[BugFix][Model] Fix commandr RoPE max_position_embeddings (vllm-proje…

…ct#3919)
xjpang · Apr 8, 2024 · d036198 · d036198
1 parent 59a6abf
commit d036198
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/vllm/model_executor/models/commandr.py b/vllm/model_executor/models/commandr.py
@@ -140,7 +140,9 @@ def __init__(
         self.q_size = self.num_heads * self.head_dim
         self.kv_size = self.num_kv_heads * self.head_dim
         self.scaling = self.head_dim**-0.5
-        self.max_position_embeddings = config.max_position_embeddings
+        self.max_position_embeddings = getattr(
+            config, "model_max_length", None) or getattr(
+                config, "max_position_embeddings", 8192)
         self.rope_theta = config.rope_theta
         self.rope_scaling = getattr(config, "rope_scaling", None)
         self.use_qk_norm = getattr(config, "use_qk_norm", False)