|
About the FX (Functional Transformations) category
|
|
0
|
734
|
February 25, 2021
|
|
What's the recommended way to achieve asynchronicity for dynamo backend?
|
|
2
|
106
|
May 8, 2023
|
|
Can we benefit from fx.graph in different language (different framework)?
|
|
4
|
209
|
May 3, 2023
|
|
Extract graph for editing
|
|
2
|
114
|
April 19, 2023
|
|
Torch.fx.symbolic_trace removes some of the keys from module state_dict
|
|
3
|
93
|
April 19, 2023
|
|
Conversion of torch GraphModule back to nn.Module
|
|
2
|
132
|
April 5, 2023
|
|
Using TorchTRT-FX backend on C++
|
|
5
|
268
|
March 23, 2023
|
|
FX.graph tracing computation graph in DGL
|
|
0
|
106
|
February 8, 2023
|
|
FX.graph generate CUDA code for customized layers
|
|
1
|
121
|
February 7, 2023
|
|
Node fusion via leaf nodes
|
|
1
|
139
|
February 7, 2023
|
|
How to use shape information during FX pattern rewriting?
|
|
0
|
126
|
January 30, 2023
|
|
How to determine if two GraphModule nodes can be "fused"
|
|
0
|
113
|
January 27, 2023
|
|
Script function as target
|
|
0
|
124
|
January 17, 2023
|
|
How to prevent fx to trace into my custom defined module
|
|
2
|
449
|
January 17, 2023
|
|
Torch.fx.trace with real tensors, similar to torch.jit.trace
|
|
0
|
117
|
January 17, 2023
|
|
How to specify a built-in target when creating new node
|
|
1
|
131
|
January 15, 2023
|
|
I can't find the source code of _scaled_dot_product_attention
|
|
2
|
262
|
January 2, 2023
|
|
Using FX for creating pt file
|
|
0
|
127
|
December 29, 2022
|
|
Extracting hierarchy/structure information for operators with opcode of "call_function" or "call_method"
|
|
0
|
159
|
December 22, 2022
|
|
Why does training a model with frozen weights take longer than training the model with unfrozen weights?
|
|
5
|
268
|
December 21, 2022
|
|
Questions about dispatch to Aten library and control flow
|
|
0
|
186
|
November 14, 2022
|
|
Reproducibility issue w.r.t. Functional Transformations with randomness
|
|
0
|
172
|
November 14, 2022
|
|
Torch.fx usage question
|
|
1
|
290
|
October 25, 2022
|
|
‘len’ is not supported in symbolic tracing by default
|
|
5
|
1015
|
October 19, 2022
|
|
Get arguments for torch.nn.functional methods
|
|
0
|
200
|
October 12, 2022
|
|
Make_fx and vjp
|
|
0
|
346
|
October 10, 2022
|
|
Class name difference between original and quantized version of Sigmoid
|
|
1
|
205
|
September 30, 2022
|
|
How can Torch.FX work with autograd.Function?
|
|
1
|
559
|
September 3, 2022
|
|
TORCH.NN.FUNCTIONAL.GRID_SAMPLE invariant to sign of the grid
|
|
2
|
395
|
August 29, 2022
|
|
Constants consume too much gpu memory in torch.fx
|
|
2
|
300
|
August 16, 2022
|