Update README.md

README.md

@@ -10,7 +10,7 @@ tags:
 inference: true
 ---
 
-I used nearly 200,000 data for training, learning_rate 5e-6, training with a6000x1, epoch 10, because I am too busy recently, I should not be able to actively do it, and the funds are slightly insufficient
+I used nearly 200,000 data for training, learning_rate 2.5e-6, training with a6000x1, because I am too busy recently, I should not be able to actively do it, and the funds are slightly insufficient
 ,Forget it, I'm overtraining, take it as an interesting model,(Warning: above 768x832 is recommended, I found that the results below seem to be less than ideal)
 
 (It's not finished yet, this is just an early version