[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

【ChatStream】大容量のLLMの推論に必要なGPUサーバー構成

【ChatStream】大容量のLLMの推論に必要なGPUサーバー構成

大容量のLLM(大規模言語モデル)の推論に必要なGPUサーバー構成とChatStreamとの連携について詳しく動画で解説しています。 特に、Llama3-70Bモデルを例に、そのメモリ要件(140GBのGPUメモリ)と、この要件を満たすために必要なGPUサーバーおよびGPUクラスターの構成について、取り扱いやすい NVIDIA RTX A6000 GPUを使用した例について説明します。 また、モデル並列化技術(テンソル並列化、パイプライン並列化、データ並列化)とその推論エンジンの選択についても触れ、実際に複数のGPUサーバーを使ったクラスター構成がどのように推論処理を効率的に行うかを示します。 最後に、ChatStreamのLLMノードとは何か、そしてそれをどのようにスケーリングして大量の同時リクエストに対応するかについても解説します。こ の動画は、高性能を必要とするLLMのデプロイメントと運用に興味がある方に特に有益です。

By Join us, Michele on Qualiteg's adventure to innovation
ChatStream🄬でLlama-3-Elyza-JP-8B を動かす

ChatStream🄬でLlama-3-Elyza-JP-8B を動かす

こんにちは、本日は Llama-3-Elyza-JP-8B を使ってみました。 昨日 2024年6月26日に発表(https://prtimes.jp/main/html/rd/p/000000046.000047565.html)された Llama-3-Elyza-JP-8B は 70B 版では「GPT-4」を上回る性能の日本語LLMといわれています。 今回、当社でも Playground 環境に Llama-3-Elyza-JP-8B を搭載して試してみましたのでご紹介します。 70B(700億パラメータ)版は GPT-4 を上回るとのことですので、8B(80億パラメータ)版はGPT-3.5 と比較してみることにいたしました。 (性能比較は https://note.com/elyza/n/n360b6084fdbd の記事に詳しく書いてあります。) AWQ量子化版を使用してみる 今回は、A4000

By Qualiteg プロダクト開発部
AttributeError: module 'torch._dynamo' has no attribute 'mark_static_address' が発生したときの対処法

AttributeError: module 'torch._dynamo' has no attribute 'mark_static_address' が発生したときの対処法

以下のようなエラーが出た場合の対処法 File "/venv/Lib/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/venv/Lib/site-packages/transformers/generation/utils.py", line 1744, in generate model_kwargs["past_key_values"] = self._get_cache( ^^^^^^^^^^^^^^^^ File "/venv/Lib/site-packages/transformers/

By Qualiteg プロダクト開発部
[AI新規事業創出]Qualitegセレクション:アイディア創造編①Qualiteg式オンラインブレストの活用術

[AI新規事業創出]Qualitegセレクション:アイディア創造編①Qualiteg式オンラインブレストの活用術

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。 今日は私のお気に入りのブレスト方法である「Qualiteg式オンラインブレスト」の活用術についてお話ししたいと思います。 場所を変えて気分を変えても良いアイディアは生まれない!? よく、「金曜日は1日ブレストしよう!」という上司の掛け声とともに、いつもと違う雰囲気なら良いアイディアも出るかもしれないといってホテルの会議室などを予約されて1日缶詰でブレストしたが、期待する結果が出なかったとおっしゃるクライアントが非常に多いです。 ブレインストーミングは複数の参加者が自由にアイデアを出し合うことで、新しい発想や解決策を見つける手法です。 批判や評価を一時的に排除し、量を重視して多くのアイデアを集めることが目的です。1950年代に広告業界で生まれたこの手法は

By Join us, Michele on Qualiteg's adventure to innovation