| {"step": 1, "loss": 2.31618595123291, "gradient_norm": 52.89557647705078, "seconds": 7.569801424993784, "decisions": 4, "branches": 10, "actual_branch_tokens": 1159, "padded_branch_tokens": 1159, "peak_cuda_allocated_bytes": 16437236224, "lr_factor": 0.01} |
| {"step": 2, "loss": 11.291901940538082, "gradient_norm": 194.39552307128906, "seconds": 8.439798079998582, "decisions": 4, "branches": 12, "actual_branch_tokens": 1309, "padded_branch_tokens": 1309, "peak_cuda_allocated_bytes": 16455383040, "lr_factor": 0.02} |
| {"step": 3, "loss": 0.0, "gradient_norm": 4.621360183421075e-09, "seconds": 8.751493896998, "decisions": 4, "branches": 12, "actual_branch_tokens": 1418, "padded_branch_tokens": 1418, "peak_cuda_allocated_bytes": 16473681920, "lr_factor": 0.03} |
| {"step": 4, "loss": 4.425229549407959, "gradient_norm": 41.81777572631836, "seconds": 10.466120449011214, "decisions": 4, "branches": 13, "actual_branch_tokens": 1685, "padded_branch_tokens": 1685, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.04} |
| {"step": 5, "loss": 0.24940787092782557, "gradient_norm": 29.091127395629883, "seconds": 8.057185469006072, "decisions": 4, "branches": 11, "actual_branch_tokens": 1378, "padded_branch_tokens": 1378, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.05} |
| {"step": 6, "loss": 0.00441942794714123, "gradient_norm": 0.8078902959823608, "seconds": 9.62166352600616, "decisions": 4, "branches": 14, "actual_branch_tokens": 1315, "padded_branch_tokens": 1315, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.06} |
| {"step": 7, "loss": 1.5947637943390873, "gradient_norm": 46.06306076049805, "seconds": 9.153237838996574, "decisions": 4, "branches": 12, "actual_branch_tokens": 1566, "padded_branch_tokens": 1566, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.07} |
| {"step": 8, "loss": 1.6465007066726685, "gradient_norm": 57.00694274902344, "seconds": 8.156997896003304, "decisions": 4, "branches": 10, "actual_branch_tokens": 1525, "padded_branch_tokens": 1525, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.08} |
| {"step": 9, "loss": 0.7907518148422081, "gradient_norm": 41.083614349365234, "seconds": 9.618812154993066, "decisions": 4, "branches": 13, "actual_branch_tokens": 1478, "padded_branch_tokens": 1478, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.09} |
| {"step": 10, "loss": 5.856753602089157, "gradient_norm": 68.11426544189453, "seconds": 8.596327488994575, "decisions": 4, "branches": 12, "actual_branch_tokens": 1381, "padded_branch_tokens": 1381, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.1} |
| {"step": 11, "loss": 0.008205516263842583, "gradient_norm": 2.2738704681396484, "seconds": 10.407428573991638, "decisions": 4, "branches": 15, "actual_branch_tokens": 1370, "padded_branch_tokens": 1370, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.11} |
| {"step": 12, "loss": 1.204397112129982, "gradient_norm": 95.3646240234375, "seconds": 8.759487815012108, "decisions": 4, "branches": 11, "actual_branch_tokens": 1426, "padded_branch_tokens": 1426, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.12} |
| {"step": 13, "loss": 0.9008829102967866, "gradient_norm": 50.37225341796875, "seconds": 8.866891690995544, "decisions": 4, "branches": 12, "actual_branch_tokens": 1423, "padded_branch_tokens": 1423, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.13} |
| {"step": 14, "loss": 0.16515788435935974, "gradient_norm": 19.898183822631836, "seconds": 8.842795117001515, "decisions": 4, "branches": 13, "actual_branch_tokens": 1348, "padded_branch_tokens": 1348, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.14} |
| {"step": 15, "loss": 2.1074386537075043, "gradient_norm": 41.82075881958008, "seconds": 8.362906023990945, "decisions": 4, "branches": 11, "actual_branch_tokens": 1405, "padded_branch_tokens": 1405, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.15} |
| {"step": 16, "loss": 9.000139471027069e-06, "gradient_norm": 0.0014378582127392292, "seconds": 9.300638262007851, "decisions": 4, "branches": 13, "actual_branch_tokens": 1444, "padded_branch_tokens": 1444, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.16} |
| {"step": 17, "loss": 0.03407556359888986, "gradient_norm": 5.7712554931640625, "seconds": 8.382403843002976, "decisions": 4, "branches": 12, "actual_branch_tokens": 1330, "padded_branch_tokens": 1330, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.17} |
| {"step": 18, "loss": 1.6320235118982964, "gradient_norm": 33.489078521728516, "seconds": 8.501154495010269, "decisions": 4, "branches": 12, "actual_branch_tokens": 1282, "padded_branch_tokens": 1282, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.18} |
| {"step": 19, "loss": 0.1661178022621641, "gradient_norm": 4.28471040725708, "seconds": 10.22584532800829, "decisions": 4, "branches": 15, "actual_branch_tokens": 1304, "padded_branch_tokens": 1304, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.19} |
| {"step": 20, "loss": 1.5273072719573975, "gradient_norm": 41.74165344238281, "seconds": 9.493052620004164, "decisions": 4, "branches": 14, "actual_branch_tokens": 1336, "padded_branch_tokens": 1336, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.2} |
| {"step": 21, "loss": 0.0011900107992914855, "gradient_norm": 0.15585853159427643, "seconds": 9.150072851000004, "decisions": 4, "branches": 13, "actual_branch_tokens": 1346, "padded_branch_tokens": 1346, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.21} |
| {"step": 22, "loss": 4.321321853240079e-06, "gradient_norm": 0.000585337111260742, "seconds": 9.47616612398997, "decisions": 4, "branches": 14, "actual_branch_tokens": 1367, "padded_branch_tokens": 1367, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.22} |
| {"step": 23, "loss": 0.5652832686521521, "gradient_norm": 31.015239715576172, "seconds": 8.032757008011686, "decisions": 4, "branches": 11, "actual_branch_tokens": 1154, "padded_branch_tokens": 1154, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.23} |
| {"step": 24, "loss": 0.8583411186700687, "gradient_norm": 60.29244613647461, "seconds": 9.473117874003947, "decisions": 4, "branches": 13, "actual_branch_tokens": 1283, "padded_branch_tokens": 1283, "peak_cuda_allocated_bytes": 16649119744, "lr_factor": 0.24} |
| {"step": 25, "loss": 0.03423098344455866, "gradient_norm": 2.508150815963745, "seconds": 10.122523254001862, "decisions": 4, "branches": 12, "actual_branch_tokens": 1818, "padded_branch_tokens": 1818, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.25} |
| {"step": 26, "loss": 3.2406842655618675, "gradient_norm": 38.515743255615234, "seconds": 9.817800144010107, "decisions": 4, "branches": 13, "actual_branch_tokens": 1372, "padded_branch_tokens": 1372, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.26} |
| {"step": 27, "loss": 0.43557077599689364, "gradient_norm": 17.96202850341797, "seconds": 8.166928169011953, "decisions": 4, "branches": 11, "actual_branch_tokens": 1225, "padded_branch_tokens": 1225, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.27} |
| {"step": 28, "loss": 0.0012749117286148248, "gradient_norm": 0.12147028744220734, "seconds": 10.022878604009748, "decisions": 4, "branches": 14, "actual_branch_tokens": 1450, "padded_branch_tokens": 1450, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.28} |
| {"step": 29, "loss": 0.1557716807496945, "gradient_norm": 6.705139636993408, "seconds": 8.533127572998637, "decisions": 4, "branches": 12, "actual_branch_tokens": 1282, "padded_branch_tokens": 1282, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.29} |
| {"step": 30, "loss": 0.21547929197546978, "gradient_norm": 10.085920333862305, "seconds": 8.928555799997412, "decisions": 4, "branches": 13, "actual_branch_tokens": 1305, "padded_branch_tokens": 1305, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.3} |
| {"step": 31, "loss": 1.1161180965609674, "gradient_norm": 26.214340209960938, "seconds": 9.20974643700174, "decisions": 4, "branches": 13, "actual_branch_tokens": 1332, "padded_branch_tokens": 1332, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.31} |
| {"step": 32, "loss": 2.6071371813304722, "gradient_norm": 34.83671951293945, "seconds": 9.068318688005093, "decisions": 4, "branches": 13, "actual_branch_tokens": 1353, "padded_branch_tokens": 1353, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.32} |
| {"step": 33, "loss": 0.7383988115470856, "gradient_norm": 28.386564254760742, "seconds": 9.51455305799027, "decisions": 4, "branches": 12, "actual_branch_tokens": 1521, "padded_branch_tokens": 1521, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.33} |
| {"step": 34, "loss": 1.7194390930234533, "gradient_norm": 42.854896545410156, "seconds": 8.959811183987767, "decisions": 4, "branches": 12, "actual_branch_tokens": 1272, "padded_branch_tokens": 1272, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.34} |
| {"step": 35, "loss": 6.356305129706641e-05, "gradient_norm": 0.008878614753484726, "seconds": 10.86596634700254, "decisions": 4, "branches": 14, "actual_branch_tokens": 1712, "padded_branch_tokens": 1712, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.35} |
| {"step": 36, "loss": 0.8997114369494739, "gradient_norm": 16.77312469482422, "seconds": 8.096381487994222, "decisions": 4, "branches": 10, "actual_branch_tokens": 1461, "padded_branch_tokens": 1461, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.36} |
| {"step": 37, "loss": 0.10167956212532658, "gradient_norm": 4.7068328857421875, "seconds": 10.060175599996, "decisions": 4, "branches": 15, "actual_branch_tokens": 1358, "padded_branch_tokens": 1358, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.37} |
| {"step": 38, "loss": 0.34200737811625004, "gradient_norm": 19.233367919921875, "seconds": 8.628295389993582, "decisions": 4, "branches": 12, "actual_branch_tokens": 1243, "padded_branch_tokens": 1243, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.38} |
| {"step": 39, "loss": 0.7590227595064789, "gradient_norm": 18.048818588256836, "seconds": 8.802753324998775, "decisions": 4, "branches": 13, "actual_branch_tokens": 1296, "padded_branch_tokens": 1296, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.39} |
| {"step": 40, "loss": 0.09986791973824438, "gradient_norm": 3.940859317779541, "seconds": 8.952552449001814, "decisions": 4, "branches": 13, "actual_branch_tokens": 1236, "padded_branch_tokens": 1236, "peak_cuda_allocated_bytes": 16654055424, "lr_factor": 0.4} |
|
|
