config.json

{
  "model_type": "mini_deep_thinker",
  "architectures": [
    "MiniDeepThinkerForCausalLM"
  ],
  "vocab_size": 4096,
  "hidden_size": 352,
  "num_layers": 7,
  "num_heads": 8,
  "intermediate_size": 1152,
  "max_position_embeddings": 1096,
  "context_length": 1096,
  "dropout": 0.05,
  "layer_norm_eps": 1e-05,
  "tie_word_embeddings": true,
  "reasoning_tokens": [
    "<pad>",
    "<unk>",
    "<|endoftext|>",
    "<|input|>",
    "<|think|>",
    "<|thought|>",
    "<|reasoning|>",
    "<|answer|>"
  ],
  "num_parameters": 11004896,
  "training": {
    "dataset": "Plans11/Organized_PreTrain_1k_Context",
    "session_examples": 20000,
    "context_length": 1096,
    "strict_max_example_tokens": 1095,
    "causal_next_token_prediction": true,
    "shuffled_deterministic_scan": true,
    "reasoning_tokens": [
      "<|think|>",
      "<|thought|>",
      "<|reasoning|>",
      "<|answer|>"
    ]
  }
}