config.json

{
  "architectures": [
    "VideoChatFlashQwenForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "modeling_videochat_flash.VideoChatFlashQwenConfig",
    "AutoModel": "modeling_videochat_flash.VideoChatFlashQwenForCausalLM"
  },
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "eos_token_id": 151645,
  "frame_aspect_ratio": "square",
  "frame_grid_pinpoints": null,
  "hidden_act": "silu",
  "hidden_size": 3584,
  "image_aspect_ratio": "anyres_nopad",
  "image_crop_resolution": null,
  "image_grid_pinpoints": [
    [
      448,
      448
    ],
    [
      448,
      896
    ],
    [
      448,
      1344
    ],
    [
      448,
      1792
    ],
    [
      448,
      2240
    ],
    [
      448,
      2688
    ],
    [
      896,
      448
    ],
    [
      896,
      896
    ],
    [
      896,
      1344
    ],
    [
      896,
      1792
    ],
    [
      896,
      2240
    ],
    [
      896,
      2688
    ],
    [
      1344,
      448
    ],
    [
      1344,
      896
    ],
    [
      1344,
      1344
    ],
    [
      1344,
      1792
    ],
    [
      1344,
      2240
    ],
    [
      1344,
      2688
    ],
    [
      1792,
      448
    ],
    [
      1792,
      896
    ],
    [
      1792,
      1344
    ],
    [
      1792,
      1792
    ],
    [
      1792,
      2240
    ],
    [
      1792,
      2688
    ],
    [
      2240,
      448
    ],
    [
      2240,
      896
    ],
    [
      2240,
      1344
    ],
    [
      2240,
      1792
    ],
    [
      2240,
      2240
    ],
    [
      2240,
      2688
    ],
    [
      2688,
      448
    ],
    [
      2688,
      896
    ],
    [
      2688,
      1344
    ],
    [
      2688,
      1792
    ],
    [
      2688,
      2240
    ],
    [
      2688,
      2688
    ]
  ],
  "image_split_resolution": null,
  "initializer_range": 0.02,
  "intermediate_size": 18944,
  "llm_compress_layer_list": [
    24
  ],
  "llm_compress_type": "attention",
  "llm_image_token_ratio_list": [
    1.0,

    0.5
  ],
  "max_num_pixels": 14745600000,
  "max_position_embeddings": 32768,
  "max_window_layers": 28,
  "min_slow_num_frames": 4,
  "mm_close_init": false,
  "mm_hidden_size": 1024,
  "mm_llm_compress": false,
  "mm_local_num_frames": 4,
  "mm_newline_position": "nothing",
  "mm_num_compress_latents": 128,
  "mm_num_compress_query_type": "learnable",
  "mm_patch_merge_type": "spatial_nopad",
  "mm_pos_num_frames": 8,
  "mm_projector_lr": null,
  "mm_projector_type": "tome16_mlp_hd64",
  "mm_resampler_type": null,
  "mm_spatial_pool_mode": "bilinear",
  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter",
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": false,
  "mm_vision_select_feature": "patch",
  "mm_vision_select_layer": -2,
  "mm_vision_tower": "umt-hd-large",
  "mm_vision_tower_lr": 2e-06,
  "model_type": "qwen2",
  "num_attention_heads": 28,
  "num_hidden_layers": 28,
  "num_key_value_heads": 4,
  "pos_skipping_range": 4096,
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000000.0,
  "sliding_window": 131072,
  "tie_word_embeddings": false,
  "tokenizer_model_max_length": 32768,
  "tokenizer_padding_side": "right",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.39.2",
  "use_cache": true,
  "use_mm_proj": true,
  "use_pos_skipping": false,
  "use_sliding_window": false,
  "vision_encode_type": "video_image",
  "vision_tower_pretrained": null,
  "vocab_size": 152064
}