diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01400.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01400.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2c5fe70e096c3dd58e4608a0a6f8db2bd5c4aa59 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01400.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47dc357dbe9c48bf94c06cd37b97d9dc90a4d87caf423249551a1f447e74d61b +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01500.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01500.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c586842488b88a187d0503255e58f36b38d8929 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01500.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:62511507db915d4d56937f88cc32450230abc4391a125236887715c8cad3eca1 +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01600.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01600.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d5e2af13b063f38828a20fcc8bd8d2f478203ba6 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01600.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4434234f986e8aff06efe7607bb27b98e05bcd39bb157fe7941432fd3bbec5e +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01700.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01700.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..faece964551bcb9dfb13362e464072359d724af8 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01700.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:439a2c195d79c63524e02a71a2a0b9b17165a01c366828ff87f1662c96e1ec26 +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01800.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01800.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..272a60763f2dc40aaf5d9cffee2210e902acd6cc --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01800.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8de414a009571876698aa394d148ff4f0b37af78f14d969d78a1cb5d5d07d4fc +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01900.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01900.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d084ee9205fa11bcfbf124f190ef2556d089ebd6 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_01900.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb120e07fd62f11ed04cc35299918acfd03f1b3949a619c97bb63fa6367dd7b9 +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_02000.safetensors b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_02000.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..de6b1bb398718a7a3c9ce2e8a5fd797c3d07b144 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/predictor_step_02000.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:165402a63b55269c3163a257f2cb465f8fa3f677405f38c2e1e081b81159f911 +size 654604408 diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/swanlab/run-20260719_050608-8mjozf71/debug/debug.log b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/swanlab/run-20260719_050608-8mjozf71/debug/debug.log new file mode 100644 index 0000000000000000000000000000000000000000..3a32a4f21768e1dae96b8594742795c79550f5ab --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/swanlab/run-20260719_050608-8mjozf71/debug/debug.log @@ -0,0 +1 @@ +Diagnostic log bound to file: /mnt/s3files/s3-us-west2-default/zoubin/cz/projects/HY-WorldPlay-DEV-Predictor/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/swanlab/run-20260719_050608-8mjozf71/debug/debug.log diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/torchrun.log b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/torchrun.log new file mode 100644 index 0000000000000000000000000000000000000000..e3b5ab77a45f8347728cfa53180f3c9bc6794602 --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/torchrun.log @@ -0,0 +1,279 @@ +{ + "manifest": "/mnt/local_nvme/zoubin/cz/hyworldplay_predictor_prefeature_v4/train_eval_manifest.jsonl", + "teacher_checkpoint": "/mnt/s3files/s3-us-west2-default/zoubin/cz/checkpoints/hy_worldplay/huggingface/hub/models--tencent--HY-WorldPlay/snapshots/f4c29235647707b571479a69b569e4166f9f5bf8/ar_distilled_action_model/diffusion_pytorch_model.safetensors", + "output_dir": "/mnt/s3files/s3-us-west2-default/zoubin/cz/projects/HY-WorldPlay-DEV-Predictor/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab", + "source_blocks": [ + 0, + 53 + ], + "predictor_variant": "disca", + "max_steps": 2000, + "batch_size": 16, + "gradient_accumulation_steps": 1, + "block_warmup_steps": 100, + "scheduler_warmup_steps": 100, + "fusion_lr": 0.0001, + "blocks_lr": 1e-05, + "weight_decay": 0.01, + "hidden_weight": 0.1, + "velocity_weight": 1.0, + "grad_clip": 1.0, + "log_every": 10, + "save_every": 100, + "keep_snapshots": 20, + "seed": 0, + "num_workers": 0, + "max_records": null, + "bucket_by_context": false, + "gradient_checkpointing": true, + "save_optimizer_state": true, + "resume": null, + "swanlab_project": "HY-WorldPlay-Predictor-DisCa", + "swanlab_workspace": null, + "swanlab_experiment": "predictor-disca-0-53-b16x8-2000steps", + "swanlab_mode": "cloud", + "predictor_version": "predictor_disca", + "schema_version": "predictor_disca_text_kv_v1", + "supervision_pairs": [ + [ + 0, + 1 + ], + [ + 1, + 2 + ], + [ + 2, + 3 + ] + ], + "records": 2800, + "dataset_pairs": 8400, + "steps_per_epoch_per_rank": 65, + "world_size": 8, + "per_rank_batch_size": 16, + "global_batch_size": 128, + "trainable_parameters": 176241152, + "frozen_projector_parameters": 0, + "dtype": "bfloat16", + "gpu": "NVIDIA H100 80GB HBM3" +} +swanlab: Currently logged in as: ZhuoChen to https://api.swanlab.cn. Use +`swanlab login --relogin` to force relogin +swanlab: Tracking run with swanlab version 0.9.0 +swanlab: 💾 Run data saved at +/mnt/s3files/s3-us-west2-default/zoubin/cz/projects/HY-WorldPlay-DEV-Predictor/c +heckpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/swanlab/run-20260719_050608-8mjozf +71 +swanlab: 📁 View project at +https://swanlab.cn/@ZhuoChen/HY-WorldPlay-Predictor-DisCa +swanlab: 🚀 View run predictor-disca-0-53-b16x8-2000steps at +https://swanlab.cn/@ZhuoChen/HY-WorldPlay-Predictor-DisCa/runs/8mjozf71 +{"swanlab_run_id": "8mjozf71"} +{"avg_context_frames": 14.71875, "avg_micro_time_s": 6.3683194299992465, "epoch": 1, "global_step": 1, "grad_norm_max": 0.01446533203125, "hidden_mse": 0.016903931158594787, "loss": 0.014591015642508864, "lr_blocks": 0.0, "lr_fusion": 2.0000000000000003e-06, "peak_memory_gib": 30.525243759155273, "velocity_mse": 0.012900622386951} +{"avg_context_frames": 14.180555555555555, "avg_micro_time_s": 1.6973763332352974, "epoch": 1, "global_step": 10, "grad_norm_max": 0.01385498046875, "hidden_mse": 0.016849822998564277, "loss": 0.013660540364475712, "lr_blocks": 0.0, "lr_fusion": 1.1000000000000001e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011975558080141328} +{"avg_context_frames": 14.4, "avg_micro_time_s": 1.706131999749414, "epoch": 1, "global_step": 20, "grad_norm_max": 0.0108642578125, "hidden_mse": 0.017140438850037753, "loss": 0.014446394855622203, "lr_blocks": 0.0, "lr_fusion": 2.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.012732350960141048} +{"avg_context_frames": 14.26875, "avg_micro_time_s": 1.6902024415863708, "epoch": 1, "global_step": 30, "grad_norm_max": 0.01409912109375, "hidden_mse": 0.016241216578055172, "loss": 0.013358410890214147, "lr_blocks": 0.0, "lr_fusion": 3.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011734289216110484} +{"avg_context_frames": 14.134375, "avg_micro_time_s": 1.6864288714114082, "epoch": 1, "global_step": 40, "grad_norm_max": 0.01904296875, "hidden_mse": 0.01575567766558379, "loss": 0.013327148038661108, "lr_blocks": 0.0, "lr_fusion": 4.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011751580244163052} +{"avg_context_frames": 14.484375, "avg_micro_time_s": 1.6857911082748616, "epoch": 1, "global_step": 50, "grad_norm_max": 0.0234375, "hidden_mse": 0.01519326661946252, "loss": 0.012959708529524506, "lr_blocks": 0.0, "lr_fusion": 5.1000000000000006e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011440381826832892} +{"avg_context_frames": 14.253125, "avg_micro_time_s": 1.6837490774865729, "epoch": 1, "global_step": 60, "grad_norm_max": 0.01708984375, "hidden_mse": 0.014341028686612845, "loss": 0.012393195397453384, "lr_blocks": 0.0, "lr_fusion": 6.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010959092515986413} +{"avg_context_frames": 14.234375, "avg_micro_time_s": 1.6833479166634788, "epoch": 2, "global_step": 70, "grad_norm_max": 0.0225830078125, "hidden_mse": 0.015120381687302143, "loss": 0.013338346488308162, "lr_blocks": 0.0, "lr_fusion": 7.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01182630832772702} +{"avg_context_frames": 14.403125000000001, "avg_micro_time_s": 1.6783960828634008, "epoch": 2, "global_step": 80, "grad_norm_max": 0.014404296875, "hidden_mse": 0.014159502333495767, "loss": 0.012736178463092074, "lr_blocks": 0.0, "lr_fusion": 8.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011320228286786006} +{"avg_context_frames": 14.053125000000001, "avg_micro_time_s": 1.6775284700990598, "epoch": 2, "global_step": 90, "grad_norm_max": 0.0281982421875, "hidden_mse": 0.01400059668812901, "loss": 0.012482481659390032, "lr_blocks": 0.0, "lr_fusion": 9.1e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011082421953324229} +{"avg_context_frames": 14.5625, "avg_micro_time_s": 1.678686482399644, "epoch": 2, "global_step": 100, "grad_norm_max": 0.01348876953125, "hidden_mse": 0.013929923344403506, "loss": 0.012565440696198494, "lr_blocks": 1e-05, "lr_fusion": 0.0001, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011172448418801651} +{"avg_context_frames": 14.175, "avg_micro_time_s": 1.7038727414888855, "epoch": 2, "global_step": 110, "grad_norm_max": 0.017822265625, "hidden_mse": 0.01416516478639096, "loss": 0.012822794512612746, "lr_blocks": 9.999316524962347e-06, "lr_fusion": 9.999316524962345e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011406278068898247} +{"avg_context_frames": 14.340625000000001, "avg_micro_time_s": 1.690337468202415, "epoch": 2, "global_step": 120, "grad_norm_max": 0.0233154296875, "hidden_mse": 0.013877624762244523, "loss": 0.012641063879709692, "lr_blocks": 9.99726628670463e-06, "lr_fusion": 9.997266286704631e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011253301409306006} +{"avg_context_frames": 14.09375, "avg_micro_time_s": 1.693360372800089, "epoch": 2, "global_step": 130, "grad_norm_max": 0.0167236328125, "hidden_mse": 0.014193043851992116, "loss": 0.013191334606381134, "lr_blocks": 9.993849845741525e-06, "lr_fusion": 9.993849845741524e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01177203023689799} +{"avg_context_frames": 14.20625, "avg_micro_time_s": 1.701355368750592, "epoch": 3, "global_step": 140, "grad_norm_max": 0.0213623046875, "hidden_mse": 0.013515560410451145, "loss": 0.012643122443114408, "lr_blocks": 9.989068136093873e-06, "lr_fusion": 9.989068136093873e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011291566424188205} +{"avg_context_frames": 14.30625, "avg_micro_time_s": 1.6922673213863164, "epoch": 3, "global_step": 150, "grad_norm_max": 0.0234375, "hidden_mse": 0.013559077697573231, "loss": 0.012912513717310503, "lr_blocks": 9.98292246503335e-06, "lr_fusion": 9.98292246503335e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011556605889927596} +{"avg_context_frames": 14.409375, "avg_micro_time_s": 1.6912575177138933, "epoch": 3, "global_step": 160, "grad_norm_max": 0.021728515625, "hidden_mse": 0.012874374637613074, "loss": 0.011889695067657158, "lr_blocks": 9.975414512725058e-06, "lr_fusion": 9.975414512725057e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010602257552091033} +{"avg_context_frames": 14.284375, "avg_micro_time_s": 1.6881116359512818, "epoch": 3, "global_step": 170, "grad_norm_max": 0.0235595703125, "hidden_mse": 0.013235451170476155, "loss": 0.012821374350460248, "lr_blocks": 9.966546331768192e-06, "lr_fusion": 9.966546331768191e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01149782930733636} +{"avg_context_frames": 14.55625, "avg_micro_time_s": 1.6881083145017328, "epoch": 3, "global_step": 180, "grad_norm_max": 0.0263671875, "hidden_mse": 0.01309632696211338, "loss": 0.01252620859304443, "lr_blocks": 9.956320346634877e-06, "lr_fusion": 9.956320346634876e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011216575902653859} +{"avg_context_frames": 14.293750000000001, "avg_micro_time_s": 1.6891175489261514, "epoch": 3, "global_step": 190, "grad_norm_max": 0.0162353515625, "hidden_mse": 0.012802221107995138, "loss": 0.012409800564637408, "lr_blocks": 9.944739353007344e-06, "lr_fusion": 9.944739353007344e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01112957840668969} +{"avg_context_frames": 13.793750000000001, "avg_micro_time_s": 1.6870502156383738, "epoch": 4, "global_step": 200, "grad_norm_max": 0.0150146484375, "hidden_mse": 0.012918339006137104, "loss": 0.012187451863428579, "lr_blocks": 9.931806517013612e-06, "lr_fusion": 9.931806517013612e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010895617940695957} +{"avg_context_frames": 14.200000000000001, "avg_micro_time_s": 1.677001495748482, "epoch": 4, "global_step": 210, "grad_norm_max": 0.0166015625, "hidden_mse": 0.012752376409480348, "loss": 0.012700100988149644, "lr_blocks": 9.917525374361913e-06, "lr_fusion": 9.917525374361912e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.0114248632802628} +{"avg_context_frames": 14.387500000000001, "avg_micro_time_s": 1.6790696400625167, "epoch": 4, "global_step": 220, "grad_norm_max": 0.0252685546875, "hidden_mse": 0.012769673491129653, "loss": 0.012019666697597131, "lr_blocks": 9.901899829374048e-06, "lr_fusion": 9.901899829374047e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01074269930832088} +{"avg_context_frames": 14.134375, "avg_micro_time_s": 1.6766690597360139, "epoch": 4, "global_step": 230, "grad_norm_max": 0.021240234375, "hidden_mse": 0.01252085343003273, "loss": 0.012354082963429392, "lr_blocks": 9.884934153917998e-06, "lr_fusion": 9.884934153917997e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011101997626246885} +{"avg_context_frames": 14.5875, "avg_micro_time_s": 1.6781999800128689, "epoch": 4, "global_step": 240, "grad_norm_max": 0.0157470703125, "hidden_mse": 0.011941308557288722, "loss": 0.01208062154473737, "lr_blocks": 9.86663298624003e-06, "lr_fusion": 9.86663298624003e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010886490694247186} +{"avg_context_frames": 14.359375, "avg_micro_time_s": 1.672726408374365, "epoch": 4, "global_step": 250, "grad_norm_max": 0.01422119140625, "hidden_mse": 0.013191320811165497, "loss": 0.012905225792201236, "lr_blocks": 9.847001329696653e-06, "lr_fusion": 9.847001329696653e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011586093687219546} +{"avg_context_frames": 14.171875, "avg_micro_time_s": 1.6619861684026547, "epoch": 4, "global_step": 260, "grad_norm_max": 0.01611328125, "hidden_mse": 0.012081997399218381, "loss": 0.011730625136988238, "lr_blocks": 9.826044551386743e-06, "lr_fusion": 9.826044551386744e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010522425378439948} +{"avg_context_frames": 14.453125, "avg_micro_time_s": 1.6650776747635974, "epoch": 5, "global_step": 270, "grad_norm_max": 0.0198974609375, "hidden_mse": 0.012529171502683312, "loss": 0.012534417357528582, "lr_blocks": 9.803768380684242e-06, "lr_fusion": 9.803768380684242e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011281500186305493} +{"avg_context_frames": 14.53125, "avg_micro_time_s": 1.6554049376394688, "epoch": 5, "global_step": 280, "grad_norm_max": 0.016357421875, "hidden_mse": 0.012344440072774888, "loss": 0.012352285004453734, "lr_blocks": 9.780178907671788e-06, "lr_fusion": 9.780178907671789e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011117841006489472} +{"avg_context_frames": 14.084375000000001, "avg_micro_time_s": 1.6595323448000272, "epoch": 5, "global_step": 290, "grad_norm_max": 0.021240234375, "hidden_mse": 0.012424418586306274, "loss": 0.011968210444319994, "lr_blocks": 9.755282581475769e-06, "lr_fusion": 9.755282581475769e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010725768603151664} +{"avg_context_frames": 14.171875, "avg_micro_time_s": 1.6620940328637517, "epoch": 5, "global_step": 300, "grad_norm_max": 0.0185546875, "hidden_mse": 0.01225031307549216, "loss": 0.012034611008130015, "lr_blocks": 9.729086208503174e-06, "lr_fusion": 9.729086208503174e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010809579689521343} +{"avg_context_frames": 14.259375, "avg_micro_time_s": 1.6782162721123315, "epoch": 5, "global_step": 310, "grad_norm_max": 0.02197265625, "hidden_mse": 0.01193973511690274, "loss": 0.011788620107108728, "lr_blocks": 9.701596950580807e-06, "lr_fusion": 9.701596950580806e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010594646586105228} +{"avg_context_frames": 14.334375000000001, "avg_micro_time_s": 1.6686666865873123, "epoch": 5, "global_step": 320, "grad_norm_max": 0.0155029296875, "hidden_mse": 0.012210708390921355, "loss": 0.01247087437659502, "lr_blocks": 9.672822322997305e-06, "lr_fusion": 9.672822322997305e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011249803507234903} +{"avg_context_frames": 14.221875, "avg_micro_time_s": 1.6616732870981652, "epoch": 6, "global_step": 330, "grad_norm_max": 0.014892578125, "hidden_mse": 0.012031648849369959, "loss": 0.011725777987157926, "lr_blocks": 9.642770192448537e-06, "lr_fusion": 9.642770192448536e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010522613144712523} +{"avg_context_frames": 14.475000000000001, "avg_micro_time_s": 1.663249169124174, "epoch": 6, "global_step": 340, "grad_norm_max": 0.0179443359375, "hidden_mse": 0.01219728536088951, "loss": 0.012152914865873755, "lr_blocks": 9.611448774886925e-06, "lr_fusion": 9.611448774886924e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01093318634084426} +{"avg_context_frames": 13.825000000000001, "avg_micro_time_s": 1.6703412917369862, "epoch": 6, "global_step": 350, "grad_norm_max": 0.01373291015625, "hidden_mse": 0.012240116082830356, "loss": 0.012266952311620116, "lr_blocks": 9.578866633275289e-06, "lr_fusion": 9.578866633275288e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01104294071556069} +{"avg_context_frames": 14.3, "avg_micro_time_s": 1.6716963182378095, "epoch": 6, "global_step": 360, "grad_norm_max": 0.01300048828125, "hidden_mse": 0.012173148326110095, "loss": 0.012153307441622019, "lr_blocks": 9.545032675245814e-06, "lr_fusion": 9.545032675245813e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010935992584563792} +{"avg_context_frames": 14.293750000000001, "avg_micro_time_s": 1.6808232290612068, "epoch": 6, "global_step": 370, "grad_norm_max": 0.01397705078125, "hidden_mse": 0.011603472143178806, "loss": 0.011518231267109514, "lr_blocks": 9.509956150664796e-06, "lr_fusion": 9.509956150664796e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010357884061522783} +{"avg_context_frames": 14.615625000000001, "avg_micro_time_s": 1.6778030030618538, "epoch": 6, "global_step": 380, "grad_norm_max": 0.02099609375, "hidden_mse": 0.01206234356504865, "loss": 0.01220254950458184, "lr_blocks": 9.473646649103819e-06, "lr_fusion": 9.473646649103818e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010996315145166592} +{"avg_context_frames": 14.175, "avg_micro_time_s": 1.6769782853614743, "epoch": 6, "global_step": 390, "grad_norm_max": 0.015869140625, "hidden_mse": 0.011909931572154165, "loss": 0.011906098818872125, "lr_blocks": 9.43611409721806e-06, "lr_fusion": 9.43611409721806e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010715105681447313} +{"avg_context_frames": 14.318750000000001, "avg_micro_time_s": 1.6752818771481544, "epoch": 7, "global_step": 400, "grad_norm_max": 0.01300048828125, "hidden_mse": 0.011655999638605863, "loss": 0.011750627792207525, "lr_blocks": 9.397368756032445e-06, "lr_fusion": 9.397368756032445e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010585027851630003} +{"avg_context_frames": 14.328125, "avg_micro_time_s": 1.677450421390313, "epoch": 7, "global_step": 410, "grad_norm_max": 0.0223388671875, "hidden_mse": 0.01190967960865237, "loss": 0.011954425496514888, "lr_blocks": 9.357421218136387e-06, "lr_fusion": 9.357421218136386e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010763457557186485} +{"avg_context_frames": 14.20625, "avg_micro_time_s": 1.6733501724360396, "epoch": 7, "global_step": 420, "grad_norm_max": 0.01556396484375, "hidden_mse": 0.011672888038447127, "loss": 0.011563985573593527, "lr_blocks": 9.31628240478787e-06, "lr_fusion": 9.316282404787871e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010396696761017666} +{"avg_context_frames": 14.321875, "avg_micro_time_s": 1.6728075357379568, "epoch": 7, "global_step": 430, "grad_norm_max": 0.0186767578125, "hidden_mse": 0.011731227574637161, "loss": 0.011881092737894505, "lr_blocks": 9.273963562927695e-06, "lr_fusion": 9.273963562927695e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01070796991698444} +{"avg_context_frames": 14.159375, "avg_micro_time_s": 1.6735653286392336, "epoch": 7, "global_step": 440, "grad_norm_max": 0.015869140625, "hidden_mse": 0.011862598883453757, "loss": 0.01177286847960204, "lr_blocks": 9.230476262104678e-06, "lr_fusion": 9.230476262104677e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010586608608718961} +{"avg_context_frames": 14.318750000000001, "avg_micro_time_s": 1.6664162646746263, "epoch": 7, "global_step": 450, "grad_norm_max": 0.0206298828125, "hidden_mse": 0.012222287594340742, "loss": 0.012133395049022512, "lr_blocks": 9.185832391312644e-06, "lr_fusion": 9.185832391312644e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010911166266305374} +{"avg_context_frames": 14.3, "avg_micro_time_s": 1.6672153513387458, "epoch": 8, "global_step": 460, "grad_norm_max": 0.0126953125, "hidden_mse": 0.011892647773493081, "loss": 0.012170521350344644, "lr_blocks": 9.140044155740102e-06, "lr_fusion": 9.140044155740101e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010981256602099166} +{"avg_context_frames": 14.3625, "avg_micro_time_s": 1.66273293293998, "epoch": 8, "global_step": 470, "grad_norm_max": 0.019775390625, "hidden_mse": 0.011298886651638897, "loss": 0.011162232892820612, "lr_blocks": 9.093124073433464e-06, "lr_fusion": 9.093124073433463e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010032344225328416} +{"avg_context_frames": 14.3125, "avg_micro_time_s": 1.6722241195380776, "epoch": 8, "global_step": 480, "grad_norm_max": 0.0159912109375, "hidden_mse": 0.011623429507017136, "loss": 0.011663176794536413, "lr_blocks": 9.045084971874738e-06, "lr_fusion": 9.045084971874738e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010500833892729135} +{"avg_context_frames": 14.26875, "avg_micro_time_s": 1.6604970885855437, "epoch": 8, "global_step": 490, "grad_norm_max": 0.019775390625, "hidden_mse": 0.0121043776802253, "loss": 0.012028527632355691, "lr_blocks": 8.995939984474624e-06, "lr_fusion": 8.995939984474624e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010818089940585196} +{"avg_context_frames": 14.175, "avg_micro_time_s": 1.671721852789051, "epoch": 8, "global_step": 500, "grad_norm_max": 0.0252685546875, "hidden_mse": 0.01200883725541644, "loss": 0.012602345040068031, "lr_blocks": 8.94570254698197e-06, "lr_fusion": 8.945702546981969e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011401461373316124} +{"avg_context_frames": 14.515625, "avg_micro_time_s": 1.66979133807763, "epoch": 8, "global_step": 510, "grad_norm_max": 0.01287841796875, "hidden_mse": 0.011961949995020406, "loss": 0.011663943075109274, "lr_blocks": 8.894386393810563e-06, "lr_fusion": 8.894386393810563e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010467748151859269} +{"avg_context_frames": 14.125, "avg_micro_time_s": 1.6650106007375145, "epoch": 8, "global_step": 520, "grad_norm_max": 0.01446533203125, "hidden_mse": 0.012165928492322564, "loss": 0.012186324427602813, "lr_blocks": 8.842005554284296e-06, "lr_fusion": 8.842005554284296e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010969731508521364} +{"avg_context_frames": 14.421875, "avg_micro_time_s": 1.6684493572865904, "epoch": 9, "global_step": 530, "grad_norm_max": 0.016845703125, "hidden_mse": 0.011617637728340924, "loss": 0.011304739100160078, "lr_blocks": 8.788574348801676e-06, "lr_fusion": 8.788574348801675e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010142975387861953} +{"avg_context_frames": 14.340625000000001, "avg_micro_time_s": 1.6649726299736358, "epoch": 9, "global_step": 540, "grad_norm_max": 0.0213623046875, "hidden_mse": 0.012169572908896954, "loss": 0.012478833075147123, "lr_blocks": 8.734107384920771e-06, "lr_fusion": 8.73410738492077e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011261875747004525} +{"avg_context_frames": 14.025, "avg_micro_time_s": 1.6626613621527213, "epoch": 9, "global_step": 550, "grad_norm_max": 0.0189208984375, "hidden_mse": 0.011851134733296931, "loss": 0.011942122137406842, "lr_blocks": 8.67861955336566e-06, "lr_fusion": 8.678619553365659e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010757008683867753} +{"avg_context_frames": 14.496875000000001, "avg_micro_time_s": 1.6742402617110201, "epoch": 9, "global_step": 560, "grad_norm_max": 0.0191650390625, "hidden_mse": 0.011568118509603665, "loss": 0.01129704446066171, "lr_blocks": 8.622126023955446e-06, "lr_fusion": 8.622126023955446e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010140232573030518} +{"avg_context_frames": 14.231250000000001, "avg_micro_time_s": 1.6741461727222486, "epoch": 9, "global_step": 570, "grad_norm_max": 0.0146484375, "hidden_mse": 0.011797175847459585, "loss": 0.01185271711437963, "lr_blocks": 8.564642241456986e-06, "lr_fusion": 8.564642241456986e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010672999499365688} +{"avg_context_frames": 14.106250000000001, "avg_micro_time_s": 1.6786558078725649, "epoch": 9, "global_step": 580, "grad_norm_max": 0.0166015625, "hidden_mse": 0.011530527414288373, "loss": 0.011785041278926656, "lr_blocks": 8.506183921362443e-06, "lr_fusion": 8.506183921362443e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010631988535169513} +{"avg_context_frames": 14.543750000000001, "avg_micro_time_s": 1.6829396546374484, "epoch": 10, "global_step": 590, "grad_norm_max": 0.0235595703125, "hidden_mse": 0.011740072624525056, "loss": 0.011994902434526012, "lr_blocks": 8.446767045592829e-06, "lr_fusion": 8.44676704559283e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010820895194774494} +{"avg_context_frames": 14.109375, "avg_micro_time_s": 1.671282860262727, "epoch": 10, "global_step": 600, "grad_norm_max": 0.0191650390625, "hidden_mse": 0.011970765586011112, "loss": 0.012209405086468906, "lr_blocks": 8.386407858128707e-06, "lr_fusion": 8.386407858128706e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01101232846849598} +{"avg_context_frames": 14.225000000000001, "avg_micro_time_s": 1.6784421321997798, "epoch": 10, "global_step": 610, "grad_norm_max": 0.0146484375, "hidden_mse": 0.011483470286475495, "loss": 0.011486795824021102, "lr_blocks": 8.325122860569241e-06, "lr_fusion": 8.32512286056924e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010338448768015951} +{"avg_context_frames": 14.196875, "avg_micro_time_s": 1.675524012938331, "epoch": 10, "global_step": 620, "grad_norm_max": 0.01458740234375, "hidden_mse": 0.011668070487212391, "loss": 0.011516804079292342, "lr_blocks": 8.262928807620843e-06, "lr_fusion": 8.262928807620843e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010349997063167395} +{"avg_context_frames": 14.18125, "avg_micro_time_s": 1.6747926779120463, "epoch": 10, "global_step": 630, "grad_norm_max": 0.0135498046875, "hidden_mse": 0.011277910188073293, "loss": 0.011375020310515538, "lr_blocks": 8.199842702516584e-06, "lr_fusion": 8.199842702516583e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01024722931906581} +{"avg_context_frames": 14.459375000000001, "avg_micro_time_s": 1.6676438258135022, "epoch": 10, "global_step": 640, "grad_norm_max": 0.00994873046875, "hidden_mse": 0.011443477176362649, "loss": 0.011377838911721484, "lr_blocks": 8.135881792367686e-06, "lr_fusion": 8.135881792367686e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010233491146937014} +{"avg_context_frames": 14.428125000000001, "avg_micro_time_s": 1.670836907836565, "epoch": 10, "global_step": 650, "grad_norm_max": 0.01165771484375, "hidden_mse": 0.011770792270544917, "loss": 0.011601047770818697, "lr_blocks": 8.071063563448341e-06, "lr_fusion": 8.07106356344834e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010423968464601786} +{"avg_context_frames": 14.246875000000001, "avg_micro_time_s": 1.6668739861495852, "epoch": 11, "global_step": 660, "grad_norm_max": 0.01458740234375, "hidden_mse": 0.01172603099839762, "loss": 0.011920786584960298, "lr_blocks": 8.005405736415127e-06, "lr_fusion": 8.005405736415126e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010748183535179125} +{"avg_context_frames": 14.162500000000001, "avg_micro_time_s": 1.657212829864875, "epoch": 11, "global_step": 670, "grad_norm_max": 0.022216796875, "hidden_mse": 0.011572300613624976, "loss": 0.011661209503654391, "lr_blocks": 7.938926261462366e-06, "lr_fusion": 7.938926261462366e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010503979376517238} +{"avg_context_frames": 14.271875000000001, "avg_micro_time_s": 1.6704402020586713, "epoch": 11, "global_step": 680, "grad_norm_max": 0.0166015625, "hidden_mse": 0.011474248359445483, "loss": 0.011235295224469156, "lr_blocks": 7.871643313414718e-06, "lr_fusion": 7.871643313414718e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010087870381539688} +{"avg_context_frames": 14.490625000000001, "avg_micro_time_s": 1.672706609874149, "epoch": 11, "global_step": 690, "grad_norm_max": 0.0152587890625, "hidden_mse": 0.01137218649382703, "loss": 0.011457442858954892, "lr_blocks": 7.803575286758365e-06, "lr_fusion": 7.803575286758364e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010320224193856121} +{"avg_context_frames": 14.171875, "avg_micro_time_s": 1.6699243164744986, "epoch": 11, "global_step": 700, "grad_norm_max": 0.01275634765625, "hidden_mse": 0.011979967611841858, "loss": 0.011598831223091112, "lr_blocks": 7.734740790612137e-06, "lr_fusion": 7.734740790612136e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010400834411848338} +{"avg_context_frames": 14.359375, "avg_micro_time_s": 1.6610987717267562, "epoch": 11, "global_step": 710, "grad_norm_max": 0.0164794921875, "hidden_mse": 0.01139996346901171, "loss": 0.011594607250299305, "lr_blocks": 7.66515864363997e-06, "lr_fusion": 7.66515864363997e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010454610927263276} +{"avg_context_frames": 14.24375, "avg_micro_time_s": 1.6625836039147543, "epoch": 12, "global_step": 720, "grad_norm_max": 0.022705078125, "hidden_mse": 0.011793063330696896, "loss": 0.012230632442515344, "lr_blocks": 7.594847868906076e-06, "lr_fusion": 7.594847868906076e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011051326082088054} +{"avg_context_frames": 14.387500000000001, "avg_micro_time_s": 1.6557012428886084, "epoch": 12, "global_step": 730, "grad_norm_max": 0.0216064453125, "hidden_mse": 0.011752536008134485, "loss": 0.011842861230252312, "lr_blocks": 7.52382768867422e-06, "lr_fusion": 7.52382768867422e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010667607601499186} +{"avg_context_frames": 14.40625, "avg_micro_time_s": 1.666427310848667, "epoch": 12, "global_step": 740, "grad_norm_max": 0.0137939453125, "hidden_mse": 0.011772934650070966, "loss": 0.011801917297998444, "lr_blocks": 7.452117519152542e-06, "lr_fusion": 7.452117519152542e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010624623831245118} +{"avg_context_frames": 14.381250000000001, "avg_micro_time_s": 1.6696053218744056, "epoch": 12, "global_step": 750, "grad_norm_max": 0.0130615234375, "hidden_mse": 0.01138463456882164, "loss": 0.011473681469215081, "lr_blocks": 7.379736965185369e-06, "lr_fusion": 7.379736965185368e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010335217957617715} +{"avg_context_frames": 14.01875, "avg_micro_time_s": 1.6625237152617047, "epoch": 12, "global_step": 760, "grad_norm_max": 0.0113525390625, "hidden_mse": 0.011510639847256244, "loss": 0.011572442314354704, "lr_blocks": 7.30670581489344e-06, "lr_fusion": 7.30670581489344e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010421378299361095} +{"avg_context_frames": 14.1875, "avg_micro_time_s": 1.6770685269762908, "epoch": 12, "global_step": 770, "grad_norm_max": 0.012451171875, "hidden_mse": 0.01137546225800179, "loss": 0.011072655598400162, "lr_blocks": 7.233044034264034e-06, "lr_fusion": 7.233044034264034e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.0099351093696896} +{"avg_context_frames": 14.3625, "avg_micro_time_s": 1.680464291626413, "epoch": 12, "global_step": 780, "grad_norm_max": 0.0172119140625, "hidden_mse": 0.011443223495734857, "loss": 0.011376247013686225, "lr_blocks": 7.158771761692464e-06, "lr_fusion": 7.158771761692464e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010231924639083446} +{"avg_context_frames": 14.11875, "avg_micro_time_s": 1.668213271964487, "epoch": 13, "global_step": 790, "grad_norm_max": 0.0157470703125, "hidden_mse": 0.011763761215843261, "loss": 0.012203776161186397, "lr_blocks": 7.083909302476453e-06, "lr_fusion": 7.083909302476453e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011027400114107877} +{"avg_context_frames": 14.096875, "avg_micro_time_s": 1.6663366743512598, "epoch": 13, "global_step": 800, "grad_norm_max": 0.0194091796875, "hidden_mse": 0.011545737728010864, "loss": 0.011326604709029199, "lr_blocks": 7.008477123264849e-06, "lr_fusion": 7.008477123264848e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01017203088849783} +{"avg_context_frames": 14.34375, "avg_micro_time_s": 1.6614910699252505, "epoch": 13, "global_step": 810, "grad_norm_max": 0.01507568359375, "hidden_mse": 0.01122304906602949, "loss": 0.011121913307579235, "lr_blocks": 6.932495846462262e-06, "lr_fusion": 6.932495846462261e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.0099996083881706} +{"avg_context_frames": 14.509375, "avg_micro_time_s": 1.6837449414124421, "epoch": 13, "global_step": 820, "grad_norm_max": 0.022216796875, "hidden_mse": 0.011082636169157924, "loss": 0.010782301204744727, "lr_blocks": 6.855986244591104e-06, "lr_fusion": 6.855986244591104e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009674037591321395} +{"avg_context_frames": 14.603125, "avg_micro_time_s": 1.6708717612764303, "epoch": 13, "global_step": 830, "grad_norm_max": 0.017578125, "hidden_mse": 0.011430756392655894, "loss": 0.01166371992439963, "lr_blocks": 6.778969234612583e-06, "lr_fusion": 6.778969234612584e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010520644317148255} +{"avg_context_frames": 14.296875, "avg_micro_time_s": 1.6763124754739693, "epoch": 13, "global_step": 840, "grad_norm_max": 0.012939453125, "hidden_mse": 0.011768269899766893, "loss": 0.011903852509567514, "lr_blocks": 6.701465872208216e-06, "lr_fusion": 6.701465872208216e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010727025504456833} +{"avg_context_frames": 13.93125, "avg_micro_time_s": 1.6694887150475552, "epoch": 14, "global_step": 850, "grad_norm_max": 0.017333984375, "hidden_mse": 0.0117756710620597, "loss": 0.011661562300287188, "lr_blocks": 6.6234973460234184e-06, "lr_fusion": 6.623497346023418e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010483995190588758} +{"avg_context_frames": 14.36875, "avg_micro_time_s": 1.6700406820895297, "epoch": 14, "global_step": 860, "grad_norm_max": 0.016357421875, "hidden_mse": 0.010869661986362189, "loss": 0.010810558591037989, "lr_blocks": 6.545084971874738e-06, "lr_fusion": 6.545084971874738e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009723592398222536} +{"avg_context_frames": 14.240625000000001, "avg_micro_time_s": 1.654148106536377, "epoch": 14, "global_step": 870, "grad_norm_max": 0.020751953125, "hidden_mse": 0.011577105836477131, "loss": 0.011544856429100038, "lr_blocks": 6.466250186922325e-06, "lr_fusion": 6.466250186922325e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010387145873392001} +{"avg_context_frames": 14.65625, "avg_micro_time_s": 1.6614088644604637, "epoch": 14, "global_step": 880, "grad_norm_max": 0.01531982421875, "hidden_mse": 0.011320531176170335, "loss": 0.01140951054985635, "lr_blocks": 6.387014543809224e-06, "lr_fusion": 6.387014543809223e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010277457459596918} +{"avg_context_frames": 14.146875000000001, "avg_micro_time_s": 1.6693167886885931, "epoch": 14, "global_step": 890, "grad_norm_max": 0.0108642578125, "hidden_mse": 0.01169083885033615, "loss": 0.011576085269916804, "lr_blocks": 6.3073997047691e-06, "lr_fusion": 6.307399704769099e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010407001402927564} +{"avg_context_frames": 14.21875, "avg_micro_time_s": 1.6690942833494775, "epoch": 14, "global_step": 900, "grad_norm_max": 0.016357421875, "hidden_mse": 0.011695230467012152, "loss": 0.011713925638468937, "lr_blocks": 6.227427435703997e-06, "lr_fusion": 6.227427435703997e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01054440257139504} +{"avg_context_frames": 14.293750000000001, "avg_micro_time_s": 1.6668594162874797, "epoch": 14, "global_step": 910, "grad_norm_max": 0.0142822265625, "hidden_mse": 0.01127870319178328, "loss": 0.011360114079434425, "lr_blocks": 6.147119600233758e-06, "lr_fusion": 6.147119600233758e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010232243849895895} +{"avg_context_frames": 14.275, "avg_micro_time_s": 1.6744390483105236, "epoch": 15, "global_step": 920, "grad_norm_max": 0.015380859375, "hidden_mse": 0.01180309365154244, "loss": 0.011988794663920999, "lr_blocks": 6.066498153718735e-06, "lr_fusion": 6.066498153718735e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010808485298184678} +{"avg_context_frames": 14.025, "avg_micro_time_s": 1.660698739412692, "epoch": 15, "global_step": 930, "grad_norm_max": 0.011962890625, "hidden_mse": 0.011142635776195675, "loss": 0.011125632375478745, "lr_blocks": 5.985585137257401e-06, "lr_fusion": 5.985585137257401e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010011368786217646} +{"avg_context_frames": 14.1875, "avg_micro_time_s": 1.6714014525521634, "epoch": 15, "global_step": 940, "grad_norm_max": 0.01141357421875, "hidden_mse": 0.011377936857752503, "loss": 0.011362285091308877, "lr_blocks": 5.904402671660551e-06, "lr_fusion": 5.90440267166055e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010224491398548708} +{"avg_context_frames": 14.471875, "avg_micro_time_s": 1.6828720892244746, "epoch": 15, "global_step": 950, "grad_norm_max": 0.015625, "hidden_mse": 0.011794780974742026, "loss": 0.011746052757371218, "lr_blocks": 5.82297295140367e-06, "lr_fusion": 5.8229729514036705e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010566574707627297} +{"avg_context_frames": 14.1875, "avg_micro_time_s": 1.6743098544498936, "epoch": 15, "global_step": 960, "grad_norm_max": 0.0147705078125, "hidden_mse": 0.011310246528591961, "loss": 0.01120361810317263, "lr_blocks": 5.74131823855921e-06, "lr_fusion": 5.74131823855921e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010072593454970047} +{"avg_context_frames": 14.421875, "avg_micro_time_s": 1.669582998111946, "epoch": 15, "global_step": 970, "grad_norm_max": 0.0167236328125, "hidden_mse": 0.011059153161477299, "loss": 0.010959938418818638, "lr_blocks": 5.659460856710346e-06, "lr_fusion": 5.6594608567103456e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009854023071238772} +{"avg_context_frames": 14.106250000000001, "avg_micro_time_s": 1.667396959274629, "epoch": 16, "global_step": 980, "grad_norm_max": 0.01483154296875, "hidden_mse": 0.011709797941148282, "loss": 0.012077316356590019, "lr_blocks": 5.577423184847932e-06, "lr_fusion": 5.577423184847932e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010906336503103376} +{"avg_context_frames": 14.381250000000001, "avg_micro_time_s": 1.6778730853613526, "epoch": 16, "global_step": 990, "grad_norm_max": 0.0115966796875, "hidden_mse": 0.011467087210621686, "loss": 0.011327883088961245, "lr_blocks": 5.495227651252315e-06, "lr_fusion": 5.495227651252315e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01018117435160093} +{"avg_context_frames": 14.543750000000001, "avg_micro_time_s": 1.6817755866748485, "epoch": 16, "global_step": 1000, "grad_norm_max": 0.01263427734375, "hidden_mse": 0.010680826468160376, "loss": 0.010792405833490193, "lr_blocks": 5.412896727361663e-06, "lr_fusion": 5.4128967273616625e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.00972432320122607} +{"avg_context_frames": 14.240625000000001, "avg_micro_time_s": 1.661411627598136, "epoch": 16, "global_step": 1010, "grad_norm_max": 0.0185546875, "hidden_mse": 0.011387520033167675, "loss": 0.011222705629188569, "lr_blocks": 5.3304529216284974e-06, "lr_fusion": 5.330452921628497e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010083953657886014} +{"avg_context_frames": 14.234375, "avg_micro_time_s": 1.6655827227739792, "epoch": 16, "global_step": 1020, "grad_norm_max": 0.014892578125, "hidden_mse": 0.011972595425322653, "loss": 0.011918087041703985, "lr_blocks": 5.247918773366112e-06, "lr_fusion": 5.247918773366112e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010720827511977405} +{"avg_context_frames": 14.240625000000001, "avg_micro_time_s": 1.6604705360390655, "epoch": 16, "global_step": 1030, "grad_norm_max": 0.01165771484375, "hidden_mse": 0.011471667175646872, "loss": 0.011304579180432484, "lr_blocks": 5.165316846586541e-06, "lr_fusion": 5.165316846586541e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010157412447733806} +{"avg_context_frames": 14.378125, "avg_micro_time_s": 1.6618204411388433, "epoch": 16, "global_step": 1040, "grad_norm_max": 0.01373291015625, "hidden_mse": 0.011016206297790633, "loss": 0.011066284886328505, "lr_blocks": 5.082669723831793e-06, "lr_fusion": 5.0826697238317935e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009964664198923856} +{"avg_context_frames": 14.165625, "avg_micro_time_s": 1.6536375549985678, "epoch": 17, "global_step": 1050, "grad_norm_max": 0.01190185546875, "hidden_mse": 0.011110378883313388, "loss": 0.010883377410937102, "lr_blocks": 5e-06, "lr_fusion": 5e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009772339568007738} +{"avg_context_frames": 14.240625000000001, "avg_micro_time_s": 1.6610624320619536, "epoch": 17, "global_step": 1060, "grad_norm_max": 0.01287841796875, "hidden_mse": 0.011551116046030076, "loss": 0.011778575333300978, "lr_blocks": 4.917330276168208e-06, "lr_fusion": 4.917330276168208e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010623463714728132} +{"avg_context_frames": 14.293750000000001, "avg_micro_time_s": 1.6562152313763363, "epoch": 17, "global_step": 1070, "grad_norm_max": 0.01031494140625, "hidden_mse": 0.01174976057955064, "loss": 0.011603503639344127, "lr_blocks": 4.8346831534134595e-06, "lr_fusion": 4.834683153413459e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010428527591284365} +{"avg_context_frames": 14.30625, "avg_micro_time_s": 1.6629208340753394, "epoch": 17, "global_step": 1080, "grad_norm_max": 0.018798828125, "hidden_mse": 0.011370424262713641, "loss": 0.011329973622923718, "lr_blocks": 4.752081226633888e-06, "lr_fusion": 4.7520812266338885e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010192931210622192} +{"avg_context_frames": 14.396875000000001, "avg_micro_time_s": 1.6774043477347733, "epoch": 17, "global_step": 1090, "grad_norm_max": 0.01385498046875, "hidden_mse": 0.011211473296862097, "loss": 0.011175840767100454, "lr_blocks": 4.669547078371503e-06, "lr_fusion": 4.669547078371504e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010054693405982107} +{"avg_context_frames": 14.190625, "avg_micro_time_s": 1.6634978182628402, "epoch": 17, "global_step": 1100, "grad_norm_max": 0.0118408203125, "hidden_mse": 0.011585666466271505, "loss": 0.01196135599166155, "lr_blocks": 4.587103272638339e-06, "lr_fusion": 4.5871032726383386e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010802789340959863} +{"avg_context_frames": 14.4375, "avg_micro_time_s": 1.6532553265013121, "epoch": 18, "global_step": 1110, "grad_norm_max": 0.01300048828125, "hidden_mse": 0.010795156040694566, "loss": 0.010537886351812631, "lr_blocks": 4.504772348747687e-06, "lr_fusion": 4.504772348747687e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009458370739594102} +{"avg_context_frames": 13.996875000000001, "avg_micro_time_s": 1.6608335784279917, "epoch": 18, "global_step": 1120, "grad_norm_max": 0.01458740234375, "hidden_mse": 0.01167248976416886, "loss": 0.011236506706336514, "lr_blocks": 4.42257681515207e-06, "lr_fusion": 4.4225768151520694e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010069257783470675} +{"avg_context_frames": 14.21875, "avg_micro_time_s": 1.6740720257494104, "epoch": 18, "global_step": 1130, "grad_norm_max": 0.00994873046875, "hidden_mse": 0.01124509716173634, "loss": 0.011494273459538818, "lr_blocks": 4.340539143289655e-06, "lr_fusion": 4.3405391432896555e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010369763767812401} +{"avg_context_frames": 14.287500000000001, "avg_micro_time_s": 1.6676692341370654, "epoch": 18, "global_step": 1140, "grad_norm_max": 0.013427734375, "hidden_mse": 0.011781077337218449, "loss": 0.011909169994760306, "lr_blocks": 4.25868176144079e-06, "lr_fusion": 4.2586817614407895e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010731062246486546} +{"avg_context_frames": 14.178125000000001, "avg_micro_time_s": 1.6677437865000684, "epoch": 18, "global_step": 1150, "grad_norm_max": 0.01025390625, "hidden_mse": 0.011095615645172075, "loss": 0.011239823908545077, "lr_blocks": 4.17702704859633e-06, "lr_fusion": 4.17702704859633e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010130262339953334} +{"avg_context_frames": 14.415625, "avg_micro_time_s": 1.6578270324760525, "epoch": 18, "global_step": 1160, "grad_norm_max": 0.0189208984375, "hidden_mse": 0.011448635597480461, "loss": 0.0112959076941479, "lr_blocks": 4.0955973283394525e-06, "lr_fusion": 4.095597328339452e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010151044093072416} +{"avg_context_frames": 14.553125000000001, "avg_micro_time_s": 1.6594081541130437, "epoch": 18, "global_step": 1170, "grad_norm_max": 0.0125732421875, "hidden_mse": 0.011061798536684365, "loss": 0.01104984058183618, "lr_blocks": 4.0144148627426e-06, "lr_fusion": 4.0144148627425993e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.00994366078521125} +{"avg_context_frames": 14.421875, "avg_micro_time_s": 1.6631646836503933, "epoch": 19, "global_step": 1180, "grad_norm_max": 0.01141357421875, "hidden_mse": 0.01066333387279883, "loss": 0.010685365134850145, "lr_blocks": 3.9335018462812664e-06, "lr_fusion": 3.933501846281267e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009619031671900303} +{"avg_context_frames": 14.253125, "avg_micro_time_s": 1.6555977173513385, "epoch": 19, "global_step": 1190, "grad_norm_max": 0.0111083984375, "hidden_mse": 0.011523925105575473, "loss": 0.011254557583015412, "lr_blocks": 3.852880399766243e-06, "lr_fusion": 3.852880399766243e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010102165024727584} +{"avg_context_frames": 13.9625, "avg_micro_time_s": 1.6477898450761133, "epoch": 19, "global_step": 1200, "grad_norm_max": 0.0140380859375, "hidden_mse": 0.011714816611493007, "loss": 0.011656381204375066, "lr_blocks": 3.7725725642960047e-06, "lr_fusion": 3.772572564296005e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010484899516450243} +{"avg_context_frames": 14.340625000000001, "avg_micro_time_s": 1.6606175010983861, "epoch": 19, "global_step": 1210, "grad_norm_max": 0.013916015625, "hidden_mse": 0.011326234409352764, "loss": 0.01120215451810509, "lr_blocks": 3.6926002952309015e-06, "lr_fusion": 3.6926002952309016e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010069531074259432} +{"avg_context_frames": 14.49375, "avg_micro_time_s": 1.6539430881886803, "epoch": 19, "global_step": 1220, "grad_norm_max": 0.01275634765625, "hidden_mse": 0.011486595304450021, "loss": 0.01146072629489936, "lr_blocks": 3.6129854561907786e-06, "lr_fusion": 3.612985456190778e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01031206674524583} +{"avg_context_frames": 14.39375, "avg_micro_time_s": 1.6584447426888802, "epoch": 19, "global_step": 1230, "grad_norm_max": 0.012451171875, "hidden_mse": 0.010983252251753584, "loss": 0.010907892114482821, "lr_blocks": 3.533749813077677e-06, "lr_fusion": 3.533749813077677e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009809566976036877} +{"avg_context_frames": 14.100000000000001, "avg_micro_time_s": 1.656307424226907, "epoch": 20, "global_step": 1240, "grad_norm_max": 0.01263427734375, "hidden_mse": 0.011136419005924837, "loss": 0.011541601398494096, "lr_blocks": 3.4549150281252635e-06, "lr_fusion": 3.4549150281252636e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010427959496155381} +{"avg_context_frames": 14.168750000000001, "avg_micro_time_s": 1.6521637100984663, "epoch": 20, "global_step": 1250, "grad_norm_max": 0.012451171875, "hidden_mse": 0.011190626042662188, "loss": 0.01142194886924699, "lr_blocks": 3.3765026539765832e-06, "lr_fusion": 3.3765026539765834e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010302886262070389} +{"avg_context_frames": 14.609375, "avg_micro_time_s": 1.6508932352240664, "epoch": 20, "global_step": 1260, "grad_norm_max": 0.01141357421875, "hidden_mse": 0.011175389477284626, "loss": 0.011005451972596348, "lr_blocks": 3.298534127791785e-06, "lr_fusion": 3.298534127791785e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009887913009151816} +{"avg_context_frames": 14.15625, "avg_micro_time_s": 1.6523738718875394, "epoch": 20, "global_step": 1270, "grad_norm_max": 0.01092529296875, "hidden_mse": 0.011317605827935041, "loss": 0.011162525363033637, "lr_blocks": 3.2210307653874175e-06, "lr_fusion": 3.221030765387417e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010030764807015658} +{"avg_context_frames": 14.256250000000001, "avg_micro_time_s": 1.6502642466119142, "epoch": 20, "global_step": 1280, "grad_norm_max": 0.01080322265625, "hidden_mse": 0.011835476104170085, "loss": 0.011668690148508177, "lr_blocks": 3.1440137554088957e-06, "lr_fusion": 3.144013755408895e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010485142481047661} +{"avg_context_frames": 14.240625000000001, "avg_micro_time_s": 1.6596468387386267, "epoch": 20, "global_step": 1290, "grad_norm_max": 0.009521484375, "hidden_mse": 0.011371044552652166, "loss": 0.011234781093662605, "lr_blocks": 3.06750415353774e-06, "lr_fusion": 3.0675041535377405e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010097676678560675} +{"avg_context_frames": 14.390625, "avg_micro_time_s": 1.652364479660173, "epoch": 20, "global_step": 1300, "grad_norm_max": 0.0113525390625, "hidden_mse": 0.01134974528104067, "loss": 0.011419776943512261, "lr_blocks": 2.991522876735154e-06, "lr_fusion": 2.991522876735154e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010284802451496945} +{"avg_context_frames": 14.2375, "avg_micro_time_s": 1.6571869058017912, "epoch": 21, "global_step": 1310, "grad_norm_max": 0.015625, "hidden_mse": 0.011261974036460743, "loss": 0.011258097348036246, "lr_blocks": 2.9160906975235493e-06, "lr_fusion": 2.916090697523549e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010131899925181643} +{"avg_context_frames": 14.3625, "avg_micro_time_s": 1.646345046314673, "epoch": 21, "global_step": 1320, "grad_norm_max": 0.01043701171875, "hidden_mse": 0.010938315524253995, "loss": 0.010823095840169118, "lr_blocks": 2.8412282383075362e-06, "lr_fusion": 2.8412282383075363e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009729264292400331} +{"avg_context_frames": 14.615625000000001, "avg_micro_time_s": 1.6541958835270636, "epoch": 21, "global_step": 1330, "grad_norm_max": 0.01177978515625, "hidden_mse": 0.011133276578038932, "loss": 0.011046531575266272, "lr_blocks": 2.766955965735968e-06, "lr_fusion": 2.766955965735968e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009933203872060405} +{"avg_context_frames": 13.946875, "avg_micro_time_s": 1.659248725712314, "epoch": 21, "global_step": 1340, "grad_norm_max": 0.00933837890625, "hidden_mse": 0.011702069418970496, "loss": 0.011699286807561294, "lr_blocks": 2.693294185106562e-06, "lr_fusion": 2.693294185106562e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010529079881962389} +{"avg_context_frames": 14.146875000000001, "avg_micro_time_s": 1.6543794904744573, "epoch": 21, "global_step": 1350, "grad_norm_max": 0.01434326171875, "hidden_mse": 0.010978883533971385, "loss": 0.01105476276134141, "lr_blocks": 2.6202630348146323e-06, "lr_fusion": 2.6202630348146324e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009956874430645258} +{"avg_context_frames": 14.537500000000001, "avg_micro_time_s": 1.6427724638611836, "epoch": 21, "global_step": 1360, "grad_norm_max": 0.01416015625, "hidden_mse": 0.011333403328899296, "loss": 0.01119010701077059, "lr_blocks": 2.5478824808474613e-06, "lr_fusion": 2.547882480847461e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010056766687193887} +{"avg_context_frames": 14.3, "avg_micro_time_s": 1.649842259700381, "epoch": 22, "global_step": 1370, "grad_norm_max": 0.0091552734375, "hidden_mse": 0.011966329917777331, "loss": 0.01182466876343824, "lr_blocks": 2.476172311325783e-06, "lr_fusion": 2.476172311325783e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010628035763511436} +{"avg_context_frames": 14.484375, "avg_micro_time_s": 1.6768584499135615, "epoch": 22, "global_step": 1380, "grad_norm_max": 0.0103759765625, "hidden_mse": 0.01126847570994869, "loss": 0.011210237367777154, "lr_blocks": 2.4051521310939258e-06, "lr_fusion": 2.405152131093926e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010083389858482407} +{"avg_context_frames": 14.14375, "avg_micro_time_s": 1.6518969393000589, "epoch": 22, "global_step": 1390, "grad_norm_max": 0.01116943359375, "hidden_mse": 0.011465440806932748, "loss": 0.011482517694821582, "lr_blocks": 2.3348413563600324e-06, "lr_fusion": 2.3348413563600325e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010335973603650928} +{"avg_context_frames": 13.96875, "avg_micro_time_s": 1.6484803539868154, "epoch": 22, "global_step": 1400, "grad_norm_max": 0.010498046875, "hidden_mse": 0.011556138604646549, "loss": 0.01106837273691781, "lr_blocks": 2.265259209387867e-06, "lr_fusion": 2.2652592093878666e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009912758847349323} +{"avg_context_frames": 14.165625, "avg_micro_time_s": 1.6538986773754005, "epoch": 22, "global_step": 1410, "grad_norm_max": 0.009765625, "hidden_mse": 0.010561504488578066, "loss": 0.010613942006602884, "lr_blocks": 2.1964247132416373e-06, "lr_fusion": 2.196424713241637e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009557791572296992} +{"avg_context_frames": 14.371875000000001, "avg_micro_time_s": 1.6527693050378731, "epoch": 22, "global_step": 1420, "grad_norm_max": 0.0111083984375, "hidden_mse": 0.011084572208346799, "loss": 0.010916793579235674, "lr_blocks": 2.1283566865852824e-06, "lr_fusion": 2.128356686585282e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009808336396235974} +{"avg_context_frames": 14.471875, "avg_micro_time_s": 1.6431071163879096, "epoch": 22, "global_step": 1430, "grad_norm_max": 0.00958251953125, "hidden_mse": 0.011512590287020431, "loss": 0.011875563196372242, "lr_blocks": 2.061073738537635e-06, "lr_fusion": 2.061073738537635e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010724304232280702} +{"avg_context_frames": 14.2125, "avg_micro_time_s": 1.6510546650504694, "epoch": 23, "global_step": 1440, "grad_norm_max": 0.0076904296875, "hidden_mse": 0.010945631068898365, "loss": 0.010957002855138854, "lr_blocks": 1.9945942635848745e-06, "lr_fusion": 1.9945942635848748e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009862439712742345} +{"avg_context_frames": 13.9625, "avg_micro_time_s": 1.6461858294234846, "epoch": 23, "global_step": 1450, "grad_norm_max": 0.0069580078125, "hidden_mse": 0.01127517645363696, "loss": 0.011059572256635876, "lr_blocks": 1.928936436551661e-06, "lr_fusion": 1.928936436551661e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.0099320545559749} +{"avg_context_frames": 14.3625, "avg_micro_time_s": 1.6362899592495523, "epoch": 23, "global_step": 1460, "grad_norm_max": 0.01116943359375, "hidden_mse": 0.011928864050423727, "loss": 0.012370476470096037, "lr_blocks": 1.864118207632315e-06, "lr_fusion": 1.8641182076323148e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.011177590105216951} +{"avg_context_frames": 14.5875, "avg_micro_time_s": 1.6395295883732617, "epoch": 23, "global_step": 1470, "grad_norm_max": 0.0087890625, "hidden_mse": 0.010839305043919013, "loss": 0.01071897276560776, "lr_blocks": 1.8001572974834169e-06, "lr_fusion": 1.800157297483417e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009635042271111161} +{"avg_context_frames": 14.340625000000001, "avg_micro_time_s": 1.6361626389869344, "epoch": 23, "global_step": 1480, "grad_norm_max": 0.01483154296875, "hidden_mse": 0.011283126019407064, "loss": 0.01097512152628042, "lr_blocks": 1.7370711923791567e-06, "lr_fusion": 1.7370711923791567e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009846808918518947} +{"avg_context_frames": 14.428125000000001, "avg_micro_time_s": 1.639323206937843, "epoch": 23, "global_step": 1490, "grad_norm_max": 0.01043701171875, "hidden_mse": 0.01127969539957121, "loss": 0.011254827643278988, "lr_blocks": 1.6748771394307584e-06, "lr_fusion": 1.6748771394307585e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010126858117291705} +{"avg_context_frames": 14.200000000000001, "avg_micro_time_s": 1.638041897965377, "epoch": 24, "global_step": 1500, "grad_norm_max": 0.009521484375, "hidden_mse": 0.011386207316536457, "loss": 0.011183572874870151, "lr_blocks": 1.6135921418712959e-06, "lr_fusion": 1.6135921418712956e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010044952091993764} +{"avg_context_frames": 14.3625, "avg_micro_time_s": 1.6395215454758727, "epoch": 24, "global_step": 1510, "grad_norm_max": 0.00994873046875, "hidden_mse": 0.011413303896551952, "loss": 0.011490373039850966, "lr_blocks": 1.5532329544071712e-06, "lr_fusion": 1.553232954407171e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010349042672896758} +{"avg_context_frames": 14.496875000000001, "avg_micro_time_s": 1.6431810966249032, "epoch": 24, "global_step": 1520, "grad_norm_max": 0.006927490234375, "hidden_mse": 0.011631385050714017, "loss": 0.011430102051235735, "lr_blocks": 1.4938160786375571e-06, "lr_fusion": 1.4938160786375572e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010266963561298326} +{"avg_context_frames": 13.903125000000001, "avg_micro_time_s": 1.665714266872965, "epoch": 24, "global_step": 1530, "grad_norm_max": 0.010009765625, "hidden_mse": 0.011071189382346348, "loss": 0.010993697366211565, "lr_blocks": 1.4353577585430152e-06, "lr_fusion": 1.435357758543015e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009886578464647756} +{"avg_context_frames": 14.146875000000001, "avg_micro_time_s": 1.6518768868489133, "epoch": 24, "global_step": 1540, "grad_norm_max": 0.00799560546875, "hidden_mse": 0.011262133251875639, "loss": 0.011162439128383994, "lr_blocks": 1.3778739760445552e-06, "lr_fusion": 1.3778739760445552e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01003622582065873} +{"avg_context_frames": 14.296875, "avg_micro_time_s": 1.6510583013143334, "epoch": 24, "global_step": 1550, "grad_norm_max": 0.0084228515625, "hidden_mse": 0.01136539108119905, "loss": 0.011442213400732726, "lr_blocks": 1.321380446634342e-06, "lr_fusion": 1.3213804466343421e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010305674298433588} +{"avg_context_frames": 14.434375000000001, "avg_micro_time_s": 1.6434962125258608, "epoch": 24, "global_step": 1560, "grad_norm_max": 0.00836181640625, "hidden_mse": 0.010991631337674335, "loss": 0.010861971409758553, "lr_blocks": 1.2658926150792321e-06, "lr_fusion": 1.2658926150792322e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009762808290543035} +{"avg_context_frames": 14.578125, "avg_micro_time_s": 1.6481255472146588, "epoch": 25, "global_step": 1570, "grad_norm_max": 0.007354736328125, "hidden_mse": 0.011273362330393867, "loss": 0.01115232922602445, "lr_blocks": 1.2114256511983274e-06, "lr_fusion": 1.2114256511983274e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010024992999387906} +{"avg_context_frames": 14.296875, "avg_micro_time_s": 1.6549864632390383, "epoch": 25, "global_step": 1580, "grad_norm_max": 0.00628662109375, "hidden_mse": 0.010917296970728785, "loss": 0.010790761729003862, "lr_blocks": 1.157994445715706e-06, "lr_fusion": 1.157994445715706e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009699032048229129} +{"avg_context_frames": 14.121875000000001, "avg_micro_time_s": 1.6433439763863136, "epoch": 25, "global_step": 1590, "grad_norm_max": 0.0072021484375, "hidden_mse": 0.011266779497964308, "loss": 0.01100723147392273, "lr_blocks": 1.1056136061894386e-06, "lr_fusion": 1.1056136061894384e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009880553546827287} +{"avg_context_frames": 14.346875, "avg_micro_time_s": 1.6426967713901832, "epoch": 25, "global_step": 1600, "grad_norm_max": 0.00665283203125, "hidden_mse": 0.01106002515880391, "loss": 0.010793162672780455, "lr_blocks": 1.0542974530180327e-06, "lr_fusion": 1.0542974530180327e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009687160170869902} +{"avg_context_frames": 14.18125, "avg_micro_time_s": 1.6422525028112434, "epoch": 25, "global_step": 1610, "grad_norm_max": 0.00830078125, "hidden_mse": 0.011610816064057873, "loss": 0.011938656150596217, "lr_blocks": 1.0040600155253766e-06, "lr_fusion": 1.0040600155253765e-05, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010777574474923314} +{"avg_context_frames": 14.24375, "avg_micro_time_s": 1.6406341752641311, "epoch": 25, "global_step": 1620, "grad_norm_max": 0.0068359375, "hidden_mse": 0.011537419148953633, "loss": 0.011604450049344451, "lr_blocks": 9.549150281252633e-07, "lr_fusion": 9.549150281252633e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01045070806867443} +{"avg_context_frames": 14.15, "avg_micro_time_s": 1.6369994136250172, "epoch": 26, "global_step": 1630, "grad_norm_max": 0.006378173828125, "hidden_mse": 0.011142240453045816, "loss": 0.01115040056174621, "lr_blocks": 9.068759265665384e-07, "lr_fusion": 9.068759265665384e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010036176547873766} +{"avg_context_frames": 14.28125, "avg_micro_time_s": 1.6424545892885363, "epoch": 26, "global_step": 1640, "grad_norm_max": 0.005523681640625, "hidden_mse": 0.011423144373111427, "loss": 0.011198497947771103, "lr_blocks": 8.599558442598998e-07, "lr_fusion": 8.599558442598998e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010056183492997662} +{"avg_context_frames": 14.381250000000001, "avg_micro_time_s": 1.6438027446994967, "epoch": 26, "global_step": 1650, "grad_norm_max": 0.0067138671875, "hidden_mse": 0.011288748204242438, "loss": 0.011151470913318917, "lr_blocks": 8.141676086873574e-07, "lr_fusion": 8.141676086873572e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010022596060298384} +{"avg_context_frames": 14.359375, "avg_micro_time_s": 1.6455117818499276, "epoch": 26, "global_step": 1660, "grad_norm_max": 0.00653076171875, "hidden_mse": 0.011075893428642304, "loss": 0.01110358367441222, "lr_blocks": 7.695237378953224e-07, "lr_fusion": 7.695237378953223e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009995994257042186} +{"avg_context_frames": 14.046875, "avg_micro_time_s": 1.6422561920637238, "epoch": 26, "global_step": 1670, "grad_norm_max": 0.0059814453125, "hidden_mse": 0.011189652676694096, "loss": 0.011105867236619816, "lr_blocks": 7.260364370723044e-07, "lr_fusion": 7.260364370723044e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009986901946831495} +{"avg_context_frames": 14.24375, "avg_micro_time_s": 1.6495943162015465, "epoch": 26, "global_step": 1680, "grad_norm_max": 0.006927490234375, "hidden_mse": 0.011674578482052313, "loss": 0.011711610347265379, "lr_blocks": 6.837175952121305e-07, "lr_fusion": 6.837175952121306e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010544152522925287} +{"avg_context_frames": 14.418750000000001, "avg_micro_time_s": 1.6613924485140161, "epoch": 26, "global_step": 1690, "grad_norm_max": 0.0054931640625, "hidden_mse": 0.011259417346445845, "loss": 0.011245386057998985, "lr_blocks": 6.425787818636131e-07, "lr_fusion": 6.425787818636131e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010119444341398776} +{"avg_context_frames": 14.05, "avg_micro_time_s": 1.6575991860758224, "epoch": 27, "global_step": 1700, "grad_norm_max": 0.00628662109375, "hidden_mse": 0.011074321490013972, "loss": 0.010853745066560805, "lr_blocks": 6.026312439675553e-07, "lr_fusion": 6.026312439675552e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009746312914649025} +{"avg_context_frames": 14.256250000000001, "avg_micro_time_s": 1.6475757559775959, "epoch": 27, "global_step": 1710, "grad_norm_max": 0.005950927734375, "hidden_mse": 0.01098624971928075, "loss": 0.01108252534759231, "lr_blocks": 5.63885902781941e-07, "lr_fusion": 5.6388590278194096e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.00998390032327734} +{"avg_context_frames": 14.415625, "avg_micro_time_s": 1.643883330663448, "epoch": 27, "global_step": 1720, "grad_norm_max": 0.006591796875, "hidden_mse": 0.011700686026597396, "loss": 0.011627906281501055, "lr_blocks": 5.263533508961827e-07, "lr_fusion": 5.263533508961827e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010457837616559119} +{"avg_context_frames": 14.109375, "avg_micro_time_s": 1.6416245827273088, "epoch": 27, "global_step": 1730, "grad_norm_max": 0.0067138671875, "hidden_mse": 0.011639733344782145, "loss": 0.011623957718256861, "lr_blocks": 4.900438493352056e-07, "lr_fusion": 4.900438493352055e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010459984338376672} +{"avg_context_frames": 14.4875, "avg_micro_time_s": 1.6559243969750241, "epoch": 27, "global_step": 1740, "grad_norm_max": 0.006134033203125, "hidden_mse": 0.011353890295140445, "loss": 0.011533448891714216, "lr_blocks": 4.549673247541875e-07, "lr_fusion": 4.549673247541875e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010398059844737873} +{"avg_context_frames": 14.481250000000001, "avg_micro_time_s": 1.6521249546349281, "epoch": 27, "global_step": 1750, "grad_norm_max": 0.005340576171875, "hidden_mse": 0.01085884579224512, "loss": 0.010458897368516774, "lr_blocks": 4.211333667247125e-07, "lr_fusion": 4.2113336672471245e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009373012726427987} +{"avg_context_frames": 14.253125, "avg_micro_time_s": 1.6516157042235138, "epoch": 28, "global_step": 1760, "grad_norm_max": 0.00897216796875, "hidden_mse": 0.01099745903047733, "loss": 0.010915642289910467, "lr_blocks": 3.885512251130763e-07, "lr_fusion": 3.885512251130763e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009815896424697712} +{"avg_context_frames": 14.40625, "avg_micro_time_s": 1.644893983801012, "epoch": 28, "global_step": 1770, "grad_norm_max": 0.00506591796875, "hidden_mse": 0.011815862666117028, "loss": 0.011678584961919114, "lr_blocks": 3.572298075514652e-07, "lr_fusion": 3.5722980755146517e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010496998694725335} +{"avg_context_frames": 14.200000000000001, "avg_micro_time_s": 1.6430497726985778, "epoch": 28, "global_step": 1780, "grad_norm_max": 0.005523681640625, "hidden_mse": 0.011236432031728328, "loss": 0.01140632214373909, "lr_blocks": 3.271776770026963e-07, "lr_fusion": 3.271776770026963e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.0102826789021492} +{"avg_context_frames": 14.068750000000001, "avg_micro_time_s": 1.6447425782131178, "epoch": 28, "global_step": 1790, "grad_norm_max": 0.006195068359375, "hidden_mse": 0.011388846929185093, "loss": 0.011349822208285332, "lr_blocks": 2.984030494191942e-07, "lr_fusion": 2.9840304941919415e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010210937494412065} +{"avg_context_frames": 14.421875, "avg_micro_time_s": 1.6483115719613126, "epoch": 28, "global_step": 1800, "grad_norm_max": 0.006622314453125, "hidden_mse": 0.011306738364510238, "loss": 0.01128948276746087, "lr_blocks": 2.7091379149682683e-07, "lr_fusion": 2.7091379149682685e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010158808931009844} +{"avg_context_frames": 14.34375, "avg_micro_time_s": 1.6450169983883827, "epoch": 28, "global_step": 1810, "grad_norm_max": 0.005523681640625, "hidden_mse": 0.01131106594693847, "loss": 0.011214539268985392, "lr_blocks": 2.447174185242324e-07, "lr_fusion": 2.4471741852423237e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01008343268185854} +{"avg_context_frames": 14.26875, "avg_micro_time_s": 1.6483923100764515, "epoch": 28, "global_step": 1820, "grad_norm_max": 0.006683349609375, "hidden_mse": 0.010966053180163728, "loss": 0.010649461433058605, "lr_blocks": 2.198210923282118e-07, "lr_fusion": 2.1982109232821178e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009552856074878947} +{"avg_context_frames": 14.231250000000001, "avg_micro_time_s": 1.6480346050742811, "epoch": 29, "global_step": 1830, "grad_norm_max": 0.005767822265625, "hidden_mse": 0.011051499308086932, "loss": 0.01099624038906768, "lr_blocks": 1.962316193157593e-07, "lr_fusion": 1.962316193157593e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009891090454766528} +{"avg_context_frames": 14.65625, "avg_micro_time_s": 1.6622220165001638, "epoch": 29, "global_step": 1840, "grad_norm_max": 0.005279541015625, "hidden_mse": 0.01025027733994648, "loss": 0.010120689944596962, "lr_blocks": 1.7395544861325718e-07, "lr_fusion": 1.7395544861325718e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009095662174513564} +{"avg_context_frames": 14.4, "avg_micro_time_s": 1.6728738018879086, "epoch": 29, "global_step": 1850, "grad_norm_max": 0.00616455078125, "hidden_mse": 0.011631351051619276, "loss": 0.011471145966788755, "lr_blocks": 1.5299867030334815e-07, "lr_fusion": 1.5299867030334814e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010308010876178742} +{"avg_context_frames": 14.040625, "avg_micro_time_s": 1.6607555011127262, "epoch": 29, "global_step": 1860, "grad_norm_max": 0.005706787109375, "hidden_mse": 0.011891721456777304, "loss": 0.011896548618096858, "lr_blocks": 1.333670137599713e-07, "lr_fusion": 1.333670137599713e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.01070737650152296} +{"avg_context_frames": 14.4, "avg_micro_time_s": 1.666131774499445, "epoch": 29, "global_step": 1870, "grad_norm_max": 0.00555419921875, "hidden_mse": 0.011788329185219482, "loss": 0.011521958839148284, "lr_blocks": 1.1506584608200366e-07, "lr_fusion": 1.1506584608200367e-06, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010343125910731034} +{"avg_context_frames": 14.0875, "avg_micro_time_s": 1.6488830267502637, "epoch": 29, "global_step": 1880, "grad_norm_max": 0.00457763671875, "hidden_mse": 0.011259010009234773, "loss": 0.01129114049836062, "lr_blocks": 9.810017062595322e-08, "lr_fusion": 9.810017062595322e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010165239439811558} +{"avg_context_frames": 14.334375000000001, "avg_micro_time_s": 1.658405292975658, "epoch": 30, "global_step": 1890, "grad_norm_max": 0.00701904296875, "hidden_mse": 0.01104553140467033, "loss": 0.011151871737092735, "lr_blocks": 8.247462563808816e-08, "lr_fusion": 8.247462563808817e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010047318600118161} +{"avg_context_frames": 14.3375, "avg_micro_time_s": 1.6700679287012463, "epoch": 30, "global_step": 1900, "grad_norm_max": 0.00677490234375, "hidden_mse": 0.011436432937625796, "loss": 0.011454507318558172, "lr_blocks": 6.819348298638839e-08, "lr_fusion": 6.819348298638839e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010310864035272971} +{"avg_context_frames": 14.175, "avg_micro_time_s": 1.6541614500634751, "epoch": 30, "global_step": 1910, "grad_norm_max": 0.0052490234375, "hidden_mse": 0.011174249934265391, "loss": 0.010894450556952507, "lr_blocks": 5.526064699265754e-08, "lr_fusion": 5.526064699265753e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009777025505900384} +{"avg_context_frames": 14.200000000000001, "avg_micro_time_s": 1.6462076358369815, "epoch": 30, "global_step": 1920, "grad_norm_max": 0.0081787109375, "hidden_mse": 0.011725715646753088, "loss": 0.011740783043205739, "lr_blocks": 4.367965336512403e-08, "lr_fusion": 4.367965336512403e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010568211472127588} +{"avg_context_frames": 14.15, "avg_micro_time_s": 1.6475436234864902, "epoch": 30, "global_step": 1930, "grad_norm_max": 0.00604248046875, "hidden_mse": 0.011032538558356465, "loss": 0.010990633897017689, "lr_blocks": 3.345366823180929e-08, "lr_fusion": 3.3453668231809286e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009887380053987727} +{"avg_context_frames": 14.137500000000001, "avg_micro_time_s": 1.6395516200249405, "epoch": 30, "global_step": 1940, "grad_norm_max": 0.004638671875, "hidden_mse": 0.011372241860954091, "loss": 0.011327299562981353, "lr_blocks": 2.4585487274942922e-08, "lr_fusion": 2.458548727494292e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010190075432183221} +{"avg_context_frames": 14.5875, "avg_micro_time_s": 1.6594994173243323, "epoch": 30, "global_step": 1950, "grad_norm_max": 0.006378173828125, "hidden_mse": 0.010940364468842745, "loss": 0.010809329786570744, "lr_blocks": 1.7077534966650767e-08, "lr_fusion": 1.7077534966650766e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009715293342014776} +{"avg_context_frames": 14.481250000000001, "avg_micro_time_s": 1.6560149330744025, "epoch": 31, "global_step": 1960, "grad_norm_max": 0.00433349609375, "hidden_mse": 0.011649387062061579, "loss": 0.011777725949650632, "lr_blocks": 1.0931863906127327e-08, "lr_fusion": 1.0931863906127327e-07, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010612787236459554} +{"avg_context_frames": 14.256250000000001, "avg_micro_time_s": 1.6530842161268084, "epoch": 31, "global_step": 1970, "grad_norm_max": 0.007781982421875, "hidden_mse": 0.01121381939155981, "loss": 0.010996508109383286, "lr_blocks": 6.150154258476315e-09, "lr_fusion": 6.150154258476315e-08, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009875126113183797} +{"avg_context_frames": 14.396875000000001, "avg_micro_time_s": 1.6463287756854699, "epoch": 31, "global_step": 1980, "grad_norm_max": 0.00567626953125, "hidden_mse": 0.010809205658733846, "loss": 0.010781579179456458, "lr_blocks": 2.7337132953697555e-09, "lr_fusion": 2.7337132953697554e-08, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009700658690417186} +{"avg_context_frames": 14.309375000000001, "avg_micro_time_s": 1.6505390620754044, "epoch": 31, "global_step": 1990, "grad_norm_max": 0.005279541015625, "hidden_mse": 0.011214246426243336, "loss": 0.01118017845437862, "lr_blocks": 6.834750376549793e-10, "lr_fusion": 6.834750376549792e-09, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.010058753815246747} +{"avg_context_frames": 14.203125, "avg_micro_time_s": 1.6564274414478861, "epoch": 31, "global_step": 2000, "grad_norm_max": 0.00537109375, "hidden_mse": 0.011120558704715223, "loss": 0.011053164611803369, "lr_blocks": 0.0, "lr_fusion": 0.0, "peak_memory_gib": 31.181796550750732, "velocity_mse": 0.009941108722705394} +swanlab: 📁 View project at +https://swanlab.cn/@ZhuoChen/HY-WorldPlay-Predictor-DisCa +swanlab: 🚀 View run predictor-disca-0-53-b16x8-2000steps at +https://swanlab.cn/@ZhuoChen/HY-WorldPlay-Predictor-DisCa/runs/8mjozf71 +swanlab: Upload complete: 25337 records diff --git a/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/training_latest.pt b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/training_latest.pt new file mode 100644 index 0000000000000000000000000000000000000000..24475c202c7c1def9ee707c4fe954b7c49faa72c --- /dev/null +++ b/checkpoints/predictor_disca_blocks0-53_bf16_8gpu_flr1e-4_blr1e-5_schedsteps2000_swarmup100_bwarmup100_swanlab/training_latest.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:975c4d8ffdafc2d22e7b5e7ea69025287f61b6d17320515f58dd3e9014e1d96b +size 1359629862 diff --git a/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/launcher.nohup.log b/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/launcher.nohup.log new file mode 100644 index 0000000000000000000000000000000000000000..500eb0b67794f51eac9b620e8aab24904d3d6914 --- /dev/null +++ b/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/launcher.nohup.log @@ -0,0 +1,296 @@ +/local/zoubin/cz/envs/hy_worldplay_july20260720/lib/python3.10/site-packages/transformers/utils/hub.py:111: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. + warnings.warn( +/local/zoubin/cz/envs/hy_worldplay_july20260720/lib/python3.10/site-packages/transformers/utils/hub.py:111: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. + warnings.warn( +/local/zoubin/cz/envs/hy_worldplay_july20260720/lib/python3.10/site-packages/transformers/utils/hub.py:111: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. + warnings.warn( +/local/zoubin/cz/envs/hy_worldplay_july20260720/lib/python3.10/site-packages/transformers/utils/hub.py:111: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. + warnings.warn( +{ + "manifest": "/local/zoubin/cz/projects/HY-WorldPlay-DEV-Predictor/datasets/hyworldplay_predictor_prefeature_v4/train_eval_manifest.jsonl", + "teacher_checkpoint": "/local/zoubin/cz/checkpoints/hy_worldplay/huggingface/hub/models--tencent--HY-WorldPlay/snapshots/f4c29235647707b571479a69b569e4166f9f5bf8/ar_distilled_action_model/diffusion_pytorch_model.safetensors", + "output_dir": "/local/zoubin/cz/projects/HY-WorldPlay-DEV-Predictor/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000", + "source_blocks": [ + 1, + 52 + ], + "predictor_variant": "v4", + "previous_condition_mode": "same_step", + "predictor_fusion_type": "concat", + "gate_hidden_dim": 256, + "gate_initial_probability": 0.8, + "predictor_arch": "atc", + "atc_variant": "full", + "atc_mlp_hidden_dim": 4096, + "atc_gate_hidden_dim": 512, + "atc_gate_initial_probability": 0.3, + "atc_transport_heads": null, + "atc_transport_mode": "global", + "atc_previous_scope": "chunk", + "atc_transport_window": 3, + "atc_transport_query_chunk_size": 128, + "atc_transport_initial_scale": 0.1, + "max_steps": 2000, + "batch_size": 8, + "gradient_accumulation_steps": 2, + "block_warmup_steps": 100, + "block_warmup_mode": "delayed", + "scheduler_warmup_steps": 100, + "fusion_lr": 0.0001, + "blocks_lr": 1e-05, + "weight_decay": 0.01, + "hidden_weight": 0.1, + "velocity_weight": 1.0, + "grad_clip": 1.0, + "log_every": 10, + "save_every": 100, + "keep_snapshots": 20, + "seed": 0, + "num_workers": 0, + "max_records": null, + "bucket_by_context": true, + "gradient_checkpointing": true, + "fp32_trainable_params": true, + "train_delta_only": false, + "train_fusion_residual_only": false, + "save_optimizer_state": true, + "init_training_state": null, + "resume": null, + "swanlab_project": "HY-WorldPlay-Predictor-v4", + "swanlab_workspace": null, + "swanlab_experiment": "predictor-v4-atc-full-global-chunk-blocks1-52-fulltrain100-4gpu-bs8x2", + "swanlab_mode": "disabled", + "predictor_version": "v4_prefeature", + "schema_version": "predictor_context_prefeature_bf16_v1", + "supervision_pairs": [ + [ + 0, + 1 + ], + [ + 1, + 2 + ], + [ + 2, + 3 + ] + ], + "previous_condition_source": "offline_full_teacher_hidden", + "records": 2800, + "dataset_pairs": 8400, + "steps_per_epoch_per_rank": 260, + "world_size": 4, + "per_rank_batch_size": 8, + "global_batch_size": 64, + "trainable_parameters": 291649026, + "frozen_projector_parameters": 16785664, + "dtype": "bfloat16", + "activation_dtype": "bfloat16", + "trainable_parameter_dtype": "float32", + "optimizer_moment_dtype": "float32", + "inference_checkpoint_dtype": "bfloat16", + "gpu": "NVIDIA H100 80GB HBM3" +} +{"swanlab_run_id": "ry693607"} +{"anchor_distance_mean": 111.84375, "avg_context_frames": 20.0, "avg_micro_time_s": 3.3916018948802957, "delta_h_c_norm": 0.0, "delta_h_d_norm": 0.0, "epoch": 1, "gate_p_mean": 0.30317118391394615, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.29931640625, "gate_p_p50": 0.302734375, "gate_p_p90": 0.306884765625, "gate_p_std": 0.003097614331636578, "gated_delta_h_c_norm": 0.0, "global_step": 1, "grad_norm_max": 0.04661368578672409, "hidden_mse": 0.012811526888981462, "loss": 0.007839966448955238, "lr_blocks": 0.0, "lr_fusion": 2.0000000000000003e-06, "peak_memory_gib": 40.41654920578003, "transport_attention_entropy": 8.700663566589355, "transport_residual_norm": 0.025727889966219664, "velocity_mse": 0.006558813736774027} +{"anchor_distance_mean": 125.80729166666666, "avg_context_frames": 13.777777777777777, "avg_micro_time_s": 2.5044178074701144, "delta_h_c_norm": 0.00997479192382242, "delta_h_d_norm": 0.003670008891074556, "epoch": 1, "gate_p_mean": 0.3034007317490048, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.2998589409722222, "gate_p_p50": 0.3033311631944444, "gate_p_p90": 0.30704752604166663, "gate_p_std": 0.0028908914763532165, "gated_delta_h_c_norm": 0.003027750549436961, "global_step": 10, "grad_norm_max": 0.08532179892063141, "hidden_mse": 0.01726151206013229, "loss": 0.014685589318913924, "lr_blocks": 0.0, "lr_fusion": 1.1000000000000001e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700913416014776, "transport_residual_norm": 0.025541054809259042, "velocity_mse": 0.012959438182103137} +{"anchor_distance_mean": 124.57421875, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.539255334097834, "delta_h_c_norm": 0.022952683083713055, "delta_h_d_norm": 0.00608184410375543, "epoch": 1, "gate_p_mean": 0.29058273881673813, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.2864501953125, "gate_p_p50": 0.2902587890625, "gate_p_p90": 0.2950439453125, "gate_p_std": 0.0033349570992868395, "gated_delta_h_c_norm": 0.006657086504856125, "global_step": 20, "grad_norm_max": 0.07027794420719147, "hidden_mse": 0.017141231457935648, "loss": 0.014751947904005648, "lr_blocks": 0.0, "lr_fusion": 2.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700921440124512, "transport_residual_norm": 0.02545799491927028, "velocity_mse": 0.013037824665661902} +{"anchor_distance_mean": 121.45234375000001, "avg_context_frames": 15.0, "avg_micro_time_s": 2.6014328160017612, "delta_h_c_norm": 0.03039693357422948, "delta_h_d_norm": 0.009237261430826038, "epoch": 1, "gate_p_mean": 0.26862451545894145, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.26031494140625, "gate_p_p50": 0.2681640625, "gate_p_p90": 0.27763671875, "gate_p_std": 0.006609475819277577, "gated_delta_h_c_norm": 0.008192429994232952, "global_step": 30, "grad_norm_max": 0.07879240810871124, "hidden_mse": 0.015000058233272285, "loss": 0.012391855125315487, "lr_blocks": 0.0, "lr_fusion": 3.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700650000572205, "transport_residual_norm": 0.025503298174589874, "velocity_mse": 0.010891849230392837} +{"anchor_distance_mean": 126.46015625000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6276170833647488, "delta_h_c_norm": 0.033589290780946614, "delta_h_d_norm": 0.012352850113529712, "epoch": 1, "gate_p_mean": 0.24227235317230225, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.22742919921875002, "gate_p_p50": 0.2413818359375, "gate_p_p90": 0.25870361328125, "gate_p_std": 0.011819115577964112, "gated_delta_h_c_norm": 0.008287024806486442, "global_step": 40, "grad_norm_max": 0.0587887167930603, "hidden_mse": 0.01564578412217088, "loss": 0.012874998210463673, "lr_blocks": 0.0, "lr_fusion": 4.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70101239681244, "transport_residual_norm": 0.025698478100821376, "velocity_mse": 0.011310419753135648} +{"anchor_distance_mean": 128.990625, "avg_context_frames": 15.8, "avg_micro_time_s": 2.708179184785695, "delta_h_c_norm": 0.03730703087057918, "delta_h_d_norm": 0.013927469297777863, "epoch": 1, "gate_p_mean": 0.22535368371754885, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.20020751953125002, "gate_p_p50": 0.2228271484375, "gate_p_p90": 0.255224609375, "gate_p_std": 0.020604861131869257, "gated_delta_h_c_norm": 0.00883385858614929, "global_step": 50, "grad_norm_max": 0.041065044701099396, "hidden_mse": 0.015060250839451329, "loss": 0.013035302184289322, "lr_blocks": 0.0, "lr_fusion": 5.1000000000000006e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70089852809906, "transport_residual_norm": 0.025752485753037038, "velocity_mse": 0.0115292771224631} +{"anchor_distance_mean": 124.54218750000001, "avg_context_frames": 16.400000000000002, "avg_micro_time_s": 2.7534286990441617, "delta_h_c_norm": 0.04436152405105531, "delta_h_d_norm": 0.016780297516379505, "epoch": 1, "gate_p_mean": 0.20458340328186753, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.16337280273437502, "gate_p_p50": 0.1978759765625, "gate_p_p90": 0.25955810546875, "gate_p_std": 0.03580526039004326, "gated_delta_h_c_norm": 0.010145669779740275, "global_step": 60, "grad_norm_max": 0.03292519971728325, "hidden_mse": 0.014389093441423029, "loss": 0.012210481296642685, "lr_blocks": 0.0, "lr_fusion": 6.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70082403421402, "transport_residual_norm": 0.025599378719925882, "velocity_mse": 0.0107715719321277} +{"anchor_distance_mean": 119.18046875, "avg_context_frames": 13.8, "avg_micro_time_s": 2.4716390638015584, "delta_h_c_norm": 0.04784528408199549, "delta_h_d_norm": 0.018238633021246644, "epoch": 1, "gate_p_mean": 0.1840870136395097, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0, "gate_p_p10": 0.12093505859375, "gate_p_p50": 0.17048339843750002, "gate_p_p90": 0.27371826171875, "gate_p_std": 0.05804847930558026, "gated_delta_h_c_norm": 0.01107517054770142, "global_step": 70, "grad_norm_max": 0.04321914166212082, "hidden_mse": 0.014457606984069572, "loss": 0.01299729972088244, "lr_blocks": 0.0, "lr_fusion": 7.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70079131126404, "transport_residual_norm": 0.025568203045986594, "velocity_mse": 0.011551539003266954} +{"anchor_distance_mean": 123.95625000000001, "avg_context_frames": 12.200000000000001, "avg_micro_time_s": 2.2920315945782934, "delta_h_c_norm": 0.047533654142171144, "delta_h_d_norm": 0.01987655949778855, "epoch": 1, "gate_p_mean": 0.17116718580946327, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.0023700421644207385, "gate_p_p10": 0.0868060290813446, "gate_p_p50": 0.146533203125, "gate_p_p90": 0.30340576171875, "gate_p_std": 0.08492751198355108, "gated_delta_h_c_norm": 0.011952080181799829, "global_step": 80, "grad_norm_max": 0.05090194195508957, "hidden_mse": 0.014643171295756475, "loss": 0.013282303983578459, "lr_blocks": 0.0, "lr_fusion": 8.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701030111312866, "transport_residual_norm": 0.02552230970468372, "velocity_mse": 0.011817986864480191} +{"anchor_distance_mean": 126.29375, "avg_context_frames": 13.4, "avg_micro_time_s": 2.418264391613775, "delta_h_c_norm": 0.04717459180392325, "delta_h_d_norm": 0.02282619383186102, "epoch": 1, "gate_p_mean": 0.10493078203871847, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.31404523224821335, "gate_p_p10": 0.04160125702619553, "gate_p_p50": 0.07586517333984376, "gate_p_p90": 0.2195751667022705, "gate_p_std": 0.07696795002557338, "gated_delta_h_c_norm": 0.009221456036902964, "global_step": 90, "grad_norm_max": 0.04668376222252846, "hidden_mse": 0.015448533091694118, "loss": 0.013303781766444445, "lr_blocks": 0.0, "lr_fusion": 9.1e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701134538650512, "transport_residual_norm": 0.025644757691770794, "velocity_mse": 0.01175892845785711} +{"anchor_distance_mean": 126.78203125, "avg_context_frames": 13.0, "avg_micro_time_s": 2.359297069949389, "delta_h_c_norm": 0.047289361059665684, "delta_h_d_norm": 0.023216025158762935, "epoch": 1, "gate_p_mean": 0.08504255732987076, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.5318134183064104, "gate_p_p10": 0.024285888671875002, "gate_p_p50": 0.053881835937500004, "gate_p_p90": 0.19698419570922854, "gate_p_std": 0.07988582017133013, "gated_delta_h_c_norm": 0.008903939131414518, "global_step": 100, "grad_norm_max": 0.06549614667892456, "hidden_mse": 0.014653777156490834, "loss": 0.013311180018354208, "lr_blocks": 1e-05, "lr_fusion": 0.0001, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701389598846436, "transport_residual_norm": 0.025689266971312463, "velocity_mse": 0.011845802285824903} +{"anchor_distance_mean": 122.59218750000001, "avg_context_frames": 13.8, "avg_micro_time_s": 2.4595028760653808, "delta_h_c_norm": 0.045384027343243365, "delta_h_d_norm": 0.023730936297215524, "epoch": 1, "gate_p_mean": 0.06513788932934404, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.673690426722169, "gate_p_p10": 0.01361083984375, "gate_p_p50": 0.034098815917968754, "gate_p_p90": 0.16410150527954104, "gate_p_std": 0.0762962760811206, "gated_delta_h_c_norm": 0.007292107934335945, "global_step": 110, "grad_norm_max": 0.06186863034963608, "hidden_mse": 0.014536809321725741, "loss": 0.012818049592897297, "lr_blocks": 9.999316524962347e-06, "lr_fusion": 9.999316524962345e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701248478889466, "transport_residual_norm": 0.026201228285208347, "velocity_mse": 0.011364368617068977} +{"anchor_distance_mean": 126.88125000000001, "avg_context_frames": 16.6, "avg_micro_time_s": 2.781841829048062, "delta_h_c_norm": 0.04880318744108081, "delta_h_d_norm": 0.024842618848197162, "epoch": 1, "gate_p_mean": 0.05634526829235256, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7096289299428463, "gate_p_p10": 0.011680221557617188, "gate_p_p50": 0.030135345458984376, "gate_p_p90": 0.13969268798828124, "gate_p_std": 0.06797815505415201, "gated_delta_h_c_norm": 0.007392013185744873, "global_step": 120, "grad_norm_max": 0.08276909589767456, "hidden_mse": 0.013440203480422498, "loss": 0.011876265768660233, "lr_blocks": 9.99726628670463e-06, "lr_fusion": 9.997266286704631e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701385080814362, "transport_residual_norm": 0.02633053897880018, "velocity_mse": 0.01053224545903504} +{"anchor_distance_mean": 127.59453125, "avg_context_frames": 12.200000000000001, "avg_micro_time_s": 2.287554034231289, "delta_h_c_norm": 0.05330531457439065, "delta_h_d_norm": 0.02785128252580762, "epoch": 1, "gate_p_mean": 0.06753015303984285, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.6613637015223504, "gate_p_p10": 0.012387466058135033, "gate_p_p50": 0.032259368896484376, "gate_p_p90": 0.17074451446533204, "gate_p_std": 0.08562655816785991, "gated_delta_h_c_norm": 0.009928627495537513, "global_step": 130, "grad_norm_max": 0.035727184265851974, "hidden_mse": 0.015052694323821926, "loss": 0.013554053226835096, "lr_blocks": 9.993849845741525e-06, "lr_fusion": 9.993849845741524e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701139283180238, "transport_residual_norm": 0.02618990105111152, "velocity_mse": 0.012048783745558468} +{"anchor_distance_mean": 126.81562500000001, "avg_context_frames": 16.0, "avg_micro_time_s": 2.6903871252623506, "delta_h_c_norm": 0.054359220247715716, "delta_h_d_norm": 0.027648662473075094, "epoch": 2, "gate_p_mean": 0.04782442124560476, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7462517768144608, "gate_p_p10": 0.009492874145507812, "gate_p_p50": 0.02564239501953125, "gate_p_p90": 0.11377716064453125, "gate_p_std": 0.05783966393209994, "gated_delta_h_c_norm": 0.006741287760814885, "global_step": 140, "grad_norm_max": 0.033778734505176544, "hidden_mse": 0.013616233336506411, "loss": 0.012111890042433515, "lr_blocks": 9.989068136093873e-06, "lr_fusion": 9.989068136093873e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701544928550721, "transport_residual_norm": 0.027187892422080043, "velocity_mse": 0.010750266641844065} +{"anchor_distance_mean": 125.38593750000001, "avg_context_frames": 12.8, "avg_micro_time_s": 2.343650365938083, "delta_h_c_norm": 0.05648717880249024, "delta_h_d_norm": 0.030456845578737558, "epoch": 2, "gate_p_mean": 0.061732819688040765, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7109207447618247, "gate_p_p10": 0.010462570190429689, "gate_p_p50": 0.029850006103515625, "gate_p_p90": 0.15647151470184328, "gate_p_std": 0.08082308233715595, "gated_delta_h_c_norm": 0.00896596709280857, "global_step": 150, "grad_norm_max": 0.038248006254434586, "hidden_mse": 0.01378734323079698, "loss": 0.012403545813867822, "lr_blocks": 9.98292246503335e-06, "lr_fusion": 9.98292246503335e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701370680332184, "transport_residual_norm": 0.027544553601183, "velocity_mse": 0.011024811497190968} +{"anchor_distance_mean": 123.6328125, "avg_context_frames": 15.0, "avg_micro_time_s": 2.588700639344461, "delta_h_c_norm": 0.056849720329046255, "delta_h_d_norm": 0.032354711345396935, "epoch": 2, "gate_p_mean": 0.05848325380356983, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7120906189084053, "gate_p_p10": 0.012528228759765626, "gate_p_p50": 0.031022644042968752, "gate_p_p90": 0.14336395263671875, "gate_p_std": 0.07111960340989754, "gated_delta_h_c_norm": 0.007654975829791511, "global_step": 160, "grad_norm_max": 0.0582103356719017, "hidden_mse": 0.014547444524941966, "loss": 0.01337206966127269, "lr_blocks": 9.975414512725058e-06, "lr_fusion": 9.975414512725057e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701601231098175, "transport_residual_norm": 0.02883081831969321, "velocity_mse": 0.01191732520237565} +{"anchor_distance_mean": 120.61171875000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.609243828036415, "delta_h_c_norm": 0.057860920578241354, "delta_h_d_norm": 0.033041059714742006, "epoch": 2, "gate_p_mean": 0.045176344877108934, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7604151863604784, "gate_p_p10": 0.01070709228515625, "gate_p_p50": 0.02453460693359375, "gate_p_p90": 0.10254535675048829, "gate_p_std": 0.05610842191963457, "gated_delta_h_c_norm": 0.005948154463840183, "global_step": 170, "grad_norm_max": 0.03626026585698128, "hidden_mse": 0.013719683879753575, "loss": 0.012855283805401997, "lr_blocks": 9.966546331768192e-06, "lr_fusion": 9.966546331768191e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701769173145294, "transport_residual_norm": 0.029950046865269545, "velocity_mse": 0.011483315433724784} +{"anchor_distance_mean": 123.43671875000001, "avg_context_frames": 16.8, "avg_micro_time_s": 2.802227162248164, "delta_h_c_norm": 0.06344779334031046, "delta_h_d_norm": 0.03263235741760582, "epoch": 2, "gate_p_mean": 0.05024055056273938, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7127784673124552, "gate_p_p10": 0.012176132202148438, "gate_p_p50": 0.0292694091796875, "gate_p_p90": 0.11561279296875, "gate_p_std": 0.055957942095119506, "gated_delta_h_c_norm": 0.0075106066986336375, "global_step": 180, "grad_norm_max": 0.025230951607227325, "hidden_mse": 0.012641142023494468, "loss": 0.011307318593026139, "lr_blocks": 9.956320346634877e-06, "lr_fusion": 9.956320346634876e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701630127429963, "transport_residual_norm": 0.030206872266717255, "velocity_mse": 0.010043204388057348} +{"anchor_distance_mean": 129.02265625, "avg_context_frames": 13.0, "avg_micro_time_s": 2.3841907729467495, "delta_h_c_norm": 0.0645554750226438, "delta_h_d_norm": 0.03248714942019433, "epoch": 2, "gate_p_mean": 0.042375249078031633, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7896179147064686, "gate_p_p10": 0.008902740478515626, "gate_p_p50": 0.021332550048828128, "gate_p_p90": 0.09999279975891114, "gate_p_std": 0.055479835462756456, "gated_delta_h_c_norm": 0.0071524462851812135, "global_step": 190, "grad_norm_max": 0.02912726253271103, "hidden_mse": 0.013364255707710982, "loss": 0.011932299192994834, "lr_blocks": 9.944739353007344e-06, "lr_fusion": 9.944739353007344e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701686239242553, "transport_residual_norm": 0.030628428561612964, "velocity_mse": 0.010595873621059583} +{"anchor_distance_mean": 125.515625, "avg_context_frames": 12.4, "avg_micro_time_s": 2.3011850478418636, "delta_h_c_norm": 0.06875569028779864, "delta_h_d_norm": 0.03420122137758881, "epoch": 2, "gate_p_mean": 0.05247711356496439, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7356037355959416, "gate_p_p10": 0.009444808959960938, "gate_p_p50": 0.0266754150390625, "gate_p_p90": 0.13108062744140625, "gate_p_std": 0.06633843354065903, "gated_delta_h_c_norm": 0.008689464366761967, "global_step": 200, "grad_norm_max": 0.028081020340323448, "hidden_mse": 0.014369747182354332, "loss": 0.013182323408545927, "lr_blocks": 9.931806517013612e-06, "lr_fusion": 9.931806517013612e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701787495613098, "transport_residual_norm": 0.031094087543897333, "velocity_mse": 0.01174534869787749} +{"anchor_distance_mean": 125.15703125, "avg_context_frames": 16.6, "avg_micro_time_s": 2.7605856346854125, "delta_h_c_norm": 0.07116367779672146, "delta_h_d_norm": 0.03666707840748131, "epoch": 2, "gate_p_mean": 0.04512293294537813, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7689623657613993, "gate_p_p10": 0.008864974975585938, "gate_p_p50": 0.023056793212890628, "gate_p_p90": 0.10902066230773927, "gate_p_std": 0.05625875624828041, "gated_delta_h_c_norm": 0.008017951759393327, "global_step": 210, "grad_norm_max": 0.04490853101015091, "hidden_mse": 0.01293622823432088, "loss": 0.011786308721639216, "lr_blocks": 9.917525374361913e-06, "lr_fusion": 9.917525374361912e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701880311965942, "transport_residual_norm": 0.031989802280440925, "velocity_mse": 0.010492685838835315} +{"anchor_distance_mean": 124.8015625, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.5584401001178776, "delta_h_c_norm": 0.07336803604848684, "delta_h_d_norm": 0.03391947434283793, "epoch": 2, "gate_p_mean": 0.03174663808895275, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8517673548310996, "gate_p_p10": 0.0054019162431359295, "gate_p_p50": 0.015155792236328125, "gate_p_p90": 0.0772568702697754, "gate_p_std": 0.04533045875723474, "gated_delta_h_c_norm": 0.0063726135114848155, "global_step": 220, "grad_norm_max": 0.045251134783029556, "hidden_mse": 0.012441512307850644, "loss": 0.011287520788027906, "lr_blocks": 9.901899829374048e-06, "lr_fusion": 9.901899829374047e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701839220523835, "transport_residual_norm": 0.032853246503509584, "velocity_mse": 0.010043369620689191} +{"anchor_distance_mean": 125.48203125, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.504839202530275, "delta_h_c_norm": 0.07506553898565471, "delta_h_d_norm": 0.03396783375646919, "epoch": 2, "gate_p_mean": 0.030358277732739226, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8656490679830313, "gate_p_p10": 0.004883766174316406, "gate_p_p50": 0.014351081848144532, "gate_p_p90": 0.07403378486633301, "gate_p_std": 0.04368014740757645, "gated_delta_h_c_norm": 0.006342572106950684, "global_step": 230, "grad_norm_max": 0.03213318809866905, "hidden_mse": 0.014076661923900247, "loss": 0.01286342016828712, "lr_blocks": 9.884934153917998e-06, "lr_fusion": 9.884934153917997e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702045381069183, "transport_residual_norm": 0.034476458793506026, "velocity_mse": 0.011455753943300806} +{"anchor_distance_mean": 123.24062500000001, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.509506347407296, "delta_h_c_norm": 0.07586349942721426, "delta_h_d_norm": 0.03696638948749751, "epoch": 2, "gate_p_mean": 0.028896749933483082, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8635860145092011, "gate_p_p10": 0.0051326751708984375, "gate_p_p50": 0.014488601684570314, "gate_p_p90": 0.06887800693511963, "gate_p_std": 0.040187523036729546, "gated_delta_h_c_norm": 0.005609195213764906, "global_step": 240, "grad_norm_max": 0.046186238527297974, "hidden_mse": 0.013516320887720213, "loss": 0.012190171156544239, "lr_blocks": 9.86663298624003e-06, "lr_fusion": 9.86663298624003e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701927721500397, "transport_residual_norm": 0.03684651078656316, "velocity_mse": 0.010838539034011774} +{"anchor_distance_mean": 125.74296875, "avg_context_frames": 11.4, "avg_micro_time_s": 2.208755862565886, "delta_h_c_norm": 0.07782657416537404, "delta_h_d_norm": 0.040381841966882354, "epoch": 2, "gate_p_mean": 0.0389973629731685, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7972756631672383, "gate_p_p10": 0.00704345703125, "gate_p_p50": 0.02066459655761719, "gate_p_p90": 0.09181900024414064, "gate_p_std": 0.05100602825405076, "gated_delta_h_c_norm": 0.007298467074724613, "global_step": 250, "grad_norm_max": 0.031774863600730896, "hidden_mse": 0.013964945607585834, "loss": 0.013243870803853497, "lr_blocks": 9.847001329696653e-06, "lr_fusion": 9.847001329696653e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702219188213348, "transport_residual_norm": 0.03645712174475193, "velocity_mse": 0.011847376226796769} +{"anchor_distance_mean": 124.99453125000001, "avg_context_frames": 13.8, "avg_micro_time_s": 2.457255856954726, "delta_h_c_norm": 0.0799748877529055, "delta_h_d_norm": 0.03981192954815924, "epoch": 2, "gate_p_mean": 0.038590700773056597, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8015259653329849, "gate_p_p10": 0.005680656433105469, "gate_p_p50": 0.016368865966796875, "gate_p_p90": 0.09919891357421876, "gate_p_std": 0.05643967702053487, "gated_delta_h_c_norm": 0.009278489332064056, "global_step": 260, "grad_norm_max": 0.04130718857049942, "hidden_mse": 0.013163421547506006, "loss": 0.012030859634978698, "lr_blocks": 9.826044551386743e-06, "lr_fusion": 9.826044551386744e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70205055475235, "transport_residual_norm": 0.035850451048463586, "velocity_mse": 0.010714517510496081} +{"anchor_distance_mean": 128.56953125, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.5004441065029823, "delta_h_c_norm": 0.07772279679775239, "delta_h_d_norm": 0.03943826127797365, "epoch": 3, "gate_p_mean": 0.03144416407449171, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8436468608677388, "gate_p_p10": 0.004585456848144531, "gate_p_p50": 0.01262664794921875, "gate_p_p90": 0.08294339179992677, "gate_p_std": 0.04636198218213394, "gated_delta_h_c_norm": 0.00697747731464915, "global_step": 270, "grad_norm_max": 0.031412165611982346, "hidden_mse": 0.012811407243134454, "loss": 0.011699121119454503, "lr_blocks": 9.803768380684242e-06, "lr_fusion": 9.803768380684242e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701973021030426, "transport_residual_norm": 0.03615835523232818, "velocity_mse": 0.010417980345664546} +{"anchor_distance_mean": 124.86562500000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4037478524056497, "delta_h_c_norm": 0.0812199360691011, "delta_h_d_norm": 0.03940788230393082, "epoch": 3, "gate_p_mean": 0.037592848244821654, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8036183148622513, "gate_p_p10": 0.005355529673397541, "gate_p_p50": 0.016916275024414062, "gate_p_p90": 0.09806332588195801, "gate_p_std": 0.05271991329500452, "gated_delta_h_c_norm": 0.00798715755736339, "global_step": 280, "grad_norm_max": 0.031035277992486954, "hidden_mse": 0.013956211262848229, "loss": 0.013469975639600307, "lr_blocks": 9.780178907671788e-06, "lr_fusion": 9.780178907671789e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70261925458908, "transport_residual_norm": 0.037364014191553, "velocity_mse": 0.01207435458782129} +{"anchor_distance_mean": 124.60703125, "avg_context_frames": 16.0, "avg_micro_time_s": 2.691459426970687, "delta_h_c_norm": 0.08519548522308469, "delta_h_d_norm": 0.04071873670909554, "epoch": 3, "gate_p_mean": 0.033276081981603055, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8239360705018044, "gate_p_p10": 0.004834175109863281, "gate_p_p50": 0.014537429809570313, "gate_p_p90": 0.08610501289367677, "gate_p_std": 0.04723041186807678, "gated_delta_h_c_norm": 0.007917539096524706, "global_step": 290, "grad_norm_max": 0.03082125261425972, "hidden_mse": 0.013293361401883886, "loss": 0.011601669166702779, "lr_blocks": 9.755282581475769e-06, "lr_fusion": 9.755282581475769e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702582967281343, "transport_residual_norm": 0.03779523042030633, "velocity_mse": 0.010272333020111546} +{"anchor_distance_mean": 123.17578125, "avg_context_frames": 13.8, "avg_micro_time_s": 2.4731935670090026, "delta_h_c_norm": 0.08512154342606665, "delta_h_d_norm": 0.03896382234524936, "epoch": 3, "gate_p_mean": 0.03331125916447491, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8321614846587182, "gate_p_p10": 0.004003715515136719, "gate_p_p50": 0.01443634033203125, "gate_p_p90": 0.08876953125, "gate_p_std": 0.04770951481768862, "gated_delta_h_c_norm": 0.007722857908083825, "global_step": 300, "grad_norm_max": 0.05298294126987457, "hidden_mse": 0.013096798432525248, "loss": 0.011808986225514673, "lr_blocks": 9.729086208503174e-06, "lr_fusion": 9.729086208503174e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702684581279755, "transport_residual_norm": 0.03759643747471273, "velocity_mse": 0.010499306343263015} +{"anchor_distance_mean": 125.38515625000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.632139423755871, "delta_h_c_norm": 0.08640313567593694, "delta_h_d_norm": 0.04063097126781941, "epoch": 3, "gate_p_mean": 0.027869600721169265, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8625641278922558, "gate_p_p10": 0.003509616851806641, "gate_p_p50": 0.012207412719726564, "gate_p_p90": 0.07377302646636963, "gate_p_std": 0.040450853720540185, "gated_delta_h_c_norm": 0.006470773777982686, "global_step": 310, "grad_norm_max": 0.022413307800889015, "hidden_mse": 0.01179097190615721, "loss": 0.010800209167064168, "lr_blocks": 9.701596950580807e-06, "lr_fusion": 9.701596950580806e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702717578411102, "transport_residual_norm": 0.03773105256259442, "velocity_mse": 0.009621111903106794} +{"anchor_distance_mean": 123.53359375000001, "avg_context_frames": 13.0, "avg_micro_time_s": 2.385848616831936, "delta_h_c_norm": 0.09134956458583475, "delta_h_d_norm": 0.04331219564191997, "epoch": 3, "gate_p_mean": 0.032703211647458375, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8351715505123138, "gate_p_p10": 0.0035376927815377714, "gate_p_p50": 0.0131988525390625, "gate_p_p90": 0.08848140239715577, "gate_p_std": 0.048794481274671855, "gated_delta_h_c_norm": 0.008268705572118052, "global_step": 320, "grad_norm_max": 0.023785248398780823, "hidden_mse": 0.013471112959086895, "loss": 0.012711388431489468, "lr_blocks": 9.672822322997305e-06, "lr_fusion": 9.672822322997305e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70261037349701, "transport_residual_norm": 0.03797716032713652, "velocity_mse": 0.011364277137909086} +{"anchor_distance_mean": 125.25625000000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.441276194159582, "delta_h_c_norm": 0.09453743286430837, "delta_h_d_norm": 0.045107904961332684, "epoch": 3, "gate_p_mean": 0.034259983233641836, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8270665816962719, "gate_p_p10": 0.0036357879638671876, "gate_p_p50": 0.013880348205566407, "gate_p_p90": 0.09431025981903077, "gate_p_std": 0.05051141881849617, "gated_delta_h_c_norm": 0.009006524793221615, "global_step": 330, "grad_norm_max": 0.03133885934948921, "hidden_mse": 0.013715448079165072, "loss": 0.013054096142877825, "lr_blocks": 9.642770192448537e-06, "lr_fusion": 9.642770192448536e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70267858505249, "transport_residual_norm": 0.03865136564709246, "velocity_mse": 0.01168255136872176} +{"anchor_distance_mean": 121.87734375000001, "avg_context_frames": 16.6, "avg_micro_time_s": 2.782007356415852, "delta_h_c_norm": 0.09649814125150442, "delta_h_d_norm": 0.043247889587655665, "epoch": 3, "gate_p_mean": 0.02797943878103979, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8543928012251855, "gate_p_p10": 0.00279083251953125, "gate_p_p50": 0.0112701416015625, "gate_p_p90": 0.07616729736328126, "gate_p_std": 0.041136164264753465, "gated_delta_h_c_norm": 0.007595414738170803, "global_step": 340, "grad_norm_max": 0.02135653607547283, "hidden_mse": 0.01192093031713739, "loss": 0.011057994764996693, "lr_blocks": 9.611448774886925e-06, "lr_fusion": 9.611448774886924e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702652752399445, "transport_residual_norm": 0.03902203994803131, "velocity_mse": 0.009865901741432027} +{"anchor_distance_mean": 124.28046875000001, "avg_context_frames": 12.200000000000001, "avg_micro_time_s": 2.312545049477194, "delta_h_c_norm": 0.09826634312048554, "delta_h_d_norm": 0.04661269546486438, "epoch": 3, "gate_p_mean": 0.0368833584128879, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8077566877007485, "gate_p_p10": 0.003362006787210703, "gate_p_p50": 0.014136695861816408, "gate_p_p90": 0.10161546468734742, "gate_p_std": 0.05573405317263678, "gated_delta_h_c_norm": 0.010448777649435215, "global_step": 350, "grad_norm_max": 0.03365763649344444, "hidden_mse": 0.013092700607376174, "loss": 0.012416656786808745, "lr_blocks": 9.578866633275289e-06, "lr_fusion": 9.578866633275288e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703021597862245, "transport_residual_norm": 0.0387838120572269, "velocity_mse": 0.011107386730145663} +{"anchor_distance_mean": 124.18359375, "avg_context_frames": 13.4, "avg_micro_time_s": 2.423264932188613, "delta_h_c_norm": 0.09478169362992049, "delta_h_d_norm": 0.044269200344569985, "epoch": 3, "gate_p_mean": 0.03240951013867743, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8337490245699883, "gate_p_p10": 0.0026590347290039065, "gate_p_p50": 0.011769294738769531, "gate_p_p90": 0.09195828437805176, "gate_p_std": 0.05059328392380849, "gated_delta_h_c_norm": 0.008758816489716992, "global_step": 360, "grad_norm_max": 0.025236016139388084, "hidden_mse": 0.013239140040241182, "loss": 0.012650595244485885, "lr_blocks": 9.545032675245814e-06, "lr_fusion": 9.545032675245813e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702556824684143, "transport_residual_norm": 0.03868572735227645, "velocity_mse": 0.011326681196806021} +{"anchor_distance_mean": 128.6984375, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.5281798244279345, "delta_h_c_norm": 0.09853337416425348, "delta_h_d_norm": 0.04300247374922037, "epoch": 3, "gate_p_mean": 0.023310313344700264, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8867928966879846, "gate_p_p10": 0.0019213199615478516, "gate_p_p50": 0.007889556884765626, "gate_p_p90": 0.06526222229003907, "gate_p_std": 0.03918320141092409, "gated_delta_h_c_norm": 0.006772257117700065, "global_step": 370, "grad_norm_max": 0.020360825583338737, "hidden_mse": 0.01143520359764807, "loss": 0.010849366610636936, "lr_blocks": 9.509956150664796e-06, "lr_fusion": 9.509956150664796e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702843034267426, "transport_residual_norm": 0.03941694661043585, "velocity_mse": 0.009705846224096605} +{"anchor_distance_mean": 120.96796875000001, "avg_context_frames": 15.8, "avg_micro_time_s": 2.702755107378471, "delta_h_c_norm": 0.10375783918425441, "delta_h_d_norm": 0.04534613741561771, "epoch": 3, "gate_p_mean": 0.02274757420527749, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8882609955966473, "gate_p_p10": 0.002256917953491211, "gate_p_p50": 0.009273147583007813, "gate_p_p90": 0.060927081108093264, "gate_p_std": 0.03447842581954319, "gated_delta_h_c_norm": 0.006431832295857021, "global_step": 380, "grad_norm_max": 0.016434332355856895, "hidden_mse": 0.011230534914648161, "loss": 0.010453642345964909, "lr_blocks": 9.473646649103819e-06, "lr_fusion": 9.473646649103818e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70299940109253, "transport_residual_norm": 0.040361832221969964, "velocity_mse": 0.009330588841112332} +{"anchor_distance_mean": 126.2625, "avg_context_frames": 14.8, "avg_micro_time_s": 2.5847810013830896, "delta_h_c_norm": 0.0998958175070584, "delta_h_d_norm": 0.04568759370595217, "epoch": 3, "gate_p_mean": 0.027090775340911935, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8712157219648362, "gate_p_p10": 0.002134561538696289, "gate_p_p50": 0.009336280822753907, "gate_p_p90": 0.07853260040283204, "gate_p_std": 0.04432942512212321, "gated_delta_h_c_norm": 0.008740647130616707, "global_step": 390, "grad_norm_max": 0.029003798961639404, "hidden_mse": 0.012296237976988778, "loss": 0.011625287146307528, "lr_blocks": 9.43611409721806e-06, "lr_fusion": 9.43611409721806e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703137290477754, "transport_residual_norm": 0.0398570123128593, "velocity_mse": 0.010395663403323853} +{"anchor_distance_mean": 123.53515625, "avg_context_frames": 11.0, "avg_micro_time_s": 2.174077699468762, "delta_h_c_norm": 0.10002230973914267, "delta_h_d_norm": 0.04647535383701325, "epoch": 4, "gate_p_mean": 0.03368478333577514, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8232800230383873, "gate_p_p10": 0.002810192108154297, "gate_p_p50": 0.013163948059082033, "gate_p_p90": 0.0948421597480774, "gate_p_std": 0.052031598694156855, "gated_delta_h_c_norm": 0.009592664884985426, "global_step": 400, "grad_norm_max": 0.02631869725883007, "hidden_mse": 0.013110433367546649, "loss": 0.012611593597102911, "lr_blocks": 9.397368756032445e-06, "lr_fusion": 9.397368756032445e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70314885377884, "transport_residual_norm": 0.039898452628403906, "velocity_mse": 0.011300550217856654} +{"anchor_distance_mean": 122.625, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.621169192600064, "delta_h_c_norm": 0.1063119161874056, "delta_h_d_norm": 0.047389440611004835, "epoch": 4, "gate_p_mean": 0.029519124602666126, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8478565953671933, "gate_p_p10": 0.002479076385498047, "gate_p_p50": 0.010724067687988281, "gate_p_p90": 0.08300018310546875, "gate_p_std": 0.04654600077774376, "gated_delta_h_c_norm": 0.009603426457033493, "global_step": 410, "grad_norm_max": 0.022217130288481712, "hidden_mse": 0.012233574583660811, "loss": 0.011557986988918856, "lr_blocks": 9.357421218136387e-06, "lr_fusion": 9.357421218136386e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703378033638002, "transport_residual_norm": 0.04070999780669809, "velocity_mse": 0.010334629548015074} +{"anchor_distance_mean": 123.79453125, "avg_context_frames": 14.4, "avg_micro_time_s": 2.554096806587768, "delta_h_c_norm": 0.11048556426540018, "delta_h_d_norm": 0.047132975468412046, "epoch": 4, "gate_p_mean": 0.026935733313439417, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8663777321577073, "gate_p_p10": 0.002042484283447266, "gate_p_p50": 0.009806442260742189, "gate_p_p90": 0.07638931274414062, "gate_p_std": 0.04285629790974781, "gated_delta_h_c_norm": 0.008620003602845828, "global_step": 420, "grad_norm_max": 0.01920047029852867, "hidden_mse": 0.012511257902951912, "loss": 0.012182446516817436, "lr_blocks": 9.31628240478787e-06, "lr_fusion": 9.316282404787871e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702891194820404, "transport_residual_norm": 0.04091003742069006, "velocity_mse": 0.010931320799863899} +{"anchor_distance_mean": 123.30781250000001, "avg_context_frames": 14.8, "avg_micro_time_s": 2.5825302017328795, "delta_h_c_norm": 0.1108503919094801, "delta_h_d_norm": 0.04918666719458997, "epoch": 4, "gate_p_mean": 0.02821205026120879, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8510592244565487, "gate_p_p10": 0.002040338516235352, "gate_p_p50": 0.010316562652587891, "gate_p_p90": 0.07881317138671876, "gate_p_std": 0.04563800527830608, "gated_delta_h_c_norm": 0.008997123196604662, "global_step": 430, "grad_norm_max": 0.021192902699112892, "hidden_mse": 0.01197292550932616, "loss": 0.011530985543504358, "lr_blocks": 9.273963562927695e-06, "lr_fusion": 9.273963562927695e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703501343727112, "transport_residual_norm": 0.04116274686530233, "velocity_mse": 0.010333692983840593} +{"anchor_distance_mean": 126.16328125000001, "avg_context_frames": 16.0, "avg_micro_time_s": 2.7118577869579896, "delta_h_c_norm": 0.10742650032043458, "delta_h_d_norm": 0.04553067232482136, "epoch": 4, "gate_p_mean": 0.02281554805231281, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8837700605392457, "gate_p_p10": 0.001787853240966797, "gate_p_p50": 0.007862472534179687, "gate_p_p90": 0.06436119079589844, "gate_p_std": 0.037177806725958364, "gated_delta_h_c_norm": 0.0071698388870572675, "global_step": 440, "grad_norm_max": 0.02410629391670227, "hidden_mse": 0.011045475269202144, "loss": 0.01059189329680521, "lr_blocks": 9.230476262104678e-06, "lr_fusion": 9.230476262104677e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703341102600097, "transport_residual_norm": 0.04157367204315961, "velocity_mse": 0.00948734585836064} +{"anchor_distance_mean": 128.34140625, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4452929514678545, "delta_h_c_norm": 0.11123060202226043, "delta_h_d_norm": 0.04898357638157905, "epoch": 4, "gate_p_mean": 0.030386892607202756, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8348908491432667, "gate_p_p10": 0.0024024963378906253, "gate_p_p50": 0.012067222595214845, "gate_p_p90": 0.08489761352539063, "gate_p_std": 0.04531522264005616, "gated_delta_h_c_norm": 0.009347341890679673, "global_step": 450, "grad_norm_max": 0.022731559351086617, "hidden_mse": 0.012265758676221594, "loss": 0.011833904156810605, "lr_blocks": 9.185832391312644e-06, "lr_fusion": 9.185832391312644e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703083312511444, "transport_residual_norm": 0.04153583385050297, "velocity_mse": 0.010607328257174232} +{"anchor_distance_mean": 125.74296875, "avg_context_frames": 14.8, "avg_micro_time_s": 2.5642772476523534, "delta_h_c_norm": 0.11180831622332335, "delta_h_d_norm": 0.048905190918594604, "epoch": 4, "gate_p_mean": 0.02874150507268496, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8542485728859902, "gate_p_p10": 0.0015736579895019533, "gate_p_p50": 0.008021736145019531, "gate_p_p90": 0.08719792366027833, "gate_p_std": 0.04884456892614253, "gated_delta_h_c_norm": 0.010203022310452071, "global_step": 460, "grad_norm_max": 0.02285076305270195, "hidden_mse": 0.012351714249234647, "loss": 0.011499692685902119, "lr_blocks": 9.140044155740102e-06, "lr_fusion": 9.140044155740101e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703203773498535, "transport_residual_norm": 0.04162424448877573, "velocity_mse": 0.010264521301724018} +{"anchor_distance_mean": 125.353125, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.4973248058129687, "delta_h_c_norm": 0.10761532941833139, "delta_h_d_norm": 0.05031369426287711, "epoch": 4, "gate_p_mean": 0.028043231216724963, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8524091318249702, "gate_p_p10": 0.0018127441406250002, "gate_p_p50": 0.009852695465087892, "gate_p_p90": 0.0828094482421875, "gate_p_std": 0.04274746457813308, "gated_delta_h_c_norm": 0.008201093983370811, "global_step": 470, "grad_norm_max": 0.024406779557466507, "hidden_mse": 0.01215514283394441, "loss": 0.011039524385705591, "lr_blocks": 9.093124073433464e-06, "lr_fusion": 9.093124073433463e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70342506170273, "transport_residual_norm": 0.04220685004256666, "velocity_mse": 0.009824010063312016} +{"anchor_distance_mean": 124.11953125000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6211049844583614, "delta_h_c_norm": 0.10741831297054888, "delta_h_d_norm": 0.04808014906011522, "epoch": 4, "gate_p_mean": 0.024227660160977396, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8704347215592861, "gate_p_p10": 0.0014694690704345704, "gate_p_p50": 0.008067131042480469, "gate_p_p90": 0.07017593383789063, "gate_p_std": 0.039806720247725025, "gated_delta_h_c_norm": 0.00743150904745562, "global_step": 480, "grad_norm_max": 0.0220582727342844, "hidden_mse": 0.011598977952962742, "loss": 0.011055564909474925, "lr_blocks": 9.045084971874738e-06, "lr_fusion": 9.045084971874738e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703603494167329, "transport_residual_norm": 0.04202401395887137, "velocity_mse": 0.00989566715725232} +{"anchor_distance_mean": 121.65, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.549194080222515, "delta_h_c_norm": 0.11493653990328312, "delta_h_d_norm": 0.05090167326852679, "epoch": 4, "gate_p_mean": 0.029579623084282504, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8453954108059407, "gate_p_p10": 0.0017007827758789063, "gate_p_p50": 0.009797286987304688, "gate_p_p90": 0.08823642730712891, "gate_p_std": 0.04635096022393555, "gated_delta_h_c_norm": 0.010058953402767657, "global_step": 490, "grad_norm_max": 0.025631099939346313, "hidden_mse": 0.011866300262045116, "loss": 0.010911043267697096, "lr_blocks": 8.995939984474624e-06, "lr_fusion": 8.995939984474624e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703456270694733, "transport_residual_norm": 0.041157411085441714, "velocity_mse": 0.009724413228104823} +{"anchor_distance_mean": 123.17500000000001, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.3866983174841154, "delta_h_c_norm": 0.1142569174990058, "delta_h_d_norm": 0.05026759901084006, "epoch": 4, "gate_p_mean": 0.026511165878036994, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8630544118583203, "gate_p_p10": 0.0015011787414550783, "gate_p_p50": 0.008347034454345703, "gate_p_p90": 0.07757954597473145, "gate_p_std": 0.044232423865469175, "gated_delta_h_c_norm": 0.009444151548814263, "global_step": 500, "grad_norm_max": 0.033091139048337936, "hidden_mse": 0.01198805258027278, "loss": 0.011321621094248259, "lr_blocks": 8.94570254698197e-06, "lr_fusion": 8.945702546981969e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703399968147279, "transport_residual_norm": 0.04204175896011293, "velocity_mse": 0.010122815813519993} +{"anchor_distance_mean": 128.2125, "avg_context_frames": 14.4, "avg_micro_time_s": 2.542909958913515, "delta_h_c_norm": 0.11850062115117908, "delta_h_d_norm": 0.05188623992726207, "epoch": 4, "gate_p_mean": 0.02377685899846256, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8782121643424035, "gate_p_p10": 0.001370382308959961, "gate_p_p50": 0.007573318481445313, "gate_p_p90": 0.07033538818359375, "gate_p_std": 0.03912990307435393, "gated_delta_h_c_norm": 0.00889493281574687, "global_step": 510, "grad_norm_max": 0.03324056416749954, "hidden_mse": 0.011761842895066366, "loss": 0.01128033324785065, "lr_blocks": 8.894386393810563e-06, "lr_fusion": 8.894386393810563e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703658759593964, "transport_residual_norm": 0.042084207572042946, "velocity_mse": 0.010104148945538328} +{"anchor_distance_mean": 126.78046875000001, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.5633915127196816, "delta_h_c_norm": 0.11437936872243881, "delta_h_d_norm": 0.04997560069896281, "epoch": 4, "gate_p_mean": 0.02725226672482677, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8627076089382172, "gate_p_p10": 0.0015118598937988283, "gate_p_p50": 0.008024311065673828, "gate_p_p90": 0.08351526260375977, "gate_p_std": 0.04577681628870778, "gated_delta_h_c_norm": 0.009597504105477129, "global_step": 520, "grad_norm_max": 0.02666490338742733, "hidden_mse": 0.012287548329913989, "loss": 0.011851613403996453, "lr_blocks": 8.842005554284296e-06, "lr_fusion": 8.842005554284296e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703446841239929, "transport_residual_norm": 0.041174311935901646, "velocity_mse": 0.010622858541319148} +{"anchor_distance_mean": 126.00312500000001, "avg_context_frames": 12.8, "avg_micro_time_s": 2.3869871570088437, "delta_h_c_norm": 0.11425890261307359, "delta_h_d_norm": 0.05121397851034999, "epoch": 5, "gate_p_mean": 0.02802812572626863, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8467127695679665, "gate_p_p10": 0.0016654967796057463, "gate_p_p50": 0.008957481384277344, "gate_p_p90": 0.08354120254516602, "gate_p_std": 0.044655832799617205, "gated_delta_h_c_norm": 0.009449811796366704, "global_step": 530, "grad_norm_max": 0.02442295104265213, "hidden_mse": 0.012854953831993044, "loss": 0.01186972778523341, "lr_blocks": 8.788574348801676e-06, "lr_fusion": 8.788574348801675e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703398048877716, "transport_residual_norm": 0.04096401012502611, "velocity_mse": 0.010584232502151281} +{"anchor_distance_mean": 125.31875000000001, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.6570763659939987, "delta_h_c_norm": 0.11860605077818037, "delta_h_d_norm": 0.05230778073891998, "epoch": 5, "gate_p_mean": 0.024631775481975638, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.869432620704174, "gate_p_p10": 0.0017627239227294923, "gate_p_p50": 0.008446407318115235, "gate_p_p90": 0.07203636169433594, "gate_p_std": 0.0388124206539942, "gated_delta_h_c_norm": 0.00843233081213839, "global_step": 540, "grad_norm_max": 0.02424430102109909, "hidden_mse": 0.011338632309343667, "loss": 0.01064014234871138, "lr_blocks": 8.734107384920771e-06, "lr_fusion": 8.73410738492077e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703717839717866, "transport_residual_norm": 0.0420270293019712, "velocity_mse": 0.009506279122433625} +{"anchor_distance_mean": 123.30625, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.551804949983489, "delta_h_c_norm": 0.12305296305567026, "delta_h_d_norm": 0.050221567647531634, "epoch": 5, "gate_p_mean": 0.027664801111677663, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8512187242507935, "gate_p_p10": 0.001523113250732422, "gate_p_p50": 0.008370208740234376, "gate_p_p90": 0.08137292861938478, "gate_p_std": 0.04547325099119917, "gated_delta_h_c_norm": 0.010267541524081027, "global_step": 550, "grad_norm_max": 0.030247187241911888, "hidden_mse": 0.012458556162891911, "loss": 0.012068655260372908, "lr_blocks": 8.67861955336566e-06, "lr_fusion": 8.678619553365659e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703580856323242, "transport_residual_norm": 0.04214459573850036, "velocity_mse": 0.010822799627203495} +{"anchor_distance_mean": 126.42500000000001, "avg_context_frames": 14.4, "avg_micro_time_s": 2.535691819734348, "delta_h_c_norm": 0.12008428517729045, "delta_h_d_norm": 0.05118598518893123, "epoch": 5, "gate_p_mean": 0.02806141014734749, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8477268889546394, "gate_p_p10": 0.0015626907348632814, "gate_p_p50": 0.009404945373535157, "gate_p_p90": 0.08400247395038606, "gate_p_std": 0.04361232834053226, "gated_delta_h_c_norm": 0.009295186278177426, "global_step": 560, "grad_norm_max": 0.021833473816514015, "hidden_mse": 0.011841632088180632, "loss": 0.011702356798923575, "lr_blocks": 8.622126023955446e-06, "lr_fusion": 8.622126023955446e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703773367404938, "transport_residual_norm": 0.04225303526036442, "velocity_mse": 0.010518193594180048} +{"anchor_distance_mean": 125.90625, "avg_context_frames": 16.2, "avg_micro_time_s": 2.726909366955806, "delta_h_c_norm": 0.118103049043566, "delta_h_d_norm": 0.05081854602321983, "epoch": 5, "gate_p_mean": 0.02098318034113618, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8880381397902966, "gate_p_p10": 0.0008576512336730957, "gate_p_p50": 0.005148839950561524, "gate_p_p90": 0.06250288486480714, "gate_p_std": 0.03826332924654707, "gated_delta_h_c_norm": 0.008198262692167191, "global_step": 570, "grad_norm_max": 0.020829254761338234, "hidden_mse": 0.011867230117786676, "loss": 0.011323085305048153, "lr_blocks": 8.564642241456986e-06, "lr_fusion": 8.564642241456986e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703350770473481, "transport_residual_norm": 0.04225750761106611, "velocity_mse": 0.010136362269986422} +{"anchor_distance_mean": 126.00312500000001, "avg_context_frames": 12.4, "avg_micro_time_s": 2.307725720430608, "delta_h_c_norm": 0.11983012752607465, "delta_h_d_norm": 0.04967220234684647, "epoch": 5, "gate_p_mean": 0.02625572074321099, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8589813984930516, "gate_p_p10": 0.0011458158493041993, "gate_p_p50": 0.007444000244140625, "gate_p_p90": 0.08201522827148439, "gate_p_std": 0.04468674865784124, "gated_delta_h_c_norm": 0.009590926085365936, "global_step": 580, "grad_norm_max": 0.02104531228542328, "hidden_mse": 0.01155740823596716, "loss": 0.010974727891152725, "lr_blocks": 8.506183921362443e-06, "lr_fusion": 8.506183921362443e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703921341896057, "transport_residual_norm": 0.04106026142835617, "velocity_mse": 0.009818987085600384} +{"anchor_distance_mean": 124.44375000000001, "avg_context_frames": 12.0, "avg_micro_time_s": 2.258997691127297, "delta_h_c_norm": 0.12200861005112529, "delta_h_d_norm": 0.054378376342356205, "epoch": 5, "gate_p_mean": 0.02801791619276628, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8563041128218174, "gate_p_p10": 0.0010315489606000483, "gate_p_p50": 0.007282066345214844, "gate_p_p90": 0.08664779663085938, "gate_p_std": 0.04887788774503861, "gated_delta_h_c_norm": 0.011200811586604686, "global_step": 590, "grad_norm_max": 0.024945871904492378, "hidden_mse": 0.012746391334803776, "loss": 0.012366789390216583, "lr_blocks": 8.446767045592829e-06, "lr_fusion": 8.44676704559283e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70394184589386, "transport_residual_norm": 0.04117898447439075, "velocity_mse": 0.011092150310287253} +{"anchor_distance_mean": 126.48828125, "avg_context_frames": 16.0, "avg_micro_time_s": 2.709427968610544, "delta_h_c_norm": 0.11349966060370208, "delta_h_d_norm": 0.04945375160314143, "epoch": 5, "gate_p_mean": 0.018460078322095796, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.9023375168442727, "gate_p_p10": 0.0008529901504516602, "gate_p_p50": 0.004498481750488281, "gate_p_p90": 0.05335049629211426, "gate_p_std": 0.0350904660532251, "gated_delta_h_c_norm": 0.006970201832154999, "global_step": 600, "grad_norm_max": 0.01684447005391121, "hidden_mse": 0.010297420975985006, "loss": 0.009851792757399381, "lr_blocks": 8.386407858128707e-06, "lr_fusion": 8.386407858128706e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703947961330414, "transport_residual_norm": 0.04154843389987946, "velocity_mse": 0.00882205069938209} +{"anchor_distance_mean": 126.16484375, "avg_context_frames": 16.0, "avg_micro_time_s": 2.705890699000156, "delta_h_c_norm": 0.11495247129350901, "delta_h_d_norm": 0.05109813697636128, "epoch": 5, "gate_p_mean": 0.01933139872562606, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.9005236133933068, "gate_p_p10": 0.001025843620300293, "gate_p_p50": 0.005321121215820313, "gate_p_p90": 0.057401978969573976, "gate_p_std": 0.03303498222667258, "gated_delta_h_c_norm": 0.00670583105456899, "global_step": 610, "grad_norm_max": 0.019726835191249847, "hidden_mse": 0.010779428246314638, "loss": 0.010325037714210339, "lr_blocks": 8.325122860569241e-06, "lr_fusion": 8.32512286056924e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703876197338104, "transport_residual_norm": 0.042256176192313435, "velocity_mse": 0.009247094925376587} +{"anchor_distance_mean": 121.5546875, "avg_context_frames": 11.4, "avg_micro_time_s": 2.1950611387961545, "delta_h_c_norm": 0.13120215386152267, "delta_h_d_norm": 0.054730232991278174, "epoch": 5, "gate_p_mean": 0.029424212221056224, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8229006692767143, "gate_p_p10": 0.0016769886016845704, "gate_p_p50": 0.01133098602294922, "gate_p_p90": 0.08655743598937989, "gate_p_std": 0.04253805003827438, "gated_delta_h_c_norm": 0.009691972860309761, "global_step": 620, "grad_norm_max": 0.018521513789892197, "hidden_mse": 0.013325474754674362, "loss": 0.012826383905485274, "lr_blocks": 8.262928807620843e-06, "lr_fusion": 8.262928807620843e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703971815109254, "transport_residual_norm": 0.04269666662439704, "velocity_mse": 0.011493836477166043} +{"anchor_distance_mean": 122.07421875, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4438838971094814, "delta_h_c_norm": 0.1392668561078608, "delta_h_d_norm": 0.05347658726386726, "epoch": 5, "gate_p_mean": 0.026077269110828638, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8496714994311333, "gate_p_p10": 0.001187252998352051, "gate_p_p50": 0.008634662628173828, "gate_p_p90": 0.07769149541854858, "gate_p_std": 0.0412054842163343, "gated_delta_h_c_norm": 0.01010455036885105, "global_step": 630, "grad_norm_max": 0.02600320242345333, "hidden_mse": 0.01201830497593619, "loss": 0.011793932912405582, "lr_blocks": 8.199842702516584e-06, "lr_fusion": 8.199842702516583e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70367033481598, "transport_residual_norm": 0.042865849845111374, "velocity_mse": 0.010592102410737426} +{"anchor_distance_mean": 128.18046875000002, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6096932063912393, "delta_h_c_norm": 0.1334919217042625, "delta_h_d_norm": 0.05297484528273344, "epoch": 5, "gate_p_mean": 0.02089140011521522, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.889479947090149, "gate_p_p10": 0.0007473945617675782, "gate_p_p50": 0.004884052276611328, "gate_p_p90": 0.0628434181213379, "gate_p_std": 0.03801634153351188, "gated_delta_h_c_norm": 0.009364405534870458, "global_step": 640, "grad_norm_max": 0.01636059395968914, "hidden_mse": 0.011282653664238751, "loss": 0.010822400928009302, "lr_blocks": 8.135881792367686e-06, "lr_fusion": 8.135881792367686e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.704022943973541, "transport_residual_norm": 0.042375073721632364, "velocity_mse": 0.009694135602330789} +{"anchor_distance_mean": 121.321875, "avg_context_frames": 15.8, "avg_micro_time_s": 2.684451166499639, "delta_h_c_norm": 0.13335950179025532, "delta_h_d_norm": 0.052019649650901556, "epoch": 5, "gate_p_mean": 0.023020902634016238, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8768362157046795, "gate_p_p10": 0.0012968825176358223, "gate_p_p50": 0.007397556304931641, "gate_p_p90": 0.06600178182125092, "gate_p_std": 0.03798849315789994, "gated_delta_h_c_norm": 0.009156512754270808, "global_step": 650, "grad_norm_max": 0.02050534449517727, "hidden_mse": 0.01108599966391921, "loss": 0.010270757711259649, "lr_blocks": 8.071063563448341e-06, "lr_fusion": 8.07106356344834e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.704257571697235, "transport_residual_norm": 0.04185098786838353, "velocity_mse": 0.00916215771867428} +{"anchor_distance_mean": 121.45703125, "avg_context_frames": 14.0, "avg_micro_time_s": 2.498486790643074, "delta_h_c_norm": 0.1340931651182473, "delta_h_d_norm": 0.05656354171223939, "epoch": 6, "gate_p_mean": 0.029963309830054644, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8309237517416478, "gate_p_p10": 0.001363573013804853, "gate_p_p50": 0.009370040893554688, "gate_p_p90": 0.09025039672851563, "gate_p_std": 0.04840789155568928, "gated_delta_h_c_norm": 0.011710750808560988, "global_step": 660, "grad_norm_max": 0.02080424316227436, "hidden_mse": 0.012682802142808215, "loss": 0.01199083934770897, "lr_blocks": 8.005405736415127e-06, "lr_fusion": 8.005405736415126e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703916251659393, "transport_residual_norm": 0.04207512075081468, "velocity_mse": 0.01072255910257809} +{"anchor_distance_mean": 124.66953125, "avg_context_frames": 14.0, "avg_micro_time_s": 2.486254331028613, "delta_h_c_norm": 0.12833140473812818, "delta_h_d_norm": 0.05427114367485047, "epoch": 6, "gate_p_mean": 0.021218744170619176, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8855456504970789, "gate_p_p10": 0.0009641003329306842, "gate_p_p50": 0.0066265583038330085, "gate_p_p90": 0.061822509765625006, "gate_p_std": 0.03589212148799561, "gated_delta_h_c_norm": 0.007970611182099675, "global_step": 670, "grad_norm_max": 0.023931629955768585, "hidden_mse": 0.011341245414223523, "loss": 0.01115665729448665, "lr_blocks": 7.938926261462366e-06, "lr_fusion": 7.938926261462366e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703433680534364, "transport_residual_norm": 0.04308734061196447, "velocity_mse": 0.010022532744915226} +{"anchor_distance_mean": 122.69140625, "avg_context_frames": 15.8, "avg_micro_time_s": 2.681947380716156, "delta_h_c_norm": 0.1368733993731439, "delta_h_d_norm": 0.052373165078461174, "epoch": 6, "gate_p_mean": 0.016952627782302443, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.9135915242135525, "gate_p_p10": 0.0006178379058837891, "gate_p_p50": 0.004228401184082031, "gate_p_p90": 0.051710331439971925, "gate_p_std": 0.03120165852305945, "gated_delta_h_c_norm": 0.007432899634295609, "global_step": 680, "grad_norm_max": 0.016307979822158813, "hidden_mse": 0.010479091567685828, "loss": 0.009842504883999938, "lr_blocks": 7.871643313414718e-06, "lr_fusion": 7.871643313414718e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703311908245087, "transport_residual_norm": 0.04480669633485377, "velocity_mse": 0.008794595740619115} +{"anchor_distance_mean": 124.63984375000001, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.6578954583732415, "delta_h_c_norm": 0.139674330316484, "delta_h_d_norm": 0.056358848046511416, "epoch": 6, "gate_p_mean": 0.02996676145703532, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8396827660501004, "gate_p_p10": 0.0014117574552074076, "gate_p_p50": 0.0091644287109375, "gate_p_p90": 0.09227375984191895, "gate_p_std": 0.04708224586793222, "gated_delta_h_c_norm": 0.012495446913089837, "global_step": 690, "grad_norm_max": 0.021534258499741554, "hidden_mse": 0.01299889405781869, "loss": 0.01221597051480785, "lr_blocks": 7.803575286758365e-06, "lr_fusion": 7.803575286758364e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703536927700043, "transport_residual_norm": 0.04261181093752384, "velocity_mse": 0.010916081145114732} +{"anchor_distance_mean": 125.70937500000001, "avg_context_frames": 13.8, "avg_micro_time_s": 2.4705268745092326, "delta_h_c_norm": 0.13355242796242237, "delta_h_d_norm": 0.05298158307559789, "epoch": 6, "gate_p_mean": 0.02549795288941823, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8687149703502656, "gate_p_p10": 0.000796198844909668, "gate_p_p50": 0.00538330078125, "gate_p_p90": 0.08204846382141114, "gate_p_std": 0.04698595065856353, "gated_delta_h_c_norm": 0.01108267149757012, "global_step": 700, "grad_norm_max": 0.02319071628153324, "hidden_mse": 0.012164107541320846, "loss": 0.011699333123397083, "lr_blocks": 7.734740790612137e-06, "lr_fusion": 7.734740790612136e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703713750839233, "transport_residual_norm": 0.04187183324247599, "velocity_mse": 0.010482922338997015} +{"anchor_distance_mean": 129.18437500000002, "avg_context_frames": 15.8, "avg_micro_time_s": 2.683608693497081, "delta_h_c_norm": 0.1300450298935175, "delta_h_d_norm": 0.05109789143316448, "epoch": 6, "gate_p_mean": 0.02385606285533868, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8720806084573269, "gate_p_p10": 0.0010449886322021484, "gate_p_p50": 0.006102561950683594, "gate_p_p90": 0.07440729141235351, "gate_p_std": 0.0415030942473095, "gated_delta_h_c_norm": 0.009426701555639738, "global_step": 710, "grad_norm_max": 0.013394320383667946, "hidden_mse": 0.010625481011811645, "loss": 0.010118101318948903, "lr_blocks": 7.66515864363997e-06, "lr_fusion": 7.66515864363997e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703607404232026, "transport_residual_norm": 0.04234149069525302, "velocity_mse": 0.009055553216603585} +{"anchor_distance_mean": 124.0515625, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4149277940145115, "delta_h_c_norm": 0.12949657244607807, "delta_h_d_norm": 0.0558471582364291, "epoch": 6, "gate_p_mean": 0.024009274429408835, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8681791119277478, "gate_p_p10": 0.0008131504058837891, "gate_p_p50": 0.005597209930419922, "gate_p_p90": 0.0751858413219452, "gate_p_std": 0.043801916734082624, "gated_delta_h_c_norm": 0.009654774749651552, "global_step": 720, "grad_norm_max": 0.01737804152071476, "hidden_mse": 0.011077903473051265, "loss": 0.010627077880781145, "lr_blocks": 7.594847868906076e-06, "lr_fusion": 7.594847868906076e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70352989435196, "transport_residual_norm": 0.04268462378531695, "velocity_mse": 0.009519287591683679} +{"anchor_distance_mean": 125.12421875000001, "avg_context_frames": 12.4, "avg_micro_time_s": 2.325140466337325, "delta_h_c_norm": 0.12647595275193455, "delta_h_d_norm": 0.05659400932490826, "epoch": 6, "gate_p_mean": 0.02053226285497658, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8880832117050886, "gate_p_p10": 0.00055159330368042, "gate_p_p50": 0.004742765426635743, "gate_p_p90": 0.06325364112854004, "gate_p_std": 0.037996029343048576, "gated_delta_h_c_norm": 0.008345044300949667, "global_step": 730, "grad_norm_max": 0.01630614884197712, "hidden_mse": 0.011715155304409565, "loss": 0.011248504492687062, "lr_blocks": 7.52382768867422e-06, "lr_fusion": 7.52382768867422e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703278613090516, "transport_residual_norm": 0.043207241455093026, "velocity_mse": 0.010076988954097033} +{"anchor_distance_mean": 119.99531250000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4142893907337566, "delta_h_c_norm": 0.12904534013941885, "delta_h_d_norm": 0.05819487250410021, "epoch": 6, "gate_p_mean": 0.029170895740389825, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8307812750339508, "gate_p_p10": 0.0011406802572309972, "gate_p_p50": 0.00923452377319336, "gate_p_p90": 0.0879339098930359, "gate_p_std": 0.0462025287328288, "gated_delta_h_c_norm": 0.010294122122286354, "global_step": 740, "grad_norm_max": 0.020795900374650955, "hidden_mse": 0.011638219008455054, "loss": 0.011283854977227748, "lr_blocks": 7.452117519152542e-06, "lr_fusion": 7.452117519152542e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703236961364746, "transport_residual_norm": 0.04260387411341071, "velocity_mse": 0.010120033047860489} +{"anchor_distance_mean": 128.56875, "avg_context_frames": 15.4, "avg_micro_time_s": 2.6222933443554215, "delta_h_c_norm": 0.12716153739020228, "delta_h_d_norm": 0.052995657082647085, "epoch": 6, "gate_p_mean": 0.0219279506476596, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8803388193249703, "gate_p_p10": 0.000932531338185072, "gate_p_p50": 0.0058921813964843755, "gate_p_p90": 0.06470210552215576, "gate_p_std": 0.03871944739366882, "gated_delta_h_c_norm": 0.008423958204002702, "global_step": 750, "grad_norm_max": 0.019116029143333435, "hidden_mse": 0.010530787042807788, "loss": 0.010264814118272625, "lr_blocks": 7.379736965185369e-06, "lr_fusion": 7.379736965185368e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703436315059662, "transport_residual_norm": 0.0432699756231159, "velocity_mse": 0.009211735462304206} +{"anchor_distance_mean": 128.50390625, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4412227277076455, "delta_h_c_norm": 0.1348069304600358, "delta_h_d_norm": 0.05459047788754106, "epoch": 6, "gate_p_mean": 0.023937737999949606, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.865582710504532, "gate_p_p10": 0.0007938789669424296, "gate_p_p50": 0.0061512947082519535, "gate_p_p90": 0.07472000122070313, "gate_p_std": 0.041242740047164266, "gated_delta_h_c_norm": 0.009601797470531893, "global_step": 760, "grad_norm_max": 0.030197974294424057, "hidden_mse": 0.01065809324500151, "loss": 0.010446040550596081, "lr_blocks": 7.30670581489344e-06, "lr_fusion": 7.30670581489344e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703099691867829, "transport_residual_norm": 0.04368960945867002, "velocity_mse": 0.009380231247632764} +{"anchor_distance_mean": 123.76015625000001, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.710041312893736, "delta_h_c_norm": 0.13027209537103773, "delta_h_d_norm": 0.05456866528838873, "epoch": 6, "gate_p_mean": 0.018812380157760347, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8972623966634274, "gate_p_p10": 0.000596153736114502, "gate_p_p50": 0.004343175888061524, "gate_p_p90": 0.056766510009765625, "gate_p_std": 0.03619167153374292, "gated_delta_h_c_norm": 0.008575864891463425, "global_step": 770, "grad_norm_max": 0.018458286300301552, "hidden_mse": 0.010839572525583208, "loss": 0.01041844263090752, "lr_blocks": 7.233044034264034e-06, "lr_fusion": 7.233044034264034e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703182816505432, "transport_residual_norm": 0.04233629298396409, "velocity_mse": 0.009334485454019159} +{"anchor_distance_mean": 124.73671875000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.439629926766793, "delta_h_c_norm": 0.13749724673107266, "delta_h_d_norm": 0.05587770021520555, "epoch": 6, "gate_p_mean": 0.031509934284258634, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8227699581533671, "gate_p_p10": 0.0012943267822265625, "gate_p_p50": 0.009550476074218751, "gate_p_p90": 0.1000626564025879, "gate_p_std": 0.04994965101941489, "gated_delta_h_c_norm": 0.012174289831455099, "global_step": 780, "grad_norm_max": 0.021267589181661606, "hidden_mse": 0.011870726599590854, "loss": 0.011733738507609816, "lr_blocks": 7.158771761692464e-06, "lr_fusion": 7.158771761692464e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70296094417572, "transport_residual_norm": 0.04202585392631591, "velocity_mse": 0.010546665819128977} +{"anchor_distance_mean": 125.74375, "avg_context_frames": 13.8, "avg_micro_time_s": 2.508046032910352, "delta_h_c_norm": 0.13457976849749684, "delta_h_d_norm": 0.055214715143665674, "epoch": 7, "gate_p_mean": 0.021498885648907163, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8812397576868535, "gate_p_p10": 0.0006949782371520997, "gate_p_p50": 0.005265426635742188, "gate_p_p90": 0.06676476001739502, "gate_p_std": 0.03852054365561344, "gated_delta_h_c_norm": 0.008366903077694587, "global_step": 790, "grad_norm_max": 0.016480036079883575, "hidden_mse": 0.010779445961816236, "loss": 0.010315197880845518, "lr_blocks": 7.083909302476453e-06, "lr_fusion": 7.083909302476453e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703286302089692, "transport_residual_norm": 0.0436361774802208, "velocity_mse": 0.009237253246828914} +{"anchor_distance_mean": 121.32734375000001, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.52982666320022, "delta_h_c_norm": 0.13939105141907931, "delta_h_d_norm": 0.05891137029975653, "epoch": 7, "gate_p_mean": 0.03091972953407094, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8259500458836556, "gate_p_p10": 0.0011241436004638672, "gate_p_p50": 0.008912467956542968, "gate_p_p90": 0.09745874404907227, "gate_p_std": 0.049235336959827694, "gated_delta_h_c_norm": 0.01184702080936404, "global_step": 800, "grad_norm_max": 0.01832185685634613, "hidden_mse": 0.01269606000278145, "loss": 0.012061346974223853, "lr_blocks": 7.008477123264849e-06, "lr_fusion": 7.008477123264848e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703179860115052, "transport_residual_norm": 0.044859326351434, "velocity_mse": 0.010791740927379579} +{"anchor_distance_mean": 126.328125, "avg_context_frames": 17.400000000000002, "avg_micro_time_s": 2.854499574804504, "delta_h_c_norm": 0.13232997665181756, "delta_h_d_norm": 0.0552015055436641, "epoch": 7, "gate_p_mean": 0.017144265596289187, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.9058306090533734, "gate_p_p10": 0.000641322135925293, "gate_p_p50": 0.004461050033569336, "gate_p_p90": 0.049582517147064215, "gate_p_std": 0.031723690775106665, "gated_delta_h_c_norm": 0.006906842954776948, "global_step": 810, "grad_norm_max": 0.011102044954895973, "hidden_mse": 0.010148722195299343, "loss": 0.009897459213971161, "lr_blocks": 6.932495846462262e-06, "lr_fusion": 6.932495846462261e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703298521041871, "transport_residual_norm": 0.04652409018017352, "velocity_mse": 0.008882587021798827} +{"anchor_distance_mean": 124.73750000000001, "avg_context_frames": 12.600000000000001, "avg_micro_time_s": 2.3244358776966694, "delta_h_c_norm": 0.1412114454433322, "delta_h_d_norm": 0.06018670182675123, "epoch": 7, "gate_p_mean": 0.03463125150301494, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8085361827164889, "gate_p_p10": 0.0012285947799682618, "gate_p_p50": 0.009840011596679688, "gate_p_p90": 0.10693262815475464, "gate_p_std": 0.05669327831128612, "gated_delta_h_c_norm": 0.014362973142124247, "global_step": 820, "grad_norm_max": 0.017931928858160973, "hidden_mse": 0.013368888694094494, "loss": 0.01257837365847081, "lr_blocks": 6.855986244591104e-06, "lr_fusion": 6.855986244591104e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703149783611298, "transport_residual_norm": 0.04150461154058576, "velocity_mse": 0.011241484785568901} +{"anchor_distance_mean": 125.3515625, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.6853507046282177, "delta_h_c_norm": 0.137673581764102, "delta_h_d_norm": 0.054009838681668046, "epoch": 7, "gate_p_mean": 0.02188439475139603, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8771690003573895, "gate_p_p10": 0.0010380268096923829, "gate_p_p50": 0.0062316894531250005, "gate_p_p90": 0.06768476963043213, "gate_p_std": 0.03673180478508584, "gated_delta_h_c_norm": 0.008633215243753512, "global_step": 830, "grad_norm_max": 0.016543859615921974, "hidden_mse": 0.01008911631652154, "loss": 0.009668853561743164, "lr_blocks": 6.778969234612583e-06, "lr_fusion": 6.778969234612584e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703220975399018, "transport_residual_norm": 0.044181434018537406, "velocity_mse": 0.00865994195046369} +{"anchor_distance_mean": 121.81093750000001, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.5514583519398, "delta_h_c_norm": 0.14241708461195232, "delta_h_d_norm": 0.057871487084776164, "epoch": 7, "gate_p_mean": 0.024830176975228824, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8655749455094338, "gate_p_p10": 0.0009836435317993164, "gate_p_p50": 0.006796646118164063, "gate_p_p90": 0.07651842832565309, "gate_p_std": 0.04148339854436927, "gated_delta_h_c_norm": 0.010107993408018956, "global_step": 840, "grad_norm_max": 0.011773426085710526, "hidden_mse": 0.01091602149535902, "loss": 0.010904198774369434, "lr_blocks": 6.701465872208216e-06, "lr_fusion": 6.701465872208216e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.703038430213928, "transport_residual_norm": 0.04460356286726892, "velocity_mse": 0.009812596591655166} +{"anchor_distance_mean": 122.915625, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.6150466405786577, "delta_h_c_norm": 0.14085388779640198, "delta_h_d_norm": 0.05797204980626702, "epoch": 7, "gate_p_mean": 0.021854246266593692, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.886638904362917, "gate_p_p10": 0.0005798816680908204, "gate_p_p50": 0.004647970199584961, "gate_p_p90": 0.06792826652526855, "gate_p_std": 0.041482301775249655, "gated_delta_h_c_norm": 0.010612553944156389, "global_step": 850, "grad_norm_max": 0.014665540307760239, "hidden_mse": 0.011255597131093965, "loss": 0.010768115919199774, "lr_blocks": 6.6234973460234184e-06, "lr_fusion": 6.623497346023418e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702477383613587, "transport_residual_norm": 0.045079361740499736, "velocity_mse": 0.009642556213657373} +{"anchor_distance_mean": 130.128125, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.442881924456742, "delta_h_c_norm": 0.13380577005445957, "delta_h_d_norm": 0.05568466833792627, "epoch": 7, "gate_p_mean": 0.024583762120164467, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8680226124823094, "gate_p_p10": 0.0008062362205237151, "gate_p_p50": 0.005471229553222656, "gate_p_p90": 0.07851829528808595, "gate_p_std": 0.04411374964838615, "gated_delta_h_c_norm": 0.010443723716161913, "global_step": 860, "grad_norm_max": 0.01910233125090599, "hidden_mse": 0.01173247168189846, "loss": 0.01171289112244267, "lr_blocks": 6.545084971874738e-06, "lr_fusion": 6.545084971874738e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70244083404541, "transport_residual_norm": 0.043396948883309965, "velocity_mse": 0.010539643955416978} +{"anchor_distance_mean": 120.44609375, "avg_context_frames": 14.8, "avg_micro_time_s": 2.6161989359126894, "delta_h_c_norm": 0.13483081748709083, "delta_h_d_norm": 0.055535512557253246, "epoch": 7, "gate_p_mean": 0.025441233642050066, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8527952469885349, "gate_p_p10": 0.001097044930793345, "gate_p_p50": 0.008185100555419923, "gate_p_p90": 0.0766077756881714, "gate_p_std": 0.040504771433188584, "gated_delta_h_c_norm": 0.008977261176914908, "global_step": 870, "grad_norm_max": 0.018643271178007126, "hidden_mse": 0.010540183208649979, "loss": 0.010145299212308601, "lr_blocks": 6.466250186922325e-06, "lr_fusion": 6.466250186922325e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702550578117371, "transport_residual_norm": 0.04519855887629092, "velocity_mse": 0.009091280910070054} +{"anchor_distance_mean": 125.97187500000001, "avg_context_frames": 14.0, "avg_micro_time_s": 2.5209289587073727, "delta_h_c_norm": 0.14082164065912367, "delta_h_d_norm": 0.055869794869795444, "epoch": 7, "gate_p_mean": 0.02600673286651727, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.860521361976862, "gate_p_p10": 0.0009404182434082031, "gate_p_p50": 0.006698894500732422, "gate_p_p90": 0.08132588863372803, "gate_p_std": 0.04463661765912548, "gated_delta_h_c_norm": 0.010872547100734665, "global_step": 880, "grad_norm_max": 0.02274242974817753, "hidden_mse": 0.011673388816416264, "loss": 0.011462437090813182, "lr_blocks": 6.387014543809224e-06, "lr_fusion": 6.387014543809223e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702339434623719, "transport_residual_norm": 0.04624709254130721, "velocity_mse": 0.010295098228380085} +{"anchor_distance_mean": 125.71015625000001, "avg_context_frames": 14.0, "avg_micro_time_s": 2.5072797623419323, "delta_h_c_norm": 0.13934360882267358, "delta_h_d_norm": 0.05682678907178343, "epoch": 7, "gate_p_mean": 0.023143651001737455, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8785704903304578, "gate_p_p10": 0.0008985042572021485, "gate_p_p50": 0.005558490753173828, "gate_p_p90": 0.06997482776641846, "gate_p_std": 0.042594177211867645, "gated_delta_h_c_norm": 0.010199358443060191, "global_step": 890, "grad_norm_max": 0.017507487908005714, "hidden_mse": 0.011300088575808331, "loss": 0.010872181510785596, "lr_blocks": 6.3073997047691e-06, "lr_fusion": 6.307399704769099e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70206286907196, "transport_residual_norm": 0.04512330978177488, "velocity_mse": 0.009742172629921698} +{"anchor_distance_mean": 125.67734375, "avg_context_frames": 13.0, "avg_micro_time_s": 2.41455020672729, "delta_h_c_norm": 0.13930992474779488, "delta_h_d_norm": 0.056907635368406775, "epoch": 7, "gate_p_mean": 0.02049857875099406, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8877584382891656, "gate_p_p10": 0.0006565928459167481, "gate_p_p50": 0.005081939697265625, "gate_p_p90": 0.06212221682071686, "gate_p_std": 0.0368069447344169, "gated_delta_h_c_norm": 0.008609529458044564, "global_step": 900, "grad_norm_max": 0.012805980630218983, "hidden_mse": 0.011077630985528232, "loss": 0.010957702167797834, "lr_blocks": 6.227427435703997e-06, "lr_fusion": 6.227427435703997e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702196848392488, "transport_residual_norm": 0.04654788235202432, "velocity_mse": 0.009849939029663801} +{"anchor_distance_mean": 126.49062500000001, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.579020023919293, "delta_h_c_norm": 0.14377288613468409, "delta_h_d_norm": 0.05990515346638858, "epoch": 7, "gate_p_mean": 0.028146718855714428, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8441584274172783, "gate_p_p10": 0.001460862159729004, "gate_p_p50": 0.009154891967773438, "gate_p_p90": 0.08076553344726563, "gate_p_std": 0.044854932336602364, "gated_delta_h_c_norm": 0.011594373441766948, "global_step": 910, "grad_norm_max": 0.019434046000242233, "hidden_mse": 0.011072538187727334, "loss": 0.01046431913273409, "lr_blocks": 6.147119600233758e-06, "lr_fusion": 6.147119600233758e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702604031562805, "transport_residual_norm": 0.0462577257771045, "velocity_mse": 0.009357065334916115} +{"anchor_distance_mean": 123.7625, "avg_context_frames": 15.8, "avg_micro_time_s": 2.7151444433635334, "delta_h_c_norm": 0.14151196694001555, "delta_h_d_norm": 0.05861211698502303, "epoch": 8, "gate_p_mean": 0.02645665557065513, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8520062372088433, "gate_p_p10": 0.0010415077209472658, "gate_p_p50": 0.007007598876953125, "gate_p_p90": 0.0821527659893036, "gate_p_std": 0.04320755244116299, "gated_delta_h_c_norm": 0.010649379188544118, "global_step": 920, "grad_norm_max": 0.018089545890688896, "hidden_mse": 0.0114126754575409, "loss": 0.011004329650313595, "lr_blocks": 6.066498153718735e-06, "lr_fusion": 6.066498153718735e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.702517104148866, "transport_residual_norm": 0.04507526620291174, "velocity_mse": 0.009863062159274705} +{"anchor_distance_mean": 122.23281250000001, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.567783433555451, "delta_h_c_norm": 0.1377837909385562, "delta_h_d_norm": 0.05532682915218175, "epoch": 8, "gate_p_mean": 0.0204882328864187, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8846512161195279, "gate_p_p10": 0.0006855368614196778, "gate_p_p50": 0.005455636978149415, "gate_p_p90": 0.06279678344726562, "gate_p_std": 0.03581919845892117, "gated_delta_h_c_norm": 0.008181244276056533, "global_step": 930, "grad_norm_max": 0.01810227520763874, "hidden_mse": 0.010243507532868535, "loss": 0.00997915804910008, "lr_blocks": 5.985585137257401e-06, "lr_fusion": 5.985585137257401e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701849448680878, "transport_residual_norm": 0.04405210944823921, "velocity_mse": 0.008954807309783064} +{"anchor_distance_mean": 125.70859375, "avg_context_frames": 12.600000000000001, "avg_micro_time_s": 2.339871819682594, "delta_h_c_norm": 0.14101366605609655, "delta_h_d_norm": 0.0590547310654074, "epoch": 8, "gate_p_mean": 0.02805236887652427, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8465955816209316, "gate_p_p10": 0.0011700749397277833, "gate_p_p50": 0.008159446716308595, "gate_p_p90": 0.08711667060852052, "gate_p_std": 0.04597044095280581, "gated_delta_h_c_norm": 0.010950952213897837, "global_step": 940, "grad_norm_max": 0.015718238428235054, "hidden_mse": 0.010958899816614576, "loss": 0.01026369848405011, "lr_blocks": 5.904402671660551e-06, "lr_fusion": 5.90440267166055e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701484060287475, "transport_residual_norm": 0.04368233704008162, "velocity_mse": 0.009167808544589207} +{"anchor_distance_mean": 129.63984375, "avg_context_frames": 14.0, "avg_micro_time_s": 2.488183892927191, "delta_h_c_norm": 0.13869488332420588, "delta_h_d_norm": 0.056394630484282976, "epoch": 8, "gate_p_mean": 0.01956274795229547, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8925348334014416, "gate_p_p10": 0.0005702352384105325, "gate_p_p50": 0.003941726684570312, "gate_p_p90": 0.06284439563751221, "gate_p_std": 0.037414134241407736, "gated_delta_h_c_norm": 0.008219949204067234, "global_step": 950, "grad_norm_max": 0.014135763980448246, "hidden_mse": 0.010373797899228522, "loss": 0.010294985031941907, "lr_blocks": 5.82297295140367e-06, "lr_fusion": 5.8229729514036705e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70158793926239, "transport_residual_norm": 0.04594136280938983, "velocity_mse": 0.009257605293532834} +{"anchor_distance_mean": 125.77656250000001, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.581919111730531, "delta_h_c_norm": 0.14343107752501966, "delta_h_d_norm": 0.060000224923715, "epoch": 8, "gate_p_mean": 0.027429060204303825, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8484881058335305, "gate_p_p10": 0.0008247565943747759, "gate_p_p50": 0.007233524322509766, "gate_p_p90": 0.08611106872558594, "gate_p_std": 0.04621565983979963, "gated_delta_h_c_norm": 0.011409406822349411, "global_step": 960, "grad_norm_max": 0.015277489088475704, "hidden_mse": 0.012058824463747442, "loss": 0.01198961203917861, "lr_blocks": 5.74131823855921e-06, "lr_fusion": 5.74131823855921e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701134884357453, "transport_residual_norm": 0.047527520358562475, "velocity_mse": 0.010783729577087798} +{"anchor_distance_mean": 124.57421875, "avg_context_frames": 15.4, "avg_micro_time_s": 2.6761638981886793, "delta_h_c_norm": 0.14074708316475154, "delta_h_d_norm": 0.0588449374306947, "epoch": 8, "gate_p_mean": 0.023072469321778047, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8718432255089283, "gate_p_p10": 0.0007384252035990358, "gate_p_p50": 0.0059375762939453125, "gate_p_p90": 0.06976454257965088, "gate_p_std": 0.040183951816288754, "gated_delta_h_c_norm": 0.009395908471924487, "global_step": 970, "grad_norm_max": 0.015359570272266865, "hidden_mse": 0.010630739753833041, "loss": 0.01035889839695301, "lr_blocks": 5.659460856710346e-06, "lr_fusion": 5.6594608567103456e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700672841072082, "transport_residual_norm": 0.05159490900114179, "velocity_mse": 0.00929582440876402} +{"anchor_distance_mean": 122.78828125000001, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4832127829256936, "delta_h_c_norm": 0.1422758823260665, "delta_h_d_norm": 0.05937856957316399, "epoch": 8, "gate_p_mean": 0.02818510058859829, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8441138587892056, "gate_p_p10": 0.0008773446083068849, "gate_p_p50": 0.0071197509765625, "gate_p_p90": 0.09002012014389038, "gate_p_std": 0.047889568412210794, "gated_delta_h_c_norm": 0.011297858020407149, "global_step": 980, "grad_norm_max": 0.016295311972498894, "hidden_mse": 0.011485196423018353, "loss": 0.011010718840407208, "lr_blocks": 5.577423184847932e-06, "lr_fusion": 5.577423184847932e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700743412971496, "transport_residual_norm": 0.04940062537789345, "velocity_mse": 0.009862199172494002} +{"anchor_distance_mean": 123.20937500000001, "avg_context_frames": 15.8, "avg_micro_time_s": 2.7135161175290707, "delta_h_c_norm": 0.14293673979118468, "delta_h_d_norm": 0.060765408119186765, "epoch": 8, "gate_p_mean": 0.02854387240950018, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8528368137776852, "gate_p_p10": 0.0010430216789245605, "gate_p_p50": 0.006920528411865235, "gate_p_p90": 0.09053354263305664, "gate_p_std": 0.049873627105262136, "gated_delta_h_c_norm": 0.012409540046064649, "global_step": 990, "grad_norm_max": 0.015892498195171356, "hidden_mse": 0.01145184708875604, "loss": 0.010712801801855677, "lr_blocks": 5.495227651252315e-06, "lr_fusion": 5.495227651252315e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701129221916199, "transport_residual_norm": 0.04764553881250322, "velocity_mse": 0.00956761706620455} +{"anchor_distance_mean": 121.51796875000001, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.6101204663980755, "delta_h_c_norm": 0.13729049609974028, "delta_h_d_norm": 0.05784498737193644, "epoch": 8, "gate_p_mean": 0.024619219431770034, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8659638166427612, "gate_p_p10": 0.001047670841217041, "gate_p_p50": 0.007479381561279298, "gate_p_p90": 0.07403912544250489, "gate_p_std": 0.04076145342551172, "gated_delta_h_c_norm": 0.009713429624389392, "global_step": 1000, "grad_norm_max": 0.017514578998088837, "hidden_mse": 0.011250478343572469, "loss": 0.010719137327396312, "lr_blocks": 5.412896727361663e-06, "lr_fusion": 5.4128967273616625e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700800740718842, "transport_residual_norm": 0.04819754557684064, "velocity_mse": 0.009594089482561685} +{"anchor_distance_mean": 122.85234375, "avg_context_frames": 12.600000000000001, "avg_micro_time_s": 2.3823870391439415, "delta_h_c_norm": 0.139105736091733, "delta_h_d_norm": 0.057907414808869365, "epoch": 8, "gate_p_mean": 0.028327132176491434, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8379968218505383, "gate_p_p10": 0.0011072635650634765, "gate_p_p50": 0.008791732788085939, "gate_p_p90": 0.08646173477172853, "gate_p_std": 0.045571636117529124, "gated_delta_h_c_norm": 0.010480597258720081, "global_step": 1010, "grad_norm_max": 0.01801569014787674, "hidden_mse": 0.011563936184393242, "loss": 0.011373769532656298, "lr_blocks": 5.3304529216284974e-06, "lr_fusion": 5.330452921628497e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700276839733124, "transport_residual_norm": 0.04710358115844429, "velocity_mse": 0.010217375884531067} +{"anchor_distance_mean": 124.05390625000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4728054219827755, "delta_h_c_norm": 0.14140552561730146, "delta_h_d_norm": 0.059613375365734106, "epoch": 8, "gate_p_mean": 0.029115126130636784, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8442550845444203, "gate_p_p10": 0.0010157108306884765, "gate_p_p50": 0.0075855255126953125, "gate_p_p90": 0.08981654644012452, "gate_p_std": 0.05048780051292852, "gated_delta_h_c_norm": 0.012475755988270977, "global_step": 1020, "grad_norm_max": 0.013774142600595951, "hidden_mse": 0.011097752855857835, "loss": 0.010805494070518763, "lr_blocks": 5.247918773366112e-06, "lr_fusion": 5.247918773366112e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70028042793274, "transport_residual_norm": 0.04509169710800052, "velocity_mse": 0.00969571877794806} +{"anchor_distance_mean": 127.3671875, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4916095851309366, "delta_h_c_norm": 0.14197700498625637, "delta_h_d_norm": 0.06038243672810495, "epoch": 8, "gate_p_mean": 0.026108083833241837, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.859450650960207, "gate_p_p10": 0.0010563135147094727, "gate_p_p50": 0.007160758972167969, "gate_p_p90": 0.08084015846252442, "gate_p_std": 0.04519442323362455, "gated_delta_h_c_norm": 0.010769091128895526, "global_step": 1030, "grad_norm_max": 0.017298651859164238, "hidden_mse": 0.011209641036111863, "loss": 0.010787643678486348, "lr_blocks": 5.165316846586541e-06, "lr_fusion": 5.165316846586541e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701074397563934, "transport_residual_norm": 0.0473057406488806, "velocity_mse": 0.009666679569636473} +{"anchor_distance_mean": 128.01796875, "avg_context_frames": 15.4, "avg_micro_time_s": 2.687570607432281, "delta_h_c_norm": 0.1441552412696183, "delta_h_d_norm": 0.059506481327116494, "epoch": 8, "gate_p_mean": 0.02856694953807164, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8457084111869335, "gate_p_p10": 0.0010456562042236329, "gate_p_p50": 0.00793294906616211, "gate_p_p90": 0.08933305740356445, "gate_p_std": 0.046156948251882574, "gated_delta_h_c_norm": 0.011634484357637122, "global_step": 1040, "grad_norm_max": 0.015069256536662579, "hidden_mse": 0.010664668687968516, "loss": 0.010275989814545028, "lr_blocks": 5.082669723831793e-06, "lr_fusion": 5.0826697238317935e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.701090013980865, "transport_residual_norm": 0.05214160173200071, "velocity_mse": 0.009209522983292117} +{"anchor_distance_mean": 127.2359375, "avg_context_frames": 14.8, "avg_micro_time_s": 2.616650238641887, "delta_h_c_norm": 0.1411490734666586, "delta_h_d_norm": 0.058434705482795835, "epoch": 9, "gate_p_mean": 0.020757657689682674, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8858781836926938, "gate_p_p10": 0.0005568981170654297, "gate_p_p50": 0.004907751083374024, "gate_p_p90": 0.06687850952148437, "gate_p_std": 0.037893601602991114, "gated_delta_h_c_norm": 0.009282157744019059, "global_step": 1050, "grad_norm_max": 0.01631907932460308, "hidden_mse": 0.010561533411964774, "loss": 0.00999965150258504, "lr_blocks": 5e-06, "lr_fusion": 5e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700011909008026, "transport_residual_norm": 0.051320256758481266, "velocity_mse": 0.008943498163716868} +{"anchor_distance_mean": 121.65078125000001, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.455780752406281, "delta_h_c_norm": 0.14522933810949326, "delta_h_d_norm": 0.06069244556128979, "epoch": 9, "gate_p_mean": 0.026083475834457207, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8578929081559181, "gate_p_p10": 0.0009184096590615809, "gate_p_p50": 0.007104110717773438, "gate_p_p90": 0.07875900268554688, "gate_p_std": 0.04503987208590843, "gated_delta_h_c_norm": 0.0107168583170278, "global_step": 1060, "grad_norm_max": 0.019005918875336647, "hidden_mse": 0.010921708645764739, "loss": 0.010423576214816422, "lr_blocks": 4.917330276168208e-06, "lr_fusion": 4.917330276168208e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69981782436371, "transport_residual_norm": 0.04865289223380387, "velocity_mse": 0.009331405302509666} +{"anchor_distance_mean": 122.65546875000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6667582734269675, "delta_h_c_norm": 0.14421578869223595, "delta_h_d_norm": 0.06057363362051547, "epoch": 9, "gate_p_mean": 0.02826415984891355, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8512264903634787, "gate_p_p10": 0.0009688138496130706, "gate_p_p50": 0.007618427276611328, "gate_p_p90": 0.08687787055969239, "gate_p_std": 0.049401048128493136, "gated_delta_h_c_norm": 0.01233311591859092, "global_step": 1070, "grad_norm_max": 0.013592449948191643, "hidden_mse": 0.011309747115592473, "loss": 0.010904293751809747, "lr_blocks": 4.8346831534134595e-06, "lr_fusion": 4.834683153413459e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.70042074918747, "transport_residual_norm": 0.04599577020853758, "velocity_mse": 0.009773319002124482} +{"anchor_distance_mean": 123.89296875000001, "avg_context_frames": 14.0, "avg_micro_time_s": 2.5311090882925784, "delta_h_c_norm": 0.1454218288883567, "delta_h_d_norm": 0.061272704321891074, "epoch": 9, "gate_p_mean": 0.031172640348086134, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8279359742999077, "gate_p_p10": 0.0011411666870117189, "gate_p_p50": 0.009245777130126953, "gate_p_p90": 0.09568952322006226, "gate_p_std": 0.050258938508341094, "gated_delta_h_c_norm": 0.012532745461794548, "global_step": 1080, "grad_norm_max": 0.013188556768000126, "hidden_mse": 0.011542299261782319, "loss": 0.011008931050309912, "lr_blocks": 4.752081226633888e-06, "lr_fusion": 4.7520812266338885e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.699573206901551, "transport_residual_norm": 0.047773818997666244, "velocity_mse": 0.009854701100266539} +{"anchor_distance_mean": 129.89921875000002, "avg_context_frames": 17.0, "avg_micro_time_s": 2.8678302616768634, "delta_h_c_norm": 0.1425568327307701, "delta_h_d_norm": 0.05748238237574697, "epoch": 9, "gate_p_mean": 0.02359184077504324, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8745395414531232, "gate_p_p10": 0.0009042024612426758, "gate_p_p50": 0.005749797821044922, "gate_p_p90": 0.07152531147003174, "gate_p_std": 0.04301955746341264, "gated_delta_h_c_norm": 0.010574302880559118, "global_step": 1090, "grad_norm_max": 0.018376562744379044, "hidden_mse": 0.010016473921132274, "loss": 0.009708093557856047, "lr_blocks": 4.669547078371503e-06, "lr_fusion": 4.669547078371504e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.700151765346527, "transport_residual_norm": 0.047637251578271395, "velocity_mse": 0.008706446146243252} +{"anchor_distance_mean": 127.75703125000001, "avg_context_frames": 16.400000000000002, "avg_micro_time_s": 2.784936450450914, "delta_h_c_norm": 0.1434620313346386, "delta_h_d_norm": 0.05850400007329881, "epoch": 9, "gate_p_mean": 0.02533233491703868, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8679452396929265, "gate_p_p10": 0.0009916305541992187, "gate_p_p50": 0.006953239440917969, "gate_p_p90": 0.0754786252975464, "gate_p_std": 0.04234219219652005, "gated_delta_h_c_norm": 0.010283124410489109, "global_step": 1100, "grad_norm_max": 0.013739962130784988, "hidden_mse": 0.010344920633360745, "loss": 0.01048077306477353, "lr_blocks": 4.587103272638339e-06, "lr_fusion": 4.5871032726383386e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.699808275699615, "transport_residual_norm": 0.05011852746829391, "velocity_mse": 0.009446281066630036} +{"anchor_distance_mean": 122.13593750000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.648476370854769, "delta_h_c_norm": 0.1416145962662995, "delta_h_d_norm": 0.05959206586703658, "epoch": 9, "gate_p_mean": 0.022588812990579757, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8760211631655693, "gate_p_p10": 0.000788724422454834, "gate_p_p50": 0.0064526557922363285, "gate_p_p90": 0.06730520725250244, "gate_p_std": 0.039858675486175346, "gated_delta_h_c_norm": 0.009246761528629577, "global_step": 1110, "grad_norm_max": 0.018206002190709114, "hidden_mse": 0.010454221750842409, "loss": 0.010104150325059891, "lr_blocks": 4.504772348747687e-06, "lr_fusion": 4.504772348747687e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.699904763698578, "transport_residual_norm": 0.05022795391269028, "velocity_mse": 0.009058728092350065} +{"anchor_distance_mean": 120.80625, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4811112852054067, "delta_h_c_norm": 0.14531509941443801, "delta_h_d_norm": 0.05984457265585661, "epoch": 9, "gate_p_mean": 0.031268925609765574, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8282770216464996, "gate_p_p10": 0.0010846327990293504, "gate_p_p50": 0.008460712432861329, "gate_p_p90": 0.09874191284179688, "gate_p_std": 0.05177538186544553, "gated_delta_h_c_norm": 0.01251621198025532, "global_step": 1120, "grad_norm_max": 0.014414284378290176, "hidden_mse": 0.0115008813503664, "loss": 0.010700579118565657, "lr_blocks": 4.42257681515207e-06, "lr_fusion": 4.4225768151520694e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.699020421504974, "transport_residual_norm": 0.04782557925209403, "velocity_mse": 0.009550491010304541} +{"anchor_distance_mean": 121.9765625, "avg_context_frames": 14.0, "avg_micro_time_s": 2.496743742225226, "delta_h_c_norm": 0.1446023155003786, "delta_h_d_norm": 0.06277434569783509, "epoch": 9, "gate_p_mean": 0.034299135048058815, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8146875225007535, "gate_p_p10": 0.0013538360595703126, "gate_p_p50": 0.01004934310913086, "gate_p_p90": 0.10598602294921876, "gate_p_std": 0.055525660782586785, "gated_delta_h_c_norm": 0.013496210567245726, "global_step": 1130, "grad_norm_max": 0.01892944611608982, "hidden_mse": 0.011780242907116191, "loss": 0.010936707103974187, "lr_blocks": 4.340539143289655e-06, "lr_fusion": 4.3405391432896555e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698611438274384, "transport_residual_norm": 0.04915836770087481, "velocity_mse": 0.009758682802203112} +{"anchor_distance_mean": 124.63984375000001, "avg_context_frames": 13.0, "avg_micro_time_s": 2.3915714628092246, "delta_h_c_norm": 0.14343811729922892, "delta_h_d_norm": 0.061522575281560424, "epoch": 9, "gate_p_mean": 0.03487627725116909, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8051895782351495, "gate_p_p10": 0.0011788415256887676, "gate_p_p50": 0.009756088256835938, "gate_p_p90": 0.10935592651367188, "gate_p_std": 0.05576424178434536, "gated_delta_h_c_norm": 0.013454842111968901, "global_step": 1140, "grad_norm_max": 0.011746343225240707, "hidden_mse": 0.011277893779333682, "loss": 0.011024634510977195, "lr_blocks": 4.25868176144079e-06, "lr_fusion": 4.2586817614407895e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.699302804470063, "transport_residual_norm": 0.05106086963787675, "velocity_mse": 0.00989684517553542} +{"anchor_distance_mean": 128.24296875000002, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.429539430003206, "delta_h_c_norm": 0.1469388084486127, "delta_h_d_norm": 0.060520680435001854, "epoch": 9, "gate_p_mean": 0.027852202818030494, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8465843133628369, "gate_p_p10": 0.0008840322494506836, "gate_p_p50": 0.006654548645019531, "gate_p_p90": 0.08834495544433595, "gate_p_std": 0.04847579528577626, "gated_delta_h_c_norm": 0.011456700430426282, "global_step": 1150, "grad_norm_max": 0.01303714606910944, "hidden_mse": 0.010369369038380684, "loss": 0.010271764235221782, "lr_blocks": 4.17702704859633e-06, "lr_fusion": 4.17702704859633e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698854768276215, "transport_residual_norm": 0.05051109427586198, "velocity_mse": 0.009234827372711153} +{"anchor_distance_mean": 124.89687500000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4524847850509106, "delta_h_c_norm": 0.1473617585375905, "delta_h_d_norm": 0.059627007367089396, "epoch": 9, "gate_p_mean": 0.023565620701992886, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8729695275425912, "gate_p_p10": 0.0007212996482849122, "gate_p_p50": 0.006207466125488281, "gate_p_p90": 0.07036006450653076, "gate_p_std": 0.04258602481568233, "gated_delta_h_c_norm": 0.009747256548143924, "global_step": 1160, "grad_norm_max": 0.009606238454580307, "hidden_mse": 0.01076387749053538, "loss": 0.010841847487608903, "lr_blocks": 4.0955973283394525e-06, "lr_fusion": 4.095597328339452e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698449122905732, "transport_residual_norm": 0.05185060971416533, "velocity_mse": 0.00976545975136105} +{"anchor_distance_mean": 127.72578125000001, "avg_context_frames": 13.0, "avg_micro_time_s": 2.422159293665027, "delta_h_c_norm": 0.1490206453949213, "delta_h_d_norm": 0.06312820762395858, "epoch": 9, "gate_p_mean": 0.025101371848722923, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8612214803695679, "gate_p_p10": 0.0007865547901019454, "gate_p_p50": 0.006750583648681641, "gate_p_p90": 0.07600964903831482, "gate_p_std": 0.044105006475001575, "gated_delta_h_c_norm": 0.01107452342985198, "global_step": 1170, "grad_norm_max": 0.008727669715881348, "hidden_mse": 0.010766622208757327, "loss": 0.010591177910100669, "lr_blocks": 4.0144148627426e-06, "lr_fusion": 4.0144148627425993e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698557257652283, "transport_residual_norm": 0.05283584077842534, "velocity_mse": 0.009514515631599352} +{"anchor_distance_mean": 127.56328125, "avg_context_frames": 16.8, "avg_micro_time_s": 2.8232711190052218, "delta_h_c_norm": 0.14733787737786772, "delta_h_d_norm": 0.06186861954629422, "epoch": 10, "gate_p_mean": 0.025439823351916857, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8638236455619336, "gate_p_p10": 0.0009691238403320313, "gate_p_p50": 0.007359504699707031, "gate_p_p90": 0.07513275146484376, "gate_p_std": 0.04383794616442174, "gated_delta_h_c_norm": 0.010827883212186862, "global_step": 1180, "grad_norm_max": 0.013276665471494198, "hidden_mse": 0.009964375296840446, "loss": 0.009825124754570425, "lr_blocks": 3.9335018462812664e-06, "lr_fusion": 3.933501846281267e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698844611644745, "transport_residual_norm": 0.054599323682487017, "velocity_mse": 0.008828687213826925} +{"anchor_distance_mean": 124.7046875, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.4481206572818337, "delta_h_c_norm": 0.14691514652222395, "delta_h_d_norm": 0.06366977724246681, "epoch": 10, "gate_p_mean": 0.032980342028895396, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8203748743981123, "gate_p_p10": 0.001151132583618164, "gate_p_p50": 0.009153175354003907, "gate_p_p90": 0.10489239692687989, "gate_p_std": 0.05316910336841829, "gated_delta_h_c_norm": 0.012806334550259636, "global_step": 1190, "grad_norm_max": 0.01193115022033453, "hidden_mse": 0.011769907455891371, "loss": 0.011238736985251309, "lr_blocks": 3.852880399766243e-06, "lr_fusion": 3.852880399766243e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698756957054139, "transport_residual_norm": 0.05206119427457452, "velocity_mse": 0.010061746250721627} +{"anchor_distance_mean": 123.33906250000001, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.5442452416376913, "delta_h_c_norm": 0.14589389711618425, "delta_h_d_norm": 0.06085667139850557, "epoch": 10, "gate_p_mean": 0.02941367079038173, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8458944566547871, "gate_p_p10": 0.0011943817138671876, "gate_p_p50": 0.008092117309570313, "gate_p_p90": 0.09088935852050782, "gate_p_std": 0.05023346686502919, "gated_delta_h_c_norm": 0.01205271235885448, "global_step": 1200, "grad_norm_max": 0.010602176189422607, "hidden_mse": 0.010358651241404004, "loss": 0.010110359321697616, "lr_blocks": 3.7725725642960047e-06, "lr_fusion": 3.772572564296005e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.698723077774048, "transport_residual_norm": 0.05355264134705067, "velocity_mse": 0.009074494219385087} +{"anchor_distance_mean": 126.26328125, "avg_context_frames": 11.8, "avg_micro_time_s": 2.293019932125753, "delta_h_c_norm": 0.1452974813990295, "delta_h_d_norm": 0.061021966766566044, "epoch": 10, "gate_p_mean": 0.031305778695968914, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8333533920347691, "gate_p_p10": 0.0009080648422241212, "gate_p_p50": 0.007978248596191407, "gate_p_p90": 0.09806332588195801, "gate_p_std": 0.05327767967246473, "gated_delta_h_c_norm": 0.012862822580791545, "global_step": 1210, "grad_norm_max": 0.013905427418649197, "hidden_mse": 0.011780155444284901, "loss": 0.011613227444468067, "lr_blocks": 3.6926002952309015e-06, "lr_fusion": 3.6926002952309016e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.697843945026397, "transport_residual_norm": 0.05134320030920208, "velocity_mse": 0.010435211914591493} +{"anchor_distance_mean": 121.1640625, "avg_context_frames": 15.8, "avg_micro_time_s": 2.7385553105938016, "delta_h_c_norm": 0.14260845826938748, "delta_h_d_norm": 0.06107559595257044, "epoch": 10, "gate_p_mean": 0.025392599400947805, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8595873683691025, "gate_p_p10": 0.0008770585060119629, "gate_p_p50": 0.0075856208801269535, "gate_p_p90": 0.0757406234741211, "gate_p_std": 0.04195708476472646, "gated_delta_h_c_norm": 0.009678740681920317, "global_step": 1220, "grad_norm_max": 0.014013596810400486, "hidden_mse": 0.009609022369841114, "loss": 0.008820208394899965, "lr_blocks": 3.6129854561907786e-06, "lr_fusion": 3.612985456190778e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.697135293483734, "transport_residual_norm": 0.06057617422193289, "velocity_mse": 0.007859306150930934} +{"anchor_distance_mean": 124.63906250000001, "avg_context_frames": 15.0, "avg_micro_time_s": 2.649493669231015, "delta_h_c_norm": 0.1431208174675703, "delta_h_d_norm": 0.06264443425461651, "epoch": 10, "gate_p_mean": 0.029498736187815667, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8372245825827123, "gate_p_p10": 0.00091007468290627, "gate_p_p50": 0.007631874084472657, "gate_p_p90": 0.09072554111480713, "gate_p_std": 0.04907558912236709, "gated_delta_h_c_norm": 0.011431857546267566, "global_step": 1230, "grad_norm_max": 0.014891402795910835, "hidden_mse": 0.01106209956924431, "loss": 0.010513548037852162, "lr_blocks": 3.533749813077677e-06, "lr_fusion": 3.533749813077677e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.697860419750214, "transport_residual_norm": 0.05647684871219099, "velocity_mse": 0.0094073380518239} +{"anchor_distance_mean": 121.87578125, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.571406935267442, "delta_h_c_norm": 0.1455438463948667, "delta_h_d_norm": 0.06158151593990624, "epoch": 10, "gate_p_mean": 0.029633655917132275, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8398465313017369, "gate_p_p10": 0.0010123968124389649, "gate_p_p50": 0.008086109161376953, "gate_p_p90": 0.09183999300003053, "gate_p_std": 0.050942283269250765, "gated_delta_h_c_norm": 0.011938002955866979, "global_step": 1240, "grad_norm_max": 0.010296285152435303, "hidden_mse": 0.010122096166014671, "loss": 0.009603919641813265, "lr_blocks": 3.4549150281252635e-06, "lr_fusion": 3.4549150281252636e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.697522652149202, "transport_residual_norm": 0.05638028760440648, "velocity_mse": 0.008591710013570264} +{"anchor_distance_mean": 123.95625000000001, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4809931258496363, "delta_h_c_norm": 0.1444373967126012, "delta_h_d_norm": 0.0617225097026676, "epoch": 10, "gate_p_mean": 0.027954107889672744, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8505614265799523, "gate_p_p10": 0.000975751830264926, "gate_p_p50": 0.007884025573730469, "gate_p_p90": 0.08561193346977235, "gate_p_std": 0.04586662820074708, "gated_delta_h_c_norm": 0.011055405469232938, "global_step": 1250, "grad_norm_max": 0.011169354431331158, "hidden_mse": 0.010930393222952263, "loss": 0.010153968387749047, "lr_blocks": 3.3765026539765832e-06, "lr_fusion": 3.3765026539765834e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69725034236908, "transport_residual_norm": 0.05496516567654908, "velocity_mse": 0.009060929052066059} +{"anchor_distance_mean": 127.78984375, "avg_context_frames": 13.4, "avg_micro_time_s": 2.45548514645634, "delta_h_c_norm": 0.1449946208857, "delta_h_d_norm": 0.06362114911898971, "epoch": 10, "gate_p_mean": 0.034882575424853714, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8194118350744248, "gate_p_p10": 0.0012555837631225586, "gate_p_p50": 0.01000051498413086, "gate_p_p90": 0.10760135650634767, "gate_p_std": 0.05764384333742783, "gated_delta_h_c_norm": 0.013873935965239071, "global_step": 1260, "grad_norm_max": 0.028508616611361504, "hidden_mse": 0.011199425265658647, "loss": 0.010942241456359626, "lr_blocks": 3.298534127791785e-06, "lr_fusion": 3.298534127791785e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69742865562439, "transport_residual_norm": 0.05032236566767097, "velocity_mse": 0.009822298964718357} +{"anchor_distance_mean": 123.69765625000001, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.6735912053336506, "delta_h_c_norm": 0.14342881608754396, "delta_h_d_norm": 0.0632528178859502, "epoch": 10, "gate_p_mean": 0.02904430728958687, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8418284498155117, "gate_p_p10": 0.0009226035792380572, "gate_p_p50": 0.007052373886108399, "gate_p_p90": 0.09113933444023133, "gate_p_std": 0.05069023971445859, "gated_delta_h_c_norm": 0.011706820924882778, "global_step": 1270, "grad_norm_max": 0.012787621468305588, "hidden_mse": 0.011054444866022096, "loss": 0.010805656461161562, "lr_blocks": 3.2210307653874175e-06, "lr_fusion": 3.221030765387417e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.697267270088195, "transport_residual_norm": 0.05740726855583489, "velocity_mse": 0.009700211975723506} +{"anchor_distance_mean": 124.31484375000001, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.524957071799145, "delta_h_c_norm": 0.1451248621568084, "delta_h_d_norm": 0.06384616643190384, "epoch": 10, "gate_p_mean": 0.03279526128026191, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.827932720631361, "gate_p_p10": 0.0009998941095545888, "gate_p_p50": 0.008397197723388672, "gate_p_p90": 0.10221633911132813, "gate_p_std": 0.05515884851338342, "gated_delta_h_c_norm": 0.012886028119828552, "global_step": 1280, "grad_norm_max": 0.01402959879487753, "hidden_mse": 0.011212144931778313, "loss": 0.011017974937567489, "lr_blocks": 3.1440137554088957e-06, "lr_fusion": 3.144013755408895e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696896278858185, "transport_residual_norm": 0.05348019972443581, "velocity_mse": 0.00989676041644998} +{"anchor_distance_mean": 122.71953125, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.5051538060331953, "delta_h_c_norm": 0.1430636014789343, "delta_h_d_norm": 0.060562001960352066, "epoch": 10, "gate_p_mean": 0.029445636243326592, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8479782909154893, "gate_p_p10": 0.0011672258377075195, "gate_p_p50": 0.00828084945678711, "gate_p_p90": 0.09240825176239015, "gate_p_std": 0.05100008072331548, "gated_delta_h_c_norm": 0.01147303802863462, "global_step": 1290, "grad_norm_max": 0.011088326573371887, "hidden_mse": 0.01069898487185128, "loss": 0.011107209173496814, "lr_blocks": 3.06750415353774e-06, "lr_fusion": 3.0675041535377405e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696709609031677, "transport_residual_norm": 0.05441704201512039, "velocity_mse": 0.010037310689222069} +{"anchor_distance_mean": 130.4171875, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.469162545618019, "delta_h_c_norm": 0.14186388356611132, "delta_h_d_norm": 0.05978843518532813, "epoch": 10, "gate_p_mean": 0.025619506248040127, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8659688286483288, "gate_p_p10": 0.001001286506652832, "gate_p_p50": 0.007311630249023438, "gate_p_p90": 0.07936248779296876, "gate_p_std": 0.043616982101229954, "gated_delta_h_c_norm": 0.010326765392528615, "global_step": 1300, "grad_norm_max": 0.011383263394236565, "hidden_mse": 0.010179219726705924, "loss": 0.009845191310159863, "lr_blocks": 2.991522876735154e-06, "lr_fusion": 2.991522876735154e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696065676212312, "transport_residual_norm": 0.052998769329860806, "velocity_mse": 0.008827269333414734} +{"anchor_distance_mean": 125.028125, "avg_context_frames": 15.0, "avg_micro_time_s": 2.6091846724550125, "delta_h_c_norm": 0.14320094119757415, "delta_h_d_norm": 0.059603389864787464, "epoch": 11, "gate_p_mean": 0.025696128921117636, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8567375533282757, "gate_p_p10": 0.0009949207305908203, "gate_p_p50": 0.007464885711669922, "gate_p_p90": 0.07806973457336426, "gate_p_std": 0.04331308398395777, "gated_delta_h_c_norm": 0.009349952643970028, "global_step": 1310, "grad_norm_max": 0.014943193644285202, "hidden_mse": 0.009468811162514613, "loss": 0.009478193655377255, "lr_blocks": 2.9160906975235493e-06, "lr_fusion": 2.916090697523549e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696650636196138, "transport_residual_norm": 0.05605415385216475, "velocity_mse": 0.008531312551349401} +{"anchor_distance_mean": 123.046875, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.5648514446016635, "delta_h_c_norm": 0.14527597241103649, "delta_h_d_norm": 0.06443535415455699, "epoch": 11, "gate_p_mean": 0.03348515655670781, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8225518565624953, "gate_p_p10": 0.001358938217163086, "gate_p_p50": 0.010395050048828125, "gate_p_p90": 0.10212445259094238, "gate_p_std": 0.05334594471205492, "gated_delta_h_c_norm": 0.01249324329692172, "global_step": 1320, "grad_norm_max": 0.010332166217267513, "hidden_mse": 0.010666257760021836, "loss": 0.0103682099172147, "lr_blocks": 2.8412282383075362e-06, "lr_fusion": 2.8412282383075363e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696724665164949, "transport_residual_norm": 0.05493600820191205, "velocity_mse": 0.009301584129570984} +{"anchor_distance_mean": 126.196875, "avg_context_frames": 12.200000000000001, "avg_micro_time_s": 2.3185341106785926, "delta_h_c_norm": 0.1439916651695967, "delta_h_d_norm": 0.06419077422469854, "epoch": 11, "gate_p_mean": 0.034745596541324635, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8167666040360928, "gate_p_p10": 0.001199328899383545, "gate_p_p50": 0.009125423431396485, "gate_p_p90": 0.11096560955047607, "gate_p_std": 0.05846179436775856, "gated_delta_h_c_norm": 0.013386590506706854, "global_step": 1330, "grad_norm_max": 0.011919054202735424, "hidden_mse": 0.011540495668305084, "loss": 0.011406175669981167, "lr_blocks": 2.766955965735968e-06, "lr_fusion": 2.766955965735968e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696160650253296, "transport_residual_norm": 0.05106422048993409, "velocity_mse": 0.010252126105478966} +{"anchor_distance_mean": 125.353125, "avg_context_frames": 14.200000000000001, "avg_micro_time_s": 2.545066675383714, "delta_h_c_norm": 0.14519022442400456, "delta_h_d_norm": 0.0625525692012161, "epoch": 11, "gate_p_mean": 0.03379856716492213, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8286318391561509, "gate_p_p10": 0.0011394023895263672, "gate_p_p50": 0.008615398406982422, "gate_p_p90": 0.10683789253234864, "gate_p_std": 0.057514448487199846, "gated_delta_h_c_norm": 0.01347761588258436, "global_step": 1340, "grad_norm_max": 0.00895714107900858, "hidden_mse": 0.010985690192319453, "loss": 0.010682093631476165, "lr_blocks": 2.693294185106562e-06, "lr_fusion": 2.693294185106562e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.696257841587068, "transport_residual_norm": 0.05565097443759442, "velocity_mse": 0.009583524614572526} +{"anchor_distance_mean": 124.44296875, "avg_context_frames": 16.2, "avg_micro_time_s": 2.7661110426037343, "delta_h_c_norm": 0.14391446774825453, "delta_h_d_norm": 0.061741086281836034, "epoch": 11, "gate_p_mean": 0.02488212555181235, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.865684124082327, "gate_p_p10": 0.000789844966493547, "gate_p_p50": 0.00708765983581543, "gate_p_p90": 0.07243773341178894, "gate_p_std": 0.04404180340352468, "gated_delta_h_c_norm": 0.010190461850288558, "global_step": 1350, "grad_norm_max": 0.008798321709036827, "hidden_mse": 0.010069099592510612, "loss": 0.0095059391023824, "lr_blocks": 2.6202630348146323e-06, "lr_fusion": 2.6202630348146324e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69545944929123, "transport_residual_norm": 0.0653115968219936, "velocity_mse": 0.008499029150698334} +{"anchor_distance_mean": 127.234375, "avg_context_frames": 14.8, "avg_micro_time_s": 2.6000541486151634, "delta_h_c_norm": 0.14386058328673243, "delta_h_d_norm": 0.06292488942854106, "epoch": 11, "gate_p_mean": 0.02703351185191423, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.86231998950243, "gate_p_p10": 0.0009106874465942383, "gate_p_p50": 0.006861972808837891, "gate_p_p90": 0.08161036968231201, "gate_p_std": 0.047626853329711595, "gated_delta_h_c_norm": 0.011421328311553226, "global_step": 1360, "grad_norm_max": 0.010866137221455574, "hidden_mse": 0.01010189727821853, "loss": 0.009872941381763667, "lr_blocks": 2.5478824808474613e-06, "lr_fusion": 2.547882480847461e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.695314884185791, "transport_residual_norm": 0.0621905142441392, "velocity_mse": 0.00886275164084509} +{"anchor_distance_mean": 122.10390625000001, "avg_context_frames": 15.4, "avg_micro_time_s": 2.6761451652258987, "delta_h_c_norm": 0.1459822637960315, "delta_h_d_norm": 0.06507538929581642, "epoch": 11, "gate_p_mean": 0.0330805323086679, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8412282392382622, "gate_p_p10": 0.0010080337524414062, "gate_p_p50": 0.007516384124755859, "gate_p_p90": 0.10565762519836426, "gate_p_std": 0.05809285204159096, "gated_delta_h_c_norm": 0.013926184151205235, "global_step": 1370, "grad_norm_max": 0.01616928167641163, "hidden_mse": 0.010961850872263312, "loss": 0.009994722332339735, "lr_blocks": 2.476172311325783e-06, "lr_fusion": 2.476172311325783e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.695717573165894, "transport_residual_norm": 0.06128365979529918, "velocity_mse": 0.008898537213099189} +{"anchor_distance_mean": 127.10703125, "avg_context_frames": 14.8, "avg_micro_time_s": 2.5976878527857483, "delta_h_c_norm": 0.14111941400915384, "delta_h_d_norm": 0.06255100229755044, "epoch": 11, "gate_p_mean": 0.027088157657999545, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8566842205822468, "gate_p_p10": 0.0011359453201293945, "gate_p_p50": 0.007598304748535157, "gate_p_p90": 0.08235702514648438, "gate_p_std": 0.04564181800815277, "gated_delta_h_c_norm": 0.010367563478939702, "global_step": 1380, "grad_norm_max": 0.008912174962460995, "hidden_mse": 0.010107378655811773, "loss": 0.01014233889291063, "lr_blocks": 2.4051521310939258e-06, "lr_fusion": 2.405152131093926e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.695559179782867, "transport_residual_norm": 0.06251524505205452, "velocity_mse": 0.009131600972614251} +{"anchor_distance_mean": 127.92109375000001, "avg_context_frames": 16.400000000000002, "avg_micro_time_s": 2.778375037129445, "delta_h_c_norm": 0.14492652136832476, "delta_h_d_norm": 0.06321079833433033, "epoch": 11, "gate_p_mean": 0.03182051681214944, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8304827965795994, "gate_p_p10": 0.0011561155319213868, "gate_p_p50": 0.008806800842285157, "gate_p_p90": 0.09707946777343751, "gate_p_std": 0.05332826197263785, "gated_delta_h_c_norm": 0.012850019397592406, "global_step": 1390, "grad_norm_max": 0.012509177438914776, "hidden_mse": 0.01056769690476358, "loss": 0.010324075291282498, "lr_blocks": 2.3348413563600324e-06, "lr_fusion": 2.3348413563600325e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.695342600345612, "transport_residual_norm": 0.06358261751011014, "velocity_mse": 0.009267305600224063} +{"anchor_distance_mean": 124.996875, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4319449037182497, "delta_h_c_norm": 0.14464688012376428, "delta_h_d_norm": 0.06512474231421948, "epoch": 11, "gate_p_mean": 0.03232743750850205, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.831343425065279, "gate_p_p10": 0.0010244416538625956, "gate_p_p50": 0.00816030502319336, "gate_p_p90": 0.10122604370117189, "gate_p_std": 0.05623119995580055, "gated_delta_h_c_norm": 0.013249474418989849, "global_step": 1400, "grad_norm_max": 0.008809905499219894, "hidden_mse": 0.011741711862850935, "loss": 0.011549836568883622, "lr_blocks": 2.265259209387867e-06, "lr_fusion": 2.2652592093878666e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.694714558124543, "transport_residual_norm": 0.056912311678752305, "velocity_mse": 0.010375665346509778} +{"anchor_distance_mean": 125.06093750000001, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4438314619794257, "delta_h_c_norm": 0.14535442851483824, "delta_h_d_norm": 0.06274851229973137, "epoch": 11, "gate_p_mean": 0.029771375612472187, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8407231867313385, "gate_p_p10": 0.000987100601196289, "gate_p_p50": 0.008704853057861329, "gate_p_p90": 0.09011478424072267, "gate_p_std": 0.050645754858851436, "gated_delta_h_c_norm": 0.011780537383310731, "global_step": 1410, "grad_norm_max": 0.01119321770966053, "hidden_mse": 0.010227996937464923, "loss": 0.00974440727441106, "lr_blocks": 2.1964247132416373e-06, "lr_fusion": 2.196424713241637e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.694360411167144, "transport_residual_norm": 0.06254954976029695, "velocity_mse": 0.008721607603365556} +{"anchor_distance_mean": 122.43046875, "avg_context_frames": 11.8, "avg_micro_time_s": 2.261301682268095, "delta_h_c_norm": 0.14900918789207937, "delta_h_d_norm": 0.06619204068556428, "epoch": 11, "gate_p_mean": 0.040055176097666845, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7881059937179089, "gate_p_p10": 0.0013541698455810547, "gate_p_p50": 0.011804389953613282, "gate_p_p90": 0.12581329345703127, "gate_p_std": 0.06367760041612201, "gated_delta_h_c_norm": 0.015440036670770497, "global_step": 1420, "grad_norm_max": 0.015816880390048027, "hidden_mse": 0.011610778601607308, "loss": 0.010744776090723463, "lr_blocks": 2.1283566865852824e-06, "lr_fusion": 2.128356686585282e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.694801783561706, "transport_residual_norm": 0.054672161163762216, "velocity_mse": 0.00958369822765235} +{"anchor_distance_mean": 122.10546875, "avg_context_frames": 13.8, "avg_micro_time_s": 2.482654240913689, "delta_h_c_norm": 0.1444759632460773, "delta_h_d_norm": 0.0639038231689483, "epoch": 11, "gate_p_mean": 0.03250563464825973, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8211603797972202, "gate_p_p10": 0.0012958049774169922, "gate_p_p50": 0.0101715087890625, "gate_p_p90": 0.09695205688476563, "gate_p_std": 0.053188829356804496, "gated_delta_h_c_norm": 0.011923219580785372, "global_step": 1430, "grad_norm_max": 0.008968343958258629, "hidden_mse": 0.01020921160816215, "loss": 0.010052936463034713, "lr_blocks": 2.061073738537635e-06, "lr_fusion": 2.061073738537635e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.695065057277679, "transport_residual_norm": 0.05941012757830322, "velocity_mse": 0.009032015294360462} +{"anchor_distance_mean": 123.9890625, "avg_context_frames": 15.4, "avg_micro_time_s": 2.7381764402845876, "delta_h_c_norm": 0.13987314673140647, "delta_h_d_norm": 0.06400921107269823, "epoch": 12, "gate_p_mean": 0.032586681045359, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8191782079637051, "gate_p_p10": 0.001391267776489258, "gate_p_p50": 0.00992584228515625, "gate_p_p90": 0.10033506155014038, "gate_p_std": 0.05217370904283598, "gated_delta_h_c_norm": 0.011467977691791021, "global_step": 1440, "grad_norm_max": 0.011451596394181252, "hidden_mse": 0.01038553234830033, "loss": 0.010122939263237641, "lr_blocks": 1.9945942635848745e-06, "lr_fusion": 1.9945942635848748e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.694369864463807, "transport_residual_norm": 0.06150251454673708, "velocity_mse": 0.00908438601763919} +{"anchor_distance_mean": 124.57343750000001, "avg_context_frames": 15.8, "avg_micro_time_s": 2.7589638563396877, "delta_h_c_norm": 0.14242526646703482, "delta_h_d_norm": 0.06354712946340442, "epoch": 12, "gate_p_mean": 0.036840360559290276, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8156766075640918, "gate_p_p10": 0.001379823684692383, "gate_p_p50": 0.010059356689453125, "gate_p_p90": 0.11641710996627808, "gate_p_std": 0.059766834165202454, "gated_delta_h_c_norm": 0.013958364549762337, "global_step": 1450, "grad_norm_max": 0.008997643366456032, "hidden_mse": 0.010222396228346043, "loss": 0.009439592526177876, "lr_blocks": 1.928936436551661e-06, "lr_fusion": 1.928936436551661e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69443805217743, "transport_residual_norm": 0.06014189720153809, "velocity_mse": 0.008417352946707979} +{"anchor_distance_mean": 121.48828125, "avg_context_frames": 12.0, "avg_micro_time_s": 2.360845357850485, "delta_h_c_norm": 0.1461672380566597, "delta_h_d_norm": 0.06820742087438703, "epoch": 12, "gate_p_mean": 0.03735123707156163, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8010717388242483, "gate_p_p10": 0.0012501001358032227, "gate_p_p50": 0.010788631439208985, "gate_p_p90": 0.11656100749969484, "gate_p_std": 0.06027926759561524, "gated_delta_h_c_norm": 0.014224503844161518, "global_step": 1460, "grad_norm_max": 0.010418052785098553, "hidden_mse": 0.011669803684344516, "loss": 0.011124833507346922, "lr_blocks": 1.864118207632315e-06, "lr_fusion": 1.8641182076323148e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693687856197357, "transport_residual_norm": 0.05643729190342128, "velocity_mse": 0.009957853151718155} +{"anchor_distance_mean": 122.46328125000001, "avg_context_frames": 15.4, "avg_micro_time_s": 2.7246366750754536, "delta_h_c_norm": 0.14229313991963863, "delta_h_d_norm": 0.06508913021534682, "epoch": 12, "gate_p_mean": 0.03282481617061422, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.82523214481771, "gate_p_p10": 0.0012591123580932618, "gate_p_p50": 0.010293006896972656, "gate_p_p90": 0.09772322177886963, "gate_p_std": 0.052734960248926654, "gated_delta_h_c_norm": 0.01159204259311082, "global_step": 1470, "grad_norm_max": 0.010075051337480545, "hidden_mse": 0.010158935861545615, "loss": 0.009692414608434775, "lr_blocks": 1.8001572974834169e-06, "lr_fusion": 1.800157297483417e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.694438481330872, "transport_residual_norm": 0.06003924724645913, "velocity_mse": 0.008676520988228732} +{"anchor_distance_mean": 122.43046875, "avg_context_frames": 12.4, "avg_micro_time_s": 2.4080637150618713, "delta_h_c_norm": 0.14653508476912977, "delta_h_d_norm": 0.06600901004858316, "epoch": 12, "gate_p_mean": 0.037675821891753, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.799397061020136, "gate_p_p10": 0.0013570308685302735, "gate_p_p50": 0.01049032211303711, "gate_p_p90": 0.11674704551696778, "gate_p_std": 0.06168326426413842, "gated_delta_h_c_norm": 0.014136297231016216, "global_step": 1480, "grad_norm_max": 0.01125259418040514, "hidden_mse": 0.0114752555382438, "loss": 0.011056927163735963, "lr_blocks": 1.7370711923791567e-06, "lr_fusion": 1.7370711923791567e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.694206702709199, "transport_residual_norm": 0.05939450049772859, "velocity_mse": 0.009909401583718136} +{"anchor_distance_mean": 124.63828125, "avg_context_frames": 11.4, "avg_micro_time_s": 2.2944430057686986, "delta_h_c_norm": 0.1441903503611684, "delta_h_d_norm": 0.06296480339951814, "epoch": 12, "gate_p_mean": 0.035596875345800076, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8167553335428238, "gate_p_p10": 0.001241612364538014, "gate_p_p50": 0.009121131896972657, "gate_p_p90": 0.11264648437500001, "gate_p_std": 0.06074884668923915, "gated_delta_h_c_norm": 0.014238467155519175, "global_step": 1490, "grad_norm_max": 0.010500326752662659, "hidden_mse": 0.011105439631501214, "loss": 0.010921176779083908, "lr_blocks": 1.6748771394307584e-06, "lr_fusion": 1.6748771394307585e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693312585353851, "transport_residual_norm": 0.055153130646795036, "velocity_mse": 0.009810632793232799} +{"anchor_distance_mean": 127.2703125, "avg_context_frames": 14.8, "avg_micro_time_s": 2.6150738059091965, "delta_h_c_norm": 0.14086839836090803, "delta_h_d_norm": 0.0634054196998477, "epoch": 12, "gate_p_mean": 0.03066912368522026, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8416506700217724, "gate_p_p10": 0.0009822845458984375, "gate_p_p50": 0.007649421691894531, "gate_p_p90": 0.09557080268859863, "gate_p_std": 0.053189439751440665, "gated_delta_h_c_norm": 0.012056168470007834, "global_step": 1500, "grad_norm_max": 0.012843739241361618, "hidden_mse": 0.010334987309761345, "loss": 0.009631934357457794, "lr_blocks": 1.6135921418712959e-06, "lr_fusion": 1.6135921418712956e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69378000497818, "transport_residual_norm": 0.06239772797562182, "velocity_mse": 0.008598435623571278} +{"anchor_distance_mean": 128.46953125000002, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.7023334297991823, "delta_h_c_norm": 0.14017630415037274, "delta_h_d_norm": 0.0625812079757452, "epoch": 12, "gate_p_mean": 0.028127973957452923, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8516419015824795, "gate_p_p10": 0.001118793454952538, "gate_p_p50": 0.008580684661865234, "gate_p_p90": 0.08386621475219727, "gate_p_std": 0.046858222241280605, "gated_delta_h_c_norm": 0.010079818801023066, "global_step": 1510, "grad_norm_max": 0.005049944389611483, "hidden_mse": 0.00933148642652668, "loss": 0.009398649979266338, "lr_blocks": 1.5532329544071712e-06, "lr_fusion": 1.553232954407171e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693485152721406, "transport_residual_norm": 0.06700174463912845, "velocity_mse": 0.008465501319733448} +{"anchor_distance_mean": 125.77578125000001, "avg_context_frames": 12.600000000000001, "avg_micro_time_s": 2.344391565420665, "delta_h_c_norm": 0.14381116861477494, "delta_h_d_norm": 0.06347116348333658, "epoch": 12, "gate_p_mean": 0.03755027006263845, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8048129729926586, "gate_p_p10": 0.0014303063973784447, "gate_p_p50": 0.010863971710205079, "gate_p_p90": 0.11468375921249391, "gate_p_std": 0.060929273301735525, "gated_delta_h_c_norm": 0.013978121038235258, "global_step": 1520, "grad_norm_max": 0.01023966632783413, "hidden_mse": 0.011025074712233619, "loss": 0.010841031841118821, "lr_blocks": 1.4938160786375571e-06, "lr_fusion": 1.4938160786375572e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69306869506836, "transport_residual_norm": 0.05502278655767441, "velocity_mse": 0.009738524415297434} +{"anchor_distance_mean": 127.821875, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6243611344951208, "delta_h_c_norm": 0.1447678778320551, "delta_h_d_norm": 0.0645699331536889, "epoch": 12, "gate_p_mean": 0.03597369624185376, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8140367329120637, "gate_p_p10": 0.0012042760848999024, "gate_p_p50": 0.009581184387207032, "gate_p_p90": 0.11391611099243165, "gate_p_std": 0.060998886876041075, "gated_delta_h_c_norm": 0.013806287336046809, "global_step": 1530, "grad_norm_max": 0.007911360822618008, "hidden_mse": 0.010535601247102023, "loss": 0.010438098484883085, "lr_blocks": 1.4353577585430152e-06, "lr_fusion": 1.435357758543015e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693739008903504, "transport_residual_norm": 0.05851141111925245, "velocity_mse": 0.009384538332233206} +{"anchor_distance_mean": 123.79218750000001, "avg_context_frames": 17.2, "avg_micro_time_s": 2.859565482432663, "delta_h_c_norm": 0.1415436377748847, "delta_h_d_norm": 0.06491621807217599, "epoch": 12, "gate_p_mean": 0.03062902966339607, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8484012186527252, "gate_p_p10": 0.0010864973068237306, "gate_p_p50": 0.007895565032958985, "gate_p_p90": 0.09312418699264527, "gate_p_std": 0.05307279750704766, "gated_delta_h_c_norm": 0.012314594372583088, "global_step": 1540, "grad_norm_max": 0.010397516191005707, "hidden_mse": 0.010417618503561245, "loss": 0.010044112682226115, "lr_blocks": 1.3778739760445552e-06, "lr_fusion": 1.3778739760445552e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693620574474336, "transport_residual_norm": 0.06602197838947177, "velocity_mse": 0.009002350785885937} +{"anchor_distance_mean": 123.8609375, "avg_context_frames": 14.8, "avg_micro_time_s": 2.59588410428114, "delta_h_c_norm": 0.1441071566194296, "delta_h_d_norm": 0.06625798381865025, "epoch": 12, "gate_p_mean": 0.03574146665632725, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8201920852065087, "gate_p_p10": 0.0012927055358886719, "gate_p_p50": 0.010346412658691406, "gate_p_p90": 0.11048669815063478, "gate_p_std": 0.05988231576629915, "gated_delta_h_c_norm": 0.01374094680941198, "global_step": 1550, "grad_norm_max": 0.007774202153086662, "hidden_mse": 0.010945526129216888, "loss": 0.010515357006806881, "lr_blocks": 1.321380446634342e-06, "lr_fusion": 1.3213804466343421e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.6934321641922, "transport_residual_norm": 0.05896453144960106, "velocity_mse": 0.009420804373803549} +{"anchor_distance_mean": 127.03984375, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4604899397774718, "delta_h_c_norm": 0.1413728398270905, "delta_h_d_norm": 0.062319235131144526, "epoch": 12, "gate_p_mean": 0.028824166106642225, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8485827587544919, "gate_p_p10": 0.0012352466583251953, "gate_p_p50": 0.008556652069091798, "gate_p_p90": 0.08488769531250001, "gate_p_std": 0.0496261040243553, "gated_delta_h_c_norm": 0.010828969261638122, "global_step": 1560, "grad_norm_max": 0.007709342986345291, "hidden_mse": 0.009485918609425426, "loss": 0.009180615976219997, "lr_blocks": 1.2658926150792321e-06, "lr_fusion": 1.2658926150792322e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693526935577394, "transport_residual_norm": 0.05873315497301519, "velocity_mse": 0.008232024096651002} +{"anchor_distance_mean": 123.11328125, "avg_context_frames": 13.8, "avg_micro_time_s": 2.5001676389656495, "delta_h_c_norm": 0.14282034579664468, "delta_h_d_norm": 0.06540395310148597, "epoch": 13, "gate_p_mean": 0.03792013247730211, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8017771169543266, "gate_p_p10": 0.0015448570251464845, "gate_p_p50": 0.011759471893310548, "gate_p_p90": 0.11787917017936707, "gate_p_std": 0.059713844431098555, "gated_delta_h_c_norm": 0.013763298020057846, "global_step": 1570, "grad_norm_max": 0.007401987910270691, "hidden_mse": 0.01086143939755857, "loss": 0.010324895021039993, "lr_blocks": 1.2114256511983274e-06, "lr_fusion": 1.2114256511983274e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692709124088287, "transport_residual_norm": 0.06273416625335813, "velocity_mse": 0.009238751116208732} +{"anchor_distance_mean": 123.01484375000001, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4748011695031895, "delta_h_c_norm": 0.1444636414758861, "delta_h_d_norm": 0.06644805576652289, "epoch": 13, "gate_p_mean": 0.04541965196840465, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7762505274266005, "gate_p_p10": 0.001710033416748047, "gate_p_p50": 0.012359237670898438, "gate_p_p90": 0.14478540420532227, "gate_p_std": 0.07294244188815356, "gated_delta_h_c_norm": 0.01666179527528584, "global_step": 1580, "grad_norm_max": 0.007409753277897835, "hidden_mse": 0.011841671454021707, "loss": 0.011075413343496621, "lr_blocks": 1.157994445715706e-06, "lr_fusion": 1.157994445715706e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693587911128999, "transport_residual_norm": 0.05785968876443803, "velocity_mse": 0.009891246166080237} +{"anchor_distance_mean": 123.37265625, "avg_context_frames": 12.8, "avg_micro_time_s": 2.3834462915867336, "delta_h_c_norm": 0.14234045296907424, "delta_h_d_norm": 0.06556431660428644, "epoch": 13, "gate_p_mean": 0.03720473845605739, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8093226909637452, "gate_p_p10": 0.001226186752319336, "gate_p_p50": 0.010919189453125, "gate_p_p90": 0.11524276733398438, "gate_p_std": 0.060143312241416426, "gated_delta_h_c_norm": 0.01366716412594542, "global_step": 1590, "grad_norm_max": 0.008119405247271061, "hidden_mse": 0.011434891758835875, "loss": 0.010683572513516993, "lr_blocks": 1.1056136061894386e-06, "lr_fusion": 1.1056136061894384e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692628347873688, "transport_residual_norm": 0.057925927592441445, "velocity_mse": 0.009540083288447932} +{"anchor_distance_mean": 128.209375, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.689641509485955, "delta_h_c_norm": 0.14095409596338868, "delta_h_d_norm": 0.06181156844832003, "epoch": 13, "gate_p_mean": 0.02948827258660458, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8455649282783271, "gate_p_p10": 0.0013683509547263385, "gate_p_p50": 0.009028244018554689, "gate_p_p90": 0.08588409423828125, "gate_p_std": 0.04948119106702507, "gated_delta_h_c_norm": 0.010633992702059914, "global_step": 1600, "grad_norm_max": 0.007462357636541128, "hidden_mse": 0.009053043031599373, "loss": 0.009081950821564534, "lr_blocks": 1.0542974530180327e-06, "lr_fusion": 1.0542974530180327e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693000864982606, "transport_residual_norm": 0.06274244775995612, "velocity_mse": 0.008176646576612257} +{"anchor_distance_mean": 122.75390625, "avg_context_frames": 15.4, "avg_micro_time_s": 2.667710428363353, "delta_h_c_norm": 0.14456945518031716, "delta_h_d_norm": 0.06552408514544368, "epoch": 13, "gate_p_mean": 0.03380357507267036, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8300168067216873, "gate_p_p10": 0.0011116027832031251, "gate_p_p50": 0.008691978454589844, "gate_p_p90": 0.10556182861328126, "gate_p_std": 0.05725167814525776, "gated_delta_h_c_norm": 0.012848352376022377, "global_step": 1610, "grad_norm_max": 0.009010273963212967, "hidden_mse": 0.01031633061647881, "loss": 0.009589677158510314, "lr_blocks": 1.0040600155253766e-06, "lr_fusion": 1.0040600155253765e-05, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693616950511933, "transport_residual_norm": 0.06331087690778077, "velocity_mse": 0.00855804407328833} +{"anchor_distance_mean": 128.24453125, "avg_context_frames": 14.0, "avg_micro_time_s": 2.519199224583281, "delta_h_c_norm": 0.14316014591604473, "delta_h_d_norm": 0.06547745172865689, "epoch": 13, "gate_p_mean": 0.034181148125207984, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8315036818385124, "gate_p_p10": 0.0012046527583152056, "gate_p_p50": 0.008693790435791016, "gate_p_p90": 0.10552518963813783, "gate_p_std": 0.0589150981977582, "gated_delta_h_c_norm": 0.01329569733934477, "global_step": 1620, "grad_norm_max": 0.01135197188705206, "hidden_mse": 0.010749437293270603, "loss": 0.010447995283175261, "lr_blocks": 9.549150281252633e-07, "lr_fusion": 9.549150281252633e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692975389957429, "transport_residual_norm": 0.06381927807815373, "velocity_mse": 0.009373051568400115} +{"anchor_distance_mean": 126.5546875, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.643709483004932, "delta_h_c_norm": 0.14156744433566928, "delta_h_d_norm": 0.0665685118176043, "epoch": 13, "gate_p_mean": 0.0352926741441479, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8196930348873139, "gate_p_p10": 0.0010887765558436513, "gate_p_p50": 0.009688377380371094, "gate_p_p90": 0.11004300117492677, "gate_p_std": 0.05858848800999113, "gated_delta_h_c_norm": 0.01333081058910466, "global_step": 1630, "grad_norm_max": 0.03117949329316616, "hidden_mse": 0.010647887998493389, "loss": 0.010391313975560479, "lr_blocks": 9.068759265665384e-07, "lr_fusion": 9.068759265665384e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692650723457337, "transport_residual_norm": 0.06852464932017029, "velocity_mse": 0.00932652518968098} +{"anchor_distance_mean": 126.29609375000001, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.670517112067319, "delta_h_c_norm": 0.14163071056827903, "delta_h_d_norm": 0.06487284344621003, "epoch": 13, "gate_p_mean": 0.03487199998926371, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8156390488147736, "gate_p_p10": 0.0012486696243286133, "gate_p_p50": 0.010739231109619142, "gate_p_p90": 0.1061453104019165, "gate_p_std": 0.05565511987661012, "gated_delta_h_c_norm": 0.012146681078593248, "global_step": 1640, "grad_norm_max": 0.01168541144579649, "hidden_mse": 0.010439298339770177, "loss": 0.010428104433231056, "lr_blocks": 8.599558442598998e-07, "lr_fusion": 8.599558442598998e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692883813381195, "transport_residual_norm": 0.06369241708889604, "velocity_mse": 0.009384174586739392} +{"anchor_distance_mean": 124.184375, "avg_context_frames": 12.8, "avg_micro_time_s": 2.3456961355914245, "delta_h_c_norm": 0.14250470139086246, "delta_h_d_norm": 0.06558881564997136, "epoch": 13, "gate_p_mean": 0.035483009641757234, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8144554037600756, "gate_p_p10": 0.0013701677322387697, "gate_p_p50": 0.011188507080078125, "gate_p_p90": 0.10619655847549439, "gate_p_std": 0.05744470906211063, "gated_delta_h_c_norm": 0.012627450494619552, "global_step": 1650, "grad_norm_max": 0.007316017057746649, "hidden_mse": 0.010640696552582086, "loss": 0.01037751990952529, "lr_blocks": 8.141676086873574e-07, "lr_fusion": 8.141676086873572e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693048250675202, "transport_residual_norm": 0.05990185565315187, "velocity_mse": 0.009313450223999098} +{"anchor_distance_mean": 121.06640625, "avg_context_frames": 13.4, "avg_micro_time_s": 2.4135365220397946, "delta_h_c_norm": 0.14370002057403328, "delta_h_d_norm": 0.0661979456897825, "epoch": 13, "gate_p_mean": 0.03711045032250695, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8039127849042416, "gate_p_p10": 0.0013277673162519932, "gate_p_p50": 0.011696624755859376, "gate_p_p90": 0.11135525703430177, "gate_p_std": 0.060953352559590716, "gated_delta_h_c_norm": 0.014223469406715594, "global_step": 1660, "grad_norm_max": 0.00890736561268568, "hidden_mse": 0.010525041300570593, "loss": 0.010101042554015295, "lr_blocks": 7.695237378953224e-07, "lr_fusion": 7.695237378953223e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692795956134796, "transport_residual_norm": 0.06503112809732557, "velocity_mse": 0.009048538375645876} +{"anchor_distance_mean": 126.29375, "avg_context_frames": 15.0, "avg_micro_time_s": 2.600719206202484, "delta_h_c_norm": 0.1397578202188015, "delta_h_d_norm": 0.06277872044593096, "epoch": 13, "gate_p_mean": 0.029630556283518673, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.845125225931406, "gate_p_p10": 0.0011027669766917826, "gate_p_p50": 0.008222293853759766, "gate_p_p90": 0.08867340087890625, "gate_p_std": 0.05194451889256016, "gated_delta_h_c_norm": 0.011012695908721072, "global_step": 1670, "grad_norm_max": 0.006753512192517519, "hidden_mse": 0.009513741749105976, "loss": 0.009366970794508234, "lr_blocks": 7.260364370723044e-07, "lr_fusion": 7.260364370723044e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692505586147309, "transport_residual_norm": 0.06626980751752853, "velocity_mse": 0.008415596606209875} +{"anchor_distance_mean": 122.85000000000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6162974901933924, "delta_h_c_norm": 0.13944709012284875, "delta_h_d_norm": 0.0649287368170917, "epoch": 13, "gate_p_mean": 0.028960380237549543, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8465617768466474, "gate_p_p10": 0.001248788833618164, "gate_p_p50": 0.0086456298828125, "gate_p_p90": 0.0838772475719452, "gate_p_std": 0.04954256118508056, "gated_delta_h_c_norm": 0.010318352290778422, "global_step": 1680, "grad_norm_max": 0.006081636529415846, "hidden_mse": 0.009364181582350284, "loss": 0.009050313488114626, "lr_blocks": 6.837175952121305e-07, "lr_fusion": 6.837175952121306e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69271125793457, "transport_residual_norm": 0.06999021680094301, "velocity_mse": 0.008113895368296653} +{"anchor_distance_mean": 126.0015625, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4271987941872797, "delta_h_c_norm": 0.14125694883987308, "delta_h_d_norm": 0.06543839937075972, "epoch": 13, "gate_p_mean": 0.03520988104864955, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8201685473322868, "gate_p_p10": 0.0012963675893843175, "gate_p_p50": 0.009435558319091798, "gate_p_p90": 0.10648465156555176, "gate_p_std": 0.06095476693008095, "gated_delta_h_c_norm": 0.013135749385401142, "global_step": 1690, "grad_norm_max": 0.008193200454115868, "hidden_mse": 0.010726260486990213, "loss": 0.010164930240716786, "lr_blocks": 6.425787818636131e-07, "lr_fusion": 6.425787818636131e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.693297171592713, "transport_residual_norm": 0.056016549980267887, "velocity_mse": 0.009092304177465848} +{"anchor_distance_mean": 126.52109375, "avg_context_frames": 12.8, "avg_micro_time_s": 2.3540422678765025, "delta_h_c_norm": 0.1435885031707585, "delta_h_d_norm": 0.06461555194109679, "epoch": 14, "gate_p_mean": 0.03756780846160837, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8107857841998339, "gate_p_p10": 0.001318359375, "gate_p_p50": 0.011030006408691407, "gate_p_p90": 0.11813635826110841, "gate_p_std": 0.0603349456214346, "gated_delta_h_c_norm": 0.013640097789175343, "global_step": 1700, "grad_norm_max": 0.008983015082776546, "hidden_mse": 0.010501214166288264, "loss": 0.01014297781221103, "lr_blocks": 6.026312439675553e-07, "lr_fusion": 6.026312439675552e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692474067211151, "transport_residual_norm": 0.0583140566945076, "velocity_mse": 0.009092856434290298} +{"anchor_distance_mean": 123.66328125000001, "avg_context_frames": 15.4, "avg_micro_time_s": 2.6447777815847076, "delta_h_c_norm": 0.1437319554388523, "delta_h_d_norm": 0.06675708745606244, "epoch": 14, "gate_p_mean": 0.03621476140106097, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8207537315785886, "gate_p_p10": 0.0013047933578491213, "gate_p_p50": 0.009731674194335939, "gate_p_p90": 0.11333322525024414, "gate_p_std": 0.0627629850176163, "gated_delta_h_c_norm": 0.014502920794620879, "global_step": 1710, "grad_norm_max": 0.00795330386608839, "hidden_mse": 0.01047207301016897, "loss": 0.009738185713649728, "lr_blocks": 5.63885902781941e-07, "lr_fusion": 5.6388590278194096e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692992544174194, "transport_residual_norm": 0.06531827473081649, "velocity_mse": 0.008690978423692286} +{"anchor_distance_mean": 126.03359375000001, "avg_context_frames": 15.200000000000001, "avg_micro_time_s": 2.6249043478019303, "delta_h_c_norm": 0.14146722219884397, "delta_h_d_norm": 0.06110543664544821, "epoch": 14, "gate_p_mean": 0.032384221733082086, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8302922442555428, "gate_p_p10": 0.0011391162872314454, "gate_p_p50": 0.00892477035522461, "gate_p_p90": 0.09820542335510254, "gate_p_std": 0.05479953882750124, "gated_delta_h_c_norm": 0.011886395646433812, "global_step": 1720, "grad_norm_max": 0.005537689197808504, "hidden_mse": 0.009320869081420824, "loss": 0.009095242148032413, "lr_blocks": 5.263533508961827e-07, "lr_fusion": 5.263533508961827e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692549562454223, "transport_residual_norm": 0.0675477446988225, "velocity_mse": 0.00816315523115918} +{"anchor_distance_mean": 123.14296875000001, "avg_context_frames": 14.0, "avg_micro_time_s": 2.4944509517197733, "delta_h_c_norm": 0.1435165110975504, "delta_h_d_norm": 0.06471430878154934, "epoch": 14, "gate_p_mean": 0.037980547349434346, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8064480926841497, "gate_p_p10": 0.001387882186099887, "gate_p_p50": 0.011869430541992188, "gate_p_p90": 0.11403088569641114, "gate_p_std": 0.06084474214585498, "gated_delta_h_c_norm": 0.013678513854392805, "global_step": 1730, "grad_norm_max": 0.009013469330966473, "hidden_mse": 0.010555854142876343, "loss": 0.010041087525314652, "lr_blocks": 4.900438493352056e-07, "lr_fusion": 4.900438493352055e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69208505153656, "transport_residual_norm": 0.06390772769227625, "velocity_mse": 0.00898550212441478} +{"anchor_distance_mean": 124.67109375000001, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.4099077696810127, "delta_h_c_norm": 0.1426487172022462, "delta_h_d_norm": 0.06589449057355523, "epoch": 14, "gate_p_mean": 0.04087619662168436, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.791329400241375, "gate_p_p10": 0.0013425111304968596, "gate_p_p50": 0.010707473754882813, "gate_p_p90": 0.12834956645965576, "gate_p_std": 0.06777921337634325, "gated_delta_h_c_norm": 0.014811432480928489, "global_step": 1740, "grad_norm_max": 0.0077249412424862385, "hidden_mse": 0.01145855950890109, "loss": 0.011506579950219021, "lr_blocks": 4.549673247541875e-07, "lr_fusion": 4.549673247541875e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692091715335847, "transport_residual_norm": 0.058440913353115326, "velocity_mse": 0.010360723902704195} +{"anchor_distance_mean": 121.8453125, "avg_context_frames": 12.600000000000001, "avg_micro_time_s": 2.336279150811606, "delta_h_c_norm": 0.14296163171529772, "delta_h_d_norm": 0.06519273985177279, "epoch": 14, "gate_p_mean": 0.038472344173351304, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7972506269812585, "gate_p_p10": 0.001485938997939229, "gate_p_p50": 0.011281490325927734, "gate_p_p90": 0.12119293212890625, "gate_p_std": 0.06147990777390078, "gated_delta_h_c_norm": 0.013586358522297816, "global_step": 1750, "grad_norm_max": 0.007994987070560455, "hidden_mse": 0.010841048654401676, "loss": 0.010454758486594074, "lr_blocks": 4.211333667247125e-07, "lr_fusion": 4.2113336672471245e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691912877559663, "transport_residual_norm": 0.059880614699795845, "velocity_mse": 0.00937065361358691} +{"anchor_distance_mean": 124.44296875, "avg_context_frames": 16.0, "avg_micro_time_s": 2.6955266967925127, "delta_h_c_norm": 0.140648153424263, "delta_h_d_norm": 0.0664528142195195, "epoch": 14, "gate_p_mean": 0.031440391333308074, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8335462003946305, "gate_p_p10": 0.0011779308319091798, "gate_p_p50": 0.008822441101074219, "gate_p_p90": 0.09277572631835938, "gate_p_std": 0.054303699301090094, "gated_delta_h_c_norm": 0.012028753470804078, "global_step": 1760, "grad_norm_max": 0.005378834903240204, "hidden_mse": 0.010171208815881983, "loss": 0.00986579676391557, "lr_blocks": 3.885512251130763e-07, "lr_fusion": 3.885512251130763e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.6918781042099, "transport_residual_norm": 0.06817689873278142, "velocity_mse": 0.008848675887566061} +{"anchor_distance_mean": 128.4390625, "avg_context_frames": 14.0, "avg_micro_time_s": 2.490123038137972, "delta_h_c_norm": 0.14139664862304926, "delta_h_d_norm": 0.06615008423104883, "epoch": 14, "gate_p_mean": 0.033118065274902624, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8366954907774926, "gate_p_p10": 0.0012230967171490193, "gate_p_p50": 0.008578777313232422, "gate_p_p90": 0.10288422107696533, "gate_p_std": 0.05872005197452382, "gated_delta_h_c_norm": 0.012743864548974672, "global_step": 1770, "grad_norm_max": 0.006281412206590176, "hidden_mse": 0.01027346380869858, "loss": 0.009747607348253952, "lr_blocks": 3.572298075514652e-07, "lr_fusion": 3.5722980755146517e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691983783245087, "transport_residual_norm": 0.062332799006253485, "velocity_mse": 0.008720260925474578} +{"anchor_distance_mean": 121.77968750000001, "avg_context_frames": 15.600000000000001, "avg_micro_time_s": 2.667939361119352, "delta_h_c_norm": 0.14216662757098675, "delta_h_d_norm": 0.06635986538603902, "epoch": 14, "gate_p_mean": 0.03298056882340461, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8266358934342861, "gate_p_p10": 0.001241445541381836, "gate_p_p50": 0.009034538269042968, "gate_p_p90": 0.10134941935539246, "gate_p_std": 0.05605688190553337, "gated_delta_h_c_norm": 0.011897874664282427, "global_step": 1780, "grad_norm_max": 0.008346100337803364, "hidden_mse": 0.010278082941658795, "loss": 0.00999935923027806, "lr_blocks": 3.271776770026963e-07, "lr_fusion": 3.271776770026963e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692710268497468, "transport_residual_norm": 0.06871715979650617, "velocity_mse": 0.008971550973365083} +{"anchor_distance_mean": 124.89921875, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.404363570676651, "delta_h_c_norm": 0.1429867004044354, "delta_h_d_norm": 0.06671208455227316, "epoch": 14, "gate_p_mean": 0.04161598390201107, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7974945161491633, "gate_p_p10": 0.0015956878196448088, "gate_p_p50": 0.012286186218261719, "gate_p_p90": 0.12801098823547363, "gate_p_std": 0.06721794557524845, "gated_delta_h_c_norm": 0.01533953876933083, "global_step": 1790, "grad_norm_max": 0.00464914133772254, "hidden_mse": 0.011365152732469143, "loss": 0.010732666481635534, "lr_blocks": 2.984030494191942e-07, "lr_fusion": 2.9840304941919415e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691777122020723, "transport_residual_norm": 0.058517938433215025, "velocity_mse": 0.009596151189180092} +{"anchor_distance_mean": 119.73515625, "avg_context_frames": 16.8, "avg_micro_time_s": 2.8008428039684077, "delta_h_c_norm": 0.14316369015723468, "delta_h_d_norm": 0.06801497978158295, "epoch": 14, "gate_p_mean": 0.03744155138265342, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8034232512116433, "gate_p_p10": 0.0015152215957641602, "gate_p_p50": 0.012140274047851562, "gate_p_p90": 0.11240990161895753, "gate_p_std": 0.05774451026227326, "gated_delta_h_c_norm": 0.012881294895487373, "global_step": 1800, "grad_norm_max": 0.005514129064977169, "hidden_mse": 0.010163118049968035, "loss": 0.009463976361439565, "lr_blocks": 2.7091379149682683e-07, "lr_fusion": 2.7091379149682685e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692448949813842, "transport_residual_norm": 0.071227763639763, "velocity_mse": 0.008447664542472921} +{"anchor_distance_mean": 125.19140625, "avg_context_frames": 12.600000000000001, "avg_micro_time_s": 2.3262952591467183, "delta_h_c_norm": 0.14315422642976047, "delta_h_d_norm": 0.06556109422817827, "epoch": 14, "gate_p_mean": 0.03822977090603672, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8018865454941988, "gate_p_p10": 0.0013128518592566253, "gate_p_p50": 0.011067581176757813, "gate_p_p90": 0.11832642555236816, "gate_p_std": 0.06297685360186733, "gated_delta_h_c_norm": 0.013805620533821639, "global_step": 1810, "grad_norm_max": 0.007135092280805111, "hidden_mse": 0.010555987714906224, "loss": 0.0103174614981981, "lr_blocks": 2.447174185242324e-07, "lr_fusion": 2.4471741852423237e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691703748703004, "transport_residual_norm": 0.06058671297505498, "velocity_mse": 0.009261862724088132} +{"anchor_distance_mean": 130.8125, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.5545062459306793, "delta_h_c_norm": 0.1407848495990038, "delta_h_d_norm": 0.06715112682431937, "epoch": 14, "gate_p_mean": 0.03414176686492283, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8287948489189149, "gate_p_p10": 0.001251645013689995, "gate_p_p50": 0.009541320800781251, "gate_p_p90": 0.10272493362426759, "gate_p_std": 0.05855310820625164, "gated_delta_h_c_norm": 0.012694219031254762, "global_step": 1820, "grad_norm_max": 0.0059729949571192265, "hidden_mse": 0.009928793887957, "loss": 0.009489995989133604, "lr_blocks": 2.198210923282118e-07, "lr_fusion": 2.1982109232821178e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69198236465454, "transport_residual_norm": 0.06333978432230651, "velocity_mse": 0.008497116598300635} +{"anchor_distance_mean": 121.84453125, "avg_context_frames": 11.600000000000001, "avg_micro_time_s": 2.226930264668772, "delta_h_c_norm": 0.14452478988096118, "delta_h_d_norm": 0.06716786818578839, "epoch": 15, "gate_p_mean": 0.04249224087689072, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.7847621437162161, "gate_p_p10": 0.0014559984207153322, "gate_p_p50": 0.011662673950195313, "gate_p_p90": 0.13603363037109376, "gate_p_std": 0.06988036204129458, "gated_delta_h_c_norm": 0.015458242155727931, "global_step": 1830, "grad_norm_max": 0.007656003814190626, "hidden_mse": 0.0115425567724742, "loss": 0.011101023916853593, "lr_blocks": 1.962316193157593e-07, "lr_fusion": 1.962316193157593e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691834592819214, "transport_residual_norm": 0.05553537025116384, "velocity_mse": 0.009946768192457968} +{"anchor_distance_mean": 124.0875, "avg_context_frames": 14.8, "avg_micro_time_s": 2.577214251825353, "delta_h_c_norm": 0.1405804619193077, "delta_h_d_norm": 0.06705828784033656, "epoch": 15, "gate_p_mean": 0.03545438448200002, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8133816875517369, "gate_p_p10": 0.0012052154168486597, "gate_p_p50": 0.009301948547363282, "gate_p_p90": 0.10927863121032716, "gate_p_std": 0.06098011991707608, "gated_delta_h_c_norm": 0.013197299893363379, "global_step": 1840, "grad_norm_max": 0.0056404112838208675, "hidden_mse": 0.010329595359507948, "loss": 0.009956879302626476, "lr_blocks": 1.7395544861325718e-07, "lr_fusion": 1.7395544861325718e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691772067546845, "transport_residual_norm": 0.06703304369002581, "velocity_mse": 0.008923919755034149} +{"anchor_distance_mean": 124.37890625, "avg_context_frames": 15.4, "avg_micro_time_s": 2.640759675142181, "delta_h_c_norm": 0.14322862084954976, "delta_h_d_norm": 0.06625800896435977, "epoch": 15, "gate_p_mean": 0.03700831406167709, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8159382767975331, "gate_p_p10": 0.001480412483215332, "gate_p_p50": 0.009899520874023439, "gate_p_p90": 0.1134335994720459, "gate_p_std": 0.06249171544332058, "gated_delta_h_c_norm": 0.013983241618552711, "global_step": 1850, "grad_norm_max": 0.006719102151691914, "hidden_mse": 0.010379394362098539, "loss": 0.009976123401429504, "lr_blocks": 1.5299867030334815e-07, "lr_fusion": 1.5299867030334814e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692216038703918, "transport_residual_norm": 0.06558769061230123, "velocity_mse": 0.008938183935242705} +{"anchor_distance_mean": 125.5484375, "avg_context_frames": 14.4, "avg_micro_time_s": 2.517447467474267, "delta_h_c_norm": 0.1413101867772639, "delta_h_d_norm": 0.06790373935364187, "epoch": 15, "gate_p_mean": 0.036218799388734624, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8183043133467436, "gate_p_p10": 0.0013306903652846815, "gate_p_p50": 0.009799385070800783, "gate_p_p90": 0.1142887592315674, "gate_p_std": 0.06100086622172967, "gated_delta_h_c_norm": 0.013446813717018814, "global_step": 1860, "grad_norm_max": 0.004700628109276295, "hidden_mse": 0.010539782579871827, "loss": 0.009867401045630686, "lr_blocks": 1.333670137599713e-07, "lr_fusion": 1.333670137599713e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691716229915619, "transport_residual_norm": 0.06569369798526169, "velocity_mse": 0.008813422813545913} +{"anchor_distance_mean": 123.6296875, "avg_context_frames": 13.200000000000001, "avg_micro_time_s": 2.4086661923618524, "delta_h_c_norm": 0.1425080167129636, "delta_h_d_norm": 0.0651101325172931, "epoch": 15, "gate_p_mean": 0.03843713882379234, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8067175246775151, "gate_p_p10": 0.0014422273263335228, "gate_p_p50": 0.010707283020019533, "gate_p_p90": 0.11830317974090576, "gate_p_std": 0.06502026828238741, "gated_delta_h_c_norm": 0.014239987911423668, "global_step": 1870, "grad_norm_max": 0.005073883105069399, "hidden_mse": 0.010463360778521747, "loss": 0.010162824095459655, "lr_blocks": 1.1506584608200366e-07, "lr_fusion": 1.1506584608200367e-06, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692511665821076, "transport_residual_norm": 0.05793416821397841, "velocity_mse": 0.00911648803157732} +{"anchor_distance_mean": 125.22109375000001, "avg_context_frames": 11.8, "avg_micro_time_s": 2.2493542150885335, "delta_h_c_norm": 0.1425988276489079, "delta_h_d_norm": 0.06439000531099737, "epoch": 15, "gate_p_mean": 0.03921799895470031, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8054277073591948, "gate_p_p10": 0.0014035319909453392, "gate_p_p50": 0.01272726058959961, "gate_p_p90": 0.12060489654541016, "gate_p_std": 0.0625863373046741, "gated_delta_h_c_norm": 0.013661489753576462, "global_step": 1880, "grad_norm_max": 0.004061415325850248, "hidden_mse": 0.010543550056172536, "loss": 0.010371940062032082, "lr_blocks": 9.810017062595322e-08, "lr_fusion": 9.810017062595322e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691524195671082, "transport_residual_norm": 0.05935611906461418, "velocity_mse": 0.009317585136159323} +{"anchor_distance_mean": 127.10781250000001, "avg_context_frames": 14.0, "avg_micro_time_s": 2.478096133320651, "delta_h_c_norm": 0.14182132668793201, "delta_h_d_norm": 0.0653932779096067, "epoch": 15, "gate_p_mean": 0.03810298062744551, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8024952676147223, "gate_p_p10": 0.0013253116281703116, "gate_p_p50": 0.011650562286376953, "gate_p_p90": 0.11796231269836427, "gate_p_std": 0.06078981920145452, "gated_delta_h_c_norm": 0.013417408070381498, "global_step": 1890, "grad_norm_max": 0.005406858399510384, "hidden_mse": 0.010593694203998895, "loss": 0.00987786817131564, "lr_blocks": 8.247462563808816e-08, "lr_fusion": 8.247462563808817e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69169272184372, "transport_residual_norm": 0.06386055490002036, "velocity_mse": 0.008818498693290168} +{"anchor_distance_mean": 125.80703125000001, "avg_context_frames": 15.0, "avg_micro_time_s": 2.594486193242483, "delta_h_c_norm": 0.14104599794372916, "delta_h_d_norm": 0.06423986125737428, "epoch": 15, "gate_p_mean": 0.035874913778388876, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8213617060333491, "gate_p_p10": 0.001252865791320801, "gate_p_p50": 0.01029033660888672, "gate_p_p90": 0.11185562610626221, "gate_p_std": 0.06033796835690737, "gated_delta_h_c_norm": 0.013230244834267069, "global_step": 1900, "grad_norm_max": 0.004896499216556549, "hidden_mse": 0.01076923911459744, "loss": 0.010614202771103011, "lr_blocks": 6.819348298638839e-08, "lr_fusion": 6.819348298638839e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69146250486374, "transport_residual_norm": 0.06687471861951054, "velocity_mse": 0.00953727885207627} +{"anchor_distance_mean": 123.11171875000001, "avg_context_frames": 18.400000000000002, "avg_micro_time_s": 2.9698089701239954, "delta_h_c_norm": 0.13928372133523226, "delta_h_d_norm": 0.06503545017912984, "epoch": 15, "gate_p_mean": 0.03143622559728101, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8358301036059856, "gate_p_p10": 0.0011606454849243165, "gate_p_p50": 0.008624458312988281, "gate_p_p90": 0.09470901489257813, "gate_p_std": 0.05396560595836491, "gated_delta_h_c_norm": 0.011887404607841746, "global_step": 1910, "grad_norm_max": 0.006939377635717392, "hidden_mse": 0.009729971352498978, "loss": 0.009157144278287888, "lr_blocks": 5.526064699265754e-08, "lr_fusion": 5.526064699265753e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692211663722992, "transport_residual_norm": 0.07729351809248329, "velocity_mse": 0.008184147189604118} +{"anchor_distance_mean": 128.1796875, "avg_context_frames": 15.4, "avg_micro_time_s": 2.6261373986591936, "delta_h_c_norm": 0.14174036998301745, "delta_h_d_norm": 0.06562835942022503, "epoch": 15, "gate_p_mean": 0.031782829837175085, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8400506094098091, "gate_p_p10": 0.0012105703353881836, "gate_p_p50": 0.008818912506103517, "gate_p_p90": 0.09671630859375001, "gate_p_std": 0.05474699921905995, "gated_delta_h_c_norm": 0.012116907874587924, "global_step": 1920, "grad_norm_max": 0.007330034859478474, "hidden_mse": 0.009542921665706673, "loss": 0.008913772259256802, "lr_blocks": 4.367965336512403e-08, "lr_fusion": 4.367965336512403e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69249186515808, "transport_residual_norm": 0.06792828002944588, "velocity_mse": 0.007959480083081871} +{"anchor_distance_mean": 123.27265625000001, "avg_context_frames": 11.600000000000001, "avg_micro_time_s": 2.224096468303469, "delta_h_c_norm": 0.14115401823073626, "delta_h_d_norm": 0.0658903154078871, "epoch": 15, "gate_p_mean": 0.03484603494871408, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8221076466143131, "gate_p_p10": 0.001422119140625, "gate_p_p50": 0.010629749298095703, "gate_p_p90": 0.10616269111633302, "gate_p_std": 0.057405780418775976, "gated_delta_h_c_norm": 0.012105253803019878, "global_step": 1930, "grad_norm_max": 0.011694157496094704, "hidden_mse": 0.01060591790592298, "loss": 0.01035737147612963, "lr_blocks": 3.345366823180929e-08, "lr_fusion": 3.3453668231809286e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691126537323, "transport_residual_norm": 0.060324323596432805, "velocity_mse": 0.009296779631404207} +{"anchor_distance_mean": 128.40703125000002, "avg_context_frames": 15.0, "avg_micro_time_s": 2.5970872980047717, "delta_h_c_norm": 0.14136814298108222, "delta_h_d_norm": 0.06570565956644714, "epoch": 15, "gate_p_mean": 0.03477297404315323, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8217580877244473, "gate_p_p10": 0.0011553525924682618, "gate_p_p50": 0.009398174285888673, "gate_p_p90": 0.1067582130432129, "gate_p_std": 0.058890018064994366, "gated_delta_h_c_norm": 0.012621946302533616, "global_step": 1940, "grad_norm_max": 0.007130275946110487, "hidden_mse": 0.010258641478139909, "loss": 0.010090042746742257, "lr_blocks": 2.4585487274942922e-08, "lr_fusion": 2.458548727494292e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692333233356477, "transport_residual_norm": 0.06623481092974544, "velocity_mse": 0.009064178590779194} +{"anchor_distance_mean": 121.49062500000001, "avg_context_frames": 15.4, "avg_micro_time_s": 2.62764158420905, "delta_h_c_norm": 0.1422020279802382, "delta_h_d_norm": 0.06708167307078838, "epoch": 15, "gate_p_mean": 0.03751733936951496, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8087432645261288, "gate_p_p10": 0.0013039827346801758, "gate_p_p50": 0.010567665100097656, "gate_p_p90": 0.11458129882812501, "gate_p_std": 0.0609739008708857, "gated_delta_h_c_norm": 0.013479097583331169, "global_step": 1950, "grad_norm_max": 0.009140651673078537, "hidden_mse": 0.01042317248065956, "loss": 0.009756203732104042, "lr_blocks": 1.7077534966650767e-08, "lr_fusion": 1.7077534966650766e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692381167411805, "transport_residual_norm": 0.06853321166709066, "velocity_mse": 0.008713886482291855} +{"anchor_distance_mean": 123.50234375000001, "avg_context_frames": 14.600000000000001, "avg_micro_time_s": 2.5451702435282644, "delta_h_c_norm": 0.1429362401366234, "delta_h_d_norm": 0.06762755992822349, "epoch": 16, "gate_p_mean": 0.04016138701117598, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8052759684622288, "gate_p_p10": 0.001464223861694336, "gate_p_p50": 0.010452461242675782, "gate_p_p90": 0.12683334350585937, "gate_p_std": 0.06695701134740376, "gated_delta_h_c_norm": 0.014940886966360268, "global_step": 1960, "grad_norm_max": 0.005991573911160231, "hidden_mse": 0.011143166679539719, "loss": 0.010476734832627699, "lr_blocks": 1.0931863906127327e-08, "lr_fusion": 1.0931863906127327e-07, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692288529872895, "transport_residual_norm": 0.06374651831574739, "velocity_mse": 0.009362418172531762} +{"anchor_distance_mean": 126.00078125, "avg_context_frames": 18.2, "avg_micro_time_s": 2.9455070269294086, "delta_h_c_norm": 0.13756417986005545, "delta_h_d_norm": 0.06396158835850656, "epoch": 16, "gate_p_mean": 0.02627640145365149, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8638609558343888, "gate_p_p10": 0.0010303974151611329, "gate_p_p50": 0.007155227661132813, "gate_p_p90": 0.07808592319488526, "gate_p_std": 0.04667906932299957, "gated_delta_h_c_norm": 0.009650876047089696, "global_step": 1970, "grad_norm_max": 0.005146809387952089, "hidden_mse": 0.008810659882146866, "loss": 0.008625803515315056, "lr_blocks": 6.150154258476315e-09, "lr_fusion": 6.150154258476315e-08, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692161679267883, "transport_residual_norm": 0.07643249831162394, "velocity_mse": 0.0077447375020710755} +{"anchor_distance_mean": 122.265625, "avg_context_frames": 14.4, "avg_micro_time_s": 2.530630034361093, "delta_h_c_norm": 0.14139191806316376, "delta_h_d_norm": 0.06665236153639853, "epoch": 16, "gate_p_mean": 0.040134805283742025, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8050536096096039, "gate_p_p10": 0.0014531612396240234, "gate_p_p50": 0.010640907287597657, "gate_p_p90": 0.12415161132812501, "gate_p_std": 0.06661342532606795, "gated_delta_h_c_norm": 0.014669717295328157, "global_step": 1980, "grad_norm_max": 0.005900868214666843, "hidden_mse": 0.011178363725775853, "loss": 0.010459155487478711, "lr_blocks": 2.7337132953697555e-09, "lr_fusion": 2.7337132953697554e-08, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.692018175125122, "transport_residual_norm": 0.06383700659498573, "velocity_mse": 0.00934131913818419} +{"anchor_distance_mean": 121.97578125000001, "avg_context_frames": 13.600000000000001, "avg_micro_time_s": 2.4541102713075817, "delta_h_c_norm": 0.14260861231014132, "delta_h_d_norm": 0.0641879737842828, "epoch": 16, "gate_p_mean": 0.03881859953980893, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.8018832873553038, "gate_p_p10": 0.0014590740203857422, "gate_p_p50": 0.012335586547851564, "gate_p_p90": 0.11598045825958253, "gate_p_std": 0.06261561290593819, "gated_delta_h_c_norm": 0.013468718749936671, "global_step": 1990, "grad_norm_max": 0.00832299329340458, "hidden_mse": 0.01081486045732163, "loss": 0.010269785349373707, "lr_blocks": 6.834750376549793e-10, "lr_fusion": 6.834750376549792e-09, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.69172682762146, "transport_residual_norm": 0.060816779872402554, "velocity_mse": 0.009188299308880233} +{"anchor_distance_mean": 124.96406250000001, "avg_context_frames": 15.0, "avg_micro_time_s": 2.6007237769503266, "delta_h_c_norm": 0.14247616585344078, "delta_h_d_norm": 0.0646928537171334, "epoch": 16, "gate_p_mean": 0.03433765485533513, "gate_p_near_one": 0.0, "gate_p_near_zero": 0.818176107108593, "gate_p_p10": 0.0011486958712339402, "gate_p_p50": 0.00918283462524414, "gate_p_p90": 0.10571666955947877, "gate_p_std": 0.05776586307911202, "gated_delta_h_c_norm": 0.012530222797067836, "global_step": 2000, "grad_norm_max": 0.004970205947756767, "hidden_mse": 0.01013188952638302, "loss": 0.009942494358983823, "lr_blocks": 0.0, "lr_fusion": 0.0, "peak_memory_gib": 42.58950757980347, "transport_attention_entropy": 8.691831862926483, "transport_residual_norm": 0.06853882661089301, "velocity_mse": 0.008929305418860167} diff --git a/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/predictor_step_00100.safetensors b/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/predictor_step_00100.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1f1aa70e1986477c1fbefbfc1adc89dd4f7f1ca9 --- /dev/null +++ b/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/predictor_step_00100.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74bff71310408ec86e7f81030497ee3aeb2e871c663912533d34f9d4b390bc36 +size 885425804 diff --git a/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/predictor_step_00200.safetensors b/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/predictor_step_00200.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fb61ef6fae1bf050ba0abc222efbf22390c1419d --- /dev/null +++ b/checkpoints/predictor_v4_atc-full_transport-global-chunk_blocks1-52_fulltrain100_fp32master_4gpu_bs8x2_flr1e-4_blr1e-5_steps2000/predictor_step_00200.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eea4729a38c187ba50a034aedc6256c9aae2b9afef8fbe72300f1f2817b6579d +size 885425804