Create app.py

app.py

@@ -0,0 +1,17 @@
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
+
+model_id = "bartowski/cognitivecomputations_Dolphin-Mistral-24B-Venice-Edition-GGUF"
+tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=False)
+model     = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
+chat      = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=200)
+
+def respond(prompt):
+    return chat(prompt)[0]["generated_text"]
+
+gr.Interface(
+    fn=respond,
+    inputs=gr.Textbox(lines=2, placeholder="say something…"),
+    outputs="text",
+    title="dolphin mistral 24b venice chat",
+).launch()