mirror/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2026-05-08 16:09:08 -04:00
Code Issues Packages Projects Releases Wiki Activity

chore(model gallery): 🤖 add new models via gallery agent (#6522)

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 23:18:28 +02:00
committed by GitHub
parent cce185b345
commit e293b65ad9
1 changed files with 22 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
gallery/index.yaml
View File

@@ -22057,3 +22057,25 @@
- filename: Aevum-0.6B-Finetuned.Q4_K_M.gguf
sha256: 6904b789894a7dae459042a28318e70dbe222cb3e6f892f3fc42e591d4a341a3
uri: huggingface://mradermacher/Aevum-0.6B-Finetuned-GGUF/Aevum-0.6B-Finetuned.Q4_K_M.gguf
- !!merge <<: *hermes-2-pro-mistral
name: "tlacuilo-12b-i1"
urls:
- https://huggingface.co/mradermacher/Tlacuilo-12B-i1-GGUF
description: |
**Tlacuilo-12B** is a high-quality, instruction-tuned language model designed for creative and roleplay writing. Built on the foundation of **Mistral-Nemo-Base-2407** and **Muse-12B**, it excels in narrative generation, storytelling, and interactive dialogue, with notable improvements in prose style and consistency over previous versions.
Trained using a multi-stage LoRA process:
- **Stage 1**: Fine-tuned on literary texts (28M tokens/epoch) to enhance stylistic richness.
- **Stage 2**: Optimized for roleplay using RP datasets (4M tokens), improving character and scenario handling.
- **Stage 3**: Instruct-tuned on curated data (1.2M tokens) to ensure strong response quality and alignment.
The model uses **ChatML** formatting and performs best at moderate temperature (1.0–1.3) with low min-p values. Ideal for writers, game masters, and creative professionals seeking expressive, coherent, and imaginative text generation.
> **Note**: The GGUF versions in `mradermacher/Tlacuilo-12B-i1-GGUF` are quantized derivatives. The original, full-precision model is hosted at `allura-org/Tlacuilo-12B`.
overrides:
parameters:
model: Tlacuilo-12B.i1-Q4_K_M.gguf
files:
- filename: Tlacuilo-12B.i1-Q4_K_M.gguf
sha256: 94218112aa02113c8e21cd2c1d10818bea39bc6aee7e67be6014f86e80e76cb1
uri: huggingface://mradermacher/Tlacuilo-12B-i1-GGUF/Tlacuilo-12B.i1-Q4_K_M.gguf