[ChatStream] Llama2 対応の ChatPrompt実装

[ChatStream] Llama2 対応の ChatPrompt実装

こんにちは! (株)Qualiteg プロダクト開発部 です!

本稿では、 ChatStream にLlama2 対応の ChatPrompt を同梱いたしましたのでご紹介いたします!

現在の ChatPrompt は以下とおりとなっております。旧バージョンのChatStreamをご利用の場合も、以下コードにて Llama2対応可能です。(もちろん最新バージョンのChatStreamには同梱済です)

from chatstream import AbstractChatPrompt
from chatstream.chat_prompt.role_type import RoleType

SYSTEM_PROMPT = """\
You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.\
"""


class ChatPromptMetaLlamaLlama2Chat(AbstractChatPrompt):
    """
    meta-llama/Llama-2-7b-chat

    Prompt Guide from
    https://huggingface.co/blog/llama2
    """

    def __init__(self):
        super().__init__()  # Call the initialization of the base class
        self.set_system(f"<s>[INST] <<SYS>>\n{SYSTEM_PROMPT}\n<</SYS>>\n\n")
        self.set_requester("")
        self.set_responder("")

    def get_stop_strs(self):
        if not self.chat_mode:
            return None
        return []

    def get_custom_skip_echo_len(self, skip_echo_len):
        # modify skip_echo_len when using llama2
        num_turn = self.get_turn()
        if num_turn >= 2:
            modified_skip_echo_len = skip_echo_len + 1 * self.get_turn()
            return modified_skip_echo_len
        return skip_echo_len

    def get_replacement_when_input(self):
        return None

    def get_replacement_when_output(self):  # replace when response_text gotten
        return None

    def create_prompt(self, opts={}):
        if self.chat_mode == False:
            return self.get_requester_last_msg()

        # Chat Mode == True の場合のプロンプトを構築する
        ret = self.system

        for chat_content in self.get_contents(opts):

            chat_content_role_type = chat_content.get_role_type()
            chat_content_message = chat_content.get_message()

            if chat_content_message:
                merged_message = ""
                if chat_content_role_type == RoleType.REQUESTER:
                    merged_message = f"{chat_content_message} [/INST] "
                elif chat_content_role_type == RoleType.RESPONDER:
                    merged_message = f"{chat_content_message} </s><s>[INST] "
                ret += merged_message
            else:
                pass

        return ret

    def build_initial_prompt(self, chat_prompt):
        # 初期プロンプトは実装しない
        pass


本実装については、1点だけトリッキーな実装を含んでいます。
というのも、ChatStreamはストリーミングチャットが基本ですので新たに生成された1トークン分を更新していくことでリアルタイムで流れるようなチャットを実現しています。そのため、出力が期待と1文字でもズレてしまうと、ストリーミング時に1文字かけたり、全体が1文字ずつズレてしまったりする現象が発生しやすいです。

Llama2の出力の場合、ターンごとに1文字ずつ出力に空白文字が発生する現象が確認されており、 なにがしかのSpecial Token や空トークン、またコード変換の影響を疑っていましたが根本原因の特定に至らず、現象をキャンセルするための実装を入れています。

現象としましては、

以下のプロンプトを入力プロンプトとしたとき、

<s>[INST] <<SYS>>
You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
<</SYS>>

Do you know the movie titanic [/INST] Hello! Yes, I'm familiar with the movie Titanic. It's a classic film directed by James Cameron, released in 1997, and starring Leonardo DiCaprio and Kate Winslet. The movie is based on the true story of the RMS Titanic, a British passenger liner that sank in the North Atlantic Ocean in 1912 after colliding with an iceberg. The film follows the story of Jack Dawson (played by DiCaprio) and Rose DeWitt Bukater (played by Winslet), who come from </s><s>[INST] Who is the director [/INST] The director of the movie "Titanic" is James Cameron. </s><s>[INST] Who is starred [/INST] ]  The movie "Titanic" features a star-studded cast, including:
* Leonardo DiCaprio as Jack Dawson
* Kate Winslet as Rose DeWitt Bukater
* Billy Zane as Cal Hockley
* Kathy Bates as Molly Brown
* Frances Fisher as Ruth DeWitt Bukater
* Bernard Hill as Captain Edward John Smith
* Jonathan Hyde as J. Bruce Ismay
* Eric Braeden as John Jacob Astor IV
* Gloria Stuart as Old Rose

These actors brought the characters from the movie to life and </s><s>[INST] 

出力は以下のようになります

[INST] <<SYS>>
You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
<</SYS>>

Do you know the movie titanic [/INST] Hello! Yes, I'm familiar with the movie Titanic. It's a classic film directed by James Cameron, released in 1997, and starring Leonardo DiCaprio and Kate Winslet. The movie is based on the true story of the RMS Titanic, a British passenger liner that sank in the North Atlantic Ocean in 1912 after colliding with an iceberg. The film follows the story of Jack Dawson (played by DiCaprio) and Rose DeWitt Bukater (played by Winslet), who come from  [INST] Who is the director [/INST] The director of the movie "Titanic" is James Cameron.  [INST] Who is starred [/INST]

Input:
Do you know the movie titanic [/INST] Hello! Yes, I'm familiar with the movie Titanic. It's a classic film directed by James Cameron, released in 1997, and starring Leonardo DiCaprio and Kate Winslet. The movie is based on the true story of the RMS Titanic, a British passenger liner that sank in the North Atlantic Ocean in 1912 after colliding with an iceberg. The film follows the story of Jack Dawson (played by DiCaprio) and Rose DeWitt Bukater (played by Winslet), who come from [INST] Who is the director [/INST] The director of the movie "Titanic" is James Cameron. [INST] Who is starred [/INST] 

つまり入力プロンプトが [INST] つまり 半角スペース1つ[INST] でも出力で、 [INST]半角スペース2つ + [INST] となって戻ってくるため、新規生成文章の切り出しで1レスポンスにたいして1文字ズレの発生が観測されます。

それならばと、入力プロンプトを 半角スペース2つ[INST] で入力したが、その場合は出力生成文は 半角スペース3つ[INST]
となり、ズレが解消されないようです。

このとき、通常のテキスト処理で trim していく方法もありますが、当ChatPromptではなるべくもとの入力と出力を変形しないことをポリシーとして、

    def get_custom_skip_echo_len(self, skip_echo_len):
        # modify skip_echo_len when using llama2
        num_turn = self.get_turn()
        if num_turn >= 2:
            modified_skip_echo_len = skip_echo_len + 1 * self.get_turn()
            return modified_skip_echo_len
        return skip_echo_len

のように、 skip_echo_len をターンにあわせて調整することでピッタリの切り出しに対応しました。約1000通りの入出力において、問題なく切り出せているため、本手法を workaround として採用しています。ちょっとした見落としの可能性も高そうなので根治のために研究部側で引き続き原因調査しています。

Llama2 をベースモデルとして採用しているほかモデルも同様の現象が発生していますが、本手法にて解決できました。

Read more

Startup JAPAN 2025 に出展いたしました

Startup JAPAN 2025 に出展いたしました

こんにちは! 2025年5月8日(木)-5月9日(金)に東京ビッグサイトで開催された Startup JAPAN 2025 に出展いたしましたので、簡単にレポートいたします😊 開催概要 出展概要 今回は当社が開発するアバター動画生成AI「MotionVox™」を中心に出展させていただきました! 展示会について簡単にふりかえってみたいとおもいます 当社ブース 当社ブースはこんなかんじです。 今回は、ブースというか、このイーゼルのような雰囲気の木枠にポスターをくっつけるというスタイルでの展示方式でした。 こういう方式ははじめてなので斬新でした。おそらくこの方式で相当なコストダウンを図れておりスタートアップにはうれしいですね。セットアップも数分で終わりました。 会場 今回の会場はビッグサイトの南ホールでした。南ホールは、ビッグサイト入口からすぐそこなので駅から会場までたいして歩かず、疲れずに行くことができアクセスがとても良いです。 ホールは広めですが、ところせましと400社の出展会社がひしめきあっておりスタートアップの勢いのある会場となっており

By Qualiteg ビジネス開発本部 | マーケティング部
GPUサービスで「Segmentation Fault 」に出会ったら~分析から解決までの実践アプローチ~

GPUサービスで「Segmentation Fault 」に出会ったら~分析から解決までの実践アプローチ~

こんにちは! 今日は仮想環境+GPUなサービスにおける「Segmentation Fault」について、分析と対処法について書いてみたいと思います。 Segmentation Faultの本質と特徴 Segmentation Faultは、プログラムが保護されたメモリ領域にアクセスしようとした際にOSが発生させる例外です。 今回は複数のGPUサービス(つまりGPUを使うプロセス)が動作していて、そのうちの1つを再起動したときに発生しました。 毎回発生するわけではありません。むしろ数百回の起動に1回程度ですが、1回でも発生すると絶望的な結果につながります。というのも、1つのGPUサービスの停止が SPOF となってサービス全体に影響が発生します。かつ、1回でも「Segmentation Fault」が発生してしまうと、その原因となったプロセスが二度と起動しなくなる、というやっかいな現象でした。 このように「普段は正常に動作しているのに突然動かなくなる」というのがデバッグを非常に難しくします。 とくにGPU+仮想化の組み合わせで従来のC++アプリよりも発生確率がぐっとあがる印象

By Qualiteg プロダクト開発部
シェルスクリプトからcondaコマンドを活用したいとき

シェルスクリプトからcondaコマンドを活用したいとき

こんにちは! 今日はみんな大好きcondaコマンドについてです。 condaコマンドで仮想環境に入って、何らかの処理をして、戻ってくる ようなシェルスクリプト、バッチタスクをやるときのTipsです。 AI開発において、Anacondaとその中核であるcondaパッケージマネージャーはとっても重宝します。 しかし、シェルスクリプトから自動的にcondaを利用しようとすると、意外なハードルがあります。 本記事では、シェルスクリプトからcondaコマンドを正しく呼び出す方法について解説します。 condaと非対話モードの課題 AnacondaがインストールされているLinux環境において、condaコマンドは通常、.bashrcや.bash_profileなどの設定ファイルによって初期化されます。 なんとなくシェルをつかっていると、このcondaコマンドの初期化を忘れてしまいますが、これらの設定は多くの場合シェルの「対話モード」でのみ有効になるように設計されています。 ゆえにシェルスクリプトのような非対話モードでは、condaコマンドが正しく機能してくれません 例えば、.b

By Qualiteg プロダクト開発部
Node.jsで大容量ファイルを扱う:AIモデルのような大きなデータ保存はストリーム処理使いましょう

Node.jsで大容量ファイルを扱う:AIモデルのような大きなデータ保存はストリーム処理使いましょう

こんにちは!今日はAIシステムのフロントサーバーとしてもよく使用するNode.jsについてのお話です。 AIモデルの普及に伴い、大容量のデータファイルを扱う機会が急増しています。LLMなどのモデルファイルやトレーニングデータセットは数GB、場合によっては数十、数百GBにも達することがあります。 一方、Node.jsはWebアプリケーションのフロントサーバーとして広く採用されており、データマネジメントやPythonで書かれたAIバックエンドとの橋渡し役としてもかなりお役立ちな存在です。 本記事では、Node.js v20LTSで5GB程度のファイルを処理しようとして遭遇した問題と、その解決方法について解説します。 Node.jsのバッファサイズ制限の変遷 Node.jsのバッファサイズ制限は、バージョンによって大きく変化してきました Node.jsバージョン サポート終了日 バッファサイズ上限 備考 Node.js 0.12.x 2016年12月31日 ~1GB 初期のバッファサイズ制限(smalloc.kMaxLength使用) Node.js 4.

By Qualiteg プロダクト開発部