Commits

Branch selector

 main

User selector

byshiue

Datepicker

All time

Commit History

Commits on May 1, 2023

• Fix/gpt early stop (NVIDIA#584)

  Show description for c6e8f60
  byshiue

  authored
  c6e8f60

Commits on Apr 20, 2023

• Update unfused_attention_kernels.cu

  Show description for d7ccf83
  byshiue

  authored
  d7ccf83

Commits on Apr 18, 2023

• Update cublasMMWrapper.cc

  byshiue

  authored
  c6ba315

• Update cublasMMWrapper.cc

  Show description for a6ef7af
  byshiue

  authored
  a6ef7af

Commits on Apr 17, 2023

• Update T5DecodingWeight.cc

  Show description for 0c12805
  byshiue

  authored
  0c12805

Commits on Apr 6, 2023

• fix: fix bug of gpt buffer and gpt gemm overflow

  byshiue

  committed
  4402759

Commits on Mar 14, 2023

• fix: change int of some kernels to int64_t to prevent overflow

  byshiue

  committed
  bb94e2d

Commits on Mar 10, 2023

• Update beam_search_topk_kernels.cu

  Show description for 72d3dce
  byshiue

  authored
  72d3dce

Commits on Mar 6, 2023

• Merge branch 'main' of https://github.com/NVIDIA/FasterTransformer into main

  byshiue

  committed
  303e052

• fix: fix custom_ar_kernel threshold

  byshiue

  committed
  f7a4418

Commits on Feb 24, 2023

• Update SelfAttentionFP8Layer.cc

  Show description for 38847d0
  byshiue

  authored
  38847d0

Commits on Feb 21, 2023

• Feat/perf opt (NVIDIA#462)

  Show description for fba7567
  byshiue

  authored
  fba7567

Commits on Feb 15, 2023

• Feat/cuda 12 support (NVIDIA#450)

  Show description for 35989aa
  byshiue

  authored
  35989aa

Commits on Feb 6, 2023

• Update: set fmha, and context_qk_half_accumulation as optional

  byshiue

  committed
  7921074

• fix: malloc host opt

  byshiue

  committed
  90cc5a7

• fix: fix perf issue of small model due to calling fmtstr

  byshiue

  committed
  652e61e

Commits on Jan 30, 2023

• Update gpt.py

  Show description for 4e77a71
  byshiue

  authored
  4e77a71

Commits on Jan 23, 2023

• Update CMakeLists.txt

  byshiue

  authored
  dbdae81

• feat: Update v5.3 (NVIDIA#418)

  byshiue

  authored
  6ea1c77

Commits on Jan 10, 2023

• fix: fix bug of outputing cum_log_probs in t5 sampling

  byshiue

  committed
  ab00df0

Commits on Jan 6, 2023

• fix: fix bug of ia3

  byshiue

  committed
  6ca2d58

Commits on Jan 1, 2023

• fix: fix bug of t5 beam search (NVIDIA#410)

  Show description for f0b5b86
  byshiue

  authored
  f0b5b86

Commits on Dec 16, 2022

• fix: fmha unrollList

  byshiue

  committed
  83776ae

Commits on Dec 14, 2022

• update: fmha setup and gpt fmha

  byshiue

  committed
  18aaefc

Commits on Dec 8, 2022

• Update fused_multihead_attention_v2.h

  Show description for d2923c0
  byshiue

  authored
  d2923c0

• Update fused_multihead_attention_v2.h

  Show description for 20f6181
  byshiue

  authored
  20f6181

Commits on Dec 6, 2022

• fix: fix fmha kernel assert bug

  byshiue

  committed
  46e1f4a

• fix: fix submodule version issue (NVIDIA#387)

  byshiue

  authored
  aa2ceb5

Commits on Dec 3, 2022

• Fix/cutlass submodule fix (NVIDIA#382)

  Show description for 87b1ae0
  byshiue

  authored
  87b1ae0

Commits on Dec 2, 2022

• feat: update v5.2 (NVIDIA#379)

  byshiue

  authored
  b672f49

Commits on Nov 16, 2022

• Update T5DecodingWeight.cc

  Show description for 9770d30
  byshiue

  authored
  9770d30

Commits on Nov 8, 2022

• Update sampling_topk_kernels.cu

  Show description for d2ad747
  byshiue

  authored
  d2ad747

• Update unfused_attention_kernels.cu

  Show description for c1b5180
  byshiue

  authored
  c1b5180

• Update ParallelGptOp.h

  Show description for 169aad9
  byshiue

  authored
  169aad9

Commits on Oct 27, 2022

• fix: fix bug of BertWeight

  byshiue

  authored
  c254713

Pagination

PreviousNext