Text Generation
Transformers
Safetensors
English
Russian
llama
llama3
llama3-12b
instruct
Merge
conversational
Inference Endpoints
text-generation-inference

What merge method was used?

#1
by rombodawg - opened

How did you merge the model together? And then how did you upscale them?

What order was it done in? Were the models merged then the resulting model upscaled?
Or
Were all the models upscaled then merged?

If you can share the mergekit .yml configs that would be great. And was the resulting model trained at all, or just merged?

Sign up or log in to comment