Create app.py

app.py

@@ -0,0 +1,44 @@
+from huggingface_hub import hf_hub_download
+from llama_cpp import Llama
+import gradio as gr
+
+model_name_or_path = "TheBloke/Llama-2-13B-chat-GGML"
+model_basename = "llama-2-13b-chat.ggmlv3.q2_K.bin" # the model is in bin format    #llama-2-13b-chat.ggmlv3.q2_K.bin  -> menos pesado   | "llama-2-13b-chat.ggmlv3.q5_1.bin" -> más pesado
+
+model_path = hf_hub_download(repo_id=model_name_or_path, filename=model_basename)
+
+# GPU
+lcpp_llm = None
+lcpp_llm = Llama(
+    model_path=model_path,
+    n_threads=2, # CPU cores
+    n_batch=512, # Should be between 1 and n_ctx, consider the amount of VRAM in your GPU.
+    n_gpu_layers=32 # Change this value based on your model and your GPU VRAM pool.
+    )
+
+def CustomChatGPT(Ask):
+    prompt = Ask
+    prompt_template=f'''SYSTEM: You are an NBA expert that helps answering questions about the NBA, its teams and its players summarizing the most important information limiting to no more that one hundred tokens.
+
+    USER: {prompt}
+
+    ASSISTANT:
+    '''
+    response=lcpp_llm(prompt=prompt_template, max_tokens=100, temperature=0.3, top_p=0.95,
+                  repeat_penalty=1.2, top_k=150,
+                  echo=True)
+    
+    reply=response["choices"][0]["text"]
+    # Split the response by lines
+    #response_lines = reply.split("\n")
+
+    # Find the index of the ASSISTANT line
+    #assistant_index = response_lines.index("ASSISTANT:")
+
+    # Extract the assistant response by joining the lines after the ASSISTANT line
+    #assistant_response = "\n".join(response_lines[assistant_index + 1:])
+    return reply
+
+gui=gr.Interface(fn=CustomChatGPT, inputs="text", outputs="text", examples=["Who is the greatest basketball player in NBA history","What is the winning record in a season?"] , title="Ask the AI coach", description="Ask the AI coach all you want about NBA Teams and Players:")
+
+gui.launch()