Update app.py

app.py

@@ -11,34 +11,35 @@ USE_LORA = False
 USE_QLORA = True
 
 processor = AutoProcessor.from_pretrained("HuggingFaceM4/idefics2-8b", do_image_splitting=False)
+model = AutoModelForPreTraining.from_pretrained("HuggingFaceM4/idefics2-8b")
 
-if USE_QLORA or USE_LORA:
-    lora_config = LoraConfig(
-        r=8,
-        lora_alpha=8,
-        lora_dropout=0.1,
-        target_modules='.*(text_model|modality_projection|perceiver_resampler).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*',
-        use_dora=False if USE_QLORA else True,
-        init_lora_weights="gaussian"
-    )
-    bnb_config = BitsAndBytesConfig(
-        load_in_4bit=True,
-        bnb_4bit_quant_type="nf4",
-        bnb_4bit_compute_dtype=torch.float16
-    ) if USE_QLORA else None
-    model = Idefics2ForConditionalGeneration.from_pretrained(
-        "HuggingFaceM4/idefics2-8b",
-        torch_dtype=torch.float16,
-        quantization_config=bnb_config,
-    )
-    model.add_adapter(lora_config)
-    model.enable_adapters()
-else:
-    model = Idefics2ForConditionalGeneration.from_pretrained(
-        "HuggingFaceM4/idefics2-8b",
-        torch_dtype=torch.float16,
-        _attn_implementation="flash_attention_2"
-    ).to(DEVICE)
+# if USE_QLORA or USE_LORA:
+#     lora_config = LoraConfig(
+#         r=8,
+#         lora_alpha=8,
+#         lora_dropout=0.1,
+#         target_modules='.*(text_model|modality_projection|perceiver_resampler).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*',
+#         use_dora=False if USE_QLORA else True,
+#         init_lora_weights="gaussian"
+#     )
+#     bnb_config = BitsAndBytesConfig(
+#         load_in_4bit=True,
+#         bnb_4bit_quant_type="nf4",
+#         bnb_4bit_compute_dtype=torch.float16
+#     ) if USE_QLORA else None
+#     model = Idefics2ForConditionalGeneration.from_pretrained(
+#         "HuggingFaceM4/idefics2-8b",
+#         torch_dtype=torch.float16,
+#         quantization_config=bnb_config,
+#     )
+#     model.add_adapter(lora_config)
+#     model.enable_adapters()
+# else:
+#     model = Idefics2ForConditionalGeneration.from_pretrained(
+#         "HuggingFaceM4/idefics2-8b",
+#         torch_dtype=torch.float16,
+#         _attn_implementation="flash_attention_2"
+#     ).to(DEVICE)
 
 def model_inference(image, text):
     resulting_messages = [{"role": "user", "content": [{"type": "image"}] + [{"type": "text", "text": text}]}]