|
Review process for data corruption issue, process?
|
|
1
|
78
|
August 24, 2026
|
|
Patterns for making a long-running multimodal data pipeline restartable?
|
|
0
|
46
|
August 22, 2026
|
|
Remote Data Collection - egocentric video recording
|
|
0
|
70
|
August 21, 2026
|
|
Patterns for making a long-running multimodal data loader more reliable?
|
|
0
|
44
|
August 21, 2026
|
|
Installation problems
|
|
1
|
79
|
August 20, 2026
|
|
Should `torch.clip` decompose to `torch.clamp`
|
|
2
|
93
|
August 20, 2026
|
|
How are you structuring region-aware multimodal datasets in PyTorch?
|
|
0
|
45
|
August 20, 2026
|
|
What is the `torch.cond` way to write conditionals in PyTorch?
|
|
3
|
146
|
August 19, 2026
|
|
Faster Attenion with MPS backend
|
|
1
|
135
|
August 17, 2026
|
|
Understanding the out of place index_put operation
|
|
3
|
93
|
August 16, 2026
|
|
Is 32× A100 40 GB with 1 GPU per node a reasonable distributed-training testbed?
|
|
2
|
134
|
August 16, 2026
|
|
Open Closed State-sum Network - My Implementation of 2D TQFT State Sum in PyTorch (Proof of Concept)
|
|
0
|
66
|
August 16, 2026
|
|
Built a 380M param LLM from scratch (no PyTorch/TF/JAX) and found a nasty residual connection bug via a custom debugger (that works across all major ml frameworks)
|
|
0
|
64
|
August 14, 2026
|
|
29.6x Less Checkpoint Work for PyTorch Model Saving — Exact Reload
|
|
0
|
48
|
August 13, 2026
|
|
Clarify on why TRITON_MAX_TILE_NUMEL limits (2**17)
|
|
0
|
39
|
August 12, 2026
|
|
Orchestration layer
|
|
0
|
70
|
August 11, 2026
|
|
+40% CUDA +160% MPS Throughput Improvement on PyTorch Adam Optimizer
|
|
0
|
53
|
August 10, 2026
|
|
Editing data labels
|
|
2
|
95
|
August 9, 2026
|
|
[Project Showcase] Disk Sparse Adam (DSA) — A disk-backed SparseAdam for large embeddings and GNNs
|
|
0
|
58
|
August 8, 2026
|
|
PyTorch support for Intel GPUs on Mac
|
|
11
|
14914
|
August 8, 2026
|
|
Torch.compile dependancy on system architecture / environment
|
|
3
|
123
|
August 7, 2026
|
|
Profiling in PyTorch (Part 1): A Beginner's Guide to torch.profiler
|
|
2
|
178
|
August 7, 2026
|
|
How to implement permanent weight dropout in a layer?
|
|
3
|
708
|
August 6, 2026
|
|
Supporting Continuous Profiling in PyTorch Profiler
|
|
0
|
79
|
August 5, 2026
|
|
PyTorch support for sm_120: NVIDIA GeForce RTX 5060
|
|
16
|
10132
|
August 4, 2026
|
|
Mps seems much slower than cpu on an M1 Mac
|
|
2
|
136
|
August 4, 2026
|
|
Does autograd differentiate with respect to the linspace arguments
|
|
2
|
120
|
August 3, 2026
|
|
PR Review Request #419
|
|
0
|
56
|
July 31, 2026
|
|
I stuck at customizing autograd.Function for multiple output
|
|
1
|
98
|
July 30, 2026
|
|
PR review help: is_differentiable_type addition stalled 3+ weeks (#189586)
|
|
0
|
46
|
July 30, 2026
|