|
|
@@ -24,7 +24,7 @@ from peft import (
|
|
|
MICRO_BATCH_SIZE = 4 # this could actually be 5 but i like powers of 2
|
|
|
BATCH_SIZE = 128
|
|
|
GRADIENT_ACCUMULATION_STEPS = BATCH_SIZE // MICRO_BATCH_SIZE
|
|
|
-EPOCHS = 5 # remember, we're loading the best checkpoint with the val set
|
|
|
+EPOCHS = 3 # remember, we're loading the best checkpoint with the val set
|
|
|
LEARNING_RATE = 3e-4 # the Karpathy constant
|
|
|
CUTOFF_LEN = 256 # 256 accounts for about 96% of the data
|
|
|
LORA_R = 8
|