facebookresearch / fairring
Fairring (FAIR + Herring) is a plug-in for PyTorch that provides a process group for distributed training that outperforms NCCL at large scales
Summary
   |
|||
| Main Code: 2,500 LOC (15 files) = CC (61%) + H (22%) + PY (14%) + YAML (1%) Secondary code: Test: 0 LOC (0); Generated: 0 LOC (0); Build & Deploy: 30 LOC (2); Other: 87 LOC (1); |
|||
| Duplication: 13% | |||
| File Size: 0% long (>1000 LOC), 34% short (<= 200 LOC) | |||
| Unit Size: 13% long (>100 LOC), 32% short (<= 10 LOC) | |||
| Conditional Complexity: 0% complex (McCabe index > 50), 63% simple (McCabe index <= 5) | |||
|
|
Logical Component Decomposition: primary (4 components) | ||
|
|
1 month old
|
||
|
|
0% of code updated more than 50 times Also see temporal dependencies for files frequently changed in same commits. |
||
| Goals: Keep the system simple and easy to change (4) | |||
|
Features of interest:
TODOs
1 file |
|||
Latest commit date: 2022-01-24
|
21
commits
(30 days)
1
contributors
(30 days) |
|