Update app.py

app.py

@@ -1,27 +1,37 @@
 import gradio as gr
-from transformers import AutoModelForCausalLM, AutoTokenizer
+from transformers import AutoTokenizer, AutoModel
 import torch
+import json
 
 # Load the tokenizer
 model_name = "TuringsSolutions/TechLegalV1"
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 
-# Load the model
-# Assuming it's a CausalLM model, you might need to adjust based on your model's architecture
-model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True)
+# Load adapter configuration manually
+adapter_config_path = "https://huggingface.co/TuringsSolutions/TechLegalV1/resolve/main/adapter_config.json"
+adapter_model_path = "https://huggingface.co/TuringsSolutions/TechLegalV1/resolve/main/adapter_model.safetensors"
+
+with open(adapter_config_path, 'r') as f:
+    adapter_config = json.load(f)
+
+# Initialize the model with the adapter configuration
+model = AutoModel.from_pretrained(model_name, trust_remote_code=True)
+
+# Load adapter weights
+model.load_adapter(adapter_model_path, config=adapter_config)
 
 # Function to make predictions
 def predict(text):
     inputs = tokenizer(text, return_tensors="pt")
     with torch.no_grad():
-        outputs = model.generate(**inputs)
-    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+        outputs = model(**inputs)
+    return outputs.last_hidden_state.mean(dim=1).squeeze().tolist()
 
 # Create a Gradio interface
 iface = gr.Interface(
     fn=predict,
     inputs=gr.inputs.Textbox(lines=2, placeholder="Enter text here..."),
-    outputs="text",
+    outputs="json",
     title="Tech Legal Model",
     description="A model for analyzing tech legal documents."
 )