Publish complete reported oscillator-flow study artifacts (part 3)

.gitattributes

@@ -141,3 +141,21 @@ study/main/cifar10-dit-s101/evaluations/test-nfe16/grid.png filter=lfs diff=lfs
 study/main/cifar10-dit-s101/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
 study/main/cifar10-dit-s101/evaluations/test-nfe64/grid.png filter=lfs diff=lfs merge=lfs -text
 study/main/cifar10-dit-s101/evaluations/test-nfe8/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-dit-s102/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-dit-s103/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-dit-s104/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-dit-s105/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-dit-s106/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s101/evaluations/identity/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s101/evaluations/shuffle/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s101/evaluations/test-nfe16/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s101/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s101/evaluations/test-nfe64/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s101/evaluations/test-nfe8/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s102/evaluations/identity/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s102/evaluations/shuffle/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s102/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s103/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s104/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s105/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text
+study/main/cifar10-onn-s106/evaluations/test-nfe32/grid.png filter=lfs diff=lfs merge=lfs -text

study-records/checks/cifar10-metric-smoke/evaluations/validation/repeat/control.json

@@ -0,0 +1,19 @@
+{
+  "kind": "validation-metric-reproducibility",
+  "status": "completed",
+  "generated_count": 128,
+  "real_count": 5000,
+  "first_metrics": {
+    "frechet_inception_distance": 487.66587247902373,
+    "kernel_inception_distance_mean": 0.6097407078742981,
+    "kernel_inception_distance_std": 0.005960302714145461
+  },
+  "repeat_metrics": {
+    "frechet_inception_distance": 487.66587247902373,
+    "kernel_inception_distance_mean": 0.6097407078742981,
+    "kernel_inception_distance_std": 0.005960302714145461
+  },
+  "repeat_max_absolute_difference": 0.0,
+  "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97",
+  "source_evaluation": "runs/checks/cifar10-metric-smoke/evaluations/validation"
+}

study-records/checks/cifar10-metric-smoke/evaluations/validation/repeat/run.log

@@ -0,0 +1 @@
+2026-10-16T23:39:40Z INFO {"event": "metric_reproducibility", "kind": "validation-metric-reproducibility", "status": "completed", "generated_count": 128, "real_count": 5000, "first_metrics": {"frechet_inception_distance": 487.66587247902373, "kernel_inception_distance_mean": 0.6097407078742981, "kernel_inception_distance_std": 0.005960302714145461}, "repeat_metrics": {"frechet_inception_distance": 487.66587247902373, "kernel_inception_distance_mean": 0.6097407078742981, "kernel_inception_distance_std": 0.005960302714145461}, "repeat_max_absolute_difference": 0.0, "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_evaluation": "runs/checks/cifar10-metric-smoke/evaluations/validation"}

study-records/checks/final-export-reloads/control.json

@@ -0,0 +1,356 @@
+{
+  "status": "completed",
+  "kind": "all-reported-checkpoint-export-reloads",
+  "device": "cpu",
+  "source_revision": "b49e52bb12834bd0f151a4f45a36462182130a8d",
+  "publication": "exports/20261017T191030.039367Z/publication.json",
+  "compared_checkpoints": 31,
+  "all_state_tensors_equal": true,
+  "prediction_max_absolute_difference": 0.0,
+  "cpu_threads": 4,
+  "procedure": "Strictly load original EMA checkpoints and exported safetensors; compare all state tensors and configurations; compare two fixed Gaussian inputs at times 0.25 and 0.75 per checkpoint. Full inputs and both outputs are preserved.",
+  "tensor_artifact_sha256": "050793bd048da810ce6135f14dd5d23d4328f0b3488d8b7ecbb18a80ec123d43",
+  "checks": [
+    {
+      "source": "runs/checks/cifar10-metric-smoke",
+      "export": "exports/20261017T191030.039367Z/cifar10/checks/cifar10-metric-smoke",
+      "tensor_prefix": "run_00",
+      "input_seed": 20261016,
+      "checkpoint_sha256": "8e8cafa8870ab0ff9c618a39ba1f368be3966ff67459009f4ca195f437cd1aec",
+      "exported_weights_sha256": "c6ef33c9235e17feb0c2656fcc29364fa02d5dd9dabf27b05c314c01420090d9",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/checks/mnist-smoke",
+      "export": "exports/20261017T191030.039367Z/mnist/checks/mnist-smoke",
+      "tensor_prefix": "run_01",
+      "input_seed": 20261017,
+      "checkpoint_sha256": "8f38d4c79f09a651583cc67b6596ab438a75269ad8fcb6a89db4135257a2a9db",
+      "exported_weights_sha256": "89560b05f7e7318c1e71d7f8436e6458e469d07b9544ab602c756368d1482404",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/ablations/cifar10-no_coupling-s101",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_coupling-s101",
+      "tensor_prefix": "run_02",
+      "input_seed": 20261018,
+      "checkpoint_sha256": "3ec78d32218e855970086aafa4e6930fbfe4204b39228bfb133280ba95086f0d",
+      "exported_weights_sha256": "90e80072c9a90978f8f2a30d80412ba0fb7976a049d69dd09c97210c746a5313",
+      "state_tensors": 61,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/ablations/cifar10-no_coupling-s102",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_coupling-s102",
+      "tensor_prefix": "run_03",
+      "input_seed": 20261019,
+      "checkpoint_sha256": "0b3cbc2ac4115efd57aed785b957b2270de7683d46d26a20ed6b0908bd329ad2",
+      "exported_weights_sha256": "7b4006f783489f1f3eb11be8f7441aa92b5df640cf7a775ad5498422fb2ac54d",
+      "state_tensors": 61,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/ablations/cifar10-no_frequency-s101",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_frequency-s101",
+      "tensor_prefix": "run_04",
+      "input_seed": 20261020,
+      "checkpoint_sha256": "5de69e1afb911e81cba8030f52489023a73a601e54690a5babf6d0663d9d1b26",
+      "exported_weights_sha256": "e414bd1148c47d7eb29817e4d725bdbffb03b37df5abbbc41e78924c149ca58a",
+      "state_tensors": 61,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/ablations/cifar10-no_frequency-s102",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_frequency-s102",
+      "tensor_prefix": "run_05",
+      "input_seed": 20261021,
+      "checkpoint_sha256": "14fa697dfda45dc97daac9ff222051777ca99779e234e429f4c6f1ef4fc2f315",
+      "exported_weights_sha256": "395b21f842524dfe11c55360d2093be3d562406a87a0f3af51b154887f33285a",
+      "state_tensors": 61,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/checks/fixed-batch",
+      "export": "exports/20261017T191030.039367Z/mnist/study/checks/fixed-batch",
+      "tensor_prefix": "run_06",
+      "input_seed": 20261022,
+      "checkpoint_sha256": "5d36c5c4faf91992c6cba4b3bdfb01880b396f817a2ee17924878ada5fe36de0",
+      "exported_weights_sha256": "dbc1290adf03cf30958883e06410e5867c99f943e25bdd637684610e552a369f",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/controls/untrained-conv",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/controls/untrained-conv",
+      "tensor_prefix": "run_07",
+      "input_seed": 20261023,
+      "checkpoint_sha256": "c3ef26854d2b6b534035ebf5993c7814172283e3a3a7febf89f703798283e6e9",
+      "exported_weights_sha256": "a09c911b1517c717c7527fa511e42da3abe44ad546b1ac8c47726cf2dc96af82",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/controls/untrained-dit",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/controls/untrained-dit",
+      "tensor_prefix": "run_08",
+      "input_seed": 20261024,
+      "checkpoint_sha256": "f88d4f36bb900a4e46ba2852061a0c8c8d1dcb0f465a8def494aee59c03793ca",
+      "exported_weights_sha256": "e0e8e1d0ed79b23b982ae432972278986797979ab46c66c70acb115217b00c18",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/controls/untrained-onn",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/controls/untrained-onn",
+      "tensor_prefix": "run_09",
+      "input_seed": 20261025,
+      "checkpoint_sha256": "cd7618119dc6162268c54e9a8b4484d41d52ca10dbdaf76798f2a2afbce56a04",
+      "exported_weights_sha256": "c6ef33c9235e17feb0c2656fcc29364fa02d5dd9dabf27b05c314c01420090d9",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-conv-s101",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s101",
+      "tensor_prefix": "run_10",
+      "input_seed": 20261026,
+      "checkpoint_sha256": "d283104b77bb38b474d78f48ee64bf89e90cd8ecb2fc32d2cced963261e28cbd",
+      "exported_weights_sha256": "d20f3947bcf79794427f8215c597a57c4b5eef3b308ccaccf1960beb86b4188e",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-conv-s102",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s102",
+      "tensor_prefix": "run_11",
+      "input_seed": 20261027,
+      "checkpoint_sha256": "ebb6eafd617b874a4fd559f32996118fdc64418f9206aa38faeb0645c4545b82",
+      "exported_weights_sha256": "7c81f89410f40e356ad05be8cb60016c65181208541c303d8ba1665778d1ce3c",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-conv-s103",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s103",
+      "tensor_prefix": "run_12",
+      "input_seed": 20261028,
+      "checkpoint_sha256": "4cb949dcbdf984ec4658182017a6f2be36da6eb9eb8d6891f6ebaf7119156a43",
+      "exported_weights_sha256": "66eb47d34559e0a1127012efc6888a22546eb0b9e9b94e7e28821dbd6c404441",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-conv-s104",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s104",
+      "tensor_prefix": "run_13",
+      "input_seed": 20261029,
+      "checkpoint_sha256": "384185eaffe897a164f9680f1506f8de0136c16b0d60ff9c491226f04488f574",
+      "exported_weights_sha256": "c896d46723b1f6e52eb8b6ad1e8fbdfc9c889784b5e114fe73606f538ed3d1ac",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-conv-s105",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s105",
+      "tensor_prefix": "run_14",
+      "input_seed": 20261030,
+      "checkpoint_sha256": "1f90d6fd0182341298cc1812aca672144633c8ea7c6ab57816bf9949e597352f",
+      "exported_weights_sha256": "d872edbf5e8bf7c11f03b6637adfc2762a009a6b0ea659f48e595e4d853cd5b7",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-conv-s106",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s106",
+      "tensor_prefix": "run_15",
+      "input_seed": 20261031,
+      "checkpoint_sha256": "fc27f7e6924da1c7426ca800f5916d5836ea6f8285d7fc926ba4ab981c156194",
+      "exported_weights_sha256": "2057aaaf7d16b4220dfdc8bc9d8ef32a6e91388223e30e65c2d72252f4562cd4",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-dit-s101",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s101",
+      "tensor_prefix": "run_16",
+      "input_seed": 20261032,
+      "checkpoint_sha256": "030d935262170666e1c5f3da7020026a573aa3de03b9f86a9c67f036afbe0dce",
+      "exported_weights_sha256": "cc227d4e18a7e1d94a2a9e0d34e809f86a8798e0c8e283af84968e529a1c8a67",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-dit-s102",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s102",
+      "tensor_prefix": "run_17",
+      "input_seed": 20261033,
+      "checkpoint_sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15",
+      "exported_weights_sha256": "2e584a96062f1a454dce28bb6322e17731df8999e32045b8ed7f9fe8fb4453cc",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-dit-s103",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s103",
+      "tensor_prefix": "run_18",
+      "input_seed": 20261034,
+      "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8",
+      "exported_weights_sha256": "75bca1cf3738374c420f34e36b0679e6369cf3fa86343aa762f0ff85bd6df64b",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-dit-s104",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s104",
+      "tensor_prefix": "run_19",
+      "input_seed": 20261035,
+      "checkpoint_sha256": "931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a",
+      "exported_weights_sha256": "3aba95b3af00dd2cad93f206c5af6d73f07431b491ed2e358a2feb8c8a1991c1",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-dit-s105",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s105",
+      "tensor_prefix": "run_20",
+      "input_seed": 20261036,
+      "checkpoint_sha256": "eec459e41060cbff6d07971d0c2fd363ccc1f525e23cc72c6b5c872e2cf832ac",
+      "exported_weights_sha256": "38b99f5c030c04a12c83c9b41ee55c16a4cde20c557e29b274950dbeb6d4a292",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-dit-s106",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s106",
+      "tensor_prefix": "run_21",
+      "input_seed": 20261037,
+      "checkpoint_sha256": "c05ea7f6e8db40d1fee8137f96251a02fd2c1ee199d91e2214aa46a17c51cc5e",
+      "exported_weights_sha256": "f4b8c7eddf76c637e828ccb902517b2f2efbfb2e269986f81bd5ac965e541f3e",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-onn-s101",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s101",
+      "tensor_prefix": "run_22",
+      "input_seed": 20261038,
+      "checkpoint_sha256": "e3eb0aabeaf1f372fbcb2f23bf94625b9fd2a1424504818a02b5d37a4522d6b1",
+      "exported_weights_sha256": "bb7094d51557bb6d7ccfdbaebd393c3cc33bbfcb48963aba33e58c48af187f99",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-onn-s102",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s102",
+      "tensor_prefix": "run_23",
+      "input_seed": 20261039,
+      "checkpoint_sha256": "e5cf67558e832212ded9b04225e34091cd62715d9df0773c5d22685da64e8366",
+      "exported_weights_sha256": "06a34f3c2e2dae9947112142592c6747376ba44959be8efb0ba37bc3687276cd",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-onn-s103",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s103",
+      "tensor_prefix": "run_24",
+      "input_seed": 20261040,
+      "checkpoint_sha256": "51090d973909f69682a6f2c126cf1377fbda43f249646733df76a4256e4566a1",
+      "exported_weights_sha256": "93b25b900dff3df0f60701b0456a0b45c89eba9e2da7f539f2c290f575f5e4e7",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-onn-s104",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s104",
+      "tensor_prefix": "run_25",
+      "input_seed": 20261041,
+      "checkpoint_sha256": "a0262716020a4118b0f750cf092274f8748ae86cb5ea23b3a700d75bc86e6322",
+      "exported_weights_sha256": "b306ed41c2b415cf5be073c0bbf0c4fcd4fdfee714ac4eab05aff3061021f620",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-onn-s105",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s105",
+      "tensor_prefix": "run_26",
+      "input_seed": 20261042,
+      "checkpoint_sha256": "6ae577c27cfe5120f2824baff526173d912511a47c56665da170c0b5a227251a",
+      "exported_weights_sha256": "12d9c6bfed79fd4a812af0b18f053486c5852171826f6e0eb453c4868fa3bd00",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/main/cifar10-onn-s106",
+      "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s106",
+      "tensor_prefix": "run_27",
+      "input_seed": 20261043,
+      "checkpoint_sha256": "e9595ec0d0fe7e945b48301e630d059d2077a641c3da27f0b6c205899f4da0b8",
+      "exported_weights_sha256": "c8b04e345990d920092c9b209471de71092020c363ac4004d631fddb476b3984",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/pilot/mnist-conv-s101",
+      "export": "exports/20261017T191030.039367Z/mnist/study/pilot/mnist-conv-s101",
+      "tensor_prefix": "run_28",
+      "input_seed": 20261044,
+      "checkpoint_sha256": "3fffc15466b2561ab3b0ebe9ad66272c4c4e3fe99e0f13602910d1f7da80b651",
+      "exported_weights_sha256": "bffd1bfd5e174490442d518b745e2f1da3416b0e2594b714f47a5dd9f16c1e31",
+      "state_tensors": 69,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/pilot/mnist-dit-s101",
+      "export": "exports/20261017T191030.039367Z/mnist/study/pilot/mnist-dit-s101",
+      "tensor_prefix": "run_29",
+      "input_seed": 20261045,
+      "checkpoint_sha256": "628a9d5c15326d0a89bb13ecab677c36bbe22b0de447d3e9bba981f82e164ee0",
+      "exported_weights_sha256": "a72223e84aa002fcb68459de860540f70499c1964584c2b1b37b437d2e327366",
+      "state_tensors": 52,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    },
+    {
+      "source": "runs/study/pilot/mnist-onn-s101",
+      "export": "exports/20261017T191030.039367Z/mnist/study/pilot/mnist-onn-s101",
+      "tensor_prefix": "run_30",
+      "input_seed": 20261046,
+      "checkpoint_sha256": "f11cdc5df1a58a9be3dd5973d540755c240dcb1c67246627de8aaf8229e0b7e6",
+      "exported_weights_sha256": "06201135c2255dcc8b2ee988f9ce128ce6fe2e6eb35f3fd142b3a4db415fec96",
+      "state_tensors": 65,
+      "all_state_tensors_equal": true,
+      "prediction_max_absolute_difference": 0.0
+    }
+  ]
+}

study-records/checks/final-export-reloads/fixed_batch.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:050793bd048da810ce6135f14dd5d23d4328f0b3488d8b7ecbb18a80ec123d43
+size 2021512

study-records/checks/final-export-reloads/run.log

@@ -0,0 +1,32 @@
+2026-10-17T19:13:25Z INFO {"event": "export_reload_verified", "source": "runs/checks/cifar10-metric-smoke", "export": "exports/20261017T191030.039367Z/cifar10/checks/cifar10-metric-smoke", "tensor_prefix": "run_00", "input_seed": 20261016, "checkpoint_sha256": "8e8cafa8870ab0ff9c618a39ba1f368be3966ff67459009f4ca195f437cd1aec", "exported_weights_sha256": "c6ef33c9235e17feb0c2656fcc29364fa02d5dd9dabf27b05c314c01420090d9", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:25Z INFO {"event": "export_reload_verified", "source": "runs/checks/mnist-smoke", "export": "exports/20261017T191030.039367Z/mnist/checks/mnist-smoke", "tensor_prefix": "run_01", "input_seed": 20261017, "checkpoint_sha256": "8f38d4c79f09a651583cc67b6596ab438a75269ad8fcb6a89db4135257a2a9db", "exported_weights_sha256": "89560b05f7e7318c1e71d7f8436e6458e469d07b9544ab602c756368d1482404", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:26Z INFO {"event": "export_reload_verified", "source": "runs/study/ablations/cifar10-no_coupling-s101", "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_coupling-s101", "tensor_prefix": "run_02", "input_seed": 20261018, "checkpoint_sha256": "3ec78d32218e855970086aafa4e6930fbfe4204b39228bfb133280ba95086f0d", "exported_weights_sha256": "90e80072c9a90978f8f2a30d80412ba0fb7976a049d69dd09c97210c746a5313", "state_tensors": 61, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:26Z INFO {"event": "export_reload_verified", "source": "runs/study/ablations/cifar10-no_coupling-s102", "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_coupling-s102", "tensor_prefix": "run_03", "input_seed": 20261019, "checkpoint_sha256": "0b3cbc2ac4115efd57aed785b957b2270de7683d46d26a20ed6b0908bd329ad2", "exported_weights_sha256": "7b4006f783489f1f3eb11be8f7441aa92b5df640cf7a775ad5498422fb2ac54d", "state_tensors": 61, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:26Z INFO {"event": "export_reload_verified", "source": "runs/study/ablations/cifar10-no_frequency-s101", "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_frequency-s101", "tensor_prefix": "run_04", "input_seed": 20261020, "checkpoint_sha256": "5de69e1afb911e81cba8030f52489023a73a601e54690a5babf6d0663d9d1b26", "exported_weights_sha256": "e414bd1148c47d7eb29817e4d725bdbffb03b37df5abbbc41e78924c149ca58a", "state_tensors": 61, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:27Z INFO {"event": "export_reload_verified", "source": "runs/study/ablations/cifar10-no_frequency-s102", "export": "exports/20261017T191030.039367Z/cifar10/study/ablations/cifar10-no_frequency-s102", "tensor_prefix": "run_05", "input_seed": 20261021, "checkpoint_sha256": "14fa697dfda45dc97daac9ff222051777ca99779e234e429f4c6f1ef4fc2f315", "exported_weights_sha256": "395b21f842524dfe11c55360d2093be3d562406a87a0f3af51b154887f33285a", "state_tensors": 61, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:27Z INFO {"event": "export_reload_verified", "source": "runs/study/checks/fixed-batch", "export": "exports/20261017T191030.039367Z/mnist/study/checks/fixed-batch", "tensor_prefix": "run_06", "input_seed": 20261022, "checkpoint_sha256": "5d36c5c4faf91992c6cba4b3bdfb01880b396f817a2ee17924878ada5fe36de0", "exported_weights_sha256": "dbc1290adf03cf30958883e06410e5867c99f943e25bdd637684610e552a369f", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:27Z INFO {"event": "export_reload_verified", "source": "runs/study/controls/untrained-conv", "export": "exports/20261017T191030.039367Z/cifar10/study/controls/untrained-conv", "tensor_prefix": "run_07", "input_seed": 20261023, "checkpoint_sha256": "c3ef26854d2b6b534035ebf5993c7814172283e3a3a7febf89f703798283e6e9", "exported_weights_sha256": "a09c911b1517c717c7527fa511e42da3abe44ad546b1ac8c47726cf2dc96af82", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:27Z INFO {"event": "export_reload_verified", "source": "runs/study/controls/untrained-dit", "export": "exports/20261017T191030.039367Z/cifar10/study/controls/untrained-dit", "tensor_prefix": "run_08", "input_seed": 20261024, "checkpoint_sha256": "f88d4f36bb900a4e46ba2852061a0c8c8d1dcb0f465a8def494aee59c03793ca", "exported_weights_sha256": "e0e8e1d0ed79b23b982ae432972278986797979ab46c66c70acb115217b00c18", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:27Z INFO {"event": "export_reload_verified", "source": "runs/study/controls/untrained-onn", "export": "exports/20261017T191030.039367Z/cifar10/study/controls/untrained-onn", "tensor_prefix": "run_09", "input_seed": 20261025, "checkpoint_sha256": "cd7618119dc6162268c54e9a8b4484d41d52ca10dbdaf76798f2a2afbce56a04", "exported_weights_sha256": "c6ef33c9235e17feb0c2656fcc29364fa02d5dd9dabf27b05c314c01420090d9", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:28Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-conv-s101", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s101", "tensor_prefix": "run_10", "input_seed": 20261026, "checkpoint_sha256": "d283104b77bb38b474d78f48ee64bf89e90cd8ecb2fc32d2cced963261e28cbd", "exported_weights_sha256": "d20f3947bcf79794427f8215c597a57c4b5eef3b308ccaccf1960beb86b4188e", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:28Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-conv-s102", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s102", "tensor_prefix": "run_11", "input_seed": 20261027, "checkpoint_sha256": "ebb6eafd617b874a4fd559f32996118fdc64418f9206aa38faeb0645c4545b82", "exported_weights_sha256": "7c81f89410f40e356ad05be8cb60016c65181208541c303d8ba1665778d1ce3c", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:29Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-conv-s103", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s103", "tensor_prefix": "run_12", "input_seed": 20261028, "checkpoint_sha256": "4cb949dcbdf984ec4658182017a6f2be36da6eb9eb8d6891f6ebaf7119156a43", "exported_weights_sha256": "66eb47d34559e0a1127012efc6888a22546eb0b9e9b94e7e28821dbd6c404441", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:29Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-conv-s104", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s104", "tensor_prefix": "run_13", "input_seed": 20261029, "checkpoint_sha256": "384185eaffe897a164f9680f1506f8de0136c16b0d60ff9c491226f04488f574", "exported_weights_sha256": "c896d46723b1f6e52eb8b6ad1e8fbdfc9c889784b5e114fe73606f538ed3d1ac", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:29Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-conv-s105", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s105", "tensor_prefix": "run_14", "input_seed": 20261030, "checkpoint_sha256": "1f90d6fd0182341298cc1812aca672144633c8ea7c6ab57816bf9949e597352f", "exported_weights_sha256": "d872edbf5e8bf7c11f03b6637adfc2762a009a6b0ea659f48e595e4d853cd5b7", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:30Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-conv-s106", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-conv-s106", "tensor_prefix": "run_15", "input_seed": 20261031, "checkpoint_sha256": "fc27f7e6924da1c7426ca800f5916d5836ea6f8285d7fc926ba4ab981c156194", "exported_weights_sha256": "2057aaaf7d16b4220dfdc8bc9d8ef32a6e91388223e30e65c2d72252f4562cd4", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:30Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-dit-s101", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s101", "tensor_prefix": "run_16", "input_seed": 20261032, "checkpoint_sha256": "030d935262170666e1c5f3da7020026a573aa3de03b9f86a9c67f036afbe0dce", "exported_weights_sha256": "cc227d4e18a7e1d94a2a9e0d34e809f86a8798e0c8e283af84968e529a1c8a67", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:31Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-dit-s102", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s102", "tensor_prefix": "run_17", "input_seed": 20261033, "checkpoint_sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15", "exported_weights_sha256": "2e584a96062f1a454dce28bb6322e17731df8999e32045b8ed7f9fe8fb4453cc", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:31Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-dit-s103", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s103", "tensor_prefix": "run_18", "input_seed": 20261034, "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8", "exported_weights_sha256": "75bca1cf3738374c420f34e36b0679e6369cf3fa86343aa762f0ff85bd6df64b", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:31Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-dit-s104", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s104", "tensor_prefix": "run_19", "input_seed": 20261035, "checkpoint_sha256": "931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a", "exported_weights_sha256": "3aba95b3af00dd2cad93f206c5af6d73f07431b491ed2e358a2feb8c8a1991c1", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:31Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-dit-s105", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s105", "tensor_prefix": "run_20", "input_seed": 20261036, "checkpoint_sha256": "eec459e41060cbff6d07971d0c2fd363ccc1f525e23cc72c6b5c872e2cf832ac", "exported_weights_sha256": "38b99f5c030c04a12c83c9b41ee55c16a4cde20c557e29b274950dbeb6d4a292", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:31Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-dit-s106", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-dit-s106", "tensor_prefix": "run_21", "input_seed": 20261037, "checkpoint_sha256": "c05ea7f6e8db40d1fee8137f96251a02fd2c1ee199d91e2214aa46a17c51cc5e", "exported_weights_sha256": "f4b8c7eddf76c637e828ccb902517b2f2efbfb2e269986f81bd5ac965e541f3e", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:31Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-onn-s101", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s101", "tensor_prefix": "run_22", "input_seed": 20261038, "checkpoint_sha256": "e3eb0aabeaf1f372fbcb2f23bf94625b9fd2a1424504818a02b5d37a4522d6b1", "exported_weights_sha256": "bb7094d51557bb6d7ccfdbaebd393c3cc33bbfcb48963aba33e58c48af187f99", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:32Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-onn-s102", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s102", "tensor_prefix": "run_23", "input_seed": 20261039, "checkpoint_sha256": "e5cf67558e832212ded9b04225e34091cd62715d9df0773c5d22685da64e8366", "exported_weights_sha256": "06a34f3c2e2dae9947112142592c6747376ba44959be8efb0ba37bc3687276cd", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:32Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-onn-s103", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s103", "tensor_prefix": "run_24", "input_seed": 20261040, "checkpoint_sha256": "51090d973909f69682a6f2c126cf1377fbda43f249646733df76a4256e4566a1", "exported_weights_sha256": "93b25b900dff3df0f60701b0456a0b45c89eba9e2da7f539f2c290f575f5e4e7", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:32Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-onn-s104", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s104", "tensor_prefix": "run_25", "input_seed": 20261041, "checkpoint_sha256": "a0262716020a4118b0f750cf092274f8748ae86cb5ea23b3a700d75bc86e6322", "exported_weights_sha256": "b306ed41c2b415cf5be073c0bbf0c4fcd4fdfee714ac4eab05aff3061021f620", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:32Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-onn-s105", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s105", "tensor_prefix": "run_26", "input_seed": 20261042, "checkpoint_sha256": "6ae577c27cfe5120f2824baff526173d912511a47c56665da170c0b5a227251a", "exported_weights_sha256": "12d9c6bfed79fd4a812af0b18f053486c5852171826f6e0eb453c4868fa3bd00", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:33Z INFO {"event": "export_reload_verified", "source": "runs/study/main/cifar10-onn-s106", "export": "exports/20261017T191030.039367Z/cifar10/study/main/cifar10-onn-s106", "tensor_prefix": "run_27", "input_seed": 20261043, "checkpoint_sha256": "e9595ec0d0fe7e945b48301e630d059d2077a641c3da27f0b6c205899f4da0b8", "exported_weights_sha256": "c8b04e345990d920092c9b209471de71092020c363ac4004d631fddb476b3984", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:33Z INFO {"event": "export_reload_verified", "source": "runs/study/pilot/mnist-conv-s101", "export": "exports/20261017T191030.039367Z/mnist/study/pilot/mnist-conv-s101", "tensor_prefix": "run_28", "input_seed": 20261044, "checkpoint_sha256": "3fffc15466b2561ab3b0ebe9ad66272c4c4e3fe99e0f13602910d1f7da80b651", "exported_weights_sha256": "bffd1bfd5e174490442d518b745e2f1da3416b0e2594b714f47a5dd9f16c1e31", "state_tensors": 69, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:33Z INFO {"event": "export_reload_verified", "source": "runs/study/pilot/mnist-dit-s101", "export": "exports/20261017T191030.039367Z/mnist/study/pilot/mnist-dit-s101", "tensor_prefix": "run_29", "input_seed": 20261045, "checkpoint_sha256": "628a9d5c15326d0a89bb13ecab677c36bbe22b0de447d3e9bba981f82e164ee0", "exported_weights_sha256": "a72223e84aa002fcb68459de860540f70499c1964584c2b1b37b437d2e327366", "state_tensors": 52, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:33Z INFO {"event": "export_reload_verified", "source": "runs/study/pilot/mnist-onn-s101", "export": "exports/20261017T191030.039367Z/mnist/study/pilot/mnist-onn-s101", "tensor_prefix": "run_30", "input_seed": 20261046, "checkpoint_sha256": "f11cdc5df1a58a9be3dd5973d540755c240dcb1c67246627de8aaf8229e0b7e6", "exported_weights_sha256": "06201135c2255dcc8b2ee988f9ce128ce6fe2e6eb35f3fd142b3a4db415fec96", "state_tensors": 65, "all_state_tensors_equal": true, "prediction_max_absolute_difference": 0.0}
+2026-10-17T19:13:33Z INFO {"event": "all_export_reloads_complete", "compared_checkpoints": 31, "prediction_max_absolute_difference": 0.0}

study-records/checks/inception-cache-failure/failure.json

@@ -0,0 +1,8 @@
+{
+  "status": "failed",
+  "kind": "CPU evaluator-cache preparation",
+  "error": "ModuleNotFoundError: No module named 'ofm.terminal'",
+  "cause": "A read/import command overlapped the transient checkout performed by gh pr merge --delete-branch.",
+  "resolution": "Waited for the merge and checkout to finish, then reran the unchanged cache preparation successfully. Future merges during active runs retain the current checkout.",
+  "gpu_seconds": 0
+}

study-records/checks/progress-monitor-failure/failure.json

@@ -0,0 +1,10 @@
+{
+  "kind": "read-only-progress-monitor",
+  "status": "failed",
+  "timestamp": "2026-10-17T19:09:58.889005+00:00",
+  "error": "IndexError: list index out of range",
+  "traceback": "File <stdin>, line 20, in <module>",
+  "context": "The separate read-only monitor exited during the final real-versus-real control, whose newly created run.log had no progress entries. The study process continued independently.",
+  "impact": "No training, sampling, metric, checkpoint, or budget state was modified by the monitor.",
+  "resolution": "Use direct artifact/process checks that tolerate an empty log; no scientific run is restarted."
+}

study-records/study/controls/real-versus-real/run.log

@@ -0,0 +1,2 @@
+2026-10-17T19:09:57Z INFO {"event": "real_control_complete", "metrics": {"frechet_inception_distance": 10.272234758028844, "kernel_inception_distance_mean": -2.55584716796875e-05, "kernel_inception_distance_std": 0.00010886578191701542}, "repeat_max_absolute_difference": 0.0}
+2026-10-17T19:09:57Z INFO {"event": "study_phase_complete", "phase": "evaluate", "root": "runs/study"}

study-records/study/main/cifar10-onn-s101/evaluations/identity/control.json

@@ -0,0 +1,3 @@
+{
+  "identity_exact_equal": true
+}

study-records/study/main/cifar10-onn-s101/evaluations/identity/run.log

@@ -0,0 +1,11 @@
+2026-10-17T18:50:17Z INFO {"event": "evaluation_start", "dataset": "cifar10", "family": "onn", "training_seed": 101, "split": "test", "intervention": "identity", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "e3eb0aabeaf1f372fbcb2f23bf94625b9fd2a1424504818a02b5d37a4522d6b1", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "b49e52bb12834bd0f151a4f45a36462182130a8d", "primary_eligible": false, "device": "cuda", "blas_threads": 4}
+2026-10-17T18:50:19Z INFO {"event": "sampling_progress", "completed": 32, "total": 5000, "nfe": 32}
+2026-10-17T18:50:38Z INFO {"event": "sampling_progress", "completed": 672, "total": 5000, "nfe": 32}
+2026-10-17T18:50:52Z INFO {"event": "sampling_progress", "completed": 1312, "total": 5000, "nfe": 32}
+2026-10-17T18:51:10Z INFO {"event": "sampling_progress", "completed": 1952, "total": 5000, "nfe": 32}
+2026-10-17T18:51:25Z INFO {"event": "sampling_progress", "completed": 2592, "total": 5000, "nfe": 32}
+2026-10-17T18:51:44Z INFO {"event": "sampling_progress", "completed": 3232, "total": 5000, "nfe": 32}
+2026-10-17T18:51:59Z INFO {"event": "sampling_progress", "completed": 3872, "total": 5000, "nfe": 32}
+2026-10-17T18:52:18Z INFO {"event": "sampling_progress", "completed": 4512, "total": 5000, "nfe": 32}
+2026-10-17T18:52:29Z INFO {"event": "sampling_complete", "directory": "runs/study/main/cifar10-onn-s101/evaluations/identity", "count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 116.22197621400119, "seconds_per_image": 0.02324439524280024, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.012888411059975624, "peak_memory_bytes": 60721664}
+2026-10-17T18:52:53Z INFO {"event": "evaluation_complete", "dataset": "cifar10", "family": "onn", "training_seed": 101, "split": "test", "intervention": "identity", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "e3eb0aabeaf1f372fbcb2f23bf94625b9fd2a1424504818a02b5d37a4522d6b1", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "b49e52bb12834bd0f151a4f45a36462182130a8d", "primary_eligible": false, "device": "cuda", "blas_threads": 4, "status": "completed", "sampling": {"count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 116.22197621400119, "seconds_per_image": 0.02324439524280024, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.012888411059975624, "peak_memory_bytes": 60721664}, "metrics": {"frechet_inception_distance": 176.70453192569718, "kernel_inception_distance_mean": 0.18851056098937988, "kernel_inception_distance_std": 0.0027402690946605403}, "validation_velocity_mse": 0.21588426921516657, "metric_seconds": 19.394039164006244}

study-records/study/main/cifar10-onn-s102/evaluations/identity/control.json

@@ -0,0 +1,3 @@
+{
+  "identity_exact_equal": true
+}

study-records/study/main/cifar10-onn-s102/evaluations/identity/run.log

@@ -0,0 +1,11 @@
+2026-10-17T18:55:43Z INFO {"event": "evaluation_start", "dataset": "cifar10", "family": "onn", "training_seed": 102, "split": "test", "intervention": "identity", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "e5cf67558e832212ded9b04225e34091cd62715d9df0773c5d22685da64e8366", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "b49e52bb12834bd0f151a4f45a36462182130a8d", "primary_eligible": false, "device": "cuda", "blas_threads": 4}
+2026-10-17T18:55:49Z INFO {"event": "sampling_progress", "completed": 32, "total": 5000, "nfe": 32}
+2026-10-17T18:56:03Z INFO {"event": "sampling_progress", "completed": 672, "total": 5000, "nfe": 32}
+2026-10-17T18:56:24Z INFO {"event": "sampling_progress", "completed": 1312, "total": 5000, "nfe": 32}
+2026-10-17T18:56:41Z INFO {"event": "sampling_progress", "completed": 1952, "total": 5000, "nfe": 32}
+2026-10-17T18:57:02Z INFO {"event": "sampling_progress", "completed": 2592, "total": 5000, "nfe": 32}
+2026-10-17T18:57:22Z INFO {"event": "sampling_progress", "completed": 3232, "total": 5000, "nfe": 32}
+2026-10-17T18:57:43Z INFO {"event": "sampling_progress", "completed": 3872, "total": 5000, "nfe": 32}
+2026-10-17T18:57:58Z INFO {"event": "sampling_progress", "completed": 4512, "total": 5000, "nfe": 32}
+2026-10-17T18:58:16Z INFO {"event": "sampling_complete", "directory": "runs/study/main/cifar10-onn-s102/evaluations/identity", "count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 133.63555448500847, "seconds_per_image": 0.026727110897001694, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.013204231858253479, "peak_memory_bytes": 60721664}
+2026-10-17T18:58:37Z INFO {"event": "evaluation_complete", "dataset": "cifar10", "family": "onn", "training_seed": 102, "split": "test", "intervention": "identity", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "e5cf67558e832212ded9b04225e34091cd62715d9df0773c5d22685da64e8366", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "b49e52bb12834bd0f151a4f45a36462182130a8d", "primary_eligible": false, "device": "cuda", "blas_threads": 4, "status": "completed", "sampling": {"count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 133.63555448500847, "seconds_per_image": 0.026727110897001694, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.013204231858253479, "peak_memory_bytes": 60721664}, "metrics": {"frechet_inception_distance": 170.89254169365023, "kernel_inception_distance_mean": 0.17922114610671996, "kernel_inception_distance_std": 0.002754562183066269}, "validation_velocity_mse": 0.22063074633479118, "metric_seconds": 20.13504967300105}

study-records/study/pilot/mnist-onn-s101/cpu-public-check/control.json

@@ -0,0 +1,13 @@
+{
+  "kind": "public-checkpoint-reload",
+  "status": "completed",
+  "repository": "BurnyCoder/oscillatory-flow-matching-mnist",
+  "path": "study/pilot/mnist-onn-s101",
+  "checkpoint_sha256": "f11cdc5df1a58a9be3dd5973d540755c240dcb1c67246627de8aaf8229e0b7e6",
+  "public_weights_sha256": "06201135c2255dcc8b2ee988f9ce128ce6fe2e6eb35f3fd142b3a4db415fec96",
+  "prediction_max_absolute_difference": 0.0,
+  "device": "cpu",
+  "seed": 20261016,
+  "sample_count": 64,
+  "nfe": 32
+}

study-records/study/pilot/mnist-onn-s101/cpu-public-check/run.log

@@ -0,0 +1,3 @@
+2026-10-16T23:39:59Z INFO {"event": "sampling_progress", "completed": 32, "total": 64, "nfe": 32}
+2026-10-16T23:40:04Z INFO {"event": "sampling_complete", "directory": "runs/study/pilot/mnist-onn-s101/cpu-public-check", "count": 64, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 10.12635529399995, "seconds_per_image": 0.15822430146874922, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.35869738459587097, "peak_memory_bytes": null}
+2026-10-16T23:40:30Z INFO {"event": "public_checkpoint_verified", "kind": "public-checkpoint-reload", "status": "completed", "repository": "BurnyCoder/oscillatory-flow-matching-mnist", "path": "study/pilot/mnist-onn-s101", "checkpoint_sha256": "f11cdc5df1a58a9be3dd5973d540755c240dcb1c67246627de8aaf8229e0b7e6", "public_weights_sha256": "06201135c2255dcc8b2ee988f9ce128ce6fe2e6eb35f3fd142b3a4db415fec96", "prediction_max_absolute_difference": 0.0, "device": "cpu", "seed": 20261016, "sample_count": 64, "nfe": 32}

study/main/cifar10-dit-s102/evaluations/test-nfe32/run.log

@@ -0,0 +1,11 @@
+2026-10-17T17:56:25Z INFO {"event": "evaluation_start", "dataset": "cifar10", "family": "dit", "training_seed": 102, "split": "test", "intervention": "none", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "0afa17670bc7339fd28a2beea4396e47d30b1d9c", "primary_eligible": true, "device": "cuda", "blas_threads": 4}
+2026-10-17T17:56:26Z INFO {"event": "sampling_progress", "completed": 32, "total": 5000, "nfe": 32}
+2026-10-17T17:56:32Z INFO {"event": "sampling_progress", "completed": 672, "total": 5000, "nfe": 32}
+2026-10-17T17:56:38Z INFO {"event": "sampling_progress", "completed": 1312, "total": 5000, "nfe": 32}
+2026-10-17T17:56:48Z INFO {"event": "sampling_progress", "completed": 1952, "total": 5000, "nfe": 32}
+2026-10-17T17:56:54Z INFO {"event": "sampling_progress", "completed": 2592, "total": 5000, "nfe": 32}
+2026-10-17T17:57:00Z INFO {"event": "sampling_progress", "completed": 3232, "total": 5000, "nfe": 32}
+2026-10-17T17:57:06Z INFO {"event": "sampling_progress", "completed": 3872, "total": 5000, "nfe": 32}
+2026-10-17T17:57:13Z INFO {"event": "sampling_progress", "completed": 4512, "total": 5000, "nfe": 32}
+2026-10-17T17:57:22Z INFO {"event": "sampling_complete", "directory": "runs/study/main/cifar10-dit-s102/evaluations/test-nfe32", "count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 48.89814832000411, "seconds_per_image": 0.009779629664000823, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.015701953321695328, "peak_memory_bytes": 58955776}
+2026-10-17T17:57:41Z INFO {"event": "evaluation_complete", "dataset": "cifar10", "family": "dit", "training_seed": 102, "split": "test", "intervention": "none", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "0afa17670bc7339fd28a2beea4396e47d30b1d9c", "primary_eligible": true, "device": "cuda", "blas_threads": 4, "status": "completed", "sampling": {"count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 48.89814832000411, "seconds_per_image": 0.009779629664000823, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.015701953321695328, "peak_memory_bytes": 58955776}, "metrics": {"frechet_inception_distance": 103.12542282658637, "kernel_inception_distance_mean": 0.08995173454284668, "kernel_inception_distance_std": 0.00169647956062002}, "validation_velocity_mse": 0.1970291268080473, "metric_seconds": 18.674224670001422}

study/main/cifar10-dit-s102/evaluations/test-nfe32/samples.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3661fb6283e36c43925a06b4f09aeb9c6b6131ebfd298b0e5bc3953a28493456
+size 61440088

study/main/cifar10-dit-s102/evaluations/test-nfe32/sampling.json

@@ -0,0 +1,12 @@
+{
+  "count": 5000,
+  "seed": 20261016,
+  "nfe": 32,
+  "batch_size": 32,
+  "seconds": 48.89814832000411,
+  "seconds_per_image": 0.009779629664000823,
+  "precision": "float32",
+  "timing_excludes": "model load and warmup",
+  "clipped_fraction": 0.015701953321695328,
+  "peak_memory_bytes": 58955776
+}

study/main/cifar10-dit-s102/manifest.json

@@ -0,0 +1,62 @@
+{
+  "README.md": {
+    "sha256": "8c5fd64a57f4309fd3032502f60e20781104069a3ba91c0fa4aaf072a597d25a",
+    "bytes": 2917
+  },
+  "checkpoint.pt": {
+    "sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15",
+    "bytes": 21305024
+  },
+  "config.json": {
+    "sha256": "9d667fda7aa2bf3596e2646565131838edd3c326812d3499b326bc7d4c584b93",
+    "bytes": 606
+  },
+  "data_manifest.json": {
+    "sha256": "d9d38bfe55c2a03eada50a4ef09f3152c987d290330edf4ed3dd452c50614350",
+    "bytes": 539473
+  },
+  "environment.json": {
+    "sha256": "01c239a46dd1e6a0fce629335d1332fa6144e2ca4e8778b2a33d817c7938b82f",
+    "bytes": 3695
+  },
+  "evaluations/test-nfe32/evaluation.json": {
+    "sha256": "b772ffe8254f2c0e01aadee4e8258a3d41e7f7bf34cbda901e8c4beacae540fc",
+    "bytes": 1233
+  },
+  "evaluations/test-nfe32/evaluation_config.json": {
+    "sha256": "f74b2655d85b04491f5043a7e3ae29ed520bc7854174696282c196ef9bc747bc",
+    "bytes": 593
+  },
+  "evaluations/test-nfe32/grid.png": {
+    "sha256": "4713e05130d45fd1e9f234d53ffafb28035d9303b8513a5b94f28bae47678762",
+    "bytes": 153977
+  },
+  "evaluations/test-nfe32/run.log": {
+    "sha256": "89bf2c8939ac08e67851b9d8a2a4e5ca136cfb8db08b7875b88bedcf6ca5c1ab",
+    "bytes": 3011
+  },
+  "evaluations/test-nfe32/samples.safetensors": {
+    "sha256": "3661fb6283e36c43925a06b4f09aeb9c6b6131ebfd298b0e5bc3953a28493456",
+    "bytes": 61440088
+  },
+  "evaluations/test-nfe32/sampling.json": {
+    "sha256": "81ff920306444775a76e0f43fcc0c76a9a52ddc0972e11485d0d7ff6ff9f178b",
+    "bytes": 299
+  },
+  "model.safetensors": {
+    "sha256": "2e584a96062f1a454dce28bb6322e17731df8999e32045b8ed7f9fe8fb4453cc",
+    "bytes": 5233488
+  },
+  "planned-config.json": {
+    "sha256": "c63673cd38edc8bc847b9d610641ff92f517c53c4bd987bf21257741bf5735c1",
+    "bytes": 604
+  },
+  "run.log": {
+    "sha256": "44d3011e4b4fd503d7710a338d9929c54dacd152945e9f1d01b993e247bd33c8",
+    "bytes": 81227
+  },
+  "training.json": {
+    "sha256": "fc1aba71169ba453885e940d126fa0e2b387878bdc3893cecd02f570ac8f738d",
+    "bytes": 543
+  }
+}

study/main/cifar10-dit-s102/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e584a96062f1a454dce28bb6322e17731df8999e32045b8ed7f9fe8fb4453cc
+size 5233488

study/main/cifar10-dit-s102/planned-config.json

@@ -0,0 +1,32 @@
+{
+  "model": {
+    "family": "dit",
+    "channels": 3,
+    "size": 32,
+    "width": 128,
+    "depth": 4,
+    "patch": 4,
+    "rotating_dim": 4,
+    "inner_steps": 4,
+    "step_size": 0.1,
+    "ffn_hidden": 0
+  },
+  "dataset": "cifar10",
+  "seed": 102,
+  "device": "cuda",
+  "batch_size": 128,
+  "microbatch": 32,
+  "learning_rate": 0.0003,
+  "weight_decay": 0.01,
+  "clip_grad": 1.0,
+  "ema_decay": 0.999,
+  "minutes": 45,
+  "max_steps": 0,
+  "validation_size": 5000,
+  "data_root": "data",
+  "runs_root": "runs",
+  "budget_hours": 24.0,
+  "log_every": 50,
+  "checkpoint_seconds": 300.0,
+  "amp": true
+}

study/main/cifar10-dit-s102/run.log

@@ -0,0 +1,350 @@
+2026-10-17T03:46:01Z INFO {"event": "training_configuration", "model": {"family": "dit", "channels": 3, "size": 32, "width": 128, "depth": 4, "patch": 4, "rotating_dim": 4, "inner_steps": 4, "step_size": 0.1, "ffn_hidden": 576}, "dataset": "cifar10", "seed": 102, "device": "cuda", "batch_size": 128, "microbatch": 32, "learning_rate": 0.0003, "weight_decay": 0.01, "clip_grad": 1.0, "ema_decay": 0.999, "minutes": 45, "max_steps": 0, "validation_size": 5000, "data_root": "data", "runs_root": "runs", "budget_hours": 24.0, "log_every": 50, "checkpoint_seconds": 300.0, "amp": true, "resume": false, "source_revisions": ["c749856d1c9fe07d6ce2f39d01de78f664aea9a8"]}
+2026-10-17T03:46:03Z INFO {"event": "training_start", "parameters": 1298864, "step": 0, "validation_mse": 1.7011075615882874}
+2026-10-17T03:46:04Z INFO {"event": "training_progress", "step": 1, "epoch": 0, "loss": 1.7102091312408447, "grad_norm": 1.3451812267303467, "train_seconds": 1.0475090869877022, "remaining_budget_seconds": 2757.2743219849945}
+2026-10-17T03:46:12Z INFO {"event": "training_progress", "step": 50, "epoch": 0, "loss": 1.2182015369133072, "grad_norm": 0.25371554493904114, "train_seconds": 8.5650643830013, "remaining_budget_seconds": 2749.7567655400053}
+2026-10-17T03:46:23Z INFO {"event": "training_progress", "step": 100, "epoch": 0, "loss": 0.60513593390584, "grad_norm": 0.43706008791923523, "train_seconds": 16.68258377398888, "remaining_budget_seconds": 2741.639252378998}
+2026-10-17T03:46:31Z INFO {"event": "training_progress", "step": 150, "epoch": 0, "loss": 0.4186970530450344, "grad_norm": 0.5762000679969788, "train_seconds": 24.71570428898849, "remaining_budget_seconds": 2733.606130700995}
+2026-10-17T03:46:38Z INFO {"event": "training_progress", "step": 200, "epoch": 0, "loss": 0.36617414459586145, "grad_norm": 0.7147449254989624, "train_seconds": 32.52099970600102, "remaining_budget_seconds": 2725.800834834008}
+2026-10-17T03:46:46Z INFO {"event": "training_progress", "step": 250, "epoch": 0, "loss": 0.3353129304945469, "grad_norm": 0.6435403823852539, "train_seconds": 40.492871151989675, "remaining_budget_seconds": 2717.8289656849956}
+2026-10-17T03:46:57Z INFO {"event": "training_progress", "step": 300, "epoch": 0, "loss": 0.3171289286017418, "grad_norm": 0.8954771161079407, "train_seconds": 48.242379249000805, "remaining_budget_seconds": 2710.079453686005}
+2026-10-17T03:47:05Z INFO {"event": "training_progress", "step": 350, "epoch": 0, "loss": 0.3121790202707052, "grad_norm": 0.7943933606147766, "train_seconds": 56.29138593899552, "remaining_budget_seconds": 2702.030450862003}
+2026-10-17T03:47:13Z INFO {"event": "training_progress", "step": 400, "epoch": 1, "loss": 0.3015691643953323, "grad_norm": 0.4489187002182007, "train_seconds": 63.885808261999045, "remaining_budget_seconds": 2694.4360269470053}
+2026-10-17T03:47:21Z INFO {"event": "training_progress", "step": 450, "epoch": 1, "loss": 0.2923025009781122, "grad_norm": 0.8282820582389832, "train_seconds": 71.74789295998926, "remaining_budget_seconds": 2686.5739401580067}
+2026-10-17T03:47:31Z INFO {"event": "training_progress", "step": 500, "epoch": 1, "loss": 0.2864245757460594, "grad_norm": 0.9439363479614258, "train_seconds": 79.67472876299871, "remaining_budget_seconds": 2678.6471066119993}
+2026-10-17T03:47:40Z INFO {"event": "training_progress", "step": 550, "epoch": 1, "loss": 0.28568645924329755, "grad_norm": 0.7066501975059509, "train_seconds": 87.76341227199009, "remaining_budget_seconds": 2670.5584223640035}
+2026-10-17T03:47:47Z INFO {"event": "training_progress", "step": 600, "epoch": 1, "loss": 0.2811491507291794, "grad_norm": 0.7494161128997803, "train_seconds": 95.70134381999378, "remaining_budget_seconds": 2662.6204925689963}
+2026-10-17T03:47:55Z INFO {"event": "training_progress", "step": 650, "epoch": 1, "loss": 0.2825243843346834, "grad_norm": 0.5063631534576416, "train_seconds": 103.48544407899317, "remaining_budget_seconds": 2654.8363908570027}
+2026-10-17T03:48:06Z INFO {"event": "training_progress", "step": 700, "epoch": 1, "loss": 0.2709879020601511, "grad_norm": 0.40633171796798706, "train_seconds": 111.22072044599918, "remaining_budget_seconds": 2647.101116538004}
+2026-10-17T03:48:14Z INFO {"event": "training_progress", "step": 750, "epoch": 2, "loss": 0.2684937585890293, "grad_norm": 0.3910692036151886, "train_seconds": 118.96756210899912, "remaining_budget_seconds": 2639.354274440004}
+2026-10-17T03:48:22Z INFO {"event": "training_progress", "step": 800, "epoch": 2, "loss": 0.26776341468095777, "grad_norm": 0.4853197932243347, "train_seconds": 126.93491645199538, "remaining_budget_seconds": 2631.386919798999}
+2026-10-17T03:48:29Z INFO {"event": "training_progress", "step": 850, "epoch": 2, "loss": 0.26979686945676806, "grad_norm": 0.6423327922821045, "train_seconds": 134.65833485798794, "remaining_budget_seconds": 2623.6635013990017}
+2026-10-17T03:48:40Z INFO {"event": "training_progress", "step": 900, "epoch": 2, "loss": 0.2637389501929283, "grad_norm": 0.5961043238639832, "train_seconds": 142.7314313579991, "remaining_budget_seconds": 2615.5904048530065}
+2026-10-17T03:48:48Z INFO {"event": "training_progress", "step": 950, "epoch": 2, "loss": 0.2647680766135454, "grad_norm": 0.6882957816123962, "train_seconds": 150.68219607300125, "remaining_budget_seconds": 2607.6396385349944}
+2026-10-17T03:48:56Z INFO {"event": "training_progress", "step": 1000, "epoch": 2, "loss": 0.2626547655463219, "grad_norm": 0.7978100776672363, "train_seconds": 158.53024411699153, "remaining_budget_seconds": 2599.791589344997}
+2026-10-17T03:49:04Z INFO {"event": "training_progress", "step": 1050, "epoch": 2, "loss": 0.25881574407219887, "grad_norm": 0.6521762013435364, "train_seconds": 166.25238407099096, "remaining_budget_seconds": 2592.0694504239946}
+2026-10-17T03:49:12Z INFO {"event": "training_progress", "step": 1100, "epoch": 3, "loss": 0.25782738529145716, "grad_norm": 0.6857454776763916, "train_seconds": 173.9919749119872, "remaining_budget_seconds": 2584.3298598840047}
+2026-10-17T03:49:22Z INFO {"event": "training_progress", "step": 1150, "epoch": 3, "loss": 0.25558778896927836, "grad_norm": 0.7858656048774719, "train_seconds": 181.96853535500122, "remaining_budget_seconds": 2576.353301370007}
+2026-10-17T03:49:30Z INFO {"event": "training_progress", "step": 1200, "epoch": 3, "loss": 0.256722389832139, "grad_norm": 0.5612936615943909, "train_seconds": 189.6345383089938, "remaining_budget_seconds": 2568.6872944479983}
+2026-10-17T03:49:38Z INFO {"event": "training_progress", "step": 1250, "epoch": 3, "loss": 0.25521011382341385, "grad_norm": 0.7020533680915833, "train_seconds": 197.66618099498737, "remaining_budget_seconds": 2560.6556526009954}
+2026-10-17T03:49:46Z INFO {"event": "training_progress", "step": 1300, "epoch": 3, "loss": 0.25028772130608556, "grad_norm": 0.5511124730110168, "train_seconds": 205.37090428698866, "remaining_budget_seconds": 2552.9509310150024}
+2026-10-17T03:49:57Z INFO {"event": "training_progress", "step": 1350, "epoch": 3, "loss": 0.24993446037173273, "grad_norm": 0.3987192213535309, "train_seconds": 213.38389921099588, "remaining_budget_seconds": 2544.937928927}
+2026-10-17T03:50:05Z INFO {"event": "training_progress", "step": 1400, "epoch": 3, "loss": 0.24921218171715737, "grad_norm": 0.7277663350105286, "train_seconds": 221.31062315798772, "remaining_budget_seconds": 2537.011210951998}
+2026-10-17T03:50:13Z INFO {"event": "training_progress", "step": 1450, "epoch": 4, "loss": 0.2531578131020069, "grad_norm": 0.8757371306419373, "train_seconds": 229.28008292899176, "remaining_budget_seconds": 2529.041752721998}
+2026-10-17T03:50:21Z INFO {"event": "training_progress", "step": 1500, "epoch": 4, "loss": 0.2458353666216135, "grad_norm": 0.573822557926178, "train_seconds": 237.18998252598976, "remaining_budget_seconds": 2521.1318524259987}
+2026-10-17T03:50:31Z INFO {"event": "training_progress", "step": 1550, "epoch": 4, "loss": 0.24778997644782066, "grad_norm": 0.6197834610939026, "train_seconds": 244.91147105798882, "remaining_budget_seconds": 2513.410364519994}
+2026-10-17T03:50:39Z INFO {"event": "training_progress", "step": 1600, "epoch": 4, "loss": 0.24746915705502034, "grad_norm": 0.5454726815223694, "train_seconds": 252.68730956799118, "remaining_budget_seconds": 2505.6345255619963}
+2026-10-17T03:50:47Z INFO {"event": "training_progress", "step": 1650, "epoch": 4, "loss": 0.24901846751570703, "grad_norm": 0.49615347385406494, "train_seconds": 260.4547721229901, "remaining_budget_seconds": 2497.867034089999}
+2026-10-17T03:50:55Z INFO {"event": "training_progress", "step": 1700, "epoch": 4, "loss": 0.24455307759344577, "grad_norm": 0.5862202048301697, "train_seconds": 268.48740352298773, "remaining_budget_seconds": 2489.8344331170083}
+2026-10-17T03:51:06Z INFO {"event": "training_progress", "step": 1750, "epoch": 4, "loss": 0.24331811271607875, "grad_norm": 0.5813148021697998, "train_seconds": 276.39509287998953, "remaining_budget_seconds": 2481.9267400730023}
+2026-10-17T03:51:14Z INFO {"event": "training_progress", "step": 1800, "epoch": 5, "loss": 0.24527225084602833, "grad_norm": 0.4104244112968445, "train_seconds": 284.0255044079968, "remaining_budget_seconds": 2474.2963321370044}
+2026-10-17T03:51:21Z INFO {"event": "training_progress", "step": 1850, "epoch": 5, "loss": 0.24289261177182198, "grad_norm": 0.5602385997772217, "train_seconds": 291.7260743849911, "remaining_budget_seconds": 2466.5957608180033}
+2026-10-17T03:51:29Z INFO {"event": "training_progress", "step": 1900, "epoch": 5, "loss": 0.24364950336515903, "grad_norm": 0.4701775312423706, "train_seconds": 299.60994632999063, "remaining_budget_seconds": 2458.7118889709964}
+2026-10-17T03:51:40Z INFO {"event": "training_progress", "step": 1950, "epoch": 5, "loss": 0.23947740778326987, "grad_norm": 0.42057105898857117, "train_seconds": 307.7549246409908, "remaining_budget_seconds": 2450.5669101480016}
+2026-10-17T03:51:48Z INFO {"event": "training_progress", "step": 2000, "epoch": 5, "loss": 0.2368535566329956, "grad_norm": 0.3872612416744232, "train_seconds": 315.3748166029982, "remaining_budget_seconds": 2442.9470182829973}
+2026-10-17T03:51:56Z INFO {"event": "training_progress", "step": 2050, "epoch": 5, "loss": 0.24142697796225548, "grad_norm": 0.37970659136772156, "train_seconds": 323.1375227690005, "remaining_budget_seconds": 2435.184311801}
+2026-10-17T03:52:03Z INFO {"event": "training_progress", "step": 2100, "epoch": 5, "loss": 0.23348697260022164, "grad_norm": 0.44649308919906616, "train_seconds": 330.8412108569901, "remaining_budget_seconds": 2427.480624984004}
+2026-10-17T03:52:14Z INFO {"event": "training_progress", "step": 2150, "epoch": 6, "loss": 0.24119653016328813, "grad_norm": 0.5410599112510681, "train_seconds": 338.48303648398723, "remaining_budget_seconds": 2419.8388005089946}
+2026-10-17T03:52:22Z INFO {"event": "training_progress", "step": 2200, "epoch": 6, "loss": 0.23356275498867035, "grad_norm": 0.5065106749534607, "train_seconds": 346.3844115129905, "remaining_budget_seconds": 2411.937425283002}
+2026-10-17T03:52:30Z INFO {"event": "training_progress", "step": 2250, "epoch": 6, "loss": 0.23867858946323395, "grad_norm": 0.4252563416957855, "train_seconds": 354.0890373449947, "remaining_budget_seconds": 2404.2327981900016}
+2026-10-17T03:52:37Z INFO {"event": "training_progress", "step": 2300, "epoch": 6, "loss": 0.23412383802235126, "grad_norm": 0.794821560382843, "train_seconds": 361.901018894001, "remaining_budget_seconds": 2396.420816372003}
+2026-10-17T03:52:48Z INFO {"event": "training_progress", "step": 2350, "epoch": 6, "loss": 0.2405099907517433, "grad_norm": 0.342065691947937, "train_seconds": 369.5176268869982, "remaining_budget_seconds": 2388.8042074480036}
+2026-10-17T03:52:56Z INFO {"event": "training_progress", "step": 2400, "epoch": 6, "loss": 0.23195613913238047, "grad_norm": 0.39763814210891724, "train_seconds": 377.20030980498996, "remaining_budget_seconds": 2381.1215251880058}
+2026-10-17T03:53:03Z INFO {"event": "training_progress", "step": 2450, "epoch": 6, "loss": 0.23616296470165252, "grad_norm": 0.4935756325721741, "train_seconds": 385.092698715991, "remaining_budget_seconds": 2373.2291354939953}
+2026-10-17T03:53:11Z INFO {"event": "training_progress", "step": 2500, "epoch": 7, "loss": 0.23028223507106305, "grad_norm": 0.4087282121181488, "train_seconds": 392.9251897929935, "remaining_budget_seconds": 2365.3966460759984}
+2026-10-17T03:53:22Z INFO {"event": "training_progress", "step": 2550, "epoch": 7, "loss": 0.2324374606460333, "grad_norm": 0.3904835879802704, "train_seconds": 400.7959010069899, "remaining_budget_seconds": 2357.525934727004}
+2026-10-17T03:53:30Z INFO {"event": "training_progress", "step": 2600, "epoch": 7, "loss": 0.2309181544929743, "grad_norm": 0.32559341192245483, "train_seconds": 408.79235587400035, "remaining_budget_seconds": 2349.5294789659965}
+2026-10-17T03:53:37Z INFO {"event": "training_progress", "step": 2650, "epoch": 7, "loss": 0.23549096263945102, "grad_norm": 0.5988760590553284, "train_seconds": 416.0627775999892, "remaining_budget_seconds": 2342.2590576280054}
+2026-10-17T03:53:45Z INFO {"event": "training_progress", "step": 2700, "epoch": 7, "loss": 0.22949353851377963, "grad_norm": 0.5599440932273865, "train_seconds": 423.5986522309977, "remaining_budget_seconds": 2334.723178444008}
+2026-10-17T03:53:53Z INFO {"event": "training_progress", "step": 2750, "epoch": 7, "loss": 0.22825357206165792, "grad_norm": 0.413958340883255, "train_seconds": 431.41924708199804, "remaining_budget_seconds": 2326.9025894210063}
+2026-10-17T03:54:04Z INFO {"event": "training_progress", "step": 2800, "epoch": 7, "loss": 0.229015199765563, "grad_norm": 0.3836515247821808, "train_seconds": 439.5745449059905, "remaining_budget_seconds": 2318.747288532002}
+2026-10-17T03:54:12Z INFO {"event": "training_progress", "step": 2850, "epoch": 8, "loss": 0.230223825648427, "grad_norm": 0.33936822414398193, "train_seconds": 447.45398111699615, "remaining_budget_seconds": 2310.8678535530053}
+2026-10-17T03:54:20Z INFO {"event": "training_progress", "step": 2900, "epoch": 8, "loss": 0.22933498866856097, "grad_norm": 0.4638117849826813, "train_seconds": 455.2872456719924, "remaining_budget_seconds": 2303.034589799994}
+2026-10-17T03:54:27Z INFO {"event": "training_progress", "step": 2950, "epoch": 8, "loss": 0.2290338983386755, "grad_norm": 0.5244609713554382, "train_seconds": 462.8320114800008, "remaining_budget_seconds": 2295.4898252809944}
+2026-10-17T03:54:38Z INFO {"event": "training_progress", "step": 3000, "epoch": 8, "loss": 0.22946545623242856, "grad_norm": 0.2612833082675934, "train_seconds": 470.75344095699256, "remaining_budget_seconds": 2287.5683904150064}
+2026-10-17T03:54:46Z INFO {"event": "training_progress", "step": 3050, "epoch": 8, "loss": 0.23123068794608115, "grad_norm": 0.7766280174255371, "train_seconds": 478.359046491998, "remaining_budget_seconds": 2279.962790286998}
+2026-10-17T03:54:53Z INFO {"event": "training_progress", "step": 3100, "epoch": 8, "loss": 0.23037957362830638, "grad_norm": 0.47285857796669006, "train_seconds": 485.91529949299, "remaining_budget_seconds": 2272.4065343020047}
+2026-10-17T03:55:01Z INFO {"event": "training_progress", "step": 3150, "epoch": 8, "loss": 0.2298418479412794, "grad_norm": 0.46852678060531616, "train_seconds": 493.70026312999835, "remaining_budget_seconds": 2264.6215739129984}
+2026-10-17T03:55:12Z INFO {"event": "training_progress", "step": 3200, "epoch": 9, "loss": 0.22839245028793811, "grad_norm": 0.4697372019290924, "train_seconds": 501.796848375001, "remaining_budget_seconds": 2256.524985433003}
+2026-10-17T03:55:20Z INFO {"event": "training_progress", "step": 3250, "epoch": 9, "loss": 0.23097094900906087, "grad_norm": 0.46524712443351746, "train_seconds": 509.77707105199806, "remaining_budget_seconds": 2248.5447651430004}
+2026-10-17T03:55:28Z INFO {"event": "training_progress", "step": 3300, "epoch": 9, "loss": 0.2286774291098118, "grad_norm": 0.3661056458950043, "train_seconds": 517.418930653992, "remaining_budget_seconds": 2240.9029050700046}
+2026-10-17T03:55:35Z INFO {"event": "training_progress", "step": 3350, "epoch": 9, "loss": 0.2254503657668829, "grad_norm": 0.4035114645957947, "train_seconds": 524.9639057969907, "remaining_budget_seconds": 2233.357929674996}
+2026-10-17T03:55:46Z INFO {"event": "training_progress", "step": 3400, "epoch": 9, "loss": 0.22484782733023168, "grad_norm": 0.3980918526649475, "train_seconds": 532.7969151149882, "remaining_budget_seconds": 2225.5249212049966}
+2026-10-17T03:55:54Z INFO {"event": "training_progress", "step": 3450, "epoch": 9, "loss": 0.22531763449311257, "grad_norm": 0.3730781376361847, "train_seconds": 540.5949077819969, "remaining_budget_seconds": 2217.7269286820083}
+2026-10-17T03:56:01Z INFO {"event": "training_progress", "step": 3500, "epoch": 9, "loss": 0.22735451720654964, "grad_norm": 0.4551793336868286, "train_seconds": 548.0730959039938, "remaining_budget_seconds": 2210.2487388470036}
+2026-10-17T03:56:09Z INFO {"event": "training_progress", "step": 3550, "epoch": 10, "loss": 0.2241469468921423, "grad_norm": 0.366840124130249, "train_seconds": 555.8608918039972, "remaining_budget_seconds": 2202.4609448050032}
+2026-10-17T03:56:20Z INFO {"event": "training_progress", "step": 3600, "epoch": 10, "loss": 0.23021844476461412, "grad_norm": 0.3977644443511963, "train_seconds": 564.0320246079937, "remaining_budget_seconds": 2194.289811544004}
+2026-10-17T03:56:28Z INFO {"event": "training_progress", "step": 3650, "epoch": 10, "loss": 0.2254459173977375, "grad_norm": 0.2870456874370575, "train_seconds": 571.6687382249947, "remaining_budget_seconds": 2186.6530960300006}
+2026-10-17T03:56:35Z INFO {"event": "training_progress", "step": 3700, "epoch": 10, "loss": 0.22565270893275738, "grad_norm": 0.3958859443664551, "train_seconds": 579.0534677549877, "remaining_budget_seconds": 2179.2683687440003}
+2026-10-17T03:56:43Z INFO {"event": "training_progress", "step": 3750, "epoch": 10, "loss": 0.22105053029954433, "grad_norm": 0.3439747095108032, "train_seconds": 586.6161270459997, "remaining_budget_seconds": 2171.705709449001}
+2026-10-17T03:56:54Z INFO {"event": "training_progress", "step": 3800, "epoch": 10, "loss": 0.22162507683038712, "grad_norm": 0.334031879901886, "train_seconds": 594.6242707939964, "remaining_budget_seconds": 2163.697563623995}
+2026-10-17T03:57:02Z INFO {"event": "training_progress", "step": 3850, "epoch": 10, "loss": 0.223569995239377, "grad_norm": 0.46721187233924866, "train_seconds": 602.6255122209986, "remaining_budget_seconds": 2155.6963242509955}
+2026-10-17T03:57:09Z INFO {"event": "training_progress", "step": 3900, "epoch": 11, "loss": 0.22280083172023296, "grad_norm": 0.38908547163009644, "train_seconds": 610.3119678549992, "remaining_budget_seconds": 2148.009867671004}
+2026-10-17T03:57:17Z INFO {"event": "training_progress", "step": 3950, "epoch": 11, "loss": 0.22489800073206426, "grad_norm": 0.32635924220085144, "train_seconds": 618.0349892119993, "remaining_budget_seconds": 2140.2868472150003}
+2026-10-17T03:57:25Z INFO {"event": "training_progress", "step": 4000, "epoch": 11, "loss": 0.22181923925876618, "grad_norm": 0.2884049713611603, "train_seconds": 625.7536075979879, "remaining_budget_seconds": 2132.568223633003}
+2026-10-17T03:57:36Z INFO {"event": "training_progress", "step": 4050, "epoch": 11, "loss": 0.2218459939956665, "grad_norm": 0.3524567186832428, "train_seconds": 633.6408477769874, "remaining_budget_seconds": 2124.6809871720034}
+2026-10-17T03:57:43Z INFO {"event": "training_progress", "step": 4100, "epoch": 11, "loss": 0.2216662535071373, "grad_norm": 0.39295804500579834, "train_seconds": 641.465680592999, "remaining_budget_seconds": 2116.856156331007}
+2026-10-17T03:57:51Z INFO {"event": "training_progress", "step": 4150, "epoch": 11, "loss": 0.22077991373836994, "grad_norm": 0.313662052154541, "train_seconds": 649.2737053869932, "remaining_budget_seconds": 2109.0481296100043}
+2026-10-17T03:57:59Z INFO {"event": "training_progress", "step": 4200, "epoch": 11, "loss": 0.22154499284923077, "grad_norm": 0.27141043543815613, "train_seconds": 657.2216755109985, "remaining_budget_seconds": 2101.100160394999}
+2026-10-17T03:58:10Z INFO {"event": "training_progress", "step": 4250, "epoch": 12, "loss": 0.22320055343210699, "grad_norm": 0.3563011884689331, "train_seconds": 665.018669180994, "remaining_budget_seconds": 2093.3031629540055}
+2026-10-17T03:58:18Z INFO {"event": "training_progress", "step": 4300, "epoch": 12, "loss": 0.22273461155593396, "grad_norm": 0.22989892959594727, "train_seconds": 672.8586085889983, "remaining_budget_seconds": 2085.4632278200006}
+2026-10-17T03:58:25Z INFO {"event": "training_progress", "step": 4350, "epoch": 12, "loss": 0.22284090965986253, "grad_norm": 0.36647114157676697, "train_seconds": 680.313178878001, "remaining_budget_seconds": 2078.008656521997}
+2026-10-17T03:58:33Z INFO {"event": "training_progress", "step": 4400, "epoch": 12, "loss": 0.21992925234138966, "grad_norm": 0.3362022042274475, "train_seconds": 688.047629171997, "remaining_budget_seconds": 2070.274201955006}
+2026-10-17T03:58:44Z INFO {"event": "training_progress", "step": 4450, "epoch": 12, "loss": 0.2233125678449869, "grad_norm": 0.2912153899669647, "train_seconds": 695.9934053949983, "remaining_budget_seconds": 2062.3284298759972}
+2026-10-17T03:58:52Z INFO {"event": "training_progress", "step": 4500, "epoch": 12, "loss": 0.22189435482025147, "grad_norm": 0.4446864128112793, "train_seconds": 703.7602820629982, "remaining_budget_seconds": 2054.5615530299983}
+2026-10-17T03:58:59Z INFO {"event": "training_progress", "step": 4550, "epoch": 12, "loss": 0.22302159786224365, "grad_norm": 0.2930428981781006, "train_seconds": 711.4029577509937, "remaining_budget_seconds": 2046.9188789830077}
+2026-10-17T03:59:07Z INFO {"event": "training_progress", "step": 4600, "epoch": 13, "loss": 0.2208305049687624, "grad_norm": 0.35772034525871277, "train_seconds": 719.2710717049922, "remaining_budget_seconds": 2039.0507636230031}
+2026-10-17T03:59:18Z INFO {"event": "training_progress", "step": 4650, "epoch": 13, "loss": 0.2222816248983145, "grad_norm": 0.4955046772956848, "train_seconds": 727.0552881299955, "remaining_budget_seconds": 2031.2665454060043}
+2026-10-17T03:59:26Z INFO {"event": "training_progress", "step": 4700, "epoch": 13, "loss": 0.21849496431648732, "grad_norm": 0.27598199248313904, "train_seconds": 734.9063177679927, "remaining_budget_seconds": 2023.4155159810034}
+2026-10-17T03:59:34Z INFO {"event": "training_progress", "step": 4750, "epoch": 13, "loss": 0.2238991356641054, "grad_norm": 0.29004335403442383, "train_seconds": 742.7180941699917, "remaining_budget_seconds": 2015.603742708001}
+2026-10-17T03:59:42Z INFO {"event": "training_progress", "step": 4800, "epoch": 13, "loss": 0.220438726618886, "grad_norm": 0.29530641436576843, "train_seconds": 750.4801842839952, "remaining_budget_seconds": 2007.8416522420011}
+2026-10-17T03:59:53Z INFO {"event": "training_progress", "step": 4850, "epoch": 13, "loss": 0.21997170239686967, "grad_norm": 0.4255085289478302, "train_seconds": 758.6039118219924, "remaining_budget_seconds": 1999.7179221179977}
+2026-10-17T04:00:01Z INFO {"event": "training_progress", "step": 4900, "epoch": 13, "loss": 0.21904529206454754, "grad_norm": 0.28269851207733154, "train_seconds": 766.5360138359974, "remaining_budget_seconds": 1991.7858228889963}
+2026-10-17T04:00:08Z INFO {"event": "training_progress", "step": 4950, "epoch": 14, "loss": 0.2196836384385824, "grad_norm": 0.3604487180709839, "train_seconds": 774.0026329309912, "remaining_budget_seconds": 1984.3192020989954}
+2026-10-17T04:00:16Z INFO {"event": "training_progress", "step": 5000, "epoch": 14, "loss": 0.22161536678671837, "grad_norm": 0.6659505367279053, "train_seconds": 781.5521558229957, "remaining_budget_seconds": 1976.769677963006}
+2026-10-17T04:00:26Z INFO {"event": "training_progress", "step": 5050, "epoch": 14, "loss": 0.2212749530375004, "grad_norm": 0.32987791299819946, "train_seconds": 789.3520595309965, "remaining_budget_seconds": 1968.9697762090073}
+2026-10-17T04:00:34Z INFO {"event": "training_progress", "step": 5100, "epoch": 14, "loss": 0.2198681078851223, "grad_norm": 0.34914904832839966, "train_seconds": 797.1751148429903, "remaining_budget_seconds": 1961.1467192489945}
+2026-10-17T04:00:42Z INFO {"event": "training_progress", "step": 5150, "epoch": 14, "loss": 0.21723262019455433, "grad_norm": 0.4405971169471741, "train_seconds": 804.7872387199895, "remaining_budget_seconds": 1953.534595966994}
+2026-10-17T04:00:50Z INFO {"event": "training_progress", "step": 5200, "epoch": 14, "loss": 0.22147177189588546, "grad_norm": 0.2907277047634125, "train_seconds": 812.4724804399884, "remaining_budget_seconds": 1945.8493524600053}
+2026-10-17T04:01:00Z INFO {"event": "training_progress", "step": 5250, "epoch": 14, "loss": 0.21746282890439034, "grad_norm": 0.36081433296203613, "train_seconds": 820.306955121996, "remaining_budget_seconds": 1938.0148791429965}
+2026-10-17T04:01:08Z INFO {"event": "training_progress", "step": 5300, "epoch": 15, "loss": 0.21793650314211846, "grad_norm": 0.40886375308036804, "train_seconds": 828.2668339369993, "remaining_budget_seconds": 1930.0550025779958}
+2026-10-17T04:01:16Z INFO {"event": "training_progress", "step": 5350, "epoch": 15, "loss": 0.2197545877844095, "grad_norm": 0.40450724959373474, "train_seconds": 835.8440800149983, "remaining_budget_seconds": 1922.477755151005}
+2026-10-17T04:01:24Z INFO {"event": "training_progress", "step": 5400, "epoch": 15, "loss": 0.21932302169501783, "grad_norm": 0.3138900101184845, "train_seconds": 843.6534106549952, "remaining_budget_seconds": 1914.668426231001}
+2026-10-17T04:01:32Z INFO {"event": "training_progress", "step": 5450, "epoch": 15, "loss": 0.21692535206675528, "grad_norm": 0.2490702122449875, "train_seconds": 851.4089029710012, "remaining_budget_seconds": 1906.9129336990009}
+2026-10-17T04:01:41Z INFO {"event": "training_progress", "step": 5500, "epoch": 15, "loss": 0.21994528509676456, "grad_norm": 0.3319157361984253, "train_seconds": 859.0554112929967, "remaining_budget_seconds": 1899.2664250350063}
+2026-10-17T04:01:49Z INFO {"event": "training_progress", "step": 5550, "epoch": 15, "loss": 0.21900013759732245, "grad_norm": 0.3277781903743744, "train_seconds": 866.5904747619934, "remaining_budget_seconds": 1891.7313603399962}
+2026-10-17T04:01:56Z INFO {"event": "training_progress", "step": 5600, "epoch": 15, "loss": 0.21626815855503081, "grad_norm": 0.33965423703193665, "train_seconds": 873.9013954229886, "remaining_budget_seconds": 1884.4204395169945}
+2026-10-17T04:02:04Z INFO {"event": "training_progress", "step": 5650, "epoch": 16, "loss": 0.21617043375968933, "grad_norm": 0.3029640018939972, "train_seconds": 881.7657526879921, "remaining_budget_seconds": 1876.5560836320074}
+2026-10-17T04:02:15Z INFO {"event": "training_progress", "step": 5700, "epoch": 16, "loss": 0.21758351117372512, "grad_norm": 0.24940519034862518, "train_seconds": 889.7597383289976, "remaining_budget_seconds": 1868.5620959929947}
+2026-10-17T04:02:23Z INFO {"event": "training_progress", "step": 5750, "epoch": 16, "loss": 0.2173467715829611, "grad_norm": 0.34982332587242126, "train_seconds": 897.6539546799904, "remaining_budget_seconds": 1860.6678816890053}
+2026-10-17T04:02:31Z INFO {"event": "training_progress", "step": 5800, "epoch": 16, "loss": 0.21773493908345698, "grad_norm": 0.30662330985069275, "train_seconds": 905.5395190009876, "remaining_budget_seconds": 1852.782317981997}
+2026-10-17T04:02:39Z INFO {"event": "training_progress", "step": 5850, "epoch": 16, "loss": 0.2151334735006094, "grad_norm": 0.37888336181640625, "train_seconds": 913.4551282789907, "remaining_budget_seconds": 1844.8667066890048}
+2026-10-17T04:02:50Z INFO {"event": "training_progress", "step": 5900, "epoch": 16, "loss": 0.2174584672600031, "grad_norm": 0.32703760266304016, "train_seconds": 921.3964516129927, "remaining_budget_seconds": 1836.9253850079986}
+2026-10-17T04:02:58Z INFO {"event": "training_progress", "step": 5950, "epoch": 16, "loss": 0.21562115371227264, "grad_norm": 0.5272236466407776, "train_seconds": 929.0775505989877, "remaining_budget_seconds": 1829.2442844570032}
+2026-10-17T04:03:05Z INFO {"event": "training_progress", "step": 6000, "epoch": 17, "loss": 0.21789717718958854, "grad_norm": 0.40285295248031616, "train_seconds": 936.8162944209907, "remaining_budget_seconds": 1821.5055399410048}
+2026-10-17T04:03:13Z INFO {"event": "training_progress", "step": 6050, "epoch": 17, "loss": 0.21542367368936538, "grad_norm": 0.2874428927898407, "train_seconds": 944.4713530089939, "remaining_budget_seconds": 1813.8504786620033}
+2026-10-17T04:03:24Z INFO {"event": "training_progress", "step": 6100, "epoch": 17, "loss": 0.21502131074666977, "grad_norm": 0.23543153703212738, "train_seconds": 952.1459115089965, "remaining_budget_seconds": 1806.1759221139946}
+2026-10-17T04:03:32Z INFO {"event": "training_progress", "step": 6150, "epoch": 17, "loss": 0.21437974125146866, "grad_norm": 0.3237047493457794, "train_seconds": 960.086726397989, "remaining_budget_seconds": 1798.235106292006}
+2026-10-17T04:03:39Z INFO {"event": "training_progress", "step": 6200, "epoch": 17, "loss": 0.2139349552989006, "grad_norm": 0.18924817442893982, "train_seconds": 967.7325660309871, "remaining_budget_seconds": 1790.589269011005}
+2026-10-17T04:03:47Z INFO {"event": "training_progress", "step": 6250, "epoch": 17, "loss": 0.21625405430793762, "grad_norm": 0.34389325976371765, "train_seconds": 975.6470755740011, "remaining_budget_seconds": 1782.6747612000036}
+2026-10-17T04:03:58Z INFO {"event": "training_progress", "step": 6300, "epoch": 17, "loss": 0.21517898343503475, "grad_norm": 0.4151454269886017, "train_seconds": 983.5430716789997, "remaining_budget_seconds": 1774.778763484006}
+2026-10-17T04:04:06Z INFO {"event": "training_progress", "step": 6350, "epoch": 18, "loss": 0.21563126638531685, "grad_norm": 0.3418363034725189, "train_seconds": 991.4824961279955, "remaining_budget_seconds": 1766.83933884}
+2026-10-17T04:04:14Z INFO {"event": "training_progress", "step": 6400, "epoch": 18, "loss": 0.21397394098341466, "grad_norm": 0.2993028461933136, "train_seconds": 999.1218924829882, "remaining_budget_seconds": 1759.1999428599956}
+2026-10-17T04:04:21Z INFO {"event": "training_progress", "step": 6450, "epoch": 18, "loss": 0.2147428645938635, "grad_norm": 0.2922501862049103, "train_seconds": 1006.6411179089919, "remaining_budget_seconds": 1751.6807176400034}
+2026-10-17T04:04:32Z INFO {"event": "training_progress", "step": 6500, "epoch": 18, "loss": 0.21873606115579605, "grad_norm": 0.3565116226673126, "train_seconds": 1014.2347380539868, "remaining_budget_seconds": 1744.0870968630043}
+2026-10-17T04:04:39Z INFO {"event": "training_progress", "step": 6550, "epoch": 18, "loss": 0.2156849554181099, "grad_norm": 0.28380316495895386, "train_seconds": 1021.7940089959884, "remaining_budget_seconds": 1736.5278244939982}
+2026-10-17T04:04:47Z INFO {"event": "training_progress", "step": 6600, "epoch": 18, "loss": 0.21615735933184624, "grad_norm": 0.3619202673435211, "train_seconds": 1029.318713752, "remaining_budget_seconds": 1729.003122492999}
+2026-10-17T04:04:55Z INFO {"event": "training_progress", "step": 6650, "epoch": 18, "loss": 0.21398403897881507, "grad_norm": 0.2969871163368225, "train_seconds": 1036.8598087749997, "remaining_budget_seconds": 1721.4620264889963}
+2026-10-17T04:05:03Z INFO {"event": "training_progress", "step": 6700, "epoch": 19, "loss": 0.21142808258533477, "grad_norm": 0.2804839611053467, "train_seconds": 1044.8744632999878, "remaining_budget_seconds": 1713.4473716560024}
+2026-10-17T04:05:14Z INFO {"event": "training_progress", "step": 6750, "epoch": 19, "loss": 0.21454619318246843, "grad_norm": 0.29788145422935486, "train_seconds": 1052.811600727, "remaining_budget_seconds": 1705.510234048008}
+2026-10-17T04:05:21Z INFO {"event": "training_progress", "step": 6800, "epoch": 19, "loss": 0.21666650481522085, "grad_norm": 0.24832245707511902, "train_seconds": 1060.0762381339882, "remaining_budget_seconds": 1698.245595690998}
+2026-10-17T04:05:29Z INFO {"event": "training_progress", "step": 6850, "epoch": 19, "loss": 0.21477059997618198, "grad_norm": 0.3523756265640259, "train_seconds": 1068.003264354993, "remaining_budget_seconds": 1690.3185722820053}
+2026-10-17T04:05:37Z INFO {"event": "training_progress", "step": 6900, "epoch": 19, "loss": 0.2147611553966999, "grad_norm": 0.5198025107383728, "train_seconds": 1075.9106873579876, "remaining_budget_seconds": 1682.4111471039942}
+2026-10-17T04:05:48Z INFO {"event": "training_progress", "step": 6950, "epoch": 19, "loss": 0.21694304257631303, "grad_norm": 0.3194659948348999, "train_seconds": 1084.0235748859996, "remaining_budget_seconds": 1674.2982602470001}
+2026-10-17T04:05:55Z INFO {"event": "training_progress", "step": 7000, "epoch": 19, "loss": 0.2135413057357073, "grad_norm": 0.4402426779270172, "train_seconds": 1091.6234369219892, "remaining_budget_seconds": 1666.6983978979988}
+2026-10-17T04:06:03Z INFO {"event": "training_progress", "step": 7050, "epoch": 20, "loss": 0.21766627244651318, "grad_norm": 0.31235089898109436, "train_seconds": 1099.374298912997, "remaining_budget_seconds": 1658.9475360869983}
+2026-10-17T04:06:10Z INFO {"event": "training_progress", "step": 7100, "epoch": 20, "loss": 0.21267986677587034, "grad_norm": 0.31476157903671265, "train_seconds": 1106.7527203319914, "remaining_budget_seconds": 1651.5691166190081}
+2026-10-17T04:06:21Z INFO {"event": "training_progress", "step": 7150, "epoch": 20, "loss": 0.21076386354863644, "grad_norm": 0.2951830327510834, "train_seconds": 1114.442969782991, "remaining_budget_seconds": 1643.8788666820037}
+2026-10-17T04:06:28Z INFO {"event": "training_progress", "step": 7200, "epoch": 20, "loss": 0.21327918708324434, "grad_norm": 0.18330152332782745, "train_seconds": 1121.7203869219957, "remaining_budget_seconds": 1636.6014499349985}
+2026-10-17T04:06:36Z INFO {"event": "training_progress", "step": 7250, "epoch": 20, "loss": 0.2103935207426548, "grad_norm": 0.3269781470298767, "train_seconds": 1129.2792172069894, "remaining_budget_seconds": 1629.0426171889994}
+2026-10-17T04:06:44Z INFO {"event": "training_progress", "step": 7300, "epoch": 20, "loss": 0.21639143161475657, "grad_norm": 0.6085897088050842, "train_seconds": 1137.0292937129998, "remaining_budget_seconds": 1621.2925403169938}
+2026-10-17T04:06:55Z INFO {"event": "training_progress", "step": 7350, "epoch": 20, "loss": 0.21413941010832788, "grad_norm": 0.3294401466846466, "train_seconds": 1145.1066068859946, "remaining_budget_seconds": 1613.2152279809961}
+2026-10-17T04:07:03Z INFO {"event": "training_progress", "step": 7400, "epoch": 21, "loss": 0.21409860655665397, "grad_norm": 0.37616610527038574, "train_seconds": 1153.0034437930008, "remaining_budget_seconds": 1605.318388365995}
+2026-10-17T04:07:10Z INFO {"event": "training_progress", "step": 7450, "epoch": 21, "loss": 0.2148568043112755, "grad_norm": 0.30314815044403076, "train_seconds": 1160.6795455319952, "remaining_budget_seconds": 1597.6422915449948}
+2026-10-17T04:07:18Z INFO {"event": "training_progress", "step": 7500, "epoch": 21, "loss": 0.21370703846216202, "grad_norm": 0.2392214834690094, "train_seconds": 1168.4110350419942, "remaining_budget_seconds": 1589.9107979719993}
+2026-10-17T04:07:29Z INFO {"event": "training_progress", "step": 7550, "epoch": 21, "loss": 0.21350899375975133, "grad_norm": 0.23006859421730042, "train_seconds": 1176.1663259179913, "remaining_budget_seconds": 1582.155509179007}
+2026-10-17T04:07:37Z INFO {"event": "training_progress", "step": 7600, "epoch": 21, "loss": 0.21544615797698496, "grad_norm": 0.3027198016643524, "train_seconds": 1183.9288353589945, "remaining_budget_seconds": 1574.392999749005}
+2026-10-17T04:07:44Z INFO {"event": "training_progress", "step": 7650, "epoch": 21, "loss": 0.2147886786609888, "grad_norm": 0.3550317883491516, "train_seconds": 1191.574774074994, "remaining_budget_seconds": 1566.7470612010075}
+2026-10-17T04:07:52Z INFO {"event": "training_progress", "step": 7700, "epoch": 21, "loss": 0.2158916885405779, "grad_norm": 0.2358863204717636, "train_seconds": 1199.1939207549876, "remaining_budget_seconds": 1559.127912423006}
+2026-10-17T04:08:03Z INFO {"event": "training_progress", "step": 7750, "epoch": 22, "loss": 0.21291930735111236, "grad_norm": 0.18737532198429108, "train_seconds": 1207.088233328992, "remaining_budget_seconds": 1551.2336010430008}
+2026-10-17T04:08:11Z INFO {"event": "training_progress", "step": 7800, "epoch": 22, "loss": 0.21512381210923195, "grad_norm": 0.21198752522468567, "train_seconds": 1214.795930731998, "remaining_budget_seconds": 1543.5259056890063}
+2026-10-17T04:08:18Z INFO {"event": "training_progress", "step": 7850, "epoch": 22, "loss": 0.2132995054870844, "grad_norm": 0.30421558022499084, "train_seconds": 1222.6823277999938, "remaining_budget_seconds": 1535.6395070739964}
+2026-10-17T04:08:26Z INFO {"event": "training_progress", "step": 7900, "epoch": 22, "loss": 0.21161525279283525, "grad_norm": 0.2890901565551758, "train_seconds": 1230.4013572149997, "remaining_budget_seconds": 1527.9204797130078}
+2026-10-17T04:08:34Z INFO {"event": "training_progress", "step": 7950, "epoch": 22, "loss": 0.2111891806870699, "grad_norm": 0.20565500855445862, "train_seconds": 1238.0509042099875, "remaining_budget_seconds": 1520.2709299160051}
+2026-10-17T04:08:44Z INFO {"event": "training_progress", "step": 8000, "epoch": 22, "loss": 0.21140068113803864, "grad_norm": 0.1962038278579712, "train_seconds": 1245.697219624999, "remaining_budget_seconds": 1512.6246157139976}
+2026-10-17T04:08:52Z INFO {"event": "training_progress", "step": 8050, "epoch": 22, "loss": 0.21494184851646422, "grad_norm": 0.2660854160785675, "train_seconds": 1253.2684135819873, "remaining_budget_seconds": 1505.0534230009944}
+2026-10-17T04:09:00Z INFO {"event": "training_progress", "step": 8100, "epoch": 23, "loss": 0.21095135249197483, "grad_norm": 0.2654925286769867, "train_seconds": 1260.8700262299972, "remaining_budget_seconds": 1497.4518088680052}
+2026-10-17T04:09:07Z INFO {"event": "training_progress", "step": 8150, "epoch": 23, "loss": 0.2122656473517418, "grad_norm": 0.3145787715911865, "train_seconds": 1268.6533213359944, "remaining_budget_seconds": 1489.6685137640015}
+2026-10-17T04:09:18Z INFO {"event": "training_progress", "step": 8200, "epoch": 23, "loss": 0.21402065746486187, "grad_norm": 0.23672500252723694, "train_seconds": 1276.3649375620007, "remaining_budget_seconds": 1481.9568927500077}
+2026-10-17T04:09:26Z INFO {"event": "training_progress", "step": 8250, "epoch": 23, "loss": 0.2135489685833454, "grad_norm": 0.283255010843277, "train_seconds": 1284.1676059569872, "remaining_budget_seconds": 1474.154229392996}
+2026-10-17T04:09:34Z INFO {"event": "training_progress", "step": 8300, "epoch": 23, "loss": 0.21000600464642047, "grad_norm": 0.2620209753513336, "train_seconds": 1291.692848753999, "remaining_budget_seconds": 1466.628986294003}
+2026-10-17T04:09:41Z INFO {"event": "training_progress", "step": 8350, "epoch": 23, "loss": 0.21077331110835076, "grad_norm": 0.18734794855117798, "train_seconds": 1299.3914980629925, "remaining_budget_seconds": 1458.930338702994}
+2026-10-17T04:09:52Z INFO {"event": "training_progress", "step": 8400, "epoch": 23, "loss": 0.21178479686379434, "grad_norm": 0.2683074474334717, "train_seconds": 1307.1540672879928, "remaining_budget_seconds": 1451.167767114006}
+2026-10-17T04:10:00Z INFO {"event": "training_progress", "step": 8450, "epoch": 24, "loss": 0.21156985260546207, "grad_norm": 0.2895117700099945, "train_seconds": 1314.9125828409888, "remaining_budget_seconds": 1443.4092514109943}
+2026-10-17T04:10:07Z INFO {"event": "training_progress", "step": 8500, "epoch": 24, "loss": 0.2084442099928856, "grad_norm": 0.2583526372909546, "train_seconds": 1322.5652359539963, "remaining_budget_seconds": 1435.7566004050022}
+2026-10-17T04:10:15Z INFO {"event": "training_progress", "step": 8550, "epoch": 24, "loss": 0.210680635496974, "grad_norm": 0.34920305013656616, "train_seconds": 1330.3292170599889, "remaining_budget_seconds": 1427.9926174730062}
+2026-10-17T04:10:26Z INFO {"event": "training_progress", "step": 8600, "epoch": 24, "loss": 0.2129681457579136, "grad_norm": 0.2752314805984497, "train_seconds": 1338.0781599779875, "remaining_budget_seconds": 1420.2436766290048}
+2026-10-17T04:10:34Z INFO {"event": "training_progress", "step": 8650, "epoch": 24, "loss": 0.2114828223735094, "grad_norm": 0.25787147879600525, "train_seconds": 1345.6931234829972, "remaining_budget_seconds": 1412.6287134449958}
+2026-10-17T04:10:41Z INFO {"event": "training_progress", "step": 8700, "epoch": 24, "loss": 0.20974436581134795, "grad_norm": 0.3201660215854645, "train_seconds": 1353.3435578469944, "remaining_budget_seconds": 1404.9782759539958}
+2026-10-17T04:10:49Z INFO {"event": "training_progress", "step": 8750, "epoch": 24, "loss": 0.2113457352668047, "grad_norm": 0.4056096076965332, "train_seconds": 1360.7676983649872, "remaining_budget_seconds": 1397.5541378890048}
+2026-10-17T04:10:59Z INFO {"event": "training_progress", "step": 8800, "epoch": 25, "loss": 0.21290438413619994, "grad_norm": 0.23211809992790222, "train_seconds": 1368.4760849519953, "remaining_budget_seconds": 1389.845749799002}
+2026-10-17T04:11:07Z INFO {"event": "training_progress", "step": 8850, "epoch": 25, "loss": 0.2116118536144495, "grad_norm": 0.22702480852603912, "train_seconds": 1376.36113209599, "remaining_budget_seconds": 1381.9606989349995}
+2026-10-17T04:11:15Z INFO {"event": "training_progress", "step": 8900, "epoch": 25, "loss": 0.21322670295834542, "grad_norm": 0.26616641879081726, "train_seconds": 1384.0168207529932, "remaining_budget_seconds": 1374.3050131030031}
+2026-10-17T04:11:23Z INFO {"event": "training_progress", "step": 8950, "epoch": 25, "loss": 0.2109166006743908, "grad_norm": 0.21228192746639252, "train_seconds": 1391.8255218299892, "remaining_budget_seconds": 1366.4963122619956}
+2026-10-17T04:11:30Z INFO {"event": "training_progress", "step": 9000, "epoch": 25, "loss": 0.20991788499057293, "grad_norm": 0.18925383687019348, "train_seconds": 1399.233434262991, "remaining_budget_seconds": 1359.0883996570046}
+2026-10-17T04:11:41Z INFO {"event": "training_progress", "step": 9050, "epoch": 25, "loss": 0.20939844250679016, "grad_norm": 0.25925150513648987, "train_seconds": 1407.370663597001, "remaining_budget_seconds": 1350.9511723339965}
+2026-10-17T04:11:49Z INFO {"event": "training_progress", "step": 9100, "epoch": 25, "loss": 0.2076793273538351, "grad_norm": 0.288874089717865, "train_seconds": 1415.053396349991, "remaining_budget_seconds": 1343.2684383040032}
+2026-10-17T04:11:57Z INFO {"event": "training_progress", "step": 9150, "epoch": 26, "loss": 0.2083969806134701, "grad_norm": 0.2412095069885254, "train_seconds": 1422.7334826799924, "remaining_budget_seconds": 1335.5883524579986}
+2026-10-17T04:12:04Z INFO {"event": "training_progress", "step": 9200, "epoch": 26, "loss": 0.21277480609714985, "grad_norm": 0.21001605689525604, "train_seconds": 1430.4471712449886, "remaining_budget_seconds": 1327.8746652919945}
+2026-10-17T04:12:15Z INFO {"event": "training_progress", "step": 9250, "epoch": 26, "loss": 0.20991496451199054, "grad_norm": 0.21817699074745178, "train_seconds": 1438.2537415299885, "remaining_budget_seconds": 1320.0680949929956}
+2026-10-17T04:12:23Z INFO {"event": "training_progress", "step": 9300, "epoch": 26, "loss": 0.21052162520587445, "grad_norm": 0.2136511355638504, "train_seconds": 1446.1444144839916, "remaining_budget_seconds": 1312.177420699998}
+2026-10-17T04:12:31Z INFO {"event": "training_progress", "step": 9350, "epoch": 26, "loss": 0.20912824481725692, "grad_norm": 0.19794027507305145, "train_seconds": 1453.8239051209966, "remaining_budget_seconds": 1304.4979297920072}
+2026-10-17T04:12:39Z INFO {"event": "training_progress", "step": 9400, "epoch": 26, "loss": 0.2090521477162838, "grad_norm": 0.22232241928577423, "train_seconds": 1461.7532556009974, "remaining_budget_seconds": 1296.5685802020016}
+2026-10-17T04:12:50Z INFO {"event": "training_progress", "step": 9450, "epoch": 26, "loss": 0.21150196492671966, "grad_norm": 0.23561358451843262, "train_seconds": 1469.4952608849999, "remaining_budget_seconds": 1288.8265741000068}
+2026-10-17T04:12:57Z INFO {"event": "training_progress", "step": 9500, "epoch": 27, "loss": 0.2089910740405321, "grad_norm": 0.25248047709465027, "train_seconds": 1477.2469319279917, "remaining_budget_seconds": 1281.0749047160061}
+2026-10-17T04:13:05Z INFO {"event": "training_progress", "step": 9550, "epoch": 27, "loss": 0.20931351102888585, "grad_norm": 0.2125513255596161, "train_seconds": 1484.7878080730006, "remaining_budget_seconds": 1273.534028562004}
+2026-10-17T04:13:13Z INFO {"event": "training_progress", "step": 9600, "epoch": 27, "loss": 0.20823630146682262, "grad_norm": 0.21613138914108276, "train_seconds": 1492.514353214996, "remaining_budget_seconds": 1265.8074822229974}
+2026-10-17T04:13:24Z INFO {"event": "training_progress", "step": 9650, "epoch": 27, "loss": 0.20836987391114234, "grad_norm": 0.20630857348442078, "train_seconds": 1500.7114703239931, "remaining_budget_seconds": 1257.6103637019987}
+2026-10-17T04:13:32Z INFO {"event": "training_progress", "step": 9700, "epoch": 27, "loss": 0.20715567335486412, "grad_norm": 0.21679754555225372, "train_seconds": 1508.928949093999, "remaining_budget_seconds": 1249.392887666996}
+2026-10-17T04:13:40Z INFO {"event": "training_progress", "step": 9750, "epoch": 27, "loss": 0.21116173230111598, "grad_norm": 0.3416095972061157, "train_seconds": 1516.7424823149922, "remaining_budget_seconds": 1241.579352796005}
+2026-10-17T04:13:48Z INFO {"event": "training_progress", "step": 9800, "epoch": 27, "loss": 0.2092953157424927, "grad_norm": 0.22785528004169464, "train_seconds": 1524.566332213988, "remaining_budget_seconds": 1233.755504724002}
+2026-10-17T04:13:59Z INFO {"event": "training_progress", "step": 9850, "epoch": 28, "loss": 0.2085603677481413, "grad_norm": 0.30455249547958374, "train_seconds": 1532.6864945929992, "remaining_budget_seconds": 1225.6353414729965}
+2026-10-17T04:14:07Z INFO {"event": "training_progress", "step": 9900, "epoch": 28, "loss": 0.20748061783611774, "grad_norm": 0.41693025827407837, "train_seconds": 1540.6405236529972, "remaining_budget_seconds": 1217.6813098489947}
+2026-10-17T04:14:15Z INFO {"event": "training_progress", "step": 9950, "epoch": 28, "loss": 0.20914514563977718, "grad_norm": 0.20932404696941376, "train_seconds": 1548.7304899169976, "remaining_budget_seconds": 1209.5913466100028}
+2026-10-17T04:14:23Z INFO {"event": "training_progress", "step": 10000, "epoch": 28, "loss": 0.2082276027649641, "grad_norm": 0.24646010994911194, "train_seconds": 1556.5190599869966, "remaining_budget_seconds": 1201.8027767169988}
+2026-10-17T04:14:34Z INFO {"event": "training_progress", "step": 10050, "epoch": 28, "loss": 0.21005257971584798, "grad_norm": 0.2576836347579956, "train_seconds": 1564.5457903809875, "remaining_budget_seconds": 1193.7760406309972}
+2026-10-17T04:14:42Z INFO {"event": "training_progress", "step": 10100, "epoch": 28, "loss": 0.2101259721070528, "grad_norm": 0.16294877231121063, "train_seconds": 1572.5354334339936, "remaining_budget_seconds": 1185.7864033060032}
+2026-10-17T04:14:50Z INFO {"event": "training_progress", "step": 10150, "epoch": 28, "loss": 0.2099949499964714, "grad_norm": 0.40265214443206787, "train_seconds": 1580.383606878997, "remaining_budget_seconds": 1177.9382281599974}
+2026-10-17T04:14:57Z INFO {"event": "training_progress", "step": 10200, "epoch": 29, "loss": 0.2099622456729412, "grad_norm": 0.3233582377433777, "train_seconds": 1588.1237823580013, "remaining_budget_seconds": 1170.1980543149984}
+2026-10-17T04:15:08Z INFO {"event": "training_progress", "step": 10250, "epoch": 29, "loss": 0.20696513637900352, "grad_norm": 0.20458567142486572, "train_seconds": 1596.2306264529907, "remaining_budget_seconds": 1162.091205418008}
+2026-10-17T04:15:16Z INFO {"event": "training_progress", "step": 10300, "epoch": 29, "loss": 0.2093918165564537, "grad_norm": 0.1842639148235321, "train_seconds": 1604.1473213399877, "remaining_budget_seconds": 1154.1745136270038}
+2026-10-17T04:15:24Z INFO {"event": "training_progress", "step": 10350, "epoch": 29, "loss": 0.2082627146691084, "grad_norm": 0.21888545155525208, "train_seconds": 1611.891518532997, "remaining_budget_seconds": 1146.4303170740022}
+2026-10-17T04:15:32Z INFO {"event": "training_progress", "step": 10400, "epoch": 29, "loss": 0.20596413962543011, "grad_norm": 0.20765799283981323, "train_seconds": 1619.5792428369896, "remaining_budget_seconds": 1138.7425923950068}
+2026-10-17T04:15:43Z INFO {"event": "training_progress", "step": 10450, "epoch": 29, "loss": 0.20989764764904975, "grad_norm": 0.281139612197876, "train_seconds": 1627.532574058001, "remaining_budget_seconds": 1130.7892624390079}
+2026-10-17T04:15:51Z INFO {"event": "training_progress", "step": 10500, "epoch": 29, "loss": 0.20924369253218175, "grad_norm": 0.22210009396076202, "train_seconds": 1635.6696168239869, "remaining_budget_seconds": 1122.6522194860008}
+2026-10-17T04:15:59Z INFO {"event": "training_progress", "step": 10550, "epoch": 30, "loss": 0.21237926445901395, "grad_norm": 0.22749651968479156, "train_seconds": 1643.5468705599924, "remaining_budget_seconds": 1114.7749643340067}
+2026-10-17T04:16:06Z INFO {"event": "training_progress", "step": 10600, "epoch": 30, "loss": 0.205811255723238, "grad_norm": 0.2237374633550644, "train_seconds": 1651.209910733989, "remaining_budget_seconds": 1107.1119260900014}
+2026-10-17T04:16:17Z INFO {"event": "training_progress", "step": 10650, "epoch": 30, "loss": 0.20670947566628456, "grad_norm": 0.19510771334171295, "train_seconds": 1658.7235736619914, "remaining_budget_seconds": 1099.5982614240056}
+2026-10-17T04:16:25Z INFO {"event": "training_progress", "step": 10700, "epoch": 30, "loss": 0.20650099016726017, "grad_norm": 0.21572937071323395, "train_seconds": 1666.8297820879961, "remaining_budget_seconds": 1091.4920531310054}
+2026-10-17T04:16:33Z INFO {"event": "training_progress", "step": 10750, "epoch": 30, "loss": 0.206681310236454, "grad_norm": 0.21620577573776245, "train_seconds": 1674.3972846479883, "remaining_budget_seconds": 1083.9245492189948}
+2026-10-17T04:16:40Z INFO {"event": "training_progress", "step": 10800, "epoch": 30, "loss": 0.20788042820990085, "grad_norm": 0.204581618309021, "train_seconds": 1682.0134250939882, "remaining_budget_seconds": 1076.3084106089955}
+2026-10-17T04:16:48Z INFO {"event": "training_progress", "step": 10850, "epoch": 30, "loss": 0.20552432648837565, "grad_norm": 0.19663642346858978, "train_seconds": 1689.3982440779946, "remaining_budget_seconds": 1068.9235909380077}
+2026-10-17T04:16:59Z INFO {"event": "training_progress", "step": 10900, "epoch": 31, "loss": 0.2076883028447628, "grad_norm": 0.21407783031463623, "train_seconds": 1697.5942786089872, "remaining_budget_seconds": 1060.727553292003}
+2026-10-17T04:17:07Z INFO {"event": "training_progress", "step": 10950, "epoch": 31, "loss": 0.20501218989491463, "grad_norm": 0.313326358795166, "train_seconds": 1705.1818180299888, "remaining_budget_seconds": 1053.1400159379991}
+2026-10-17T04:17:14Z INFO {"event": "training_progress", "step": 11000, "epoch": 31, "loss": 0.20808349043130875, "grad_norm": 0.22111499309539795, "train_seconds": 1713.0668861969898, "remaining_budget_seconds": 1045.254948436006}
+2026-10-17T04:17:22Z INFO {"event": "training_progress", "step": 11050, "epoch": 31, "loss": 0.2075783322006464, "grad_norm": 0.1735968291759491, "train_seconds": 1720.641412853991, "remaining_budget_seconds": 1037.680424006001}
+2026-10-17T04:17:32Z INFO {"event": "training_progress", "step": 11100, "epoch": 31, "loss": 0.20652272462844848, "grad_norm": 0.3095574378967285, "train_seconds": 1728.3819353259896, "remaining_budget_seconds": 1029.939898919998}
+2026-10-17T04:17:40Z INFO {"event": "training_progress", "step": 11150, "epoch": 31, "loss": 0.204665200561285, "grad_norm": 0.2805726230144501, "train_seconds": 1735.8822853929887, "remaining_budget_seconds": 1022.4395499789971}
+2026-10-17T04:17:47Z INFO {"event": "training_progress", "step": 11200, "epoch": 31, "loss": 0.20487669840455056, "grad_norm": 0.18167486786842346, "train_seconds": 1743.237969017995, "remaining_budget_seconds": 1015.0838678510045}
+2026-10-17T04:17:55Z INFO {"event": "training_progress", "step": 11250, "epoch": 31, "loss": 0.20690724074840547, "grad_norm": 0.17999602854251862, "train_seconds": 1750.888376185001, "remaining_budget_seconds": 1007.4334591130028}
+2026-10-17T04:18:06Z INFO {"event": "training_progress", "step": 11300, "epoch": 32, "loss": 0.2058569233864546, "grad_norm": 0.22722332179546356, "train_seconds": 1758.8876233789924, "remaining_budget_seconds": 999.4342114720057}
+2026-10-17T04:18:14Z INFO {"event": "training_progress", "step": 11350, "epoch": 32, "loss": 0.2061556202918291, "grad_norm": 0.1712944656610489, "train_seconds": 1767.0462330929877, "remaining_budget_seconds": 991.2756026499992}
+2026-10-17T04:18:22Z INFO {"event": "training_progress", "step": 11400, "epoch": 32, "loss": 0.20515856318175793, "grad_norm": 0.29912757873535156, "train_seconds": 1775.2068917259894, "remaining_budget_seconds": 983.1149446380005}
+2026-10-17T04:18:30Z INFO {"event": "training_progress", "step": 11450, "epoch": 32, "loss": 0.2091503657400608, "grad_norm": 0.3186807930469513, "train_seconds": 1783.122672462996, "remaining_budget_seconds": 975.1991638720065}
+2026-10-17T04:18:42Z INFO {"event": "training_progress", "step": 11500, "epoch": 32, "loss": 0.20946510069072247, "grad_norm": 0.15606242418289185, "train_seconds": 1791.2882309749984, "remaining_budget_seconds": 967.0335994780035}
+2026-10-17T04:18:49Z INFO {"event": "training_progress", "step": 11550, "epoch": 32, "loss": 0.20702410086989403, "grad_norm": 0.2816813290119171, "train_seconds": 1798.8508195229952, "remaining_budget_seconds": 959.4710147669975}
+2026-10-17T04:18:57Z INFO {"event": "training_progress", "step": 11600, "epoch": 32, "loss": 0.20766438640654086, "grad_norm": 0.2217205911874771, "train_seconds": 1806.3688316359912, "remaining_budget_seconds": 951.953005332005}
+2026-10-17T04:19:04Z INFO {"event": "training_progress", "step": 11650, "epoch": 33, "loss": 0.20695838943123818, "grad_norm": 0.18589890003204346, "train_seconds": 1813.9847739539982, "remaining_budget_seconds": 944.337061615006}
+2026-10-17T04:19:15Z INFO {"event": "training_progress", "step": 11700, "epoch": 33, "loss": 0.20876008652150632, "grad_norm": 0.21559444069862366, "train_seconds": 1822.12098159999, "remaining_budget_seconds": 936.2008548000013}
+2026-10-17T04:19:23Z INFO {"event": "training_progress", "step": 11750, "epoch": 33, "loss": 0.20574559964239597, "grad_norm": 0.17442794144153595, "train_seconds": 1829.7554733939905, "remaining_budget_seconds": 928.5663635380042}
+2026-10-17T04:19:31Z INFO {"event": "training_progress", "step": 11800, "epoch": 33, "loss": 0.2107481199502945, "grad_norm": 0.21642164885997772, "train_seconds": 1837.6287685279967, "remaining_budget_seconds": 920.6930661500082}
+2026-10-17T04:19:39Z INFO {"event": "training_progress", "step": 11850, "epoch": 33, "loss": 0.20352963373064994, "grad_norm": 0.16930916905403137, "train_seconds": 1845.4342768589995, "remaining_budget_seconds": 912.8875583949994}
+2026-10-17T04:19:50Z INFO {"event": "training_progress", "step": 11900, "epoch": 33, "loss": 0.20694829009473323, "grad_norm": 0.13937650620937347, "train_seconds": 1853.2927069889993, "remaining_budget_seconds": 905.0291255159973}
+2026-10-17T04:19:58Z INFO {"event": "training_progress", "step": 11950, "epoch": 33, "loss": 0.2057694187760353, "grad_norm": 0.15992262959480286, "train_seconds": 1861.2606206989876, "remaining_budget_seconds": 897.0612150429952}
+2026-10-17T04:20:06Z INFO {"event": "training_progress", "step": 12000, "epoch": 34, "loss": 0.20988979294896126, "grad_norm": 0.22632242739200592, "train_seconds": 1869.2037155419966, "remaining_budget_seconds": 889.1181207680056}
+2026-10-17T04:20:13Z INFO {"event": "training_progress", "step": 12050, "epoch": 34, "loss": 0.20601653844118117, "grad_norm": 0.23349906504154205, "train_seconds": 1876.9566163540003, "remaining_budget_seconds": 881.3652190309949}
+2026-10-17T04:20:25Z INFO {"event": "training_progress", "step": 12100, "epoch": 34, "loss": 0.2067984552681446, "grad_norm": 0.1699834167957306, "train_seconds": 1885.2085546559974, "remaining_budget_seconds": 873.1132793249999}
+2026-10-17T04:20:32Z INFO {"event": "training_progress", "step": 12150, "epoch": 34, "loss": 0.20529855251312257, "grad_norm": 0.1913667619228363, "train_seconds": 1892.8957960529951, "remaining_budget_seconds": 865.4260408069968}
+2026-10-17T04:20:40Z INFO {"event": "training_progress", "step": 12200, "epoch": 34, "loss": 0.20446116544306278, "grad_norm": 0.24787098169326782, "train_seconds": 1900.8358690140012, "remaining_budget_seconds": 857.4859660609945}
+2026-10-17T04:20:48Z INFO {"event": "training_progress", "step": 12250, "epoch": 34, "loss": 0.20452409252524376, "grad_norm": 0.24422083795070648, "train_seconds": 1908.5492635049886, "remaining_budget_seconds": 849.7725713929976}
+2026-10-17T04:20:59Z INFO {"event": "training_progress", "step": 12300, "epoch": 34, "loss": 0.20638249725103378, "grad_norm": 0.2754463851451874, "train_seconds": 1916.6716891329997, "remaining_budget_seconds": 841.6501471499942}
+2026-10-17T04:21:08Z INFO {"event": "training_progress", "step": 12350, "epoch": 35, "loss": 0.20277065314352513, "grad_norm": 0.19966790080070496, "train_seconds": 1925.0466986329993, "remaining_budget_seconds": 833.2751338539965}
+2026-10-17T04:21:15Z INFO {"event": "training_progress", "step": 12400, "epoch": 35, "loss": 0.20755073621869088, "grad_norm": 0.18065983057022095, "train_seconds": 1932.9141016599897, "remaining_budget_seconds": 825.4077332519955}
+2026-10-17T04:21:23Z INFO {"event": "training_progress", "step": 12450, "epoch": 35, "loss": 0.20402595899999143, "grad_norm": 0.3261452615261078, "train_seconds": 1940.4459463499952, "remaining_budget_seconds": 817.8758894009952}
+2026-10-17T04:21:34Z INFO {"event": "training_progress", "step": 12500, "epoch": 35, "loss": 0.20431903272867202, "grad_norm": 0.2531786561012268, "train_seconds": 1948.1077541699924, "remaining_budget_seconds": 810.2140825300012}
+2026-10-17T04:21:42Z INFO {"event": "training_progress", "step": 12550, "epoch": 35, "loss": 0.20418029420077802, "grad_norm": 0.20533309876918793, "train_seconds": 1955.977107062994, "remaining_budget_seconds": 802.3447283550049}
+2026-10-17T04:21:49Z INFO {"event": "training_progress", "step": 12600, "epoch": 35, "loss": 0.2036767289042473, "grad_norm": 0.2080046832561493, "train_seconds": 1963.7016508719971, "remaining_budget_seconds": 794.6201856359985}
+2026-10-17T04:21:57Z INFO {"event": "training_progress", "step": 12650, "epoch": 35, "loss": 0.20317092038691043, "grad_norm": 0.2522684931755066, "train_seconds": 1971.5902848419937, "remaining_budget_seconds": 786.7315514590009}
+2026-10-17T04:22:05Z INFO {"event": "training_progress", "step": 12700, "epoch": 36, "loss": 0.20515918917953968, "grad_norm": 0.1660490781068802, "train_seconds": 1979.5348472969927, "remaining_budget_seconds": 778.7869891330047}
+2026-10-17T04:22:16Z INFO {"event": "training_progress", "step": 12750, "epoch": 36, "loss": 0.20194629967212677, "grad_norm": 0.2054073065519333, "train_seconds": 1987.750192513995, "remaining_budget_seconds": 770.5716428949963}
+2026-10-17T04:22:24Z INFO {"event": "training_progress", "step": 12800, "epoch": 36, "loss": 0.20454927355051042, "grad_norm": 0.17777623236179352, "train_seconds": 1995.4208752429986, "remaining_budget_seconds": 762.9009574869997}
+2026-10-17T04:22:32Z INFO {"event": "training_progress", "step": 12850, "epoch": 36, "loss": 0.2042512045800686, "grad_norm": 0.13130199909210205, "train_seconds": 2003.3077001519996, "remaining_budget_seconds": 755.0141367449978}
+2026-10-17T04:22:40Z INFO {"event": "training_progress", "step": 12900, "epoch": 36, "loss": 0.20236358486115932, "grad_norm": 0.23816508054733276, "train_seconds": 2010.9477503909875, "remaining_budget_seconds": 747.3740854159987}
+2026-10-17T04:22:51Z INFO {"event": "training_progress", "step": 12950, "epoch": 36, "loss": 0.20844044648110865, "grad_norm": 0.26335757970809937, "train_seconds": 2019.3843142409896, "remaining_budget_seconds": 738.9375218479981}
+2026-10-17T04:22:59Z INFO {"event": "training_progress", "step": 13000, "epoch": 36, "loss": 0.20223753973841668, "grad_norm": 0.129208043217659, "train_seconds": 2027.1994108029903, "remaining_budget_seconds": 731.1224250850064}
+2026-10-17T04:23:07Z INFO {"event": "training_progress", "step": 13050, "epoch": 37, "loss": 0.20483313404023648, "grad_norm": 0.2250521034002304, "train_seconds": 2034.9790891579905, "remaining_budget_seconds": 723.3427457399957}
+2026-10-17T04:23:15Z INFO {"event": "training_progress", "step": 13100, "epoch": 37, "loss": 0.20396253667771816, "grad_norm": 0.18548943102359772, "train_seconds": 2042.9783738659899, "remaining_budget_seconds": 715.343460974007}
+2026-10-17T04:23:25Z INFO {"event": "training_progress", "step": 13150, "epoch": 37, "loss": 0.204786242172122, "grad_norm": 0.14071564376354218, "train_seconds": 2051.13507714399, "remaining_budget_seconds": 707.1867555320059}
+2026-10-17T04:23:33Z INFO {"event": "training_progress", "step": 13200, "epoch": 37, "loss": 0.20380452446639538, "grad_norm": 0.29681527614593506, "train_seconds": 2058.7011598339886, "remaining_budget_seconds": 699.6206738790061}
+2026-10-17T04:23:41Z INFO {"event": "training_progress", "step": 13250, "epoch": 37, "loss": 0.2058169861882925, "grad_norm": 0.22865130007266998, "train_seconds": 2066.333674921989, "remaining_budget_seconds": 691.9881624429981}
+2026-10-17T04:23:48Z INFO {"event": "training_progress", "step": 13300, "epoch": 37, "loss": 0.20265643030405045, "grad_norm": 0.3003791868686676, "train_seconds": 2073.9614632059966, "remaining_budget_seconds": 684.3603563240031}
+2026-10-17T04:24:00Z INFO {"event": "training_progress", "step": 13350, "epoch": 37, "loss": 0.2039161168038845, "grad_norm": 0.1822301596403122, "train_seconds": 2081.953406296001, "remaining_budget_seconds": 676.3684002439986}
+2026-10-17T04:24:07Z INFO {"event": "training_progress", "step": 13400, "epoch": 38, "loss": 0.2051972659677267, "grad_norm": 0.2477494478225708, "train_seconds": 2089.7842302520003, "remaining_budget_seconds": 668.5376027260063}
+2026-10-17T04:24:15Z INFO {"event": "training_progress", "step": 13450, "epoch": 38, "loss": 0.20415040329098702, "grad_norm": 0.21553632616996765, "train_seconds": 2097.7590171749907, "remaining_budget_seconds": 660.5628194019955}
+2026-10-17T04:24:23Z INFO {"event": "training_progress", "step": 13500, "epoch": 38, "loss": 0.20891239523887634, "grad_norm": 0.16984611749649048, "train_seconds": 2105.455898643995, "remaining_budget_seconds": 652.8659378400043}
+2026-10-17T04:24:34Z INFO {"event": "training_progress", "step": 13550, "epoch": 38, "loss": 0.20486515410244466, "grad_norm": 0.2042544186115265, "train_seconds": 2113.23648927499, "remaining_budget_seconds": 645.0853461420047}
+2026-10-17T04:24:42Z INFO {"event": "training_progress", "step": 13600, "epoch": 38, "loss": 0.2046149132400751, "grad_norm": 0.1931714415550232, "train_seconds": 2121.2003271759895, "remaining_budget_seconds": 637.1215074189968}
+2026-10-17T04:24:50Z INFO {"event": "training_progress", "step": 13650, "epoch": 38, "loss": 0.2048764503747225, "grad_norm": 0.22288557887077332, "train_seconds": 2128.9336061509966, "remaining_budget_seconds": 629.3882286890002}
+2026-10-17T04:24:58Z INFO {"event": "training_progress", "step": 13700, "epoch": 38, "loss": 0.2040010969340801, "grad_norm": 0.34049686789512634, "train_seconds": 2137.052007539998, "remaining_budget_seconds": 621.2698291700071}
+2026-10-17T04:25:09Z INFO {"event": "training_progress", "step": 13750, "epoch": 39, "loss": 0.20148540429770948, "grad_norm": 0.1756831258535385, "train_seconds": 2144.9882336169976, "remaining_budget_seconds": 613.3335996999958}
+2026-10-17T04:25:16Z INFO {"event": "training_progress", "step": 13800, "epoch": 39, "loss": 0.20325117126107217, "grad_norm": 0.21022135019302368, "train_seconds": 2152.6422894069983, "remaining_budget_seconds": 605.6795429159974}
+2026-10-17T04:25:24Z INFO {"event": "training_progress", "step": 13850, "epoch": 39, "loss": 0.20132556043565272, "grad_norm": 0.1752561628818512, "train_seconds": 2160.2498112479952, "remaining_budget_seconds": 598.0720238039939}
+2026-10-17T04:25:32Z INFO {"event": "training_progress", "step": 13900, "epoch": 39, "loss": 0.207040508762002, "grad_norm": 0.2137463241815567, "train_seconds": 2167.9320212919993, "remaining_budget_seconds": 590.3898133949988}
+2026-10-17T04:25:43Z INFO {"event": "training_progress", "step": 13950, "epoch": 39, "loss": 0.20194717541337012, "grad_norm": 0.2540922164916992, "train_seconds": 2175.7053182859963, "remaining_budget_seconds": 582.6165185340069}
+2026-10-17T04:25:50Z INFO {"event": "training_progress", "step": 14000, "epoch": 39, "loss": 0.20384908474981786, "grad_norm": 0.1525818109512329, "train_seconds": 2183.4481048499874, "remaining_budget_seconds": 574.8737309999997}
+2026-10-17T04:25:58Z INFO {"event": "training_progress", "step": 14050, "epoch": 39, "loss": 0.20406619407236576, "grad_norm": 0.18791253864765167, "train_seconds": 2191.3946628609992, "remaining_budget_seconds": 566.9271738560055}
+2026-10-17T04:26:06Z INFO {"event": "training_progress", "step": 14100, "epoch": 40, "loss": 0.20187776669859886, "grad_norm": 0.1721971482038498, "train_seconds": 2199.3852887039975, "remaining_budget_seconds": 558.9365461760026}
+2026-10-17T04:26:17Z INFO {"event": "training_progress", "step": 14150, "epoch": 40, "loss": 0.2046895184367895, "grad_norm": 0.2355424016714096, "train_seconds": 2207.259818074992, "remaining_budget_seconds": 551.0620186720043}
+2026-10-17T04:26:25Z INFO {"event": "training_progress", "step": 14200, "epoch": 40, "loss": 0.20306445240974427, "grad_norm": 0.23813016712665558, "train_seconds": 2215.050344311996, "remaining_budget_seconds": 543.2714907680056}
+2026-10-17T04:26:33Z INFO {"event": "training_progress", "step": 14250, "epoch": 40, "loss": 0.20454083062708378, "grad_norm": 0.13687604665756226, "train_seconds": 2222.9779303789983, "remaining_budget_seconds": 535.3439055449999}
+2026-10-17T04:26:41Z INFO {"event": "training_progress", "step": 14300, "epoch": 40, "loss": 0.2022160694003105, "grad_norm": 0.14537237584590912, "train_seconds": 2230.9506242200005, "remaining_budget_seconds": 527.3712126010068}
+2026-10-17T04:26:52Z INFO {"event": "training_progress", "step": 14350, "epoch": 40, "loss": 0.2015623326599598, "grad_norm": 0.13822270929813385, "train_seconds": 2238.8879786049947, "remaining_budget_seconds": 519.4338569789979}
+2026-10-17T04:27:00Z INFO {"event": "training_progress", "step": 14400, "epoch": 40, "loss": 0.2024122104048729, "grad_norm": 0.28817880153656006, "train_seconds": 2247.117988262995, "remaining_budget_seconds": 511.2038474590081}
+2026-10-17T04:27:08Z INFO {"event": "training_progress", "step": 14450, "epoch": 41, "loss": 0.20500035412609577, "grad_norm": 0.25615260004997253, "train_seconds": 2254.9260932139878, "remaining_budget_seconds": 503.3957404420071}
+2026-10-17T04:27:16Z INFO {"event": "training_progress", "step": 14500, "epoch": 41, "loss": 0.2048893515765667, "grad_norm": 0.1311032772064209, "train_seconds": 2262.6469923449913, "remaining_budget_seconds": 495.67484441200213}
+2026-10-17T04:27:23Z INFO {"event": "training_progress", "step": 14550, "epoch": 41, "loss": 0.2033272498846054, "grad_norm": 0.17486201226711273, "train_seconds": 2270.3666347729886, "remaining_budget_seconds": 487.9551996109949}
+2026-10-17T04:27:34Z INFO {"event": "training_progress", "step": 14600, "epoch": 41, "loss": 0.2048669384419918, "grad_norm": 0.16264063119888306, "train_seconds": 2278.130419590001, "remaining_budget_seconds": 480.19141518000106}
+2026-10-17T04:27:42Z INFO {"event": "training_progress", "step": 14650, "epoch": 41, "loss": 0.20339644126594067, "grad_norm": 0.20440377295017242, "train_seconds": 2285.9361563529965, "remaining_budget_seconds": 472.3856804810057}
+2026-10-17T04:27:50Z INFO {"event": "training_progress", "step": 14700, "epoch": 41, "loss": 0.20104929439723493, "grad_norm": 0.19542422890663147, "train_seconds": 2293.6681576139963, "remaining_budget_seconds": 464.65367753000464}
+2026-10-17T04:27:58Z INFO {"event": "training_progress", "step": 14750, "epoch": 41, "loss": 0.20425425708293915, "grad_norm": 0.19769257307052612, "train_seconds": 2301.436949689989, "remaining_budget_seconds": 456.8848857109988}
+2026-10-17T04:28:09Z INFO {"event": "training_progress", "step": 14800, "epoch": 42, "loss": 0.20421387903392316, "grad_norm": 0.24155256152153015, "train_seconds": 2309.259062557001, "remaining_budget_seconds": 449.0627709959954}
+2026-10-17T04:28:16Z INFO {"event": "training_progress", "step": 14850, "epoch": 42, "loss": 0.20276544675230979, "grad_norm": 0.27624017000198364, "train_seconds": 2317.0282936779986, "remaining_budget_seconds": 441.2935431030055}
+2026-10-17T04:28:24Z INFO {"event": "training_progress", "step": 14900, "epoch": 42, "loss": 0.20225549310445787, "grad_norm": 0.23883283138275146, "train_seconds": 2324.9405534299876, "remaining_budget_seconds": 433.38128143100766}
+2026-10-17T04:28:32Z INFO {"event": "training_progress", "step": 14950, "epoch": 42, "loss": 0.20520699240267276, "grad_norm": 0.1793350726366043, "train_seconds": 2332.672629513996, "remaining_budget_seconds": 425.6492056040006}
+2026-10-17T04:28:43Z INFO {"event": "training_progress", "step": 15000, "epoch": 42, "loss": 0.2052006187289953, "grad_norm": 0.20015664398670197, "train_seconds": 2340.470110230992, "remaining_budget_seconds": 417.8517265279952}
+2026-10-17T04:28:51Z INFO {"event": "training_progress", "step": 15050, "epoch": 42, "loss": 0.20337512098252775, "grad_norm": 0.22750115394592285, "train_seconds": 2348.22815455799, "remaining_budget_seconds": 410.0936817490001}
+2026-10-17T04:28:58Z INFO {"event": "training_progress", "step": 15100, "epoch": 42, "loss": 0.2040988090634346, "grad_norm": 0.14135335385799408, "train_seconds": 2355.8829823319975, "remaining_budget_seconds": 402.43885426099587}
+2026-10-17T04:29:06Z INFO {"event": "training_progress", "step": 15150, "epoch": 43, "loss": 0.2008038141578436, "grad_norm": 0.1750342696905136, "train_seconds": 2363.7299181699927, "remaining_budget_seconds": 394.59191855099925}
+2026-10-17T04:29:17Z INFO {"event": "training_progress", "step": 15200, "epoch": 43, "loss": 0.2034083677828312, "grad_norm": 0.16881364583969116, "train_seconds": 2371.675510604997, "remaining_budget_seconds": 386.64632511200034}
+2026-10-17T04:29:25Z INFO {"event": "training_progress", "step": 15250, "epoch": 43, "loss": 0.2055370032787323, "grad_norm": 0.19815604388713837, "train_seconds": 2379.6708797149913, "remaining_budget_seconds": 378.650954961995}
+2026-10-17T04:29:33Z INFO {"event": "training_progress", "step": 15300, "epoch": 43, "loss": 0.20088803246617318, "grad_norm": 0.2271074801683426, "train_seconds": 2387.6892739569885, "remaining_budget_seconds": 370.6325599910051}
+2026-10-17T04:29:41Z INFO {"event": "training_progress", "step": 15350, "epoch": 43, "loss": 0.20149704985320568, "grad_norm": 0.13866619765758514, "train_seconds": 2395.3912523939944, "remaining_budget_seconds": 362.9305845570052}
+2026-10-17T04:29:52Z INFO {"event": "training_progress", "step": 15400, "epoch": 43, "loss": 0.20255502715706825, "grad_norm": 0.14590679109096527, "train_seconds": 2402.996528179996, "remaining_budget_seconds": 355.32530879099795}
+2026-10-17T04:29:59Z INFO {"event": "training_progress", "step": 15450, "epoch": 43, "loss": 0.2009129036962986, "grad_norm": 0.17647479474544525, "train_seconds": 2410.783432414988, "remaining_budget_seconds": 347.53840198800026}
+2026-10-17T04:30:07Z INFO {"event": "training_progress", "step": 15500, "epoch": 44, "loss": 0.20324138812720777, "grad_norm": 0.19626981019973755, "train_seconds": 2418.4476623709925, "remaining_budget_seconds": 339.87417395299417}
+2026-10-17T04:30:15Z INFO {"event": "training_progress", "step": 15550, "epoch": 44, "loss": 0.20036308720707893, "grad_norm": 0.24269011616706848, "train_seconds": 2426.329693290987, "remaining_budget_seconds": 331.9921437659941}
+2026-10-17T04:30:26Z INFO {"event": "training_progress", "step": 15600, "epoch": 44, "loss": 0.20179530426859857, "grad_norm": 0.16291356086730957, "train_seconds": 2434.2034384579893, "remaining_budget_seconds": 324.11839539899665}
+2026-10-17T04:30:34Z INFO {"event": "training_progress", "step": 15650, "epoch": 44, "loss": 0.202247779071331, "grad_norm": 0.21417425572872162, "train_seconds": 2441.9946632879874, "remaining_budget_seconds": 316.32716944700223}
+2026-10-17T04:30:41Z INFO {"event": "training_progress", "step": 15700, "epoch": 44, "loss": 0.20176410645246506, "grad_norm": 0.16213227808475494, "train_seconds": 2449.573057347996, "remaining_budget_seconds": 308.7487779009971}
+2026-10-17T04:30:49Z INFO {"event": "training_progress", "step": 15750, "epoch": 44, "loss": 0.20806347377598286, "grad_norm": 0.23035374283790588, "train_seconds": 2457.2221401939896, "remaining_budget_seconds": 301.0996963420039}
+2026-10-17T04:31:00Z INFO {"event": "training_progress", "step": 15800, "epoch": 44, "loss": 0.20325741097331046, "grad_norm": 0.20756259560585022, "train_seconds": 2465.0042803359975, "remaining_budget_seconds": 293.31755289199646}
+2026-10-17T04:31:08Z INFO {"event": "training_progress", "step": 15850, "epoch": 45, "loss": 0.2037798783183098, "grad_norm": 0.19836583733558655, "train_seconds": 2472.750122616999, "remaining_budget_seconds": 285.5717124799994}
+2026-10-17T04:31:16Z INFO {"event": "training_progress", "step": 15900, "epoch": 45, "loss": 0.19847053959965705, "grad_norm": 0.17224258184432983, "train_seconds": 2480.584320366994, "remaining_budget_seconds": 277.7375153720059}
+2026-10-17T04:31:23Z INFO {"event": "training_progress", "step": 15950, "epoch": 45, "loss": 0.20316697962582111, "grad_norm": 0.15670692920684814, "train_seconds": 2488.5218736349925, "remaining_budget_seconds": 269.79996086700703}
+2026-10-17T04:31:35Z INFO {"event": "training_progress", "step": 16000, "epoch": 45, "loss": 0.20343928471207617, "grad_norm": 0.1294861137866974, "train_seconds": 2496.461160578998, "remaining_budget_seconds": 261.8606759450049}
+2026-10-17T04:31:42Z INFO {"event": "training_progress", "step": 16050, "epoch": 45, "loss": 0.20192311391234397, "grad_norm": 0.1697823703289032, "train_seconds": 2503.9899668289872, "remaining_budget_seconds": 254.33186944000772}
+2026-10-17T04:31:50Z INFO {"event": "training_progress", "step": 16100, "epoch": 45, "loss": 0.20183641038835048, "grad_norm": 0.20433315634727478, "train_seconds": 2511.701557215987, "remaining_budget_seconds": 246.6202792510012}
+2026-10-17T04:31:58Z INFO {"event": "training_progress", "step": 16150, "epoch": 45, "loss": 0.2040775366127491, "grad_norm": 0.14372092485427856, "train_seconds": 2519.480498356992, "remaining_budget_seconds": 238.84133441200538}
+2026-10-17T04:32:06Z INFO {"event": "training_progress", "step": 16200, "epoch": 46, "loss": 0.2047281301766634, "grad_norm": 0.26334816217422485, "train_seconds": 2527.6802643509873, "remaining_budget_seconds": 230.64157101999444}
+2026-10-17T04:32:17Z INFO {"event": "training_progress", "step": 16250, "epoch": 46, "loss": 0.20522519573569298, "grad_norm": 0.26509347558021545, "train_seconds": 2535.519028685987, "remaining_budget_seconds": 222.80280366299849}
+2026-10-17T04:32:24Z INFO {"event": "training_progress", "step": 16300, "epoch": 46, "loss": 0.20482657231390478, "grad_norm": 0.2136716991662979, "train_seconds": 2543.251587681996, "remaining_budget_seconds": 215.07024922900018}
+2026-10-17T04:32:32Z INFO {"event": "training_progress", "step": 16350, "epoch": 46, "loss": 0.1977387433499098, "grad_norm": 0.12940342724323273, "train_seconds": 2551.1493304039905, "remaining_budget_seconds": 207.172504896007}
+2026-10-17T04:32:40Z INFO {"event": "training_progress", "step": 16400, "epoch": 46, "loss": 0.20379791855812074, "grad_norm": 0.19397981464862823, "train_seconds": 2558.9351640989917, "remaining_budget_seconds": 199.38667147500382}
+2026-10-17T04:32:51Z INFO {"event": "training_progress", "step": 16450, "epoch": 46, "loss": 0.2044042294472456, "grad_norm": 0.16187992691993713, "train_seconds": 2566.7365628549887, "remaining_budget_seconds": 191.58527328900527}
+2026-10-17T04:32:59Z INFO {"event": "training_progress", "step": 16500, "epoch": 46, "loss": 0.19983828254044056, "grad_norm": 0.13921356201171875, "train_seconds": 2574.777229204992, "remaining_budget_seconds": 183.54460462700808}
+2026-10-17T04:33:07Z INFO {"event": "training_progress", "step": 16550, "epoch": 47, "loss": 0.2018457204848528, "grad_norm": 0.13830426335334778, "train_seconds": 2582.4592588109954, "remaining_budget_seconds": 175.86257793099503}
+2026-10-17T04:33:14Z INFO {"event": "training_progress", "step": 16600, "epoch": 47, "loss": 0.20291364520788194, "grad_norm": 0.16542242467403412, "train_seconds": 2589.9993136659905, "remaining_budget_seconds": 168.3225211550016}
+2026-10-17T04:33:25Z INFO {"event": "training_progress", "step": 16650, "epoch": 47, "loss": 0.20187486797571183, "grad_norm": 0.23571257293224335, "train_seconds": 2597.6685390519997, "remaining_budget_seconds": 160.65329634900263}
+2026-10-17T04:33:33Z INFO {"event": "training_progress", "step": 16700, "epoch": 47, "loss": 0.20129220075905324, "grad_norm": 0.2754823565483093, "train_seconds": 2605.3209184129955, "remaining_budget_seconds": 153.00091839600645}
+2026-10-17T04:33:40Z INFO {"event": "training_progress", "step": 16750, "epoch": 47, "loss": 0.2037076147645712, "grad_norm": 0.2162092626094818, "train_seconds": 2612.9154628739925, "remaining_budget_seconds": 145.4063703569991}
+2026-10-17T04:33:48Z INFO {"event": "training_progress", "step": 16800, "epoch": 47, "loss": 0.20377071656286716, "grad_norm": 0.26763251423835754, "train_seconds": 2620.7856999270007, "remaining_budget_seconds": 137.53613552100433}
+2026-10-17T04:33:59Z INFO {"event": "training_progress", "step": 16850, "epoch": 47, "loss": 0.20208512231707573, "grad_norm": 0.19020387530326843, "train_seconds": 2628.685072424996, "remaining_budget_seconds": 129.6366234110028}
+2026-10-17T04:34:08Z INFO {"event": "training_progress", "step": 16900, "epoch": 48, "loss": 0.2041041664034128, "grad_norm": 0.14761796593666077, "train_seconds": 2636.876884833997, "remaining_budget_seconds": 121.44495210700552}
+2026-10-17T04:34:15Z INFO {"event": "training_progress", "step": 16950, "epoch": 48, "loss": 0.20115386255085468, "grad_norm": 0.1790865659713745, "train_seconds": 2644.573771516996, "remaining_budget_seconds": 113.74806341000658}
+2026-10-17T04:34:23Z INFO {"event": "training_progress", "step": 17000, "epoch": 48, "loss": 0.1986386686563492, "grad_norm": 0.15265612304210663, "train_seconds": 2652.6046220989956, "remaining_budget_seconds": 105.71721286399406}
+2026-10-17T04:34:34Z INFO {"event": "training_progress", "step": 17050, "epoch": 48, "loss": 0.20079386129975318, "grad_norm": 0.14744126796722412, "train_seconds": 2660.3530424489873, "remaining_budget_seconds": 97.96879400699981}
+2026-10-17T04:34:42Z INFO {"event": "training_progress", "step": 17100, "epoch": 48, "loss": 0.20248331934213637, "grad_norm": 0.20612840354442596, "train_seconds": 2668.344439923996, "remaining_budget_seconds": 89.97739742300473}
+2026-10-17T04:34:50Z INFO {"event": "training_progress", "step": 17150, "epoch": 48, "loss": 0.20396406933665276, "grad_norm": 0.2249777615070343, "train_seconds": 2676.349584809999, "remaining_budget_seconds": 81.97225012999843}
+2026-10-17T04:34:58Z INFO {"event": "training_progress", "step": 17200, "epoch": 48, "loss": 0.20195546470582484, "grad_norm": 0.20627085864543915, "train_seconds": 2683.939542588996, "remaining_budget_seconds": 74.38229409100313}
+2026-10-17T04:35:09Z INFO {"event": "training_progress", "step": 17250, "epoch": 49, "loss": 0.2005927813053131, "grad_norm": 0.15732496976852417, "train_seconds": 2691.8102008229907, "remaining_budget_seconds": 66.51163395399635}
+2026-10-17T04:35:17Z INFO {"event": "training_progress", "step": 17300, "epoch": 49, "loss": 0.20044718973338604, "grad_norm": 0.18730229139328003, "train_seconds": 2699.6349081929948, "remaining_budget_seconds": 58.686926377005875}
+2026-10-17T04:35:17Z INFO {"event": "training_finished", "status": "completed", "dataset": "cifar10", "family": "dit", "seed": 102, "step": 17303, "train_seconds": 2700.0645746579976, "parameters": 1298864, "device": "cuda", "initial_validation_mse": 1.7011075615882874, "final_validation_mse": 0.1970291268080473, "images_per_second": 820.2707523321121, "source_revisions": ["c749856d1c9fe07d6ce2f39d01de78f664aea9a8"], "peak_memory_bytes": 161050112, "checkpoint_sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15", "failure": null}

study/main/cifar10-dit-s102/training.json

@@ -0,0 +1,19 @@
+{
+  "status": "completed",
+  "dataset": "cifar10",
+  "family": "dit",
+  "seed": 102,
+  "step": 17303,
+  "train_seconds": 2700.0645746579976,
+  "parameters": 1298864,
+  "device": "cuda",
+  "initial_validation_mse": 1.7011075615882874,
+  "final_validation_mse": 0.1970291268080473,
+  "images_per_second": 820.2707523321121,
+  "source_revisions": [
+    "c749856d1c9fe07d6ce2f39d01de78f664aea9a8"
+  ],
+  "peak_memory_bytes": 161050112,
+  "checkpoint_sha256": "a87b3e659c6398bf2b2791e05297293a8c154ada863778310148018b4f940c15",
+  "failure": null
+}

study/main/cifar10-dit-s103/README.md

@@ -0,0 +1,30 @@
+---
+license: mit
+pipeline_tag: unconditional-image-generation
+tags:
+- pytorch
+- flow-matching
+- oscillatory-neural-networks
+---
+# cifar10 / dit / seed 103
+
+This is an unconditional pixel-space velocity network from the [oscillatory flow matching study](https://github.com/BurnyCoder/flow-matching-with-oscillatory-neural-networks-instead-of-transformer).
+Recorded role: **completed**; status: **completed**. The model is a research artifact; sample quality and limitations are documented in its accompanying evaluations and the public reports. Fixed-batch diagnostics deliberately memorize a few examples and are not general-purpose generators. No performance advantage is assumed.
+
+## Loading and sampling
+
+Install the locked environment from the source repository. Download this entire run subdirectory, including `config.json` and `model.safetensors`, then use:
+
+```bash
+ofm sample --run /path/to/downloaded/run --device cpu --count 64 --nfe 32
+```
+
+The public Python loader is `ofm.checkpoints.load_generator(path, device="cpu")`; models accept `forward(images, times)` and return image-shaped velocities. This export contains **EMA weights**. `checkpoint.pt` additionally preserves the available training/diagnostic state; ordinary training checkpoints are resumable with `ofm train --resume /path/to/run` after downloading the official dataset. Untrained and fixed-batch diagnostic checkpoints do not contain a resumable ordinary training loop.
+
+## Training and evaluation
+
+Dataset: cifar10; architecture: dit; seed: 103. This is a from-scratch generator, with no text conditioning or pretrained image encoder. The exact architecture, optimizer, data manifests, source revisions, and metrics are included as JSON. Generated tensors are complete and unclipped; PNG grids show the first fixed samples after display clipping. KID/FID are CIFAR diagnostics and are not classification accuracy. Untrained, fixed-batch, or step-limited checkpoints are explicitly labeled and must not be treated as final-budget primary results.
+
+## Limitations and provenance
+
+Small native-resolution datasets and a short local compute budget limit image quality and external generalization. Six paired seeds are exploratory; inference interventions alone do not establish an oscillator mechanism. Dataset content and evaluator biases remain relevant. See the [protocol](https://github.com/BurnyCoder/flow-matching-with-oscillatory-neural-networks-instead-of-transformer/blob/main/docs/protocol.md) and [reports](https://github.com/BurnyCoder/flow-matching-with-oscillatory-neural-networks-instead-of-transformer/tree/main/reports) for all results, including failures. This export contains no downloaded dataset or credentials. `manifest.json` records file hashes; the original checkpoint SHA-256 is `e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8`. Export code revision: `13c72f9474e2fb46be8f209e15dba220f91ef4d5`.

study/main/cifar10-dit-s103/checkpoint.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8
+size 21305024

study/main/cifar10-dit-s103/config.json

@@ -0,0 +1,32 @@
+{
+  "model": {
+    "family": "dit",
+    "channels": 3,
+    "size": 32,
+    "width": 128,
+    "depth": 4,
+    "patch": 4,
+    "rotating_dim": 4,
+    "inner_steps": 4,
+    "step_size": 0.1,
+    "ffn_hidden": 576
+  },
+  "dataset": "cifar10",
+  "seed": 103,
+  "device": "cuda",
+  "batch_size": 128,
+  "microbatch": 32,
+  "learning_rate": 0.0003,
+  "weight_decay": 0.01,
+  "clip_grad": 1.0,
+  "ema_decay": 0.999,
+  "minutes": 45,
+  "max_steps": 0,
+  "validation_size": 5000,
+  "data_root": "data",
+  "runs_root": "runs",
+  "budget_hours": 24.0,
+  "log_every": 50,
+  "checkpoint_seconds": 300.0,
+  "amp": true
+}

study/main/cifar10-dit-s103/environment.json

@@ -0,0 +1,130 @@
+{
+  "python": "3.12.3 (main, Aug 31 2026, 10:18:26) [GCC 13.3.0]",
+  "torch": "2.11.0+cu128",
+  "cuda_build": "12.8",
+  "device": "NVIDIA GeForce RTX 5070 Laptop GPU",
+  "git_revision": "c749856d1c9fe07d6ce2f39d01de78f664aea9a8",
+  "deterministic_algorithms": true,
+  "float32_matmul_precision": "high",
+  "packages": {
+    "Jinja2": "3.1.6",
+    "MarkupSafe": "3.0.4",
+    "POT": "0.9.7.post1",
+    "PyYAML": "6.0.3",
+    "Pygments": "2.21.0",
+    "absl-py": "2.5.1",
+    "aiohappyeyeballs": "2.7.1",
+    "aiohttp": "3.14.5",
+    "aiosignal": "1.4.0",
+    "anyio": "4.15.1",
+    "asttokens": "3.0.2",
+    "attrs": "26.1.0",
+    "click": "8.5.0",
+    "cloudpickle": "3.1.2",
+    "comm": "0.2.3",
+    "contourpy": "1.4.0",
+    "cuda-bindings": "12.9.10",
+    "cuda-pathfinder": "1.8.3",
+    "cuda-toolkit": "12.8.1",
+    "cycler": "0.12.1",
+    "debugpy": "1.8.22",
+    "executing": "2.3.0",
+    "filelock": "4.2.0",
+    "fonttools": "4.67.0",
+    "frozenlist": "1.8.0",
+    "fsspec": "2026.9.0",
+    "h11": "0.16.0",
+    "hf-xet": "1.7.0",
+    "httpcore2": "2.13.1",
+    "httpx2": "2.13.1",
+    "huggingface_hub": "2.2.0",
+    "idna": "3.20",
+    "iniconfig": "2.3.1",
+    "ipykernel": "7.4.0",
+    "ipython": "9.17.1",
+    "ipython_pygments_lexers": "1.1.1",
+    "ipywidgets": "8.1.9",
+    "jedi": "0.20.1",
+    "joblib": "1.6.0",
+    "jupyter_client": "8.10.0",
+    "jupyter_core": "5.9.1",
+    "jupyterlab_widgets": "3.0.17",
+    "kiwisolver": "1.5.1",
+    "lightning-utilities": "0.15.3",
+    "matplotlib": "3.11.2",
+    "matplotlib-inline": "0.2.2",
+    "mpmath": "1.3.0",
+    "multidict": "7.1.0",
+    "narwhals": "2.27.1",
+    "nest-asyncio2": "1.7.4",
+    "networkx": "3.7",
+    "numpy": "2.5.4",
+    "nvidia-cublas-cu12": "12.8.4.1",
+    "nvidia-cuda-cupti-cu12": "12.8.90",
+    "nvidia-cuda-nvrtc-cu12": "12.8.93",
+    "nvidia-cuda-runtime-cu12": "12.8.90",
+    "nvidia-cudnn-cu12": "9.19.0.56",
+    "nvidia-cufft-cu12": "11.3.3.83",
+    "nvidia-cufile-cu12": "1.13.1.3",
+    "nvidia-curand-cu12": "10.3.9.90",
+    "nvidia-cusolver-cu12": "11.7.3.90",
+    "nvidia-cusparse-cu12": "12.5.8.93",
+    "nvidia-cusparselt-cu12": "0.7.1",
+    "nvidia-nccl-cu12": "2.28.9",
+    "nvidia-nvjitlink-cu12": "12.8.93",
+    "nvidia-nvshmem-cu12": "3.4.5",
+    "nvidia-nvtx-cu12": "12.8.90",
+    "oscillatory-flow-matching": "0.1.0",
+    "packaging": "26.3",
+    "pandas": "3.0.6",
+    "parso": "0.8.7",
+    "pastel": "0.2.1",
+    "pexpect": "4.9.0",
+    "pillow": "12.3.0",
+    "platformdirs": "4.13.3",
+    "pluggy": "1.7.0",
+    "poethepoet": "0.10.0",
+    "prompt_toolkit": "3.0.53",
+    "propcache": "0.5.4",
+    "psutil": "7.2.2",
+    "ptyprocess": "0.7.0",
+    "pure_eval": "0.2.4",
+    "pypandoc_binary": "1.17",
+    "pyparsing": "3.3.3",
+    "pytest": "9.1.1",
+    "python-dateutil": "2.9.0.post0",
+    "python-dotenv": "1.2.4",
+    "pytorch-lightning": "2.6.6",
+    "pyzmq": "27.2.0",
+    "ruff": "0.17.1",
+    "safetensors": "0.8.0",
+    "scikit-learn": "1.9.1",
+    "scipy": "1.18.1",
+    "setuptools": "81.0.0",
+    "six": "1.17.0",
+    "stack-data": "0.6.3",
+    "sympy": "1.14.0",
+    "threadpoolctl": "3.7.0",
+    "tomlkit": "0.15.2",
+    "torch": "2.11.0+cu128",
+    "torch_fidelity": "0.4.0",
+    "torchcde": "0.2.5",
+    "torchcfm": "1.0.7",
+    "torchdiffeq": "0.2.5",
+    "torchdyn": "1.0.6",
+    "torchmetrics": "1.9.0",
+    "torchsde": "0.2.6",
+    "torchvision": "0.26.0+cu128",
+    "tornado": "6.5.10",
+    "tqdm": "4.70.1",
+    "traitlets": "5.16.1",
+    "trampoline": "0.1.2",
+    "triton": "3.6.0",
+    "truststore": "0.10.5",
+    "typing_extensions": "4.16.0",
+    "typst": "0.15.0",
+    "wcwidth": "0.9.2",
+    "widgetsnbextension": "4.0.16",
+    "yarl": "1.25.1"
+  }
+}

study/main/cifar10-dit-s103/evaluations/test-nfe32/evaluation.json

@@ -0,0 +1,40 @@
+{
+  "dataset": "cifar10",
+  "family": "dit",
+  "training_seed": 103,
+  "split": "test",
+  "intervention": "none",
+  "generated_count": 5000,
+  "real_count": 10000,
+  "sampling_seed": 20261016,
+  "metric_seed": 20261016,
+  "nfe": 32,
+  "kid_subsets": 100,
+  "kid_subset_size": 1000,
+  "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8",
+  "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97",
+  "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06",
+  "primary_eligible": true,
+  "device": "cuda",
+  "blas_threads": 4,
+  "status": "completed",
+  "sampling": {
+    "count": 5000,
+    "seed": 20261016,
+    "nfe": 32,
+    "batch_size": 32,
+    "seconds": 50.870260541996686,
+    "seconds_per_image": 0.010174052108399338,
+    "precision": "float32",
+    "timing_excludes": "model load and warmup",
+    "clipped_fraction": 0.016197331249713898,
+    "peak_memory_bytes": 58955776
+  },
+  "metrics": {
+    "frechet_inception_distance": 105.04701183050412,
+    "kernel_inception_distance_mean": 0.09203234910964966,
+    "kernel_inception_distance_std": 0.0018412378148906523
+  },
+  "validation_velocity_mse": 0.20122884213924408,
+  "metric_seconds": 18.536182572002872
+}

study/main/cifar10-dit-s103/evaluations/test-nfe32/evaluation_config.json

@@ -0,0 +1,20 @@
+{
+  "dataset": "cifar10",
+  "family": "dit",
+  "training_seed": 103,
+  "split": "test",
+  "intervention": "none",
+  "generated_count": 5000,
+  "real_count": 10000,
+  "sampling_seed": 20261016,
+  "metric_seed": 20261016,
+  "nfe": 32,
+  "kid_subsets": 100,
+  "kid_subset_size": 1000,
+  "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8",
+  "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97",
+  "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06",
+  "primary_eligible": true,
+  "device": "cuda",
+  "blas_threads": 4
+}

study/main/cifar10-dit-s103/evaluations/test-nfe32/run.log

@@ -0,0 +1,11 @@
+2026-10-17T18:02:50Z INFO {"event": "evaluation_start", "dataset": "cifar10", "family": "dit", "training_seed": 103, "split": "test", "intervention": "none", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06", "primary_eligible": true, "device": "cuda", "blas_threads": 4}
+2026-10-17T18:02:51Z INFO {"event": "sampling_progress", "completed": 32, "total": 5000, "nfe": 32}
+2026-10-17T18:02:58Z INFO {"event": "sampling_progress", "completed": 672, "total": 5000, "nfe": 32}
+2026-10-17T18:03:04Z INFO {"event": "sampling_progress", "completed": 1312, "total": 5000, "nfe": 32}
+2026-10-17T18:03:10Z INFO {"event": "sampling_progress", "completed": 1952, "total": 5000, "nfe": 32}
+2026-10-17T18:03:20Z INFO {"event": "sampling_progress", "completed": 2592, "total": 5000, "nfe": 32}
+2026-10-17T18:03:27Z INFO {"event": "sampling_progress", "completed": 3232, "total": 5000, "nfe": 32}
+2026-10-17T18:03:34Z INFO {"event": "sampling_progress", "completed": 3872, "total": 5000, "nfe": 32}
+2026-10-17T18:03:40Z INFO {"event": "sampling_progress", "completed": 4512, "total": 5000, "nfe": 32}
+2026-10-17T18:03:46Z INFO {"event": "sampling_complete", "directory": "runs/study/main/cifar10-dit-s103/evaluations/test-nfe32", "count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 50.870260541996686, "seconds_per_image": 0.010174052108399338, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.016197331249713898, "peak_memory_bytes": 58955776}
+2026-10-17T18:04:08Z INFO {"event": "evaluation_complete", "dataset": "cifar10", "family": "dit", "training_seed": 103, "split": "test", "intervention": "none", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06", "primary_eligible": true, "device": "cuda", "blas_threads": 4, "status": "completed", "sampling": {"count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 50.870260541996686, "seconds_per_image": 0.010174052108399338, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.016197331249713898, "peak_memory_bytes": 58955776}, "metrics": {"frechet_inception_distance": 105.04701183050412, "kernel_inception_distance_mean": 0.09203234910964966, "kernel_inception_distance_std": 0.0018412378148906523}, "validation_velocity_mse": 0.20122884213924408, "metric_seconds": 18.536182572002872}

study/main/cifar10-dit-s103/evaluations/test-nfe32/samples.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0f32af8d145aecdf4702ba11457aa15982a63209984f91c323966e64826cffd
+size 61440088

study/main/cifar10-dit-s103/evaluations/test-nfe32/sampling.json

@@ -0,0 +1,12 @@
+{
+  "count": 5000,
+  "seed": 20261016,
+  "nfe": 32,
+  "batch_size": 32,
+  "seconds": 50.870260541996686,
+  "seconds_per_image": 0.010174052108399338,
+  "precision": "float32",
+  "timing_excludes": "model load and warmup",
+  "clipped_fraction": 0.016197331249713898,
+  "peak_memory_bytes": 58955776
+}

study/main/cifar10-dit-s103/manifest.json

@@ -0,0 +1,62 @@
+{
+  "README.md": {
+    "sha256": "0adfbad0b044ddec61745a896e5258a4ef6ea3a9c18c70766c8601a3ccc24f8e",
+    "bytes": 2917
+  },
+  "checkpoint.pt": {
+    "sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8",
+    "bytes": 21305024
+  },
+  "config.json": {
+    "sha256": "c2935cce20fdc5ddce31152f75afea1cbae16862a27c64221e3eff4a2e1b16f0",
+    "bytes": 606
+  },
+  "data_manifest.json": {
+    "sha256": "d9d38bfe55c2a03eada50a4ef09f3152c987d290330edf4ed3dd452c50614350",
+    "bytes": 539473
+  },
+  "environment.json": {
+    "sha256": "01c239a46dd1e6a0fce629335d1332fa6144e2ca4e8778b2a33d817c7938b82f",
+    "bytes": 3695
+  },
+  "evaluations/test-nfe32/evaluation.json": {
+    "sha256": "23f2cbacdbbda45c31bdfcbe277ca1c14412a8beb7c7fdbb825d4dcb647df88c",
+    "bytes": 1237
+  },
+  "evaluations/test-nfe32/evaluation_config.json": {
+    "sha256": "6ec2b00f8a022dd81c45d7c497f276df6252dc7e76090f1c99c641e99d0203e4",
+    "bytes": 593
+  },
+  "evaluations/test-nfe32/grid.png": {
+    "sha256": "c1fb71f27637a2585ccce6cf94602aa5adb507a5de512bf1320bb29142cba54d",
+    "bytes": 154094
+  },
+  "evaluations/test-nfe32/run.log": {
+    "sha256": "6a30e7c329b8705eee0081444146422da2479f99f46191a6df243f8a79f02288",
+    "bytes": 3016
+  },
+  "evaluations/test-nfe32/samples.safetensors": {
+    "sha256": "f0f32af8d145aecdf4702ba11457aa15982a63209984f91c323966e64826cffd",
+    "bytes": 61440088
+  },
+  "evaluations/test-nfe32/sampling.json": {
+    "sha256": "f217de56b290912d43ba23f7e5a2b9e06e13d9fdac8d5c8869b825718487f4bc",
+    "bytes": 300
+  },
+  "model.safetensors": {
+    "sha256": "75bca1cf3738374c420f34e36b0679e6369cf3fa86343aa762f0ff85bd6df64b",
+    "bytes": 5233488
+  },
+  "planned-config.json": {
+    "sha256": "78efcf937758037f97e824101752678b300259fafe087bee674aa72fd0be9a8c",
+    "bytes": 604
+  },
+  "run.log": {
+    "sha256": "8b7c633bc1425d2d314b5baa753a50ac5d2230fce4e29599e2acb29e559233eb",
+    "bytes": 80315
+  },
+  "training.json": {
+    "sha256": "013d0fe9179b627007a12bcdaf38e842ce5d460b37b9dc62591263c7e90210a1",
+    "bytes": 542
+  }
+}

study/main/cifar10-dit-s103/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75bca1cf3738374c420f34e36b0679e6369cf3fa86343aa762f0ff85bd6df64b
+size 5233488

study/main/cifar10-dit-s103/planned-config.json

@@ -0,0 +1,32 @@
+{
+  "model": {
+    "family": "dit",
+    "channels": 3,
+    "size": 32,
+    "width": 128,
+    "depth": 4,
+    "patch": 4,
+    "rotating_dim": 4,
+    "inner_steps": 4,
+    "step_size": 0.1,
+    "ffn_hidden": 0
+  },
+  "dataset": "cifar10",
+  "seed": 103,
+  "device": "cuda",
+  "batch_size": 128,
+  "microbatch": 32,
+  "learning_rate": 0.0003,
+  "weight_decay": 0.01,
+  "clip_grad": 1.0,
+  "ema_decay": 0.999,
+  "minutes": 45,
+  "max_steps": 0,
+  "validation_size": 5000,
+  "data_root": "data",
+  "runs_root": "runs",
+  "budget_hours": 24.0,
+  "log_every": 50,
+  "checkpoint_seconds": 300.0,
+  "amp": true
+}

study/main/cifar10-dit-s103/run.log

@@ -0,0 +1,346 @@
+2026-10-17T06:12:52Z INFO {"event": "training_configuration", "model": {"family": "dit", "channels": 3, "size": 32, "width": 128, "depth": 4, "patch": 4, "rotating_dim": 4, "inner_steps": 4, "step_size": 0.1, "ffn_hidden": 576}, "dataset": "cifar10", "seed": 103, "device": "cuda", "batch_size": 128, "microbatch": 32, "learning_rate": 0.0003, "weight_decay": 0.01, "clip_grad": 1.0, "ema_decay": 0.999, "minutes": 45, "max_steps": 0, "validation_size": 5000, "data_root": "data", "runs_root": "runs", "budget_hours": 24.0, "log_every": 50, "checkpoint_seconds": 300.0, "amp": true, "resume": false, "source_revisions": ["c749856d1c9fe07d6ce2f39d01de78f664aea9a8"]}
+2026-10-17T06:12:54Z INFO {"event": "training_start", "parameters": 1298864, "step": 0, "validation_mse": 1.594213217496872}
+2026-10-17T06:12:55Z INFO {"event": "training_progress", "step": 1, "epoch": 0, "loss": 1.5858024060726166, "grad_norm": 1.1447488069534302, "train_seconds": 1.1093979769939324, "remaining_budget_seconds": 2757.1118136859877}
+2026-10-17T06:13:02Z INFO {"event": "training_progress", "step": 50, "epoch": 0, "loss": 1.1657225693366966, "grad_norm": 0.25531885027885437, "train_seconds": 8.89960309000162, "remaining_budget_seconds": 2749.321612376996}
+2026-10-17T06:13:13Z INFO {"event": "training_progress", "step": 100, "epoch": 0, "loss": 0.5596751123666763, "grad_norm": 0.9169033765792847, "train_seconds": 16.903101167001296, "remaining_budget_seconds": 2741.318117502}
+2026-10-17T06:13:21Z INFO {"event": "training_progress", "step": 150, "epoch": 0, "loss": 0.40701583728194235, "grad_norm": 0.5373302698135376, "train_seconds": 24.719850269000744, "remaining_budget_seconds": 2733.501367577992}
+2026-10-17T06:13:29Z INFO {"event": "training_progress", "step": 200, "epoch": 0, "loss": 0.36346321791410446, "grad_norm": 0.7082643508911133, "train_seconds": 32.72035880399926, "remaining_budget_seconds": 2725.5008597029955}
+2026-10-17T06:13:37Z INFO {"event": "training_progress", "step": 250, "epoch": 0, "loss": 0.3313479420542717, "grad_norm": 0.7261473536491394, "train_seconds": 40.51681599499716, "remaining_budget_seconds": 2717.704403676995}
+2026-10-17T06:13:48Z INFO {"event": "training_progress", "step": 300, "epoch": 0, "loss": 0.3218912038952112, "grad_norm": 0.6696639657020569, "train_seconds": 48.83902604099421, "remaining_budget_seconds": 2709.382193925994}
+2026-10-17T06:13:56Z INFO {"event": "training_progress", "step": 350, "epoch": 0, "loss": 0.31165684685111045, "grad_norm": 0.549306333065033, "train_seconds": 56.60705974699522, "remaining_budget_seconds": 2701.614160344994}
+2026-10-17T06:14:03Z INFO {"event": "training_progress", "step": 400, "epoch": 1, "loss": 0.2996305550634861, "grad_norm": 0.7370758056640625, "train_seconds": 64.41853809199529, "remaining_budget_seconds": 2693.802678672}
+2026-10-17T06:14:11Z INFO {"event": "training_progress", "step": 450, "epoch": 1, "loss": 0.2978191590309143, "grad_norm": 0.5665438175201416, "train_seconds": 72.30074656500074, "remaining_budget_seconds": 2685.920469263001}
+2026-10-17T06:14:22Z INFO {"event": "training_progress", "step": 500, "epoch": 1, "loss": 0.2881073398888111, "grad_norm": 0.7129167318344116, "train_seconds": 80.4670163510018, "remaining_budget_seconds": 2677.7542015809886}
+2026-10-17T06:14:30Z INFO {"event": "training_progress", "step": 550, "epoch": 1, "loss": 0.28837867863476274, "grad_norm": 0.501752495765686, "train_seconds": 88.2101274329907, "remaining_budget_seconds": 2670.011090547996}
+2026-10-17T06:14:38Z INFO {"event": "training_progress", "step": 600, "epoch": 1, "loss": 0.2830388394743204, "grad_norm": 0.48170122504234314, "train_seconds": 96.50876470899675, "remaining_budget_seconds": 2661.712455273999}
+2026-10-17T06:14:46Z INFO {"event": "training_progress", "step": 650, "epoch": 1, "loss": 0.2765868289023638, "grad_norm": 0.5665010809898376, "train_seconds": 104.37119753399747, "remaining_budget_seconds": 2653.850017429999}
+2026-10-17T06:14:57Z INFO {"event": "training_progress", "step": 700, "epoch": 1, "loss": 0.2799556093662977, "grad_norm": 0.7932279109954834, "train_seconds": 112.20621758900234, "remaining_budget_seconds": 2646.015000507992}
+2026-10-17T06:15:05Z INFO {"event": "training_progress", "step": 750, "epoch": 2, "loss": 0.2713454516977072, "grad_norm": 0.5437489748001099, "train_seconds": 120.24805785699573, "remaining_budget_seconds": 2637.9731600629893}
+2026-10-17T06:15:13Z INFO {"event": "training_progress", "step": 800, "epoch": 2, "loss": 0.267328729480505, "grad_norm": 0.7559741139411926, "train_seconds": 127.93736295698909, "remaining_budget_seconds": 2630.283854914989}
+2026-10-17T06:15:21Z INFO {"event": "training_progress", "step": 850, "epoch": 2, "loss": 0.26908930391073227, "grad_norm": 0.6857976913452148, "train_seconds": 135.98099386499962, "remaining_budget_seconds": 2622.240225329995}
+2026-10-17T06:15:31Z INFO {"event": "training_progress", "step": 900, "epoch": 2, "loss": 0.2664102631062269, "grad_norm": 0.5317476391792297, "train_seconds": 143.85549412899127, "remaining_budget_seconds": 2614.365726125994}
+2026-10-17T06:15:39Z INFO {"event": "training_progress", "step": 950, "epoch": 2, "loss": 0.2644197683781385, "grad_norm": 0.48710930347442627, "train_seconds": 152.06733656299184, "remaining_budget_seconds": 2606.1538825099997}
+2026-10-17T06:15:47Z INFO {"event": "training_progress", "step": 1000, "epoch": 2, "loss": 0.258511461019516, "grad_norm": 0.5359388589859009, "train_seconds": 159.74682801899326, "remaining_budget_seconds": 2598.4743904859934}
+2026-10-17T06:15:55Z INFO {"event": "training_progress", "step": 1050, "epoch": 2, "loss": 0.26546950921416285, "grad_norm": 0.4923234283924103, "train_seconds": 167.64112890599063, "remaining_budget_seconds": 2590.5800893470005}
+2026-10-17T06:16:06Z INFO {"event": "training_progress", "step": 1100, "epoch": 3, "loss": 0.259367870092392, "grad_norm": 0.5759830474853516, "train_seconds": 175.46796600800008, "remaining_budget_seconds": 2582.7532510109886}
+2026-10-17T06:16:13Z INFO {"event": "training_progress", "step": 1150, "epoch": 3, "loss": 0.2555141791701317, "grad_norm": 0.43287357687950134, "train_seconds": 183.24664693899103, "remaining_budget_seconds": 2574.9745658979955}
+2026-10-17T06:16:21Z INFO {"event": "training_progress", "step": 1200, "epoch": 3, "loss": 0.25495643079280855, "grad_norm": 0.5788059234619141, "train_seconds": 190.73366184199404, "remaining_budget_seconds": 2567.487556565}
+2026-10-17T06:16:29Z INFO {"event": "training_progress", "step": 1250, "epoch": 3, "loss": 0.2528892236948013, "grad_norm": 0.5327805876731873, "train_seconds": 198.944605081997, "remaining_budget_seconds": 2559.2766120179876}
+2026-10-17T06:16:40Z INFO {"event": "training_progress", "step": 1300, "epoch": 3, "loss": 0.24987026616930963, "grad_norm": 0.5859372615814209, "train_seconds": 207.00150906699128, "remaining_budget_seconds": 2551.2197092959977}
+2026-10-17T06:16:48Z INFO {"event": "training_progress", "step": 1350, "epoch": 3, "loss": 0.25114845678210257, "grad_norm": 0.46870237588882446, "train_seconds": 215.05307725499733, "remaining_budget_seconds": 2543.1681393229956}
+2026-10-17T06:16:56Z INFO {"event": "training_progress", "step": 1400, "epoch": 3, "loss": 0.24749716266989708, "grad_norm": 0.7022876739501953, "train_seconds": 222.58840013999725, "remaining_budget_seconds": 2535.6328182219877}
+2026-10-17T06:17:03Z INFO {"event": "training_progress", "step": 1450, "epoch": 4, "loss": 0.2500631644576788, "grad_norm": 0.43558770418167114, "train_seconds": 230.2677556869894, "remaining_budget_seconds": 2527.9534599759936}
+2026-10-17T06:17:14Z INFO {"event": "training_progress", "step": 1500, "epoch": 4, "loss": 0.25021994598209857, "grad_norm": 0.5498150587081909, "train_seconds": 238.16692778699507, "remaining_budget_seconds": 2520.054290521992}
+2026-10-17T06:17:22Z INFO {"event": "training_progress", "step": 1550, "epoch": 4, "loss": 0.24929071314632892, "grad_norm": 0.4663602411746979, "train_seconds": 245.93860444999882, "remaining_budget_seconds": 2512.282615240998}
+2026-10-17T06:17:29Z INFO {"event": "training_progress", "step": 1600, "epoch": 4, "loss": 0.24522738866508006, "grad_norm": 0.4618476331233978, "train_seconds": 253.65732551499968, "remaining_budget_seconds": 2504.563892899998}
+2026-10-17T06:17:38Z INFO {"event": "training_progress", "step": 1650, "epoch": 4, "loss": 0.24152128726243974, "grad_norm": 0.5242508053779602, "train_seconds": 261.7224400159903, "remaining_budget_seconds": 2496.4987768489955}
+2026-10-17T06:17:49Z INFO {"event": "training_progress", "step": 1700, "epoch": 4, "loss": 0.24401607491075994, "grad_norm": 0.47633522748947144, "train_seconds": 269.93991573799576, "remaining_budget_seconds": 2488.281304142991}
+2026-10-17T06:17:56Z INFO {"event": "training_progress", "step": 1750, "epoch": 4, "loss": 0.24733651667833328, "grad_norm": 0.40939226746559143, "train_seconds": 277.85371196299093, "remaining_budget_seconds": 2480.3675074519997}
+2026-10-17T06:18:04Z INFO {"event": "training_progress", "step": 1800, "epoch": 5, "loss": 0.24594629168510437, "grad_norm": 0.4144447147846222, "train_seconds": 285.6933178229956, "remaining_budget_seconds": 2472.5279020409944}
+2026-10-17T06:18:12Z INFO {"event": "training_progress", "step": 1850, "epoch": 5, "loss": 0.2368610467761755, "grad_norm": 0.4092739522457123, "train_seconds": 293.6113533829921, "remaining_budget_seconds": 2464.6098649740015}
+2026-10-17T06:18:23Z INFO {"event": "training_progress", "step": 1900, "epoch": 5, "loss": 0.24035702586174013, "grad_norm": 0.38784483075141907, "train_seconds": 301.8605836919887, "remaining_budget_seconds": 2456.3606362739956}
+2026-10-17T06:18:31Z INFO {"event": "training_progress", "step": 1950, "epoch": 5, "loss": 0.23633535146713258, "grad_norm": 0.3784784972667694, "train_seconds": 310.0069431820011, "remaining_budget_seconds": 2448.2142755729874}
+2026-10-17T06:18:39Z INFO {"event": "training_progress", "step": 2000, "epoch": 5, "loss": 0.23702838465571405, "grad_norm": 0.3886110484600067, "train_seconds": 317.79530229698867, "remaining_budget_seconds": 2440.4259165989934}
+2026-10-17T06:18:47Z INFO {"event": "training_progress", "step": 2050, "epoch": 5, "loss": 0.2401875351369381, "grad_norm": 0.6384299993515015, "train_seconds": 325.5639061839902, "remaining_budget_seconds": 2432.6573138559907}
+2026-10-17T06:18:58Z INFO {"event": "training_progress", "step": 2100, "epoch": 5, "loss": 0.24170859850943088, "grad_norm": 0.3459925353527069, "train_seconds": 333.67934680100007, "remaining_budget_seconds": 2424.5418713789986}
+2026-10-17T06:19:06Z INFO {"event": "training_progress", "step": 2150, "epoch": 6, "loss": 0.23443863317370414, "grad_norm": 0.28909772634506226, "train_seconds": 341.595917284998, "remaining_budget_seconds": 2416.625299923995}
+2026-10-17T06:19:13Z INFO {"event": "training_progress", "step": 2200, "epoch": 6, "loss": 0.23649297505617142, "grad_norm": 0.3782138526439667, "train_seconds": 349.32196098400163, "remaining_budget_seconds": 2408.899257033001}
+2026-10-17T06:19:22Z INFO {"event": "training_progress", "step": 2250, "epoch": 6, "loss": 0.23255741193890572, "grad_norm": 0.3699982166290283, "train_seconds": 357.5707889639889, "remaining_budget_seconds": 2400.650429117988}
+2026-10-17T06:19:32Z INFO {"event": "training_progress", "step": 2300, "epoch": 6, "loss": 0.23195066228508948, "grad_norm": 0.5726337432861328, "train_seconds": 365.416113964995, "remaining_budget_seconds": 2392.805104332001}
+2026-10-17T06:19:40Z INFO {"event": "training_progress", "step": 2350, "epoch": 6, "loss": 0.23521249562501909, "grad_norm": 0.4563778340816498, "train_seconds": 373.5735043859895, "remaining_budget_seconds": 2384.6477153759915}
+2026-10-17T06:19:48Z INFO {"event": "training_progress", "step": 2400, "epoch": 6, "loss": 0.23175752215087414, "grad_norm": 0.37046200037002563, "train_seconds": 381.3004812029976, "remaining_budget_seconds": 2376.920737330991}
+2026-10-17T06:19:56Z INFO {"event": "training_progress", "step": 2450, "epoch": 6, "loss": 0.23276604041457177, "grad_norm": 0.37460559606552124, "train_seconds": 389.5152460199897, "remaining_budget_seconds": 2368.705973581993}
+2026-10-17T06:20:07Z INFO {"event": "training_progress", "step": 2500, "epoch": 7, "loss": 0.23026653416454793, "grad_norm": 0.4962834417819977, "train_seconds": 397.5793896809919, "remaining_budget_seconds": 2360.6418280200014}
+2026-10-17T06:20:14Z INFO {"event": "training_progress", "step": 2550, "epoch": 7, "loss": 0.23353782542049883, "grad_norm": 0.43206289410591125, "train_seconds": 405.28961028400226, "remaining_budget_seconds": 2352.9316071939975}
+2026-10-17T06:20:22Z INFO {"event": "training_progress", "step": 2600, "epoch": 7, "loss": 0.23521389670670032, "grad_norm": 0.3576737940311432, "train_seconds": 412.7502793349995, "remaining_budget_seconds": 2345.4709408569906}
+2026-10-17T06:20:30Z INFO {"event": "training_progress", "step": 2650, "epoch": 7, "loss": 0.2284833437204361, "grad_norm": 0.4273090362548828, "train_seconds": 420.68508532398846, "remaining_budget_seconds": 2337.536132919995}
+2026-10-17T06:20:37Z INFO {"event": "training_progress", "step": 2700, "epoch": 7, "loss": 0.228330737426877, "grad_norm": 0.34203803539276123, "train_seconds": 428.2994178529916, "remaining_budget_seconds": 2329.9218012209894}
+2026-10-17T06:20:48Z INFO {"event": "training_progress", "step": 2750, "epoch": 7, "loss": 0.23243201471865177, "grad_norm": 0.3263692855834961, "train_seconds": 436.55705744799343, "remaining_budget_seconds": 2321.6641600329895}
+2026-10-17T06:20:56Z INFO {"event": "training_progress", "step": 2800, "epoch": 7, "loss": 0.2268905485421419, "grad_norm": 0.3280545473098755, "train_seconds": 444.2458421289921, "remaining_budget_seconds": 2313.975378215997}
+2026-10-17T06:21:04Z INFO {"event": "training_progress", "step": 2850, "epoch": 8, "loss": 0.23023489162325858, "grad_norm": 0.46501192450523376, "train_seconds": 452.1488154249964, "remaining_budget_seconds": 2306.072404551989}
+2026-10-17T06:21:12Z INFO {"event": "training_progress", "step": 2900, "epoch": 8, "loss": 0.2276052101701498, "grad_norm": 0.3357611894607544, "train_seconds": 459.962149820989, "remaining_budget_seconds": 2298.2590702179878}
+2026-10-17T06:21:22Z INFO {"event": "training_progress", "step": 2950, "epoch": 8, "loss": 0.22960343651473522, "grad_norm": 0.47332116961479187, "train_seconds": 467.78257468299125, "remaining_budget_seconds": 2290.438645131988}
+2026-10-17T06:21:30Z INFO {"event": "training_progress", "step": 3000, "epoch": 8, "loss": 0.22828908033668996, "grad_norm": 0.4681585133075714, "train_seconds": 475.48973585299973, "remaining_budget_seconds": 2282.7314830039977}
+2026-10-17T06:21:38Z INFO {"event": "training_progress", "step": 3050, "epoch": 8, "loss": 0.22836946815252304, "grad_norm": 0.44736719131469727, "train_seconds": 483.40557999900193, "remaining_budget_seconds": 2274.8156385069888}
+2026-10-17T06:21:46Z INFO {"event": "training_progress", "step": 3100, "epoch": 8, "loss": 0.22799053288996218, "grad_norm": 0.4848003387451172, "train_seconds": 491.14953085398884, "remaining_budget_seconds": 2267.0716881359986}
+2026-10-17T06:21:57Z INFO {"event": "training_progress", "step": 3150, "epoch": 8, "loss": 0.22419281348586081, "grad_norm": 0.5007691383361816, "train_seconds": 499.0091629660019, "remaining_budget_seconds": 2259.2120533179987}
+2026-10-17T06:22:05Z INFO {"event": "training_progress", "step": 3200, "epoch": 9, "loss": 0.22625502377748488, "grad_norm": 0.6219082474708557, "train_seconds": 507.40365570699214, "remaining_budget_seconds": 2250.8175624219875}
+2026-10-17T06:22:13Z INFO {"event": "training_progress", "step": 3250, "epoch": 9, "loss": 0.22761659011244773, "grad_norm": 0.304275244474411, "train_seconds": 515.3742627369938, "remaining_budget_seconds": 2242.8469567779975}
+2026-10-17T06:22:21Z INFO {"event": "training_progress", "step": 3300, "epoch": 9, "loss": 0.2233795889467001, "grad_norm": 0.45222702622413635, "train_seconds": 523.1404202929989, "remaining_budget_seconds": 2235.080799463991}
+2026-10-17T06:22:31Z INFO {"event": "training_progress", "step": 3350, "epoch": 9, "loss": 0.2270749957859516, "grad_norm": 0.45801252126693726, "train_seconds": 530.955374386991, "remaining_budget_seconds": 2227.265844221998}
+2026-10-17T06:22:39Z INFO {"event": "training_progress", "step": 3400, "epoch": 9, "loss": 0.22660564221441745, "grad_norm": 0.4578585922718048, "train_seconds": 538.7965619719907, "remaining_budget_seconds": 2219.424655912997}
+2026-10-17T06:22:47Z INFO {"event": "training_progress", "step": 3450, "epoch": 9, "loss": 0.22066663138568401, "grad_norm": 0.35886430740356445, "train_seconds": 546.8847156970005, "remaining_budget_seconds": 2211.33650055599}
+2026-10-17T06:22:55Z INFO {"event": "training_progress", "step": 3500, "epoch": 9, "loss": 0.22484881162643433, "grad_norm": 0.39434459805488586, "train_seconds": 554.6958909579989, "remaining_budget_seconds": 2203.52532181899}
+2026-10-17T06:23:06Z INFO {"event": "training_progress", "step": 3550, "epoch": 10, "loss": 0.22516201093792915, "grad_norm": 0.2976161241531372, "train_seconds": 562.5748311779898, "remaining_budget_seconds": 2195.646388807989}
+2026-10-17T06:23:14Z INFO {"event": "training_progress", "step": 3600, "epoch": 10, "loss": 0.22430625572800636, "grad_norm": 0.45903125405311584, "train_seconds": 570.5946364059928, "remaining_budget_seconds": 2187.626583716992}
+2026-10-17T06:23:22Z INFO {"event": "training_progress", "step": 3650, "epoch": 10, "loss": 0.2187049974501133, "grad_norm": 0.40333452820777893, "train_seconds": 578.6608084089967, "remaining_budget_seconds": 2179.560409480997}
+2026-10-17T06:23:30Z INFO {"event": "training_progress", "step": 3700, "epoch": 10, "loss": 0.22649382025003434, "grad_norm": 0.3392047584056854, "train_seconds": 586.7165378739883, "remaining_budget_seconds": 2171.504680392987}
+2026-10-17T06:23:41Z INFO {"event": "training_progress", "step": 3750, "epoch": 10, "loss": 0.2196612397581339, "grad_norm": 0.32081568241119385, "train_seconds": 594.6509519899992, "remaining_budget_seconds": 2163.5702675629873}
+2026-10-17T06:23:49Z INFO {"event": "training_progress", "step": 3800, "epoch": 10, "loss": 0.22242018193006516, "grad_norm": 0.40472421050071716, "train_seconds": 602.7207385520014, "remaining_budget_seconds": 2155.5004798449954}
+2026-10-17T06:23:57Z INFO {"event": "training_progress", "step": 3850, "epoch": 10, "loss": 0.22197414949536323, "grad_norm": 0.3029738664627075, "train_seconds": 610.6465785589971, "remaining_budget_seconds": 2147.574641788}
+2026-10-17T06:24:05Z INFO {"event": "training_progress", "step": 3900, "epoch": 11, "loss": 0.2249407770484686, "grad_norm": 0.5050836801528931, "train_seconds": 618.7807029219985, "remaining_budget_seconds": 2139.4405155509885}
+2026-10-17T06:24:16Z INFO {"event": "training_progress", "step": 3950, "epoch": 11, "loss": 0.22174503870308399, "grad_norm": 0.3507663309574127, "train_seconds": 626.8719755369966, "remaining_budget_seconds": 2131.3492437469977}
+2026-10-17T06:24:23Z INFO {"event": "training_progress", "step": 4000, "epoch": 11, "loss": 0.22617063827812672, "grad_norm": 0.3266119360923767, "train_seconds": 634.7928381959937, "remaining_budget_seconds": 2123.4283812629874}
+2026-10-17T06:24:31Z INFO {"event": "training_progress", "step": 4050, "epoch": 11, "loss": 0.22236737102270127, "grad_norm": 0.3884757161140442, "train_seconds": 642.633714274998, "remaining_budget_seconds": 2115.587505546995}
+2026-10-17T06:24:39Z INFO {"event": "training_progress", "step": 4100, "epoch": 11, "loss": 0.22096656911075116, "grad_norm": 0.3370361924171448, "train_seconds": 650.5870550889958, "remaining_budget_seconds": 2107.6341635439894}
+2026-10-17T06:24:50Z INFO {"event": "training_progress", "step": 4150, "epoch": 11, "loss": 0.22029311262071133, "grad_norm": 0.28956443071365356, "train_seconds": 658.511136991001, "remaining_budget_seconds": 2099.710080053992}
+2026-10-17T06:24:58Z INFO {"event": "training_progress", "step": 4200, "epoch": 11, "loss": 0.22083521753549576, "grad_norm": 0.37593013048171997, "train_seconds": 666.3864306609903, "remaining_budget_seconds": 2091.834787150001}
+2026-10-17T06:25:06Z INFO {"event": "training_progress", "step": 4250, "epoch": 12, "loss": 0.21979545205831527, "grad_norm": 0.2975398898124695, "train_seconds": 674.34134752299, "remaining_budget_seconds": 2083.879871981}
+2026-10-17T06:25:13Z INFO {"event": "training_progress", "step": 4300, "epoch": 12, "loss": 0.21920039802789687, "grad_norm": 0.3509204685688019, "train_seconds": 682.0125237289903, "remaining_budget_seconds": 2076.208696115995}
+2026-10-17T06:25:24Z INFO {"event": "training_progress", "step": 4350, "epoch": 12, "loss": 0.22070457577705382, "grad_norm": 0.34450697898864746, "train_seconds": 689.942477367993, "remaining_budget_seconds": 2068.2787422279944}
+2026-10-17T06:25:32Z INFO {"event": "training_progress", "step": 4400, "epoch": 12, "loss": 0.22187306962907313, "grad_norm": 0.356631338596344, "train_seconds": 697.7941616359894, "remaining_budget_seconds": 2060.4270562549937}
+2026-10-17T06:25:40Z INFO {"event": "training_progress", "step": 4450, "epoch": 12, "loss": 0.21966786608099936, "grad_norm": 0.26416870951652527, "train_seconds": 705.7873432740016, "remaining_budget_seconds": 2052.4338744499983}
+2026-10-17T06:25:48Z INFO {"event": "training_progress", "step": 4500, "epoch": 12, "loss": 0.21762670360505582, "grad_norm": 0.31004515290260315, "train_seconds": 713.5471682189964, "remaining_budget_seconds": 2044.6740501819877}
+2026-10-17T06:25:59Z INFO {"event": "training_progress", "step": 4550, "epoch": 12, "loss": 0.22110140047967433, "grad_norm": 0.3795686364173889, "train_seconds": 721.4695359369944, "remaining_budget_seconds": 2036.7516838729935}
+2026-10-17T06:26:07Z INFO {"event": "training_progress", "step": 4600, "epoch": 13, "loss": 0.22088508442044258, "grad_norm": 0.3754642605781555, "train_seconds": 729.4922079129901, "remaining_budget_seconds": 2028.729011782998}
+2026-10-17T06:26:15Z INFO {"event": "training_progress", "step": 4650, "epoch": 13, "loss": 0.22165796056389808, "grad_norm": 0.26521599292755127, "train_seconds": 737.1305047209898, "remaining_budget_seconds": 2021.0907148219994}
+2026-10-17T06:26:22Z INFO {"event": "training_progress", "step": 4700, "epoch": 13, "loss": 0.21576032608747484, "grad_norm": 0.3360327482223511, "train_seconds": 745.0182190929918, "remaining_budget_seconds": 2013.2030000939994}
+2026-10-17T06:26:33Z INFO {"event": "training_progress", "step": 4750, "epoch": 13, "loss": 0.2179314721375704, "grad_norm": 0.2595120072364807, "train_seconds": 752.9275016669999, "remaining_budget_seconds": 2005.293716931992}
+2026-10-17T06:26:41Z INFO {"event": "training_progress", "step": 4800, "epoch": 13, "loss": 0.21904217034578324, "grad_norm": 0.2852647602558136, "train_seconds": 760.8422218049964, "remaining_budget_seconds": 1997.3789980000001}
+2026-10-17T06:26:49Z INFO {"event": "training_progress", "step": 4850, "epoch": 13, "loss": 0.21807369999587536, "grad_norm": 0.2661583125591278, "train_seconds": 768.5785933299921, "remaining_budget_seconds": 1989.6426248149946}
+2026-10-17T06:26:57Z INFO {"event": "training_progress", "step": 4900, "epoch": 13, "loss": 0.22032808914780616, "grad_norm": 0.26635926961898804, "train_seconds": 776.1246975779941, "remaining_budget_seconds": 1982.0965204859967}
+2026-10-17T06:27:04Z INFO {"event": "training_progress", "step": 4950, "epoch": 14, "loss": 0.21697402730584145, "grad_norm": 0.2398473471403122, "train_seconds": 783.4518110369972, "remaining_budget_seconds": 1974.769407423999}
+2026-10-17T06:27:15Z INFO {"event": "training_progress", "step": 5000, "epoch": 14, "loss": 0.21662791088223457, "grad_norm": 0.3562970459461212, "train_seconds": 791.5267033609998, "remaining_budget_seconds": 1966.694516907999}
+2026-10-17T06:27:23Z INFO {"event": "training_progress", "step": 5050, "epoch": 14, "loss": 0.21890034899115562, "grad_norm": 0.444506973028183, "train_seconds": 799.5762423409906, "remaining_budget_seconds": 1958.6449775939982}
+2026-10-17T06:27:31Z INFO {"event": "training_progress", "step": 5100, "epoch": 14, "loss": 0.21985600829124452, "grad_norm": 0.25608178973197937, "train_seconds": 807.4401934310008, "remaining_budget_seconds": 1950.7810239819955}
+2026-10-17T06:27:39Z INFO {"event": "training_progress", "step": 5150, "epoch": 14, "loss": 0.21900405824184419, "grad_norm": 0.5257599949836731, "train_seconds": 815.0230543200014, "remaining_budget_seconds": 1943.198165763999}
+2026-10-17T06:27:49Z INFO {"event": "training_progress", "step": 5200, "epoch": 14, "loss": 0.22095485627651215, "grad_norm": 0.358149915933609, "train_seconds": 822.9534938069992, "remaining_budget_seconds": 1935.2677237079915}
+2026-10-17T06:27:57Z INFO {"event": "training_progress", "step": 5250, "epoch": 14, "loss": 0.21590333737432957, "grad_norm": 0.4232774078845978, "train_seconds": 830.9310384799901, "remaining_budget_seconds": 1927.2901802150009}
+2026-10-17T06:28:05Z INFO {"event": "training_progress", "step": 5300, "epoch": 15, "loss": 0.21534299075603486, "grad_norm": 0.432813823223114, "train_seconds": 838.4716667009925, "remaining_budget_seconds": 1919.7495535449998}
+2026-10-17T06:28:12Z INFO {"event": "training_progress", "step": 5350, "epoch": 15, "loss": 0.21747671745717526, "grad_norm": 0.31067749857902527, "train_seconds": 845.9670338919968, "remaining_budget_seconds": 1912.254186361999}
+2026-10-17T06:28:24Z INFO {"event": "training_progress", "step": 5400, "epoch": 15, "loss": 0.21838203243911267, "grad_norm": 0.4284810423851013, "train_seconds": 854.118107383998, "remaining_budget_seconds": 1904.1031090139877}
+2026-10-17T06:28:32Z INFO {"event": "training_progress", "step": 5450, "epoch": 15, "loss": 0.21866775512695313, "grad_norm": 0.4132545590400696, "train_seconds": 862.2800148319948, "remaining_budget_seconds": 1895.9412034369889}
+2026-10-17T06:28:40Z INFO {"event": "training_progress", "step": 5500, "epoch": 15, "loss": 0.2167457301914692, "grad_norm": 0.25229963660240173, "train_seconds": 869.8667010229983, "remaining_budget_seconds": 1888.3545189889992}
+2026-10-17T06:28:47Z INFO {"event": "training_progress", "step": 5550, "epoch": 15, "loss": 0.2184431102871895, "grad_norm": 0.3536131978034973, "train_seconds": 877.4874744619883, "remaining_budget_seconds": 1880.7337433039938}
+2026-10-17T06:28:58Z INFO {"event": "training_progress", "step": 5600, "epoch": 15, "loss": 0.21813512712717056, "grad_norm": 0.4633287489414215, "train_seconds": 885.2120020110015, "remaining_budget_seconds": 1873.0092149279953}
+2026-10-17T06:29:05Z INFO {"event": "training_progress", "step": 5650, "epoch": 16, "loss": 0.2149638119339943, "grad_norm": 0.331958144903183, "train_seconds": 893.0496498559951, "remaining_budget_seconds": 1865.1715682349895}
+2026-10-17T06:29:13Z INFO {"event": "training_progress", "step": 5700, "epoch": 16, "loss": 0.21578671015799045, "grad_norm": 0.24876010417938232, "train_seconds": 900.9600049719884, "remaining_budget_seconds": 1857.2612124789885}
+2026-10-17T06:29:21Z INFO {"event": "training_progress", "step": 5750, "epoch": 16, "loss": 0.21443996228277684, "grad_norm": 0.26986733078956604, "train_seconds": 908.6872599189956, "remaining_budget_seconds": 1849.533956603991}
+2026-10-17T06:29:32Z INFO {"event": "training_progress", "step": 5800, "epoch": 16, "loss": 0.21777682840824128, "grad_norm": 0.2711006700992584, "train_seconds": 916.7837736419897, "remaining_budget_seconds": 1841.4374460260005}
+2026-10-17T06:29:40Z INFO {"event": "training_progress", "step": 5850, "epoch": 16, "loss": 0.2169272170960903, "grad_norm": 0.24431927502155304, "train_seconds": 924.7597131959919, "remaining_budget_seconds": 1833.4615065809921}
+2026-10-17T06:29:49Z INFO {"event": "training_progress", "step": 5900, "epoch": 16, "loss": 0.2135427714139223, "grad_norm": 0.5062280297279358, "train_seconds": 932.9391810369998, "remaining_budget_seconds": 1825.2820368279936}
+2026-10-17T06:29:56Z INFO {"event": "training_progress", "step": 5950, "epoch": 16, "loss": 0.21666086301207543, "grad_norm": 0.5161721706390381, "train_seconds": 940.6140677270014, "remaining_budget_seconds": 1817.6071521580016}
+2026-10-17T06:30:07Z INFO {"event": "training_progress", "step": 6000, "epoch": 17, "loss": 0.21441231653094292, "grad_norm": 0.3891008198261261, "train_seconds": 948.4656887960009, "remaining_budget_seconds": 1809.75553072599}
+2026-10-17T06:30:15Z INFO {"event": "training_progress", "step": 6050, "epoch": 17, "loss": 0.2118591333180666, "grad_norm": 0.2517883777618408, "train_seconds": 956.3341775159934, "remaining_budget_seconds": 1801.8870428189985}
+2026-10-17T06:30:23Z INFO {"event": "training_progress", "step": 6100, "epoch": 17, "loss": 0.2163904131948948, "grad_norm": 0.29660964012145996, "train_seconds": 964.047685847996, "remaining_budget_seconds": 1794.173531620996}
+2026-10-17T06:30:30Z INFO {"event": "training_progress", "step": 6150, "epoch": 17, "loss": 0.21628059163689614, "grad_norm": 0.3079354465007782, "train_seconds": 971.5731122699945, "remaining_budget_seconds": 1786.6481051269948}
+2026-10-17T06:30:41Z INFO {"event": "training_progress", "step": 6200, "epoch": 17, "loss": 0.21293036803603171, "grad_norm": 0.4128168225288391, "train_seconds": 979.5427955859923, "remaining_budget_seconds": 1778.6784241290006}
+2026-10-17T06:30:50Z INFO {"event": "training_progress", "step": 6250, "epoch": 17, "loss": 0.21732972212135793, "grad_norm": 0.35838252305984497, "train_seconds": 987.6214082359947, "remaining_budget_seconds": 1770.5998087279877}
+2026-10-17T06:30:57Z INFO {"event": "training_progress", "step": 6300, "epoch": 17, "loss": 0.21474817119538783, "grad_norm": 0.3027653098106384, "train_seconds": 995.1935697849985, "remaining_budget_seconds": 1763.0276491989935}
+2026-10-17T06:31:05Z INFO {"event": "training_progress", "step": 6350, "epoch": 18, "loss": 0.2170732882618904, "grad_norm": 0.23623588681221008, "train_seconds": 1002.7794739359961, "remaining_budget_seconds": 1755.4417462739948}
+2026-10-17T06:31:15Z INFO {"event": "training_progress", "step": 6400, "epoch": 18, "loss": 0.21364225596189498, "grad_norm": 0.28234952688217163, "train_seconds": 1010.3593756560003, "remaining_budget_seconds": 1747.861843131992}
+2026-10-17T06:31:23Z INFO {"event": "training_progress", "step": 6450, "epoch": 18, "loss": 0.21390661716461182, "grad_norm": 0.24649281799793243, "train_seconds": 1018.3334046669916, "remaining_budget_seconds": 1739.8878134599945}
+2026-10-17T06:31:31Z INFO {"event": "training_progress", "step": 6500, "epoch": 18, "loss": 0.21604178041219713, "grad_norm": 0.2529313564300537, "train_seconds": 1026.002345817993, "remaining_budget_seconds": 1732.2188740949932}
+2026-10-17T06:31:39Z INFO {"event": "training_progress", "step": 6550, "epoch": 18, "loss": 0.2154915354400873, "grad_norm": 0.23569811880588531, "train_seconds": 1033.6815443319938, "remaining_budget_seconds": 1724.5396756979899}
+2026-10-17T06:31:46Z INFO {"event": "training_progress", "step": 6600, "epoch": 18, "loss": 0.21301396816968918, "grad_norm": 0.28058719635009766, "train_seconds": 1041.3645258289907, "remaining_budget_seconds": 1716.8566938680015}
+2026-10-17T06:31:58Z INFO {"event": "training_progress", "step": 6650, "epoch": 18, "loss": 0.21463014669716357, "grad_norm": 0.3410484790802002, "train_seconds": 1049.5480205549975, "remaining_budget_seconds": 1708.6731994919974}
+2026-10-17T06:32:06Z INFO {"event": "training_progress", "step": 6700, "epoch": 19, "loss": 0.21202912256121637, "grad_norm": 0.2347462773323059, "train_seconds": 1057.557144958002, "remaining_budget_seconds": 1700.664073232998}
+2026-10-17T06:32:14Z INFO {"event": "training_progress", "step": 6750, "epoch": 19, "loss": 0.21363258503377439, "grad_norm": 0.39894700050354004, "train_seconds": 1065.2013304769935, "remaining_budget_seconds": 1693.0198881099932}
+2026-10-17T06:32:21Z INFO {"event": "training_progress", "step": 6800, "epoch": 19, "loss": 0.21273767828941345, "grad_norm": 0.2518003284931183, "train_seconds": 1072.7879386610002, "remaining_budget_seconds": 1685.4332804939913}
+2026-10-17T06:32:32Z INFO {"event": "training_progress", "step": 6850, "epoch": 19, "loss": 0.21206520803272724, "grad_norm": 0.2709713876247406, "train_seconds": 1080.8757093919994, "remaining_budget_seconds": 1677.3455092849908}
+2026-10-17T06:32:41Z INFO {"event": "training_progress", "step": 6900, "epoch": 19, "loss": 0.21198268584907054, "grad_norm": 0.36465102434158325, "train_seconds": 1089.0305606049951, "remaining_budget_seconds": 1669.1906575030007}
+2026-10-17T06:32:48Z INFO {"event": "training_progress", "step": 6950, "epoch": 19, "loss": 0.21190824881196021, "grad_norm": 0.2723015546798706, "train_seconds": 1096.9220626439928, "remaining_budget_seconds": 1661.2991569949954}
+2026-10-17T06:32:56Z INFO {"event": "training_progress", "step": 7000, "epoch": 19, "loss": 0.21476556770503522, "grad_norm": 0.2623066008090973, "train_seconds": 1104.8191740449984, "remaining_budget_seconds": 1653.4020441629982}
+2026-10-17T06:33:07Z INFO {"event": "training_progress", "step": 7050, "epoch": 20, "loss": 0.21184806406497955, "grad_norm": 0.24717234075069427, "train_seconds": 1112.7572312669945, "remaining_budget_seconds": 1645.4639849660016}
+2026-10-17T06:33:15Z INFO {"event": "training_progress", "step": 7100, "epoch": 20, "loss": 0.20931062363088132, "grad_norm": 0.36191606521606445, "train_seconds": 1120.8722741459933, "remaining_budget_seconds": 1637.3489452919894}
+2026-10-17T06:33:23Z INFO {"event": "training_progress", "step": 7150, "epoch": 20, "loss": 0.20896621443331242, "grad_norm": 0.17356930673122406, "train_seconds": 1129.1010476969968, "remaining_budget_seconds": 1629.1201687189896}
+2026-10-17T06:33:31Z INFO {"event": "training_progress", "step": 7200, "epoch": 20, "loss": 0.21159220896661282, "grad_norm": 0.21727073192596436, "train_seconds": 1136.9237193000008, "remaining_budget_seconds": 1621.2974989889917}
+2026-10-17T06:33:42Z INFO {"event": "training_progress", "step": 7250, "epoch": 20, "loss": 0.21354237526655198, "grad_norm": 0.19906765222549438, "train_seconds": 1144.8545873909898, "remaining_budget_seconds": 1613.3666320060001}
+2026-10-17T06:33:50Z INFO {"event": "training_progress", "step": 7300, "epoch": 20, "loss": 0.210043710693717, "grad_norm": 0.3381940424442291, "train_seconds": 1152.678713700996, "remaining_budget_seconds": 1605.542505127989}
+2026-10-17T06:33:57Z INFO {"event": "training_progress", "step": 7350, "epoch": 20, "loss": 0.21247626639902592, "grad_norm": 0.3457295596599579, "train_seconds": 1160.4936575049942, "remaining_budget_seconds": 1597.727562629996}
+2026-10-17T06:34:05Z INFO {"event": "training_progress", "step": 7400, "epoch": 21, "loss": 0.21362198017537593, "grad_norm": 0.26587045192718506, "train_seconds": 1168.2810809099901, "remaining_budget_seconds": 1589.940137794998}
+2026-10-17T06:34:16Z INFO {"event": "training_progress", "step": 7450, "epoch": 21, "loss": 0.21017569452524185, "grad_norm": 0.21149902045726776, "train_seconds": 1175.9413869150012, "remaining_budget_seconds": 1582.279833210996}
+2026-10-17T06:34:23Z INFO {"event": "training_progress", "step": 7500, "epoch": 21, "loss": 0.21378100469708441, "grad_norm": 0.309298574924469, "train_seconds": 1183.7483515149943, "remaining_budget_seconds": 1574.4728681799897}
+2026-10-17T06:34:31Z INFO {"event": "training_progress", "step": 7550, "epoch": 21, "loss": 0.21444288767874242, "grad_norm": 0.28780344128608704, "train_seconds": 1191.7338567190018, "remaining_budget_seconds": 1566.487361513995}
+2026-10-17T06:34:39Z INFO {"event": "training_progress", "step": 7600, "epoch": 21, "loss": 0.20844829760491848, "grad_norm": 0.23348569869995117, "train_seconds": 1199.5081117559894, "remaining_budget_seconds": 1558.7131039939995}
+2026-10-17T06:34:50Z INFO {"event": "training_progress", "step": 7650, "epoch": 21, "loss": 0.21059861436486244, "grad_norm": 0.2881143391132355, "train_seconds": 1207.4051772309904, "remaining_budget_seconds": 1550.8160392589925}
+2026-10-17T06:34:58Z INFO {"event": "training_progress", "step": 7700, "epoch": 21, "loss": 0.21166083037853242, "grad_norm": 0.24514462053775787, "train_seconds": 1215.2980003720004, "remaining_budget_seconds": 1542.9232182809938}
+2026-10-17T06:35:06Z INFO {"event": "training_progress", "step": 7750, "epoch": 22, "loss": 0.21556138440966607, "grad_norm": 0.16975194215774536, "train_seconds": 1223.2441348499997, "remaining_budget_seconds": 1534.9770825239975}
+2026-10-17T06:35:14Z INFO {"event": "training_progress", "step": 7800, "epoch": 22, "loss": 0.21349207431077957, "grad_norm": 0.17212574183940887, "train_seconds": 1231.139628298988, "remaining_budget_seconds": 1527.0815921780013}
+2026-10-17T06:35:25Z INFO {"event": "training_progress", "step": 7850, "epoch": 22, "loss": 0.20977546989917756, "grad_norm": 0.20683804154396057, "train_seconds": 1239.285113484002, "remaining_budget_seconds": 1518.9361049709987}
+2026-10-17T06:35:32Z INFO {"event": "training_progress", "step": 7900, "epoch": 22, "loss": 0.2131910376995802, "grad_norm": 0.2890412211418152, "train_seconds": 1247.224016327993, "remaining_budget_seconds": 1510.9972008869954}
+2026-10-17T06:35:41Z INFO {"event": "training_progress", "step": 7950, "epoch": 22, "loss": 0.21142928540706635, "grad_norm": 0.23489676415920258, "train_seconds": 1255.2954016149888, "remaining_budget_seconds": 1502.9258148919907}
+2026-10-17T06:35:48Z INFO {"event": "training_progress", "step": 8000, "epoch": 22, "loss": 0.20862337008118628, "grad_norm": 0.23820820450782776, "train_seconds": 1263.189306811997, "remaining_budget_seconds": 1495.0319094000006}
+2026-10-17T06:35:59Z INFO {"event": "training_progress", "step": 8050, "epoch": 22, "loss": 0.21064610473811626, "grad_norm": 0.17739921808242798, "train_seconds": 1270.8059393769945, "remaining_budget_seconds": 1487.4152801949967}
+2026-10-17T06:36:07Z INFO {"event": "training_progress", "step": 8100, "epoch": 23, "loss": 0.21179107315838336, "grad_norm": 0.29210153222084045, "train_seconds": 1278.560498093997, "remaining_budget_seconds": 1479.6607217579876}
+2026-10-17T06:36:15Z INFO {"event": "training_progress", "step": 8150, "epoch": 23, "loss": 0.21046625688672066, "grad_norm": 0.29260578751564026, "train_seconds": 1286.5133894799947, "remaining_budget_seconds": 1471.7078284399904}
+2026-10-17T06:36:22Z INFO {"event": "training_progress", "step": 8200, "epoch": 23, "loss": 0.21308900199830533, "grad_norm": 0.23153471946716309, "train_seconds": 1294.3605744689994, "remaining_budget_seconds": 1463.8606439399882}
+2026-10-17T06:36:33Z INFO {"event": "training_progress", "step": 8250, "epoch": 23, "loss": 0.2102259775996208, "grad_norm": 0.3016645610332489, "train_seconds": 1302.197912292002, "remaining_budget_seconds": 1456.0233044929919}
+2026-10-17T06:36:41Z INFO {"event": "training_progress", "step": 8300, "epoch": 23, "loss": 0.21103645876049995, "grad_norm": 0.33297526836395264, "train_seconds": 1310.0228520529927, "remaining_budget_seconds": 1448.1983679459954}
+2026-10-17T06:36:49Z INFO {"event": "training_progress", "step": 8350, "epoch": 23, "loss": 0.21245137684047222, "grad_norm": 0.2007666379213333, "train_seconds": 1317.9571042119933, "remaining_budget_seconds": 1440.2641138829931}
+2026-10-17T06:36:57Z INFO {"event": "training_progress", "step": 8400, "epoch": 23, "loss": 0.21224195025861264, "grad_norm": 0.25062859058380127, "train_seconds": 1325.713042089992, "remaining_budget_seconds": 1432.5081769749959}
+2026-10-17T06:37:07Z INFO {"event": "training_progress", "step": 8450, "epoch": 24, "loss": 0.21264866031706334, "grad_norm": 0.25242772698402405, "train_seconds": 1333.466451752989, "remaining_budget_seconds": 1424.7547672379878}
+2026-10-17T06:37:15Z INFO {"event": "training_progress", "step": 8500, "epoch": 24, "loss": 0.21025844402611255, "grad_norm": 0.24382533133029938, "train_seconds": 1341.384392455002, "remaining_budget_seconds": 1416.8368255989917}
+2026-10-17T06:37:23Z INFO {"event": "training_progress", "step": 8550, "epoch": 24, "loss": 0.21102447651326656, "grad_norm": 0.18141523003578186, "train_seconds": 1349.028785175993, "remaining_budget_seconds": 1409.1924331689952}
+2026-10-17T06:37:30Z INFO {"event": "training_progress", "step": 8600, "epoch": 24, "loss": 0.2094156504422426, "grad_norm": 0.20076628029346466, "train_seconds": 1356.543468140997, "remaining_budget_seconds": 1401.6777516250004}
+2026-10-17T06:37:38Z INFO {"event": "training_progress", "step": 8650, "epoch": 24, "loss": 0.2059295117110014, "grad_norm": 0.3899737000465393, "train_seconds": 1364.4036618190003, "remaining_budget_seconds": 1393.8175569039886}
+2026-10-17T06:37:48Z INFO {"event": "training_progress", "step": 8700, "epoch": 24, "loss": 0.2118746581673622, "grad_norm": 0.23736672103405, "train_seconds": 1372.0406321359915, "remaining_budget_seconds": 1386.1805867849907}
+2026-10-17T06:37:57Z INFO {"event": "training_progress", "step": 8750, "epoch": 24, "loss": 0.2112443158775568, "grad_norm": 0.2045997977256775, "train_seconds": 1380.1720160429977, "remaining_budget_seconds": 1378.049202661001}
+2026-10-17T06:38:04Z INFO {"event": "training_progress", "step": 8800, "epoch": 25, "loss": 0.20690092124044895, "grad_norm": 0.19909073412418365, "train_seconds": 1387.8698936479923, "remaining_budget_seconds": 1370.3513246229995}
+2026-10-17T06:38:12Z INFO {"event": "training_progress", "step": 8850, "epoch": 25, "loss": 0.21027196444571017, "grad_norm": 0.28534066677093506, "train_seconds": 1395.6840967280004, "remaining_budget_seconds": 1362.537121165995}
+2026-10-17T06:38:22Z INFO {"event": "training_progress", "step": 8900, "epoch": 25, "loss": 0.21155953459441662, "grad_norm": 0.3083724081516266, "train_seconds": 1403.3021984959923, "remaining_budget_seconds": 1354.9190186759952}
+2026-10-17T06:38:31Z INFO {"event": "training_progress", "step": 8950, "epoch": 25, "loss": 0.20987045288085937, "grad_norm": 0.1705908626317978, "train_seconds": 1411.3982536739932, "remaining_budget_seconds": 1346.8229647199914}
+2026-10-17T06:38:38Z INFO {"event": "training_progress", "step": 9000, "epoch": 25, "loss": 0.20699717156589031, "grad_norm": 0.2439018338918686, "train_seconds": 1419.0009563939966, "remaining_budget_seconds": 1339.2202608939988}
+2026-10-17T06:38:46Z INFO {"event": "training_progress", "step": 9050, "epoch": 25, "loss": 0.20848954677581788, "grad_norm": 0.19655311107635498, "train_seconds": 1426.7149759079912, "remaining_budget_seconds": 1331.5062445840013}
+2026-10-17T06:38:56Z INFO {"event": "training_progress", "step": 9100, "epoch": 25, "loss": 0.2094003726541996, "grad_norm": 0.23187188804149628, "train_seconds": 1434.5290789009887, "remaining_budget_seconds": 1323.6921393169905}
+2026-10-17T06:39:04Z INFO {"event": "training_progress", "step": 9150, "epoch": 26, "loss": 0.20932894945144653, "grad_norm": 0.24344025552272797, "train_seconds": 1442.3943787229946, "remaining_budget_seconds": 1315.8268393809994}
+2026-10-17T06:39:12Z INFO {"event": "training_progress", "step": 9200, "epoch": 26, "loss": 0.20785485677421092, "grad_norm": 0.24847282469272614, "train_seconds": 1449.964227561999, "remaining_budget_seconds": 1308.256990540991}
+2026-10-17T06:39:19Z INFO {"event": "training_progress", "step": 9250, "epoch": 26, "loss": 0.2083331260085106, "grad_norm": 0.23448261618614197, "train_seconds": 1457.548790389992, "remaining_budget_seconds": 1300.6724278139882}
+2026-10-17T06:39:30Z INFO {"event": "training_progress", "step": 9300, "epoch": 26, "loss": 0.2091069394350052, "grad_norm": 0.18608133494853973, "train_seconds": 1465.0746408389969, "remaining_budget_seconds": 1293.1465789409995}
+2026-10-17T06:39:38Z INFO {"event": "training_progress", "step": 9350, "epoch": 26, "loss": 0.20806187354028224, "grad_norm": 0.168502539396286, "train_seconds": 1472.7863955179928, "remaining_budget_seconds": 1285.4348227149894}
+2026-10-17T06:39:45Z INFO {"event": "training_progress", "step": 9400, "epoch": 26, "loss": 0.20902829885482788, "grad_norm": 0.20174887776374817, "train_seconds": 1480.6690902569972, "remaining_budget_seconds": 1277.5521298999956}
+2026-10-17T06:39:53Z INFO {"event": "training_progress", "step": 9450, "epoch": 26, "loss": 0.20884532928466798, "grad_norm": 0.18616576492786407, "train_seconds": 1488.543697022993, "remaining_budget_seconds": 1269.6775215199887}
+2026-10-17T06:40:04Z INFO {"event": "training_progress", "step": 9500, "epoch": 27, "loss": 0.20734742119908334, "grad_norm": 0.1633692979812622, "train_seconds": 1496.6906066910014, "remaining_budget_seconds": 1261.5306136309955}
+2026-10-17T06:40:12Z INFO {"event": "training_progress", "step": 9550, "epoch": 27, "loss": 0.20967667877674104, "grad_norm": 0.26126667857170105, "train_seconds": 1504.6241762309946, "remaining_budget_seconds": 1253.5970416700002}
+2026-10-17T06:40:20Z INFO {"event": "training_progress", "step": 9600, "epoch": 27, "loss": 0.20971706219017505, "grad_norm": 0.23650963604450226, "train_seconds": 1512.5093808739912, "remaining_budget_seconds": 1245.711839244992}
+2026-10-17T06:40:28Z INFO {"event": "training_progress", "step": 9650, "epoch": 27, "loss": 0.20963012851774693, "grad_norm": 0.2437625229358673, "train_seconds": 1520.006094240991, "remaining_budget_seconds": 1238.2151238139922}
+2026-10-17T06:40:38Z INFO {"event": "training_progress", "step": 9700, "epoch": 27, "loss": 0.20671652212738992, "grad_norm": 0.250139057636261, "train_seconds": 1527.539682289993, "remaining_budget_seconds": 1230.6815377489984}
+2026-10-17T06:40:46Z INFO {"event": "training_progress", "step": 9750, "epoch": 27, "loss": 0.20731545232236384, "grad_norm": 0.240016907453537, "train_seconds": 1535.3160248679924, "remaining_budget_seconds": 1222.905195157}
+2026-10-17T06:40:54Z INFO {"event": "training_progress", "step": 9800, "epoch": 27, "loss": 0.20977834552526475, "grad_norm": 0.18026690185070038, "train_seconds": 1543.2424171559978, "remaining_budget_seconds": 1214.9788016669918}
+2026-10-17T06:41:02Z INFO {"event": "training_progress", "step": 9850, "epoch": 28, "loss": 0.21067674949765205, "grad_norm": 0.29966339468955994, "train_seconds": 1551.1231336490018, "remaining_budget_seconds": 1207.0980844809965}
+2026-10-17T06:41:12Z INFO {"event": "training_progress", "step": 9900, "epoch": 28, "loss": 0.20992211379110814, "grad_norm": 0.17998231947422028, "train_seconds": 1558.8533035500004, "remaining_budget_seconds": 1199.3679157430015}
+2026-10-17T06:41:20Z INFO {"event": "training_progress", "step": 9950, "epoch": 28, "loss": 0.20471700675785542, "grad_norm": 0.198791965842247, "train_seconds": 1566.4210878009908, "remaining_budget_seconds": 1191.8001301759941}
+2026-10-17T06:41:28Z INFO {"event": "training_progress", "step": 10000, "epoch": 28, "loss": 0.20792209669947626, "grad_norm": 0.38189563155174255, "train_seconds": 1574.2945478420006, "remaining_budget_seconds": 1183.9266719820007}
+2026-10-17T06:41:36Z INFO {"event": "training_progress", "step": 10050, "epoch": 28, "loss": 0.21199581816792487, "grad_norm": 0.22019319236278534, "train_seconds": 1582.3076597899926, "remaining_budget_seconds": 1175.9135599059955}
+2026-10-17T06:41:44Z INFO {"event": "training_progress", "step": 10100, "epoch": 28, "loss": 0.20892803251743317, "grad_norm": 0.26406148076057434, "train_seconds": 1590.4591422889935, "remaining_budget_seconds": 1167.7620780409925}
+2026-10-17T06:41:55Z INFO {"event": "training_progress", "step": 10150, "epoch": 28, "loss": 0.20910865187644959, "grad_norm": 0.16735641658306122, "train_seconds": 1598.1227415780013, "remaining_budget_seconds": 1160.0984787689958}
+2026-10-17T06:42:02Z INFO {"event": "training_progress", "step": 10200, "epoch": 29, "loss": 0.2081767102330923, "grad_norm": 0.20143333077430725, "train_seconds": 1605.9476321949915, "remaining_budget_seconds": 1152.2735861789988}
+2026-10-17T06:42:10Z INFO {"event": "training_progress", "step": 10250, "epoch": 29, "loss": 0.2072686606645584, "grad_norm": 0.20326614379882812, "train_seconds": 1614.035556361996, "remaining_budget_seconds": 1144.1856634680007}
+2026-10-17T06:42:18Z INFO {"event": "training_progress", "step": 10300, "epoch": 29, "loss": 0.20747787952423097, "grad_norm": 0.22896775603294373, "train_seconds": 1621.910922540992, "remaining_budget_seconds": 1136.3102951399924}
+2026-10-17T06:42:29Z INFO {"event": "training_progress", "step": 10350, "epoch": 29, "loss": 0.20990434676408767, "grad_norm": 0.252536803483963, "train_seconds": 1629.92143183999, "remaining_budget_seconds": 1128.2997863469936}
+2026-10-17T06:42:37Z INFO {"event": "training_progress", "step": 10400, "epoch": 29, "loss": 0.206152770370245, "grad_norm": 0.2248266190290451, "train_seconds": 1637.7176829029922, "remaining_budget_seconds": 1120.5035366230004}
+2026-10-17T06:42:45Z INFO {"event": "training_progress", "step": 10450, "epoch": 29, "loss": 0.20922394767403601, "grad_norm": 0.23835115134716034, "train_seconds": 1646.176101043995, "remaining_budget_seconds": 1112.045116562993}
+2026-10-17T06:42:53Z INFO {"event": "training_progress", "step": 10500, "epoch": 29, "loss": 0.208537323102355, "grad_norm": 0.18309664726257324, "train_seconds": 1654.0169961319916, "remaining_budget_seconds": 1104.2042238860013}
+2026-10-17T06:43:04Z INFO {"event": "training_progress", "step": 10550, "epoch": 30, "loss": 0.20564709931612016, "grad_norm": 0.14978639781475067, "train_seconds": 1661.8232914299879, "remaining_budget_seconds": 1096.3979273079894}
+2026-10-17T06:43:12Z INFO {"event": "training_progress", "step": 10600, "epoch": 30, "loss": 0.20418001748621464, "grad_norm": 0.16789186000823975, "train_seconds": 1669.7267072999966, "remaining_budget_seconds": 1088.4945128159889}
+2026-10-17T06:43:20Z INFO {"event": "training_progress", "step": 10650, "epoch": 30, "loss": 0.20872004479169845, "grad_norm": 0.26974886655807495, "train_seconds": 1677.5427511399903, "remaining_budget_seconds": 1080.6784664699953}
+2026-10-17T06:43:28Z INFO {"event": "training_progress", "step": 10700, "epoch": 30, "loss": 0.20655062921345235, "grad_norm": 0.20656718313694, "train_seconds": 1685.515436228001, "remaining_budget_seconds": 1072.7057840749912}
+2026-10-17T06:43:38Z INFO {"event": "training_progress", "step": 10750, "epoch": 30, "loss": 0.20864452823996543, "grad_norm": 0.19486947357654572, "train_seconds": 1693.334200348996, "remaining_budget_seconds": 1064.8870197329961}
+2026-10-17T06:43:46Z INFO {"event": "training_progress", "step": 10800, "epoch": 30, "loss": 0.20710798919200898, "grad_norm": 0.1921011209487915, "train_seconds": 1701.1045345729945, "remaining_budget_seconds": 1057.1166835179902}
+2026-10-17T06:43:54Z INFO {"event": "training_progress", "step": 10850, "epoch": 30, "loss": 0.20745471209287644, "grad_norm": 0.17478099465370178, "train_seconds": 1708.7335673310008, "remaining_budget_seconds": 1049.487650015988}
+2026-10-17T06:44:02Z INFO {"event": "training_progress", "step": 10900, "epoch": 31, "loss": 0.20606576785445213, "grad_norm": 0.21477928757667542, "train_seconds": 1716.5625163599907, "remaining_budget_seconds": 1041.6586775319884}
+2026-10-17T06:44:12Z INFO {"event": "training_progress", "step": 10950, "epoch": 31, "loss": 0.20829749993979932, "grad_norm": 0.3015783429145813, "train_seconds": 1724.1594639029936, "remaining_budget_seconds": 1034.0617545329878}
+2026-10-17T06:44:20Z INFO {"event": "training_progress", "step": 11000, "epoch": 31, "loss": 0.2086262845993042, "grad_norm": 0.14239469170570374, "train_seconds": 1732.1916953179898, "remaining_budget_seconds": 1026.029520718992}
+2026-10-17T06:44:28Z INFO {"event": "training_progress", "step": 11050, "epoch": 31, "loss": 0.20854291014373302, "grad_norm": 0.23979739844799042, "train_seconds": 1739.9159708200023, "remaining_budget_seconds": 1018.3052447480004}
+2026-10-17T06:44:36Z INFO {"event": "training_progress", "step": 11100, "epoch": 31, "loss": 0.20619553521275522, "grad_norm": 0.20393475890159607, "train_seconds": 1747.8777994089905, "remaining_budget_seconds": 1010.343420389996}
+2026-10-17T06:44:47Z INFO {"event": "training_progress", "step": 11150, "epoch": 31, "loss": 0.20619359143078328, "grad_norm": 0.209952712059021, "train_seconds": 1755.9344881369907, "remaining_budget_seconds": 1002.2867319639918}
+2026-10-17T06:44:55Z INFO {"event": "training_progress", "step": 11200, "epoch": 31, "loss": 0.20738924257457256, "grad_norm": 0.17169706523418427, "train_seconds": 1763.6188626829971, "remaining_budget_seconds": 994.6023553879932}
+2026-10-17T06:45:02Z INFO {"event": "training_progress", "step": 11250, "epoch": 31, "loss": 0.2088747302442789, "grad_norm": 0.19977518916130066, "train_seconds": 1771.2066325429914, "remaining_budget_seconds": 987.0145870359993}
+2026-10-17T06:45:10Z INFO {"event": "training_progress", "step": 11300, "epoch": 32, "loss": 0.20649185813963414, "grad_norm": 0.18649625778198242, "train_seconds": 1778.6324829629884, "remaining_budget_seconds": 979.5887364459923}
+2026-10-17T06:45:21Z INFO {"event": "training_progress", "step": 11350, "epoch": 32, "loss": 0.20543449580669404, "grad_norm": 0.15457011759281158, "train_seconds": 1786.528188309996, "remaining_budget_seconds": 971.693029778995}
+2026-10-17T06:45:29Z INFO {"event": "training_progress", "step": 11400, "epoch": 32, "loss": 0.2084383987635374, "grad_norm": 0.22192299365997314, "train_seconds": 1794.371386090992, "remaining_budget_seconds": 963.8498340020014}
+2026-10-17T06:45:36Z INFO {"event": "training_progress", "step": 11450, "epoch": 32, "loss": 0.2041545283049345, "grad_norm": 0.17271234095096588, "train_seconds": 1802.011733677995, "remaining_budget_seconds": 956.2094830689894}
+2026-10-17T06:45:44Z INFO {"event": "training_progress", "step": 11500, "epoch": 32, "loss": 0.20593056157231332, "grad_norm": 0.13901636004447937, "train_seconds": 1809.819895060995, "remaining_budget_seconds": 948.4013252879959}
+2026-10-17T06:45:55Z INFO {"event": "training_progress", "step": 11550, "epoch": 32, "loss": 0.2060591322183609, "grad_norm": 0.21440042555332184, "train_seconds": 1817.5558555919997, "remaining_budget_seconds": 940.6653641859884}
+2026-10-17T06:46:03Z INFO {"event": "training_progress", "step": 11600, "epoch": 32, "loss": 0.20499382205307484, "grad_norm": 0.18244387209415436, "train_seconds": 1825.6548963879904, "remaining_budget_seconds": 932.5663237229892}
+2026-10-17T06:46:11Z INFO {"event": "training_progress", "step": 11650, "epoch": 33, "loss": 0.2089501679688692, "grad_norm": 0.25282150506973267, "train_seconds": 1833.4721418019908, "remaining_budget_seconds": 924.7490768749994}
+2026-10-17T06:46:19Z INFO {"event": "training_progress", "step": 11700, "epoch": 33, "loss": 0.2043175957351923, "grad_norm": 0.1754787266254425, "train_seconds": 1841.7543280289974, "remaining_budget_seconds": 916.4668871319882}
+2026-10-17T06:46:30Z INFO {"event": "training_progress", "step": 11750, "epoch": 33, "loss": 0.20896793551743031, "grad_norm": 0.21432793140411377, "train_seconds": 1849.6491937789979, "remaining_budget_seconds": 908.5720237219939}
+2026-10-17T06:46:38Z INFO {"event": "training_progress", "step": 11800, "epoch": 33, "loss": 0.20122682251036167, "grad_norm": 0.15520066022872925, "train_seconds": 1857.8812360159936, "remaining_budget_seconds": 900.3399839129997}
+2026-10-17T06:46:46Z INFO {"event": "training_progress", "step": 11850, "epoch": 33, "loss": 0.20525543697178364, "grad_norm": 0.19260427355766296, "train_seconds": 1865.7787357880006, "remaining_budget_seconds": 892.4424839349958}
+2026-10-17T06:46:54Z INFO {"event": "training_progress", "step": 11900, "epoch": 33, "loss": 0.20380798898637295, "grad_norm": 0.20593677461147308, "train_seconds": 1873.066616393, "remaining_budget_seconds": 885.1546042319969}
+2026-10-17T06:47:01Z INFO {"event": "training_progress", "step": 11950, "epoch": 33, "loss": 0.2058164206147194, "grad_norm": 0.233494833111763, "train_seconds": 1880.7946742519998, "remaining_budget_seconds": 877.4265460039896}
+2026-10-17T06:47:12Z INFO {"event": "training_progress", "step": 12000, "epoch": 34, "loss": 0.2065538839995861, "grad_norm": 0.2510746121406555, "train_seconds": 1888.5159004689922, "remaining_budget_seconds": 869.7053200779919}
+2026-10-17T06:47:19Z INFO {"event": "training_progress", "step": 12050, "epoch": 34, "loss": 0.20548509329557418, "grad_norm": 0.20480316877365112, "train_seconds": 1896.070133065994, "remaining_budget_seconds": 862.1510871439968}
+2026-10-17T06:47:27Z INFO {"event": "training_progress", "step": 12100, "epoch": 34, "loss": 0.20893418677151204, "grad_norm": 0.18582095205783844, "train_seconds": 1903.8689504159993, "remaining_budget_seconds": 854.3522699199966}
+2026-10-17T06:47:34Z INFO {"event": "training_progress", "step": 12150, "epoch": 34, "loss": 0.20343999035656452, "grad_norm": 0.21617741882801056, "train_seconds": 1911.1321164039982, "remaining_budget_seconds": 847.0891025809979}
+2026-10-17T06:47:46Z INFO {"event": "training_progress", "step": 12200, "epoch": 34, "loss": 0.2035939521342516, "grad_norm": 0.18571901321411133, "train_seconds": 1919.1143258519878, "remaining_budget_seconds": 839.1068908640009}
+2026-10-17T06:47:55Z INFO {"event": "training_progress", "step": 12250, "epoch": 34, "loss": 0.20439254909753798, "grad_norm": 0.13488374650478363, "train_seconds": 1928.141580932992, "remaining_budget_seconds": 830.079637087998}
+2026-10-17T06:48:03Z INFO {"event": "training_progress", "step": 12300, "epoch": 34, "loss": 0.2052023034542799, "grad_norm": 0.13511128723621368, "train_seconds": 1936.2917478829913, "remaining_budget_seconds": 821.9294722349878}
+2026-10-17T06:48:11Z INFO {"event": "training_progress", "step": 12350, "epoch": 35, "loss": 0.20561822153627873, "grad_norm": 0.16673240065574646, "train_seconds": 1944.2464866419905, "remaining_budget_seconds": 813.9746666909923}
+2026-10-17T06:48:21Z INFO {"event": "training_progress", "step": 12400, "epoch": 35, "loss": 0.20605925634503364, "grad_norm": 0.15861380100250244, "train_seconds": 1951.7870919429988, "remaining_budget_seconds": 806.4341260949877}
+2026-10-17T06:48:28Z INFO {"event": "training_progress", "step": 12450, "epoch": 35, "loss": 0.20506505139172077, "grad_norm": 0.183832049369812, "train_seconds": 1959.363436707994, "remaining_budget_seconds": 798.8577835289907}
+2026-10-17T06:48:36Z INFO {"event": "training_progress", "step": 12500, "epoch": 35, "loss": 0.20606060557067393, "grad_norm": 0.2135191559791565, "train_seconds": 1966.9179381759895, "remaining_budget_seconds": 791.3032757129986}
+2026-10-17T06:48:43Z INFO {"event": "training_progress", "step": 12550, "epoch": 35, "loss": 0.2031797133386135, "grad_norm": 0.16003946959972382, "train_seconds": 1974.6656954889913, "remaining_budget_seconds": 783.5555227129953}
+2026-10-17T06:48:55Z INFO {"event": "training_progress", "step": 12600, "epoch": 35, "loss": 0.20552744887769223, "grad_norm": 0.26339080929756165, "train_seconds": 1982.6442521979916, "remaining_budget_seconds": 775.5769645989931}
+2026-10-17T06:49:04Z INFO {"event": "training_progress", "step": 12650, "epoch": 35, "loss": 0.20582616589963437, "grad_norm": 0.26691147685050964, "train_seconds": 1991.3267411580018, "remaining_budget_seconds": 766.8944765449996}
+2026-10-17T06:49:12Z INFO {"event": "training_progress", "step": 12700, "epoch": 36, "loss": 0.20587863631546496, "grad_norm": 0.2733739912509918, "train_seconds": 1999.4040638759907, "remaining_budget_seconds": 758.8171544119978}
+2026-10-17T06:49:20Z INFO {"event": "training_progress", "step": 12750, "epoch": 36, "loss": 0.2057138828933239, "grad_norm": 0.19187605381011963, "train_seconds": 2007.426594006989, "remaining_budget_seconds": 750.7946260009921}
+2026-10-17T06:49:29Z INFO {"event": "training_progress", "step": 12800, "epoch": 36, "loss": 0.20542144760489464, "grad_norm": 0.35593998432159424, "train_seconds": 2015.0953422019957, "remaining_budget_seconds": 743.1258755009912}
+2026-10-17T06:49:37Z INFO {"event": "training_progress", "step": 12850, "epoch": 36, "loss": 0.20542322605848312, "grad_norm": 0.14651162922382355, "train_seconds": 2022.9930012789991, "remaining_budget_seconds": 735.2282184529904}
+2026-10-17T06:49:45Z INFO {"event": "training_progress", "step": 12900, "epoch": 36, "loss": 0.20174520395696163, "grad_norm": 0.17968371510505676, "train_seconds": 2030.8991770239954, "remaining_budget_seconds": 727.3220403909945}
+2026-10-17T06:49:53Z INFO {"event": "training_progress", "step": 12950, "epoch": 36, "loss": 0.20304487131536006, "grad_norm": 0.22086384892463684, "train_seconds": 2038.7764210909954, "remaining_budget_seconds": 719.4447948059969}
+2026-10-17T06:50:04Z INFO {"event": "training_progress", "step": 13000, "epoch": 36, "loss": 0.20826566092669963, "grad_norm": 0.2281903326511383, "train_seconds": 2046.8157588859904, "remaining_budget_seconds": 711.4054590960004}
+2026-10-17T06:50:12Z INFO {"event": "training_progress", "step": 13050, "epoch": 37, "loss": 0.20679211303591727, "grad_norm": 0.23708920180797577, "train_seconds": 2055.1567234269896, "remaining_budget_seconds": 703.0644959829951}
+2026-10-17T06:50:20Z INFO {"event": "training_progress", "step": 13100, "epoch": 37, "loss": 0.2024442505091429, "grad_norm": 0.11819595098495483, "train_seconds": 2062.8477186689997, "remaining_budget_seconds": 695.3734974299878}
+2026-10-17T06:50:28Z INFO {"event": "training_progress", "step": 13150, "epoch": 37, "loss": 0.2050893994420767, "grad_norm": 0.27389460802078247, "train_seconds": 2070.478651641999, "remaining_budget_seconds": 687.7425667200005}
+2026-10-17T06:50:38Z INFO {"event": "training_progress", "step": 13200, "epoch": 37, "loss": 0.20434360310435296, "grad_norm": 0.1685611605644226, "train_seconds": 2078.3495821699908, "remaining_budget_seconds": 679.8716376699886}
+2026-10-17T06:50:46Z INFO {"event": "training_progress", "step": 13250, "epoch": 37, "loss": 0.20292616792023183, "grad_norm": 0.15167318284511566, "train_seconds": 2086.174242448993, "remaining_budget_seconds": 672.046977688995}
+2026-10-17T06:50:54Z INFO {"event": "training_progress", "step": 13300, "epoch": 37, "loss": 0.20179653137922288, "grad_norm": 0.14049096405506134, "train_seconds": 2094.031874744993, "remaining_budget_seconds": 664.1893440489948}
+2026-10-17T06:51:02Z INFO {"event": "training_progress", "step": 13350, "epoch": 37, "loss": 0.20362573444843293, "grad_norm": 0.16399073600769043, "train_seconds": 2102.275443161998, "remaining_budget_seconds": 655.9457757489872}
+2026-10-17T06:51:13Z INFO {"event": "training_progress", "step": 13400, "epoch": 38, "loss": 0.2056318850070238, "grad_norm": 0.26136475801467896, "train_seconds": 2110.341620678999, "remaining_budget_seconds": 647.8795952169894}
+2026-10-17T06:51:21Z INFO {"event": "training_progress", "step": 13450, "epoch": 38, "loss": 0.20414499714970588, "grad_norm": 0.2768358290195465, "train_seconds": 2118.2566050000023, "remaining_budget_seconds": 639.9643705669878}
+2026-10-17T06:51:29Z INFO {"event": "training_progress", "step": 13500, "epoch": 38, "loss": 0.2023255718499422, "grad_norm": 0.11922867596149445, "train_seconds": 2125.9023437089927, "remaining_budget_seconds": 632.3188763929938}
+2026-10-17T06:51:37Z INFO {"event": "training_progress", "step": 13550, "epoch": 38, "loss": 0.20637989051640035, "grad_norm": 0.17125394940376282, "train_seconds": 2133.6770146729978, "remaining_budget_seconds": 624.5442039409972}
+2026-10-17T06:51:48Z INFO {"event": "training_progress", "step": 13600, "epoch": 38, "loss": 0.20419657088816165, "grad_norm": 0.14482194185256958, "train_seconds": 2141.6163924769935, "remaining_budget_seconds": 616.6048262819968}
+2026-10-17T06:51:56Z INFO {"event": "training_progress", "step": 13650, "epoch": 38, "loss": 0.20774060502648353, "grad_norm": 0.21590781211853027, "train_seconds": 2149.4728274029912, "remaining_budget_seconds": 608.7483928500005}
+2026-10-17T06:52:03Z INFO {"event": "training_progress", "step": 13700, "epoch": 38, "loss": 0.20463906422257425, "grad_norm": 0.166121244430542, "train_seconds": 2157.2439062799967, "remaining_budget_seconds": 600.9773107469955}
+2026-10-17T06:52:11Z INFO {"event": "training_progress", "step": 13750, "epoch": 39, "loss": 0.2046719191968441, "grad_norm": 0.1526232212781906, "train_seconds": 2164.585919584002, "remaining_budget_seconds": 593.6353001309908}
+2026-10-17T06:52:22Z INFO {"event": "training_progress", "step": 13800, "epoch": 39, "loss": 0.20533158518373967, "grad_norm": 0.17545640468597412, "train_seconds": 2172.3327806669986, "remaining_budget_seconds": 585.8884389929881}
+2026-10-17T06:52:29Z INFO {"event": "training_progress", "step": 13850, "epoch": 39, "loss": 0.2053325640410185, "grad_norm": 0.17603451013565063, "train_seconds": 2180.1669082989974, "remaining_budget_seconds": 578.0543101259973}
+2026-10-17T06:52:37Z INFO {"event": "training_progress", "step": 13900, "epoch": 39, "loss": 0.2059980207681656, "grad_norm": 0.1634254902601242, "train_seconds": 2188.0479923749954, "remaining_budget_seconds": 570.1732262479927}
+2026-10-17T06:52:45Z INFO {"event": "training_progress", "step": 13950, "epoch": 39, "loss": 0.20442092426121236, "grad_norm": 0.17376577854156494, "train_seconds": 2195.9901029569883, "remaining_budget_seconds": 562.2311139989906}
+2026-10-17T06:52:56Z INFO {"event": "training_progress", "step": 14000, "epoch": 39, "loss": 0.2047081658244133, "grad_norm": 0.1326722502708435, "train_seconds": 2203.63158628199, "remaining_budget_seconds": 554.589633326992}
+2026-10-17T06:53:04Z INFO {"event": "training_progress", "step": 14050, "epoch": 39, "loss": 0.20415146455168723, "grad_norm": 0.21919021010398865, "train_seconds": 2211.3737211349944, "remaining_budget_seconds": 546.8474980389874}
+2026-10-17T06:53:11Z INFO {"event": "training_progress", "step": 14100, "epoch": 40, "loss": 0.20190833888947965, "grad_norm": 0.23007231950759888, "train_seconds": 2218.8530785259936, "remaining_budget_seconds": 539.3681399259949}
+2026-10-17T06:53:19Z INFO {"event": "training_progress", "step": 14150, "epoch": 40, "loss": 0.20308267667889596, "grad_norm": 0.16996777057647705, "train_seconds": 2226.675207176988, "remaining_budget_seconds": 531.5460131049913}
+2026-10-17T06:53:27Z INFO {"event": "training_progress", "step": 14200, "epoch": 40, "loss": 0.2053612072020769, "grad_norm": 0.17176616191864014, "train_seconds": 2234.3467604920006, "remaining_budget_seconds": 523.8744589589915}
+2026-10-17T06:53:38Z INFO {"event": "training_progress", "step": 14250, "epoch": 40, "loss": 0.2064024694263935, "grad_norm": 0.2542005181312561, "train_seconds": 2242.1570656149997, "remaining_budget_seconds": 516.0641525029932}
+2026-10-17T06:53:46Z INFO {"event": "training_progress", "step": 14300, "epoch": 40, "loss": 0.20313367649912834, "grad_norm": 0.29254356026649475, "train_seconds": 2249.900426127002, "remaining_budget_seconds": 508.3207941639994}
+2026-10-17T06:53:53Z INFO {"event": "training_progress", "step": 14350, "epoch": 40, "loss": 0.20446258641779422, "grad_norm": 0.18399208784103394, "train_seconds": 2257.504559174995, "remaining_budget_seconds": 500.71666057499533}
+2026-10-17T06:54:01Z INFO {"event": "training_progress", "step": 14400, "epoch": 40, "loss": 0.20040781863033771, "grad_norm": 0.18320538103580475, "train_seconds": 2265.474116696001, "remaining_budget_seconds": 492.7471016299969}
+2026-10-17T06:54:12Z INFO {"event": "training_progress", "step": 14450, "epoch": 41, "loss": 0.2026386158913374, "grad_norm": 0.160194531083107, "train_seconds": 2273.000575883998, "remaining_budget_seconds": 485.2206417409907}
+2026-10-17T06:54:20Z INFO {"event": "training_progress", "step": 14500, "epoch": 41, "loss": 0.20519260734319686, "grad_norm": 0.13754484057426453, "train_seconds": 2280.914593798996, "remaining_budget_seconds": 477.30662631298765}
+2026-10-17T06:54:28Z INFO {"event": "training_progress", "step": 14550, "epoch": 41, "loss": 0.2025008487701416, "grad_norm": 0.20801329612731934, "train_seconds": 2288.9178140250006, "remaining_budget_seconds": 469.30340378399706}
+2026-10-17T06:54:36Z INFO {"event": "training_progress", "step": 14600, "epoch": 41, "loss": 0.2045880326628685, "grad_norm": 0.18149444460868835, "train_seconds": 2296.61447421099, "remaining_budget_seconds": 461.60674423999444}
+2026-10-17T06:54:47Z INFO {"event": "training_progress", "step": 14650, "epoch": 41, "loss": 0.20460483759641648, "grad_norm": 0.13798978924751282, "train_seconds": 2305.625122294994, "remaining_budget_seconds": 452.5960947269923}
+2026-10-17T06:54:55Z INFO {"event": "training_progress", "step": 14700, "epoch": 41, "loss": 0.2038745207339525, "grad_norm": 0.15859320759773254, "train_seconds": 2313.3736816210003, "remaining_budget_seconds": 444.8475381839962}
+2026-10-17T06:55:02Z INFO {"event": "training_progress", "step": 14750, "epoch": 41, "loss": 0.20369226552546024, "grad_norm": 0.15507939457893372, "train_seconds": 2321.0958147799975, "remaining_budget_seconds": 437.12540459498996}
+2026-10-17T06:55:10Z INFO {"event": "training_progress", "step": 14800, "epoch": 42, "loss": 0.20376790225505828, "grad_norm": 0.2823501229286194, "train_seconds": 2329.091587206989, "remaining_budget_seconds": 429.1296314270003}
+2026-10-17T06:55:20Z INFO {"event": "training_progress", "step": 14850, "epoch": 42, "loss": 0.20244880311191082, "grad_norm": 0.18490184843540192, "train_seconds": 2336.7360359899903, "remaining_budget_seconds": 421.48518252499343}
+2026-10-17T06:55:28Z INFO {"event": "training_progress", "step": 14900, "epoch": 42, "loss": 0.20176301695406437, "grad_norm": 0.1892932504415512, "train_seconds": 2344.223819492996, "remaining_budget_seconds": 413.9973987019912}
+2026-10-17T06:55:35Z INFO {"event": "training_progress", "step": 14950, "epoch": 42, "loss": 0.20469888888299465, "grad_norm": 0.1484113186597824, "train_seconds": 2351.9064786980016, "remaining_budget_seconds": 406.3147409519879}
+2026-10-17T06:55:43Z INFO {"event": "training_progress", "step": 15000, "epoch": 42, "loss": 0.20252390004694462, "grad_norm": 0.19353589415550232, "train_seconds": 2359.9067787999957, "remaining_budget_seconds": 398.3144395959971}
+2026-10-17T06:55:54Z INFO {"event": "training_progress", "step": 15050, "epoch": 42, "loss": 0.2058820912986994, "grad_norm": 0.4541766047477722, "train_seconds": 2367.8331197389925, "remaining_budget_seconds": 390.3880969439924}
+2026-10-17T06:56:02Z INFO {"event": "training_progress", "step": 15100, "epoch": 42, "loss": 0.2038658981025219, "grad_norm": 0.1260318160057068, "train_seconds": 2375.534023845001, "remaining_budget_seconds": 382.687195282997}
+2026-10-17T06:56:10Z INFO {"event": "training_progress", "step": 15150, "epoch": 43, "loss": 0.20284746058285236, "grad_norm": 0.23423700034618378, "train_seconds": 2383.325947072997, "remaining_budget_seconds": 374.8952732359903}
+2026-10-17T06:56:18Z INFO {"event": "training_progress", "step": 15200, "epoch": 43, "loss": 0.20378739163279533, "grad_norm": 0.2349720448255539, "train_seconds": 2391.344669150989, "remaining_budget_seconds": 366.8765489159996}
+2026-10-17T06:56:28Z INFO {"event": "training_progress", "step": 15250, "epoch": 43, "loss": 0.2025800547003746, "grad_norm": 0.16816389560699463, "train_seconds": 2399.267247098993, "remaining_budget_seconds": 358.9539713809936}
+2026-10-17T06:56:36Z INFO {"event": "training_progress", "step": 15300, "epoch": 43, "loss": 0.19726849488914014, "grad_norm": 0.17712712287902832, "train_seconds": 2407.313981690997, "remaining_budget_seconds": 350.90723761999107}
+2026-10-17T06:56:44Z INFO {"event": "training_progress", "step": 15350, "epoch": 43, "loss": 0.20235530026257037, "grad_norm": 0.1613466441631317, "train_seconds": 2415.139091330988, "remaining_budget_seconds": 343.0821269559965}
+2026-10-17T06:56:52Z INFO {"event": "training_progress", "step": 15400, "epoch": 43, "loss": 0.20516164287924765, "grad_norm": 0.17841172218322754, "train_seconds": 2422.9507403219905, "remaining_budget_seconds": 335.27047987699916}
+2026-10-17T06:57:03Z INFO {"event": "training_progress", "step": 15450, "epoch": 43, "loss": 0.20319763116538525, "grad_norm": 0.18814611434936523, "train_seconds": 2431.3652004879987, "remaining_budget_seconds": 326.85601715400117}
+2026-10-17T06:57:11Z INFO {"event": "training_progress", "step": 15500, "epoch": 44, "loss": 0.20023954957723616, "grad_norm": 0.18509891629219055, "train_seconds": 2439.323554978997, "remaining_budget_seconds": 318.8976628859964}
+2026-10-17T06:57:19Z INFO {"event": "training_progress", "step": 15550, "epoch": 44, "loss": 0.2011230855435133, "grad_norm": 0.16458310186862946, "train_seconds": 2447.1737292799953, "remaining_budget_seconds": 311.0474895919906}
+2026-10-17T06:57:27Z INFO {"event": "training_progress", "step": 15600, "epoch": 44, "loss": 0.2023391305655241, "grad_norm": 0.20387966930866241, "train_seconds": 2455.0064516669954, "remaining_budget_seconds": 303.2147669209953}
+2026-10-17T06:57:37Z INFO {"event": "training_progress", "step": 15650, "epoch": 44, "loss": 0.1997825662791729, "grad_norm": 0.16172705590724945, "train_seconds": 2462.6945404019934, "remaining_budget_seconds": 295.52667931700125}
+2026-10-17T06:57:45Z INFO {"event": "training_progress", "step": 15700, "epoch": 44, "loss": 0.2012400320917368, "grad_norm": 0.2068691849708557, "train_seconds": 2470.574156785995, "remaining_budget_seconds": 287.6470631199918}
+2026-10-17T06:57:53Z INFO {"event": "training_progress", "step": 15750, "epoch": 44, "loss": 0.20396058388054372, "grad_norm": 0.16739360988140106, "train_seconds": 2478.6234246209933, "remaining_budget_seconds": 279.59779253898887}
+2026-10-17T06:58:01Z INFO {"event": "training_progress", "step": 15800, "epoch": 44, "loss": 0.20159869879484177, "grad_norm": 0.16892848908901215, "train_seconds": 2486.3330643619993, "remaining_budget_seconds": 271.88815578199865}
+2026-10-17T06:58:11Z INFO {"event": "training_progress", "step": 15850, "epoch": 45, "loss": 0.20155722193419934, "grad_norm": 0.14428381621837616, "train_seconds": 2493.9846429460013, "remaining_budget_seconds": 264.23657460400136}
+2026-10-17T06:58:19Z INFO {"event": "training_progress", "step": 15900, "epoch": 45, "loss": 0.205288890004158, "grad_norm": 0.127704456448555, "train_seconds": 2502.048992971002, "remaining_budget_seconds": 256.17222569099977}
+2026-10-17T06:58:27Z INFO {"event": "training_progress", "step": 15950, "epoch": 45, "loss": 0.2012444518506527, "grad_norm": 0.16653671860694885, "train_seconds": 2509.7541814180004, "remaining_budget_seconds": 248.46703864299343}
+2026-10-17T06:58:34Z INFO {"event": "training_progress", "step": 16000, "epoch": 45, "loss": 0.2039499846100807, "grad_norm": 0.1666935831308365, "train_seconds": 2517.5272540509905, "remaining_budget_seconds": 240.6939662579971}
+2026-10-17T06:58:42Z INFO {"event": "training_progress", "step": 16050, "epoch": 45, "loss": 0.20215526059269906, "grad_norm": 0.3045313060283661, "train_seconds": 2525.273723004997, "remaining_budget_seconds": 232.94749722100096}
+2026-10-17T06:58:53Z INFO {"event": "training_progress", "step": 16100, "epoch": 45, "loss": 0.20405189126729964, "grad_norm": 0.24798619747161865, "train_seconds": 2533.3859256119904, "remaining_budget_seconds": 224.8352855629928}
+2026-10-17T06:59:01Z INFO {"event": "training_progress", "step": 16150, "epoch": 45, "loss": 0.2045450873672962, "grad_norm": 0.17237797379493713, "train_seconds": 2541.398203784993, "remaining_budget_seconds": 216.82301611099683}
+2026-10-17T06:59:09Z INFO {"event": "training_progress", "step": 16200, "epoch": 46, "loss": 0.19947660386562346, "grad_norm": 0.1839933842420578, "train_seconds": 2549.11766532599, "remaining_budget_seconds": 209.10355430499476}
+2026-10-17T06:59:17Z INFO {"event": "training_progress", "step": 16250, "epoch": 46, "loss": 0.199812021702528, "grad_norm": 0.19961605966091156, "train_seconds": 2557.147813424992, "remaining_budget_seconds": 201.07340392698825}
+2026-10-17T06:59:28Z INFO {"event": "training_progress", "step": 16300, "epoch": 46, "loss": 0.20179016679525374, "grad_norm": 0.12725019454956055, "train_seconds": 2565.3370482509927, "remaining_budget_seconds": 192.88416930299718}
+2026-10-17T06:59:35Z INFO {"event": "training_progress", "step": 16350, "epoch": 46, "loss": 0.20330994755029677, "grad_norm": 0.17625784873962402, "train_seconds": 2573.217689565994, "remaining_budget_seconds": 185.00353001098847}
+2026-10-17T06:59:44Z INFO {"event": "training_progress", "step": 16400, "epoch": 46, "loss": 0.20164781250059605, "grad_norm": 0.18062390387058258, "train_seconds": 2581.223138767993, "remaining_budget_seconds": 176.99808129399025}
+2026-10-17T06:59:51Z INFO {"event": "training_progress", "step": 16450, "epoch": 46, "loss": 0.20233974061906337, "grad_norm": 0.17482997477054596, "train_seconds": 2589.076577971995, "remaining_budget_seconds": 169.14463969199278}
+2026-10-17T07:00:02Z INFO {"event": "training_progress", "step": 16500, "epoch": 46, "loss": 0.2022358613461256, "grad_norm": 0.19365696609020233, "train_seconds": 2597.226409222989, "remaining_budget_seconds": 160.99481031698815}
+2026-10-17T07:00:10Z INFO {"event": "training_progress", "step": 16550, "epoch": 47, "loss": 0.2029309691488743, "grad_norm": 0.2084585428237915, "train_seconds": 2605.030664318998, "remaining_budget_seconds": 153.19055436000053}
+2026-10-17T07:00:18Z INFO {"event": "training_progress", "step": 16600, "epoch": 47, "loss": 0.20201003238558768, "grad_norm": 0.1370936781167984, "train_seconds": 2612.5909103209997, "remaining_budget_seconds": 145.63030825699389}
+2026-10-17T07:00:25Z INFO {"event": "training_progress", "step": 16650, "epoch": 47, "loss": 0.20161313585937024, "grad_norm": 0.1483106017112732, "train_seconds": 2620.353469993992, "remaining_budget_seconds": 137.8677497919998}
+2026-10-17T07:00:36Z INFO {"event": "training_progress", "step": 16700, "epoch": 47, "loss": 0.20065634690225123, "grad_norm": 0.1851779818534851, "train_seconds": 2628.3085665239923, "remaining_budget_seconds": 129.91265353299968}
+2026-10-17T07:00:44Z INFO {"event": "training_progress", "step": 16750, "epoch": 47, "loss": 0.20168462835252285, "grad_norm": 0.18821412324905396, "train_seconds": 2636.1105996699916, "remaining_budget_seconds": 122.11061900199275}
+2026-10-17T07:00:52Z INFO {"event": "training_progress", "step": 16800, "epoch": 47, "loss": 0.2007866182178259, "grad_norm": 0.13557882606983185, "train_seconds": 2644.225238692001, "remaining_budget_seconds": 113.99597901800007}
+2026-10-17T07:01:01Z INFO {"event": "training_progress", "step": 16850, "epoch": 47, "loss": 0.20400909669697284, "grad_norm": 0.1410677433013916, "train_seconds": 2652.994396021997, "remaining_budget_seconds": 105.22682368398819}
+2026-10-17T07:01:12Z INFO {"event": "training_progress", "step": 16900, "epoch": 48, "loss": 0.20166947588324546, "grad_norm": 0.1040268987417221, "train_seconds": 2661.0455315610016, "remaining_budget_seconds": 97.17568569499417}
+2026-10-17T07:01:20Z INFO {"event": "training_progress", "step": 16950, "epoch": 48, "loss": 0.20066443838179113, "grad_norm": 0.1651110053062439, "train_seconds": 2668.9644984629995, "remaining_budget_seconds": 89.25672010600101}
+2026-10-17T07:01:28Z INFO {"event": "training_progress", "step": 17000, "epoch": 48, "loss": 0.2013324923068285, "grad_norm": 0.232211172580719, "train_seconds": 2676.8832903669972, "remaining_budget_seconds": 81.3379281929956}
+2026-10-17T07:01:36Z INFO {"event": "training_progress", "step": 17050, "epoch": 48, "loss": 0.20297404542565345, "grad_norm": 0.18932557106018066, "train_seconds": 2684.8543510659947, "remaining_budget_seconds": 73.36686748499051}
+2026-10-17T07:01:46Z INFO {"event": "training_progress", "step": 17100, "epoch": 48, "loss": 0.20096327379345894, "grad_norm": 0.13374604284763336, "train_seconds": 2692.6224750260008, "remaining_budget_seconds": 65.59873993799556}
+2026-10-17T07:01:54Z INFO {"event": "training_finished", "status": "completed", "dataset": "cifar10", "family": "dit", "seed": 103, "step": 17148, "train_seconds": 2700.118430960996, "parameters": 1298864, "device": "cuda", "initial_validation_mse": 1.594213217496872, "final_validation_mse": 0.20122884213924408, "images_per_second": 812.9065654423165, "source_revisions": ["c749856d1c9fe07d6ce2f39d01de78f664aea9a8"], "peak_memory_bytes": 161050112, "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8", "failure": null}

study/main/cifar10-dit-s103/training.json

@@ -0,0 +1,19 @@
+{
+  "status": "completed",
+  "dataset": "cifar10",
+  "family": "dit",
+  "seed": 103,
+  "step": 17148,
+  "train_seconds": 2700.118430960996,
+  "parameters": 1298864,
+  "device": "cuda",
+  "initial_validation_mse": 1.594213217496872,
+  "final_validation_mse": 0.20122884213924408,
+  "images_per_second": 812.9065654423165,
+  "source_revisions": [
+    "c749856d1c9fe07d6ce2f39d01de78f664aea9a8"
+  ],
+  "peak_memory_bytes": 161050112,
+  "checkpoint_sha256": "e737abf235c2b51f00a3e768fc58f991acaee1881c8e43ad60fa46de0b8ecdd8",
+  "failure": null
+}

study/main/cifar10-dit-s104/README.md

@@ -0,0 +1,30 @@
+---
+license: mit
+pipeline_tag: unconditional-image-generation
+tags:
+- pytorch
+- flow-matching
+- oscillatory-neural-networks
+---
+# cifar10 / dit / seed 104
+
+This is an unconditional pixel-space velocity network from the [oscillatory flow matching study](https://github.com/BurnyCoder/flow-matching-with-oscillatory-neural-networks-instead-of-transformer).
+Recorded role: **completed**; status: **completed**. The model is a research artifact; sample quality and limitations are documented in its accompanying evaluations and the public reports. Fixed-batch diagnostics deliberately memorize a few examples and are not general-purpose generators. No performance advantage is assumed.
+
+## Loading and sampling
+
+Install the locked environment from the source repository. Download this entire run subdirectory, including `config.json` and `model.safetensors`, then use:
+
+```bash
+ofm sample --run /path/to/downloaded/run --device cpu --count 64 --nfe 32
+```
+
+The public Python loader is `ofm.checkpoints.load_generator(path, device="cpu")`; models accept `forward(images, times)` and return image-shaped velocities. This export contains **EMA weights**. `checkpoint.pt` additionally preserves the available training/diagnostic state; ordinary training checkpoints are resumable with `ofm train --resume /path/to/run` after downloading the official dataset. Untrained and fixed-batch diagnostic checkpoints do not contain a resumable ordinary training loop.
+
+## Training and evaluation
+
+Dataset: cifar10; architecture: dit; seed: 104. This is a from-scratch generator, with no text conditioning or pretrained image encoder. The exact architecture, optimizer, data manifests, source revisions, and metrics are included as JSON. Generated tensors are complete and unclipped; PNG grids show the first fixed samples after display clipping. KID/FID are CIFAR diagnostics and are not classification accuracy. Untrained, fixed-batch, or step-limited checkpoints are explicitly labeled and must not be treated as final-budget primary results.
+
+## Limitations and provenance
+
+Small native-resolution datasets and a short local compute budget limit image quality and external generalization. Six paired seeds are exploratory; inference interventions alone do not establish an oscillator mechanism. Dataset content and evaluator biases remain relevant. See the [protocol](https://github.com/BurnyCoder/flow-matching-with-oscillatory-neural-networks-instead-of-transformer/blob/main/docs/protocol.md) and [reports](https://github.com/BurnyCoder/flow-matching-with-oscillatory-neural-networks-instead-of-transformer/tree/main/reports) for all results, including failures. This export contains no downloaded dataset or credentials. `manifest.json` records file hashes; the original checkpoint SHA-256 is `931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a`. Export code revision: `13c72f9474e2fb46be8f209e15dba220f91ef4d5`.

study/main/cifar10-dit-s104/checkpoint.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a
+size 21305024

study/main/cifar10-dit-s104/config.json

@@ -0,0 +1,32 @@
+{
+  "model": {
+    "family": "dit",
+    "channels": 3,
+    "size": 32,
+    "width": 128,
+    "depth": 4,
+    "patch": 4,
+    "rotating_dim": 4,
+    "inner_steps": 4,
+    "step_size": 0.1,
+    "ffn_hidden": 576
+  },
+  "dataset": "cifar10",
+  "seed": 104,
+  "device": "cuda",
+  "batch_size": 128,
+  "microbatch": 32,
+  "learning_rate": 0.0003,
+  "weight_decay": 0.01,
+  "clip_grad": 1.0,
+  "ema_decay": 0.999,
+  "minutes": 45,
+  "max_steps": 0,
+  "validation_size": 5000,
+  "data_root": "data",
+  "runs_root": "runs",
+  "budget_hours": 24.0,
+  "log_every": 50,
+  "checkpoint_seconds": 300.0,
+  "amp": true
+}

study/main/cifar10-dit-s104/environment.json

@@ -0,0 +1,130 @@
+{
+  "python": "3.12.3 (main, Aug 31 2026, 10:18:26) [GCC 13.3.0]",
+  "torch": "2.11.0+cu128",
+  "cuda_build": "12.8",
+  "device": "NVIDIA GeForce RTX 5070 Laptop GPU",
+  "git_revision": "c749856d1c9fe07d6ce2f39d01de78f664aea9a8",
+  "deterministic_algorithms": true,
+  "float32_matmul_precision": "high",
+  "packages": {
+    "Jinja2": "3.1.6",
+    "MarkupSafe": "3.0.4",
+    "POT": "0.9.7.post1",
+    "PyYAML": "6.0.3",
+    "Pygments": "2.21.0",
+    "absl-py": "2.5.1",
+    "aiohappyeyeballs": "2.7.1",
+    "aiohttp": "3.14.5",
+    "aiosignal": "1.4.0",
+    "anyio": "4.15.1",
+    "asttokens": "3.0.2",
+    "attrs": "26.1.0",
+    "click": "8.5.0",
+    "cloudpickle": "3.1.2",
+    "comm": "0.2.3",
+    "contourpy": "1.4.0",
+    "cuda-bindings": "12.9.10",
+    "cuda-pathfinder": "1.8.3",
+    "cuda-toolkit": "12.8.1",
+    "cycler": "0.12.1",
+    "debugpy": "1.8.22",
+    "executing": "2.3.0",
+    "filelock": "4.2.0",
+    "fonttools": "4.67.0",
+    "frozenlist": "1.8.0",
+    "fsspec": "2026.9.0",
+    "h11": "0.16.0",
+    "hf-xet": "1.7.0",
+    "httpcore2": "2.13.1",
+    "httpx2": "2.13.1",
+    "huggingface_hub": "2.2.0",
+    "idna": "3.20",
+    "iniconfig": "2.3.1",
+    "ipykernel": "7.4.0",
+    "ipython": "9.17.1",
+    "ipython_pygments_lexers": "1.1.1",
+    "ipywidgets": "8.1.9",
+    "jedi": "0.20.1",
+    "joblib": "1.6.0",
+    "jupyter_client": "8.10.0",
+    "jupyter_core": "5.9.1",
+    "jupyterlab_widgets": "3.0.17",
+    "kiwisolver": "1.5.1",
+    "lightning-utilities": "0.15.3",
+    "matplotlib": "3.11.2",
+    "matplotlib-inline": "0.2.2",
+    "mpmath": "1.3.0",
+    "multidict": "7.1.0",
+    "narwhals": "2.27.1",
+    "nest-asyncio2": "1.7.4",
+    "networkx": "3.7",
+    "numpy": "2.5.4",
+    "nvidia-cublas-cu12": "12.8.4.1",
+    "nvidia-cuda-cupti-cu12": "12.8.90",
+    "nvidia-cuda-nvrtc-cu12": "12.8.93",
+    "nvidia-cuda-runtime-cu12": "12.8.90",
+    "nvidia-cudnn-cu12": "9.19.0.56",
+    "nvidia-cufft-cu12": "11.3.3.83",
+    "nvidia-cufile-cu12": "1.13.1.3",
+    "nvidia-curand-cu12": "10.3.9.90",
+    "nvidia-cusolver-cu12": "11.7.3.90",
+    "nvidia-cusparse-cu12": "12.5.8.93",
+    "nvidia-cusparselt-cu12": "0.7.1",
+    "nvidia-nccl-cu12": "2.28.9",
+    "nvidia-nvjitlink-cu12": "12.8.93",
+    "nvidia-nvshmem-cu12": "3.4.5",
+    "nvidia-nvtx-cu12": "12.8.90",
+    "oscillatory-flow-matching": "0.1.0",
+    "packaging": "26.3",
+    "pandas": "3.0.6",
+    "parso": "0.8.7",
+    "pastel": "0.2.1",
+    "pexpect": "4.9.0",
+    "pillow": "12.3.0",
+    "platformdirs": "4.13.3",
+    "pluggy": "1.7.0",
+    "poethepoet": "0.10.0",
+    "prompt_toolkit": "3.0.53",
+    "propcache": "0.5.4",
+    "psutil": "7.2.2",
+    "ptyprocess": "0.7.0",
+    "pure_eval": "0.2.4",
+    "pypandoc_binary": "1.17",
+    "pyparsing": "3.3.3",
+    "pytest": "9.1.1",
+    "python-dateutil": "2.9.0.post0",
+    "python-dotenv": "1.2.4",
+    "pytorch-lightning": "2.6.6",
+    "pyzmq": "27.2.0",
+    "ruff": "0.17.1",
+    "safetensors": "0.8.0",
+    "scikit-learn": "1.9.1",
+    "scipy": "1.18.1",
+    "setuptools": "81.0.0",
+    "six": "1.17.0",
+    "stack-data": "0.6.3",
+    "sympy": "1.14.0",
+    "threadpoolctl": "3.7.0",
+    "tomlkit": "0.15.2",
+    "torch": "2.11.0+cu128",
+    "torch_fidelity": "0.4.0",
+    "torchcde": "0.2.5",
+    "torchcfm": "1.0.7",
+    "torchdiffeq": "0.2.5",
+    "torchdyn": "1.0.6",
+    "torchmetrics": "1.9.0",
+    "torchsde": "0.2.6",
+    "torchvision": "0.26.0+cu128",
+    "tornado": "6.5.10",
+    "tqdm": "4.70.1",
+    "traitlets": "5.16.1",
+    "trampoline": "0.1.2",
+    "triton": "3.6.0",
+    "truststore": "0.10.5",
+    "typing_extensions": "4.16.0",
+    "typst": "0.15.0",
+    "wcwidth": "0.9.2",
+    "widgetsnbextension": "4.0.16",
+    "yarl": "1.25.1"
+  }
+}

study/main/cifar10-dit-s104/evaluations/test-nfe32/evaluation.json

@@ -0,0 +1,40 @@
+{
+  "dataset": "cifar10",
+  "family": "dit",
+  "training_seed": 104,
+  "split": "test",
+  "intervention": "none",
+  "generated_count": 5000,
+  "real_count": 10000,
+  "sampling_seed": 20261016,
+  "metric_seed": 20261016,
+  "nfe": 32,
+  "kid_subsets": 100,
+  "kid_subset_size": 1000,
+  "checkpoint_sha256": "931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a",
+  "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97",
+  "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06",
+  "primary_eligible": true,
+  "device": "cuda",
+  "blas_threads": 4,
+  "status": "completed",
+  "sampling": {
+    "count": 5000,
+    "seed": 20261016,
+    "nfe": 32,
+    "batch_size": 32,
+    "seconds": 67.29390516399872,
+    "seconds_per_image": 0.013458781032799743,
+    "precision": "float32",
+    "timing_excludes": "model load and warmup",
+    "clipped_fraction": 0.01615540310740471,
+    "peak_memory_bytes": 58955776
+  },
+  "metrics": {
+    "frechet_inception_distance": 105.1405045558987,
+    "kernel_inception_distance_mean": 0.0926866340637207,
+    "kernel_inception_distance_std": 0.0019192328223508978
+  },
+  "validation_velocity_mse": 0.1988416574895382,
+  "metric_seconds": 19.559638132996042
+}

study/main/cifar10-dit-s104/evaluations/test-nfe32/evaluation_config.json

@@ -0,0 +1,20 @@
+{
+  "dataset": "cifar10",
+  "family": "dit",
+  "training_seed": 104,
+  "split": "test",
+  "intervention": "none",
+  "generated_count": 5000,
+  "real_count": 10000,
+  "sampling_seed": 20261016,
+  "metric_seed": 20261016,
+  "nfe": 32,
+  "kid_subsets": 100,
+  "kid_subset_size": 1000,
+  "checkpoint_sha256": "931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a",
+  "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97",
+  "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06",
+  "primary_eligible": true,
+  "device": "cuda",
+  "blas_threads": 4
+}

study/main/cifar10-dit-s104/evaluations/test-nfe32/run.log

@@ -0,0 +1,11 @@
+2026-10-17T18:09:07Z INFO {"event": "evaluation_start", "dataset": "cifar10", "family": "dit", "training_seed": 104, "split": "test", "intervention": "none", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06", "primary_eligible": true, "device": "cuda", "blas_threads": 4}
+2026-10-17T18:09:09Z INFO {"event": "sampling_progress", "completed": 32, "total": 5000, "nfe": 32}
+2026-10-17T18:09:20Z INFO {"event": "sampling_progress", "completed": 672, "total": 5000, "nfe": 32}
+2026-10-17T18:09:27Z INFO {"event": "sampling_progress", "completed": 1312, "total": 5000, "nfe": 32}
+2026-10-17T18:09:35Z INFO {"event": "sampling_progress", "completed": 1952, "total": 5000, "nfe": 32}
+2026-10-17T18:09:42Z INFO {"event": "sampling_progress", "completed": 2592, "total": 5000, "nfe": 32}
+2026-10-17T18:09:53Z INFO {"event": "sampling_progress", "completed": 3232, "total": 5000, "nfe": 32}
+2026-10-17T18:10:01Z INFO {"event": "sampling_progress", "completed": 3872, "total": 5000, "nfe": 32}
+2026-10-17T18:10:14Z INFO {"event": "sampling_progress", "completed": 4512, "total": 5000, "nfe": 32}
+2026-10-17T18:10:27Z INFO {"event": "sampling_complete", "directory": "runs/study/main/cifar10-dit-s104/evaluations/test-nfe32", "count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 67.29390516399872, "seconds_per_image": 0.013458781032799743, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.01615540310740471, "peak_memory_bytes": 58955776}
+2026-10-17T18:10:47Z INFO {"event": "evaluation_complete", "dataset": "cifar10", "family": "dit", "training_seed": 104, "split": "test", "intervention": "none", "generated_count": 5000, "real_count": 10000, "sampling_seed": 20261016, "metric_seed": 20261016, "nfe": 32, "kid_subsets": 100, "kid_subset_size": 1000, "checkpoint_sha256": "931c047b6ac7ef2a657e898a26f94ab432497b05a07e2c281eef36f50e20588a", "data_sha256": "51ca348f81e4c280d10942cf2fa0d4dd1ce688a16b0088283069b54bbaaf6f97", "source_revision": "5cb68c27c6dddcba6b49df92776493e18f828f06", "primary_eligible": true, "device": "cuda", "blas_threads": 4, "status": "completed", "sampling": {"count": 5000, "seed": 20261016, "nfe": 32, "batch_size": 32, "seconds": 67.29390516399872, "seconds_per_image": 0.013458781032799743, "precision": "float32", "timing_excludes": "model load and warmup", "clipped_fraction": 0.01615540310740471, "peak_memory_bytes": 58955776}, "metrics": {"frechet_inception_distance": 105.1405045558987, "kernel_inception_distance_mean": 0.0926866340637207, "kernel_inception_distance_std": 0.0019192328223508978}, "validation_velocity_mse": 0.1988416574895382, "metric_seconds": 19.559638132996042}