issues
search
spcl
/
daceml
A Data-Centric Compiler for Machine Learning
https://daceml.readthedocs.io
BSD 3-Clause "New" or "Revised" License
81
stars
14
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Backward file is compiled even when only forward is called
#81
tbennun
opened
3 years ago
0
[FPGA] Add example script for VGG Convolutions
#80
manuelburger
closed
3 years ago
5
FPGA Convolution (Im2col with aligned vectors and tiling)
#79
manuelburger
closed
3 years ago
2
MaxPool FPGA: support vectorized output and Xilinx
#78
manuelburger
closed
3 years ago
1
Add padding support to FPGA im2col convolution
#77
manuelburger
closed
3 years ago
1
Add ONNXEinsum to symbolic shape inference
#76
tbennun
closed
3 years ago
1
DaCeML 0.2.0a
#75
orausch
closed
2 years ago
1
Replacement Transformations
#74
tbennun
closed
3 years ago
1
fix memlet path in backward pass
#73
tbennun
closed
3 years ago
1
Pure BatchNormalization
#72
orausch
closed
3 years ago
1
CUDNN op implementations, EfficientNet forward pass
#71
orausch
closed
3 years ago
1
Deterministic order of constant code generation
#70
tbennun
closed
3 years ago
1
Add InputToConstant transformation
#69
tbennun
closed
3 years ago
1
TaskletFusion and Mish example
#68
orausch
closed
3 years ago
1
PyTorch C++ Codegen for differentiable functions
#67
orausch
closed
3 years ago
2
Custom backward implementations and setup.py fixes
#66
tbennun
closed
3 years ago
1
Always download bert ONNX file in test
#65
orausch
closed
3 years ago
0
Fix automatic CUDA detection
#64
orausch
closed
3 years ago
1
Add profiling infra from substation
#63
orausch
closed
3 years ago
1
Add norm printouts
#62
orausch
closed
3 years ago
0
Gallery
#61
orausch
closed
3 years ago
1
Bump dace
#60
orausch
closed
3 years ago
1
Some autodiff tests fail with pure implementations
#59
orausch
closed
2 years ago
0
Add Manual C++ implementations for Layer Normalization and Softmax
#58
orausch
closed
2 years ago
1
Tutorial
#57
tbennun
closed
2 years ago
1
Generate forward C++ Operator code for wrapped PyTorch modules
#56
orausch
closed
3 years ago
1
Autodiff Improvements
#55
orausch
closed
3 years ago
1
Make cuda models expect GPU inputs, update parameter_to_transient to handle this
#54
orausch
closed
3 years ago
1
Add pure Flatten support via Reshape
#53
manuelburger
closed
3 years ago
1
Flatten support for FPGA
#52
manuelburger
closed
3 years ago
6
Small CI tweaks
#51
orausch
closed
3 years ago
1
New ORT API Interface, large constant folding speedup
#50
orausch
closed
3 years ago
1
Im2col is broken
#49
orausch
opened
3 years ago
0
Implement IMG ops, Fix autogenerated docs, Update symbolic shape inference, Support BatchNorm
#48
orausch
closed
3 years ago
1
Apply transformations to softmax
#47
and-ivanov
closed
2 years ago
1
Add Pure sum, ONNX opset versioning related fixes
#46
orausch
closed
3 years ago
1
Update transformation interface for pytorch modules, Add GPU backward support, Change gradient summation in autodiff
#45
orausch
closed
3 years ago
1
Add Parameter to transient transformation for Persistent GPU arrays
#44
orausch
closed
3 years ago
1
Add a check for opset version in ONNXModel
#43
orausch
closed
3 years ago
0
Support CUDA streams, add Einsum Node, Support basic auto-optimization, Make einsums use cuBLAS
#42
orausch
closed
3 years ago
1
Update README.md
#41
tbennun
closed
3 years ago
1
Initial FPGA Support
#40
TizianoDeMatteis
closed
3 years ago
7
Rename DaceML -> DaCeML
#39
orausch
closed
3 years ago
1
Fix gpu scalar issues
#38
orausch
closed
3 years ago
2
Binary operators for FPGA
#37
manuelburger
closed
3 years ago
0
Matmul fixes
#36
orausch
closed
3 years ago
0
Add DACE_VERSION variable
#35
orausch
closed
3 years ago
1
Add `sdfg_name` fixture and give each sdfg in test suite a unique name
#34
orausch
closed
3 years ago
2
test
#33
tbennun
closed
3 years ago
1
Update to latest dace version
#32
orausch
closed
3 years ago
1
Previous
Next