【Streamlit】gemma-3-1b-ptモデルを使ったチャットアプリ

pip install streamlit transformers torch

import streamlit as st
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

# モデル名
model_name = "google/gemma-3-1b-pt"

# トークナイザーとモデルの読み込み
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

# Streamlitアプリの設定
st.title("Gemma-3-1b-pt チャットアプリ")
st.write("ユーザー入力に対してGemma-3-1b-ptモデルが応答します。")

# チャット履歴を保存するためのセッション状態を初期化
if "chat_history" not in st.session_state:
    st.session_state.chat_history = []

# ユーザー入力の取得
user_input = st.text_input("あなた: ", "")

# ユーザーが入力を送信した場合
if user_input:
    # モデルへの入力を準備
    prompt = user_input
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

    # モデルによるテキスト生成
    with torch.no_grad():
        outputs = model.generate(
            inputs.input_ids,
            max_new_tokens=50,
            pad_token_id=tokenizer.eos_token_id
        )

    # モデルの応答をデコード
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)

    # チャット履歴にユーザー入力とモデルの応答を追加
    st.session_state.chat_history.append((user_input, response))

# チャット履歴の表示
for i, (user_msg, bot_msg) in enumerate(st.session_state.chat_history):
    st.write(f"**あなた:** {user_msg}")
    st.write(f"**Gemma-3-1b-pt:** {bot_msg}")

model_name = "google/gemma-3-1b-pt"

# トークナイザーとモデルをHugging Faceからロード
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

st.title("Gemma-3-1b-pt チャットアプリ")
st.write("ユーザー入力に対してGemma-3-1b-ptモデルが応答します。")

if "chat_history" not in st.session_state:
    st.session_state.chat_history = []

inputs = tokenizer(user_input, return_tensors="pt").to(model.device)

with torch.no_grad():
    outputs = model.generate(
        inputs.input_ids,
        max_new_tokens=50,
        pad_token_id=tokenizer.eos_token_id
    )

response = tokenizer.decode(outputs[0], skip_special_tokens=True)

for i, (user_msg, bot_msg) in enumerate(st.session_state.chat_history):
    st.write(f"**あなた:** {user_msg}")
    st.write(f"**Gemma-3-1b-pt:** {bot_msg}")

OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-3-1b-pt is not the path to a directory containing a file named config.json.

huggingface-cli login

ValueError: The checkpoint you are trying to load has model type gemma3_text but Transformers does not recognize this architecture. This could be because of an issue with the checkpoint, or because your version of Transformers is out of date. You can update Transformers with the command pip install --upgrade transformers. If this does not work, and the checkpoint is very new, then there may not be a release version that supports this model yet. In this case, you can get the most up-to-date code by installing Transformers from source with the command pip install git+https://github.com/huggingface/transformers.git

pip uninstall transformers
pip install git+https://github.com/huggingface/transformers.git