Finetuning Query for gpt-oss
I would say at first attempts it gives good output, short queries it works, for longer and complex examples it happens.
Coming to Data - I prepared it with GPT 5.2 with longer and shorter samples with analysis as well.
For EOS <|return|>, channel <|end|>, is handled proper and extra validation checks are there also I used harmony github repo creating conversation.
Temperature: Tried 1 and 0,0.3 as well, repeating penalties, top k=20,30,top p=0.9, num predict.
Thank you for your support and guidance
Please let me know how should I tackle this and your guidance to do a successful fine tuning with custom dataset.
My question:
Your guidance on the above thinking loop.
Do I add analysis channel data into training or not.
Custom chat template to work with inference and training later ollama compatibility
My max sample token is 8k most of the samples have 3,4k token and smaller as well.
Which quant is best q4, 5,8 or any other
Discussion in the ATmosphere