Create app.py

app.py

@@ -0,0 +1,77 @@
+import os
+import json
+import torch
+import random
+from transformers import pipeline, set_seed
+import gradio as gr
+
+# Model setup
+CACHE_DIR = "./model_cache"
+os.makedirs(CACHE_DIR, exist_ok=True)
+generator = pipeline(
+    "text-generation",
+    model="openai-community/openai-gpt",
+    cache_dir=CACHE_DIR,
+    device_map="cpu",
+    torch_dtype=torch.float32
+)
+
+# State for chat history
+def init_history():
+    return []
+
+def generate_and_record(prompt, max_length, temperature, top_k, top_p, repetition_penalty, seed, num_return_sequences, history):
+    if seed is not None:
+        set_seed(seed)
+    outputs = generator(
+        prompt,
+        max_length=max_length,
+        temperature=temperature,
+        top_k=top_k,
+        top_p=top_p,
+        repetition_penalty=repetition_penalty,
+        num_return_sequences=num_return_sequences,
+        do_sample=True
+    )
+    texts = [out["generated_text"] for out in outputs]
+    entry = {"prompt": prompt, "results": texts}
+    history.append(entry)
+    return "\n\n---\n\n".join(texts), history
+
+def export_history(history):
+    path = "chat_history.json"
+    with open(path, "w", encoding="utf-8") as f:
+        json.dump(history, f, ensure_ascii=False, indent=2)
+    return path
+
+with gr.Blocks(title="GPT Text Generation") as demo:
+    gr.Markdown("## Text Generation with openai-community/openai-gpt (CPU)")
+
+    with gr.Row():
+        prompt_input = gr.Textbox(label="Prompt", placeholder="Enter your prompt here", lines=2)
+        max_length = gr.Slider(32, 1024, value=128, step=32, label="Max Length")
+    with gr.Row():
+        temperature = gr.Slider(0.1, 1.5, value=1.0, step=0.1, label="Temperature")
+        top_k = gr.Slider(0, 100, value=50, step=1, label="Top-K Sampling")
+    with gr.Row():
+        top_p = gr.Slider(0.0, 1.0, value=0.9, step=0.05, label="Top-P (Nucleus) Sampling")
+        repetition_penalty = gr.Slider(0.5, 2.0, value=1.1, step=0.1, label="Repetition Penalty")
+    seed_input = gr.Number(value=None, precision=0, label="Seed (optional)")
+    num_seq = gr.Dropdown(choices=[1, 2, 3, 5], value=1, label="Number of Generations")
+
+    generate_btn = gr.Button("Generate")
+    clear_btn = gr.Button("Clear History")
+    export_btn = gr.Button("Export History")
+
+    output_text = gr.TextArea(label="Generated Text", interactive=False, lines=10)
+    history_state = gr.State(init_history())
+
+    generate_btn.click(
+        fn=generate_and_record,
+        inputs=[prompt_input, max_length, temperature, top_k, top_p, repetition_penalty, seed_input, num_seq, history_state],
+        outputs=[output_text, history_state]
+    )
+    clear_btn.click(lambda _: [], inputs=[history_state], outputs=[history_state, output_text], _js="() => {document.querySelectorAll('textarea')[1].value='';}")
+    export_btn.click(fn=export_history, inputs=[history_state], outputs=[])
+
+demo.queue().launch(server_name="0.0.0.0", server_port=7860)