| { | |
| "architectures": [ | |
| "InternS1ForConditionalGeneration" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_interns1.InternS1Config", | |
| "AutoModel": "modeling_interns1.InternS1Model", | |
| "AutoModelForCausalLM": "modeling_interns1.InternS1ForConditionalGeneration" | |
| }, | |
| "downsample_ratio": 0.5, | |
| "image_seq_length": 256, | |
| "image_token_id": 152957, | |
| "model_type": "interns1", | |
| "projector_hidden_act": "gelu", | |
| "text_config": { | |
| "architectures": [ | |
| "Qwen3MoeForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 151643, | |
| "decoder_sparse_step": 1, | |
| "eos_token_id": 151645, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 4096, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 8192, | |
| "max_position_embeddings": 65536, | |
| "max_window_layers": 94, | |
| "mlp_only_layers": [], | |
| "model_type": "qwen3_moe", | |
| "moe_intermediate_size": 1536, | |
| "norm_topk_prob": true, | |
| "num_attention_heads": 64, | |
| "num_experts": 128, | |
| "num_experts_per_tok": 8, | |
| "num_hidden_layers": 94, | |
| "num_key_value_heads": 4, | |
| "output_router_logits": false, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 1000000.0, | |
| "router_aux_loss_coef": 0.001, | |
| "sliding_window": null, | |
| "torch_dtype": "bfloat16", | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "vocab_size": 153216 | |
| }, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.53.0", | |
| "vision_config": { | |
| "architectures": [ | |
| "InternVisionModel" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_interns1.InternS1VisionConfig", | |
| "AutoModel": "modeling_interns1.InternS1VisionModel" | |
| }, | |
| "drop_path_rate": 0.1, | |
| "dropout": 0.0, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.0, | |
| "hidden_size": 3200, | |
| "image_size": [ | |
| 448, | |
| 448 | |
| ], | |
| "initializer_factor": 0.1, | |
| "initializer_range": 1e-10, | |
| "intermediate_size": 12800, | |
| "layer_norm_eps": 1e-06, | |
| "layer_scale_init_value": 0.1, | |
| "model_type": "interns1_vision", | |
| "norm_type": "rms_norm", | |
| "num_attention_heads": 25, | |
| "num_channels": 3, | |
| "num_hidden_layers": 45, | |
| "patch_size": [ | |
| 14, | |
| 14 | |
| ], | |
| "projection_dropout": 0.0, | |
| "torch_dtype": "bfloat16", | |
| "use_absolute_position_embeddings": true, | |
| "use_bfloat16": true, | |
| "use_flash_attn": true, | |
| "use_mask_token": false, | |
| "use_mean_pooling": true, | |
| "use_qk_norm": true | |
| }, | |
| "vision_feature_layer": -1, | |
| "vision_feature_select_strategy": "default" | |
| } | |