316949626298273,
197
+ "step": 180
198
+ },
199
+ {
200
+ "epoch": 0.79,
201
+ "eval_CR": 72.59,
202
+ "eval_FakeCLSDev": 0.9902200488997555,
203
+ "eval_FakeCLSTest": 0.5392670157068062,
204
+ "eval_FakeCLSTrain": 0.555858310626703,
205
+ "eval_FakePairs": -0.06334422882340904,
206
+ "eval_FakePairsNLI": 60.53,
207
+ "eval_MPQA": 77.32,
208
+ "eval_MR": 62.44,
209
+ "eval_MRPC": 70.41,
210
+ "eval_SST2": 69.72,
211
+ "eval_SUBJ": 82.33,
212
+ "eval_TREC": 62.16,
213
+ "eval_avg_sts": 0.5994183846840018,
214
+ "eval_avg_transfer": 70.99571428571427,
215
+ "eval_sickr_spearman": 0.5617657105013845,
216
+ "eval_stsb_indo": 0.3362704444969455,
217
+ "eval_stsb_spearman": 0.6370710588666191,
218
+ "step": 200
219
+ },
220
+ {
221
+ "epoch": 0.87,
222
+ "eval_CR": 71.93,
223
+ "eval_FakeCLSDev": 0.9864066193853428,
224
+ "eval_FakeCLSTest": 0.44529750479846447,
225
+ "eval_FakeCLSTrain": 0.4653846153846154,
226
+ "eval_FakePairs": -0.0781218862412283,
227
+ "eval_FakePairsNLI": 60.69,
228
+ "eval_MPQA": 77.74,
229
+ "eval_MR": 62.75,
230
+ "eval_MRPC": 69.95,
231
+ "eval_SST2": 69.72,
232
+ "eval_SUBJ": 81.79,
233
+ "eval_TREC": 62.12,
234
+ "eval_avg_sts": 0.6017830377412869,
235
+ "eval_avg_transfer": 70.85714285714286,
236
+ "eval_sickr_spearman": 0.5577540568470704,
237
+ "eval_stsb_indo": 0.3330327185293657,
238
+ "eval_stsb_spearman": 0.6458120186355034,
239
+ "step": 220
240
+ },
241
+ {
242
+ "epoch": 0.95,
243
+ "eval_CR": 72.07,
244
+ "eval_FakeCLSDev": 0.9851767388825542,
245
+ "eval_FakeCLSTest": 0.4206081081081081,
246
+ "eval_FakeCLSTrain": 0.5619596541786743,
247
+ "eval_FakePairs": -0.06254543653055394,
248
+ "eval_FakePairsNLI": 62.4,
249
+ "eval_MPQA": 77.63,
250
+ "eval_MR": 62.42,
251
+ "eval_MRPC": 70.39,
252
+ "eval_SST2": 67.89,
253
+ "eval_SUBJ": 81.1,
254
+ "eval_TREC": 63.5,
255
+ "eval_avg_sts": 0.6093138534071756,
256
+ "eval_avg_transfer": 70.71428571428571,
257
+ "eval_sickr_spearman": 0.5638283100904413,
258
+ "eval_stsb_indo": 0.3179406446290444,
259
+ "eval_stsb_spearman": 0.6547993967239099,
260
+ "step": 240
261
+ },
262
+ {
263
+ "epoch": 1.03,
264
+ "eval_CR": 71.91,
265
+ "eval_FakeCLSDev": 0.9855156431054461,
266
+ "eval_FakeCLSTest": 0.4250871080139373,
267
+ "eval_FakeCLSTrain": 0.5616045845272206,
268
+ "eval_FakePairs": -0.0714120309812455,
269
+ "eval_FakePairsNLI": 62.71,
270
+ "eval_MPQA": 77.76,
271
+ "eval_MR": 62.24,
272
+ "eval_MRPC": 70.29,
273
+ "eval_SST2": 69.27,
274
+ "eval_SUBJ": 81.23,
275
+ "eval_TREC": 62.22,
276
+ "eval_avg_sts": 0.6098143668968183,
277
+ "eval_avg_transfer": 70.70285714285714,
278
+ "eval_sickr_spearman": 0.5708000244641833,
279
+ "eval_stsb_indo": 0.3233259049654498,
280
+ "eval_stsb_spearman": 0.6488287093294535,
281
+ "step": 260
282
+ },
283
+ {
284
+ "epoch": 1.11,
285
+ "eval_CR": 71.6,
286
+ "eval_FakeCLSDev": 0.9872,
287
+ "eval_FakeCLSTest": 0.4154929577464789,
288
+ "eval_FakeCLSTrain": 0.5055187637969095,
289
+ "eval_FakePairs": -0.06158689255421782,
290
+ "eval_FakePairsNLI": 61.31,
291
+ "eval_MPQA": 77.74,
292
+ "eval_MR": 62.29,
293
+ "eval_MRPC": 69.85,
294
+ "eval_SST2": 69.84,
295
+ "eval_SUBJ": 81.13,
296
+ "eval_TREC": 61.83,
297
+ "eval_avg_sts": 0.6152679773595253,
298
+ "eval_avg_transfer": 70.61142857142856,
299
+ "eval_sickr_spearman": 0.5785564709732213,
300
+ "eval_stsb_indo": 0.3262186264651655,
301
+ "eval_stsb_spearman": 0.6519794837458294,
302
+ "step": 280
303
+ },
304
+ {
305
+ "epoch": 1.19,
306
+ "eval_CR": 71.99,
307
+ "eval_FakeCLSDev": 0.9887798036465638,
308
+ "eval_FakeCLSTest": 0.3881118881118881,
309
+ "eval_FakeCLSTrain": 0.5628415300546448,
310
+ "eval_FakePairs": -0.05176174057701016,
311
+ "eval_FakePairsNLI": 63.18,
312
+ "eval_MPQA": 77.82,
313
+ "eval_MR": 62.44,
314
+ "eval_MRPC": 70.49,
315
+ "eval_SST2": 69.38,
316
+ "eval_SUBJ": 81.6,
317
+ "eval_TREC": 62.64,
318
+ "eval_avg_sts": 0.6081735394724535,
319
+ "eval_avg_transfer": 70.90857142857143,
320
+ "eval_sickr_spearman": 0.5737140233573322,
321
+ "eval_stsb_indo": 0.33079265734267466,
322
+ "eval_stsb_spearman": 0.6426330555875746,
323
+ "step": 300
324
+ },
325
+ {
326
+ "epoch": 1.27,
327
+ "eval_CR": 71.69,
328
+ "eval_FakeCLSDev": 0.9950738916256158,
329
+ "eval_FakeCLSTest": 0.32575757575757575,
330
+ "eval_FakeCLSTrain": 0.5660919540229885,
331
+ "eval_FakePairs": -0.03634504932490683,
332
+ "eval_FakePairsNLI": 61.31,
333
+ "eval_MPQA": 78.3,
334
+ "eval_MR": 61.9,
335
+ "eval_MRPC": 70.41,
336
+ "eval_SST2": 68.81,
337
+ "eval_SUBJ": 81.27,
338
+ "eval_TREC": 64.03,
339
+ "eval_avg_sts": 0.6034605442857701,
340
+ "eval_avg_transfer": 70.91571428571429,
341
+ "eval_sickr_spearman": 0.5857587346338663,
342
+ "eval_stsb_indo": 0.3175442989614004,
343
+ "eval_stsb_spearman": 0.621162353937674,
344
+ "step": 320
345
+ },
346
+ {
347
+ "epoch": 1.35,
348
+ "eval_CR": 72.16,
349
+ "eval_FakeCLSDev": 0.8333333333333334,
350
+ "eval_FakeCLSTest": 0.2903225806451613,
351
+ "eval_FakeCLSTrain": 0.5706051873198847,
352
+ "eval_FakePairs": -0.009345869826404611,
353
+ "eval_FakePairsNLI": 60.53,
354
+ "eval_MPQA": 77.78,
355
+ "eval_MR": 61.92,
356
+ "eval_MRPC": 69.87,
357
+ "eval_SST2": 68.35,
358
+ "eval_SUBJ": 81.05,
359
+ "eval_TREC": 63.74,
360
+ "eval_avg_sts": 0.5978999565018066,
361
+ "eval_avg_transfer": 70.69571428571429,
362
+ "eval_sickr_spearman": 0.5846008969025991,
363
+ "eval_stsb_indo": 0.2896616256780842,
364
+ "eval_stsb_spearman": 0.6111990161010141,
365
+ "step": 340
366
+ },
367
+ {
368
+ "epoch": 1.43,
369
+ "eval_CR": 72.4,
370
+ "eval_FakeCLSDev": 0.75,
371
+ "eval_FakeCLSTest": 0.6,
372
+ "eval_FakeCLSTrain": 0.5866261398176292,
373
+ "eval_FakePairs": 0.009186111367833594,
374
+ "eval_FakePairsNLI": 61.31,
375
+ "eval_MPQA": 77.71,
376
+ "eval_MR": 62.08,
377
+ "eval_MRPC": 70.66,
378
+ "eval_SST2": 68.92,
379
+ "eval_SUBJ": 81.17,
380
+ "eval_TREC": 62.38,
381
+ "eval_avg_sts": 0.6025418380141683,
382
+ "eval_avg_transfer": 70.76,
383
+ "eval_sickr_spearman": 0.5790834682182003,
384
+ "eval_stsb_indo": 0.2856567184812426,
385
+ "eval_stsb_spearman": 0.6260002078101363,
386
+ "step": 360
387
+ },
388
+ {
389
+ "epoch": 1.51,
390
+ "eval_CR": 71.56,
391
+ "eval_FakeCLSDev": 1.0,
392
+ "eval_FakeCLSTest": 0.4,
393
+ "eval_FakeCLSTrain": 0.5331695331695332,
394
+ "eval_FakePairs": 0.01469777818853375,
395
+ "eval_FakePairsNLI": 60.53,
396
+ "eval_MPQA": 77.58,
397
+ "eval_MR": 62.1,
398
+ "eval_MRPC": 69.77,
399
+ "eval_SST2": 69.5,
400
+ "eval_SUBJ": 80.9,
401
+ "eval_TREC": 60.66,
402
+ "eval_avg_sts": 0.6008793834242059,
403
+ "eval_avg_transfer": 70.29571428571428,
404
+ "eval_sickr_spearman": 0.5750952057102866,
405
+ "eval_stsb_indo": 0.2851042737082779,
406
+ "eval_stsb_spearman": 0.626663561138125,
407
+ "step": 380
408
+ },
409
+ {
410
+ "epoch": 1.59,
411
+ "eval_CR": 72.58,
412
+ "eval_FakeCLSDev": 0.75,
413
+ "eval_FakeCLSTest": 0.2558139534883721,
414
+ "eval_FakeCLSTrain": 0.5617977528089888,
415
+ "eval_FakePairs": -0.02212654651208613,
416
+ "eval_FakePairsNLI": 60.53,
417
+ "eval_MPQA": 77.56,
418
+ "eval_MR": 61.98,
419
+ "eval_MRPC": 70.56,
420
+ "eval_SST2": 69.5,
421
+ "eval_SUBJ": 81.99,
422
+ "eval_TREC": 60.35,
423
+ "eval_avg_sts": 0.6094671498051143,
424
+ "eval_avg_transfer": 70.64571428571429,
425
+ "eval_sickr_spearman": 0.5732523003792077,
426
+ "eval_stsb_indo": 0.29926655176105116,
427
+ "eval_stsb_spearman": 0.645681999231021,
428
+ "step": 400
429
+ },
430
+ {
431
+ "epoch": 1.67,
432
+ "eval_CR": 72.26,
433
+ "eval_FakeCLSDev": 0.75,
434
+ "eval_FakeCLSTest": 0.24489795918367346,
435
+ "eval_FakeCLSTrain": 0.5461538461538461,
436
+ "eval_FakePairs": -0.04640983221488102,
437
+ "eval_FakePairsNLI": 62.56,
438
+ "eval_MPQA": 78.1,
439
+ "eval_MR": 62.29,
440
+ "eval_MRPC": 69.97,
441
+ "eval_SST2": 70.07,
442
+ "eval_SUBJ": 82.35,
443
+ "eval_TREC": 58.6,
444
+ "eval_avg_sts": 0.6147454616026657,
445
+ "eval_avg_transfer": 70.52,
446
+ "eval_sickr_spearman": 0.5760541946815999,
447
+ "eval_stsb_indo": 0.3099846776429508,
448
+ "eval_stsb_spearman": 0.6534367285237316,
449
+ "step": 420
450
+ },
451
+ {
452
+ "epoch": 1.75,
453
+ "eval_CR": 72.46,
454
+ "eval_FakeCLSDev": 0.6666666666666666,
455
+ "eval_FakeCLSTest": 0.2564102564102564,
456
+ "eval_FakeCLSTrain": 0.5638888888888889,
457
+ "eval_FakePairs": -0.05903075044199152,
458
+ "eval_FakePairsNLI": 60.84,
459
+ "eval_MPQA": 77.68,
460
+ "eval_MR": 62.84,
461
+ "eval_MRPC": 70.51,
462
+ "eval_SST2": 70.18,
463
+ "eval_SUBJ": 82.59,
464
+ "eval_TREC": 60.55,
465
+ "eval_avg_sts": 0.6163739495549794,
466
+ "eval_avg_transfer": 70.97285714285715,
467
+ "eval_sickr_spearman": 0.5752962639009065,
468
+ "eval_stsb_indo": 0.3144443089302872,
469
+ "eval_stsb_spearman": 0.6574516352090524,
470
+ "step": 440
471
+ },
472
+ {
473
+ "epoch": 1.83,
474
+ "eval_CR": 72.66,
475
+ "eval_FakeCLSDev": 0.8,
476
+ "eval_FakeCLSTest": 0.21621621621621623,
477
+ "eval_FakeCLSTrain": 0.5635838150289018,
478
+ "eval_FakePairs": -0.06614000184840188,
479
+ "eval_FakePairsNLI": 61.62,
480
+ "eval_MPQA": 77.6,
481
+ "eval_MR": 63.18,
482
+ "eval_MRPC": 70.58,
483
+ "eval_SST2": 68.69,
484
+ "eval_SUBJ": 82.96,
485
+ "eval_TREC": 60.49,
486
+ "eval_avg_sts": 0.6191142419028671,
487
+ "eval_avg_transfer": 70.88,
488
+ "eval_sickr_spearman": 0.5765413741434867,
489
+ "eval_stsb_indo": 0.32117950245618027,
490
+ "eval_stsb_spearman": 0.6616871096622475,
491
+ "step": 460
492
+ },
493
+ {
494
+ "epoch": 1.9,
495
+ "eval_CR": 72.22,
496
+ "eval_FakeCLSDev": 0.9850746268656716,
497
+ "eval_FakeCLSTest": 0.2713178294573643,
498
+ "eval_FakeCLSTrain": 0.5596590909090909,
499
+ "eval_FakePairs": -0.06198628192555537,
500
+ "eval_FakePairsNLI": 62.4,
501
+ "eval_MPQA": 77.47,
502
+ "eval_MR": 63.01,
503
+ "eval_MRPC": 70.22,
504
+ "eval_SST2": 69.04,
505
+ "eval_SUBJ": 82.91,
506
+ "eval_TREC": 60.24,
507
+ "eval_avg_sts": 0.6214756182164681,
508
+ "eval_avg_transfer": 70.73,
509
+ "eval_sickr_spearman": 0.5775538697618021,
510
+ "eval_stsb_indo": 0.32220638413969804,
511
+ "eval_stsb_spearman": 0.6653973666711341,
512
+ "step": 480
513
+ },
514
+ {
515
+ "epoch": 1.98,
516
+ "learning_rate": 1.693121693121693e-05,
517
+ "loss": 0.039,
518
+ "step": 500
519
+ },
520
+ {
521
+ "epoch": 1.98,
522
+ "eval_CR": 72.6,
523
+ "eval_FakeCLSDev": 0.9814814814814815,
524
+ "eval_FakeCLSTest": 0.23529411764705882,
525
+ "eval_FakeCLSTrain": 0.5516304347826086,
526
+ "eval_FakePairs": -0.05535630589485809,
527
+ "eval_FakePairsNLI": 61.31,
528
+ "eval_MPQA": 77.9,
529
+ "eval_MR": 62.95,
530
+ "eval_MRPC": 70.24,
531
+ "eval_SST2": 69.27,
532
+ "eval_SUBJ": 82.79,
533
+ "eval_TREC": 61.37,
534
+ "eval_avg_sts": 0.623177260213142,
535
+ "eval_avg_transfer": 71.01714285714286,
536
+ "eval_sickr_spearman": 0.5778529113993559,
537
+ "eval_stsb_indo": 0.32376765191924134,
538
+ "eval_stsb_spearman": 0.668501609026928,
539
+ "step": 500
540
+ },
541
+ {
542
+ "epoch": 2.06,
543
+ "eval_CR": 72.46,
544
+ "eval_FakeCLSDev": 0.9794871794871794,
545
+ "eval_FakeCLSTest": 0.29891304347826086,
546
+ "eval_FakeCLSTrain": 0.5808383233532934,
547
+ "eval_FakePairs": -0.046010436068453474,
548
+ "eval_FakePairsNLI": 60.53,
549
+ "eval_MPQA": 77.7,
550
+ "eval_MR": 63.16,
551
+ "eval_MRPC": 70.07,
552
+ "eval_SST2": 69.72,
553
+ "eval_SUBJ": 82.93,
554
+ "eval_TREC": 61.23,
555
+ "eval_avg_sts": 0.6209015004562825,
556
+ "eval_avg_transfer": 71.03857142857143,
557
+ "eval_sickr_spearman": 0.5749312275299768,
558
+ "eval_stsb_indo": 0.32549874112208793,
559
+ "eval_stsb_spearman": 0.6668717733825881,
560
+ "step": 520
561
+ },
562
+ {
563
+ "epoch": 2.14,
564
+ "eval_CR": 72.09,
565
+ "eval_FakeCLSDev": 0.9809523809523809,
566
+ "eval_FakeCLSTest": 0.2903225806451613,
567
+ "eval_FakeCLSTrain": 0.5580110497237569,
568
+ "eval_FakePairs": -0.050563552137727515,
569
+ "eval_FakePairsNLI": 60.53,
570
+ "eval_MPQA": 77.54,
571
+ "eval_MR": 63.24,
572
+ "eval_MRPC": 70.68,
573
+ "eval_SST2": 70.41,
574
+ "eval_SUBJ": 83.0,
575
+ "eval_TREC": 62.14,
576
+ "eval_avg_sts": 0.6184279691881094,
577
+ "eval_avg_transfer": 71.3,
578
+ "eval_sickr_spearman": 0.5744775257457924,
579
+ "eval_stsb_indo": 0.32854134925494444,
580
+ "eval_stsb_spearman": 0.6623784126304263,
581
+ "step": 540
582
+ },
583
+ {
584
+ "epoch": 2.22,
585
+ "eval_CR": 72.34,
586
+ "eval_FakeCLSDev": 0.9838709677419355,
587
+ "eval_FakeCLSTest": 0.2857142857142857,
588
+ "eval_FakeCLSTrain": 0.5187793427230047,
589
+ "eval_FakePairs": -0.056554494334140726,
590
+ "eval_FakePairsNLI": 60.53,
591
+ "eval_MPQA": 77.73,
592
+ "eval_MR": 63.1,
593
+ "eval_MRPC": 70.66,
594
+ "eval_SST2": 70.3,
595
+ "eval_SUBJ": 83.1,
596
+ "eval_TREC": 62.11,
597
+ "eval_avg_sts": 0.6199240402637818,
598
+ "eval_avg_transfer": 71.33428571428571,
599
+ "eval_sickr_spearman": 0.5761467506140712,
600
+ "eval_stsb_indo": 0.33424233003408477,
601
+ "eval_stsb_spearman": 0.6637013299134924,
602
+ "step": 560
603
+ },
604
+ {
605
+ "epoch": 2.3,
606
+ "eval_CR": 72.49,
607
+ "eval_FakeCLSDev": 0.9806451612903225,
608
+ "eval_FakeCLSTest": 0.3,
609
+ "eval_FakeCLSTrain": 0.553133514986376,
610
+ "eval_FakePairs": -0.04856657140558978,
611
+ "eval_FakePairsNLI": 60.53,
612
+ "eval_MPQA": 77.9,
613
+ "eval_MR": 62.85,
614
+ "eval_MRPC": 70.53,
615
+ "eval_SST2": 70.07,
616
+ "eval_SUBJ": 83.09,
617
+ "eval_TREC": 62.14,
618
+ "eval_avg_sts": 0.6214529155922321,
619
+ "eval_avg_transfer": 71.2957142857143,
620
+ "eval_sickr_spearman": 0.5773829751028853,
621
+ "eval_stsb_indo": 0.3387023269485864,
622
+ "eval_stsb_spearman": 0.6655228560815791,
623
+ "step": 580
624
+ },
625
+ {
626
+ "epoch": 2.38,
627
+ "eval_CR": 72.1,
628
+ "eval_FakeCLSDev": 0.9838056680161943,
629
+ "eval_FakeCLSTest": 0.30857142857142855,
630
+ "eval_FakeCLSTrain": 0.525,
631
+ "eval_FakePairs": -0.04832693371773324,
632
+ "eval_FakePairsNLI": 60.69,
633
+ "eval_MPQA": 77.61,
634
+ "eval_MR": 63.27,
635
+ "eval_MRPC": 70.66,
636
+ "eval_SST2": 70.07,
637
+ "eval_SUBJ": 82.87,
638
+ "eval_TREC": 62.12,
639
+ "eval_avg_sts": 0.6206915906216466,
640
+ "eval_avg_transfer": 71.24285714285715,
641
+ "eval_sickr_spearman": 0.5774246660989336,
642
+ "eval_stsb_indo": 0.33551007709822284,
643
+ "eval_stsb_spearman": 0.6639585151443594,
644
+ "step": 600
645
+ },
646
+ {
647
+ "epoch": 2.46,
648
+ "eval_CR": 72.62,
649
+ "eval_FakeCLSDev": 0.9790575916230366,
650
+ "eval_FakeCLSTest": 0.3,
651
+ "eval_FakeCLSTrain": 0.5488126649076517,
652
+ "eval_FakePairs": -0.04353417996060268,
653
+ "eval_FakePairsNLI": 60.53,
654
+ "eval_MPQA": 77.77,
655
+ "eval_MR": 62.79,
656
+ "eval_MRPC": 70.09,
657
+ "eval_SST2": 70.07,
658
+ "eval_SUBJ": 82.63,
659
+ "eval_TREC": 62.73,
660
+ "eval_avg_sts": 0.6199540161828849,
661
+ "eval_avg_transfer": 71.24285714285715,
662
+ "eval_sickr_spearman": 0.5757875260064156,
663
+ "eval_stsb_indo": 0.33257213690590665,
664
+ "eval_stsb_spearman": 0.6641205063593542,
665
+ "step": 620
666
+ },
667
+ {
668
+ "epoch": 2.54,
669
+ "eval_CR": 72.58,
670
+ "eval_FakeCLSDev": 0.9788359788359788,
671
+ "eval_FakeCLSTest": 0.32,
672
+ "eval_FakeCLSTrain": 0.5694050991501416,
673
+ "eval_FakePairs": -0.03690420392990539,
674
+ "eval_FakePairsNLI": 60.53,
675
+ "eval_MPQA": 77.76,
676
+ "eval_MR": 63.21,
677
+ "eval_MRPC": 70.39,
678
+ "eval_SST2": 69.61,
679
+ "eval_SUBJ": 82.64,
680
+ "eval_TREC": 62.91,
681
+ "eval_avg_sts": 0.6197236868817497,
682
+ "eval_avg_transfer": 71.3,
683
+ "eval_sickr_spearman": 0.575295111154472,
684
+ "eval_stsb_indo": 0.33277347755938264,
685
+ "eval_stsb_spearman": 0.6641522626090275,
686
+ "step": 640
687
+ },
688
+ {
689
+ "epoch": 2.62,
690
+ "eval_CR": 72.37,
691
+ "eval_FakeCLSDev": 0.9813664596273292,
692
+ "eval_FakeCLSTest": 0.30714285714285716,
693
+ "eval_FakeCLSTrain": 0.5673352435530086,
694
+ "eval_FakePairs": -0.04185671614560698,
695
+ "eval_FakePairsNLI": 60.53,
696
+ "eval_MPQA": 77.7,
697
+ "eval_MR": 63.44,
698
+ "eval_MRPC": 70.41,
699
+ "eval_SST2": 70.07,
700
+ "eval_SUBJ": 82.42,
701
+ "eval_TREC": 62.01,
702
+ "eval_avg_sts": 0.6192603163969024,
703
+ "eval_avg_transfer": 71.20285714285714,
704
+ "eval_sickr_spearman": 0.5760667788301768,
705
+ "eval_stsb_indo": 0.3335736566547829,
706
+ "eval_stsb_spearman": 0.662453853963628,
707
+ "step": 660
708
+ },
709
+ {
710
+ "epoch": 2.7,
711
+ "eval_CR": 72.76,
712
+ "eval_FakeCLSDev": 0.9801324503311258,
713
+ "eval_FakeCLSTest": 0.31386861313868614,
714
+ "eval_FakeCLSTrain": 0.5516304347826086,
715
+ "eval_FakePairs": -0.037702996222760486,
716
+ "eval_FakePairsNLI": 60.53,
717
+ "eval_MPQA": 77.76,
718
+ "eval_MR": 63.35,
719
+ "eval_MRPC": 70.26,
720
+ "eval_SST2": 69.72,
721
+ "eval_SUBJ": 82.5,
722
+ "eval_TREC": 62.16,
723
+ "eval_avg_sts": 0.6187790665182422,
724
+ "eval_avg_transfer": 71.21571428571428,
725
+ "eval_sickr_spearman": 0.5762393065465424,
726
+ "eval_stsb_indo": 0.33152148371183193,
727
+ "eval_stsb_spearman": 0.661318826489942,
728
+ "step": 680
729
+ },
730
+ {
731
+ "epoch": 2.78,
732
+ "eval_CR": 72.7,
733
+ "eval_FakeCLSDev": 0.9854014598540146,
734
+ "eval_FakeCLSTest": 0.3153846153846154,
735
+ "eval_FakeCLSTrain": 0.5739644970414202,
736
+ "eval_FakePairs": -0.03586577394919376,
737
+ "eval_FakePairsNLI": 61.15,
738
+ "eval_MPQA": 77.57,
739
+ "eval_MR": 63.55,
740
+ "eval_MRPC": 70.49,
741
+ "eval_SST2": 70.3,
742
+ "eval_SUBJ": 82.3,
743
+ "eval_TREC": 62.05,
744
+ "eval_avg_sts": 0.6190232984727024,
745
+ "eval_avg_transfer": 71.28,
746
+ "eval_sickr_spearman": 0.5767789359711997,
747
+ "eval_stsb_indo": 0.33084746387873076,
748
+ "eval_stsb_spearman": 0.6612676609742051,
749
+ "step": 700
750
+ },
751
+ {
752
+ "epoch": 2.86,
753
+ "eval_CR": 72.72,
754
+ "eval_FakeCLSDev": 0.9819277108433735,
755
+ "eval_FakeCLSTest": 0.31690140845070425,
756
+ "eval_FakeCLSTrain": 0.5483870967741935,
757
+ "eval_FakePairs": -0.032590725548487874,
758
+ "eval_FakePairsNLI": 60.53,
759
+ "eval_MPQA": 77.62,
760
+ "eval_MR": 63.28,
761
+ "eval_MRPC": 70.22,
762
+ "eval_SST2": 70.3,
763
+ "eval_SUBJ": 82.42,
764
+ "eval_TREC": 62.27,
765
+ "eval_avg_sts": 0.619679371820255,
766
+ "eval_avg_transfer": 71.26142857142858,
767
+ "eval_sickr_spearman": 0.5773890270216666,
768
+ "eval_stsb_indo": 0.3306356462208514,
769
+ "eval_stsb_spearman": 0.6619697166188434,
770
+ "step": 720
771
+ },
772
+ {
773
+ "epoch": 2.94,
774
+ "eval_CR": 72.56,
775
+ "eval_FakeCLSDev": 0.9830508474576272,
776
+ "eval_FakeCLSTest": 0.31543624161073824,
777
+ "eval_FakeCLSTrain": 0.5803571428571429,
778
+ "eval_FakePairs": -0.03251084631920237,
779
+ "eval_FakePairsNLI": 60.53,
780
+ "eval_MPQA": 77.88,
781
+ "eval_MR": 63.26,
782
+ "eval_MRPC": 70.31,
783
+ "eval_SST2": 70.3,
784
+ "eval_SUBJ": 82.49,
785
+ "eval_TREC": 62.01,
786
+ "eval_avg_sts": 0.6197486139988778,
787
+ "eval_avg_transfer": 71.25857142857143,
788
+ "eval_sickr_spearman": 0.5774980576219311,
789
+ "eval_stsb_indo": 0.3309848917313693,
790
+ "eval_stsb_spearman": 0.6619991703758246,
791
+ "step": 740
792
+ },
793
+ {
794
+ "epoch": 3.0,
795
+ "step": 756,
796
+ "train_runtime": 51381.4965,
797
+ "train_samples_per_second": 0.015
798
+ }
799
+ ],
800
+ "max_steps": 756,
801
+ "num_train_epochs": 3,
802
+ "total_flos": 24352519106985984,
803
+ "trial_name": null,
804
+ "trial_params": null
805
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:621e596ff80e2b3728101db6c682993f8850d656cd74f64199f617edbf032813
3
+ size 2808
vocab.txt ADDED
The diff for this file is too large to render. See raw diff