Rucci-2053183's workspace
Runs
7
State
Notes
User
Tags
Created
Runtime
Sweep
actor_hidden_dim
actor_input_dim
actor_lr
actor_output_dim
baseline_tile_size
critique_hidden_dim
critique_input_dim
critique_lr
current_seq
decoder_output_dim
description
enable_gs_lr_scheduler
enable_on_plateau_phase2
enable_wandb
encoder_hidden_dim
encoder_input_dim
error_window_size
exp_name
exploration
exploration_iter_threshold
gamma
global_iterations
iteration_first_timestep
iteration_others_timestep
lambda1_correction_weight
lambda_correction_weight
lambda_keep_high
lambda_keep_low
lr_cam_c
lr_cam_m
lr_log_scales
lr_logit_opacities
lr_means3D
lr_rgb_colors
lr_seg_colors
lr_unnorm_rotations
max_lr_log_scales
max_lr_logit_opacities
max_lr_means3D
max_lr_rgb_colors
max_lr_seg_colors
max_lr_unnorm_rotations
memory_capacity
memory_sample_batchsize
Crashed
-
rucci-2053183
1m 1s
-
-
-
-
-
-
-
-
-
basketball
12
-
false
false
true
128
14
-
Baseline Fixed LR
-
-
-
20
500
10
1
0.5
-
-
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
-
-
-
-
-
-
-
-
Finished
-
rucci-2053183
9m 28s
-
128
5
0.0001
1
40
128
6
0.0001
basketball
-
Like Before + Enabled keep high and low
-
-
true
-
-
10
RL 06
true
10
0.99
20
500
10
-
-
1
1
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
0.1
0.5
0.016
0.25
0
0.1
100
16
Finished
-
rucci-2053183
10m 2s
-
128
5
0.0001
1
40
128
6
0.0001
basketball
-
Like Before + enable exploration on the first 10 global iters
-
-
true
-
-
10
RL 05
true
10
0.99
20
500
10
-
-
0
0
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
0.1
0.5
0.016
0.25
0
0.1
100
16
Finished
-
rucci-2053183
9m 45s
-
128
5
0.0001
1
40
128
6
0.0001
basketball
-
Like Before new reward formula (slightly improved)
-
-
true
-
-
10
RL 04
false
10
0.99
20
500
10
-
-
0
0
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
0.1
0.5
0.016
0.25
0
0.1
100
16
Finished
-
rucci-2053183
9m 47s
-
128
5
0.0001
1
40
128
6
0.0001
basketball
-
Lr only for one params (means3D), input actor 4 (error, density, tile size, iteration embedding, entropy of tiles of ground truth)
-
-
true
-
-
10
RL 03
false
10
0.99
20
500
10
-
-
0
0
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
0.1
0.5
0.016
0.25
0
0.1
100
16
Finished
-
rucci-2053183
9m 54s
-
128
4
0.0001
1
40
128
5
0.0001
basketball
-
Lr only for one params (means3D), input actor 4 (error, density, tile size, iteration embedding)
-
-
true
-
-
10
RL 02
false
10
0.99
20
500
10
-
-
0
0
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
0.1
0.5
0.016
0.25
0
0.1
100
16
Finished
-
rucci-2053183
9m 51s
-
128
3
0.0001
1
40
128
4
0.0001
basketball
-
Lr only for one params (means3D), input actor 3 (error, density, tile size)
-
-
true
-
-
10
RL 01
false
10
0.99
20
500
10
-
-
0
0
0.0001
0.0001
0.001
0.05
0.00016
0.0025
0
0.001
0.1
0.5
0.016
0.25
0
0.1
100
16
1-7
of 7