hub / github.com/LTH14/rcg / create_model_and_diffusion

Function create_model_and_diffusion

pixel_generator/guided_diffusion/script_util.py:76–133 · view source on GitHub ↗

(
    image_size,
    class_cond,
    rep_cond,
    rep_dim,
    learn_sigma,
    num_channels,
    num_res_blocks,
    channel_mult,
    num_heads,
    num_head_channels,
    num_heads_upsample,
    attention_resolutions,
    dropout,
    diffusion_steps,
    noise_schedule,
    timestep_respacing,
    use_kl,
    predict_xstart,
    rescale_timesteps,
    rescale_learned_sigmas,
    use_checkpoint,
    use_scale_shift_norm,
    resblock_updown,
    use_fp16,
    use_new_attention_order,
)

Source from the content-addressed store, hash-verified

74
75
76	def create_model_and_diffusion(
77	image_size,
78	class_cond,
79	rep_cond,
80	rep_dim,
81	learn_sigma,
82	num_channels,
83	num_res_blocks,
84	channel_mult,
85	num_heads,
86	num_head_channels,
87	num_heads_upsample,
88	attention_resolutions,
89	dropout,
90	diffusion_steps,
91	noise_schedule,
92	timestep_respacing,
93	use_kl,
94	predict_xstart,
95	rescale_timesteps,
96	rescale_learned_sigmas,
97	use_checkpoint,
98	use_scale_shift_norm,
99	resblock_updown,
100	use_fp16,
101	use_new_attention_order,
102	):
103	model = create_model(
104	image_size,
105	num_channels,
106	num_res_blocks,
107	channel_mult=channel_mult,
108	learn_sigma=learn_sigma,
109	class_cond=class_cond,
110	rep_cond=rep_cond,
111	rep_dim=rep_dim,
112	use_checkpoint=use_checkpoint,
113	attention_resolutions=attention_resolutions,
114	num_heads=num_heads,
115	num_head_channels=num_head_channels,
116	num_heads_upsample=num_heads_upsample,
117	use_scale_shift_norm=use_scale_shift_norm,
118	dropout=dropout,
119	resblock_updown=resblock_updown,
120	use_fp16=use_fp16,
121	use_new_attention_order=use_new_attention_order,
122	)
123	diffusion = create_gaussian_diffusion(
124	steps=diffusion_steps,
125	learn_sigma=learn_sigma,
126	noise_schedule=noise_schedule,
127	use_kl=use_kl,
128	predict_xstart=predict_xstart,
129	rescale_timesteps=rescale_timesteps,
130	rescale_learned_sigmas=rescale_learned_sigmas,
131	timestep_respacing=timestep_respacing,
132	)
133	return model, diffusion

Callers 1

mainFunction · 0.90

Calls 2

create_modelFunction · 0.85

create_gaussian_diffusionFunction · 0.85

Tested by

no test coverage detected