Malaysian-gemma-3-1b-it
/
lora-embedding-128-gemma3-1b-malaysian-8k
/checkpoint-200
/trainer_state.json
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.08378718056137412, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00041893590280687055, | |
| "grad_norm": 8.333969116210938, | |
| "learning_rate": 2e-05, | |
| "loss": 3.3261, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008378718056137411, | |
| "grad_norm": 7.348582744598389, | |
| "learning_rate": 1.9998324256388776e-05, | |
| "loss": 3.2312, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0012568077084206116, | |
| "grad_norm": 3.1899449825286865, | |
| "learning_rate": 1.9996648512777547e-05, | |
| "loss": 2.9318, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0016757436112274822, | |
| "grad_norm": 2.177359104156494, | |
| "learning_rate": 1.9994972769166317e-05, | |
| "loss": 2.8876, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0020946795140343527, | |
| "grad_norm": 2.1708009243011475, | |
| "learning_rate": 1.999329702555509e-05, | |
| "loss": 3.068, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0025136154168412233, | |
| "grad_norm": 1.77362060546875, | |
| "learning_rate": 1.9991621281943865e-05, | |
| "loss": 2.6598, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.002932551319648094, | |
| "grad_norm": 1.709831714630127, | |
| "learning_rate": 1.9989945538332636e-05, | |
| "loss": 2.767, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0033514872224549644, | |
| "grad_norm": 1.5493128299713135, | |
| "learning_rate": 1.998826979472141e-05, | |
| "loss": 2.7182, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.003770423125261835, | |
| "grad_norm": 1.39557945728302, | |
| "learning_rate": 1.998659405111018e-05, | |
| "loss": 2.8695, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0041893590280687055, | |
| "grad_norm": 1.1357755661010742, | |
| "learning_rate": 1.9984918307498955e-05, | |
| "loss": 2.5497, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004608294930875576, | |
| "grad_norm": 1.0683544874191284, | |
| "learning_rate": 1.9983242563887726e-05, | |
| "loss": 2.7363, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.005027230833682447, | |
| "grad_norm": 0.9199109673500061, | |
| "learning_rate": 1.99815668202765e-05, | |
| "loss": 2.4438, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005446166736489317, | |
| "grad_norm": 0.9708887338638306, | |
| "learning_rate": 1.997989107666527e-05, | |
| "loss": 2.6098, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.005865102639296188, | |
| "grad_norm": 0.9026995301246643, | |
| "learning_rate": 1.9978215333054045e-05, | |
| "loss": 2.4444, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.006284038542103058, | |
| "grad_norm": 1.2418183088302612, | |
| "learning_rate": 1.9976539589442816e-05, | |
| "loss": 2.5436, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.006702974444909929, | |
| "grad_norm": 0.8761052489280701, | |
| "learning_rate": 1.997486384583159e-05, | |
| "loss": 2.3926, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.007121910347716799, | |
| "grad_norm": 0.8849633932113647, | |
| "learning_rate": 1.997318810222036e-05, | |
| "loss": 2.5935, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00754084625052367, | |
| "grad_norm": 0.8368175029754639, | |
| "learning_rate": 1.9971512358609135e-05, | |
| "loss": 2.4684, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.007959782153330541, | |
| "grad_norm": 0.697807788848877, | |
| "learning_rate": 1.9969836614997905e-05, | |
| "loss": 2.3537, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.008378718056137411, | |
| "grad_norm": 0.7476556301116943, | |
| "learning_rate": 1.996816087138668e-05, | |
| "loss": 2.4659, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008797653958944282, | |
| "grad_norm": 0.7885666489601135, | |
| "learning_rate": 1.9966485127775454e-05, | |
| "loss": 2.4466, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.009216589861751152, | |
| "grad_norm": 0.7271686792373657, | |
| "learning_rate": 1.9964809384164224e-05, | |
| "loss": 2.3659, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.009635525764558023, | |
| "grad_norm": 0.7286465764045715, | |
| "learning_rate": 1.9963133640552995e-05, | |
| "loss": 2.4473, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.010054461667364893, | |
| "grad_norm": 0.8572853207588196, | |
| "learning_rate": 1.996145789694177e-05, | |
| "loss": 2.3595, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.010473397570171765, | |
| "grad_norm": 0.8283334374427795, | |
| "learning_rate": 1.9959782153330543e-05, | |
| "loss": 2.5291, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.010892333472978634, | |
| "grad_norm": 0.6586313843727112, | |
| "learning_rate": 1.9958106409719314e-05, | |
| "loss": 2.3985, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.011311269375785506, | |
| "grad_norm": 0.6430657505989075, | |
| "learning_rate": 1.9956430666108085e-05, | |
| "loss": 2.3578, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.011730205278592375, | |
| "grad_norm": 0.6550448536872864, | |
| "learning_rate": 1.995475492249686e-05, | |
| "loss": 2.4077, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.012149141181399247, | |
| "grad_norm": 0.7592840194702148, | |
| "learning_rate": 1.9953079178885633e-05, | |
| "loss": 2.5008, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.012568077084206116, | |
| "grad_norm": 0.7858672738075256, | |
| "learning_rate": 1.9951403435274407e-05, | |
| "loss": 2.4674, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.012987012987012988, | |
| "grad_norm": 0.6130352020263672, | |
| "learning_rate": 1.9949727691663178e-05, | |
| "loss": 2.3526, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.013405948889819858, | |
| "grad_norm": 0.6684207320213318, | |
| "learning_rate": 1.994805194805195e-05, | |
| "loss": 2.3732, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.013824884792626729, | |
| "grad_norm": 0.8275600671768188, | |
| "learning_rate": 1.9946376204440723e-05, | |
| "loss": 2.135, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.014243820695433599, | |
| "grad_norm": 0.5858725309371948, | |
| "learning_rate": 1.9944700460829494e-05, | |
| "loss": 2.1368, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01466275659824047, | |
| "grad_norm": 0.7133444547653198, | |
| "learning_rate": 1.9943024717218268e-05, | |
| "loss": 2.304, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01508169250104734, | |
| "grad_norm": 0.5466803312301636, | |
| "learning_rate": 1.994134897360704e-05, | |
| "loss": 2.1682, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.015500628403854211, | |
| "grad_norm": 0.5196086168289185, | |
| "learning_rate": 1.9939673229995813e-05, | |
| "loss": 2.1546, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.015919564306661083, | |
| "grad_norm": 0.5088497400283813, | |
| "learning_rate": 1.9937997486384583e-05, | |
| "loss": 2.1018, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.016338500209467952, | |
| "grad_norm": 0.6117899417877197, | |
| "learning_rate": 1.9936321742773358e-05, | |
| "loss": 2.2346, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.016757436112274822, | |
| "grad_norm": 0.5710458159446716, | |
| "learning_rate": 1.993464599916213e-05, | |
| "loss": 2.2147, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01717637201508169, | |
| "grad_norm": 0.5152861475944519, | |
| "learning_rate": 1.9932970255550902e-05, | |
| "loss": 2.2716, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.017595307917888565, | |
| "grad_norm": 0.6851192712783813, | |
| "learning_rate": 1.9931294511939673e-05, | |
| "loss": 2.3158, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.018014243820695434, | |
| "grad_norm": 0.5485531687736511, | |
| "learning_rate": 1.9929618768328447e-05, | |
| "loss": 2.2679, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.018433179723502304, | |
| "grad_norm": 0.48592010140419006, | |
| "learning_rate": 1.992794302471722e-05, | |
| "loss": 2.1303, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.018852115626309174, | |
| "grad_norm": 0.5533665418624878, | |
| "learning_rate": 1.9926267281105992e-05, | |
| "loss": 2.1981, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.019271051529116047, | |
| "grad_norm": 0.5932656526565552, | |
| "learning_rate": 1.9924591537494763e-05, | |
| "loss": 2.3737, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.019689987431922917, | |
| "grad_norm": 0.5236673951148987, | |
| "learning_rate": 1.9922915793883537e-05, | |
| "loss": 2.2694, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.020108923334729786, | |
| "grad_norm": 0.5357316732406616, | |
| "learning_rate": 1.992124005027231e-05, | |
| "loss": 2.2368, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.020527859237536656, | |
| "grad_norm": 0.5500349998474121, | |
| "learning_rate": 1.9919564306661082e-05, | |
| "loss": 2.213, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.02094679514034353, | |
| "grad_norm": 0.48040810227394104, | |
| "learning_rate": 1.9917888563049853e-05, | |
| "loss": 2.1892, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0213657310431504, | |
| "grad_norm": 0.5716186165809631, | |
| "learning_rate": 1.9916212819438627e-05, | |
| "loss": 2.2039, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.02178466694595727, | |
| "grad_norm": 0.5564374923706055, | |
| "learning_rate": 1.99145370758274e-05, | |
| "loss": 2.1411, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.022203602848764138, | |
| "grad_norm": 0.4996980130672455, | |
| "learning_rate": 1.9912861332216175e-05, | |
| "loss": 2.1521, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.02262253875157101, | |
| "grad_norm": 0.5239240527153015, | |
| "learning_rate": 1.9911185588604946e-05, | |
| "loss": 2.0742, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.02304147465437788, | |
| "grad_norm": 0.4403076767921448, | |
| "learning_rate": 1.9909509844993716e-05, | |
| "loss": 1.9841, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.02346041055718475, | |
| "grad_norm": 0.5169032216072083, | |
| "learning_rate": 1.990783410138249e-05, | |
| "loss": 2.0327, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.02387934645999162, | |
| "grad_norm": 0.4901898503303528, | |
| "learning_rate": 1.9906158357771265e-05, | |
| "loss": 2.0063, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.024298282362798494, | |
| "grad_norm": 0.6581910252571106, | |
| "learning_rate": 1.9904482614160035e-05, | |
| "loss": 2.1385, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.024717218265605363, | |
| "grad_norm": 0.4522070586681366, | |
| "learning_rate": 1.9902806870548806e-05, | |
| "loss": 1.9944, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.025136154168412233, | |
| "grad_norm": 0.5315820574760437, | |
| "learning_rate": 1.990113112693758e-05, | |
| "loss": 2.1579, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.025555090071219103, | |
| "grad_norm": 0.4661259353160858, | |
| "learning_rate": 1.9899455383326354e-05, | |
| "loss": 2.1193, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.025974025974025976, | |
| "grad_norm": 0.4940222203731537, | |
| "learning_rate": 1.9897779639715125e-05, | |
| "loss": 2.0844, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.026392961876832845, | |
| "grad_norm": 0.46520665287971497, | |
| "learning_rate": 1.98961038961039e-05, | |
| "loss": 1.9306, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.026811897779639715, | |
| "grad_norm": 0.5645989179611206, | |
| "learning_rate": 1.989442815249267e-05, | |
| "loss": 2.1236, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.027230833682446585, | |
| "grad_norm": 0.47880157828330994, | |
| "learning_rate": 1.989275240888144e-05, | |
| "loss": 2.0206, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.027649769585253458, | |
| "grad_norm": 0.6371349692344666, | |
| "learning_rate": 1.9891076665270215e-05, | |
| "loss": 2.019, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.028068705488060328, | |
| "grad_norm": 0.5742272734642029, | |
| "learning_rate": 1.988940092165899e-05, | |
| "loss": 2.0899, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.028487641390867197, | |
| "grad_norm": 0.5579768419265747, | |
| "learning_rate": 1.988772517804776e-05, | |
| "loss": 2.081, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.028906577293674067, | |
| "grad_norm": 0.5897182822227478, | |
| "learning_rate": 1.988604943443653e-05, | |
| "loss": 1.9601, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.02932551319648094, | |
| "grad_norm": 0.46881428360939026, | |
| "learning_rate": 1.9884373690825305e-05, | |
| "loss": 1.9085, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02974444909928781, | |
| "grad_norm": 0.6095844507217407, | |
| "learning_rate": 1.988269794721408e-05, | |
| "loss": 1.9762, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.03016338500209468, | |
| "grad_norm": 0.599513053894043, | |
| "learning_rate": 1.988102220360285e-05, | |
| "loss": 1.8723, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.03058232090490155, | |
| "grad_norm": 0.585457980632782, | |
| "learning_rate": 1.987934645999162e-05, | |
| "loss": 1.9209, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.031001256807708422, | |
| "grad_norm": 0.42224225401878357, | |
| "learning_rate": 1.9877670716380394e-05, | |
| "loss": 1.9186, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.03142019271051529, | |
| "grad_norm": 0.4566991329193115, | |
| "learning_rate": 1.987599497276917e-05, | |
| "loss": 2.018, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.031839128613322165, | |
| "grad_norm": 0.47718995809555054, | |
| "learning_rate": 1.9874319229157943e-05, | |
| "loss": 2.0119, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.03225806451612903, | |
| "grad_norm": 0.4412285089492798, | |
| "learning_rate": 1.9872643485546713e-05, | |
| "loss": 1.9211, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.032677000418935905, | |
| "grad_norm": 0.4711454212665558, | |
| "learning_rate": 1.9870967741935484e-05, | |
| "loss": 1.9051, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.03309593632174277, | |
| "grad_norm": 0.4665948450565338, | |
| "learning_rate": 1.9869291998324258e-05, | |
| "loss": 1.9571, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.033514872224549644, | |
| "grad_norm": 0.46011775732040405, | |
| "learning_rate": 1.9867616254713032e-05, | |
| "loss": 1.9599, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03393380812735652, | |
| "grad_norm": 0.46272069215774536, | |
| "learning_rate": 1.9865940511101803e-05, | |
| "loss": 1.9161, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.03435274403016338, | |
| "grad_norm": 0.5554195046424866, | |
| "learning_rate": 1.9864264767490574e-05, | |
| "loss": 2.0202, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.034771679932970256, | |
| "grad_norm": 0.5324104428291321, | |
| "learning_rate": 1.9862589023879348e-05, | |
| "loss": 1.9356, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.03519061583577713, | |
| "grad_norm": 0.5279750823974609, | |
| "learning_rate": 1.9860913280268122e-05, | |
| "loss": 1.9511, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.035609551738583996, | |
| "grad_norm": 0.5002080202102661, | |
| "learning_rate": 1.9859237536656893e-05, | |
| "loss": 1.9248, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.03602848764139087, | |
| "grad_norm": 0.5625497102737427, | |
| "learning_rate": 1.9857561793045667e-05, | |
| "loss": 2.0023, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.036447423544197735, | |
| "grad_norm": 0.6030247807502747, | |
| "learning_rate": 1.9855886049434438e-05, | |
| "loss": 1.893, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.03686635944700461, | |
| "grad_norm": 0.4760509729385376, | |
| "learning_rate": 1.9854210305823212e-05, | |
| "loss": 1.8902, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.03728529534981148, | |
| "grad_norm": 0.6618624925613403, | |
| "learning_rate": 1.9852534562211983e-05, | |
| "loss": 1.9173, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.03770423125261835, | |
| "grad_norm": 0.47204822301864624, | |
| "learning_rate": 1.9850858818600757e-05, | |
| "loss": 1.9266, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03812316715542522, | |
| "grad_norm": 0.5421533584594727, | |
| "learning_rate": 1.9849183074989527e-05, | |
| "loss": 1.9796, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.038542103058232094, | |
| "grad_norm": 0.48972201347351074, | |
| "learning_rate": 1.98475073313783e-05, | |
| "loss": 1.91, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.03896103896103896, | |
| "grad_norm": 0.5566658973693848, | |
| "learning_rate": 1.9845831587767072e-05, | |
| "loss": 1.8992, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.03937997486384583, | |
| "grad_norm": 0.4685937464237213, | |
| "learning_rate": 1.9844155844155846e-05, | |
| "loss": 1.9231, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0397989107666527, | |
| "grad_norm": 0.6744531393051147, | |
| "learning_rate": 1.9842480100544617e-05, | |
| "loss": 1.9109, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.04021784666945957, | |
| "grad_norm": 0.6984325051307678, | |
| "learning_rate": 1.984080435693339e-05, | |
| "loss": 1.9566, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.040636782572266446, | |
| "grad_norm": 0.6627328991889954, | |
| "learning_rate": 1.9839128613322162e-05, | |
| "loss": 1.9933, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.04105571847507331, | |
| "grad_norm": 0.4586343765258789, | |
| "learning_rate": 1.9837452869710936e-05, | |
| "loss": 1.7939, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.041474654377880185, | |
| "grad_norm": 0.6211162805557251, | |
| "learning_rate": 1.983577712609971e-05, | |
| "loss": 1.9164, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.04189359028068706, | |
| "grad_norm": 0.9397639632225037, | |
| "learning_rate": 1.983410138248848e-05, | |
| "loss": 2.0262, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.042312526183493925, | |
| "grad_norm": 0.7698065638542175, | |
| "learning_rate": 1.9832425638877252e-05, | |
| "loss": 1.8979, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.0427314620863008, | |
| "grad_norm": 0.5800043940544128, | |
| "learning_rate": 1.9830749895266026e-05, | |
| "loss": 1.9483, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.043150397989107664, | |
| "grad_norm": 0.7634892463684082, | |
| "learning_rate": 1.98290741516548e-05, | |
| "loss": 1.7777, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.04356933389191454, | |
| "grad_norm": 0.5963580012321472, | |
| "learning_rate": 1.982739840804357e-05, | |
| "loss": 1.9406, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.04398826979472141, | |
| "grad_norm": 0.6970496773719788, | |
| "learning_rate": 1.982572266443234e-05, | |
| "loss": 1.967, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.044407205697528276, | |
| "grad_norm": 0.5826534032821655, | |
| "learning_rate": 1.9824046920821116e-05, | |
| "loss": 1.8577, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.04482614160033515, | |
| "grad_norm": 0.60413658618927, | |
| "learning_rate": 1.982237117720989e-05, | |
| "loss": 1.9528, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.04524507750314202, | |
| "grad_norm": 0.7267922759056091, | |
| "learning_rate": 1.982069543359866e-05, | |
| "loss": 1.7067, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.04566401340594889, | |
| "grad_norm": 0.6376165747642517, | |
| "learning_rate": 1.9819019689987435e-05, | |
| "loss": 1.9888, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.04608294930875576, | |
| "grad_norm": 0.5887031555175781, | |
| "learning_rate": 1.9817343946376205e-05, | |
| "loss": 1.7777, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04650188521156263, | |
| "grad_norm": 0.6548938155174255, | |
| "learning_rate": 1.981566820276498e-05, | |
| "loss": 1.9222, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.0469208211143695, | |
| "grad_norm": 0.5757064819335938, | |
| "learning_rate": 1.981399245915375e-05, | |
| "loss": 1.844, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.047339757017176375, | |
| "grad_norm": 0.7597166895866394, | |
| "learning_rate": 1.9812316715542524e-05, | |
| "loss": 1.8678, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.04775869291998324, | |
| "grad_norm": 0.5536984801292419, | |
| "learning_rate": 1.9810640971931295e-05, | |
| "loss": 1.8502, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.048177628822790114, | |
| "grad_norm": 0.5753149390220642, | |
| "learning_rate": 1.980896522832007e-05, | |
| "loss": 1.8613, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.04859656472559699, | |
| "grad_norm": 0.6214611530303955, | |
| "learning_rate": 1.980728948470884e-05, | |
| "loss": 1.8803, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.04901550062840385, | |
| "grad_norm": 0.5892764329910278, | |
| "learning_rate": 1.9805613741097614e-05, | |
| "loss": 1.8241, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.049434436531210726, | |
| "grad_norm": 0.5623623132705688, | |
| "learning_rate": 1.9803937997486388e-05, | |
| "loss": 1.8584, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.04985337243401759, | |
| "grad_norm": 0.5206480622291565, | |
| "learning_rate": 1.980226225387516e-05, | |
| "loss": 1.8058, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.050272308336824466, | |
| "grad_norm": 0.7416813373565674, | |
| "learning_rate": 1.980058651026393e-05, | |
| "loss": 1.7422, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05069124423963134, | |
| "grad_norm": 0.6095878481864929, | |
| "learning_rate": 1.9798910766652704e-05, | |
| "loss": 1.7567, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.051110180142438205, | |
| "grad_norm": 0.5830249786376953, | |
| "learning_rate": 1.9797235023041478e-05, | |
| "loss": 1.8099, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.05152911604524508, | |
| "grad_norm": 0.7603867053985596, | |
| "learning_rate": 1.979555927943025e-05, | |
| "loss": 1.8074, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 0.607905387878418, | |
| "learning_rate": 1.979388353581902e-05, | |
| "loss": 1.8359, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.05236698785085882, | |
| "grad_norm": 0.5446661710739136, | |
| "learning_rate": 1.9792207792207794e-05, | |
| "loss": 1.7291, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.05278592375366569, | |
| "grad_norm": 0.5527285933494568, | |
| "learning_rate": 1.9790532048596568e-05, | |
| "loss": 1.7841, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.05320485965647256, | |
| "grad_norm": 0.6565405130386353, | |
| "learning_rate": 1.978885630498534e-05, | |
| "loss": 1.9168, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.05362379555927943, | |
| "grad_norm": 0.5959405899047852, | |
| "learning_rate": 1.978718056137411e-05, | |
| "loss": 1.8735, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.0540427314620863, | |
| "grad_norm": 0.7685437202453613, | |
| "learning_rate": 1.9785504817762883e-05, | |
| "loss": 1.782, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.05446166736489317, | |
| "grad_norm": 0.5747430324554443, | |
| "learning_rate": 1.9783829074151657e-05, | |
| "loss": 1.7831, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05488060326770004, | |
| "grad_norm": 0.7328975200653076, | |
| "learning_rate": 1.9782153330540428e-05, | |
| "loss": 1.7451, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.055299539170506916, | |
| "grad_norm": 0.5662095546722412, | |
| "learning_rate": 1.9780477586929202e-05, | |
| "loss": 1.6731, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.05571847507331378, | |
| "grad_norm": 0.6165090799331665, | |
| "learning_rate": 1.9778801843317973e-05, | |
| "loss": 1.7787, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.056137410976120655, | |
| "grad_norm": 0.6399924755096436, | |
| "learning_rate": 1.9777126099706747e-05, | |
| "loss": 1.8697, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.05655634687892752, | |
| "grad_norm": 0.6513495445251465, | |
| "learning_rate": 1.9775450356095518e-05, | |
| "loss": 1.7885, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.056975282781734395, | |
| "grad_norm": 0.652104914188385, | |
| "learning_rate": 1.9773774612484292e-05, | |
| "loss": 1.8462, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.05739421868454127, | |
| "grad_norm": 0.8418712615966797, | |
| "learning_rate": 1.9772098868873063e-05, | |
| "loss": 1.8771, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.057813154587348134, | |
| "grad_norm": 0.6133163571357727, | |
| "learning_rate": 1.9770423125261837e-05, | |
| "loss": 1.7998, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.05823209049015501, | |
| "grad_norm": 0.6718358993530273, | |
| "learning_rate": 1.9768747381650608e-05, | |
| "loss": 1.7633, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.05865102639296188, | |
| "grad_norm": 0.6728368997573853, | |
| "learning_rate": 1.9767071638039382e-05, | |
| "loss": 1.8313, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.059069962295768746, | |
| "grad_norm": 0.588307797908783, | |
| "learning_rate": 1.9765395894428156e-05, | |
| "loss": 1.7395, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.05948889819857562, | |
| "grad_norm": 0.889776885509491, | |
| "learning_rate": 1.9763720150816927e-05, | |
| "loss": 1.7591, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.059907834101382486, | |
| "grad_norm": 0.5996978878974915, | |
| "learning_rate": 1.9762044407205697e-05, | |
| "loss": 1.6881, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.06032677000418936, | |
| "grad_norm": 0.6324535012245178, | |
| "learning_rate": 1.976036866359447e-05, | |
| "loss": 1.7142, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.06074570590699623, | |
| "grad_norm": 0.6198902130126953, | |
| "learning_rate": 1.9758692919983246e-05, | |
| "loss": 1.6867, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.0611646418098031, | |
| "grad_norm": 0.6651074886322021, | |
| "learning_rate": 1.9757017176372016e-05, | |
| "loss": 1.7773, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.06158357771260997, | |
| "grad_norm": 0.6877864599227905, | |
| "learning_rate": 1.9755341432760787e-05, | |
| "loss": 1.792, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.062002513615416845, | |
| "grad_norm": 0.5880079865455627, | |
| "learning_rate": 1.975366568914956e-05, | |
| "loss": 1.7761, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.06242144951822371, | |
| "grad_norm": 0.6574519872665405, | |
| "learning_rate": 1.9751989945538335e-05, | |
| "loss": 1.7769, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.06284038542103058, | |
| "grad_norm": 0.5385615825653076, | |
| "learning_rate": 1.9750314201927106e-05, | |
| "loss": 1.7226, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06325932132383745, | |
| "grad_norm": 0.6323086023330688, | |
| "learning_rate": 1.9748638458315877e-05, | |
| "loss": 1.7043, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.06367825722664433, | |
| "grad_norm": 0.5224570035934448, | |
| "learning_rate": 1.974696271470465e-05, | |
| "loss": 1.7103, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.0640971931294512, | |
| "grad_norm": 0.5448949933052063, | |
| "learning_rate": 1.9745286971093425e-05, | |
| "loss": 1.7691, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 0.47703322768211365, | |
| "learning_rate": 1.97436112274822e-05, | |
| "loss": 1.7107, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 0.555005669593811, | |
| "learning_rate": 1.974193548387097e-05, | |
| "loss": 1.7493, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.06535400083787181, | |
| "grad_norm": 0.5083591938018799, | |
| "learning_rate": 1.974025974025974e-05, | |
| "loss": 1.621, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.06577293674067868, | |
| "grad_norm": 0.5037218928337097, | |
| "learning_rate": 1.9738583996648515e-05, | |
| "loss": 1.6905, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.06619187264348554, | |
| "grad_norm": 0.48570749163627625, | |
| "learning_rate": 1.973690825303729e-05, | |
| "loss": 1.73, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.06661080854629242, | |
| "grad_norm": 0.5082443952560425, | |
| "learning_rate": 1.973523250942606e-05, | |
| "loss": 1.7107, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.06702974444909929, | |
| "grad_norm": 0.5330659747123718, | |
| "learning_rate": 1.973355676581483e-05, | |
| "loss": 1.7756, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06744868035190615, | |
| "grad_norm": 0.6401243209838867, | |
| "learning_rate": 1.9731881022203605e-05, | |
| "loss": 1.7171, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.06786761625471303, | |
| "grad_norm": 0.5555040836334229, | |
| "learning_rate": 1.9730205278592375e-05, | |
| "loss": 1.7117, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.0682865521575199, | |
| "grad_norm": 0.5580366253852844, | |
| "learning_rate": 1.972852953498115e-05, | |
| "loss": 1.7173, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.06870548806032677, | |
| "grad_norm": 0.6296005249023438, | |
| "learning_rate": 1.9726853791369924e-05, | |
| "loss": 1.7316, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.06912442396313365, | |
| "grad_norm": 0.681302547454834, | |
| "learning_rate": 1.9725178047758694e-05, | |
| "loss": 1.7849, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.06954335986594051, | |
| "grad_norm": 0.6220213770866394, | |
| "learning_rate": 1.9723502304147465e-05, | |
| "loss": 1.656, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.06996229576874738, | |
| "grad_norm": 0.6640814542770386, | |
| "learning_rate": 1.972182656053624e-05, | |
| "loss": 1.8749, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.07038123167155426, | |
| "grad_norm": 0.5521919131278992, | |
| "learning_rate": 1.9720150816925013e-05, | |
| "loss": 1.763, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.07080016757436113, | |
| "grad_norm": 0.6188511848449707, | |
| "learning_rate": 1.9718475073313784e-05, | |
| "loss": 1.6431, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.07121910347716799, | |
| "grad_norm": 0.5388275980949402, | |
| "learning_rate": 1.9716799329702555e-05, | |
| "loss": 1.7444, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.07163803937997486, | |
| "grad_norm": 1.0150574445724487, | |
| "learning_rate": 1.971512358609133e-05, | |
| "loss": 1.6986, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.07205697528278174, | |
| "grad_norm": 0.6714919805526733, | |
| "learning_rate": 1.9713447842480103e-05, | |
| "loss": 1.7461, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.0724759111855886, | |
| "grad_norm": 0.6587640047073364, | |
| "learning_rate": 1.9711772098868874e-05, | |
| "loss": 1.6743, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.07289484708839547, | |
| "grad_norm": 0.6181256175041199, | |
| "learning_rate": 1.9710096355257648e-05, | |
| "loss": 1.8552, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.07331378299120235, | |
| "grad_norm": 0.5564039945602417, | |
| "learning_rate": 1.970842061164642e-05, | |
| "loss": 1.741, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.07373271889400922, | |
| "grad_norm": 0.5421382188796997, | |
| "learning_rate": 1.9706744868035193e-05, | |
| "loss": 1.6919, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.07415165479681608, | |
| "grad_norm": 0.6172086000442505, | |
| "learning_rate": 1.9705069124423967e-05, | |
| "loss": 1.6686, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.07457059069962296, | |
| "grad_norm": 0.5004185438156128, | |
| "learning_rate": 1.9703393380812738e-05, | |
| "loss": 1.6309, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.07498952660242983, | |
| "grad_norm": 0.5099078416824341, | |
| "learning_rate": 1.970171763720151e-05, | |
| "loss": 1.674, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.0754084625052367, | |
| "grad_norm": 0.5554249882698059, | |
| "learning_rate": 1.9700041893590283e-05, | |
| "loss": 1.6227, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07582739840804358, | |
| "grad_norm": 0.6313985586166382, | |
| "learning_rate": 1.9698366149979057e-05, | |
| "loss": 1.7098, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.07624633431085044, | |
| "grad_norm": 0.8186052441596985, | |
| "learning_rate": 1.9696690406367827e-05, | |
| "loss": 1.7933, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.07666527021365731, | |
| "grad_norm": 0.5017969608306885, | |
| "learning_rate": 1.9695014662756598e-05, | |
| "loss": 1.6853, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.07708420611646419, | |
| "grad_norm": 0.6917697787284851, | |
| "learning_rate": 1.9693338919145372e-05, | |
| "loss": 1.7031, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.07750314201927105, | |
| "grad_norm": 0.5040557980537415, | |
| "learning_rate": 1.9691663175534146e-05, | |
| "loss": 1.6353, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.07792207792207792, | |
| "grad_norm": 0.5733162760734558, | |
| "learning_rate": 1.9689987431922917e-05, | |
| "loss": 1.7278, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.07834101382488479, | |
| "grad_norm": 0.7823026776313782, | |
| "learning_rate": 1.968831168831169e-05, | |
| "loss": 1.6954, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.07875994972769167, | |
| "grad_norm": 0.5559296607971191, | |
| "learning_rate": 1.9686635944700462e-05, | |
| "loss": 1.7417, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.07917888563049853, | |
| "grad_norm": 0.6399711966514587, | |
| "learning_rate": 1.9684960201089236e-05, | |
| "loss": 1.6529, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.0795978215333054, | |
| "grad_norm": 0.6075267195701599, | |
| "learning_rate": 1.9683284457478007e-05, | |
| "loss": 1.687, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.08001675743611228, | |
| "grad_norm": 0.5875303149223328, | |
| "learning_rate": 1.968160871386678e-05, | |
| "loss": 1.6569, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.08043569333891915, | |
| "grad_norm": 0.6546170711517334, | |
| "learning_rate": 1.9679932970255552e-05, | |
| "loss": 1.6844, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.08085462924172601, | |
| "grad_norm": 0.5778879523277283, | |
| "learning_rate": 1.9678257226644322e-05, | |
| "loss": 1.6705, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.08127356514453289, | |
| "grad_norm": 0.5718396306037903, | |
| "learning_rate": 1.9676581483033097e-05, | |
| "loss": 1.6519, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.08169250104733976, | |
| "grad_norm": 0.6596693992614746, | |
| "learning_rate": 1.967490573942187e-05, | |
| "loss": 1.6547, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.08211143695014662, | |
| "grad_norm": 0.5057825446128845, | |
| "learning_rate": 1.9673229995810645e-05, | |
| "loss": 1.6214, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.0825303728529535, | |
| "grad_norm": 0.6651629209518433, | |
| "learning_rate": 1.9671554252199416e-05, | |
| "loss": 1.6835, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.08294930875576037, | |
| "grad_norm": 0.5056618452072144, | |
| "learning_rate": 1.9669878508588186e-05, | |
| "loss": 1.6368, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.08336824465856724, | |
| "grad_norm": 0.4693203568458557, | |
| "learning_rate": 1.966820276497696e-05, | |
| "loss": 1.6322, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.08378718056137412, | |
| "grad_norm": 0.5992833375930786, | |
| "learning_rate": 1.9666527021365735e-05, | |
| "loss": 1.7568, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 11935, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.103598585675776e+17, | |
| "train_batch_size": 12, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |