SequentialLearning
/

SuperLinear

mixture-of-experts

Model card Files Files and versions

razmars commited on May 13

Commit

0b9df35

·

verified ·

1 Parent(s): d714dab

Update zero_shot/config.json

Files changed (1) hide show

zero_shot/config.json +2 -1

zero_shot/config.json CHANGED Viewed

@@ -34,5 +34,6 @@
   "top_k_experts": 12,
   "torch_dtype": "float32",
   "transformers_version": "4.40.1",
-  "use_fft": 1
 }

   "top_k_experts": 12,
   "torch_dtype": "float32",
   "transformers_version": "4.40.1",
+  "use_fft": 1,
+  "misc_moe" :10,
 }