hf-transformers-bot HF Staff
Update tiny models for Qwen2_5_VLForConditionalGeneration
e05c980 verified | { | |
| "architectures": [ | |
| "Qwen2_5_VLForConditionalGeneration" | |
| ], | |
| "dtype": "float32", | |
| "image_token_id": 4, | |
| "model_type": "qwen2_5_vl", | |
| "text_config": { | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 0, | |
| "eos_token_id": 151645, | |
| "hidden_act": "silu", | |
| "hidden_size": 32, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 37, | |
| "layer_types": [ | |
| "full_attention", | |
| "full_attention" | |
| ], | |
| "max_position_embeddings": 512, | |
| "max_window_layers": 3, | |
| "model_type": "qwen2_5_vl_text", | |
| "num_attention_heads": 4, | |
| "num_hidden_layers": 2, | |
| "num_key_value_heads": 2, | |
| "pad_token_id": 151643, | |
| "rms_norm_eps": 1e-05, | |
| "rope_parameters": { | |
| "mrope_section": [ | |
| 2, | |
| 1, | |
| 1 | |
| ], | |
| "rope_theta": 10000, | |
| "rope_type": "default", | |
| "type": "mrope" | |
| }, | |
| "sliding_window": null, | |
| "tie_word_embeddings": true, | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "vocab_size": 151657 | |
| }, | |
| "tie_word_embeddings": false, | |
| "transformers_version": "5.16.0.dev0", | |
| "video_token_id": 5, | |
| "vision_config": { | |
| "depth": 2, | |
| "fullatt_block_indexes": [ | |
| 7, | |
| 15, | |
| 23, | |
| 31 | |
| ], | |
| "hidden_act": "silu", | |
| "hidden_size": 32, | |
| "in_channels": 3, | |
| "in_chans": 3, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 32, | |
| "model_type": "qwen2_5_vl_vision", | |
| "num_heads": 4, | |
| "out_hidden_size": 32, | |
| "patch_size": 14, | |
| "spatial_merge_size": 1, | |
| "spatial_patch_size": 14, | |
| "temporal_patch_size": 2, | |
| "tokens_per_second": 1, | |
| "window_size": 112 | |
| }, | |
| "vision_end_token_id": 151653, | |
| "vision_start_token_id": 3 | |
| } | |