search

wanmeihuali / taichi_3d_gaussian_splatting

An unofficial implementation of paper 3D Gaussian Splatting for Real-Time Radiance Field Rendering by taichi lang.
Apache License 2.0
662 stars 62 forks source link

Backward formula #89

Open wanmeihuali opened 1 year ago

github-actions[bot] commented 1 year ago

Running experiment on sagemaker with git sha 12362802f4f90f1b5e8899a65a0388f2b8755e2c

github-actions[bot] commented 1 year ago

Training job backward-formula-1236280-230710-082042-tat-truck created

github-actions[bot] commented 1 year ago

Training job backward-formula-1236280-230710-082042-tat-truck completed.

Model url: s3://taichi-3d-gaussian-splatting-log/tat-truck/backward-formula-1236280-230710-082042-tat-truck/output/model.tar.gz,

tensorboard output path: s3://taichi-3d-gaussian-splatting-log/tat-truck/backward-formula-1236280-230710-082042-tat-truck/output/output.tar.gz

github-actions[bot] commented 1 year ago

Training job backward-formula-1236280-230710-082042-tat-truck final metrics:

Latest Metrics

val:loss train:ssimloss train:num_valid_points val:psnr train:iteration train:ssim train:loss train:l1loss train:psnr val:ssim
0.060406848788261414 0.14444714784622192 432782.0 24.491334915161133 30000.0 0.8555528521537781 0.05315975472331047 0.030337903648614883 25.80133819580078 0.8402186036109924

Max Metrics

val:5kpsnr val:7kpsnr val:psnr val:ssim train:ssim train:7kpsnr train:5kpsnr val:5kssim train:5kssim train:psnr train:7kssim val:7kssim
22.438127517700195 23.08782196044922 24.491334915161133 0.8402186036109924 0.8555528521537781 25.7239933013916 23.490097045898438 0.780360221862793 0.7940565943717957 25.80133819580078 0.8378960490226746 0.8028812408447266
github-actions[bot] commented 1 year ago

Running experiment on sagemaker with git sha 88da747c674391b180424092a28e990352bdb165

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033341-tat-truck-baseline created

github-actions[bot] commented 1 year ago

Running experiment on sagemaker with git sha 88da747c674391b180424092a28e990352bdb165

github-actions[bot] commented 1 year ago

Running experiment on sagemaker with git sha 88da747c674391b180424092a28e990352bdb165

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033427-stump created

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033429-tat-train-baseline created

github-actions[bot] commented 1 year ago

Running experiment on sagemaker with git sha 88da747c674391b180424092a28e990352bdb165

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-034213-bicycle created

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033427-stump completed.

Model url: s3://taichi-3d-gaussian-splatting-log/stump/backward-formula-88da747-230712-033427-stump/output/model.tar.gz,

tensorboard output path: s3://taichi-3d-gaussian-splatting-log/stump/backward-formula-88da747-230712-033427-stump/output/output.tar.gz

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033427-stump final metrics:

Latest Metrics

train:iteration train:l1loss train:loss train:num_valid_points train:psnr train:ssim train:ssimloss val:loss val:psnr val:ssim
30000.0 0.02205265685915947 0.046825431287288666 420490.0 30.080543518066406 0.8540834784507751 0.14591652154922485 0.08314770460128784 25.551284790039062 0.7309532761573792

Max Metrics

train:5kpsnr train:5kssim train:7kpsnr train:7kssim train:psnr train:ssim val:5kpsnr val:5kssim val:7kpsnr val:7kssim val:psnr val:ssim
25.814929962158203 0.6433213949203491 18.704259872436523 0.5691277384757996 30.080543518066406 0.8540834784507751 23.71395492553711 0.613716185092926 24.206401824951172 0.6512601375579834 25.551284790039062 0.7309532761573792
github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033341-tat-truck-baseline completed.

Model url: s3://taichi-3d-gaussian-splatting-log/tat-truck-baseline/backward-formula-88da747-230712-033341-tat-truck-baseline/output/model.tar.gz,

tensorboard output path: s3://taichi-3d-gaussian-splatting-log/tat-truck-baseline/backward-formula-88da747-230712-033341-tat-truck-baseline/output/output.tar.gz

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033341-tat-truck-baseline final metrics:

Latest Metrics

train:iteration train:l1loss train:loss train:num_valid_points train:psnr train:ssim train:ssimloss val:loss val:psnr val:ssim
30000.0 0.03276287019252777 0.054251909255981445 462186.0 24.921741485595703 0.8597919344902039 0.14020806550979614 0.053635016083717346 25.108041763305664 0.8649178743362427

Max Metrics

train:5kpsnr train:5kssim train:7kpsnr train:7kssim train:psnr train:ssim val:5kpsnr val:5kssim val:7kpsnr val:7kssim val:psnr val:ssim
23.197551727294922 0.8108317255973816 24.970890045166016 0.8554181456565857 24.921741485595703 0.8597919344902039 23.022335052490234 0.8153651356697083 23.773136138916016 0.8345668911933899 25.108041763305664 0.8649178743362427
github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-034213-bicycle completed.

Model url: s3://taichi-3d-gaussian-splatting-log/bicycle/backward-formula-88da747-230712-034213-bicycle/output/model.tar.gz,

tensorboard output path: s3://taichi-3d-gaussian-splatting-log/bicycle/backward-formula-88da747-230712-034213-bicycle/output/output.tar.gz

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-034213-bicycle final metrics:

Latest Metrics

train:iteration train:l1loss train:loss train:num_valid_points train:psnr train:ssim train:ssimloss val:loss val:psnr val:ssim
30000.0 0.049789585173130035 0.10326465219259262 597084.0 19.983814239501953 0.6828351020812988 0.31716489791870117 0.08860167115926743 24.64862060546875 0.7237744927406311

Max Metrics

train:5kpsnr train:5kssim train:7kpsnr train:7kssim train:psnr train:ssim val:5kpsnr val:5kssim val:7kpsnr val:7kssim val:psnr val:ssim
14.369088172912598 0.5607634782791138 23.881305694580078 0.6051310300827026 19.983814239501953 0.6828351020812988 22.625579833984375 0.5718045830726624 23.113313674926758 0.6138212084770203 24.64862060546875 0.7237744927406311
github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033429-tat-train-baseline completed.

Model url: s3://taichi-3d-gaussian-splatting-log/tat-train-baseline/backward-formula-88da747-230712-033429-tat-train-baseline/output/model.tar.gz,

tensorboard output path: s3://taichi-3d-gaussian-splatting-log/tat-train-baseline/backward-formula-88da747-230712-033429-tat-train-baseline/output/output.tar.gz

github-actions[bot] commented 1 year ago

Training job backward-formula-88da747-230712-033429-tat-train-baseline final metrics:

Latest Metrics

train:iteration train:l1loss train:loss train:num_valid_points train:psnr train:ssim train:ssimloss val:loss val:psnr val:ssim
30000.0 0.03160819038748741 0.05087752267718315 690155.0 25.805309295654297 0.8720451593399048 0.12795484066009521 0.10523243248462677 20.447111129760742 0.7820889353752136

Max Metrics

train:5kpsnr train:5kssim train:7kpsnr train:7kssim train:psnr train:ssim val:5kpsnr val:5kssim val:7kpsnr val:7kssim val:psnr val:ssim
23.588279724121094 0.8426668047904968 22.427139282226562 0.7562798261642456 25.805309295654297 0.8720451593399048 19.09138298034668 0.7266534566879272 19.642362594604492 0.7492522597312927 20.447111129760742 0.7820889353752136