Decision Workspace
mil_convert vs bananamendr vs ruvector-attention-wasm
Side-by-side comparison of Rust crates
43
mil_convert
experimentalv0.1.0
safetensors → MIL: config-driven transformer graph generation (Qwen-class) into .mlpackage. No Python, no coremltools.
54
bananamendr
experimentalv1.0.4
CPU inference for BananaMind-2 (Nano/Mini/Pro) decoder checkpoints, with a CLI
57
ruvector-attention-wasm
growingv2.3.0
High-performance WebAssembly attention mechanisms: Multi-Head, Flash, Hyperbolic, MoE, CGT Sheaf Attention with GPU acceleration for transformers and LLMs
Core Metrics
| mil_convert | bananamendr | ruvector-attention-wasm | |
|---|---|---|---|
| Health Score | 43 | 54 | 57 |
| Total Downloads | 13 | 90 | 124 |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 2 | 0 | 0 |
| Releases | 1 | 5 | 6 |
| Last Updated | 2d ago | 53d ago | 13d ago |
| Age | 2d | 1m | 10m |
Health Breakdown
mil_convert
Maintenance
13
Quality
13
Community
4
Popularity
1
Documentation
12
bananamendr
Maintenance
16
Quality
18
Community
6
Popularity
2
Documentation
12
ruvector-attention-wasm
Maintenance
18
Quality
17
Community
6
Popularity
3
Documentation
13
Technical Details
| mil_convert | bananamendr | ruvector-attention-wasm | |
|---|---|---|---|
| Version | 0.1.0 | 1.0.4 | 2.3.0 |
| Stable (≥1.0) | ✗ No | ✓ Yes | ✓ Yes |
| License | MIT OR Apache-2.0 | Apache-2.0 | MIT |
| Dependencies | 4 | 10 | 9 |
| Crate Size | 16KB | 31KB | 86KB |
| Features | 0 | 6 | 1 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2021 | 2024 | 2021 |
| MSRV | 1.80 | 1.87 | — |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •ruvector-attention-wasm leads with a health score of 57/100, but none of the options score above 80.
- •mil_convert is pre-1.0 — API may change.