search

jina-ai / jerboa

LLM finetuning
Apache License 2.0
42 stars 4 forks source link

feat(finetune): added low footprint Llama model for debugging #3

Closed sebastian-weisshaar closed 1 year ago

sebastian-weisshaar commented 1 year ago

If debug==True a low footprint version of LlamaForCausalLM gets initialised without the trained weights. Memory footprint is 10x smaller.