|
|
{ |
|
|
"best_global_step": 400, |
|
|
"best_metric": 0.16972463, |
|
|
"best_model_checkpoint": "/Qwen/Qwen-VL-master/ckp/checkpoints_IO_3B/v0-20250504-043938/checkpoint-400", |
|
|
"epoch": 0.9996103390050656, |
|
|
"eval_steps": 100, |
|
|
"global_step": 481, |
|
|
"is_hyper_param_search": false, |
|
|
"is_local_process_zero": true, |
|
|
"is_world_process_zero": true, |
|
|
"log_history": [ |
|
|
{ |
|
|
"epoch": 0.002078191972983504, |
|
|
"grad_norm": 49.5, |
|
|
"learning_rate": 2.0000000000000002e-07, |
|
|
"loss": 2.024962902069092, |
|
|
"memory(GiB)": 30.5, |
|
|
"step": 1, |
|
|
"token_acc": 0.5925925925925926, |
|
|
"train_speed(iter/s)": 0.096714 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.010390959864917521, |
|
|
"grad_norm": 48.25, |
|
|
"learning_rate": 1.0000000000000002e-06, |
|
|
"loss": 1.9107239246368408, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 5, |
|
|
"token_acc": 0.6030968247745983, |
|
|
"train_speed(iter/s)": 0.104169 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.020781919729835042, |
|
|
"grad_norm": 46.0, |
|
|
"learning_rate": 2.0000000000000003e-06, |
|
|
"loss": 1.8856124877929688, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 10, |
|
|
"token_acc": 0.6060725241878392, |
|
|
"train_speed(iter/s)": 0.10431 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.031172879594752564, |
|
|
"grad_norm": 41.25, |
|
|
"learning_rate": 3e-06, |
|
|
"loss": 1.6671241760253905, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 15, |
|
|
"token_acc": 0.6276759587247805, |
|
|
"train_speed(iter/s)": 0.104357 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.041563839459670085, |
|
|
"grad_norm": 35.75, |
|
|
"learning_rate": 4.000000000000001e-06, |
|
|
"loss": 1.4283610343933106, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 20, |
|
|
"token_acc": 0.644587791063562, |
|
|
"train_speed(iter/s)": 0.104096 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.05195479932458761, |
|
|
"grad_norm": 24.125, |
|
|
"learning_rate": 5e-06, |
|
|
"loss": 0.8581354141235351, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 25, |
|
|
"token_acc": 0.7361944076641483, |
|
|
"train_speed(iter/s)": 0.103919 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.06234575918950513, |
|
|
"grad_norm": 10.4375, |
|
|
"learning_rate": 4.998516877114918e-06, |
|
|
"loss": 0.3941912889480591, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 30, |
|
|
"token_acc": 0.8509957048028114, |
|
|
"train_speed(iter/s)": 0.103913 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.07273671905442265, |
|
|
"grad_norm": 3.6875, |
|
|
"learning_rate": 4.994069268182461e-06, |
|
|
"loss": 0.24166102409362794, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 35, |
|
|
"token_acc": 0.9107004193197702, |
|
|
"train_speed(iter/s)": 0.103983 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.08312767891934017, |
|
|
"grad_norm": 1.828125, |
|
|
"learning_rate": 4.986662450283107e-06, |
|
|
"loss": 0.2008065700531006, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 40, |
|
|
"token_acc": 0.9205524090309566, |
|
|
"train_speed(iter/s)": 0.103895 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.0935186387842577, |
|
|
"grad_norm": 2.265625, |
|
|
"learning_rate": 4.976305211593758e-06, |
|
|
"loss": 0.19015437364578247, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 45, |
|
|
"token_acc": 0.9180647678501755, |
|
|
"train_speed(iter/s)": 0.103752 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.10390959864917522, |
|
|
"grad_norm": 1.9921875, |
|
|
"learning_rate": 4.963009840960598e-06, |
|
|
"loss": 0.19012608528137206, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 50, |
|
|
"token_acc": 0.9183419043166589, |
|
|
"train_speed(iter/s)": 0.103736 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.11430055851409274, |
|
|
"grad_norm": 1.9921875, |
|
|
"learning_rate": 4.9467921133183864e-06, |
|
|
"loss": 0.1876815915107727, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 55, |
|
|
"token_acc": 0.9174203696421549, |
|
|
"train_speed(iter/s)": 0.103693 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.12469151837901025, |
|
|
"grad_norm": 1.640625, |
|
|
"learning_rate": 4.927671270973533e-06, |
|
|
"loss": 0.18475005626678467, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 60, |
|
|
"token_acc": 0.9194869417400711, |
|
|
"train_speed(iter/s)": 0.103749 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.13508247824392777, |
|
|
"grad_norm": 2.03125, |
|
|
"learning_rate": 4.905670000773126e-06, |
|
|
"loss": 0.18700193166732787, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 65, |
|
|
"token_acc": 0.9183423809159722, |
|
|
"train_speed(iter/s)": 0.103789 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.1454734381088453, |
|
|
"grad_norm": 1.8984375, |
|
|
"learning_rate": 4.880814407187037e-06, |
|
|
"loss": 0.1833273410797119, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 70, |
|
|
"token_acc": 0.9198803956145059, |
|
|
"train_speed(iter/s)": 0.103777 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.15586439797376284, |
|
|
"grad_norm": 1.8046875, |
|
|
"learning_rate": 4.8531339813349995e-06, |
|
|
"loss": 0.17667144536972046, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 75, |
|
|
"token_acc": 0.9221237549996079, |
|
|
"train_speed(iter/s)": 0.101758 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.16625535783868034, |
|
|
"grad_norm": 1.8203125, |
|
|
"learning_rate": 4.822661565995454e-06, |
|
|
"loss": 0.17785824537277223, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 80, |
|
|
"token_acc": 0.9208750290900628, |
|
|
"train_speed(iter/s)": 0.096873 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.17664631770359787, |
|
|
"grad_norm": 1.5390625, |
|
|
"learning_rate": 4.789433316637644e-06, |
|
|
"loss": 0.18061867952346802, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 85, |
|
|
"token_acc": 0.9204729886389984, |
|
|
"train_speed(iter/s)": 0.092995 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.1870372775685154, |
|
|
"grad_norm": 1.8828125, |
|
|
"learning_rate": 4.753488658523212e-06, |
|
|
"loss": 0.17701296806335448, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 90, |
|
|
"token_acc": 0.9225134926754048, |
|
|
"train_speed(iter/s)": 0.089657 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.1974282374334329, |
|
|
"grad_norm": 3.21875, |
|
|
"learning_rate": 4.714870239928195e-06, |
|
|
"loss": 0.1849340558052063, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 95, |
|
|
"token_acc": 0.9169326586220319, |
|
|
"train_speed(iter/s)": 0.086893 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.20781919729835044, |
|
|
"grad_norm": 3.3125, |
|
|
"learning_rate": 4.673623881540917e-06, |
|
|
"loss": 0.1811493992805481, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 100, |
|
|
"token_acc": 0.9194093158432174, |
|
|
"train_speed(iter/s)": 0.084511 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.20781919729835044, |
|
|
"eval_loss": 0.17230889201164246, |
|
|
"eval_runtime": 1320.3184, |
|
|
"eval_samples_per_second": 60.881, |
|
|
"eval_steps_per_second": 3.805, |
|
|
"eval_token_acc": 0.9252880661892655, |
|
|
"step": 100 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.21821015716326797, |
|
|
"grad_norm": 1.9375, |
|
|
"learning_rate": 4.629798522095818e-06, |
|
|
"loss": 0.1755787253379822, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 105, |
|
|
"token_acc": 0.925254316998113, |
|
|
"train_speed(iter/s)": 0.03748 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.22860111702818547, |
|
|
"grad_norm": 1.6484375, |
|
|
"learning_rate": 4.583446160307729e-06, |
|
|
"loss": 0.16902718544006348, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 110, |
|
|
"token_acc": 0.9260815437286025, |
|
|
"train_speed(iter/s)": 0.0381 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.238992076893103, |
|
|
"grad_norm": 3.0, |
|
|
"learning_rate": 4.534621793175488e-06, |
|
|
"loss": 0.17990721464157106, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 115, |
|
|
"token_acc": 0.9196615695102073, |
|
|
"train_speed(iter/s)": 0.038616 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.2493830367580205, |
|
|
"grad_norm": 1.984375, |
|
|
"learning_rate": 4.4833833507280884e-06, |
|
|
"loss": 0.17740323543548583, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 120, |
|
|
"token_acc": 0.9212245219170536, |
|
|
"train_speed(iter/s)": 0.039106 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.25977399662293804, |
|
|
"grad_norm": 2.03125, |
|
|
"learning_rate": 4.429791627290802e-06, |
|
|
"loss": 0.17831547260284425, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 125, |
|
|
"token_acc": 0.9183422870299309, |
|
|
"train_speed(iter/s)": 0.039569 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.27016495648785555, |
|
|
"grad_norm": 1.5546875, |
|
|
"learning_rate": 4.373910209352816e-06, |
|
|
"loss": 0.1763285517692566, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 130, |
|
|
"token_acc": 0.9201423817998917, |
|
|
"train_speed(iter/s)": 0.040384 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.2805559163527731, |
|
|
"grad_norm": 1.625, |
|
|
"learning_rate": 4.315805400121963e-06, |
|
|
"loss": 0.17515565156936647, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 135, |
|
|
"token_acc": 0.9202636948673678, |
|
|
"train_speed(iter/s)": 0.041134 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.2909468762176906, |
|
|
"grad_norm": 2.09375, |
|
|
"learning_rate": 4.255546140856087e-06, |
|
|
"loss": 0.18066967725753785, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 140, |
|
|
"token_acc": 0.9194688203774171, |
|
|
"train_speed(iter/s)": 0.041899 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.3013378360826081, |
|
|
"grad_norm": 1.140625, |
|
|
"learning_rate": 4.1932039290643534e-06, |
|
|
"loss": 0.18014354705810548, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 145, |
|
|
"token_acc": 0.9178993503952414, |
|
|
"train_speed(iter/s)": 0.042363 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.31172879594752567, |
|
|
"grad_norm": 1.3515625, |
|
|
"learning_rate": 4.128852733675572e-06, |
|
|
"loss": 0.17886133193969728, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 150, |
|
|
"token_acc": 0.9210793106137409, |
|
|
"train_speed(iter/s)": 0.042968 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.3221197558124432, |
|
|
"grad_norm": 1.7734375, |
|
|
"learning_rate": 4.062568907274197e-06, |
|
|
"loss": 0.18089871406555175, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 155, |
|
|
"token_acc": 0.9191895486192599, |
|
|
"train_speed(iter/s)": 0.043361 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.3325107156773607, |
|
|
"grad_norm": 1.7421875, |
|
|
"learning_rate": 3.994431095508102e-06, |
|
|
"loss": 0.17732292413711548, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 160, |
|
|
"token_acc": 0.9188111067657411, |
|
|
"train_speed(iter/s)": 0.043692 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.34290167554227824, |
|
|
"grad_norm": 1.78125, |
|
|
"learning_rate": 3.9245201437756655e-06, |
|
|
"loss": 0.1741779327392578, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 165, |
|
|
"token_acc": 0.9216403392732083, |
|
|
"train_speed(iter/s)": 0.043984 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.35329263540719574, |
|
|
"grad_norm": 1.875, |
|
|
"learning_rate": 3.852919001302833e-06, |
|
|
"loss": 0.17619532346725464, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 170, |
|
|
"token_acc": 0.9227966824277187, |
|
|
"train_speed(iter/s)": 0.044279 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.36368359527211325, |
|
|
"grad_norm": 3.71875, |
|
|
"learning_rate": 3.779712622724003e-06, |
|
|
"loss": 0.17437649965286256, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 175, |
|
|
"token_acc": 0.9230951628201075, |
|
|
"train_speed(iter/s)": 0.044551 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.3740745551370308, |
|
|
"grad_norm": 1.84375, |
|
|
"learning_rate": 3.704987867283499e-06, |
|
|
"loss": 0.17724401950836183, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 180, |
|
|
"token_acc": 0.9202978056426332, |
|
|
"train_speed(iter/s)": 0.044808 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.3844655150019483, |
|
|
"grad_norm": 2.1875, |
|
|
"learning_rate": 3.628833395777224e-06, |
|
|
"loss": 0.18116596937179566, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 185, |
|
|
"token_acc": 0.9170886075949367, |
|
|
"train_speed(iter/s)": 0.045056 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.3948564748668658, |
|
|
"grad_norm": 2.046875, |
|
|
"learning_rate": 3.551339565356769e-06, |
|
|
"loss": 0.17593677043914796, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 190, |
|
|
"token_acc": 0.9202586206896551, |
|
|
"train_speed(iter/s)": 0.045325 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.4052474347317834, |
|
|
"grad_norm": 2.0, |
|
|
"learning_rate": 3.4725983223208155e-06, |
|
|
"loss": 0.1823140025138855, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 195, |
|
|
"token_acc": 0.917639683410391, |
|
|
"train_speed(iter/s)": 0.04554 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.4156383945967009, |
|
|
"grad_norm": 1.7109375, |
|
|
"learning_rate": 3.392703093021e-06, |
|
|
"loss": 0.18079570531845093, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 200, |
|
|
"token_acc": 0.9180621543733936, |
|
|
"train_speed(iter/s)": 0.045874 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.4156383945967009, |
|
|
"eval_loss": 0.17064262926578522, |
|
|
"eval_runtime": 1310.4171, |
|
|
"eval_samples_per_second": 61.341, |
|
|
"eval_steps_per_second": 3.834, |
|
|
"eval_token_acc": 0.9238987477359346, |
|
|
"step": 200 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.4260293544616184, |
|
|
"grad_norm": 1.7890625, |
|
|
"learning_rate": 3.3117486730117092e-06, |
|
|
"loss": 0.181607449054718, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 205, |
|
|
"token_acc": 0.9238299582974754, |
|
|
"train_speed(iter/s)": 0.034354 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.43642031432653594, |
|
|
"grad_norm": 2.0625, |
|
|
"learning_rate": 3.229831114575315e-06, |
|
|
"loss": 0.17884130477905275, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 210, |
|
|
"token_acc": 0.9180160352145889, |
|
|
"train_speed(iter/s)": 0.034671 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.44681127419145344, |
|
|
"grad_norm": 1.484375, |
|
|
"learning_rate": 3.147047612756302e-06, |
|
|
"loss": 0.174602210521698, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 215, |
|
|
"token_acc": 0.920204921214003, |
|
|
"train_speed(iter/s)": 0.034976 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.45720223405637095, |
|
|
"grad_norm": 1.234375, |
|
|
"learning_rate": 3.063496390039516e-06, |
|
|
"loss": 0.1783522129058838, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 220, |
|
|
"token_acc": 0.9198404629702042, |
|
|
"train_speed(iter/s)": 0.035377 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.46759319392128845, |
|
|
"grad_norm": 1.78125, |
|
|
"learning_rate": 2.9792765798093466e-06, |
|
|
"loss": 0.179638671875, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 225, |
|
|
"token_acc": 0.9207520783155932, |
|
|
"train_speed(iter/s)": 0.035838 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.477984153786206, |
|
|
"grad_norm": 1.921875, |
|
|
"learning_rate": 2.8944881087281375e-06, |
|
|
"loss": 0.18421536684036255, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 230, |
|
|
"token_acc": 0.916609076249712, |
|
|
"train_speed(iter/s)": 0.036335 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.4883751136511235, |
|
|
"grad_norm": 1.7890625, |
|
|
"learning_rate": 2.80923157817337e-06, |
|
|
"loss": 0.17683053016662598, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 235, |
|
|
"token_acc": 0.9213091922005571, |
|
|
"train_speed(iter/s)": 0.036825 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.498766073516041, |
|
|
"grad_norm": 1.84375, |
|
|
"learning_rate": 2.723608144874298e-06, |
|
|
"loss": 0.17713736295700072, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 240, |
|
|
"token_acc": 0.92035536159601, |
|
|
"train_speed(iter/s)": 0.037143 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5091570333809585, |
|
|
"grad_norm": 1.5234375, |
|
|
"learning_rate": 2.637719400889664e-06, |
|
|
"loss": 0.17735506296157838, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 245, |
|
|
"token_acc": 0.9191973969631236, |
|
|
"train_speed(iter/s)": 0.037405 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5195479932458761, |
|
|
"grad_norm": 1.3828125, |
|
|
"learning_rate": 2.5516672530688864e-06, |
|
|
"loss": 0.18297756910324098, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 250, |
|
|
"token_acc": 0.9193918758413177, |
|
|
"train_speed(iter/s)": 0.037652 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5299389531107936, |
|
|
"grad_norm": 2.203125, |
|
|
"learning_rate": 2.4655538021397592e-06, |
|
|
"loss": 0.17588781118392943, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 255, |
|
|
"token_acc": 0.9216273752266814, |
|
|
"train_speed(iter/s)": 0.037896 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5403299129757111, |
|
|
"grad_norm": 1.6796875, |
|
|
"learning_rate": 2.3794812215661134e-06, |
|
|
"loss": 0.18080203533172606, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 260, |
|
|
"token_acc": 0.921203216826477, |
|
|
"train_speed(iter/s)": 0.038125 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5507208728406287, |
|
|
"grad_norm": 2.6875, |
|
|
"learning_rate": 2.2935516363191695e-06, |
|
|
"loss": 0.18104695081710814, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 265, |
|
|
"token_acc": 0.9180170831005029, |
|
|
"train_speed(iter/s)": 0.038349 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5611118327055462, |
|
|
"grad_norm": 1.7578125, |
|
|
"learning_rate": 2.2078670017064366e-06, |
|
|
"loss": 0.1768229365348816, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 270, |
|
|
"token_acc": 0.9203719621786356, |
|
|
"train_speed(iter/s)": 0.038567 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5715027925704637, |
|
|
"grad_norm": 1.34375, |
|
|
"learning_rate": 2.1225289824019077e-06, |
|
|
"loss": 0.17726024389266967, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 275, |
|
|
"token_acc": 0.9191293295201779, |
|
|
"train_speed(iter/s)": 0.038795 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5818937524353812, |
|
|
"grad_norm": 1.71875, |
|
|
"learning_rate": 2.037638831821104e-06, |
|
|
"loss": 0.18150538206100464, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 280, |
|
|
"token_acc": 0.9174561267277528, |
|
|
"train_speed(iter/s)": 0.03901 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.5922847123002988, |
|
|
"grad_norm": 1.546875, |
|
|
"learning_rate": 1.953297271984061e-06, |
|
|
"loss": 0.1750028371810913, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 285, |
|
|
"token_acc": 0.9228211275230067, |
|
|
"train_speed(iter/s)": 0.039295 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6026756721652162, |
|
|
"grad_norm": 1.234375, |
|
|
"learning_rate": 1.8696043740088236e-06, |
|
|
"loss": 0.17536247968673707, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 290, |
|
|
"token_acc": 0.9208355815925383, |
|
|
"train_speed(iter/s)": 0.039587 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6130666320301338, |
|
|
"grad_norm": 1.5390625, |
|
|
"learning_rate": 1.7866594393772375e-06, |
|
|
"loss": 0.1767476439476013, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 295, |
|
|
"token_acc": 0.9220417633410672, |
|
|
"train_speed(iter/s)": 0.039803 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6234575918950513, |
|
|
"grad_norm": 1.9140625, |
|
|
"learning_rate": 1.7045608821139045e-06, |
|
|
"loss": 0.17351619005203248, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 300, |
|
|
"token_acc": 0.9228345246330029, |
|
|
"train_speed(iter/s)": 0.040008 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6234575918950513, |
|
|
"eval_loss": 0.16996929049491882, |
|
|
"eval_runtime": 1338.6589, |
|
|
"eval_samples_per_second": 60.047, |
|
|
"eval_steps_per_second": 3.753, |
|
|
"eval_token_acc": 0.9252277899263542, |
|
|
"step": 300 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6338485517599688, |
|
|
"grad_norm": 1.7109375, |
|
|
"learning_rate": 1.6234061120181144e-06, |
|
|
"loss": 0.1729782223701477, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 305, |
|
|
"token_acc": 0.9252146572948351, |
|
|
"train_speed(iter/s)": 0.033422 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6442395116248864, |
|
|
"grad_norm": 2.296875, |
|
|
"learning_rate": 1.5432914190872757e-06, |
|
|
"loss": 0.1758588194847107, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 310, |
|
|
"token_acc": 0.9202410211686355, |
|
|
"train_speed(iter/s)": 0.033685 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6546304714898039, |
|
|
"grad_norm": 2.03125, |
|
|
"learning_rate": 1.464311859269003e-06, |
|
|
"loss": 0.17637253999710084, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 315, |
|
|
"token_acc": 0.9202949713658115, |
|
|
"train_speed(iter/s)": 0.034017 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6650214313547214, |
|
|
"grad_norm": 2.296875, |
|
|
"learning_rate": 1.3865611416773921e-06, |
|
|
"loss": 0.1803775191307068, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 320, |
|
|
"token_acc": 0.9197953561589925, |
|
|
"train_speed(iter/s)": 0.034317 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6754123912196389, |
|
|
"grad_norm": 2.046875, |
|
|
"learning_rate": 1.3101315174073162e-06, |
|
|
"loss": 0.17441234588623047, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 325, |
|
|
"token_acc": 0.9233814625058221, |
|
|
"train_speed(iter/s)": 0.034642 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6858033510845565, |
|
|
"grad_norm": 1.640625, |
|
|
"learning_rate": 1.235113670078658e-06, |
|
|
"loss": 0.17859526872634887, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 330, |
|
|
"token_acc": 0.9203665665981988, |
|
|
"train_speed(iter/s)": 0.034906 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.6961943109494739, |
|
|
"grad_norm": 1.375, |
|
|
"learning_rate": 1.161596608240349e-06, |
|
|
"loss": 0.17650686502456664, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 335, |
|
|
"token_acc": 0.9200734956361966, |
|
|
"train_speed(iter/s)": 0.035109 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7065852708143915, |
|
|
"grad_norm": 1.7734375, |
|
|
"learning_rate": 1.0896675597618725e-06, |
|
|
"loss": 0.18125290870666505, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 340, |
|
|
"token_acc": 0.9172004012655297, |
|
|
"train_speed(iter/s)": 0.035306 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.716976230679309, |
|
|
"grad_norm": 1.046875, |
|
|
"learning_rate": 1.0194118683375502e-06, |
|
|
"loss": 0.1810152769088745, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 345, |
|
|
"token_acc": 0.9158885657431501, |
|
|
"train_speed(iter/s)": 0.035495 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7273671905442265, |
|
|
"grad_norm": 2.203125, |
|
|
"learning_rate": 9.509128922263886e-07, |
|
|
"loss": 0.17332950830459595, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 350, |
|
|
"token_acc": 0.9222935923546921, |
|
|
"train_speed(iter/s)": 0.035686 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.737758150409144, |
|
|
"grad_norm": 1.28125, |
|
|
"learning_rate": 8.842519053476476e-07, |
|
|
"loss": 0.17291887998580932, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 355, |
|
|
"token_acc": 0.9233901951744276, |
|
|
"train_speed(iter/s)": 0.035869 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7481491102740616, |
|
|
"grad_norm": 1.5859375, |
|
|
"learning_rate": 8.195080008494744e-07, |
|
|
"loss": 0.17527254819869995, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 360, |
|
|
"token_acc": 0.9194406927782632, |
|
|
"train_speed(iter/s)": 0.03605 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7585400701389791, |
|
|
"grad_norm": 1.8984375, |
|
|
"learning_rate": 7.567579972650116e-07, |
|
|
"loss": 0.1758326292037964, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 365, |
|
|
"token_acc": 0.9222837796244998, |
|
|
"train_speed(iter/s)": 0.036228 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7689310300038966, |
|
|
"grad_norm": 1.4453125, |
|
|
"learning_rate": 6.960763473673451e-07, |
|
|
"loss": 0.17296208143234254, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 370, |
|
|
"token_acc": 0.9199661069172701, |
|
|
"train_speed(iter/s)": 0.036459 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7793219898688142, |
|
|
"grad_norm": 2.15625, |
|
|
"learning_rate": 6.375350498314075e-07, |
|
|
"loss": 0.17531843185424806, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 375, |
|
|
"token_acc": 0.921526620645878, |
|
|
"train_speed(iter/s)": 0.036681 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.7897129497337316, |
|
|
"grad_norm": 1.453125, |
|
|
"learning_rate": 5.812035638076785e-07, |
|
|
"loss": 0.17355836629867555, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 380, |
|
|
"token_acc": 0.9211911683821796, |
|
|
"train_speed(iter/s)": 0.036859 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8001039095986492, |
|
|
"grad_norm": 2.03125, |
|
|
"learning_rate": 5.271487265090163e-07, |
|
|
"loss": 0.1834742784500122, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 385, |
|
|
"token_acc": 0.9166990442147797, |
|
|
"train_speed(iter/s)": 0.037022 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8104948694635667, |
|
|
"grad_norm": 1.78125, |
|
|
"learning_rate": 4.754346739084173e-07, |
|
|
"loss": 0.17866382598876954, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 390, |
|
|
"token_acc": 0.9190055507778907, |
|
|
"train_speed(iter/s)": 0.037179 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8208858293284842, |
|
|
"grad_norm": 1.7265625, |
|
|
"learning_rate": 4.2612276464179673e-07, |
|
|
"loss": 0.16837071180343627, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 395, |
|
|
"token_acc": 0.925252047889099, |
|
|
"train_speed(iter/s)": 0.037337 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8312767891934018, |
|
|
"grad_norm": 1.625, |
|
|
"learning_rate": 3.7927150720606596e-07, |
|
|
"loss": 0.17911137342453004, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 400, |
|
|
"token_acc": 0.920388196307401, |
|
|
"train_speed(iter/s)": 0.037494 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8312767891934018, |
|
|
"eval_loss": 0.16972462832927704, |
|
|
"eval_runtime": 1325.8316, |
|
|
"eval_samples_per_second": 60.628, |
|
|
"eval_steps_per_second": 3.789, |
|
|
"eval_token_acc": 0.9239007240068497, |
|
|
"step": 400 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8416677490583193, |
|
|
"grad_norm": 1.546875, |
|
|
"learning_rate": 3.3493649053890325e-07, |
|
|
"loss": 0.17124630212783815, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 405, |
|
|
"token_acc": 0.9238681922697054, |
|
|
"train_speed(iter/s)": 0.032973 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8520587089232368, |
|
|
"grad_norm": 2.25, |
|
|
"learning_rate": 2.931703180625736e-07, |
|
|
"loss": 0.17123721837997435, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 410, |
|
|
"token_acc": 0.9216517506672947, |
|
|
"train_speed(iter/s)": 0.033214 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8624496687881543, |
|
|
"grad_norm": 1.5859375, |
|
|
"learning_rate": 2.5402254527005286e-07, |
|
|
"loss": 0.17746515274047853, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 415, |
|
|
"token_acc": 0.9180823830579474, |
|
|
"train_speed(iter/s)": 0.033458 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8728406286530719, |
|
|
"grad_norm": 1.3984375, |
|
|
"learning_rate": 2.1753962092752545e-07, |
|
|
"loss": 0.1777236580848694, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 420, |
|
|
"token_acc": 0.9190801349972529, |
|
|
"train_speed(iter/s)": 0.033705 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8832315885179893, |
|
|
"grad_norm": 1.53125, |
|
|
"learning_rate": 1.837648319629956e-07, |
|
|
"loss": 0.17560485601425171, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 425, |
|
|
"token_acc": 0.9201813013751249, |
|
|
"train_speed(iter/s)": 0.033868 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.8936225483829069, |
|
|
"grad_norm": 1.7578125, |
|
|
"learning_rate": 1.5273825210642608e-07, |
|
|
"loss": 0.17915327548980714, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 430, |
|
|
"token_acc": 0.9182488623882002, |
|
|
"train_speed(iter/s)": 0.034029 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9040135082478243, |
|
|
"grad_norm": 1.3671875, |
|
|
"learning_rate": 1.2449669434232202e-07, |
|
|
"loss": 0.17606736421585084, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 435, |
|
|
"token_acc": 0.9228688461241209, |
|
|
"train_speed(iter/s)": 0.034183 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9144044681127419, |
|
|
"grad_norm": 1.546875, |
|
|
"learning_rate": 9.907366723118678e-08, |
|
|
"loss": 0.17585525512695313, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 440, |
|
|
"token_acc": 0.9183866393571766, |
|
|
"train_speed(iter/s)": 0.034333 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9247954279776595, |
|
|
"grad_norm": 1.5, |
|
|
"learning_rate": 7.649933515167407e-08, |
|
|
"loss": 0.16930280923843383, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 445, |
|
|
"token_acc": 0.9238362821778419, |
|
|
"train_speed(iter/s)": 0.034485 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9351863878425769, |
|
|
"grad_norm": 2.078125, |
|
|
"learning_rate": 5.6800482510601937e-08, |
|
|
"loss": 0.1705850124359131, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 450, |
|
|
"token_acc": 0.925625, |
|
|
"train_speed(iter/s)": 0.03464 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9455773477074945, |
|
|
"grad_norm": 1.5703125, |
|
|
"learning_rate": 4.000048196330014e-08, |
|
|
"loss": 0.1817856550216675, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 455, |
|
|
"token_acc": 0.9208294062205467, |
|
|
"train_speed(iter/s)": 0.03485 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.955968307572412, |
|
|
"grad_norm": 1.234375, |
|
|
"learning_rate": 2.611926668199316e-08, |
|
|
"loss": 0.1783198595046997, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 460, |
|
|
"token_acc": 0.9211605415860735, |
|
|
"train_speed(iter/s)": 0.035001 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9663592674373295, |
|
|
"grad_norm": 1.59375, |
|
|
"learning_rate": 1.517330670512629e-08, |
|
|
"loss": 0.17677730321884155, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 465, |
|
|
"token_acc": 0.9180868609125893, |
|
|
"train_speed(iter/s)": 0.03514 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.976750227302247, |
|
|
"grad_norm": 1.984375, |
|
|
"learning_rate": 7.175589395692351e-09, |
|
|
"loss": 0.17313094139099122, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 470, |
|
|
"token_acc": 0.9224636548382054, |
|
|
"train_speed(iter/s)": 0.03528 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9871411871671646, |
|
|
"grad_norm": 2.0625, |
|
|
"learning_rate": 2.1356040317474512e-09, |
|
|
"loss": 0.1789810061454773, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 475, |
|
|
"token_acc": 0.918305714728202, |
|
|
"train_speed(iter/s)": 0.035421 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.997532147032082, |
|
|
"grad_norm": 1.296875, |
|
|
"learning_rate": 5.933054739837296e-11, |
|
|
"loss": 0.17718052864074707, |
|
|
"memory(GiB)": 36.99, |
|
|
"step": 480, |
|
|
"token_acc": 0.9196187450357427, |
|
|
"train_speed(iter/s)": 0.03556 |
|
|
}, |
|
|
{ |
|
|
"epoch": 0.9996103390050656, |
|
|
"eval_loss": 0.16983823478221893, |
|
|
"eval_runtime": 1324.7962, |
|
|
"eval_samples_per_second": 60.675, |
|
|
"eval_steps_per_second": 3.792, |
|
|
"eval_token_acc": 0.9239886680625727, |
|
|
"step": 481 |
|
|
} |
|
|
], |
|
|
"logging_steps": 5, |
|
|
"max_steps": 481, |
|
|
"num_input_tokens_seen": 0, |
|
|
"num_train_epochs": 1, |
|
|
"save_steps": 100, |
|
|
"stateful_callbacks": { |
|
|
"TrainerControl": { |
|
|
"args": { |
|
|
"should_epoch_stop": false, |
|
|
"should_evaluate": false, |
|
|
"should_log": false, |
|
|
"should_save": true, |
|
|
"should_training_stop": true |
|
|
}, |
|
|
"attributes": {} |
|
|
} |
|
|
}, |
|
|
"total_flos": 5.1002858766178714e+17, |
|
|
"train_batch_size": 16, |
|
|
"trial_name": null, |
|
|
"trial_params": null |
|
|
} |
|
|
|
