TypePredictor / figures /eval_metrics.json
hadikhamoud's picture
Add final model card with evaluation curves and release artifacts
984c462 verified
Raw
History Blame Contribute Delete
35.4 kB
[
{
"step": 500,
"eval_index": 1,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.08,
"eval_accuracy": 0.4996031746031746,
"eval_loss": 1.6034170389175415,
"eval_macro_f1": 0.16675213835131356,
"eval_macro_precision": 0.20112826287340668,
"eval_macro_recall": 0.1656798423386093,
"eval_micro_f1": 0.4996031746031746,
"eval_micro_precision": 0.4996031746031746,
"eval_micro_recall": 0.4996031746031746,
"eval_runtime": 50.3632,
"eval_samples_per_second": 250.183,
"eval_steps_per_second": 31.273,
"eval_weighted_f1": 0.4648136683073095,
"eval_weighted_precision": 0.4663299729027011,
"eval_weighted_recall": 0.4996031746031746
},
{
"step": 1000,
"eval_index": 2,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.16,
"eval_accuracy": 0.7561111111111111,
"eval_loss": 0.8336577415466309,
"eval_macro_f1": 0.3742962053337755,
"eval_macro_precision": 0.4440270170632841,
"eval_macro_recall": 0.36141326244744065,
"eval_micro_f1": 0.7561111111111111,
"eval_micro_precision": 0.7561111111111111,
"eval_micro_recall": 0.7561111111111111,
"eval_runtime": 50.358,
"eval_samples_per_second": 250.208,
"eval_steps_per_second": 31.276,
"eval_weighted_f1": 0.7347115811510523,
"eval_weighted_precision": 0.739827037528709,
"eval_weighted_recall": 0.7561111111111111
},
{
"step": 1500,
"eval_index": 3,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.24,
"eval_accuracy": 0.8311111111111111,
"eval_loss": 0.5750956535339355,
"eval_macro_f1": 0.5919220318807322,
"eval_macro_precision": 0.6615551279169473,
"eval_macro_recall": 0.5725095316587115,
"eval_micro_f1": 0.8311111111111111,
"eval_micro_precision": 0.8311111111111111,
"eval_micro_recall": 0.8311111111111111,
"eval_runtime": 50.1384,
"eval_samples_per_second": 251.304,
"eval_steps_per_second": 31.413,
"eval_weighted_f1": 0.8258660663106083,
"eval_weighted_precision": 0.8313369703842328,
"eval_weighted_recall": 0.8311111111111111
},
{
"step": 2000,
"eval_index": 4,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.32,
"eval_accuracy": 0.8685714285714285,
"eval_loss": 0.43678638339042664,
"eval_macro_f1": 0.704721130059806,
"eval_macro_precision": 0.8084736830738437,
"eval_macro_recall": 0.6569171685565485,
"eval_micro_f1": 0.8685714285714285,
"eval_micro_precision": 0.8685714285714285,
"eval_micro_recall": 0.8685714285714285,
"eval_runtime": 54.0525,
"eval_samples_per_second": 233.107,
"eval_steps_per_second": 29.138,
"eval_weighted_f1": 0.8650850638163624,
"eval_weighted_precision": 0.8704622381617023,
"eval_weighted_recall": 0.8685714285714285
},
{
"step": 2500,
"eval_index": 5,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.4,
"eval_accuracy": 0.8876190476190476,
"eval_loss": 0.3863752782344818,
"eval_macro_f1": 0.7708534077145469,
"eval_macro_precision": 0.8424204092431431,
"eval_macro_recall": 0.7602036693520349,
"eval_micro_f1": 0.8876190476190476,
"eval_micro_precision": 0.8876190476190476,
"eval_micro_recall": 0.8876190476190476,
"eval_runtime": 58.1835,
"eval_samples_per_second": 216.556,
"eval_steps_per_second": 27.07,
"eval_weighted_f1": 0.8866826747603135,
"eval_weighted_precision": 0.894918351088462,
"eval_weighted_recall": 0.8876190476190476
},
{
"step": 3000,
"eval_index": 6,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.48,
"eval_accuracy": 0.9049206349206349,
"eval_loss": 0.3749255836009979,
"eval_macro_f1": 0.83146311466218,
"eval_macro_precision": 0.8822616279099216,
"eval_macro_recall": 0.8044703495421244,
"eval_micro_f1": 0.9049206349206349,
"eval_micro_precision": 0.9049206349206349,
"eval_micro_recall": 0.9049206349206349,
"eval_runtime": 51.2979,
"eval_samples_per_second": 245.624,
"eval_steps_per_second": 30.703,
"eval_weighted_f1": 0.9042368915226878,
"eval_weighted_precision": 0.9091398319969811,
"eval_weighted_recall": 0.9049206349206349
},
{
"step": 3500,
"eval_index": 7,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.56,
"eval_accuracy": 0.9016666666666666,
"eval_loss": 0.34068959951400757,
"eval_macro_f1": 0.852643844128256,
"eval_macro_precision": 0.8869002266347256,
"eval_macro_recall": 0.8344066000317897,
"eval_micro_f1": 0.9016666666666666,
"eval_micro_precision": 0.9016666666666666,
"eval_micro_recall": 0.9016666666666666,
"eval_runtime": 55.1712,
"eval_samples_per_second": 228.38,
"eval_steps_per_second": 28.547,
"eval_weighted_f1": 0.9056553567113477,
"eval_weighted_precision": 0.9166483950530401,
"eval_weighted_recall": 0.9016666666666666
},
{
"step": 4000,
"eval_index": 8,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.63,
"eval_accuracy": 0.9234126984126985,
"eval_loss": 0.27018845081329346,
"eval_macro_f1": 0.8629184237562926,
"eval_macro_precision": 0.896103415544095,
"eval_macro_recall": 0.8427102453711095,
"eval_micro_f1": 0.9234126984126985,
"eval_micro_precision": 0.9234126984126985,
"eval_micro_recall": 0.9234126984126985,
"eval_runtime": 45.4117,
"eval_samples_per_second": 277.462,
"eval_steps_per_second": 34.683,
"eval_weighted_f1": 0.9225121978819937,
"eval_weighted_precision": 0.9237214610954837,
"eval_weighted_recall": 0.9234126984126985
},
{
"step": 4500,
"eval_index": 9,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.71,
"eval_accuracy": 0.9284126984126985,
"eval_loss": 0.25139176845550537,
"eval_macro_f1": 0.8858543195103097,
"eval_macro_precision": 0.9001288293009205,
"eval_macro_recall": 0.8748161720063445,
"eval_micro_f1": 0.9284126984126985,
"eval_micro_precision": 0.9284126984126985,
"eval_micro_recall": 0.9284126984126985,
"eval_runtime": 65.3815,
"eval_samples_per_second": 192.715,
"eval_steps_per_second": 24.089,
"eval_weighted_f1": 0.9284553859947943,
"eval_weighted_precision": 0.9293969786750571,
"eval_weighted_recall": 0.9284126984126985
},
{
"step": 5000,
"eval_index": 10,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.79,
"eval_accuracy": 0.9337301587301587,
"eval_loss": 0.24514073133468628,
"eval_macro_f1": 0.8968955709213094,
"eval_macro_precision": 0.9239801593127872,
"eval_macro_recall": 0.8775447083795846,
"eval_micro_f1": 0.9337301587301587,
"eval_micro_precision": 0.9337301587301587,
"eval_micro_recall": 0.9337301587301587,
"eval_runtime": 55.1353,
"eval_samples_per_second": 228.529,
"eval_steps_per_second": 28.566,
"eval_weighted_f1": 0.9333867608778208,
"eval_weighted_precision": 0.9342365821570298,
"eval_weighted_recall": 0.9337301587301587
},
{
"step": 5500,
"eval_index": 11,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.87,
"eval_accuracy": 0.9361904761904762,
"eval_loss": 0.2405300885438919,
"eval_macro_f1": 0.8852157589128977,
"eval_macro_precision": 0.8966852505493809,
"eval_macro_recall": 0.8879515092023682,
"eval_micro_f1": 0.9361904761904762,
"eval_micro_precision": 0.9361904761904762,
"eval_micro_recall": 0.9361904761904762,
"eval_runtime": 50.0976,
"eval_samples_per_second": 251.509,
"eval_steps_per_second": 31.439,
"eval_weighted_f1": 0.9373120231008413,
"eval_weighted_precision": 0.940336582906886,
"eval_weighted_recall": 0.9361904761904762
},
{
"step": 6000,
"eval_index": 12,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 0.95,
"eval_accuracy": 0.941984126984127,
"eval_loss": 0.2432594746351242,
"eval_macro_f1": 0.9103364146476316,
"eval_macro_precision": 0.9298312623198658,
"eval_macro_recall": 0.8954247781763811,
"eval_micro_f1": 0.941984126984127,
"eval_micro_precision": 0.941984126984127,
"eval_micro_recall": 0.941984126984127,
"eval_runtime": 49.9297,
"eval_samples_per_second": 252.355,
"eval_steps_per_second": 31.544,
"eval_weighted_f1": 0.9417305390461214,
"eval_weighted_precision": 0.9424591918065203,
"eval_weighted_recall": 0.941984126984127
},
{
"step": 6500,
"eval_index": 13,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.03,
"eval_accuracy": 0.9456349206349206,
"eval_loss": 0.22848264873027802,
"eval_macro_f1": 0.909507315203273,
"eval_macro_precision": 0.9296991239988623,
"eval_macro_recall": 0.8961337328597196,
"eval_micro_f1": 0.9456349206349206,
"eval_micro_precision": 0.9456349206349206,
"eval_micro_recall": 0.9456349206349206,
"eval_runtime": 48.6038,
"eval_samples_per_second": 259.239,
"eval_steps_per_second": 32.405,
"eval_weighted_f1": 0.9454117579717388,
"eval_weighted_precision": 0.9459939034429095,
"eval_weighted_recall": 0.9456349206349206
},
{
"step": 7000,
"eval_index": 14,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.11,
"eval_accuracy": 0.9484126984126984,
"eval_loss": 0.20768418908119202,
"eval_macro_f1": 0.9125337197557731,
"eval_macro_precision": 0.9190348121850224,
"eval_macro_recall": 0.9113443135094732,
"eval_micro_f1": 0.9484126984126984,
"eval_micro_precision": 0.9484126984126984,
"eval_micro_recall": 0.9484126984126984,
"eval_runtime": 51.0028,
"eval_samples_per_second": 247.045,
"eval_steps_per_second": 30.881,
"eval_weighted_f1": 0.9481753539970462,
"eval_weighted_precision": 0.9487731067127975,
"eval_weighted_recall": 0.9484126984126984
},
{
"step": 7500,
"eval_index": 15,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.19,
"eval_accuracy": 0.9442857142857143,
"eval_loss": 0.2635294198989868,
"eval_macro_f1": 0.9057940212609807,
"eval_macro_precision": 0.911842211214998,
"eval_macro_recall": 0.9069588500794394,
"eval_micro_f1": 0.9442857142857143,
"eval_micro_precision": 0.9442857142857143,
"eval_micro_recall": 0.9442857142857143,
"eval_runtime": 53.2935,
"eval_samples_per_second": 236.427,
"eval_steps_per_second": 29.553,
"eval_weighted_f1": 0.9443844599149406,
"eval_weighted_precision": 0.945921955299144,
"eval_weighted_recall": 0.9442857142857143
},
{
"step": 8000,
"eval_index": 16,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.27,
"eval_accuracy": 0.9494444444444444,
"eval_loss": 0.21279270946979523,
"eval_macro_f1": 0.9101912478331972,
"eval_macro_precision": 0.9273715052159579,
"eval_macro_recall": 0.901402436657646,
"eval_micro_f1": 0.9494444444444444,
"eval_micro_precision": 0.9494444444444444,
"eval_micro_recall": 0.9494444444444444,
"eval_runtime": 48.6141,
"eval_samples_per_second": 259.184,
"eval_steps_per_second": 32.398,
"eval_weighted_f1": 0.9491690418479931,
"eval_weighted_precision": 0.9504200289531454,
"eval_weighted_recall": 0.9494444444444444
},
{
"step": 8500,
"eval_index": 17,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.35,
"eval_accuracy": 0.9530952380952381,
"eval_loss": 0.21746815741062164,
"eval_macro_f1": 0.9156584913591659,
"eval_macro_precision": 0.9241711528138504,
"eval_macro_recall": 0.9152950181018766,
"eval_micro_f1": 0.9530952380952381,
"eval_micro_precision": 0.9530952380952381,
"eval_micro_recall": 0.9530952380952381,
"eval_runtime": 49.6088,
"eval_samples_per_second": 253.987,
"eval_steps_per_second": 31.748,
"eval_weighted_f1": 0.9528707065770654,
"eval_weighted_precision": 0.9533906189807044,
"eval_weighted_recall": 0.9530952380952381
},
{
"step": 9000,
"eval_index": 18,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.43,
"eval_accuracy": 0.9544444444444444,
"eval_loss": 0.21344760060310364,
"eval_macro_f1": 0.9316646003398935,
"eval_macro_precision": 0.9481060096209216,
"eval_macro_recall": 0.9195331693514776,
"eval_micro_f1": 0.9544444444444444,
"eval_micro_precision": 0.9544444444444444,
"eval_micro_recall": 0.9544444444444444,
"eval_runtime": 49.9732,
"eval_samples_per_second": 252.135,
"eval_steps_per_second": 31.517,
"eval_weighted_f1": 0.9541756739346082,
"eval_weighted_precision": 0.9548010972085054,
"eval_weighted_recall": 0.9544444444444444
},
{
"step": 9500,
"eval_index": 19,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.51,
"eval_accuracy": 0.9568253968253968,
"eval_loss": 0.19153176248073578,
"eval_macro_f1": 0.921610631007778,
"eval_macro_precision": 0.9404398497869574,
"eval_macro_recall": 0.905748060531071,
"eval_micro_f1": 0.9568253968253968,
"eval_micro_precision": 0.9568253968253968,
"eval_micro_recall": 0.9568253968253968,
"eval_runtime": 49.0787,
"eval_samples_per_second": 256.731,
"eval_steps_per_second": 32.091,
"eval_weighted_f1": 0.9566293314445607,
"eval_weighted_precision": 0.9568515996032587,
"eval_weighted_recall": 0.9568253968253968
},
{
"step": 10000,
"eval_index": 20,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.59,
"eval_accuracy": 0.959920634920635,
"eval_loss": 0.19577036798000336,
"eval_macro_f1": 0.9235742546691592,
"eval_macro_precision": 0.9273628410860382,
"eval_macro_recall": 0.9271792802963574,
"eval_micro_f1": 0.959920634920635,
"eval_micro_precision": 0.959920634920635,
"eval_micro_recall": 0.959920634920635,
"eval_runtime": 60.9394,
"eval_samples_per_second": 206.763,
"eval_steps_per_second": 25.845,
"eval_weighted_f1": 0.9599028886198321,
"eval_weighted_precision": 0.960305322421165,
"eval_weighted_recall": 0.959920634920635
},
{
"step": 10500,
"eval_index": 21,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.67,
"eval_accuracy": 0.960952380952381,
"eval_loss": 0.17503954470157623,
"eval_macro_f1": 0.9170932665272469,
"eval_macro_precision": 0.9116456444630479,
"eval_macro_recall": 0.9260931287646225,
"eval_micro_f1": 0.960952380952381,
"eval_micro_precision": 0.960952380952381,
"eval_micro_recall": 0.960952380952381,
"eval_runtime": 45.8116,
"eval_samples_per_second": 275.04,
"eval_steps_per_second": 34.38,
"eval_weighted_f1": 0.9609014768411174,
"eval_weighted_precision": 0.9611475161993005,
"eval_weighted_recall": 0.960952380952381
},
{
"step": 11000,
"eval_index": 22,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.75,
"eval_accuracy": 0.9624603174603175,
"eval_loss": 0.17708104848861694,
"eval_macro_f1": 0.9319695940105357,
"eval_macro_precision": 0.9462998611668713,
"eval_macro_recall": 0.9234620410796542,
"eval_micro_f1": 0.9624603174603175,
"eval_micro_precision": 0.9624603174603175,
"eval_micro_recall": 0.9624603174603175,
"eval_runtime": 45.9838,
"eval_samples_per_second": 274.009,
"eval_steps_per_second": 34.251,
"eval_weighted_f1": 0.9623559428261603,
"eval_weighted_precision": 0.962723564261779,
"eval_weighted_recall": 0.9624603174603175
},
{
"step": 11500,
"eval_index": 23,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.83,
"eval_accuracy": 0.9632539682539683,
"eval_loss": 0.17640231549739838,
"eval_macro_f1": 0.9383726570780089,
"eval_macro_precision": 0.9361695510382104,
"eval_macro_recall": 0.9420156784780935,
"eval_micro_f1": 0.9632539682539683,
"eval_micro_precision": 0.9632539682539683,
"eval_micro_recall": 0.9632539682539683,
"eval_runtime": 46.3964,
"eval_samples_per_second": 271.573,
"eval_steps_per_second": 33.947,
"eval_weighted_f1": 0.9633197036349216,
"eval_weighted_precision": 0.9636434749852494,
"eval_weighted_recall": 0.9632539682539683
},
{
"step": 12000,
"eval_index": 24,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.9,
"eval_accuracy": 0.9657142857142857,
"eval_loss": 0.16351820528507233,
"eval_macro_f1": 0.9426202066832035,
"eval_macro_precision": 0.9507506384547523,
"eval_macro_recall": 0.9366009019904619,
"eval_micro_f1": 0.9657142857142857,
"eval_micro_precision": 0.9657142857142857,
"eval_micro_recall": 0.9657142857142857,
"eval_runtime": 46.8483,
"eval_samples_per_second": 268.953,
"eval_steps_per_second": 33.619,
"eval_weighted_f1": 0.9656532805046822,
"eval_weighted_precision": 0.9659512006173707,
"eval_weighted_recall": 0.9657142857142857
},
{
"step": 12500,
"eval_index": 25,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 1.98,
"eval_accuracy": 0.9681746031746031,
"eval_loss": 0.1587391495704651,
"eval_macro_f1": 0.9466181342033713,
"eval_macro_precision": 0.947442786667393,
"eval_macro_recall": 0.9465717167703007,
"eval_micro_f1": 0.9681746031746031,
"eval_micro_precision": 0.9681746031746031,
"eval_micro_recall": 0.9681746031746031,
"eval_runtime": 66.4354,
"eval_samples_per_second": 189.658,
"eval_steps_per_second": 23.707,
"eval_weighted_f1": 0.9681650956489429,
"eval_weighted_precision": 0.9683707150909681,
"eval_weighted_recall": 0.9681746031746031
},
{
"step": 13000,
"eval_index": 26,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.06,
"eval_accuracy": 0.9689682539682539,
"eval_loss": 0.16660018265247345,
"eval_macro_f1": 0.9431082887845699,
"eval_macro_precision": 0.9435320164238983,
"eval_macro_recall": 0.9440342115841122,
"eval_micro_f1": 0.9689682539682539,
"eval_micro_precision": 0.9689682539682539,
"eval_micro_recall": 0.9689682539682539,
"eval_runtime": 55.6815,
"eval_samples_per_second": 226.287,
"eval_steps_per_second": 28.286,
"eval_weighted_f1": 0.9690350615547291,
"eval_weighted_precision": 0.9692930831154812,
"eval_weighted_recall": 0.9689682539682539
},
{
"step": 13500,
"eval_index": 27,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.14,
"eval_accuracy": 0.9687301587301588,
"eval_loss": 0.18048988282680511,
"eval_macro_f1": 0.9422923846160581,
"eval_macro_precision": 0.9532870485294106,
"eval_macro_recall": 0.934424229646169,
"eval_micro_f1": 0.9687301587301588,
"eval_micro_precision": 0.9687301587301588,
"eval_micro_recall": 0.9687301587301588,
"eval_runtime": 48.6144,
"eval_samples_per_second": 259.182,
"eval_steps_per_second": 32.398,
"eval_weighted_f1": 0.9686287726395876,
"eval_weighted_precision": 0.9689670561915049,
"eval_weighted_recall": 0.9687301587301588
},
{
"step": 14000,
"eval_index": 28,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.22,
"eval_accuracy": 0.9709523809523809,
"eval_loss": 0.17584019899368286,
"eval_macro_f1": 0.9548845957042732,
"eval_macro_precision": 0.9616225902344291,
"eval_macro_recall": 0.9490903584972133,
"eval_micro_f1": 0.9709523809523809,
"eval_micro_precision": 0.9709523809523809,
"eval_micro_recall": 0.9709523809523809,
"eval_runtime": 65.6446,
"eval_samples_per_second": 191.943,
"eval_steps_per_second": 23.993,
"eval_weighted_f1": 0.9709045086147281,
"eval_weighted_precision": 0.9709764613384323,
"eval_weighted_recall": 0.9709523809523809
},
{
"step": 14500,
"eval_index": 29,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.3,
"eval_accuracy": 0.9695238095238096,
"eval_loss": 0.177631214261055,
"eval_macro_f1": 0.9506224965454739,
"eval_macro_precision": 0.9661964085054093,
"eval_macro_recall": 0.9377514378889877,
"eval_micro_f1": 0.9695238095238096,
"eval_micro_precision": 0.9695238095238096,
"eval_micro_recall": 0.9695238095238096,
"eval_runtime": 55.0712,
"eval_samples_per_second": 228.795,
"eval_steps_per_second": 28.599,
"eval_weighted_f1": 0.9694153409312327,
"eval_weighted_precision": 0.9696628969782775,
"eval_weighted_recall": 0.9695238095238096
},
{
"step": 15000,
"eval_index": 30,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.38,
"eval_accuracy": 0.97,
"eval_loss": 0.17878225445747375,
"eval_macro_f1": 0.952375532557884,
"eval_macro_precision": 0.9642924920360566,
"eval_macro_recall": 0.9429523998338022,
"eval_micro_f1": 0.97,
"eval_micro_precision": 0.97,
"eval_micro_recall": 0.97,
"eval_runtime": 51.0169,
"eval_samples_per_second": 246.977,
"eval_steps_per_second": 30.872,
"eval_weighted_f1": 0.9699429015929246,
"eval_weighted_precision": 0.9701562195790373,
"eval_weighted_recall": 0.97
},
{
"step": 15500,
"eval_index": 31,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.46,
"eval_accuracy": 0.9695238095238096,
"eval_loss": 0.17210516333580017,
"eval_macro_f1": 0.9462875984373381,
"eval_macro_precision": 0.9580763571099492,
"eval_macro_recall": 0.9378514622035224,
"eval_micro_f1": 0.9695238095238096,
"eval_micro_precision": 0.9695238095238096,
"eval_micro_recall": 0.9695238095238096,
"eval_runtime": 49.2932,
"eval_samples_per_second": 255.613,
"eval_steps_per_second": 31.952,
"eval_weighted_f1": 0.9694666934898747,
"eval_weighted_precision": 0.9696918541700538,
"eval_weighted_recall": 0.9695238095238096
},
{
"step": 16000,
"eval_index": 32,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.54,
"eval_accuracy": 0.9714285714285714,
"eval_loss": 0.15938520431518555,
"eval_macro_f1": 0.951264365385694,
"eval_macro_precision": 0.9601219800413174,
"eval_macro_recall": 0.9443849087252134,
"eval_micro_f1": 0.9714285714285714,
"eval_micro_precision": 0.9714285714285714,
"eval_micro_recall": 0.9714285714285714,
"eval_runtime": 64.8873,
"eval_samples_per_second": 194.183,
"eval_steps_per_second": 24.273,
"eval_weighted_f1": 0.971362978431738,
"eval_weighted_precision": 0.9714994225414888,
"eval_weighted_recall": 0.9714285714285714
},
{
"step": 16500,
"eval_index": 33,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.62,
"eval_accuracy": 0.9723015873015873,
"eval_loss": 0.1643410474061966,
"eval_macro_f1": 0.9561822135615655,
"eval_macro_precision": 0.9665323672953172,
"eval_macro_recall": 0.947527505766003,
"eval_micro_f1": 0.9723015873015873,
"eval_micro_precision": 0.9723015873015873,
"eval_micro_recall": 0.9723015873015873,
"eval_runtime": 52.441,
"eval_samples_per_second": 240.27,
"eval_steps_per_second": 30.034,
"eval_weighted_f1": 0.9722751933392187,
"eval_weighted_precision": 0.9724025203500458,
"eval_weighted_recall": 0.9723015873015873
},
{
"step": 17000,
"eval_index": 34,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.7,
"eval_accuracy": 0.973015873015873,
"eval_loss": 0.16805726289749146,
"eval_macro_f1": 0.9491028977586443,
"eval_macro_precision": 0.960433813735491,
"eval_macro_recall": 0.9406320914982489,
"eval_micro_f1": 0.973015873015873,
"eval_micro_precision": 0.973015873015873,
"eval_micro_recall": 0.973015873015873,
"eval_runtime": 50.2558,
"eval_samples_per_second": 250.717,
"eval_steps_per_second": 31.34,
"eval_weighted_f1": 0.9729576202329057,
"eval_weighted_precision": 0.9730875205196509,
"eval_weighted_recall": 0.973015873015873
},
{
"step": 17500,
"eval_index": 35,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.78,
"eval_accuracy": 0.9734126984126984,
"eval_loss": 0.16671566665172577,
"eval_macro_f1": 0.9471381395941902,
"eval_macro_precision": 0.9549885122141759,
"eval_macro_recall": 0.9428154605456508,
"eval_micro_f1": 0.9734126984126984,
"eval_micro_precision": 0.9734126984126984,
"eval_micro_recall": 0.9734126984126984,
"eval_runtime": 64.8418,
"eval_samples_per_second": 194.319,
"eval_steps_per_second": 24.29,
"eval_weighted_f1": 0.9733825407985148,
"eval_weighted_precision": 0.973532406246267,
"eval_weighted_recall": 0.9734126984126984
},
{
"step": 18000,
"eval_index": 36,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.86,
"eval_accuracy": 0.9738095238095238,
"eval_loss": 0.17004433274269104,
"eval_macro_f1": 0.9529234191380291,
"eval_macro_precision": 0.962677256883139,
"eval_macro_recall": 0.9454556144210745,
"eval_micro_f1": 0.9738095238095238,
"eval_micro_precision": 0.9738095238095238,
"eval_micro_recall": 0.9738095238095238,
"eval_runtime": 48.8808,
"eval_samples_per_second": 257.77,
"eval_steps_per_second": 32.221,
"eval_weighted_f1": 0.9737837762311287,
"eval_weighted_precision": 0.9739539079223564,
"eval_weighted_recall": 0.9738095238095238
},
{
"step": 18500,
"eval_index": 37,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 2.94,
"eval_accuracy": 0.975,
"eval_loss": 0.16028916835784912,
"eval_macro_f1": 0.9540199515604112,
"eval_macro_precision": 0.9647089557155375,
"eval_macro_recall": 0.945480105008737,
"eval_micro_f1": 0.975,
"eval_micro_precision": 0.975,
"eval_micro_recall": 0.975,
"eval_runtime": 60.6955,
"eval_samples_per_second": 207.594,
"eval_steps_per_second": 25.949,
"eval_weighted_f1": 0.9749518139909482,
"eval_weighted_precision": 0.9750584916764344,
"eval_weighted_recall": 0.975
},
{
"step": 19000,
"eval_index": 38,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.02,
"eval_accuracy": 0.9762698412698413,
"eval_loss": 0.14878861606121063,
"eval_macro_f1": 0.954347658252044,
"eval_macro_precision": 0.9607715131633574,
"eval_macro_recall": 0.9501403976665715,
"eval_micro_f1": 0.9762698412698413,
"eval_micro_precision": 0.9762698412698413,
"eval_micro_recall": 0.9762698412698413,
"eval_runtime": 51.7414,
"eval_samples_per_second": 243.519,
"eval_steps_per_second": 30.44,
"eval_weighted_f1": 0.9762667198823463,
"eval_weighted_precision": 0.9764195140172701,
"eval_weighted_recall": 0.9762698412698413
},
{
"step": 19500,
"eval_index": 39,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.1,
"eval_accuracy": 0.9756349206349206,
"eval_loss": 0.16310656070709229,
"eval_macro_f1": 0.957125861216084,
"eval_macro_precision": 0.9727967123798239,
"eval_macro_recall": 0.9442922894357393,
"eval_micro_f1": 0.9756349206349206,
"eval_micro_precision": 0.9756349206349206,
"eval_micro_recall": 0.9756349206349206,
"eval_runtime": 47.4743,
"eval_samples_per_second": 265.407,
"eval_steps_per_second": 33.176,
"eval_weighted_f1": 0.9755432972728447,
"eval_weighted_precision": 0.9756807849259415,
"eval_weighted_recall": 0.9756349206349206
},
{
"step": 20000,
"eval_index": 40,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.17,
"eval_accuracy": 0.9767460317460317,
"eval_loss": 0.16326671838760376,
"eval_macro_f1": 0.9567534320462188,
"eval_macro_precision": 0.9698508452291146,
"eval_macro_recall": 0.9464068389671931,
"eval_micro_f1": 0.9767460317460317,
"eval_micro_precision": 0.9767460317460317,
"eval_micro_recall": 0.9767460317460317,
"eval_runtime": 65.9869,
"eval_samples_per_second": 190.947,
"eval_steps_per_second": 23.868,
"eval_weighted_f1": 0.9766894759901187,
"eval_weighted_precision": 0.9767954117631042,
"eval_weighted_recall": 0.9767460317460317
},
{
"step": 20500,
"eval_index": 41,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.25,
"eval_accuracy": 0.9761111111111112,
"eval_loss": 0.1650867611169815,
"eval_macro_f1": 0.955655463952606,
"eval_macro_precision": 0.9677917142854108,
"eval_macro_recall": 0.9461865102151874,
"eval_micro_f1": 0.9761111111111112,
"eval_micro_precision": 0.9761111111111112,
"eval_micro_recall": 0.9761111111111112,
"eval_runtime": 63.0368,
"eval_samples_per_second": 199.883,
"eval_steps_per_second": 24.985,
"eval_weighted_f1": 0.9760636712967664,
"eval_weighted_precision": 0.9761579320691899,
"eval_weighted_recall": 0.9761111111111112
},
{
"step": 21000,
"eval_index": 42,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.33,
"eval_accuracy": 0.9763492063492063,
"eval_loss": 0.17093592882156372,
"eval_macro_f1": 0.956419648478896,
"eval_macro_precision": 0.9707661139733746,
"eval_macro_recall": 0.9448057653586319,
"eval_micro_f1": 0.9763492063492063,
"eval_micro_precision": 0.9763492063492063,
"eval_micro_recall": 0.9763492063492063,
"eval_runtime": 52.3722,
"eval_samples_per_second": 240.586,
"eval_steps_per_second": 30.073,
"eval_weighted_f1": 0.9762812381183971,
"eval_weighted_precision": 0.976400813243511,
"eval_weighted_recall": 0.9763492063492063
},
{
"step": 21500,
"eval_index": 43,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.41,
"eval_accuracy": 0.9769047619047619,
"eval_loss": 0.16325406730175018,
"eval_macro_f1": 0.9557812168870482,
"eval_macro_precision": 0.9696293003435598,
"eval_macro_recall": 0.9445357837826058,
"eval_micro_f1": 0.9769047619047619,
"eval_micro_precision": 0.9769047619047619,
"eval_micro_recall": 0.9769047619047619,
"eval_runtime": 48.9392,
"eval_samples_per_second": 257.462,
"eval_steps_per_second": 32.183,
"eval_weighted_f1": 0.9768367289110993,
"eval_weighted_precision": 0.9769254523389815,
"eval_weighted_recall": 0.9769047619047619
},
{
"step": 22000,
"eval_index": 44,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.49,
"eval_accuracy": 0.9756349206349206,
"eval_loss": 0.16351020336151123,
"eval_macro_f1": 0.9551717284616833,
"eval_macro_precision": 0.9681364681283637,
"eval_macro_recall": 0.945098492018414,
"eval_micro_f1": 0.9756349206349206,
"eval_micro_precision": 0.9756349206349206,
"eval_micro_recall": 0.9756349206349206,
"eval_runtime": 48.5537,
"eval_samples_per_second": 259.506,
"eval_steps_per_second": 32.438,
"eval_weighted_f1": 0.9755767607894801,
"eval_weighted_precision": 0.9756980555491296,
"eval_weighted_recall": 0.9756349206349206
},
{
"step": 22500,
"eval_index": 45,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.57,
"eval_accuracy": 0.9765873015873016,
"eval_loss": 0.16308797895908356,
"eval_macro_f1": 0.9559072153039692,
"eval_macro_precision": 0.9680547193198462,
"eval_macro_recall": 0.9463389894990366,
"eval_micro_f1": 0.9765873015873016,
"eval_micro_precision": 0.9765873015873016,
"eval_micro_recall": 0.9765873015873016,
"eval_runtime": 49.1975,
"eval_samples_per_second": 256.11,
"eval_steps_per_second": 32.014,
"eval_weighted_f1": 0.9765228546965284,
"eval_weighted_precision": 0.9766302422786851,
"eval_weighted_recall": 0.9765873015873016
},
{
"step": 23000,
"eval_index": 46,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.65,
"eval_accuracy": 0.9774603174603175,
"eval_loss": 0.16377675533294678,
"eval_macro_f1": 0.9572334679389451,
"eval_macro_precision": 0.9700362196030153,
"eval_macro_recall": 0.9471315195900166,
"eval_micro_f1": 0.9774603174603175,
"eval_micro_precision": 0.9774603174603175,
"eval_micro_recall": 0.9774603174603175,
"eval_runtime": 64.2461,
"eval_samples_per_second": 196.121,
"eval_steps_per_second": 24.515,
"eval_weighted_f1": 0.9773904798618511,
"eval_weighted_precision": 0.9775254966358302,
"eval_weighted_recall": 0.9774603174603175
},
{
"step": 23500,
"eval_index": 47,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.73,
"eval_accuracy": 0.9770634920634921,
"eval_loss": 0.16450679302215576,
"eval_macro_f1": 0.957808445275978,
"eval_macro_precision": 0.9706746434129168,
"eval_macro_recall": 0.9475056428847263,
"eval_micro_f1": 0.9770634920634921,
"eval_micro_precision": 0.9770634920634921,
"eval_micro_recall": 0.9770634920634921,
"eval_runtime": 48.1744,
"eval_samples_per_second": 261.55,
"eval_steps_per_second": 32.694,
"eval_weighted_f1": 0.9769934673940778,
"eval_weighted_precision": 0.9770703311250897,
"eval_weighted_recall": 0.9770634920634921
},
{
"step": 24000,
"eval_index": 48,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.81,
"eval_accuracy": 0.9775396825396825,
"eval_loss": 0.16624021530151367,
"eval_macro_f1": 0.9583518371874029,
"eval_macro_precision": 0.9715119738925228,
"eval_macro_recall": 0.947699963206565,
"eval_micro_f1": 0.9775396825396825,
"eval_micro_precision": 0.9775396825396825,
"eval_micro_recall": 0.9775396825396825,
"eval_runtime": 46.101,
"eval_samples_per_second": 273.313,
"eval_steps_per_second": 34.164,
"eval_weighted_f1": 0.9774702480661045,
"eval_weighted_precision": 0.9775614944397367,
"eval_weighted_recall": 0.9775396825396825
},
{
"step": 24500,
"eval_index": 49,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.89,
"eval_accuracy": 0.9776984126984127,
"eval_loss": 0.16271832585334778,
"eval_macro_f1": 0.9581437912558748,
"eval_macro_precision": 0.9706729136929291,
"eval_macro_recall": 0.9480541459491892,
"eval_micro_f1": 0.9776984126984127,
"eval_micro_precision": 0.9776984126984127,
"eval_micro_recall": 0.9776984126984127,
"eval_runtime": 46.9842,
"eval_samples_per_second": 268.175,
"eval_steps_per_second": 33.522,
"eval_weighted_f1": 0.9776414544812937,
"eval_weighted_precision": 0.9777223020348919,
"eval_weighted_recall": 0.9776984126984127
},
{
"step": 25000,
"eval_index": 50,
"seen_at_utc": "2026-07-12T19:07:28+00:00",
"epoch": 3.97,
"eval_accuracy": 0.9778571428571429,
"eval_loss": 0.16299165785312653,
"eval_macro_f1": 0.9587281193044483,
"eval_macro_precision": 0.9725024753902682,
"eval_macro_recall": 0.9475659044801065,
"eval_micro_f1": 0.9778571428571429,
"eval_micro_precision": 0.9778571428571429,
"eval_micro_recall": 0.9778571428571429,
"eval_runtime": 45.6663,
"eval_samples_per_second": 275.914,
"eval_steps_per_second": 34.489,
"eval_weighted_f1": 0.9777906938936711,
"eval_weighted_precision": 0.977879248064854,
"eval_weighted_recall": 0.9778571428571429
}
]