Zeta Code Edit Prediction Training Dataset

Zeta is an open dataset used to train Zeta, an edit prediction model for Zed's predictive coding feature. The dataset is derived from Qwen2.5-Coder-7B and is designed to predict a developer's next code edit based on recent programming patterns and cursor position. It was published by zed-industries and last updated on February 27, 2025.

Use Cases

• Fine-tuning a model to predict the next code edit based on recent programming patterns.
• Training a model for intelligent code completion using cursor position as a feature.
• Conducting Direct Preference Optimization (DPO) using the provided dpo.jsonl data.
• Analyzing patterns in developer edits for supervised fine-tuning via train.jsonl.

Strengths

• Dataset is derived from the established Qwen2.5-Coder-7B model.
• Last updated on February 27, 2025, indicating recent maintenance.

Limitations

• Specific row counts, column names, and dataset size are unknown.
• The dataset's temporal and geographic coverage are not specified.

Provenance

Source

zed-industries

Collection Method

Derived from Qwen2.5-Coder-7B for training an edit prediction model.

Freshness

Last updated on 2025-02-27.