Saumya-Mundra commited on
Commit
0ff762b
·
verified ·
1 Parent(s): 54b039d

End of training

Browse files
Files changed (5) hide show
  1. README.md +3 -1
  2. all_results.json +19 -0
  3. eval_results.json +14 -0
  4. train_results.json +8 -0
  5. trainer_state.json +1217 -0
README.md CHANGED
@@ -3,6 +3,8 @@ library_name: transformers
3
  license: other
4
  base_model: nvidia/mit-b0
5
  tags:
 
 
6
  - generated_from_trainer
7
  model-index:
8
  - name: segformer-finetuned-tt-225-2k
@@ -14,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
14
 
15
  # segformer-finetuned-tt-225-2k
16
 
17
- This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
  - Loss: 0.1299
20
  - Mean Iou: 0.4851
 
3
  license: other
4
  base_model: nvidia/mit-b0
5
  tags:
6
+ - image-segmentation
7
+ - vision
8
  - generated_from_trainer
9
  model-index:
10
  - name: segformer-finetuned-tt-225-2k
 
16
 
17
  # segformer-finetuned-tt-225-2k
18
 
19
+ This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the Saumya-Mundra/text255 dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.1299
22
  - Mean Iou: 0.4851
all_results.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 68.96551724137932,
3
+ "eval_accuracy_no_text": 0.9701784379830324,
4
+ "eval_accuracy_text": NaN,
5
+ "eval_iou_no_text": 0.9701784379830324,
6
+ "eval_iou_text": 0.0,
7
+ "eval_loss": 0.12989187240600586,
8
+ "eval_mean_accuracy": 0.9701784379830324,
9
+ "eval_mean_iou": 0.4850892189915162,
10
+ "eval_overall_accuracy": 0.9701784379830324,
11
+ "eval_runtime": 2.4511,
12
+ "eval_samples_per_second": 30.599,
13
+ "eval_steps_per_second": 4.08,
14
+ "total_flos": 2.7210398151632486e+17,
15
+ "train_loss": 0.11668173718452454,
16
+ "train_runtime": 724.9682,
17
+ "train_samples_per_second": 22.07,
18
+ "train_steps_per_second": 2.759
19
+ }
eval_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 68.96551724137932,
3
+ "eval_accuracy_no_text": 0.9701784379830324,
4
+ "eval_accuracy_text": NaN,
5
+ "eval_iou_no_text": 0.9701784379830324,
6
+ "eval_iou_text": 0.0,
7
+ "eval_loss": 0.12989187240600586,
8
+ "eval_mean_accuracy": 0.9701784379830324,
9
+ "eval_mean_iou": 0.4850892189915162,
10
+ "eval_overall_accuracy": 0.9701784379830324,
11
+ "eval_runtime": 2.4511,
12
+ "eval_samples_per_second": 30.599,
13
+ "eval_steps_per_second": 4.08
14
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 68.96551724137932,
3
+ "total_flos": 2.7210398151632486e+17,
4
+ "train_loss": 0.11668173718452454,
5
+ "train_runtime": 724.9682,
6
+ "train_samples_per_second": 22.07,
7
+ "train_steps_per_second": 2.759
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 68.96551724137932,
5
+ "eval_steps": 500,
6
+ "global_step": 2000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 1.0,
13
+ "eval_accuracy_no_text": 0.9411412367275587,
14
+ "eval_accuracy_text": NaN,
15
+ "eval_iou_no_text": 0.9411412367275587,
16
+ "eval_iou_text": 0.0,
17
+ "eval_loss": 0.26036718487739563,
18
+ "eval_mean_accuracy": 0.9411412367275587,
19
+ "eval_mean_iou": 0.47057061836377934,
20
+ "eval_overall_accuracy": 0.9411412367275587,
21
+ "eval_runtime": 2.53,
22
+ "eval_samples_per_second": 29.644,
23
+ "eval_steps_per_second": 3.953,
24
+ "step": 29
25
+ },
26
+ {
27
+ "epoch": 2.0,
28
+ "eval_accuracy_no_text": 0.9556761981853922,
29
+ "eval_accuracy_text": NaN,
30
+ "eval_iou_no_text": 0.9556761981853922,
31
+ "eval_iou_text": 0.0,
32
+ "eval_loss": 0.20939157903194427,
33
+ "eval_mean_accuracy": 0.9556761981853922,
34
+ "eval_mean_iou": 0.4778380990926961,
35
+ "eval_overall_accuracy": 0.9556761981853922,
36
+ "eval_runtime": 2.5445,
37
+ "eval_samples_per_second": 29.475,
38
+ "eval_steps_per_second": 3.93,
39
+ "step": 58
40
+ },
41
+ {
42
+ "epoch": 3.0,
43
+ "eval_accuracy_no_text": 0.9646688438230138,
44
+ "eval_accuracy_text": NaN,
45
+ "eval_iou_no_text": 0.9646688438230138,
46
+ "eval_iou_text": 0.0,
47
+ "eval_loss": 0.18179728090763092,
48
+ "eval_mean_accuracy": 0.9646688438230138,
49
+ "eval_mean_iou": 0.4823344219115069,
50
+ "eval_overall_accuracy": 0.9646688438230138,
51
+ "eval_runtime": 2.5227,
52
+ "eval_samples_per_second": 29.73,
53
+ "eval_steps_per_second": 3.964,
54
+ "step": 87
55
+ },
56
+ {
57
+ "epoch": 3.4482758620689653,
58
+ "grad_norm": 69.57986450195312,
59
+ "learning_rate": 5.7005e-05,
60
+ "loss": 0.323,
61
+ "step": 100
62
+ },
63
+ {
64
+ "epoch": 4.0,
65
+ "eval_accuracy_no_text": 0.9392420264366191,
66
+ "eval_accuracy_text": NaN,
67
+ "eval_iou_no_text": 0.9392420264366191,
68
+ "eval_iou_text": 0.0,
69
+ "eval_loss": 0.20981624722480774,
70
+ "eval_mean_accuracy": 0.9392420264366191,
71
+ "eval_mean_iou": 0.46962101321830957,
72
+ "eval_overall_accuracy": 0.9392420264366191,
73
+ "eval_runtime": 2.5287,
74
+ "eval_samples_per_second": 29.66,
75
+ "eval_steps_per_second": 3.955,
76
+ "step": 116
77
+ },
78
+ {
79
+ "epoch": 5.0,
80
+ "eval_accuracy_no_text": 0.9698847328744703,
81
+ "eval_accuracy_text": NaN,
82
+ "eval_iou_no_text": 0.9698847328744703,
83
+ "eval_iou_text": 0.0,
84
+ "eval_loss": 0.17170381546020508,
85
+ "eval_mean_accuracy": 0.9698847328744703,
86
+ "eval_mean_iou": 0.48494236643723515,
87
+ "eval_overall_accuracy": 0.9698847328744703,
88
+ "eval_runtime": 2.4809,
89
+ "eval_samples_per_second": 30.231,
90
+ "eval_steps_per_second": 4.031,
91
+ "step": 145
92
+ },
93
+ {
94
+ "epoch": 6.0,
95
+ "eval_accuracy_no_text": 0.9680704296500666,
96
+ "eval_accuracy_text": NaN,
97
+ "eval_iou_no_text": 0.9680704296500666,
98
+ "eval_iou_text": 0.0,
99
+ "eval_loss": 0.14835301041603088,
100
+ "eval_mean_accuracy": 0.9680704296500666,
101
+ "eval_mean_iou": 0.4840352148250333,
102
+ "eval_overall_accuracy": 0.9680704296500666,
103
+ "eval_runtime": 2.518,
104
+ "eval_samples_per_second": 29.786,
105
+ "eval_steps_per_second": 3.971,
106
+ "step": 174
107
+ },
108
+ {
109
+ "epoch": 6.896551724137931,
110
+ "grad_norm": 2.1472854614257812,
111
+ "learning_rate": 5.401e-05,
112
+ "loss": 0.2027,
113
+ "step": 200
114
+ },
115
+ {
116
+ "epoch": 7.0,
117
+ "eval_accuracy_no_text": 0.9630674570564881,
118
+ "eval_accuracy_text": NaN,
119
+ "eval_iou_no_text": 0.9630674570564881,
120
+ "eval_iou_text": 0.0,
121
+ "eval_loss": 0.16156023740768433,
122
+ "eval_mean_accuracy": 0.9630674570564881,
123
+ "eval_mean_iou": 0.48153372852824405,
124
+ "eval_overall_accuracy": 0.9630674570564881,
125
+ "eval_runtime": 2.5744,
126
+ "eval_samples_per_second": 29.133,
127
+ "eval_steps_per_second": 3.884,
128
+ "step": 203
129
+ },
130
+ {
131
+ "epoch": 8.0,
132
+ "eval_accuracy_no_text": 0.9727854516339448,
133
+ "eval_accuracy_text": NaN,
134
+ "eval_iou_no_text": 0.9727854516339448,
135
+ "eval_iou_text": 0.0,
136
+ "eval_loss": 0.15030205249786377,
137
+ "eval_mean_accuracy": 0.9727854516339448,
138
+ "eval_mean_iou": 0.4863927258169724,
139
+ "eval_overall_accuracy": 0.9727854516339448,
140
+ "eval_runtime": 2.6278,
141
+ "eval_samples_per_second": 28.541,
142
+ "eval_steps_per_second": 3.806,
143
+ "step": 232
144
+ },
145
+ {
146
+ "epoch": 9.0,
147
+ "eval_accuracy_no_text": 0.9566264239140725,
148
+ "eval_accuracy_text": NaN,
149
+ "eval_iou_no_text": 0.9566264239140725,
150
+ "eval_iou_text": 0.0,
151
+ "eval_loss": 0.16279229521751404,
152
+ "eval_mean_accuracy": 0.9566264239140725,
153
+ "eval_mean_iou": 0.47831321195703624,
154
+ "eval_overall_accuracy": 0.9566264239140725,
155
+ "eval_runtime": 2.4976,
156
+ "eval_samples_per_second": 30.029,
157
+ "eval_steps_per_second": 4.004,
158
+ "step": 261
159
+ },
160
+ {
161
+ "epoch": 10.0,
162
+ "eval_accuracy_no_text": 0.9670189360247539,
163
+ "eval_accuracy_text": NaN,
164
+ "eval_iou_no_text": 0.9670189360247539,
165
+ "eval_iou_text": 0.0,
166
+ "eval_loss": 0.142441064119339,
167
+ "eval_mean_accuracy": 0.9670189360247539,
168
+ "eval_mean_iou": 0.48350946801237693,
169
+ "eval_overall_accuracy": 0.9670189360247539,
170
+ "eval_runtime": 2.5057,
171
+ "eval_samples_per_second": 29.931,
172
+ "eval_steps_per_second": 3.991,
173
+ "step": 290
174
+ },
175
+ {
176
+ "epoch": 10.344827586206897,
177
+ "grad_norm": 0.8168253302574158,
178
+ "learning_rate": 5.1015e-05,
179
+ "loss": 0.1693,
180
+ "step": 300
181
+ },
182
+ {
183
+ "epoch": 11.0,
184
+ "eval_accuracy_no_text": 0.9806479024118233,
185
+ "eval_accuracy_text": NaN,
186
+ "eval_iou_no_text": 0.9806479024118233,
187
+ "eval_iou_text": 0.0,
188
+ "eval_loss": 0.12359502166509628,
189
+ "eval_mean_accuracy": 0.9806479024118233,
190
+ "eval_mean_iou": 0.49032395120591166,
191
+ "eval_overall_accuracy": 0.9806479024118233,
192
+ "eval_runtime": 2.5703,
193
+ "eval_samples_per_second": 29.18,
194
+ "eval_steps_per_second": 3.891,
195
+ "step": 319
196
+ },
197
+ {
198
+ "epoch": 12.0,
199
+ "eval_accuracy_no_text": 0.9632355222732392,
200
+ "eval_accuracy_text": NaN,
201
+ "eval_iou_no_text": 0.9632355222732392,
202
+ "eval_iou_text": 0.0,
203
+ "eval_loss": 0.13876928389072418,
204
+ "eval_mean_accuracy": 0.9632355222732392,
205
+ "eval_mean_iou": 0.4816177611366196,
206
+ "eval_overall_accuracy": 0.9632355222732392,
207
+ "eval_runtime": 2.4816,
208
+ "eval_samples_per_second": 30.223,
209
+ "eval_steps_per_second": 4.03,
210
+ "step": 348
211
+ },
212
+ {
213
+ "epoch": 13.0,
214
+ "eval_accuracy_no_text": 0.9576819795385817,
215
+ "eval_accuracy_text": NaN,
216
+ "eval_iou_no_text": 0.9576819795385817,
217
+ "eval_iou_text": 0.0,
218
+ "eval_loss": 0.1554112434387207,
219
+ "eval_mean_accuracy": 0.9576819795385817,
220
+ "eval_mean_iou": 0.47884098976929085,
221
+ "eval_overall_accuracy": 0.9576819795385817,
222
+ "eval_runtime": 2.4981,
223
+ "eval_samples_per_second": 30.023,
224
+ "eval_steps_per_second": 4.003,
225
+ "step": 377
226
+ },
227
+ {
228
+ "epoch": 13.793103448275861,
229
+ "grad_norm": 3.8616011142730713,
230
+ "learning_rate": 4.8020000000000004e-05,
231
+ "loss": 0.1502,
232
+ "step": 400
233
+ },
234
+ {
235
+ "epoch": 14.0,
236
+ "eval_accuracy_no_text": 0.9521418075512791,
237
+ "eval_accuracy_text": NaN,
238
+ "eval_iou_no_text": 0.9521418075512791,
239
+ "eval_iou_text": 0.0,
240
+ "eval_loss": 0.15367722511291504,
241
+ "eval_mean_accuracy": 0.9521418075512791,
242
+ "eval_mean_iou": 0.47607090377563954,
243
+ "eval_overall_accuracy": 0.9521418075512791,
244
+ "eval_runtime": 2.5218,
245
+ "eval_samples_per_second": 29.741,
246
+ "eval_steps_per_second": 3.965,
247
+ "step": 406
248
+ },
249
+ {
250
+ "epoch": 15.0,
251
+ "eval_accuracy_no_text": 0.9562213523275425,
252
+ "eval_accuracy_text": NaN,
253
+ "eval_iou_no_text": 0.9562213523275425,
254
+ "eval_iou_text": 0.0,
255
+ "eval_loss": 0.14778192341327667,
256
+ "eval_mean_accuracy": 0.9562213523275425,
257
+ "eval_mean_iou": 0.47811067616377123,
258
+ "eval_overall_accuracy": 0.9562213523275425,
259
+ "eval_runtime": 2.4931,
260
+ "eval_samples_per_second": 30.083,
261
+ "eval_steps_per_second": 4.011,
262
+ "step": 435
263
+ },
264
+ {
265
+ "epoch": 16.0,
266
+ "eval_accuracy_no_text": 0.9650662991610505,
267
+ "eval_accuracy_text": NaN,
268
+ "eval_iou_no_text": 0.9650662991610505,
269
+ "eval_iou_text": 0.0,
270
+ "eval_loss": 0.1367165446281433,
271
+ "eval_mean_accuracy": 0.9650662991610505,
272
+ "eval_mean_iou": 0.48253314958052523,
273
+ "eval_overall_accuracy": 0.9650662991610505,
274
+ "eval_runtime": 2.5189,
275
+ "eval_samples_per_second": 29.774,
276
+ "eval_steps_per_second": 3.97,
277
+ "step": 464
278
+ },
279
+ {
280
+ "epoch": 17.0,
281
+ "eval_accuracy_no_text": 0.9702622167164582,
282
+ "eval_accuracy_text": NaN,
283
+ "eval_iou_no_text": 0.9702622167164582,
284
+ "eval_iou_text": 0.0,
285
+ "eval_loss": 0.13298067450523376,
286
+ "eval_mean_accuracy": 0.9702622167164582,
287
+ "eval_mean_iou": 0.4851311083582291,
288
+ "eval_overall_accuracy": 0.9702622167164582,
289
+ "eval_runtime": 2.4829,
290
+ "eval_samples_per_second": 30.207,
291
+ "eval_steps_per_second": 4.028,
292
+ "step": 493
293
+ },
294
+ {
295
+ "epoch": 17.24137931034483,
296
+ "grad_norm": 0.5257724523544312,
297
+ "learning_rate": 4.5025000000000003e-05,
298
+ "loss": 0.127,
299
+ "step": 500
300
+ },
301
+ {
302
+ "epoch": 18.0,
303
+ "eval_accuracy_no_text": 0.9695612961884569,
304
+ "eval_accuracy_text": NaN,
305
+ "eval_iou_no_text": 0.9695612961884569,
306
+ "eval_iou_text": 0.0,
307
+ "eval_loss": 0.13220080733299255,
308
+ "eval_mean_accuracy": 0.9695612961884569,
309
+ "eval_mean_iou": 0.48478064809422844,
310
+ "eval_overall_accuracy": 0.9695612961884569,
311
+ "eval_runtime": 2.4852,
312
+ "eval_samples_per_second": 30.179,
313
+ "eval_steps_per_second": 4.024,
314
+ "step": 522
315
+ },
316
+ {
317
+ "epoch": 19.0,
318
+ "eval_accuracy_no_text": 0.9684219618305219,
319
+ "eval_accuracy_text": NaN,
320
+ "eval_iou_no_text": 0.9684219618305219,
321
+ "eval_iou_text": 0.0,
322
+ "eval_loss": 0.13408276438713074,
323
+ "eval_mean_accuracy": 0.9684219618305219,
324
+ "eval_mean_iou": 0.48421098091526094,
325
+ "eval_overall_accuracy": 0.9684219618305219,
326
+ "eval_runtime": 2.5013,
327
+ "eval_samples_per_second": 29.984,
328
+ "eval_steps_per_second": 3.998,
329
+ "step": 551
330
+ },
331
+ {
332
+ "epoch": 20.0,
333
+ "eval_accuracy_no_text": 0.9764442974101032,
334
+ "eval_accuracy_text": NaN,
335
+ "eval_iou_no_text": 0.9764442974101032,
336
+ "eval_iou_text": 0.0,
337
+ "eval_loss": 0.12438876926898956,
338
+ "eval_mean_accuracy": 0.9764442974101032,
339
+ "eval_mean_iou": 0.4882221487050516,
340
+ "eval_overall_accuracy": 0.9764442974101032,
341
+ "eval_runtime": 2.4814,
342
+ "eval_samples_per_second": 30.225,
343
+ "eval_steps_per_second": 4.03,
344
+ "step": 580
345
+ },
346
+ {
347
+ "epoch": 20.689655172413794,
348
+ "grad_norm": 0.9491066336631775,
349
+ "learning_rate": 4.203e-05,
350
+ "loss": 0.1173,
351
+ "step": 600
352
+ },
353
+ {
354
+ "epoch": 21.0,
355
+ "eval_accuracy_no_text": 0.9792908561802923,
356
+ "eval_accuracy_text": NaN,
357
+ "eval_iou_no_text": 0.9792908561802923,
358
+ "eval_iou_text": 0.0,
359
+ "eval_loss": 0.1200159564614296,
360
+ "eval_mean_accuracy": 0.9792908561802923,
361
+ "eval_mean_iou": 0.48964542809014616,
362
+ "eval_overall_accuracy": 0.9792908561802923,
363
+ "eval_runtime": 2.4555,
364
+ "eval_samples_per_second": 30.544,
365
+ "eval_steps_per_second": 4.072,
366
+ "step": 609
367
+ },
368
+ {
369
+ "epoch": 22.0,
370
+ "eval_accuracy_no_text": 0.9652650550383967,
371
+ "eval_accuracy_text": NaN,
372
+ "eval_iou_no_text": 0.9652650550383967,
373
+ "eval_iou_text": 0.0,
374
+ "eval_loss": 0.13735905289649963,
375
+ "eval_mean_accuracy": 0.9652650550383967,
376
+ "eval_mean_iou": 0.48263252751919833,
377
+ "eval_overall_accuracy": 0.9652650550383967,
378
+ "eval_runtime": 2.4618,
379
+ "eval_samples_per_second": 30.465,
380
+ "eval_steps_per_second": 4.062,
381
+ "step": 638
382
+ },
383
+ {
384
+ "epoch": 23.0,
385
+ "eval_accuracy_no_text": 0.9768179449195166,
386
+ "eval_accuracy_text": NaN,
387
+ "eval_iou_no_text": 0.9768179449195166,
388
+ "eval_iou_text": 0.0,
389
+ "eval_loss": 0.1247735321521759,
390
+ "eval_mean_accuracy": 0.9768179449195166,
391
+ "eval_mean_iou": 0.4884089724597583,
392
+ "eval_overall_accuracy": 0.9768179449195166,
393
+ "eval_runtime": 2.5434,
394
+ "eval_samples_per_second": 29.489,
395
+ "eval_steps_per_second": 3.932,
396
+ "step": 667
397
+ },
398
+ {
399
+ "epoch": 24.0,
400
+ "eval_accuracy_no_text": 0.9713726093301188,
401
+ "eval_accuracy_text": NaN,
402
+ "eval_iou_no_text": 0.9713726093301188,
403
+ "eval_iou_text": 0.0,
404
+ "eval_loss": 0.12621533870697021,
405
+ "eval_mean_accuracy": 0.9713726093301188,
406
+ "eval_mean_iou": 0.4856863046650594,
407
+ "eval_overall_accuracy": 0.9713726093301188,
408
+ "eval_runtime": 2.5006,
409
+ "eval_samples_per_second": 29.993,
410
+ "eval_steps_per_second": 3.999,
411
+ "step": 696
412
+ },
413
+ {
414
+ "epoch": 24.137931034482758,
415
+ "grad_norm": 1.288313627243042,
416
+ "learning_rate": 3.9035e-05,
417
+ "loss": 0.1193,
418
+ "step": 700
419
+ },
420
+ {
421
+ "epoch": 25.0,
422
+ "eval_accuracy_no_text": 0.9719735595445641,
423
+ "eval_accuracy_text": NaN,
424
+ "eval_iou_no_text": 0.9719735595445641,
425
+ "eval_iou_text": 0.0,
426
+ "eval_loss": 0.12352190911769867,
427
+ "eval_mean_accuracy": 0.9719735595445641,
428
+ "eval_mean_iou": 0.48598677977228205,
429
+ "eval_overall_accuracy": 0.9719735595445641,
430
+ "eval_runtime": 2.4888,
431
+ "eval_samples_per_second": 30.135,
432
+ "eval_steps_per_second": 4.018,
433
+ "step": 725
434
+ },
435
+ {
436
+ "epoch": 26.0,
437
+ "eval_accuracy_no_text": 0.9676620294808619,
438
+ "eval_accuracy_text": NaN,
439
+ "eval_iou_no_text": 0.9676620294808619,
440
+ "eval_iou_text": 0.0,
441
+ "eval_loss": 0.1323447823524475,
442
+ "eval_mean_accuracy": 0.9676620294808619,
443
+ "eval_mean_iou": 0.48383101474043094,
444
+ "eval_overall_accuracy": 0.9676620294808619,
445
+ "eval_runtime": 2.5267,
446
+ "eval_samples_per_second": 29.683,
447
+ "eval_steps_per_second": 3.958,
448
+ "step": 754
449
+ },
450
+ {
451
+ "epoch": 27.0,
452
+ "eval_accuracy_no_text": 0.970691829548133,
453
+ "eval_accuracy_text": NaN,
454
+ "eval_iou_no_text": 0.970691829548133,
455
+ "eval_iou_text": 0.0,
456
+ "eval_loss": 0.12354418635368347,
457
+ "eval_mean_accuracy": 0.970691829548133,
458
+ "eval_mean_iou": 0.4853459147740665,
459
+ "eval_overall_accuracy": 0.970691829548133,
460
+ "eval_runtime": 2.5317,
461
+ "eval_samples_per_second": 29.625,
462
+ "eval_steps_per_second": 3.95,
463
+ "step": 783
464
+ },
465
+ {
466
+ "epoch": 27.586206896551722,
467
+ "grad_norm": 0.8687916398048401,
468
+ "learning_rate": 3.604e-05,
469
+ "loss": 0.0912,
470
+ "step": 800
471
+ },
472
+ {
473
+ "epoch": 28.0,
474
+ "eval_accuracy_no_text": 0.9631846908666286,
475
+ "eval_accuracy_text": NaN,
476
+ "eval_iou_no_text": 0.9631846908666286,
477
+ "eval_iou_text": 0.0,
478
+ "eval_loss": 0.13487361371517181,
479
+ "eval_mean_accuracy": 0.9631846908666286,
480
+ "eval_mean_iou": 0.4815923454333143,
481
+ "eval_overall_accuracy": 0.9631846908666286,
482
+ "eval_runtime": 2.5216,
483
+ "eval_samples_per_second": 29.743,
484
+ "eval_steps_per_second": 3.966,
485
+ "step": 812
486
+ },
487
+ {
488
+ "epoch": 29.0,
489
+ "eval_accuracy_no_text": 0.9631912351986673,
490
+ "eval_accuracy_text": NaN,
491
+ "eval_iou_no_text": 0.9631912351986673,
492
+ "eval_iou_text": 0.0,
493
+ "eval_loss": 0.140753835439682,
494
+ "eval_mean_accuracy": 0.9631912351986673,
495
+ "eval_mean_iou": 0.48159561759933367,
496
+ "eval_overall_accuracy": 0.9631912351986673,
497
+ "eval_runtime": 2.4836,
498
+ "eval_samples_per_second": 30.198,
499
+ "eval_steps_per_second": 4.026,
500
+ "step": 841
501
+ },
502
+ {
503
+ "epoch": 30.0,
504
+ "eval_accuracy_no_text": 0.9753996302001064,
505
+ "eval_accuracy_text": NaN,
506
+ "eval_iou_no_text": 0.9753996302001064,
507
+ "eval_iou_text": 0.0,
508
+ "eval_loss": 0.1205972284078598,
509
+ "eval_mean_accuracy": 0.9753996302001064,
510
+ "eval_mean_iou": 0.4876998151000532,
511
+ "eval_overall_accuracy": 0.9753996302001064,
512
+ "eval_runtime": 2.4858,
513
+ "eval_samples_per_second": 30.171,
514
+ "eval_steps_per_second": 4.023,
515
+ "step": 870
516
+ },
517
+ {
518
+ "epoch": 31.0,
519
+ "eval_accuracy_no_text": 0.9646436819946582,
520
+ "eval_accuracy_text": NaN,
521
+ "eval_iou_no_text": 0.9646436819946582,
522
+ "eval_iou_text": 0.0,
523
+ "eval_loss": 0.13470463454723358,
524
+ "eval_mean_accuracy": 0.9646436819946582,
525
+ "eval_mean_iou": 0.4823218409973291,
526
+ "eval_overall_accuracy": 0.9646436819946582,
527
+ "eval_runtime": 2.4993,
528
+ "eval_samples_per_second": 30.009,
529
+ "eval_steps_per_second": 4.001,
530
+ "step": 899
531
+ },
532
+ {
533
+ "epoch": 31.03448275862069,
534
+ "grad_norm": 1.7799224853515625,
535
+ "learning_rate": 3.3045000000000006e-05,
536
+ "loss": 0.1005,
537
+ "step": 900
538
+ },
539
+ {
540
+ "epoch": 32.0,
541
+ "eval_accuracy_no_text": 0.9757926850390138,
542
+ "eval_accuracy_text": NaN,
543
+ "eval_iou_no_text": 0.9757926850390138,
544
+ "eval_iou_text": 0.0,
545
+ "eval_loss": 0.11876140534877777,
546
+ "eval_mean_accuracy": 0.9757926850390138,
547
+ "eval_mean_iou": 0.4878963425195069,
548
+ "eval_overall_accuracy": 0.9757926850390138,
549
+ "eval_runtime": 2.4793,
550
+ "eval_samples_per_second": 30.25,
551
+ "eval_steps_per_second": 4.033,
552
+ "step": 928
553
+ },
554
+ {
555
+ "epoch": 33.0,
556
+ "eval_accuracy_no_text": 0.9781089272474506,
557
+ "eval_accuracy_text": NaN,
558
+ "eval_iou_no_text": 0.9781089272474506,
559
+ "eval_iou_text": 0.0,
560
+ "eval_loss": 0.12336363643407822,
561
+ "eval_mean_accuracy": 0.9781089272474506,
562
+ "eval_mean_iou": 0.4890544636237253,
563
+ "eval_overall_accuracy": 0.9781089272474506,
564
+ "eval_runtime": 2.4798,
565
+ "eval_samples_per_second": 30.245,
566
+ "eval_steps_per_second": 4.033,
567
+ "step": 957
568
+ },
569
+ {
570
+ "epoch": 34.0,
571
+ "eval_accuracy_no_text": 0.9848704069931378,
572
+ "eval_accuracy_text": NaN,
573
+ "eval_iou_no_text": 0.9848704069931378,
574
+ "eval_iou_text": 0.0,
575
+ "eval_loss": 0.12059501558542252,
576
+ "eval_mean_accuracy": 0.9848704069931378,
577
+ "eval_mean_iou": 0.4924352034965689,
578
+ "eval_overall_accuracy": 0.9848704069931378,
579
+ "eval_runtime": 2.5063,
580
+ "eval_samples_per_second": 29.925,
581
+ "eval_steps_per_second": 3.99,
582
+ "step": 986
583
+ },
584
+ {
585
+ "epoch": 34.48275862068966,
586
+ "grad_norm": 1.1413416862487793,
587
+ "learning_rate": 3.0050000000000002e-05,
588
+ "loss": 0.0933,
589
+ "step": 1000
590
+ },
591
+ {
592
+ "epoch": 35.0,
593
+ "eval_accuracy_no_text": 0.9789616085787618,
594
+ "eval_accuracy_text": NaN,
595
+ "eval_iou_no_text": 0.9789616085787618,
596
+ "eval_iou_text": 0.0,
597
+ "eval_loss": 0.12342917174100876,
598
+ "eval_mean_accuracy": 0.9789616085787618,
599
+ "eval_mean_iou": 0.4894808042893809,
600
+ "eval_overall_accuracy": 0.9789616085787618,
601
+ "eval_runtime": 2.4727,
602
+ "eval_samples_per_second": 30.332,
603
+ "eval_steps_per_second": 4.044,
604
+ "step": 1015
605
+ },
606
+ {
607
+ "epoch": 36.0,
608
+ "eval_accuracy_no_text": 0.9753304069638011,
609
+ "eval_accuracy_text": NaN,
610
+ "eval_iou_no_text": 0.9753304069638011,
611
+ "eval_iou_text": 0.0,
612
+ "eval_loss": 0.1192336305975914,
613
+ "eval_mean_accuracy": 0.9753304069638011,
614
+ "eval_mean_iou": 0.48766520348190057,
615
+ "eval_overall_accuracy": 0.9753304069638011,
616
+ "eval_runtime": 2.4878,
617
+ "eval_samples_per_second": 30.147,
618
+ "eval_steps_per_second": 4.02,
619
+ "step": 1044
620
+ },
621
+ {
622
+ "epoch": 37.0,
623
+ "eval_accuracy_no_text": 0.9614615005408664,
624
+ "eval_accuracy_text": NaN,
625
+ "eval_iou_no_text": 0.9614615005408664,
626
+ "eval_iou_text": 0.0,
627
+ "eval_loss": 0.13877929747104645,
628
+ "eval_mean_accuracy": 0.9614615005408664,
629
+ "eval_mean_iou": 0.4807307502704332,
630
+ "eval_overall_accuracy": 0.9614615005408664,
631
+ "eval_runtime": 2.4791,
632
+ "eval_samples_per_second": 30.253,
633
+ "eval_steps_per_second": 4.034,
634
+ "step": 1073
635
+ },
636
+ {
637
+ "epoch": 37.93103448275862,
638
+ "grad_norm": 0.8322708606719971,
639
+ "learning_rate": 2.7054999999999998e-05,
640
+ "loss": 0.0937,
641
+ "step": 1100
642
+ },
643
+ {
644
+ "epoch": 38.0,
645
+ "eval_accuracy_no_text": 0.9719491311327302,
646
+ "eval_accuracy_text": NaN,
647
+ "eval_iou_no_text": 0.9719491311327302,
648
+ "eval_iou_text": 0.0,
649
+ "eval_loss": 0.12549874186515808,
650
+ "eval_mean_accuracy": 0.9719491311327302,
651
+ "eval_mean_iou": 0.4859745655663651,
652
+ "eval_overall_accuracy": 0.9719491311327302,
653
+ "eval_runtime": 2.5079,
654
+ "eval_samples_per_second": 29.905,
655
+ "eval_steps_per_second": 3.987,
656
+ "step": 1102
657
+ },
658
+ {
659
+ "epoch": 39.0,
660
+ "eval_accuracy_no_text": 0.9754536209394253,
661
+ "eval_accuracy_text": NaN,
662
+ "eval_iou_no_text": 0.9754536209394253,
663
+ "eval_iou_text": 0.0,
664
+ "eval_loss": 0.11986999213695526,
665
+ "eval_mean_accuracy": 0.9754536209394253,
666
+ "eval_mean_iou": 0.48772681046971267,
667
+ "eval_overall_accuracy": 0.9754536209394253,
668
+ "eval_runtime": 2.5348,
669
+ "eval_samples_per_second": 29.588,
670
+ "eval_steps_per_second": 3.945,
671
+ "step": 1131
672
+ },
673
+ {
674
+ "epoch": 40.0,
675
+ "eval_accuracy_no_text": 0.9798514684860512,
676
+ "eval_accuracy_text": NaN,
677
+ "eval_iou_no_text": 0.9798514684860512,
678
+ "eval_iou_text": 0.0,
679
+ "eval_loss": 0.11944077908992767,
680
+ "eval_mean_accuracy": 0.9798514684860512,
681
+ "eval_mean_iou": 0.4899257342430256,
682
+ "eval_overall_accuracy": 0.9798514684860512,
683
+ "eval_runtime": 2.5215,
684
+ "eval_samples_per_second": 29.744,
685
+ "eval_steps_per_second": 3.966,
686
+ "step": 1160
687
+ },
688
+ {
689
+ "epoch": 41.0,
690
+ "eval_accuracy_no_text": 0.9687282478532616,
691
+ "eval_accuracy_text": NaN,
692
+ "eval_iou_no_text": 0.9687282478532616,
693
+ "eval_iou_text": 0.0,
694
+ "eval_loss": 0.12739954888820648,
695
+ "eval_mean_accuracy": 0.9687282478532616,
696
+ "eval_mean_iou": 0.4843641239266308,
697
+ "eval_overall_accuracy": 0.9687282478532616,
698
+ "eval_runtime": 2.483,
699
+ "eval_samples_per_second": 30.206,
700
+ "eval_steps_per_second": 4.027,
701
+ "step": 1189
702
+ },
703
+ {
704
+ "epoch": 41.37931034482759,
705
+ "grad_norm": 0.581297755241394,
706
+ "learning_rate": 2.406e-05,
707
+ "loss": 0.0923,
708
+ "step": 1200
709
+ },
710
+ {
711
+ "epoch": 42.0,
712
+ "eval_accuracy_no_text": 0.9703110735401259,
713
+ "eval_accuracy_text": NaN,
714
+ "eval_iou_no_text": 0.9703110735401259,
715
+ "eval_iou_text": 0.0,
716
+ "eval_loss": 0.13015305995941162,
717
+ "eval_mean_accuracy": 0.9703110735401259,
718
+ "eval_mean_iou": 0.48515553677006296,
719
+ "eval_overall_accuracy": 0.9703110735401259,
720
+ "eval_runtime": 2.4521,
721
+ "eval_samples_per_second": 30.586,
722
+ "eval_steps_per_second": 4.078,
723
+ "step": 1218
724
+ },
725
+ {
726
+ "epoch": 43.0,
727
+ "eval_accuracy_no_text": 0.9708830820102969,
728
+ "eval_accuracy_text": NaN,
729
+ "eval_iou_no_text": 0.9708830820102969,
730
+ "eval_iou_text": 0.0,
731
+ "eval_loss": 0.1296657919883728,
732
+ "eval_mean_accuracy": 0.9708830820102969,
733
+ "eval_mean_iou": 0.48544154100514847,
734
+ "eval_overall_accuracy": 0.9708830820102969,
735
+ "eval_runtime": 2.5553,
736
+ "eval_samples_per_second": 29.351,
737
+ "eval_steps_per_second": 3.913,
738
+ "step": 1247
739
+ },
740
+ {
741
+ "epoch": 44.0,
742
+ "eval_accuracy_no_text": 0.9700965774158937,
743
+ "eval_accuracy_text": NaN,
744
+ "eval_iou_no_text": 0.9700965774158937,
745
+ "eval_iou_text": 0.0,
746
+ "eval_loss": 0.12993517518043518,
747
+ "eval_mean_accuracy": 0.9700965774158937,
748
+ "eval_mean_iou": 0.48504828870794686,
749
+ "eval_overall_accuracy": 0.9700965774158937,
750
+ "eval_runtime": 2.572,
751
+ "eval_samples_per_second": 29.16,
752
+ "eval_steps_per_second": 3.888,
753
+ "step": 1276
754
+ },
755
+ {
756
+ "epoch": 44.827586206896555,
757
+ "grad_norm": 0.5657644867897034,
758
+ "learning_rate": 2.1065e-05,
759
+ "loss": 0.0812,
760
+ "step": 1300
761
+ },
762
+ {
763
+ "epoch": 45.0,
764
+ "eval_accuracy_no_text": 0.9733241742773675,
765
+ "eval_accuracy_text": NaN,
766
+ "eval_iou_no_text": 0.9733241742773675,
767
+ "eval_iou_text": 0.0,
768
+ "eval_loss": 0.12587976455688477,
769
+ "eval_mean_accuracy": 0.9733241742773675,
770
+ "eval_mean_iou": 0.48666208713868375,
771
+ "eval_overall_accuracy": 0.9733241742773675,
772
+ "eval_runtime": 2.476,
773
+ "eval_samples_per_second": 30.291,
774
+ "eval_steps_per_second": 4.039,
775
+ "step": 1305
776
+ },
777
+ {
778
+ "epoch": 46.0,
779
+ "eval_accuracy_no_text": 0.9766478487031672,
780
+ "eval_accuracy_text": NaN,
781
+ "eval_iou_no_text": 0.9766478487031672,
782
+ "eval_iou_text": 0.0,
783
+ "eval_loss": 0.12515687942504883,
784
+ "eval_mean_accuracy": 0.9766478487031672,
785
+ "eval_mean_iou": 0.4883239243515836,
786
+ "eval_overall_accuracy": 0.9766478487031672,
787
+ "eval_runtime": 2.468,
788
+ "eval_samples_per_second": 30.389,
789
+ "eval_steps_per_second": 4.052,
790
+ "step": 1334
791
+ },
792
+ {
793
+ "epoch": 47.0,
794
+ "eval_accuracy_no_text": 0.9761687020479585,
795
+ "eval_accuracy_text": NaN,
796
+ "eval_iou_no_text": 0.9761687020479585,
797
+ "eval_iou_text": 0.0,
798
+ "eval_loss": 0.12227589637041092,
799
+ "eval_mean_accuracy": 0.9761687020479585,
800
+ "eval_mean_iou": 0.48808435102397923,
801
+ "eval_overall_accuracy": 0.9761687020479585,
802
+ "eval_runtime": 2.4339,
803
+ "eval_samples_per_second": 30.815,
804
+ "eval_steps_per_second": 4.109,
805
+ "step": 1363
806
+ },
807
+ {
808
+ "epoch": 48.0,
809
+ "eval_accuracy_no_text": 0.9757235746360194,
810
+ "eval_accuracy_text": NaN,
811
+ "eval_iou_no_text": 0.9757235746360194,
812
+ "eval_iou_text": 0.0,
813
+ "eval_loss": 0.12267401069402695,
814
+ "eval_mean_accuracy": 0.9757235746360194,
815
+ "eval_mean_iou": 0.4878617873180097,
816
+ "eval_overall_accuracy": 0.9757235746360194,
817
+ "eval_runtime": 2.4543,
818
+ "eval_samples_per_second": 30.559,
819
+ "eval_steps_per_second": 4.075,
820
+ "step": 1392
821
+ },
822
+ {
823
+ "epoch": 48.275862068965516,
824
+ "grad_norm": 1.296958565711975,
825
+ "learning_rate": 1.807e-05,
826
+ "loss": 0.0901,
827
+ "step": 1400
828
+ },
829
+ {
830
+ "epoch": 49.0,
831
+ "eval_accuracy_no_text": 0.9759114985155085,
832
+ "eval_accuracy_text": NaN,
833
+ "eval_iou_no_text": 0.9759114985155085,
834
+ "eval_iou_text": 0.0,
835
+ "eval_loss": 0.12235772609710693,
836
+ "eval_mean_accuracy": 0.9759114985155085,
837
+ "eval_mean_iou": 0.48795574925775426,
838
+ "eval_overall_accuracy": 0.9759114985155085,
839
+ "eval_runtime": 2.4545,
840
+ "eval_samples_per_second": 30.556,
841
+ "eval_steps_per_second": 4.074,
842
+ "step": 1421
843
+ },
844
+ {
845
+ "epoch": 50.0,
846
+ "eval_accuracy_no_text": 0.9636262076126153,
847
+ "eval_accuracy_text": NaN,
848
+ "eval_iou_no_text": 0.9636262076126153,
849
+ "eval_iou_text": 0.0,
850
+ "eval_loss": 0.13496458530426025,
851
+ "eval_mean_accuracy": 0.9636262076126153,
852
+ "eval_mean_iou": 0.48181310380630765,
853
+ "eval_overall_accuracy": 0.9636262076126153,
854
+ "eval_runtime": 2.4536,
855
+ "eval_samples_per_second": 30.567,
856
+ "eval_steps_per_second": 4.076,
857
+ "step": 1450
858
+ },
859
+ {
860
+ "epoch": 51.0,
861
+ "eval_accuracy_no_text": 0.9718740969809078,
862
+ "eval_accuracy_text": NaN,
863
+ "eval_iou_no_text": 0.9718740969809078,
864
+ "eval_iou_text": 0.0,
865
+ "eval_loss": 0.12845410406589508,
866
+ "eval_mean_accuracy": 0.9718740969809078,
867
+ "eval_mean_iou": 0.4859370484904539,
868
+ "eval_overall_accuracy": 0.9718740969809078,
869
+ "eval_runtime": 2.4972,
870
+ "eval_samples_per_second": 30.034,
871
+ "eval_steps_per_second": 4.005,
872
+ "step": 1479
873
+ },
874
+ {
875
+ "epoch": 51.724137931034484,
876
+ "grad_norm": 1.6991603374481201,
877
+ "learning_rate": 1.5075000000000002e-05,
878
+ "loss": 0.083,
879
+ "step": 1500
880
+ },
881
+ {
882
+ "epoch": 52.0,
883
+ "eval_accuracy_no_text": 0.9695387295262546,
884
+ "eval_accuracy_text": NaN,
885
+ "eval_iou_no_text": 0.9695387295262546,
886
+ "eval_iou_text": 0.0,
887
+ "eval_loss": 0.12856803834438324,
888
+ "eval_mean_accuracy": 0.9695387295262546,
889
+ "eval_mean_iou": 0.4847693647631273,
890
+ "eval_overall_accuracy": 0.9695387295262546,
891
+ "eval_runtime": 2.5249,
892
+ "eval_samples_per_second": 29.704,
893
+ "eval_steps_per_second": 3.961,
894
+ "step": 1508
895
+ },
896
+ {
897
+ "epoch": 53.0,
898
+ "eval_accuracy_no_text": 0.9700392580939001,
899
+ "eval_accuracy_text": NaN,
900
+ "eval_iou_no_text": 0.9700392580939001,
901
+ "eval_iou_text": 0.0,
902
+ "eval_loss": 0.12850378453731537,
903
+ "eval_mean_accuracy": 0.9700392580939001,
904
+ "eval_mean_iou": 0.48501962904695006,
905
+ "eval_overall_accuracy": 0.9700392580939001,
906
+ "eval_runtime": 2.4685,
907
+ "eval_samples_per_second": 30.383,
908
+ "eval_steps_per_second": 4.051,
909
+ "step": 1537
910
+ },
911
+ {
912
+ "epoch": 54.0,
913
+ "eval_accuracy_no_text": 0.9775246199463794,
914
+ "eval_accuracy_text": NaN,
915
+ "eval_iou_no_text": 0.9775246199463794,
916
+ "eval_iou_text": 0.0,
917
+ "eval_loss": 0.12516801059246063,
918
+ "eval_mean_accuracy": 0.9775246199463794,
919
+ "eval_mean_iou": 0.4887623099731897,
920
+ "eval_overall_accuracy": 0.9775246199463794,
921
+ "eval_runtime": 2.4581,
922
+ "eval_samples_per_second": 30.511,
923
+ "eval_steps_per_second": 4.068,
924
+ "step": 1566
925
+ },
926
+ {
927
+ "epoch": 55.0,
928
+ "eval_accuracy_no_text": 0.9734067118443722,
929
+ "eval_accuracy_text": NaN,
930
+ "eval_iou_no_text": 0.9734067118443722,
931
+ "eval_iou_text": 0.0,
932
+ "eval_loss": 0.1259094923734665,
933
+ "eval_mean_accuracy": 0.9734067118443722,
934
+ "eval_mean_iou": 0.4867033559221861,
935
+ "eval_overall_accuracy": 0.9734067118443722,
936
+ "eval_runtime": 2.4476,
937
+ "eval_samples_per_second": 30.642,
938
+ "eval_steps_per_second": 4.086,
939
+ "step": 1595
940
+ },
941
+ {
942
+ "epoch": 55.172413793103445,
943
+ "grad_norm": 0.6328563690185547,
944
+ "learning_rate": 1.2079999999999998e-05,
945
+ "loss": 0.0822,
946
+ "step": 1600
947
+ },
948
+ {
949
+ "epoch": 56.0,
950
+ "eval_accuracy_no_text": 0.9642389489080613,
951
+ "eval_accuracy_text": NaN,
952
+ "eval_iou_no_text": 0.9642389489080613,
953
+ "eval_iou_text": 0.0,
954
+ "eval_loss": 0.13731960952281952,
955
+ "eval_mean_accuracy": 0.9642389489080613,
956
+ "eval_mean_iou": 0.48211947445403064,
957
+ "eval_overall_accuracy": 0.9642389489080613,
958
+ "eval_runtime": 2.4564,
959
+ "eval_samples_per_second": 30.532,
960
+ "eval_steps_per_second": 4.071,
961
+ "step": 1624
962
+ },
963
+ {
964
+ "epoch": 57.0,
965
+ "eval_accuracy_no_text": 0.9719260003039729,
966
+ "eval_accuracy_text": NaN,
967
+ "eval_iou_no_text": 0.9719260003039729,
968
+ "eval_iou_text": 0.0,
969
+ "eval_loss": 0.12484534084796906,
970
+ "eval_mean_accuracy": 0.9719260003039729,
971
+ "eval_mean_iou": 0.48596300015198646,
972
+ "eval_overall_accuracy": 0.9719260003039729,
973
+ "eval_runtime": 2.4497,
974
+ "eval_samples_per_second": 30.616,
975
+ "eval_steps_per_second": 4.082,
976
+ "step": 1653
977
+ },
978
+ {
979
+ "epoch": 58.0,
980
+ "eval_accuracy_no_text": 0.9761464174690337,
981
+ "eval_accuracy_text": NaN,
982
+ "eval_iou_no_text": 0.9761464174690337,
983
+ "eval_iou_text": 0.0,
984
+ "eval_loss": 0.12278713285923004,
985
+ "eval_mean_accuracy": 0.9761464174690337,
986
+ "eval_mean_iou": 0.4880732087345169,
987
+ "eval_overall_accuracy": 0.9761464174690337,
988
+ "eval_runtime": 2.4362,
989
+ "eval_samples_per_second": 30.786,
990
+ "eval_steps_per_second": 4.105,
991
+ "step": 1682
992
+ },
993
+ {
994
+ "epoch": 58.62068965517241,
995
+ "grad_norm": 1.0782545804977417,
996
+ "learning_rate": 9.085000000000002e-06,
997
+ "loss": 0.0777,
998
+ "step": 1700
999
+ },
1000
+ {
1001
+ "epoch": 59.0,
1002
+ "eval_accuracy_no_text": 0.967698982390218,
1003
+ "eval_accuracy_text": NaN,
1004
+ "eval_iou_no_text": 0.967698982390218,
1005
+ "eval_iou_text": 0.0,
1006
+ "eval_loss": 0.13151277601718903,
1007
+ "eval_mean_accuracy": 0.967698982390218,
1008
+ "eval_mean_iou": 0.483849491195109,
1009
+ "eval_overall_accuracy": 0.967698982390218,
1010
+ "eval_runtime": 2.4544,
1011
+ "eval_samples_per_second": 30.557,
1012
+ "eval_steps_per_second": 4.074,
1013
+ "step": 1711
1014
+ },
1015
+ {
1016
+ "epoch": 60.0,
1017
+ "eval_accuracy_no_text": 0.9766246614577545,
1018
+ "eval_accuracy_text": NaN,
1019
+ "eval_iou_no_text": 0.9766246614577545,
1020
+ "eval_iou_text": 0.0,
1021
+ "eval_loss": 0.12540759146213531,
1022
+ "eval_mean_accuracy": 0.9766246614577545,
1023
+ "eval_mean_iou": 0.48831233072887725,
1024
+ "eval_overall_accuracy": 0.9766246614577545,
1025
+ "eval_runtime": 2.449,
1026
+ "eval_samples_per_second": 30.624,
1027
+ "eval_steps_per_second": 4.083,
1028
+ "step": 1740
1029
+ },
1030
+ {
1031
+ "epoch": 61.0,
1032
+ "eval_accuracy_no_text": 0.9682487062814643,
1033
+ "eval_accuracy_text": NaN,
1034
+ "eval_iou_no_text": 0.9682487062814643,
1035
+ "eval_iou_text": 0.0,
1036
+ "eval_loss": 0.13411304354667664,
1037
+ "eval_mean_accuracy": 0.9682487062814643,
1038
+ "eval_mean_iou": 0.48412435314073216,
1039
+ "eval_overall_accuracy": 0.9682487062814643,
1040
+ "eval_runtime": 2.4399,
1041
+ "eval_samples_per_second": 30.739,
1042
+ "eval_steps_per_second": 4.099,
1043
+ "step": 1769
1044
+ },
1045
+ {
1046
+ "epoch": 62.0,
1047
+ "eval_accuracy_no_text": 0.9726401787460179,
1048
+ "eval_accuracy_text": NaN,
1049
+ "eval_iou_no_text": 0.9726401787460179,
1050
+ "eval_iou_text": 0.0,
1051
+ "eval_loss": 0.12687736749649048,
1052
+ "eval_mean_accuracy": 0.9726401787460179,
1053
+ "eval_mean_iou": 0.48632008937300897,
1054
+ "eval_overall_accuracy": 0.9726401787460179,
1055
+ "eval_runtime": 2.4427,
1056
+ "eval_samples_per_second": 30.703,
1057
+ "eval_steps_per_second": 4.094,
1058
+ "step": 1798
1059
+ },
1060
+ {
1061
+ "epoch": 62.06896551724138,
1062
+ "grad_norm": 0.44511547684669495,
1063
+ "learning_rate": 6.0899999999999984e-06,
1064
+ "loss": 0.079,
1065
+ "step": 1800
1066
+ },
1067
+ {
1068
+ "epoch": 63.0,
1069
+ "eval_accuracy_no_text": 0.9755465955876985,
1070
+ "eval_accuracy_text": NaN,
1071
+ "eval_iou_no_text": 0.9755465955876985,
1072
+ "eval_iou_text": 0.0,
1073
+ "eval_loss": 0.12655200064182281,
1074
+ "eval_mean_accuracy": 0.9755465955876985,
1075
+ "eval_mean_iou": 0.48777329779384926,
1076
+ "eval_overall_accuracy": 0.9755465955876985,
1077
+ "eval_runtime": 2.433,
1078
+ "eval_samples_per_second": 30.826,
1079
+ "eval_steps_per_second": 4.11,
1080
+ "step": 1827
1081
+ },
1082
+ {
1083
+ "epoch": 64.0,
1084
+ "eval_accuracy_no_text": 0.9705849764026055,
1085
+ "eval_accuracy_text": NaN,
1086
+ "eval_iou_no_text": 0.9705849764026055,
1087
+ "eval_iou_text": 0.0,
1088
+ "eval_loss": 0.13036541640758514,
1089
+ "eval_mean_accuracy": 0.9705849764026055,
1090
+ "eval_mean_iou": 0.4852924882013028,
1091
+ "eval_overall_accuracy": 0.9705849764026055,
1092
+ "eval_runtime": 2.4456,
1093
+ "eval_samples_per_second": 30.668,
1094
+ "eval_steps_per_second": 4.089,
1095
+ "step": 1856
1096
+ },
1097
+ {
1098
+ "epoch": 65.0,
1099
+ "eval_accuracy_no_text": 0.974686523674516,
1100
+ "eval_accuracy_text": NaN,
1101
+ "eval_iou_no_text": 0.974686523674516,
1102
+ "eval_iou_text": 0.0,
1103
+ "eval_loss": 0.12534736096858978,
1104
+ "eval_mean_accuracy": 0.974686523674516,
1105
+ "eval_mean_iou": 0.487343261837258,
1106
+ "eval_overall_accuracy": 0.974686523674516,
1107
+ "eval_runtime": 2.4366,
1108
+ "eval_samples_per_second": 30.78,
1109
+ "eval_steps_per_second": 4.104,
1110
+ "step": 1885
1111
+ },
1112
+ {
1113
+ "epoch": 65.51724137931035,
1114
+ "grad_norm": 0.8290705680847168,
1115
+ "learning_rate": 3.0950000000000026e-06,
1116
+ "loss": 0.0781,
1117
+ "step": 1900
1118
+ },
1119
+ {
1120
+ "epoch": 66.0,
1121
+ "eval_accuracy_no_text": 0.9731443179796158,
1122
+ "eval_accuracy_text": NaN,
1123
+ "eval_iou_no_text": 0.9731443179796158,
1124
+ "eval_iou_text": 0.0,
1125
+ "eval_loss": 0.12833286821842194,
1126
+ "eval_mean_accuracy": 0.9731443179796158,
1127
+ "eval_mean_iou": 0.4865721589898079,
1128
+ "eval_overall_accuracy": 0.9731443179796158,
1129
+ "eval_runtime": 2.4332,
1130
+ "eval_samples_per_second": 30.823,
1131
+ "eval_steps_per_second": 4.11,
1132
+ "step": 1914
1133
+ },
1134
+ {
1135
+ "epoch": 67.0,
1136
+ "eval_accuracy_no_text": 0.9784451705142637,
1137
+ "eval_accuracy_text": NaN,
1138
+ "eval_iou_no_text": 0.9784451705142637,
1139
+ "eval_iou_text": 0.0,
1140
+ "eval_loss": 0.12897755205631256,
1141
+ "eval_mean_accuracy": 0.9784451705142637,
1142
+ "eval_mean_iou": 0.48922258525713186,
1143
+ "eval_overall_accuracy": 0.9784451705142637,
1144
+ "eval_runtime": 2.443,
1145
+ "eval_samples_per_second": 30.7,
1146
+ "eval_steps_per_second": 4.093,
1147
+ "step": 1943
1148
+ },
1149
+ {
1150
+ "epoch": 68.0,
1151
+ "eval_accuracy_no_text": 0.9669106724628387,
1152
+ "eval_accuracy_text": NaN,
1153
+ "eval_iou_no_text": 0.9669106724628387,
1154
+ "eval_iou_text": 0.0,
1155
+ "eval_loss": 0.13632425665855408,
1156
+ "eval_mean_accuracy": 0.9669106724628387,
1157
+ "eval_mean_iou": 0.48345533623141934,
1158
+ "eval_overall_accuracy": 0.9669106724628387,
1159
+ "eval_runtime": 2.4464,
1160
+ "eval_samples_per_second": 30.657,
1161
+ "eval_steps_per_second": 4.088,
1162
+ "step": 1972
1163
+ },
1164
+ {
1165
+ "epoch": 68.96551724137932,
1166
+ "grad_norm": 2.6724536418914795,
1167
+ "learning_rate": 1e-07,
1168
+ "loss": 0.0826,
1169
+ "step": 2000
1170
+ },
1171
+ {
1172
+ "epoch": 68.96551724137932,
1173
+ "eval_accuracy_no_text": 0.9701784379830324,
1174
+ "eval_accuracy_text": NaN,
1175
+ "eval_iou_no_text": 0.9701784379830324,
1176
+ "eval_iou_text": 0.0,
1177
+ "eval_loss": 0.12989187240600586,
1178
+ "eval_mean_accuracy": 0.9701784379830324,
1179
+ "eval_mean_iou": 0.4850892189915162,
1180
+ "eval_overall_accuracy": 0.9701784379830324,
1181
+ "eval_runtime": 2.4612,
1182
+ "eval_samples_per_second": 30.473,
1183
+ "eval_steps_per_second": 4.063,
1184
+ "step": 2000
1185
+ },
1186
+ {
1187
+ "epoch": 68.96551724137932,
1188
+ "step": 2000,
1189
+ "total_flos": 2.7210398151632486e+17,
1190
+ "train_loss": 0.11668173718452454,
1191
+ "train_runtime": 724.9682,
1192
+ "train_samples_per_second": 22.07,
1193
+ "train_steps_per_second": 2.759
1194
+ }
1195
+ ],
1196
+ "logging_steps": 100,
1197
+ "max_steps": 2000,
1198
+ "num_input_tokens_seen": 0,
1199
+ "num_train_epochs": 69,
1200
+ "save_steps": 500,
1201
+ "stateful_callbacks": {
1202
+ "TrainerControl": {
1203
+ "args": {
1204
+ "should_epoch_stop": false,
1205
+ "should_evaluate": false,
1206
+ "should_log": false,
1207
+ "should_save": true,
1208
+ "should_training_stop": true
1209
+ },
1210
+ "attributes": {}
1211
+ }
1212
+ },
1213
+ "total_flos": 2.7210398151632486e+17,
1214
+ "train_batch_size": 8,
1215
+ "trial_name": null,
1216
+ "trial_params": null
1217
+ }