ChartLlama is trained on our proposed dataset, based on LLaVA-1.5. You can first download the Lora weights and then combine them using your local LLaVA-1.5 scripts.
- Downloads last month
- 121
Inference API (serverless) is not available, repository is disabled.