Modelo de lenguaje Transformer autoregresivo en español, entrenado completamente desde cero (sin pesos preentrenados, arquitectura propia con RoPE + atención causal + SwiGLU + RMSNorm). - Parámetros: ~18.80M - dmodel: 320 | capas: 6 | heads: 8 - contextlength: 768 - model.safetensors: pesos del modelo - config.json: hiperparámetros de arquitectura (para reconstruir MiloX1) - tokenizer.json: tokenizer BPE byte-level (formato nativo de la librería tokenizers)
Organization
BLACKDAQ
blackdaqlabs
Models in Library1
Datasets in Library0
Models on Hugging Face2
Followers1
Compare the published evidence
This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Reported licenses: mit. Open a record to review its specific license and conditions.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.