fix attn_bias default value

InternLM · Nov 8, 2024 · 4c55c8d · 4c55c8d
1 parent 48d1a5c
commit 4c55c8d
Showing 1 changed file with 4 additions and 2 deletions.
diff --git a/lmdeploy/turbomind/deploy/source_model/llava.py b/lmdeploy/turbomind/deploy/source_model/llava.py
@@ -33,7 +33,7 @@ def __init__(self, model_path: str, tokenizer_path: str, **kwargs):
         config = getattr(config, 'text_config', config)
         arch = config.architectures[0]
         _readers = dict(Qwen2ForCausalLM=LlavaReader,
-                        LlamaForCausalL=LlavaReader)
+                        LlamaForCausalLM=LlavaReader)
         self.Reader = _readers[arch]
         self.arch = arch
 
@@ -63,7 +63,9 @@ def model_info(self):
             hidden_units = model_arg.get('hidden_size', 4096)
             vocab_size = model_arg.get('vocab_size', 152000)
             intermediate_size = model_arg.get('intermediate_size', 11008)
-            attn_bias = int(model_arg.get('attn_bias', 1))
+            attn_bias = 1 if model_arg['architectures'][0] \
+                == 'Qwen2ForCausalLM' else 0
+            attn_bias = int(model_arg.get('attn_bias', attn_bias))
             use_logn_attn = int(model_arg.get('use_logn_attn', 0))
 
             if isinstance(rope_scaling, dict):