mirror of
				https://github.com/Wan-Video/Wan2.1.git
				synced 2025-11-04 06:15:17 +00:00 
			
		
		
		
	
		
			
				
	
	
		
			267 lines
		
	
	
		
			12 KiB
		
	
	
	
		
			Python
		
	
	
	
	
	
			
		
		
	
	
			267 lines
		
	
	
		
			12 KiB
		
	
	
	
		
			Python
		
	
	
	
	
	
# thanks Comfyui for the rgb factors (https://github.com/comfyanonymous/ComfyUI/blob/master/comfy/latent_formats.py)
 | 
						|
def get_rgb_factors(model_family, model_type = None): 
 | 
						|
    if model_family in ["wan", "qwen"]:
 | 
						|
        if model_type =="ti2v_2_2":
 | 
						|
            latent_channels = 48
 | 
						|
            latent_dimensions = 3            
 | 
						|
            latent_rgb_factors = [
 | 
						|
                [ 0.0119,  0.0103,  0.0046],
 | 
						|
                [-0.1062, -0.0504,  0.0165],
 | 
						|
                [ 0.0140,  0.0409,  0.0491],
 | 
						|
                [-0.0813, -0.0677,  0.0607],
 | 
						|
                [ 0.0656,  0.0851,  0.0808],
 | 
						|
                [ 0.0264,  0.0463,  0.0912],
 | 
						|
                [ 0.0295,  0.0326,  0.0590],
 | 
						|
                [-0.0244, -0.0270,  0.0025],
 | 
						|
                [ 0.0443, -0.0102,  0.0288],
 | 
						|
                [-0.0465, -0.0090, -0.0205],
 | 
						|
                [ 0.0359,  0.0236,  0.0082],
 | 
						|
                [-0.0776,  0.0854,  0.1048],
 | 
						|
                [ 0.0564,  0.0264,  0.0561],
 | 
						|
                [ 0.0006,  0.0594,  0.0418],
 | 
						|
                [-0.0319, -0.0542, -0.0637],
 | 
						|
                [-0.0268,  0.0024,  0.0260],
 | 
						|
                [ 0.0539,  0.0265,  0.0358],
 | 
						|
                [-0.0359, -0.0312, -0.0287],
 | 
						|
                [-0.0285, -0.1032, -0.1237],
 | 
						|
                [ 0.1041,  0.0537,  0.0622],
 | 
						|
                [-0.0086, -0.0374, -0.0051],
 | 
						|
                [ 0.0390,  0.0670,  0.2863],
 | 
						|
                [ 0.0069,  0.0144,  0.0082],
 | 
						|
                [ 0.0006, -0.0167,  0.0079],
 | 
						|
                [ 0.0313, -0.0574, -0.0232],
 | 
						|
                [-0.1454, -0.0902, -0.0481],
 | 
						|
                [ 0.0714,  0.0827,  0.0447],
 | 
						|
                [-0.0304, -0.0574, -0.0196],
 | 
						|
                [ 0.0401,  0.0384,  0.0204],
 | 
						|
                [-0.0758, -0.0297, -0.0014],
 | 
						|
                [ 0.0568,  0.1307,  0.1372],
 | 
						|
                [-0.0055, -0.0310, -0.0380],
 | 
						|
                [ 0.0239, -0.0305,  0.0325],
 | 
						|
                [-0.0663, -0.0673, -0.0140],
 | 
						|
                [-0.0416, -0.0047, -0.0023],
 | 
						|
                [ 0.0166,  0.0112, -0.0093],
 | 
						|
                [-0.0211,  0.0011,  0.0331],
 | 
						|
                [ 0.1833,  0.1466,  0.2250],
 | 
						|
                [-0.0368,  0.0370,  0.0295],
 | 
						|
                [-0.3441, -0.3543, -0.2008],
 | 
						|
                [-0.0479, -0.0489, -0.0420],
 | 
						|
                [-0.0660, -0.0153,  0.0800],
 | 
						|
                [-0.0101,  0.0068,  0.0156],
 | 
						|
                [-0.0690, -0.0452, -0.0927],
 | 
						|
                [-0.0145,  0.0041,  0.0015],
 | 
						|
                [ 0.0421,  0.0451,  0.0373],
 | 
						|
                [ 0.0504, -0.0483, -0.0356],
 | 
						|
                [-0.0837,  0.0168,  0.0055]
 | 
						|
        ]
 | 
						|
        else:
 | 
						|
            latent_channels = 16
 | 
						|
            latent_dimensions = 3
 | 
						|
            latent_rgb_factors = [
 | 
						|
                    [-0.1299, -0.1692,  0.2932],
 | 
						|
                    [ 0.0671,  0.0406,  0.0442],
 | 
						|
                    [ 0.3568,  0.2548,  0.1747],
 | 
						|
                    [ 0.0372,  0.2344,  0.1420],
 | 
						|
                    [ 0.0313,  0.0189, -0.0328],
 | 
						|
                    [ 0.0296, -0.0956, -0.0665],
 | 
						|
                    [-0.3477, -0.4059, -0.2925],
 | 
						|
                    [ 0.0166,  0.1902,  0.1975],
 | 
						|
                    [-0.0412,  0.0267, -0.1364],
 | 
						|
                    [-0.1293,  0.0740,  0.1636],
 | 
						|
                    [ 0.0680,  0.3019,  0.1128],
 | 
						|
                    [ 0.0032,  0.0581,  0.0639],
 | 
						|
                    [-0.1251,  0.0927,  0.1699],
 | 
						|
                    [ 0.0060, -0.0633,  0.0005],
 | 
						|
                    [ 0.3477,  0.2275,  0.2950],
 | 
						|
                    [ 0.1984,  0.0913,  0.1861]
 | 
						|
                ]
 | 
						|
    
 | 
						|
        latent_rgb_factors_bias = [-0.1835, -0.0868, -0.3360]
 | 
						|
 | 
						|
        # latent_rgb_factors_bias = [0.0259, -0.0192, -0.0761]
 | 
						|
    elif model_family =="flux":
 | 
						|
        scale_factor = 0.3611
 | 
						|
        shift_factor = 0.1159
 | 
						|
        latent_rgb_factors =[
 | 
						|
            [-0.0346,  0.0244,  0.0681],
 | 
						|
            [ 0.0034,  0.0210,  0.0687],
 | 
						|
            [ 0.0275, -0.0668, -0.0433],
 | 
						|
            [-0.0174,  0.0160,  0.0617],
 | 
						|
            [ 0.0859,  0.0721,  0.0329],
 | 
						|
            [ 0.0004,  0.0383,  0.0115],
 | 
						|
            [ 0.0405,  0.0861,  0.0915],
 | 
						|
            [-0.0236, -0.0185, -0.0259],
 | 
						|
            [-0.0245,  0.0250,  0.1180],
 | 
						|
            [ 0.1008,  0.0755, -0.0421],
 | 
						|
            [-0.0515,  0.0201,  0.0011],
 | 
						|
            [ 0.0428, -0.0012, -0.0036],
 | 
						|
            [ 0.0817,  0.0765,  0.0749],
 | 
						|
            [-0.1264, -0.0522, -0.1103],
 | 
						|
            [-0.0280, -0.0881, -0.0499],
 | 
						|
            [-0.1262, -0.0982, -0.0778]
 | 
						|
        ]
 | 
						|
        latent_rgb_factors_bias = [-0.0329, -0.0718, -0.0851]
 | 
						|
 | 
						|
    elif model_family == "ltxv":
 | 
						|
        latent_channels = 128
 | 
						|
        latent_dimensions = 3
 | 
						|
 | 
						|
        latent_rgb_factors = [
 | 
						|
            [ 1.1202e-02, -6.3815e-04, -1.0021e-02],
 | 
						|
            [ 8.6031e-02,  6.5813e-02,  9.5409e-04],
 | 
						|
            [-1.2576e-02, -7.5734e-03, -4.0528e-03],
 | 
						|
            [ 9.4063e-03, -2.1688e-03,  2.6093e-03],
 | 
						|
            [ 3.7636e-03,  1.2765e-02,  9.1548e-03],
 | 
						|
            [ 2.1024e-02, -5.2973e-03,  3.4373e-03],
 | 
						|
            [-8.8896e-03, -1.9703e-02, -1.8761e-02],
 | 
						|
            [-1.3160e-02, -1.0523e-02,  1.9709e-03],
 | 
						|
            [-1.5152e-03, -6.9891e-03, -7.5810e-03],
 | 
						|
            [-1.7247e-03,  4.6560e-04, -3.3839e-03],
 | 
						|
            [ 1.3617e-02,  4.7077e-03, -2.0045e-03],
 | 
						|
            [ 1.0256e-02,  7.7318e-03,  1.3948e-02],
 | 
						|
            [-1.6108e-02, -6.2151e-03,  1.1561e-03],
 | 
						|
            [ 7.3407e-03,  1.5628e-02,  4.4865e-04],
 | 
						|
            [ 9.5357e-04, -2.9518e-03, -1.4760e-02],
 | 
						|
            [ 1.9143e-02,  1.0868e-02,  1.2264e-02],
 | 
						|
            [ 4.4575e-03,  3.6682e-05, -6.8508e-03],
 | 
						|
            [-4.5681e-04,  3.2570e-03,  7.7929e-03],
 | 
						|
            [ 3.3902e-02,  3.3405e-02,  3.7454e-02],
 | 
						|
            [-2.3001e-02, -2.4877e-03, -3.1033e-03],
 | 
						|
            [ 5.0265e-02,  3.8841e-02,  3.3539e-02],
 | 
						|
            [-4.1018e-03, -1.1095e-03,  1.5859e-03],
 | 
						|
            [-1.2689e-01, -1.3107e-01, -2.1005e-01],
 | 
						|
            [ 2.6276e-02,  1.4189e-02, -3.5963e-03],
 | 
						|
            [-4.8679e-03,  8.8486e-03,  7.8029e-03],
 | 
						|
            [-1.6610e-03, -4.8597e-03, -5.2060e-03],
 | 
						|
            [-2.1010e-03,  2.3610e-03,  9.3796e-03],
 | 
						|
            [-2.2482e-02, -2.1305e-02, -1.5087e-02],
 | 
						|
            [-1.5753e-02, -1.0646e-02, -6.5083e-03],
 | 
						|
            [-4.6975e-03,  5.0288e-03, -6.7390e-03],
 | 
						|
            [ 1.1951e-02,  2.0712e-02,  1.6191e-02],
 | 
						|
            [-6.3704e-03, -8.4827e-03, -9.5483e-03],
 | 
						|
            [ 7.2610e-03, -9.9326e-03, -2.2978e-02],
 | 
						|
            [-9.1904e-04,  6.2882e-03,  9.5720e-03],
 | 
						|
            [-3.7178e-02, -3.7123e-02, -5.6713e-02],
 | 
						|
            [-1.3373e-01, -1.0720e-01, -5.3801e-02],
 | 
						|
            [-5.3702e-03,  8.1256e-03,  8.8397e-03],
 | 
						|
            [-1.5247e-01, -2.1437e-01, -2.1843e-01],
 | 
						|
            [ 3.1441e-02,  7.0335e-03, -9.7541e-03],
 | 
						|
            [ 2.1528e-03, -8.9817e-03, -2.1023e-02],
 | 
						|
            [ 3.8461e-03, -5.8957e-03, -1.5014e-02],
 | 
						|
            [-4.3470e-03, -1.2940e-02, -1.5972e-02],
 | 
						|
            [-5.4781e-03, -1.0842e-02, -3.0204e-03],
 | 
						|
            [-6.5347e-03,  3.0806e-03, -1.0163e-02],
 | 
						|
            [-5.0414e-03, -7.1503e-03, -8.9686e-04],
 | 
						|
            [-8.5851e-03, -2.4351e-03,  1.0674e-03],
 | 
						|
            [-9.0016e-03, -9.6493e-03,  1.5692e-03],
 | 
						|
            [ 5.0914e-03,  1.2099e-02,  1.9968e-02],
 | 
						|
            [ 1.3758e-02,  1.1669e-02,  8.1958e-03],
 | 
						|
            [-1.0518e-02, -1.1575e-02, -4.1307e-03],
 | 
						|
            [-2.8410e-02, -3.1266e-02, -2.2149e-02],
 | 
						|
            [ 2.9336e-03,  3.6511e-02,  1.8717e-02],
 | 
						|
            [-1.6703e-02, -1.6696e-02, -4.4529e-03],
 | 
						|
            [ 4.8818e-02,  4.0063e-02,  8.7410e-03],
 | 
						|
            [-1.5066e-02, -5.7328e-04,  2.9785e-03],
 | 
						|
            [-1.7613e-02, -8.1034e-03,  1.3086e-02],
 | 
						|
            [-9.2633e-03,  1.0803e-02, -6.3489e-03],
 | 
						|
            [ 3.0851e-03,  4.7750e-04,  1.2347e-02],
 | 
						|
            [-2.2785e-02, -2.3043e-02, -2.6005e-02],
 | 
						|
            [-2.4787e-02, -1.5389e-02, -2.2104e-02],
 | 
						|
            [-2.3572e-02,  1.0544e-03,  1.2361e-02],
 | 
						|
            [-7.8915e-03, -1.2271e-03, -6.0968e-03],
 | 
						|
            [-1.1478e-02, -1.2543e-03,  6.2679e-03],
 | 
						|
            [-5.4229e-02,  2.6644e-02,  6.3394e-03],
 | 
						|
            [ 4.4216e-03, -7.3338e-03, -1.0464e-02],
 | 
						|
            [-4.5013e-03,  1.6082e-03,  1.4420e-02],
 | 
						|
            [ 1.3673e-02,  8.8877e-03,  4.1253e-03],
 | 
						|
            [-1.0145e-02,  9.0072e-03,  1.5695e-02],
 | 
						|
            [-5.6234e-03,  1.1847e-03,  8.1261e-03],
 | 
						|
            [-3.7171e-03, -5.3538e-03,  1.2590e-03],
 | 
						|
            [ 2.9476e-02,  2.1424e-02,  3.0424e-02],
 | 
						|
            [-3.4925e-02, -2.4340e-02, -2.5316e-02],
 | 
						|
            [-3.4127e-02, -2.2406e-02, -1.0589e-02],
 | 
						|
            [-1.7342e-02, -1.3249e-02, -1.0719e-02],
 | 
						|
            [-2.1478e-03, -8.6051e-03, -2.9878e-03],
 | 
						|
            [ 1.2089e-03, -4.2391e-03, -6.8569e-03],
 | 
						|
            [ 9.0411e-04, -6.6886e-03, -6.7547e-05],
 | 
						|
            [ 1.6048e-02, -1.0057e-02, -2.8929e-02],
 | 
						|
            [ 1.2290e-03,  1.0163e-02,  1.8861e-02],
 | 
						|
            [ 1.7264e-02,  2.7257e-04,  1.3785e-02],
 | 
						|
            [-1.3482e-02, -3.6427e-03,  6.7481e-04],
 | 
						|
            [ 4.6782e-03, -5.2423e-03,  2.4467e-03],
 | 
						|
            [-5.9113e-03, -6.2244e-03, -1.8162e-03],
 | 
						|
            [ 1.5496e-02,  1.4582e-02,  1.9514e-03],
 | 
						|
            [ 7.4958e-03,  1.5886e-03, -8.2305e-03],
 | 
						|
            [ 1.9086e-02,  1.6360e-03, -3.9674e-03],
 | 
						|
            [-5.7021e-03, -2.7307e-03, -4.1066e-03],
 | 
						|
            [ 1.7450e-03,  1.4602e-02,  2.5794e-02],
 | 
						|
            [-8.2788e-04,  2.2902e-03,  4.5161e-03],
 | 
						|
            [ 1.1632e-02,  8.9193e-03, -7.2813e-03],
 | 
						|
            [ 7.5721e-03,  2.6784e-03,  1.1393e-02],
 | 
						|
            [ 5.1939e-03,  3.6903e-03,  1.4049e-02],
 | 
						|
            [-1.8383e-02, -2.2529e-02, -2.4477e-02],
 | 
						|
            [ 5.8842e-04, -5.7874e-03, -1.4770e-02],
 | 
						|
            [-1.6125e-02, -8.6101e-03, -1.4533e-02],
 | 
						|
            [ 2.0540e-02,  2.0729e-02,  6.4338e-03],
 | 
						|
            [ 3.3587e-03, -1.1226e-02, -1.6444e-02],
 | 
						|
            [-1.4742e-03, -1.0489e-02,  1.7097e-03],
 | 
						|
            [ 2.8130e-02,  2.3546e-02,  3.2791e-02],
 | 
						|
            [-1.8532e-02, -1.2842e-02, -8.7756e-03],
 | 
						|
            [-8.0533e-03, -1.0771e-02, -1.7536e-02],
 | 
						|
            [-3.9009e-03,  1.6150e-02,  3.3359e-02],
 | 
						|
            [-7.4554e-03, -1.4154e-02, -6.1910e-03],
 | 
						|
            [ 3.4734e-03, -1.1370e-02, -1.0581e-02],
 | 
						|
            [ 1.1476e-02,  3.9281e-03,  2.8231e-03],
 | 
						|
            [ 7.1639e-03, -1.4741e-03, -3.8066e-03],
 | 
						|
            [ 2.2250e-03, -8.7552e-03, -9.5719e-03],
 | 
						|
            [ 2.4146e-02,  2.1696e-02,  2.8056e-02],
 | 
						|
            [-5.4365e-03, -2.4291e-02, -1.7802e-02],
 | 
						|
            [ 7.4263e-03,  1.0510e-02,  1.2705e-02],
 | 
						|
            [ 6.2669e-03,  6.2658e-03,  1.9211e-02],
 | 
						|
            [ 1.6378e-02,  9.4933e-03,  6.6971e-03],
 | 
						|
            [ 1.7173e-02,  2.3601e-02,  2.3296e-02],
 | 
						|
            [-1.4568e-02, -9.8279e-03, -1.1556e-02],
 | 
						|
            [ 1.4431e-02,  1.4430e-02,  6.6362e-03],
 | 
						|
            [-6.8230e-03,  1.8863e-02,  1.4555e-02],
 | 
						|
            [ 6.1156e-03,  3.4700e-03, -2.6662e-03],
 | 
						|
            [-2.6983e-03, -5.9402e-03, -9.2276e-03],
 | 
						|
            [ 1.0235e-02,  7.4173e-03, -7.6243e-03],
 | 
						|
            [-1.3255e-02,  1.9322e-02, -9.2153e-04],
 | 
						|
            [ 2.4222e-03, -4.8039e-03, -1.5759e-02],
 | 
						|
            [ 2.6244e-02,  2.5951e-02,  2.0249e-02],
 | 
						|
            [ 1.5711e-02,  1.8498e-02,  2.7407e-03],
 | 
						|
            [-2.1714e-03,  4.7214e-03, -2.2443e-02],
 | 
						|
            [-7.4747e-03,  7.4166e-03,  1.4430e-02],
 | 
						|
            [-8.3906e-03, -7.9776e-03,  9.7927e-03],
 | 
						|
            [ 3.8321e-02,  9.6622e-03, -1.9268e-02],
 | 
						|
            [-1.4605e-02, -6.7032e-03,  3.9675e-03]
 | 
						|
        ]
 | 
						|
        latent_rgb_factors_bias = [-0.0571, -0.1657, -0.2512]    
 | 
						|
 | 
						|
    elif model_family == "hunyuan":
 | 
						|
        latent_channels = 16
 | 
						|
        latent_dimensions = 3
 | 
						|
        scale_factor = 0.476986
 | 
						|
        latent_rgb_factors = [
 | 
						|
            [-0.0395, -0.0331,  0.0445],
 | 
						|
            [ 0.0696,  0.0795,  0.0518],
 | 
						|
            [ 0.0135, -0.0945, -0.0282],
 | 
						|
            [ 0.0108, -0.0250, -0.0765],
 | 
						|
            [-0.0209,  0.0032,  0.0224],
 | 
						|
            [-0.0804, -0.0254, -0.0639],
 | 
						|
            [-0.0991,  0.0271, -0.0669],
 | 
						|
            [-0.0646, -0.0422, -0.0400],
 | 
						|
            [-0.0696, -0.0595, -0.0894],
 | 
						|
            [-0.0799, -0.0208, -0.0375],
 | 
						|
            [ 0.1166,  0.1627,  0.0962],
 | 
						|
            [ 0.1165,  0.0432,  0.0407],
 | 
						|
            [-0.2315, -0.1920, -0.1355],
 | 
						|
            [-0.0270,  0.0401, -0.0821],
 | 
						|
            [-0.0616, -0.0997, -0.0727],
 | 
						|
            [ 0.0249, -0.0469, -0.1703]
 | 
						|
        ]
 | 
						|
 | 
						|
        latent_rgb_factors_bias = [ 0.0259, -0.0192, -0.0761]
 | 
						|
    else:
 | 
						|
        latent_rgb_factors_bias = latent_rgb_factors = None
 | 
						|
    return latent_rgb_factors, latent_rgb_factors_bias |