issues
search
ROCm
/
apex
A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch
BSD 3-Clause "New" or "Revised" License
17
stars
14
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Add a wrapper to skip flaky tests and un-skip some MLP unit tests
#86
hubertlu-tw
closed
2 years ago
1
Skip the failing unit tests from the FusedRMSNorm PR
#85
hubertlu-tw
closed
2 years ago
2
Skip some failed unit tests from the FusedRMSNorm PR
#84
hubertlu-tw
closed
2 years ago
0
test_multi_device (test_lamb.TestFusedMixedPrecisionLamb)
#83
jithunnair-amd
opened
2 years ago
0
Unit tests failing with "AssertionError: If capturable=False, state_steps should not be CUDA tensors."
#82
jithunnair-amd
closed
1 year ago
1
test_autocast (test_fused_layer_norm.TestAutocastFusedRMSNorm)
#81
jithunnair-amd
opened
2 years ago
0
IFU-master-2022-07-29
#80
hubertlu-tw
closed
2 years ago
3
Make rocblas_gemm_flags_fp16_alt_impl backward-compat for new naming
#79
hubertlu-tw
closed
2 years ago
0
Enable FusedRMSNorm
#78
hubertlu-tw
closed
2 years ago
6
Apex transformer
#77
hubertlu-tw
closed
2 years ago
0
Cherry-picked the commit from upstream for faster --fast_multihead_attn build
#76
hubertlu-tw
closed
2 years ago
1
IFU-master-2022-02-09
#75
hubertlu-tw
closed
2 years ago
1
Make rocblas_gemm_flags_fp16_alt_impl in MHA and MLP backward compatible with old PyTorch versions
#74
hubertlu-tw
closed
2 years ago
2
python setup.py clean does not fully clean up build folder
#73
another-pjohnson
opened
2 years ago
0
Added support for memory format API(torch.channels_last) in GBN
#72
Mahathi-Vatsal
closed
2 years ago
0
Add rocblas_alt_impl flag for backprop in MLP
#71
hubertlu-tw
closed
2 years ago
0
Add rocblas_alt_impl falg for bwd rocblas calls in MHA
#70
athitten
closed
2 years ago
3
Cherry-pick b2fdf9c from upstream Apex and resolve conflicts
#68
jithunnair-amd
closed
2 years ago
1
Fix bn_addrelu's bitmask type error
#67
sryap
closed
2 years ago
3
Optimize layer normalization for AMD GPUs
#66
hubertlu-tw
closed
2 years ago
2
Failing DDP race condition test on MI50
#65
hubertlu-tw
closed
2 years ago
1
IFU-master-2021-12-08
#64
hubertlu-tw
closed
2 years ago
1
NaNs in test_half (test_fused_optimizer.TestFusedAdam)
#63
jithunnair-amd
opened
2 years ago
2
Unit test failures: "if cached_x.grad_fn.next_functions[1][0].variable is not x: IndexError: tuple index out of range"
#62
jithunnair-amd
closed
1 year ago
3
Skip failing unit tests
#61
hubertlu-tw
closed
2 years ago
0
Document on extension build options
#60
sunway513
closed
2 years ago
0
Enable all supported CUDA extensions using --cuda_ext flag
#59
jithunnair-amd
closed
2 years ago
0
Add more unit tests for both distributed and extensions
#58
hubertlu-tw
closed
2 years ago
2
Enable Distributed FusedLAMB
#57
athitten
closed
2 years ago
4
Enable multihead atten
#56
hubertlu-tw
closed
2 years ago
0
IFU-2021-10-15 (+ remove redundant __HIP_NO_HALF_OPERATORS__ defines + C10_CUDA_CHECK)
#55
hubertlu-tw
closed
2 years ago
1
in multi tensor apply, skip empty tensors
#54
jeffdaily
closed
2 years ago
0
work around hipify not finding headers
#53
jeffdaily
closed
3 years ago
0
add distributed fused lamb
#52
jeffdaily
closed
3 years ago
0
Enable group batch norm (--bnp) on ROCm (only bn_group = 1)
#51
sryap
closed
2 years ago
7
Make torch version check numeric
#50
jithunnair-amd
closed
3 years ago
0
delete the redundant instructions & calculation of Welford online algorithm in fused_layer_norm
#49
alexshuang
closed
3 years ago
0
IFU-2020-03-04
#48
jeffdaily
closed
3 years ago
0
Revert "pass all TensorListMetadata as pointer to pinned host memory (#13)
#47
jeffdaily
closed
3 years ago
1
fix bugs in syncbn
#46
jeffdaily
closed
3 years ago
2
fix cross-compiled ROCm builds when no GPUs detected
#45
jeffdaily
closed
3 years ago
0
use __launch_bounds__ for multi_tensor_apply
#44
jeffdaily
closed
3 years ago
0
IFU-2021-01-18
#43
jeffdaily
closed
3 years ago
0
Fix reduce_block_into_lanes for multi_tensor_l2norm for ROCm
#42
sryap
closed
3 years ago
0
Skip the unit tests
#41
lcskrishna
closed
3 years ago
6
[DONT-MERGE] Dummy PR for testing Jenkins
#40
lcskrishna
closed
3 years ago
3
Revert "Hipify revamp changes for apex extensions on ROCm."
#39
jeffdaily
closed
3 years ago
0
Hipify revamp changes for apex extensions on ROCm.
#38
lcskrishna
closed
3 years ago
8
Use correct warp size of 64 for ROCm in layernorm kernel
#37
ashishfarmer
closed
2 years ago
1
Fix LayerNorm op on ROCm
#36
ashishfarmer
closed
3 years ago
1
Previous
Next