README.md

metadata

license: apache-2.0
task_categories:
  - text-generation
language:
  - en
tags:
  - agent
  - coding
  - terminal
  - shell
pretty_name: WebTerminal

Terminal/CLI Web Text

A filtered extract of terminal and command-line content from two large web-text corpora.

Sources

• DCLM (Zyphra/dclm-dedup)
• FineWeb (Salesforce/fineweb_deduplicated)

How it was built

1. Fast filter: skip any document that doesn't contain obvious CLI indicators ($, sudo, pip install, ```bash, root@, etc.)
2. Score: remaining docs are scored (0-34) across five signals, each with a per-match point value and a cap:

Signal	What it detects	Points	Max
Prompt patterns	Shell prompts like $ cmd, user@host:~$, >>>, root@, PS C:\	2 per match	10
CLI commands	Known commands: sudo, apt-get, pip install, git clone, docker run, curl, ssh, gcc, etc. (30+ patterns)	1 per unique match	8
stdout patterns	Output indicators: "successfully installed", "cloning into", drwx (ls output), "packets transmitted", "traceback", version strings	2 per match	6
Code blocks	Terminal-flavored code blocks: ```bash, ```shell, <pre><code>, terminal/console div classes	2 per match	6
Indented blocks	3+ consecutive lines indented 4+ spaces (code/output blocks)	1 per match	4

Documents scoring ≥5 are kept.

3. Dedup: exact dedup across both datasets using xxhash64 on full text.

Stats

	Chunks	Size	Rows
DCLM	13,144	~229 GB	~18.8M
FineWeb	8,800	~669 GB	~47.5M

Combined ~66M rows before dedup.

Schema

text: string
term_score: int32

Use case

Pre-training or fine-tuning on terminal/CLI-flavored web text, useful for models that need to understand shell commands, error output, package management, and general command-line workflows.