File size: 2,070 Bytes
3557791 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
MODEL_NAME = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
# Load tokenizer and model
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
if tokenizer.pad_token is None:
tokenizer.pad_token = tokenizer.eos_token
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME,
torch_dtype=torch.bfloat16 if torch.cuda.is_available() else torch.float32
)
if torch.cuda.is_available():
model.to("cuda")
model.eval()
def generate_text(prompt, max_new_tokens=100, temperature=0.7, top_k=50):
if not prompt:
return "Please enter a prompt."
messages = [{"role": "user", "content": prompt}]
encoded = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt",
padding=True,
return_attention_mask=True,
)
input_ids = encoded["input_ids"]
attention_mask = encoded["attention_mask"]
if torch.cuda.is_available():
input_ids = input_ids.to("cuda")
attention_mask = attention_mask.to("cuda")
output_ids = model.generate(
input_ids,
attention_mask=attention_mask,
max_new_tokens=max_new_tokens,
do_sample=True,
temperature=temperature,
top_k=top_k,
pad_token_id=tokenizer.eos_token_id
)
response = tokenizer.decode(output_ids[0][input_ids.shape[-1]:], skip_special_tokens=True)
return response
# Gradio interface
demo = gr.Interface(
fn=generate_text,
inputs=[
gr.Textbox(label="Prompt"),
gr.Slider(minimum=10, maximum=500, value=100, label="Max New Tokens"),
gr.Slider(minimum=0.1, maximum=1.0, value=0.7, step=0.05, label="Temperature"),
gr.Slider(minimum=0, maximum=100, value=50, step=1, label="Top K")
],
outputs=gr.Textbox(label="Generated Text"),
title="TinyLlama Gradio API",
description="Use this via UI or API via `/run/predict`"
)
if __name__ == "__main__":
demo.launch(server_name="0.0.0.0", server_port=7860)
|