Datasets:
|
Download README.md from OvertureMMXXI/HDL-RepoBench-Sample: direct link, hf CLI and curl.
- Browser
- Download file 897 Bytes
-
https://huggingface.co/datasets/OvertureMMXXI/HDL-RepoBench-Sample/resolve/main/README.md
- Command line
-
hf download hf://datasets/OvertureMMXXI/HDL-RepoBench-Sample/README.md
-
curl -L -o README.md https://huggingface.co/datasets/OvertureMMXXI/HDL-RepoBench-Sample/resolve/main/README.md
897 Bytes
metadata
license: other
task_categories:
- text-generation
language:
- code
tags:
- hardware
- verilog
- systemverilog
- vhdl
- chisel
- hls
- code-completion
size_categories:
- 1K<n<10K
HDL-RepoBench Sample
This is a compact, deterministic sample of MHRC-Bench for lightweight inspection and upload tests. It preserves the original language directories and JSONL file names while keeping the output below 2 GB.
Sampling policy:
- Languages: chisel, hls, systemverilog, vhdl.
- Validation and test splits: copied in full by default.
- Training split: repository-stratified deterministic sample.
- Raw repository archive: omitted because
original_repo/collected_repos.tar.gzis about 33 GB in the source dataset.
This generated sample contains 6,537 JSONL records and occupies
217.0 MiB on disk. See sample_manifest.json for exact
per-file counts, byte sizes, and generation arguments.