fix internvl-4b (modelscope#3178)

Jintao-Huang · Feb 19, 2025 · 12ec92f · 12ec92f
1 parent 7989581
commit 12ec92f
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 4 deletions.
diff --git a/swift/llm/infer/infer_engine/pt_engine.py b/swift/llm/infer/infer_engine/pt_engine.py
@@ -420,6 +420,15 @@ def _infer(
                 except Exception as e:
                     error_list.append((i, e))
                     continue
+        if len(batched_inputs) == 0:
+            if request_config.stream:
+
+                def _gen_wrapper():
+                    yield self._add_error_list([], error_list)
+
+                return _gen_wrapper()
+            else:
+                return self._add_error_list([], error_list)
         template_inputs = [inputs.pop('template_inputs') for inputs in batched_inputs]
         inputs = to_device(template.data_collator(batched_inputs), self.model.device)
         template.debug_logger(inputs)  # debug

diff --git a/swift/llm/template/base.py b/swift/llm/template/base.py
@@ -673,10 +673,10 @@ def _swift_encode(self, inputs: StdTemplateInputs):
         res_context_types: List[ContextType] = []
         sep_token = None
         if template_meta.auto_add_bos:
-            all_tokens = self.tokenizer.encode('0')
-            single_zero = self.tokenizer.encode('0', add_special_tokens=False)
-            assert len(single_zero) == 1
-            idx = all_tokens.index(single_zero[0])
+            all_tokens = self.tokenizer.encode('a')
+            single_token = self.tokenizer.encode('a', add_special_tokens=False)
+            assert len(single_token) == 1
+            idx = all_tokens.index(single_token[0])
             bos_token = all_tokens[:idx]
             sep_token = all_tokens[idx + 1:]
             if bos_token: