rafmacalaba/lfm2.5-Encoder-350M-datause
018
1{2 "architectures": [3 "Lfm2BidirectionalModel"4 ],5 "auto_map": {6 "AutoModel": "modeling_lfm2_bidirectional.Lfm2BidirectionalModel",7 "AutoModelForMaskedLM": "modeling_lfm2_bidirectional.Lfm2BidirectionalForMaskedLM"8 },9 "block_auto_adjust_ff_dim": true,10 "block_dim": 1024,11 "block_ffn_dim_multiplier": 1.0,12 "block_mlp_init_scale": 1.0,13 "block_multiple_of": 256,14 "block_norm_eps": 1e-05,15 "block_out_init_scale": 1.0,16 "block_use_swiglu": true,17 "block_use_xavier_init": true,18 "bos_token_id": 1,19 "conv_L_cache": 3,20 "conv_bias": false,21 "conv_dim": 1024,22 "conv_dim_out": 1024,23 "conv_use_xavier_init": true,24 "dtype": "float32",25 "eos_token_id": 7,26 "full_attn_idxs": null,27 "hidden_size": 1024,28 "id2label": {29 "0": "O",30 "1": "B-NAMED_DATA",31 "2": "I-NAMED_DATA",32 "3": "B-DESCRIPTIVE_DATA",33 "4": "I-DESCRIPTIVE_DATA",34 "5": "B-VAGUE_DATA",35 "6": "I-VAGUE_DATA"36 },37 "initializer_range": 0.02,38 "intermediate_size": 6656,39 "label2id": {40 "B-DESCRIPTIVE_DATA": 3,41 "B-NAMED_DATA": 1,42 "B-VAGUE_DATA": 5,43 "I-DESCRIPTIVE_DATA": 4,44 "I-NAMED_DATA": 2,45 "I-VAGUE_DATA": 6,46 "O": 047 },48 "layer_types": [49 "conv",50 "conv",51 "full_attention",52 "conv",53 "conv",54 "full_attention",55 "conv",56 "conv",57 "full_attention",58 "conv",59 "full_attention",60 "conv",61 "full_attention",62 "conv",63 "full_attention",64 "conv"65 ],66 "max_position_embeddings": 128000,67 "model_type": "lfm2",68 "norm_eps": 1e-05,69 "num_attention_heads": 16,70 "num_heads": 16,71 "num_hidden_layers": 16,72 "num_key_value_heads": 8,73 "pad_token_id": 0,74 "rope_parameters": {75 "rope_theta": 1000000.0,76 "rope_type": "default"77 },78 "tie_word_embeddings": true,79 "transformers_version": "5.13.1",80 "use_cache": false,81 "use_pos_enc": true,82 "vocab_size": 6553683}84 