All my finetuning jobs are failing when epochs=8
My data is just 150 rows out if which 90 are for training ,60 for validation
batch_szie=4,prompt loss weight=0.1,learning rate multiplier =0.
finetunes.follow gives Fine tune will exceed billing hard limit
Also When I reduced it to 1epoch the jib was successful but it took 2 hrs which is too long for 150 rows
Can I do something for getting queue number 1
Please help me asap on this