Nishith Jain
Nishith Jain
How we can fine tune Multimodal LLM like Phi 3 using unsloth package ? ``` config.json: 100% 684/684 [00:00 4 model, tokenizer = FastLanguageModel.from_pretrained( 5 model_name = "HuggingFaceM4/idefics2-8b", 6 max_seq_length...
I updated prompt of AI to get better search result and understands users demand and generate useful query. Why I updated this? Because current model is good in generating query...
I am a beginner, so I was not able to complete it entirely, but I have created a structure for it. Remaining Task: Display Assistants - there should be 6-8...
Added input type option From which user can manually select input_type or default auto mode. Also, added 2 new types of input type: bytes and base64.