MCPcopy Create free account

hub / github.com/LTH14/mar / functions

Functions149 in github.com/LTH14/mar

⨍ Functions 149◇ Types & classes 25
Method__init__
( self, root: str, )
util/loader.py:30
Method__init__
(self, window_size=20, fmt=None)
util/misc.py:25
Method__init__
(self, delimiter="\t")
util/misc.py:83
Method__init__
(self)
util/misc.py:250
Method__str__
(self)
util/misc.py:73
Method__str__
(self)
util/misc.py:104
Method_basic_init
(module)
models/diffloss.py:195
Method_init_weights
(self, m)
models/mar.py:119
Method_scale_timesteps
(self, t)
diffusion/respace.py:112
Functionadd_weight_decay
(model, weight_decay=1e-5, skip_list=())
util/misc.py:291
Functionadjust_learning_rate
Decay the learning rate with half-cycle cosine after warmup
util/lr_sched.py:4
Functionall_reduce_mean
(x)
util/misc.py:332
Methodavg
(self)
util/misc.py:57
Methodcalc_bpd_loop
Compute the entire variational lower-bound, measured in bits-per-dim, as well as other related quantities. :param model: the
diffusion/gaussian_diffusion.py:809
Methodcondition_mean
(self, cond_fn, *args, **kwargs)
diffusion/respace.py:99
Methodcondition_score
(self, cond_fn, *args, **kwargs)
diffusion/respace.py:102
Methodddim_reverse_sample
Sample x_{t+1} from the model using DDIM reverse ODE.
diffusion/gaussian_diffusion.py:568
Methodddim_sample_loop
Generate samples from the model using DDIM. Same usage as p_sample_loop().
diffusion/gaussian_diffusion.py:606
Methodforward
(self, x)
models/vae.py:28
Methodforward
(self, x)
models/vae.py:45
Methodforward
(self, x, temb)
models/vae.py:92
Methodforward
(self, x)
models/vae.py:134
Methodforward
(self, x)
models/vae.py:245
Methodforward
(self, z)
models/vae.py:365
Methodforward
(self, inputs, disable=True, train=True, optimizer_idx=0)
models/vae.py:486
Methodforward
(self, target, z, mask=None)
models/diffloss.py:26
Methodforward
(self, t)
models/diffloss.py:93
Methodforward
(self, x, y)
models/diffloss.py:124
Methodforward
(self, x, c)
models/diffloss.py:144
Methodforward
(self, imgs, labels)
models/mar.py:240
Methodforward_with_cfg
(self, x, t, c, cfg_scale)
models/diffloss.py:240
Functiongenerate_image
(seed, num_ar_steps, class_labels, cfg_scale, cfg_schedule)
demo/gradio_app.py:44
Functionget_named_beta_schedule
Get a pre-defined beta schedule for the given name. The beta schedule library consists of beta schedules which remain similar in the limi
diffusion/gaussian_diffusion.py:98
Methodglobal_avg
(self)
util/misc.py:62
Functioninit_distributed_mode
(args)
util/misc.py:212
Methodis_vb
(self)
diffusion/gaussian_diffusion.py:54
Methodkl
(self, other=None)
models/vae.py:418
Functionmar_base
(**kwargs)
models/mar.py:332
Functionmar_huge
(**kwargs)
models/mar.py:348
Functionmar_large
(**kwargs)
models/mar.py:340
Methodmedian
(self)
util/misc.py:52
Methodmode
(self)
models/vae.py:446
Methodnll
(self, sample, dims=[1, 2, 3])
models/vae.py:437
Methodp_mean_variance
( self, model, *args, **kwargs )
diffusion/respace.py:89
Methodprocess_xstart
(x)
diffusion/gaussian_diffusion.py:310
Functionsave_model
(args, epoch, model, model_without_ddp, optimizer, loss_scaler, ema_params=None, epoch_name=None)
util/misc.py:306
Methodsynchronize_between_processes
Warning: does not synchronize the deque!
util/misc.py:38
Methodtraining_losses
Compute training losses for a single timestep. :param model: the model to evaluate loss on. :param x_start: the [N x C x ...]
diffusion/gaussian_diffusion.py:719
Methodvalue
(self)
util/misc.py:70
← previous101–149 of 149, ranked by callers