| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 91, | |
| "global_step": 362, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0027624309392265192, | |
| "grad_norm": 0.42503952980041504, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.7314, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0027624309392265192, | |
| "eval_loss": 1.3109512329101562, | |
| "eval_runtime": 138.9562, | |
| "eval_samples_per_second": 10.241, | |
| "eval_steps_per_second": 0.324, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0055248618784530384, | |
| "grad_norm": 0.4957458972930908, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.8536, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008287292817679558, | |
| "grad_norm": 0.4254479706287384, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7545, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.011049723756906077, | |
| "grad_norm": 0.4761411249637604, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7814, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013812154696132596, | |
| "grad_norm": 0.47434934973716736, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7703, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.016574585635359115, | |
| "grad_norm": 0.4699034094810486, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.83, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.019337016574585635, | |
| "grad_norm": 0.4335905909538269, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.6648, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.022099447513812154, | |
| "grad_norm": 0.4264945685863495, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.7164, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.024861878453038673, | |
| "grad_norm": 0.417251318693161, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.7058, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.027624309392265192, | |
| "grad_norm": 0.42205414175987244, | |
| "learning_rate": 2e-05, | |
| "loss": 1.71, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03038674033149171, | |
| "grad_norm": 0.414096862077713, | |
| "learning_rate": 1.9999601726381415e-05, | |
| "loss": 1.6954, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03314917127071823, | |
| "grad_norm": 0.412890762090683, | |
| "learning_rate": 1.9998406937250035e-05, | |
| "loss": 1.6941, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03591160220994475, | |
| "grad_norm": 0.4428125321865082, | |
| "learning_rate": 1.9996415727776456e-05, | |
| "loss": 1.6439, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03867403314917127, | |
| "grad_norm": 0.4378657341003418, | |
| "learning_rate": 1.999362825656992e-05, | |
| "loss": 1.7518, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04143646408839779, | |
| "grad_norm": 0.4670448303222656, | |
| "learning_rate": 1.9990044745665672e-05, | |
| "loss": 1.6461, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04419889502762431, | |
| "grad_norm": 0.44707468152046204, | |
| "learning_rate": 1.998566548050729e-05, | |
| "loss": 1.7201, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04696132596685083, | |
| "grad_norm": 0.4317484200000763, | |
| "learning_rate": 1.9980490809923928e-05, | |
| "loss": 1.7697, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.049723756906077346, | |
| "grad_norm": 0.390918105840683, | |
| "learning_rate": 1.9974521146102535e-05, | |
| "loss": 1.6336, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.052486187845303865, | |
| "grad_norm": 0.44283199310302734, | |
| "learning_rate": 1.9967756964555044e-05, | |
| "loss": 1.7058, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.055248618784530384, | |
| "grad_norm": 0.40721943974494934, | |
| "learning_rate": 1.9960198804080462e-05, | |
| "loss": 1.7619, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.058011049723756904, | |
| "grad_norm": 0.3747328817844391, | |
| "learning_rate": 1.995184726672197e-05, | |
| "loss": 1.6998, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06077348066298342, | |
| "grad_norm": 0.3749975860118866, | |
| "learning_rate": 1.9942703017718977e-05, | |
| "loss": 1.7488, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06353591160220995, | |
| "grad_norm": 0.36129963397979736, | |
| "learning_rate": 1.99327667854541e-05, | |
| "loss": 1.6649, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06629834254143646, | |
| "grad_norm": 0.3862815797328949, | |
| "learning_rate": 1.9922039361395186e-05, | |
| "loss": 1.626, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06906077348066299, | |
| "grad_norm": 0.4051111042499542, | |
| "learning_rate": 1.991052160003223e-05, | |
| "loss": 1.6523, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0718232044198895, | |
| "grad_norm": 0.42131727933883667, | |
| "learning_rate": 1.989821441880933e-05, | |
| "loss": 1.8082, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07458563535911603, | |
| "grad_norm": 0.39152222871780396, | |
| "learning_rate": 1.9885118798051607e-05, | |
| "loss": 1.6831, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07734806629834254, | |
| "grad_norm": 0.40526384115219116, | |
| "learning_rate": 1.9871235780887114e-05, | |
| "loss": 1.6613, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08011049723756906, | |
| "grad_norm": 0.3861314654350281, | |
| "learning_rate": 1.9856566473163747e-05, | |
| "loss": 1.6981, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08287292817679558, | |
| "grad_norm": 0.38509440422058105, | |
| "learning_rate": 1.984111204336116e-05, | |
| "loss": 1.6309, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0856353591160221, | |
| "grad_norm": 0.3785051107406616, | |
| "learning_rate": 1.9824873722497694e-05, | |
| "loss": 1.675, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08839779005524862, | |
| "grad_norm": 0.4013555645942688, | |
| "learning_rate": 1.9807852804032306e-05, | |
| "loss": 1.6033, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.09116022099447514, | |
| "grad_norm": 0.36987578868865967, | |
| "learning_rate": 1.9790050643761552e-05, | |
| "loss": 1.643, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09392265193370165, | |
| "grad_norm": 0.34048470854759216, | |
| "learning_rate": 1.9771468659711595e-05, | |
| "loss": 1.7037, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09668508287292818, | |
| "grad_norm": 0.370815247297287, | |
| "learning_rate": 1.975210833202524e-05, | |
| "loss": 1.7189, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09944751381215469, | |
| "grad_norm": 0.3450065553188324, | |
| "learning_rate": 1.9731971202844036e-05, | |
| "loss": 1.6371, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.10220994475138122, | |
| "grad_norm": 0.3642140328884125, | |
| "learning_rate": 1.9711058876185446e-05, | |
| "loss": 1.6962, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10497237569060773, | |
| "grad_norm": 0.3298833668231964, | |
| "learning_rate": 1.9689373017815076e-05, | |
| "loss": 1.5134, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10773480662983426, | |
| "grad_norm": 0.35403338074684143, | |
| "learning_rate": 1.9666915355113976e-05, | |
| "loss": 1.7167, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.11049723756906077, | |
| "grad_norm": 0.36911338567733765, | |
| "learning_rate": 1.964368767694107e-05, | |
| "loss": 1.6674, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1132596685082873, | |
| "grad_norm": 0.38089630007743835, | |
| "learning_rate": 1.9619691833490645e-05, | |
| "loss": 1.5777, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11602209944751381, | |
| "grad_norm": 0.3290576934814453, | |
| "learning_rate": 1.9594929736144978e-05, | |
| "loss": 1.6851, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11878453038674033, | |
| "grad_norm": 0.3029409348964691, | |
| "learning_rate": 1.956940335732209e-05, | |
| "loss": 1.2519, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.12154696132596685, | |
| "grad_norm": 0.3706592321395874, | |
| "learning_rate": 1.954311473031864e-05, | |
| "loss": 1.6061, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.12430939226519337, | |
| "grad_norm": 0.36404719948768616, | |
| "learning_rate": 1.9516065949147945e-05, | |
| "loss": 1.5515, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1270718232044199, | |
| "grad_norm": 0.33177605271339417, | |
| "learning_rate": 1.9488259168373198e-05, | |
| "loss": 1.5692, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1298342541436464, | |
| "grad_norm": 0.3582829535007477, | |
| "learning_rate": 1.9459696602935838e-05, | |
| "loss": 1.5336, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.13259668508287292, | |
| "grad_norm": 0.3565369248390198, | |
| "learning_rate": 1.9430380527979124e-05, | |
| "loss": 1.6251, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13535911602209943, | |
| "grad_norm": 0.3471137285232544, | |
| "learning_rate": 1.94003132786669e-05, | |
| "loss": 1.6245, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13812154696132597, | |
| "grad_norm": 0.4011210501194, | |
| "learning_rate": 1.936949724999762e-05, | |
| "loss": 1.3687, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1408839779005525, | |
| "grad_norm": 0.36058053374290466, | |
| "learning_rate": 1.9337934896613516e-05, | |
| "loss": 1.5283, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.143646408839779, | |
| "grad_norm": 0.39054054021835327, | |
| "learning_rate": 1.930562873260514e-05, | |
| "loss": 1.5784, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1464088397790055, | |
| "grad_norm": 0.39804011583328247, | |
| "learning_rate": 1.927258133131105e-05, | |
| "loss": 1.605, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14917127071823205, | |
| "grad_norm": 0.42572247982025146, | |
| "learning_rate": 1.9238795325112867e-05, | |
| "loss": 1.4554, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.15193370165745856, | |
| "grad_norm": 0.4022100269794464, | |
| "learning_rate": 1.9204273405225588e-05, | |
| "loss": 1.5905, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.15469613259668508, | |
| "grad_norm": 0.43215224146842957, | |
| "learning_rate": 1.9169018321483198e-05, | |
| "loss": 1.6233, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1574585635359116, | |
| "grad_norm": 0.38549479842185974, | |
| "learning_rate": 1.9133032882119656e-05, | |
| "loss": 1.5868, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.16022099447513813, | |
| "grad_norm": 0.42155200242996216, | |
| "learning_rate": 1.9096319953545186e-05, | |
| "loss": 1.4519, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.16298342541436464, | |
| "grad_norm": 0.4292687177658081, | |
| "learning_rate": 1.9058882460117972e-05, | |
| "loss": 1.6053, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.16574585635359115, | |
| "grad_norm": 0.44173118472099304, | |
| "learning_rate": 1.9020723383911214e-05, | |
| "loss": 1.6328, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1685082872928177, | |
| "grad_norm": 0.4508700966835022, | |
| "learning_rate": 1.8981845764475585e-05, | |
| "loss": 1.6, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1712707182320442, | |
| "grad_norm": 0.42624732851982117, | |
| "learning_rate": 1.8942252698597113e-05, | |
| "loss": 1.6086, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.17403314917127072, | |
| "grad_norm": 0.4029814600944519, | |
| "learning_rate": 1.890194734005053e-05, | |
| "loss": 1.5955, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.17679558011049723, | |
| "grad_norm": 0.4393555223941803, | |
| "learning_rate": 1.8860932899348028e-05, | |
| "loss": 1.5735, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.17955801104972377, | |
| "grad_norm": 0.40672072768211365, | |
| "learning_rate": 1.881921264348355e-05, | |
| "loss": 1.5975, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.18232044198895028, | |
| "grad_norm": 0.4309908151626587, | |
| "learning_rate": 1.8776789895672557e-05, | |
| "loss": 1.5539, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1850828729281768, | |
| "grad_norm": 0.41728851199150085, | |
| "learning_rate": 1.8733668035087302e-05, | |
| "loss": 1.5678, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1878453038674033, | |
| "grad_norm": 0.46532243490219116, | |
| "learning_rate": 1.8689850496587674e-05, | |
| "loss": 1.5277, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.19060773480662985, | |
| "grad_norm": 0.43453872203826904, | |
| "learning_rate": 1.8645340770447595e-05, | |
| "loss": 1.5192, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.19337016574585636, | |
| "grad_norm": 0.46961575746536255, | |
| "learning_rate": 1.8600142402077006e-05, | |
| "loss": 1.4962, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19613259668508287, | |
| "grad_norm": 0.47958558797836304, | |
| "learning_rate": 1.8554258991739454e-05, | |
| "loss": 1.502, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.19889502762430938, | |
| "grad_norm": 0.4123697280883789, | |
| "learning_rate": 1.850769419426531e-05, | |
| "loss": 1.4997, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.20165745856353592, | |
| "grad_norm": 0.5050204396247864, | |
| "learning_rate": 1.8460451718760653e-05, | |
| "loss": 1.5283, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.20441988950276244, | |
| "grad_norm": 0.44690632820129395, | |
| "learning_rate": 1.8412535328311813e-05, | |
| "loss": 1.5429, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.20718232044198895, | |
| "grad_norm": 0.45188507437705994, | |
| "learning_rate": 1.8363948839685638e-05, | |
| "loss": 1.532, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.20994475138121546, | |
| "grad_norm": 0.4514811038970947, | |
| "learning_rate": 1.8314696123025456e-05, | |
| "loss": 1.4147, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.212707182320442, | |
| "grad_norm": 0.5114113688468933, | |
| "learning_rate": 1.8264781101542797e-05, | |
| "loss": 1.5552, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2154696132596685, | |
| "grad_norm": 0.51650071144104, | |
| "learning_rate": 1.8214207751204917e-05, | |
| "loss": 1.5554, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.21823204419889503, | |
| "grad_norm": 0.48588883876800537, | |
| "learning_rate": 1.816298010041806e-05, | |
| "loss": 1.5688, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.22099447513812154, | |
| "grad_norm": 0.4869782328605652, | |
| "learning_rate": 1.8111102229706593e-05, | |
| "loss": 1.6062, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.22375690607734808, | |
| "grad_norm": 0.4736042320728302, | |
| "learning_rate": 1.805857827138798e-05, | |
| "loss": 1.4834, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2265193370165746, | |
| "grad_norm": 0.5085079073905945, | |
| "learning_rate": 1.8005412409243604e-05, | |
| "loss": 1.4549, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2292817679558011, | |
| "grad_norm": 0.5268322229385376, | |
| "learning_rate": 1.7951608878185533e-05, | |
| "loss": 1.4762, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.23204419889502761, | |
| "grad_norm": 0.5247999429702759, | |
| "learning_rate": 1.789717196391916e-05, | |
| "loss": 1.4625, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.23480662983425415, | |
| "grad_norm": 0.5090611577033997, | |
| "learning_rate": 1.7842106002601854e-05, | |
| "loss": 1.4634, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.23756906077348067, | |
| "grad_norm": 0.500877857208252, | |
| "learning_rate": 1.778641538049755e-05, | |
| "loss": 1.5259, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.24033149171270718, | |
| "grad_norm": 0.5528935790061951, | |
| "learning_rate": 1.773010453362737e-05, | |
| "loss": 1.4783, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2430939226519337, | |
| "grad_norm": 0.4904051125049591, | |
| "learning_rate": 1.7673177947416258e-05, | |
| "loss": 1.5713, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.24585635359116023, | |
| "grad_norm": 0.4856030344963074, | |
| "learning_rate": 1.7615640156335713e-05, | |
| "loss": 1.5448, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.24861878453038674, | |
| "grad_norm": 0.4999338388442993, | |
| "learning_rate": 1.7557495743542586e-05, | |
| "loss": 1.5286, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2513812154696133, | |
| "grad_norm": 0.6124939322471619, | |
| "learning_rate": 1.749874934051401e-05, | |
| "loss": 1.5815, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2513812154696133, | |
| "eval_loss": 1.121329426765442, | |
| "eval_runtime": 153.7165, | |
| "eval_samples_per_second": 9.257, | |
| "eval_steps_per_second": 0.293, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2541436464088398, | |
| "grad_norm": 0.5473253726959229, | |
| "learning_rate": 1.7439405626678496e-05, | |
| "loss": 1.5358, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2569060773480663, | |
| "grad_norm": 0.515661895275116, | |
| "learning_rate": 1.7379469329043166e-05, | |
| "loss": 1.5705, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2596685082872928, | |
| "grad_norm": 0.5037718415260315, | |
| "learning_rate": 1.7318945221817255e-05, | |
| "loss": 1.5362, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.26243093922651933, | |
| "grad_norm": 0.5043213367462158, | |
| "learning_rate": 1.7257838126031797e-05, | |
| "loss": 1.5082, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.26519337016574585, | |
| "grad_norm": 0.5211421847343445, | |
| "learning_rate": 1.719615290915563e-05, | |
| "loss": 1.5486, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.26795580110497236, | |
| "grad_norm": 0.5876896381378174, | |
| "learning_rate": 1.7133894484707657e-05, | |
| "loss": 1.4926, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.27071823204419887, | |
| "grad_norm": 0.5891074538230896, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 1.5895, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.27348066298342544, | |
| "grad_norm": 0.539527952671051, | |
| "learning_rate": 1.7007677895070358e-05, | |
| "loss": 1.4588, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.27624309392265195, | |
| "grad_norm": 0.4934506416320801, | |
| "learning_rate": 1.694372978362861e-05, | |
| "loss": 1.5116, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.27900552486187846, | |
| "grad_norm": 0.5579091906547546, | |
| "learning_rate": 1.6879228571309377e-05, | |
| "loss": 1.4786, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.281767955801105, | |
| "grad_norm": 0.5706738233566284, | |
| "learning_rate": 1.6814179395938915e-05, | |
| "loss": 1.549, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2845303867403315, | |
| "grad_norm": 0.5942355394363403, | |
| "learning_rate": 1.6748587438991303e-05, | |
| "loss": 1.4979, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.287292817679558, | |
| "grad_norm": 0.5835041403770447, | |
| "learning_rate": 1.6682457925175762e-05, | |
| "loss": 1.4915, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2900552486187845, | |
| "grad_norm": 0.6035829782485962, | |
| "learning_rate": 1.6615796122020443e-05, | |
| "loss": 1.5754, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.292817679558011, | |
| "grad_norm": 0.5598711967468262, | |
| "learning_rate": 1.6548607339452853e-05, | |
| "loss": 1.5017, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2955801104972376, | |
| "grad_norm": 0.5402417778968811, | |
| "learning_rate": 1.6480896929376905e-05, | |
| "loss": 1.5558, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2983425414364641, | |
| "grad_norm": 0.6131693720817566, | |
| "learning_rate": 1.641267028524661e-05, | |
| "loss": 1.532, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3011049723756906, | |
| "grad_norm": 0.5868387222290039, | |
| "learning_rate": 1.6343932841636455e-05, | |
| "loss": 1.4656, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.30386740331491713, | |
| "grad_norm": 0.5688586831092834, | |
| "learning_rate": 1.627469007380852e-05, | |
| "loss": 1.5049, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.30662983425414364, | |
| "grad_norm": 0.5797792673110962, | |
| "learning_rate": 1.6204947497276346e-05, | |
| "loss": 1.5484, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.30939226519337015, | |
| "grad_norm": 0.5769445300102234, | |
| "learning_rate": 1.6134710667365598e-05, | |
| "loss": 1.4402, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.31215469613259667, | |
| "grad_norm": 0.6333233118057251, | |
| "learning_rate": 1.6063985178771555e-05, | |
| "loss": 1.4678, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3149171270718232, | |
| "grad_norm": 0.558728814125061, | |
| "learning_rate": 1.599277666511347e-05, | |
| "loss": 1.4787, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.31767955801104975, | |
| "grad_norm": 0.5599526166915894, | |
| "learning_rate": 1.592109079848583e-05, | |
| "loss": 1.5417, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.32044198895027626, | |
| "grad_norm": 0.5926061272621155, | |
| "learning_rate": 1.584893328900653e-05, | |
| "loss": 1.3919, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.32320441988950277, | |
| "grad_norm": 0.589717447757721, | |
| "learning_rate": 1.577630988436206e-05, | |
| "loss": 1.5362, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3259668508287293, | |
| "grad_norm": 0.6871376037597656, | |
| "learning_rate": 1.5703226369349642e-05, | |
| "loss": 1.4358, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3287292817679558, | |
| "grad_norm": 0.607738733291626, | |
| "learning_rate": 1.562968856541648e-05, | |
| "loss": 1.5095, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3314917127071823, | |
| "grad_norm": 0.635498583316803, | |
| "learning_rate": 1.5555702330196024e-05, | |
| "loss": 1.4874, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3342541436464088, | |
| "grad_norm": 0.6137527227401733, | |
| "learning_rate": 1.5481273557041402e-05, | |
| "loss": 1.4166, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3370165745856354, | |
| "grad_norm": 0.6075506210327148, | |
| "learning_rate": 1.5406408174555978e-05, | |
| "loss": 1.5638, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3397790055248619, | |
| "grad_norm": 0.6399998068809509, | |
| "learning_rate": 1.5331112146121104e-05, | |
| "loss": 1.503, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3425414364640884, | |
| "grad_norm": 0.5782871246337891, | |
| "learning_rate": 1.525539146942113e-05, | |
| "loss": 1.481, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3453038674033149, | |
| "grad_norm": 0.6390048265457153, | |
| "learning_rate": 1.5179252175965632e-05, | |
| "loss": 1.4298, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.34806629834254144, | |
| "grad_norm": 0.5457322001457214, | |
| "learning_rate": 1.5102700330609e-05, | |
| "loss": 1.4868, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.35082872928176795, | |
| "grad_norm": 0.5977615118026733, | |
| "learning_rate": 1.5025742031067316e-05, | |
| "loss": 1.4753, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.35359116022099446, | |
| "grad_norm": 0.6722098588943481, | |
| "learning_rate": 1.4948383407432678e-05, | |
| "loss": 1.5022, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.356353591160221, | |
| "grad_norm": 0.6676556468009949, | |
| "learning_rate": 1.4870630621684873e-05, | |
| "loss": 1.4862, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.35911602209944754, | |
| "grad_norm": 0.6139523386955261, | |
| "learning_rate": 1.479248986720057e-05, | |
| "loss": 1.4543, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36187845303867405, | |
| "grad_norm": 0.621616542339325, | |
| "learning_rate": 1.4713967368259981e-05, | |
| "loss": 1.4795, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.36464088397790057, | |
| "grad_norm": 0.6133718490600586, | |
| "learning_rate": 1.4635069379551054e-05, | |
| "loss": 1.4821, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3674033149171271, | |
| "grad_norm": 0.7033741474151611, | |
| "learning_rate": 1.4555802185671297e-05, | |
| "loss": 1.5079, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3701657458563536, | |
| "grad_norm": 0.6731663942337036, | |
| "learning_rate": 1.4476172100627127e-05, | |
| "loss": 1.4438, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3729281767955801, | |
| "grad_norm": 0.6156182885169983, | |
| "learning_rate": 1.4396185467330974e-05, | |
| "loss": 1.5067, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3756906077348066, | |
| "grad_norm": 0.6311376690864563, | |
| "learning_rate": 1.4315848657096006e-05, | |
| "loss": 1.4958, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3784530386740331, | |
| "grad_norm": 0.6299065947532654, | |
| "learning_rate": 1.4235168069128657e-05, | |
| "loss": 1.4514, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3812154696132597, | |
| "grad_norm": 0.7021288275718689, | |
| "learning_rate": 1.4154150130018867e-05, | |
| "loss": 1.4633, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3839779005524862, | |
| "grad_norm": 0.6808854937553406, | |
| "learning_rate": 1.407280129322819e-05, | |
| "loss": 1.5116, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3867403314917127, | |
| "grad_norm": 0.6397327780723572, | |
| "learning_rate": 1.3991128038575741e-05, | |
| "loss": 1.4773, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.38950276243093923, | |
| "grad_norm": 0.6835840344429016, | |
| "learning_rate": 1.3909136871722066e-05, | |
| "loss": 1.4515, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.39226519337016574, | |
| "grad_norm": 0.6662083268165588, | |
| "learning_rate": 1.3826834323650899e-05, | |
| "loss": 1.4796, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.39502762430939226, | |
| "grad_norm": 0.6650438904762268, | |
| "learning_rate": 1.374422695014897e-05, | |
| "loss": 1.5343, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.39779005524861877, | |
| "grad_norm": 0.6556645035743713, | |
| "learning_rate": 1.3661321331283796e-05, | |
| "loss": 1.5149, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4005524861878453, | |
| "grad_norm": 0.7009831666946411, | |
| "learning_rate": 1.3578124070879534e-05, | |
| "loss": 1.3801, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.40331491712707185, | |
| "grad_norm": 0.6953743696212769, | |
| "learning_rate": 1.3494641795990986e-05, | |
| "loss": 1.3673, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.40607734806629836, | |
| "grad_norm": 0.7101163268089294, | |
| "learning_rate": 1.3410881156375684e-05, | |
| "loss": 1.4491, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4088397790055249, | |
| "grad_norm": 0.6939963102340698, | |
| "learning_rate": 1.3326848823964243e-05, | |
| "loss": 1.4347, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.4116022099447514, | |
| "grad_norm": 0.7195811867713928, | |
| "learning_rate": 1.3242551492328875e-05, | |
| "loss": 1.4454, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.4143646408839779, | |
| "grad_norm": 0.7362983226776123, | |
| "learning_rate": 1.3157995876150252e-05, | |
| "loss": 1.4402, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4171270718232044, | |
| "grad_norm": 0.706754744052887, | |
| "learning_rate": 1.3073188710682612e-05, | |
| "loss": 1.2887, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.4198895027624309, | |
| "grad_norm": 0.7431464195251465, | |
| "learning_rate": 1.2988136751217292e-05, | |
| "loss": 1.4228, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.42265193370165743, | |
| "grad_norm": 0.7429276704788208, | |
| "learning_rate": 1.2902846772544625e-05, | |
| "loss": 1.5767, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.425414364640884, | |
| "grad_norm": 0.672516405582428, | |
| "learning_rate": 1.2817325568414299e-05, | |
| "loss": 1.4013, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.4281767955801105, | |
| "grad_norm": 0.7116546630859375, | |
| "learning_rate": 1.27315799509942e-05, | |
| "loss": 1.4784, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.430939226519337, | |
| "grad_norm": 0.7850514054298401, | |
| "learning_rate": 1.2645616750327792e-05, | |
| "loss": 1.4335, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.43370165745856354, | |
| "grad_norm": 0.7218865752220154, | |
| "learning_rate": 1.2559442813790077e-05, | |
| "loss": 1.4628, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.43646408839779005, | |
| "grad_norm": 0.746462881565094, | |
| "learning_rate": 1.2473065005542155e-05, | |
| "loss": 1.4531, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.43922651933701656, | |
| "grad_norm": 0.7100480794906616, | |
| "learning_rate": 1.2386490205984488e-05, | |
| "loss": 1.4729, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.4419889502762431, | |
| "grad_norm": 0.7361482977867126, | |
| "learning_rate": 1.2299725311208807e-05, | |
| "loss": 1.5175, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4447513812154696, | |
| "grad_norm": 0.7085789442062378, | |
| "learning_rate": 1.2212777232448837e-05, | |
| "loss": 1.4351, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.44751381215469616, | |
| "grad_norm": 0.6986429691314697, | |
| "learning_rate": 1.2125652895529766e-05, | |
| "loss": 1.484, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.45027624309392267, | |
| "grad_norm": 0.656014084815979, | |
| "learning_rate": 1.2038359240316589e-05, | |
| "loss": 1.4858, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4530386740331492, | |
| "grad_norm": 0.7168847322463989, | |
| "learning_rate": 1.1950903220161286e-05, | |
| "loss": 1.4222, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4558011049723757, | |
| "grad_norm": 0.6796424388885498, | |
| "learning_rate": 1.186329180134898e-05, | |
| "loss": 1.5088, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4585635359116022, | |
| "grad_norm": 0.6602550148963928, | |
| "learning_rate": 1.1775531962543036e-05, | |
| "loss": 1.2236, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4613259668508287, | |
| "grad_norm": 0.7656331062316895, | |
| "learning_rate": 1.1687630694229159e-05, | |
| "loss": 1.4906, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.46408839779005523, | |
| "grad_norm": 0.7794011235237122, | |
| "learning_rate": 1.1599594998158602e-05, | |
| "loss": 1.4335, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.46685082872928174, | |
| "grad_norm": 0.7844555377960205, | |
| "learning_rate": 1.1511431886790407e-05, | |
| "loss": 1.3969, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4696132596685083, | |
| "grad_norm": 0.6949150562286377, | |
| "learning_rate": 1.1423148382732854e-05, | |
| "loss": 1.4488, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4723756906077348, | |
| "grad_norm": 0.866104781627655, | |
| "learning_rate": 1.1334751518184062e-05, | |
| "loss": 1.3399, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.47513812154696133, | |
| "grad_norm": 0.763599157333374, | |
| "learning_rate": 1.124624833437186e-05, | |
| "loss": 1.5064, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.47790055248618785, | |
| "grad_norm": 0.7990726232528687, | |
| "learning_rate": 1.1157645880992901e-05, | |
| "loss": 1.4328, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.48066298342541436, | |
| "grad_norm": 0.782218873500824, | |
| "learning_rate": 1.1068951215651132e-05, | |
| "loss": 1.3896, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.48342541436464087, | |
| "grad_norm": 0.741022527217865, | |
| "learning_rate": 1.098017140329561e-05, | |
| "loss": 1.3803, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4861878453038674, | |
| "grad_norm": 0.8042486310005188, | |
| "learning_rate": 1.089131351565776e-05, | |
| "loss": 1.4333, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4889502762430939, | |
| "grad_norm": 0.7474654316902161, | |
| "learning_rate": 1.080238463068808e-05, | |
| "loss": 1.4138, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.49171270718232046, | |
| "grad_norm": 0.7193496227264404, | |
| "learning_rate": 1.0713391831992324e-05, | |
| "loss": 1.4721, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.494475138121547, | |
| "grad_norm": 0.7830453515052795, | |
| "learning_rate": 1.0624342208267293e-05, | |
| "loss": 1.475, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4972375690607735, | |
| "grad_norm": 0.6899390816688538, | |
| "learning_rate": 1.0535242852736152e-05, | |
| "loss": 1.3806, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.779139518737793, | |
| "learning_rate": 1.0446100862583459e-05, | |
| "loss": 1.4362, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5027624309392266, | |
| "grad_norm": 0.7901574969291687, | |
| "learning_rate": 1.0356923338389807e-05, | |
| "loss": 1.5008, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.5027624309392266, | |
| "eval_loss": 1.09184730052948, | |
| "eval_runtime": 154.1688, | |
| "eval_samples_per_second": 9.23, | |
| "eval_steps_per_second": 0.292, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.505524861878453, | |
| "grad_norm": 0.7908266186714172, | |
| "learning_rate": 1.0267717383566247e-05, | |
| "loss": 1.4892, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5082872928176796, | |
| "grad_norm": 0.7942709922790527, | |
| "learning_rate": 1.0178490103788462e-05, | |
| "loss": 1.391, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.511049723756906, | |
| "grad_norm": 0.76580411195755, | |
| "learning_rate": 1.0089248606430775e-05, | |
| "loss": 1.382, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5138121546961326, | |
| "grad_norm": 0.7731238603591919, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4122, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.5165745856353591, | |
| "grad_norm": 0.8293752074241638, | |
| "learning_rate": 9.910751393569228e-06, | |
| "loss": 1.5193, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5193370165745856, | |
| "grad_norm": 0.9073200225830078, | |
| "learning_rate": 9.82150989621154e-06, | |
| "loss": 1.3044, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5220994475138122, | |
| "grad_norm": 0.7942474484443665, | |
| "learning_rate": 9.732282616433756e-06, | |
| "loss": 1.4997, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5248618784530387, | |
| "grad_norm": 0.7780297994613647, | |
| "learning_rate": 9.643076661610197e-06, | |
| "loss": 1.4907, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5276243093922652, | |
| "grad_norm": 0.8634337782859802, | |
| "learning_rate": 9.553899137416546e-06, | |
| "loss": 1.3418, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5303867403314917, | |
| "grad_norm": 0.7954802513122559, | |
| "learning_rate": 9.464757147263849e-06, | |
| "loss": 1.4508, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5331491712707183, | |
| "grad_norm": 0.8032029271125793, | |
| "learning_rate": 9.37565779173271e-06, | |
| "loss": 1.4302, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5359116022099447, | |
| "grad_norm": 0.7397053837776184, | |
| "learning_rate": 9.286608168007678e-06, | |
| "loss": 1.5004, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5386740331491713, | |
| "grad_norm": 0.9212160706520081, | |
| "learning_rate": 9.197615369311926e-06, | |
| "loss": 1.4778, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5414364640883977, | |
| "grad_norm": 0.779822587966919, | |
| "learning_rate": 9.108686484342241e-06, | |
| "loss": 1.4482, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5441988950276243, | |
| "grad_norm": 0.7739251852035522, | |
| "learning_rate": 9.019828596704394e-06, | |
| "loss": 1.4599, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5469613259668509, | |
| "grad_norm": 0.7287247776985168, | |
| "learning_rate": 8.931048784348875e-06, | |
| "loss": 1.4667, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5497237569060773, | |
| "grad_norm": 0.758631706237793, | |
| "learning_rate": 8.8423541190071e-06, | |
| "loss": 1.3432, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5524861878453039, | |
| "grad_norm": 0.7681453824043274, | |
| "learning_rate": 8.753751665628141e-06, | |
| "loss": 1.5143, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5552486187845304, | |
| "grad_norm": 0.8416883945465088, | |
| "learning_rate": 8.665248481815941e-06, | |
| "loss": 1.362, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5580110497237569, | |
| "grad_norm": 0.7365747094154358, | |
| "learning_rate": 8.576851617267151e-06, | |
| "loss": 1.4192, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5607734806629834, | |
| "grad_norm": 0.8545944690704346, | |
| "learning_rate": 8.488568113209593e-06, | |
| "loss": 1.4204, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.56353591160221, | |
| "grad_norm": 0.8342128396034241, | |
| "learning_rate": 8.4004050018414e-06, | |
| "loss": 1.3981, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5662983425414365, | |
| "grad_norm": 0.7602656483650208, | |
| "learning_rate": 8.312369305770843e-06, | |
| "loss": 1.437, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.569060773480663, | |
| "grad_norm": 0.8728726506233215, | |
| "learning_rate": 8.224468037456969e-06, | |
| "loss": 1.332, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5718232044198895, | |
| "grad_norm": 0.845259428024292, | |
| "learning_rate": 8.136708198651022e-06, | |
| "loss": 1.4522, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.574585635359116, | |
| "grad_norm": 0.8054398894309998, | |
| "learning_rate": 8.04909677983872e-06, | |
| "loss": 1.3441, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5773480662983426, | |
| "grad_norm": 0.7871323227882385, | |
| "learning_rate": 7.961640759683416e-06, | |
| "loss": 1.4925, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.580110497237569, | |
| "grad_norm": 0.8431714177131653, | |
| "learning_rate": 7.874347104470234e-06, | |
| "loss": 1.3429, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5828729281767956, | |
| "grad_norm": 0.8485934138298035, | |
| "learning_rate": 7.787222767551164e-06, | |
| "loss": 1.4277, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.585635359116022, | |
| "grad_norm": 0.8744496703147888, | |
| "learning_rate": 7.700274688791196e-06, | |
| "loss": 1.436, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5883977900552486, | |
| "grad_norm": 0.8178616762161255, | |
| "learning_rate": 7.613509794015517e-06, | |
| "loss": 1.4106, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5911602209944752, | |
| "grad_norm": 0.8785072565078735, | |
| "learning_rate": 7.5269349944578454e-06, | |
| "loss": 1.461, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5939226519337016, | |
| "grad_norm": 0.7647153735160828, | |
| "learning_rate": 7.440557186209927e-06, | |
| "loss": 1.3674, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5966850828729282, | |
| "grad_norm": 0.7870045900344849, | |
| "learning_rate": 7.354383249672212e-06, | |
| "loss": 1.4552, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5994475138121547, | |
| "grad_norm": 0.791970431804657, | |
| "learning_rate": 7.268420049005806e-06, | |
| "loss": 1.4316, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.6022099447513812, | |
| "grad_norm": 0.8452421426773071, | |
| "learning_rate": 7.182674431585703e-06, | |
| "loss": 1.4539, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.6049723756906077, | |
| "grad_norm": 0.9066639542579651, | |
| "learning_rate": 7.097153227455379e-06, | |
| "loss": 1.4379, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.6077348066298343, | |
| "grad_norm": 0.8649947643280029, | |
| "learning_rate": 7.011863248782711e-06, | |
| "loss": 1.3483, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6104972375690608, | |
| "grad_norm": 0.8475552201271057, | |
| "learning_rate": 6.92681128931739e-06, | |
| "loss": 1.4798, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.6132596685082873, | |
| "grad_norm": 0.8348631262779236, | |
| "learning_rate": 6.8420041238497525e-06, | |
| "loss": 1.4084, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.6160220994475138, | |
| "grad_norm": 0.9010429382324219, | |
| "learning_rate": 6.7574485076711285e-06, | |
| "loss": 1.4011, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.6187845303867403, | |
| "grad_norm": 0.8535116910934448, | |
| "learning_rate": 6.673151176035762e-06, | |
| "loss": 1.3661, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.6215469613259669, | |
| "grad_norm": 0.8269426822662354, | |
| "learning_rate": 6.589118843624316e-06, | |
| "loss": 1.3894, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6243093922651933, | |
| "grad_norm": 0.76876300573349, | |
| "learning_rate": 6.505358204009018e-06, | |
| "loss": 1.4496, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6270718232044199, | |
| "grad_norm": 0.7975645065307617, | |
| "learning_rate": 6.421875929120469e-06, | |
| "loss": 1.3987, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6298342541436464, | |
| "grad_norm": 0.8084754347801208, | |
| "learning_rate": 6.33867866871621e-06, | |
| "loss": 1.4141, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6325966850828729, | |
| "grad_norm": 0.7989472150802612, | |
| "learning_rate": 6.25577304985103e-06, | |
| "loss": 1.2641, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.6353591160220995, | |
| "grad_norm": 0.8317887783050537, | |
| "learning_rate": 6.173165676349103e-06, | |
| "loss": 1.3783, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.638121546961326, | |
| "grad_norm": 0.883558452129364, | |
| "learning_rate": 6.090863128277938e-06, | |
| "loss": 1.3727, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6408839779005525, | |
| "grad_norm": 0.7857452630996704, | |
| "learning_rate": 6.008871961424259e-06, | |
| "loss": 1.4159, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.643646408839779, | |
| "grad_norm": 0.7758345007896423, | |
| "learning_rate": 5.927198706771813e-06, | |
| "loss": 1.4271, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.6464088397790055, | |
| "grad_norm": 0.8178966045379639, | |
| "learning_rate": 5.845849869981137e-06, | |
| "loss": 1.4526, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.649171270718232, | |
| "grad_norm": 0.8668413758277893, | |
| "learning_rate": 5.764831930871346e-06, | |
| "loss": 1.4209, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6519337016574586, | |
| "grad_norm": 0.866632342338562, | |
| "learning_rate": 5.684151342903992e-06, | |
| "loss": 1.4785, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6546961325966851, | |
| "grad_norm": 0.8386754393577576, | |
| "learning_rate": 5.603814532669032e-06, | |
| "loss": 1.4133, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6574585635359116, | |
| "grad_norm": 0.8321985006332397, | |
| "learning_rate": 5.523827899372876e-06, | |
| "loss": 1.4049, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6602209944751382, | |
| "grad_norm": 0.8785558342933655, | |
| "learning_rate": 5.444197814328707e-06, | |
| "loss": 1.3697, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.6629834254143646, | |
| "grad_norm": 0.7912325263023376, | |
| "learning_rate": 5.364930620448946e-06, | |
| "loss": 1.4928, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6657458563535912, | |
| "grad_norm": 0.8722138404846191, | |
| "learning_rate": 5.286032631740023e-06, | |
| "loss": 1.4547, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6685082872928176, | |
| "grad_norm": 0.8593927621841431, | |
| "learning_rate": 5.207510132799436e-06, | |
| "loss": 1.4188, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6712707182320442, | |
| "grad_norm": 0.7740910649299622, | |
| "learning_rate": 5.129369378315128e-06, | |
| "loss": 1.4642, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6740331491712708, | |
| "grad_norm": 0.8590474724769592, | |
| "learning_rate": 5.051616592567323e-06, | |
| "loss": 1.2543, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.6767955801104972, | |
| "grad_norm": 0.862617015838623, | |
| "learning_rate": 4.974257968932687e-06, | |
| "loss": 1.4393, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6795580110497238, | |
| "grad_norm": 0.8476159572601318, | |
| "learning_rate": 4.897299669391006e-06, | |
| "loss": 1.4986, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6823204419889503, | |
| "grad_norm": 0.8105822205543518, | |
| "learning_rate": 4.820747824034369e-06, | |
| "loss": 1.5027, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6850828729281768, | |
| "grad_norm": 0.9459635615348816, | |
| "learning_rate": 4.744608530578872e-06, | |
| "loss": 1.3336, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6878453038674033, | |
| "grad_norm": 0.8690259456634521, | |
| "learning_rate": 4.668887853878896e-06, | |
| "loss": 1.3479, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6906077348066298, | |
| "grad_norm": 0.8513994812965393, | |
| "learning_rate": 4.593591825444028e-06, | |
| "loss": 1.5261, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6933701657458563, | |
| "grad_norm": 0.8456645011901855, | |
| "learning_rate": 4.518726442958599e-06, | |
| "loss": 1.4669, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6961325966850829, | |
| "grad_norm": 0.8776260018348694, | |
| "learning_rate": 4.444297669803981e-06, | |
| "loss": 1.4177, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6988950276243094, | |
| "grad_norm": 0.860339343547821, | |
| "learning_rate": 4.370311434583525e-06, | |
| "loss": 1.454, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.7016574585635359, | |
| "grad_norm": 0.902574360370636, | |
| "learning_rate": 4.296773630650358e-06, | |
| "loss": 1.3987, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.7044198895027625, | |
| "grad_norm": 0.8529084920883179, | |
| "learning_rate": 4.223690115637944e-06, | |
| "loss": 1.4428, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7071823204419889, | |
| "grad_norm": 0.8335188627243042, | |
| "learning_rate": 4.15106671099347e-06, | |
| "loss": 1.4812, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.7099447513812155, | |
| "grad_norm": 0.8945775628089905, | |
| "learning_rate": 4.078909201514172e-06, | |
| "loss": 1.3559, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.712707182320442, | |
| "grad_norm": 0.9545679688453674, | |
| "learning_rate": 4.007223334886531e-06, | |
| "loss": 1.3806, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.7154696132596685, | |
| "grad_norm": 0.7985166311264038, | |
| "learning_rate": 3.936014821228448e-06, | |
| "loss": 1.3433, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.7182320441988951, | |
| "grad_norm": 0.8410876393318176, | |
| "learning_rate": 3.865289332634407e-06, | |
| "loss": 1.4393, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7209944751381215, | |
| "grad_norm": 0.8620865941047668, | |
| "learning_rate": 3.7950525027236585e-06, | |
| "loss": 1.2242, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.7237569060773481, | |
| "grad_norm": 0.8445248603820801, | |
| "learning_rate": 3.7253099261914794e-06, | |
| "loss": 1.403, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.7265193370165746, | |
| "grad_norm": 0.8497399687767029, | |
| "learning_rate": 3.6560671583635467e-06, | |
| "loss": 1.4453, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7292817679558011, | |
| "grad_norm": 0.83712238073349, | |
| "learning_rate": 3.5873297147533913e-06, | |
| "loss": 1.503, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7320441988950276, | |
| "grad_norm": 0.8133417963981628, | |
| "learning_rate": 3.5191030706230967e-06, | |
| "loss": 1.3821, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7348066298342542, | |
| "grad_norm": 0.8870840668678284, | |
| "learning_rate": 3.4513926605471504e-06, | |
| "loss": 1.4533, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7375690607734806, | |
| "grad_norm": 0.9070349335670471, | |
| "learning_rate": 3.3842038779795594e-06, | |
| "loss": 1.4055, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.7403314917127072, | |
| "grad_norm": 0.8012280464172363, | |
| "learning_rate": 3.3175420748242405e-06, | |
| "loss": 1.2719, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.7430939226519337, | |
| "grad_norm": 0.8506048321723938, | |
| "learning_rate": 3.2514125610086957e-06, | |
| "loss": 1.3588, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.7458563535911602, | |
| "grad_norm": 0.7189351916313171, | |
| "learning_rate": 3.1858206040610883e-06, | |
| "loss": 1.2039, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7486187845303868, | |
| "grad_norm": 0.8134811520576477, | |
| "learning_rate": 3.1207714286906253e-06, | |
| "loss": 1.2114, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7513812154696132, | |
| "grad_norm": 0.8526425957679749, | |
| "learning_rate": 3.0562702163713954e-06, | |
| "loss": 1.5044, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7541436464088398, | |
| "grad_norm": 0.9256671071052551, | |
| "learning_rate": 2.9923221049296448e-06, | |
| "loss": 1.3762, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.7541436464088398, | |
| "eval_loss": 1.0853379964828491, | |
| "eval_runtime": 155.2763, | |
| "eval_samples_per_second": 9.164, | |
| "eval_steps_per_second": 0.29, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.7569060773480663, | |
| "grad_norm": 1.0012723207473755, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 1.4391, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7596685082872928, | |
| "grad_norm": 0.9301908016204834, | |
| "learning_rate": 2.8661055152923456e-06, | |
| "loss": 1.3913, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7624309392265194, | |
| "grad_norm": 0.9066746830940247, | |
| "learning_rate": 2.8038470908443717e-06, | |
| "loss": 1.3527, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7651933701657458, | |
| "grad_norm": 0.8317292332649231, | |
| "learning_rate": 2.742161873968202e-06, | |
| "loss": 1.4783, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7679558011049724, | |
| "grad_norm": 0.892299473285675, | |
| "learning_rate": 2.681054778182748e-06, | |
| "loss": 1.3643, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7707182320441989, | |
| "grad_norm": 0.8800353407859802, | |
| "learning_rate": 2.6205306709568358e-06, | |
| "loss": 1.4015, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.7734806629834254, | |
| "grad_norm": 0.8604403138160706, | |
| "learning_rate": 2.5605943733215044e-06, | |
| "loss": 1.4247, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7762430939226519, | |
| "grad_norm": 0.8747833967208862, | |
| "learning_rate": 2.501250659485992e-06, | |
| "loss": 1.3748, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7790055248618785, | |
| "grad_norm": 0.8854520320892334, | |
| "learning_rate": 2.4425042564574186e-06, | |
| "loss": 1.4518, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7817679558011049, | |
| "grad_norm": 0.8976154327392578, | |
| "learning_rate": 2.38435984366429e-06, | |
| "loss": 1.3765, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7845303867403315, | |
| "grad_norm": 0.8289183378219604, | |
| "learning_rate": 2.3268220525837436e-06, | |
| "loss": 1.4559, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.787292817679558, | |
| "grad_norm": 0.8268212080001831, | |
| "learning_rate": 2.26989546637263e-06, | |
| "loss": 1.4105, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7900552486187845, | |
| "grad_norm": 0.7702677249908447, | |
| "learning_rate": 2.213584619502451e-06, | |
| "loss": 1.4579, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7928176795580111, | |
| "grad_norm": 0.9030853509902954, | |
| "learning_rate": 2.157893997398146e-06, | |
| "loss": 1.3844, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7955801104972375, | |
| "grad_norm": 0.8461067080497742, | |
| "learning_rate": 2.1028280360808405e-06, | |
| "loss": 1.4283, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7983425414364641, | |
| "grad_norm": 0.860618531703949, | |
| "learning_rate": 2.0483911218144713e-06, | |
| "loss": 1.4576, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.8011049723756906, | |
| "grad_norm": 0.9135028123855591, | |
| "learning_rate": 1.994587590756397e-06, | |
| "loss": 1.3549, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8038674033149171, | |
| "grad_norm": 0.936713457107544, | |
| "learning_rate": 1.941421728612023e-06, | |
| "loss": 1.4119, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.8066298342541437, | |
| "grad_norm": 0.9670917987823486, | |
| "learning_rate": 1.8888977702934086e-06, | |
| "loss": 1.4523, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.8093922651933702, | |
| "grad_norm": 0.927078902721405, | |
| "learning_rate": 1.8370198995819432e-06, | |
| "loss": 1.405, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.8121546961325967, | |
| "grad_norm": 0.8833804130554199, | |
| "learning_rate": 1.7857922487950873e-06, | |
| "loss": 1.4527, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.8149171270718232, | |
| "grad_norm": 0.821182131767273, | |
| "learning_rate": 1.7352188984572026e-06, | |
| "loss": 1.3541, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.8176795580110497, | |
| "grad_norm": 0.8782601952552795, | |
| "learning_rate": 1.6853038769745466e-06, | |
| "loss": 1.45, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.8204419889502762, | |
| "grad_norm": 0.8677568435668945, | |
| "learning_rate": 1.6360511603143648e-06, | |
| "loss": 1.4061, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.8232044198895028, | |
| "grad_norm": 0.9125531911849976, | |
| "learning_rate": 1.587464671688187e-06, | |
| "loss": 1.4138, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.8259668508287292, | |
| "grad_norm": 0.8667160272598267, | |
| "learning_rate": 1.5395482812393513e-06, | |
| "loss": 1.4394, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.8287292817679558, | |
| "grad_norm": 0.9092755913734436, | |
| "learning_rate": 1.492305805734693e-06, | |
| "loss": 1.4223, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8314917127071824, | |
| "grad_norm": 0.813664436340332, | |
| "learning_rate": 1.4457410082605483e-06, | |
| "loss": 1.4421, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8342541436464088, | |
| "grad_norm": 0.9390541911125183, | |
| "learning_rate": 1.3998575979229944e-06, | |
| "loss": 1.3292, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8370165745856354, | |
| "grad_norm": 0.9930654168128967, | |
| "learning_rate": 1.3546592295524075e-06, | |
| "loss": 1.4802, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.8397790055248618, | |
| "grad_norm": 0.8965404629707336, | |
| "learning_rate": 1.3101495034123313e-06, | |
| "loss": 1.4596, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.8425414364640884, | |
| "grad_norm": 0.9070413112640381, | |
| "learning_rate": 1.2663319649127025e-06, | |
| "loss": 1.4076, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8453038674033149, | |
| "grad_norm": 0.7749320268630981, | |
| "learning_rate": 1.2232101043274437e-06, | |
| "loss": 1.2553, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.8480662983425414, | |
| "grad_norm": 0.949621856212616, | |
| "learning_rate": 1.1807873565164507e-06, | |
| "loss": 1.3909, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.850828729281768, | |
| "grad_norm": 0.8850001096725464, | |
| "learning_rate": 1.139067100651976e-06, | |
| "loss": 1.3992, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8535911602209945, | |
| "grad_norm": 0.8339431285858154, | |
| "learning_rate": 1.0980526599494733e-06, | |
| "loss": 1.4141, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.856353591160221, | |
| "grad_norm": 0.8964627385139465, | |
| "learning_rate": 1.0577473014028872e-06, | |
| "loss": 1.3828, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8591160220994475, | |
| "grad_norm": 0.8783918023109436, | |
| "learning_rate": 1.0181542355244167e-06, | |
| "loss": 1.3791, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.861878453038674, | |
| "grad_norm": 0.8714961409568787, | |
| "learning_rate": 9.792766160887868e-07, | |
| "loss": 1.4513, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8646408839779005, | |
| "grad_norm": 0.8861087560653687, | |
| "learning_rate": 9.411175398820271e-07, | |
| "loss": 1.3653, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8674033149171271, | |
| "grad_norm": 0.8015440702438354, | |
| "learning_rate": 9.036800464548157e-07, | |
| "loss": 1.4818, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8701657458563536, | |
| "grad_norm": 0.8085585832595825, | |
| "learning_rate": 8.669671178803485e-07, | |
| "loss": 1.4441, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8729281767955801, | |
| "grad_norm": 0.8612155318260193, | |
| "learning_rate": 8.309816785168035e-07, | |
| "loss": 1.3938, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8756906077348067, | |
| "grad_norm": 0.866340160369873, | |
| "learning_rate": 7.957265947744131e-07, | |
| "loss": 1.3557, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.8784530386740331, | |
| "grad_norm": 0.8810187578201294, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 1.3672, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8812154696132597, | |
| "grad_norm": 0.9028764367103577, | |
| "learning_rate": 7.274186686889539e-07, | |
| "loss": 1.411, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8839779005524862, | |
| "grad_norm": 0.9450622797012329, | |
| "learning_rate": 6.943712673948643e-07, | |
| "loss": 1.3494, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8867403314917127, | |
| "grad_norm": 0.7640666365623474, | |
| "learning_rate": 6.620651033864844e-07, | |
| "loss": 1.4108, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8895027624309392, | |
| "grad_norm": 0.8005813360214233, | |
| "learning_rate": 6.305027500023841e-07, | |
| "loss": 1.4023, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8922651933701657, | |
| "grad_norm": 0.8076258301734924, | |
| "learning_rate": 5.996867213330993e-07, | |
| "loss": 1.4143, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.8950276243093923, | |
| "grad_norm": 0.9263111352920532, | |
| "learning_rate": 5.696194720208792e-07, | |
| "loss": 1.4247, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8977900552486188, | |
| "grad_norm": 0.7982999682426453, | |
| "learning_rate": 5.403033970641647e-07, | |
| "loss": 1.4022, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.9005524861878453, | |
| "grad_norm": 0.8366706967353821, | |
| "learning_rate": 5.117408316268047e-07, | |
| "loss": 1.5076, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.9033149171270718, | |
| "grad_norm": 0.8495275974273682, | |
| "learning_rate": 4.839340508520563e-07, | |
| "loss": 1.4251, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.9060773480662984, | |
| "grad_norm": 0.8747670650482178, | |
| "learning_rate": 4.5688526968136193e-07, | |
| "loss": 1.3898, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.9088397790055248, | |
| "grad_norm": 0.8361295461654663, | |
| "learning_rate": 4.305966426779118e-07, | |
| "loss": 1.4397, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.9116022099447514, | |
| "grad_norm": 0.8162137866020203, | |
| "learning_rate": 4.0507026385502747e-07, | |
| "loss": 1.4023, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.914364640883978, | |
| "grad_norm": 0.8866202235221863, | |
| "learning_rate": 3.8030816650935777e-07, | |
| "loss": 1.3697, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.9171270718232044, | |
| "grad_norm": 0.865460216999054, | |
| "learning_rate": 3.5631232305893047e-07, | |
| "loss": 1.4998, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.919889502762431, | |
| "grad_norm": 0.9784380197525024, | |
| "learning_rate": 3.3308464488602587e-07, | |
| "loss": 1.4262, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.9226519337016574, | |
| "grad_norm": 0.8989951610565186, | |
| "learning_rate": 3.106269821849273e-07, | |
| "loss": 1.3742, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.925414364640884, | |
| "grad_norm": 0.8595617413520813, | |
| "learning_rate": 2.889411238145545e-07, | |
| "loss": 1.4038, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.9281767955801105, | |
| "grad_norm": 0.8496826887130737, | |
| "learning_rate": 2.6802879715596585e-07, | |
| "loss": 1.3938, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.930939226519337, | |
| "grad_norm": 0.8510773777961731, | |
| "learning_rate": 2.478916679747623e-07, | |
| "loss": 1.4617, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.9337016574585635, | |
| "grad_norm": 0.8786448240280151, | |
| "learning_rate": 2.2853134028840594e-07, | |
| "loss": 1.3615, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.93646408839779, | |
| "grad_norm": 0.9002432823181152, | |
| "learning_rate": 2.099493562384469e-07, | |
| "loss": 1.4047, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.9392265193370166, | |
| "grad_norm": 0.8121012449264526, | |
| "learning_rate": 1.921471959676957e-07, | |
| "loss": 1.4307, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9419889502762431, | |
| "grad_norm": 0.8053271770477295, | |
| "learning_rate": 1.7512627750230772e-07, | |
| "loss": 1.4443, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.9447513812154696, | |
| "grad_norm": 0.9050955176353455, | |
| "learning_rate": 1.5888795663883904e-07, | |
| "loss": 1.4102, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.9475138121546961, | |
| "grad_norm": 0.9110968112945557, | |
| "learning_rate": 1.4343352683625412e-07, | |
| "loss": 1.4198, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.9502762430939227, | |
| "grad_norm": 0.929118812084198, | |
| "learning_rate": 1.2876421911288906e-07, | |
| "loss": 1.4203, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.9530386740331491, | |
| "grad_norm": 0.9060593843460083, | |
| "learning_rate": 1.148812019483958e-07, | |
| "loss": 1.3268, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9558011049723757, | |
| "grad_norm": 0.8292207717895508, | |
| "learning_rate": 1.0178558119067316e-07, | |
| "loss": 1.3877, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9585635359116023, | |
| "grad_norm": 0.8450669646263123, | |
| "learning_rate": 8.947839996777286e-08, | |
| "loss": 1.3981, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.9613259668508287, | |
| "grad_norm": 0.9085505604743958, | |
| "learning_rate": 7.796063860481595e-08, | |
| "loss": 1.353, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.9640883977900553, | |
| "grad_norm": 0.8617837429046631, | |
| "learning_rate": 6.723321454590093e-08, | |
| "loss": 1.4215, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.9668508287292817, | |
| "grad_norm": 0.8624401092529297, | |
| "learning_rate": 5.7296982281026534e-08, | |
| "loss": 1.3668, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9696132596685083, | |
| "grad_norm": 0.8622123003005981, | |
| "learning_rate": 4.815273327803183e-08, | |
| "loss": 1.3723, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9723756906077348, | |
| "grad_norm": 0.7842567563056946, | |
| "learning_rate": 3.980119591954101e-08, | |
| "loss": 1.3255, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9751381215469613, | |
| "grad_norm": 0.8146648406982422, | |
| "learning_rate": 3.224303544495766e-08, | |
| "loss": 1.3913, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9779005524861878, | |
| "grad_norm": 0.8488260507583618, | |
| "learning_rate": 2.547885389746485e-08, | |
| "loss": 1.385, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9806629834254144, | |
| "grad_norm": 0.8932926058769226, | |
| "learning_rate": 1.9509190076074657e-08, | |
| "loss": 1.3992, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9834254143646409, | |
| "grad_norm": 0.9198254346847534, | |
| "learning_rate": 1.4334519492711362e-08, | |
| "loss": 1.3375, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.9861878453038674, | |
| "grad_norm": 0.9434136748313904, | |
| "learning_rate": 9.955254334328424e-09, | |
| "loss": 1.3329, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.988950276243094, | |
| "grad_norm": 0.8083603978157043, | |
| "learning_rate": 6.371743430082511e-09, | |
| "loss": 1.4152, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9917127071823204, | |
| "grad_norm": 0.8143340349197388, | |
| "learning_rate": 3.5842722235468475e-09, | |
| "loss": 1.3503, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.994475138121547, | |
| "grad_norm": 0.8487191200256348, | |
| "learning_rate": 1.593062749967178e-09, | |
| "loss": 1.4374, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9972375690607734, | |
| "grad_norm": 0.8385122418403625, | |
| "learning_rate": 3.982736185859093e-10, | |
| "loss": 1.384, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8801374435424805, | |
| "learning_rate": 0.0, | |
| "loss": 1.368, | |
| "step": 362 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 362, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6738557053055795e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |