Create app.py

app.py

@@ -0,0 +1,36 @@
+import gradio as gr
+from huggingface_hub import snapshot_download
+from pathlib import Path
+from mistral.cli.chat import load_model, generate_stream
+
+# Download the model
+mistral_models_path = Path.home().joinpath('mistral_models', 'mamba-codestral-7B-v0.1')
+mistral_models_path.mkdir(parents=True, exist_ok=True)
+
+snapshot_download(repo_id="mistralai/mamba-codestral-7B-v0.1", 
+                  allow_patterns=["params.json", "consolidated.safetensors", "tokenizer.model.v3"], 
+                  local_dir=mistral_models_path)
+
+# Load the model
+model = load_model(str(mistral_models_path))
+
+def generate_response(message, history):
+    history_mistral_format = [
+        {"role": "user" if i % 2 == 0 else "assistant", "content": m}
+        for i, m in enumerate(sum(history, []))
+    ]
+    history_mistral_format.append({"role": "user", "content": message})
+    
+    response = ""
+    for chunk in generate_stream(model, history_mistral_format, max_tokens=256):
+        response += chunk
+        yield response
+
+iface = gr.ChatInterface(
+    generate_response,
+    title="Mamba Codestral Chat",
+    description="Chat with the Mamba Codestral 7B model.",
+)
+
+if __name__ == "__main__":
+    iface.launch()