chore: make format

cli_chat.py

@@ -132,8 +132,12 @@ def chat(args, tokenizer, vl_chat_processor, vl_gpt, generation_config):
 
                 while cur_img_idx < num_images:
                     try:
-                        image_file = input(f"({cur_img_idx + 1}/{num_images}) Input the image file path: ")
+                        image_file = input(
-                        image_file = image_file.strip() # trim whitespaces around path, enables drop-in from for example Dolphin
+                            f"({cur_img_idx + 1}/{num_images}) Input the image file path: "
+                        )
+                        image_file = (
+                            image_file.strip()
+                        )  # trim whitespaces around path, enables drop-in from for example Dolphin
 
                     except KeyboardInterrupt:
                         print()

deepseek_vl/__init__.py

@@ -20,10 +20,12 @@
 
 # check if python version is above 3.10
 import sys
-if sys.version_info >=(3, 10):
+
+if sys.version_info >= (3, 10):
     print("Python version is above 3.10, patching the collections module.")
     # Monkey patch collections
     import collections
     import collections.abc
+
     for type_name in collections.abc.__all__:
         setattr(collections, type_name, getattr(collections.abc, type_name))

deepseek_vl/models/siglip_vit.py

@@ -338,9 +338,9 @@ class VisionTransformer(nn.Module):
 
         self.num_classes = num_classes
         self.global_pool = global_pool
-        self.num_features = (
+        self.num_features = self.embed_dim = (
-            self.embed_dim
+            embed_dim  # num_features for consistency with other models
-        ) = embed_dim  # num_features for consistency with other models
+        )
         self.num_prefix_tokens = 1 if class_token else 0
         self.num_prefix_tokens += reg_tokens
         self.num_reg_tokens = reg_tokens

deepseek_vl/serve/inference.py

@@ -50,9 +50,11 @@ def convert_conversation_to_prompts(conversation: Conversation):
     for i in range(0, len(messages), 2):
         prompt = {
             "role": messages[i][0],
-            "content": messages[i][1][0]
+            "content": (
+                messages[i][1][0]
                 if isinstance(messages[i][1], tuple)
-            else messages[i][1],
+                else messages[i][1]
+            ),
             "images": [messages[i][1][1]] if isinstance(messages[i][1], tuple) else [],
         }
         response = {"role": messages[i + 1][0], "content": messages[i + 1][1]}

pyproject.toml

@@ -10,7 +10,7 @@ authors = [{name = "DeepSeek-AI"}]
 license = {file = "LICENSE-CODE"}
 urls = {homepage = "https://github.com/deepseek-ai/DeepSeek-VL"}
 readme = "README.md"
-requires-python = ">=3.8, <3.10"
+requires-python = ">=3.8"
 dependencies = [
     "torch>=2.0.1",
     "transformers>=4.38.2",