Update app.py

app.py

@@ -5,7 +5,7 @@ import torch
 
 from PIL import Image
 from decord import VideoReader, cpu
-from transformers import AutoModel, AutoTokenizer
+from transformers import AutoModel, AutoTokenizer, AutoProcessor
 
 # Pre-Initialize
 DEVICE = "auto"
@@ -18,8 +18,12 @@ DEFAULT_INPUT = "Describe in one paragraph."
 MAX_FRAMES = 64
 
 repo_name = "openbmb/MiniCPM-o-2_6" # "openbmb/MiniCPM-V-2_6-int4" # "openbmb/MiniCPM-V-2_6"
-repo = AutoModel.from_pretrained(repo_name, trust_remote_code=True, attn_implementation="sdpa", torch_dtype=torch.bfloat16)
+repo = AutoModel.from_pretrained(repo_name, 
+            init_vision=True,
+            init_audio=False,
+            init_tts=False, trust_remote_code=True, attn_implementation="sdpa", torch_dtype=torch.bfloat16)
 tokenizer = AutoTokenizer.from_pretrained(repo_name, trust_remote_code=True)
+processor = AutoProcessor.from_pretrained(repo_name, trust_remote_code=True)
 
 css = '''
 .gradio-container{max-width: 560px !important}
@@ -72,6 +76,8 @@ def generate(image, video, audio, instruction=DEFAULT_INPUT, sampling=False, tem
         return "No input provided."
     
     parameters = {
+        "tokenizer": tokenizer,
+        "processor": processor,
         "sampling": sampling,
         "temperature": temperature,
         "top_p": top_p,
@@ -80,7 +86,7 @@ def generate(image, video, audio, instruction=DEFAULT_INPUT, sampling=False, tem
         "max_new_tokens": max_tokens,
     }
     
-    output = repo.chat(image=None, msgs=inputs, tokenizer=tokenizer, **parameters)
+    output = repo.chat(msgs=inputs, tokenizer=tokenizer, **parameters)
     
     print(output)