issues
search
microsoft
/
nnfusion
A flexible and efficient deep neural network (DNN) compiler that generates high-performance executable from a DNN model description.
MIT License
948
stars
158
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
[Fix/Feat] Correct the fp16 inference of resnet50.onnx
#433
LeiWang1999
closed
2 years ago
0
Add ArgMax Operator and HLSL kernel.
#432
wenxcs
closed
2 years ago
0
[BUG] Incorrect output for resnet50-float16
#431
LeiWang1999
opened
2 years ago
2
Support cumsum
#430
mzmssg
closed
2 years ago
0
General TopK Kernel for HLSL devices
#429
wenxcs
closed
2 years ago
1
General TopK Kernel for HLSL devices
#428
wenxcs
closed
2 years ago
0
fix some script bugs and add estimated time information
#427
jlxue
closed
2 years ago
0
add history log for ansor time
#426
jlxue
closed
2 years ago
0
[BUG] Unsupported Operator Prod, Einsum, and BroadcastTo in Tensorflow Frozen PB model format
#425
donglinz
opened
2 years ago
0
Avoid hanging on xbox
#424
mzmssg
closed
2 years ago
0
Fix e2e ansor kernel injection
#423
LittleQili
closed
2 years ago
0
add missing files
#422
Cjkkkk
closed
2 years ago
1
fix roller script_v1 elementwise.sh
#421
LittleQili
closed
2 years ago
0
fix test_op_mp.py
#420
LittleQili
closed
2 years ago
1
code change to adapt roller kernels and e2e models
#419
jlxue
closed
2 years ago
0
fix default_kernel1 entry
#418
Cjkkkk
opened
2 years ago
0
[BUG] Failed to generate code for bert fp16 model
#417
Binyang2014
opened
2 years ago
3
[Typo] An automatic github action solution for typo inspection.
#416
LeiWang1999
closed
2 years ago
9
[Is there any plan to support windows system?]
#415
CnybTseng
opened
2 years ago
0
fix: typo of scattermin op register.
#414
LeiWang1999
closed
2 years ago
7
Align with onnx pad's sematics
#413
zyeric
closed
2 years ago
2
[BUG] Unsupported type for GPT model
#412
Binyang2014
opened
2 years ago
0
Load fp16 weight from onnx
#411
mzmssg
closed
2 years ago
0
Fix workspace return type
#410
jlxue
opened
2 years ago
0
Add README.md for script kernel
#409
wenxcs
closed
2 years ago
0
TopK
#408
wenxcs
closed
2 years ago
0
Topk
#407
wenxcs
closed
2 years ago
0
TopK support
#406
wenxcs
closed
2 years ago
0
hi , when i use your plugin it appears:cudaScaleRunner.cpp (119) - Cuda Error in execute: 700,
#405
scuizhibin
closed
2 years ago
3
[BUG] Exception: Unknown symbilic dimension: seq_length
#404
Binyang2014
closed
2 years ago
1
Customized script operator support
#403
wenxcs
closed
2 years ago
0
Fix memcpy in fused IR; fix kernel elimination rules; fix kernel shape check; optimize GEMM op convert
#402
jlxue
closed
2 years ago
0
Support onnxruntime >= 1.8
#401
mzmssg
closed
2 years ago
1
Export parameters to inputs in ONNX in Python trainer to avoid folding parameters
#400
xysmlx
closed
2 years ago
0
[BUG] Limitation of the number of arguments in ctypes in NNFusion Python interface
#399
xysmlx
opened
2 years ago
0
Support JIT for class and class method + nnf config + 3-lvl-signature
#398
siahuat0727
closed
2 years ago
0
Resolve UB in gcc9
#397
mzmssg
closed
2 years ago
1
Fix FP16 codegen for ONNXmodels
#396
jlxue
closed
2 years ago
0
Optimize EmbeddingGrad for the backward of gather op
#395
xysmlx
closed
2 years ago
2
Optimize EmbeddingGrad for the backward of gather op
#394
xysmlx
closed
2 years ago
2
Add docstring
#393
siahuat0727
closed
2 years ago
1
Python API integration
#392
siahuat0727
closed
2 years ago
4
Update antares template
#391
mzmssg
closed
2 years ago
0
Fix mapping error bug in dtypes.py
#390
siahuat0727
closed
2 years ago
2
Fix exit status is always zero
#389
siahuat0727
closed
2 years ago
0
[BUG] Error if running nnfusion with onnxruntime >= 1.9
#388
mzmssg
closed
2 years ago
0
[BUG] Segment fault if compiling NNFusion with GCC-9
#387
mzmssg
closed
2 years ago
0
[BUG] import Pad op
#386
jlxue
opened
2 years ago
3
[BUG] shape size mismatch for Constant op with FP16
#385
jlxue
opened
2 years ago
0
Support tuning kernels synchronously
#384
jlxue
closed
2 years ago
0
Previous
Next