Ling-3.0-tiny-UBTree / config.json
Rookie-Liu's picture
Update config.json
775cef4 verified
Raw History Blame Contribute Delete
2.36 kB
{
"architectures": [
"Ling3UBTreeModel"
],
"attention_bias": false,
"attention_dropout": 0.0,
"block_size": 16,
"bos_token_id": null,
"dflash_config": {
"block_size": 16,
"contract": "flashflow_anchor_inclusive_v1",
"mask_token_id": 156892,
"num_target_layers": 24,
"target_layer_ids": [
4,
8,
12,
16,
22
]
},
"dflash_contract": "flashflow_anchor_inclusive_v1",
"dtype": "bfloat16",
"eos_token_id": 156895,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 1536,
"initializer_range": 0.02,
"intermediate_size": 4608,
"layer_types": [
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention"
],
"mask_token_id": 156892,
"max_position_embeddings": 131072,
"max_window_layers": 20,
"model_type": "qwen3",
"num_attention_heads": 16,
"num_hidden_layers": 5,
"num_key_value_heads": 16,
"num_target_layers": 24,
"pad_token_id": 156892,
"partial_rotary_factor": 1.0,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"partial_rotary_factor": 1.0,
"rope_theta": 6000000,
"rope_type": "default"
},
"rope_theta": 6000000.0,
"sliding_window": null,
"target_layer_ids": [
4,
8,
12,
16,
22
],
"tie_word_embeddings": false,
"transformers_version": "5.12.1",
"ubtree_config": {
"adds_unary_logits": true,
"architecture": "cp",
"candidate_k": 64,
"candidate_source": "frozen_dflash_unary_topk_ids_only",
"depth_condition_count": 15,
"depth_conditioning": "unary_correction_two_scale",
"depth_scale_parameterization": "exp_log_scale",
"factorizer_hidden_width": 512,
"hidden_projection_depth_count": 15,
"hidden_projection_type": "shared",
"inference_hidden_projection": "materialized_per_depth_affine",
"kl_temperature": 1.0,
"objective": "kl",
"rank": 256,
"target_embedding_dim": 1536,
"token_factorizer": "target_embedding_mlp",
"train_backbone": true,
"trained": true,
"training_mode": "joint_backbone_selector",
"tree_budget": 64,
"tree_builder": "pruned",
"tree_depth_bonus": 0.0,
"tree_supertree_width": 12,
"type": "residual_bigram",
"version": 1
},
"use_cache": true,
"use_qk_norm": true,
"use_sliding_window": false,
"vocab_size": 157184
}