AnnyNguyen commited on
Commit
b89042b
Β·
verified Β·
1 Parent(s): d274307

Upload logs/evaluation_log_20251031_142658.txt with huggingface_hub

Browse files
logs/evaluation_log_20251031_142658.txt ADDED
@@ -0,0 +1,220 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2025-10-31 14:26:58,096] [INFO] Logging to outputs/runs/bilstm/logs/evaluation_log_20251031_142658.txt
2
+ [2025-10-31 14:26:58,096] [INFO] Arguments: {'model_name': 'vinai/phobert-base', 'arch': 'bilstm', 'train_file': 'data/ViHOS_train.csv', 'valid_file': 'data/ViHOS_dev.csv', 'test_file': 'data/ViHOS_test.csv', 'output_dir': None, 'run_name': 'bilstm', 'max_length': 64, 'batch_size': 32, 'learning_rate': 5e-06, 'epochs': 100, 'early_stopping_patience': 5, 'seed': 42, 'task': 'hate-speech-span', 'push_to_hub': False, 'hub_repo': None, 'use_trainer': False}
3
+ [2025-10-31 14:26:58,108] [INFO] Using device: cuda
4
+ [2025-10-31 14:27:03,703] [INFO] Epoch: 1
5
+ [2025-10-31 14:27:08,033] [INFO] Training loss: 0.651537
6
+ [2025-10-31 14:27:08,033] [INFO] Validation loss: 0.640087
7
+ [2025-10-31 14:27:08,033] [INFO] Span F1-score: 0.516371
8
+ [2025-10-31 14:27:08,033] [INFO] Training progress β€” Best F1 so far: 0.516371
9
+ [2025-10-31 14:27:08,033] [INFO] Epoch: 2
10
+ [2025-10-31 14:27:09,994] [INFO] Training loss: 0.606350
11
+ [2025-10-31 14:27:09,994] [INFO] Validation loss: 0.594915
12
+ [2025-10-31 14:27:09,994] [INFO] Span F1-score: 0.599841
13
+ [2025-10-31 14:27:09,994] [INFO] Training progress β€” Best F1 so far: 0.599841
14
+ [2025-10-31 14:27:09,994] [INFO] Epoch: 3
15
+ [2025-10-31 14:27:11,957] [INFO] Training loss: 0.510911
16
+ [2025-10-31 14:27:11,958] [INFO] Validation loss: 0.457474
17
+ [2025-10-31 14:27:11,958] [INFO] Span F1-score: 0.720883
18
+ [2025-10-31 14:27:11,958] [INFO] Training progress β€” Best F1 so far: 0.720883
19
+ [2025-10-31 14:27:11,958] [INFO] Epoch: 4
20
+ [2025-10-31 14:27:13,914] [INFO] Training loss: 0.398343
21
+ [2025-10-31 14:27:13,914] [INFO] Validation loss: 0.431433
22
+ [2025-10-31 14:27:13,915] [INFO] Span F1-score: 0.754515
23
+ [2025-10-31 14:27:13,915] [INFO] Training progress β€” Best F1 so far: 0.754515
24
+ [2025-10-31 14:27:13,915] [INFO] Epoch: 5
25
+ [2025-10-31 14:27:15,876] [INFO] Training loss: 0.360899
26
+ [2025-10-31 14:27:15,876] [INFO] Validation loss: 0.409769
27
+ [2025-10-31 14:27:15,876] [INFO] Span F1-score: 0.788672
28
+ [2025-10-31 14:27:15,876] [INFO] Training progress β€” Best F1 so far: 0.788672
29
+ [2025-10-31 14:27:15,877] [INFO] Epoch: 6
30
+ [2025-10-31 14:27:19,724] [INFO] Training loss: 0.350446
31
+ [2025-10-31 14:27:19,724] [INFO] Validation loss: 0.406164
32
+ [2025-10-31 14:27:19,724] [INFO] Span F1-score: 0.789374
33
+ [2025-10-31 14:27:19,724] [INFO] Training progress β€” Best F1 so far: 0.789374
34
+ [2025-10-31 14:27:19,724] [INFO] Epoch: 7
35
+ [2025-10-31 14:27:21,709] [INFO] Training loss: 0.345478
36
+ [2025-10-31 14:27:21,709] [INFO] Validation loss: 0.396200
37
+ [2025-10-31 14:27:21,709] [INFO] Span F1-score: 0.795195
38
+ [2025-10-31 14:27:21,709] [INFO] Training progress β€” Best F1 so far: 0.795195
39
+ [2025-10-31 14:27:21,709] [INFO] Epoch: 8
40
+ [2025-10-31 14:27:23,689] [INFO] Training loss: 0.341237
41
+ [2025-10-31 14:27:23,689] [INFO] Validation loss: 0.392313
42
+ [2025-10-31 14:27:23,689] [INFO] Span F1-score: 0.797404
43
+ [2025-10-31 14:27:23,689] [INFO] Training progress β€” Best F1 so far: 0.797404
44
+ [2025-10-31 14:27:23,689] [INFO] Epoch: 9
45
+ [2025-10-31 14:27:25,672] [INFO] Training loss: 0.337608
46
+ [2025-10-31 14:27:25,672] [INFO] Validation loss: 0.388612
47
+ [2025-10-31 14:27:25,672] [INFO] Span F1-score: 0.799451
48
+ [2025-10-31 14:27:25,672] [INFO] Training progress β€” Best F1 so far: 0.799451
49
+ [2025-10-31 14:27:25,672] [INFO] Epoch: 10
50
+ [2025-10-31 14:27:27,657] [INFO] Training loss: 0.334552
51
+ [2025-10-31 14:27:27,657] [INFO] Validation loss: 0.386967
52
+ [2025-10-31 14:27:27,657] [INFO] Span F1-score: 0.800001
53
+ [2025-10-31 14:27:27,657] [INFO] Training progress β€” Best F1 so far: 0.800001
54
+ [2025-10-31 14:27:27,657] [INFO] Epoch: 11
55
+ [2025-10-31 14:27:29,642] [INFO] Training loss: 0.331772
56
+ [2025-10-31 14:27:29,642] [INFO] Validation loss: 0.382622
57
+ [2025-10-31 14:27:29,642] [INFO] Span F1-score: 0.803232
58
+ [2025-10-31 14:27:29,642] [INFO] Training progress β€” Best F1 so far: 0.803232
59
+ [2025-10-31 14:27:29,642] [INFO] Epoch: 12
60
+ [2025-10-31 14:27:31,808] [INFO] Training loss: 0.329042
61
+ [2025-10-31 14:27:31,808] [INFO] Validation loss: 0.378945
62
+ [2025-10-31 14:27:31,808] [INFO] Span F1-score: 0.804720
63
+ [2025-10-31 14:27:31,808] [INFO] Training progress β€” Best F1 so far: 0.804720
64
+ [2025-10-31 14:27:31,808] [INFO] Epoch: 13
65
+ [2025-10-31 14:27:33,768] [INFO] Training loss: 0.326495
66
+ [2025-10-31 14:27:33,768] [INFO] Validation loss: 0.377685
67
+ [2025-10-31 14:27:33,768] [INFO] Span F1-score: 0.803954
68
+ [2025-10-31 14:27:33,768] [INFO] Training progress β€” Best F1 so far: 0.804720
69
+ [2025-10-31 14:27:33,768] [INFO] Epoch: 14
70
+ [2025-10-31 14:27:35,723] [INFO] Training loss: 0.324020
71
+ [2025-10-31 14:27:35,723] [INFO] Validation loss: 0.374984
72
+ [2025-10-31 14:27:35,724] [INFO] Span F1-score: 0.806405
73
+ [2025-10-31 14:27:35,724] [INFO] Training progress β€” Best F1 so far: 0.806405
74
+ [2025-10-31 14:27:35,724] [INFO] Epoch: 15
75
+ [2025-10-31 14:27:37,684] [INFO] Training loss: 0.322127
76
+ [2025-10-31 14:27:37,684] [INFO] Validation loss: 0.371529
77
+ [2025-10-31 14:27:37,684] [INFO] Span F1-score: 0.809109
78
+ [2025-10-31 14:27:37,685] [INFO] Training progress β€” Best F1 so far: 0.809109
79
+ [2025-10-31 14:27:37,685] [INFO] Epoch: 16
80
+ [2025-10-31 14:27:39,644] [INFO] Training loss: 0.320118
81
+ [2025-10-31 14:27:39,644] [INFO] Validation loss: 0.372944
82
+ [2025-10-31 14:27:39,644] [INFO] Span F1-score: 0.807766
83
+ [2025-10-31 14:27:39,644] [INFO] Training progress β€” Best F1 so far: 0.809109
84
+ [2025-10-31 14:27:39,645] [INFO] Epoch: 17
85
+ [2025-10-31 14:27:43,487] [INFO] Training loss: 0.318397
86
+ [2025-10-31 14:27:43,487] [INFO] Validation loss: 0.370239
87
+ [2025-10-31 14:27:43,487] [INFO] Span F1-score: 0.809545
88
+ [2025-10-31 14:27:43,487] [INFO] Training progress β€” Best F1 so far: 0.809545
89
+ [2025-10-31 14:27:43,487] [INFO] Epoch: 18
90
+ [2025-10-31 14:27:46,149] [INFO] Training loss: 0.316116
91
+ [2025-10-31 14:27:46,150] [INFO] Validation loss: 0.368311
92
+ [2025-10-31 14:27:46,150] [INFO] Span F1-score: 0.809918
93
+ [2025-10-31 14:27:46,150] [INFO] Training progress β€” Best F1 so far: 0.809918
94
+ [2025-10-31 14:27:46,150] [INFO] Epoch: 19
95
+ [2025-10-31 14:27:49,116] [INFO] Training loss: 0.314194
96
+ [2025-10-31 14:27:49,116] [INFO] Validation loss: 0.365634
97
+ [2025-10-31 14:27:49,116] [INFO] Span F1-score: 0.812161
98
+ [2025-10-31 14:27:49,116] [INFO] Training progress β€” Best F1 so far: 0.812161
99
+ [2025-10-31 14:27:49,116] [INFO] Epoch: 20
100
+ [2025-10-31 14:27:52,312] [INFO] Training loss: 0.311269
101
+ [2025-10-31 14:27:52,312] [INFO] Validation loss: 0.370024
102
+ [2025-10-31 14:27:52,312] [INFO] Span F1-score: 0.803181
103
+ [2025-10-31 14:27:52,313] [INFO] Training progress β€” Best F1 so far: 0.812161
104
+ [2025-10-31 14:27:52,313] [INFO] Epoch: 21
105
+ [2025-10-31 14:27:54,269] [INFO] Training loss: 0.308337
106
+ [2025-10-31 14:27:54,269] [INFO] Validation loss: 0.360934
107
+ [2025-10-31 14:27:54,269] [INFO] Span F1-score: 0.812684
108
+ [2025-10-31 14:27:54,269] [INFO] Training progress β€” Best F1 so far: 0.812684
109
+ [2025-10-31 14:27:54,269] [INFO] Epoch: 22
110
+ [2025-10-31 14:27:56,228] [INFO] Training loss: 0.304993
111
+ [2025-10-31 14:27:56,228] [INFO] Validation loss: 0.360967
112
+ [2025-10-31 14:27:56,228] [INFO] Span F1-score: 0.812092
113
+ [2025-10-31 14:27:56,229] [INFO] Training progress β€” Best F1 so far: 0.812684
114
+ [2025-10-31 14:27:56,229] [INFO] Epoch: 23
115
+ [2025-10-31 14:27:58,188] [INFO] Training loss: 0.302593
116
+ [2025-10-31 14:27:58,188] [INFO] Validation loss: 0.366468
117
+ [2025-10-31 14:27:58,188] [INFO] Span F1-score: 0.808721
118
+ [2025-10-31 14:27:58,188] [INFO] Training progress β€” Best F1 so far: 0.812684
119
+ [2025-10-31 14:27:58,188] [INFO] Epoch: 24
120
+ [2025-10-31 14:28:00,147] [INFO] Training loss: 0.301002
121
+ [2025-10-31 14:28:00,147] [INFO] Validation loss: 0.359821
122
+ [2025-10-31 14:28:00,147] [INFO] Span F1-score: 0.812440
123
+ [2025-10-31 14:28:00,147] [INFO] Training progress β€” Best F1 so far: 0.812684
124
+ [2025-10-31 14:28:00,147] [INFO] Epoch: 25
125
+ [2025-10-31 14:28:02,106] [INFO] Training loss: 0.299599
126
+ [2025-10-31 14:28:02,106] [INFO] Validation loss: 0.362088
127
+ [2025-10-31 14:28:02,106] [INFO] Span F1-score: 0.812726
128
+ [2025-10-31 14:28:02,106] [INFO] Training progress β€” Best F1 so far: 0.812726
129
+ [2025-10-31 14:28:02,106] [INFO] Epoch: 26
130
+ [2025-10-31 14:28:04,062] [INFO] Training loss: 0.298221
131
+ [2025-10-31 14:28:04,062] [INFO] Validation loss: 0.360622
132
+ [2025-10-31 14:28:04,062] [INFO] Span F1-score: 0.813593
133
+ [2025-10-31 14:28:04,062] [INFO] Training progress β€” Best F1 so far: 0.813593
134
+ [2025-10-31 14:28:04,062] [INFO] Epoch: 27
135
+ [2025-10-31 14:28:06,021] [INFO] Training loss: 0.297082
136
+ [2025-10-31 14:28:06,021] [INFO] Validation loss: 0.357963
137
+ [2025-10-31 14:28:06,021] [INFO] Span F1-score: 0.812870
138
+ [2025-10-31 14:28:06,021] [INFO] Training progress β€” Best F1 so far: 0.813593
139
+ [2025-10-31 14:28:06,021] [INFO] Epoch: 28
140
+ [2025-10-31 14:28:07,980] [INFO] Training loss: 0.295896
141
+ [2025-10-31 14:28:07,980] [INFO] Validation loss: 0.356999
142
+ [2025-10-31 14:28:07,980] [INFO] Span F1-score: 0.811567
143
+ [2025-10-31 14:28:07,980] [INFO] Training progress β€” Best F1 so far: 0.813593
144
+ [2025-10-31 14:28:07,980] [INFO] Epoch: 29
145
+ [2025-10-31 14:28:09,940] [INFO] Training loss: 0.294845
146
+ [2025-10-31 14:28:09,940] [INFO] Validation loss: 0.353802
147
+ [2025-10-31 14:28:09,940] [INFO] Span F1-score: 0.812522
148
+ [2025-10-31 14:28:09,940] [INFO] Training progress β€” Best F1 so far: 0.813593
149
+ [2025-10-31 14:28:09,940] [INFO] Epoch: 30
150
+ [2025-10-31 14:28:12,142] [INFO] Training loss: 0.293840
151
+ [2025-10-31 14:28:12,142] [INFO] Validation loss: 0.352834
152
+ [2025-10-31 14:28:12,142] [INFO] Span F1-score: 0.814580
153
+ [2025-10-31 14:28:12,142] [INFO] Training progress β€” Best F1 so far: 0.814580
154
+ [2025-10-31 14:28:12,142] [INFO] Epoch: 31
155
+ [2025-10-31 14:28:17,057] [INFO] Training loss: 0.292787
156
+ [2025-10-31 14:28:17,057] [INFO] Validation loss: 0.351322
157
+ [2025-10-31 14:28:17,057] [INFO] Span F1-score: 0.813985
158
+ [2025-10-31 14:28:17,057] [INFO] Training progress β€” Best F1 so far: 0.814580
159
+ [2025-10-31 14:28:17,057] [INFO] Epoch: 32
160
+ [2025-10-31 14:28:19,846] [INFO] Training loss: 0.291731
161
+ [2025-10-31 14:28:19,846] [INFO] Validation loss: 0.352527
162
+ [2025-10-31 14:28:19,846] [INFO] Span F1-score: 0.816096
163
+ [2025-10-31 14:28:19,846] [INFO] Training progress β€” Best F1 so far: 0.816096
164
+ [2025-10-31 14:28:19,846] [INFO] Epoch: 33
165
+ [2025-10-31 14:28:21,804] [INFO] Training loss: 0.290901
166
+ [2025-10-31 14:28:21,804] [INFO] Validation loss: 0.355154
167
+ [2025-10-31 14:28:21,804] [INFO] Span F1-score: 0.813174
168
+ [2025-10-31 14:28:21,804] [INFO] Training progress β€” Best F1 so far: 0.816096
169
+ [2025-10-31 14:28:21,804] [INFO] Epoch: 34
170
+ [2025-10-31 14:28:23,761] [INFO] Training loss: 0.290108
171
+ [2025-10-31 14:28:23,761] [INFO] Validation loss: 0.351310
172
+ [2025-10-31 14:28:23,761] [INFO] Span F1-score: 0.816379
173
+ [2025-10-31 14:28:23,762] [INFO] Training progress β€” Best F1 so far: 0.816379
174
+ [2025-10-31 14:28:23,762] [INFO] Epoch: 35
175
+ [2025-10-31 14:28:25,723] [INFO] Training loss: 0.289010
176
+ [2025-10-31 14:28:25,723] [INFO] Validation loss: 0.350996
177
+ [2025-10-31 14:28:25,723] [INFO] Span F1-score: 0.814537
178
+ [2025-10-31 14:28:25,723] [INFO] Training progress β€” Best F1 so far: 0.816379
179
+ [2025-10-31 14:28:25,723] [INFO] Epoch: 36
180
+ [2025-10-31 14:28:28,048] [INFO] Training loss: 0.287984
181
+ [2025-10-31 14:28:28,048] [INFO] Validation loss: 0.347109
182
+ [2025-10-31 14:28:28,048] [INFO] Span F1-score: 0.814592
183
+ [2025-10-31 14:28:28,048] [INFO] Training progress β€” Best F1 so far: 0.816379
184
+ [2025-10-31 14:28:28,048] [INFO] Epoch: 37
185
+ [2025-10-31 14:28:30,175] [INFO] Training loss: 0.287533
186
+ [2025-10-31 14:28:30,175] [INFO] Validation loss: 0.352668
187
+ [2025-10-31 14:28:30,175] [INFO] Span F1-score: 0.811936
188
+ [2025-10-31 14:28:30,175] [INFO] Training progress β€” Best F1 so far: 0.816379
189
+ [2025-10-31 14:28:30,175] [INFO] Epoch: 38
190
+ [2025-10-31 14:28:32,151] [INFO] Training loss: 0.286601
191
+ [2025-10-31 14:28:32,151] [INFO] Validation loss: 0.346407
192
+ [2025-10-31 14:28:32,151] [INFO] Span F1-score: 0.817902
193
+ [2025-10-31 14:28:32,151] [INFO] Training progress β€” Best F1 so far: 0.817902
194
+ [2025-10-31 14:28:32,151] [INFO] Epoch: 39
195
+ [2025-10-31 14:28:34,111] [INFO] Training loss: 0.285633
196
+ [2025-10-31 14:28:34,111] [INFO] Validation loss: 0.347593
197
+ [2025-10-31 14:28:34,111] [INFO] Span F1-score: 0.815733
198
+ [2025-10-31 14:28:34,111] [INFO] Training progress β€” Best F1 so far: 0.817902
199
+ [2025-10-31 14:28:34,111] [INFO] Epoch: 40
200
+ [2025-10-31 14:28:36,073] [INFO] Training loss: 0.284899
201
+ [2025-10-31 14:28:36,073] [INFO] Validation loss: 0.347043
202
+ [2025-10-31 14:28:36,073] [INFO] Span F1-score: 0.816380
203
+ [2025-10-31 14:28:36,073] [INFO] Training progress β€” Best F1 so far: 0.817902
204
+ [2025-10-31 14:28:36,073] [INFO] Epoch: 41
205
+ [2025-10-31 14:28:38,030] [INFO] Training loss: 0.284384
206
+ [2025-10-31 14:28:38,030] [INFO] Validation loss: 0.346493
207
+ [2025-10-31 14:28:38,030] [INFO] Span F1-score: 0.815939
208
+ [2025-10-31 14:28:38,030] [INFO] Training progress β€” Best F1 so far: 0.817902
209
+ [2025-10-31 14:28:38,030] [INFO] Epoch: 42
210
+ [2025-10-31 14:28:39,991] [INFO] Training loss: 0.283211
211
+ [2025-10-31 14:28:39,991] [INFO] Validation loss: 0.346017
212
+ [2025-10-31 14:28:39,991] [INFO] Span F1-score: 0.816070
213
+ [2025-10-31 14:28:39,991] [INFO] Training progress β€” Best F1 so far: 0.817902
214
+ [2025-10-31 14:28:39,991] [INFO] Epoch: 43
215
+ [2025-10-31 14:28:43,209] [INFO] Training loss: 0.282417
216
+ [2025-10-31 14:28:43,209] [INFO] Validation loss: 0.345447
217
+ [2025-10-31 14:28:43,209] [INFO] Span F1-score: 0.813230
218
+ [2025-10-31 14:28:43,209] [INFO] Early stopping triggered after 43 epochs. Best F1: 0.8179
219
+ [2025-10-31 14:28:43,339] [INFO] Final metrics: {'f1': 0.6667522486048217, 'precision': 0.741786573692187, 'recall': 0.6362395175632489, 'accuracy': 0.8769072106690777, 'exact_match': 0.012658227848101266}
220
+ [2025-10-31 14:28:43,339] [INFO] Saving artifacts ...