issues
search
NVIDIA
/
apex
A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch
BSD 3-Clause "New" or "Revised" License
8.34k
stars
1.39k
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Make distributed fused lamb test names friendly to keyword filtering
#1698
crcrpar
opened
1 year ago
0
Fail to install apex: TypeError: unsupported operand type(s) for +: 'NoneType' and 'str'
#1697
yu1679959321
closed
1 year ago
2
Bf16lamb
#1696
yuanzhedong
closed
1 year ago
0
Fast CUDA NHWC Group Norm
#1695
alpha0422
closed
1 year ago
5
Using nvidia_dlprof_pytorch_nvtx.init() with apex errors out as "ModuleNotFoundError: No module named 'xentropy_cuda' "
#1694
nipunagarwala
opened
1 year ago
1
Use `torch.testing.assert_close` in test_index_mul_2d.py
#1693
crcrpar
closed
1 year ago
0
Add custom build backend to support build args
#1692
janEbert
opened
1 year ago
3
[Transformer][UCC] Fix async p2p ops
#1691
Aidyn-A
closed
1 year ago
0
Fix installation command
#1690
janEbert
closed
1 year ago
2
Use a modern tensor constructor in cudnn_gbn
#1689
crcrpar
opened
1 year ago
0
A FasterRMSNorm implementation (based on FasterLayerNorm)
#1688
Njuapp
opened
1 year ago
0
data_file = open("myways.json","r") data = json.loads(data_file.read()) print(data['intents']) KeyError Traceback (most recent call last) Cell In[72], line 3 1 data_file = open("myways.json","r") 2 data = json.loads(data_file.read()) ----> 3 print(data['intents']) KeyError: 'intents' This key error is coming though I have created a json file with intents as an object
#1687
PushkarSri
opened
1 year ago
1
sequence parallel with rmsnorm/layernorm
#1686
wlike
opened
1 year ago
0
Tkurth/sgbn fixes
#1685
azrael417
closed
1 year ago
3
Tkurth/mplamb fixed
#1684
azrael417
closed
1 year ago
0
Backprop through TransducerLoss creates NaN gradients
#1683
TheoEhrenborg
opened
1 year ago
0
ERROR: Could not build wheels for apex, which is required to install pyproject.toml-based projects
#1682
PeytonTse
opened
1 year ago
4
ERROR: Directory './' is not installable. Neither 'setup.py' nor 'pyproject.toml' found.
#1681
abbas695
closed
1 year ago
1
Updating missing build dependency in pyproject.toml
#1680
loadams
opened
1 year ago
5
`pyproject.toml` missing `packaging` dependency
#1679
calebho
opened
1 year ago
46
Tkurth/new gbn
#1678
azrael417
closed
1 year ago
0
scaled_upper_triang_masked_softmax_cuda: undefined symbol
#1677
TheGravityZero
opened
1 year ago
1
Issue Installing Apex in WSL Environment
#1676
l8g
opened
1 year ago
5
[Transformer] Do not use batch_isend_irecv for UCC
#1675
Aidyn-A
closed
1 year ago
0
I might have some pip issue while running autogpt in vs code
#1674
KTH1881
closed
1 year ago
0
[Test][Transformer] Pre-parse container version
#1673
Aidyn-A
closed
1 year ago
1
current code cannot build due to tensor.type()
#1672
ycsos
closed
1 year ago
0
AttributeError: module 'torch.distributed' has no attribute '_all_gather_base'
#1671
HloveMM
opened
1 year ago
1
bf16 support for FusedDense preventing apex build on CUDA 10.2
#1670
minostauros
opened
1 year ago
6
Add `pyproject.toml`
#1669
crcrpar
closed
1 year ago
0
Please publish versions tags to Github
#1668
h-vetinari
opened
1 year ago
1
Update setup.py
#1667
RedaGrace
closed
1 year ago
1
The phase of Channel Permutation operation from ASP
#1666
erwangccc
closed
1 year ago
5
Install apex without ninja
#1665
ustcwhy
opened
1 year ago
1
Support for FP32 input dtype
#1664
jacob-crux
closed
1 year ago
1
Update distributed optimizer with new coalescing manager API in PyTorch
#1663
timmoon10
closed
1 year ago
2
install error
#1662
yz103
opened
1 year ago
4
allow for custom directory of xml reports
#1661
crcrpar
closed
1 year ago
0
get compatible with the latest `torch.distributed.distributed_c10d._coalescing_manager`
#1660
crcrpar
closed
1 year ago
1
sparsity test part1 failed
#1659
jackzhou121
opened
1 year ago
0
[DDP][Master Weight] For DDP + Master weight, is it necessary to set torch seed for training?
#1658
zejun-chen
opened
1 year ago
3
it didn't use correct nvcc path when installing apex
#1657
yz103
closed
1 year ago
6
Help me, I'm dying soon，error: command '/opt/rh/devtoolset-7/root/usr/bin/gcc' failed with exit code 1 error: subprocess-exited-with-error
#1656
listwebit
opened
1 year ago
0
PyTorch FSDP compatibility with Apex
#1655
conceptofmind
closed
1 year ago
3
[BUG] CUDA error: an illegal memory access was encountered with Adam optimizer on H100
#1654
szhengac
opened
1 year ago
6
ERROR: Could not build wheels for apex, which is required to install pyproject.toml-based projects
#1653
Rainbowman0
closed
1 year ago
29
Questions about numeric precision of FusedRMSNorm
#1652
yingtongxiong
opened
1 year ago
4
[Transformer][test/L0] Increase max world_size for BERT and GPT
#1651
Aidyn-A
closed
1 year ago
0
[Transformer] Update p2p communication routine
#1650
Aidyn-A
closed
1 year ago
0
report VS type alignment error when installed with cuda118
#1649
ljf841
opened
1 year ago
2
Previous
Next