| { |
| "_attn_implementation": "magi", |
| "architectures": [ |
| "LocateAnythingForConditionalGeneration" |
| ], |
| "auto_map": { |
| "AutoConfig": "configuration_locateanything.LocateAnythingConfig", |
| "AutoModel": "modeling_locateanything.LocateAnythingForConditionalGeneration" |
| }, |
| "box_end_token_id": 151669, |
| "box_start_token_id": 151668, |
| "coord_end_token_id": 152677, |
| "coord_start_token_id": 151677, |
| "downsample_ratio": 0.5, |
| "dtype": "bfloat16", |
| "image_token_index": 151665, |
| "loss_version": "v1", |
| "mlp_checkpoint": false, |
| "mlp_connector_layers": 2, |
| "model_type": "locateanything", |
| "none_token_id": 4064, |
| "output_attentions": false, |
| "quantization_config": { |
| "config_groups": { |
| "group_0": { |
| "input_activations": { |
| "actorder": null, |
| "block_structure": null, |
| "dynamic": true, |
| "group_size": null, |
| "num_bits": 8, |
| "observer": null, |
| "observer_kwargs": {}, |
| "strategy": "token", |
| "symmetric": true, |
| "type": "float" |
| }, |
| "output_activations": null, |
| "targets": [ |
| "Linear" |
| ], |
| "weights": { |
| "actorder": null, |
| "block_structure": null, |
| "dynamic": false, |
| "group_size": null, |
| "num_bits": 8, |
| "observer": "minmax", |
| "observer_kwargs": {}, |
| "strategy": "channel", |
| "symmetric": true, |
| "type": "float" |
| } |
| } |
| }, |
| "format": "float-quantized", |
| "global_compression_ratio": null, |
| "ignore": [ |
| "vision_model.encoder.blocks.0.mlp.fc0", |
| "vision_model.encoder.blocks.0.mlp.fc1", |
| "vision_model.encoder.blocks.0.wqkv", |
| "vision_model.encoder.blocks.0.wo", |
| "vision_model.encoder.blocks.1.mlp.fc0", |
| "vision_model.encoder.blocks.1.mlp.fc1", |
| "vision_model.encoder.blocks.1.wqkv", |
| "vision_model.encoder.blocks.1.wo", |
| "vision_model.encoder.blocks.2.mlp.fc0", |
| "vision_model.encoder.blocks.2.mlp.fc1", |
| "vision_model.encoder.blocks.2.wqkv", |
| "vision_model.encoder.blocks.2.wo", |
| "vision_model.encoder.blocks.3.mlp.fc0", |
| "vision_model.encoder.blocks.3.mlp.fc1", |
| "vision_model.encoder.blocks.3.wqkv", |
| "vision_model.encoder.blocks.3.wo", |
| "vision_model.encoder.blocks.4.mlp.fc0", |
| "vision_model.encoder.blocks.4.mlp.fc1", |
| "vision_model.encoder.blocks.4.wqkv", |
| "vision_model.encoder.blocks.4.wo", |
| "vision_model.encoder.blocks.5.mlp.fc0", |
| "vision_model.encoder.blocks.5.mlp.fc1", |
| "vision_model.encoder.blocks.5.wqkv", |
| "vision_model.encoder.blocks.5.wo", |
| "vision_model.encoder.blocks.6.mlp.fc0", |
| "vision_model.encoder.blocks.6.mlp.fc1", |
| "vision_model.encoder.blocks.6.wqkv", |
| "vision_model.encoder.blocks.6.wo", |
| "vision_model.encoder.blocks.7.mlp.fc0", |
| "vision_model.encoder.blocks.7.mlp.fc1", |
| "vision_model.encoder.blocks.7.wqkv", |
| "vision_model.encoder.blocks.7.wo", |
| "vision_model.encoder.blocks.8.mlp.fc0", |
| "vision_model.encoder.blocks.8.mlp.fc1", |
| "vision_model.encoder.blocks.8.wqkv", |
| "vision_model.encoder.blocks.8.wo", |
| "vision_model.encoder.blocks.9.mlp.fc0", |
| "vision_model.encoder.blocks.9.mlp.fc1", |
| "vision_model.encoder.blocks.9.wqkv", |
| "vision_model.encoder.blocks.9.wo", |
| "vision_model.encoder.blocks.10.mlp.fc0", |
| "vision_model.encoder.blocks.10.mlp.fc1", |
| "vision_model.encoder.blocks.10.wqkv", |
| "vision_model.encoder.blocks.10.wo", |
| "vision_model.encoder.blocks.11.mlp.fc0", |
| "vision_model.encoder.blocks.11.mlp.fc1", |
| "vision_model.encoder.blocks.11.wqkv", |
| "vision_model.encoder.blocks.11.wo", |
| "vision_model.encoder.blocks.12.mlp.fc0", |
| "vision_model.encoder.blocks.12.mlp.fc1", |
| "vision_model.encoder.blocks.12.wqkv", |
| "vision_model.encoder.blocks.12.wo", |
| "vision_model.encoder.blocks.13.mlp.fc0", |
| "vision_model.encoder.blocks.13.mlp.fc1", |
| "vision_model.encoder.blocks.13.wqkv", |
| "vision_model.encoder.blocks.13.wo", |
| "vision_model.encoder.blocks.14.mlp.fc0", |
| "vision_model.encoder.blocks.14.mlp.fc1", |
| "vision_model.encoder.blocks.14.wqkv", |
| "vision_model.encoder.blocks.14.wo", |
| "vision_model.encoder.blocks.15.mlp.fc0", |
| "vision_model.encoder.blocks.15.mlp.fc1", |
| "vision_model.encoder.blocks.15.wqkv", |
| "vision_model.encoder.blocks.15.wo", |
| "vision_model.encoder.blocks.16.mlp.fc0", |
| "vision_model.encoder.blocks.16.mlp.fc1", |
| "vision_model.encoder.blocks.16.wqkv", |
| "vision_model.encoder.blocks.16.wo", |
| "vision_model.encoder.blocks.17.mlp.fc0", |
| "vision_model.encoder.blocks.17.mlp.fc1", |
| "vision_model.encoder.blocks.17.wqkv", |
| "vision_model.encoder.blocks.17.wo", |
| "vision_model.encoder.blocks.18.mlp.fc0", |
| "vision_model.encoder.blocks.18.mlp.fc1", |
| "vision_model.encoder.blocks.18.wqkv", |
| "vision_model.encoder.blocks.18.wo", |
| "vision_model.encoder.blocks.19.mlp.fc0", |
| "vision_model.encoder.blocks.19.mlp.fc1", |
| "vision_model.encoder.blocks.19.wqkv", |
| "vision_model.encoder.blocks.19.wo", |
| "vision_model.encoder.blocks.20.mlp.fc0", |
| "vision_model.encoder.blocks.20.mlp.fc1", |
| "vision_model.encoder.blocks.20.wqkv", |
| "vision_model.encoder.blocks.20.wo", |
| "vision_model.encoder.blocks.21.mlp.fc0", |
| "vision_model.encoder.blocks.21.mlp.fc1", |
| "vision_model.encoder.blocks.21.wqkv", |
| "vision_model.encoder.blocks.21.wo", |
| "vision_model.encoder.blocks.22.mlp.fc0", |
| "vision_model.encoder.blocks.22.mlp.fc1", |
| "vision_model.encoder.blocks.22.wqkv", |
| "vision_model.encoder.blocks.22.wo", |
| "vision_model.encoder.blocks.23.mlp.fc0", |
| "vision_model.encoder.blocks.23.mlp.fc1", |
| "vision_model.encoder.blocks.23.wqkv", |
| "vision_model.encoder.blocks.23.wo", |
| "vision_model.encoder.blocks.24.mlp.fc0", |
| "vision_model.encoder.blocks.24.mlp.fc1", |
| "vision_model.encoder.blocks.24.wqkv", |
| "vision_model.encoder.blocks.24.wo", |
| "vision_model.encoder.blocks.25.mlp.fc0", |
| "vision_model.encoder.blocks.25.mlp.fc1", |
| "vision_model.encoder.blocks.25.wqkv", |
| "vision_model.encoder.blocks.25.wo", |
| "vision_model.encoder.blocks.26.mlp.fc0", |
| "vision_model.encoder.blocks.26.mlp.fc1", |
| "vision_model.encoder.blocks.26.wqkv", |
| "vision_model.encoder.blocks.26.wo", |
| "language_model.lm_head", |
| "mlp1.1", |
| "mlp1.3" |
| ], |
| "kv_cache_scheme": null, |
| "quant_method": "compressed-tensors", |
| "quantization_status": "compressed", |
| "run_compressed": false |
| }, |
| "ref_end_token_id": 151673, |
| "ref_start_token_id": 151672, |
| "template": null, |
| "text_config": { |
| "_attn_implementation_autoset": true, |
| "_name_or_path": "Qwen/Qwen2.5-3B-Instruct", |
| "architectures": [ |
| "Qwen2ForCausalLM" |
| ], |
| "attention_dropout": 0.0, |
| "block_size": 6, |
| "bos_token_id": 151643, |
| "causal_attn": false, |
| "dtype": "bfloat16", |
| "eos_token_id": 151645, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 11008, |
| "layer_types": [ |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention", |
| "full_attention" |
| ], |
| "max_position_embeddings": 32768, |
| "max_window_layers": 70, |
| "model_type": "qwen2", |
| "null_token_id": 152678, |
| "num_attention_heads": 16, |
| "num_hidden_layers": 36, |
| "num_key_value_heads": 2, |
| "rms_norm_eps": 1e-06, |
| "rope_scaling": null, |
| "rope_theta": 1000000.0, |
| "sliding_window": null, |
| "switch_token_id": 152679, |
| "text_mask_token_id": 151676, |
| "tie_word_embeddings": true, |
| "use_cache": false, |
| "use_sliding_window": false, |
| "vocab_size": 152681 |
| }, |
| "tie_word_embeddings": true, |
| "transformers_version": null, |
| "use_backbone_lora": 0, |
| "use_llm_lora": 0, |
| "vision_config": { |
| "_attn_implementation_autoset": true, |
| "_name_or_path": "moonshotai/MoonViT-SO-400M", |
| "auto_map": { |
| "AutoConfig": "moonshotai/MoonViT-SO-400M--configuration_moonvit.MoonViTConfig", |
| "AutoModel": "moonshotai/MoonViT-SO-400M--modeling_moonvit.MoonVitPretrainedModel" |
| }, |
| "dtype": "bfloat16", |
| "hidden_size": 1152, |
| "init_pos_emb_height": 64, |
| "init_pos_emb_width": 64, |
| "intermediate_size": 4304, |
| "merge_kernel_size": [ |
| 2, |
| 2 |
| ], |
| "model_type": "moonvit", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 27, |
| "patch_size": 14 |
| }, |
| "torch_dtype": "bfloat16" |
| } |
