SequentialLearning
/

SuperLinear

mixture-of-experts

Model card Files Files and versions

razmars commited on May 15

Commit

f8fb71c

·

verified ·

1 Parent(s): 2d25803

Update full_shot/config.json

Files changed (1) hide show

full_shot/config.json +1 -1

full_shot/config.json CHANGED Viewed

@@ -34,7 +34,7 @@
   "torch_dtype": "float32",
   "transformers_version": "4.40.1",
   "use_fft": 1,
-  "train_epochs": 10,
   "patience": 5,
   "lradj": "type1",
   "learning_rate": 0.005,

   "torch_dtype": "float32",
   "transformers_version": "4.40.1",
   "use_fft": 1,
+  "train_epochs": 30,
   "patience": 5,
   "lradj": "type1",
   "learning_rate": 0.005,