{
  "_name_or_path": "seed_encoder_3_decoder_layers",
  "activation_dropout": 0.0,
  "activation_fn": "gelu",
  "adaptive_input": false,
  "adaptive_softmax_cutoff": null,
  "adaptive_softmax_dropout": 0,
  "architectures": [
    "SEEDEncoderDot_NLL_LN"
  ],
  "attention_dropout": 0.1,
  "cross_self_attention": false,
  "decoder_atten_window": 8,
  "decoder_attention_heads": 12,
  "decoder_embed_dim": 768,
  "decoder_embed_path": null,
  "decoder_ffn_embed_dim": 3072,
  "decoder_input_dim": 768,
  "decoder_layerdrop": 0,
  "decoder_layers": 3,
  "decoder_layers_to_keep": null,
  "decoder_learned_pos": true,
  "decoder_normalize_before": true,
  "decoder_output_dim": 768,
  "dropout": 0.1,
  "encoder_attention_heads": 12,
  "encoder_embed_dim": 768,
  "encoder_ffn_embed_dim": 3072,
  "encoder_layerdrop": 0.0,
  "encoder_layers": 12,
  "encoder_layers_to_keep": null,
  "finetuning_task": "msmarco",
  "layernorm_embedding": true,
  "max_positions": 512,
  "max_source_positions": 512,
  "max_target_positions": 512,
  "model_type": "seed_encoder",
  "no_cross_attention": false,
  "no_scale_embedding": true,
  "no_token_positional_embeddings": false,
  "pad_token_id": 1,
  "pooler_activation_fn": "tanh",
  "pooler_dropout": 0.0,
  "quant_noise_pq": 0.0,
  "quant_noise_pq_block_size": 8,
  "share_all_embeddings": true,
  "share_decoder_input_output_embed": true,
  "tie_adaptive_weights": true,
  "train_ratio": "0.5:0.5",
  "vocab_size": 32769
}