Failed to build yolo_v3_tiny_pan_lstm from issue 3114

[peijason]

If you want to report a bug - provide:
* description of a bug: Failed to build yolo_v3_tiny_pan_lstm from issue 3114
* what command do you use? darknet detector train data/self_driving.data cfg/yolo_v3_tiny_pan_lstm.cfg yolov3-tiny.conv.14 -map -dont_show
* do you use Win/Linux/Mac? Linux Ubuntu 20.04

⋊> ....../darknet on master ◦ nvidia-smi                                                                                                                                                         15:21:28
Fri Aug 21 15:21:31 2020       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 450.57       Driver Version: 450.57       CUDA Version: 11.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  GeForce GTX 1050Ti  Off  | 00000000:01:00.0 Off |                  N/A |
| N/A   47C    P0    N/A /  N/A |    816MiB /  4042MiB |      2%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A      1377      G   /usr/lib/xorg/Xorg                 80MiB |
|    0   N/A  N/A      2043      G   /usr/lib/xorg/Xorg                323MiB |
|    0   N/A  N/A      2234      G   /usr/bin/gnome-shell              121MiB |
|    0   N/A  N/A      2406      G   seadrive-gui                        1MiB |
|    0   N/A  N/A      2689      G   ...AAAAAAAAA= --shared-files       24MiB |
|    0   N/A  N/A      3216      G   ...token=7574150999326190285      141MiB |
|    0   N/A  N/A      3872      G   ...AAAAAAAAA= --shared-files      113MiB |
+-----------------------------------------------------------------------------+

CUDNN version: 8.0.2

⋊> ....../darknet on master ◦ ./darknet detector train data/self_driving.data cfg/yolo_v3_tiny_pan_lstm.cfg yolov3-tiny.conv.14 -map -dont_show 
......
[yolo] params: iou loss: mse (2), iou_norm: 0.75, cls_norm: 1.00, scale_x_y: 1.00
Total BFLOPS 23.447 
avg_outputs = 689534 
Loading weights from yolov3-tiny.conv.14...
 seen 64, trained: 0 K-images (0 Kilo-batches_64) 
Done! Loaded 14 layers from weights-file 
Learning Rate: 0.001, Momentum: 0.9, Decay: 0.0005
 Detection layer: 46 - type = 27 
 Detection layer: 53 - type = 27 
 Detection layer: 59 - type = 27 

 Tracking! batch = 6, subdiv = 16, time_steps = 3, mini_batch = 2 
 Create 4 permanent cpu-threads 
 sequential_subdivisions = 4, sequence = 4 
Loaded: 1.888440 seconds
v3 (mse loss, Normalizer: (iou: 0.75, cls: 1.00) Region 46 Avg (IOU: 0.319579, GIOU: 0.258581), Class: 0.519559, Obj: 0.416531, No Obj: 0.489394, .5R: 0.000000, .75R: 0.000000, count: 9, class_loss = 3612.610107, iou_loss = 3.847982, total_loss = 3616.458008 
v3 (mse loss, Normalizer: (iou: 0.75, cls: 1.00) Region 53 Avg (IOU: 0.000000, GIOU: 0.000000), Class: 0.000000, Obj: 0.000000, No Obj: 0.562382, .5R: 0.000000, .75R: 0.000000, count: 1, class_loss = 1152.953857, iou_loss = 0.000000, total_loss = 1152.953857 
v3 (mse loss, Normalizer: (iou: 0.75, cls: 1.00) Region 59 Avg (IOU: 0.000000, GIOU: 0.000000), Class: 0.000000, Obj: 0.000000, No Obj: 0.554808, .5R: 0.000000, .75R: 0.000000, count: 1, class_loss = 282.553375, iou_loss = 0.000000, total_loss = 282.553375 
CUDA status Error: file: ....../darknet/src/maxpool_layer_kernels.cu : () : line: 248 : build time: Aug  4 2020 - 12:41:13 

 CUDA Error: an illegal memory access was encountered
CUDA Error: an illegal memory access was encountered: Success

And some times, it's showing:

 cuDNN status Error in: file: ....../darknet/src/convolutional_kernels.cu : () : line: 821 : build time: Aug  4 2020 - 12:41:04

Cheers

[arnaud-nt2i]

hi @peijason
I have the same issue on W10, have you found an answer?

[arnaud-nt2i]

Ok I have found a walkaround, by using "Yolo v3 optimal" here: https://github.com/AlexeyAB/darknet/releases/tag/darknet_yolo_v3_optimal