Neel Gupta
Is it compulsory to use `TPOT cuML` for GPU acceleration, or would the vanilla "pip" install use GPU anyways?
Ahh, right. I was wondering whether it would use a standalone piece of code that would run the model. Still, thanx a ton for the info!
it is not an issue, per se. it is more of a feature request/question about how we can resume TPOT training even after shutting down our computer. apparently, it's not...
@Manidills I had the same problem. I recommend you buy Colab PRO for extra RAM which would prevent Colab from crashing. Even then, is there any way to save on...
Yep, upgrading `jaxlib` from `0.4.25` -> `0.4.26` works locally. So this is definitely a version issue. However, on my docker image on a `TPU v3-8` I have libs: ```py jax:...
+1 its pretty annoying. What I feel users want is: - Cmd/Ctrl + Tab ==> Switch to last active tab. Pressing the shortcut again should bring to back to the...
With some basic formatting (for quick copy-and-paste): ```py def count_params(model: eqx.Module): num_params = sum(x.size for x in jax.tree_util.tree_leaves(eqx.filter(model, eqx.is_array))) num_millions = num_params / 1_000_000 print(f"Model # of parameters: {num_millions:.2f}M") count_params(model)...
@imawizard thank you for the awesome bindings. I'm having a problem with using `tab` for co-pilot completions. While it's not really crucial for my workflow, it's a bit annoying not...
I'm on the latest commit for those `keybindings.json`, so its different and these lines aren't there. However, I commented out this snippet: ```py { "key": "tab", "command": "dance.run", "args": {...
@nashid You haven't installed CUDA, which is required to run TensorFlow ops in GPU and utilize its computational cores. However, CUDA is proprietary to Nvidia (whereas AMD uses OpenCL). So...