【Gradio 徹底解説】AIモデルをWebで公開するためのPython最強ライブラリ

pip install gradio

import gradio as gr

!pip install gradio

!pip install ultralytics

#YOLOv10-Nをロード
from ultralytics import YOLO
import cv2
import numpy as np
from google.colab.patches import cv2_imshow

#YOLOの一連コードを関数化
def detect_objects_and_visualize(image: np.array):
    cv2.imwrite('data.jpg', image)
    model = YOLO("yolov10n.pt")
    results = model('data.jpg')
    for r in results:
      boxes = r.boxes
      for box in boxes:
        x1, y1, x2, y2 = box.xyxy[0]
        cv2.rectangle(image, (int(x1), int(y1)), (int(x2), int(y2)), (0, 255, 0), 2)
        cv2.putText(image, f"{r.names[int(box.cls.item())]}", (int(x1), int(y1)-10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0,255,0), 2)
        cv2.putText(image, f"{box.conf}", (int(x1)+300, int(y1)-10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0,255,0), 2)
        return image

import gradio as gr
yolo_ui=gr.Interface(fn=detect_objects_and_visualize, inputs="image", outputs="image")
yolo_ui.launch()

!pip install diffusers
!pip install accelerate

#画像生成モデルの用意
from diffusers import AutoPipelineForText2Image
import torch
pipe = AutoPipelineForText2Image.from_pretrained('dataautogpt3/OpenDalle', torch_dtype=torch.float16).to('cuda')

#画像生成の関数化
def gen_img(text: str):
  image = pipe(text).images[0]
  return image

import gradio as gr
with gr.Blocks() as jap_img_gen:
    input = gr.Textbox(label="英語のプロンプト")
    output = gr.Image(label="Output Box")
    button = gr.Button("画像を生成する")
    button.click(fn=gen_img, inputs=input, outputs=output)

jap_img_gen.launch()

!pip install -U g4f
!pip install curl_cffi --upgrade
!pip install git+https://github.com/openai/whisper.git

#WhisperとGPT4FREEの用意
import whisper
import numpy as np
from g4f.client import Client
import time
client = Client()
model = whisper.load_model("medium")

#文字起こし・要約の流れを関数化
def transcribe(file):
    moji = model.transcribe(file)["text"]
    prompt = "続く文章を全文日本語訳・要約して出力だけを返してください。:"+moji
    chat_completion = client.chat.completions.create(model="gpt-4o",messages=[{"role": "system", "content": "あなたは最強最高の翻訳家で入力を綺麗な日本語で返します。"},{"role": "user", "content": prompt}])
    nyan=chat_completion.choices[0].message.content or ""
    time.sleep(1)
    return [moji,nyan]

import gradio as gr
with gr.Blocks() as voice_script_summarise:
    input = gr.File(label="音声ファイルアップロード")
    output = [gr.Textbox("文字起こし"),gr.Textbox("要約")]
    button = gr.Button("要約する")
    button.click(fn=transcribe, inputs=input, outputs=output)

voice_script_summarise.launch()

!pip install openai langchain chromadb tiktoken -U langchain-community

#APIキー入力
import os
os.environ["OPENAI_API_KEY"] = "任意のAPIキー"

#モジュールインポート
from langchain.chains import RetrievalQA
from langchain.document_loaders import TextLoader
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.llms import OpenAI as LangChainOpenAI
from openai import OpenAI
from langchain.text_splitter import CharacterTextSplitter
from langchain.vectorstores import Chroma

%cd sample_data
#テキストをチャンクに分ける
loader = TextLoader("WEEL_Zenkiji.txt",encoding="utf-8")
documents = loader.load()
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
texts = text_splitter.split_documents(documents)

#ベクトル空間に埋め込む
embeddings = OpenAIEmbeddings(disallowed_special=())
docsearch = Chroma.from_documents(texts, embeddings)

chain_type_kwargs = {"verbose": True}
qa = RetrievalQA.from_chain_type(
    llm=LangChainOpenAI(),
    chain_type="stuff",
    retriever=docsearch.as_retriever(),
    verbose=True,
    chain_type_kwargs=chain_type_kwargs,
)

#RAGあり・RAGなしの回答出力の関数化
def rag(text):
    response = client.chat.completions.create(
        model="gpt-3.5-turbo-1106",
        messages=[
            {"role": "system", "content": "あなたは最強最高のアシスタントです。"},
            {"role": "user", "content": text},
        ],
        temperature=0.2,
    )
    no_rag = response.choices[0]
    yes_rag = qa.run(text)
    return [no_rag, yes_rag]

import gradio as gr
with gr.Blocks() as rag_chatbot:
    with gr.Column():
        with gr.Row():
            input = gr.Textbox(label="質問を入力")
            with gr.Column():
                output = [gr.Textbox(label="RAGなしの回答"),gr.Textbox(label="RAGありの回答")]
    button = gr.Button("回答を生成する")
    button.click(fn=rag, inputs=input, outputs=output)

rag_chatbot.launch()

ChatGPTの有料モデルGPT-4を無料で使えるツールについて教えて。