issues
search
bytedance
/
lightseq
LightSeq: A High Performance Library for Sequence Processing and Generation
Other
3.21k
stars
329
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
beam search output
#429
kevinmgyu
closed
1 year ago
0
README里面的【Speedup of Transformer Inference】是哪个模型的，这边用里面的VIT用例测试的性能只有2.64倍？
#428
Wayne-Bfx
opened
2 years ago
4
Multi GPU triton failures
#427
Csinclair0
opened
2 years ago
4
fix bug for hard_gate_moe proto parse
#426
Anychnn
closed
2 years ago
0
Hard gate moe
#425
Anychnn
closed
2 years ago
0
Hard gate moe
#424
Anychnn
closed
2 years ago
0
Hard gate moe
#423
Anychnn
closed
2 years ago
0
What should I do if I want to use lightseq to accelerate MarianMTModel
#422
Youggls
opened
2 years ago
7
Hard gate moe
#421
Anychnn
closed
2 years ago
0
Hard gate moe
#420
Anychnn
closed
2 years ago
0
transformer new arch
#419
hexisyztem
closed
2 years ago
1
How to use test/ls_quant_bert.py?
#418
gjm441
opened
2 years ago
1
Update README and doc
#417
godweiyang
closed
2 years ago
0
Is it possible to accelerate mbart by lightseq?
#416
GongCQ
opened
2 years ago
2
Fix gpu name
#415
godweiyang
closed
2 years ago
0
Does LightSeq support ONNX export and Triton Inference Server?
#414
stevezheng23
opened
2 years ago
1
does lightsq supportting distributed inference?
#413
dearowen
opened
2 years ago
1
[WIP] Decoder layer
#412
hexisyztem
closed
2 years ago
0
AttributeError: module 'lightseq.inference' has no attribute 'MT5'
#411
huiyangzhou
opened
2 years ago
2
No acceleration compared with timm vit block
#410
woolpeeker
closed
2 years ago
11
compatible gcq params
#409
HandH1998
closed
2 years ago
0
add auto igemm for gpt, vit
#408
godweiyang
closed
2 years ago
0
为什么使用beam_search推理多条句子一样呢
#407
duiyady
opened
2 years ago
1
RuntimeError: Error building extension 'lightseq_layers_new'
#406
xiaotingyun
opened
2 years ago
10
'Namespace' object has no attribute 'enable_GCQ' with lightseq 3.0.0
#405
stephanpeitz
closed
2 years ago
5
update to 3.0.0
#404
godweiyang
closed
2 years ago
0
Fix train bug
#403
godweiyang
closed
2 years ago
0
What is the requirement of lightseq3.0
#402
Fizzmy
opened
2 years ago
9
fix auto gemm bug
#401
godweiyang
closed
2 years ago
0
I Failed to execute the shell hf_bart_export.py
#400
YINGPENGZH
closed
2 years ago
0
fix auto gemm & new arch pad mask bug
#399
godweiyang
closed
2 years ago
0
This may not be a problem. Sorry.
#398
jundaf2
closed
2 years ago
0
AttributeError: module 'lightseq' has no attribute 'training'
#397
wangyong1122
opened
2 years ago
4
Does any plans to support the Pegasus for summary
#396
lileilai
opened
2 years ago
1
fix cudaStreamSynchronize error
#395
hexisyztem
closed
2 years ago
0
New version requirement
#394
GrittyChen
opened
2 years ago
1
support auto gemm for int8 inference and training
#393
godweiyang
closed
2 years ago
0
update version to 3.0.0.dev20220926, update dockerfile to manylinux2014
#392
godweiyang
closed
2 years ago
0
Bert crf
#391
godweiyang
closed
2 years ago
0
add bert-crf files
#390
godweiyang
closed
2 years ago
0
编译过程碰到的问题
#389
lileilai
opened
2 years ago
14
add bert-crf files
#388
hexisyztem
closed
2 years ago
0
Revert "support bert+linear+crf"
#387
hexisyztem
closed
2 years ago
0
Revert "support bert+linear+crf"
#386
hexisyztem
closed
2 years ago
0
CUBLAS_STATUS_NOT_SUPPORTED
#385
lovelyllwang
opened
2 years ago
1
Is there any plan to support MarianMTModel？
#384
dengcunqin
opened
2 years ago
5
support bert+linear+crf
#383
godweiyang
closed
2 years ago
1
split total buffer into serveral mini-buffer
#382
hexisyztem
closed
2 years ago
0
to run quant train, which cuda version should use?
#381
tricky61
closed
2 years ago
5
Fix error
#380
hexisyztem
closed
2 years ago
0
Previous
Next