This is a test project, performance will be bad
The base model is OpenLlama-3B, this model was finetuned on 50k examples from the following dataset. https://www.kaggle.com/datasets/gauravduttakiit/bitcoin-tweets-16m-tweets-with-sentiment-tagged/ I used H2O Studio for fine tuning this model.
This project should be used on MLC Chat (Android) https://llm.mlc.ai/docs/index.html#getting-started.