== Fri Jul 31 23:11:56 UTC 2026 elapsed=254s == load:0.44 0.22 0.10 2/361 481 == gpu:0 %, 1 MiB, 24564 MiB == log tail: vcodec_image_loader.cpp:737 [NvCodecImageLoader] Initializing: device=0, pool=2, fallback=true [23:11:42.750] [info] nvcodec_image_loader.cpp:785 [NvCodecImageLoader] 2 decoders ready [23:11:42.751] [info] nvcodec_image_loader.cpp:799 [NvCodecImageLoader] Accounted nvImageCodec init VRAM: 6.0 MiB [23:11:42.780] [info] pipelined_image_loader.cpp:376 [PipelinedImageLoader] Shutting down... [23:11:42.801] [info] pipelined_image_loader.cpp:1483 [PipelinedImageLoader] shutdown reconciliation: accepted=6 succeeded=3 failed=0 cancelled=3 mask=3/3/0 depth=0/0/0 normal=0/0/0 [23:11:42.808] [info] pipelined_image_loader.cpp:416 [PipelinedImageLoader] Done: 3 loaded, 6 hits, 0 misses [23:11:42.808] [info] pipelined_image_loader.cpp:420 [PipelinedImageLoader] compressed cache retained: 9 entries, 3.9 MiB RAM, fs=/tmp/LichtFeld/pipeline_cache/ppl_j2k_unified_v1 [23:11:42.808] [info] trainer.cpp:6042 Saving recovery model at iteration 0... [23:11:43.231] [info] checkpoint.cpp:240 Checkpoint saved: /out8/checkpoints/checkpoint.resume (167171 Gaussians, iter 0) Training [█████████████ ] 100% [00m:00s<00m:00s] Initializing... Training [█████████████ ] 100% [00m:00s<00m:00s] Initializing... ✓ Training completed in 0.507s (avg 59174.9 iter/s) ✓ Final splats: 167171 [23:11:43.232] [error] application.cpp:350 Training error: lfs::Error[InvalidArgument/Training] severity=Error retryability=NotRetryable user_message: Distorted images detected. Use --gut or --undistort to train on cameras with distortion. frame[0] ./src/training/trainer.cpp:3759 lfs::training::Trainer::train_step(int, lfs::core::Camera*, lfs::core::Tensor, lfs::training::RenderMode, std::stop_token):: frame[1] ./src/training/trainer.cpp:5572 lfs::Result lfs::training::Trainer::train_step(int, lfs::core::Camera*, lfs::core::Tensor, lfs::training::RenderMode, std::stop_token) — train_step iteration=1 mutation_epoch=0 step_phase=Forward persistent_commit=false frame[2] ./src/training/trainer.cpp:5601 lfs::training::Trainer::train(std::stop_token):: — train iteration=0 mutation_epoch=0 step_phase=TerminalCleanup persistent_commit=false [23:11:43.413] [info] memory_pool.hpp:77 Shutting down CudaMemoryPool... [mark] v8-01-train.txt [FAILED] train