vcasecnikovs / yet-another-yolov4-pytorch Goto Github PK
View Code? Open in Web Editor NEWYOLOv4 Pytorch implementation with all freebies and specials and 15+ more exclusive improvements. Easy to use!
License: MIT License
YOLOv4 Pytorch implementation with all freebies and specials and 15+ more exclusive improvements. Easy to use!
License: MIT License
loss_conf_noobj = F.binary_cross_entropy(pred_conf[noobj_mask], tconf[noobj_mask])
RuntimeError: copy_if failed to synchronize: cudaErrorAssert: device-side assert triggered
pred_conf
tensor([[[[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
...,
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan]],
[[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
...,
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan]],
[[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
...,
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan],
[nan, nan, nan, ..., nan, nan, nan]]],
@VCasecnikovs Thank you for answering so fast.
[I cannot reopen the issue, this is why I am creating a new issue, see https://newbedev.com/how-to-re-open-an-issue-in-github]
I reverted the code and reinstall the pytorch-lightning library
$ conda list | grep pytorch-lightning
pytorch-lightning 0.7.6 pypi_0 pypi
Running r = t.lr_find(m, min_lr=1e-10, max_lr=1, early_stop_threshold=None)
gives me the following error:
| Name | Type | Params
--------------------------------------------------------------------------------
0 | model | YOLOv4 | 70 M
1 | model.backbone | Backbone | 26 M
2 | model.backbone.d1 | DownSampleFirst | 61 K
3 | model.backbone.d1.c1 | ConvBlock | 928
4 | model.backbone.d1.c1.module | Sequential | 928
5 | model.backbone.d1.c1.module.0 | Conv2d | 864
6 | model.backbone.d1.c1.module.1 | BatchNorm2d | 64
7 | model.backbone.d1.c1.module.2 | Mish | 0
8 | model.backbone.d1.c2 | ConvBlock | 18 K
9 | model.backbone.d1.c2.module | Sequential | 18 K
10 | model.backbone.d1.c2.module.0 | Conv2d | 18 K
11 | model.backbone.d1.c2.module.1 | BatchNorm2d | 128
12 | model.backbone.d1.c2.module.2 | Mish | 0
13 | model.backbone.d1.c3 | ConvBlock | 4 K
14 | model.backbone.d1.c3.module | Sequential | 4 K
15 | model.backbone.d1.c3.module.0 | Conv2d | 4 K
16 | model.backbone.d1.c3.module.1 | BatchNorm2d | 128
17 | model.backbone.d1.c3.module.2 | Mish | 0
18 | model.backbone.d1.c4 | ConvBlock | 2 K
19 | model.backbone.d1.c4.module | Sequential | 2 K
20 | model.backbone.d1.c4.module.0 | Conv2d | 2 K
21 | model.backbone.d1.c4.module.1 | BatchNorm2d | 64
22 | model.backbone.d1.c4.module.2 | Mish | 0
23 | model.backbone.d1.c5 | ConvBlock | 18 K
24 | model.backbone.d1.c5.module | Sequential | 18 K
25 | model.backbone.d1.c5.module.0 | Conv2d | 18 K
26 | model.backbone.d1.c5.module.1 | BatchNorm2d | 128
27 | model.backbone.d1.c5.module.2 | Mish | 0
28 | model.backbone.d1.c6 | ConvBlock | 4 K
29 | model.backbone.d1.c6.module | Sequential | 4 K
30 | model.backbone.d1.c6.module.0 | Conv2d | 4 K
31 | model.backbone.d1.c6.module.1 | BatchNorm2d | 128
32 | model.backbone.d1.c6.module.2 | Mish | 0
33 | model.backbone.d1.dense_c3_c6 | ConvBlock | 4 K
34 | model.backbone.d1.dense_c3_c6.module | Sequential | 4 K
35 | model.backbone.d1.dense_c3_c6.module.0 | Conv2d | 4 K
36 | model.backbone.d1.dense_c3_c6.module.1 | BatchNorm2d | 128
37 | model.backbone.d1.dense_c3_c6.module.2 | Mish | 0
38 | model.backbone.d1.c7 | ConvBlock | 8 K
39 | model.backbone.d1.c7.module | Sequential | 8 K
40 | model.backbone.d1.c7.module.0 | Conv2d | 8 K
41 | model.backbone.d1.c7.module.1 | BatchNorm2d | 128
42 | model.backbone.d1.c7.module.2 | Mish | 0
43 | model.backbone.d2 | DownSampleBlock | 193 K
44 | model.backbone.d2.c1 | ConvBlock | 73 K
45 | model.backbone.d2.c1.module | Sequential | 73 K
46 | model.backbone.d2.c1.module.0 | Conv2d | 73 K
47 | model.backbone.d2.c1.module.1 | BatchNorm2d | 256
48 | model.backbone.d2.c1.module.2 | Mish | 0
49 | model.backbone.d2.c2 | ConvBlock | 8 K
50 | model.backbone.d2.c2.module | Sequential | 8 K
51 | model.backbone.d2.c2.module.0 | Conv2d | 8 K
52 | model.backbone.d2.c2.module.1 | BatchNorm2d | 128
53 | model.backbone.d2.c2.module.2 | Mish | 0
54 | model.backbone.d2.r3 | ResBlock | 82 K
55 | model.backbone.d2.r3.module_list | ModuleList | 82 K
56 | model.backbone.d2.r3.module_list.0 | ModuleList | 41 K
57 | model.backbone.d2.r3.module_list.0.0 | ConvBlock | 4 K
58 | model.backbone.d2.r3.module_list.0.0.module | Sequential | 4 K
59 | model.backbone.d2.r3.module_list.0.0.module.0 | Conv2d | 4 K
60 | model.backbone.d2.r3.module_list.0.0.module.1 | BatchNorm2d | 128
61 | model.backbone.d2.r3.module_list.0.0.module.2 | Mish | 0
62 | model.backbone.d2.r3.module_list.0.1 | ConvBlock | 36 K
63 | model.backbone.d2.r3.module_list.0.1.module | Sequential | 36 K
64 | model.backbone.d2.r3.module_list.0.1.module.0 | Conv2d | 36 K
65 | model.backbone.d2.r3.module_list.0.1.module.1 | BatchNorm2d | 128
66 | model.backbone.d2.r3.module_list.0.1.module.2 | Mish | 0
67 | model.backbone.d2.r3.module_list.1 | ModuleList | 41 K
68 | model.backbone.d2.r3.module_list.1.0 | ConvBlock | 4 K
69 | model.backbone.d2.r3.module_list.1.0.module | Sequential | 4 K
70 | model.backbone.d2.r3.module_list.1.0.module.0 | Conv2d | 4 K
71 | model.backbone.d2.r3.module_list.1.0.module.1 | BatchNorm2d | 128
72 | model.backbone.d2.r3.module_list.1.0.module.2 | Mish | 0
73 | model.backbone.d2.r3.module_list.1.1 | ConvBlock | 36 K
74 | model.backbone.d2.r3.module_list.1.1.module | Sequential | 36 K
75 | model.backbone.d2.r3.module_list.1.1.module.0 | Conv2d | 36 K
76 | model.backbone.d2.r3.module_list.1.1.module.1 | BatchNorm2d | 128
77 | model.backbone.d2.r3.module_list.1.1.module.2 | Mish | 0
78 | model.backbone.d2.c4 | ConvBlock | 4 K
79 | model.backbone.d2.c4.module | Sequential | 4 K
80 | model.backbone.d2.c4.module.0 | Conv2d | 4 K
81 | model.backbone.d2.c4.module.1 | BatchNorm2d | 128
82 | model.backbone.d2.c4.module.2 | Mish | 0
83 | model.backbone.d2.dense_c2_c4 | ConvBlock | 8 K
84 | model.backbone.d2.dense_c2_c4.module | Sequential | 8 K
85 | model.backbone.d2.dense_c2_c4.module.0 | Conv2d | 8 K
86 | model.backbone.d2.dense_c2_c4.module.1 | BatchNorm2d | 128
87 | model.backbone.d2.dense_c2_c4.module.2 | Mish | 0
88 | model.backbone.d2.c5 | ConvBlock | 16 K
89 | model.backbone.d2.c5.module | Sequential | 16 K
90 | model.backbone.d2.c5.module.0 | Conv2d | 16 K
91 | model.backbone.d2.c5.module.1 | BatchNorm2d | 256
92 | model.backbone.d2.c5.module.2 | Mish | 0
93 | model.backbone.d3 | DownSampleBlock | 1 M
94 | model.backbone.d3.c1 | ConvBlock | 295 K
95 | model.backbone.d3.c1.module | Sequential | 295 K
96 | model.backbone.d3.c1.module.0 | Conv2d | 294 K
97 | model.backbone.d3.c1.module.1 | BatchNorm2d | 512
98 | model.backbone.d3.c1.module.2 | Mish | 0
99 | model.backbone.d3.c2 | ConvBlock | 33 K
100 | model.backbone.d3.c2.module | Sequential | 33 K
101 | model.backbone.d3.c2.module.0 | Conv2d | 32 K
102 | model.backbone.d3.c2.module.1 | BatchNorm2d | 256
103 | model.backbone.d3.c2.module.2 | Mish | 0
104 | model.backbone.d3.r3 | ResBlock | 1 M
105 | model.backbone.d3.r3.module_list | ModuleList | 1 M
106 | model.backbone.d3.r3.module_list.0 | ModuleList | 164 K
107 | model.backbone.d3.r3.module_list.0.0 | ConvBlock | 16 K
108 | model.backbone.d3.r3.module_list.0.0.module | Sequential | 16 K
109 | model.backbone.d3.r3.module_list.0.0.module.0 | Conv2d | 16 K
110 | model.backbone.d3.r3.module_list.0.0.module.1 | BatchNorm2d | 256
111 | model.backbone.d3.r3.module_list.0.0.module.2 | Mish | 0
112 | model.backbone.d3.r3.module_list.0.1 | ConvBlock | 147 K
113 | model.backbone.d3.r3.module_list.0.1.module | Sequential | 147 K
114 | model.backbone.d3.r3.module_list.0.1.module.0 | Conv2d | 147 K
115 | model.backbone.d3.r3.module_list.0.1.module.1 | BatchNorm2d | 256
116 | model.backbone.d3.r3.module_list.0.1.module.2 | Mish | 0
117 | model.backbone.d3.r3.module_list.1 | ModuleList | 164 K
118 | model.backbone.d3.r3.module_list.1.0 | ConvBlock | 16 K
119 | model.backbone.d3.r3.module_list.1.0.module | Sequential | 16 K
120 | model.backbone.d3.r3.module_list.1.0.module.0 | Conv2d | 16 K
121 | model.backbone.d3.r3.module_list.1.0.module.1 | BatchNorm2d | 256
122 | model.backbone.d3.r3.module_list.1.0.module.2 | Mish | 0
123 | model.backbone.d3.r3.module_list.1.1 | ConvBlock | 147 K
124 | model.backbone.d3.r3.module_list.1.1.module | Sequential | 147 K
125 | model.backbone.d3.r3.module_list.1.1.module.0 | Conv2d | 147 K
126 | model.backbone.d3.r3.module_list.1.1.module.1 | BatchNorm2d | 256
127 | model.backbone.d3.r3.module_list.1.1.module.2 | Mish | 0
128 | model.backbone.d3.r3.module_list.2 | ModuleList | 164 K
129 | model.backbone.d3.r3.module_list.2.0 | ConvBlock | 16 K
130 | model.backbone.d3.r3.module_list.2.0.module | Sequential | 16 K
131 | model.backbone.d3.r3.module_list.2.0.module.0 | Conv2d | 16 K
132 | model.backbone.d3.r3.module_list.2.0.module.1 | BatchNorm2d | 256
133 | model.backbone.d3.r3.module_list.2.0.module.2 | Mish | 0
134 | model.backbone.d3.r3.module_list.2.1 | ConvBlock | 147 K
135 | model.backbone.d3.r3.module_list.2.1.module | Sequential | 147 K
136 | model.backbone.d3.r3.module_list.2.1.module.0 | Conv2d | 147 K
137 | model.backbone.d3.r3.module_list.2.1.module.1 | BatchNorm2d | 256
138 | model.backbone.d3.r3.module_list.2.1.module.2 | Mish | 0
139 | model.backbone.d3.r3.module_list.3 | ModuleList | 164 K
140 | model.backbone.d3.r3.module_list.3.0 | ConvBlock | 16 K
141 | model.backbone.d3.r3.module_list.3.0.module | Sequential | 16 K
142 | model.backbone.d3.r3.module_list.3.0.module.0 | Conv2d | 16 K
143 | model.backbone.d3.r3.module_list.3.0.module.1 | BatchNorm2d | 256
144 | model.backbone.d3.r3.module_list.3.0.module.2 | Mish | 0
145 | model.backbone.d3.r3.module_list.3.1 | ConvBlock | 147 K
146 | model.backbone.d3.r3.module_list.3.1.module | Sequential | 147 K
147 | model.backbone.d3.r3.module_list.3.1.module.0 | Conv2d | 147 K
148 | model.backbone.d3.r3.module_list.3.1.module.1 | BatchNorm2d | 256
149 | model.backbone.d3.r3.module_list.3.1.module.2 | Mish | 0
150 | model.backbone.d3.r3.module_list.4 | ModuleList | 164 K
151 | model.backbone.d3.r3.module_list.4.0 | ConvBlock | 16 K
152 | model.backbone.d3.r3.module_list.4.0.module | Sequential | 16 K
153 | model.backbone.d3.r3.module_list.4.0.module.0 | Conv2d | 16 K
154 | model.backbone.d3.r3.module_list.4.0.module.1 | BatchNorm2d | 256
155 | model.backbone.d3.r3.module_list.4.0.module.2 | Mish | 0
156 | model.backbone.d3.r3.module_list.4.1 | ConvBlock | 147 K
157 | model.backbone.d3.r3.module_list.4.1.module | Sequential | 147 K
158 | model.backbone.d3.r3.module_list.4.1.module.0 | Conv2d | 147 K
159 | model.backbone.d3.r3.module_list.4.1.module.1 | BatchNorm2d | 256
160 | model.backbone.d3.r3.module_list.4.1.module.2 | Mish | 0
161 | model.backbone.d3.r3.module_list.5 | ModuleList | 164 K
162 | model.backbone.d3.r3.module_list.5.0 | ConvBlock | 16 K
163 | model.backbone.d3.r3.module_list.5.0.module | Sequential | 16 K
164 | model.backbone.d3.r3.module_list.5.0.module.0 | Conv2d | 16 K
165 | model.backbone.d3.r3.module_list.5.0.module.1 | BatchNorm2d | 256
166 | model.backbone.d3.r3.module_list.5.0.module.2 | Mish | 0
167 | model.backbone.d3.r3.module_list.5.1 | ConvBlock | 147 K
168 | model.backbone.d3.r3.module_list.5.1.module | Sequential | 147 K
169 | model.backbone.d3.r3.module_list.5.1.module.0 | Conv2d | 147 K
170 | model.backbone.d3.r3.module_list.5.1.module.1 | BatchNorm2d | 256
171 | model.backbone.d3.r3.module_list.5.1.module.2 | Mish | 0
172 | model.backbone.d3.r3.module_list.6 | ModuleList | 164 K
173 | model.backbone.d3.r3.module_list.6.0 | ConvBlock | 16 K
174 | model.backbone.d3.r3.module_list.6.0.module | Sequential | 16 K
175 | model.backbone.d3.r3.module_list.6.0.module.0 | Conv2d | 16 K
176 | model.backbone.d3.r3.module_list.6.0.module.1 | BatchNorm2d | 256
177 | model.backbone.d3.r3.module_list.6.0.module.2 | Mish | 0
178 | model.backbone.d3.r3.module_list.6.1 | ConvBlock | 147 K
179 | model.backbone.d3.r3.module_list.6.1.module | Sequential | 147 K
180 | model.backbone.d3.r3.module_list.6.1.module.0 | Conv2d | 147 K
181 | model.backbone.d3.r3.module_list.6.1.module.1 | BatchNorm2d | 256
182 | model.backbone.d3.r3.module_list.6.1.module.2 | Mish | 0
183 | model.backbone.d3.r3.module_list.7 | ModuleList | 164 K
184 | model.backbone.d3.r3.module_list.7.0 | ConvBlock | 16 K
185 | model.backbone.d3.r3.module_list.7.0.module | Sequential | 16 K
186 | model.backbone.d3.r3.module_list.7.0.module.0 | Conv2d | 16 K
187 | model.backbone.d3.r3.module_list.7.0.module.1 | BatchNorm2d | 256
188 | model.backbone.d3.r3.module_list.7.0.module.2 | Mish | 0
189 | model.backbone.d3.r3.module_list.7.1 | ConvBlock | 147 K
190 | model.backbone.d3.r3.module_list.7.1.module | Sequential | 147 K
191 | model.backbone.d3.r3.module_list.7.1.module.0 | Conv2d | 147 K
192 | model.backbone.d3.r3.module_list.7.1.module.1 | BatchNorm2d | 256
193 | model.backbone.d3.r3.module_list.7.1.module.2 | Mish | 0
194 | model.backbone.d3.c4 | ConvBlock | 16 K
195 | model.backbone.d3.c4.module | Sequential | 16 K
196 | model.backbone.d3.c4.module.0 | Conv2d | 16 K
197 | model.backbone.d3.c4.module.1 | BatchNorm2d | 256
198 | model.backbone.d3.c4.module.2 | Mish | 0
199 | model.backbone.d3.dense_c2_c4 | ConvBlock | 33 K
200 | model.backbone.d3.dense_c2_c4.module | Sequential | 33 K
201 | model.backbone.d3.dense_c2_c4.module.0 | Conv2d | 32 K
202 | model.backbone.d3.dense_c2_c4.module.1 | BatchNorm2d | 256
203 | model.backbone.d3.dense_c2_c4.module.2 | Mish | 0
204 | model.backbone.d3.c5 | ConvBlock | 66 K
205 | model.backbone.d3.c5.module | Sequential | 66 K
206 | model.backbone.d3.c5.module.0 | Conv2d | 65 K
207 | model.backbone.d3.c5.module.1 | BatchNorm2d | 512
208 | model.backbone.d3.c5.module.2 | Mish | 0
209 | model.backbone.d4 | DownSampleBlock | 7 M
210 | model.backbone.d4.c1 | ConvBlock | 1 M
211 | model.backbone.d4.c1.module | Sequential | 1 M
212 | model.backbone.d4.c1.module.0 | Conv2d | 1 M
213 | model.backbone.d4.c1.module.1 | BatchNorm2d | 1 K
214 | model.backbone.d4.c1.module.2 | Mish | 0
215 | model.backbone.d4.c2 | ConvBlock | 131 K
216 | model.backbone.d4.c2.module | Sequential | 131 K
217 | model.backbone.d4.c2.module.0 | Conv2d | 131 K
218 | model.backbone.d4.c2.module.1 | BatchNorm2d | 512
219 | model.backbone.d4.c2.module.2 | Mish | 0
220 | model.backbone.d4.r3 | ResBlock | 5 M
221 | model.backbone.d4.r3.module_list | ModuleList | 5 M
222 | model.backbone.d4.r3.module_list.0 | ModuleList | 656 K
223 | model.backbone.d4.r3.module_list.0.0 | ConvBlock | 66 K
224 | model.backbone.d4.r3.module_list.0.0.module | Sequential | 66 K
225 | model.backbone.d4.r3.module_list.0.0.module.0 | Conv2d | 65 K
226 | model.backbone.d4.r3.module_list.0.0.module.1 | BatchNorm2d | 512
227 | model.backbone.d4.r3.module_list.0.0.module.2 | Mish | 0
228 | model.backbone.d4.r3.module_list.0.1 | ConvBlock | 590 K
229 | model.backbone.d4.r3.module_list.0.1.module | Sequential | 590 K
230 | model.backbone.d4.r3.module_list.0.1.module.0 | Conv2d | 589 K
231 | model.backbone.d4.r3.module_list.0.1.module.1 | BatchNorm2d | 512
232 | model.backbone.d4.r3.module_list.0.1.module.2 | Mish | 0
233 | model.backbone.d4.r3.module_list.1 | ModuleList | 656 K
234 | model.backbone.d4.r3.module_list.1.0 | ConvBlock | 66 K
235 | model.backbone.d4.r3.module_list.1.0.module | Sequential | 66 K
236 | model.backbone.d4.r3.module_list.1.0.module.0 | Conv2d | 65 K
237 | model.backbone.d4.r3.module_list.1.0.module.1 | BatchNorm2d | 512
238 | model.backbone.d4.r3.module_list.1.0.module.2 | Mish | 0
239 | model.backbone.d4.r3.module_list.1.1 | ConvBlock | 590 K
240 | model.backbone.d4.r3.module_list.1.1.module | Sequential | 590 K
241 | model.backbone.d4.r3.module_list.1.1.module.0 | Conv2d | 589 K
242 | model.backbone.d4.r3.module_list.1.1.module.1 | BatchNorm2d | 512
243 | model.backbone.d4.r3.module_list.1.1.module.2 | Mish | 0
244 | model.backbone.d4.r3.module_list.2 | ModuleList | 656 K
245 | model.backbone.d4.r3.module_list.2.0 | ConvBlock | 66 K
246 | model.backbone.d4.r3.module_list.2.0.module | Sequential | 66 K
247 | model.backbone.d4.r3.module_list.2.0.module.0 | Conv2d | 65 K
248 | model.backbone.d4.r3.module_list.2.0.module.1 | BatchNorm2d | 512
249 | model.backbone.d4.r3.module_list.2.0.module.2 | Mish | 0
250 | model.backbone.d4.r3.module_list.2.1 | ConvBlock | 590 K
251 | model.backbone.d4.r3.module_list.2.1.module | Sequential | 590 K
252 | model.backbone.d4.r3.module_list.2.1.module.0 | Conv2d | 589 K
253 | model.backbone.d4.r3.module_list.2.1.module.1 | BatchNorm2d | 512
254 | model.backbone.d4.r3.module_list.2.1.module.2 | Mish | 0
255 | model.backbone.d4.r3.module_list.3 | ModuleList | 656 K
256 | model.backbone.d4.r3.module_list.3.0 | ConvBlock | 66 K
257 | model.backbone.d4.r3.module_list.3.0.module | Sequential | 66 K
258 | model.backbone.d4.r3.module_list.3.0.module.0 | Conv2d | 65 K
259 | model.backbone.d4.r3.module_list.3.0.module.1 | BatchNorm2d | 512
260 | model.backbone.d4.r3.module_list.3.0.module.2 | Mish | 0
261 | model.backbone.d4.r3.module_list.3.1 | ConvBlock | 590 K
262 | model.backbone.d4.r3.module_list.3.1.module | Sequential | 590 K
263 | model.backbone.d4.r3.module_list.3.1.module.0 | Conv2d | 589 K
264 | model.backbone.d4.r3.module_list.3.1.module.1 | BatchNorm2d | 512
265 | model.backbone.d4.r3.module_list.3.1.module.2 | Mish | 0
266 | model.backbone.d4.r3.module_list.4 | ModuleList | 656 K
267 | model.backbone.d4.r3.module_list.4.0 | ConvBlock | 66 K
268 | model.backbone.d4.r3.module_list.4.0.module | Sequential | 66 K
269 | model.backbone.d4.r3.module_list.4.0.module.0 | Conv2d | 65 K
270 | model.backbone.d4.r3.module_list.4.0.module.1 | BatchNorm2d | 512
271 | model.backbone.d4.r3.module_list.4.0.module.2 | Mish | 0
272 | model.backbone.d4.r3.module_list.4.1 | ConvBlock | 590 K
273 | model.backbone.d4.r3.module_list.4.1.module | Sequential | 590 K
274 | model.backbone.d4.r3.module_list.4.1.module.0 | Conv2d | 589 K
275 | model.backbone.d4.r3.module_list.4.1.module.1 | BatchNorm2d | 512
276 | model.backbone.d4.r3.module_list.4.1.module.2 | Mish | 0
277 | model.backbone.d4.r3.module_list.5 | ModuleList | 656 K
278 | model.backbone.d4.r3.module_list.5.0 | ConvBlock | 66 K
279 | model.backbone.d4.r3.module_list.5.0.module | Sequential | 66 K
280 | model.backbone.d4.r3.module_list.5.0.module.0 | Conv2d | 65 K
281 | model.backbone.d4.r3.module_list.5.0.module.1 | BatchNorm2d | 512
282 | model.backbone.d4.r3.module_list.5.0.module.2 | Mish | 0
283 | model.backbone.d4.r3.module_list.5.1 | ConvBlock | 590 K
284 | model.backbone.d4.r3.module_list.5.1.module | Sequential | 590 K
285 | model.backbone.d4.r3.module_list.5.1.module.0 | Conv2d | 589 K
286 | model.backbone.d4.r3.module_list.5.1.module.1 | BatchNorm2d | 512
287 | model.backbone.d4.r3.module_list.5.1.module.2 | Mish | 0
288 | model.backbone.d4.r3.module_list.6 | ModuleList | 656 K
289 | model.backbone.d4.r3.module_list.6.0 | ConvBlock | 66 K
290 | model.backbone.d4.r3.module_list.6.0.module | Sequential | 66 K
291 | model.backbone.d4.r3.module_list.6.0.module.0 | Conv2d | 65 K
292 | model.backbone.d4.r3.module_list.6.0.module.1 | BatchNorm2d | 512
293 | model.backbone.d4.r3.module_list.6.0.module.2 | Mish | 0
294 | model.backbone.d4.r3.module_list.6.1 | ConvBlock | 590 K
295 | model.backbone.d4.r3.module_list.6.1.module | Sequential | 590 K
296 | model.backbone.d4.r3.module_list.6.1.module.0 | Conv2d | 589 K
297 | model.backbone.d4.r3.module_list.6.1.module.1 | BatchNorm2d | 512
298 | model.backbone.d4.r3.module_list.6.1.module.2 | Mish | 0
299 | model.backbone.d4.r3.module_list.7 | ModuleList | 656 K
300 | model.backbone.d4.r3.module_list.7.0 | ConvBlock | 66 K
301 | model.backbone.d4.r3.module_list.7.0.module | Sequential | 66 K
302 | model.backbone.d4.r3.module_list.7.0.module.0 | Conv2d | 65 K
303 | model.backbone.d4.r3.module_list.7.0.module.1 | BatchNorm2d | 512
304 | model.backbone.d4.r3.module_list.7.0.module.2 | Mish | 0
305 | model.backbone.d4.r3.module_list.7.1 | ConvBlock | 590 K
306 | model.backbone.d4.r3.module_list.7.1.module | Sequential | 590 K
307 | model.backbone.d4.r3.module_list.7.1.module.0 | Conv2d | 589 K
308 | model.backbone.d4.r3.module_list.7.1.module.1 | BatchNorm2d | 512
309 | model.backbone.d4.r3.module_list.7.1.module.2 | Mish | 0
310 | model.backbone.d4.c4 | ConvBlock | 66 K
311 | model.backbone.d4.c4.module | Sequential | 66 K
312 | model.backbone.d4.c4.module.0 | Conv2d | 65 K
313 | model.backbone.d4.c4.module.1 | BatchNorm2d | 512
314 | model.backbone.d4.c4.module.2 | Mish | 0
315 | model.backbone.d4.dense_c2_c4 | ConvBlock | 131 K
316 | model.backbone.d4.dense_c2_c4.module | Sequential | 131 K
317 | model.backbone.d4.dense_c2_c4.module.0 | Conv2d | 131 K
318 | model.backbone.d4.dense_c2_c4.module.1 | BatchNorm2d | 512
319 | model.backbone.d4.dense_c2_c4.module.2 | Mish | 0
320 | model.backbone.d4.c5 | ConvBlock | 263 K
321 | model.backbone.d4.c5.module | Sequential | 263 K
322 | model.backbone.d4.c5.module.0 | Conv2d | 262 K
323 | model.backbone.d4.c5.module.1 | BatchNorm2d | 1 K
324 | model.backbone.d4.c5.module.2 | Mish | 0
325 | model.backbone.d5 | DownSampleBlock | 17 M
326 | model.backbone.d5.c1 | ConvBlock | 4 M
327 | model.backbone.d5.c1.module | Sequential | 4 M
328 | model.backbone.d5.c1.module.0 | Conv2d | 4 M
329 | model.backbone.d5.c1.module.1 | BatchNorm2d | 2 K
330 | model.backbone.d5.c1.module.2 | Mish | 0
331 | model.backbone.d5.c2 | ConvBlock | 525 K
332 | model.backbone.d5.c2.module | Sequential | 525 K
333 | model.backbone.d5.c2.module.0 | Conv2d | 524 K
334 | model.backbone.d5.c2.module.1 | BatchNorm2d | 1 K
335 | model.backbone.d5.c2.module.2 | Mish | 0
336 | model.backbone.d5.r3 | ResBlock | 10 M
337 | model.backbone.d5.r3.module_list | ModuleList | 10 M
338 | model.backbone.d5.r3.module_list.0 | ModuleList | 2 M
339 | model.backbone.d5.r3.module_list.0.0 | ConvBlock | 263 K
340 | model.backbone.d5.r3.module_list.0.0.module | Sequential | 263 K
341 | model.backbone.d5.r3.module_list.0.0.module.0 | Conv2d | 262 K
342 | model.backbone.d5.r3.module_list.0.0.module.1 | BatchNorm2d | 1 K
343 | model.backbone.d5.r3.module_list.0.0.module.2 | Mish | 0
344 | model.backbone.d5.r3.module_list.0.1 | ConvBlock | 2 M
345 | model.backbone.d5.r3.module_list.0.1.module | Sequential | 2 M
346 | model.backbone.d5.r3.module_list.0.1.module.0 | Conv2d | 2 M
347 | model.backbone.d5.r3.module_list.0.1.module.1 | BatchNorm2d | 1 K
348 | model.backbone.d5.r3.module_list.0.1.module.2 | Mish | 0
349 | model.backbone.d5.r3.module_list.1 | ModuleList | 2 M
350 | model.backbone.d5.r3.module_list.1.0 | ConvBlock | 263 K
351 | model.backbone.d5.r3.module_list.1.0.module | Sequential | 263 K
352 | model.backbone.d5.r3.module_list.1.0.module.0 | Conv2d | 262 K
353 | model.backbone.d5.r3.module_list.1.0.module.1 | BatchNorm2d | 1 K
354 | model.backbone.d5.r3.module_list.1.0.module.2 | Mish | 0
355 | model.backbone.d5.r3.module_list.1.1 | ConvBlock | 2 M
356 | model.backbone.d5.r3.module_list.1.1.module | Sequential | 2 M
357 | model.backbone.d5.r3.module_list.1.1.module.0 | Conv2d | 2 M
358 | model.backbone.d5.r3.module_list.1.1.module.1 | BatchNorm2d | 1 K
359 | model.backbone.d5.r3.module_list.1.1.module.2 | Mish | 0
360 | model.backbone.d5.r3.module_list.2 | ModuleList | 2 M
361 | model.backbone.d5.r3.module_list.2.0 | ConvBlock | 263 K
362 | model.backbone.d5.r3.module_list.2.0.module | Sequential | 263 K
363 | model.backbone.d5.r3.module_list.2.0.module.0 | Conv2d | 262 K
364 | model.backbone.d5.r3.module_list.2.0.module.1 | BatchNorm2d | 1 K
365 | model.backbone.d5.r3.module_list.2.0.module.2 | Mish | 0
366 | model.backbone.d5.r3.module_list.2.1 | ConvBlock | 2 M
367 | model.backbone.d5.r3.module_list.2.1.module | Sequential | 2 M
368 | model.backbone.d5.r3.module_list.2.1.module.0 | Conv2d | 2 M
369 | model.backbone.d5.r3.module_list.2.1.module.1 | BatchNorm2d | 1 K
370 | model.backbone.d5.r3.module_list.2.1.module.2 | Mish | 0
371 | model.backbone.d5.r3.module_list.3 | ModuleList | 2 M
372 | model.backbone.d5.r3.module_list.3.0 | ConvBlock | 263 K
373 | model.backbone.d5.r3.module_list.3.0.module | Sequential | 263 K
374 | model.backbone.d5.r3.module_list.3.0.module.0 | Conv2d | 262 K
375 | model.backbone.d5.r3.module_list.3.0.module.1 | BatchNorm2d | 1 K
376 | model.backbone.d5.r3.module_list.3.0.module.2 | Mish | 0
377 | model.backbone.d5.r3.module_list.3.1 | ConvBlock | 2 M
378 | model.backbone.d5.r3.module_list.3.1.module | Sequential | 2 M
379 | model.backbone.d5.r3.module_list.3.1.module.0 | Conv2d | 2 M
380 | model.backbone.d5.r3.module_list.3.1.module.1 | BatchNorm2d | 1 K
381 | model.backbone.d5.r3.module_list.3.1.module.2 | Mish | 0
382 | model.backbone.d5.c4 | ConvBlock | 263 K
383 | model.backbone.d5.c4.module | Sequential | 263 K
384 | model.backbone.d5.c4.module.0 | Conv2d | 262 K
385 | model.backbone.d5.c4.module.1 | BatchNorm2d | 1 K
386 | model.backbone.d5.c4.module.2 | Mish | 0
387 | model.backbone.d5.dense_c2_c4 | ConvBlock | 525 K
388 | model.backbone.d5.dense_c2_c4.module | Sequential | 525 K
389 | model.backbone.d5.dense_c2_c4.module.0 | Conv2d | 524 K
390 | model.backbone.d5.dense_c2_c4.module.1 | BatchNorm2d | 1 K
391 | model.backbone.d5.dense_c2_c4.module.2 | Mish | 0
392 | model.backbone.d5.c5 | ConvBlock | 1 M
393 | model.backbone.d5.c5.module | Sequential | 1 M
394 | model.backbone.d5.c5.module.0 | Conv2d | 1 M
395 | model.backbone.d5.c5.module.1 | BatchNorm2d | 2 K
396 | model.backbone.d5.c5.module.2 | Mish | 0
397 | model.neck | Neck | 21 M
398 | model.neck.c1 | ConvBlock | 525 K
399 | model.neck.c1.module | Sequential | 525 K
400 | model.neck.c1.module.0 | Conv2d | 524 K
401 | model.neck.c1.module.1 | BatchNorm2d | 1 K
402 | model.neck.c1.module.2 | LeakyReLU | 0
403 | model.neck.c2 | ConvBlock | 4 M
404 | model.neck.c2.module | Sequential | 4 M
405 | model.neck.c2.module.0 | Conv2d | 4 M
406 | model.neck.c2.module.1 | BatchNorm2d | 2 K
407 | model.neck.c2.module.2 | LeakyReLU | 0
408 | model.neck.c3 | ConvBlock | 525 K
409 | model.neck.c3.module | Sequential | 525 K
410 | model.neck.c3.module.0 | Conv2d | 524 K
411 | model.neck.c3.module.1 | BatchNorm2d | 1 K
412 | model.neck.c3.module.2 | LeakyReLU | 0
413 | model.neck.mp4_1 | MaxPool2d | 0
414 | model.neck.mp4_2 | MaxPool2d | 0
415 | model.neck.mp4_3 | MaxPool2d | 0
416 | model.neck.c5 | ConvBlock | 1 M
417 | model.neck.c5.module | Sequential | 1 M
418 | model.neck.c5.module.0 | Conv2d | 1 M
419 | model.neck.c5.module.1 | BatchNorm2d | 1 K
420 | model.neck.c5.module.2 | LeakyReLU | 0
421 | model.neck.c6 | ConvBlock | 4 M
422 | model.neck.c6.module | Sequential | 4 M
423 | model.neck.c6.module.0 | Conv2d | 4 M
424 | model.neck.c6.module.1 | BatchNorm2d | 2 K
425 | model.neck.c6.module.2 | LeakyReLU | 0
426 | model.neck.c7 | ConvBlock | 525 K
427 | model.neck.c7.module | Sequential | 525 K
428 | model.neck.c7.module.0 | Conv2d | 524 K
429 | model.neck.c7.module.1 | BatchNorm2d | 1 K
430 | model.neck.c7.module.2 | LeakyReLU | 0
431 | model.neck.PAN8 | PAN_Layer | 3 M
432 | model.neck.PAN8.c1 | ConvBlock | 131 K
433 | model.neck.PAN8.c1.module | Sequential | 131 K
434 | model.neck.PAN8.c1.module.0 | Conv2d | 131 K
435 | model.neck.PAN8.c1.module.1 | BatchNorm2d | 512
436 | model.neck.PAN8.c1.module.2 | LeakyReLU | 0
437 | model.neck.PAN8.u2 | Upsample | 0
438 | model.neck.PAN8.c2_from_upsampled | ConvBlock | 131 K
439 | model.neck.PAN8.c2_from_upsampled.module | Sequential | 131 K
440 | model.neck.PAN8.c2_from_upsampled.module.0 | Conv2d | 131 K
441 | model.neck.PAN8.c2_from_upsampled.module.1 | BatchNorm2d | 512
442 | model.neck.PAN8.c2_from_upsampled.module.2 | LeakyReLU | 0
443 | model.neck.PAN8.c3 | ConvBlock | 131 K
444 | model.neck.PAN8.c3.module | Sequential | 131 K
445 | model.neck.PAN8.c3.module.0 | Conv2d | 131 K
446 | model.neck.PAN8.c3.module.1 | BatchNorm2d | 512
447 | model.neck.PAN8.c3.module.2 | LeakyReLU | 0
448 | model.neck.PAN8.c4 | ConvBlock | 1 M
449 | model.neck.PAN8.c4.module | Sequential | 1 M
450 | model.neck.PAN8.c4.module.0 | Conv2d | 1 M
451 | model.neck.PAN8.c4.module.1 | BatchNorm2d | 1 K
452 | model.neck.PAN8.c4.module.2 | LeakyReLU | 0
453 | model.neck.PAN8.c5 | ConvBlock | 131 K
454 | model.neck.PAN8.c5.module | Sequential | 131 K
455 | model.neck.PAN8.c5.module.0 | Conv2d | 131 K
456 | model.neck.PAN8.c5.module.1 | BatchNorm2d | 512
457 | model.neck.PAN8.c5.module.2 | LeakyReLU | 0
458 | model.neck.PAN8.c6 | ConvBlock | 1 M
459 | model.neck.PAN8.c6.module | Sequential | 1 M
460 | model.neck.PAN8.c6.module.0 | Conv2d | 1 M
461 | model.neck.PAN8.c6.module.1 | BatchNorm2d | 1 K
462 | model.neck.PAN8.c6.module.2 | LeakyReLU | 0
463 | model.neck.PAN8.c7 | ConvBlock | 131 K
464 | model.neck.PAN8.c7.module | Sequential | 131 K
465 | model.neck.PAN8.c7.module.0 | Conv2d | 131 K
466 | model.neck.PAN8.c7.module.1 | BatchNorm2d | 512
467 | model.neck.PAN8.c7.module.2 | LeakyReLU | 0
468 | model.neck.PAN9 | PAN_Layer | 755 K
469 | model.neck.PAN9.c1 | ConvBlock | 33 K
470 | model.neck.PAN9.c1.module | Sequential | 33 K
471 | model.neck.PAN9.c1.module.0 | Conv2d | 32 K
472 | model.neck.PAN9.c1.module.1 | BatchNorm2d | 256
473 | model.neck.PAN9.c1.module.2 | LeakyReLU | 0
474 | model.neck.PAN9.u2 | Upsample | 0
475 | model.neck.PAN9.c2_from_upsampled | ConvBlock | 33 K
476 | model.neck.PAN9.c2_from_upsampled.module | Sequential | 33 K
477 | model.neck.PAN9.c2_from_upsampled.module.0 | Conv2d | 32 K
478 | model.neck.PAN9.c2_from_upsampled.module.1 | BatchNorm2d | 256
479 | model.neck.PAN9.c2_from_upsampled.module.2 | LeakyReLU | 0
480 | model.neck.PAN9.c3 | ConvBlock | 33 K
481 | model.neck.PAN9.c3.module | Sequential | 33 K
482 | model.neck.PAN9.c3.module.0 | Conv2d | 32 K
483 | model.neck.PAN9.c3.module.1 | BatchNorm2d | 256
484 | model.neck.PAN9.c3.module.2 | LeakyReLU | 0
485 | model.neck.PAN9.c4 | ConvBlock | 295 K
486 | model.neck.PAN9.c4.module | Sequential | 295 K
487 | model.neck.PAN9.c4.module.0 | Conv2d | 294 K
488 | model.neck.PAN9.c4.module.1 | BatchNorm2d | 512
489 | model.neck.PAN9.c4.module.2 | LeakyReLU | 0
490 | model.neck.PAN9.c5 | ConvBlock | 33 K
491 | model.neck.PAN9.c5.module | Sequential | 33 K
492 | model.neck.PAN9.c5.module.0 | Conv2d | 32 K
493 | model.neck.PAN9.c5.module.1 | BatchNorm2d | 256
494 | model.neck.PAN9.c5.module.2 | LeakyReLU | 0
495 | model.neck.PAN9.c6 | ConvBlock | 295 K
496 | model.neck.PAN9.c6.module | Sequential | 295 K
497 | model.neck.PAN9.c6.module.0 | Conv2d | 294 K
498 | model.neck.PAN9.c6.module.1 | BatchNorm2d | 512
499 | model.neck.PAN9.c6.module.2 | LeakyReLU | 0
500 | model.neck.PAN9.c7 | ConvBlock | 33 K
501 | model.neck.PAN9.c7.module | Sequential | 33 K
502 | model.neck.PAN9.c7.module.0 | Conv2d | 32 K
503 | model.neck.PAN9.c7.module.1 | BatchNorm2d | 256
504 | model.neck.PAN9.c7.module.2 | LeakyReLU | 0
505 | model.neck.ACFF_0 | ACFF | 4 M
506 | model.neck.ACFF_0.stride_level_1 | ConvBlock | 1 M
507 | model.neck.ACFF_0.stride_level_1.module | Sequential | 1 M
508 | model.neck.ACFF_0.stride_level_1.module.0 | Conv2d | 1 M
509 | model.neck.ACFF_0.stride_level_1.module.1 | BatchNorm2d | 1 K
510 | model.neck.ACFF_0.stride_level_1.module.2 | LeakyReLU | 0
511 | model.neck.ACFF_0.stride_level_2 | ConvBlock | 590 K
512 | model.neck.ACFF_0.stride_level_2.module | Sequential | 590 K
513 | model.neck.ACFF_0.stride_level_2.module.0 | Conv2d | 589 K
514 | model.neck.ACFF_0.stride_level_2.module.1 | BatchNorm2d | 1 K
515 | model.neck.ACFF_0.stride_level_2.module.2 | LeakyReLU | 0
516 | model.neck.ACFF_0.expand | ConvBlock | 2 M
517 | model.neck.ACFF_0.expand.module | Sequential | 2 M
518 | model.neck.ACFF_0.expand.module.0 | Conv2d | 2 M
519 | model.neck.ACFF_0.expand.module.1 | BatchNorm2d | 1 K
520 | model.neck.ACFF_0.expand.module.2 | LeakyReLU | 0
521 | model.neck.ACFF_1 | ACFF | 1 M
522 | model.neck.ACFF_1.compress_level_0 | ConvBlock | 131 K
523 | model.neck.ACFF_1.compress_level_0.module | Sequential | 131 K
524 | model.neck.ACFF_1.compress_level_0.module.0 | Conv2d | 131 K
525 | model.neck.ACFF_1.compress_level_0.module.1 | BatchNorm2d | 512
526 | model.neck.ACFF_1.compress_level_0.module.2 | LeakyReLU | 0
527 | model.neck.ACFF_1.stride_level_2 | ConvBlock | 295 K
528 | model.neck.ACFF_1.stride_level_2.module | Sequential | 295 K
529 | model.neck.ACFF_1.stride_level_2.module.0 | Conv2d | 294 K
530 | model.neck.ACFF_1.stride_level_2.module.1 | BatchNorm2d | 512
531 | model.neck.ACFF_1.stride_level_2.module.2 | LeakyReLU | 0
532 | model.neck.ACFF_1.expand | ConvBlock | 590 K
533 | model.neck.ACFF_1.expand.module | Sequential | 590 K
534 | model.neck.ACFF_1.expand.module.0 | Conv2d | 589 K
535 | model.neck.ACFF_1.expand.module.1 | BatchNorm2d | 512
536 | model.neck.ACFF_1.expand.module.2 | LeakyReLU | 0
537 | model.neck.ACFF_2 | ACFF | 247 K
538 | model.neck.ACFF_2.compress_level_0 | ConvBlock | 65 K
539 | model.neck.ACFF_2.compress_level_0.module | Sequential | 65 K
540 | model.neck.ACFF_2.compress_level_0.module.0 | Conv2d | 65 K
541 | model.neck.ACFF_2.compress_level_0.module.1 | BatchNorm2d | 256
542 | model.neck.ACFF_2.compress_level_0.module.2 | LeakyReLU | 0
543 | model.neck.ACFF_2.compress_level_1 | ConvBlock | 33 K
544 | model.neck.ACFF_2.compress_level_1.module | Sequential | 33 K
545 | model.neck.ACFF_2.compress_level_1.module.0 | Conv2d | 32 K
546 | model.neck.ACFF_2.compress_level_1.module.1 | BatchNorm2d | 256
547 | model.neck.ACFF_2.compress_level_1.module.2 | LeakyReLU | 0
548 | model.neck.ACFF_2.expand | ConvBlock | 147 K
549 | model.neck.ACFF_2.expand.module | Sequential | 147 K
550 | model.neck.ACFF_2.expand.module.0 | Conv2d | 147 K
551 | model.neck.ACFF_2.expand.module.1 | BatchNorm2d | 256
552 | model.neck.ACFF_2.expand.module.2 | LeakyReLU | 0
553 | model.head | Head | 22 M
554 | model.head.ho1 | HeadOutput | 430 K
555 | model.head.ho1.c1 | ConvBlock | 295 K
556 | model.head.ho1.c1.module | Sequential | 295 K
557 | model.head.ho1.c1.module.0 | Conv2d | 294 K
558 | model.head.ho1.c1.module.1 | BatchNorm2d | 512
559 | model.head.ho1.c1.module.2 | LeakyReLU | 0
560 | model.head.ho1.c2 | ConvBlock | 134 K
561 | model.head.ho1.c2.module | Sequential | 134 K
562 | model.head.ho1.c2.module.0 | Conv2d | 134 K
563 | model.head.hp2 | HeadPreprocessing | 3 M
564 | model.head.hp2.c1 | ConvBlock | 295 K
565 | model.head.hp2.c1.module | Sequential | 295 K
566 | model.head.hp2.c1.module.0 | Conv2d | 294 K
567 | model.head.hp2.c1.module.1 | BatchNorm2d | 512
568 | model.head.hp2.c1.module.2 | LeakyReLU | 0
569 | model.head.hp2.c2 | ConvBlock | 131 K
570 | model.head.hp2.c2.module | Sequential | 131 K
571 | model.head.hp2.c2.module.0 | Conv2d | 131 K
572 | model.head.hp2.c2.module.1 | BatchNorm2d | 512
573 | model.head.hp2.c2.module.2 | LeakyReLU | 0
574 | model.head.hp2.c3 | ConvBlock | 1 M
575 | model.head.hp2.c3.module | Sequential | 1 M
576 | model.head.hp2.c3.module.0 | Conv2d | 1 M
577 | model.head.hp2.c3.module.1 | BatchNorm2d | 1 K
578 | model.head.hp2.c3.module.2 | LeakyReLU | 0
579 | model.head.hp2.c4 | ConvBlock | 131 K
580 | model.head.hp2.c4.module | Sequential | 131 K
581 | model.head.hp2.c4.module.0 | Conv2d | 131 K
582 | model.head.hp2.c4.module.1 | BatchNorm2d | 512
583 | model.head.hp2.c4.module.2 | LeakyReLU | 0
584 | model.head.hp2.c5 | ConvBlock | 1 M
585 | model.head.hp2.c5.module | Sequential | 1 M
586 | model.head.hp2.c5.module.0 | Conv2d | 1 M
587 | model.head.hp2.c5.module.1 | BatchNorm2d | 1 K
588 | model.head.hp2.c5.module.2 | LeakyReLU | 0
589 | model.head.hp2.c6 | ConvBlock | 131 K
590 | model.head.hp2.c6.module | Sequential | 131 K
591 | model.head.hp2.c6.module.0 | Conv2d | 131 K
592 | model.head.hp2.c6.module.1 | BatchNorm2d | 512
593 | model.head.hp2.c6.module.2 | LeakyReLU | 0
594 | model.head.ho2 | HeadOutput | 1 M
595 | model.head.ho2.c1 | ConvBlock | 1 M
596 | model.head.ho2.c1.module | Sequential | 1 M
597 | model.head.ho2.c1.module.0 | Conv2d | 1 M
598 | model.head.ho2.c1.module.1 | BatchNorm2d | 1 K
599 | model.head.ho2.c1.module.2 | LeakyReLU | 0
600 | model.head.ho2.c2 | ConvBlock | 269 K
601 | model.head.ho2.c2.module | Sequential | 269 K
602 | model.head.ho2.c2.module.0 | Conv2d | 269 K
603 | model.head.hp3 | HeadPreprocessing | 12 M
604 | model.head.hp3.c1 | ConvBlock | 1 M
605 | model.head.hp3.c1.module | Sequential | 1 M
606 | model.head.hp3.c1.module.0 | Conv2d | 1 M
607 | model.head.hp3.c1.module.1 | BatchNorm2d | 1 K
608 | model.head.hp3.c1.module.2 | LeakyReLU | 0
609 | model.head.hp3.c2 | ConvBlock | 525 K
610 | model.head.hp3.c2.module | Sequential | 525 K
611 | model.head.hp3.c2.module.0 | Conv2d | 524 K
612 | model.head.hp3.c2.module.1 | BatchNorm2d | 1 K
613 | model.head.hp3.c2.module.2 | LeakyReLU | 0
614 | model.head.hp3.c3 | ConvBlock | 4 M
615 | model.head.hp3.c3.module | Sequential | 4 M
616 | model.head.hp3.c3.module.0 | Conv2d | 4 M
617 | model.head.hp3.c3.module.1 | BatchNorm2d | 2 K
618 | model.head.hp3.c3.module.2 | LeakyReLU | 0
619 | model.head.hp3.c4 | ConvBlock | 525 K
620 | model.head.hp3.c4.module | Sequential | 525 K
621 | model.head.hp3.c4.module.0 | Conv2d | 524 K
622 | model.head.hp3.c4.module.1 | BatchNorm2d | 1 K
623 | model.head.hp3.c4.module.2 | LeakyReLU | 0
624 | model.head.hp3.c5 | ConvBlock | 4 M
625 | model.head.hp3.c5.module | Sequential | 4 M
626 | model.head.hp3.c5.module.0 | Conv2d | 4 M
627 | model.head.hp3.c5.module.1 | BatchNorm2d | 2 K
628 | model.head.hp3.c5.module.2 | LeakyReLU | 0
629 | model.head.hp3.c6 | ConvBlock | 525 K
630 | model.head.hp3.c6.module | Sequential | 525 K
631 | model.head.hp3.c6.module.0 | Conv2d | 524 K
632 | model.head.hp3.c6.module.1 | BatchNorm2d | 1 K
633 | model.head.hp3.c6.module.2 | LeakyReLU | 0
634 | model.head.ho3 | HeadOutput | 5 M
635 | model.head.ho3.c1 | ConvBlock | 4 M
636 | model.head.ho3.c1.module | Sequential | 4 M
637 | model.head.ho3.c1.module.0 | Conv2d | 4 M
638 | model.head.ho3.c1.module.1 | BatchNorm2d | 2 K
639 | model.head.ho3.c1.module.2 | LeakyReLU | 0
640 | model.head.ho3.c2 | ConvBlock | 538 K
641 | model.head.ho3.c2.module | Sequential | 538 K
642 | model.head.ho3.c2.module.0 | Conv2d | 538 K
643 | model.yolo1 | YOLOLayer | 0
644 | model.yolo2 | YOLOLayer | 0
645 | model.yolo3 | YOLOLayer | 0
Ranger optimizer loaded.
Gradient Centralization usage = True
GC applied to both conv and fc layers
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-10-0524224c2dd3> in <module>
----> 1 r = t.lr_find(m, min_lr=1e-10, max_lr=1, early_stop_threshold=None)
2 r.plot()
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/pytorch_lightning/trainer/lr_finder.py in lr_find(self, model, train_dataloader, val_dataloaders, min_lr, max_lr, num_training, mode, early_stop_threshold, num_accumulation_steps)
168 self.fit(model,
169 train_dataloader=train_dataloader,
--> 170 val_dataloaders=val_dataloaders)
171
172 # Prompt if we stopped early
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/pytorch_lightning/trainer/trainer.py in fit(self, model, train_dataloader, val_dataloaders)
885 self.optimizers, self.lr_schedulers, self.optimizer_frequencies = self.init_optimizers(model)
886
--> 887 self.run_pretrain_routine(model)
888
889 # return 1 when finished
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/pytorch_lightning/trainer/trainer.py in run_pretrain_routine(self, model)
999 self.val_dataloaders,
1000 self.num_sanity_val_steps,
-> 1001 False)
1002 _, _, _, callback_metrics, _ = self.process_output(eval_results)
1003
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/pytorch_lightning/trainer/evaluation_loop.py in _evaluate(self, model, dataloaders, max_batches, test_mode)
254 dataloader = dataloader.per_device_loader(device)
255
--> 256 for batch_idx, batch in enumerate(dataloader):
257 if batch is None:
258 continue
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/torch/utils/data/dataloader.py in __next__(self)
519 if self._sampler_iter is None:
520 self._reset()
--> 521 data = self._next_data()
522 self._num_yielded += 1
523 if self._dataset_kind == _DatasetKind.Iterable and \
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/torch/utils/data/dataloader.py in _next_data(self)
1201 else:
1202 del self._task_info[idx]
-> 1203 return self._process_data(data)
1204
1205 def _try_put_index(self):
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/torch/utils/data/dataloader.py in _process_data(self, data)
1227 self._try_put_index()
1228 if isinstance(data, ExceptionWrapper):
-> 1229 data.reraise()
1230 return data
1231
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/torch/_utils.py in reraise(self)
423 # have message field
424 raise self.exc_type(message=msg)
--> 425 raise self.exc_type(msg)
426
427
TypeError: function takes exactly 5 arguments (1 given)
Originally posted by @a-haja in #18 (comment)
Hello, after reading your code carefully, I feel I have benefited a lot. There are some problems in the ipynb of operation training, which may be related to the different operation environment. Can you tell me the specific operation environment?
Hello I was wondering how do you deal with training samples that don't have bboxes?
I'm trying to run your code, but it never seems to get into the training_step
callback of the module and also I'm always getting nan
for the loss. I'm using version 1.0.5
if that helps, also where have you downloaded the dataset for training?
I have a model in darknet repository which I want to convert that weights and retrain using this repo, can we do that?
Hi,
can you explain why the IoU ignore threshold is only applied to the no-object mask, but not to the object mask (or why the no-object mask is not simply the inversion of the object mask in general):
Yet-Another-YOLOv4-Pytorch/model.py
Line 834 in 47b0450
I would have expected
obj_mask[b[i], anchor_ious < ignore_thres, gj[i], gi[i]] = 0
in addition.
please help! thanks
I am getting the following error when running video_demo.py. Can somebody please suggest how to fix it?
qt.qpa.xcb: could not connect to display
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "/home/orazayy/.local/lib/python3.8/site-packages/cv2/qt/plugins" even though it was found.
This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.
Available platform plugins are: xcb.
This is a really great repo and I'm enjoying reading it. I've noticed it throws an error in lines 925:926 of model.py, since the target tensors are masks (either 0. or 1.).
The complaint is that this function can't be autocast, and one should use F.binary_cross_entropy_with_logits instead. I'm curious if you have encountered this and whether it is worth updating your model.py to prevent this error?
Thanks!
Thanks for this repo. !
I was wondering if you had any numbers you could share comparing what AP this repo. is able to train to on coco or similar and maybe FPS. - just so we can compare to AlexeyAB's darknet.
Which pyTorch version are you using? Can you please provide us with a requirments.txt file?
I am trying to run "Training YOLOv4 .ipynb". I am facing different errors. for example: I needed to change from pytorch_lightning.callbacks import LearningRateLogger
to from pytorch_lightning.callbacks import LearningRateMonitor
as I have the following pytorch-ligthning version
$ conda list | grep pytorch-lightning
pytorch-lightning 1.3.8 pypi_0 pypi
Now, I am getting the following error:
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
<ipython-input-27-44f6feb2fd1c> in <module>
----> 1 m = YOLOv4PL(hparams2)
<ipython-input-25-96522b0b7356> in __init__(self, hparams)
9 print (hparams)
10
---> 11 self.hparams = hparams
12
13 self.train_ds = ListDataset(hparams.train_ds, train=True, img_extensions=hparams.img_extensions)
~/anaconda3/envs/django_pyimagej_tf/lib/python3.7/site-packages/torch/nn/modules/module.py in __setattr__(self, name, value)
1176 buffers[name] = value
1177 else:
-> 1178 object.__setattr__(self, name, value)
1179
1180 def __delattr__(self, name):
AttributeError: can't set attribute
Do you have any idea why the hparams cannot be setted? I would assume that it is because of the different versions of pytorch versions.
Would you please give a concrete example?
Dear @VCasecnikovs,
How to get the confidence value next to object labels in the following picture: https://github.com/VCasecnikovs/Yet-Another-YOLOv4-Pytorch/blob/master/github_imgs/from_net.png?
I want to get these values when running this code for example:
bboxes, labels = utils.get_bboxes_from_anchors(anchors, 0.4, 0.5, coco_dict)
Allow to download weights automatically by using PyTorch Hub
Thanks for your nice work!
But I have a question about dropblock.The original paper writes "We only sample mask from shaded green region in which each sampled entry can expanded to a mask fully contained inside the feature map"
So the mask sample point will smaller than the input and then expanded to the input size. In your code, it seems like the sample points get from the input size.
Am I right? Thanks again for your GREAT work!
What are the step to run the model on sample image?
What are the required package list?
Could you please clearly state these instructions?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.