{
  "model_type": "qwen",
  "model_name": "ThatAI",
  "architectures": ["QwenForCausalLM"],
  "hidden_size": 1024,
  "intermediate_size": 4096,
  "num_attention_heads": 16,
  "num_hidden_layers": 24,
  "vocab_size": 65024,
  "max_position_embeddings": 2048,
  "rotary_dim": 64,
  "layer_norm_eps": 1e-6,
  "hidden_act": "gelu",
  "bos_token_id": 1,
  "eos_token_id": 2,
  "pad_token_id": 0,
  "tie_word_embeddings": true,
  "torch_dtype": "float32"
}