Adding `safetensors` variant of this model

config.json

@@ -2,7 +2,8 @@
   "architectures": ["MoEForCausalLM"],
   "auto_map": {
     "AutoConfig": "moe.MoELMConfig",
-    "AutoModelForCausalLM": "moe.MoEForCausalLM"
+    "AutoModelForCausalLM": "moe.MoEForCausalLM",
+    "AutoModel": "moe.MoEModel"
   }
   ,
   "attn_pdrop": 0.1,

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb78b4c625012ce0d10bc2099f96699ceeb704042961edb7a909027166c35962
+size 96683760

moe.py

@@ -7,7 +7,7 @@ import torch.nn as nn
 import torch.nn.functional as F
 
 from transformers import PreTrainedModel, PretrainedConfig
-from transformers import AutoConfig, AutoModelForCausalLM
+from transformers import AutoConfig, AutoModelForCausalLM, AutoModel
 
 
 from transformers.modeling_outputs import (
@@ -553,6 +553,7 @@ class MoEForCausalLM(PreTrainedModel):
 
 AutoConfig.register("moe_lm", MoELMConfig)
 AutoModelForCausalLM.register(MoELMConfig, MoEForCausalLM)
+AutoModel.register(MoELMConfig, MoEModel)