config.json
| 1 | { |
| 2 | "architectures": [ |
| 3 | "Qwen2VLForConditionalGeneration" |
| 4 | ], |
| 5 | "attention_dropout": 0.0, |
| 6 | "bos_token_id": 151643, |
| 7 | "eos_token_id": 151645, |
| 8 | "vision_start_token_id": 151652, |
| 9 | "vision_end_token_id": 151653, |
| 10 | "vision_token_id": 151654, |
| 11 | "image_token_id": 151655, |
| 12 | "video_token_id": 151656, |
| 13 | "hidden_act": "silu", |
| 14 | "hidden_size": 3584, |
| 15 | "initializer_range": 0.02, |
| 16 | "intermediate_size": 18944, |
| 17 | "max_position_embeddings": 32768, |
| 18 | "max_window_layers": 28, |
| 19 | "model_type": "qwen2_vl", |
| 20 | "num_attention_heads": 28, |
| 21 | "num_hidden_layers": 28, |
| 22 | "num_key_value_heads": 4, |
| 23 | "rms_norm_eps": 1e-06, |
| 24 | "rope_theta": 1000000.0, |
| 25 | "sliding_window": 32768, |
| 26 | "tie_word_embeddings": false, |
| 27 | "torch_dtype": "bfloat16", |
| 28 | "transformers_version": "4.41.2", |
| 29 | "use_cache": true, |
| 30 | "use_sliding_window": false, |
| 31 | "vision_config": { |
| 32 | "depth": 32, |
| 33 | "embed_dim": 1280, |
| 34 | "mlp_ratio": 4, |
| 35 | "num_heads": 16, |
| 36 | "in_chans": 3, |
| 37 | "hidden_size": 3584, |
| 38 | "patch_size": 14, |
| 39 | "spatial_merge_size": 2, |
| 40 | "spatial_patch_size": 14, |
| 41 | "temporal_patch_size": 2 |
| 42 | }, |
| 43 | "rope_scaling": { |
| 44 | "type": "mrope", |
| 45 | "mrope_section": [ |
| 46 | 16, |
| 47 | 24, |
| 48 | 24 |
| 49 | ] |
| 50 | }, |
| 51 | "vocab_size": 152064 |
| 52 | } |