danthe3rd
commented
11 months ago Hi, For simplicity, our example was only meant to work with llama/llama V2/codellama architectures. I'm not familiar with TinyLlama, but it might have a different architecture
Hello, I saw this example code with a llama model, I tried to replicate it with my own tinyllama model (leaded as PEFT model), but during inference it seems to generate random words. Is there a example script with PEFT models or tinyllama? Thank you.