Update app_dialogue.py

app_dialogue.py

@@ -19,12 +19,12 @@ subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENT
 
 DEVICE = torch.device("cuda")
 MODELS = {
-    "282 - mix1 fixed - opt 23'000": AutoModelForCausalLM.from_pretrained(
+    "288 - mix8 - opt 5'800": AutoModelForCausalLM.from_pretrained(
         "HuggingFaceM4/idefics2",
         trust_remote_code=True,
         torch_dtype=torch.bfloat16,
         token=os.environ["HF_AUTH_TOKEN"],
-        revision="a1bc6a2b0f74cde25844144f602dde2808a564d9",
+        revision="25bb7ad6d9ab9e43d5002d30f857d4106ed964f3",
     ).to(DEVICE),
     "282 - mix 5 - opt 18'500": AutoModelForCausalLM.from_pretrained(
         "HuggingFaceM4/idefics2",