Decision Workspace
aegir vs torsh-autograd vs gtensor
Side-by-side comparison of Rust crates
44
aegir
growingv2.0.0
Strongly-typed, reverse-mode autodiff library in Rust
61
torsh-autograd
experimentalv0.1.1
Automatic differentiation engine for ToRSh with PyTorch-compatible API
45
gtensor
growingv1.0.0
Reverse-mode autodifferentiation of computational graphs with tensors and more for machine learning.
Core Metrics
| aegir | torsh-autograd | gtensor | |
|---|---|---|---|
| Health Score | 44 | 61 | 45 |
| Total Downloads | 4.7K | 1.7K | 5.1K |
| 30d Downloads | 5 | 305 | 6 |
| Dependents | 0 | 37 | 0 |
| Releases | 3 | 6 | 4 |
| Last Updated | 1184d ago | 7d ago | 851d ago |
| Age | 5y 2m | 5m | 3y 3m |
Health Breakdown
aegir
Maintenance
4
Quality
19
Community
3
Popularity
4
Documentation
14
torsh-autograd
Maintenance
20
Quality
12
Community
10
Popularity
4
Documentation
15
gtensor
Maintenance
4
Quality
18
Community
6
Popularity
4
Documentation
13
Technical Details
| aegir | torsh-autograd | gtensor | |
|---|---|---|---|
| Version | 2.0.0 | 0.1.1 | 1.0.0 |
| Stable (≥1.0) | ✓ Yes | ✗ No | ✓ Yes |
| License | MIT | Apache-2.0 | MIT OR Apache-2.0 |
| Dependencies | 10 | 23 | 12 |
| Crate Size | 46KB | 788KB | 28KB |
| Features | 1 | 9 | 0 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2018 | 2021 | 2021 |
| MSRV | 1.66 | 1.77 | — |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •torsh-autograd leads with a health score of 61/100, but none of the options score above 80.
- •gtensor has the most downloads (5.1K), suggesting wider adoption.
- •⚠ aegir, gtensor have not been updated in over a year.
- •torsh-autograd is pre-1.0 — API may change.