Skip to content

Issues: NVIDIA/TransformerEngine

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

156 Open 212 Closed
156 Open 212 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

_NoopCatFunc in transformer layer
#1384 opened Dec 22, 2024 by robot-transformer
thd qkv-format in transformer layer
#1383 opened Dec 22, 2024 by robot-transformer
AttributeError: module 'transformer_engine' has no attribute 'pytorch'
#1379 opened Dec 17, 2024 by carrot0117
1
ViT Support
#1377 opened Dec 16, 2024 by cnut1648
TypeError: initialize_ub() got an unexpected keyword argument 'tp_size'
#1376 opened Dec 13, 2024 by wccccp
@ksivaman
3
Should cublasLtHandle_t be Destroyed?
#1372 opened Dec 13, 2024 by shenzhenghai
How to use thd format qkv with cp + packed_seq_params
#1368 opened Dec 12, 2024 by Wraythh
4
support float8 in flash-attn v3
#1359 opened Dec 5, 2024 by Monekyzoon
overlapping issue about backward of LayerNormLinear
#1353 opened Dec 3, 2024 by cos120
3
Support more than 1 shape/attention_params for DotProductAttention decision cache
#1349 opened Nov 29, 2024 by parthmannan
1
@cyanguwa
[Bug] attention_backend update throttle
#1348 opened Nov 27, 2024 by Jianbing-D
How to setup TP Overlap configs
#1344 opened Nov 21, 2024 by TJ-Solergibert
@denera
1
the max error of moe_permute/unpermute.grad could reach 3.6e+00
#1336 opened Nov 15, 2024 by NiuMa-1234
1
[TP comm overlap unit test]CUDA Error: misaligned address error when testing with recent cublas (or pytorch container)
#1332 opened Nov 14, 2024 by erhoo82
@denera
4
TransformerEngine doesn't work with uv
#1323 opened Nov 8, 2024 by jennifgcrl
1
2
How can I use fp8_gemm to realize the function of "torch.mm()"?
#1318 opened Nov 6, 2024 by duomicoding
Linear does not support TP comm overlap for Column Parallel mode
#1312 opened Nov 5, 2024 by parthmannan
1
[PyTorch] transformer engine modifies root logger
#1294 opened Oct 28, 2024 by Marks101
1
@cyanguwa
2
Installing TE is so hard. bug Something isn't working build Build system
#1284 opened Oct 23, 2024 by ZihaoZheng98
5
BUG: fused_amax_and_scale_update_after_reduction(): incompatible function arguments. The following argument types are supported:
#1275 opened Oct 20, 2024 by cassanof
@ksivaman
2
Failed to build transformer-engine build Build system
#1270 opened Oct 18, 2024 by jaefan11
3
Isn't the memory consumption should be dropped when using fp8? question Further information is requested
#1261 opened Oct 16, 2024 by JayC1208
1
[PyTorch] fused attention and cu_seqlens
#1259 opened Oct 16, 2024 by Marks101
@cyanguwa
2
TransformerEngine install fails with no clear cause bug Something isn't working build Build system
#1249 opened Oct 14, 2024 by sytelus
1
FSDP with FP8 is not working
#1247 opened Oct 14, 2024 by eljandoubi
1
1
ProTip! Adding no:label will show everything without a label.