DeepseekV3ForCausalLM

The diff reflects that most differences between [modeling_glm4_moe_lite.py](https://github.com/huggingface/transformers/blob/main/src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py) and [modeling_deepseek_v3.p](https://github.com/huggingface/transformers/blob/main/src/transformers/models/deepseek_v3/modeling_deepseek_v3.py)y are just naming changes.

Question: we simply use `DeepseekV3ForCausalLM` here?

Files changed (1) hide show

config.json +1 -1

config.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "architectures": [
-    "Glm4MoeLiteForCausalLM"
   ],
   "attention_bias": false,
   "attention_dropout": 0.0,

 {
   "architectures": [
+    "DeepseekV3ForCausalLM"
   ],
   "attention_bias": false,
   "attention_dropout": 0.0,