Datasetpublic

tomjnet/SeqAtom-Coder-Dataset

SeqAtom Coder Dataset SeqAtom Coder Dataset is the training dataset for: tomjnet/SeqAtom-Coder-1.5B-Instruct The dataset is designed to teach the SeqAtom model the Seq programming language, including syntax, execution flow, AI instructions, compiler behavior, and code translation. Base Model The initial SeqAtom model is based on: Qwen/Qwen2.5-Coder-1.5B-Instruct Dataset Format The dataset uses conversational instruction examples. Each record… See the full description on the dataset page: https://huggingface.co/datasets/tomjnet/SeqAtom-Coder-Dataset.

sourceHugging Faceapache-2.0updated 5d agoView on Hugging Face

0likes45downloads

Dataset Card

SeqAtom Coder Dataset

SeqAtom Coder Dataset is the training dataset for:

tomjnet/SeqAtom-Coder-1.5B-Instruct

The dataset is designed to teach the SeqAtom model the Seq programming language, including syntax, execution flow, AI instructions, compiler behavior, and code translation.

Base Model

The initial SeqAtom model is based on:

Qwen/Qwen2.5-Coder-1.5B-Instruct

Dataset Format

The dataset uses conversational instruction examples.

Each record contains:

json

{
  "messages": [
    {
      "role": "system",
      "content": "You are SeqAtom, an expert in the Seq programming language."
    },
    {
      "role": "user",
      "content": "Create a Seq program that prints Hello World."
    },
    {
      "role": "assistant",
      "content": "name = \"hello\"\n\nstep step1():\n    print(\"Hello World\")"
    }
  ]
}