-
NVIDIA
- Bay Area, CA, USA
- https://ericrxw.github.io/xiaoweiren/
Block or Report
Block or report xrennvidia
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePopular repositories
-
-
apex
apex PublicForked from NVIDIA/apex
A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch
Python
-
NeMo-Megatron-Launcher
NeMo-Megatron-Launcher PublicForked from NVIDIA/NeMo-Megatron-Launcher
NeMo Megatron launcher and tools
Python
-
flash-attention
flash-attention PublicForked from Dao-AILab/flash-attention
Fast and memory-efficient exact attention
Python
-
-
TransformerEngine
TransformerEngine PublicForked from NVIDIA/TransformerEngine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper GPUs, to provide better performance with lower memory utilization in bot…
Python
97 contributions in the last year
| Day of Week | December Dec | January Jan | February Feb | March Mar | April Apr | May May | June Jun | July Jul | August Aug | September Sep | October Oct | November Nov | |||||||||||||||||||||||||||||||||||||||||
| Sunday Sun | |||||||||||||||||||||||||||||||||||||||||||||||||||||
| Monday Mon | |||||||||||||||||||||||||||||||||||||||||||||||||||||
| Tuesday Tue | |||||||||||||||||||||||||||||||||||||||||||||||||||||
| Wednesday Wed | |||||||||||||||||||||||||||||||||||||||||||||||||||||
| Thursday Thu | |||||||||||||||||||||||||||||||||||||||||||||||||||||
| Friday Fri | |||||||||||||||||||||||||||||||||||||||||||||||||||||
| Saturday Sat | |||||||||||||||||||||||||||||||||||||||||||||||||||||
Contribution activity
November 2023
Created 3 commits in 3 repositories
Created 2 repositories
-
xrennvidia/praxis
Python
This contribution was made on Nov 20
-
xrennvidia/paxml
Python
This contribution was made on Nov 10
Created a pull request in NVIDIA/TransformerEngine that received 3 comments
fix global cu_seqlens setting
While virutal_pipeline_parallel_size > 1, model in each VP stage usually only has 1 transformer layer (i..e, layer number is always 1). Hence, we c…
Opened 1 other pull request in 1 repository
NVIDIA/NeMo
1
merged
-
fix tp_overlap config var name
This contribution was made on Nov 22

