On Data Engineering for Scaling LLM Terminal Capabilities
Paper • 2602.21193 • Published • 103
conversations list | agent string | model string | model_provider string | date string | task string | episode string | run_id string | trial_name string | enable_thinking bool | source string | difficulty string |
|---|---|---|---|---|---|---|---|---|---|---|---|
[
{
"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 | 2025-12-30T15:09:23.418048 | file_operations_task_3765 | episode-7 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_3765__WX6UcHS | true | file_operations | 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 | 2025-12-30T11:20:49.392913 | file_operations_task_4118 | episode-5 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_4118__YspxPn3 | true | file_operations | 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 | 2025-12-30T18:07:11.128096 | file_operations_task_1314 | episode-6 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_1314__y8NEnjo | true | file_operations | 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 | 2025-12-30T09:09:15.828573 | file_operations_task_0834 | episode-8 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_0834__JbitxYi | true | file_operations | 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 | 2025-12-30T11:54:59.711375 | file_operations_task_2622 | episode-5 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_2622__6Aq9Wgs | true | file_operations | 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 | 2025-12-30T15:17:24.982713 | file_operations_task_3123 | episode-7 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_3123__ZwW6Mcx | true | file_operations | 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 | 2025-12-30T15:35:28.822694 | file_operations_task_0242 | episode-6 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_0242__6pZzvzF | true | file_operations | 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 | 2025-12-30T19:05:11.381315 | file_operations_task_0619 | episode-7 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_0619__UTeMvBt | true | file_operations | 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 | 2025-12-30T10:47:46.830124 | file_operations_task_4620 | episode-7 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_4620__XNnPAT5 | true | file_operations | 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 | 2025-12-30T19:05:39.526840 | file_operations_task_3651 | episode-4 | 8e1fee79-6a63-40b9-84e4-044bcf65861a | file_operations_task_3651__jS45QTk | true | file_operations | easy |
Per-source partition of nvidia/Nemotron-Terminal-Corpus,
filtered to source == "file_operations". 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 ("file_operations" 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.