mirror/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2026-01-07 02:59:54 -06:00
Code Issues Packages Projects Releases Wiki Activity

chore(model gallery): 🤖 add new models via gallery agent (#6512)

Browse Source 
Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: mudler <2420543+mudler@users.noreply.github.com>
This commit is contained in:
LocalAI [bot]
2025-10-17 16:04:54 +02:00
committed by GitHub
parent 33c48164d7
commit c54f5cdf12
1 changed files with 27 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
gallery/index.yaml
View File

@@ -21933,3 +21933,30 @@
- filename: Magidonia-24B-v4.2.0.i1-Q4_K_M.gguf
sha256: f89fbe09ea9edd4b91aa89516cbfaabdf0d956e0458cfc4b44b8054a1546b559
uri: huggingface://mradermacher/Magidonia-24B-v4.2.0-i1-GGUF/Magidonia-24B-v4.2.0.i1-Q4_K_M.gguf
- !!merge <<: *qwen25coder
name: "strand-rust-coder-14b-v1-i1"
urls:
- https://huggingface.co/mradermacher/Strand-Rust-Coder-14B-v1-i1-GGUF
description: |
**Strand-Rust-Coder-14B-v1** is a high-performance, domain-specialized large language model fine-tuned for Rust programming. Built on **Qwen2.5-Coder-14B-Instruct** using LoRA and trained on a 191K-example synthetic dataset generated via peer-reviewed Swarm Inference, it achieves state-of-the-art performance on Rust-specific benchmarks — outperforming larger proprietary models like GPT-5 Codex on tasks such as code generation, test creation, and refactoring.
Key highlights:
- **Specialized for Rust**: Optimized for ownership, lifetimes, and idiomatic syntax.
- **High accuracy**: 48% pass rate on hold-out test sets and 43% on RustEvo².
- **Efficient training**: Fine-tuned with LoRA (r=64, α=16), updating only ~1% of parameters.
- **Extended context**: 32,768 tokens for complex code reasoning.
- **Open & decentralized**: Part of the Fortytwo Network, enabling collaborative, peer-validated inference.
Ideal for developers, AI code assistants, and researchers working on Rust. Available in GGUF quantized versions for local deployment (Q4_K_M to Q8_0).
> **Base model**: Qwen2.5-Coder-14B-Instruct
> **Dataset**: Fortytwo-Network/Strandset-Rust-v1
> **License**: Apache 2.0
> **Repository**: [Fortytwo-Network/Strand-Rust-Coder-14B-v1](https://huggingface.co/Fortytwo-Network/Strand-Rust-Coder-14B-v1)
overrides:
parameters:
model: Strand-Rust-Coder-14B-v1.i1-Q4_K_M.gguf
files:
- filename: Strand-Rust-Coder-14B-v1.i1-Q4_K_M.gguf
sha256: 56bacda1fd25cdebd349df839c3ef3248a3e7cb98da6ce1ee1f4e8c2153e777e
uri: huggingface://mradermacher/Strand-Rust-Coder-14B-v1-i1-GGUF/Strand-Rust-Coder-14B-v1.i1-Q4_K_M.gguf