Files
model-training/llm/Qwen3-1.7B-ru-cpt/checkpoint-1600/trainer_state.json
T
2026-07-19 23:52:25 +04:00

595 lines
14 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.198110214284056,
"eval_steps": 500,
"global_step": 1600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0024763776785507,
"grad_norm": 1.5859375,
"learning_rate": 7.818930041152265e-07,
"loss": 1.9104843139648438,
"step": 20
},
{
"epoch": 0.0049527553571014,
"grad_norm": 1.5703125,
"learning_rate": 1.6049382716049383e-06,
"loss": 1.9554376602172852,
"step": 40
},
{
"epoch": 0.0074291330356521,
"grad_norm": 1.5078125,
"learning_rate": 2.4279835390946504e-06,
"loss": 1.9402402877807616,
"step": 60
},
{
"epoch": 0.0099055107142028,
"grad_norm": 1.515625,
"learning_rate": 3.2510288065843625e-06,
"loss": 1.9371170043945312,
"step": 80
},
{
"epoch": 0.0123818883927535,
"grad_norm": 1.6171875,
"learning_rate": 4.074074074074074e-06,
"loss": 1.9214637756347657,
"step": 100
},
{
"epoch": 0.0148582660713042,
"grad_norm": 1.5234375,
"learning_rate": 4.897119341563787e-06,
"loss": 1.9509721755981446,
"step": 120
},
{
"epoch": 0.0173346437498549,
"grad_norm": 1.46875,
"learning_rate": 5.720164609053498e-06,
"loss": 1.9089406967163085,
"step": 140
},
{
"epoch": 0.0198110214284056,
"grad_norm": 1.4453125,
"learning_rate": 6.543209876543211e-06,
"loss": 1.937429428100586,
"step": 160
},
{
"epoch": 0.0222873991069563,
"grad_norm": 1.4375,
"learning_rate": 7.3662551440329225e-06,
"loss": 1.9281219482421874,
"step": 180
},
{
"epoch": 0.024763776785507,
"grad_norm": 1.4140625,
"learning_rate": 8.189300411522634e-06,
"loss": 1.9310754776000976,
"step": 200
},
{
"epoch": 0.0272401544640577,
"grad_norm": 1.515625,
"learning_rate": 9.012345679012346e-06,
"loss": 1.930306625366211,
"step": 220
},
{
"epoch": 0.0297165321426084,
"grad_norm": 1.4296875,
"learning_rate": 9.835390946502057e-06,
"loss": 1.9268671035766602,
"step": 240
},
{
"epoch": 0.0321929098211591,
"grad_norm": 1.4296875,
"learning_rate": 9.999897077317936e-06,
"loss": 1.933909797668457,
"step": 260
},
{
"epoch": 0.0346692874997098,
"grad_norm": 1.3828125,
"learning_rate": 9.999478961184079e-06,
"loss": 1.9370132446289063,
"step": 280
},
{
"epoch": 0.0371456651782605,
"grad_norm": 1.4296875,
"learning_rate": 9.998739245806088e-06,
"loss": 1.9290205001831056,
"step": 300
},
{
"epoch": 0.0396220428568112,
"grad_norm": 1.3671875,
"learning_rate": 9.997677978767299e-06,
"loss": 1.9694570541381835,
"step": 320
},
{
"epoch": 0.0420984205353619,
"grad_norm": 1.4296875,
"learning_rate": 9.996295228335362e-06,
"loss": 1.915028953552246,
"step": 340
},
{
"epoch": 0.0445747982139126,
"grad_norm": 1.421875,
"learning_rate": 9.994591083457846e-06,
"loss": 1.8872411727905274,
"step": 360
},
{
"epoch": 0.0470511758924633,
"grad_norm": 1.4140625,
"learning_rate": 9.992565653756513e-06,
"loss": 1.9496313095092774,
"step": 380
},
{
"epoch": 0.049527553571014,
"grad_norm": 1.484375,
"learning_rate": 9.990219069520271e-06,
"loss": 1.9227249145507812,
"step": 400
},
{
"epoch": 0.0520039312495647,
"grad_norm": 1.4609375,
"learning_rate": 9.987551481696798e-06,
"loss": 1.951949691772461,
"step": 420
},
{
"epoch": 0.0544803089281154,
"grad_norm": 1.4453125,
"learning_rate": 9.984563061882819e-06,
"loss": 1.9391481399536132,
"step": 440
},
{
"epoch": 0.0569566866066661,
"grad_norm": 1.421875,
"learning_rate": 9.981254002313077e-06,
"loss": 1.910933494567871,
"step": 460
},
{
"epoch": 0.0594330642852168,
"grad_norm": 1.3984375,
"learning_rate": 9.977624515847966e-06,
"loss": 1.8651504516601562,
"step": 480
},
{
"epoch": 0.0619094419637675,
"grad_norm": 1.484375,
"learning_rate": 9.973674835959834e-06,
"loss": 1.9343826293945312,
"step": 500
},
{
"epoch": 0.0643858196423182,
"grad_norm": 1.4765625,
"learning_rate": 9.969405216717972e-06,
"loss": 1.9207374572753906,
"step": 520
},
{
"epoch": 0.0668621973208689,
"grad_norm": 1.375,
"learning_rate": 9.964815932772264e-06,
"loss": 1.9276981353759766,
"step": 540
},
{
"epoch": 0.0693385749994196,
"grad_norm": 1.3828125,
"learning_rate": 9.959907279335525e-06,
"loss": 1.8973264694213867,
"step": 560
},
{
"epoch": 0.07181495267797029,
"grad_norm": 1.3984375,
"learning_rate": 9.954679572164504e-06,
"loss": 1.9223299026489258,
"step": 580
},
{
"epoch": 0.074291330356521,
"grad_norm": 1.46875,
"learning_rate": 9.94913314753958e-06,
"loss": 1.9345746994018556,
"step": 600
},
{
"epoch": 0.0767677080350717,
"grad_norm": 1.3984375,
"learning_rate": 9.943268362243126e-06,
"loss": 1.9141740798950195,
"step": 620
},
{
"epoch": 0.0792440857136224,
"grad_norm": 1.390625,
"learning_rate": 9.937085593536558e-06,
"loss": 1.9250791549682618,
"step": 640
},
{
"epoch": 0.08172046339217309,
"grad_norm": 1.4296875,
"learning_rate": 9.930585239136073e-06,
"loss": 1.9468931198120116,
"step": 660
},
{
"epoch": 0.0841968410707238,
"grad_norm": 1.4296875,
"learning_rate": 9.923767717187051e-06,
"loss": 1.9057769775390625,
"step": 680
},
{
"epoch": 0.0866732187492745,
"grad_norm": 1.3671875,
"learning_rate": 9.916633466237181e-06,
"loss": 1.9412412643432617,
"step": 700
},
{
"epoch": 0.0891495964278252,
"grad_norm": 1.375,
"learning_rate": 9.909182945208224e-06,
"loss": 1.9465457916259765,
"step": 720
},
{
"epoch": 0.0916259741063759,
"grad_norm": 1.390625,
"learning_rate": 9.901416633366511e-06,
"loss": 1.9030138015747071,
"step": 740
},
{
"epoch": 0.0941023517849266,
"grad_norm": 1.3828125,
"learning_rate": 9.893335030292104e-06,
"loss": 1.9357397079467773,
"step": 760
},
{
"epoch": 0.0965787294634773,
"grad_norm": 1.40625,
"learning_rate": 9.884938655846667e-06,
"loss": 1.9043821334838866,
"step": 780
},
{
"epoch": 0.099055107142028,
"grad_norm": 1.3828125,
"learning_rate": 9.876228050140017e-06,
"loss": 1.8731830596923829,
"step": 800
},
{
"epoch": 0.1015314848205787,
"grad_norm": 1.4375,
"learning_rate": 9.867203773495389e-06,
"loss": 1.927505111694336,
"step": 820
},
{
"epoch": 0.1040078624991294,
"grad_norm": 1.40625,
"learning_rate": 9.85786640641338e-06,
"loss": 1.9416738510131837,
"step": 840
},
{
"epoch": 0.1064842401776801,
"grad_norm": 1.4375,
"learning_rate": 9.848216549534626e-06,
"loss": 1.9252901077270508,
"step": 860
},
{
"epoch": 0.1089606178562308,
"grad_norm": 1.4453125,
"learning_rate": 9.838254823601146e-06,
"loss": 1.913006591796875,
"step": 880
},
{
"epoch": 0.1114369955347815,
"grad_norm": 1.4609375,
"learning_rate": 9.827981869416423e-06,
"loss": 1.890475082397461,
"step": 900
},
{
"epoch": 0.1139133732133322,
"grad_norm": 1.4375,
"learning_rate": 9.81739834780418e-06,
"loss": 1.853487777709961,
"step": 920
},
{
"epoch": 0.1163897508918829,
"grad_norm": 1.4921875,
"learning_rate": 9.806504939565873e-06,
"loss": 1.9159151077270509,
"step": 940
},
{
"epoch": 0.1188661285704336,
"grad_norm": 1.4375,
"learning_rate": 9.795302345436895e-06,
"loss": 1.954474639892578,
"step": 960
},
{
"epoch": 0.1213425062489843,
"grad_norm": 1.4609375,
"learning_rate": 9.7837912860415e-06,
"loss": 1.8987157821655274,
"step": 980
},
{
"epoch": 0.123818883927535,
"grad_norm": 1.3828125,
"learning_rate": 9.771972501846451e-06,
"loss": 1.9053752899169922,
"step": 1000
},
{
"epoch": 0.1262952616060857,
"grad_norm": 1.3828125,
"learning_rate": 9.759846753113389e-06,
"loss": 1.9074106216430664,
"step": 1020
},
{
"epoch": 0.1287716392846364,
"grad_norm": 1.40625,
"learning_rate": 9.74741481984992e-06,
"loss": 1.9128997802734375,
"step": 1040
},
{
"epoch": 0.1312480169631871,
"grad_norm": 1.3984375,
"learning_rate": 9.734677501759444e-06,
"loss": 1.9284509658813476,
"step": 1060
},
{
"epoch": 0.1337243946417378,
"grad_norm": 1.4296875,
"learning_rate": 9.721635618189721e-06,
"loss": 1.921483039855957,
"step": 1080
},
{
"epoch": 0.1362007723202885,
"grad_norm": 1.3359375,
"learning_rate": 9.708290008080158e-06,
"loss": 1.9045528411865233,
"step": 1100
},
{
"epoch": 0.1386771499988392,
"grad_norm": 1.3828125,
"learning_rate": 9.694641529907833e-06,
"loss": 1.888129425048828,
"step": 1120
},
{
"epoch": 0.1411535276773899,
"grad_norm": 1.5625,
"learning_rate": 9.680691061632293e-06,
"loss": 1.9138872146606445,
"step": 1140
},
{
"epoch": 0.14362990535594058,
"grad_norm": 1.46875,
"learning_rate": 9.666439500639062e-06,
"loss": 1.9142763137817382,
"step": 1160
},
{
"epoch": 0.1461062830344913,
"grad_norm": 1.359375,
"learning_rate": 9.651887763681921e-06,
"loss": 1.9339773178100585,
"step": 1180
},
{
"epoch": 0.148582660713042,
"grad_norm": 1.3984375,
"learning_rate": 9.637036786823939e-06,
"loss": 1.8940820693969727,
"step": 1200
},
{
"epoch": 0.1510590383915927,
"grad_norm": 1.40625,
"learning_rate": 9.621887525377249e-06,
"loss": 1.924995231628418,
"step": 1220
},
{
"epoch": 0.1535354160701434,
"grad_norm": 1.484375,
"learning_rate": 9.606440953841611e-06,
"loss": 1.9189504623413085,
"step": 1240
},
{
"epoch": 0.1560117937486941,
"grad_norm": 1.4609375,
"learning_rate": 9.590698065841713e-06,
"loss": 1.9165498733520507,
"step": 1260
},
{
"epoch": 0.1584881714272448,
"grad_norm": 1.390625,
"learning_rate": 9.57465987406326e-06,
"loss": 1.944415283203125,
"step": 1280
},
{
"epoch": 0.1609645491057955,
"grad_norm": 1.34375,
"learning_rate": 9.558327410187836e-06,
"loss": 1.8826276779174804,
"step": 1300
},
{
"epoch": 0.16344092678434619,
"grad_norm": 1.3984375,
"learning_rate": 9.541701724826527e-06,
"loss": 1.9053705215454102,
"step": 1320
},
{
"epoch": 0.1659173044628969,
"grad_norm": 1.390625,
"learning_rate": 9.524783887452349e-06,
"loss": 1.8804658889770507,
"step": 1340
},
{
"epoch": 0.1683936821414476,
"grad_norm": 1.3515625,
"learning_rate": 9.50757498633145e-06,
"loss": 1.8912498474121093,
"step": 1360
},
{
"epoch": 0.1708700598199983,
"grad_norm": 1.359375,
"learning_rate": 9.490076128453107e-06,
"loss": 1.894525146484375,
"step": 1380
},
{
"epoch": 0.173346437498549,
"grad_norm": 1.4296875,
"learning_rate": 9.472288439458514e-06,
"loss": 1.8989877700805664,
"step": 1400
},
{
"epoch": 0.1758228151770997,
"grad_norm": 1.46875,
"learning_rate": 9.454213063568377e-06,
"loss": 1.9056276321411132,
"step": 1420
},
{
"epoch": 0.1782991928556504,
"grad_norm": 1.3203125,
"learning_rate": 9.4358511635093e-06,
"loss": 1.8861434936523438,
"step": 1440
},
{
"epoch": 0.1807755705342011,
"grad_norm": 1.4296875,
"learning_rate": 9.417203920439008e-06,
"loss": 1.9283615112304688,
"step": 1460
},
{
"epoch": 0.1832519482127518,
"grad_norm": 1.421875,
"learning_rate": 9.398272533870357e-06,
"loss": 1.8878576278686523,
"step": 1480
},
{
"epoch": 0.1857283258913025,
"grad_norm": 1.515625,
"learning_rate": 9.379058221594168e-06,
"loss": 1.909010887145996,
"step": 1500
},
{
"epoch": 0.1882047035698532,
"grad_norm": 1.390625,
"learning_rate": 9.359562219600903e-06,
"loss": 1.9048601150512696,
"step": 1520
},
{
"epoch": 0.1906810812484039,
"grad_norm": 1.3671875,
"learning_rate": 9.339785782001156e-06,
"loss": 1.8977005004882812,
"step": 1540
},
{
"epoch": 0.1931574589269546,
"grad_norm": 1.40625,
"learning_rate": 9.319730180944966e-06,
"loss": 1.9012252807617187,
"step": 1560
},
{
"epoch": 0.1956338366055053,
"grad_norm": 1.359375,
"learning_rate": 9.299396706540002e-06,
"loss": 1.886746597290039,
"step": 1580
},
{
"epoch": 0.198110214284056,
"grad_norm": 1.359375,
"learning_rate": 9.278786666768566e-06,
"loss": 1.883855438232422,
"step": 1600
}
],
"logging_steps": 20,
"max_steps": 8077,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.433620651474944e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}