# okto_version: "1.1"
PROJECT "MixedDatasetModel"
DESCRIPTION "Training with multiple datasets using weighted mixing"
VERSION "1.0"
AUTHOR "OktoSeek"

DATASET {
    mix_datasets: [
        { path: "dataset/general.jsonl", weight: 50 },
        { path: "dataset/technical.jsonl", weight: 30 },
        { path: "dataset/creative.jsonl", weight: 20 }
    ]
    dataset_percent: 80
    sampling: "weighted"
    shuffle: true
    format: "jsonl"
    type: "generation"
    language: "en"
}

MODEL {
    base: "oktoseek/base-mini"
    architecture: "transformer"
    parameters: 120M
    context_window: 2048
    precision: "fp16"
}

TRAIN {
    epochs: 10
    batch_size: 32
    learning_rate: 0.0001
    optimizer: "adamw"
    scheduler: "cosine"
    device: "cuda"
    early_stopping: true
}

METRICS {
    loss
    perplexity
    bleu
    rouge_l
}

MONITOR {
    level: "basic"
    log_metrics: ["loss", "perplexity"]
    log_system: ["gpu_memory_used", "cpu_usage"]
    refresh_interval: 5s
    export_to: "runs/mixed-dataset/system.json"
    dashboard: false
}

EXPORT {
    format: ["okm"]
    path: "export/"
}