Functions — github.com/WendyShang/flare

CodeHub MCPcopy Create free account

hub / github.com/WendyShang/flare / functions

Functions121 in github.com/WendyShang/flare

⨍ Functions 121◇ Types & classes 15

↓ 14 callersMethodlog

(self, L, step, log_freq=LOG_FREQ)

↓ 11 callersMethodlog_param

(self, key, param, step)

↓ 9 callersFunctionnow

data_augs.py:302

↓ 8 callersMethodlog

(self, key, value, step, n=1)

↓ 8 callersMethodlog_histogram

(self, key, histogram, step)

↓ 8 callersFunctionsecs

data_augs.py:304

↓ 5 callersMethodstep

↓ 4 callersMethod_check_input

(self, value, name, center=1, bound=(0, float('inf')), clip_first_on_zero=True)

TransformLayer.py:99

↓ 4 callersMethodreset

↓ 4 callersMethodsample_action

↓ 4 callersMethodtrain

(self, training=True)

↓ 3 callersMethoddump

↓ 3 callersMethodsave

(self, file_name)

↓ 3 callersMethodtransform

TransformLayer.py:179

↓ 2 callersMethod__init__

( self, obs_shape, action_shape, hidden_dim, encoder_type, encoder_feature_dim, log_std_min, l

↓ 2 callersMethod__init__

(self, obs_shape, feature_dim, channels=[16, 32, 32], num_layers=2, num_filters=32,

↓ 2 callersMethod_get_obs

↓ 2 callersMethod_get_qpos

↓ 2 callersFunctioncenter_crop_images

(image, output_size)

↓ 2 callersMethodforward_conv

(self, obs,flatten=True)

↓ 2 callersMethodforward_conv

(self, obs,flatten=True)

↓ 2 callersMethodload

(self, model_dir, step)

↓ 2 callersMethodlog_image

(self, key, image, step)

↓ 2 callersFunctionmake_encoder

( encoder_type, obs_shape, feature_dim, num_layers, num_filters, output_logits=False, channels=[

↓ 2 callersMethodselect_action

↓ 2 callersFunctiontie_weights

↓ 2 callersMethodupdate

(self, value, n=1)

↓ 1 callersMethod__init__

(self, *models)

↓ 1 callersMethod_dump_to_console

(self, data, prefix)

↓ 1 callersMethod_dump_to_file

↓ 1 callersMethod_format

(self, key, value, ty)

↓ 1 callersMethod_prime_meters

↓ 1 callersMethod_try_sw_log

(self, key, value, step)

↓ 1 callersMethod_try_sw_log_histogram

(self, key, histogram, step)

↓ 1 callersMethod_try_sw_log_image

(self, key, image, step)

↓ 1 callersMethod_try_sw_log_video

(self, key, frames, step)

↓ 1 callersMethodadd

(self, obs, action, reward, next_obs, done, eps_done)

↓ 1 callersMethodcopy_conv_weights_from

Tie convolutional layers

↓ 1 callersFunctionevaluate

(env, agent, video, num_episodes, L, step, args)

train_ddp.py:100

↓ 1 callersFunctiongaussian_logprob

Compute Gaussian log probability.

↓ 1 callersFunctiongrayscale

data_augs.py:82

↓ 1 callersMethodinit

(self, enabled=True)

↓ 1 callersFunctionmain

train_ddp.py:219

↓ 1 callersFunctionmake_agent

(obs_shape, action_shape, action_range, args, device, image_channel=3)

train_ddp.py:181

↓ 1 callersFunctionparse_args

train_ddp.py:35

↓ 1 callersFunctionrandom_color_jitter

inputs np array outputs tensor

data_augs.py:279

↓ 1 callersFunctionrandom_convolution

random covolution in "network randomization" (imbs): B x (C x stack) x H x W, note: imgs should be normalized and torch tensor

data_augs.py:249

↓ 1 callersFunctionrandom_crop

args: imgs: np.array shape (B,C,H,W) out: output size (e.g. 84) returns np.array

data_augs.py:64

↓ 1 callersFunctionrandom_cutout

args: imgs: np.array shape (B,C,H,W) min / max cut: int, min / max size of cutout returns np.array

data_augs.py:127

↓ 1 callersFunctionrandom_cutout_color

args: imgs: shape (B,C,H,W) out: output size (e.g. 84)

data_augs.py:147

↓ 1 callersFunctionrandom_flip

args: imgs: torch.tensor shape (B,C,H,W) device: cpu or gpu, p: prob of applying aug, returns torch.tensor

data_augs.py:173

↓ 1 callersFunctionrandom_grayscale

args: imgs: torch.tensor shape (B,C,H,W) device: cpu or cuda returns torch.tensor

data_augs.py:97

↓ 1 callersFunctionrandom_rotation

args: imgs: torch.tensor shape (B,C,H,W) device: str, cpu or gpu, p: float, prob of applying aug, returns to

data_augs.py:206

↓ 1 callersMethodrecord

↓ 1 callersFunctionrun_eval_loop

(sample_stochastically=True)

train_ddp.py:103

↓ 1 callersMethodsample_proprio

↓ 1 callersMethodsample_rad

(self,aug_funcs)

↓ 1 callersMethodsave

(self, save_dir)

↓ 1 callersFunctionsquash

Apply squashing function. See appendix C from https://arxiv.org/pdf/1812.05905.pdf.

↓ 1 callersMethodupdate_actor_and_alpha

(self, obs, L, step)

↓ 1 callersMethodupdate_critic

(self, obs, action, reward, next_obs, not_done, L, step)

↓ 1 callersMethodvalue

Method__enter__

Method__getitem__

(self, obs_shape, action_shape, capacity, batch_size, device, image_size=84, pre_image_size=100, trans

(self, file_name, formating)

(self, log_dir, use_tb=True, config='rl')

(self, obs_dim, action_dim, hidden_dim)

( self, obs_shape, action_shape, hidden_dim, encoder_type, encoder_feature_dim, num_layers, nu

( self, obs_shape, action_shape, device, hidden_dim=256, disco

(self, dir_name, height=256, width=256, camera_id=0, fps=30)

(self, obs_shape, feature_dim, num_layers, num_filters,*args)

(self, obs_shape, feature_dim, channels=[16, 32, 32], num_layers=2, num_filters=32,

(self, brightness=0, contrast=0, saturation=0, hue=0, p=0, batch_size=128, stack_size=3)

TransformLayer.py:88

Methodadjust_brightness

Args: x: torch tensor img (hsv type) Factor: torch tensor with same length as x

TransformLayer.py:143

Methodadjust_contrast

Args: x: torch tensor img (rgb type) Factor: torch tensor with same length as x 0 gives g

TransformLayer.py:117

Methodadjust_hue

TransformLayer.py:133

Methodadjust_saturate

Args: x: torch tensor img (hsv type) Factor: torch tensor with same length as x

TransformLayer.py:161

Functioncenter_crop_image

(image, output_size)

Functioncenter_crop_images

(image, output_size)

train_ddp.py:25

Functioncenter_translate

Functioncenter_window

(imgs, window_size, in_place=False)

data_augs.py:29

Methodcopy_conv_weights_from

Methodcopy_conv_weights_from

Tie convolutional layers

(self, step, prefix)

( self, obs, compute_pi=True, compute_log_pi=True, detach_encoder=False )

(self, obs, action)

(self, obs, action, detach_encoder=False)

(self, obs, detach=False)

(self, obs, detach=False)

(self, obs, detach=False)

TransformLayer.py:193

Functionhsv2rgb

TransformLayer.py:41

(self, save_dir)

(self, key, value, n=1)

next →1–100 of 121, ranked by callers