Random Old Comic: Blanche Has a Bat'leth Now https://www.toyboxcomix.com/2018/07/24/blanche-has-a-batleth-now/ Blanche Has a Bat’leth Now #BlancheDevereaux #Buzzer #DC #GIJoe #GoldenGirls #Ripper #Robin #Torch
Random Old Comic: Blanche Has a Bat'leth Now https://www.toyboxcomix.com/2018/07/24/blanche-has-a-batleth-now/ Blanche Has a Bat’leth Now #BlancheDevereaux #Buzzer #DC #GIJoe #GoldenGirls #Ripper #Robin #Torch
Import GPU: Python Programming with CUDA https://hackaday.com/2025/02/25/import-gpu-python-programming-with-cuda/ #graphicsprocessing #parallelprocessing #developer #pytorch #NVIDIA #python #torch #News #CUDA #gpu
Powerful Flashlight Gets Active Air Cooling - LEDs were once little more than weedy little indicators with low light output. Tod... - https://hackaday.com/2025/02/09/powerful-flashlight-gets-active-air-cooling/ #ledhacks #torch #led
Now training models on the #AirQualityInMadrid dataset! Using LSTM, ARIMA, and other models with #Torch and #Darts. Excited to see how each model performs for time series forecasting. #MachineLearning #TimeSeries #Python #DataTalksClub #zoomcamp #Machinelearning
Hey, I'm stuck with this PyTorch problem and no one has answered on Stack Overflow yet. Need help creating a new tensor by using values from a 3D tensor as indices for a 1D tensor, efficiently and preferably without loops. Any PyTorch experts around?
https://stackoverflow.com/q/79389115/21239182
Boosts appreciated - thanks!
Melting 925 silver.
In the dungeon of dwarves and dragons
Preparing for new handmade jewelry designs. Rough and beautiful!
#artisan #working #process #handmade #jewelryaddict #jewelrylover #torch
(I posted it with pixelfed account too, don’t know right now, which one to keep still #newhere any suggestions?)
US troops on board their landing craft going in to land at Oran, Algeria as Operation Torch gets underway - North Africa, Nov 1942. CREDIT : US Army Archives #ww2 #oran #worldwar2 #history #torch #algeria #northafrica #usarmy #landingcraft
This is a very useful #torch function if your thinking tends to gravitate towards hypernetworks (as mine does for some reason):
https://pytorch.org/docs/stable/generated/torch.func.functional_call.html#torch.func.functional_call
It allows you to call the forward of any module, but to replace all of its parameters by a set of tensors of your own choosing.
The tensors get gradients so they can be derived from some upstream network.
The posts going around with various "go bag" ideas have the right idea but I'd throw in this - when buying batteries to go with your torch (flashlight) get ones with a long shelf life. Lithium chemistry AAs can have a 20 year shelf life, and CR123s (different chemistry according to the internet) can go 10.
(on flashlights I'd recommend having at least a pair, one that takes a generally easy to come by cell size like AA, and a good high capacity rechargeable. I am very normal about flashlights)