Open whranclast opened 2 weeks ago
So to maintain the model's "old" knowledge, I would try adding some generic datasets to the finetune (concat), or reduce the number of steps. Another "trick" is before you merge the lora adapters, simply scale all of them by some fraction (say 0.1) to reduce the effect of the adapters. It might not work very well, but it should function (hopefully)
So to maintain the model's "old" knowledge, I would try adding some generic datasets to the finetune (concat), or reduce the number of steps. Another "trick" is before you merge the lora adapters, simply scale all of them by some fraction (say 0.1) to reduce the effect of the adapters. It might not work very well, but it should function (hopefully)
Thank you, in your second suggestion how to exactly scale them in terms of code ?
Oh in the adapter_config.json
file, change lora_alpha
literally
Hello,
I am currently playing the the unsloth library and its performing amazingly, even on my local machine. Unfortunately, I have an issue with the model kind of "forgetting" its generic purpose as I've been training it on "custom task" dataset e.g.
Currently my dataset is around 30k entries and I've been using r=8 and lora_alpha=16 to try to prevent overfitting.
However, now when I ask "What is the capital of France", I get hallucinations of the sort of "Apologies for the issue with order 3124, the reason for the issue is {"id": 3124}".
Correct me if I am wrong but running
if False: model.push_to_hub_gguf("hf/model", tokenizer, quantization_method = "f16", token = "")
merges the lora adapters with the pre-trained weights so in a sense we should keep the "old" knowledge of the model.Maybe reducing the amount of target modules can help ?