atft-pipeline

Official

Automate ATFT data ETL, keep training flowing.

Authorwer-inc
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Manually managing complex financial data ingestion, feature engineering, and caching for the ATFT-GAT-FAN model is prone to errors, API quota issues, and training stalls. This Skill automates the entire data pipeline, ensuring data quality and availability.

Core Features & Use Cases

  • Automated Dataset Provisioning: Generate fresh or historical Parquet datasets with GPU-accelerated ETL, ensuring data readiness for model training.
  • Deterministic Feature Graph Management: Maintain hundreds of engineered factors with high determinism, crucial for consistent model accuracy.
  • Use Case: Before a new training run, use this Skill to automatically refresh the 5-year dataset, verify cache integrity, and ensure all J-Quants API quotas are respected, preventing any data-related training interruptions.

Quick Start

Example: Refresh dataset in background and monitor

make dataset-bg tail -f _logs/dataset/*.log

Dependency Matrix

Required Modules

polarspyarrowcudfjquants-api-client

Components

scripts

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: atft-pipeline
Download link: https://github.com/wer-inc/gogooku3/archive/main.zip#atft-pipeline

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.