| { | |
| "best_metric": 0.279557466506958, | |
| "best_model_checkpoint": "output/output_minicpmv26_upsampled_new/checkpoint-100", | |
| "epoch": 0.9720534629404617, | |
| "eval_steps": 100, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002430133657351154, | |
| "grad_norm": 3.186406373977661, | |
| "learning_rate": 0.0, | |
| "loss": 0.2724, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004860267314702308, | |
| "grad_norm": 3.186406373977661, | |
| "learning_rate": 0.0, | |
| "loss": 0.475, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007290400972053463, | |
| "grad_norm": 3.186406373977661, | |
| "learning_rate": 0.0, | |
| "loss": 0.6319, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.009720534629404616, | |
| "grad_norm": 3.186406373977661, | |
| "learning_rate": 0.0, | |
| "loss": 0.5354, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012150668286755772, | |
| "grad_norm": 9.746641159057617, | |
| "learning_rate": 5.017166594399687e-06, | |
| "loss": 0.6389, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014580801944106925, | |
| "grad_norm": 10.395784378051758, | |
| "learning_rate": 7.952020911994375e-06, | |
| "loss": 0.6399, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01701093560145808, | |
| "grad_norm": 3.7885968685150146, | |
| "learning_rate": 1.0034333188799373e-05, | |
| "loss": 0.3931, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.019441069258809233, | |
| "grad_norm": 3.202871799468994, | |
| "learning_rate": 1.164950007226698e-05, | |
| "loss": 0.4881, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02187120291616039, | |
| "grad_norm": 4.895542144775391, | |
| "learning_rate": 1.2969187506394062e-05, | |
| "loss": 0.3696, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.024301336573511544, | |
| "grad_norm": 2.880216360092163, | |
| "learning_rate": 1.4084967333570947e-05, | |
| "loss": 0.4236, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.026731470230862697, | |
| "grad_norm": 1.7968358993530273, | |
| "learning_rate": 1.505149978319906e-05, | |
| "loss": 0.2569, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02916160388821385, | |
| "grad_norm": 2.6668541431427, | |
| "learning_rate": 1.590404182398875e-05, | |
| "loss": 0.5213, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.031591737545565005, | |
| "grad_norm": 2.0502774715423584, | |
| "learning_rate": 1.666666666666667e-05, | |
| "loss": 0.3246, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03402187120291616, | |
| "grad_norm": 1.6508930921554565, | |
| "learning_rate": 1.7356544752637084e-05, | |
| "loss": 0.1883, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03645200486026731, | |
| "grad_norm": 1.4152283668518066, | |
| "learning_rate": 1.7986354100793748e-05, | |
| "loss": 0.2409, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.038882138517618466, | |
| "grad_norm": 2.176948308944702, | |
| "learning_rate": 1.8565722538447282e-05, | |
| "loss": 0.2042, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.041312272174969626, | |
| "grad_norm": 2.2497684955596924, | |
| "learning_rate": 1.9102133927970633e-05, | |
| "loss": 0.2568, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04374240583232078, | |
| "grad_norm": 2.5578770637512207, | |
| "learning_rate": 1.9601520984261358e-05, | |
| "loss": 0.1914, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.046172539489671933, | |
| "grad_norm": 1.6818372011184692, | |
| "learning_rate": 2.0068666377598747e-05, | |
| "loss": 0.2355, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04860267314702309, | |
| "grad_norm": 1.2788841724395752, | |
| "learning_rate": 2.0507482022971233e-05, | |
| "loss": 0.2149, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05103280680437424, | |
| "grad_norm": 1.4694865942001343, | |
| "learning_rate": 2.0921208418388435e-05, | |
| "loss": 0.1945, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.053462940461725394, | |
| "grad_norm": 1.3222826719284058, | |
| "learning_rate": 2.1312560015880482e-05, | |
| "loss": 0.271, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05589307411907655, | |
| "grad_norm": 1.3958441019058228, | |
| "learning_rate": 2.1683833261066357e-05, | |
| "loss": 0.2217, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0583232077764277, | |
| "grad_norm": 1.6378260850906372, | |
| "learning_rate": 2.2036988245565324e-05, | |
| "loss": 0.235, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.060753341433778855, | |
| "grad_norm": 1.7126753330230713, | |
| "learning_rate": 2.2373711347036773e-05, | |
| "loss": 0.1788, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06318347509113001, | |
| "grad_norm": 1.4920830726623535, | |
| "learning_rate": 2.269546393362655e-05, | |
| "loss": 0.2211, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06561360874848117, | |
| "grad_norm": 1.6582273244857788, | |
| "learning_rate": 2.3003520695193437e-05, | |
| "loss": 0.1942, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06804374240583232, | |
| "grad_norm": 1.3107630014419556, | |
| "learning_rate": 2.329900014453396e-05, | |
| "loss": 0.1218, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07047387606318348, | |
| "grad_norm": 1.492126226425171, | |
| "learning_rate": 2.3582889132846968e-05, | |
| "loss": 0.1473, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07290400972053462, | |
| "grad_norm": 2.5211129188537598, | |
| "learning_rate": 2.3856062735983123e-05, | |
| "loss": 0.3478, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07533414337788578, | |
| "grad_norm": 1.6695655584335327, | |
| "learning_rate": 2.4119300522370322e-05, | |
| "loss": 0.1309, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07776427703523693, | |
| "grad_norm": 1.708465576171875, | |
| "learning_rate": 2.4373299964982603e-05, | |
| "loss": 0.2028, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08019441069258809, | |
| "grad_norm": 1.2873278856277466, | |
| "learning_rate": 2.4618687578661044e-05, | |
| "loss": 0.1629, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08262454434993925, | |
| "grad_norm": 1.6095136404037476, | |
| "learning_rate": 2.4856028230571212e-05, | |
| "loss": 0.2027, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0850546780072904, | |
| "grad_norm": 2.230327844619751, | |
| "learning_rate": 2.5085832971998436e-05, | |
| "loss": 0.2505, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08748481166464156, | |
| "grad_norm": 0.9581132531166077, | |
| "learning_rate": 2.530856566463146e-05, | |
| "loss": 0.1652, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0899149453219927, | |
| "grad_norm": 2.2543814182281494, | |
| "learning_rate": 2.552464861737092e-05, | |
| "loss": 0.2478, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09234507897934387, | |
| "grad_norm": 1.2664082050323486, | |
| "learning_rate": 2.5734467405837933e-05, | |
| "loss": 0.1083, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09477521263669501, | |
| "grad_norm": 1.69247567653656, | |
| "learning_rate": 2.5938375012788124e-05, | |
| "loss": 0.3731, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09720534629404617, | |
| "grad_norm": 1.6350871324539185, | |
| "learning_rate": 2.6136695401116585e-05, | |
| "loss": 0.3767, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09963547995139732, | |
| "grad_norm": 1.7677048444747925, | |
| "learning_rate": 2.6329726610280168e-05, | |
| "loss": 0.1207, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10206561360874848, | |
| "grad_norm": 1.1837761402130127, | |
| "learning_rate": 2.651774345044166e-05, | |
| "loss": 0.119, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10449574726609964, | |
| "grad_norm": 1.1551034450531006, | |
| "learning_rate": 2.6700999855466042e-05, | |
| "loss": 0.1566, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.10692588092345079, | |
| "grad_norm": 1.4947562217712402, | |
| "learning_rate": 2.687973094532893e-05, | |
| "loss": 0.2224, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10935601458080195, | |
| "grad_norm": 1.3775482177734375, | |
| "learning_rate": 2.7054154839965013e-05, | |
| "loss": 0.158, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1117861482381531, | |
| "grad_norm": 1.6749316453933716, | |
| "learning_rate": 2.722447425965978e-05, | |
| "loss": 0.1058, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.11421628189550426, | |
| "grad_norm": 0.8986448049545288, | |
| "learning_rate": 2.739087794143646e-05, | |
| "loss": 0.1299, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1166464155528554, | |
| "grad_norm": 1.2980709075927734, | |
| "learning_rate": 2.755354189625573e-05, | |
| "loss": 0.1029, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11907654921020656, | |
| "grad_norm": 1.271391749382019, | |
| "learning_rate": 2.771263052802624e-05, | |
| "loss": 0.1576, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12150668286755771, | |
| "grad_norm": 1.4220929145812988, | |
| "learning_rate": 2.7868297632261957e-05, | |
| "loss": 0.1338, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12393681652490887, | |
| "grad_norm": 0.962883472442627, | |
| "learning_rate": 2.8020687289593123e-05, | |
| "loss": 0.1546, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12636695018226002, | |
| "grad_norm": 0.8322131633758545, | |
| "learning_rate": 2.8169934667141895e-05, | |
| "loss": 0.1877, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12879708383961117, | |
| "grad_norm": 1.5774791240692139, | |
| "learning_rate": 2.8316166738933646e-05, | |
| "loss": 0.1473, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.13122721749696234, | |
| "grad_norm": 1.2331901788711548, | |
| "learning_rate": 2.845950293496561e-05, | |
| "loss": 0.226, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1336573511543135, | |
| "grad_norm": 2.1382243633270264, | |
| "learning_rate": 2.8600055727246657e-05, | |
| "loss": 0.2057, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.13608748481166463, | |
| "grad_norm": 0.8981903195381165, | |
| "learning_rate": 2.8737931160013153e-05, | |
| "loss": 0.1151, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1385176184690158, | |
| "grad_norm": 2.6175801753997803, | |
| "learning_rate": 2.8873229330382812e-05, | |
| "loss": 0.187, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.14094775212636695, | |
| "grad_norm": 2.0045104026794434, | |
| "learning_rate": 2.9006044824904066e-05, | |
| "loss": 0.1901, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1433778857837181, | |
| "grad_norm": 1.3968229293823242, | |
| "learning_rate": 2.913646711677001e-05, | |
| "loss": 0.1513, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.14580801944106925, | |
| "grad_norm": 8.191553115844727, | |
| "learning_rate": 2.926458092787486e-05, | |
| "loss": 0.1286, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14823815309842042, | |
| "grad_norm": 2.5880401134490967, | |
| "learning_rate": 2.939046655938229e-05, | |
| "loss": 0.1799, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.15066828675577157, | |
| "grad_norm": 1.9319133758544922, | |
| "learning_rate": 2.951420019403574e-05, | |
| "loss": 0.2835, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.15309842041312272, | |
| "grad_norm": 2.653153419494629, | |
| "learning_rate": 2.963585417306073e-05, | |
| "loss": 0.1936, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.15552855407047386, | |
| "grad_norm": 1.4456685781478882, | |
| "learning_rate": 2.9755497250179453e-05, | |
| "loss": 0.2256, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.15795868772782504, | |
| "grad_norm": 1.168137788772583, | |
| "learning_rate": 2.98731948249709e-05, | |
| "loss": 0.1318, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16038882138517618, | |
| "grad_norm": 1.0261298418045044, | |
| "learning_rate": 2.9989009157559694e-05, | |
| "loss": 0.1738, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.16281895504252733, | |
| "grad_norm": 2.1802849769592285, | |
| "learning_rate": 3.010299956639812e-05, | |
| "loss": 0.1304, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1652490886998785, | |
| "grad_norm": 0.9934577941894531, | |
| "learning_rate": 3.021522261071426e-05, | |
| "loss": 0.1472, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.16767922235722965, | |
| "grad_norm": 1.7603391408920288, | |
| "learning_rate": 3.0325732259031143e-05, | |
| "loss": 0.0965, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1701093560145808, | |
| "grad_norm": 1.1827404499053955, | |
| "learning_rate": 3.043458004501377e-05, | |
| "loss": 0.0957, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17253948967193194, | |
| "grad_norm": 1.950810432434082, | |
| "learning_rate": 3.054181521177061e-05, | |
| "loss": 0.1916, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.17496962332928312, | |
| "grad_norm": 1.1083086729049683, | |
| "learning_rate": 3.064748484562093e-05, | |
| "loss": 0.1189, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.17739975698663427, | |
| "grad_norm": 1.8297152519226074, | |
| "learning_rate": 3.0751634000237615e-05, | |
| "loss": 0.1802, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1798298906439854, | |
| "grad_norm": 1.1761589050292969, | |
| "learning_rate": 3.085430581198459e-05, | |
| "loss": 0.1516, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1822600243013366, | |
| "grad_norm": 1.0804104804992676, | |
| "learning_rate": 3.095554160718781e-05, | |
| "loss": 0.1177, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.18469015795868773, | |
| "grad_norm": 1.3176584243774414, | |
| "learning_rate": 3.10553810020076e-05, | |
| "loss": 0.2232, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.18712029161603888, | |
| "grad_norm": 0.8258953094482422, | |
| "learning_rate": 3.115386199551628e-05, | |
| "loss": 0.0939, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.18955042527339003, | |
| "grad_norm": 1.2120673656463623, | |
| "learning_rate": 3.1251021056528336e-05, | |
| "loss": 0.0949, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1919805589307412, | |
| "grad_norm": 2.156952142715454, | |
| "learning_rate": 3.134689320467986e-05, | |
| "loss": 0.1487, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.19441069258809235, | |
| "grad_norm": 1.1182252168655396, | |
| "learning_rate": 3.144151208620804e-05, | |
| "loss": 0.1251, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1968408262454435, | |
| "grad_norm": 0.9124034643173218, | |
| "learning_rate": 3.1534910044841344e-05, | |
| "loss": 0.0881, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.19927095990279464, | |
| "grad_norm": 1.1577218770980835, | |
| "learning_rate": 3.1627118188174024e-05, | |
| "loss": 0.1401, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.20170109356014582, | |
| "grad_norm": 0.8857666254043579, | |
| "learning_rate": 3.171816644986573e-05, | |
| "loss": 0.1202, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.20413122721749696, | |
| "grad_norm": 0.8261206746101379, | |
| "learning_rate": 3.18080836479775e-05, | |
| "loss": 0.1156, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2065613608748481, | |
| "grad_norm": 2.775029182434082, | |
| "learning_rate": 3.1896897539728616e-05, | |
| "loss": 0.1777, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20899149453219928, | |
| "grad_norm": 0.9188308715820312, | |
| "learning_rate": 3.198463487293457e-05, | |
| "loss": 0.0676, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.21142162818955043, | |
| "grad_norm": 1.3610761165618896, | |
| "learning_rate": 3.207132143436469e-05, | |
| "loss": 0.1576, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.21385176184690158, | |
| "grad_norm": 1.8237483501434326, | |
| "learning_rate": 3.215698209523821e-05, | |
| "loss": 0.0858, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.21628189550425272, | |
| "grad_norm": 1.4661478996276855, | |
| "learning_rate": 3.224164085405946e-05, | |
| "loss": 0.1596, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2187120291616039, | |
| "grad_norm": 0.8811361193656921, | |
| "learning_rate": 3.232532087697698e-05, | |
| "loss": 0.1661, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22114216281895505, | |
| "grad_norm": 1.274592638015747, | |
| "learning_rate": 3.240804453583615e-05, | |
| "loss": 0.0793, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2235722964763062, | |
| "grad_norm": 1.125353455543518, | |
| "learning_rate": 3.248983344408188e-05, | |
| "loss": 0.2034, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.22600243013365734, | |
| "grad_norm": 1.1274268627166748, | |
| "learning_rate": 3.2570708490655414e-05, | |
| "loss": 0.117, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2284325637910085, | |
| "grad_norm": 0.8626168370246887, | |
| "learning_rate": 3.265068987201822e-05, | |
| "loss": 0.0469, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.23086269744835966, | |
| "grad_norm": 0.9799898862838745, | |
| "learning_rate": 3.2729797122425925e-05, | |
| "loss": 0.1822, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2332928311057108, | |
| "grad_norm": 1.093001127243042, | |
| "learning_rate": 3.280804914256559e-05, | |
| "loss": 0.1182, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.23572296476306198, | |
| "grad_norm": 0.8328416347503662, | |
| "learning_rate": 3.288546422666164e-05, | |
| "loss": 0.1668, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.23815309842041313, | |
| "grad_norm": 1.4609471559524536, | |
| "learning_rate": 3.2962060088147464e-05, | |
| "loss": 0.1545, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.24058323207776428, | |
| "grad_norm": 1.207360029220581, | |
| "learning_rate": 3.3037853883992805e-05, | |
| "loss": 0.2099, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.24301336573511542, | |
| "grad_norm": 0.9304331541061401, | |
| "learning_rate": 3.3112862237770756e-05, | |
| "loss": 0.0543, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24301336573511542, | |
| "eval_loss": 0.279557466506958, | |
| "eval_runtime": 507.19, | |
| "eval_samples_per_second": 5.418, | |
| "eval_steps_per_second": 0.678, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2454434993924666, | |
| "grad_norm": 0.8421845436096191, | |
| "learning_rate": 3.3187101261541584e-05, | |
| "loss": 0.117, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.24787363304981774, | |
| "grad_norm": 1.1542550325393677, | |
| "learning_rate": 3.326058657662584e-05, | |
| "loss": 0.172, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2503037667071689, | |
| "grad_norm": 1.1323540210723877, | |
| "learning_rate": 3.333333333333334e-05, | |
| "loss": 0.1541, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.25273390036452004, | |
| "grad_norm": 0.9894705414772034, | |
| "learning_rate": 3.340535622971072e-05, | |
| "loss": 0.0689, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2551640340218712, | |
| "grad_norm": 1.0771222114562988, | |
| "learning_rate": 3.3476669529365295e-05, | |
| "loss": 0.1796, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.25759416767922233, | |
| "grad_norm": 0.9119389653205872, | |
| "learning_rate": 3.3547287078419544e-05, | |
| "loss": 0.1092, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2600243013365735, | |
| "grad_norm": 1.194972038269043, | |
| "learning_rate": 3.361722232164634e-05, | |
| "loss": 0.1077, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2624544349939247, | |
| "grad_norm": 1.0005968809127808, | |
| "learning_rate": 3.3686488317832306e-05, | |
| "loss": 0.2024, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2648845686512758, | |
| "grad_norm": 1.0404926538467407, | |
| "learning_rate": 3.375509775441284e-05, | |
| "loss": 0.0672, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.267314702308627, | |
| "grad_norm": 0.9089542031288147, | |
| "learning_rate": 3.382306296142016e-05, | |
| "loss": 0.0831, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.26974483596597815, | |
| "grad_norm": 1.2017402648925781, | |
| "learning_rate": 3.38903959247825e-05, | |
| "loss": 0.1336, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.27217496962332927, | |
| "grad_norm": 0.9206855893135071, | |
| "learning_rate": 3.395710829901039e-05, | |
| "loss": 0.0721, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.27460510328068044, | |
| "grad_norm": 1.4966216087341309, | |
| "learning_rate": 3.402321141930376e-05, | |
| "loss": 0.124, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2770352369380316, | |
| "grad_norm": 0.7973242998123169, | |
| "learning_rate": 3.4088716313110955e-05, | |
| "loss": 0.0597, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.27946537059538273, | |
| "grad_norm": 0.941777765750885, | |
| "learning_rate": 3.415363371116969e-05, | |
| "loss": 0.0737, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2818955042527339, | |
| "grad_norm": 0.9339935183525085, | |
| "learning_rate": 3.4217974058057e-05, | |
| "loss": 0.1308, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.284325637910085, | |
| "grad_norm": 1.0190166234970093, | |
| "learning_rate": 3.428174752227455e-05, | |
| "loss": 0.117, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2867557715674362, | |
| "grad_norm": 1.7908596992492676, | |
| "learning_rate": 3.434496400589353e-05, | |
| "loss": 0.1853, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2891859052247874, | |
| "grad_norm": 1.1210516691207886, | |
| "learning_rate": 3.440763315378198e-05, | |
| "loss": 0.1775, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2916160388821385, | |
| "grad_norm": 0.916373610496521, | |
| "learning_rate": 3.446976436243603e-05, | |
| "loss": 0.0902, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.29404617253948967, | |
| "grad_norm": 1.1219305992126465, | |
| "learning_rate": 3.4531366788435425e-05, | |
| "loss": 0.1742, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.29647630619684084, | |
| "grad_norm": 0.6167263388633728, | |
| "learning_rate": 3.459244935654219e-05, | |
| "loss": 0.0674, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.29890643985419196, | |
| "grad_norm": 1.4466544389724731, | |
| "learning_rate": 3.465302076746041e-05, | |
| "loss": 0.1252, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.30133657351154314, | |
| "grad_norm": 1.1142324209213257, | |
| "learning_rate": 3.471308950527417e-05, | |
| "loss": 0.2289, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3037667071688943, | |
| "grad_norm": 1.2791327238082886, | |
| "learning_rate": 3.477266384457914e-05, | |
| "loss": 0.0885, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.30619684082624543, | |
| "grad_norm": 0.87410968542099, | |
| "learning_rate": 3.48317518573233e-05, | |
| "loss": 0.1188, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3086269744835966, | |
| "grad_norm": 1.3250837326049805, | |
| "learning_rate": 3.489036141937059e-05, | |
| "loss": 0.1572, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3110571081409477, | |
| "grad_norm": 1.2806202173233032, | |
| "learning_rate": 3.494850021680094e-05, | |
| "loss": 0.1576, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3134872417982989, | |
| "grad_norm": 1.1694215536117554, | |
| "learning_rate": 3.500617575195938e-05, | |
| "loss": 0.0828, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3159173754556501, | |
| "grad_norm": 1.1224102973937988, | |
| "learning_rate": 3.5063395349265945e-05, | |
| "loss": 0.197, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3183475091130012, | |
| "grad_norm": 5.477957248687744, | |
| "learning_rate": 3.5120166160797804e-05, | |
| "loss": 0.1175, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.32077764277035237, | |
| "grad_norm": 0.8001928925514221, | |
| "learning_rate": 3.517649517165415e-05, | |
| "loss": 0.0827, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.32320777642770354, | |
| "grad_norm": 0.8747657537460327, | |
| "learning_rate": 3.523238920511395e-05, | |
| "loss": 0.0611, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.32563791008505466, | |
| "grad_norm": 0.7952237725257874, | |
| "learning_rate": 3.528785492759607e-05, | |
| "loss": 0.0875, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.32806804374240583, | |
| "grad_norm": 1.1550030708312988, | |
| "learning_rate": 3.5342898853430836e-05, | |
| "loss": 0.1051, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.330498177399757, | |
| "grad_norm": 1.0629292726516724, | |
| "learning_rate": 3.539752734945143e-05, | |
| "loss": 0.0615, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.33292831105710813, | |
| "grad_norm": 1.4681527614593506, | |
| "learning_rate": 3.5451746639413466e-05, | |
| "loss": 0.1419, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3353584447144593, | |
| "grad_norm": 1.924991488456726, | |
| "learning_rate": 3.550556280825011e-05, | |
| "loss": 0.1506, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3377885783718105, | |
| "grad_norm": 1.485156774520874, | |
| "learning_rate": 3.55589818061703e-05, | |
| "loss": 0.1149, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3402187120291616, | |
| "grad_norm": 1.5923975706100464, | |
| "learning_rate": 3.561200945260678e-05, | |
| "loss": 0.1039, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.34264884568651277, | |
| "grad_norm": 1.22085440158844, | |
| "learning_rate": 3.5664651440020616e-05, | |
| "loss": 0.1192, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3450789793438639, | |
| "grad_norm": 1.2707493305206299, | |
| "learning_rate": 3.571691333756825e-05, | |
| "loss": 0.1605, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.34750911300121506, | |
| "grad_norm": 1.3047083616256714, | |
| "learning_rate": 3.5768800594637304e-05, | |
| "loss": 0.085, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.34993924665856624, | |
| "grad_norm": 1.43068265914917, | |
| "learning_rate": 3.582031854425634e-05, | |
| "loss": 0.1063, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.35236938031591736, | |
| "grad_norm": 1.2630374431610107, | |
| "learning_rate": 3.587147240638428e-05, | |
| "loss": 0.0877, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.35479951397326853, | |
| "grad_norm": 0.7847033143043518, | |
| "learning_rate": 3.5922267291084366e-05, | |
| "loss": 0.0309, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3572296476306197, | |
| "grad_norm": 1.0574724674224854, | |
| "learning_rate": 3.5972708201587496e-05, | |
| "loss": 0.1964, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3596597812879708, | |
| "grad_norm": 1.662292718887329, | |
| "learning_rate": 3.6022800037249585e-05, | |
| "loss": 0.0853, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.362089914945322, | |
| "grad_norm": 2.2652666568756104, | |
| "learning_rate": 3.607254759640729e-05, | |
| "loss": 0.1134, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3645200486026732, | |
| "grad_norm": 2.5058281421661377, | |
| "learning_rate": 3.612195557913627e-05, | |
| "loss": 0.0678, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3669501822600243, | |
| "grad_norm": 1.5377355813980103, | |
| "learning_rate": 3.6171028589915954e-05, | |
| "loss": 0.2186, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.36938031591737547, | |
| "grad_norm": 1.1862763166427612, | |
| "learning_rate": 3.6219771140204575e-05, | |
| "loss": 0.1957, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3718104495747266, | |
| "grad_norm": 0.6389171481132507, | |
| "learning_rate": 3.626818765092802e-05, | |
| "loss": 0.1293, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.37424058323207776, | |
| "grad_norm": 0.7620474696159363, | |
| "learning_rate": 3.6316282454886157e-05, | |
| "loss": 0.1237, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.37667071688942894, | |
| "grad_norm": 2.63171124458313, | |
| "learning_rate": 3.636405979907955e-05, | |
| "loss": 0.1276, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.37910085054678005, | |
| "grad_norm": 1.1460998058319092, | |
| "learning_rate": 3.6411523846959985e-05, | |
| "loss": 0.0713, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.38153098420413123, | |
| "grad_norm": 0.7793575525283813, | |
| "learning_rate": 3.645867868060772e-05, | |
| "loss": 0.0684, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.3839611178614824, | |
| "grad_norm": 0.973029613494873, | |
| "learning_rate": 3.6505528302838193e-05, | |
| "loss": 0.0667, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3863912515188335, | |
| "grad_norm": 2.509524345397949, | |
| "learning_rate": 3.6552076639241027e-05, | |
| "loss": 0.2404, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3888213851761847, | |
| "grad_norm": 1.7816277742385864, | |
| "learning_rate": 3.65983275401539e-05, | |
| "loss": 0.0928, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.39125151883353587, | |
| "grad_norm": 0.7536255121231079, | |
| "learning_rate": 3.664428478257371e-05, | |
| "loss": 0.1489, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.393681652490887, | |
| "grad_norm": 0.8865494132041931, | |
| "learning_rate": 3.668995207200753e-05, | |
| "loss": 0.1395, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.39611178614823817, | |
| "grad_norm": 1.5424951314926147, | |
| "learning_rate": 3.673533304426541e-05, | |
| "loss": 0.2781, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3985419198055893, | |
| "grad_norm": 1.0128049850463867, | |
| "learning_rate": 3.67804312671975e-05, | |
| "loss": 0.161, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.40097205346294046, | |
| "grad_norm": 1.1372522115707397, | |
| "learning_rate": 3.682525024237719e-05, | |
| "loss": 0.1299, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.40340218712029163, | |
| "grad_norm": 1.0910232067108154, | |
| "learning_rate": 3.6869793406732636e-05, | |
| "loss": 0.0693, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.40583232077764275, | |
| "grad_norm": 0.8842087388038635, | |
| "learning_rate": 3.69140641341283e-05, | |
| "loss": 0.136, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.4082624544349939, | |
| "grad_norm": 1.3073561191558838, | |
| "learning_rate": 3.695806573689844e-05, | |
| "loss": 0.0366, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4106925880923451, | |
| "grad_norm": 1.3095803260803223, | |
| "learning_rate": 3.700180146733426e-05, | |
| "loss": 0.1218, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4131227217496962, | |
| "grad_norm": 1.3316633701324463, | |
| "learning_rate": 3.704527451912639e-05, | |
| "loss": 0.14, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4155528554070474, | |
| "grad_norm": 0.7762789130210876, | |
| "learning_rate": 3.708848802876438e-05, | |
| "loss": 0.1055, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.41798298906439857, | |
| "grad_norm": 1.0042897462844849, | |
| "learning_rate": 3.7131445076894564e-05, | |
| "loss": 0.0627, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4204131227217497, | |
| "grad_norm": 2.6669344902038574, | |
| "learning_rate": 3.717414868963791e-05, | |
| "loss": 0.1243, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.42284325637910086, | |
| "grad_norm": 0.6236822605133057, | |
| "learning_rate": 3.721660183986924e-05, | |
| "loss": 0.0497, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.425273390036452, | |
| "grad_norm": 1.0069197416305542, | |
| "learning_rate": 3.725880744845915e-05, | |
| "loss": 0.1257, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.42770352369380316, | |
| "grad_norm": 0.8991190195083618, | |
| "learning_rate": 3.730076838547993e-05, | |
| "loss": 0.1139, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.43013365735115433, | |
| "grad_norm": 1.1606578826904297, | |
| "learning_rate": 3.734248747137666e-05, | |
| "loss": 0.0454, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.43256379100850545, | |
| "grad_norm": 1.6333225965499878, | |
| "learning_rate": 3.738396747810492e-05, | |
| "loss": 0.0885, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4349939246658566, | |
| "grad_norm": 1.1033371686935425, | |
| "learning_rate": 3.7425211130235834e-05, | |
| "loss": 0.139, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4374240583232078, | |
| "grad_norm": 0.5453643202781677, | |
| "learning_rate": 3.7466221106030115e-05, | |
| "loss": 0.0543, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4398541919805589, | |
| "grad_norm": 0.8695247769355774, | |
| "learning_rate": 3.750700003848157e-05, | |
| "loss": 0.1419, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4422843256379101, | |
| "grad_norm": 1.7627193927764893, | |
| "learning_rate": 3.7547550516331555e-05, | |
| "loss": 0.1566, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.44471445929526127, | |
| "grad_norm": 0.7778682708740234, | |
| "learning_rate": 3.75878750850551e-05, | |
| "loss": 0.0559, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4471445929526124, | |
| "grad_norm": 0.8081510663032532, | |
| "learning_rate": 3.7627976247819744e-05, | |
| "loss": 0.0729, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.44957472660996356, | |
| "grad_norm": 0.6429978013038635, | |
| "learning_rate": 3.766785646641792e-05, | |
| "loss": 0.0474, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4520048602673147, | |
| "grad_norm": 1.088581919670105, | |
| "learning_rate": 3.770751816217383e-05, | |
| "loss": 0.0644, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.45443499392466585, | |
| "grad_norm": 1.4708858728408813, | |
| "learning_rate": 3.7746963716825615e-05, | |
| "loss": 0.1651, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.456865127582017, | |
| "grad_norm": 0.8583030700683594, | |
| "learning_rate": 3.778619547338356e-05, | |
| "loss": 0.0822, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.45929526123936815, | |
| "grad_norm": 1.0450993776321411, | |
| "learning_rate": 3.782521573696528e-05, | |
| "loss": 0.0731, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4617253948967193, | |
| "grad_norm": 1.1490970849990845, | |
| "learning_rate": 3.786402677560832e-05, | |
| "loss": 0.1354, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4641555285540705, | |
| "grad_norm": 0.5680958032608032, | |
| "learning_rate": 3.790263082106134e-05, | |
| "loss": 0.0836, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4665856622114216, | |
| "grad_norm": 0.7936691641807556, | |
| "learning_rate": 3.794103006955407e-05, | |
| "loss": 0.0526, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4690157958687728, | |
| "grad_norm": 1.0569026470184326, | |
| "learning_rate": 3.797922668254715e-05, | |
| "loss": 0.1512, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.47144592952612396, | |
| "grad_norm": 1.2363556623458862, | |
| "learning_rate": 3.801722278746213e-05, | |
| "loss": 0.1316, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4738760631834751, | |
| "grad_norm": 1.1622111797332764, | |
| "learning_rate": 3.8055020478392495e-05, | |
| "loss": 0.1432, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.47630619684082626, | |
| "grad_norm": 2.1137237548828125, | |
| "learning_rate": 3.809262181679623e-05, | |
| "loss": 0.1273, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4787363304981774, | |
| "grad_norm": 1.0623483657836914, | |
| "learning_rate": 3.813002883217044e-05, | |
| "loss": 0.1066, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.48116646415552855, | |
| "grad_norm": 1.0300410985946655, | |
| "learning_rate": 3.816724352270863e-05, | |
| "loss": 0.1443, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4835965978128797, | |
| "grad_norm": 0.6886430382728577, | |
| "learning_rate": 3.8204267855941266e-05, | |
| "loss": 0.0969, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.48602673147023084, | |
| "grad_norm": 0.6511848568916321, | |
| "learning_rate": 3.824110376935989e-05, | |
| "loss": 0.0791, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.48602673147023084, | |
| "eval_loss": 0.311313271522522, | |
| "eval_runtime": 505.4888, | |
| "eval_samples_per_second": 5.436, | |
| "eval_steps_per_second": 0.681, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.488456865127582, | |
| "grad_norm": 0.9628679752349854, | |
| "learning_rate": 3.827775317102552e-05, | |
| "loss": 0.0952, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4908869987849332, | |
| "grad_norm": 1.1625686883926392, | |
| "learning_rate": 3.831421794016178e-05, | |
| "loss": 0.1665, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4933171324422843, | |
| "grad_norm": 1.7337137460708618, | |
| "learning_rate": 3.835049992773302e-05, | |
| "loss": 0.141, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.4957472660996355, | |
| "grad_norm": 1.1475183963775635, | |
| "learning_rate": 3.838660095700815e-05, | |
| "loss": 0.0996, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.49817739975698666, | |
| "grad_norm": 0.6650044918060303, | |
| "learning_rate": 3.84225228241104e-05, | |
| "loss": 0.0611, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5006075334143378, | |
| "grad_norm": 0.6770364046096802, | |
| "learning_rate": 3.8458267298553554e-05, | |
| "loss": 0.051, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.503037667071689, | |
| "grad_norm": 1.22215735912323, | |
| "learning_rate": 3.8493836123764984e-05, | |
| "loss": 0.0724, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5054678007290401, | |
| "grad_norm": 1.1168265342712402, | |
| "learning_rate": 3.852923101759591e-05, | |
| "loss": 0.116, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5078979343863913, | |
| "grad_norm": 0.7812952995300293, | |
| "learning_rate": 3.856445367281923e-05, | |
| "loss": 0.0726, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5103280680437424, | |
| "grad_norm": 0.7324075698852539, | |
| "learning_rate": 3.859950575761529e-05, | |
| "loss": 0.0359, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5127582017010935, | |
| "grad_norm": 0.8063955903053284, | |
| "learning_rate": 3.8634388916046025e-05, | |
| "loss": 0.0912, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5151883353584447, | |
| "grad_norm": 4.057889938354492, | |
| "learning_rate": 3.866910476851757e-05, | |
| "loss": 0.2407, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5176184690157959, | |
| "grad_norm": 0.7670718431472778, | |
| "learning_rate": 3.870365491223199e-05, | |
| "loss": 0.058, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.520048602673147, | |
| "grad_norm": 0.717004120349884, | |
| "learning_rate": 3.8738040921628215e-05, | |
| "loss": 0.0308, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5224787363304981, | |
| "grad_norm": 1.2053970098495483, | |
| "learning_rate": 3.877226434881253e-05, | |
| "loss": 0.091, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5249088699878494, | |
| "grad_norm": 1.3508349657058716, | |
| "learning_rate": 3.880632672397897e-05, | |
| "loss": 0.1201, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5273390036452005, | |
| "grad_norm": 0.6895163059234619, | |
| "learning_rate": 3.884022955581985e-05, | |
| "loss": 0.0403, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5297691373025516, | |
| "grad_norm": 0.7722972631454468, | |
| "learning_rate": 3.887397433192676e-05, | |
| "loss": 0.0984, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5321992709599028, | |
| "grad_norm": 0.897186279296875, | |
| "learning_rate": 3.890756251918219e-05, | |
| "loss": 0.1564, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.534629404617254, | |
| "grad_norm": 0.5847256183624268, | |
| "learning_rate": 3.894099556414216e-05, | |
| "loss": 0.0905, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5370595382746051, | |
| "grad_norm": 0.7907549142837524, | |
| "learning_rate": 3.897427489341009e-05, | |
| "loss": 0.075, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5394896719319563, | |
| "grad_norm": 0.8477361798286438, | |
| "learning_rate": 3.900740191400198e-05, | |
| "loss": 0.028, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5419198055893074, | |
| "grad_norm": 1.1732168197631836, | |
| "learning_rate": 3.904037801370344e-05, | |
| "loss": 0.0921, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5443499392466585, | |
| "grad_norm": 0.9721484184265137, | |
| "learning_rate": 3.9073204561418514e-05, | |
| "loss": 0.1273, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5467800729040098, | |
| "grad_norm": 0.6401745676994324, | |
| "learning_rate": 3.9105882907510644e-05, | |
| "loss": 0.0433, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5492102065613609, | |
| "grad_norm": 1.2300125360488892, | |
| "learning_rate": 3.913841438413601e-05, | |
| "loss": 0.0574, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.551640340218712, | |
| "grad_norm": 0.9645660519599915, | |
| "learning_rate": 3.917080030556938e-05, | |
| "loss": 0.0491, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5540704738760632, | |
| "grad_norm": 1.5600403547286987, | |
| "learning_rate": 3.9203041968522716e-05, | |
| "loss": 0.1232, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5565006075334143, | |
| "grad_norm": 1.07868230342865, | |
| "learning_rate": 3.923514065245669e-05, | |
| "loss": 0.0674, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5589307411907655, | |
| "grad_norm": 0.933269739151001, | |
| "learning_rate": 3.926709761988538e-05, | |
| "loss": 0.0753, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5613608748481167, | |
| "grad_norm": 2.12733793258667, | |
| "learning_rate": 3.929891411667424e-05, | |
| "loss": 0.1144, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5637910085054678, | |
| "grad_norm": 1.380554437637329, | |
| "learning_rate": 3.933059137233147e-05, | |
| "loss": 0.0843, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5662211421628189, | |
| "grad_norm": 1.0420360565185547, | |
| "learning_rate": 3.9362130600293214e-05, | |
| "loss": 0.111, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.56865127582017, | |
| "grad_norm": 0.6409225463867188, | |
| "learning_rate": 3.9393532998202405e-05, | |
| "loss": 0.0425, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5710814094775213, | |
| "grad_norm": 0.8021889328956604, | |
| "learning_rate": 3.942479974818166e-05, | |
| "loss": 0.0386, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5735115431348724, | |
| "grad_norm": 1.5499992370605469, | |
| "learning_rate": 3.945593201710032e-05, | |
| "loss": 0.0404, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5759416767922235, | |
| "grad_norm": 0.7606313824653625, | |
| "learning_rate": 3.9486930956835724e-05, | |
| "loss": 0.0789, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5783718104495748, | |
| "grad_norm": 0.9025908708572388, | |
| "learning_rate": 3.951779770452894e-05, | |
| "loss": 0.0617, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5808019441069259, | |
| "grad_norm": 1.1557445526123047, | |
| "learning_rate": 3.954853338283512e-05, | |
| "loss": 0.0751, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.583232077764277, | |
| "grad_norm": 1.5104789733886719, | |
| "learning_rate": 3.9579139100168404e-05, | |
| "loss": 0.0732, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5856622114216282, | |
| "grad_norm": 0.9768268465995789, | |
| "learning_rate": 3.960961595094187e-05, | |
| "loss": 0.0797, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5880923450789793, | |
| "grad_norm": 0.7394477725028992, | |
| "learning_rate": 3.96399650158023e-05, | |
| "loss": 0.0397, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5905224787363305, | |
| "grad_norm": 0.742852509021759, | |
| "learning_rate": 3.96701873618601e-05, | |
| "loss": 0.0742, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5929526123936817, | |
| "grad_norm": 0.7512255311012268, | |
| "learning_rate": 3.970028404291448e-05, | |
| "loss": 0.0281, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5953827460510328, | |
| "grad_norm": 0.6248149871826172, | |
| "learning_rate": 3.9730256099673865e-05, | |
| "loss": 0.0375, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5978128797083839, | |
| "grad_norm": 0.8596628904342651, | |
| "learning_rate": 3.976010455997187e-05, | |
| "loss": 0.1213, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6002430133657352, | |
| "grad_norm": 0.7119196057319641, | |
| "learning_rate": 3.978983043897883e-05, | |
| "loss": 0.0849, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6026731470230863, | |
| "grad_norm": 0.8290873765945435, | |
| "learning_rate": 3.981943473940888e-05, | |
| "loss": 0.0531, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6051032806804374, | |
| "grad_norm": 0.4561799168586731, | |
| "learning_rate": 3.984891845172299e-05, | |
| "loss": 0.0327, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6075334143377886, | |
| "grad_norm": 0.8212061524391174, | |
| "learning_rate": 3.987828255432777e-05, | |
| "loss": 0.0769, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6099635479951397, | |
| "grad_norm": 0.6895563006401062, | |
| "learning_rate": 3.9907528013770276e-05, | |
| "loss": 0.0405, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6123936816524909, | |
| "grad_norm": 1.4649648666381836, | |
| "learning_rate": 3.993665578492894e-05, | |
| "loss": 0.1539, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6148238153098421, | |
| "grad_norm": 1.6070215702056885, | |
| "learning_rate": 3.9965666811200624e-05, | |
| "loss": 0.2167, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6172539489671932, | |
| "grad_norm": 0.9879380464553833, | |
| "learning_rate": 3.999456202468397e-05, | |
| "loss": 0.1, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6196840826245443, | |
| "grad_norm": 0.8622350096702576, | |
| "learning_rate": 4.002334234635907e-05, | |
| "loss": 0.1184, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6221142162818954, | |
| "grad_norm": 1.95242440700531, | |
| "learning_rate": 4.005200868626364e-05, | |
| "loss": 0.1317, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6245443499392467, | |
| "grad_norm": 1.3259199857711792, | |
| "learning_rate": 4.008056194366564e-05, | |
| "loss": 0.1522, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6269744835965978, | |
| "grad_norm": 0.9809350371360779, | |
| "learning_rate": 4.010900300723259e-05, | |
| "loss": 0.0581, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6294046172539489, | |
| "grad_norm": 1.1465263366699219, | |
| "learning_rate": 4.013733275519749e-05, | |
| "loss": 0.1303, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6318347509113001, | |
| "grad_norm": 1.4853742122650146, | |
| "learning_rate": 4.016555205552158e-05, | |
| "loss": 0.0983, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6342648845686513, | |
| "grad_norm": 0.4413992166519165, | |
| "learning_rate": 4.0193661766053834e-05, | |
| "loss": 0.0473, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6366950182260024, | |
| "grad_norm": 0.9577488899230957, | |
| "learning_rate": 4.022166273468753e-05, | |
| "loss": 0.15, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6391251518833536, | |
| "grad_norm": 0.7132108211517334, | |
| "learning_rate": 4.024955579951363e-05, | |
| "loss": 0.1153, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6415552855407047, | |
| "grad_norm": 0.7191299796104431, | |
| "learning_rate": 4.027734178897136e-05, | |
| "loss": 0.0538, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6439854191980559, | |
| "grad_norm": 0.6709555983543396, | |
| "learning_rate": 4.030502152199576e-05, | |
| "loss": 0.1552, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6464155528554071, | |
| "grad_norm": 0.8649526834487915, | |
| "learning_rate": 4.033259580816264e-05, | |
| "loss": 0.1492, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6488456865127582, | |
| "grad_norm": 0.6234789490699768, | |
| "learning_rate": 4.036006544783052e-05, | |
| "loss": 0.0777, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6512758201701093, | |
| "grad_norm": 1.0768671035766602, | |
| "learning_rate": 4.0387431232280135e-05, | |
| "loss": 0.177, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6537059538274606, | |
| "grad_norm": 0.7391580939292908, | |
| "learning_rate": 4.041469394385112e-05, | |
| "loss": 0.1217, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6561360874848117, | |
| "grad_norm": 0.5944250226020813, | |
| "learning_rate": 4.0441854356076257e-05, | |
| "loss": 0.0678, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6585662211421628, | |
| "grad_norm": 0.5040566325187683, | |
| "learning_rate": 4.046891323381315e-05, | |
| "loss": 0.0865, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.660996354799514, | |
| "grad_norm": 1.0286433696746826, | |
| "learning_rate": 4.049587133337347e-05, | |
| "loss": 0.0643, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6634264884568651, | |
| "grad_norm": 1.6537009477615356, | |
| "learning_rate": 4.0522729402649793e-05, | |
| "loss": 0.1008, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6658566221142163, | |
| "grad_norm": 0.7121666669845581, | |
| "learning_rate": 4.0549488181240096e-05, | |
| "loss": 0.0865, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6682867557715675, | |
| "grad_norm": 0.8037539720535278, | |
| "learning_rate": 4.057614840056998e-05, | |
| "loss": 0.0976, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6707168894289186, | |
| "grad_norm": 0.6083033680915833, | |
| "learning_rate": 4.06027107840126e-05, | |
| "loss": 0.1118, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6731470230862697, | |
| "grad_norm": 1.8657127618789673, | |
| "learning_rate": 4.0629176047006474e-05, | |
| "loss": 0.0523, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.675577156743621, | |
| "grad_norm": 0.6102950572967529, | |
| "learning_rate": 4.065554489717105e-05, | |
| "loss": 0.0926, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6780072904009721, | |
| "grad_norm": 0.7026309370994568, | |
| "learning_rate": 4.068181803442029e-05, | |
| "loss": 0.0968, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6804374240583232, | |
| "grad_norm": 0.6937738656997681, | |
| "learning_rate": 4.0707996151074147e-05, | |
| "loss": 0.0554, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6828675577156743, | |
| "grad_norm": 1.4066294431686401, | |
| "learning_rate": 4.073407993196794e-05, | |
| "loss": 0.0964, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6852976913730255, | |
| "grad_norm": 0.6314956545829773, | |
| "learning_rate": 4.076007005455996e-05, | |
| "loss": 0.1161, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6877278250303767, | |
| "grad_norm": 0.9460674524307251, | |
| "learning_rate": 4.0785967189036986e-05, | |
| "loss": 0.044, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6901579586877278, | |
| "grad_norm": 0.7385574579238892, | |
| "learning_rate": 4.0811771998418e-05, | |
| "loss": 0.118, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.692588092345079, | |
| "grad_norm": 0.7021672129631042, | |
| "learning_rate": 4.083748513865602e-05, | |
| "loss": 0.085, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6950182260024301, | |
| "grad_norm": 0.6057882308959961, | |
| "learning_rate": 4.086310725873818e-05, | |
| "loss": 0.0389, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6974483596597812, | |
| "grad_norm": 0.7348142862319946, | |
| "learning_rate": 4.0888639000783966e-05, | |
| "loss": 0.0888, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6998784933171325, | |
| "grad_norm": 0.8555133938789368, | |
| "learning_rate": 4.0914081000141844e-05, | |
| "loss": 0.0917, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7023086269744836, | |
| "grad_norm": 0.8100624084472656, | |
| "learning_rate": 4.0939433885484055e-05, | |
| "loss": 0.0811, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7047387606318347, | |
| "grad_norm": 0.5672865509986877, | |
| "learning_rate": 4.0964698278899874e-05, | |
| "loss": 0.0389, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.707168894289186, | |
| "grad_norm": 0.842689573764801, | |
| "learning_rate": 4.0989874795987185e-05, | |
| "loss": 0.0887, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7095990279465371, | |
| "grad_norm": 1.099148154258728, | |
| "learning_rate": 4.1014964045942465e-05, | |
| "loss": 0.0592, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7120291616038882, | |
| "grad_norm": 1.0394737720489502, | |
| "learning_rate": 4.103996663164927e-05, | |
| "loss": 0.0568, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7144592952612394, | |
| "grad_norm": 0.5482613444328308, | |
| "learning_rate": 4.106488314976513e-05, | |
| "loss": 0.0724, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7168894289185905, | |
| "grad_norm": 0.6032484173774719, | |
| "learning_rate": 4.108971419080698e-05, | |
| "loss": 0.0448, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7193195625759417, | |
| "grad_norm": 0.7295458316802979, | |
| "learning_rate": 4.111446033923516e-05, | |
| "loss": 0.093, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7217496962332929, | |
| "grad_norm": 0.5472877621650696, | |
| "learning_rate": 4.113912217353596e-05, | |
| "loss": 0.0799, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.724179829890644, | |
| "grad_norm": 0.682966411113739, | |
| "learning_rate": 4.116370026630272e-05, | |
| "loss": 0.0575, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7266099635479951, | |
| "grad_norm": 0.4737589657306671, | |
| "learning_rate": 4.118819518431564e-05, | |
| "loss": 0.0949, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7290400972053463, | |
| "grad_norm": 0.6645620465278625, | |
| "learning_rate": 4.121260748862021e-05, | |
| "loss": 0.0233, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7290400972053463, | |
| "eval_loss": 0.3244495689868927, | |
| "eval_runtime": 503.9011, | |
| "eval_samples_per_second": 5.453, | |
| "eval_steps_per_second": 0.683, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7314702308626975, | |
| "grad_norm": 0.6423314809799194, | |
| "learning_rate": 4.123693773460426e-05, | |
| "loss": 0.0774, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.7339003645200486, | |
| "grad_norm": 0.6238884329795837, | |
| "learning_rate": 4.126118647207383e-05, | |
| "loss": 0.0638, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7363304981773997, | |
| "grad_norm": 2.3978679180145264, | |
| "learning_rate": 4.1285354245327715e-05, | |
| "loss": 0.1022, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7387606318347509, | |
| "grad_norm": 0.49952298402786255, | |
| "learning_rate": 4.1309441593230726e-05, | |
| "loss": 0.0477, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.741190765492102, | |
| "grad_norm": 1.5169883966445923, | |
| "learning_rate": 4.133344904928585e-05, | |
| "loss": 0.1601, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7436208991494532, | |
| "grad_norm": 0.5875476002693176, | |
| "learning_rate": 4.1357377141705084e-05, | |
| "loss": 0.0866, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7460510328068044, | |
| "grad_norm": 0.9874062538146973, | |
| "learning_rate": 4.1381226393479236e-05, | |
| "loss": 0.0901, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7484811664641555, | |
| "grad_norm": 1.2504879236221313, | |
| "learning_rate": 4.1404997322446435e-05, | |
| "loss": 0.1227, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7509113001215066, | |
| "grad_norm": 0.6713179349899292, | |
| "learning_rate": 4.142869044135967e-05, | |
| "loss": 0.0347, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7533414337788579, | |
| "grad_norm": 0.8156313896179199, | |
| "learning_rate": 4.145230625795311e-05, | |
| "loss": 0.0548, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.755771567436209, | |
| "grad_norm": 0.9566905498504639, | |
| "learning_rate": 4.14758452750074e-05, | |
| "loss": 0.1255, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7582017010935601, | |
| "grad_norm": 0.8393445611000061, | |
| "learning_rate": 4.149930799041392e-05, | |
| "loss": 0.0587, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7606318347509113, | |
| "grad_norm": 0.637996256351471, | |
| "learning_rate": 4.152269489723788e-05, | |
| "loss": 0.0881, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7630619684082625, | |
| "grad_norm": 0.8390913605690002, | |
| "learning_rate": 4.1546006483780626e-05, | |
| "loss": 0.0881, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7654921020656136, | |
| "grad_norm": 0.7430179715156555, | |
| "learning_rate": 4.156924323364072e-05, | |
| "loss": 0.0409, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7679222357229648, | |
| "grad_norm": 0.7785168886184692, | |
| "learning_rate": 4.1592405625774144e-05, | |
| "loss": 0.0511, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7703523693803159, | |
| "grad_norm": 0.8135663866996765, | |
| "learning_rate": 4.161549413455358e-05, | |
| "loss": 0.0703, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.772782503037667, | |
| "grad_norm": 0.8496614694595337, | |
| "learning_rate": 4.163850922982668e-05, | |
| "loss": 0.1502, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7752126366950183, | |
| "grad_norm": 0.8001265525817871, | |
| "learning_rate": 4.16614513769734e-05, | |
| "loss": 0.0917, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7776427703523694, | |
| "grad_norm": 0.5384124517440796, | |
| "learning_rate": 4.1684321036962526e-05, | |
| "loss": 0.0574, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7800729040097205, | |
| "grad_norm": 0.6082786917686462, | |
| "learning_rate": 4.170711866640721e-05, | |
| "loss": 0.0285, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7825030376670717, | |
| "grad_norm": 0.6169834136962891, | |
| "learning_rate": 4.1729844717619684e-05, | |
| "loss": 0.0529, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7849331713244229, | |
| "grad_norm": 1.1811317205429077, | |
| "learning_rate": 4.17524996386651e-05, | |
| "loss": 0.0449, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.787363304981774, | |
| "grad_norm": 0.7238284945487976, | |
| "learning_rate": 4.177508387341454e-05, | |
| "loss": 0.046, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7897934386391251, | |
| "grad_norm": 1.2236160039901733, | |
| "learning_rate": 4.179759786159719e-05, | |
| "loss": 0.0871, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7922235722964763, | |
| "grad_norm": 0.8143868446350098, | |
| "learning_rate": 4.182004203885172e-05, | |
| "loss": 0.0629, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7946537059538274, | |
| "grad_norm": 0.33017951250076294, | |
| "learning_rate": 4.184241683677687e-05, | |
| "loss": 0.0117, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7970838396111786, | |
| "grad_norm": 0.3505575358867645, | |
| "learning_rate": 4.1864722682981245e-05, | |
| "loss": 0.0158, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7995139732685298, | |
| "grad_norm": 1.4309784173965454, | |
| "learning_rate": 4.188696000113232e-05, | |
| "loss": 0.0492, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8019441069258809, | |
| "grad_norm": 1.023452639579773, | |
| "learning_rate": 4.190912921100477e-05, | |
| "loss": 0.0623, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.804374240583232, | |
| "grad_norm": 0.7057633996009827, | |
| "learning_rate": 4.1931230728527994e-05, | |
| "loss": 0.0468, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8068043742405833, | |
| "grad_norm": 0.8735628724098206, | |
| "learning_rate": 4.195326496583291e-05, | |
| "loss": 0.0709, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8092345078979344, | |
| "grad_norm": 3.259680986404419, | |
| "learning_rate": 4.1975232331298125e-05, | |
| "loss": 0.0491, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8116646415552855, | |
| "grad_norm": 1.0592741966247559, | |
| "learning_rate": 4.1997133229595316e-05, | |
| "loss": 0.044, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8140947752126367, | |
| "grad_norm": 0.5978744029998779, | |
| "learning_rate": 4.201896806173394e-05, | |
| "loss": 0.0563, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8165249088699879, | |
| "grad_norm": 0.8989129662513733, | |
| "learning_rate": 4.2040737225105335e-05, | |
| "loss": 0.0582, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.818955042527339, | |
| "grad_norm": 0.9518970251083374, | |
| "learning_rate": 4.206244111352608e-05, | |
| "loss": 0.0764, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.8213851761846902, | |
| "grad_norm": 0.48601099848747253, | |
| "learning_rate": 4.2084080117280756e-05, | |
| "loss": 0.0495, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8238153098420413, | |
| "grad_norm": 0.6095461249351501, | |
| "learning_rate": 4.210565462316407e-05, | |
| "loss": 0.0234, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8262454434993924, | |
| "grad_norm": 0.6459489464759827, | |
| "learning_rate": 4.2127165014522315e-05, | |
| "loss": 0.0236, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8286755771567437, | |
| "grad_norm": 0.8500393629074097, | |
| "learning_rate": 4.214861167129425e-05, | |
| "loss": 0.1073, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.8311057108140948, | |
| "grad_norm": 0.7379801869392395, | |
| "learning_rate": 4.2169994970051365e-05, | |
| "loss": 0.0779, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.8335358444714459, | |
| "grad_norm": 0.7911482453346252, | |
| "learning_rate": 4.219131528403759e-05, | |
| "loss": 0.083, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.8359659781287971, | |
| "grad_norm": 1.379714012145996, | |
| "learning_rate": 4.22125729832083e-05, | |
| "loss": 0.0527, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.8383961117861483, | |
| "grad_norm": 0.7738386988639832, | |
| "learning_rate": 4.2233768434268914e-05, | |
| "loss": 0.0439, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8408262454434994, | |
| "grad_norm": 0.9524173140525818, | |
| "learning_rate": 4.225490200071284e-05, | |
| "loss": 0.0635, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.8432563791008505, | |
| "grad_norm": 0.843377947807312, | |
| "learning_rate": 4.227597404285883e-05, | |
| "loss": 0.0531, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.8456865127582017, | |
| "grad_norm": 2.120123863220215, | |
| "learning_rate": 4.229698491788791e-05, | |
| "loss": 0.068, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.8481166464155528, | |
| "grad_norm": 0.804137647151947, | |
| "learning_rate": 4.231793497987961e-05, | |
| "loss": 0.0577, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.850546780072904, | |
| "grad_norm": 0.8609597086906433, | |
| "learning_rate": 4.2338824579847904e-05, | |
| "loss": 0.0204, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8529769137302552, | |
| "grad_norm": 0.7206110954284668, | |
| "learning_rate": 4.235965406577636e-05, | |
| "loss": 0.0682, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8554070473876063, | |
| "grad_norm": 0.8393117785453796, | |
| "learning_rate": 4.2380423782653e-05, | |
| "loss": 0.0442, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8578371810449574, | |
| "grad_norm": 0.5761812329292297, | |
| "learning_rate": 4.240113407250459e-05, | |
| "loss": 0.0985, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8602673147023087, | |
| "grad_norm": 0.6769473552703857, | |
| "learning_rate": 4.24217852744304e-05, | |
| "loss": 0.0347, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8626974483596598, | |
| "grad_norm": 0.6801010370254517, | |
| "learning_rate": 4.244237772463552e-05, | |
| "loss": 0.0541, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8651275820170109, | |
| "grad_norm": 0.9833145141601562, | |
| "learning_rate": 4.246291175646371e-05, | |
| "loss": 0.0909, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8675577156743621, | |
| "grad_norm": 0.8934769034385681, | |
| "learning_rate": 4.24833877004298e-05, | |
| "loss": 0.0769, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8699878493317132, | |
| "grad_norm": 0.523007869720459, | |
| "learning_rate": 4.250380588425157e-05, | |
| "loss": 0.0413, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8724179829890644, | |
| "grad_norm": 2.014488935470581, | |
| "learning_rate": 4.2524166632881255e-05, | |
| "loss": 0.0651, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8748481166464156, | |
| "grad_norm": 0.8834489583969116, | |
| "learning_rate": 4.254447026853656e-05, | |
| "loss": 0.0288, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8772782503037667, | |
| "grad_norm": 0.8947015404701233, | |
| "learning_rate": 4.2564717110731244e-05, | |
| "loss": 0.1693, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8797083839611178, | |
| "grad_norm": 1.9122964143753052, | |
| "learning_rate": 4.258490747630532e-05, | |
| "loss": 0.0738, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8821385176184691, | |
| "grad_norm": 0.8690800070762634, | |
| "learning_rate": 4.260504167945479e-05, | |
| "loss": 0.0472, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8845686512758202, | |
| "grad_norm": 0.7206094861030579, | |
| "learning_rate": 4.2625120031760965e-05, | |
| "loss": 0.0881, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8869987849331713, | |
| "grad_norm": 0.6016607284545898, | |
| "learning_rate": 4.264514284221944e-05, | |
| "loss": 0.0403, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8894289185905225, | |
| "grad_norm": 0.8117083311080933, | |
| "learning_rate": 4.266511041726854e-05, | |
| "loss": 0.0533, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8918590522478737, | |
| "grad_norm": 0.7658905982971191, | |
| "learning_rate": 4.26850230608176e-05, | |
| "loss": 0.067, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8942891859052248, | |
| "grad_norm": 0.8042786717414856, | |
| "learning_rate": 4.2704881074274584e-05, | |
| "loss": 0.08, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8967193195625759, | |
| "grad_norm": 0.5545147657394409, | |
| "learning_rate": 4.272468475657351e-05, | |
| "loss": 0.0534, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8991494532199271, | |
| "grad_norm": 0.7363461256027222, | |
| "learning_rate": 4.2744434404201497e-05, | |
| "loss": 0.0394, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9015795868772782, | |
| "grad_norm": 0.5137555003166199, | |
| "learning_rate": 4.27641303112253e-05, | |
| "loss": 0.0602, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.9040097205346294, | |
| "grad_norm": 0.7932950258255005, | |
| "learning_rate": 4.278377276931767e-05, | |
| "loss": 0.0711, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9064398541919806, | |
| "grad_norm": 2.459850788116455, | |
| "learning_rate": 4.2803362067783256e-05, | |
| "loss": 0.1969, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9088699878493317, | |
| "grad_norm": 0.6758642792701721, | |
| "learning_rate": 4.2822898493584104e-05, | |
| "loss": 0.1131, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9113001215066828, | |
| "grad_norm": 0.6686858534812927, | |
| "learning_rate": 4.284238233136496e-05, | |
| "loss": 0.0951, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.913730255164034, | |
| "grad_norm": 0.8299110531806946, | |
| "learning_rate": 4.286181386347813e-05, | |
| "loss": 0.046, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.9161603888213852, | |
| "grad_norm": 2.4468648433685303, | |
| "learning_rate": 4.288119337000801e-05, | |
| "loss": 0.222, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.9185905224787363, | |
| "grad_norm": 1.187517523765564, | |
| "learning_rate": 4.2900521128795315e-05, | |
| "loss": 0.1061, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9210206561360875, | |
| "grad_norm": 0.6776664853096008, | |
| "learning_rate": 4.291979741546102e-05, | |
| "loss": 0.0287, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.9234507897934386, | |
| "grad_norm": 0.6103045344352722, | |
| "learning_rate": 4.293902250342989e-05, | |
| "loss": 0.0439, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9258809234507898, | |
| "grad_norm": 0.44634753465652466, | |
| "learning_rate": 4.295819666395376e-05, | |
| "loss": 0.03, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.928311057108141, | |
| "grad_norm": 0.8639276027679443, | |
| "learning_rate": 4.297732016613454e-05, | |
| "loss": 0.1479, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9307411907654921, | |
| "grad_norm": 0.7611154913902283, | |
| "learning_rate": 4.299639327694684e-05, | |
| "loss": 0.0403, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.9331713244228432, | |
| "grad_norm": 0.8710222244262695, | |
| "learning_rate": 4.3015416261260325e-05, | |
| "loss": 0.0522, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.9356014580801945, | |
| "grad_norm": 0.7666921615600586, | |
| "learning_rate": 4.303438938186182e-05, | |
| "loss": 0.0303, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9380315917375456, | |
| "grad_norm": 0.5829209089279175, | |
| "learning_rate": 4.305331289947705e-05, | |
| "loss": 0.0315, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.9404617253948967, | |
| "grad_norm": 0.9368737936019897, | |
| "learning_rate": 4.3072187072792184e-05, | |
| "loss": 0.0661, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.9428918590522479, | |
| "grad_norm": 0.43866387009620667, | |
| "learning_rate": 4.309101215847502e-05, | |
| "loss": 0.0335, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.945321992709599, | |
| "grad_norm": 0.6128959059715271, | |
| "learning_rate": 4.3109788411195924e-05, | |
| "loss": 0.108, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.9477521263669502, | |
| "grad_norm": 0.5031237602233887, | |
| "learning_rate": 4.312851608364853e-05, | |
| "loss": 0.0332, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9501822600243013, | |
| "grad_norm": 0.6804950833320618, | |
| "learning_rate": 4.314719542657013e-05, | |
| "loss": 0.0259, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.9526123936816525, | |
| "grad_norm": 0.9872898459434509, | |
| "learning_rate": 4.3165826688761796e-05, | |
| "loss": 0.0281, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.9550425273390036, | |
| "grad_norm": 1.9561671018600464, | |
| "learning_rate": 4.318441011710833e-05, | |
| "loss": 0.0342, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.9574726609963548, | |
| "grad_norm": 0.6444841027259827, | |
| "learning_rate": 4.3202945956597786e-05, | |
| "loss": 0.035, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.959902794653706, | |
| "grad_norm": 0.49825266003608704, | |
| "learning_rate": 4.3221434450340956e-05, | |
| "loss": 0.0248, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9623329283110571, | |
| "grad_norm": 0.548538863658905, | |
| "learning_rate": 4.323987583959045e-05, | |
| "loss": 0.076, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.9647630619684082, | |
| "grad_norm": 0.6437348127365112, | |
| "learning_rate": 4.325827036375957e-05, | |
| "loss": 0.0749, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.9671931956257594, | |
| "grad_norm": 0.5854353308677673, | |
| "learning_rate": 4.327661826044101e-05, | |
| "loss": 0.0413, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9696233292831106, | |
| "grad_norm": 1.0665239095687866, | |
| "learning_rate": 4.329491976542521e-05, | |
| "loss": 0.0574, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.9720534629404617, | |
| "grad_norm": 0.7473851442337036, | |
| "learning_rate": 4.331317511271859e-05, | |
| "loss": 0.1447, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9720534629404617, | |
| "eval_loss": 0.31331515312194824, | |
| "eval_runtime": 503.3031, | |
| "eval_samples_per_second": 5.46, | |
| "eval_steps_per_second": 0.683, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9720534629404617, | |
| "step": 400, | |
| "total_flos": 1.1378390115664527e+18, | |
| "train_loss": 0.11899106367724016, | |
| "train_runtime": 12503.2939, | |
| "train_samples_per_second": 25.593, | |
| "train_steps_per_second": 0.8 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 100, | |
| "total_flos": 1.1378390115664527e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |