Skip to content

Commit cf60757

Browse files
kaiyuxkaiyux
committed
Table of Contents
Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> add alltoall optimization part Signed-off-by: Dongxu Yang <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Minor updates Signed-off-by: Kaiyu Xie <[email protected]> doc: add FP8 context FMHA support part. Signed-off-by: Yuxian Qiu <[email protected]> Add lowprecision all2all and fuse shared expert into local reduction Signed-off-by: Zongfei Jing <[email protected]> Add MTP LM head tensor parallelism Signed-off-by: Kaiyu Xie <[email protected]> Polish Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Add images Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> AI polishment Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]> Update Signed-off-by: Kaiyu Xie <[email protected]>
1 parent 1e0fbb7 commit cf60757

9 files changed

+239
-0
lines changed

‎docs/source/blogs/media/tech_blog14_MTP_parallel_1.png‎

236 KB
Loading

‎docs/source/blogs/media/tech_blog14_MTP_parallel_2.png‎

354 KB
Loading

‎docs/source/blogs/media/tech_blog14_alltoall_dataflow.png‎

77.4 KB
Loading

‎docs/source/blogs/media/tech_blog14_overview_after_opt.png‎

196 KB
Loading

‎docs/source/blogs/media/tech_blog14_overview_before_opt.png‎

190 KB
Loading

‎docs/source/blogs/media/tech_blog14_pdloff.png‎

150 KB
Loading

‎docs/source/blogs/media/tech_blog14_pdlon.png‎

168 KB
Loading

‎docs/source/blogs/media/tech_blog14_perf.png‎

400 KB
Loading

‎docs/source/blogs/tech_blog/blog14_Scaling_Expert_Parallelism_in_TensorRT-LLM_part3.md‎

Lines changed: 239 additions & 0 deletions
Large diffs are not rendered by default.

0 commit comments

Comments
 (0)