{
  "bos_token_id": 1,
  "class_token_index": 64402,
  "dropout": 0.1,
  "embed_ent_token": true,
  "encoder_config": {
    "transformers_version": "5.8.1",
    "architectures": [
      "Lfm2BiModel"
    ],
    "output_hidden_states": false,
    "return_dict": true,
    "dtype": "float32",
    "chunk_size_feed_forward": 0,
    "is_encoder_decoder": false,
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "problem_type": null,
    "vocab_size": 64404,
    "hidden_size": 1024,
    "intermediate_size": 6656,
    "num_hidden_layers": 16,
    "num_attention_heads": 16,
    "num_key_value_heads": 8,
    "max_position_embeddings": 128000,
    "initializer_range": 0.02,
    "norm_eps": 1e-05,
    "use_cache": true,
    "pad_token_id": 0,
    "bos_token_id": 1,
    "eos_token_id": 7,
    "tie_word_embeddings": true,
    "rope_parameters": {
      "rope_theta": 1000000.0,
      "rope_type": "default"
    },
    "conv_bias": false,
    "conv_L_cache": 3,
    "block_multiple_of": 256,
    "block_ffn_dim_multiplier": 1.0,
    "block_auto_adjust_ff_dim": true,
    "full_attn_idxs": null,
    "layer_types": [
      "conv",
      "conv",
      "full_attention",
      "conv",
      "conv",
      "full_attention",
      "conv",
      "conv",
      "full_attention",
      "conv",
      "full_attention",
      "conv",
      "full_attention",
      "conv",
      "full_attention",
      "conv"
    ],
    "_name_or_path": "/run/determined/workdir/mmontebovi/gliner_boost/models_lfm2_bi_mlm_v2/ckpt22500_backbone",
    "block_dim": 1024,
    "block_mlp_init_scale": 1.0,
    "block_norm_eps": 1e-05,
    "block_out_init_scale": 1.0,
    "block_use_swiglu": true,
    "block_use_xavier_init": true,
    "conv_dim": 1024,
    "conv_use_xavier_init": true,
    "model_type": "lfm2",
    "num_heads": 16,
    "use_pos_enc": true,
    "output_attentions": false
  },
  "ent_token": "<<ENT>>",
  "eos_token_id": 7,
  "fine_tune": true,
  "fuse_layers": true,
  "hidden_size": 1024,
  "labels_decoder": null,
  "labels_encoder": null,
  "max_len": 1024,
  "max_neg_type_ratio": 1,
  "max_types": 100,
  "max_width": 12,
  "model_name": "/run/determined/workdir/mmontebovi/gliner_boost/models_lfm2_bi_mlm_v2/ckpt22500_backbone",
  "model_type": null,
  "moe_aux_loss_coef": 0.0,
  "moe_bilinear": false,
  "moe_bilinear_gate_init_std": 0.02,
  "moe_bilinear_init_std": 0.01,
  "moe_bilinear_num_experts": 8,
  "moe_bilinear_rank": 32,
  "moe_bilinear_reg_coef": 0.001,
  "moe_drop_upcycle_p": 0.5,
  "moe_expert_dim": null,
  "moe_gate_init_std": 0.02,
  "moe_num_experts": null,
  "moe_num_topics": null,
  "moe_post_encoder": false,
  "moe_residual_scale": 0.1,
  "moe_shared_expert_dim": null,
  "moe_top_k": 2,
  "moe_topic_loss_coef": 0.1,
  "moe_topic_routed": false,
  "moe_use_shared_expert": false,
  "moe_weight_scale_exp": 0.3333333333333333,
  "moe_zloss_coef": 0.0,
  "name": "LFM2.5-350M +MLM-v2(ckpt22500) DENSE GLiNER \u2014 STAGE 2",
  "neg_spans_ratio": 1.0,
  "num_post_fusion_layers": 1,
  "num_rnn_layers": 1,
  "pad_token_id": 0,
  "post_fusion_schema": null,
  "represent_spans": false,
  "sep_token": "<<SEP>>",
  "span_loss_coef": 1.0,
  "span_mode": "markerV1",
  "subtoken_pooling": "first",
  "token_loss_coef": 1.0,
  "transformers_version": "5.8.1",
  "use_cache": false,
  "vocab_size": 64404,
  "words_splitter_type": "whitespace"
}