Transformers
Safetensors
PEFT
mtrini
compiwer-ai
qwen3-vl
lora
reasoning
coding
mathematics
darija
moroccan-arabic
Instructions to use CompiwerAI/Mtrini-SVL-1.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use CompiwerAI/Mtrini-SVL-1.1 with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("CompiwerAI/Mtrini-SVL-1.1", device_map="auto") - PEFT
How to use CompiwerAI/Mtrini-SVL-1.1 with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
|
Download README.md from CompiwerAI/Mtrini-SVL-1.1: direct link, hf CLI and curl.
- Browser
- Download file 2.13 kB
-
https://huggingface.co/CompiwerAI/Mtrini-SVL-1.1/resolve/main/README.md
- Command line
-
hf download hf://CompiwerAI/Mtrini-SVL-1.1/README.md
-
curl -L -o README.md https://huggingface.co/CompiwerAI/Mtrini-SVL-1.1/resolve/main/README.md
2.13 kB
| language: | |
| - en | |
| - ar | |
| - fr | |
| - ary | |
| license: apache-2.0 | |
| library_name: transformers | |
| base_model: Qwen/Qwen3-VL-8B-Instruct | |
| tags: | |
| - mtrini | |
| - compiwer-ai | |
| - qwen3-vl | |
| - lora | |
| - peft | |
| - reasoning | |
| - coding | |
| - mathematics | |
| - darija | |
| - moroccan-arabic | |
| # ๐ช Mtrini-SVL-1.1 | |
| > **Small lab. Big ambitions.** | |
| Mtrini-SVL-1.1 is an 8B model adaptation developed by **Compiwer AI**, built on **Qwen3-VL-8B-Instruct** and trained with a focus on reasoning, coding, mathematics, and Moroccan Darija. | |
| ## ๐ง SVL โ Self-Verifying Loop | |
| **SVL = Self-Verifying Loop** | |
| The idea: | |
| **Think โ Verify โ Improve โ Repeat** | |
| SVL is Compiwer AI's approach to encouraging models to check and refine their responses instead of immediately stopping at the first answer. | |
| ## ๐ Training Data | |
| The final dataset contained approximately **58.6K unique examples**: | |
| - ๐ฒ๐ฆ Moroccan Darija QA + Reasoning โ 2,170 | |
| - ๐ป OpenCodeReasoning-2 โ 30,000 | |
| - ๐งฎ OpenMathInstruct-2 โ 30,000 | |
| ## โ๏ธ Model Details | |
| | Property | Value | | |
| |---|---| | |
| | Model | Mtrini-SVL-1.1 | | |
| | Base | Qwen3-VL-8B-Instruct | | |
| | Parameters | ~8B | | |
| | Architecture | Qwen3-VL | | |
| | Training | LoRA / PEFT | | |
| | Precision | BF16 | | |
| | Training Context | 4096 | | |
| | Developer | Compiwer AI | | |
| | Origin | ๐ฒ๐ฆ Salรฉ, Morocco | | |
| ## ๐งช Training | |
| - 3,661 training steps | |
| - 1 epoch | |
| - Training loss: **0.5855** | |
| - Mean token accuracy: **83.55%** | |
| These are training metrics and are not standardized benchmark results. | |
| ## ๐ฆ Repository | |
| This repository contains the **LoRA adapter** for Mtrini-SVL-1.1. | |
| For the merged model: | |
| **CompiwerAI/Mtrini-SVL-1.1-Merged** | |
| For local GGUF inference: | |
| **CompiwerAI/Mtrini-SVL-1.1-GGUF** | |
| ## โ ๏ธ Limitations | |
| Mtrini-SVL-1.1 is an experimental model and can still hallucinate, make coding mistakes, misunderstand Darija, or produce incorrect mathematical answers. | |
| Always verify important outputs. | |
| ## ๐ฒ๐ฆ Built by Compiwer AI | |
| Mtrini is part of Compiwer AI's effort to build open AI systems focused on: | |
| **Reasoning โข Coding โข Mathematics โข Moroccan Darija โข Open AI** | |
| > **Building AI For Everyone.** |