| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 181, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0055248618784530384, | |
| "grad_norm": 3.9804399013519287, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 5.2929, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011049723756906077, | |
| "grad_norm": 5.95102071762085, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 5.6811, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.016574585635359115, | |
| "grad_norm": 6.399249076843262, | |
| "learning_rate": 3e-06, | |
| "loss": 6.0649, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.022099447513812154, | |
| "grad_norm": 6.8701982498168945, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 6.054, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.027624309392265192, | |
| "grad_norm": 7.215997219085693, | |
| "learning_rate": 5e-06, | |
| "loss": 6.1508, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03314917127071823, | |
| "grad_norm": 9.195732116699219, | |
| "learning_rate": 6e-06, | |
| "loss": 5.9273, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03867403314917127, | |
| "grad_norm": 12.597426414489746, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 5.8757, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.04419889502762431, | |
| "grad_norm": 11.52719783782959, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 6.3232, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.049723756906077346, | |
| "grad_norm": 9.948908805847168, | |
| "learning_rate": 9e-06, | |
| "loss": 7.0539, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.055248618784530384, | |
| "grad_norm": 15.710564613342285, | |
| "learning_rate": 1e-05, | |
| "loss": 6.4505, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06077348066298342, | |
| "grad_norm": 12.096623420715332, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 7.0895, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06629834254143646, | |
| "grad_norm": 72.50323486328125, | |
| "learning_rate": 1.2e-05, | |
| "loss": 13.0317, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0718232044198895, | |
| "grad_norm": 4.915622234344482, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 4.458, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07734806629834254, | |
| "grad_norm": 4.010397911071777, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 6.3289, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08287292817679558, | |
| "grad_norm": 5.764005661010742, | |
| "learning_rate": 1.5e-05, | |
| "loss": 5.8784, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08839779005524862, | |
| "grad_norm": 6.722602844238281, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 5.7637, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09392265193370165, | |
| "grad_norm": 6.578387260437012, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 5.9894, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.09944751381215469, | |
| "grad_norm": 6.852511405944824, | |
| "learning_rate": 1.8e-05, | |
| "loss": 6.0283, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10497237569060773, | |
| "grad_norm": 11.264864921569824, | |
| "learning_rate": 1.9e-05, | |
| "loss": 5.5976, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11049723756906077, | |
| "grad_norm": 12.024802207946777, | |
| "learning_rate": 2e-05, | |
| "loss": 5.6752, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11602209944751381, | |
| "grad_norm": 7.510622024536133, | |
| "learning_rate": 2.1e-05, | |
| "loss": 6.5538, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12154696132596685, | |
| "grad_norm": 12.385478973388672, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 6.0905, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1270718232044199, | |
| "grad_norm": 14.713839530944824, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 5.7133, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.13259668508287292, | |
| "grad_norm": 23.432165145874023, | |
| "learning_rate": 2.4e-05, | |
| "loss": 9.1648, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.13812154696132597, | |
| "grad_norm": 86.79972076416016, | |
| "learning_rate": 2.5e-05, | |
| "loss": 12.0242, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.143646408839779, | |
| "grad_norm": 3.4584054946899414, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 4.8302, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.14917127071823205, | |
| "grad_norm": 4.496112823486328, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 5.0922, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15469613259668508, | |
| "grad_norm": 6.2076029777526855, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 5.0634, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.16022099447513813, | |
| "grad_norm": 5.173417091369629, | |
| "learning_rate": 2.9e-05, | |
| "loss": 5.2037, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16574585635359115, | |
| "grad_norm": 6.735776424407959, | |
| "learning_rate": 3e-05, | |
| "loss": 4.8899, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1712707182320442, | |
| "grad_norm": 7.046944618225098, | |
| "learning_rate": 3.1e-05, | |
| "loss": 4.5185, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.17679558011049723, | |
| "grad_norm": 9.225940704345703, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 3.9559, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18232044198895028, | |
| "grad_norm": 8.940814018249512, | |
| "learning_rate": 3.3e-05, | |
| "loss": 4.024, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1878453038674033, | |
| "grad_norm": 6.720527648925781, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 5.5122, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.19337016574585636, | |
| "grad_norm": 14.003108978271484, | |
| "learning_rate": 3.5e-05, | |
| "loss": 4.6331, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.19889502762430938, | |
| "grad_norm": 12.830449104309082, | |
| "learning_rate": 3.6e-05, | |
| "loss": 5.3631, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.20441988950276244, | |
| "grad_norm": 96.01248931884766, | |
| "learning_rate": 3.7e-05, | |
| "loss": 8.2067, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.20994475138121546, | |
| "grad_norm": 3.922679901123047, | |
| "learning_rate": 3.8e-05, | |
| "loss": 3.702, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2154696132596685, | |
| "grad_norm": 3.2943949699401855, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 4.3623, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.22099447513812154, | |
| "grad_norm": 4.222253322601318, | |
| "learning_rate": 4e-05, | |
| "loss": 3.9072, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2265193370165746, | |
| "grad_norm": 3.905756711959839, | |
| "learning_rate": 4.1e-05, | |
| "loss": 3.5621, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.23204419889502761, | |
| "grad_norm": 4.060239315032959, | |
| "learning_rate": 4.2e-05, | |
| "loss": 3.7462, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.23756906077348067, | |
| "grad_norm": 3.8542888164520264, | |
| "learning_rate": 4.3e-05, | |
| "loss": 3.596, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2430939226519337, | |
| "grad_norm": 3.7129149436950684, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 3.0085, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.24861878453038674, | |
| "grad_norm": 4.574946403503418, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.3354, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2541436464088398, | |
| "grad_norm": 4.760665416717529, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 3.0683, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2596685082872928, | |
| "grad_norm": 8.342724800109863, | |
| "learning_rate": 4.7e-05, | |
| "loss": 4.2298, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.26519337016574585, | |
| "grad_norm": 10.613301277160645, | |
| "learning_rate": 4.8e-05, | |
| "loss": 3.0263, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.27071823204419887, | |
| "grad_norm": 12.508091926574707, | |
| "learning_rate": 4.9e-05, | |
| "loss": 3.7188, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.27624309392265195, | |
| "grad_norm": 49.06351089477539, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7781, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.281767955801105, | |
| "grad_norm": 4.32130241394043, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 3.2349, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.287292817679558, | |
| "grad_norm": 5.024357318878174, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 3.238, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.292817679558011, | |
| "grad_norm": 4.892840385437012, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 3.0299, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2983425414364641, | |
| "grad_norm": 3.730414628982544, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 3.2957, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.30386740331491713, | |
| "grad_norm": 3.0157148838043213, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.5627, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.30939226519337015, | |
| "grad_norm": 3.5059242248535156, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 2.5583, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3149171270718232, | |
| "grad_norm": 4.939709663391113, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 1.6391, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.32044198895027626, | |
| "grad_norm": 3.7843198776245117, | |
| "learning_rate": 5.8e-05, | |
| "loss": 2.2175, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3259668508287293, | |
| "grad_norm": 7.142724990844727, | |
| "learning_rate": 5.9e-05, | |
| "loss": 2.8714, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3314917127071823, | |
| "grad_norm": 9.160916328430176, | |
| "learning_rate": 6e-05, | |
| "loss": 1.865, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3370165745856354, | |
| "grad_norm": 8.80765438079834, | |
| "learning_rate": 6.1e-05, | |
| "loss": 2.2662, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3425414364640884, | |
| "grad_norm": 88.81080627441406, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.3548, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.34806629834254144, | |
| "grad_norm": 2.8465707302093506, | |
| "learning_rate": 6.3e-05, | |
| "loss": 2.7783, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.35359116022099446, | |
| "grad_norm": 2.7563679218292236, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 2.8983, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.35911602209944754, | |
| "grad_norm": 3.4825193881988525, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.8698, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.36464088397790057, | |
| "grad_norm": 3.067711353302002, | |
| "learning_rate": 6.6e-05, | |
| "loss": 2.7357, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.3701657458563536, | |
| "grad_norm": 3.0403783321380615, | |
| "learning_rate": 6.7e-05, | |
| "loss": 2.0308, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.3756906077348066, | |
| "grad_norm": 2.707793712615967, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 2.2543, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3812154696132597, | |
| "grad_norm": 2.735945224761963, | |
| "learning_rate": 6.9e-05, | |
| "loss": 2.0137, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3867403314917127, | |
| "grad_norm": 3.2104952335357666, | |
| "learning_rate": 7e-05, | |
| "loss": 1.6004, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.39226519337016574, | |
| "grad_norm": 3.5461206436157227, | |
| "learning_rate": 7.1e-05, | |
| "loss": 2.4595, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.39779005524861877, | |
| "grad_norm": 4.658211708068848, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.713, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.40331491712707185, | |
| "grad_norm": 5.711689472198486, | |
| "learning_rate": 7.3e-05, | |
| "loss": 0.9919, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4088397790055249, | |
| "grad_norm": 18.51163101196289, | |
| "learning_rate": 7.4e-05, | |
| "loss": 3.5525, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4143646408839779, | |
| "grad_norm": 89.78804016113281, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.6067, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4198895027624309, | |
| "grad_norm": 3.574091672897339, | |
| "learning_rate": 7.6e-05, | |
| "loss": 2.9171, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.425414364640884, | |
| "grad_norm": 3.860858678817749, | |
| "learning_rate": 7.7e-05, | |
| "loss": 2.7496, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.430939226519337, | |
| "grad_norm": 3.6850547790527344, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 2.4272, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.43646408839779005, | |
| "grad_norm": 3.1749162673950195, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 2.3552, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.4419889502762431, | |
| "grad_norm": 2.6579089164733887, | |
| "learning_rate": 8e-05, | |
| "loss": 2.202, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.44751381215469616, | |
| "grad_norm": 2.2511134147644043, | |
| "learning_rate": 8.1e-05, | |
| "loss": 1.7343, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.4530386740331492, | |
| "grad_norm": 2.7213551998138428, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.3241, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4585635359116022, | |
| "grad_norm": 2.9692206382751465, | |
| "learning_rate": 8.3e-05, | |
| "loss": 1.6291, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.46408839779005523, | |
| "grad_norm": 4.021595478057861, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.8124, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4696132596685083, | |
| "grad_norm": 3.9451472759246826, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.1126, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.47513812154696133, | |
| "grad_norm": 4.617029190063477, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.4704, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.48066298342541436, | |
| "grad_norm": 47.51131057739258, | |
| "learning_rate": 8.7e-05, | |
| "loss": 1.5657, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4861878453038674, | |
| "grad_norm": 2.8982062339782715, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 2.7824, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.49171270718232046, | |
| "grad_norm": 3.4071297645568848, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 3.1011, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.4972375690607735, | |
| "grad_norm": 3.2673630714416504, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4384, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5027624309392266, | |
| "grad_norm": 2.663628101348877, | |
| "learning_rate": 9.1e-05, | |
| "loss": 2.4114, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5082872928176796, | |
| "grad_norm": 2.7086033821105957, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 2.1603, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5138121546961326, | |
| "grad_norm": 2.409043788909912, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 1.8341, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5193370165745856, | |
| "grad_norm": 2.636979341506958, | |
| "learning_rate": 9.4e-05, | |
| "loss": 2.0164, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5248618784530387, | |
| "grad_norm": 3.0589420795440674, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.4738, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5303867403314917, | |
| "grad_norm": 2.772052764892578, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.7229, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5359116022099447, | |
| "grad_norm": 2.64821457862854, | |
| "learning_rate": 9.7e-05, | |
| "loss": 1.312, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5414364640883977, | |
| "grad_norm": 2.601024627685547, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.5236, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5469613259668509, | |
| "grad_norm": 7.427682876586914, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 2.5924, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5524861878453039, | |
| "grad_norm": 78.61561584472656, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8103, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5580110497237569, | |
| "grad_norm": 3.5797324180603027, | |
| "learning_rate": 9.996239762521151e-05, | |
| "loss": 2.8864, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.56353591160221, | |
| "grad_norm": 4.539000988006592, | |
| "learning_rate": 9.98496470583896e-05, | |
| "loss": 2.4452, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.569060773480663, | |
| "grad_norm": 4.56262731552124, | |
| "learning_rate": 9.966191788709716e-05, | |
| "loss": 2.5186, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.574585635359116, | |
| "grad_norm": 3.57301926612854, | |
| "learning_rate": 9.939949247384046e-05, | |
| "loss": 2.3074, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.580110497237569, | |
| "grad_norm": 2.9676594734191895, | |
| "learning_rate": 9.906276553136923e-05, | |
| "loss": 2.2185, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.585635359116022, | |
| "grad_norm": 2.439439535140991, | |
| "learning_rate": 9.865224352899119e-05, | |
| "loss": 1.7985, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5911602209944752, | |
| "grad_norm": 2.2659783363342285, | |
| "learning_rate": 9.816854393079403e-05, | |
| "loss": 1.4258, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5966850828729282, | |
| "grad_norm": 2.8905081748962402, | |
| "learning_rate": 9.761239426692077e-05, | |
| "loss": 1.9446, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6022099447513812, | |
| "grad_norm": 3.475658655166626, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 2.1245, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6077348066298343, | |
| "grad_norm": 3.568830728530884, | |
| "learning_rate": 9.628619846344454e-05, | |
| "loss": 0.651, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6132596685082873, | |
| "grad_norm": 4.190188407897949, | |
| "learning_rate": 9.551814704830734e-05, | |
| "loss": 1.3732, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6187845303867403, | |
| "grad_norm": 42.48178482055664, | |
| "learning_rate": 9.468163201617062e-05, | |
| "loss": 3.0859, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6243093922651933, | |
| "grad_norm": 3.153904676437378, | |
| "learning_rate": 9.377791156510455e-05, | |
| "loss": 3.0313, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6298342541436464, | |
| "grad_norm": 2.8389861583709717, | |
| "learning_rate": 9.280834497651334e-05, | |
| "loss": 2.6934, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6353591160220995, | |
| "grad_norm": 2.8842246532440186, | |
| "learning_rate": 9.177439057064683e-05, | |
| "loss": 2.2787, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6408839779005525, | |
| "grad_norm": 2.4654698371887207, | |
| "learning_rate": 9.067760351314838e-05, | |
| "loss": 2.4482, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6464088397790055, | |
| "grad_norm": 2.3591573238372803, | |
| "learning_rate": 8.951963347593797e-05, | |
| "loss": 1.9383, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6519337016574586, | |
| "grad_norm": 2.334392786026001, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 2.2049, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6574585635359116, | |
| "grad_norm": 2.0469932556152344, | |
| "learning_rate": 8.702720065545024e-05, | |
| "loss": 1.3291, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6629834254143646, | |
| "grad_norm": 2.4725210666656494, | |
| "learning_rate": 8.569648672789497e-05, | |
| "loss": 1.2557, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6685082872928176, | |
| "grad_norm": 2.755540132522583, | |
| "learning_rate": 8.43120818934367e-05, | |
| "loss": 2.008, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6740331491712708, | |
| "grad_norm": 2.401703357696533, | |
| "learning_rate": 8.28760684284532e-05, | |
| "loss": 0.8254, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6795580110497238, | |
| "grad_norm": 2.2195017337799072, | |
| "learning_rate": 8.139060623360493e-05, | |
| "loss": 0.4399, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6850828729281768, | |
| "grad_norm": 8.88201904296875, | |
| "learning_rate": 7.985792958513931e-05, | |
| "loss": 2.0833, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6906077348066298, | |
| "grad_norm": 50.49702072143555, | |
| "learning_rate": 7.828034377432693e-05, | |
| "loss": 0.4008, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6961325966850829, | |
| "grad_norm": 2.5161263942718506, | |
| "learning_rate": 7.666022164008457e-05, | |
| "loss": 2.5859, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7016574585635359, | |
| "grad_norm": 2.9786441326141357, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.3956, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7071823204419889, | |
| "grad_norm": 2.978400707244873, | |
| "learning_rate": 7.330217598512695e-05, | |
| "loss": 2.2787, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.712707182320442, | |
| "grad_norm": 2.7484891414642334, | |
| "learning_rate": 7.156930328406268e-05, | |
| "loss": 2.2322, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7182320441988951, | |
| "grad_norm": 2.5018727779388428, | |
| "learning_rate": 6.980398830195785e-05, | |
| "loss": 1.975, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7237569060773481, | |
| "grad_norm": 2.3781044483184814, | |
| "learning_rate": 6.800888624023553e-05, | |
| "loss": 1.8442, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7292817679558011, | |
| "grad_norm": 1.9898909330368042, | |
| "learning_rate": 6.618669710291606e-05, | |
| "loss": 1.3109, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7348066298342542, | |
| "grad_norm": 2.00718355178833, | |
| "learning_rate": 6.434016163555452e-05, | |
| "loss": 1.0545, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7403314917127072, | |
| "grad_norm": 2.8020102977752686, | |
| "learning_rate": 6.247205720289907e-05, | |
| "loss": 1.9756, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7458563535911602, | |
| "grad_norm": 2.493680953979492, | |
| "learning_rate": 6.058519361147055e-05, | |
| "loss": 0.9292, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7513812154696132, | |
| "grad_norm": 4.338136196136475, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 1.3196, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7569060773480663, | |
| "grad_norm": 54.399234771728516, | |
| "learning_rate": 5.6766564987506566e-05, | |
| "loss": 1.849, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7624309392265194, | |
| "grad_norm": 2.2929675579071045, | |
| "learning_rate": 5.484054353515896e-05, | |
| "loss": 2.1105, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7679558011049724, | |
| "grad_norm": 2.054893970489502, | |
| "learning_rate": 5.290724144552379e-05, | |
| "loss": 2.4684, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7734806629834254, | |
| "grad_norm": 2.0970098972320557, | |
| "learning_rate": 5.096956658859122e-05, | |
| "loss": 2.0751, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7790055248618785, | |
| "grad_norm": 2.0389559268951416, | |
| "learning_rate": 4.903043341140879e-05, | |
| "loss": 2.2336, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7845303867403315, | |
| "grad_norm": 1.9463435411453247, | |
| "learning_rate": 4.709275855447621e-05, | |
| "loss": 2.1493, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7900552486187845, | |
| "grad_norm": 2.0549769401550293, | |
| "learning_rate": 4.515945646484105e-05, | |
| "loss": 1.6908, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.7955801104972375, | |
| "grad_norm": 2.507564067840576, | |
| "learning_rate": 4.323343501249346e-05, | |
| "loss": 0.9681, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8011049723756906, | |
| "grad_norm": 2.2302608489990234, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 1.0328, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8066298342541437, | |
| "grad_norm": 2.4775586128234863, | |
| "learning_rate": 3.941480638852948e-05, | |
| "loss": 1.9766, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.8121546961325967, | |
| "grad_norm": 1.9927563667297363, | |
| "learning_rate": 3.752794279710094e-05, | |
| "loss": 0.6734, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8176795580110497, | |
| "grad_norm": 1.3751696348190308, | |
| "learning_rate": 3.5659838364445505e-05, | |
| "loss": 0.248, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8232044198895028, | |
| "grad_norm": 8.684656143188477, | |
| "learning_rate": 3.381330289708396e-05, | |
| "loss": 2.9323, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8287292817679558, | |
| "grad_norm": 65.27632141113281, | |
| "learning_rate": 3.199111375976449e-05, | |
| "loss": 0.5837, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8342541436464088, | |
| "grad_norm": 1.8862208127975464, | |
| "learning_rate": 3.019601169804216e-05, | |
| "loss": 2.6431, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8397790055248618, | |
| "grad_norm": 2.046152353286743, | |
| "learning_rate": 2.8430696715937337e-05, | |
| "loss": 2.1444, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8453038674033149, | |
| "grad_norm": 2.135226249694824, | |
| "learning_rate": 2.6697824014873075e-05, | |
| "loss": 2.2409, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.850828729281768, | |
| "grad_norm": 2.0305089950561523, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.8083, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.856353591160221, | |
| "grad_norm": 2.2512691020965576, | |
| "learning_rate": 2.333977835991545e-05, | |
| "loss": 1.7166, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.861878453038674, | |
| "grad_norm": 1.9329092502593994, | |
| "learning_rate": 2.171965622567308e-05, | |
| "loss": 1.3238, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8674033149171271, | |
| "grad_norm": 1.8936848640441895, | |
| "learning_rate": 2.0142070414860704e-05, | |
| "loss": 1.2302, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.8729281767955801, | |
| "grad_norm": 2.046297073364258, | |
| "learning_rate": 1.8609393766395085e-05, | |
| "loss": 1.2754, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8784530386740331, | |
| "grad_norm": 2.7109391689300537, | |
| "learning_rate": 1.7123931571546827e-05, | |
| "loss": 1.6343, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8839779005524862, | |
| "grad_norm": 1.6708534955978394, | |
| "learning_rate": 1.5687918106563326e-05, | |
| "loss": 0.4472, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8895027624309392, | |
| "grad_norm": 3.256784439086914, | |
| "learning_rate": 1.4303513272105057e-05, | |
| "loss": 1.0192, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.8950276243093923, | |
| "grad_norm": 54.5963020324707, | |
| "learning_rate": 1.297279934454978e-05, | |
| "loss": 0.8035, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9005524861878453, | |
| "grad_norm": 1.8667521476745605, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 2.2657, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9060773480662984, | |
| "grad_norm": 1.7495869398117065, | |
| "learning_rate": 1.0480366524062042e-05, | |
| "loss": 2.3223, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9116022099447514, | |
| "grad_norm": 1.6687796115875244, | |
| "learning_rate": 9.322396486851626e-06, | |
| "loss": 2.2108, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9171270718232044, | |
| "grad_norm": 1.8657686710357666, | |
| "learning_rate": 8.225609429353187e-06, | |
| "loss": 2.0596, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9226519337016574, | |
| "grad_norm": 1.7651612758636475, | |
| "learning_rate": 7.191655023486682e-06, | |
| "loss": 1.6356, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9281767955801105, | |
| "grad_norm": 1.8384994268417358, | |
| "learning_rate": 6.222088434895462e-06, | |
| "loss": 1.644, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9337016574585635, | |
| "grad_norm": 1.8519909381866455, | |
| "learning_rate": 5.318367983829392e-06, | |
| "loss": 1.4412, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9392265193370166, | |
| "grad_norm": 2.047981023788452, | |
| "learning_rate": 4.4818529516926726e-06, | |
| "loss": 1.2117, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9447513812154696, | |
| "grad_norm": 2.5267603397369385, | |
| "learning_rate": 3.7138015365554833e-06, | |
| "loss": 2.0226, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9502762430939227, | |
| "grad_norm": 2.1057608127593994, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.8825, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9558011049723757, | |
| "grad_norm": 1.898366093635559, | |
| "learning_rate": 2.3876057330792346e-06, | |
| "loss": 0.454, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9613259668508287, | |
| "grad_norm": 8.366283416748047, | |
| "learning_rate": 1.8314560692059835e-06, | |
| "loss": 2.6083, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9668508287292817, | |
| "grad_norm": 62.182796478271484, | |
| "learning_rate": 1.3477564710088098e-06, | |
| "loss": 0.5136, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9723756906077348, | |
| "grad_norm": 1.588686466217041, | |
| "learning_rate": 9.372344686307655e-07, | |
| "loss": 2.2373, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9779005524861878, | |
| "grad_norm": 1.7134623527526855, | |
| "learning_rate": 6.005075261595494e-07, | |
| "loss": 1.9072, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9834254143646409, | |
| "grad_norm": 1.8341338634490967, | |
| "learning_rate": 3.380821129028489e-07, | |
| "loss": 1.6802, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.988950276243094, | |
| "grad_norm": 2.2128942012786865, | |
| "learning_rate": 1.503529416103988e-07, | |
| "loss": 1.7029, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.994475138121547, | |
| "grad_norm": 2.1348469257354736, | |
| "learning_rate": 3.760237478849793e-08, | |
| "loss": 0.6997, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 24.64427375793457, | |
| "learning_rate": 0.0, | |
| "loss": 1.6622, | |
| "step": 181 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 181, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.7596352046025933e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |