--- title: AlpineLLM Live Demo emoji: 🏔️ colorFrom: indigo colorTo: blue sdk: gradio sdk_version: 5.42.0 app_file: app.py pinned: false hf_oauth: true hf_oauth_scopes: - inference-api license: mit short_description: A domain-specific language model for alpine storytelling. --- # AlpineLLM Live Demo A domain-specific language model for alpine storytelling. Try asking about mountain adventures! 🏔️ ## About AlpineLLM AlpineLLM-Tiny-10M-Base is a lightweight base language model with ~10.8 million trainable parameters. It was pre-trained from scratch on raw text corpora drawn primarily from public-domain literature on alpinism, including expedition narratives and climbing essays. This demo showcases the model’s text generation capabilities within its specialized domain. Please note that AlpineLLM is a base model, and it has not been fine-tuned for downstream tasks such as summarization or dialogue. Its outputs reflect patterns learned directly from the training texts. This space shows a free CPU-only demo of the model, so inference may take a few seconds. Text generation of the tiny model may lack full coherence due to its limited size and character-level tokenization. For improved results, consider using the source repository to load larger pretrained weights and run inference on a GPU. Complete source code and full model documentation is available in the related repositories. ### Related Repositories - [**🤗 AlpineLLM Model Page @ HuggingFace**](https://huggingface.co/Borzyszkowski/AlpineLLM-Tiny-10M-Base) - [**⛏️ AlpineLLM Source Code @ GitHub**](https://github.com/Borzyszkowski/AlpineLLM) ### How to install? The software has been tested on Ubuntu 20.04 with CUDA 12.2 and Python3.10. Please use a Python virtual environment to install the dependencies: python3.10 -m venv venv_AlpineLLM source venv_AlpineLLM/bin/activate pip install -r requirements.txt ### How to start? The application starts automatically upon pushing changes to the Hugging Face Space. For local development, please run: ``` python app.py ``` ### Contact and technical support - Bartek Borzyszkowski
Web: borzyszkowski.github.io