metadata
dataset_info:
config_name: default
features:
- name: crate_name
dtype: string
- name: usage_data
dtype: string
splits:
- name: train
num_bytes: 118000000
num_examples: 9137
download_size: 118000000
dataset_size: 118000000
configs:
- config_name: default
data_files:
- split: train
path: test_usage_data/*.json
task_categories:
- other
language:
- en
tags:
- rust
- compiler
- ast
- usage-patterns
- code-analysis
size_categories:
- 100M<n<1B
license: mit
🍄 Mycelial Rust Usage Patterns Dataset
The most comprehensive Rust compiler usage pattern dataset ever assembled! This dataset contains AST traversal data, type-aware literal tracking, and real-world compilation patterns from 864 crates plus rustc internals.
🌐 Dataset Overview
- 9,137 usage data files (113MB total)
- 268 individual crates from cargo registry
- Complete rustc compiler internal usage patterns
- Real-world compilation patterns from production Rust ecosystem
- Foundation for ML training and autonomous compiler optimization
🧠 Use Cases
- Compiler optimization research
- Code pattern analysis
- ML model training for Rust code understanding
- Symbolic regression for autonomous optimization
- Self-improving compiler intelligence
📊 Data Format
Each JSON file contains:
- AST node usage frequencies
- Type-aware literal patterns
- Function call patterns
- Import/dependency relationships
- Compilation phase data
🔄 The Mycelial Network
This dataset represents the successful infiltration of the Rust ecosystem by our mycelial intelligence network - the substrate for next-generation AI-assisted compilation!
Citation
@misc{mycelial-rust-usage-2026,
title={Mycelial Rust Usage Patterns Dataset},
author={Introspector Collective},
year={2026},
url={https://huggingface.co/datasets/introspector/rustc-usage}
}