Oliver Guy
Oliver Guy
Setting `torch.backends.cudnn.benchmark = True` at the beginning of the script fixed it for me.
Hi John, No problem at all, and thanks for checking back ! I was indeed able to train, and I'm now working on crafting my own coco dataset. Regards, Oliver
Hey all, I'm having a similar error as @NicolaGugole using the training dataset downloaded through the provided shell script. Any tips on how to fix this would be greatly appreciated...
@NicolaGugole I only modified those in the json, but I'm having issues with CudNN not finding the convolution algorithm (see #54).
I would love to see this fixed, especially as iterating over RaggedTensors seems to not properly work in a lot of cases.
To add more data to this, the following config is supposed to work using OSC 52: ```lua if os.getenv 'SSH_CLIENT' ~= nil then local function noop() return {} end vim.g.clipboard...
It unfortunately did not work within Neovim; the test I ran shows that the terminal emulator itself is capable of OSC 52 (it also works within tmux). Be aware that...
Apologies for the noise, but this issue on the nvim v0.11 roadmap seems related: https://github.com/neovim/neovim/issues/28792
Can confirm, also requires setting `front_end = "OpenGL"`. I cannot overstate how unusable WezTerm is with `__NV_DISABLE_EXPLICIT_SYNC=1`, as characters input during a lag spike will be duplicated; this basically makes...
Hi, getting the same issue after a full system upgrade (Arch/Wayland), except setting `VideoBackend=0` does not prevent crashes.