Tag: experiment-tracking
Structured State Management for Multi-Phase Experiments
This skill maintains a canonical, machine-readable JSON progress file to track the state of complex, multi-phase research experiments. It enforces strict lifecycle updates, ensuring dashboards and CLIs can reliably poll the experiment's cur…
Automated experiment benchmarking and metric extraction
Defines comparison metrics and extracts baseline values from notebook outputs to record them in a structured JSON log for downstream evaluation.
Manage and track ENCODE experiments locally
This skill enables local management of ENCODE experiments by tracking metadata, publications, and data provenance in a SQLite database. It supports experiment comparison, citation export in multiple formats, and the logging of derived files…
Research Analysis and Implementation Planning
Extracts actionable implementation details from research sources like PDFs, notebooks, or text files and records structured findings into an experiment log. It identifies method summaries, requirements, and compatibility analysis to facilit…
Machine-readable experiment progress tracking skill
This skill maintains a machine-readable JSON file to track the real-time status, phases, and activities of an ongoing experiment. It allows external dashboards and CLIs to monitor progress, failures, and current operations through a standar…
Research Method Implementation Skill
Automates the generation of Jupyter notebooks for implementing new research methods, ensuring parity with baseline experiments through consistent data loading and metric computation. It manages dependency updates and logs implementation det…
Automated Experiment Comparison and Reporting
Automates the comparison of baseline and new implementation results by calculating performance metrics and generating structured JSON reports. It handles the systematic updating of experiment logs, summaries, and comparison datasets.
Experiment Benchmarking and Metric Extraction
Defines comparison metrics and extracts baseline values from notebook outputs to update experiment logs for downstream evaluation.
Analyze Current Baseline Implementation
Extracts model architectures, preprocessing steps, and hyperparameters from existing Jupyter notebooks to create a structured experiment log. It also identifies dependencies and generates a requirements file from the notebook's import state…
Extracting Actionable Research Insights
This skill parses research materials such as PDFs, notebooks, or text ideas to extract method summaries, implementation requirements, and compatibility analysis. It records findings as structured JSON entries within an experiment log to fac…
Machine-readable experiment progress tracker
Maintains a machine-readable JSON file to track the real-time status, phases, and activity of autonomous experiments. It ensures consistent state reporting for dashboards and external monitoring tools.
Automated Experiment Benchmarking Skill
Defines comparison metrics and extracts baseline values from notebook outputs to record them in a structured JSON log for downstream evaluation.
Analyse Current Experiment Baseline
Analyses existing Jupyter notebooks to extract model architectures, preprocessing steps, and hyperparameters, recording the findings in a structured experiment log.