# Catalan chat model MiniCPM5-2B-catalan-chat-v2 released by community member

> Original title: 🌍 “Hola!” from the MiniCPM community!

- Source: OpenBMB
- Published: 2026-11-03T13:30:10.000Z
- Bad Influence: https://badinfluence.news/items/yb65reo6lh5qs7u87w7bx7vzu
- Original: https://x.com/OpenBMB/status/2117610288552038581

## Summary

A community member, luispoveda93, released MiniCPM5-2B-catalan-chat-v2, a Catalan conversational model built from OpenBMB's MiniCPM5-2B. The model was trained in two LoRA fine-tuning stages: 165,100 single-turn instruction examples from Projecte AINA's InstruCAT dataset, then about 23.4K multi-turn chat examples from BSC-LT/ALIA-2606-SFT. Merged BF16 weights, LoRA adapters, and community GGUF versions for llama.cpp and Ollama are available.

## Post

🌍 “Hola!” from the MiniCPM community!

Meet MiniCPM5-2B-catalan-chat-v2, a Catalan conversational model built by our ambassador luispoveda93 through two-stage LoRA fine-tuning of OpenBMB’s MiniCPM5-2B.

📚 Stage 1 — Single-turn instruction tuning
165,100 examples from Projecte AINA’s InstruCAT dataset, covering Q&A, summarization, paraphrasing, and more.

💬 Stage 2 — Multi-turn conversation
Continued fine-tuning on a ~23.4K chat-focused subset of BSC-LT/ALIA-2606-SFT, focusing on dialogue flow and instruction following across turns.

💻 Ready to explore locally
Merged BF16 weights and LoRA adapters are available, alongside community-contributed GGUF versions for local use with llama.cpp and Ollama.

🛠️ Building for your language? This community project offers a concrete example of adapting MiniCPM with language-specific instructions and dialogue data.

Explore the model 👇
http://huggingface.co/luispoveda93/MiniCPM5-2B-catalan-chat-v2

🚀 Build with MiniCPM5-2B:
GitHub: https://github.com/OpenBMB/MiniCPM
Hugging Face: https://huggingface.co/openbmb/MiniCPM5-2B

#MiniCPM #OpenBMB #Catalan #LoRA #LocalAI

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWNDVacacAAf4X-.jpg&mode=thumb&exp=1793923200&sig=8589e1cecfe3f1d2)