sam1120 commited on
Commit
3e7274b
1 Parent(s): 7137d3e

Training in progress, step 200

Browse files
checkpoint-200/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "nvidia/mit-b0",
3
+ "architectures": [
4
+ "SegformerForSemanticSegmentation"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "classifier_dropout_prob": 0.1,
8
+ "decoder_hidden_size": 256,
9
+ "depths": [
10
+ 2,
11
+ 2,
12
+ 2,
13
+ 2
14
+ ],
15
+ "downsampling_rates": [
16
+ 1,
17
+ 4,
18
+ 8,
19
+ 16
20
+ ],
21
+ "drop_path_rate": 0.1,
22
+ "hidden_act": "gelu",
23
+ "hidden_dropout_prob": 0.0,
24
+ "hidden_sizes": [
25
+ 32,
26
+ 64,
27
+ 160,
28
+ 256
29
+ ],
30
+ "id2label": {
31
+ "0": "unlabeled",
32
+ "1": "dropoff",
33
+ "2": "undropoff"
34
+ },
35
+ "image_size": 224,
36
+ "initializer_range": 0.02,
37
+ "label2id": {
38
+ "dropoff": 1,
39
+ "undropoff": 2,
40
+ "unlabeled": 0
41
+ },
42
+ "layer_norm_eps": 1e-06,
43
+ "mlp_ratios": [
44
+ 4,
45
+ 4,
46
+ 4,
47
+ 4
48
+ ],
49
+ "model_type": "segformer",
50
+ "num_attention_heads": [
51
+ 1,
52
+ 2,
53
+ 5,
54
+ 8
55
+ ],
56
+ "num_channels": 3,
57
+ "num_encoder_blocks": 4,
58
+ "patch_sizes": [
59
+ 7,
60
+ 3,
61
+ 3,
62
+ 3
63
+ ],
64
+ "reshape_last_stage": true,
65
+ "semantic_loss_ignore_index": 255,
66
+ "sr_ratios": [
67
+ 8,
68
+ 4,
69
+ 2,
70
+ 1
71
+ ],
72
+ "strides": [
73
+ 4,
74
+ 2,
75
+ 2,
76
+ 2
77
+ ],
78
+ "torch_dtype": "float32",
79
+ "transformers_version": "4.30.2"
80
+ }
checkpoint-200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a675cd2a73a3d25a1946520cf54ccdc55edf6fbd37620a7ce4bd190347830ed
3
+ size 29838853
checkpoint-200/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84bf59eeba83524aa2f95d95c0fb00af94b856d9a46f7c9107d87d3865e0dd25
3
+ size 14932813
checkpoint-200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83edd066879d3b986e6c1ca5f686b4fce3a2aa6be7381480c9ed4b3261dc5eff
3
+ size 14575
checkpoint-200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8345e28c36423dbe7e6531bc7b30508fece52ffd391c1ee37236387d75104d8c
3
+ size 627
checkpoint-200/trainer_state.json ADDED
@@ -0,0 +1,1556 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.4188476800918579,
3
+ "best_model_checkpoint": "/robodata/smodak/Projects/nspl/scripts/terrainseg/training/models/dropoff-utcustom-train-SF-RGB-b0_3/checkpoint-200",
4
+ "epoch": 66.66666666666667,
5
+ "global_step": 200,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.33,
12
+ "learning_rate": 1.111111111111111e-06,
13
+ "loss": 1.1153,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.67,
18
+ "learning_rate": 2.222222222222222e-06,
19
+ "loss": 1.1199,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 1.0,
24
+ "learning_rate": 3.3333333333333333e-06,
25
+ "loss": 1.1156,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 1.33,
30
+ "learning_rate": 4.444444444444444e-06,
31
+ "loss": 1.1182,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 1.67,
36
+ "learning_rate": 5.555555555555557e-06,
37
+ "loss": 1.1161,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 2.0,
42
+ "learning_rate": 6.666666666666667e-06,
43
+ "loss": 1.1106,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 2.33,
48
+ "learning_rate": 7.77777777777778e-06,
49
+ "loss": 1.1141,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 2.67,
54
+ "learning_rate": 8.888888888888888e-06,
55
+ "loss": 1.1085,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 3.0,
60
+ "learning_rate": 1e-05,
61
+ "loss": 1.0992,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 3.33,
66
+ "learning_rate": 1.1111111111111113e-05,
67
+ "loss": 1.1015,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 3.33,
72
+ "eval_accuracy_dropoff": 0.5975256737197602,
73
+ "eval_accuracy_undropoff": 0.31695782815139334,
74
+ "eval_accuracy_unlabeled": NaN,
75
+ "eval_iou_dropoff": 0.04272275097162494,
76
+ "eval_iou_undropoff": 0.3123742688218148,
77
+ "eval_iou_unlabeled": 0.0,
78
+ "eval_loss": 1.0990227460861206,
79
+ "eval_mean_accuracy": 0.45724175093557673,
80
+ "eval_mean_iou": 0.1183656732644799,
81
+ "eval_overall_accuracy": 0.32938079833984374,
82
+ "eval_runtime": 1.6795,
83
+ "eval_samples_per_second": 8.931,
84
+ "eval_steps_per_second": 0.595,
85
+ "step": 10
86
+ },
87
+ {
88
+ "epoch": 3.67,
89
+ "learning_rate": 1.2222222222222224e-05,
90
+ "loss": 1.102,
91
+ "step": 11
92
+ },
93
+ {
94
+ "epoch": 4.0,
95
+ "learning_rate": 1.3333333333333333e-05,
96
+ "loss": 1.0912,
97
+ "step": 12
98
+ },
99
+ {
100
+ "epoch": 4.33,
101
+ "learning_rate": 1.4444444444444446e-05,
102
+ "loss": 1.0939,
103
+ "step": 13
104
+ },
105
+ {
106
+ "epoch": 4.67,
107
+ "learning_rate": 1.555555555555556e-05,
108
+ "loss": 1.0846,
109
+ "step": 14
110
+ },
111
+ {
112
+ "epoch": 5.0,
113
+ "learning_rate": 1.6666666666666667e-05,
114
+ "loss": 1.0804,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 5.33,
119
+ "learning_rate": 1.7777777777777777e-05,
120
+ "loss": 1.0734,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 5.67,
125
+ "learning_rate": 1.888888888888889e-05,
126
+ "loss": 1.0642,
127
+ "step": 17
128
+ },
129
+ {
130
+ "epoch": 6.0,
131
+ "learning_rate": 2e-05,
132
+ "loss": 1.0945,
133
+ "step": 18
134
+ },
135
+ {
136
+ "epoch": 6.33,
137
+ "learning_rate": 1.994152046783626e-05,
138
+ "loss": 1.0588,
139
+ "step": 19
140
+ },
141
+ {
142
+ "epoch": 6.67,
143
+ "learning_rate": 1.9883040935672515e-05,
144
+ "loss": 1.0478,
145
+ "step": 20
146
+ },
147
+ {
148
+ "epoch": 6.67,
149
+ "eval_accuracy_dropoff": 0.8648482551060261,
150
+ "eval_accuracy_undropoff": 0.5515245664509166,
151
+ "eval_accuracy_unlabeled": NaN,
152
+ "eval_iou_dropoff": 0.08793587759511783,
153
+ "eval_iou_undropoff": 0.5482206998703943,
154
+ "eval_iou_unlabeled": 0.0,
155
+ "eval_loss": 1.0755951404571533,
156
+ "eval_mean_accuracy": 0.7081864107784714,
157
+ "eval_mean_iou": 0.21205219248850404,
158
+ "eval_overall_accuracy": 0.5653978983561198,
159
+ "eval_runtime": 1.8074,
160
+ "eval_samples_per_second": 8.299,
161
+ "eval_steps_per_second": 0.553,
162
+ "step": 20
163
+ },
164
+ {
165
+ "epoch": 7.0,
166
+ "learning_rate": 1.9824561403508773e-05,
167
+ "loss": 1.0748,
168
+ "step": 21
169
+ },
170
+ {
171
+ "epoch": 7.33,
172
+ "learning_rate": 1.976608187134503e-05,
173
+ "loss": 1.0343,
174
+ "step": 22
175
+ },
176
+ {
177
+ "epoch": 7.67,
178
+ "learning_rate": 1.970760233918129e-05,
179
+ "loss": 1.0348,
180
+ "step": 23
181
+ },
182
+ {
183
+ "epoch": 8.0,
184
+ "learning_rate": 1.9649122807017544e-05,
185
+ "loss": 1.0366,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 8.33,
190
+ "learning_rate": 1.9590643274853802e-05,
191
+ "loss": 1.0351,
192
+ "step": 25
193
+ },
194
+ {
195
+ "epoch": 8.67,
196
+ "learning_rate": 1.953216374269006e-05,
197
+ "loss": 1.0085,
198
+ "step": 26
199
+ },
200
+ {
201
+ "epoch": 9.0,
202
+ "learning_rate": 1.9473684210526318e-05,
203
+ "loss": 1.0048,
204
+ "step": 27
205
+ },
206
+ {
207
+ "epoch": 9.33,
208
+ "learning_rate": 1.9415204678362573e-05,
209
+ "loss": 1.0001,
210
+ "step": 28
211
+ },
212
+ {
213
+ "epoch": 9.67,
214
+ "learning_rate": 1.935672514619883e-05,
215
+ "loss": 0.9877,
216
+ "step": 29
217
+ },
218
+ {
219
+ "epoch": 10.0,
220
+ "learning_rate": 1.929824561403509e-05,
221
+ "loss": 1.0451,
222
+ "step": 30
223
+ },
224
+ {
225
+ "epoch": 10.0,
226
+ "eval_accuracy_dropoff": 0.8842327750591586,
227
+ "eval_accuracy_undropoff": 0.7264455095352592,
228
+ "eval_accuracy_unlabeled": NaN,
229
+ "eval_iou_dropoff": 0.1312558881078339,
230
+ "eval_iou_undropoff": 0.7226098464796189,
231
+ "eval_iou_unlabeled": 0.0,
232
+ "eval_loss": 1.026917576789856,
233
+ "eval_mean_accuracy": 0.8053391422972089,
234
+ "eval_mean_iou": 0.2846219115291509,
235
+ "eval_overall_accuracy": 0.7334320068359375,
236
+ "eval_runtime": 1.7831,
237
+ "eval_samples_per_second": 8.412,
238
+ "eval_steps_per_second": 0.561,
239
+ "step": 30
240
+ },
241
+ {
242
+ "epoch": 10.33,
243
+ "learning_rate": 1.9239766081871347e-05,
244
+ "loss": 0.9774,
245
+ "step": 31
246
+ },
247
+ {
248
+ "epoch": 10.67,
249
+ "learning_rate": 1.9181286549707602e-05,
250
+ "loss": 0.9797,
251
+ "step": 32
252
+ },
253
+ {
254
+ "epoch": 11.0,
255
+ "learning_rate": 1.912280701754386e-05,
256
+ "loss": 0.9325,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 11.33,
261
+ "learning_rate": 1.9064327485380118e-05,
262
+ "loss": 0.9775,
263
+ "step": 34
264
+ },
265
+ {
266
+ "epoch": 11.67,
267
+ "learning_rate": 1.9005847953216376e-05,
268
+ "loss": 0.9393,
269
+ "step": 35
270
+ },
271
+ {
272
+ "epoch": 12.0,
273
+ "learning_rate": 1.894736842105263e-05,
274
+ "loss": 0.9431,
275
+ "step": 36
276
+ },
277
+ {
278
+ "epoch": 12.33,
279
+ "learning_rate": 1.888888888888889e-05,
280
+ "loss": 0.9369,
281
+ "step": 37
282
+ },
283
+ {
284
+ "epoch": 12.67,
285
+ "learning_rate": 1.8830409356725147e-05,
286
+ "loss": 0.9234,
287
+ "step": 38
288
+ },
289
+ {
290
+ "epoch": 13.0,
291
+ "learning_rate": 1.8771929824561405e-05,
292
+ "loss": 1.0184,
293
+ "step": 39
294
+ },
295
+ {
296
+ "epoch": 13.33,
297
+ "learning_rate": 1.871345029239766e-05,
298
+ "loss": 0.9095,
299
+ "step": 40
300
+ },
301
+ {
302
+ "epoch": 13.33,
303
+ "eval_accuracy_dropoff": 0.7349346382704988,
304
+ "eval_accuracy_undropoff": 0.8459984055569215,
305
+ "eval_accuracy_unlabeled": NaN,
306
+ "eval_iou_dropoff": 0.17233702408257653,
307
+ "eval_iou_undropoff": 0.835792668708584,
308
+ "eval_iou_unlabeled": 0.0,
309
+ "eval_loss": 0.947559118270874,
310
+ "eval_mean_accuracy": 0.7904665219137101,
311
+ "eval_mean_iou": 0.33604323093038685,
312
+ "eval_overall_accuracy": 0.8410807291666667,
313
+ "eval_runtime": 1.7664,
314
+ "eval_samples_per_second": 8.492,
315
+ "eval_steps_per_second": 0.566,
316
+ "step": 40
317
+ },
318
+ {
319
+ "epoch": 13.67,
320
+ "learning_rate": 1.8654970760233918e-05,
321
+ "loss": 0.9068,
322
+ "step": 41
323
+ },
324
+ {
325
+ "epoch": 14.0,
326
+ "learning_rate": 1.8596491228070176e-05,
327
+ "loss": 0.9478,
328
+ "step": 42
329
+ },
330
+ {
331
+ "epoch": 14.33,
332
+ "learning_rate": 1.8538011695906434e-05,
333
+ "loss": 0.9028,
334
+ "step": 43
335
+ },
336
+ {
337
+ "epoch": 14.67,
338
+ "learning_rate": 1.847953216374269e-05,
339
+ "loss": 0.8937,
340
+ "step": 44
341
+ },
342
+ {
343
+ "epoch": 15.0,
344
+ "learning_rate": 1.8421052631578947e-05,
345
+ "loss": 0.7988,
346
+ "step": 45
347
+ },
348
+ {
349
+ "epoch": 15.33,
350
+ "learning_rate": 1.8362573099415205e-05,
351
+ "loss": 0.8873,
352
+ "step": 46
353
+ },
354
+ {
355
+ "epoch": 15.67,
356
+ "learning_rate": 1.8304093567251464e-05,
357
+ "loss": 0.8282,
358
+ "step": 47
359
+ },
360
+ {
361
+ "epoch": 16.0,
362
+ "learning_rate": 1.824561403508772e-05,
363
+ "loss": 0.868,
364
+ "step": 48
365
+ },
366
+ {
367
+ "epoch": 16.33,
368
+ "learning_rate": 1.8187134502923976e-05,
369
+ "loss": 0.8684,
370
+ "step": 49
371
+ },
372
+ {
373
+ "epoch": 16.67,
374
+ "learning_rate": 1.8128654970760235e-05,
375
+ "loss": 0.8091,
376
+ "step": 50
377
+ },
378
+ {
379
+ "epoch": 16.67,
380
+ "eval_accuracy_dropoff": 0.5974624945436166,
381
+ "eval_accuracy_undropoff": 0.9315254285997107,
382
+ "eval_accuracy_unlabeled": NaN,
383
+ "eval_iou_dropoff": 0.2429297387681515,
384
+ "eval_iou_undropoff": 0.914512280023459,
385
+ "eval_iou_unlabeled": 0.0,
386
+ "eval_loss": 0.8424645662307739,
387
+ "eval_mean_accuracy": 0.7644939615716636,
388
+ "eval_mean_iou": 0.3858140062638702,
389
+ "eval_overall_accuracy": 0.9167338053385417,
390
+ "eval_runtime": 1.7428,
391
+ "eval_samples_per_second": 8.607,
392
+ "eval_steps_per_second": 0.574,
393
+ "step": 50
394
+ },
395
+ {
396
+ "epoch": 17.0,
397
+ "learning_rate": 1.8070175438596493e-05,
398
+ "loss": 0.7826,
399
+ "step": 51
400
+ },
401
+ {
402
+ "epoch": 17.33,
403
+ "learning_rate": 1.8011695906432747e-05,
404
+ "loss": 0.8106,
405
+ "step": 52
406
+ },
407
+ {
408
+ "epoch": 17.67,
409
+ "learning_rate": 1.7953216374269006e-05,
410
+ "loss": 0.7961,
411
+ "step": 53
412
+ },
413
+ {
414
+ "epoch": 18.0,
415
+ "learning_rate": 1.7894736842105264e-05,
416
+ "loss": 0.7195,
417
+ "step": 54
418
+ },
419
+ {
420
+ "epoch": 18.33,
421
+ "learning_rate": 1.7836257309941522e-05,
422
+ "loss": 0.7777,
423
+ "step": 55
424
+ },
425
+ {
426
+ "epoch": 18.67,
427
+ "learning_rate": 1.7777777777777777e-05,
428
+ "loss": 0.7961,
429
+ "step": 56
430
+ },
431
+ {
432
+ "epoch": 19.0,
433
+ "learning_rate": 1.7719298245614035e-05,
434
+ "loss": 0.9039,
435
+ "step": 57
436
+ },
437
+ {
438
+ "epoch": 19.33,
439
+ "learning_rate": 1.7660818713450293e-05,
440
+ "loss": 0.7125,
441
+ "step": 58
442
+ },
443
+ {
444
+ "epoch": 19.67,
445
+ "learning_rate": 1.760233918128655e-05,
446
+ "loss": 0.7704,
447
+ "step": 59
448
+ },
449
+ {
450
+ "epoch": 20.0,
451
+ "learning_rate": 1.754385964912281e-05,
452
+ "loss": 0.8094,
453
+ "step": 60
454
+ },
455
+ {
456
+ "epoch": 20.0,
457
+ "eval_accuracy_dropoff": 0.5280687848921359,
458
+ "eval_accuracy_undropoff": 0.9608414146478016,
459
+ "eval_accuracy_unlabeled": NaN,
460
+ "eval_iou_dropoff": 0.2866027007818053,
461
+ "eval_iou_undropoff": 0.9402828328280236,
462
+ "eval_iou_unlabeled": 0.0,
463
+ "eval_loss": 0.7488501071929932,
464
+ "eval_mean_accuracy": 0.7444550997699688,
465
+ "eval_mean_iou": 0.40896184453660966,
466
+ "eval_overall_accuracy": 0.9416791280110677,
467
+ "eval_runtime": 1.8293,
468
+ "eval_samples_per_second": 8.2,
469
+ "eval_steps_per_second": 0.547,
470
+ "step": 60
471
+ },
472
+ {
473
+ "epoch": 20.33,
474
+ "learning_rate": 1.7485380116959064e-05,
475
+ "loss": 0.7233,
476
+ "step": 61
477
+ },
478
+ {
479
+ "epoch": 20.67,
480
+ "learning_rate": 1.7426900584795322e-05,
481
+ "loss": 0.7181,
482
+ "step": 62
483
+ },
484
+ {
485
+ "epoch": 21.0,
486
+ "learning_rate": 1.736842105263158e-05,
487
+ "loss": 0.9374,
488
+ "step": 63
489
+ },
490
+ {
491
+ "epoch": 21.33,
492
+ "learning_rate": 1.7309941520467838e-05,
493
+ "loss": 0.7476,
494
+ "step": 64
495
+ },
496
+ {
497
+ "epoch": 21.67,
498
+ "learning_rate": 1.7251461988304093e-05,
499
+ "loss": 0.7718,
500
+ "step": 65
501
+ },
502
+ {
503
+ "epoch": 22.0,
504
+ "learning_rate": 1.719298245614035e-05,
505
+ "loss": 0.7803,
506
+ "step": 66
507
+ },
508
+ {
509
+ "epoch": 22.33,
510
+ "learning_rate": 1.713450292397661e-05,
511
+ "loss": 0.761,
512
+ "step": 67
513
+ },
514
+ {
515
+ "epoch": 22.67,
516
+ "learning_rate": 1.7076023391812867e-05,
517
+ "loss": 0.6901,
518
+ "step": 68
519
+ },
520
+ {
521
+ "epoch": 23.0,
522
+ "learning_rate": 1.7017543859649125e-05,
523
+ "loss": 0.9088,
524
+ "step": 69
525
+ },
526
+ {
527
+ "epoch": 23.33,
528
+ "learning_rate": 1.695906432748538e-05,
529
+ "loss": 0.6945,
530
+ "step": 70
531
+ },
532
+ {
533
+ "epoch": 23.33,
534
+ "eval_accuracy_dropoff": 0.5298435453856227,
535
+ "eval_accuracy_undropoff": 0.9645688777057901,
536
+ "eval_accuracy_unlabeled": NaN,
537
+ "eval_iou_dropoff": 0.30036630036630035,
538
+ "eval_iou_undropoff": 0.9440065001523473,
539
+ "eval_iou_unlabeled": 0.0,
540
+ "eval_loss": 0.7005425095558167,
541
+ "eval_mean_accuracy": 0.7472062115457064,
542
+ "eval_mean_iou": 0.4147909335062159,
543
+ "eval_overall_accuracy": 0.9453201293945312,
544
+ "eval_runtime": 1.7223,
545
+ "eval_samples_per_second": 8.71,
546
+ "eval_steps_per_second": 0.581,
547
+ "step": 70
548
+ },
549
+ {
550
+ "epoch": 23.67,
551
+ "learning_rate": 1.690058479532164e-05,
552
+ "loss": 0.6711,
553
+ "step": 71
554
+ },
555
+ {
556
+ "epoch": 24.0,
557
+ "learning_rate": 1.6842105263157896e-05,
558
+ "loss": 0.7554,
559
+ "step": 72
560
+ },
561
+ {
562
+ "epoch": 24.33,
563
+ "learning_rate": 1.6783625730994155e-05,
564
+ "loss": 0.6708,
565
+ "step": 73
566
+ },
567
+ {
568
+ "epoch": 24.67,
569
+ "learning_rate": 1.672514619883041e-05,
570
+ "loss": 0.6926,
571
+ "step": 74
572
+ },
573
+ {
574
+ "epoch": 25.0,
575
+ "learning_rate": 1.6666666666666667e-05,
576
+ "loss": 0.9247,
577
+ "step": 75
578
+ },
579
+ {
580
+ "epoch": 25.33,
581
+ "learning_rate": 1.6608187134502926e-05,
582
+ "loss": 0.6425,
583
+ "step": 76
584
+ },
585
+ {
586
+ "epoch": 25.67,
587
+ "learning_rate": 1.6549707602339184e-05,
588
+ "loss": 0.6567,
589
+ "step": 77
590
+ },
591
+ {
592
+ "epoch": 26.0,
593
+ "learning_rate": 1.649122807017544e-05,
594
+ "loss": 0.8971,
595
+ "step": 78
596
+ },
597
+ {
598
+ "epoch": 26.33,
599
+ "learning_rate": 1.6432748538011697e-05,
600
+ "loss": 0.6993,
601
+ "step": 79
602
+ },
603
+ {
604
+ "epoch": 26.67,
605
+ "learning_rate": 1.6374269005847955e-05,
606
+ "loss": 0.6337,
607
+ "step": 80
608
+ },
609
+ {
610
+ "epoch": 26.67,
611
+ "eval_accuracy_dropoff": 0.49581294369012335,
612
+ "eval_accuracy_undropoff": 0.9709405298276873,
613
+ "eval_accuracy_unlabeled": NaN,
614
+ "eval_iou_dropoff": 0.3046961862238145,
615
+ "eval_iou_undropoff": 0.9487783537255239,
616
+ "eval_iou_unlabeled": NaN,
617
+ "eval_loss": 0.6331071257591248,
618
+ "eval_mean_accuracy": 0.7333767367589054,
619
+ "eval_mean_iou": 0.6267372699746692,
620
+ "eval_overall_accuracy": 0.9499028523763021,
621
+ "eval_runtime": 1.7079,
622
+ "eval_samples_per_second": 8.783,
623
+ "eval_steps_per_second": 0.586,
624
+ "step": 80
625
+ },
626
+ {
627
+ "epoch": 27.0,
628
+ "learning_rate": 1.6315789473684213e-05,
629
+ "loss": 0.7322,
630
+ "step": 81
631
+ },
632
+ {
633
+ "epoch": 27.33,
634
+ "learning_rate": 1.625730994152047e-05,
635
+ "loss": 0.6814,
636
+ "step": 82
637
+ },
638
+ {
639
+ "epoch": 27.67,
640
+ "learning_rate": 1.6198830409356726e-05,
641
+ "loss": 0.6695,
642
+ "step": 83
643
+ },
644
+ {
645
+ "epoch": 28.0,
646
+ "learning_rate": 1.6140350877192984e-05,
647
+ "loss": 0.6004,
648
+ "step": 84
649
+ },
650
+ {
651
+ "epoch": 28.33,
652
+ "learning_rate": 1.6081871345029242e-05,
653
+ "loss": 0.647,
654
+ "step": 85
655
+ },
656
+ {
657
+ "epoch": 28.67,
658
+ "learning_rate": 1.60233918128655e-05,
659
+ "loss": 0.5803,
660
+ "step": 86
661
+ },
662
+ {
663
+ "epoch": 29.0,
664
+ "learning_rate": 1.5964912280701755e-05,
665
+ "loss": 0.8716,
666
+ "step": 87
667
+ },
668
+ {
669
+ "epoch": 29.33,
670
+ "learning_rate": 1.5906432748538013e-05,
671
+ "loss": 0.6239,
672
+ "step": 88
673
+ },
674
+ {
675
+ "epoch": 29.67,
676
+ "learning_rate": 1.584795321637427e-05,
677
+ "loss": 0.597,
678
+ "step": 89
679
+ },
680
+ {
681
+ "epoch": 30.0,
682
+ "learning_rate": 1.578947368421053e-05,
683
+ "loss": 0.603,
684
+ "step": 90
685
+ },
686
+ {
687
+ "epoch": 30.0,
688
+ "eval_accuracy_dropoff": 0.40565625933328736,
689
+ "eval_accuracy_undropoff": 0.9813906247172738,
690
+ "eval_accuracy_unlabeled": NaN,
691
+ "eval_iou_dropoff": 0.28940801416143874,
692
+ "eval_iou_undropoff": 0.9550916579222961,
693
+ "eval_iou_unlabeled": NaN,
694
+ "eval_loss": 0.5726452469825745,
695
+ "eval_mean_accuracy": 0.6935234420252806,
696
+ "eval_mean_iou": 0.6222498360418675,
697
+ "eval_overall_accuracy": 0.9558982849121094,
698
+ "eval_runtime": 1.7573,
699
+ "eval_samples_per_second": 8.536,
700
+ "eval_steps_per_second": 0.569,
701
+ "step": 90
702
+ },
703
+ {
704
+ "epoch": 30.33,
705
+ "learning_rate": 1.5730994152046787e-05,
706
+ "loss": 0.5913,
707
+ "step": 91
708
+ },
709
+ {
710
+ "epoch": 30.67,
711
+ "learning_rate": 1.5672514619883042e-05,
712
+ "loss": 0.6683,
713
+ "step": 92
714
+ },
715
+ {
716
+ "epoch": 31.0,
717
+ "learning_rate": 1.56140350877193e-05,
718
+ "loss": 0.8263,
719
+ "step": 93
720
+ },
721
+ {
722
+ "epoch": 31.33,
723
+ "learning_rate": 1.555555555555556e-05,
724
+ "loss": 0.5791,
725
+ "step": 94
726
+ },
727
+ {
728
+ "epoch": 31.67,
729
+ "learning_rate": 1.5497076023391816e-05,
730
+ "loss": 0.6118,
731
+ "step": 95
732
+ },
733
+ {
734
+ "epoch": 32.0,
735
+ "learning_rate": 1.543859649122807e-05,
736
+ "loss": 0.6546,
737
+ "step": 96
738
+ },
739
+ {
740
+ "epoch": 32.33,
741
+ "learning_rate": 1.538011695906433e-05,
742
+ "loss": 0.6008,
743
+ "step": 97
744
+ },
745
+ {
746
+ "epoch": 32.67,
747
+ "learning_rate": 1.5321637426900587e-05,
748
+ "loss": 0.5931,
749
+ "step": 98
750
+ },
751
+ {
752
+ "epoch": 33.0,
753
+ "learning_rate": 1.5263157894736846e-05,
754
+ "loss": 0.8761,
755
+ "step": 99
756
+ },
757
+ {
758
+ "epoch": 33.33,
759
+ "learning_rate": 1.52046783625731e-05,
760
+ "loss": 0.5903,
761
+ "step": 100
762
+ },
763
+ {
764
+ "epoch": 33.33,
765
+ "eval_accuracy_dropoff": 0.4546373515289361,
766
+ "eval_accuracy_undropoff": 0.9756530777115378,
767
+ "eval_accuracy_unlabeled": NaN,
768
+ "eval_iou_dropoff": 0.2980214980892679,
769
+ "eval_iou_undropoff": 0.9516094455131633,
770
+ "eval_iou_unlabeled": NaN,
771
+ "eval_loss": 0.5840517282485962,
772
+ "eval_mean_accuracy": 0.715145214620237,
773
+ "eval_mean_iou": 0.6248154718012155,
774
+ "eval_overall_accuracy": 0.9525835673014323,
775
+ "eval_runtime": 1.6835,
776
+ "eval_samples_per_second": 8.91,
777
+ "eval_steps_per_second": 0.594,
778
+ "step": 100
779
+ },
780
+ {
781
+ "epoch": 33.67,
782
+ "learning_rate": 1.5146198830409358e-05,
783
+ "loss": 0.5691,
784
+ "step": 101
785
+ },
786
+ {
787
+ "epoch": 34.0,
788
+ "learning_rate": 1.5087719298245615e-05,
789
+ "loss": 0.8078,
790
+ "step": 102
791
+ },
792
+ {
793
+ "epoch": 34.33,
794
+ "learning_rate": 1.5029239766081873e-05,
795
+ "loss": 0.5583,
796
+ "step": 103
797
+ },
798
+ {
799
+ "epoch": 34.67,
800
+ "learning_rate": 1.497076023391813e-05,
801
+ "loss": 0.5857,
802
+ "step": 104
803
+ },
804
+ {
805
+ "epoch": 35.0,
806
+ "learning_rate": 1.4912280701754388e-05,
807
+ "loss": 0.5148,
808
+ "step": 105
809
+ },
810
+ {
811
+ "epoch": 35.33,
812
+ "learning_rate": 1.4853801169590644e-05,
813
+ "loss": 0.5866,
814
+ "step": 106
815
+ },
816
+ {
817
+ "epoch": 35.67,
818
+ "learning_rate": 1.4795321637426902e-05,
819
+ "loss": 0.5276,
820
+ "step": 107
821
+ },
822
+ {
823
+ "epoch": 36.0,
824
+ "learning_rate": 1.4736842105263159e-05,
825
+ "loss": 0.8399,
826
+ "step": 108
827
+ },
828
+ {
829
+ "epoch": 36.33,
830
+ "learning_rate": 1.4678362573099417e-05,
831
+ "loss": 0.5761,
832
+ "step": 109
833
+ },
834
+ {
835
+ "epoch": 36.67,
836
+ "learning_rate": 1.4619883040935675e-05,
837
+ "loss": 0.5514,
838
+ "step": 110
839
+ },
840
+ {
841
+ "epoch": 36.67,
842
+ "eval_accuracy_dropoff": 0.3781273692191054,
843
+ "eval_accuracy_undropoff": 0.9853985522286546,
844
+ "eval_accuracy_unlabeled": NaN,
845
+ "eval_iou_dropoff": 0.28751293775466086,
846
+ "eval_iou_undropoff": 0.9578033403813699,
847
+ "eval_iou_unlabeled": NaN,
848
+ "eval_loss": 0.5156698822975159,
849
+ "eval_mean_accuracy": 0.68176296072388,
850
+ "eval_mean_iou": 0.6226581390680154,
851
+ "eval_overall_accuracy": 0.9585098266601563,
852
+ "eval_runtime": 1.8267,
853
+ "eval_samples_per_second": 8.212,
854
+ "eval_steps_per_second": 0.547,
855
+ "step": 110
856
+ },
857
+ {
858
+ "epoch": 37.0,
859
+ "learning_rate": 1.4561403508771931e-05,
860
+ "loss": 0.6281,
861
+ "step": 111
862
+ },
863
+ {
864
+ "epoch": 37.33,
865
+ "learning_rate": 1.4502923976608188e-05,
866
+ "loss": 0.5795,
867
+ "step": 112
868
+ },
869
+ {
870
+ "epoch": 37.67,
871
+ "learning_rate": 1.4444444444444446e-05,
872
+ "loss": 0.5427,
873
+ "step": 113
874
+ },
875
+ {
876
+ "epoch": 38.0,
877
+ "learning_rate": 1.4385964912280704e-05,
878
+ "loss": 0.8123,
879
+ "step": 114
880
+ },
881
+ {
882
+ "epoch": 38.33,
883
+ "learning_rate": 1.432748538011696e-05,
884
+ "loss": 0.5282,
885
+ "step": 115
886
+ },
887
+ {
888
+ "epoch": 38.67,
889
+ "learning_rate": 1.4269005847953217e-05,
890
+ "loss": 0.6357,
891
+ "step": 116
892
+ },
893
+ {
894
+ "epoch": 39.0,
895
+ "learning_rate": 1.4210526315789475e-05,
896
+ "loss": 0.5648,
897
+ "step": 117
898
+ },
899
+ {
900
+ "epoch": 39.33,
901
+ "learning_rate": 1.4152046783625733e-05,
902
+ "loss": 0.5177,
903
+ "step": 118
904
+ },
905
+ {
906
+ "epoch": 39.67,
907
+ "learning_rate": 1.409356725146199e-05,
908
+ "loss": 0.5163,
909
+ "step": 119
910
+ },
911
+ {
912
+ "epoch": 40.0,
913
+ "learning_rate": 1.4035087719298246e-05,
914
+ "loss": 0.6464,
915
+ "step": 120
916
+ },
917
+ {
918
+ "epoch": 40.0,
919
+ "eval_accuracy_dropoff": 0.3941002136604866,
920
+ "eval_accuracy_undropoff": 0.983633542058492,
921
+ "eval_accuracy_unlabeled": NaN,
922
+ "eval_iou_dropoff": 0.29122208357737656,
923
+ "eval_iou_undropoff": 0.9567759549263578,
924
+ "eval_iou_unlabeled": NaN,
925
+ "eval_loss": 0.5141220092773438,
926
+ "eval_mean_accuracy": 0.6888668778594893,
927
+ "eval_mean_iou": 0.6239990192518672,
928
+ "eval_overall_accuracy": 0.9575302124023437,
929
+ "eval_runtime": 1.7894,
930
+ "eval_samples_per_second": 8.383,
931
+ "eval_steps_per_second": 0.559,
932
+ "step": 120
933
+ },
934
+ {
935
+ "epoch": 40.33,
936
+ "learning_rate": 1.3976608187134504e-05,
937
+ "loss": 0.5223,
938
+ "step": 121
939
+ },
940
+ {
941
+ "epoch": 40.67,
942
+ "learning_rate": 1.3918128654970762e-05,
943
+ "loss": 0.5158,
944
+ "step": 122
945
+ },
946
+ {
947
+ "epoch": 41.0,
948
+ "learning_rate": 1.385964912280702e-05,
949
+ "loss": 0.7741,
950
+ "step": 123
951
+ },
952
+ {
953
+ "epoch": 41.33,
954
+ "learning_rate": 1.3801169590643275e-05,
955
+ "loss": 0.5087,
956
+ "step": 124
957
+ },
958
+ {
959
+ "epoch": 41.67,
960
+ "learning_rate": 1.3742690058479533e-05,
961
+ "loss": 0.5203,
962
+ "step": 125
963
+ },
964
+ {
965
+ "epoch": 42.0,
966
+ "learning_rate": 1.3684210526315791e-05,
967
+ "loss": 0.5061,
968
+ "step": 126
969
+ },
970
+ {
971
+ "epoch": 42.33,
972
+ "learning_rate": 1.362573099415205e-05,
973
+ "loss": 0.4992,
974
+ "step": 127
975
+ },
976
+ {
977
+ "epoch": 42.67,
978
+ "learning_rate": 1.3567251461988304e-05,
979
+ "loss": 0.5293,
980
+ "step": 128
981
+ },
982
+ {
983
+ "epoch": 43.0,
984
+ "learning_rate": 1.3508771929824562e-05,
985
+ "loss": 0.7778,
986
+ "step": 129
987
+ },
988
+ {
989
+ "epoch": 43.33,
990
+ "learning_rate": 1.345029239766082e-05,
991
+ "loss": 0.5198,
992
+ "step": 130
993
+ },
994
+ {
995
+ "epoch": 43.33,
996
+ "eval_accuracy_dropoff": 0.3656810715188274,
997
+ "eval_accuracy_undropoff": 0.9866212601635103,
998
+ "eval_accuracy_unlabeled": NaN,
999
+ "eval_iou_dropoff": 0.28375204453179664,
1000
+ "eval_iou_undropoff": 0.9584546122881992,
1001
+ "eval_iou_unlabeled": 0.0,
1002
+ "eval_loss": 0.48896968364715576,
1003
+ "eval_mean_accuracy": 0.6761511658411689,
1004
+ "eval_mean_iou": 0.41406888560666527,
1005
+ "eval_overall_accuracy": 0.9591272989908854,
1006
+ "eval_runtime": 1.7404,
1007
+ "eval_samples_per_second": 8.619,
1008
+ "eval_steps_per_second": 0.575,
1009
+ "step": 130
1010
+ },
1011
+ {
1012
+ "epoch": 43.67,
1013
+ "learning_rate": 1.3391812865497079e-05,
1014
+ "loss": 0.4739,
1015
+ "step": 131
1016
+ },
1017
+ {
1018
+ "epoch": 44.0,
1019
+ "learning_rate": 1.3333333333333333e-05,
1020
+ "loss": 0.7511,
1021
+ "step": 132
1022
+ },
1023
+ {
1024
+ "epoch": 44.33,
1025
+ "learning_rate": 1.3274853801169591e-05,
1026
+ "loss": 0.4826,
1027
+ "step": 133
1028
+ },
1029
+ {
1030
+ "epoch": 44.67,
1031
+ "learning_rate": 1.321637426900585e-05,
1032
+ "loss": 0.5534,
1033
+ "step": 134
1034
+ },
1035
+ {
1036
+ "epoch": 45.0,
1037
+ "learning_rate": 1.3157894736842108e-05,
1038
+ "loss": 0.5636,
1039
+ "step": 135
1040
+ },
1041
+ {
1042
+ "epoch": 45.33,
1043
+ "learning_rate": 1.3099415204678362e-05,
1044
+ "loss": 0.5102,
1045
+ "step": 136
1046
+ },
1047
+ {
1048
+ "epoch": 45.67,
1049
+ "learning_rate": 1.304093567251462e-05,
1050
+ "loss": 0.485,
1051
+ "step": 137
1052
+ },
1053
+ {
1054
+ "epoch": 46.0,
1055
+ "learning_rate": 1.2982456140350879e-05,
1056
+ "loss": 0.486,
1057
+ "step": 138
1058
+ },
1059
+ {
1060
+ "epoch": 46.33,
1061
+ "learning_rate": 1.2923976608187137e-05,
1062
+ "loss": 0.4934,
1063
+ "step": 139
1064
+ },
1065
+ {
1066
+ "epoch": 46.67,
1067
+ "learning_rate": 1.2865497076023392e-05,
1068
+ "loss": 0.5077,
1069
+ "step": 140
1070
+ },
1071
+ {
1072
+ "epoch": 46.67,
1073
+ "eval_accuracy_dropoff": 0.3572322925999954,
1074
+ "eval_accuracy_undropoff": 0.986647337503579,
1075
+ "eval_accuracy_unlabeled": NaN,
1076
+ "eval_iou_dropoff": 0.27731853040841803,
1077
+ "eval_iou_undropoff": 0.95811562022686,
1078
+ "eval_iou_unlabeled": 0.0,
1079
+ "eval_loss": 0.48549026250839233,
1080
+ "eval_mean_accuracy": 0.6719398150517872,
1081
+ "eval_mean_iou": 0.4118113835450927,
1082
+ "eval_overall_accuracy": 0.9587781270345052,
1083
+ "eval_runtime": 1.7229,
1084
+ "eval_samples_per_second": 8.706,
1085
+ "eval_steps_per_second": 0.58,
1086
+ "step": 140
1087
+ },
1088
+ {
1089
+ "epoch": 47.0,
1090
+ "learning_rate": 1.280701754385965e-05,
1091
+ "loss": 0.647,
1092
+ "step": 141
1093
+ },
1094
+ {
1095
+ "epoch": 47.33,
1096
+ "learning_rate": 1.2748538011695908e-05,
1097
+ "loss": 0.4891,
1098
+ "step": 142
1099
+ },
1100
+ {
1101
+ "epoch": 47.67,
1102
+ "learning_rate": 1.2690058479532166e-05,
1103
+ "loss": 0.4703,
1104
+ "step": 143
1105
+ },
1106
+ {
1107
+ "epoch": 48.0,
1108
+ "learning_rate": 1.263157894736842e-05,
1109
+ "loss": 0.7596,
1110
+ "step": 144
1111
+ },
1112
+ {
1113
+ "epoch": 48.33,
1114
+ "learning_rate": 1.2573099415204679e-05,
1115
+ "loss": 0.4825,
1116
+ "step": 145
1117
+ },
1118
+ {
1119
+ "epoch": 48.67,
1120
+ "learning_rate": 1.2514619883040937e-05,
1121
+ "loss": 0.4702,
1122
+ "step": 146
1123
+ },
1124
+ {
1125
+ "epoch": 49.0,
1126
+ "learning_rate": 1.2456140350877195e-05,
1127
+ "loss": 0.6684,
1128
+ "step": 147
1129
+ },
1130
+ {
1131
+ "epoch": 49.33,
1132
+ "learning_rate": 1.239766081871345e-05,
1133
+ "loss": 0.4681,
1134
+ "step": 148
1135
+ },
1136
+ {
1137
+ "epoch": 49.67,
1138
+ "learning_rate": 1.2339181286549708e-05,
1139
+ "loss": 0.497,
1140
+ "step": 149
1141
+ },
1142
+ {
1143
+ "epoch": 50.0,
1144
+ "learning_rate": 1.2280701754385966e-05,
1145
+ "loss": 0.4817,
1146
+ "step": 150
1147
+ },
1148
+ {
1149
+ "epoch": 50.0,
1150
+ "eval_accuracy_dropoff": 0.36020745744020954,
1151
+ "eval_accuracy_undropoff": 0.9863945469620963,
1152
+ "eval_accuracy_unlabeled": NaN,
1153
+ "eval_iou_dropoff": 0.27843880695086975,
1154
+ "eval_iou_undropoff": 0.9579983692751803,
1155
+ "eval_iou_unlabeled": NaN,
1156
+ "eval_loss": 0.47096526622772217,
1157
+ "eval_mean_accuracy": 0.673301002201153,
1158
+ "eval_mean_iou": 0.6182185881130251,
1159
+ "eval_overall_accuracy": 0.9586682637532552,
1160
+ "eval_runtime": 1.7512,
1161
+ "eval_samples_per_second": 8.565,
1162
+ "eval_steps_per_second": 0.571,
1163
+ "step": 150
1164
+ },
1165
+ {
1166
+ "epoch": 50.33,
1167
+ "learning_rate": 1.2222222222222224e-05,
1168
+ "loss": 0.4865,
1169
+ "step": 151
1170
+ },
1171
+ {
1172
+ "epoch": 50.67,
1173
+ "learning_rate": 1.216374269005848e-05,
1174
+ "loss": 0.4611,
1175
+ "step": 152
1176
+ },
1177
+ {
1178
+ "epoch": 51.0,
1179
+ "learning_rate": 1.2105263157894737e-05,
1180
+ "loss": 0.5011,
1181
+ "step": 153
1182
+ },
1183
+ {
1184
+ "epoch": 51.33,
1185
+ "learning_rate": 1.2046783625730995e-05,
1186
+ "loss": 0.4967,
1187
+ "step": 154
1188
+ },
1189
+ {
1190
+ "epoch": 51.67,
1191
+ "learning_rate": 1.1988304093567253e-05,
1192
+ "loss": 0.4445,
1193
+ "step": 155
1194
+ },
1195
+ {
1196
+ "epoch": 52.0,
1197
+ "learning_rate": 1.192982456140351e-05,
1198
+ "loss": 0.7343,
1199
+ "step": 156
1200
+ },
1201
+ {
1202
+ "epoch": 52.33,
1203
+ "learning_rate": 1.1871345029239766e-05,
1204
+ "loss": 0.4597,
1205
+ "step": 157
1206
+ },
1207
+ {
1208
+ "epoch": 52.67,
1209
+ "learning_rate": 1.1812865497076024e-05,
1210
+ "loss": 0.5019,
1211
+ "step": 158
1212
+ },
1213
+ {
1214
+ "epoch": 53.0,
1215
+ "learning_rate": 1.1754385964912282e-05,
1216
+ "loss": 0.4556,
1217
+ "step": 159
1218
+ },
1219
+ {
1220
+ "epoch": 53.33,
1221
+ "learning_rate": 1.1695906432748539e-05,
1222
+ "loss": 0.4713,
1223
+ "step": 160
1224
+ },
1225
+ {
1226
+ "epoch": 53.33,
1227
+ "eval_accuracy_dropoff": 0.3479047487766214,
1228
+ "eval_accuracy_undropoff": 0.9886552926888718,
1229
+ "eval_accuracy_unlabeled": NaN,
1230
+ "eval_iou_dropoff": 0.27947052255677257,
1231
+ "eval_iou_undropoff": 0.9596627946110987,
1232
+ "eval_iou_unlabeled": NaN,
1233
+ "eval_loss": 0.46686360239982605,
1234
+ "eval_mean_accuracy": 0.6682800207327466,
1235
+ "eval_mean_iou": 0.6195666585839357,
1236
+ "eval_overall_accuracy": 0.960284169514974,
1237
+ "eval_runtime": 1.8139,
1238
+ "eval_samples_per_second": 8.27,
1239
+ "eval_steps_per_second": 0.551,
1240
+ "step": 160
1241
+ },
1242
+ {
1243
+ "epoch": 53.67,
1244
+ "learning_rate": 1.1637426900584797e-05,
1245
+ "loss": 0.4486,
1246
+ "step": 161
1247
+ },
1248
+ {
1249
+ "epoch": 54.0,
1250
+ "learning_rate": 1.1578947368421053e-05,
1251
+ "loss": 0.7127,
1252
+ "step": 162
1253
+ },
1254
+ {
1255
+ "epoch": 54.33,
1256
+ "learning_rate": 1.1520467836257312e-05,
1257
+ "loss": 0.4508,
1258
+ "step": 163
1259
+ },
1260
+ {
1261
+ "epoch": 54.67,
1262
+ "learning_rate": 1.1461988304093568e-05,
1263
+ "loss": 0.4438,
1264
+ "step": 164
1265
+ },
1266
+ {
1267
+ "epoch": 55.0,
1268
+ "learning_rate": 1.1403508771929826e-05,
1269
+ "loss": 0.464,
1270
+ "step": 165
1271
+ },
1272
+ {
1273
+ "epoch": 55.33,
1274
+ "learning_rate": 1.1345029239766083e-05,
1275
+ "loss": 0.4426,
1276
+ "step": 166
1277
+ },
1278
+ {
1279
+ "epoch": 55.67,
1280
+ "learning_rate": 1.128654970760234e-05,
1281
+ "loss": 0.4411,
1282
+ "step": 167
1283
+ },
1284
+ {
1285
+ "epoch": 56.0,
1286
+ "learning_rate": 1.1228070175438597e-05,
1287
+ "loss": 0.71,
1288
+ "step": 168
1289
+ },
1290
+ {
1291
+ "epoch": 56.33,
1292
+ "learning_rate": 1.1169590643274855e-05,
1293
+ "loss": 0.4387,
1294
+ "step": 169
1295
+ },
1296
+ {
1297
+ "epoch": 56.67,
1298
+ "learning_rate": 1.1111111111111113e-05,
1299
+ "loss": 0.4516,
1300
+ "step": 170
1301
+ },
1302
+ {
1303
+ "epoch": 56.67,
1304
+ "eval_accuracy_dropoff": 0.3265214694327659,
1305
+ "eval_accuracy_undropoff": 0.9906150846236295,
1306
+ "eval_accuracy_unlabeled": NaN,
1307
+ "eval_iou_dropoff": 0.2715210912425493,
1308
+ "eval_iou_undropoff": 0.9606413422410078,
1309
+ "eval_iou_unlabeled": 0.0,
1310
+ "eval_loss": 0.4485587179660797,
1311
+ "eval_mean_accuracy": 0.6585682770281976,
1312
+ "eval_mean_iou": 0.41072081116118575,
1313
+ "eval_overall_accuracy": 0.9612103780110677,
1314
+ "eval_runtime": 1.7967,
1315
+ "eval_samples_per_second": 8.349,
1316
+ "eval_steps_per_second": 0.557,
1317
+ "step": 170
1318
+ },
1319
+ {
1320
+ "epoch": 57.0,
1321
+ "learning_rate": 1.105263157894737e-05,
1322
+ "loss": 0.6996,
1323
+ "step": 171
1324
+ },
1325
+ {
1326
+ "epoch": 57.33,
1327
+ "learning_rate": 1.0994152046783626e-05,
1328
+ "loss": 0.456,
1329
+ "step": 172
1330
+ },
1331
+ {
1332
+ "epoch": 57.67,
1333
+ "learning_rate": 1.0935672514619884e-05,
1334
+ "loss": 0.4341,
1335
+ "step": 173
1336
+ },
1337
+ {
1338
+ "epoch": 58.0,
1339
+ "learning_rate": 1.0877192982456142e-05,
1340
+ "loss": 0.7256,
1341
+ "step": 174
1342
+ },
1343
+ {
1344
+ "epoch": 58.33,
1345
+ "learning_rate": 1.0818713450292399e-05,
1346
+ "loss": 0.4225,
1347
+ "step": 175
1348
+ },
1349
+ {
1350
+ "epoch": 58.67,
1351
+ "learning_rate": 1.0760233918128655e-05,
1352
+ "loss": 0.471,
1353
+ "step": 176
1354
+ },
1355
+ {
1356
+ "epoch": 59.0,
1357
+ "learning_rate": 1.0701754385964913e-05,
1358
+ "loss": 0.4369,
1359
+ "step": 177
1360
+ },
1361
+ {
1362
+ "epoch": 59.33,
1363
+ "learning_rate": 1.0643274853801172e-05,
1364
+ "loss": 0.4481,
1365
+ "step": 178
1366
+ },
1367
+ {
1368
+ "epoch": 59.67,
1369
+ "learning_rate": 1.0584795321637428e-05,
1370
+ "loss": 0.4331,
1371
+ "step": 179
1372
+ },
1373
+ {
1374
+ "epoch": 60.0,
1375
+ "learning_rate": 1.0526315789473684e-05,
1376
+ "loss": 0.4059,
1377
+ "step": 180
1378
+ },
1379
+ {
1380
+ "epoch": 60.0,
1381
+ "eval_accuracy_dropoff": 0.31866427734509617,
1382
+ "eval_accuracy_undropoff": 0.9909349311824317,
1383
+ "eval_accuracy_unlabeled": NaN,
1384
+ "eval_iou_dropoff": 0.26651615227572956,
1385
+ "eval_iou_undropoff": 0.9606124109353421,
1386
+ "eval_iou_unlabeled": NaN,
1387
+ "eval_loss": 0.4360799789428711,
1388
+ "eval_mean_accuracy": 0.6547996042637639,
1389
+ "eval_mean_iou": 0.6135642816055358,
1390
+ "eval_overall_accuracy": 0.9611681620279948,
1391
+ "eval_runtime": 1.686,
1392
+ "eval_samples_per_second": 8.897,
1393
+ "eval_steps_per_second": 0.593,
1394
+ "step": 180
1395
+ },
1396
+ {
1397
+ "epoch": 60.33,
1398
+ "learning_rate": 1.0467836257309943e-05,
1399
+ "loss": 0.4479,
1400
+ "step": 181
1401
+ },
1402
+ {
1403
+ "epoch": 60.67,
1404
+ "learning_rate": 1.04093567251462e-05,
1405
+ "loss": 0.4304,
1406
+ "step": 182
1407
+ },
1408
+ {
1409
+ "epoch": 61.0,
1410
+ "learning_rate": 1.0350877192982459e-05,
1411
+ "loss": 0.4971,
1412
+ "step": 183
1413
+ },
1414
+ {
1415
+ "epoch": 61.33,
1416
+ "learning_rate": 1.0292397660818714e-05,
1417
+ "loss": 0.4302,
1418
+ "step": 184
1419
+ },
1420
+ {
1421
+ "epoch": 61.67,
1422
+ "learning_rate": 1.0233918128654972e-05,
1423
+ "loss": 0.4735,
1424
+ "step": 185
1425
+ },
1426
+ {
1427
+ "epoch": 62.0,
1428
+ "learning_rate": 1.017543859649123e-05,
1429
+ "loss": 0.3974,
1430
+ "step": 186
1431
+ },
1432
+ {
1433
+ "epoch": 62.33,
1434
+ "learning_rate": 1.0116959064327488e-05,
1435
+ "loss": 0.4172,
1436
+ "step": 187
1437
+ },
1438
+ {
1439
+ "epoch": 62.67,
1440
+ "learning_rate": 1.0058479532163743e-05,
1441
+ "loss": 0.4458,
1442
+ "step": 188
1443
+ },
1444
+ {
1445
+ "epoch": 63.0,
1446
+ "learning_rate": 1e-05,
1447
+ "loss": 0.6849,
1448
+ "step": 189
1449
+ },
1450
+ {
1451
+ "epoch": 63.33,
1452
+ "learning_rate": 9.941520467836257e-06,
1453
+ "loss": 0.4142,
1454
+ "step": 190
1455
+ },
1456
+ {
1457
+ "epoch": 63.33,
1458
+ "eval_accuracy_dropoff": 0.3088657614813794,
1459
+ "eval_accuracy_undropoff": 0.9917302900545283,
1460
+ "eval_accuracy_unlabeled": NaN,
1461
+ "eval_iou_dropoff": 0.2620841577885431,
1462
+ "eval_iou_undropoff": 0.9609605443916848,
1463
+ "eval_iou_unlabeled": NaN,
1464
+ "eval_loss": 0.42670688033103943,
1465
+ "eval_mean_accuracy": 0.6502980257679538,
1466
+ "eval_mean_iou": 0.611522351090114,
1467
+ "eval_overall_accuracy": 0.9614944458007812,
1468
+ "eval_runtime": 1.7659,
1469
+ "eval_samples_per_second": 8.494,
1470
+ "eval_steps_per_second": 0.566,
1471
+ "step": 190
1472
+ },
1473
+ {
1474
+ "epoch": 63.67,
1475
+ "learning_rate": 9.883040935672515e-06,
1476
+ "loss": 0.4708,
1477
+ "step": 191
1478
+ },
1479
+ {
1480
+ "epoch": 64.0,
1481
+ "learning_rate": 9.824561403508772e-06,
1482
+ "loss": 0.6601,
1483
+ "step": 192
1484
+ },
1485
+ {
1486
+ "epoch": 64.33,
1487
+ "learning_rate": 9.76608187134503e-06,
1488
+ "loss": 0.4291,
1489
+ "step": 193
1490
+ },
1491
+ {
1492
+ "epoch": 64.67,
1493
+ "learning_rate": 9.707602339181286e-06,
1494
+ "loss": 0.4169,
1495
+ "step": 194
1496
+ },
1497
+ {
1498
+ "epoch": 65.0,
1499
+ "learning_rate": 9.649122807017545e-06,
1500
+ "loss": 0.4025,
1501
+ "step": 195
1502
+ },
1503
+ {
1504
+ "epoch": 65.33,
1505
+ "learning_rate": 9.590643274853801e-06,
1506
+ "loss": 0.415,
1507
+ "step": 196
1508
+ },
1509
+ {
1510
+ "epoch": 65.67,
1511
+ "learning_rate": 9.532163742690059e-06,
1512
+ "loss": 0.4125,
1513
+ "step": 197
1514
+ },
1515
+ {
1516
+ "epoch": 66.0,
1517
+ "learning_rate": 9.473684210526315e-06,
1518
+ "loss": 0.4011,
1519
+ "step": 198
1520
+ },
1521
+ {
1522
+ "epoch": 66.33,
1523
+ "learning_rate": 9.415204678362574e-06,
1524
+ "loss": 0.3921,
1525
+ "step": 199
1526
+ },
1527
+ {
1528
+ "epoch": 66.67,
1529
+ "learning_rate": 9.35672514619883e-06,
1530
+ "loss": 0.4393,
1531
+ "step": 200
1532
+ },
1533
+ {
1534
+ "epoch": 66.67,
1535
+ "eval_accuracy_dropoff": 0.2767822271233947,
1536
+ "eval_accuracy_undropoff": 0.9940290874101795,
1537
+ "eval_accuracy_unlabeled": NaN,
1538
+ "eval_iou_dropoff": 0.24518308597943494,
1539
+ "eval_iou_undropoff": 0.9618027430695911,
1540
+ "eval_iou_unlabeled": NaN,
1541
+ "eval_loss": 0.4188476800918579,
1542
+ "eval_mean_accuracy": 0.6354056572667871,
1543
+ "eval_mean_iou": 0.603492914524513,
1544
+ "eval_overall_accuracy": 0.9622708638509114,
1545
+ "eval_runtime": 1.8501,
1546
+ "eval_samples_per_second": 8.108,
1547
+ "eval_steps_per_second": 0.541,
1548
+ "step": 200
1549
+ }
1550
+ ],
1551
+ "max_steps": 360,
1552
+ "num_train_epochs": 120,
1553
+ "total_flos": 3.989638682247168e+16,
1554
+ "trial_name": null,
1555
+ "trial_params": null
1556
+ }
checkpoint-200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:956bf1976427a2f6d5abaeb4217db08d9dd40b4dfb95afeb46f806f35903c20d
3
+ size 4155
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c663bfae7b4964d248bdb6619226f4c73c2a34202d3e672fa6ea66b9335345f6
3
  size 14932813
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84bf59eeba83524aa2f95d95c0fb00af94b856d9a46f7c9107d87d3865e0dd25
3
  size 14932813
runs/Feb12_06-10-25_robovision.csres.utexas.edu/events.out.tfevents.1707739840.robovision.csres.utexas.edu.2691300.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:be311b5856466da709a8ad71cf450bb63b97d1ec64bb388590512693bc99399a
3
- size 42035
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca17312f008a5f24966bbfdad17d1cb46938a902d923c0d6547c89d201d560c7
3
+ size 51487