Spaces:
Runtime error
Runtime error
File size: 1,327 Bytes
679c32f 7c0fc12 679c32f 7c0fc12 679c32f |
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 |
import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel
import torch
# ๐น Load tokenizer and base model
base_model_id = "deepseek-ai/deepseek-coder-1.3b-base"
lora_model_id = "brijmansuriya/deepseek-lora" # โ
Your LoRA fine-tuned model repo
tokenizer = AutoTokenizer.from_pretrained(lora_model_id, trust_remote_code=True)
base_model = AutoModelForCausalLM.from_pretrained(
base_model_id,
device_map="auto",
torch_dtype=torch.float16,
trust_remote_code=True
)
# ๐น Load LoRA adapter
model = PeftModel.from_pretrained(base_model, lora_model_id)
# ๐น Define the function
def generate_code(prompt):
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(
**inputs,
max_new_tokens=200,
temperature=0.7,
do_sample=True,
top_p=0.95,
)
return tokenizer.decode(outputs[0], skip_special_tokens=True)
# ๐น Gradio UI
demo = gr.Interface(
fn=generate_code,
inputs=gr.Textbox(label="Enter your coding prompt"),
outputs=gr.Textbox(label="Generated Code"),
title="๐ค DeepSeek Code Generator (LoRA)",
description="This app uses DeepSeek-Coder with Brijbhai's fine-tuned LoRA model to generate code from natural language prompts."
)
demo.launch()
|