workatournd for overflow

Signed-off-by: wenhuach <[email protected]>

config.json

@@ -79,7 +79,7 @@
   "tie_word_embeddings": false,
   "topk_group": 4,
   "topk_method": "noaux_tc",
-  "torch_dtype": "bfloat16",
+  "torch_dtype": "float16",
   "transformers_version": "4.47.0",
   "use_cache": true,
   "v_head_dim": 128,

modeling_deepseek.py

@@ -386,7 +386,7 @@ class DeepseekV3MLP(nn.Module):
         self.act_fn = ACT2FN[config.hidden_act]
 
     def forward(self, x):
-        down_proj = self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
+        down_proj = self.down_proj(torch.clip(self.act_fn(self.gate_proj(x)) * self.up_proj(x),-65504,65504))
         return down_proj