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 sha256:689c70fe1e30cf49d20d1da563b8bda0043622fe7d1ba4dc0318af1c38e1b6b5
+size 96681624

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)
 
 
 

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4e4d636ecd56378378a013c5d1b55190b138b5db258d60faafb6b1a3033accdc
-size 96780961
+oid sha256:02b86bbac1775cc330983bec0b83979e6bea4e046120fc8f07c20fec7421df0f
+size 96757363