diff --git a/server/text_generation_server/models/custom_modeling/flash_dbrx_modeling.py b/server/text_generation_server/models/custom_modeling/flash_dbrx_modeling.py
index e469495f..41aa5859 100644
--- a/server/text_generation_server/models/custom_modeling/flash_dbrx_modeling.py
+++ b/server/text_generation_server/models/custom_modeling/flash_dbrx_modeling.py
@@ -711,7 +711,7 @@ class FlashDbrxForCausalLM(torch.nn.Module):
         else:
             prefix = f"{prefix}.transformer"
 
-        self.model = DbrxModel(config, weights)
+        self.model = DbrxModel(prefix, config, weights)
         self.lm_head = SpeculativeHead.load(
             config,
             prefix="lm_head",
diff --git a/server/text_generation_server/models/custom_modeling/opt_modeling.py b/server/text_generation_server/models/custom_modeling/opt_modeling.py
index 5ab02959..84a1c069 100644
--- a/server/text_generation_server/models/custom_modeling/opt_modeling.py
+++ b/server/text_generation_server/models/custom_modeling/opt_modeling.py
@@ -757,7 +757,7 @@ class OPTForCausalLM(OPTPreTrainedModel):
         else:
             prefix = f"{prefix}.model"
 
-        self.model = OPTModel(config, weights)
+        self.model = OPTModel(prefix, config, weights)
 
         self.lm_head = SpeculativeHead.load(
             config, prefix=f"{prefix}.decoder.embed_tokens", weights=weights