[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

【解説】Tekken トークナイザーとは何か? 〜 Mistral が採用する新世代トークナイザーの特徴

【解説】Tekken トークナイザーとは何か? 〜 Mistral が採用する新世代トークナイザーの特徴

こんにちは! 本日は、Tekkenについて解説いたします! 皆さま Tekken と聞いて何を思い浮かべますか? 格ゲーの鉄拳でしょうか? 私は、昔プレイした Age of Empires に登場する鉄剣戦士を思い浮かべました🤗 ちょっと古いかもしれませんが、名作です! さてつかみはこのくらいにして、、 LLMはご存じのとおり驚異的なスピードで進化しています。そんな中でひそかに注目されているのが、トークナイザーの改善です。 たとえば、Meta の Llama 系モデルのトークナイザーは Sentence Piece から BPE系へ進化するなど、LLM業界では従来よりも高効率なトークナイズ(テキスト分割)の方法を導入し始めています。 そして Mistral AI もまた、新たに「Tekken トークナイザー」という仕組みを採用し、大規模言語モデルの性能を底上げしています。 本記事では、Tekken トークナイザーの登場背景や技術的特徴、他のトークナイザーとの違い、さらには Mistral との関係などをわかりやすく解説していきます。 1. Tekken トーク

By Qualiteg プロダクト開発部
[AI新規事業創出]Qualitegオリジナル、アイディア評価、事業アイディア選定方法

[AI新規事業創出]Qualitegオリジナル、アイディア評価、事業アイディア選定方法

Qualiteg blogを訪問してくださった皆様、こんにちは。Micheleです。AIを活用した新規事業やマーケティングを手がけている私には、クライアントからよく寄せられる質問があります。AIを用いた事業展開を検討されている方々が共通して直面するであろう課題に対して、このブログを通じて私なりの解答をご提供したいと思います。 はじめに AI技術の急速な発展は、スタートアップから大企業まで、あらゆるビジネスに新たな可能性をもたらしています。クライアントとの会話の中でも、AIを活用した革新的な事業アイディアに関する相談が増えています。 しかし、多くの企業が「素晴らしいアイディアを思いついた!」と興奮しながらも、そのアイディアを具体化し、成功に導くための方法論に悩んでいるのも事実です。特にAIを用いた事業展開においては、従来のビジネスモデルとは異なる視点が必要となるため、その難しさはさらに増します。 本記事では、Qualitegオリジナルのアイディア評価、事業アイディア選定方法について解説します。特に、AIを用いた事業展開を検討されている方々が共通して直面するであろう課題に対して、

By Join us, Michele on Qualiteg's adventure to innovation
日本語対応!Mistral Small v3 解説

日本語対応!Mistral Small v3 解説

こんにちは! Mistral AIは2025年1月30日、新しい言語モデル「Mistral Small v3」を発表しました。このモデルは、24Bという比較的小規模なパラメータ数ながら、70B以上の大規模モデルに匹敵する性能を実現しています。また日本語対応も謳われており期待の高い小型モデルです! https://huggingface.co/mistralai/Mistral-Small-24B-Instruct-2501 動画 こちら本ブログの解説動画もご覧いただけます😊 きわだってるのは、レイテンシー最適化 Mistral Small 3のめだった特徴は、その処理性能とレイテンシーの絶妙なバランスではないでしょうか。 公開されている以下の性能評価のグラフによると、トークンあたり約11ミリ秒という業界最速レベルのレイテンシーを達成しています。これは、Qwen-2.5 32Bの約15ミリ秒やGemma-2 27Bの約14ミリ秒と比較して、明確な優位性を示しています。さらに注目すべきは、GPT-4o Miniと比較しても、より低いレイテンシーで同等以上の性能を実現し

By Qualiteg プロダクト開発部
[vLLM] To use CUDA with multiprocessing, you must use the 'spawn' start method の対処法

[vLLM] To use CUDA with multiprocessing, you must use the 'spawn' start method の対処法

WSLで vLLM を使用するとき、 tensor parallel を使って複数枚のGPUで1つのLLMをサーブしようとしたとき以下のようなエラーが発生しがちです RuntimeError: Cannot re-initialize CUDA in forked subprocess. To use CUDA with multiprocessing, you must use the 'spawn' start method 遭遇するシーンとしてはvLLMの起動オプションに以下のようにテンソル並列化オプションを指定したときです。 --tensor-parallel-size 2 つまり、マルチプロセッシングでCUDA使うときは、 "fork"じゃなくて"spawn" 使ってね、というエラーです。 これを vLLM に教えるために、以下の2行目のように環境変数を設定してあげるとvLLMが "spawn" を使ってくれるようになります。 export

By Qualiteg プロダクト開発部