5M Terminal Language Model Suite
Collection
Complete 5.0M parameter terminal LM suite: Chinchilla & saturated models, raw 50M token corpus, pure CLI dataset, and ChatGPT playground Space. • 5 items • Updated
text stringclasses 144
values |
|---|
User: How do I discard all unstaged local changes in git? |
Assistant: Run `git checkout -- .` |
User: How do I run an interactive container with volume mount? |
Assistant: Run `docker run -it -v $(pwd):/app -p 8080:8080 myapp:latest /bin/bash` |
User: How do I list running processes matching python? |
Assistant: Run `ps aux | grep python` |
User: How do I count lines in a text file? |
Assistant: Run `wc -l dataset.txt` |
```session |
$ ps aux | grep python |
user 12345 98.2 4.1 452104 338102 ? Rsl 00:00 12:30 python train.py |
user 12390 0.0 0.0 6200 892 pts/0 S+ 00:15 0:00 grep python |
``` |
```session |
$ curl -I https://api.github.com |
HTTP/2 200 |
server: GitHub.com |
date: Sun, 02 Aug 2026 00:00:00 GMT |
content-type: application/json; charset=utf-8 |
``` |
User: How do I force remove a folder and all contents? |
Assistant: Run `rm -rf target_folder/` |
```session |
$ curl -I https://api.github.com |
HTTP/2 200 |
server: GitHub.com |
date: Sun, 02 Aug 2026 00:00:00 GMT |
content-type: application/json; charset=utf-8 |
``` |
User: How do I create a nested directory structure? |
Assistant: Run `mkdir -p path/to/nested/directory` |
User: How do I sort lines and remove duplicates? |
Assistant: Run `sort input.txt | uniq -c` |
```session |
$ df -h |
Filesystem Size Used Avail Use% Mounted on |
/dev/sda1 99G 32G 63G 34% / |
tmpfs 7.8G 0 7.8G 0% /dev/shm |
``` |
User: How do I find all python files in the current folder? |
Assistant: Run `find . -type f -name "*.py"` |
```session |
$ mkdir project && cd project |
$ git init |
Initialized empty Git repository in /home/user/project/.git/ |
$ touch main.py README.md |
$ git status |
On branch main |
Untracked files: |
README.md |
main.py |
``` |
User: How do I SSH into a remote server with custom port? |
Assistant: Run `ssh -p 2222 user@remote-host.com` |
```session |
$ git commit -m "fix: resolve permission issue" |
[main a1b2c3d] fix: resolve permission issue |
2 files changed, 14 insertions(+), 3 deletions(-) |
$ git push origin main |
To github.com:user/repo.git |
e4f5g6h..a1b2c3d main -> main |
``` |
```session |
$ cd .. |
$ pwd |
/home/user |
$ ls -la |
total 32 |
drwxr-xr-x 4 user user 4096 Aug 2 00:00 . |
drwxr-xr-x 8 user user 4096 Aug 2 00:00 .. |
-rw-r--r-- 1 user user 220 Aug 2 00:00 .bashrc |
``` |
```session |
$ chmod +x build.sh |
$ ./build.sh |
[INFO] Building release binary... |
[SUCCESS] Build completed in 2.4s. |
``` |
User: How do I sort lines and remove duplicates? |
Assistant: Run `sort input.txt | uniq -c` |
User: How do I recursively search for text in files? |
Assistant: Run `grep -rn "search_term" .` |
End of preview. Expand in Data Studio
💻 Pure Linux Terminal & CLI Commands Dataset
This repository contains 50,000 pure Linux CLI command scenarios, shell navigation interactions, and bash command Q&A pairs.
🧐 How This Dataset Came To Exist (Source Code Linked)
Small language models (such as 5.0 Million parameter models) face a severe data scarcity and noise problem: web text dumps (like Common Crawl) contain noisy HTML artifacts, broken code, and rambling prose that dilute small neural networks.
To train a tiny 5M parameter model to navigate Linux, run commands, and assist with terminal syntax with 100% accuracy and zero noise, we engineered a Synthesized Shell Compilation Engine:
👉 View Python Generator Code: terminal_dataset.py
👉 View Full Dataset Build Code: build_cli_only.py
Engine Mechanics:
- Structured Command Pair Generation: Programmatically synthesizes natural language user queries mapped to exact, un-corrupted Linux CLI commands (
cd ..,ls -la,mkdir -p,grep -rn,chmod +x,git commit,docker run,curl,df -h,free -h). - Interactive Shell Session Logs: Generates multi-turn terminal session logs (
$ cd .. \n $ pwd \n /home/user) to teach the model causal terminal prompt mechanics. - Chained Execution Logic: Generates multi-command shell pipelines (
$ mkdir project && cd project && git init).
📊 Dataset Schema & Files
| File | Format | Description |
|---|---|---|
terminal_dataset.py |
Python | Raw generator script defining templates and dataset compilation logic. |
build_cli_only.py |
Python | Execution script that built and uploaded this dataset to HF Hub. |
terminal_cli_commands.jsonl |
JSON Lines | Structured JSON entries containing question, command, and formatted_dialogue. |
terminal_cli_commands.txt.gz |
Compressed Text | Gzip-compressed raw text stream for direct PyTorch DataLoader streaming. |
terminal_cli_sample.txt |
Plain Text | 500-line sample preview for instant browser inspection. |
JSONL Sample Entry:
{
"id": 42,
"type": "qa_pair",
"question": "How do I navigate up one directory?",
"command": "cd ..",
"formatted_dialogue": "User: How do I navigate up one directory?\nAssistant: Run `cd ..`\n"
}
🔗 Related Models & Code
- 5M Chinchilla Model:
kipasyangin5/5m-terminal-lm-chinchilla - 5M Saturated Model:
kipasyangin5/5m-terminal-lm-saturated - Complete Training Code Folder:
kipasyangin5/5m-terminal-lm-chinchilla/tree/main/training_code - Live Playground Space:
kipasyangin5/5m-terminal-lm-chat - Official Collection:
kipasyangin5/5m-terminal-language-model-suite-6a6e83a13bb0effb11d9e53d
- Downloads last month
- 21