unslothai / unsloth

Finetune Llama 3.2, Mistral, Phi & Gemma LLMs 2-5x faster with 80% less memory
https://unsloth.ai
Apache License 2.0
17.33k stars 1.2k forks source link

does unlsoth support freeze tunning #1183

Open NathanaelTamirat opened 13 hours ago

NathanaelTamirat commented 13 hours ago

does unlsoth support freeze tunning please its urgunt

danielhanchen commented 6 hours ago

Current not sorry - though you could manually set some layers to require no grads

NathanaelTamirat commented 3 hours ago

i see thank you