| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9876543209876543, | |
| "eval_steps": 500, | |
| "global_step": 363, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00823045267489712, | |
| "grad_norm": 1.4866708861109026, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 1.7565, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01646090534979424, | |
| "grad_norm": 1.4564242960595204, | |
| "learning_rate": 1.0810810810810812e-05, | |
| "loss": 1.7428, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.024691358024691357, | |
| "grad_norm": 1.5029588271629886, | |
| "learning_rate": 1.6216216216216218e-05, | |
| "loss": 1.7399, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03292181069958848, | |
| "grad_norm": 1.4275061812775922, | |
| "learning_rate": 2.1621621621621624e-05, | |
| "loss": 1.7083, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0411522633744856, | |
| "grad_norm": 1.3211892837590813, | |
| "learning_rate": 2.702702702702703e-05, | |
| "loss": 1.6472, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04938271604938271, | |
| "grad_norm": 1.2954095745224536, | |
| "learning_rate": 3.2432432432432436e-05, | |
| "loss": 1.5935, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05761316872427984, | |
| "grad_norm": 0.9115980406962145, | |
| "learning_rate": 3.783783783783784e-05, | |
| "loss": 1.4303, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06584362139917696, | |
| "grad_norm": 0.7713124020101795, | |
| "learning_rate": 4.324324324324325e-05, | |
| "loss": 1.3661, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 0.7920851957124333, | |
| "learning_rate": 4.8648648648648654e-05, | |
| "loss": 1.3093, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0823045267489712, | |
| "grad_norm": 0.8823079310051533, | |
| "learning_rate": 5.405405405405406e-05, | |
| "loss": 1.1587, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09053497942386832, | |
| "grad_norm": 0.8557863491129706, | |
| "learning_rate": 5.9459459459459466e-05, | |
| "loss": 1.0485, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.09876543209876543, | |
| "grad_norm": 0.8802985405555526, | |
| "learning_rate": 6.486486486486487e-05, | |
| "loss": 0.9429, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.10699588477366255, | |
| "grad_norm": 0.8436384566917278, | |
| "learning_rate": 7.027027027027028e-05, | |
| "loss": 0.8218, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11522633744855967, | |
| "grad_norm": 0.6676398141327192, | |
| "learning_rate": 7.567567567567568e-05, | |
| "loss": 0.7054, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 0.6471842089986185, | |
| "learning_rate": 8.108108108108109e-05, | |
| "loss": 0.6758, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.13168724279835392, | |
| "grad_norm": 0.4142479004737112, | |
| "learning_rate": 8.64864864864865e-05, | |
| "loss": 0.6222, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.13991769547325103, | |
| "grad_norm": 0.42850243250289555, | |
| "learning_rate": 9.18918918918919e-05, | |
| "loss": 0.5962, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.5568208902875614, | |
| "learning_rate": 9.729729729729731e-05, | |
| "loss": 0.5984, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.15637860082304528, | |
| "grad_norm": 0.5540687236420326, | |
| "learning_rate": 0.0001027027027027027, | |
| "loss": 0.4938, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1646090534979424, | |
| "grad_norm": 0.33332950971206693, | |
| "learning_rate": 0.00010810810810810812, | |
| "loss": 0.5161, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1728395061728395, | |
| "grad_norm": 0.2653066423031725, | |
| "learning_rate": 0.00011351351351351351, | |
| "loss": 0.5136, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.18106995884773663, | |
| "grad_norm": 0.2272101382108708, | |
| "learning_rate": 0.00011891891891891893, | |
| "loss": 0.4873, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.18930041152263374, | |
| "grad_norm": 0.24006404965358885, | |
| "learning_rate": 0.00012432432432432433, | |
| "loss": 0.4844, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.20608476733616163, | |
| "learning_rate": 0.00012972972972972974, | |
| "loss": 0.4821, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.205761316872428, | |
| "grad_norm": 0.16150407452920948, | |
| "learning_rate": 0.00013513513513513514, | |
| "loss": 0.4167, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2139917695473251, | |
| "grad_norm": 0.1538751616515209, | |
| "learning_rate": 0.00014054054054054056, | |
| "loss": 0.446, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.16472183357224798, | |
| "learning_rate": 0.00014594594594594595, | |
| "loss": 0.4142, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.23045267489711935, | |
| "grad_norm": 0.1552702492617925, | |
| "learning_rate": 0.00015135135135135137, | |
| "loss": 0.4909, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.23868312757201646, | |
| "grad_norm": 0.14727903417905372, | |
| "learning_rate": 0.00015675675675675676, | |
| "loss": 0.4065, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 0.14784235244019464, | |
| "learning_rate": 0.00016216216216216218, | |
| "loss": 0.3886, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2551440329218107, | |
| "grad_norm": 0.1450616035553425, | |
| "learning_rate": 0.00016756756756756757, | |
| "loss": 0.3902, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.26337448559670784, | |
| "grad_norm": 0.153342097650188, | |
| "learning_rate": 0.000172972972972973, | |
| "loss": 0.3888, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2716049382716049, | |
| "grad_norm": 0.15900713930790533, | |
| "learning_rate": 0.00017837837837837839, | |
| "loss": 0.3524, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.27983539094650206, | |
| "grad_norm": 0.1546499892536795, | |
| "learning_rate": 0.0001837837837837838, | |
| "loss": 0.3648, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2880658436213992, | |
| "grad_norm": 0.14320964717963003, | |
| "learning_rate": 0.0001891891891891892, | |
| "loss": 0.3658, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.13103727806567708, | |
| "learning_rate": 0.00019459459459459462, | |
| "loss": 0.3496, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3045267489711934, | |
| "grad_norm": 0.13308383037195082, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3572, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.31275720164609055, | |
| "grad_norm": 0.1387871382009263, | |
| "learning_rate": 0.00019999535665248002, | |
| "loss": 0.3214, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.32098765432098764, | |
| "grad_norm": 0.14270131722019058, | |
| "learning_rate": 0.0001999814270411335, | |
| "loss": 0.3445, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3292181069958848, | |
| "grad_norm": 0.1545337969749864, | |
| "learning_rate": 0.000199958212459561, | |
| "loss": 0.32, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3374485596707819, | |
| "grad_norm": 0.13802182056866283, | |
| "learning_rate": 0.00019992571506363, | |
| "loss": 0.3449, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.345679012345679, | |
| "grad_norm": 0.14162173979281106, | |
| "learning_rate": 0.00019988393787127441, | |
| "loss": 0.3262, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.35390946502057613, | |
| "grad_norm": 0.13594846523403772, | |
| "learning_rate": 0.0001998328847622148, | |
| "loss": 0.2958, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.36213991769547327, | |
| "grad_norm": 0.12596629667831052, | |
| "learning_rate": 0.00019977256047759765, | |
| "loss": 0.286, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 0.13835455454793713, | |
| "learning_rate": 0.00019970297061955533, | |
| "loss": 0.2878, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3786008230452675, | |
| "grad_norm": 0.14644104388608956, | |
| "learning_rate": 0.00019962412165068573, | |
| "loss": 0.2952, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3868312757201646, | |
| "grad_norm": 0.1361181054099992, | |
| "learning_rate": 0.00019953602089345217, | |
| "loss": 0.267, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.1545760777310844, | |
| "learning_rate": 0.0001994386765295032, | |
| "loss": 0.2823, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.40329218106995884, | |
| "grad_norm": 0.16048561280797768, | |
| "learning_rate": 0.00019933209759891317, | |
| "loss": 0.2598, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.411522633744856, | |
| "grad_norm": 0.14585143788717048, | |
| "learning_rate": 0.00019921629399934223, | |
| "loss": 0.2834, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.41975308641975306, | |
| "grad_norm": 0.15590757372746555, | |
| "learning_rate": 0.00019909127648511755, | |
| "loss": 0.2619, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4279835390946502, | |
| "grad_norm": 0.142205678471833, | |
| "learning_rate": 0.0001989570566662345, | |
| "loss": 0.2477, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.43621399176954734, | |
| "grad_norm": 0.14334288636686987, | |
| "learning_rate": 0.00019881364700727823, | |
| "loss": 0.2958, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.1584716514858334, | |
| "learning_rate": 0.0001986610608262665, | |
| "loss": 0.2708, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "grad_norm": 0.15229425624208448, | |
| "learning_rate": 0.00019849931229341258, | |
| "loss": 0.2776, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.4609053497942387, | |
| "grad_norm": 0.1497514774482752, | |
| "learning_rate": 0.00019832841642980945, | |
| "loss": 0.2325, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.4691358024691358, | |
| "grad_norm": 0.15302882739571017, | |
| "learning_rate": 0.00019814838910603481, | |
| "loss": 0.2755, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.4773662551440329, | |
| "grad_norm": 0.20080266795074445, | |
| "learning_rate": 0.00019795924704067721, | |
| "loss": 0.2421, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.48559670781893005, | |
| "grad_norm": 0.15063814452987448, | |
| "learning_rate": 0.00019776100779878345, | |
| "loss": 0.2152, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 0.1424158898404743, | |
| "learning_rate": 0.00019755368979022732, | |
| "loss": 0.2424, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5020576131687243, | |
| "grad_norm": 0.15708199535695755, | |
| "learning_rate": 0.00019733731226800015, | |
| "loss": 0.2439, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5102880658436214, | |
| "grad_norm": 0.1469505519285418, | |
| "learning_rate": 0.00019711189532642243, | |
| "loss": 0.2174, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 0.15443629035666892, | |
| "learning_rate": 0.00019687745989927823, | |
| "loss": 0.2201, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5267489711934157, | |
| "grad_norm": 0.16091185410714737, | |
| "learning_rate": 0.00019663402775787066, | |
| "loss": 0.2153, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5349794238683128, | |
| "grad_norm": 0.16812247962269722, | |
| "learning_rate": 0.00019638162150900027, | |
| "loss": 0.2245, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5432098765432098, | |
| "grad_norm": 0.15476006323853086, | |
| "learning_rate": 0.00019612026459286578, | |
| "loss": 0.2168, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.551440329218107, | |
| "grad_norm": 0.14319948869013133, | |
| "learning_rate": 0.00019584998128088684, | |
| "loss": 0.2102, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5596707818930041, | |
| "grad_norm": 0.1416595781349322, | |
| "learning_rate": 0.0001955707966734505, | |
| "loss": 0.2109, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5679012345679012, | |
| "grad_norm": 0.15627936343326146, | |
| "learning_rate": 0.00019528273669757972, | |
| "loss": 0.221, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5761316872427984, | |
| "grad_norm": 0.1355242418955904, | |
| "learning_rate": 0.0001949858281045261, | |
| "loss": 0.1934, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5843621399176955, | |
| "grad_norm": 0.15837694932534985, | |
| "learning_rate": 0.00019468009846728513, | |
| "loss": 0.2106, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.14575785598463487, | |
| "learning_rate": 0.00019436557617803595, | |
| "loss": 0.1958, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6008230452674898, | |
| "grad_norm": 0.15004026647926874, | |
| "learning_rate": 0.00019404229044550433, | |
| "loss": 0.2111, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6090534979423868, | |
| "grad_norm": 0.1531758472218286, | |
| "learning_rate": 0.00019371027129225042, | |
| "loss": 0.1796, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 0.14522275863691966, | |
| "learning_rate": 0.0001933695495518804, | |
| "loss": 0.1879, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6255144032921811, | |
| "grad_norm": 0.14778202822977027, | |
| "learning_rate": 0.00019302015686618326, | |
| "loss": 0.1783, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6337448559670782, | |
| "grad_norm": 0.14202535229938665, | |
| "learning_rate": 0.0001926621256821922, | |
| "loss": 0.1672, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6419753086419753, | |
| "grad_norm": 0.15532295497905474, | |
| "learning_rate": 0.00019229548924917146, | |
| "loss": 0.1894, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6502057613168725, | |
| "grad_norm": 0.1298933712297695, | |
| "learning_rate": 0.00019192028161552847, | |
| "loss": 0.1626, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6584362139917695, | |
| "grad_norm": 0.15486589167638415, | |
| "learning_rate": 0.0001915365376256519, | |
| "loss": 0.1829, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.13198994845547862, | |
| "learning_rate": 0.00019114429291667583, | |
| "loss": 0.1827, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6748971193415638, | |
| "grad_norm": 0.14486854526136775, | |
| "learning_rate": 0.00019074358391517023, | |
| "loss": 0.1711, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6831275720164609, | |
| "grad_norm": 0.18136392511258384, | |
| "learning_rate": 0.00019033444783375804, | |
| "loss": 0.1852, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.691358024691358, | |
| "grad_norm": 0.12835927358561786, | |
| "learning_rate": 0.00018991692266765947, | |
| "loss": 0.1874, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6995884773662552, | |
| "grad_norm": 0.15923203558527596, | |
| "learning_rate": 0.00018949104719116332, | |
| "loss": 0.1754, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7078189300411523, | |
| "grad_norm": 0.1524830096667991, | |
| "learning_rate": 0.00018905686095402647, | |
| "loss": 0.1772, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7160493827160493, | |
| "grad_norm": 0.1458630884597557, | |
| "learning_rate": 0.0001886144042778006, | |
| "loss": 0.1884, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7242798353909465, | |
| "grad_norm": 0.14812531898730455, | |
| "learning_rate": 0.00018816371825208789, | |
| "loss": 0.1694, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7325102880658436, | |
| "grad_norm": 0.1371838995231899, | |
| "learning_rate": 0.0001877048447307252, | |
| "loss": 0.175, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.12597672532717027, | |
| "learning_rate": 0.00018723782632789701, | |
| "loss": 0.1663, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7489711934156379, | |
| "grad_norm": 0.13653897293999862, | |
| "learning_rate": 0.00018676270641417822, | |
| "loss": 0.1902, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.757201646090535, | |
| "grad_norm": 0.11918081783294039, | |
| "learning_rate": 0.0001862795291125063, | |
| "loss": 0.1662, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7654320987654321, | |
| "grad_norm": 0.1333312758081247, | |
| "learning_rate": 0.0001857883392940837, | |
| "loss": 0.199, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7736625514403292, | |
| "grad_norm": 0.12930618535907987, | |
| "learning_rate": 0.000185289182574211, | |
| "loss": 0.1697, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7818930041152263, | |
| "grad_norm": 0.1336640654310965, | |
| "learning_rate": 0.0001847821053080505, | |
| "loss": 0.1852, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.14974272672832492, | |
| "learning_rate": 0.00018426715458632153, | |
| "loss": 0.1819, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.7983539094650206, | |
| "grad_norm": 0.11738767625285426, | |
| "learning_rate": 0.00018374437823092724, | |
| "loss": 0.1628, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8065843621399177, | |
| "grad_norm": 0.1126501945549117, | |
| "learning_rate": 0.00018321382479051347, | |
| "loss": 0.1574, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 0.14716070780057008, | |
| "learning_rate": 0.00018267554353596025, | |
| "loss": 0.1671, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.823045267489712, | |
| "grad_norm": 0.14236534823623162, | |
| "learning_rate": 0.0001821295844558062, | |
| "loss": 0.179, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.831275720164609, | |
| "grad_norm": 0.14819744720356537, | |
| "learning_rate": 0.0001815759982516061, | |
| "loss": 0.1765, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8395061728395061, | |
| "grad_norm": 0.1359773653210936, | |
| "learning_rate": 0.00018101483633322255, | |
| "loss": 0.1736, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8477366255144033, | |
| "grad_norm": 0.12220336917197815, | |
| "learning_rate": 0.00018044615081405153, | |
| "loss": 0.1559, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8559670781893004, | |
| "grad_norm": 0.12734749296356374, | |
| "learning_rate": 0.00017986999450618295, | |
| "loss": 0.1598, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 0.13635727971275893, | |
| "learning_rate": 0.00017928642091549613, | |
| "loss": 0.1716, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.8724279835390947, | |
| "grad_norm": 0.1256041209800328, | |
| "learning_rate": 0.00017869548423669077, | |
| "loss": 0.1694, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.8806584362139918, | |
| "grad_norm": 0.156735390007985, | |
| "learning_rate": 0.00017809723934825405, | |
| "loss": 0.1711, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.13613462557969605, | |
| "learning_rate": 0.00017749174180736442, | |
| "loss": 0.1575, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.897119341563786, | |
| "grad_norm": 0.12136550882477753, | |
| "learning_rate": 0.00017687904784473188, | |
| "loss": 0.1541, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9053497942386831, | |
| "grad_norm": 0.11805606680499708, | |
| "learning_rate": 0.00017625921435937637, | |
| "loss": 0.153, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9135802469135802, | |
| "grad_norm": 0.12306976139021918, | |
| "learning_rate": 0.00017563229891334338, | |
| "loss": 0.1723, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9218106995884774, | |
| "grad_norm": 0.12639467162562257, | |
| "learning_rate": 0.00017499835972635856, | |
| "loss": 0.1637, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9300411522633745, | |
| "grad_norm": 0.11341325633796959, | |
| "learning_rate": 0.00017435745567042095, | |
| "loss": 0.1471, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.9382716049382716, | |
| "grad_norm": 0.12238460464162876, | |
| "learning_rate": 0.00017370964626433567, | |
| "loss": 0.1682, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.9465020576131687, | |
| "grad_norm": 0.12574079950832473, | |
| "learning_rate": 0.0001730549916681868, | |
| "loss": 0.1493, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9547325102880658, | |
| "grad_norm": 0.13251823016582745, | |
| "learning_rate": 0.00017239355267775018, | |
| "loss": 0.1649, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 0.11699891536949006, | |
| "learning_rate": 0.0001717253907188477, | |
| "loss": 0.1628, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.9711934156378601, | |
| "grad_norm": 0.11316448680114288, | |
| "learning_rate": 0.00017105056784164294, | |
| "loss": 0.1434, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.9794238683127572, | |
| "grad_norm": 0.11531932007410475, | |
| "learning_rate": 0.00017036914671487852, | |
| "loss": 0.1565, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.11423778544173263, | |
| "learning_rate": 0.00016968119062005642, | |
| "loss": 0.1481, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.9958847736625515, | |
| "grad_norm": 0.10808259099105172, | |
| "learning_rate": 0.00016898676344556118, | |
| "loss": 0.1393, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.9958847736625515, | |
| "eval_loss": 0.1582392454147339, | |
| "eval_runtime": 24.3633, | |
| "eval_samples_per_second": 33.534, | |
| "eval_steps_per_second": 1.067, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0041152263374487, | |
| "grad_norm": 0.12179637517786698, | |
| "learning_rate": 0.00016828592968072678, | |
| "loss": 0.1367, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0123456790123457, | |
| "grad_norm": 0.12278954055202473, | |
| "learning_rate": 0.00016757875440984768, | |
| "loss": 0.1352, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0205761316872428, | |
| "grad_norm": 0.11262318129903952, | |
| "learning_rate": 0.0001668653033061347, | |
| "loss": 0.1319, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.02880658436214, | |
| "grad_norm": 0.10685919423230769, | |
| "learning_rate": 0.00016614564262561608, | |
| "loss": 0.1483, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 0.12688594869999706, | |
| "learning_rate": 0.0001654198392009846, | |
| "loss": 0.1345, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.045267489711934, | |
| "grad_norm": 0.11380792656533252, | |
| "learning_rate": 0.0001646879604353908, | |
| "loss": 0.1435, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0534979423868314, | |
| "grad_norm": 0.11336135523754398, | |
| "learning_rate": 0.00016395007429618382, | |
| "loss": 0.1496, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.0617283950617284, | |
| "grad_norm": 0.11068219074387899, | |
| "learning_rate": 0.00016320624930859904, | |
| "loss": 0.1412, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.0699588477366255, | |
| "grad_norm": 0.10745622525472565, | |
| "learning_rate": 0.00016245655454939474, | |
| "loss": 0.1294, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0781893004115226, | |
| "grad_norm": 0.10201633324761311, | |
| "learning_rate": 0.00016170105964043695, | |
| "loss": 0.1443, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.0864197530864197, | |
| "grad_norm": 0.1137314706455847, | |
| "learning_rate": 0.0001609398347422339, | |
| "loss": 0.1389, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.0946502057613168, | |
| "grad_norm": 0.11204326746406651, | |
| "learning_rate": 0.00016017295054742046, | |
| "loss": 0.1422, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.102880658436214, | |
| "grad_norm": 0.10142584272145459, | |
| "learning_rate": 0.00015940047827419303, | |
| "loss": 0.1301, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.10382986472663408, | |
| "learning_rate": 0.00015862248965969604, | |
| "loss": 0.1388, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1193415637860082, | |
| "grad_norm": 0.11472916493277631, | |
| "learning_rate": 0.00015783905695335946, | |
| "loss": 0.1406, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.1275720164609053, | |
| "grad_norm": 0.11056711973862805, | |
| "learning_rate": 0.0001570502529101896, | |
| "loss": 0.1295, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.1358024691358024, | |
| "grad_norm": 0.11881161089888494, | |
| "learning_rate": 0.00015625615078401244, | |
| "loss": 0.1491, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.1440329218106995, | |
| "grad_norm": 0.10248132676392517, | |
| "learning_rate": 0.00015545682432067067, | |
| "loss": 0.1235, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1522633744855968, | |
| "grad_norm": 0.10737832299551002, | |
| "learning_rate": 0.0001546523477511754, | |
| "loss": 0.15, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1604938271604939, | |
| "grad_norm": 0.10807513117026502, | |
| "learning_rate": 0.00015384279578481221, | |
| "loss": 0.1302, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.168724279835391, | |
| "grad_norm": 0.11235063041052787, | |
| "learning_rate": 0.00015302824360220353, | |
| "loss": 0.1386, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.176954732510288, | |
| "grad_norm": 0.11303603244571206, | |
| "learning_rate": 0.00015220876684832638, | |
| "loss": 0.1354, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.10834449382941344, | |
| "learning_rate": 0.0001513844416254879, | |
| "loss": 0.1329, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.1934156378600824, | |
| "grad_norm": 0.1079013280016857, | |
| "learning_rate": 0.00015055534448625766, | |
| "loss": 0.1395, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2016460905349795, | |
| "grad_norm": 0.10261607273228789, | |
| "learning_rate": 0.00014972155242635852, | |
| "loss": 0.129, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2098765432098766, | |
| "grad_norm": 0.11363708966450349, | |
| "learning_rate": 0.0001488831428775164, | |
| "loss": 0.1461, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.2181069958847737, | |
| "grad_norm": 0.11418227389762935, | |
| "learning_rate": 0.00014804019370026926, | |
| "loss": 0.1408, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.2263374485596708, | |
| "grad_norm": 0.11642847183223218, | |
| "learning_rate": 0.00014719278317673655, | |
| "loss": 0.1462, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.2345679012345678, | |
| "grad_norm": 0.11219874980224723, | |
| "learning_rate": 0.0001463409900033493, | |
| "loss": 0.1302, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.242798353909465, | |
| "grad_norm": 0.11039753961672125, | |
| "learning_rate": 0.00014548489328354195, | |
| "loss": 0.1349, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.2510288065843622, | |
| "grad_norm": 0.10772127499535779, | |
| "learning_rate": 0.00014462457252040607, | |
| "loss": 0.134, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.2592592592592593, | |
| "grad_norm": 0.10693684319005581, | |
| "learning_rate": 0.00014376010760930728, | |
| "loss": 0.1314, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.2674897119341564, | |
| "grad_norm": 0.10099951149470732, | |
| "learning_rate": 0.00014289157883046568, | |
| "loss": 0.1314, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.2757201646090535, | |
| "grad_norm": 0.10714073475937398, | |
| "learning_rate": 0.0001420190668415002, | |
| "loss": 0.1168, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.2839506172839505, | |
| "grad_norm": 0.11431029472007842, | |
| "learning_rate": 0.00014114265266993846, | |
| "loss": 0.1457, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.2921810699588478, | |
| "grad_norm": 0.11035801526331707, | |
| "learning_rate": 0.00014026241770569197, | |
| "loss": 0.1496, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.300411522633745, | |
| "grad_norm": 0.10419031063352954, | |
| "learning_rate": 0.00013937844369349734, | |
| "loss": 0.1323, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.308641975308642, | |
| "grad_norm": 0.10823681149718124, | |
| "learning_rate": 0.00013849081272532544, | |
| "loss": 0.1264, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.316872427983539, | |
| "grad_norm": 0.11224858342990347, | |
| "learning_rate": 0.00013759960723275732, | |
| "loss": 0.1494, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3251028806584362, | |
| "grad_norm": 0.11222637320000997, | |
| "learning_rate": 0.00013670490997932922, | |
| "loss": 0.1446, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.11350101469976674, | |
| "learning_rate": 0.00013580680405284664, | |
| "loss": 0.1501, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.3415637860082303, | |
| "grad_norm": 0.11216990215296897, | |
| "learning_rate": 0.00013490537285766808, | |
| "loss": 0.1518, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.3497942386831276, | |
| "grad_norm": 0.10385126129598882, | |
| "learning_rate": 0.00013400070010695966, | |
| "loss": 0.1326, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.3580246913580247, | |
| "grad_norm": 0.09801204121771982, | |
| "learning_rate": 0.00013309286981492085, | |
| "loss": 0.1385, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.3662551440329218, | |
| "grad_norm": 0.10515015522554948, | |
| "learning_rate": 0.00013218196628898233, | |
| "loss": 0.1435, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.374485596707819, | |
| "grad_norm": 0.11167870227165867, | |
| "learning_rate": 0.00013126807412197665, | |
| "loss": 0.1469, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.110692438020908, | |
| "learning_rate": 0.0001303512781842824, | |
| "loss": 0.1267, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.3909465020576133, | |
| "grad_norm": 0.11426160677474222, | |
| "learning_rate": 0.00012943166361594242, | |
| "loss": 0.1308, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.3991769547325104, | |
| "grad_norm": 0.11583690935003194, | |
| "learning_rate": 0.00012850931581875723, | |
| "loss": 0.1484, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4074074074074074, | |
| "grad_norm": 0.10018605187361156, | |
| "learning_rate": 0.00012758432044835392, | |
| "loss": 0.141, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.4156378600823045, | |
| "grad_norm": 0.10604765462173282, | |
| "learning_rate": 0.0001266567634062317, | |
| "loss": 0.1291, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.4238683127572016, | |
| "grad_norm": 0.10158287607968718, | |
| "learning_rate": 0.0001257267308317845, | |
| "loss": 0.1276, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.4320987654320987, | |
| "grad_norm": 0.10299808028319524, | |
| "learning_rate": 0.00012479430909430108, | |
| "loss": 0.1317, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.4403292181069958, | |
| "grad_norm": 0.1118568576111719, | |
| "learning_rate": 0.00012385958478494487, | |
| "loss": 0.1288, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.448559670781893, | |
| "grad_norm": 0.10353944126895748, | |
| "learning_rate": 0.00012292264470871182, | |
| "loss": 0.1175, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.4567901234567902, | |
| "grad_norm": 0.11318325883515341, | |
| "learning_rate": 0.00012198357587636957, | |
| "loss": 0.1304, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.4650205761316872, | |
| "grad_norm": 0.11698708468284778, | |
| "learning_rate": 0.00012104246549637683, | |
| "loss": 0.143, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.4732510288065843, | |
| "grad_norm": 0.12408074070049697, | |
| "learning_rate": 0.00012009940096678452, | |
| "loss": 0.1583, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "grad_norm": 0.115849428171122, | |
| "learning_rate": 0.00011915446986711953, | |
| "loss": 0.1401, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.4897119341563787, | |
| "grad_norm": 0.10498841318355592, | |
| "learning_rate": 0.00011820775995025147, | |
| "loss": 0.1404, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.4979423868312758, | |
| "grad_norm": 0.11284118297477287, | |
| "learning_rate": 0.0001172593591342432, | |
| "loss": 0.1479, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.5061728395061729, | |
| "grad_norm": 0.1063063021523787, | |
| "learning_rate": 0.00011630935549418627, | |
| "loss": 0.1223, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.51440329218107, | |
| "grad_norm": 0.09875055061391125, | |
| "learning_rate": 0.00011535783725402163, | |
| "loss": 0.1177, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.522633744855967, | |
| "grad_norm": 0.10114695106701617, | |
| "learning_rate": 0.00011440489277834645, | |
| "loss": 0.1403, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5308641975308643, | |
| "grad_norm": 0.09017689726172246, | |
| "learning_rate": 0.0001134506105642081, | |
| "loss": 0.1114, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.5390946502057612, | |
| "grad_norm": 0.09727031699745013, | |
| "learning_rate": 0.00011249507923288562, | |
| "loss": 0.1368, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.5473251028806585, | |
| "grad_norm": 0.10131253354694934, | |
| "learning_rate": 0.0001115383875216598, | |
| "loss": 0.1278, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.09766756433474798, | |
| "learning_rate": 0.00011058062427557229, | |
| "loss": 0.1284, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.5637860082304527, | |
| "grad_norm": 0.09911328195816287, | |
| "learning_rate": 0.00010962187843917497, | |
| "loss": 0.1284, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5720164609053497, | |
| "grad_norm": 0.10809617890378591, | |
| "learning_rate": 0.0001086622390482699, | |
| "loss": 0.1423, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.09456158065899406, | |
| "learning_rate": 0.00010770179522164079, | |
| "loss": 0.1317, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.5884773662551441, | |
| "grad_norm": 0.09799719160217063, | |
| "learning_rate": 0.0001067406361527768, | |
| "loss": 0.1356, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.596707818930041, | |
| "grad_norm": 0.09992479118544549, | |
| "learning_rate": 0.00010577885110158958, | |
| "loss": 0.1292, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.6049382716049383, | |
| "grad_norm": 0.09693089475338096, | |
| "learning_rate": 0.00010481652938612374, | |
| "loss": 0.1391, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6131687242798354, | |
| "grad_norm": 0.09217578544631694, | |
| "learning_rate": 0.00010385376037426226, | |
| "loss": 0.1152, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.6213991769547325, | |
| "grad_norm": 0.09249375316990154, | |
| "learning_rate": 0.00010289063347542726, | |
| "loss": 0.1154, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.6296296296296298, | |
| "grad_norm": 0.09641811645171969, | |
| "learning_rate": 0.00010192723813227672, | |
| "loss": 0.1182, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.6378600823045266, | |
| "grad_norm": 0.09748485166957398, | |
| "learning_rate": 0.00010096366381239808, | |
| "loss": 0.1338, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.646090534979424, | |
| "grad_norm": 0.09383283234337292, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1235, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.654320987654321, | |
| "grad_norm": 0.09670711594290475, | |
| "learning_rate": 9.903633618760195e-05, | |
| "loss": 0.1222, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.662551440329218, | |
| "grad_norm": 0.09682675756651625, | |
| "learning_rate": 9.807276186772333e-05, | |
| "loss": 0.1271, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.6707818930041154, | |
| "grad_norm": 0.09455470748347092, | |
| "learning_rate": 9.710936652457276e-05, | |
| "loss": 0.1217, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.6790123456790123, | |
| "grad_norm": 0.10988834676545467, | |
| "learning_rate": 9.614623962573776e-05, | |
| "loss": 0.1288, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.6872427983539096, | |
| "grad_norm": 0.10415165974330459, | |
| "learning_rate": 9.518347061387628e-05, | |
| "loss": 0.1307, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.6954732510288066, | |
| "grad_norm": 0.10985786659106267, | |
| "learning_rate": 9.422114889841044e-05, | |
| "loss": 0.1405, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.7037037037037037, | |
| "grad_norm": 0.09921801439791741, | |
| "learning_rate": 9.325936384722321e-05, | |
| "loss": 0.1421, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.7119341563786008, | |
| "grad_norm": 0.09695464903081527, | |
| "learning_rate": 9.229820477835927e-05, | |
| "loss": 0.1261, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.7201646090534979, | |
| "grad_norm": 0.09747696306459716, | |
| "learning_rate": 9.133776095173015e-05, | |
| "loss": 0.1295, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.7283950617283952, | |
| "grad_norm": 0.09028526494291234, | |
| "learning_rate": 9.037812156082504e-05, | |
| "loss": 0.1148, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.736625514403292, | |
| "grad_norm": 0.11512926087416764, | |
| "learning_rate": 8.941937572442773e-05, | |
| "loss": 0.1385, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.7448559670781894, | |
| "grad_norm": 0.09719378966338557, | |
| "learning_rate": 8.846161247834024e-05, | |
| "loss": 0.1308, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.7530864197530864, | |
| "grad_norm": 0.09008557356074057, | |
| "learning_rate": 8.750492076711439e-05, | |
| "loss": 0.1142, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.7613168724279835, | |
| "grad_norm": 0.10577290101156246, | |
| "learning_rate": 8.654938943579194e-05, | |
| "loss": 0.1409, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.7695473251028808, | |
| "grad_norm": 0.10122461461639144, | |
| "learning_rate": 8.55951072216536e-05, | |
| "loss": 0.1314, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.09823573980962558, | |
| "learning_rate": 8.464216274597838e-05, | |
| "loss": 0.1219, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.786008230452675, | |
| "grad_norm": 0.10401921170876789, | |
| "learning_rate": 8.369064450581373e-05, | |
| "loss": 0.144, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.794238683127572, | |
| "grad_norm": 0.09078999043742002, | |
| "learning_rate": 8.274064086575681e-05, | |
| "loss": 0.1146, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.8024691358024691, | |
| "grad_norm": 0.09290736424935679, | |
| "learning_rate": 8.179224004974857e-05, | |
| "loss": 0.1338, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.8106995884773662, | |
| "grad_norm": 0.10089995497723929, | |
| "learning_rate": 8.084553013288048e-05, | |
| "loss": 0.1299, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8189300411522633, | |
| "grad_norm": 0.11086354657368475, | |
| "learning_rate": 7.990059903321553e-05, | |
| "loss": 0.1413, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.8271604938271606, | |
| "grad_norm": 0.10239856014254489, | |
| "learning_rate": 7.89575345036232e-05, | |
| "loss": 0.1357, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.8353909465020575, | |
| "grad_norm": 0.10469014425315878, | |
| "learning_rate": 7.801642412363041e-05, | |
| "loss": 0.1386, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.8436213991769548, | |
| "grad_norm": 0.09567702676769906, | |
| "learning_rate": 7.707735529128819e-05, | |
| "loss": 0.1191, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 0.09887833758041985, | |
| "learning_rate": 7.614041521505517e-05, | |
| "loss": 0.1219, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.860082304526749, | |
| "grad_norm": 0.11752510500534394, | |
| "learning_rate": 7.520569090569893e-05, | |
| "loss": 0.1436, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.8683127572016462, | |
| "grad_norm": 0.10388061868632578, | |
| "learning_rate": 7.427326916821557e-05, | |
| "loss": 0.1521, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.876543209876543, | |
| "grad_norm": 0.09472507284073291, | |
| "learning_rate": 7.334323659376829e-05, | |
| "loss": 0.1309, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.8847736625514404, | |
| "grad_norm": 0.09368373509928322, | |
| "learning_rate": 7.24156795516461e-05, | |
| "loss": 0.1226, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.8930041152263375, | |
| "grad_norm": 0.09798127382996634, | |
| "learning_rate": 7.149068418124281e-05, | |
| "loss": 0.134, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9012345679012346, | |
| "grad_norm": 0.09780244211128043, | |
| "learning_rate": 7.056833638405762e-05, | |
| "loss": 0.1139, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.9094650205761317, | |
| "grad_norm": 0.09955863946262253, | |
| "learning_rate": 6.964872181571764e-05, | |
| "loss": 0.1372, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.9176954732510287, | |
| "grad_norm": 0.0999251460303617, | |
| "learning_rate": 6.87319258780234e-05, | |
| "loss": 0.1302, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.925925925925926, | |
| "grad_norm": 0.10537500302863075, | |
| "learning_rate": 6.781803371101774e-05, | |
| "loss": 0.1361, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.934156378600823, | |
| "grad_norm": 0.10646426579724969, | |
| "learning_rate": 6.690713018507918e-05, | |
| "loss": 0.1382, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.9423868312757202, | |
| "grad_norm": 0.09815506133539018, | |
| "learning_rate": 6.599929989304035e-05, | |
| "loss": 0.1248, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.9506172839506173, | |
| "grad_norm": 0.10476726048050965, | |
| "learning_rate": 6.509462714233195e-05, | |
| "loss": 0.139, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.9588477366255144, | |
| "grad_norm": 0.10022243654627384, | |
| "learning_rate": 6.419319594715339e-05, | |
| "loss": 0.1285, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.9670781893004117, | |
| "grad_norm": 0.09626907084774464, | |
| "learning_rate": 6.32950900206708e-05, | |
| "loss": 0.1345, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.09684684548905444, | |
| "learning_rate": 6.240039276724272e-05, | |
| "loss": 0.1338, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9835390946502058, | |
| "grad_norm": 0.09617321255871392, | |
| "learning_rate": 6.150918727467455e-05, | |
| "loss": 0.1475, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.991769547325103, | |
| "grad_norm": 0.10288685971512879, | |
| "learning_rate": 6.062155630650265e-05, | |
| "loss": 0.13, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.09997416312505185, | |
| "learning_rate": 5.973758229430806e-05, | |
| "loss": 0.1282, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.14014942944049835, | |
| "eval_runtime": 21.1655, | |
| "eval_samples_per_second": 38.601, | |
| "eval_steps_per_second": 1.228, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.0082304526748973, | |
| "grad_norm": 0.0846538233930147, | |
| "learning_rate": 5.885734733006154e-05, | |
| "loss": 0.0975, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.016460905349794, | |
| "grad_norm": 0.09549609421876461, | |
| "learning_rate": 5.798093315849984e-05, | |
| "loss": 0.1102, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.0246913580246915, | |
| "grad_norm": 0.09559068048673255, | |
| "learning_rate": 5.710842116953438e-05, | |
| "loss": 0.1025, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.0329218106995883, | |
| "grad_norm": 0.10047776150134546, | |
| "learning_rate": 5.623989239069275e-05, | |
| "loss": 0.1167, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.0411522633744856, | |
| "grad_norm": 0.10670187619168023, | |
| "learning_rate": 5.537542747959394e-05, | |
| "loss": 0.1115, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.049382716049383, | |
| "grad_norm": 0.10560096207358394, | |
| "learning_rate": 5.451510671645807e-05, | |
| "loss": 0.117, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.05761316872428, | |
| "grad_norm": 0.10122771696509619, | |
| "learning_rate": 5.36590099966507e-05, | |
| "loss": 0.1253, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.065843621399177, | |
| "grad_norm": 0.09738370841133107, | |
| "learning_rate": 5.2807216823263484e-05, | |
| "loss": 0.1188, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.074074074074074, | |
| "grad_norm": 0.09886958144813014, | |
| "learning_rate": 5.1959806299730774e-05, | |
| "loss": 0.1237, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.0823045267489713, | |
| "grad_norm": 0.10219049090054223, | |
| "learning_rate": 5.111685712248364e-05, | |
| "loss": 0.1137, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.090534979423868, | |
| "grad_norm": 0.09836202650897583, | |
| "learning_rate": 5.0278447573641495e-05, | |
| "loss": 0.1196, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.0987654320987654, | |
| "grad_norm": 0.09213881530201376, | |
| "learning_rate": 4.944465551374238e-05, | |
| "loss": 0.0999, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.1069958847736627, | |
| "grad_norm": 0.09541725924400123, | |
| "learning_rate": 4.861555837451213e-05, | |
| "loss": 0.1042, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.1152263374485596, | |
| "grad_norm": 0.09856792963956205, | |
| "learning_rate": 4.779123315167362e-05, | |
| "loss": 0.1139, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.123456790123457, | |
| "grad_norm": 0.09596664358323789, | |
| "learning_rate": 4.6971756397796504e-05, | |
| "loss": 0.1127, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.1316872427983538, | |
| "grad_norm": 0.10398664110407933, | |
| "learning_rate": 4.61572042151878e-05, | |
| "loss": 0.1226, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.139917695473251, | |
| "grad_norm": 0.09339956051495163, | |
| "learning_rate": 4.5347652248824624e-05, | |
| "loss": 0.1099, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.148148148148148, | |
| "grad_norm": 0.09890330904002925, | |
| "learning_rate": 4.4543175679329344e-05, | |
| "loss": 0.1195, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.156378600823045, | |
| "grad_norm": 0.09997598174399476, | |
| "learning_rate": 4.3743849215987595e-05, | |
| "loss": 0.1128, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.1646090534979425, | |
| "grad_norm": 0.09613540261918692, | |
| "learning_rate": 4.294974708981041e-05, | |
| "loss": 0.11, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.09811570643997117, | |
| "learning_rate": 4.216094304664056e-05, | |
| "loss": 0.1221, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.1810699588477367, | |
| "grad_norm": 0.10130508026732313, | |
| "learning_rate": 4.137751034030399e-05, | |
| "loss": 0.1147, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.1893004115226335, | |
| "grad_norm": 0.10155247373344696, | |
| "learning_rate": 4.059952172580694e-05, | |
| "loss": 0.1258, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.197530864197531, | |
| "grad_norm": 0.10114305608112335, | |
| "learning_rate": 3.982704945257957e-05, | |
| "loss": 0.1125, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.205761316872428, | |
| "grad_norm": 0.09834925536076868, | |
| "learning_rate": 3.906016525776611e-05, | |
| "loss": 0.1178, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.213991769547325, | |
| "grad_norm": 0.10693349155074929, | |
| "learning_rate": 3.829894035956306e-05, | |
| "loss": 0.125, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.09467414992649671, | |
| "learning_rate": 3.7543445450605285e-05, | |
| "loss": 0.1054, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.230452674897119, | |
| "grad_norm": 0.09826633722499567, | |
| "learning_rate": 3.6793750691400994e-05, | |
| "loss": 0.1066, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.2386831275720165, | |
| "grad_norm": 0.10348161204951604, | |
| "learning_rate": 3.6049925703816214e-05, | |
| "loss": 0.1144, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.246913580246914, | |
| "grad_norm": 0.10479446585724977, | |
| "learning_rate": 3.53120395646092e-05, | |
| "loss": 0.1135, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.2551440329218106, | |
| "grad_norm": 0.0951654749026979, | |
| "learning_rate": 3.458016079901544e-05, | |
| "loss": 0.1074, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.263374485596708, | |
| "grad_norm": 0.11076793087444725, | |
| "learning_rate": 3.38543573743839e-05, | |
| "loss": 0.1299, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.271604938271605, | |
| "grad_norm": 0.0946857687815803, | |
| "learning_rate": 3.3134696693865316e-05, | |
| "loss": 0.106, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.279835390946502, | |
| "grad_norm": 0.09607169531950183, | |
| "learning_rate": 3.242124559015234e-05, | |
| "loss": 0.0966, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.288065843621399, | |
| "grad_norm": 0.10167592887197249, | |
| "learning_rate": 3.171407031927325e-05, | |
| "loss": 0.1162, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.2962962962962963, | |
| "grad_norm": 0.09726294827804387, | |
| "learning_rate": 3.101323655443882e-05, | |
| "loss": 0.1139, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.3045267489711936, | |
| "grad_norm": 0.09582434907073016, | |
| "learning_rate": 3.031880937994359e-05, | |
| "loss": 0.1084, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.3127572016460904, | |
| "grad_norm": 0.11021289451624465, | |
| "learning_rate": 2.9630853285121508e-05, | |
| "loss": 0.1231, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.3209876543209877, | |
| "grad_norm": 0.09391459178579925, | |
| "learning_rate": 2.894943215835708e-05, | |
| "loss": 0.1047, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.3292181069958846, | |
| "grad_norm": 0.09899700696295954, | |
| "learning_rate": 2.827460928115232e-05, | |
| "loss": 0.1101, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.337448559670782, | |
| "grad_norm": 0.09901505034589045, | |
| "learning_rate": 2.7606447322249872e-05, | |
| "loss": 0.1067, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.3456790123456788, | |
| "grad_norm": 0.10420896589962987, | |
| "learning_rate": 2.6945008331813226e-05, | |
| "loss": 0.1282, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.353909465020576, | |
| "grad_norm": 0.09729393595160353, | |
| "learning_rate": 2.629035373566433e-05, | |
| "loss": 0.1023, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.3621399176954734, | |
| "grad_norm": 0.09829406149572398, | |
| "learning_rate": 2.5642544329579088e-05, | |
| "loss": 0.1155, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.10275430043325989, | |
| "learning_rate": 2.500164027364147e-05, | |
| "loss": 0.1252, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.3786008230452675, | |
| "grad_norm": 0.10367750094719844, | |
| "learning_rate": 2.4367701086656624e-05, | |
| "loss": 0.1166, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.386831275720165, | |
| "grad_norm": 0.10274988439011652, | |
| "learning_rate": 2.3740785640623643e-05, | |
| "loss": 0.1169, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3950617283950617, | |
| "grad_norm": 0.09057411016580416, | |
| "learning_rate": 2.312095215526814e-05, | |
| "loss": 0.1026, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.403292181069959, | |
| "grad_norm": 0.104243770759663, | |
| "learning_rate": 2.2508258192635612e-05, | |
| "loss": 0.1127, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.411522633744856, | |
| "grad_norm": 0.09595429188209968, | |
| "learning_rate": 2.1902760651745958e-05, | |
| "loss": 0.1117, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.419753086419753, | |
| "grad_norm": 0.10143248145662324, | |
| "learning_rate": 2.1304515763309253e-05, | |
| "loss": 0.111, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.42798353909465, | |
| "grad_norm": 0.1074361000716321, | |
| "learning_rate": 2.0713579084503876e-05, | |
| "loss": 0.1222, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.4362139917695473, | |
| "grad_norm": 0.09389635392057138, | |
| "learning_rate": 2.013000549381706e-05, | |
| "loss": 0.1009, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.0950080424907371, | |
| "learning_rate": 1.9553849185948512e-05, | |
| "loss": 0.1111, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.4526748971193415, | |
| "grad_norm": 0.09761821744754258, | |
| "learning_rate": 1.8985163666777473e-05, | |
| "loss": 0.1013, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.460905349794239, | |
| "grad_norm": 0.10939996129330187, | |
| "learning_rate": 1.8424001748393905e-05, | |
| "loss": 0.1215, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.4691358024691357, | |
| "grad_norm": 0.09848524806568702, | |
| "learning_rate": 1.787041554419381e-05, | |
| "loss": 0.1, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.477366255144033, | |
| "grad_norm": 0.10208382321734329, | |
| "learning_rate": 1.7324456464039752e-05, | |
| "loss": 0.1156, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.48559670781893, | |
| "grad_norm": 0.10860830565047937, | |
| "learning_rate": 1.6786175209486566e-05, | |
| "loss": 0.1135, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.493827160493827, | |
| "grad_norm": 0.11254066618708665, | |
| "learning_rate": 1.6255621769072805e-05, | |
| "loss": 0.1202, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.5020576131687244, | |
| "grad_norm": 0.09949710536532337, | |
| "learning_rate": 1.5732845413678477e-05, | |
| "loss": 0.108, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.5102880658436213, | |
| "grad_norm": 0.09871004589964077, | |
| "learning_rate": 1.521789469194952e-05, | |
| "loss": 0.1048, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.5185185185185186, | |
| "grad_norm": 0.09509589762300931, | |
| "learning_rate": 1.4710817425789014e-05, | |
| "loss": 0.1108, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.526748971193416, | |
| "grad_norm": 0.09874336616077671, | |
| "learning_rate": 1.4211660705916285e-05, | |
| "loss": 0.1075, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.5349794238683128, | |
| "grad_norm": 0.10610686628994136, | |
| "learning_rate": 1.3720470887493719e-05, | |
| "loss": 0.114, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.5432098765432096, | |
| "grad_norm": 0.09371419408053047, | |
| "learning_rate": 1.3237293585821786e-05, | |
| "loss": 0.1, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.551440329218107, | |
| "grad_norm": 0.097056043473284, | |
| "learning_rate": 1.2762173672102996e-05, | |
| "loss": 0.1091, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.5596707818930042, | |
| "grad_norm": 0.10289500706688874, | |
| "learning_rate": 1.2295155269274827e-05, | |
| "loss": 0.108, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.10969046879996797, | |
| "learning_rate": 1.1836281747912125e-05, | |
| "loss": 0.1231, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.5761316872427984, | |
| "grad_norm": 0.09910058926653355, | |
| "learning_rate": 1.1385595722199438e-05, | |
| "loss": 0.1132, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.5843621399176957, | |
| "grad_norm": 0.0973576033135528, | |
| "learning_rate": 1.0943139045973549e-05, | |
| "loss": 0.1152, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.5925925925925926, | |
| "grad_norm": 0.10324605327000957, | |
| "learning_rate": 1.050895280883668e-05, | |
| "loss": 0.119, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.60082304526749, | |
| "grad_norm": 0.09940459072200475, | |
| "learning_rate": 1.0083077332340562e-05, | |
| "loss": 0.1086, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.6090534979423867, | |
| "grad_norm": 0.10054320644748665, | |
| "learning_rate": 9.665552166241964e-06, | |
| "loss": 0.1149, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.617283950617284, | |
| "grad_norm": 0.10479612628242368, | |
| "learning_rate": 9.256416084829778e-06, | |
| "loss": 0.1274, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.625514403292181, | |
| "grad_norm": 0.09669311712494659, | |
| "learning_rate": 8.855707083324183e-06, | |
| "loss": 0.1165, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.633744855967078, | |
| "grad_norm": 0.09887270600607351, | |
| "learning_rate": 8.46346237434813e-06, | |
| "loss": 0.1056, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.6419753086419755, | |
| "grad_norm": 0.10455687155682203, | |
| "learning_rate": 8.079718384471557e-06, | |
| "loss": 0.1152, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.6502057613168724, | |
| "grad_norm": 0.10011518680387445, | |
| "learning_rate": 7.704510750828542e-06, | |
| "loss": 0.1056, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.6584362139917697, | |
| "grad_norm": 0.09682701363965382, | |
| "learning_rate": 7.337874317807802e-06, | |
| "loss": 0.1034, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.10000970149912758, | |
| "learning_rate": 6.979843133816743e-06, | |
| "loss": 0.1119, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.674897119341564, | |
| "grad_norm": 0.10465713299866858, | |
| "learning_rate": 6.630450448119618e-06, | |
| "loss": 0.1183, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.6831275720164607, | |
| "grad_norm": 0.10068814150000575, | |
| "learning_rate": 6.289728707749609e-06, | |
| "loss": 0.11, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.691358024691358, | |
| "grad_norm": 0.09310993381960417, | |
| "learning_rate": 5.957709554495683e-06, | |
| "loss": 0.103, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.6995884773662553, | |
| "grad_norm": 0.10472398535370353, | |
| "learning_rate": 5.634423821964074e-06, | |
| "loss": 0.1161, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.707818930041152, | |
| "grad_norm": 0.10645507272996577, | |
| "learning_rate": 5.319901532714877e-06, | |
| "loss": 0.1217, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.7160493827160495, | |
| "grad_norm": 0.10602188133639034, | |
| "learning_rate": 5.014171895473929e-06, | |
| "loss": 0.1144, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.7242798353909468, | |
| "grad_norm": 0.10289659614865782, | |
| "learning_rate": 4.717263302420283e-06, | |
| "loss": 0.1079, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.7325102880658436, | |
| "grad_norm": 0.09777304769239446, | |
| "learning_rate": 4.429203326549525e-06, | |
| "loss": 0.0999, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.7407407407407405, | |
| "grad_norm": 0.1006102230121861, | |
| "learning_rate": 4.1500187191131466e-06, | |
| "loss": 0.099, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.748971193415638, | |
| "grad_norm": 0.10180225031662725, | |
| "learning_rate": 3.879735407134244e-06, | |
| "loss": 0.1173, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.757201646090535, | |
| "grad_norm": 0.10250402250260239, | |
| "learning_rate": 3.6183784909997187e-06, | |
| "loss": 0.1139, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.10615733085722019, | |
| "learning_rate": 3.3659722421293783e-06, | |
| "loss": 0.1133, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.7736625514403292, | |
| "grad_norm": 0.10439855178259218, | |
| "learning_rate": 3.1225401007217936e-06, | |
| "loss": 0.1119, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.7818930041152266, | |
| "grad_norm": 0.10526144305704047, | |
| "learning_rate": 2.8881046735775742e-06, | |
| "loss": 0.1219, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.7901234567901234, | |
| "grad_norm": 0.10521105916873999, | |
| "learning_rate": 2.66268773199988e-06, | |
| "loss": 0.1074, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.7983539094650207, | |
| "grad_norm": 0.10748613106096833, | |
| "learning_rate": 2.446310209772684e-06, | |
| "loss": 0.1128, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8065843621399176, | |
| "grad_norm": 0.09914523804031186, | |
| "learning_rate": 2.2389922012165944e-06, | |
| "loss": 0.1121, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.814814814814815, | |
| "grad_norm": 0.09936886240495578, | |
| "learning_rate": 2.0407529593228116e-06, | |
| "loss": 0.108, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.8230452674897117, | |
| "grad_norm": 0.09150083914451453, | |
| "learning_rate": 1.8516108939651945e-06, | |
| "loss": 0.1033, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.831275720164609, | |
| "grad_norm": 0.10108462806007419, | |
| "learning_rate": 1.6715835701905603e-06, | |
| "loss": 0.1126, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.8395061728395063, | |
| "grad_norm": 0.09387291972239033, | |
| "learning_rate": 1.5006877065874336e-06, | |
| "loss": 0.1033, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.847736625514403, | |
| "grad_norm": 0.10136947076698849, | |
| "learning_rate": 1.3389391737335112e-06, | |
| "loss": 0.1104, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.8559670781893005, | |
| "grad_norm": 0.09974875048963104, | |
| "learning_rate": 1.1863529927217732e-06, | |
| "loss": 0.1043, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.8641975308641974, | |
| "grad_norm": 0.1025752381453182, | |
| "learning_rate": 1.0429433337655115e-06, | |
| "loss": 0.1122, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.8724279835390947, | |
| "grad_norm": 0.10454016619792661, | |
| "learning_rate": 9.087235148824368e-07, | |
| "loss": 0.1184, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.8806584362139915, | |
| "grad_norm": 0.09589319596591109, | |
| "learning_rate": 7.837060006577801e-07, | |
| "loss": 0.1046, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.09451088377127861, | |
| "learning_rate": 6.679024010868618e-07, | |
| "loss": 0.1103, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.897119341563786, | |
| "grad_norm": 0.09770471400992828, | |
| "learning_rate": 5.613234704967996e-07, | |
| "loss": 0.1027, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.905349794238683, | |
| "grad_norm": 0.09543457833989027, | |
| "learning_rate": 4.639791065478738e-07, | |
| "loss": 0.1025, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.9135802469135803, | |
| "grad_norm": 0.10382486799240734, | |
| "learning_rate": 3.758783493142737e-07, | |
| "loss": 0.118, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.9218106995884776, | |
| "grad_norm": 0.1030393593220336, | |
| "learning_rate": 2.9702938044468e-07, | |
| "loss": 0.1179, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.9300411522633745, | |
| "grad_norm": 0.10073287500645238, | |
| "learning_rate": 2.2743952240236176e-07, | |
| "loss": 0.1058, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.9382716049382713, | |
| "grad_norm": 0.10064431575396734, | |
| "learning_rate": 1.6711523778520921e-07, | |
| "loss": 0.1061, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.9465020576131686, | |
| "grad_norm": 0.08907660041993512, | |
| "learning_rate": 1.1606212872559141e-07, | |
| "loss": 0.0985, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.954732510288066, | |
| "grad_norm": 0.10069703961300915, | |
| "learning_rate": 7.428493637002821e-08, | |
| "loss": 0.1107, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.09834578070623724, | |
| "learning_rate": 4.178754043898669e-08, | |
| "loss": 0.1056, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.97119341563786, | |
| "grad_norm": 0.10056360393735715, | |
| "learning_rate": 1.8572958866514e-08, | |
| "loss": 0.1123, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.9794238683127574, | |
| "grad_norm": 0.0943621498377698, | |
| "learning_rate": 4.643347520005836e-09, | |
| "loss": 0.1045, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.9876543209876543, | |
| "grad_norm": 0.09983652481727213, | |
| "learning_rate": 0.0, | |
| "loss": 0.1135, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.9876543209876543, | |
| "eval_loss": 0.13842210173606873, | |
| "eval_runtime": 21.2699, | |
| "eval_samples_per_second": 38.411, | |
| "eval_steps_per_second": 1.222, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.9876543209876543, | |
| "step": 363, | |
| "total_flos": 9.618537139981517e+16, | |
| "train_loss": 0.21143763487742953, | |
| "train_runtime": 3172.33, | |
| "train_samples_per_second": 14.668, | |
| "train_steps_per_second": 0.114 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 363, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 9.618537139981517e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |