Alex Mai

How do I accept a completion without writing a character?

4

Or, more importantly, how do I configure this behavior? I think the key used to be enter to accept a completion without typing, but now I can't seem to trigger...

Out of Bounds issue when using own feature extractor

12

I am trying to use hloc to test my own feature extractor but I'm running into a out of bounds issue in localize_sfm.main. Here is the error. There are a...

Depth on UE4.25 is very incorrect

5

- Operating System: Windows - UE4 Version: 4.25 - UnrealCV Version: Branch 4.25 - Client (python2, 3 or matlab): python3 - Problem Description: Depth is incorrect/always the same. The depth...

I changed out the matrix * vector multiplication for matrix * matrix multiplication. This fixes the error "THFloatBlas_gemv is undefined" as well as the error "THCudaBlas_Sgemv is undefined"

<11% mAP with pretrained YOLO after tiny amount of training

1

Only 200 iterations of training on COCO completely destroys the accuracy of the pretrained YOLOv3 net. I am using an unchanged repo with the COCO dataset downloaded using the script...

Fast Dev Run still triggers logging with comet.ml

1

The argument --fast_dev_run is meant to disable logging in pytorch lightning, but it doesn't. Version: pytorch_lightning: 1.3.2 comet_ml: 3.12.2

Segfault on Summing Outer Product

1

Code that triggers segfault: ``` import torch from icecream import ic import triton import triton.language as tl @triton.jit def distortion_bidir_kernel( full_weight_ptr, accum_ptr, M: tl.constexpr, stridem: tl.constexpr, BLOCK_SIZE_M: tl.constexpr): pid_m =...

Error when passing in 0 size empty in PyTorch Kernel

1

When passing in torch.empty((0), device="cuda"), I get the following error: ``` Traceback (most recent call last): File "/home/amai/gaussian-splatting/train.py", line 510, in training( File "/home/amai/gaussian-splatting/train.py", line 160, in training loss.backward() File...

Slang code: ``` [Differentiable] float3 project(in float3 xyz, in float4x4 wct) { float4 xyzw = float4(xyz, 1.f); let p_view = mul(xyzw, wct); float2 pix2d = {p_view.x / p_view.z, p_view.y /...

Crash before Second Pass

2

Not sure what kind of information you need to debug this. ``` Base mesh has 88958 triangles and 44260 vertices. Writing mesh: out/nerf_car/dmtet_mesh/mesh.obj writing 44260 vertices writing 88260 texcoords writing...