| { |
| "architectures": [ |
| "Qwen3DSparkModel" |
| ], |
| "model_type": "qwen3", |
| "hidden_size": 6144, |
| "head_dim": 64, |
| "intermediate_size": 12288, |
| "num_hidden_layers": 5, |
| "num_attention_heads": 64, |
| "num_key_value_heads": 64, |
| "vocab_size": 154880, |
| "draft_vocab_size": 32000, |
| "rms_norm_eps": 1e-05, |
| "max_position_embeddings": 40960, |
| "rope_theta": 8000000.0, |
| "num_target_layers": 5, |
| "target_hidden_size": 6144, |
| "target_num_hidden_layers": 78, |
| "target_layer_ids": [ |
| 8, |
| 23, |
| 39, |
| 55, |
| 70 |
| ], |
| "mask_token_id": 154856, |
| "block_size": 8, |
| "markov_rank": 256, |
| "markov_head_type": "vanilla", |
| "enable_confidence_head": true, |
| "confidence_head_with_markov": true, |
| "confidence_detach_input": true, |
| "moe_type": "vanilla", |
| "num_experts": 8, |
| "num_experts_per_tok": 2, |
| "norm_topk_prob": true, |
| "num_shared_experts": 1, |
| "moe_intermediate_size": 3072, |
| "griffin_fuse": true, |
| "per_layer_context_fusion": true, |
| "tie_word_embeddings": false, |
| "_torchspec_version": "0.1.0", |
| "torch_dtype": "bfloat16" |
| } |
