{
  "runtime": "tinybrain-causallm-llama-v1",
  "model_id": "HuggingFaceTB/SmolLM2-135M-Instruct",
  "bundle_name": "smollm2_135m_instruct_only_logits",
  "architecture": "smollm2",
  "sequence_length": 32,
  "vocab_size": 49152,
  "input_names": [
    "input_ids",
    "attention_mask"
  ],
  "output_name": "logits",
  "tokenizer_dir": "tokenizer",
  "requires_compiled_model": true,
  "chat_template_type": "chatml",
  "tokenizer_family": "llama",
  "default_system_prompt": "You are a helpful assistant.",
  "sampling_defaults": {
    "temperature": 0.3,
    "top_k": 12,
    "top_p": 0.75,
    "frequency_penalty": 0.55,
    "presence_penalty": 0.2
  },
  "runtime_profile": {
    "uses_chat_template": true,
    "default_system_prompt": "You are a helpful assistant.",
    "no_history_below_seq_len": 32,
    "short_history_below_seq_len": 64,
    "short_history_count": 2,
    "long_history_count": 6,
    "generation_min": 96,
    "generation_cap": 192,
    "generation_multiplier": 2,
    "block_special_tokens": true,
    "allow_eos": true,
    "sampling": {
      "temperature": 0.3,
      "top_k": 12,
      "top_p": 0.75,
      "frequency_penalty": 0.55,
      "presence_penalty": 0.2
    }
  }
}