{ "architectures": [ "SwarmMoEForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "auto_map": { "AutoConfig": "configuration_swarm_moe.SwarmMoEConfig", "AutoModelForCausalLM": "modeling_swarm_moe.SwarmMoEForCausalLM" }, "bos_token_id": 1, "deeper_router": true, "dtype": "bfloat16", "eos_token_id": 2, "head_dim": 64, "hidden_act": "silu", "hidden_size": 1024, "initializer_range": 0.02, "intermediate_size": 1792, "max_position_embeddings": 4096, "memory_injection_strength": 0.0, "model_type": "swarm_moe", "norm_topk_prob": true, "num_attention_heads": 16, "num_experts": 16, "num_experts_per_tok": 2, "num_hidden_layers": 40, "num_key_value_heads": 4, "num_shared_experts": 1, "num_unique_layers": 40, "output_router_logits": false, "pad_token_id": 0, "plasticity_strength": 0.0, "quantization": "none", "qwen35_growth_policy": "preserve_existing_phill_tensors_exactly", "qwen35_growth_source": "projected_new_capacity_only", "rms_norm_eps": 1e-06, "rope_parameters": { "rope_theta": 10000.0, "rope_type": "default" }, "rope_theta": 10000.0, "router_aux_loss_coef": 0.01, "router_bias_strength": 0.0, "router_jitter_noise": 0.0, "router_z_loss_coef": 0.001, "sliding_window": null, "swarm_app_config": "swarm_app_config.example.json", "swarm_app_enabled": false, "swarm_app_launcher": "launch_swarm_app.py", "swarm_runtime_config": "swarm_runtime_config.json", "test_time_compute_layer": -1, "test_time_compute_steps": 0, "tie_word_embeddings": true, "transformers_version": "5.3.0", "use_cache": true, "use_v_norm": true, "vision_sidecar_enabled": true, "vision_sidecar_max_blocks": 24, "vision_sidecar_path": "vision_sidecar", "vision_snapshot_policy": "retain_latest_only", "vision_source_checkpoint": "external_source_redacted_for_portable_hf_package", "vision_token_map": { "reason": "source vision token ids are outside the 32k tied embedding table; pixel tensors route through the sidecar instead", "status": "metadata_only_no_embedding_resize", "target_vocab_size": 32000, "tokens": { "<|box_end|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248050, "target_id": null }, "<|box_start|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248049, "target_id": null }, "<|image_pad|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248056, "target_id": null }, "<|object_ref_end|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248048, "target_id": null }, "<|object_ref_start|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248047, "target_id": null }, "<|quad_end|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248052, "target_id": null }, "<|quad_start|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248051, "target_id": null }, "<|video_pad|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248057, "target_id": null }, "<|vision_end|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248054, "target_id": null }, "<|vision_pad|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248055, "target_id": null }, "<|vision_start|>": { "encoding_policy": "external_processor_marker_not_embedded", "source_id": 248053, "target_id": null } } }, "vocab_size": 248077, "router_entropy_loss_coef": 0.0001, "router_confidence_loss_coef": 0.0, "thinking_consistency_loss_coef": 0.0001, "scaffold_alignment_loss_coef": 0.0001 }