sam1120 commited on
Commit
d931e68
1 Parent(s): b440a7c

Training in progress, step 40

Browse files
checkpoint-40/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": 6,
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-40/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34c273537bb1abc8910513c5232fd8c54f0fc71d87a75b8e1215240269a66de9
3
+ size 29876485
checkpoint-40/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b67fc1e9142a1425d3256b17a33712c77775bc46bfde2d9cb053361338550cf4
3
+ size 14951629
checkpoint-40/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12614450c70a1153287aefc455bca8990f9cabf36a348572795d8d1a2ba80518
3
+ size 14575
checkpoint-40/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25237a923eee17c18a7b29086e4a1cf0b19b4cee7a21985adcaa35e9c29c2ac8
3
+ size 627
checkpoint-40/trainer_state.json ADDED
@@ -0,0 +1,324 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.48036208748817444,
3
+ "best_model_checkpoint": "/robodata/smodak/Projects/nspl/scripts/terrainseg/training/models/dropoff-utcustom-train-SF-RGBD-b0_7/checkpoint-40",
4
+ "epoch": 20.0,
5
+ "global_step": 40,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.5,
12
+ "learning_rate": 6.666666666666667e-06,
13
+ "loss": 1.0264,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 1.0,
18
+ "learning_rate": 1.3333333333333333e-05,
19
+ "loss": 1.0249,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 1.5,
24
+ "learning_rate": 2e-05,
25
+ "loss": 1.0219,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 2.0,
30
+ "learning_rate": 2.6666666666666667e-05,
31
+ "loss": 1.0261,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 2.5,
36
+ "learning_rate": 3.3333333333333335e-05,
37
+ "loss": 1.0043,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 3.0,
42
+ "learning_rate": 4e-05,
43
+ "loss": 0.9882,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 3.5,
48
+ "learning_rate": 4.666666666666667e-05,
49
+ "loss": 0.9908,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 4.0,
54
+ "learning_rate": 5.333333333333333e-05,
55
+ "loss": 0.9709,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 4.5,
60
+ "learning_rate": 6.000000000000001e-05,
61
+ "loss": 0.961,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 5.0,
66
+ "learning_rate": 6.666666666666667e-05,
67
+ "loss": 0.9508,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 5.0,
72
+ "eval_accuracy_dropoff": 0.19369710314456753,
73
+ "eval_accuracy_undropoff": 0.9010801242217483,
74
+ "eval_accuracy_unlabeled": NaN,
75
+ "eval_iou_dropoff": 0.06049444267748618,
76
+ "eval_iou_undropoff": 0.8706257192603624,
77
+ "eval_iou_unlabeled": 0.0,
78
+ "eval_loss": 1.026282548904419,
79
+ "eval_mean_accuracy": 0.5473886136831579,
80
+ "eval_mean_iou": 0.31037338731261616,
81
+ "eval_overall_accuracy": 0.8716676712036133,
82
+ "eval_runtime": 2.7665,
83
+ "eval_samples_per_second": 7.229,
84
+ "eval_steps_per_second": 0.723,
85
+ "step": 10
86
+ },
87
+ {
88
+ "epoch": 5.5,
89
+ "learning_rate": 7.333333333333333e-05,
90
+ "loss": 0.9292,
91
+ "step": 11
92
+ },
93
+ {
94
+ "epoch": 6.0,
95
+ "learning_rate": 8e-05,
96
+ "loss": 0.9246,
97
+ "step": 12
98
+ },
99
+ {
100
+ "epoch": 6.5,
101
+ "learning_rate": 7.964912280701755e-05,
102
+ "loss": 0.9097,
103
+ "step": 13
104
+ },
105
+ {
106
+ "epoch": 7.0,
107
+ "learning_rate": 7.929824561403509e-05,
108
+ "loss": 0.8786,
109
+ "step": 14
110
+ },
111
+ {
112
+ "epoch": 7.5,
113
+ "learning_rate": 7.894736842105263e-05,
114
+ "loss": 0.8697,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 8.0,
119
+ "learning_rate": 7.859649122807017e-05,
120
+ "loss": 0.8421,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 8.5,
125
+ "learning_rate": 7.824561403508773e-05,
126
+ "loss": 0.8223,
127
+ "step": 17
128
+ },
129
+ {
130
+ "epoch": 9.0,
131
+ "learning_rate": 7.789473684210527e-05,
132
+ "loss": 0.8166,
133
+ "step": 18
134
+ },
135
+ {
136
+ "epoch": 9.5,
137
+ "learning_rate": 7.754385964912281e-05,
138
+ "loss": 0.7987,
139
+ "step": 19
140
+ },
141
+ {
142
+ "epoch": 10.0,
143
+ "learning_rate": 7.719298245614036e-05,
144
+ "loss": 0.7814,
145
+ "step": 20
146
+ },
147
+ {
148
+ "epoch": 10.0,
149
+ "eval_accuracy_dropoff": 0.09518566939608707,
150
+ "eval_accuracy_undropoff": 0.9726282293027602,
151
+ "eval_accuracy_unlabeled": NaN,
152
+ "eval_iou_dropoff": 0.05836274909643214,
153
+ "eval_iou_undropoff": 0.9358911019066932,
154
+ "eval_iou_unlabeled": NaN,
155
+ "eval_loss": 0.7567641139030457,
156
+ "eval_mean_accuracy": 0.5339069493494236,
157
+ "eval_mean_iou": 0.4971269255015627,
158
+ "eval_overall_accuracy": 0.9361448287963867,
159
+ "eval_runtime": 2.8406,
160
+ "eval_samples_per_second": 7.041,
161
+ "eval_steps_per_second": 0.704,
162
+ "step": 20
163
+ },
164
+ {
165
+ "epoch": 10.5,
166
+ "learning_rate": 7.68421052631579e-05,
167
+ "loss": 0.7775,
168
+ "step": 21
169
+ },
170
+ {
171
+ "epoch": 11.0,
172
+ "learning_rate": 7.649122807017544e-05,
173
+ "loss": 0.744,
174
+ "step": 22
175
+ },
176
+ {
177
+ "epoch": 11.5,
178
+ "learning_rate": 7.614035087719298e-05,
179
+ "loss": 0.7385,
180
+ "step": 23
181
+ },
182
+ {
183
+ "epoch": 12.0,
184
+ "learning_rate": 7.578947368421052e-05,
185
+ "loss": 0.7156,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 12.5,
190
+ "learning_rate": 7.543859649122808e-05,
191
+ "loss": 0.7073,
192
+ "step": 25
193
+ },
194
+ {
195
+ "epoch": 13.0,
196
+ "learning_rate": 7.508771929824562e-05,
197
+ "loss": 0.6876,
198
+ "step": 26
199
+ },
200
+ {
201
+ "epoch": 13.5,
202
+ "learning_rate": 7.473684210526316e-05,
203
+ "loss": 0.68,
204
+ "step": 27
205
+ },
206
+ {
207
+ "epoch": 14.0,
208
+ "learning_rate": 7.43859649122807e-05,
209
+ "loss": 0.6647,
210
+ "step": 28
211
+ },
212
+ {
213
+ "epoch": 14.5,
214
+ "learning_rate": 7.403508771929825e-05,
215
+ "loss": 0.656,
216
+ "step": 29
217
+ },
218
+ {
219
+ "epoch": 15.0,
220
+ "learning_rate": 7.368421052631579e-05,
221
+ "loss": 0.642,
222
+ "step": 30
223
+ },
224
+ {
225
+ "epoch": 15.0,
226
+ "eval_accuracy_dropoff": 0.10256657262781256,
227
+ "eval_accuracy_undropoff": 0.9861238615411099,
228
+ "eval_accuracy_unlabeled": NaN,
229
+ "eval_iou_dropoff": 0.07771069890623208,
230
+ "eval_iou_undropoff": 0.9491694411343236,
231
+ "eval_iou_unlabeled": NaN,
232
+ "eval_loss": 0.590660810470581,
233
+ "eval_mean_accuracy": 0.5443452170844613,
234
+ "eval_mean_iou": 0.5134400700202778,
235
+ "eval_overall_accuracy": 0.9493862152099609,
236
+ "eval_runtime": 2.7723,
237
+ "eval_samples_per_second": 7.214,
238
+ "eval_steps_per_second": 0.721,
239
+ "step": 30
240
+ },
241
+ {
242
+ "epoch": 15.5,
243
+ "learning_rate": 7.333333333333333e-05,
244
+ "loss": 0.6219,
245
+ "step": 31
246
+ },
247
+ {
248
+ "epoch": 16.0,
249
+ "learning_rate": 7.298245614035087e-05,
250
+ "loss": 0.6276,
251
+ "step": 32
252
+ },
253
+ {
254
+ "epoch": 16.5,
255
+ "learning_rate": 7.263157894736843e-05,
256
+ "loss": 0.5895,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 17.0,
261
+ "learning_rate": 7.228070175438597e-05,
262
+ "loss": 0.5987,
263
+ "step": 34
264
+ },
265
+ {
266
+ "epoch": 17.5,
267
+ "learning_rate": 7.192982456140351e-05,
268
+ "loss": 0.578,
269
+ "step": 35
270
+ },
271
+ {
272
+ "epoch": 18.0,
273
+ "learning_rate": 7.157894736842105e-05,
274
+ "loss": 0.5622,
275
+ "step": 36
276
+ },
277
+ {
278
+ "epoch": 18.5,
279
+ "learning_rate": 7.12280701754386e-05,
280
+ "loss": 0.5411,
281
+ "step": 37
282
+ },
283
+ {
284
+ "epoch": 19.0,
285
+ "learning_rate": 7.087719298245614e-05,
286
+ "loss": 0.5512,
287
+ "step": 38
288
+ },
289
+ {
290
+ "epoch": 19.5,
291
+ "learning_rate": 7.052631578947368e-05,
292
+ "loss": 0.5536,
293
+ "step": 39
294
+ },
295
+ {
296
+ "epoch": 20.0,
297
+ "learning_rate": 7.017543859649124e-05,
298
+ "loss": 0.5118,
299
+ "step": 40
300
+ },
301
+ {
302
+ "epoch": 20.0,
303
+ "eval_accuracy_dropoff": 0.20064680382577582,
304
+ "eval_accuracy_undropoff": 0.9839122686389837,
305
+ "eval_accuracy_unlabeled": NaN,
306
+ "eval_iou_dropoff": 0.146396810999508,
307
+ "eval_iou_undropoff": 0.9509353469996961,
308
+ "eval_iou_unlabeled": 0.0,
309
+ "eval_loss": 0.48036208748817444,
310
+ "eval_mean_accuracy": 0.5922795362323798,
311
+ "eval_mean_iou": 0.3657773859997347,
312
+ "eval_overall_accuracy": 0.9513446807861328,
313
+ "eval_runtime": 3.0452,
314
+ "eval_samples_per_second": 6.568,
315
+ "eval_steps_per_second": 0.657,
316
+ "step": 40
317
+ }
318
+ ],
319
+ "max_steps": 240,
320
+ "num_train_epochs": 120,
321
+ "total_flos": 2.035958276947968e+16,
322
+ "trial_name": null,
323
+ "trial_params": null
324
+ }
checkpoint-40/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8aac04d8cdad507b67b0cf9d85a115bd8b48759ff485068ff5a4f634035f8954
3
+ size 4219
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": 6,
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
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b67fc1e9142a1425d3256b17a33712c77775bc46bfde2d9cb053361338550cf4
3
+ size 14951629
runs/Feb12_06-53-01_robovision.csres.utexas.edu/events.out.tfevents.1707742388.robovision.csres.utexas.edu.2756623.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d11245e8ea3b3a52fffcc9f3d5edcf93b36a0db606421a1c6d8c55e0277a3b7
3
+ size 14056
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8aac04d8cdad507b67b0cf9d85a115bd8b48759ff485068ff5a4f634035f8954
3
+ size 4219