Open Peter-Devine opened 3 months ago
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・建築に関する質問を集めたcsvファイルの作成 ・Evaluation harnessにコミットするためにYAMLファイルの作成(現在ここ) - 現在詰まっている箇所は、YAMLファイルのdataset_pathで指定した(事前に作成しておいた)HFのurlからdataを見つけることができない状態です。そもそもHFの作り方に問題があると考えているため、そこを修正しようとしています。
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・建築に関する四択問題及び正誤問題について、Evaluation harnessにコミットするためにYAMLファイルの作成の完了 ・Google Colab上でGPUを用いて様々なaiモデルの正答率を評価(現在ここ)
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Google Colab上でGPUを用いて様々なaiモデルの正答率を評価(https://drive.google.com/drive/folders/17cyRbDM7zjLoKSwQ8Dcsj6FTuHU9jROU) ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットを利用するためにローカルで利用できるEmbedding Modelの作成(現在ここ)
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
・RAGとして用いるための性能を評価する用に、比較となるモデルのMRR、precision@のcsvの作成
これが達成されたら、モデルをより定性的に評価するためのデモを作成します。これはまた、佐藤さんが Gradio インターフェースの作成について詳しく学ぶ良い機会でもあります。Lightblue でよく行っています。 このデモは次のようになるべきです:
このインターフェースを使うことで、トレーニングされたモデルがどれほど有用か、また、ベクトルに直接基づいてトレーニングするこの方法がどれほど効果的かを完全に評価することができます。
(堀川さんは Gradio インターフェースの作成や GPT の呼び出しに経験がありますので、行き詰まったら彼に助けを求めてください。)
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
・RAGとして用いるための性能を評価する用に、比較となるモデルのMRR、precision@のcsvの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Cohere/wikipedia-2023-11-embed-multilingual-v3のデータセットをローカルで利用するためのEmbedding Modelの作成
・RAGとして用いるための性能を評価する用に、比較となるモデルのMRR、precision@のcsvの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・JSQuADを用いたMRRとPrecision@の評価 -> MRR: 0.696829, @1: 0.621937, @5: 0.786095, @10: 0.829586, @50: 0.899244, @100: 0.919362, @500: 0.953805 ・JSTSを用いたピアソンの相関係数 -> 0.791876 ・JSQuADを用いたMRRとPrecision@の評価 -> MRR: 0.601576, @1: 0.503759, @5: 0.685350, @10: 0.789474, @50: 0.906357, @100: 0.939850, @500: 0.981545
・Gradioインターフェースの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Gradioインターフェースを用いて、質問と数値を受け取り質問に適したWikiの情報を指定した数だけ出力するデモの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Gradioインターフェースを用いて、質問と数値を受け取り質問に適したWikiの情報を指定した数だけ出力するデモの作成
@Peter-Devine @shun1taniguchi 本日の勤務を終了させていただきます。 勤務内容は以下のとおりです。 ・Gradioインターフェースを用いて、質問と数値を受け取り質問に適したWikiの情報を指定した数だけ出力するデモの作成
建築評価データセット作成プロジェクト