On Data Engineering for Scaling LLM Terminal Capabilities
Paper • 2602.21193 • Published • 102
conversations listlengths 2 90 | agent stringclasses 1
value | model stringclasses 1
value | model_provider stringclasses 1
value | date stringdate 2025-12-30 08:04:16 2026-01-02 04:14:16 | task stringlengths 23 23 | episode stringclasses 26
values | run_id stringclasses 2
values | trial_name stringlengths 32 32 | enable_thinking bool 1
class | source stringclasses 1
value | difficulty stringclasses 2
values |
|---|---|---|---|---|---|---|---|---|---|---|---|
[
{
"content": "You are an AI assistant tasked with solving command-line tasks in a Linux environment. You will be given a task description and the output from previously executed commands. Your goal is to solve the task by providing batches of shell commands.\n\nFormat your response as JSON with the following st... | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T15:39:59.692761 | data_querying_task_3087 | episode-7 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_3087__Zuj2ccn | true | data_querying | easy |
[
{
"content": "You are an AI assistant tasked with solving command-line tasks in a Linux environment. You will be given a task description and the output from previously executed commands. Your goal is to solve the task by providing batches of shell commands.\n\nFormat your response as JSON with the following st... | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T17:02:53.846420 | data_querying_task_2395 | episode-4 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_2395__YdbHKAG | true | data_querying | easy |
[
{
"content": "You are an AI assistant tasked with solving command-line tasks in a Linux environment. You will be given a task description and the output from previously executed commands. Your goal is to solve the task by providing batches of shell commands.\n\nFormat your response as JSON with the following st... | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T20:33:41.813547 | data_querying_task_1768 | episode-16 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_1768__TqqDKbn | true | data_querying | easy |
[
{
"content": "You are an AI assistant tasked with solving command-line tasks in a Linux environment. You will be given a task description and the output from previously executed commands. Your goal is to solve the task by providing batches of shell commands.\n\nFormat your response as JSON with the following st... | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T11:57:26.371232 | data_querying_task_1814 | episode-6 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_1814__UnGn6Ux | true | data_querying | easy |
[
{
"content": "You are an AI assistant tasked with solving command-line tasks in a Linux environment. You will be given a task description and the output from previously executed commands. Your goal is to solve the task by providing batches of shell commands.\n\nFormat your response as JSON with the following st... | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T20:08:28.833074 | data_querying_task_2032 | episode-4 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_2032__2m8KZij | true | data_querying | easy |
[{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T13:27:01.801490 | data_querying_task_0552 | episode-8 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_0552__wDMn7pS | true | data_querying | easy |
[{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T20:13:23.886021 | data_querying_task_2003 | episode-16 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_2003__FGy6evJ | true | data_querying | easy |
[{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T11:54:47.487116 | data_querying_task_3270 | episode-11 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_3270__dM9V2SY | true | data_querying | easy |
[{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T18:27:37.429374 | data_querying_task_1609 | episode-9 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_1609__6yXYAxf | true | data_querying | easy |
[{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | terminus-2 | deepseek-ai/DeepSeek-V3.2 | hosted_vllm | 2025-12-30T12:47:17.497149 | data_querying_task_3529 | episode-17 | 9b302b8c-5d8a-42ec-8490-3feea61f7c64 | data_querying_task_3529__Cuc4LMU | true | data_querying | easy |
Per-source partition of nvidia/Nemotron-Terminal-Corpus,
filtered to source == "data_querying". The difficulty column preserves the original
easy / medium / mixed split (na for the dataset_adapters/* files, which
did not carry a difficulty label).
Partitioning scheme:
dataset_adapters/{code,math,swe}.parquetdebugging, security, …) — rows from
synthetic_tasks/skill_based/{easy,medium,mixed}/{skill}/data_filtered.parquetSame as the source dataset (conversations, agent, model, model_provider,
date, task, episode, run_id, trial_name, enable_thinking) plus:
source — the partition key ("data_querying" throughout this repo)difficulty — easy / medium / mixed / naoriginal_source — only present in adapters_code; preserves the original
source column value (OpenCodeReasoning or synthetic) from the upstream file.@misc{pi2026dataengineeringscalingllm,
title={On Data Engineering for Scaling LLM Terminal Capabilities},
author={Renjie Pi and Grace Lam and Mohammad Shoeybi and Pooya Jannaty and Bryan Catanzaro and Wei Ping},
year={2026},
eprint={2602.21193},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2602.21193},
}
Original dataset license: CC-BY-4.0.