| {"step": 14201, "loss": 3.326014280319214, "lr": 3.525264394829612e-05, "elapsed_sec": 32.856996297836304, "step_time_sec": 32.80085609498201, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 9876475392} |
| {"step": 14202, "loss": 3.430114269256592, "lr": 3.5211267605633805e-05, "elapsed_sec": 53.07335877418518, "step_time_sec": 20.21630221803207, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12536096256} |
| {"step": 14203, "loss": 3.0527024269104004, "lr": 3.516998827667058e-05, "elapsed_sec": 73.32396197319031, "step_time_sec": 20.25048040499678, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14204, "loss": 3.1614513397216797, "lr": 3.51288056206089e-05, "elapsed_sec": 93.55463981628418, "step_time_sec": 20.230452782008797, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14205, "loss": 3.193356513977051, "lr": 3.508771929824562e-05, "elapsed_sec": 113.79088258743286, "step_time_sec": 20.23602553398814, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14206, "loss": 3.2387802600860596, "lr": 3.504672897196262e-05, "elapsed_sec": 134.0397665500641, "step_time_sec": 20.248610740993172, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14207, "loss": 3.234318494796753, "lr": 3.500583430571762e-05, "elapsed_sec": 154.28360867500305, "step_time_sec": 20.243666027032305, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14208, "loss": 3.0101470947265625, "lr": 3.4965034965034965e-05, "elapsed_sec": 174.5314974784851, "step_time_sec": 20.247731239011046, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14209, "loss": 3.065960645675659, "lr": 3.492433061699651e-05, "elapsed_sec": 194.77540469169617, "step_time_sec": 20.24367621098645, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14210, "loss": 3.1022982597351074, "lr": 3.4883720930232556e-05, "elapsed_sec": 215.02032613754272, "step_time_sec": 20.244803178997245, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14211, "loss": 3.0942368507385254, "lr": 3.484320557491289e-05, "elapsed_sec": 235.26611423492432, "step_time_sec": 20.24564375198679, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14212, "loss": 2.820225715637207, "lr": 3.480278422273782e-05, "elapsed_sec": 255.51216983795166, "step_time_sec": 20.245926293020602, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14213, "loss": 3.0561769008636475, "lr": 3.476245654692932e-05, "elapsed_sec": 275.756343126297, "step_time_sec": 20.244040971039794, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14214, "loss": 3.092322587966919, "lr": 3.472222222222222e-05, "elapsed_sec": 296.00151658058167, "step_time_sec": 20.24497652001446, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14215, "loss": 3.090698480606079, "lr": 3.46820809248555e-05, "elapsed_sec": 316.2462475299835, "step_time_sec": 20.24443377298303, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14216, "loss": 2.882643461227417, "lr": 3.464203233256351e-05, "elapsed_sec": 336.49201250076294, "step_time_sec": 20.245643769972958, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14217, "loss": 3.082810878753662, "lr": 3.460207612456748e-05, "elapsed_sec": 356.73663115501404, "step_time_sec": 20.244433832995128, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14218, "loss": 3.1174566745758057, "lr": 3.456221198156682e-05, "elapsed_sec": 376.9817841053009, "step_time_sec": 20.24501561600482, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14219, "loss": 3.083685874938965, "lr": 3.452243958573073e-05, "elapsed_sec": 397.2268602848053, "step_time_sec": 20.244831698015332, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14220, "loss": 2.70790696144104, "lr": 3.4482758620689657e-05, "elapsed_sec": 417.47352623939514, "step_time_sec": 20.246631649963092, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14221, "loss": 3.0516579151153564, "lr": 3.4443168771526975e-05, "elapsed_sec": 437.71810388565063, "step_time_sec": 20.244371090026107, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14222, "loss": 3.0419559478759766, "lr": 3.4403669724770645e-05, "elapsed_sec": 457.9633803367615, "step_time_sec": 20.245050727040507, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14223, "loss": 3.26985239982605, "lr": 3.436426116838488e-05, "elapsed_sec": 478.21104621887207, "step_time_sec": 20.24745466496097, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14224, "loss": 2.711554765701294, "lr": 3.432494279176202e-05, "elapsed_sec": 498.45885014533997, "step_time_sec": 20.247747126966715, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14225, "loss": 3.0412909984588623, "lr": 3.428571428571429e-05, "elapsed_sec": 518.7062938213348, "step_time_sec": 20.247230926994234, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14226, "loss": 3.306185722351074, "lr": 3.424657534246576e-05, "elapsed_sec": 538.954452753067, "step_time_sec": 20.24796016898472, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14227, "loss": 3.004136085510254, "lr": 3.420752565564424e-05, "elapsed_sec": 559.1980767250061, "step_time_sec": 20.243452478025574, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14228, "loss": 2.963714361190796, "lr": 3.416856492027335e-05, "elapsed_sec": 579.4434659481049, "step_time_sec": 20.245247831975576, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14229, "loss": 3.0929441452026367, "lr": 3.4129692832764505e-05, "elapsed_sec": 599.689733505249, "step_time_sec": 20.246140004019253, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14230, "loss": 3.068145751953125, "lr": 3.409090909090909e-05, "elapsed_sec": 619.9347589015961, "step_time_sec": 20.244760621979367, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14231, "loss": 2.976562738418579, "lr": 3.4052213393870606e-05, "elapsed_sec": 640.1789832115173, "step_time_sec": 20.244083293015137, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14232, "loss": 2.777550220489502, "lr": 3.401360544217687e-05, "elapsed_sec": 660.4275341033936, "step_time_sec": 20.248375858005602, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14233, "loss": 3.0061192512512207, "lr": 3.3975084937712346e-05, "elapsed_sec": 680.6716969013214, "step_time_sec": 20.243960318970494, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14234, "loss": 3.287689208984375, "lr": 3.393665158371041e-05, "elapsed_sec": 700.9177694320679, "step_time_sec": 20.245918015949428, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14235, "loss": 3.089726686477661, "lr": 3.389830508474577e-05, "elapsed_sec": 721.1639926433563, "step_time_sec": 20.246067653002683, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14236, "loss": 3.1028006076812744, "lr": 3.386004514672686e-05, "elapsed_sec": 741.4096949100494, "step_time_sec": 20.245483995997347, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14237, "loss": 2.979891538619995, "lr": 3.382187147688839e-05, "elapsed_sec": 761.6535356044769, "step_time_sec": 20.24370126100257, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14238, "loss": 3.05448579788208, "lr": 3.378378378378379e-05, "elapsed_sec": 781.9008166790009, "step_time_sec": 20.247151552990545, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14239, "loss": 3.023242235183716, "lr": 3.374578177727784e-05, "elapsed_sec": 802.1452660560608, "step_time_sec": 20.244261433021165, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14240, "loss": 3.170909881591797, "lr": 3.370786516853932e-05, "elapsed_sec": 822.3919031620026, "step_time_sec": 20.24642723996658, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14241, "loss": 3.0978312492370605, "lr": 3.367003367003367e-05, "elapsed_sec": 842.6374385356903, "step_time_sec": 20.245412093005143, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14242, "loss": 2.8913803100585938, "lr": 3.36322869955157e-05, "elapsed_sec": 862.8899161815643, "step_time_sec": 20.252345580025576, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14243, "loss": 3.0536351203918457, "lr": 3.35946248600224e-05, "elapsed_sec": 883.1362988948822, "step_time_sec": 20.24619279196486, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14244, "loss": 3.104767322540283, "lr": 3.3557046979865775e-05, "elapsed_sec": 903.3832108974457, "step_time_sec": 20.246665198006667, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14245, "loss": 2.962524890899658, "lr": 3.35195530726257e-05, "elapsed_sec": 923.6354978084564, "step_time_sec": 20.25215609895531, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14246, "loss": 2.9640769958496094, "lr": 3.348214285714286e-05, "elapsed_sec": 943.8888056278229, "step_time_sec": 20.25323325203499, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14247, "loss": 3.0278351306915283, "lr": 3.344481605351171e-05, "elapsed_sec": 964.1392753124237, "step_time_sec": 20.25023515400244, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14248, "loss": 2.9767239093780518, "lr": 3.34075723830735e-05, "elapsed_sec": 984.3932750225067, "step_time_sec": 20.253768098948058, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14249, "loss": 3.2613861560821533, "lr": 3.3370411568409345e-05, "elapsed_sec": 1004.638640165329, "step_time_sec": 20.24521486501908, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14250, "loss": 3.0163660049438477, "lr": 3.3333333333333335e-05, "elapsed_sec": 1024.890727519989, "step_time_sec": 78.71003663999727, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14251, "loss": 3.051971435546875, "lr": 3.329633740288568e-05, "elapsed_sec": 1103.5936512947083, "step_time_sec": 20.244686024961993, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14252, "loss": 3.0249013900756836, "lr": 3.325942350332594e-05, "elapsed_sec": 1123.8444130420685, "step_time_sec": 20.250656382006127, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14253, "loss": 3.0835001468658447, "lr": 3.322259136212625e-05, "elapsed_sec": 1144.0907323360443, "step_time_sec": 20.24606150999898, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14254, "loss": 3.2369232177734375, "lr": 3.3185840707964604e-05, "elapsed_sec": 1164.3359801769257, "step_time_sec": 20.245110250019934, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14255, "loss": 3.2457869052886963, "lr": 3.3149171270718233e-05, "elapsed_sec": 1184.5823588371277, "step_time_sec": 20.246166658995207, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14256, "loss": 3.0576624870300293, "lr": 3.3112582781456954e-05, "elapsed_sec": 1204.83407497406, "step_time_sec": 20.251580917974934, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14257, "loss": 3.0608038902282715, "lr": 3.307607497243661e-05, "elapsed_sec": 1225.0834589004517, "step_time_sec": 20.249164732987992, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14258, "loss": 3.0436534881591797, "lr": 3.3039647577092515e-05, "elapsed_sec": 1245.3356897830963, "step_time_sec": 20.252086902037263, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14259, "loss": 3.0499989986419678, "lr": 3.3003300330033e-05, "elapsed_sec": 1265.5830528736115, "step_time_sec": 20.24718170100823, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14260, "loss": 3.061340093612671, "lr": 3.296703296703297e-05, "elapsed_sec": 1285.829606294632, "step_time_sec": 20.246476543019526, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14261, "loss": 2.6028149127960205, "lr": 3.2930845225027445e-05, "elapsed_sec": 1306.0829391479492, "step_time_sec": 20.253119079978205, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14262, "loss": 3.1862916946411133, "lr": 3.289473684210526e-05, "elapsed_sec": 1326.3271102905273, "step_time_sec": 20.243930553959217, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14263, "loss": 3.1725847721099854, "lr": 3.285870755750274e-05, "elapsed_sec": 1346.5807814598083, "step_time_sec": 20.253457829996478, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14264, "loss": 3.154231309890747, "lr": 3.282275711159738e-05, "elapsed_sec": 1366.8326392173767, "step_time_sec": 20.25174448901089, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14265, "loss": 3.0478479862213135, "lr": 3.278688524590164e-05, "elapsed_sec": 1387.07843375206, "step_time_sec": 20.24556388903875, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14266, "loss": 3.123854875564575, "lr": 3.2751091703056775e-05, "elapsed_sec": 1407.3243007659912, "step_time_sec": 20.245702241023537, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14267, "loss": 3.107759475708008, "lr": 3.271537622682661e-05, "elapsed_sec": 1427.5700001716614, "step_time_sec": 20.245564211974852, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14268, "loss": 3.1172375679016113, "lr": 3.2679738562091506e-05, "elapsed_sec": 1447.8154792785645, "step_time_sec": 20.24531163898064, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14269, "loss": 2.9831154346466064, "lr": 3.264417845484223e-05, "elapsed_sec": 1468.0632693767548, "step_time_sec": 20.247606133983936, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14270, "loss": 3.0181291103363037, "lr": 3.260869565217392e-05, "elapsed_sec": 1488.3106715679169, "step_time_sec": 20.24730821000412, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14271, "loss": 3.071887731552124, "lr": 3.2573289902280134e-05, "elapsed_sec": 1508.5580706596375, "step_time_sec": 20.24718900496373, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14272, "loss": 3.0637927055358887, "lr": 3.253796095444686e-05, "elapsed_sec": 1528.8046555519104, "step_time_sec": 20.246417469985317, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14273, "loss": 3.004138946533203, "lr": 3.250270855904659e-05, "elapsed_sec": 1549.0521178245544, "step_time_sec": 20.24721804901492, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14274, "loss": 2.815593719482422, "lr": 3.246753246753247e-05, "elapsed_sec": 1569.297596693039, "step_time_sec": 20.245289771002717, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14275, "loss": 3.067631721496582, "lr": 3.2432432432432436e-05, "elapsed_sec": 1589.5464975833893, "step_time_sec": 20.248689144966193, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14276, "loss": 2.989835262298584, "lr": 3.239740820734342e-05, "elapsed_sec": 1609.790893316269, "step_time_sec": 20.244282644009218, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14277, "loss": 3.035227060317993, "lr": 3.236245954692557e-05, "elapsed_sec": 1630.0366847515106, "step_time_sec": 20.245549344981555, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14278, "loss": 2.9704177379608154, "lr": 3.232758620689655e-05, "elapsed_sec": 1650.2835445404053, "step_time_sec": 20.246673259942327, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14279, "loss": 3.447019100189209, "lr": 3.229278794402584e-05, "elapsed_sec": 1670.537933588028, "step_time_sec": 20.254225387994666, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14280, "loss": 3.2398951053619385, "lr": 3.2258064516129034e-05, "elapsed_sec": 1690.7829399108887, "step_time_sec": 20.24488888302585, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14281, "loss": 3.095768451690674, "lr": 3.22234156820623e-05, "elapsed_sec": 1711.0299561023712, "step_time_sec": 20.24679094797466, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14282, "loss": 2.9023778438568115, "lr": 3.218884120171674e-05, "elapsed_sec": 1731.275554895401, "step_time_sec": 20.24547490902478, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14283, "loss": 3.237497091293335, "lr": 3.2154340836012867e-05, "elapsed_sec": 1751.52077293396, "step_time_sec": 20.244980972027406, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14284, "loss": 2.913965940475464, "lr": 3.2119914346895075e-05, "elapsed_sec": 1771.7665286064148, "step_time_sec": 20.245683521963656, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14285, "loss": 3.3141417503356934, "lr": 3.20855614973262e-05, "elapsed_sec": 1792.020013809204, "step_time_sec": 20.253217659017537, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14286, "loss": 2.987138271331787, "lr": 3.205128205128205e-05, "elapsed_sec": 1812.2661809921265, "step_time_sec": 20.246050766960252, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14287, "loss": 3.0936648845672607, "lr": 3.2017075773746e-05, "elapsed_sec": 1832.5192775726318, "step_time_sec": 20.25286985200364, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14288, "loss": 3.0200934410095215, "lr": 3.1982942430703626e-05, "elapsed_sec": 1852.7640359401703, "step_time_sec": 20.244714363012463, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14289, "loss": 2.915832042694092, "lr": 3.194888178913738e-05, "elapsed_sec": 1873.0158631801605, "step_time_sec": 20.251542875019368, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14290, "loss": 2.778709888458252, "lr": 3.1914893617021275e-05, "elapsed_sec": 1893.2680068016052, "step_time_sec": 20.252040933992248, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14291, "loss": 3.4536991119384766, "lr": 3.1880977683315626e-05, "elapsed_sec": 1913.5150320529938, "step_time_sec": 20.246859991981182, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14292, "loss": 3.153477907180786, "lr": 3.184713375796179e-05, "elapsed_sec": 1933.7610864639282, "step_time_sec": 20.24589644599473, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14293, "loss": 2.926630973815918, "lr": 3.181336161187699e-05, "elapsed_sec": 1954.0147671699524, "step_time_sec": 20.253573944035452, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14294, "loss": 3.347106695175171, "lr": 3.177966101694915e-05, "elapsed_sec": 1974.25962972641, "step_time_sec": 20.244592437054962, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14295, "loss": 3.104076862335205, "lr": 3.1746031746031745e-05, "elapsed_sec": 1994.505154132843, "step_time_sec": 20.24536820303183, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14296, "loss": 3.021847724914551, "lr": 3.171247357293869e-05, "elapsed_sec": 2014.758764743805, "step_time_sec": 20.253436631988734, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14297, "loss": 3.1765811443328857, "lr": 3.1678986272439285e-05, "elapsed_sec": 2035.0056371688843, "step_time_sec": 20.246638229000382, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14298, "loss": 3.041292190551758, "lr": 3.1645569620253167e-05, "elapsed_sec": 2055.255085706711, "step_time_sec": 20.2493343980168, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14299, "loss": 2.954458713531494, "lr": 3.1612223393045315e-05, "elapsed_sec": 2075.5067613124847, "step_time_sec": 20.251489585032687, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14300, "loss": 3.177778959274292, "lr": 3.157894736842106e-05, "elapsed_sec": 2095.8021941184998, "step_time_sec": 103.58231412497116, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14301, "loss": 3.526448965072632, "lr": 3.154574132492114e-05, "elapsed_sec": 2199.333794116974, "step_time_sec": 20.244487776013557, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14302, "loss": 3.030207872390747, "lr": 3.151260504201681e-05, "elapsed_sec": 2219.5788345336914, "step_time_sec": 20.244791296019685, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14303, "loss": 2.986767053604126, "lr": 3.147953830010493e-05, "elapsed_sec": 2239.8232340812683, "step_time_sec": 20.244200907996856, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14304, "loss": 3.3092713356018066, "lr": 3.144654088050315e-05, "elapsed_sec": 2260.0670635700226, "step_time_sec": 20.243650268006604, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14305, "loss": 3.1539323329925537, "lr": 3.141361256544502e-05, "elapsed_sec": 2280.309940814972, "step_time_sec": 20.242812635959126, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14306, "loss": 3.1359660625457764, "lr": 3.1380753138075316e-05, "elapsed_sec": 2300.5546414852142, "step_time_sec": 20.24446077697212, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14307, "loss": 2.834909677505493, "lr": 3.134796238244514e-05, "elapsed_sec": 2320.807281732559, "step_time_sec": 20.252532458980568, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14308, "loss": 3.025994300842285, "lr": 3.13152400835073e-05, "elapsed_sec": 2341.1029558181763, "step_time_sec": 20.2954289970221, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14309, "loss": 3.067126750946045, "lr": 3.128258602711158e-05, "elapsed_sec": 2361.353232383728, "step_time_sec": 20.250053297029808, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14310, "loss": 3.0347940921783447, "lr": 3.125e-05, "elapsed_sec": 2381.5987343788147, "step_time_sec": 20.245357868028805, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14311, "loss": 3.07893967628479, "lr": 3.12174817898023e-05, "elapsed_sec": 2401.844511270523, "step_time_sec": 20.245585469005164, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14312, "loss": 3.3123128414154053, "lr": 3.118503118503119e-05, "elapsed_sec": 2422.0956377983093, "step_time_sec": 20.251032027998008, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14313, "loss": 3.151665687561035, "lr": 3.115264797507789e-05, "elapsed_sec": 2442.340573310852, "step_time_sec": 20.244737935019657, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14314, "loss": 3.052455425262451, "lr": 3.1120331950207475e-05, "elapsed_sec": 2462.586478471756, "step_time_sec": 20.24575374298729, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14315, "loss": 2.408630132675171, "lr": 3.108808290155441e-05, "elapsed_sec": 2482.8818407058716, "step_time_sec": 20.295169452030677, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14316, "loss": 2.962158679962158, "lr": 3.1055900621118014e-05, "elapsed_sec": 2503.1349124908447, "step_time_sec": 20.252918049984146, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14317, "loss": 3.066033124923706, "lr": 3.1023784901758014e-05, "elapsed_sec": 2523.3798308372498, "step_time_sec": 20.24474267900223, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14318, "loss": 3.0686659812927246, "lr": 3.099173553719008e-05, "elapsed_sec": 2543.6750843524933, "step_time_sec": 20.295087003032677, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14319, "loss": 3.065215826034546, "lr": 3.0959752321981425e-05, "elapsed_sec": 2563.9720497131348, "step_time_sec": 20.296727553009987, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14320, "loss": 2.908726692199707, "lr": 3.0927835051546395e-05, "elapsed_sec": 2584.224764108658, "step_time_sec": 20.25260426499881, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14321, "loss": 3.1743264198303223, "lr": 3.089598352214212e-05, "elapsed_sec": 2604.47646856308, "step_time_sec": 20.25157042295905, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14322, "loss": 2.9983158111572266, "lr": 3.08641975308642e-05, "elapsed_sec": 2624.7221205234528, "step_time_sec": 20.24548942700494, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14323, "loss": 3.2986068725585938, "lr": 3.083247687564235e-05, "elapsed_sec": 2645.0175507068634, "step_time_sec": 20.29518997995183, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14324, "loss": 3.1037676334381104, "lr": 3.080082135523614e-05, "elapsed_sec": 2665.267897605896, "step_time_sec": 20.250183215015568, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14325, "loss": 2.937708616256714, "lr": 3.076923076923077e-05, "elapsed_sec": 2685.515451192856, "step_time_sec": 20.247369970020372, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14326, "loss": 3.131836414337158, "lr": 3.0737704918032794e-05, "elapsed_sec": 2705.761915206909, "step_time_sec": 20.246368686959613, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14327, "loss": 3.346839427947998, "lr": 3.070624360286592e-05, "elapsed_sec": 2726.006814479828, "step_time_sec": 20.244777419022284, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14328, "loss": 2.7623889446258545, "lr": 3.067484662576687e-05, "elapsed_sec": 2746.2539479732513, "step_time_sec": 20.246850994008128, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14329, "loss": 3.104720115661621, "lr": 3.0643513789581204e-05, "elapsed_sec": 2766.4982640743256, "step_time_sec": 20.244102438970003, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14330, "loss": 3.0003204345703125, "lr": 3.061224489795918e-05, "elapsed_sec": 2786.749739408493, "step_time_sec": 20.251415303966496, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14331, "loss": 2.9977900981903076, "lr": 3.0581039755351684e-05, "elapsed_sec": 2806.995441675186, "step_time_sec": 20.245435582008213, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14332, "loss": 2.9013006687164307, "lr": 3.054989816700611e-05, "elapsed_sec": 2827.289441347122, "step_time_sec": 20.293828447000124, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14333, "loss": 3.1282310485839844, "lr": 3.051881993896236e-05, "elapsed_sec": 2847.53524184227, "step_time_sec": 20.24561606097268, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14334, "loss": 3.2402398586273193, "lr": 3.048780487804878e-05, "elapsed_sec": 2867.7874331474304, "step_time_sec": 20.25205808400642, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14335, "loss": 3.172844886779785, "lr": 3.045685279187818e-05, "elapsed_sec": 2888.030898332596, "step_time_sec": 20.24325596698327, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14336, "loss": 3.04007625579834, "lr": 3.0425963488843817e-05, "elapsed_sec": 2908.3250510692596, "step_time_sec": 20.29407237796113, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14337, "loss": 2.982228994369507, "lr": 3.039513677811551e-05, "elapsed_sec": 2928.6193993091583, "step_time_sec": 20.29419972602045, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14338, "loss": 2.9553866386413574, "lr": 3.0364372469635632e-05, "elapsed_sec": 2948.9139606952667, "step_time_sec": 20.29437808302464, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14339, "loss": 3.155527114868164, "lr": 3.0333670374115264e-05, "elapsed_sec": 2969.2084667682648, "step_time_sec": 20.294287731987424, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14340, "loss": 3.3011703491210938, "lr": 3.0303030303030302e-05, "elapsed_sec": 2989.5042345523834, "step_time_sec": 20.295594187045936, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14341, "loss": 3.126640558242798, "lr": 3.0272452068617564e-05, "elapsed_sec": 3009.7982115745544, "step_time_sec": 20.293820610037073, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14342, "loss": 2.7945666313171387, "lr": 3.024193548387097e-05, "elapsed_sec": 3030.0921211242676, "step_time_sec": 20.293798475991935, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14343, "loss": 3.1115591526031494, "lr": 3.0211480362537764e-05, "elapsed_sec": 3050.3467211723328, "step_time_sec": 20.254470523970667, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14344, "loss": 3.035567045211792, "lr": 3.0181086519114692e-05, "elapsed_sec": 3070.6000685691833, "step_time_sec": 20.253041434974875, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14345, "loss": 2.93070125579834, "lr": 3.0150753768844224e-05, "elapsed_sec": 3090.846444606781, "step_time_sec": 20.24621760600712, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14346, "loss": 2.8092472553253174, "lr": 3.012048192771085e-05, "elapsed_sec": 3111.093688726425, "step_time_sec": 20.247143196989782, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14347, "loss": 3.0345406532287598, "lr": 3.0090270812437316e-05, "elapsed_sec": 3131.344284057617, "step_time_sec": 20.250389819033444, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14348, "loss": 3.092236042022705, "lr": 3.0060120240480963e-05, "elapsed_sec": 3151.639260292053, "step_time_sec": 20.29479996598093, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14349, "loss": 3.301471710205078, "lr": 3.003003003003003e-05, "elapsed_sec": 3171.936121702194, "step_time_sec": 20.296681175997946, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14350, "loss": 2.897218704223633, "lr": 3.0000000000000004e-05, "elapsed_sec": 3192.231118917465, "step_time_sec": 76.909539329994, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14351, "loss": 2.974977493286133, "lr": 2.9970029970029973e-05, "elapsed_sec": 3269.0883524417877, "step_time_sec": 20.242395090986975, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14352, "loss": 3.0464024543762207, "lr": 2.994011976047904e-05, "elapsed_sec": 3289.3343522548676, "step_time_sec": 20.24589339195518, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14353, "loss": 3.1339354515075684, "lr": 2.9910269192422737e-05, "elapsed_sec": 3309.57905960083, "step_time_sec": 20.244539626990445, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14354, "loss": 3.1596548557281494, "lr": 2.9880478087649403e-05, "elapsed_sec": 3329.830394744873, "step_time_sec": 20.25116685195826, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14355, "loss": 3.0190837383270264, "lr": 2.9850746268656717e-05, "elapsed_sec": 3350.0864136219025, "step_time_sec": 20.25578761898214, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14356, "loss": 3.1247596740722656, "lr": 2.9821073558648116e-05, "elapsed_sec": 3370.3394610881805, "step_time_sec": 20.25303244101815, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14357, "loss": 2.6996028423309326, "lr": 2.9791459781529296e-05, "elapsed_sec": 3390.592534303665, "step_time_sec": 20.25293517904356, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14358, "loss": 2.9066720008850098, "lr": 2.9761904761904762e-05, "elapsed_sec": 3410.889263153076, "step_time_sec": 20.296479283017106, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14359, "loss": 3.1184680461883545, "lr": 2.9732408325074334e-05, "elapsed_sec": 3431.183988571167, "step_time_sec": 20.294532208994497, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14360, "loss": 3.156154155731201, "lr": 2.9702970297029706e-05, "elapsed_sec": 3451.479638338089, "step_time_sec": 20.295530700997915, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14361, "loss": 2.8187649250030518, "lr": 2.967359050445104e-05, "elapsed_sec": 3471.7750380039215, "step_time_sec": 20.29524505196605, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14362, "loss": 2.901840925216675, "lr": 2.9644268774703563e-05, "elapsed_sec": 3492.0690548419952, "step_time_sec": 20.293799236009363, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14363, "loss": 3.0660693645477295, "lr": 2.9615004935834158e-05, "elapsed_sec": 3512.3620216846466, "step_time_sec": 20.292749168991577, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14364, "loss": 3.041766405105591, "lr": 2.9585798816568047e-05, "elapsed_sec": 3532.656890153885, "step_time_sec": 20.29480946698459, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14365, "loss": 2.6588683128356934, "lr": 2.9556650246305422e-05, "elapsed_sec": 3552.9503219127655, "step_time_sec": 20.29312630003551, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14366, "loss": 3.033107280731201, "lr": 2.9527559055118114e-05, "elapsed_sec": 3573.2449464797974, "step_time_sec": 20.29452548798872, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14367, "loss": 3.272382974624634, "lr": 2.9498525073746314e-05, "elapsed_sec": 3593.5416281223297, "step_time_sec": 20.29653985699406, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14368, "loss": 3.2442550659179688, "lr": 2.9469548133595288e-05, "elapsed_sec": 3613.8357830047607, "step_time_sec": 20.293878594005946, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14369, "loss": 2.955573797225952, "lr": 2.9440628066732092e-05, "elapsed_sec": 3634.079775571823, "step_time_sec": 20.243853824969847, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14370, "loss": 3.0011277198791504, "lr": 2.9411764705882354e-05, "elapsed_sec": 3654.3272676467896, "step_time_sec": 20.24732323898934, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14371, "loss": 2.8591678142547607, "lr": 2.9382957884427035e-05, "elapsed_sec": 3674.5796864032745, "step_time_sec": 20.252246844000183, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14372, "loss": 3.1109297275543213, "lr": 2.9354207436399218e-05, "elapsed_sec": 3694.8730714321136, "step_time_sec": 20.293139518995304, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14373, "loss": 3.4279532432556152, "lr": 2.932551319648094e-05, "elapsed_sec": 3715.168060064316, "step_time_sec": 20.294869364995975, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14374, "loss": 3.177588939666748, "lr": 2.9296875000000006e-05, "elapsed_sec": 3735.461927652359, "step_time_sec": 20.293637392984238, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14375, "loss": 3.031513214111328, "lr": 2.9268292682926833e-05, "elapsed_sec": 3755.7547335624695, "step_time_sec": 20.292729823035188, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14376, "loss": 3.2436323165893555, "lr": 2.9239766081871346e-05, "elapsed_sec": 3776.0516142845154, "step_time_sec": 20.296719784033485, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14377, "loss": 3.0722944736480713, "lr": 2.9211295034079848e-05, "elapsed_sec": 3796.3439741134644, "step_time_sec": 20.292114606010728, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14378, "loss": 3.199162721633911, "lr": 2.918287937743191e-05, "elapsed_sec": 3816.6376953125, "step_time_sec": 20.293585366976913, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14379, "loss": 3.1289453506469727, "lr": 2.915451895043732e-05, "elapsed_sec": 3836.934839963913, "step_time_sec": 20.29688300198177, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14380, "loss": 3.2863688468933105, "lr": 2.9126213592233014e-05, "elapsed_sec": 3857.227168083191, "step_time_sec": 20.29223110701423, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14381, "loss": 3.416825771331787, "lr": 2.9097963142580023e-05, "elapsed_sec": 3877.520122528076, "step_time_sec": 20.292729242995847, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14382, "loss": 3.1590774059295654, "lr": 2.9069767441860467e-05, "elapsed_sec": 3897.8153676986694, "step_time_sec": 20.295149994955864, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14383, "loss": 3.2252116203308105, "lr": 2.9041626331074545e-05, "elapsed_sec": 3918.1085307598114, "step_time_sec": 20.292959166981746, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14384, "loss": 3.4548568725585938, "lr": 2.9013539651837528e-05, "elapsed_sec": 3938.403241634369, "step_time_sec": 20.294615847989917, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14385, "loss": 3.1284658908843994, "lr": 2.898550724637681e-05, "elapsed_sec": 3958.7001984119415, "step_time_sec": 20.296725013991818, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14386, "loss": 2.85689115524292, "lr": 2.8957528957528963e-05, "elapsed_sec": 3978.9945414066315, "step_time_sec": 20.294098593003582, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14387, "loss": 3.0951666831970215, "lr": 2.8929604628736744e-05, "elapsed_sec": 3999.288489341736, "step_time_sec": 20.293835081974976, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14388, "loss": 2.9902684688568115, "lr": 2.8901734104046242e-05, "elapsed_sec": 4019.5831248760223, "step_time_sec": 20.29446854197886, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14389, "loss": 3.096475601196289, "lr": 2.887391722810395e-05, "elapsed_sec": 4039.87775015831, "step_time_sec": 20.29443489102414, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14390, "loss": 2.993753671646118, "lr": 2.8846153846153856e-05, "elapsed_sec": 4060.1715636253357, "step_time_sec": 20.293619193020277, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14391, "loss": 3.2228047847747803, "lr": 2.8818443804034583e-05, "elapsed_sec": 4080.468119621277, "step_time_sec": 20.296373952995054, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14392, "loss": 3.3127169609069824, "lr": 2.8790786948176586e-05, "elapsed_sec": 4100.763431787491, "step_time_sec": 20.29519756801892, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14393, "loss": 2.9260149002075195, "lr": 2.8763183125599234e-05, "elapsed_sec": 4121.0570368766785, "step_time_sec": 20.293476185994223, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14394, "loss": 2.8646349906921387, "lr": 2.8735632183908045e-05, "elapsed_sec": 4141.352027416229, "step_time_sec": 20.29480962699745, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14395, "loss": 2.9876654148101807, "lr": 2.870813397129187e-05, "elapsed_sec": 4161.645342588425, "step_time_sec": 20.29314596299082, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14396, "loss": 2.857205629348755, "lr": 2.868068833652008e-05, "elapsed_sec": 4181.939605474472, "step_time_sec": 20.29407900897786, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14397, "loss": 2.9529545307159424, "lr": 2.865329512893983e-05, "elapsed_sec": 4202.234663248062, "step_time_sec": 20.294822348980233, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14398, "loss": 3.137472629547119, "lr": 2.8625954198473285e-05, "elapsed_sec": 4222.527722597122, "step_time_sec": 20.292882908019237, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14399, "loss": 3.372178554534912, "lr": 2.859866539561487e-05, "elapsed_sec": 4242.821668624878, "step_time_sec": 20.293794917000923, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14400, "loss": 3.051363945007324, "lr": 2.857142857142857e-05, "elapsed_sec": 4263.117016792297, "step_time_sec": 95.44050040101865, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14401, "loss": 2.8357632160186768, "lr": 2.8544243577545198e-05, "elapsed_sec": 4358.505894184113, "step_time_sec": 20.24345093301963, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14402, "loss": 3.1297526359558105, "lr": 2.85171102661597e-05, "elapsed_sec": 4378.751817703247, "step_time_sec": 20.24585773597937, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14403, "loss": 3.1887142658233643, "lr": 2.8490028490028496e-05, "elapsed_sec": 4398.996700048447, "step_time_sec": 20.244554674020037, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14404, "loss": 2.826076030731201, "lr": 2.84629981024668e-05, "elapsed_sec": 4419.247441291809, "step_time_sec": 20.250610545044765, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14405, "loss": 3.316462993621826, "lr": 2.8436018957345974e-05, "elapsed_sec": 4439.49858546257, "step_time_sec": 20.251004901947454, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14406, "loss": 2.9499573707580566, "lr": 2.840909090909091e-05, "elapsed_sec": 4459.742914915085, "step_time_sec": 20.24413976800861, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14407, "loss": 3.1703341007232666, "lr": 2.8382213812677392e-05, "elapsed_sec": 4480.037998199463, "step_time_sec": 20.294854079023935, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14408, "loss": 3.0919349193573, "lr": 2.835538752362949e-05, "elapsed_sec": 4500.332680463791, "step_time_sec": 20.294529171020258, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14409, "loss": 2.8126282691955566, "lr": 2.8328611898016997e-05, "elapsed_sec": 4520.626079082489, "step_time_sec": 20.293216199963354, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14410, "loss": 2.6524434089660645, "lr": 2.8301886792452834e-05, "elapsed_sec": 4540.921655416489, "step_time_sec": 20.295445593015756, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14411, "loss": 3.179302215576172, "lr": 2.8275212064090483e-05, "elapsed_sec": 4561.220673322678, "step_time_sec": 20.29894226702163, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14412, "loss": 3.300379514694214, "lr": 2.824858757062147e-05, "elapsed_sec": 4581.516413450241, "step_time_sec": 20.29548402200453, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14413, "loss": 3.1554830074310303, "lr": 2.8222013170272817e-05, "elapsed_sec": 4601.813019752502, "step_time_sec": 20.296498661045916, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14414, "loss": 2.751103639602661, "lr": 2.819548872180452e-05, "elapsed_sec": 4622.107627868652, "step_time_sec": 20.29432068101596, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14415, "loss": 3.1852877140045166, "lr": 2.8169014084507043e-05, "elapsed_sec": 4642.402827978134, "step_time_sec": 20.295041332021356, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14416, "loss": 2.657370090484619, "lr": 2.8142589118198878e-05, "elapsed_sec": 4662.699469089508, "step_time_sec": 20.296405756031163, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14417, "loss": 3.0992114543914795, "lr": 2.8116213683223995e-05, "elapsed_sec": 4682.951299190521, "step_time_sec": 20.251697104016785, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14418, "loss": 3.10667085647583, "lr": 2.808988764044944e-05, "elapsed_sec": 4703.196231365204, "step_time_sec": 20.244724685966503, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14419, "loss": 3.064990758895874, "lr": 2.8063610851262868e-05, "elapsed_sec": 4723.492627382278, "step_time_sec": 20.296247807971667, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14420, "loss": 3.22350811958313, "lr": 2.8037383177570097e-05, "elapsed_sec": 4743.786318778992, "step_time_sec": 20.2935258349753, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14421, "loss": 3.098968029022217, "lr": 2.8011204481792718e-05, "elapsed_sec": 4764.080056428909, "step_time_sec": 20.293596900010016, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14422, "loss": 3.1251068115234375, "lr": 2.7985074626865675e-05, "elapsed_sec": 4784.375018835068, "step_time_sec": 20.294797073991504, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14423, "loss": 2.8113532066345215, "lr": 2.7958993476234864e-05, "elapsed_sec": 4804.66965508461, "step_time_sec": 20.294386799971107, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14424, "loss": 3.021998167037964, "lr": 2.793296089385475e-05, "elapsed_sec": 4824.964441537857, "step_time_sec": 20.294682978012133, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14425, "loss": 2.9581496715545654, "lr": 2.790697674418605e-05, "elapsed_sec": 4845.259066581726, "step_time_sec": 20.29439882701263, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14426, "loss": 3.1387124061584473, "lr": 2.788104089219331e-05, "elapsed_sec": 4865.553725719452, "step_time_sec": 20.294565346033778, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14427, "loss": 2.9339611530303955, "lr": 2.7855153203342618e-05, "elapsed_sec": 4885.849071979523, "step_time_sec": 20.295239826955367, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14428, "loss": 2.862379550933838, "lr": 2.782931354359926e-05, "elapsed_sec": 4906.100003004074, "step_time_sec": 20.25073655496817, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14429, "loss": 3.1041553020477295, "lr": 2.7803521779425396e-05, "elapsed_sec": 4926.394334316254, "step_time_sec": 20.294027141993865, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14430, "loss": 3.0310468673706055, "lr": 2.777777777777778e-05, "elapsed_sec": 4946.68866610527, "step_time_sec": 20.294227891019545, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14431, "loss": 2.9290859699249268, "lr": 2.775208140610546e-05, "elapsed_sec": 4966.9839334487915, "step_time_sec": 20.295169020013418, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14432, "loss": 3.405491828918457, "lr": 2.7726432532347505e-05, "elapsed_sec": 4987.2777326107025, "step_time_sec": 20.293607011029962, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14433, "loss": 2.9560184478759766, "lr": 2.7700831024930747e-05, "elapsed_sec": 5007.571496009827, "step_time_sec": 20.293606442050077, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14434, "loss": 2.9618799686431885, "lr": 2.767527675276753e-05, "elapsed_sec": 5027.867188453674, "step_time_sec": 20.295447216019966, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14435, "loss": 2.800428867340088, "lr": 2.7649769585253464e-05, "elapsed_sec": 5048.162499427795, "step_time_sec": 20.295189825992566, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14436, "loss": 3.294682025909424, "lr": 2.7624309392265193e-05, "elapsed_sec": 5068.406665802002, "step_time_sec": 20.24398285400821, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14437, "loss": 2.9232544898986816, "lr": 2.7598896044158237e-05, "elapsed_sec": 5088.702172756195, "step_time_sec": 20.295398643997032, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14438, "loss": 3.0188703536987305, "lr": 2.7573529411764707e-05, "elapsed_sec": 5108.99792432785, "step_time_sec": 20.295559955993667, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14439, "loss": 3.299753427505493, "lr": 2.7548209366391185e-05, "elapsed_sec": 5129.294015645981, "step_time_sec": 20.295936161011923, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14440, "loss": 3.2646186351776123, "lr": 2.752293577981652e-05, "elapsed_sec": 5149.590435266495, "step_time_sec": 20.29625399596989, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14441, "loss": 2.9656996726989746, "lr": 2.7497708524289645e-05, "elapsed_sec": 5169.885887384415, "step_time_sec": 20.29520443303045, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14442, "loss": 3.14632248878479, "lr": 2.7472527472527473e-05, "elapsed_sec": 5190.178212404251, "step_time_sec": 20.292195445043035, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14443, "loss": 3.1237659454345703, "lr": 2.744739249771272e-05, "elapsed_sec": 5210.472795009613, "step_time_sec": 20.29439823498251, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14444, "loss": 2.997528314590454, "lr": 2.742230347349178e-05, "elapsed_sec": 5230.767369270325, "step_time_sec": 20.294441453006584, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14445, "loss": 3.146052598953247, "lr": 2.7397260273972603e-05, "elapsed_sec": 5251.061493873596, "step_time_sec": 20.29391917499015, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14446, "loss": 3.088326930999756, "lr": 2.737226277372263e-05, "elapsed_sec": 5271.357444763184, "step_time_sec": 20.29574260400841, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14447, "loss": 2.9556987285614014, "lr": 2.7347310847766643e-05, "elapsed_sec": 5291.6516790390015, "step_time_sec": 20.294105749984737, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14448, "loss": 3.2207164764404297, "lr": 2.73224043715847e-05, "elapsed_sec": 5311.946282863617, "step_time_sec": 20.29439747502329, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14449, "loss": 3.1964831352233887, "lr": 2.7297543221110102e-05, "elapsed_sec": 5332.241861104965, "step_time_sec": 20.295432341983542, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14450, "loss": 3.1131680011749268, "lr": 2.727272727272727e-05, "elapsed_sec": 5352.536325216293, "step_time_sec": 105.70619719696697, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14451, "loss": 2.851229190826416, "lr": 2.7247956403269754e-05, "elapsed_sec": 5458.193310976028, "step_time_sec": 20.2449997960357, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14452, "loss": 3.040221929550171, "lr": 2.722323049001815e-05, "elapsed_sec": 5478.435982227325, "step_time_sec": 20.242433573992457, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14453, "loss": 3.0006461143493652, "lr": 2.7198549410698102e-05, "elapsed_sec": 5498.681066989899, "step_time_sec": 20.24497825599974, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14454, "loss": 3.012714147567749, "lr": 2.7173913043478262e-05, "elapsed_sec": 5518.925888776779, "step_time_sec": 20.244627312989905, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14455, "loss": 2.984611749649048, "lr": 2.714932126696833e-05, "elapsed_sec": 5539.170202732086, "step_time_sec": 20.24416334601119, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14456, "loss": 2.986966371536255, "lr": 2.7124773960217004e-05, "elapsed_sec": 5559.421022415161, "step_time_sec": 20.250667900952976, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14457, "loss": 2.6607229709625244, "lr": 2.7100271002710034e-05, "elapsed_sec": 5579.667571544647, "step_time_sec": 20.246381581004243, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14458, "loss": 3.0441668033599854, "lr": 2.7075812274368234e-05, "elapsed_sec": 5599.963038444519, "step_time_sec": 20.29523294302635, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14459, "loss": 2.67258620262146, "lr": 2.705139765554554e-05, "elapsed_sec": 5620.257573604584, "step_time_sec": 20.294491763052065, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14460, "loss": 2.778170585632324, "lr": 2.7027027027027027e-05, "elapsed_sec": 5640.554527759552, "step_time_sec": 20.296642208006233, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14461, "loss": 2.898829460144043, "lr": 2.7002700270027007e-05, "elapsed_sec": 5660.849278450012, "step_time_sec": 20.29467036598362, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14462, "loss": 3.1337296962738037, "lr": 2.697841726618706e-05, "elapsed_sec": 5681.1449239254, "step_time_sec": 20.295366070000455, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14463, "loss": 3.121311664581299, "lr": 2.6954177897574123e-05, "elapsed_sec": 5701.44184422493, "step_time_sec": 20.29685512202559, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14464, "loss": 2.914886474609375, "lr": 2.692998204667864e-05, "elapsed_sec": 5721.736476421356, "step_time_sec": 20.294456853007432, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14465, "loss": 3.0024240016937256, "lr": 2.690582959641256e-05, "elapsed_sec": 5742.033113479614, "step_time_sec": 20.296477719966788, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14466, "loss": 3.421869993209839, "lr": 2.6881720430107534e-05, "elapsed_sec": 5762.332247018814, "step_time_sec": 20.299008749017958, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14467, "loss": 3.2445359230041504, "lr": 2.6857654431512986e-05, "elapsed_sec": 5782.627683162689, "step_time_sec": 20.295257972029503, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14468, "loss": 3.060469150543213, "lr": 2.6833631484794278e-05, "elapsed_sec": 5802.9231214523315, "step_time_sec": 20.295261898019817, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14469, "loss": 3.2443220615386963, "lr": 2.680965147453083e-05, "elapsed_sec": 5823.219158887863, "step_time_sec": 20.295762012014166, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14470, "loss": 2.937112808227539, "lr": 2.6785714285714288e-05, "elapsed_sec": 5843.513719797134, "step_time_sec": 20.294444806990214, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14471, "loss": 2.9604270458221436, "lr": 2.676181980374666e-05, "elapsed_sec": 5863.810955524445, "step_time_sec": 20.29712406801991, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14472, "loss": 2.8851702213287354, "lr": 2.67379679144385e-05, "elapsed_sec": 5884.06445980072, "step_time_sec": 20.253254287992604, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14473, "loss": 2.980788469314575, "lr": 2.671415850400713e-05, "elapsed_sec": 5904.358699560165, "step_time_sec": 20.294127891014796, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14474, "loss": 3.1960718631744385, "lr": 2.6690391459074734e-05, "elapsed_sec": 5924.655257940292, "step_time_sec": 20.29629086802015, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14475, "loss": 3.316148042678833, "lr": 2.6666666666666673e-05, "elapsed_sec": 5944.952974081039, "step_time_sec": 20.29765334300464, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14476, "loss": 2.824221611022949, "lr": 2.6642984014209596e-05, "elapsed_sec": 5965.246812820435, "step_time_sec": 20.293606469000224, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14477, "loss": 3.0173988342285156, "lr": 2.6619343389529726e-05, "elapsed_sec": 5985.542961359024, "step_time_sec": 20.296018966997508, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14478, "loss": 3.0135879516601562, "lr": 2.6595744680851064e-05, "elapsed_sec": 6005.839404344559, "step_time_sec": 20.29631313000573, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14479, "loss": 3.454969882965088, "lr": 2.6572187776793625e-05, "elapsed_sec": 6026.133392572403, "step_time_sec": 20.2938220470096, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14480, "loss": 2.804536819458008, "lr": 2.6548672566371683e-05, "elapsed_sec": 6046.428466558456, "step_time_sec": 20.294816666981205, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14481, "loss": 2.634939193725586, "lr": 2.6525198938992047e-05, "elapsed_sec": 6066.725298643112, "step_time_sec": 20.296678638027515, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14482, "loss": 3.5499746799468994, "lr": 2.65017667844523e-05, "elapsed_sec": 6087.02038192749, "step_time_sec": 20.294920709973667, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14483, "loss": 2.82466983795166, "lr": 2.64783759929391e-05, "elapsed_sec": 6107.316073179245, "step_time_sec": 20.295458573964424, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14484, "loss": 2.993723154067993, "lr": 2.6455026455026456e-05, "elapsed_sec": 6127.61114025116, "step_time_sec": 20.294991886999924, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14485, "loss": 3.195629358291626, "lr": 2.6431718061674013e-05, "elapsed_sec": 6147.907293319702, "step_time_sec": 20.295869527035393, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14486, "loss": 3.0372958183288574, "lr": 2.6408450704225354e-05, "elapsed_sec": 6168.161356925964, "step_time_sec": 20.254048396018334, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14487, "loss": 3.1199958324432373, "lr": 2.6385224274406334e-05, "elapsed_sec": 6188.4572241306305, "step_time_sec": 20.295631564978976, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14488, "loss": 2.9989633560180664, "lr": 2.6362038664323377e-05, "elapsed_sec": 6208.7522439956665, "step_time_sec": 20.294812785985414, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14489, "loss": 3.1209194660186768, "lr": 2.633889376646181e-05, "elapsed_sec": 6229.048989534378, "step_time_sec": 20.296574258012697, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14490, "loss": 3.155992269515991, "lr": 2.6315789473684212e-05, "elapsed_sec": 6249.343646526337, "step_time_sec": 20.294462232035585, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14491, "loss": 2.9256680011749268, "lr": 2.629272567922875e-05, "elapsed_sec": 6269.636790037155, "step_time_sec": 20.293036771006882, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14492, "loss": 2.9574952125549316, "lr": 2.6269702276707533e-05, "elapsed_sec": 6289.933557510376, "step_time_sec": 20.29652963002445, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14493, "loss": 2.7824599742889404, "lr": 2.624671916010499e-05, "elapsed_sec": 6310.230711936951, "step_time_sec": 20.296993604977615, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14494, "loss": 2.7755773067474365, "lr": 2.6223776223776227e-05, "elapsed_sec": 6330.525745868683, "step_time_sec": 20.29484072100604, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14495, "loss": 3.1580371856689453, "lr": 2.6200873362445416e-05, "elapsed_sec": 6350.822524309158, "step_time_sec": 20.29661513096653, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14496, "loss": 2.954751968383789, "lr": 2.6178010471204188e-05, "elapsed_sec": 6371.11852145195, "step_time_sec": 20.29582857299829, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14497, "loss": 3.270009994506836, "lr": 2.615518744551003e-05, "elapsed_sec": 6391.412641048431, "step_time_sec": 20.29400794103276, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14498, "loss": 3.275869607925415, "lr": 2.613240418118467e-05, "elapsed_sec": 6411.70786690712, "step_time_sec": 20.295039348013233, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14499, "loss": 2.985745906829834, "lr": 2.6109660574412533e-05, "elapsed_sec": 6432.0042951107025, "step_time_sec": 20.29622198297875, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14500, "loss": 3.062817335128784, "lr": 2.6086956521739135e-05, "elapsed_sec": 6452.298645973206, "step_time_sec": 79.03201317699859, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14501, "loss": 3.0206918716430664, "lr": 2.6064291920069507e-05, "elapsed_sec": 6531.28768491745, "step_time_sec": 20.251074911036994, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14502, "loss": 3.111438035964966, "lr": 2.6041666666666672e-05, "elapsed_sec": 6551.5330119133, "step_time_sec": 20.245079485990573, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14503, "loss": 3.2037863731384277, "lr": 2.6019080659150047e-05, "elapsed_sec": 6571.7833297252655, "step_time_sec": 20.250251095043495, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14504, "loss": 3.175057888031006, "lr": 2.5996533795493934e-05, "elapsed_sec": 6592.030759811401, "step_time_sec": 20.247149285976775, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14505, "loss": 2.974827527999878, "lr": 2.597402597402598e-05, "elapsed_sec": 6612.273571014404, "step_time_sec": 20.24270796601195, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14506, "loss": 2.613668918609619, "lr": 2.595155709342561e-05, "elapsed_sec": 6632.569670677185, "step_time_sec": 20.295899692049716, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14507, "loss": 2.845076084136963, "lr": 2.592912705272256e-05, "elapsed_sec": 6652.864205121994, "step_time_sec": 20.294349688047078, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14508, "loss": 3.3066186904907227, "lr": 2.5906735751295337e-05, "elapsed_sec": 6673.162058830261, "step_time_sec": 20.297786896990146, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14509, "loss": 3.326589584350586, "lr": 2.5884383088869717e-05, "elapsed_sec": 6693.456743240356, "step_time_sec": 20.294506990991067, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14510, "loss": 3.3451764583587646, "lr": 2.5862068965517248e-05, "elapsed_sec": 6713.709338903427, "step_time_sec": 20.252383875020314, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14511, "loss": 3.572477340698242, "lr": 2.5839793281653753e-05, "elapsed_sec": 6734.004163980484, "step_time_sec": 20.29462328302907, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14512, "loss": 3.1919538974761963, "lr": 2.5817555938037867e-05, "elapsed_sec": 6754.301580667496, "step_time_sec": 20.297253151016776, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14513, "loss": 3.2171971797943115, "lr": 2.5795356835769564e-05, "elapsed_sec": 6774.5963315963745, "step_time_sec": 20.29452161898371, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14514, "loss": 3.167440414428711, "lr": 2.5773195876288664e-05, "elapsed_sec": 6794.893342256546, "step_time_sec": 20.296884521027096, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14515, "loss": 3.1556756496429443, "lr": 2.5751072961373394e-05, "elapsed_sec": 6815.188628435135, "step_time_sec": 20.295078555995133, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14516, "loss": 3.0201878547668457, "lr": 2.572898799313894e-05, "elapsed_sec": 6835.482971668243, "step_time_sec": 20.29411927703768, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14517, "loss": 3.177608013153076, "lr": 2.570694087403599e-05, "elapsed_sec": 6855.776812791824, "step_time_sec": 20.29377526801545, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14518, "loss": 3.1314165592193604, "lr": 2.5684931506849317e-05, "elapsed_sec": 6876.074098110199, "step_time_sec": 20.297168767021503, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14519, "loss": 3.0711357593536377, "lr": 2.566295979469633e-05, "elapsed_sec": 6896.368215084076, "step_time_sec": 20.29384084098274, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14520, "loss": 2.830352544784546, "lr": 2.564102564102564e-05, "elapsed_sec": 6916.663896560669, "step_time_sec": 20.29551890096627, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14521, "loss": 2.386747360229492, "lr": 2.5619128949615715e-05, "elapsed_sec": 6936.959090709686, "step_time_sec": 20.29507700901013, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14522, "loss": 3.0159387588500977, "lr": 2.559726962457338e-05, "elapsed_sec": 6957.259245634079, "step_time_sec": 20.30010442796629, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14523, "loss": 3.1395602226257324, "lr": 2.5575447570332485e-05, "elapsed_sec": 6977.51269030571, "step_time_sec": 20.2530938879936, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14524, "loss": 2.8143467903137207, "lr": 2.5553662691652473e-05, "elapsed_sec": 6997.758945941925, "step_time_sec": 20.246010300004855, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14525, "loss": 2.9667751789093018, "lr": 2.5531914893617022e-05, "elapsed_sec": 7018.0525143146515, "step_time_sec": 20.293461324996315, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14526, "loss": 3.2281367778778076, "lr": 2.5510204081632657e-05, "elapsed_sec": 7038.349065542221, "step_time_sec": 20.296357339946553, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14527, "loss": 3.247736692428589, "lr": 2.548853016142736e-05, "elapsed_sec": 7058.643560409546, "step_time_sec": 20.294293071026914, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14528, "loss": 3.1136934757232666, "lr": 2.5466893039049238e-05, "elapsed_sec": 7078.938697576523, "step_time_sec": 20.29505166702438, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14529, "loss": 3.0821402072906494, "lr": 2.544529262086514e-05, "elapsed_sec": 7099.23254942894, "step_time_sec": 20.293646791018546, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14530, "loss": 3.034372329711914, "lr": 2.5423728813559325e-05, "elapsed_sec": 7119.530031681061, "step_time_sec": 20.29727386299055, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14531, "loss": 3.0912961959838867, "lr": 2.5402201524132094e-05, "elapsed_sec": 7139.824479103088, "step_time_sec": 20.29425593803171, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14532, "loss": 3.044557571411133, "lr": 2.5380710659898476e-05, "elapsed_sec": 7160.120755672455, "step_time_sec": 20.296086271002423, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14533, "loss": 3.2650558948516846, "lr": 2.53592561284869e-05, "elapsed_sec": 7180.416291952133, "step_time_sec": 20.295406290970277, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14534, "loss": 3.0453450679779053, "lr": 2.533783783783784e-05, "elapsed_sec": 7200.711848020554, "step_time_sec": 20.295347281033173, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14535, "loss": 3.142735719680786, "lr": 2.5316455696202537e-05, "elapsed_sec": 7220.962723970413, "step_time_sec": 20.25073147099465, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14536, "loss": 3.0760600566864014, "lr": 2.5295109612141657e-05, "elapsed_sec": 7241.211495876312, "step_time_sec": 20.248556550010107, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14537, "loss": 2.992377758026123, "lr": 2.527379949452401e-05, "elapsed_sec": 7261.507624864578, "step_time_sec": 20.296000614995137, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14538, "loss": 2.88380765914917, "lr": 2.5252525252525256e-05, "elapsed_sec": 7281.803962945938, "step_time_sec": 20.296189444023184, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14539, "loss": 3.1121461391448975, "lr": 2.523128679562658e-05, "elapsed_sec": 7302.099179983139, "step_time_sec": 20.295063704019412, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14540, "loss": 2.9200761318206787, "lr": 2.5210084033613446e-05, "elapsed_sec": 7322.394440412521, "step_time_sec": 20.295072346983943, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14541, "loss": 2.599668264389038, "lr": 2.5188916876574308e-05, "elapsed_sec": 7342.688844442368, "step_time_sec": 20.294290792022366, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14542, "loss": 3.4101722240448, "lr": 2.516778523489933e-05, "elapsed_sec": 7362.985641717911, "step_time_sec": 20.29658061004011, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14543, "loss": 3.02226185798645, "lr": 2.5146689019279135e-05, "elapsed_sec": 7383.27952837944, "step_time_sec": 20.293642591976095, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14544, "loss": 2.9383535385131836, "lr": 2.5125628140703518e-05, "elapsed_sec": 7403.576044559479, "step_time_sec": 20.296404607011937, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14545, "loss": 2.775360584259033, "lr": 2.5104602510460254e-05, "elapsed_sec": 7423.871204853058, "step_time_sec": 20.294932289980352, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14546, "loss": 3.037692070007324, "lr": 2.508361204013378e-05, "elapsed_sec": 7444.123303890228, "step_time_sec": 20.251998613995966, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14547, "loss": 2.918139934539795, "lr": 2.5062656641604014e-05, "elapsed_sec": 7464.3671181201935, "step_time_sec": 20.243631780031137, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14548, "loss": 3.054605722427368, "lr": 2.5041736227045077e-05, "elapsed_sec": 7484.664298772812, "step_time_sec": 20.297087342012674, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14549, "loss": 3.4835281372070312, "lr": 2.5020850708924106e-05, "elapsed_sec": 7504.95764875412, "step_time_sec": 20.29314720100956, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14550, "loss": 3.039633274078369, "lr": 2.5000000000000005e-05, "elapsed_sec": 7525.25355553627, "step_time_sec": 95.67981145699741, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14551, "loss": 2.788181781768799, "lr": 2.4979184013322235e-05, "elapsed_sec": 7620.880258798599, "step_time_sec": 20.242392069019843, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14552, "loss": 2.93289852142334, "lr": 2.4958402662229623e-05, "elapsed_sec": 7641.13277721405, "step_time_sec": 20.252360929036513, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14553, "loss": 3.058677911758423, "lr": 2.493765586034913e-05, "elapsed_sec": 7661.381151437759, "step_time_sec": 20.248149065009784, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14554, "loss": 2.898536205291748, "lr": 2.4916943521594686e-05, "elapsed_sec": 7681.626301288605, "step_time_sec": 20.24505558598321, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14555, "loss": 2.7032198905944824, "lr": 2.489626556016598e-05, "elapsed_sec": 7701.870831727982, "step_time_sec": 20.244345671031624, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14556, "loss": 2.9943294525146484, "lr": 2.487562189054727e-05, "elapsed_sec": 7722.166499853134, "step_time_sec": 20.295442414004356, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14557, "loss": 3.1768298149108887, "lr": 2.4855012427506217e-05, "elapsed_sec": 7742.460569620132, "step_time_sec": 20.2938816450187, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14558, "loss": 3.137235641479492, "lr": 2.4834437086092715e-05, "elapsed_sec": 7762.756292104721, "step_time_sec": 20.295606975036208, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14559, "loss": 3.283235549926758, "lr": 2.481389578163772e-05, "elapsed_sec": 7783.05135846138, "step_time_sec": 20.294845596014056, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14560, "loss": 3.143052577972412, "lr": 2.479338842975207e-05, "elapsed_sec": 7803.34667634964, "step_time_sec": 20.295195818005595, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14561, "loss": 3.1018996238708496, "lr": 2.4772914946325354e-05, "elapsed_sec": 7823.642556667328, "step_time_sec": 20.29570782301016, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14562, "loss": 3.276235818862915, "lr": 2.4752475247524754e-05, "elapsed_sec": 7843.938555955887, "step_time_sec": 20.29576161201112, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14563, "loss": 2.9921915531158447, "lr": 2.47320692497939e-05, "elapsed_sec": 7864.233399152756, "step_time_sec": 20.294724118022714, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14564, "loss": 3.29410982131958, "lr": 2.4711696869851732e-05, "elapsed_sec": 7884.528192520142, "step_time_sec": 20.294595668965485, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14565, "loss": 2.8167800903320312, "lr": 2.4691358024691357e-05, "elapsed_sec": 7904.8248653411865, "step_time_sec": 20.296575230022427, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14566, "loss": 3.2209246158599854, "lr": 2.467105263157895e-05, "elapsed_sec": 7925.119813680649, "step_time_sec": 20.29473823000444, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14567, "loss": 2.80836820602417, "lr": 2.465078060805259e-05, "elapsed_sec": 7945.370305776596, "step_time_sec": 20.25028270803159, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14568, "loss": 3.1489205360412598, "lr": 2.4630541871921187e-05, "elapsed_sec": 7965.61695432663, "step_time_sec": 20.24653618200682, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14569, "loss": 2.7822020053863525, "lr": 2.4610336341263332e-05, "elapsed_sec": 7985.912032842636, "step_time_sec": 20.294968295958824, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14570, "loss": 3.175785779953003, "lr": 2.459016393442623e-05, "elapsed_sec": 8006.208937168121, "step_time_sec": 20.296647992974613, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14571, "loss": 3.1073997020721436, "lr": 2.4570024570024573e-05, "elapsed_sec": 8026.503829956055, "step_time_sec": 20.294657905003987, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14572, "loss": 3.123774290084839, "lr": 2.4549918166939446e-05, "elapsed_sec": 8046.796811580658, "step_time_sec": 20.292906665999908, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14573, "loss": 3.199583053588867, "lr": 2.4529844644317255e-05, "elapsed_sec": 8067.090978860855, "step_time_sec": 20.293869472981896, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14574, "loss": 3.049694776535034, "lr": 2.4509803921568626e-05, "elapsed_sec": 8087.38676404953, "step_time_sec": 20.295665101031773, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14575, "loss": 3.1841611862182617, "lr": 2.448979591836735e-05, "elapsed_sec": 8107.680796861649, "step_time_sec": 20.293791113013867, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14576, "loss": 2.7222228050231934, "lr": 2.446982055464927e-05, "elapsed_sec": 8127.977182149887, "step_time_sec": 20.296362776018213, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14577, "loss": 2.849477767944336, "lr": 2.444987775061125e-05, "elapsed_sec": 8148.273337364197, "step_time_sec": 20.296041172987316, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14578, "loss": 3.1612534523010254, "lr": 2.44299674267101e-05, "elapsed_sec": 8168.529844999313, "step_time_sec": 20.255482898035552, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14579, "loss": 2.933157205581665, "lr": 2.4410089503661513e-05, "elapsed_sec": 8188.779203653336, "step_time_sec": 20.24907582398737, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14580, "loss": 2.9066765308380127, "lr": 2.439024390243903e-05, "elapsed_sec": 8209.075446844101, "step_time_sec": 20.296178028045688, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14581, "loss": 3.037838935852051, "lr": 2.437043054427295e-05, "elapsed_sec": 8229.371141672134, "step_time_sec": 20.29543018998811, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14582, "loss": 3.228271245956421, "lr": 2.435064935064935e-05, "elapsed_sec": 8249.667850732803, "step_time_sec": 20.29653283697553, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14583, "loss": 3.1904518604278564, "lr": 2.4330900243309002e-05, "elapsed_sec": 8269.96415424347, "step_time_sec": 20.29614628799027, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14584, "loss": 3.143195152282715, "lr": 2.4311183144246355e-05, "elapsed_sec": 8290.261054992676, "step_time_sec": 20.296719723031856, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14585, "loss": 3.243663787841797, "lr": 2.429149797570851e-05, "elapsed_sec": 8310.513187408447, "step_time_sec": 20.252035573998, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14586, "loss": 3.067291498184204, "lr": 2.427184466019418e-05, "elapsed_sec": 8330.758316516876, "step_time_sec": 20.244868198991753, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14587, "loss": 3.031355857849121, "lr": 2.425222312045271e-05, "elapsed_sec": 8351.052290916443, "step_time_sec": 20.29385557299247, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14588, "loss": 3.035738706588745, "lr": 2.4232633279483043e-05, "elapsed_sec": 8371.349194049835, "step_time_sec": 20.296769020031206, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14589, "loss": 3.04557728767395, "lr": 2.4213075060532693e-05, "elapsed_sec": 8391.644371509552, "step_time_sec": 20.294926159025636, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14590, "loss": 2.940002679824829, "lr": 2.4193548387096777e-05, "elapsed_sec": 8411.940176725388, "step_time_sec": 20.29567505099112, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14591, "loss": 3.157705545425415, "lr": 2.4174053182917003e-05, "elapsed_sec": 8432.235337734222, "step_time_sec": 20.29492094402667, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14592, "loss": 2.964266061782837, "lr": 2.415458937198068e-05, "elapsed_sec": 8452.531442642212, "step_time_sec": 20.295996031025425, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14593, "loss": 2.9330856800079346, "lr": 2.4135156878519712e-05, "elapsed_sec": 8472.825898170471, "step_time_sec": 20.294227748992853, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14594, "loss": 3.095811605453491, "lr": 2.4115755627009648e-05, "elapsed_sec": 8493.12183213234, "step_time_sec": 20.295815126039088, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14595, "loss": 3.1736032962799072, "lr": 2.4096385542168674e-05, "elapsed_sec": 8513.416427135468, "step_time_sec": 20.294405575026758, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14596, "loss": 2.8593907356262207, "lr": 2.4077046548956665e-05, "elapsed_sec": 8533.711786031723, "step_time_sec": 20.295162900001742, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14597, "loss": 3.1661477088928223, "lr": 2.4057738572574183e-05, "elapsed_sec": 8553.96482872963, "step_time_sec": 20.25287985999603, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14598, "loss": 3.1825828552246094, "lr": 2.403846153846154e-05, "elapsed_sec": 8574.21240735054, "step_time_sec": 20.247375676000956, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14599, "loss": 3.140871524810791, "lr": 2.401921537229784e-05, "elapsed_sec": 8594.5063393116, "step_time_sec": 20.29380568600027, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14600, "loss": 3.48490834236145, "lr": 2.4e-05, "elapsed_sec": 8614.80274271965, "step_time_sec": 106.4968876070343, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14601, "loss": 3.269023895263672, "lr": 2.3980815347721828e-05, "elapsed_sec": 8721.248898744583, "step_time_sec": 20.24536184302997, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14602, "loss": 2.853721857070923, "lr": 2.396166134185304e-05, "elapsed_sec": 8741.49297451973, "step_time_sec": 20.243893689010292, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14603, "loss": 3.2676053047180176, "lr": 2.394253790901836e-05, "elapsed_sec": 8761.739249706268, "step_time_sec": 20.246173378021922, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14604, "loss": 3.150700092315674, "lr": 2.392344497607656e-05, "elapsed_sec": 8781.990807533264, "step_time_sec": 20.251292611996178, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14605, "loss": 3.289569616317749, "lr": 2.3904382470119523e-05, "elapsed_sec": 8802.243191480637, "step_time_sec": 20.252256217994727, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14606, "loss": 3.231201648712158, "lr": 2.3885350318471344e-05, "elapsed_sec": 8822.487558841705, "step_time_sec": 20.24414770398289, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14607, "loss": 3.6270675659179688, "lr": 2.3866348448687352e-05, "elapsed_sec": 8842.784970283508, "step_time_sec": 20.297323294973467, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14608, "loss": 3.0400238037109375, "lr": 2.3847376788553263e-05, "elapsed_sec": 8863.081142663956, "step_time_sec": 20.296078054991085, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14609, "loss": 3.2477893829345703, "lr": 2.38284352660842e-05, "elapsed_sec": 8883.377549409866, "step_time_sec": 20.29612704901956, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14610, "loss": 2.977452039718628, "lr": 2.3809523809523814e-05, "elapsed_sec": 8903.673085212708, "step_time_sec": 20.295375737012364, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14611, "loss": 2.8659679889678955, "lr": 2.379064234734338e-05, "elapsed_sec": 8923.969893455505, "step_time_sec": 20.296576225024182, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14612, "loss": 3.3190181255340576, "lr": 2.377179080824089e-05, "elapsed_sec": 8944.26451587677, "step_time_sec": 20.294500672025606, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14613, "loss": 3.042198896408081, "lr": 2.3752969121140148e-05, "elapsed_sec": 8964.561774730682, "step_time_sec": 20.297054091992322, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14614, "loss": 3.1444406509399414, "lr": 2.3734177215189877e-05, "elapsed_sec": 8984.856337308884, "step_time_sec": 20.29445535200648, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14615, "loss": 2.7256979942321777, "lr": 2.3715415019762852e-05, "elapsed_sec": 9005.151687145233, "step_time_sec": 20.29512240999611, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14616, "loss": 2.9152019023895264, "lr": 2.3696682464454976e-05, "elapsed_sec": 9025.448768854141, "step_time_sec": 20.296930917014834, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14617, "loss": 3.167424440383911, "lr": 2.3677979479084452e-05, "elapsed_sec": 9045.744606256485, "step_time_sec": 20.295754999038763, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14618, "loss": 2.9626150131225586, "lr": 2.3659305993690857e-05, "elapsed_sec": 9066.039901018143, "step_time_sec": 20.295173094025813, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14619, "loss": 2.988466262817383, "lr": 2.3640661938534283e-05, "elapsed_sec": 9086.337045431137, "step_time_sec": 20.296923456946388, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14620, "loss": 2.820420026779175, "lr": 2.3622047244094492e-05, "elapsed_sec": 9106.631124258041, "step_time_sec": 20.29384709603619, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14621, "loss": 3.145664930343628, "lr": 2.360346184107003e-05, "elapsed_sec": 9126.926582813263, "step_time_sec": 20.295364984020125, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14622, "loss": 3.246743679046631, "lr": 2.3584905660377357e-05, "elapsed_sec": 9147.222254037857, "step_time_sec": 20.295463036047295, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14623, "loss": 2.831376314163208, "lr": 2.3566378633150042e-05, "elapsed_sec": 9167.517796278, "step_time_sec": 20.295323894009925, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14624, "loss": 2.983189821243286, "lr": 2.3547880690737836e-05, "elapsed_sec": 9187.812956571579, "step_time_sec": 20.29503698303597, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14625, "loss": 2.956096649169922, "lr": 2.3529411764705887e-05, "elapsed_sec": 9208.110219717026, "step_time_sec": 20.29699692799477, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14626, "loss": 3.0319998264312744, "lr": 2.3510971786833858e-05, "elapsed_sec": 9228.406824350357, "step_time_sec": 20.296534688037354, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14627, "loss": 3.1130623817443848, "lr": 2.3492560689115116e-05, "elapsed_sec": 9248.704310655594, "step_time_sec": 20.297242737957276, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14628, "loss": 2.9921724796295166, "lr": 2.3474178403755872e-05, "elapsed_sec": 9268.999779939651, "step_time_sec": 20.295303980994504, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14629, "loss": 2.876302480697632, "lr": 2.345582486317436e-05, "elapsed_sec": 9289.248946905136, "step_time_sec": 20.249083780974615, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14630, "loss": 3.285179376602173, "lr": 2.3437500000000004e-05, "elapsed_sec": 9309.544896364212, "step_time_sec": 20.29573817399796, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14631, "loss": 3.03349232673645, "lr": 2.3419203747072603e-05, "elapsed_sec": 9329.84275317192, "step_time_sec": 20.297621512028854, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14632, "loss": 2.90791392326355, "lr": 2.34009360374415e-05, "elapsed_sec": 9350.13872551918, "step_time_sec": 20.295701447990723, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14633, "loss": 3.022428512573242, "lr": 2.3382696804364772e-05, "elapsed_sec": 9370.43648815155, "step_time_sec": 20.297668505983893, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14634, "loss": 2.8551249504089355, "lr": 2.336448598130841e-05, "elapsed_sec": 9390.732904672623, "step_time_sec": 20.296176339965314, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14635, "loss": 3.3084499835968018, "lr": 2.334630350194553e-05, "elapsed_sec": 9411.030569553375, "step_time_sec": 20.29757585201878, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14636, "loss": 3.009204387664795, "lr": 2.3328149300155525e-05, "elapsed_sec": 9431.287460803986, "step_time_sec": 20.25668368296465, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14637, "loss": 3.0563979148864746, "lr": 2.3310023310023313e-05, "elapsed_sec": 9451.585545063019, "step_time_sec": 20.29799033998279, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14638, "loss": 3.298977851867676, "lr": 2.3291925465838514e-05, "elapsed_sec": 9471.880954504013, "step_time_sec": 20.295207311981358, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14639, "loss": 3.3000547885894775, "lr": 2.3273855702094647e-05, "elapsed_sec": 9492.177595376968, "step_time_sec": 20.29639302601572, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14640, "loss": 3.2087767124176025, "lr": 2.325581395348837e-05, "elapsed_sec": 9512.473530054092, "step_time_sec": 20.295868507004343, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14641, "loss": 3.142127513885498, "lr": 2.323780015491867e-05, "elapsed_sec": 9532.771582841873, "step_time_sec": 20.297927025996614, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14642, "loss": 2.889847993850708, "lr": 2.321981424148607e-05, "elapsed_sec": 9553.067354679108, "step_time_sec": 20.295508875045925, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14643, "loss": 3.2420194149017334, "lr": 2.3201856148491884e-05, "elapsed_sec": 9573.36487197876, "step_time_sec": 20.297310610010754, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14644, "loss": 3.13321590423584, "lr": 2.3183925811437406e-05, "elapsed_sec": 9593.659631729126, "step_time_sec": 20.29459696903359, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14645, "loss": 2.6968860626220703, "lr": 2.3166023166023168e-05, "elapsed_sec": 9613.956747055054, "step_time_sec": 20.296971614996437, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14646, "loss": 3.12434458732605, "lr": 2.314814814814815e-05, "elapsed_sec": 9634.252273082733, "step_time_sec": 20.295307202031836, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14647, "loss": 2.9854819774627686, "lr": 2.3130300693909022e-05, "elapsed_sec": 9654.54947423935, "step_time_sec": 20.297091864049435, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14648, "loss": 3.037294626235962, "lr": 2.311248073959939e-05, "elapsed_sec": 9674.843921422958, "step_time_sec": 20.2942303600139, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14649, "loss": 3.036832332611084, "lr": 2.3094688221709012e-05, "elapsed_sec": 9695.142429590225, "step_time_sec": 20.298379714949988, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14650, "loss": 3.0504322052001953, "lr": 2.307692307692308e-05, "elapsed_sec": 9715.43872332573, "step_time_sec": 85.10170380200725, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14651, "loss": 3.3539180755615234, "lr": 2.305918524212145e-05, "elapsed_sec": 9800.4889690876, "step_time_sec": 20.244608784036245, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14652, "loss": 3.0668580532073975, "lr": 2.304147465437788e-05, "elapsed_sec": 9820.734976768494, "step_time_sec": 20.245861761970446, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14653, "loss": 2.9951300621032715, "lr": 2.3023791250959328e-05, "elapsed_sec": 9840.98758482933, "step_time_sec": 20.252432583016343, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14654, "loss": 2.992971420288086, "lr": 2.3006134969325158e-05, "elapsed_sec": 9861.232117652893, "step_time_sec": 20.244301781000104, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14655, "loss": 2.9484946727752686, "lr": 2.298850574712644e-05, "elapsed_sec": 9881.526780605316, "step_time_sec": 20.294510248000734, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14656, "loss": 3.0648109912872314, "lr": 2.297090352220521e-05, "elapsed_sec": 9901.822459936142, "step_time_sec": 20.295547453977633, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14657, "loss": 3.1296093463897705, "lr": 2.295332823259373e-05, "elapsed_sec": 9922.117600440979, "step_time_sec": 20.294973802985623, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14658, "loss": 2.9258711338043213, "lr": 2.2935779816513765e-05, "elapsed_sec": 9942.413573265076, "step_time_sec": 20.295714109030087, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14659, "loss": 3.204667806625366, "lr": 2.291825821237586e-05, "elapsed_sec": 9962.70977640152, "step_time_sec": 20.296080474043265, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14660, "loss": 2.9013214111328125, "lr": 2.2900763358778628e-05, "elapsed_sec": 9983.004338502884, "step_time_sec": 20.294336327991914, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14661, "loss": 3.122368335723877, "lr": 2.2883295194508012e-05, "elapsed_sec": 10003.299338579178, "step_time_sec": 20.294860395020805, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14662, "loss": 2.905921459197998, "lr": 2.286585365853659e-05, "elapsed_sec": 10023.59410572052, "step_time_sec": 20.29458068695385, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14663, "loss": 3.1862337589263916, "lr": 2.2848438690022855e-05, "elapsed_sec": 10043.888324737549, "step_time_sec": 20.29409157502232, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14664, "loss": 3.141629219055176, "lr": 2.2831050228310503e-05, "elapsed_sec": 10064.184646129608, "step_time_sec": 20.296199906035326, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14665, "loss": 2.8663833141326904, "lr": 2.2813688212927758e-05, "elapsed_sec": 10084.478262662888, "step_time_sec": 20.2933337650029, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14666, "loss": 3.000880718231201, "lr": 2.2796352583586632e-05, "elapsed_sec": 10104.773900032043, "step_time_sec": 20.29552645701915, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14667, "loss": 3.3061647415161133, "lr": 2.2779043280182233e-05, "elapsed_sec": 10125.069473981857, "step_time_sec": 20.295328327978496, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14668, "loss": 2.8989546298980713, "lr": 2.276176024279211e-05, "elapsed_sec": 10145.363253116608, "step_time_sec": 20.293674582033418, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14669, "loss": 3.0951249599456787, "lr": 2.2744503411675516e-05, "elapsed_sec": 10165.657623529434, "step_time_sec": 20.294201308046468, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14670, "loss": 2.864743232727051, "lr": 2.2727272727272733e-05, "elapsed_sec": 10185.953611135483, "step_time_sec": 20.29591862799134, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14671, "loss": 3.1557106971740723, "lr": 2.2710068130204394e-05, "elapsed_sec": 10206.248090267181, "step_time_sec": 20.294258124020416, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14672, "loss": 2.7705278396606445, "lr": 2.2692889561270805e-05, "elapsed_sec": 10226.543172121048, "step_time_sec": 20.294794412970077, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14673, "loss": 2.7376644611358643, "lr": 2.267573696145125e-05, "elapsed_sec": 10246.836607456207, "step_time_sec": 20.293360033945646, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14674, "loss": 3.1152944564819336, "lr": 2.2658610271903327e-05, "elapsed_sec": 10267.130905389786, "step_time_sec": 20.294058508006856, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14675, "loss": 3.047821283340454, "lr": 2.2641509433962268e-05, "elapsed_sec": 10287.383160829544, "step_time_sec": 20.252155375026632, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14676, "loss": 2.871577501296997, "lr": 2.2624434389140274e-05, "elapsed_sec": 10307.62866473198, "step_time_sec": 20.245297568966635, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14677, "loss": 3.0040853023529053, "lr": 2.260738507912585e-05, "elapsed_sec": 10327.924387931824, "step_time_sec": 20.295524586981628, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14678, "loss": 2.8628177642822266, "lr": 2.2590361445783133e-05, "elapsed_sec": 10348.218563079834, "step_time_sec": 20.29399499599822, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14679, "loss": 3.080033302307129, "lr": 2.2573363431151244e-05, "elapsed_sec": 10368.512791872025, "step_time_sec": 20.293995125975925, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14680, "loss": 2.8894643783569336, "lr": 2.2556390977443613e-05, "elapsed_sec": 10388.808295726776, "step_time_sec": 20.295423044008203, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14681, "loss": 2.923534631729126, "lr": 2.2539444027047338e-05, "elapsed_sec": 10409.103943109512, "step_time_sec": 20.295431350008585, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14682, "loss": 2.9545016288757324, "lr": 2.2522522522522526e-05, "elapsed_sec": 10429.399760961533, "step_time_sec": 20.295643221004866, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14683, "loss": 2.9791080951690674, "lr": 2.2505626406601654e-05, "elapsed_sec": 10449.694463014603, "step_time_sec": 20.29453801602358, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14684, "loss": 3.0833325386047363, "lr": 2.248875562218891e-05, "elapsed_sec": 10469.989173650742, "step_time_sec": 20.29455778398551, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14685, "loss": 2.9084131717681885, "lr": 2.2471910112359552e-05, "elapsed_sec": 10490.286324977875, "step_time_sec": 20.297008352004923, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14686, "loss": 3.1241586208343506, "lr": 2.2455089820359285e-05, "elapsed_sec": 10510.580510854721, "step_time_sec": 20.293949826038443, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14687, "loss": 3.0500566959381104, "lr": 2.2438294689603595e-05, "elapsed_sec": 10530.875740766525, "step_time_sec": 20.29517986200517, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14688, "loss": 2.721839189529419, "lr": 2.2421524663677132e-05, "elapsed_sec": 10551.172632932663, "step_time_sec": 20.296720482001547, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14689, "loss": 3.0690572261810303, "lr": 2.2404779686333086e-05, "elapsed_sec": 10571.47294640541, "step_time_sec": 20.30003773799399, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14690, "loss": 3.0637943744659424, "lr": 2.238805970149254e-05, "elapsed_sec": 10591.725957632065, "step_time_sec": 20.25291009101784, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14691, "loss": 3.1571192741394043, "lr": 2.2371364653243853e-05, "elapsed_sec": 10612.020830869675, "step_time_sec": 20.294686706969514, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14692, "loss": 2.9246668815612793, "lr": 2.235469448584203e-05, "elapsed_sec": 10632.313873767853, "step_time_sec": 20.29293717298424, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14693, "loss": 3.280407190322876, "lr": 2.233804914370812e-05, "elapsed_sec": 10652.609302520752, "step_time_sec": 20.295122326991986, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14694, "loss": 2.857445478439331, "lr": 2.2321428571428575e-05, "elapsed_sec": 10672.905413389206, "step_time_sec": 20.296064285037573, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14695, "loss": 2.9856057167053223, "lr": 2.230483271375465e-05, "elapsed_sec": 10693.200241088867, "step_time_sec": 20.294529716018587, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14696, "loss": 3.12935471534729, "lr": 2.2288261515601787e-05, "elapsed_sec": 10713.496294736862, "step_time_sec": 20.296002777002286, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14697, "loss": 3.332432270050049, "lr": 2.2271714922048998e-05, "elapsed_sec": 10733.78997039795, "step_time_sec": 20.29340437904466, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14698, "loss": 3.080264091491699, "lr": 2.225519287833828e-05, "elapsed_sec": 10754.084595918655, "step_time_sec": 20.294504763034638, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14699, "loss": 3.1325793266296387, "lr": 2.2238695329873987e-05, "elapsed_sec": 10774.37938284874, "step_time_sec": 20.294609375006985, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14700, "loss": 2.935532331466675, "lr": 2.2222222222222223e-05, "elapsed_sec": 10794.674332141876, "step_time_sec": 95.70591431402136, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14701, "loss": 3.1636717319488525, "lr": 2.2205773501110292e-05, "elapsed_sec": 10890.33702969551, "step_time_sec": 20.25141726998845, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14702, "loss": 3.163215398788452, "lr": 2.218934911242604e-05, "elapsed_sec": 10910.580521583557, "step_time_sec": 20.2433621819946, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14703, "loss": 3.207862377166748, "lr": 2.2172949002217298e-05, "elapsed_sec": 10930.825070381165, "step_time_sec": 20.24428377201548, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14704, "loss": 2.9647557735443115, "lr": 2.215657311669129e-05, "elapsed_sec": 10951.068953752518, "step_time_sec": 20.243774408008903, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14705, "loss": 2.95170521736145, "lr": 2.2140221402214025e-05, "elapsed_sec": 10971.364521980286, "step_time_sec": 20.295295750023797, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14706, "loss": 3.326880693435669, "lr": 2.2123893805309738e-05, "elapsed_sec": 10991.659207105637, "step_time_sec": 20.294660629995633, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14707, "loss": 3.1724460124969482, "lr": 2.2107590272660283e-05, "elapsed_sec": 11011.9526014328, "step_time_sec": 20.29320275003556, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14708, "loss": 3.2750980854034424, "lr": 2.209131075110457e-05, "elapsed_sec": 11032.24840259552, "step_time_sec": 20.295538004953414, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14709, "loss": 3.1930978298187256, "lr": 2.2075055187637972e-05, "elapsed_sec": 11052.542211294174, "step_time_sec": 20.29373359901365, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14710, "loss": 3.052170991897583, "lr": 2.2058823529411766e-05, "elapsed_sec": 11072.836043834686, "step_time_sec": 20.293639675946906, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14711, "loss": 3.1813979148864746, "lr": 2.2042615723732552e-05, "elapsed_sec": 11093.131707906723, "step_time_sec": 20.295467588002793, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14712, "loss": 2.994006872177124, "lr": 2.202643171806168e-05, "elapsed_sec": 11113.42624258995, "step_time_sec": 20.294398316997103, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14713, "loss": 3.102754831314087, "lr": 2.2010271460014675e-05, "elapsed_sec": 11133.719632148743, "step_time_sec": 20.293274604016915, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14714, "loss": 3.290703058242798, "lr": 2.1994134897360707e-05, "elapsed_sec": 11154.015748023987, "step_time_sec": 20.295803823973984, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14715, "loss": 3.056776762008667, "lr": 2.197802197802198e-05, "elapsed_sec": 11174.310015678406, "step_time_sec": 20.29420637199655, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14716, "loss": 3.0795493125915527, "lr": 2.1961932650073208e-05, "elapsed_sec": 11194.602074861526, "step_time_sec": 20.291839618002996, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14717, "loss": 3.219189167022705, "lr": 2.1945866861741044e-05, "elapsed_sec": 11214.898438692093, "step_time_sec": 20.296141814964358, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14718, "loss": 2.9603817462921143, "lr": 2.1929824561403507e-05, "elapsed_sec": 11235.191650152206, "step_time_sec": 20.293082272983156, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14719, "loss": 3.2905678749084473, "lr": 2.1913805697589485e-05, "elapsed_sec": 11255.485347747803, "step_time_sec": 20.29347997100558, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14720, "loss": 2.8524112701416016, "lr": 2.1897810218978105e-05, "elapsed_sec": 11275.779911279678, "step_time_sec": 20.29447937500663, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14721, "loss": 2.712590456008911, "lr": 2.1881838074398253e-05, "elapsed_sec": 11296.072237730026, "step_time_sec": 20.292214529006742, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14722, "loss": 2.7762606143951416, "lr": 2.1865889212827992e-05, "elapsed_sec": 11316.366557836533, "step_time_sec": 20.29405266599497, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14723, "loss": 3.1094179153442383, "lr": 2.184996358339403e-05, "elapsed_sec": 11336.662580013275, "step_time_sec": 20.29583748197183, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14724, "loss": 3.293379783630371, "lr": 2.183406113537118e-05, "elapsed_sec": 11356.957025766373, "step_time_sec": 20.294304521987215, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14725, "loss": 3.2059102058410645, "lr": 2.181818181818182e-05, "elapsed_sec": 11377.251116275787, "step_time_sec": 20.2939261419815, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14726, "loss": 2.726285219192505, "lr": 2.1802325581395353e-05, "elapsed_sec": 11397.54714679718, "step_time_sec": 20.295811487012543, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14727, "loss": 3.119354009628296, "lr": 2.178649237472767e-05, "elapsed_sec": 11417.84137415886, "step_time_sec": 20.294133846007753, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14728, "loss": 2.9841127395629883, "lr": 2.177068214804064e-05, "elapsed_sec": 11438.135333061218, "step_time_sec": 20.293786857975647, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14729, "loss": 2.937411308288574, "lr": 2.1754894851341557e-05, "elapsed_sec": 11458.431651830673, "step_time_sec": 20.296070429030806, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14730, "loss": 3.2536723613739014, "lr": 2.173913043478261e-05, "elapsed_sec": 11478.724133491516, "step_time_sec": 20.292420139012393, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14731, "loss": 3.0648787021636963, "lr": 2.1723388848660392e-05, "elapsed_sec": 11499.016632318497, "step_time_sec": 20.292231238039676, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14732, "loss": 2.8838844299316406, "lr": 2.1707670043415345e-05, "elapsed_sec": 11519.314717054367, "step_time_sec": 20.29794881399721, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14733, "loss": 2.988574743270874, "lr": 2.169197396963124e-05, "elapsed_sec": 11539.609112977982, "step_time_sec": 20.29416408902034, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14734, "loss": 2.97137713432312, "lr": 2.1676300578034685e-05, "elapsed_sec": 11559.90276145935, "step_time_sec": 20.293542158964556, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14735, "loss": 2.9742395877838135, "lr": 2.1660649819494586e-05, "elapsed_sec": 11580.198554754257, "step_time_sec": 20.295604671991896, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14736, "loss": 3.1275649070739746, "lr": 2.164502164502165e-05, "elapsed_sec": 11600.492937803268, "step_time_sec": 20.294193073990755, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14737, "loss": 3.104492664337158, "lr": 2.162941600576785e-05, "elapsed_sec": 11620.786359548569, "step_time_sec": 20.29326886497438, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14738, "loss": 2.9018821716308594, "lr": 2.1613832853025936e-05, "elapsed_sec": 11641.082486867905, "step_time_sec": 20.29592420597328, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14739, "loss": 2.501403570175171, "lr": 2.1598272138228946e-05, "elapsed_sec": 11661.376643657684, "step_time_sec": 20.29399747500429, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14740, "loss": 3.020723581314087, "lr": 2.1582733812949642e-05, "elapsed_sec": 11681.670082569122, "step_time_sec": 20.293315012997482, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14741, "loss": 3.2612829208374023, "lr": 2.1567217828900075e-05, "elapsed_sec": 11701.965436935425, "step_time_sec": 20.295108881022315, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14742, "loss": 3.00956130027771, "lr": 2.1551724137931036e-05, "elapsed_sec": 11722.259223937988, "step_time_sec": 20.293688278994523, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14743, "loss": 3.323131561279297, "lr": 2.1536252692031587e-05, "elapsed_sec": 11742.554267644882, "step_time_sec": 20.294852941995487, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14744, "loss": 2.923809051513672, "lr": 2.1520803443328555e-05, "elapsed_sec": 11762.854056596756, "step_time_sec": 20.29962008597795, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14745, "loss": 3.209899663925171, "lr": 2.1505376344086024e-05, "elapsed_sec": 11783.146491289139, "step_time_sec": 20.292185105965473, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14746, "loss": 2.9388813972473145, "lr": 2.1489971346704873e-05, "elapsed_sec": 11803.4413459301, "step_time_sec": 20.29480475903256, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14747, "loss": 3.2663214206695557, "lr": 2.1474588403722266e-05, "elapsed_sec": 11823.736037492752, "step_time_sec": 20.29442899202695, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14748, "loss": 2.9466302394866943, "lr": 2.1459227467811164e-05, "elapsed_sec": 11844.031195640564, "step_time_sec": 20.29509741900256, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14749, "loss": 3.111006021499634, "lr": 2.1443888491779845e-05, "elapsed_sec": 11864.275580883026, "step_time_sec": 20.244148313009646, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14750, "loss": 2.9698007106781006, "lr": 2.142857142857143e-05, "elapsed_sec": 11884.572651147842, "step_time_sec": 89.55087272397941, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14751, "loss": 2.9135687351226807, "lr": 2.1413276231263384e-05, "elapsed_sec": 11974.072368860245, "step_time_sec": 20.245682011998724, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14752, "loss": 2.8195302486419678, "lr": 2.139800285306705e-05, "elapsed_sec": 11994.325055599213, "step_time_sec": 20.252488749974873, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14753, "loss": 3.1901862621307373, "lr": 2.1382751247327157e-05, "elapsed_sec": 12014.576565742493, "step_time_sec": 20.25129988696426, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14754, "loss": 3.132554054260254, "lr": 2.136752136752137e-05, "elapsed_sec": 12034.872246265411, "step_time_sec": 20.29552359104855, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14755, "loss": 2.9640581607818604, "lr": 2.135231316725979e-05, "elapsed_sec": 12055.16918516159, "step_time_sec": 20.29669307405129, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14756, "loss": 3.2530176639556885, "lr": 2.1337126600284495e-05, "elapsed_sec": 12075.46211528778, "step_time_sec": 20.292818190995604, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14757, "loss": 2.9657199382781982, "lr": 2.1321961620469086e-05, "elapsed_sec": 12095.757426738739, "step_time_sec": 20.295103985001333, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14758, "loss": 2.754300594329834, "lr": 2.1306818181818183e-05, "elapsed_sec": 12116.053265810013, "step_time_sec": 20.295799440995324, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14759, "loss": 2.8972694873809814, "lr": 2.1291696238467002e-05, "elapsed_sec": 12136.34696984291, "step_time_sec": 20.293489895004313, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14760, "loss": 3.130112648010254, "lr": 2.1276595744680856e-05, "elapsed_sec": 12156.642750263214, "step_time_sec": 20.29551521904068, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14761, "loss": 3.0833888053894043, "lr": 2.1261516654854714e-05, "elapsed_sec": 12176.938275814056, "step_time_sec": 20.2954459960456, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14762, "loss": 3.055039882659912, "lr": 2.1246458923512754e-05, "elapsed_sec": 12197.232644796371, "step_time_sec": 20.294113218027633, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14763, "loss": 2.7731144428253174, "lr": 2.1231422505307857e-05, "elapsed_sec": 12217.527304649353, "step_time_sec": 20.294494414003566, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14764, "loss": 3.058884382247925, "lr": 2.121640735502122e-05, "elapsed_sec": 12237.82235121727, "step_time_sec": 20.294863273971714, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14765, "loss": 3.2585456371307373, "lr": 2.120141342756184e-05, "elapsed_sec": 12258.116617679596, "step_time_sec": 20.2941840760177, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14766, "loss": 2.758425235748291, "lr": 2.1186440677966107e-05, "elapsed_sec": 12278.412507295609, "step_time_sec": 20.29566257999977, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14767, "loss": 3.03973650932312, "lr": 2.117148906139732e-05, "elapsed_sec": 12298.707724571228, "step_time_sec": 20.294986739987507, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14768, "loss": 3.280247926712036, "lr": 2.1156558533145277e-05, "elapsed_sec": 12319.000936985016, "step_time_sec": 20.293103662028443, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14769, "loss": 3.0571706295013428, "lr": 2.1141649048625795e-05, "elapsed_sec": 12339.296663761139, "step_time_sec": 20.295472417026758, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14770, "loss": 3.140362501144409, "lr": 2.1126760563380288e-05, "elapsed_sec": 12359.592759609222, "step_time_sec": 20.296043498965446, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14771, "loss": 3.1904497146606445, "lr": 2.11118930330753e-05, "elapsed_sec": 12379.887667655945, "step_time_sec": 20.29463013098575, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14772, "loss": 3.2616758346557617, "lr": 2.1097046413502112e-05, "elapsed_sec": 12400.183535337448, "step_time_sec": 20.29574021603912, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14773, "loss": 2.9620227813720703, "lr": 2.108222066057625e-05, "elapsed_sec": 12420.479271888733, "step_time_sec": 20.29559524403885, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14774, "loss": 3.0188984870910645, "lr": 2.106741573033708e-05, "elapsed_sec": 12440.773670196533, "step_time_sec": 20.294237782014534, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14775, "loss": 2.757904052734375, "lr": 2.1052631578947372e-05, "elapsed_sec": 12461.069506406784, "step_time_sec": 20.295582314953208, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14776, "loss": 2.7883284091949463, "lr": 2.103786816269285e-05, "elapsed_sec": 12481.366441249847, "step_time_sec": 20.296809321036562, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14777, "loss": 3.1235902309417725, "lr": 2.1023125437981784e-05, "elapsed_sec": 12501.661507606506, "step_time_sec": 20.294925407972187, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14778, "loss": 2.9888718128204346, "lr": 2.1008403361344542e-05, "elapsed_sec": 12521.955903053284, "step_time_sec": 20.29430345201399, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14779, "loss": 3.0157923698425293, "lr": 2.099370188943317e-05, "elapsed_sec": 12542.249109983444, "step_time_sec": 20.292963578016497, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14780, "loss": 3.1474342346191406, "lr": 2.0979020979020984e-05, "elapsed_sec": 12562.544226646423, "step_time_sec": 20.294894782011397, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14781, "loss": 3.1357388496398926, "lr": 2.09643605870021e-05, "elapsed_sec": 12582.84052491188, "step_time_sec": 20.29619859100785, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14782, "loss": 3.085314989089966, "lr": 2.0949720670391065e-05, "elapsed_sec": 12603.135294437408, "step_time_sec": 20.29455934400903, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14783, "loss": 2.9441490173339844, "lr": 2.0935101186322405e-05, "elapsed_sec": 12623.42955994606, "step_time_sec": 20.29404649598291, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14784, "loss": 3.104279041290283, "lr": 2.0920502092050214e-05, "elapsed_sec": 12643.724556684494, "step_time_sec": 20.29488640301861, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14785, "loss": 3.123767852783203, "lr": 2.0905923344947736e-05, "elapsed_sec": 12664.019005537033, "step_time_sec": 20.29427057900466, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14786, "loss": 3.211829900741577, "lr": 2.0891364902506964e-05, "elapsed_sec": 12684.313100099564, "step_time_sec": 20.293959123955574, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14787, "loss": 3.0369837284088135, "lr": 2.087682672233821e-05, "elapsed_sec": 12704.608576774597, "step_time_sec": 20.295200675958768, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14788, "loss": 2.8031575679779053, "lr": 2.0862308762169682e-05, "elapsed_sec": 12724.904077529907, "step_time_sec": 20.29545761703048, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14789, "loss": 3.1282875537872314, "lr": 2.0847810979847118e-05, "elapsed_sec": 12745.199493646622, "step_time_sec": 20.295160514011513, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14790, "loss": 3.0840601921081543, "lr": 2.0833333333333336e-05, "elapsed_sec": 12765.495182037354, "step_time_sec": 20.29550998902414, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14791, "loss": 2.9477598667144775, "lr": 2.0818875780707845e-05, "elapsed_sec": 12785.789454460144, "step_time_sec": 20.29415146401152, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14792, "loss": 2.9953277111053467, "lr": 2.080443828016644e-05, "elapsed_sec": 12806.04055929184, "step_time_sec": 20.25097002903931, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14793, "loss": 2.7508797645568848, "lr": 2.0790020790020793e-05, "elapsed_sec": 12826.336554765701, "step_time_sec": 20.29584007500671, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14794, "loss": 2.8282971382141113, "lr": 2.0775623268698063e-05, "elapsed_sec": 12846.632610082626, "step_time_sec": 20.295753328013234, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14795, "loss": 3.116091012954712, "lr": 2.076124567474049e-05, "elapsed_sec": 12866.926549196243, "step_time_sec": 20.293846800981555, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14796, "loss": 2.9686131477355957, "lr": 2.074688796680498e-05, "elapsed_sec": 12887.22269320488, "step_time_sec": 20.29597019200446, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14797, "loss": 3.173510789871216, "lr": 2.0732550103662754e-05, "elapsed_sec": 12907.517633914948, "step_time_sec": 20.294711217982695, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14798, "loss": 3.074167251586914, "lr": 2.07182320441989e-05, "elapsed_sec": 12927.811036109924, "step_time_sec": 20.293401677976362, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14799, "loss": 3.100010395050049, "lr": 2.070393374741201e-05, "elapsed_sec": 12948.110035419464, "step_time_sec": 20.298803970974404, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14800, "loss": 3.0025177001953125, "lr": 2.0689655172413797e-05, "elapsed_sec": 12968.410388469696, "step_time_sec": 95.86004593799589, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14801, "loss": 3.1852779388427734, "lr": 2.0675396278428672e-05, "elapsed_sec": 13064.216254234314, "step_time_sec": 20.245765013969503, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14802, "loss": 3.0558085441589355, "lr": 2.066115702479339e-05, "elapsed_sec": 13084.460396766663, "step_time_sec": 20.244043672049884, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14803, "loss": 3.2099709510803223, "lr": 2.0646937370956648e-05, "elapsed_sec": 13104.757232666016, "step_time_sec": 20.296586476964876, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14804, "loss": 3.1638104915618896, "lr": 2.0632737276478684e-05, "elapsed_sec": 13125.05343079567, "step_time_sec": 20.29606372700073, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14805, "loss": 2.8769690990448, "lr": 2.061855670103093e-05, "elapsed_sec": 13145.347031116486, "step_time_sec": 20.293489175965078, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14806, "loss": 3.1275973320007324, "lr": 2.0604395604395608e-05, "elapsed_sec": 13165.644016742706, "step_time_sec": 20.296713074028958, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14807, "loss": 3.2140934467315674, "lr": 2.0590253946465342e-05, "elapsed_sec": 13185.940699577332, "step_time_sec": 20.296625778020825, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14808, "loss": 2.8155007362365723, "lr": 2.05761316872428e-05, "elapsed_sec": 13206.237083435059, "step_time_sec": 20.2961752639967, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14809, "loss": 3.2400670051574707, "lr": 2.0562028786840303e-05, "elapsed_sec": 13226.532541275024, "step_time_sec": 20.295247911999468, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14810, "loss": 3.0164499282836914, "lr": 2.0547945205479456e-05, "elapsed_sec": 13246.82799744606, "step_time_sec": 20.295272916031536, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14811, "loss": 3.140031337738037, "lr": 2.053388090349076e-05, "elapsed_sec": 13267.122593164444, "step_time_sec": 20.29441728000529, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14812, "loss": 2.9991559982299805, "lr": 2.051983584131327e-05, "elapsed_sec": 13287.418045759201, "step_time_sec": 20.2952777579776, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14813, "loss": 3.0802950859069824, "lr": 2.0505809979494193e-05, "elapsed_sec": 13307.714617967606, "step_time_sec": 20.29636112699518, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14814, "loss": 2.9286129474639893, "lr": 2.0491803278688528e-05, "elapsed_sec": 13328.05422449112, "step_time_sec": 20.30478637100896, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14815, "loss": 3.0903146266937256, "lr": 2.0477815699658708e-05, "elapsed_sec": 13348.350186347961, "step_time_sec": 20.29578008397948, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14816, "loss": 3.02911114692688, "lr": 2.046384720327422e-05, "elapsed_sec": 13368.646079063416, "step_time_sec": 20.29582256998401, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14817, "loss": 3.2195823192596436, "lr": 2.044989775051125e-05, "elapsed_sec": 13388.94290304184, "step_time_sec": 20.29667246999452, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14818, "loss": 3.2187366485595703, "lr": 2.043596730245232e-05, "elapsed_sec": 13409.240273952484, "step_time_sec": 20.29709081997862, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14819, "loss": 3.034708023071289, "lr": 2.042205582028591e-05, "elapsed_sec": 13429.539461135864, "step_time_sec": 20.299063294020016, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14820, "loss": 3.3949973583221436, "lr": 2.0408163265306126e-05, "elapsed_sec": 13449.833633899689, "step_time_sec": 20.29399180901237, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14821, "loss": 3.2538886070251465, "lr": 2.0394289598912306e-05, "elapsed_sec": 13470.130506515503, "step_time_sec": 20.296733791998122, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14822, "loss": 2.962343454360962, "lr": 2.03804347826087e-05, "elapsed_sec": 13490.427655220032, "step_time_sec": 20.296999787038658, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14823, "loss": 2.68149995803833, "lr": 2.0366598778004075e-05, "elapsed_sec": 13510.723243236542, "step_time_sec": 20.295359140029177, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14824, "loss": 2.910261869430542, "lr": 2.03527815468114e-05, "elapsed_sec": 13531.021242141724, "step_time_sec": 20.297781174012925, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14825, "loss": 2.8913795948028564, "lr": 2.0338983050847462e-05, "elapsed_sec": 13551.31826210022, "step_time_sec": 20.296902409987524, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14826, "loss": 2.87841796875, "lr": 2.0325203252032524e-05, "elapsed_sec": 13571.612529754639, "step_time_sec": 20.294083287997637, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14827, "loss": 3.1794116497039795, "lr": 2.031144211238998e-05, "elapsed_sec": 13591.908519983292, "step_time_sec": 20.295757359010167, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14828, "loss": 3.2521281242370605, "lr": 2.0297699594046013e-05, "elapsed_sec": 13612.204085350037, "step_time_sec": 20.29552069096826, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14829, "loss": 2.873164415359497, "lr": 2.0283975659229208e-05, "elapsed_sec": 13632.500267744064, "step_time_sec": 20.295901755976956, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14830, "loss": 2.798046112060547, "lr": 2.0270270270270273e-05, "elapsed_sec": 13652.797776222229, "step_time_sec": 20.297389381972607, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14831, "loss": 2.8264317512512207, "lr": 2.0256583389601622e-05, "elapsed_sec": 13673.094727754593, "step_time_sec": 20.296740697987843, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14832, "loss": 2.8732216358184814, "lr": 2.0242914979757088e-05, "elapsed_sec": 13693.389204740524, "step_time_sec": 20.29430033097742, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14833, "loss": 2.8542284965515137, "lr": 2.022926500337155e-05, "elapsed_sec": 13713.68609213829, "step_time_sec": 20.296825438039377, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14834, "loss": 2.7608649730682373, "lr": 2.0215633423180595e-05, "elapsed_sec": 13733.982067108154, "step_time_sec": 20.295718333974946, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14835, "loss": 3.20353364944458, "lr": 2.0202020202020206e-05, "elapsed_sec": 13754.276511192322, "step_time_sec": 20.294302406022325, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14836, "loss": 2.970308780670166, "lr": 2.0188425302826384e-05, "elapsed_sec": 13774.53122973442, "step_time_sec": 20.254486093006562, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14837, "loss": 3.320341110229492, "lr": 2.017484868863484e-05, "elapsed_sec": 13794.828103780746, "step_time_sec": 20.296788952022325, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14838, "loss": 3.000316619873047, "lr": 2.016129032258065e-05, "elapsed_sec": 13815.12275648117, "step_time_sec": 20.294503910001367, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14839, "loss": 3.001391887664795, "lr": 2.014775016789792e-05, "elapsed_sec": 13835.419256925583, "step_time_sec": 20.296281903982162, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14840, "loss": 3.1607043743133545, "lr": 2.0134228187919465e-05, "elapsed_sec": 13855.714679956436, "step_time_sec": 20.295168715005275, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14841, "loss": 3.0188310146331787, "lr": 2.012072434607646e-05, "elapsed_sec": 13876.007687330246, "step_time_sec": 20.292925050016493, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14842, "loss": 2.859942674636841, "lr": 2.0107238605898124e-05, "elapsed_sec": 13896.304084300995, "step_time_sec": 20.296237856033258, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14843, "loss": 2.734583616256714, "lr": 2.009377093101139e-05, "elapsed_sec": 13916.600497484207, "step_time_sec": 20.2961310549872, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14844, "loss": 2.903538703918457, "lr": 2.0080321285140566e-05, "elapsed_sec": 13936.895529031754, "step_time_sec": 20.294951389019843, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14845, "loss": 3.0242230892181396, "lr": 2.0066889632107027e-05, "elapsed_sec": 13957.193504333496, "step_time_sec": 20.297778365027625, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14846, "loss": 3.098963975906372, "lr": 2.005347593582888e-05, "elapsed_sec": 13977.489894390106, "step_time_sec": 20.296212226967327, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14847, "loss": 3.071321487426758, "lr": 2.0040080160320643e-05, "elapsed_sec": 13997.785160303116, "step_time_sec": 20.295077067043167, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14848, "loss": 3.1383237838745117, "lr": 2.002670226969293e-05, "elapsed_sec": 14018.081498384476, "step_time_sec": 20.29613250796683, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14849, "loss": 3.1606638431549072, "lr": 2.0013342228152104e-05, "elapsed_sec": 14038.378963708878, "step_time_sec": 20.297365125035867, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": false, "validation_ran": false, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
| {"step": 14850, "loss": 2.895521879196167, "lr": 2e-05, "elapsed_sec": 14058.674489736557, "step_time_sec": 86.04011881101178, "effective_batch_tokens": 239904, "sequence_length": 2499, "batch_size": 2, "grad_accum_steps": 48, "checkpoint_saved": true, "validation_ran": true, "probe_ran": false, "probe_elapsed_sec": 0.0, "warmup_excluded": false, "peak_vram_bytes": 12537172480} |
|
|
