Page 1 of 1

Off Topic DeepFaceLab error 6) train SAEHD.bat

Posted: Fri Aug 16, 2024 9:32 pm
by oliver21a
Good afternoon people! I'm using: DeepFaceLab NVIDIA RTX 3000 series build_11_20_2021 Present in this option: 6) train SAEHD.bat Can you help me. My video card is RTX 4070 Ti Super 16Gb Error: Initializing models: 20%|############6 | 1/5 [00:10<00:40, 10.24s/it] Error: OOM when allocating tensor with shape[2654208,1024] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc [[node inter_AB/dense1/weight/Initializer/random_uniform/RandomUniform (defined at D:\Fake\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\core\leras\layers\Dense.py:47) ]] Hint: If you want to see a list of allocated tensors when OOM happens, add report_tensor_allocations_upon_oom to RunOptions for current allocation info. This isn't available when running in Eager mode.

Link:
https://imgbox.com/Jt8nzjDG

Re: Off Topic DeepFaceLab error 6) train SAEHD.bat

Posted: Fri Aug 16, 2024 9:32 pm
by oliver21a
Working! I used these settings. But something strange appeared, my video card has 16.00GB and the image only shows 13.12GB

Link:
https://imgbox.com/smAoKrnH