{
  "repo_id": "ZeyuLing/HFTrainer-TinyLlama-1.1B-Chat",
  "revision": "b23a22f572ca8d25e3d10411e90932826b9875f1",
  "artifact_type": "repository-owned HFTrainer bundle",
  "verification": {
    "export_reload_exact": true,
    "output_shape": [
      1,
      3,
      32000
    ],
    "finite": true,
    "real_pretrained_reference_logits_max_abs": 0.0,
    "real_tokenizer_reference_cases": [
      "English",
      "leading whitespace",
      "CRLF and tabs",
      "Chinese and emoji",
      "decomposed Unicode",
      "special tokens"
    ]
  },
  "distribution_verification": "Fresh complete snapshot downloaded; SHA256 matched every exported file",
  "gated": false,
  "training_entrypoint_check": {
    "optimizer_step_completed": true,
    "nonzero_trained_lora_B_tensors": 155,
    "training_data": "Four repeated synthetic instruction records; execution check only"
  },
  "inference_entrypoint_check": {
    "command": "python tools/infer.py --config configs/pretrained/infer_llama.py",
    "scope": "Full downloaded weights; CLI executed successfully. Not a quality benchmark."
  }
}