In the first iteration of the project, I used fine tuned qwen model for the ai but since we're moving away to a procedural generation model and since any base model performs pretty much as good, I'm thinking we can get rid of the model weights and the hugging face model download stuff. LM Studio and Gemini setup is far more simpler as we know it.
In the first iteration of the project, I used fine tuned qwen model for the ai but since we're moving away to a procedural generation model and since any base model performs pretty much as good, I'm thinking we can get rid of the model weights and the hugging face model download stuff. LM Studio and Gemini setup is far more simpler as we know it.