sam1120 commited on
Commit
442fd78
1 Parent(s): f234d9d

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": "safe",
33
+ "2": "unsafe"
34
+ },
35
+ "image_size": 224,
36
+ "initializer_range": 0.02,
37
+ "label2id": {
38
+ "safe": 1,
39
+ "unlabeled": 0,
40
+ "unsafe": 2
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-40/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5883999b89251e67a046fd252c4ebb64a28f8e18c76d227cd00a272bfa628cfa
3
+ size 677455103
checkpoint-40/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d86ebb7678504c184b3fa37c7ba69feb6a1d5bf503afa303828067bfac013316
3
+ size 338796489
checkpoint-40/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d1b64c45dad45aec1bf9dd5d9ab268ef160b567c3dc95e67f957fd60e3048f8
3
+ size 14575
checkpoint-40/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0bc233fc24cb6677f85e2cb4ba7fcebcfd878470a515da2427abecc6928cd87
3
+ size 627
checkpoint-40/trainer_state.json ADDED
@@ -0,0 +1,324 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.0047743320465088,
3
+ "best_model_checkpoint": "/robodata/smodak/Projects/nspl/scripts/terrainseg/training/models/safety-utcustom-train-SF-RGB-b5/checkpoint-40",
4
+ "epoch": 3.6363636363636362,
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": 0.09,
12
+ "learning_rate": 1.764705882352941e-07,
13
+ "loss": 1.2293,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.18,
18
+ "learning_rate": 3.529411764705882e-07,
19
+ "loss": 1.2349,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 0.27,
24
+ "learning_rate": 5.294117647058824e-07,
25
+ "loss": 1.2357,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 0.36,
30
+ "learning_rate": 7.058823529411765e-07,
31
+ "loss": 1.2242,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 0.45,
36
+ "learning_rate": 8.823529411764706e-07,
37
+ "loss": 1.2247,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 0.55,
42
+ "learning_rate": 1.0588235294117648e-06,
43
+ "loss": 1.227,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 0.64,
48
+ "learning_rate": 1.2352941176470588e-06,
49
+ "loss": 1.2292,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 0.73,
54
+ "learning_rate": 1.411764705882353e-06,
55
+ "loss": 1.2257,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 0.82,
60
+ "learning_rate": 1.5882352941176472e-06,
61
+ "loss": 1.2173,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 0.91,
66
+ "learning_rate": 1.7647058823529412e-06,
67
+ "loss": 1.2239,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 0.91,
72
+ "eval_accuracy_safe": 0.39920432072351303,
73
+ "eval_accuracy_unlabeled": NaN,
74
+ "eval_accuracy_unsafe": 0.29510386893492613,
75
+ "eval_iou_safe": 0.031418797454432365,
76
+ "eval_iou_unlabeled": 0.0,
77
+ "eval_iou_unsafe": 0.29392505413655395,
78
+ "eval_loss": 1.1102662086486816,
79
+ "eval_mean_accuracy": 0.34715409482921955,
80
+ "eval_mean_iou": 0.10844795053032878,
81
+ "eval_overall_accuracy": 0.2981788293639226,
82
+ "eval_runtime": 9.3061,
83
+ "eval_samples_per_second": 7.2,
84
+ "eval_steps_per_second": 0.537,
85
+ "step": 10
86
+ },
87
+ {
88
+ "epoch": 1.0,
89
+ "learning_rate": 1.9411764705882353e-06,
90
+ "loss": 1.221,
91
+ "step": 11
92
+ },
93
+ {
94
+ "epoch": 1.09,
95
+ "learning_rate": 2.1176470588235296e-06,
96
+ "loss": 1.2149,
97
+ "step": 12
98
+ },
99
+ {
100
+ "epoch": 1.18,
101
+ "learning_rate": 2.2941176470588234e-06,
102
+ "loss": 1.2197,
103
+ "step": 13
104
+ },
105
+ {
106
+ "epoch": 1.27,
107
+ "learning_rate": 2.4705882352941177e-06,
108
+ "loss": 1.2064,
109
+ "step": 14
110
+ },
111
+ {
112
+ "epoch": 1.36,
113
+ "learning_rate": 2.647058823529412e-06,
114
+ "loss": 1.2042,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 1.45,
119
+ "learning_rate": 2.823529411764706e-06,
120
+ "loss": 1.2046,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 1.55,
125
+ "learning_rate": 3e-06,
126
+ "loss": 1.2054,
127
+ "step": 17
128
+ },
129
+ {
130
+ "epoch": 1.64,
131
+ "learning_rate": 2.990415335463259e-06,
132
+ "loss": 1.2002,
133
+ "step": 18
134
+ },
135
+ {
136
+ "epoch": 1.73,
137
+ "learning_rate": 2.9808306709265177e-06,
138
+ "loss": 1.1982,
139
+ "step": 19
140
+ },
141
+ {
142
+ "epoch": 1.82,
143
+ "learning_rate": 2.9712460063897764e-06,
144
+ "loss": 1.1948,
145
+ "step": 20
146
+ },
147
+ {
148
+ "epoch": 1.82,
149
+ "eval_accuracy_safe": 0.5218599774094934,
150
+ "eval_accuracy_unlabeled": NaN,
151
+ "eval_accuracy_unsafe": 0.3705393407485172,
152
+ "eval_iou_safe": 0.0439841259320199,
153
+ "eval_iou_unlabeled": 0.0,
154
+ "eval_iou_unsafe": 0.36893471456827454,
155
+ "eval_loss": 1.0963102579116821,
156
+ "eval_mean_accuracy": 0.4461996590790053,
157
+ "eval_mean_iou": 0.13763961350009815,
158
+ "eval_overall_accuracy": 0.37500910972481344,
159
+ "eval_runtime": 8.3942,
160
+ "eval_samples_per_second": 7.982,
161
+ "eval_steps_per_second": 0.596,
162
+ "step": 20
163
+ },
164
+ {
165
+ "epoch": 1.91,
166
+ "learning_rate": 2.9616613418530352e-06,
167
+ "loss": 1.1776,
168
+ "step": 21
169
+ },
170
+ {
171
+ "epoch": 2.0,
172
+ "learning_rate": 2.952076677316294e-06,
173
+ "loss": 1.1936,
174
+ "step": 22
175
+ },
176
+ {
177
+ "epoch": 2.09,
178
+ "learning_rate": 2.942492012779553e-06,
179
+ "loss": 1.1664,
180
+ "step": 23
181
+ },
182
+ {
183
+ "epoch": 2.18,
184
+ "learning_rate": 2.9329073482428116e-06,
185
+ "loss": 1.1809,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 2.27,
190
+ "learning_rate": 2.9233226837060704e-06,
191
+ "loss": 1.1731,
192
+ "step": 25
193
+ },
194
+ {
195
+ "epoch": 2.36,
196
+ "learning_rate": 2.913738019169329e-06,
197
+ "loss": 1.1794,
198
+ "step": 26
199
+ },
200
+ {
201
+ "epoch": 2.45,
202
+ "learning_rate": 2.904153354632588e-06,
203
+ "loss": 1.1684,
204
+ "step": 27
205
+ },
206
+ {
207
+ "epoch": 2.55,
208
+ "learning_rate": 2.8945686900958468e-06,
209
+ "loss": 1.1594,
210
+ "step": 28
211
+ },
212
+ {
213
+ "epoch": 2.64,
214
+ "learning_rate": 2.884984025559105e-06,
215
+ "loss": 1.146,
216
+ "step": 29
217
+ },
218
+ {
219
+ "epoch": 2.73,
220
+ "learning_rate": 2.8753993610223644e-06,
221
+ "loss": 1.1661,
222
+ "step": 30
223
+ },
224
+ {
225
+ "epoch": 2.73,
226
+ "eval_accuracy_safe": 0.5863200218966003,
227
+ "eval_accuracy_unlabeled": NaN,
228
+ "eval_accuracy_unsafe": 0.49882181393718666,
229
+ "eval_iou_safe": 0.06472096392925002,
230
+ "eval_iou_unlabeled": 0.0,
231
+ "eval_iou_unsafe": 0.49613792944888824,
232
+ "eval_loss": 1.0516184568405151,
233
+ "eval_mean_accuracy": 0.5425709179168935,
234
+ "eval_mean_iou": 0.1869529644593794,
235
+ "eval_overall_accuracy": 0.5014063707038537,
236
+ "eval_runtime": 9.3615,
237
+ "eval_samples_per_second": 7.157,
238
+ "eval_steps_per_second": 0.534,
239
+ "step": 30
240
+ },
241
+ {
242
+ "epoch": 2.82,
243
+ "learning_rate": 2.865814696485623e-06,
244
+ "loss": 1.153,
245
+ "step": 31
246
+ },
247
+ {
248
+ "epoch": 2.91,
249
+ "learning_rate": 2.856230031948882e-06,
250
+ "loss": 1.149,
251
+ "step": 32
252
+ },
253
+ {
254
+ "epoch": 3.0,
255
+ "learning_rate": 2.8466453674121407e-06,
256
+ "loss": 1.164,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 3.09,
261
+ "learning_rate": 2.8370607028753995e-06,
262
+ "loss": 1.1479,
263
+ "step": 34
264
+ },
265
+ {
266
+ "epoch": 3.18,
267
+ "learning_rate": 2.8274760383386583e-06,
268
+ "loss": 1.1428,
269
+ "step": 35
270
+ },
271
+ {
272
+ "epoch": 3.27,
273
+ "learning_rate": 2.817891373801917e-06,
274
+ "loss": 1.1398,
275
+ "step": 36
276
+ },
277
+ {
278
+ "epoch": 3.36,
279
+ "learning_rate": 2.8083067092651755e-06,
280
+ "loss": 1.1415,
281
+ "step": 37
282
+ },
283
+ {
284
+ "epoch": 3.45,
285
+ "learning_rate": 2.7987220447284347e-06,
286
+ "loss": 1.1146,
287
+ "step": 38
288
+ },
289
+ {
290
+ "epoch": 3.55,
291
+ "learning_rate": 2.7891373801916935e-06,
292
+ "loss": 1.1234,
293
+ "step": 39
294
+ },
295
+ {
296
+ "epoch": 3.64,
297
+ "learning_rate": 2.7795527156549523e-06,
298
+ "loss": 1.1112,
299
+ "step": 40
300
+ },
301
+ {
302
+ "epoch": 3.64,
303
+ "eval_accuracy_safe": 0.5458517893146133,
304
+ "eval_accuracy_unlabeled": NaN,
305
+ "eval_accuracy_unsafe": 0.5794390867479824,
306
+ "eval_iou_safe": 0.0899865428032596,
307
+ "eval_iou_unlabeled": 0.0,
308
+ "eval_iou_unsafe": 0.5753581085678546,
309
+ "eval_loss": 1.0047743320465088,
310
+ "eval_mean_accuracy": 0.5626454380312979,
311
+ "eval_mean_iou": 0.22178155045703807,
312
+ "eval_overall_accuracy": 0.5784469718363747,
313
+ "eval_runtime": 9.38,
314
+ "eval_samples_per_second": 7.143,
315
+ "eval_steps_per_second": 0.533,
316
+ "step": 40
317
+ }
318
+ ],
319
+ "max_steps": 330,
320
+ "num_train_epochs": 30,
321
+ "total_flos": 2.2633072927860326e+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 sha256:7f030ee6346022874816800d95eb680870a6f0608d7f7d70d11613eaa9ed9ab3
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": "safe",
33
+ "2": "unsafe"
34
+ },
35
+ "image_size": 224,
36
+ "initializer_range": 0.02,
37
+ "label2id": {
38
+ "safe": 1,
39
+ "unlabeled": 0,
40
+ "unsafe": 2
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
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d86ebb7678504c184b3fa37c7ba69feb6a1d5bf503afa303828067bfac013316
3
+ size 338796489
runs/Feb07_22-15-06_robovision.csres.utexas.edu/events.out.tfevents.1707365713.robovision.csres.utexas.edu.742362.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e99078c2dcc73c09e1c22cc9d70d5714bbd0bd1364af7fa0be070ba50cec358
3
+ size 13979
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f030ee6346022874816800d95eb680870a6f0608d7f7d70d11613eaa9ed9ab3
3
+ size 4155