| { | |
| "architectures": [ | |
| "MiniCPMV4_7ForConditionalGeneration" | |
| ], | |
| "downsample_mode": "16x", | |
| "image_end_id": 248079, | |
| "image_size": 448, | |
| "image_start_id": 248078, | |
| "image_token_id": 248056, | |
| "insert_layer_id": 6, | |
| "merge_kernel_size": [ | |
| 2, | |
| 2 | |
| ], | |
| "merger_times": 1, | |
| "model_type": "minicpmv4_7", | |
| "mrope_mode": "canvas", | |
| "newline_id": 198, | |
| "patch_size": 14, | |
| "slice_end_id": 248089, | |
| "slice_start_id": 248088, | |
| "text_config": { | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "attn_output_gate": true, | |
| "bos_token_id": null, | |
| "dtype": "float32", | |
| "eos_token_id": null, | |
| "full_attention_interval": 4, | |
| "head_dim": 256, | |
| "hidden_act": "silu", | |
| "hidden_size": 1024, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 3584, | |
| "layer_types": [ | |
| "linear_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "full_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "full_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "full_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "full_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "full_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "linear_attention", | |
| "full_attention" | |
| ], | |
| "linear_conv_kernel_dim": 4, | |
| "linear_key_head_dim": 128, | |
| "linear_num_key_heads": 16, | |
| "linear_num_value_heads": 16, | |
| "linear_value_head_dim": 128, | |
| "mamba_ssm_dtype": "float32", | |
| "max_position_embeddings": 262144, | |
| "mlp_only_layers": [], | |
| "model_type": "qwen3_5_text", | |
| "mtp_num_hidden_layers": 1, | |
| "mtp_use_dedicated_embeddings": false, | |
| "num_attention_heads": 8, | |
| "num_hidden_layers": 24, | |
| "num_key_value_heads": 2, | |
| "pad_token_id": null, | |
| "rms_norm_eps": 1e-06, | |
| "rope_parameters": { | |
| "mrope_interleaved": true, | |
| "mrope_section": [ | |
| 11, | |
| 11, | |
| 10 | |
| ], | |
| "partial_rotary_factor": 0.25, | |
| "rope_theta": 10000000, | |
| "rope_type": "default" | |
| }, | |
| "tie_word_embeddings": false, | |
| "use_cache": true, | |
| "vocab_size": 248094 | |
| }, | |
| "tie_word_embeddings": false, | |
| "transformers_version": "5.20.2", | |
| "version": 4.7, | |
| "video_token_id": 248057, | |
| "vision_config": { | |
| "attention_dropout": 0.0, | |
| "dtype": "float32", | |
| "hidden_act": "gelu_pytorch_tanh", | |
| "hidden_size": 1152, | |
| "image_size": 980, | |
| "insert_layer_id": 6, | |
| "intermediate_size": 4304, | |
| "layer_norm_eps": 1e-06, | |
| "model_type": "minicpmv4_7_vision", | |
| "num_attention_heads": 16, | |
| "num_channels": 3, | |
| "num_hidden_layers": 27, | |
| "patch_size": 14, | |
| "window_kernel_size": [ | |
| 2, | |
| 2 | |
| ] | |
| }, | |
| "transformers.js_config": { | |
| "use_external_data_format": { | |
| "embed_tokens_q8f16.onnx": 1, | |
| "decoder_model_merged_q8f16.onnx": 1, | |
| "embed_tokens_q4f16.onnx": 1, | |
| "decoder_model_merged_q4f16.onnx": 1, | |
| "vision_encoder_fp16.onnx": 1 | |
| }, | |
| "kv_cache_dtype": { | |
| "q8f16": "float16", | |
| "q4f16": "float16" | |
| } | |
| }, | |
| "webbrain_runtime": { | |
| "runtime": "runtime/minicpmv-runtime.mjs", | |
| "recommended": { | |
| "embed_tokens": "onnx/embed_tokens_q8f16.onnx", | |
| "decoder_model_merged": "onnx/decoder_model_merged_q8f16.onnx", | |
| "vision_encoder": "onnx/vision_encoder_fp16.onnx" | |
| }, | |
| "lite": { | |
| "embed_tokens": "onnx/embed_tokens_q4f16.onnx", | |
| "decoder_model_merged": "onnx/decoder_model_merged_q4f16.onnx", | |
| "vision_encoder": "onnx/vision_encoder_fp16.onnx" | |
| }, | |
| "vision_encoder_io": "pixel_values float32[num_crops,3,14h,14w] (h,w multiples of 4; (x-127.5)/127.5, PIL bicubic resize + MiniCPM-V slicing) -> image_features float32[num_crops,(h/4)*(w/4),1024]", | |
| "decoder_io": "Qwen3.5-0.8B topology: inputs_embeds, attention_mask, position_ids int64[3,B,S] (canvas M-RoPE); 18 GatedDeltaNet conv/recurrent + 6 KV caches in fp16; logits and decision_hidden for the last position", | |
| "chunked_prefill_tokens": 128, | |
| "processor": { | |
| "max_slice_nums": 9, | |
| "scale_resolution": 448, | |
| "patch_size": 14, | |
| "downsample_mode": "16x", | |
| "use_image_id": true | |
| } | |
| } | |
| } |