| { | |
| "best_metric": 10.219217300415039, | |
| "best_model_checkpoint": "tiny_bert_rand_50_v2/checkpoint-10000", | |
| "epoch": 25.0, | |
| "eval_steps": 10000, | |
| "global_step": 59550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 2.58121919631958, | |
| "learning_rate": 5e-06, | |
| "loss": 13.7114, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.6663143634796143, | |
| "learning_rate": 1e-05, | |
| "loss": 12.131, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.270777702331543, | |
| "learning_rate": 1.5e-05, | |
| "loss": 11.137, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.577506422996521, | |
| "learning_rate": 2e-05, | |
| "loss": 10.9382, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 1.4026228189468384, | |
| "learning_rate": 2.5e-05, | |
| "loss": 10.8204, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.661681056022644, | |
| "learning_rate": 3e-05, | |
| "loss": 10.7133, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.7125732898712158, | |
| "learning_rate": 3.5e-05, | |
| "loss": 10.6266, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 1.7105406522750854, | |
| "learning_rate": 4e-05, | |
| "loss": 10.5437, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.66139554977417, | |
| "learning_rate": 4.5e-05, | |
| "loss": 10.4697, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 1.7279640436172485, | |
| "learning_rate": 5e-05, | |
| "loss": 10.404, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.6322011947631836, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 10.3583, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 1.7196799516677856, | |
| "learning_rate": 6e-05, | |
| "loss": 10.3272, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.834273099899292, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 10.3003, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 1.8481425046920776, | |
| "learning_rate": 7e-05, | |
| "loss": 10.2768, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 1.9202250242233276, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.2173, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.084458827972412, | |
| "learning_rate": 8e-05, | |
| "loss": 10.1825, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.9900461435317993, | |
| "learning_rate": 8.5e-05, | |
| "loss": 10.1725, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.0722720623016357, | |
| "learning_rate": 9e-05, | |
| "loss": 10.1502, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.108538866043091, | |
| "learning_rate": 9.5e-05, | |
| "loss": 10.1384, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 2.6479101181030273, | |
| "learning_rate": 0.0001, | |
| "loss": 9.9992, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.14992069896213028, | |
| "eval_loss": 10.219217300415039, | |
| "eval_runtime": 0.7695, | |
| "eval_samples_per_second": 622.515, | |
| "eval_steps_per_second": 6.498, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.442643642425537, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 9.9888, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.897751808166504, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 9.9847, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.748310089111328, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 9.9902, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.0857391357421875, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 9.9478, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.875980854034424, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 9.8005, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.7046589851379395, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 9.8128, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.8039207458496094, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 9.8344, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.0976402759552, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 9.8429, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 3.0782783031463623, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 9.7596, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.946847915649414, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 9.6788, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.9093332290649414, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 9.7014, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 2.6727590560913086, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 9.7254, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.99617600440979, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 9.7411, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.994882106781006, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 9.6146, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 3.271754741668701, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 9.5749, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 3.124356508255005, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 9.6026, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 3.1813101768493652, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 9.6323, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 3.0956149101257324, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 9.6405, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 3.48783016204834, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 9.4607, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.276175022125244, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 9.4707, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15233311821319342, | |
| "eval_loss": 10.47436237335205, | |
| "eval_runtime": 0.721, | |
| "eval_samples_per_second": 664.388, | |
| "eval_steps_per_second": 6.935, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 3.416642189025879, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 9.4907, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 3.240661144256592, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 9.5221, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 3.578784704208374, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 9.497, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 4.055538177490234, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 9.3045, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 4.117690086364746, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 9.3292, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 4.141088962554932, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 9.3685, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 4.071991920471191, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 9.3732, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 4.145738124847412, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 9.2859, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 4.137731552124023, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 9.1324, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 4.4439697265625, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 9.1463, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 4.70512056350708, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 9.1796, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 4.668952465057373, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 9.2083, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 4.878523349761963, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 9.0282, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 5.5130791664123535, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 8.9275, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 5.883714199066162, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 8.9614, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 5.100078582763672, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 8.9831, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 5.795191287994385, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 9.0042, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 5.933318138122559, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 8.7401, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 5.403182506561279, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 8.7121, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 6.357689380645752, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 8.7418, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.1531745809665053, | |
| "eval_loss": 11.1043062210083, | |
| "eval_runtime": 0.7223, | |
| "eval_samples_per_second": 663.166, | |
| "eval_steps_per_second": 6.922, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 6.47633695602417, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 8.763, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 7.0516204833984375, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 8.7694, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 6.569056987762451, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 8.4566, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 6.599896430969238, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 8.4863, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 6.598655700683594, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 8.536, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 6.2865729331970215, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 8.5553, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 7.088315010070801, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 8.4595, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 8.113789558410645, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 8.2519, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 7.205976486206055, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 8.2845, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 7.893230438232422, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 8.3131, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 7.748480319976807, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 8.3371, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 7.629475116729736, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 8.1552, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 7.711606025695801, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 8.0505, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 8.223636627197266, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 8.0798, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 7.557830810546875, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 8.1166, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 7.7148356437683105, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 8.1425, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 9.080387115478516, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 7.8852, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 8.934139251708984, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 7.8681, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 9.111722946166992, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 7.8913, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 9.9085693359375, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 7.9122, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15232140854547493, | |
| "eval_loss": 11.895573616027832, | |
| "eval_runtime": 0.709, | |
| "eval_samples_per_second": 675.569, | |
| "eval_steps_per_second": 7.052, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 8.389793395996094, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 7.9564, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 8.006103515625, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 7.6512, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 9.008237838745117, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 7.6851, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 8.521134376525879, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 7.7155, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 9.356441497802734, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 7.7525, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 10.665168762207031, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 7.684, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 9.566268920898438, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 7.4961, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 9.710851669311523, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 7.5286, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 9.869644165039062, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 7.5567, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 8.710551261901855, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 7.5871, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 9.51053237915039, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 7.4638, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 10.861311912536621, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 7.3646, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 9.868949890136719, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 7.3834, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 9.439346313476562, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 7.4004, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 9.86860179901123, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 7.4242, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 9.786796569824219, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 7.2865, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 10.61046028137207, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 7.2454, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 10.151837348937988, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 7.2606, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 9.988868713378906, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 7.2772, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 10.063469886779785, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 7.2963, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1511200829898786, | |
| "eval_loss": 12.642455101013184, | |
| "eval_runtime": 0.7663, | |
| "eval_samples_per_second": 625.069, | |
| "eval_steps_per_second": 6.525, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 10.229823112487793, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 7.1302, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 10.6376953125, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 7.1522, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 10.69062614440918, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 7.1604, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 12.068879127502441, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 7.1741, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 8.581454277038574, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 7.1449, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 9.05636978149414, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 7.0424, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 9.613524436950684, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 7.053, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 11.62191390991211, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 7.0738, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 9.554201126098633, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 7.0662, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 11.097816467285156, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 7.0302, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 10.611686706542969, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.9769, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 11.101253509521484, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.9982, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 11.03598690032959, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.9746, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 11.151888847351074, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 7.0049, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 8.397727966308594, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.9599, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 9.544459342956543, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.9543, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 10.507208824157715, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.9565, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 10.332313537597656, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.9401, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 10.73017406463623, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.9322, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.053850510375936e+17, | |
| "train_loss": 8.741478605102232, | |
| "train_runtime": 12176.6113, | |
| "train_samples_per_second": 469.422, | |
| "train_steps_per_second": 4.891 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 3.053850510375936e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |