-
Notifications
You must be signed in to change notification settings - Fork 3.1k
Pull requests: karpathy/llm.c
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Replaced parameter B with the total parameters calculated outside of the kernel
#521
opened Jun 2, 2024 by
ChrisDryden
Loading…
updated Jun 2, 2024
Removing B in encoder and replacing with calculated N in kernel
#520
opened Jun 2, 2024 by
ChrisDryden
Loading…
updated Jun 2, 2024
Added kernel development file for permute_backwards
#477
opened May 27, 2024 by
ChrisDryden
Loading…
updated Jun 1, 2024
Added additional layernorm forward kernel that does not recalculate mean and rstd
#506
opened May 31, 2024 by
ChrisDryden
Loading…
updated May 31, 2024
Added constexpr for blocksizes to optimize compilation
#505
opened May 31, 2024 by
ChrisDryden
Loading…
updated May 31, 2024
Added new cuda kernel for encoder forwards using three dimensional kernels
#459
opened May 25, 2024 by
ChrisDryden
Loading…
updated May 27, 2024
Modified version of ademeure's fused gelu_forward kernel
#363
opened May 5, 2024 by
ChrisDryden
Loading…
updated May 6, 2024
Experimenting with global instantiation for the layouts
#347
opened May 3, 2024 by
ChrisDryden
•
Draft
updated May 3, 2024
Updated adamw to use packed data types
#303
opened Apr 30, 2024 by
ChrisDryden
Loading…
updated May 2, 2024
gelu_backwards cuda dev file and float4 dtype for parrallel memory read
#241
opened Apr 24, 2024 by
ChrisDryden
•
Draft
updated Apr 27, 2024
ProTip!
no:milestone will show everything without a milestone.