Create app.py

app.py

@@ -0,0 +1,47 @@
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+# 1. Setup the Model Name
+model_name = "haykgrigorian/TimeCapsuleLLM-v2-llama-1.2B"
+
+# 2. Load the Model and Tokenizer
+# We use device_map="auto" to use available CPU RAM efficiently
+print("Loading model... this usually takes 1-2 minutes on first run.")
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name)
+
+# 3. Define the Generate Function
+def generate_text(prompt, max_tokens=100, temperature=0.7):
+    # Format inputs
+    inputs = tokenizer(prompt, return_tensors="pt")
+    
+    # Generate
+    # We disable gradients to save memory and speed up inference
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs, 
+            max_new_tokens=int(max_tokens),
+            temperature=float(temperature),
+            do_sample=True, # Allows for creativity/temperature
+            pad_token_id=tokenizer.eos_token_id
+        )
+    
+    # Decode result
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+# 4. Launch the Gradio Interface
+# This creates the UI and the API endpoint automatically
+iface = gr.Interface(
+    fn=generate_text,
+    inputs=[
+        gr.Textbox(label="Prompt", placeholder="Enter your text here..."),
+        gr.Slider(minimum=10, maximum=300, value=100, label="Max New Tokens"),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.7, label="Temperature")
+    ],
+    outputs="text",
+    title="TimeCapsule LLM API",
+    description="API for n8n connection."
+)
+
+iface.launch()