sam1120 commited on
Commit
6bf8b6c
·
1 Parent(s): 216df49

Training in progress, step 40

Browse files
checkpoint-40/config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "nvidia/mit-b5",
3
+ "architectures": [
4
+ "SegformerForSemanticSegmentation"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "classifier_dropout_prob": 0.1,
8
+ "decoder_hidden_size": 768,
9
+ "depths": [
10
+ 3,
11
+ 6,
12
+ 40,
13
+ 3
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
+ 64,
26
+ 128,
27
+ 320,
28
+ 512
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:d9c86fac2ae286e5fd5f3897d2dfa79857d42be8e7847e80c07eedfce5bba06f
3
+ size 677530431
checkpoint-40/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b3b06573a479f351440110f4c1210ecc19ff56616d851c2329b184cadffb6af
3
+ size 338834121
checkpoint-40/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a6fab04f88b1971c41429d2a9c9d882c2ba5e079292f8e5d1c86a1e912eef0c
3
+ size 14575
checkpoint-40/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b34f08a64012d57ba08701a7409055bd4794bcbffb4646f47919ece74d7e77b
3
+ size 627
checkpoint-40/trainer_state.json ADDED
@@ -0,0 +1,324 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.26776206493377686,
3
+ "best_model_checkpoint": "/robodata/smodak/Projects/nspl/scripts/terrainseg/training/models/dropoff-utcustom-train-SF-RGBD-b5_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": 4.166666666666667e-06,
13
+ "loss": 1.1395,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 1.0,
18
+ "learning_rate": 8.333333333333334e-06,
19
+ "loss": 1.1418,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 1.5,
24
+ "learning_rate": 1.25e-05,
25
+ "loss": 1.1243,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 2.0,
30
+ "learning_rate": 1.6666666666666667e-05,
31
+ "loss": 1.1018,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 2.5,
36
+ "learning_rate": 2.0833333333333336e-05,
37
+ "loss": 1.0687,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 3.0,
42
+ "learning_rate": 2.5e-05,
43
+ "loss": 1.0541,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 3.5,
48
+ "learning_rate": 2.916666666666667e-05,
49
+ "loss": 1.022,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 4.0,
54
+ "learning_rate": 3.3333333333333335e-05,
55
+ "loss": 0.9945,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 4.5,
60
+ "learning_rate": 3.7500000000000003e-05,
61
+ "loss": 0.9642,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 5.0,
66
+ "learning_rate": 4.166666666666667e-05,
67
+ "loss": 0.9278,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 5.0,
72
+ "eval_accuracy_dropoff": 0.2009128649739673,
73
+ "eval_accuracy_undropoff": 0.9081783961224983,
74
+ "eval_accuracy_unlabeled": NaN,
75
+ "eval_iou_dropoff": 0.0807428683093736,
76
+ "eval_iou_undropoff": 0.8784609129926589,
77
+ "eval_iou_unlabeled": 0.0,
78
+ "eval_loss": 0.845426082611084,
79
+ "eval_mean_accuracy": 0.5545456305482328,
80
+ "eval_mean_iou": 0.3197345937673442,
81
+ "eval_overall_accuracy": 0.8787708282470703,
82
+ "eval_runtime": 2.7556,
83
+ "eval_samples_per_second": 7.258,
84
+ "eval_steps_per_second": 0.726,
85
+ "step": 10
86
+ },
87
+ {
88
+ "epoch": 5.5,
89
+ "learning_rate": 4.5833333333333334e-05,
90
+ "loss": 0.8786,
91
+ "step": 11
92
+ },
93
+ {
94
+ "epoch": 6.0,
95
+ "learning_rate": 5e-05,
96
+ "loss": 0.8476,
97
+ "step": 12
98
+ },
99
+ {
100
+ "epoch": 6.5,
101
+ "learning_rate": 4.9780701754385966e-05,
102
+ "loss": 0.7993,
103
+ "step": 13
104
+ },
105
+ {
106
+ "epoch": 7.0,
107
+ "learning_rate": 4.956140350877193e-05,
108
+ "loss": 0.7682,
109
+ "step": 14
110
+ },
111
+ {
112
+ "epoch": 7.5,
113
+ "learning_rate": 4.9342105263157894e-05,
114
+ "loss": 0.7194,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 8.0,
119
+ "learning_rate": 4.912280701754386e-05,
120
+ "loss": 0.7016,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 8.5,
125
+ "learning_rate": 4.890350877192983e-05,
126
+ "loss": 0.6592,
127
+ "step": 17
128
+ },
129
+ {
130
+ "epoch": 9.0,
131
+ "learning_rate": 4.868421052631579e-05,
132
+ "loss": 0.6162,
133
+ "step": 18
134
+ },
135
+ {
136
+ "epoch": 9.5,
137
+ "learning_rate": 4.8464912280701755e-05,
138
+ "loss": 0.5957,
139
+ "step": 19
140
+ },
141
+ {
142
+ "epoch": 10.0,
143
+ "learning_rate": 4.824561403508772e-05,
144
+ "loss": 0.5551,
145
+ "step": 20
146
+ },
147
+ {
148
+ "epoch": 10.0,
149
+ "eval_accuracy_dropoff": 0.013546182251886512,
150
+ "eval_accuracy_undropoff": 0.994846449222221,
151
+ "eval_accuracy_unlabeled": NaN,
152
+ "eval_iou_dropoff": 0.0122168669714333,
153
+ "eval_iou_undropoff": 0.9540401274092,
154
+ "eval_iou_unlabeled": 0.0,
155
+ "eval_loss": 0.4667937755584717,
156
+ "eval_mean_accuracy": 0.5041963157370537,
157
+ "eval_mean_iou": 0.32208566479354445,
158
+ "eval_overall_accuracy": 0.9540447235107422,
159
+ "eval_runtime": 2.9472,
160
+ "eval_samples_per_second": 6.786,
161
+ "eval_steps_per_second": 0.679,
162
+ "step": 20
163
+ },
164
+ {
165
+ "epoch": 10.5,
166
+ "learning_rate": 4.802631578947368e-05,
167
+ "loss": 0.5393,
168
+ "step": 21
169
+ },
170
+ {
171
+ "epoch": 11.0,
172
+ "learning_rate": 4.780701754385965e-05,
173
+ "loss": 0.5054,
174
+ "step": 22
175
+ },
176
+ {
177
+ "epoch": 11.5,
178
+ "learning_rate": 4.758771929824562e-05,
179
+ "loss": 0.4768,
180
+ "step": 23
181
+ },
182
+ {
183
+ "epoch": 12.0,
184
+ "learning_rate": 4.736842105263158e-05,
185
+ "loss": 0.4747,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 12.5,
190
+ "learning_rate": 4.714912280701755e-05,
191
+ "loss": 0.4453,
192
+ "step": 25
193
+ },
194
+ {
195
+ "epoch": 13.0,
196
+ "learning_rate": 4.6929824561403515e-05,
197
+ "loss": 0.4263,
198
+ "step": 26
199
+ },
200
+ {
201
+ "epoch": 13.5,
202
+ "learning_rate": 4.671052631578948e-05,
203
+ "loss": 0.4107,
204
+ "step": 27
205
+ },
206
+ {
207
+ "epoch": 14.0,
208
+ "learning_rate": 4.649122807017544e-05,
209
+ "loss": 0.3888,
210
+ "step": 28
211
+ },
212
+ {
213
+ "epoch": 14.5,
214
+ "learning_rate": 4.6271929824561406e-05,
215
+ "loss": 0.3656,
216
+ "step": 29
217
+ },
218
+ {
219
+ "epoch": 15.0,
220
+ "learning_rate": 4.605263157894737e-05,
221
+ "loss": 0.3667,
222
+ "step": 30
223
+ },
224
+ {
225
+ "epoch": 15.0,
226
+ "eval_accuracy_dropoff": 0.008839652285602882,
227
+ "eval_accuracy_undropoff": 0.9981533905751077,
228
+ "eval_accuracy_unlabeled": NaN,
229
+ "eval_iou_dropoff": 0.0085100181505836,
230
+ "eval_iou_undropoff": 0.9570052236822447,
231
+ "eval_iou_unlabeled": 0.0,
232
+ "eval_loss": 0.33538389205932617,
233
+ "eval_mean_accuracy": 0.5034965214303553,
234
+ "eval_mean_iou": 0.3218384139442761,
235
+ "eval_overall_accuracy": 0.9570184707641601,
236
+ "eval_runtime": 11.3505,
237
+ "eval_samples_per_second": 1.762,
238
+ "eval_steps_per_second": 0.176,
239
+ "step": 30
240
+ },
241
+ {
242
+ "epoch": 15.5,
243
+ "learning_rate": 4.5833333333333334e-05,
244
+ "loss": 0.3391,
245
+ "step": 31
246
+ },
247
+ {
248
+ "epoch": 16.0,
249
+ "learning_rate": 4.56140350877193e-05,
250
+ "loss": 0.3252,
251
+ "step": 32
252
+ },
253
+ {
254
+ "epoch": 16.5,
255
+ "learning_rate": 4.539473684210527e-05,
256
+ "loss": 0.3199,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 17.0,
261
+ "learning_rate": 4.517543859649123e-05,
262
+ "loss": 0.2957,
263
+ "step": 34
264
+ },
265
+ {
266
+ "epoch": 17.5,
267
+ "learning_rate": 4.4956140350877196e-05,
268
+ "loss": 0.2837,
269
+ "step": 35
270
+ },
271
+ {
272
+ "epoch": 18.0,
273
+ "learning_rate": 4.473684210526316e-05,
274
+ "loss": 0.2764,
275
+ "step": 36
276
+ },
277
+ {
278
+ "epoch": 18.5,
279
+ "learning_rate": 4.451754385964912e-05,
280
+ "loss": 0.239,
281
+ "step": 37
282
+ },
283
+ {
284
+ "epoch": 19.0,
285
+ "learning_rate": 4.429824561403509e-05,
286
+ "loss": 0.2586,
287
+ "step": 38
288
+ },
289
+ {
290
+ "epoch": 19.5,
291
+ "learning_rate": 4.407894736842105e-05,
292
+ "loss": 0.2539,
293
+ "step": 39
294
+ },
295
+ {
296
+ "epoch": 20.0,
297
+ "learning_rate": 4.3859649122807014e-05,
298
+ "loss": 0.2402,
299
+ "step": 40
300
+ },
301
+ {
302
+ "epoch": 20.0,
303
+ "eval_accuracy_dropoff": 0.31156219179338973,
304
+ "eval_accuracy_undropoff": 0.9867716375598645,
305
+ "eval_accuracy_unlabeled": NaN,
306
+ "eval_iou_dropoff": 0.23875964087096524,
307
+ "eval_iou_undropoff": 0.9581548537274001,
308
+ "eval_iou_unlabeled": NaN,
309
+ "eval_loss": 0.26776206493377686,
310
+ "eval_mean_accuracy": 0.6491669146766271,
311
+ "eval_mean_iou": 0.5984572472991827,
312
+ "eval_overall_accuracy": 0.9586969375610351,
313
+ "eval_runtime": 3.4868,
314
+ "eval_samples_per_second": 5.736,
315
+ "eval_steps_per_second": 0.574,
316
+ "step": 40
317
+ }
318
+ ],
319
+ "max_steps": 240,
320
+ "num_train_epochs": 120,
321
+ "total_flos": 4.630914352860365e+17,
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:cab6ccc8391de81e679a8390875ef1e1d4b854c12575abf4fbccc128549a7cb6
3
+ size 4155
config.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "nvidia/mit-b5",
3
+ "architectures": [
4
+ "SegformerForSemanticSegmentation"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "classifier_dropout_prob": 0.1,
8
+ "decoder_hidden_size": 768,
9
+ "depths": [
10
+ 3,
11
+ 6,
12
+ 40,
13
+ 3
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
+ 64,
26
+ 128,
27
+ 320,
28
+ 512
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 sha256:1b3b06573a479f351440110f4c1210ecc19ff56616d851c2329b184cadffb6af
3
+ size 338834121
runs/Feb12_07-24-26_robolidar/events.out.tfevents.1707744272.robolidar.4128056.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7841e0f7a190b2e4a8b4dadd90430060b175dca8adbcbbb2f0dfb68ac8041cd
3
+ size 14040
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cab6ccc8391de81e679a8390875ef1e1d4b854c12575abf4fbccc128549a7cb6
3
+ size 4155