qwen25_7b:
  use_mcore_models: true
  num_layers: 28
  hidden_size: 3584
  ffn_hidden_size: 18944
  num_attention_heads: 28
  rotary_base: 1000000
  max_position_embeddings: 32768
  make_vocab_size_divisible_by: 1
  padded_vocab_size: 152064
  untie_embeddings_and_output_weights: true
  add_qkv_bias: true
  disable_bias_linear: true
  group_query_attention: true
  num_query_groups: 4
  position_embedding_type: rope
  norm_epsilon: 1e-6
  normalization: RMSNorm
  swiglu: true
  attention_softmax_in_fp32: true