|
How to Persist JIT Fused CUDA Kernels State for Efficient Inference on Multi-GPU Setup?
|
|
5
|
444
|
April 4, 2024
|
|
Torch.jit.trace() memory leak workarounds
|
|
4
|
897
|
April 3, 2024
|
|
Build torch 2.2.0 with vulkan failed by ndk-26b
|
|
0
|
305
|
March 31, 2024
|
|
Cannot trace my own model using torch.jit.trace
|
|
1
|
411
|
March 30, 2024
|
|
Why torch1.13.x JIT compiler doesn't support sm90
|
|
1
|
249
|
March 27, 2024
|
|
How to add Attribute and use after building the model?
|
|
0
|
213
|
March 24, 2024
|
|
How to load constant tensors in the model onto GPU?
|
|
4
|
1477
|
March 12, 2024
|
|
Libtorch 2.1.1 build error in linux
|
|
1
|
389
|
March 9, 2024
|
|
Problem saving nn.Module as a TorchScript module (DLRM model)
|
|
1
|
1413
|
March 7, 2024
|
|
Export Detectron2 Model
|
|
0
|
380
|
March 6, 2024
|
|
When will `torch.jit.fork` create new cuda stream?
|
|
0
|
199
|
March 5, 2024
|
|
Getting _ivalue_ INTERNAL ASSERT FAILED at "../torch/csrc/jit/api/object.h
|
|
0
|
372
|
February 23, 2024
|
|
Error in loading the model
|
|
8
|
6286
|
February 23, 2024
|
|
How to avoid `intimplicit` when `torch.jit.trace` a Model?
|
|
1
|
308
|
February 22, 2024
|
|
How to trace external modules for model serialization
|
|
2
|
655
|
February 19, 2024
|
|
Can I perform jit.trace only on specific layers of a model?
|
|
1
|
321
|
February 14, 2024
|
|
LSTM h_0 & c_0 storage when using One by one forward step
|
|
5
|
311
|
February 13, 2024
|
|
Torch.jit.save not working on some models from Torchhub
|
|
0
|
264
|
February 7, 2024
|
|
How can I prevent TorchScript from including the original Python source code?
|
|
0
|
324
|
February 6, 2024
|
|
Torch.jit.trace consumes more than 40x gpu memory
|
|
1
|
503
|
January 31, 2024
|
|
Replace a specific attribute in RecursiveScriptModule
|
|
0
|
269
|
January 26, 2024
|
|
Comparison between saving the whole model, saving only state_dict and torchscript
|
|
3
|
3927
|
January 25, 2024
|
|
RuntimeError: torch.linalg.lu_factor: (Batch element 1): U[4,4] is zero and using it on lu_solve would result in a division by zero. If you still want to perform the factorization, consider calling linalg.lu(A, pivot) or linalg.lu_factor_ex(A, pivot)
|
|
0
|
271
|
January 23, 2024
|
|
Having troubles to vmap a custom function with permute
|
|
0
|
360
|
January 20, 2024
|
|
Nondeterminstic behaviour of sigmoid and tanh in JIT script
|
|
1
|
376
|
January 19, 2024
|
|
How to calculate FLOPS of a JIT model?
|
|
1
|
735
|
January 18, 2024
|
|
Converting GPT-2 to TorchScript
|
|
5
|
1209
|
December 30, 2023
|
|
About initialization of RecursiveScriptModule instance?
|
|
0
|
358
|
December 29, 2023
|
|
Access to model parameters
|
|
5
|
856
|
December 26, 2023
|
|
Jit tracing failing (torch mangling diverging operator)
|
|
0
|
477
|
December 24, 2023
|