abcd2a_control_ft_step8000 / trainer_state.json
happyhappy-jun's picture
Upload folder using huggingface_hub
1b6f8e0 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8,
"eval_steps": 500,
"global_step": 8000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"grad_norm": 0.5420440435409546,
"learning_rate": 1.8e-06,
"loss": 0.0694,
"step": 10
},
{
"grad_norm": 0.7189459204673767,
"learning_rate": 3.8e-06,
"loss": 0.0645,
"step": 20
},
{
"grad_norm": 0.21701538562774658,
"learning_rate": 5.8e-06,
"loss": 0.0317,
"step": 30
},
{
"grad_norm": 0.29029083251953125,
"learning_rate": 7.8e-06,
"loss": 0.0315,
"step": 40
},
{
"grad_norm": 0.10109149664640427,
"learning_rate": 9.800000000000001e-06,
"loss": 0.0273,
"step": 50
},
{
"grad_norm": 0.19640426337718964,
"learning_rate": 1.18e-05,
"loss": 0.0319,
"step": 60
},
{
"grad_norm": 0.1066979169845581,
"learning_rate": 1.3800000000000002e-05,
"loss": 0.0251,
"step": 70
},
{
"grad_norm": 0.12810614705085754,
"learning_rate": 1.58e-05,
"loss": 0.024,
"step": 80
},
{
"grad_norm": 0.08170574903488159,
"learning_rate": 1.78e-05,
"loss": 0.0232,
"step": 90
},
{
"grad_norm": 0.16492632031440735,
"learning_rate": 1.9800000000000004e-05,
"loss": 0.0283,
"step": 100
},
{
"grad_norm": 0.14535479247570038,
"learning_rate": 2.18e-05,
"loss": 0.0231,
"step": 110
},
{
"grad_norm": 0.17147088050842285,
"learning_rate": 2.38e-05,
"loss": 0.0315,
"step": 120
},
{
"grad_norm": 0.10768885165452957,
"learning_rate": 2.58e-05,
"loss": 0.0194,
"step": 130
},
{
"grad_norm": 0.10331500321626663,
"learning_rate": 2.7800000000000005e-05,
"loss": 0.0203,
"step": 140
},
{
"grad_norm": 0.15777619183063507,
"learning_rate": 2.98e-05,
"loss": 0.0285,
"step": 150
},
{
"grad_norm": 0.11859080195426941,
"learning_rate": 3.18e-05,
"loss": 0.0234,
"step": 160
},
{
"grad_norm": 0.20379579067230225,
"learning_rate": 3.38e-05,
"loss": 0.0246,
"step": 170
},
{
"grad_norm": 0.08920297771692276,
"learning_rate": 3.58e-05,
"loss": 0.0209,
"step": 180
},
{
"grad_norm": 0.11533823609352112,
"learning_rate": 3.7800000000000004e-05,
"loss": 0.0261,
"step": 190
},
{
"grad_norm": 0.1448919177055359,
"learning_rate": 3.9800000000000005e-05,
"loss": 0.0264,
"step": 200
},
{
"grad_norm": 0.1489187777042389,
"learning_rate": 4.18e-05,
"loss": 0.0249,
"step": 210
},
{
"grad_norm": 0.1333455890417099,
"learning_rate": 4.38e-05,
"loss": 0.0281,
"step": 220
},
{
"grad_norm": 0.09986919909715652,
"learning_rate": 4.58e-05,
"loss": 0.0241,
"step": 230
},
{
"grad_norm": 0.15185418725013733,
"learning_rate": 4.78e-05,
"loss": 0.0279,
"step": 240
},
{
"grad_norm": 0.16237834095954895,
"learning_rate": 4.9800000000000004e-05,
"loss": 0.0294,
"step": 250
},
{
"grad_norm": 0.25100767612457275,
"learning_rate": 5.1800000000000005e-05,
"loss": 0.0292,
"step": 260
},
{
"grad_norm": 0.1342630535364151,
"learning_rate": 5.380000000000001e-05,
"loss": 0.0291,
"step": 270
},
{
"grad_norm": 0.1496191918849945,
"learning_rate": 5.580000000000001e-05,
"loss": 0.0332,
"step": 280
},
{
"grad_norm": 0.1313694268465042,
"learning_rate": 5.7799999999999995e-05,
"loss": 0.0251,
"step": 290
},
{
"grad_norm": 0.13200165331363678,
"learning_rate": 5.9800000000000003e-05,
"loss": 0.0307,
"step": 300
},
{
"grad_norm": 0.16884654760360718,
"learning_rate": 6.18e-05,
"loss": 0.0266,
"step": 310
},
{
"grad_norm": 0.21465812623500824,
"learning_rate": 6.38e-05,
"loss": 0.0309,
"step": 320
},
{
"grad_norm": 0.22422099113464355,
"learning_rate": 6.58e-05,
"loss": 0.0345,
"step": 330
},
{
"grad_norm": 0.16468451917171478,
"learning_rate": 6.780000000000001e-05,
"loss": 0.0269,
"step": 340
},
{
"grad_norm": 0.1649431437253952,
"learning_rate": 6.98e-05,
"loss": 0.0285,
"step": 350
},
{
"grad_norm": 0.18796072900295258,
"learning_rate": 7.18e-05,
"loss": 0.0294,
"step": 360
},
{
"grad_norm": 0.1406984180212021,
"learning_rate": 7.38e-05,
"loss": 0.0262,
"step": 370
},
{
"grad_norm": 0.12663526833057404,
"learning_rate": 7.58e-05,
"loss": 0.0287,
"step": 380
},
{
"grad_norm": 0.17990556359291077,
"learning_rate": 7.780000000000001e-05,
"loss": 0.0358,
"step": 390
},
{
"grad_norm": 0.14493271708488464,
"learning_rate": 7.98e-05,
"loss": 0.0262,
"step": 400
},
{
"grad_norm": 0.18220767378807068,
"learning_rate": 8.18e-05,
"loss": 0.0307,
"step": 410
},
{
"grad_norm": 0.2229013890028,
"learning_rate": 8.38e-05,
"loss": 0.0393,
"step": 420
},
{
"grad_norm": 0.21163704991340637,
"learning_rate": 8.58e-05,
"loss": 0.0327,
"step": 430
},
{
"grad_norm": 0.1903618574142456,
"learning_rate": 8.78e-05,
"loss": 0.034,
"step": 440
},
{
"grad_norm": 0.19045576453208923,
"learning_rate": 8.98e-05,
"loss": 0.0334,
"step": 450
},
{
"grad_norm": 0.12919433414936066,
"learning_rate": 9.180000000000001e-05,
"loss": 0.0276,
"step": 460
},
{
"grad_norm": 0.22944214940071106,
"learning_rate": 9.38e-05,
"loss": 0.0366,
"step": 470
},
{
"grad_norm": 0.2297431230545044,
"learning_rate": 9.58e-05,
"loss": 0.031,
"step": 480
},
{
"grad_norm": 0.11911984533071518,
"learning_rate": 9.78e-05,
"loss": 0.0349,
"step": 490
},
{
"grad_norm": 0.14332818984985352,
"learning_rate": 9.98e-05,
"loss": 0.0298,
"step": 500
},
{
"grad_norm": 0.23667897284030914,
"learning_rate": 9.9999778549206e-05,
"loss": 0.0325,
"step": 510
},
{
"grad_norm": 0.1407025307416916,
"learning_rate": 9.999901304280685e-05,
"loss": 0.0275,
"step": 520
},
{
"grad_norm": 0.2208714336156845,
"learning_rate": 9.999770075521164e-05,
"loss": 0.0351,
"step": 530
},
{
"grad_norm": 0.18774442374706268,
"learning_rate": 9.99958417007713e-05,
"loss": 0.0415,
"step": 540
},
{
"grad_norm": 0.17202721536159515,
"learning_rate": 9.999343589981615e-05,
"loss": 0.0279,
"step": 550
},
{
"grad_norm": 0.1869555562734604,
"learning_rate": 9.999048337865568e-05,
"loss": 0.0325,
"step": 560
},
{
"grad_norm": 0.184967041015625,
"learning_rate": 9.998698416957815e-05,
"loss": 0.0426,
"step": 570
},
{
"grad_norm": 0.16213007271289825,
"learning_rate": 9.998293831085037e-05,
"loss": 0.0329,
"step": 580
},
{
"grad_norm": 0.1410742849111557,
"learning_rate": 9.997834584671719e-05,
"loss": 0.0313,
"step": 590
},
{
"grad_norm": 0.1963675171136856,
"learning_rate": 9.997320682740107e-05,
"loss": 0.0341,
"step": 600
},
{
"grad_norm": 0.18096579611301422,
"learning_rate": 9.996752130910149e-05,
"loss": 0.0378,
"step": 610
},
{
"grad_norm": 0.16188250482082367,
"learning_rate": 9.99612893539944e-05,
"loss": 0.0355,
"step": 620
},
{
"grad_norm": 0.20988130569458008,
"learning_rate": 9.995451103023144e-05,
"loss": 0.0327,
"step": 630
},
{
"grad_norm": 0.13830691576004028,
"learning_rate": 9.994718641193928e-05,
"loss": 0.0303,
"step": 640
},
{
"grad_norm": 0.10543777048587799,
"learning_rate": 9.993931557921874e-05,
"loss": 0.0311,
"step": 650
},
{
"grad_norm": 0.15627624094486237,
"learning_rate": 9.993089861814402e-05,
"loss": 0.0311,
"step": 660
},
{
"grad_norm": 0.14420993626117706,
"learning_rate": 9.992193562076166e-05,
"loss": 0.0256,
"step": 670
},
{
"grad_norm": 0.18450821936130524,
"learning_rate": 9.991242668508954e-05,
"loss": 0.04,
"step": 680
},
{
"grad_norm": 0.16379375755786896,
"learning_rate": 9.990237191511587e-05,
"loss": 0.0284,
"step": 690
},
{
"grad_norm": 0.14060579240322113,
"learning_rate": 9.989177142079802e-05,
"loss": 0.0276,
"step": 700
},
{
"grad_norm": 0.22451449930667877,
"learning_rate": 9.988062531806126e-05,
"loss": 0.0337,
"step": 710
},
{
"grad_norm": 0.13399986922740936,
"learning_rate": 9.986893372879762e-05,
"loss": 0.0292,
"step": 720
},
{
"grad_norm": 0.14917080104351044,
"learning_rate": 9.985669678086443e-05,
"loss": 0.0369,
"step": 730
},
{
"grad_norm": 0.2764376699924469,
"learning_rate": 9.984391460808298e-05,
"loss": 0.0334,
"step": 740
},
{
"grad_norm": 0.15942853689193726,
"learning_rate": 9.983058735023709e-05,
"loss": 0.0314,
"step": 750
},
{
"grad_norm": 0.19267232716083527,
"learning_rate": 9.98167151530715e-05,
"loss": 0.0302,
"step": 760
},
{
"grad_norm": 0.1857871413230896,
"learning_rate": 9.980229816829034e-05,
"loss": 0.0324,
"step": 770
},
{
"grad_norm": 0.09672777354717255,
"learning_rate": 9.978733655355544e-05,
"loss": 0.0336,
"step": 780
},
{
"grad_norm": 0.1687411516904831,
"learning_rate": 9.977183047248464e-05,
"loss": 0.0264,
"step": 790
},
{
"grad_norm": 0.15346349775791168,
"learning_rate": 9.975578009464992e-05,
"loss": 0.0342,
"step": 800
},
{
"grad_norm": 0.12507975101470947,
"learning_rate": 9.97391855955757e-05,
"loss": 0.0282,
"step": 810
},
{
"grad_norm": 0.1824805736541748,
"learning_rate": 9.972204715673669e-05,
"loss": 0.0475,
"step": 820
},
{
"grad_norm": 0.1841856837272644,
"learning_rate": 9.970436496555617e-05,
"loss": 0.0357,
"step": 830
},
{
"grad_norm": 0.19906769692897797,
"learning_rate": 9.968613921540373e-05,
"loss": 0.0295,
"step": 840
},
{
"grad_norm": 0.15382325649261475,
"learning_rate": 9.966737010559326e-05,
"loss": 0.0308,
"step": 850
},
{
"grad_norm": 0.16068829596042633,
"learning_rate": 9.964805784138072e-05,
"loss": 0.0353,
"step": 860
},
{
"grad_norm": 0.12692637741565704,
"learning_rate": 9.962820263396195e-05,
"loss": 0.0269,
"step": 870
},
{
"grad_norm": 0.17217664420604706,
"learning_rate": 9.960780470047033e-05,
"loss": 0.0329,
"step": 880
},
{
"grad_norm": 0.1513223648071289,
"learning_rate": 9.958686426397437e-05,
"loss": 0.0297,
"step": 890
},
{
"grad_norm": 0.14868152141571045,
"learning_rate": 9.956538155347534e-05,
"loss": 0.028,
"step": 900
},
{
"grad_norm": 0.16328009963035583,
"learning_rate": 9.95433568039047e-05,
"loss": 0.0308,
"step": 910
},
{
"grad_norm": 0.15237754583358765,
"learning_rate": 9.952079025612162e-05,
"loss": 0.0342,
"step": 920
},
{
"grad_norm": 0.17242884635925293,
"learning_rate": 9.949768215691022e-05,
"loss": 0.0404,
"step": 930
},
{
"grad_norm": 0.15023668110370636,
"learning_rate": 9.9474032758977e-05,
"loss": 0.0228,
"step": 940
},
{
"grad_norm": 0.16494718194007874,
"learning_rate": 9.944984232094794e-05,
"loss": 0.0287,
"step": 950
},
{
"grad_norm": 0.149717316031456,
"learning_rate": 9.942511110736584e-05,
"loss": 0.0337,
"step": 960
},
{
"grad_norm": 0.14683957397937775,
"learning_rate": 9.939983938868726e-05,
"loss": 0.0281,
"step": 970
},
{
"grad_norm": 0.18098339438438416,
"learning_rate": 9.93740274412797e-05,
"loss": 0.0309,
"step": 980
},
{
"grad_norm": 0.17116370797157288,
"learning_rate": 9.934767554741846e-05,
"loss": 0.0291,
"step": 990
},
{
"grad_norm": 0.13913299143314362,
"learning_rate": 9.932078399528361e-05,
"loss": 0.0267,
"step": 1000
},
{
"grad_norm": 0.1686856746673584,
"learning_rate": 9.929335307895689e-05,
"loss": 0.0363,
"step": 1010
},
{
"grad_norm": 0.13024087250232697,
"learning_rate": 9.926538309841839e-05,
"loss": 0.0285,
"step": 1020
},
{
"grad_norm": 0.17685119807720184,
"learning_rate": 9.923687435954334e-05,
"loss": 0.0287,
"step": 1030
},
{
"grad_norm": 0.20717529952526093,
"learning_rate": 9.920782717409873e-05,
"loss": 0.027,
"step": 1040
},
{
"grad_norm": 0.15997067093849182,
"learning_rate": 9.917824185973994e-05,
"loss": 0.0283,
"step": 1050
},
{
"grad_norm": 0.13489104807376862,
"learning_rate": 9.914811874000723e-05,
"loss": 0.0339,
"step": 1060
},
{
"grad_norm": 0.1213482990860939,
"learning_rate": 9.911745814432218e-05,
"loss": 0.0247,
"step": 1070
},
{
"grad_norm": 0.13703866302967072,
"learning_rate": 9.90862604079842e-05,
"loss": 0.0288,
"step": 1080
},
{
"grad_norm": 0.18348318338394165,
"learning_rate": 9.90545258721667e-05,
"loss": 0.0297,
"step": 1090
},
{
"grad_norm": 0.11433199793100357,
"learning_rate": 9.90222548839135e-05,
"loss": 0.0268,
"step": 1100
},
{
"grad_norm": 0.17286507785320282,
"learning_rate": 9.898944779613495e-05,
"loss": 0.0267,
"step": 1110
},
{
"grad_norm": 0.1405310034751892,
"learning_rate": 9.89561049676041e-05,
"loss": 0.0264,
"step": 1120
},
{
"grad_norm": 0.10237786918878555,
"learning_rate": 9.89222267629528e-05,
"loss": 0.0272,
"step": 1130
},
{
"grad_norm": 0.17079147696495056,
"learning_rate": 9.888781355266763e-05,
"loss": 0.0239,
"step": 1140
},
{
"grad_norm": 0.1960420310497284,
"learning_rate": 9.885286571308598e-05,
"loss": 0.0292,
"step": 1150
},
{
"grad_norm": 0.1341986507177353,
"learning_rate": 9.881738362639182e-05,
"loss": 0.0237,
"step": 1160
},
{
"grad_norm": 0.14957411587238312,
"learning_rate": 9.878136768061154e-05,
"loss": 0.0304,
"step": 1170
},
{
"grad_norm": 0.23255684971809387,
"learning_rate": 9.874481826960979e-05,
"loss": 0.0292,
"step": 1180
},
{
"grad_norm": 0.12505985796451569,
"learning_rate": 9.870773579308503e-05,
"loss": 0.0304,
"step": 1190
},
{
"grad_norm": 0.13537172973155975,
"learning_rate": 9.867012065656533e-05,
"loss": 0.0304,
"step": 1200
},
{
"grad_norm": 0.16207456588745117,
"learning_rate": 9.863197327140376e-05,
"loss": 0.035,
"step": 1210
},
{
"grad_norm": 0.22213716804981232,
"learning_rate": 9.859329405477403e-05,
"loss": 0.031,
"step": 1220
},
{
"grad_norm": 0.18120811879634857,
"learning_rate": 9.855408342966585e-05,
"loss": 0.0267,
"step": 1230
},
{
"grad_norm": 0.18635620176792145,
"learning_rate": 9.851434182488033e-05,
"loss": 0.0308,
"step": 1240
},
{
"grad_norm": 0.14335626363754272,
"learning_rate": 9.84740696750253e-05,
"loss": 0.0229,
"step": 1250
},
{
"grad_norm": 0.12551842629909515,
"learning_rate": 9.843326742051055e-05,
"loss": 0.0224,
"step": 1260
},
{
"grad_norm": 0.1251303106546402,
"learning_rate": 9.839193550754297e-05,
"loss": 0.0232,
"step": 1270
},
{
"grad_norm": 0.1565120369195938,
"learning_rate": 9.835007438812177e-05,
"loss": 0.03,
"step": 1280
},
{
"grad_norm": 0.15832260251045227,
"learning_rate": 9.830768452003341e-05,
"loss": 0.0271,
"step": 1290
},
{
"grad_norm": 0.13791042566299438,
"learning_rate": 9.826476636684671e-05,
"loss": 0.0276,
"step": 1300
},
{
"grad_norm": 0.10483084619045258,
"learning_rate": 9.822132039790773e-05,
"loss": 0.0248,
"step": 1310
},
{
"grad_norm": 0.14374341070652008,
"learning_rate": 9.817734708833461e-05,
"loss": 0.0269,
"step": 1320
},
{
"grad_norm": 0.16665680706501007,
"learning_rate": 9.813284691901243e-05,
"loss": 0.0307,
"step": 1330
},
{
"grad_norm": 0.21279016137123108,
"learning_rate": 9.808782037658792e-05,
"loss": 0.0333,
"step": 1340
},
{
"grad_norm": 0.1460651457309723,
"learning_rate": 9.804226795346411e-05,
"loss": 0.0242,
"step": 1350
},
{
"grad_norm": 0.20963597297668457,
"learning_rate": 9.799619014779503e-05,
"loss": 0.0321,
"step": 1360
},
{
"grad_norm": 0.19664771854877472,
"learning_rate": 9.794958746348013e-05,
"loss": 0.0257,
"step": 1370
},
{
"grad_norm": 0.2412680685520172,
"learning_rate": 9.790246041015896e-05,
"loss": 0.0265,
"step": 1380
},
{
"grad_norm": 0.2061948925256729,
"learning_rate": 9.785480950320538e-05,
"loss": 0.0289,
"step": 1390
},
{
"grad_norm": 0.1481916904449463,
"learning_rate": 9.78066352637221e-05,
"loss": 0.027,
"step": 1400
},
{
"grad_norm": 0.15597012639045715,
"learning_rate": 9.775793821853488e-05,
"loss": 0.0373,
"step": 1410
},
{
"grad_norm": 0.14834840595722198,
"learning_rate": 9.77087189001868e-05,
"loss": 0.0296,
"step": 1420
},
{
"grad_norm": 0.18620194494724274,
"learning_rate": 9.765897784693243e-05,
"loss": 0.0287,
"step": 1430
},
{
"grad_norm": 0.13414987921714783,
"learning_rate": 9.760871560273197e-05,
"loss": 0.0227,
"step": 1440
},
{
"grad_norm": 0.1416088491678238,
"learning_rate": 9.755793271724526e-05,
"loss": 0.0346,
"step": 1450
},
{
"grad_norm": 0.10656988620758057,
"learning_rate": 9.750662974582584e-05,
"loss": 0.027,
"step": 1460
},
{
"grad_norm": 0.12058360874652863,
"learning_rate": 9.745480724951473e-05,
"loss": 0.0199,
"step": 1470
},
{
"grad_norm": 0.21359531581401825,
"learning_rate": 9.740246579503447e-05,
"loss": 0.0251,
"step": 1480
},
{
"grad_norm": 0.15765275061130524,
"learning_rate": 9.734960595478284e-05,
"loss": 0.0245,
"step": 1490
},
{
"grad_norm": 0.13638831675052643,
"learning_rate": 9.729622830682657e-05,
"loss": 0.031,
"step": 1500
},
{
"grad_norm": 0.14469914138317108,
"learning_rate": 9.724233343489504e-05,
"loss": 0.0237,
"step": 1510
},
{
"grad_norm": 0.12858904898166656,
"learning_rate": 9.718792192837396e-05,
"loss": 0.028,
"step": 1520
},
{
"grad_norm": 0.18743933737277985,
"learning_rate": 9.713299438229886e-05,
"loss": 0.0266,
"step": 1530
},
{
"grad_norm": 0.17122581601142883,
"learning_rate": 9.707755139734855e-05,
"loss": 0.0249,
"step": 1540
},
{
"grad_norm": 0.125565305352211,
"learning_rate": 9.702159357983866e-05,
"loss": 0.0313,
"step": 1550
},
{
"grad_norm": 0.11599742621183395,
"learning_rate": 9.696512154171492e-05,
"loss": 0.0229,
"step": 1560
},
{
"grad_norm": 0.17274866998195648,
"learning_rate": 9.690813590054645e-05,
"loss": 0.0226,
"step": 1570
},
{
"grad_norm": 0.10541734844446182,
"learning_rate": 9.685063727951914e-05,
"loss": 0.0247,
"step": 1580
},
{
"grad_norm": 0.27482670545578003,
"learning_rate": 9.679262630742865e-05,
"loss": 0.0271,
"step": 1590
},
{
"grad_norm": 0.1570490002632141,
"learning_rate": 9.673410361867373e-05,
"loss": 0.0282,
"step": 1600
},
{
"grad_norm": 0.1765255182981491,
"learning_rate": 9.667506985324909e-05,
"loss": 0.0292,
"step": 1610
},
{
"grad_norm": 0.12382631003856659,
"learning_rate": 9.661552565673855e-05,
"loss": 0.0276,
"step": 1620
},
{
"grad_norm": 0.1488092988729477,
"learning_rate": 9.655547168030789e-05,
"loss": 0.0297,
"step": 1630
},
{
"grad_norm": 0.17621779441833496,
"learning_rate": 9.649490858069777e-05,
"loss": 0.0246,
"step": 1640
},
{
"grad_norm": 0.1317397505044937,
"learning_rate": 9.643383702021658e-05,
"loss": 0.0228,
"step": 1650
},
{
"grad_norm": 0.18702450394630432,
"learning_rate": 9.637225766673307e-05,
"loss": 0.0326,
"step": 1660
},
{
"grad_norm": 0.15152408182621002,
"learning_rate": 9.631017119366922e-05,
"loss": 0.025,
"step": 1670
},
{
"grad_norm": 0.14151324331760406,
"learning_rate": 9.624757827999273e-05,
"loss": 0.0234,
"step": 1680
},
{
"grad_norm": 0.17372380197048187,
"learning_rate": 9.618447961020971e-05,
"loss": 0.0361,
"step": 1690
},
{
"grad_norm": 0.14488667249679565,
"learning_rate": 9.612087587435707e-05,
"loss": 0.0248,
"step": 1700
},
{
"grad_norm": 0.1468273550271988,
"learning_rate": 9.605676776799508e-05,
"loss": 0.0236,
"step": 1710
},
{
"grad_norm": 0.16720327734947205,
"learning_rate": 9.599215599219973e-05,
"loss": 0.0299,
"step": 1720
},
{
"grad_norm": 0.10409308969974518,
"learning_rate": 9.592704125355505e-05,
"loss": 0.0224,
"step": 1730
},
{
"grad_norm": 0.22400495409965515,
"learning_rate": 9.586142426414538e-05,
"loss": 0.0298,
"step": 1740
},
{
"grad_norm": 0.16129139065742493,
"learning_rate": 9.57953057415476e-05,
"loss": 0.0216,
"step": 1750
},
{
"grad_norm": 0.15150457620620728,
"learning_rate": 9.572868640882328e-05,
"loss": 0.0255,
"step": 1760
},
{
"grad_norm": 0.27912425994873047,
"learning_rate": 9.56615669945108e-05,
"loss": 0.0265,
"step": 1770
},
{
"grad_norm": 0.15684591233730316,
"learning_rate": 9.55939482326173e-05,
"loss": 0.0231,
"step": 1780
},
{
"grad_norm": 0.1414581537246704,
"learning_rate": 9.552583086261069e-05,
"loss": 0.0298,
"step": 1790
},
{
"grad_norm": 0.12455716729164124,
"learning_rate": 9.545721562941168e-05,
"loss": 0.0311,
"step": 1800
},
{
"grad_norm": 0.1510583609342575,
"learning_rate": 9.538810328338543e-05,
"loss": 0.0245,
"step": 1810
},
{
"grad_norm": 0.12179003655910492,
"learning_rate": 9.531849458033349e-05,
"loss": 0.0195,
"step": 1820
},
{
"grad_norm": 0.15041644871234894,
"learning_rate": 9.524839028148547e-05,
"loss": 0.0237,
"step": 1830
},
{
"grad_norm": 0.18131820857524872,
"learning_rate": 9.517779115349077e-05,
"loss": 0.027,
"step": 1840
},
{
"grad_norm": 0.13325054943561554,
"learning_rate": 9.510669796841014e-05,
"loss": 0.0222,
"step": 1850
},
{
"grad_norm": 0.14594273269176483,
"learning_rate": 9.503511150370727e-05,
"loss": 0.0252,
"step": 1860
},
{
"grad_norm": 0.1101432591676712,
"learning_rate": 9.496303254224024e-05,
"loss": 0.0196,
"step": 1870
},
{
"grad_norm": 0.11899697780609131,
"learning_rate": 9.489046187225306e-05,
"loss": 0.0238,
"step": 1880
},
{
"grad_norm": 0.1834312528371811,
"learning_rate": 9.481740028736692e-05,
"loss": 0.027,
"step": 1890
},
{
"grad_norm": 0.1628856211900711,
"learning_rate": 9.474384858657164e-05,
"loss": 0.033,
"step": 1900
},
{
"grad_norm": 0.16760051250457764,
"learning_rate": 9.466980757421679e-05,
"loss": 0.0237,
"step": 1910
},
{
"grad_norm": 0.15349121391773224,
"learning_rate": 9.459527806000305e-05,
"loss": 0.0201,
"step": 1920
},
{
"grad_norm": 0.18761931359767914,
"learning_rate": 9.452026085897325e-05,
"loss": 0.0263,
"step": 1930
},
{
"grad_norm": 0.13588790595531464,
"learning_rate": 9.444475679150348e-05,
"loss": 0.0199,
"step": 1940
},
{
"grad_norm": 0.11293373256921768,
"learning_rate": 9.436876668329411e-05,
"loss": 0.0265,
"step": 1950
},
{
"grad_norm": 0.15826141834259033,
"learning_rate": 9.429229136536079e-05,
"loss": 0.0253,
"step": 1960
},
{
"grad_norm": 0.1520809531211853,
"learning_rate": 9.421533167402534e-05,
"loss": 0.0245,
"step": 1970
},
{
"grad_norm": 0.12347332388162613,
"learning_rate": 9.413788845090666e-05,
"loss": 0.0294,
"step": 1980
},
{
"grad_norm": 0.13870365917682648,
"learning_rate": 9.405996254291136e-05,
"loss": 0.022,
"step": 1990
},
{
"grad_norm": 0.1662704199552536,
"learning_rate": 9.398155480222474e-05,
"loss": 0.0267,
"step": 2000
},
{
"grad_norm": 0.13557152450084686,
"learning_rate": 9.390266608630128e-05,
"loss": 0.0227,
"step": 2010
},
{
"grad_norm": 0.0978504940867424,
"learning_rate": 9.38232972578553e-05,
"loss": 0.0228,
"step": 2020
},
{
"grad_norm": 0.1562250852584839,
"learning_rate": 9.374344918485164e-05,
"loss": 0.022,
"step": 2030
},
{
"grad_norm": 0.12346863746643066,
"learning_rate": 9.366312274049602e-05,
"loss": 0.0227,
"step": 2040
},
{
"grad_norm": 0.12018659710884094,
"learning_rate": 9.358231880322554e-05,
"loss": 0.0217,
"step": 2050
},
{
"grad_norm": 0.1706809550523758,
"learning_rate": 9.350103825669916e-05,
"loss": 0.025,
"step": 2060
},
{
"grad_norm": 0.1152491346001625,
"learning_rate": 9.341928198978787e-05,
"loss": 0.0226,
"step": 2070
},
{
"grad_norm": 0.1519344598054886,
"learning_rate": 9.333705089656512e-05,
"loss": 0.0242,
"step": 2080
},
{
"grad_norm": 0.16569732129573822,
"learning_rate": 9.325434587629698e-05,
"loss": 0.0196,
"step": 2090
},
{
"grad_norm": 0.16990657150745392,
"learning_rate": 9.31711678334323e-05,
"loss": 0.022,
"step": 2100
},
{
"grad_norm": 0.16231462359428406,
"learning_rate": 9.308751767759282e-05,
"loss": 0.0248,
"step": 2110
},
{
"grad_norm": 0.2376696765422821,
"learning_rate": 9.300339632356325e-05,
"loss": 0.0329,
"step": 2120
},
{
"grad_norm": 0.2505277395248413,
"learning_rate": 9.291880469128124e-05,
"loss": 0.0299,
"step": 2130
},
{
"grad_norm": 0.196861132979393,
"learning_rate": 9.283374370582732e-05,
"loss": 0.0239,
"step": 2140
},
{
"grad_norm": 0.13766983151435852,
"learning_rate": 9.274821429741482e-05,
"loss": 0.0217,
"step": 2150
},
{
"grad_norm": 0.1426539272069931,
"learning_rate": 9.266221740137961e-05,
"loss": 0.0309,
"step": 2160
},
{
"grad_norm": 0.14772582054138184,
"learning_rate": 9.257575395817001e-05,
"loss": 0.0192,
"step": 2170
},
{
"grad_norm": 0.12440606206655502,
"learning_rate": 9.248882491333637e-05,
"loss": 0.0172,
"step": 2180
},
{
"grad_norm": 0.17530487477779388,
"learning_rate": 9.240143121752076e-05,
"loss": 0.0234,
"step": 2190
},
{
"grad_norm": 0.12355060875415802,
"learning_rate": 9.23135738264467e-05,
"loss": 0.0218,
"step": 2200
},
{
"grad_norm": 0.21175123751163483,
"learning_rate": 9.222525370090849e-05,
"loss": 0.0274,
"step": 2210
},
{
"grad_norm": 0.14347153902053833,
"learning_rate": 9.213647180676088e-05,
"loss": 0.0238,
"step": 2220
},
{
"grad_norm": 0.11950457096099854,
"learning_rate": 9.204722911490846e-05,
"loss": 0.0278,
"step": 2230
},
{
"grad_norm": 0.14002735912799835,
"learning_rate": 9.1957526601295e-05,
"loss": 0.0235,
"step": 2240
},
{
"grad_norm": 0.13814957439899445,
"learning_rate": 9.186736524689281e-05,
"loss": 0.0243,
"step": 2250
},
{
"grad_norm": 0.13607311248779297,
"learning_rate": 9.177674603769204e-05,
"loss": 0.0341,
"step": 2260
},
{
"grad_norm": 0.16439314186573029,
"learning_rate": 9.168566996468983e-05,
"loss": 0.0266,
"step": 2270
},
{
"grad_norm": 0.1378590613603592,
"learning_rate": 9.159413802387951e-05,
"loss": 0.0258,
"step": 2280
},
{
"grad_norm": 0.18438801169395447,
"learning_rate": 9.150215121623974e-05,
"loss": 0.027,
"step": 2290
},
{
"grad_norm": 0.12811657786369324,
"learning_rate": 9.140971054772349e-05,
"loss": 0.0233,
"step": 2300
},
{
"grad_norm": 0.14623720943927765,
"learning_rate": 9.131681702924713e-05,
"loss": 0.0204,
"step": 2310
},
{
"grad_norm": 0.12561063468456268,
"learning_rate": 9.122347167667926e-05,
"loss": 0.0263,
"step": 2320
},
{
"grad_norm": 0.17007383704185486,
"learning_rate": 9.112967551082973e-05,
"loss": 0.0234,
"step": 2330
},
{
"grad_norm": 0.11141370236873627,
"learning_rate": 9.103542955743835e-05,
"loss": 0.0248,
"step": 2340
},
{
"grad_norm": 0.18721242249011993,
"learning_rate": 9.094073484716381e-05,
"loss": 0.0273,
"step": 2350
},
{
"grad_norm": 0.14813850820064545,
"learning_rate": 9.084559241557226e-05,
"loss": 0.0212,
"step": 2360
},
{
"grad_norm": 0.15478885173797607,
"learning_rate": 9.075000330312608e-05,
"loss": 0.0269,
"step": 2370
},
{
"grad_norm": 0.15512995421886444,
"learning_rate": 9.065396855517253e-05,
"loss": 0.0241,
"step": 2380
},
{
"grad_norm": 0.18194761872291565,
"learning_rate": 9.055748922193219e-05,
"loss": 0.0298,
"step": 2390
},
{
"grad_norm": 0.09133775532245636,
"learning_rate": 9.046056635848761e-05,
"loss": 0.0202,
"step": 2400
},
{
"grad_norm": 0.14668187499046326,
"learning_rate": 9.036320102477169e-05,
"loss": 0.0235,
"step": 2410
},
{
"grad_norm": 0.1325305700302124,
"learning_rate": 9.02653942855561e-05,
"loss": 0.0245,
"step": 2420
},
{
"grad_norm": 0.14857754111289978,
"learning_rate": 9.016714721043971e-05,
"loss": 0.0205,
"step": 2430
},
{
"grad_norm": 0.16199085116386414,
"learning_rate": 9.006846087383675e-05,
"loss": 0.0213,
"step": 2440
},
{
"grad_norm": 0.10761445760726929,
"learning_rate": 8.996933635496523e-05,
"loss": 0.0278,
"step": 2450
},
{
"grad_norm": 0.1585974097251892,
"learning_rate": 8.986977473783498e-05,
"loss": 0.0241,
"step": 2460
},
{
"grad_norm": 0.10342201590538025,
"learning_rate": 8.97697771112359e-05,
"loss": 0.0212,
"step": 2470
},
{
"grad_norm": 0.14516714215278625,
"learning_rate": 8.966934456872602e-05,
"loss": 0.0243,
"step": 2480
},
{
"grad_norm": 0.14087602496147156,
"learning_rate": 8.95684782086195e-05,
"loss": 0.0279,
"step": 2490
},
{
"grad_norm": 0.12304757535457611,
"learning_rate": 8.946717913397476e-05,
"loss": 0.027,
"step": 2500
},
{
"grad_norm": 0.1442318856716156,
"learning_rate": 8.93654484525822e-05,
"loss": 0.0239,
"step": 2510
},
{
"grad_norm": 0.15643472969532013,
"learning_rate": 8.926328727695226e-05,
"loss": 0.0205,
"step": 2520
},
{
"grad_norm": 0.09812328964471817,
"learning_rate": 8.916069672430319e-05,
"loss": 0.0214,
"step": 2530
},
{
"grad_norm": 0.1295197308063507,
"learning_rate": 8.905767791654884e-05,
"loss": 0.0302,
"step": 2540
},
{
"grad_norm": 0.13791200518608093,
"learning_rate": 8.895423198028638e-05,
"loss": 0.0199,
"step": 2550
},
{
"grad_norm": 0.16387410461902618,
"learning_rate": 8.885036004678402e-05,
"loss": 0.0289,
"step": 2560
},
{
"grad_norm": 0.15156884491443634,
"learning_rate": 8.874606325196857e-05,
"loss": 0.0265,
"step": 2570
},
{
"grad_norm": 0.12447788566350937,
"learning_rate": 8.864134273641304e-05,
"loss": 0.0243,
"step": 2580
},
{
"grad_norm": 0.11832483112812042,
"learning_rate": 8.853619964532427e-05,
"loss": 0.0219,
"step": 2590
},
{
"grad_norm": 0.15960167348384857,
"learning_rate": 8.843063512853019e-05,
"loss": 0.0262,
"step": 2600
},
{
"grad_norm": 0.16603811085224152,
"learning_rate": 8.832465034046749e-05,
"loss": 0.0228,
"step": 2610
},
{
"grad_norm": 0.1218847781419754,
"learning_rate": 8.821824644016882e-05,
"loss": 0.0205,
"step": 2620
},
{
"grad_norm": 0.1599685251712799,
"learning_rate": 8.811142459125019e-05,
"loss": 0.0259,
"step": 2630
},
{
"grad_norm": 0.10958845168352127,
"learning_rate": 8.800418596189822e-05,
"loss": 0.0236,
"step": 2640
},
{
"grad_norm": 0.12428676337003708,
"learning_rate": 8.789653172485737e-05,
"loss": 0.0208,
"step": 2650
},
{
"grad_norm": 0.12821273505687714,
"learning_rate": 8.778846305741715e-05,
"loss": 0.0205,
"step": 2660
},
{
"grad_norm": 0.11910676211118698,
"learning_rate": 8.767998114139918e-05,
"loss": 0.0261,
"step": 2670
},
{
"grad_norm": 0.1570224165916443,
"learning_rate": 8.757108716314429e-05,
"loss": 0.0206,
"step": 2680
},
{
"grad_norm": 0.1630830466747284,
"learning_rate": 8.746178231349962e-05,
"loss": 0.0203,
"step": 2690
},
{
"grad_norm": 0.07918090373277664,
"learning_rate": 8.735206778780549e-05,
"loss": 0.0222,
"step": 2700
},
{
"grad_norm": 0.1236148476600647,
"learning_rate": 8.724194478588234e-05,
"loss": 0.0194,
"step": 2710
},
{
"grad_norm": 0.15769383311271667,
"learning_rate": 8.713141451201772e-05,
"loss": 0.0252,
"step": 2720
},
{
"grad_norm": 0.10508900135755539,
"learning_rate": 8.702047817495295e-05,
"loss": 0.0171,
"step": 2730
},
{
"grad_norm": 0.12346747517585754,
"learning_rate": 8.69091369878701e-05,
"loss": 0.0191,
"step": 2740
},
{
"grad_norm": 0.1627315729856491,
"learning_rate": 8.679739216837849e-05,
"loss": 0.0228,
"step": 2750
},
{
"grad_norm": 0.111842580139637,
"learning_rate": 8.66852449385016e-05,
"loss": 0.021,
"step": 2760
},
{
"grad_norm": 0.111451655626297,
"learning_rate": 8.657269652466356e-05,
"loss": 0.0188,
"step": 2770
},
{
"grad_norm": 0.12247065454721451,
"learning_rate": 8.645974815767577e-05,
"loss": 0.02,
"step": 2780
},
{
"grad_norm": 0.18082000315189362,
"learning_rate": 8.634640107272351e-05,
"loss": 0.0254,
"step": 2790
},
{
"grad_norm": 0.14393867552280426,
"learning_rate": 8.623265650935234e-05,
"loss": 0.0231,
"step": 2800
},
{
"grad_norm": 0.14712563157081604,
"learning_rate": 8.611851571145456e-05,
"loss": 0.0189,
"step": 2810
},
{
"grad_norm": 0.11501806229352951,
"learning_rate": 8.600397992725566e-05,
"loss": 0.0179,
"step": 2820
},
{
"grad_norm": 0.14018242061138153,
"learning_rate": 8.588905040930061e-05,
"loss": 0.0275,
"step": 2830
},
{
"grad_norm": 0.15879635512828827,
"learning_rate": 8.577372841444022e-05,
"loss": 0.0213,
"step": 2840
},
{
"grad_norm": 0.14423353970050812,
"learning_rate": 8.565801520381736e-05,
"loss": 0.0209,
"step": 2850
},
{
"grad_norm": 0.13850180804729462,
"learning_rate": 8.554191204285313e-05,
"loss": 0.0211,
"step": 2860
},
{
"grad_norm": 0.13924437761306763,
"learning_rate": 8.542542020123315e-05,
"loss": 0.0238,
"step": 2870
},
{
"grad_norm": 0.09380088746547699,
"learning_rate": 8.530854095289347e-05,
"loss": 0.0186,
"step": 2880
},
{
"grad_norm": 0.1399223804473877,
"learning_rate": 8.519127557600688e-05,
"loss": 0.0247,
"step": 2890
},
{
"grad_norm": 0.13163888454437256,
"learning_rate": 8.507362535296871e-05,
"loss": 0.0215,
"step": 2900
},
{
"grad_norm": 0.12561990320682526,
"learning_rate": 8.495559157038299e-05,
"loss": 0.0194,
"step": 2910
},
{
"grad_norm": 0.12927483022212982,
"learning_rate": 8.483717551904823e-05,
"loss": 0.0221,
"step": 2920
},
{
"grad_norm": 0.12346947938203812,
"learning_rate": 8.47183784939434e-05,
"loss": 0.0203,
"step": 2930
},
{
"grad_norm": 0.13767321407794952,
"learning_rate": 8.459920179421374e-05,
"loss": 0.0247,
"step": 2940
},
{
"grad_norm": 0.17473234236240387,
"learning_rate": 8.447964672315656e-05,
"loss": 0.0277,
"step": 2950
},
{
"grad_norm": 0.13432490825653076,
"learning_rate": 8.435971458820692e-05,
"loss": 0.0234,
"step": 2960
},
{
"grad_norm": 0.1274847835302353,
"learning_rate": 8.423940670092345e-05,
"loss": 0.0212,
"step": 2970
},
{
"grad_norm": 0.0896708145737648,
"learning_rate": 8.411872437697394e-05,
"loss": 0.0188,
"step": 2980
},
{
"grad_norm": 0.13214024901390076,
"learning_rate": 8.399766893612096e-05,
"loss": 0.0231,
"step": 2990
},
{
"grad_norm": 0.13123571872711182,
"learning_rate": 8.38762417022074e-05,
"loss": 0.0316,
"step": 3000
},
{
"grad_norm": 0.13476744294166565,
"learning_rate": 8.375444400314204e-05,
"loss": 0.0203,
"step": 3010
},
{
"grad_norm": 0.12753205001354218,
"learning_rate": 8.3632277170885e-05,
"loss": 0.0222,
"step": 3020
},
{
"grad_norm": 0.10575438290834427,
"learning_rate": 8.350974254143318e-05,
"loss": 0.021,
"step": 3030
},
{
"grad_norm": 0.1535600870847702,
"learning_rate": 8.338684145480566e-05,
"loss": 0.0215,
"step": 3040
},
{
"grad_norm": 0.21555498242378235,
"learning_rate": 8.326357525502904e-05,
"loss": 0.0271,
"step": 3050
},
{
"grad_norm": 0.14917391538619995,
"learning_rate": 8.313994529012273e-05,
"loss": 0.0196,
"step": 3060
},
{
"grad_norm": 0.10461319237947464,
"learning_rate": 8.301595291208422e-05,
"loss": 0.0185,
"step": 3070
},
{
"grad_norm": 0.21023470163345337,
"learning_rate": 8.289159947687427e-05,
"loss": 0.0355,
"step": 3080
},
{
"grad_norm": 0.15654537081718445,
"learning_rate": 8.276688634440216e-05,
"loss": 0.0245,
"step": 3090
},
{
"grad_norm": 0.1336844116449356,
"learning_rate": 8.26418148785107e-05,
"loss": 0.0261,
"step": 3100
},
{
"grad_norm": 0.15251711010932922,
"learning_rate": 8.251638644696141e-05,
"loss": 0.021,
"step": 3110
},
{
"grad_norm": 0.1364535242319107,
"learning_rate": 8.23906024214195e-05,
"loss": 0.0233,
"step": 3120
},
{
"grad_norm": 0.14940956234931946,
"learning_rate": 8.226446417743897e-05,
"loss": 0.0187,
"step": 3130
},
{
"grad_norm": 0.12878847122192383,
"learning_rate": 8.213797309444742e-05,
"loss": 0.0186,
"step": 3140
},
{
"grad_norm": 0.1497132033109665,
"learning_rate": 8.201113055573105e-05,
"loss": 0.02,
"step": 3150
},
{
"grad_norm": 0.1366586536169052,
"learning_rate": 8.188393794841958e-05,
"loss": 0.0172,
"step": 3160
},
{
"grad_norm": 0.13715018332004547,
"learning_rate": 8.175639666347094e-05,
"loss": 0.0258,
"step": 3170
},
{
"grad_norm": 0.15244945883750916,
"learning_rate": 8.162850809565623e-05,
"loss": 0.023,
"step": 3180
},
{
"grad_norm": 0.20791588723659515,
"learning_rate": 8.150027364354431e-05,
"loss": 0.0182,
"step": 3190
},
{
"grad_norm": 0.11517085880041122,
"learning_rate": 8.137169470948662e-05,
"loss": 0.0238,
"step": 3200
},
{
"grad_norm": 0.1119120642542839,
"learning_rate": 8.124277269960179e-05,
"loss": 0.0192,
"step": 3210
},
{
"grad_norm": 0.10967746376991272,
"learning_rate": 8.111350902376023e-05,
"loss": 0.0187,
"step": 3220
},
{
"grad_norm": 0.1142214760184288,
"learning_rate": 8.098390509556883e-05,
"loss": 0.0183,
"step": 3230
},
{
"grad_norm": 0.12059076130390167,
"learning_rate": 8.085396233235536e-05,
"loss": 0.0164,
"step": 3240
},
{
"grad_norm": 0.10252851992845535,
"learning_rate": 8.072368215515306e-05,
"loss": 0.0198,
"step": 3250
},
{
"grad_norm": 0.13169099390506744,
"learning_rate": 8.059306598868506e-05,
"loss": 0.0166,
"step": 3260
},
{
"grad_norm": 0.12112400680780411,
"learning_rate": 8.046211526134888e-05,
"loss": 0.0206,
"step": 3270
},
{
"grad_norm": 0.10669714212417603,
"learning_rate": 8.033083140520065e-05,
"loss": 0.0197,
"step": 3280
},
{
"grad_norm": 0.19681450724601746,
"learning_rate": 8.019921585593962e-05,
"loss": 0.0202,
"step": 3290
},
{
"grad_norm": 0.1599496304988861,
"learning_rate": 8.006727005289232e-05,
"loss": 0.0282,
"step": 3300
},
{
"grad_norm": 0.14205233752727509,
"learning_rate": 7.993499543899692e-05,
"loss": 0.017,
"step": 3310
},
{
"grad_norm": 0.14852288365364075,
"learning_rate": 7.980239346078742e-05,
"loss": 0.0176,
"step": 3320
},
{
"grad_norm": 0.12392748147249222,
"learning_rate": 7.966946556837778e-05,
"loss": 0.0204,
"step": 3330
},
{
"grad_norm": 0.12351145595312119,
"learning_rate": 7.953621321544616e-05,
"loss": 0.0196,
"step": 3340
},
{
"grad_norm": 0.13077925145626068,
"learning_rate": 7.940263785921896e-05,
"loss": 0.0199,
"step": 3350
},
{
"grad_norm": 0.09509874135255814,
"learning_rate": 7.926874096045482e-05,
"loss": 0.0235,
"step": 3360
},
{
"grad_norm": 0.12468800693750381,
"learning_rate": 7.913452398342881e-05,
"loss": 0.0271,
"step": 3370
},
{
"grad_norm": 0.09213206171989441,
"learning_rate": 7.89999883959163e-05,
"loss": 0.0179,
"step": 3380
},
{
"grad_norm": 0.13299454748630524,
"learning_rate": 7.886513566917687e-05,
"loss": 0.0193,
"step": 3390
},
{
"grad_norm": 0.13831903040409088,
"learning_rate": 7.872996727793838e-05,
"loss": 0.0199,
"step": 3400
},
{
"grad_norm": 0.14821100234985352,
"learning_rate": 7.859448470038069e-05,
"loss": 0.0177,
"step": 3410
},
{
"grad_norm": 0.1522444635629654,
"learning_rate": 7.845868941811956e-05,
"loss": 0.0201,
"step": 3420
},
{
"grad_norm": 0.13491684198379517,
"learning_rate": 7.832258291619043e-05,
"loss": 0.0204,
"step": 3430
},
{
"grad_norm": 0.13357700407505035,
"learning_rate": 7.81861666830322e-05,
"loss": 0.0292,
"step": 3440
},
{
"grad_norm": 0.16105544567108154,
"learning_rate": 7.804944221047097e-05,
"loss": 0.0233,
"step": 3450
},
{
"grad_norm": 0.10065405815839767,
"learning_rate": 7.791241099370364e-05,
"loss": 0.021,
"step": 3460
},
{
"grad_norm": 0.11704110354185104,
"learning_rate": 7.777507453128163e-05,
"loss": 0.0197,
"step": 3470
},
{
"grad_norm": 0.09754088521003723,
"learning_rate": 7.763743432509451e-05,
"loss": 0.0177,
"step": 3480
},
{
"grad_norm": 0.09270471334457397,
"learning_rate": 7.749949188035353e-05,
"loss": 0.023,
"step": 3490
},
{
"grad_norm": 0.12514866888523102,
"learning_rate": 7.736124870557516e-05,
"loss": 0.0192,
"step": 3500
},
{
"grad_norm": 0.17735812067985535,
"learning_rate": 7.722270631256459e-05,
"loss": 0.0246,
"step": 3510
},
{
"grad_norm": 0.13739928603172302,
"learning_rate": 7.708386621639925e-05,
"loss": 0.0267,
"step": 3520
},
{
"grad_norm": 0.09319963306188583,
"learning_rate": 7.694472993541219e-05,
"loss": 0.0164,
"step": 3530
},
{
"grad_norm": 0.15453414618968964,
"learning_rate": 7.680529899117547e-05,
"loss": 0.0258,
"step": 3540
},
{
"grad_norm": 0.09976519644260406,
"learning_rate": 7.666557490848358e-05,
"loss": 0.0189,
"step": 3550
},
{
"grad_norm": 0.1448899358510971,
"learning_rate": 7.65255592153367e-05,
"loss": 0.0217,
"step": 3560
},
{
"grad_norm": 0.29534968733787537,
"learning_rate": 7.638525344292402e-05,
"loss": 0.0269,
"step": 3570
},
{
"grad_norm": 0.14484824240207672,
"learning_rate": 7.624465912560697e-05,
"loss": 0.0257,
"step": 3580
},
{
"grad_norm": 0.1478380411863327,
"learning_rate": 7.610377780090249e-05,
"loss": 0.0189,
"step": 3590
},
{
"grad_norm": 0.13017632067203522,
"learning_rate": 7.596261100946618e-05,
"loss": 0.0196,
"step": 3600
},
{
"grad_norm": 0.14489924907684326,
"learning_rate": 7.582116029507542e-05,
"loss": 0.0206,
"step": 3610
},
{
"grad_norm": 0.1458485722541809,
"learning_rate": 7.56794272046126e-05,
"loss": 0.0194,
"step": 3620
},
{
"grad_norm": 0.14738769829273224,
"learning_rate": 7.55374132880481e-05,
"loss": 0.0203,
"step": 3630
},
{
"grad_norm": 0.13488487899303436,
"learning_rate": 7.539512009842333e-05,
"loss": 0.0167,
"step": 3640
},
{
"grad_norm": 0.19352062046527863,
"learning_rate": 7.525254919183382e-05,
"loss": 0.0202,
"step": 3650
},
{
"grad_norm": 0.10093115270137787,
"learning_rate": 7.510970212741215e-05,
"loss": 0.0179,
"step": 3660
},
{
"grad_norm": 0.11678976565599442,
"learning_rate": 7.496658046731096e-05,
"loss": 0.02,
"step": 3670
},
{
"grad_norm": 0.12365427613258362,
"learning_rate": 7.482318577668578e-05,
"loss": 0.0169,
"step": 3680
},
{
"grad_norm": 0.12157289683818817,
"learning_rate": 7.467951962367796e-05,
"loss": 0.0184,
"step": 3690
},
{
"grad_norm": 0.14409120380878448,
"learning_rate": 7.453558357939755e-05,
"loss": 0.018,
"step": 3700
},
{
"grad_norm": 0.12636560201644897,
"learning_rate": 7.439137921790606e-05,
"loss": 0.0228,
"step": 3710
},
{
"grad_norm": 0.13573269546031952,
"learning_rate": 7.42469081161993e-05,
"loss": 0.0207,
"step": 3720
},
{
"grad_norm": 0.062106985598802567,
"learning_rate": 7.410217185419006e-05,
"loss": 0.0184,
"step": 3730
},
{
"grad_norm": 0.16812735795974731,
"learning_rate": 7.395717201469095e-05,
"loss": 0.0238,
"step": 3740
},
{
"grad_norm": 0.12150177359580994,
"learning_rate": 7.381191018339696e-05,
"loss": 0.0217,
"step": 3750
},
{
"grad_norm": 0.2145930528640747,
"learning_rate": 7.36663879488682e-05,
"loss": 0.0163,
"step": 3760
},
{
"grad_norm": 0.08534784615039825,
"learning_rate": 7.352060690251254e-05,
"loss": 0.0174,
"step": 3770
},
{
"grad_norm": 0.0786997601389885,
"learning_rate": 7.337456863856811e-05,
"loss": 0.0173,
"step": 3780
},
{
"grad_norm": 0.14394225180149078,
"learning_rate": 7.3228274754086e-05,
"loss": 0.023,
"step": 3790
},
{
"grad_norm": 0.1746888905763626,
"learning_rate": 7.308172684891267e-05,
"loss": 0.0278,
"step": 3800
},
{
"grad_norm": 0.11417494714260101,
"learning_rate": 7.293492652567255e-05,
"loss": 0.024,
"step": 3810
},
{
"grad_norm": 0.11693543940782547,
"learning_rate": 7.278787538975043e-05,
"loss": 0.0223,
"step": 3820
},
{
"grad_norm": 0.12056666612625122,
"learning_rate": 7.2640575049274e-05,
"loss": 0.0163,
"step": 3830
},
{
"grad_norm": 0.11734664440155029,
"learning_rate": 7.249302711509616e-05,
"loss": 0.0215,
"step": 3840
},
{
"grad_norm": 0.10295484960079193,
"learning_rate": 7.23452332007775e-05,
"loss": 0.0171,
"step": 3850
},
{
"grad_norm": 0.11296578496694565,
"learning_rate": 7.219719492256858e-05,
"loss": 0.0244,
"step": 3860
},
{
"grad_norm": 0.12529076635837555,
"learning_rate": 7.20489138993923e-05,
"loss": 0.023,
"step": 3870
},
{
"grad_norm": 0.11730336397886276,
"learning_rate": 7.190039175282614e-05,
"loss": 0.0193,
"step": 3880
},
{
"grad_norm": 0.12242171168327332,
"learning_rate": 7.175163010708455e-05,
"loss": 0.0223,
"step": 3890
},
{
"grad_norm": 0.11755640804767609,
"learning_rate": 7.1602630589001e-05,
"loss": 0.0198,
"step": 3900
},
{
"grad_norm": 0.11875393241643906,
"learning_rate": 7.14533948280104e-05,
"loss": 0.0163,
"step": 3910
},
{
"grad_norm": 0.12509284913539886,
"learning_rate": 7.130392445613109e-05,
"loss": 0.0183,
"step": 3920
},
{
"grad_norm": 0.14473293721675873,
"learning_rate": 7.115422110794711e-05,
"loss": 0.0205,
"step": 3930
},
{
"grad_norm": 0.1699618548154831,
"learning_rate": 7.100428642059033e-05,
"loss": 0.0221,
"step": 3940
},
{
"grad_norm": 0.1720869392156601,
"learning_rate": 7.08541220337224e-05,
"loss": 0.022,
"step": 3950
},
{
"grad_norm": 0.11222980916500092,
"learning_rate": 7.070372958951706e-05,
"loss": 0.0291,
"step": 3960
},
{
"grad_norm": 0.13873277604579926,
"learning_rate": 7.055311073264194e-05,
"loss": 0.0203,
"step": 3970
},
{
"grad_norm": 0.11736611276865005,
"learning_rate": 7.040226711024077e-05,
"loss": 0.0177,
"step": 3980
},
{
"grad_norm": 0.09165987372398376,
"learning_rate": 7.02512003719152e-05,
"loss": 0.0171,
"step": 3990
},
{
"grad_norm": 0.1122891902923584,
"learning_rate": 7.00999121697069e-05,
"loss": 0.0168,
"step": 4000
},
{
"grad_norm": 0.14379985630512238,
"learning_rate": 6.99484041580794e-05,
"loss": 0.0155,
"step": 4010
},
{
"grad_norm": 0.1120666041970253,
"learning_rate": 6.979667799390004e-05,
"loss": 0.015,
"step": 4020
},
{
"grad_norm": 0.13411420583724976,
"learning_rate": 6.964473533642185e-05,
"loss": 0.0176,
"step": 4030
},
{
"grad_norm": 0.16511785984039307,
"learning_rate": 6.949257784726539e-05,
"loss": 0.0225,
"step": 4040
},
{
"grad_norm": 0.16178478300571442,
"learning_rate": 6.934020719040056e-05,
"loss": 0.0211,
"step": 4050
},
{
"grad_norm": 0.0941905677318573,
"learning_rate": 6.918762503212848e-05,
"loss": 0.021,
"step": 4060
},
{
"grad_norm": 0.1833467185497284,
"learning_rate": 6.903483304106319e-05,
"loss": 0.0279,
"step": 4070
},
{
"grad_norm": 0.12390021234750748,
"learning_rate": 6.888183288811341e-05,
"loss": 0.0195,
"step": 4080
},
{
"grad_norm": 0.1076701432466507,
"learning_rate": 6.87286262464643e-05,
"loss": 0.0198,
"step": 4090
},
{
"grad_norm": 0.12085499614477158,
"learning_rate": 6.857521479155915e-05,
"loss": 0.0183,
"step": 4100
},
{
"grad_norm": 0.11408424377441406,
"learning_rate": 6.842160020108104e-05,
"loss": 0.0162,
"step": 4110
},
{
"grad_norm": 0.10348336398601532,
"learning_rate": 6.826778415493455e-05,
"loss": 0.0168,
"step": 4120
},
{
"grad_norm": 0.14931976795196533,
"learning_rate": 6.811376833522729e-05,
"loss": 0.0149,
"step": 4130
},
{
"grad_norm": 0.12067539244890213,
"learning_rate": 6.795955442625159e-05,
"loss": 0.0183,
"step": 4140
},
{
"grad_norm": 0.09142495691776276,
"learning_rate": 6.780514411446608e-05,
"loss": 0.0165,
"step": 4150
},
{
"grad_norm": 0.11517301201820374,
"learning_rate": 6.765053908847716e-05,
"loss": 0.0145,
"step": 4160
},
{
"grad_norm": 0.09203200042247772,
"learning_rate": 6.749574103902064e-05,
"loss": 0.0153,
"step": 4170
},
{
"grad_norm": 0.12721046805381775,
"learning_rate": 6.734075165894317e-05,
"loss": 0.0181,
"step": 4180
},
{
"grad_norm": 0.0939367413520813,
"learning_rate": 6.71855726431838e-05,
"loss": 0.0178,
"step": 4190
},
{
"grad_norm": 0.14988242089748383,
"learning_rate": 6.703020568875538e-05,
"loss": 0.0142,
"step": 4200
},
{
"grad_norm": 0.10384884476661682,
"learning_rate": 6.687465249472603e-05,
"loss": 0.016,
"step": 4210
},
{
"grad_norm": 0.07902251929044724,
"learning_rate": 6.671891476220055e-05,
"loss": 0.0163,
"step": 4220
},
{
"grad_norm": 0.11719082295894623,
"learning_rate": 6.656299419430183e-05,
"loss": 0.0162,
"step": 4230
},
{
"grad_norm": 0.11947415769100189,
"learning_rate": 6.640689249615223e-05,
"loss": 0.0144,
"step": 4240
},
{
"grad_norm": 0.12006451189517975,
"learning_rate": 6.625061137485491e-05,
"loss": 0.0211,
"step": 4250
},
{
"grad_norm": 0.10241923481225967,
"learning_rate": 6.609415253947517e-05,
"loss": 0.0138,
"step": 4260
},
{
"grad_norm": 0.1031462550163269,
"learning_rate": 6.593751770102178e-05,
"loss": 0.0181,
"step": 4270
},
{
"grad_norm": 0.13527601957321167,
"learning_rate": 6.578070857242823e-05,
"loss": 0.0186,
"step": 4280
},
{
"grad_norm": 0.19854587316513062,
"learning_rate": 6.562372686853402e-05,
"loss": 0.0294,
"step": 4290
},
{
"grad_norm": 0.10074352473020554,
"learning_rate": 6.546657430606593e-05,
"loss": 0.0157,
"step": 4300
},
{
"grad_norm": 0.13401925563812256,
"learning_rate": 6.530925260361918e-05,
"loss": 0.0175,
"step": 4310
},
{
"grad_norm": 0.1723649501800537,
"learning_rate": 6.515176348163871e-05,
"loss": 0.0158,
"step": 4320
},
{
"grad_norm": 0.16466432809829712,
"learning_rate": 6.499410866240032e-05,
"loss": 0.0183,
"step": 4330
},
{
"grad_norm": 0.09429506212472916,
"learning_rate": 6.48362898699919e-05,
"loss": 0.0131,
"step": 4340
},
{
"grad_norm": 0.11094179004430771,
"learning_rate": 6.467830883029443e-05,
"loss": 0.0148,
"step": 4350
},
{
"grad_norm": 0.1540028154850006,
"learning_rate": 6.452016727096326e-05,
"loss": 0.0246,
"step": 4360
},
{
"grad_norm": 0.13154880702495575,
"learning_rate": 6.436186692140916e-05,
"loss": 0.0161,
"step": 4370
},
{
"grad_norm": 0.09090028703212738,
"learning_rate": 6.420340951277938e-05,
"loss": 0.018,
"step": 4380
},
{
"grad_norm": 0.08051954954862595,
"learning_rate": 6.404479677793874e-05,
"loss": 0.0214,
"step": 4390
},
{
"grad_norm": 0.06494958698749542,
"learning_rate": 6.388603045145075e-05,
"loss": 0.0128,
"step": 4400
},
{
"grad_norm": 0.10332348197698593,
"learning_rate": 6.372711226955843e-05,
"loss": 0.0155,
"step": 4410
},
{
"grad_norm": 0.157622292637825,
"learning_rate": 6.356804397016564e-05,
"loss": 0.0155,
"step": 4420
},
{
"grad_norm": 0.08613033592700958,
"learning_rate": 6.340882729281779e-05,
"loss": 0.0134,
"step": 4430
},
{
"grad_norm": 0.12358319014310837,
"learning_rate": 6.324946397868294e-05,
"loss": 0.0179,
"step": 4440
},
{
"grad_norm": 0.1570926010608673,
"learning_rate": 6.308995577053276e-05,
"loss": 0.0175,
"step": 4450
},
{
"grad_norm": 0.20264583826065063,
"learning_rate": 6.293030441272347e-05,
"loss": 0.0205,
"step": 4460
},
{
"grad_norm": 0.11582266539335251,
"learning_rate": 6.277051165117677e-05,
"loss": 0.017,
"step": 4470
},
{
"grad_norm": 0.1089734211564064,
"learning_rate": 6.261057923336064e-05,
"loss": 0.0137,
"step": 4480
},
{
"grad_norm": 0.10029280930757523,
"learning_rate": 6.245050890827042e-05,
"loss": 0.0149,
"step": 4490
},
{
"grad_norm": 0.10253600031137466,
"learning_rate": 6.229030242640952e-05,
"loss": 0.0245,
"step": 4500
},
{
"grad_norm": 0.14356465637683868,
"learning_rate": 6.212996153977037e-05,
"loss": 0.0149,
"step": 4510
},
{
"grad_norm": 0.12464722245931625,
"learning_rate": 6.196948800181523e-05,
"loss": 0.022,
"step": 4520
},
{
"grad_norm": 0.09775647521018982,
"learning_rate": 6.180888356745695e-05,
"loss": 0.0163,
"step": 4530
},
{
"grad_norm": 0.096392922103405,
"learning_rate": 6.164814999303995e-05,
"loss": 0.0136,
"step": 4540
},
{
"grad_norm": 0.10405683517456055,
"learning_rate": 6.148728903632081e-05,
"loss": 0.0223,
"step": 4550
},
{
"grad_norm": 0.11791567504405975,
"learning_rate": 6.132630245644921e-05,
"loss": 0.0168,
"step": 4560
},
{
"grad_norm": 0.08310893923044205,
"learning_rate": 6.116519201394857e-05,
"loss": 0.0164,
"step": 4570
},
{
"grad_norm": 0.08378935605287552,
"learning_rate": 6.10039594706969e-05,
"loss": 0.0173,
"step": 4580
},
{
"grad_norm": 0.0762941837310791,
"learning_rate": 6.084260658990744e-05,
"loss": 0.0144,
"step": 4590
},
{
"grad_norm": 0.08985130488872528,
"learning_rate": 6.068113513610943e-05,
"loss": 0.0145,
"step": 4600
},
{
"grad_norm": 0.11911336332559586,
"learning_rate": 6.0519546875128876e-05,
"loss": 0.0215,
"step": 4610
},
{
"grad_norm": 0.08101450651884079,
"learning_rate": 6.035784357406906e-05,
"loss": 0.0191,
"step": 4620
},
{
"grad_norm": 0.08633100986480713,
"learning_rate": 6.01960270012914e-05,
"loss": 0.0129,
"step": 4630
},
{
"grad_norm": 0.1489562839269638,
"learning_rate": 6.003409892639599e-05,
"loss": 0.0166,
"step": 4640
},
{
"grad_norm": 0.16843366622924805,
"learning_rate": 5.9872061120202336e-05,
"loss": 0.0227,
"step": 4650
},
{
"grad_norm": 0.1730397492647171,
"learning_rate": 5.9709915354729914e-05,
"loss": 0.0188,
"step": 4660
},
{
"grad_norm": 0.10702052712440491,
"learning_rate": 5.9547663403178824e-05,
"loss": 0.0175,
"step": 4670
},
{
"grad_norm": 0.13497740030288696,
"learning_rate": 5.9385307039910445e-05,
"loss": 0.0132,
"step": 4680
},
{
"grad_norm": 0.08305234462022781,
"learning_rate": 5.922284804042792e-05,
"loss": 0.02,
"step": 4690
},
{
"grad_norm": 0.13063625991344452,
"learning_rate": 5.906028818135687e-05,
"loss": 0.02,
"step": 4700
},
{
"grad_norm": 0.10262671113014221,
"learning_rate": 5.889762924042585e-05,
"loss": 0.0164,
"step": 4710
},
{
"grad_norm": 0.09967511147260666,
"learning_rate": 5.873487299644699e-05,
"loss": 0.0147,
"step": 4720
},
{
"grad_norm": 0.1044909656047821,
"learning_rate": 5.857202122929649e-05,
"loss": 0.0128,
"step": 4730
},
{
"grad_norm": 0.09511242061853409,
"learning_rate": 5.840907571989518e-05,
"loss": 0.0207,
"step": 4740
},
{
"grad_norm": 0.10167711228132248,
"learning_rate": 5.824603825018904e-05,
"loss": 0.0146,
"step": 4750
},
{
"grad_norm": 0.123440220952034,
"learning_rate": 5.808291060312975e-05,
"loss": 0.0172,
"step": 4760
},
{
"grad_norm": 0.1454525589942932,
"learning_rate": 5.7919694562655083e-05,
"loss": 0.0197,
"step": 4770
},
{
"grad_norm": 0.14693108201026917,
"learning_rate": 5.775639191366954e-05,
"loss": 0.0182,
"step": 4780
},
{
"grad_norm": 0.11754677444696426,
"learning_rate": 5.75930044420247e-05,
"loss": 0.0232,
"step": 4790
},
{
"grad_norm": 0.06257650256156921,
"learning_rate": 5.74295339344998e-05,
"loss": 0.0148,
"step": 4800
},
{
"grad_norm": 0.10418052226305008,
"learning_rate": 5.726598217878211e-05,
"loss": 0.0163,
"step": 4810
},
{
"grad_norm": 0.0884476900100708,
"learning_rate": 5.71023509634474e-05,
"loss": 0.015,
"step": 4820
},
{
"grad_norm": 0.1937217265367508,
"learning_rate": 5.693864207794049e-05,
"loss": 0.017,
"step": 4830
},
{
"grad_norm": 0.11750844120979309,
"learning_rate": 5.677485731255545e-05,
"loss": 0.0154,
"step": 4840
},
{
"grad_norm": 0.09092950820922852,
"learning_rate": 5.6610998458416296e-05,
"loss": 0.0144,
"step": 4850
},
{
"grad_norm": 0.10875654965639114,
"learning_rate": 5.644706730745716e-05,
"loss": 0.016,
"step": 4860
},
{
"grad_norm": 0.1630352884531021,
"learning_rate": 5.628306565240287e-05,
"loss": 0.0201,
"step": 4870
},
{
"grad_norm": 0.08496683835983276,
"learning_rate": 5.611899528674923e-05,
"loss": 0.0173,
"step": 4880
},
{
"grad_norm": 0.11503885686397552,
"learning_rate": 5.595485800474349e-05,
"loss": 0.0152,
"step": 4890
},
{
"grad_norm": 0.1367446631193161,
"learning_rate": 5.579065560136467e-05,
"loss": 0.0149,
"step": 4900
},
{
"grad_norm": 0.12649130821228027,
"learning_rate": 5.562638987230392e-05,
"loss": 0.0183,
"step": 4910
},
{
"grad_norm": 0.09708788245916367,
"learning_rate": 5.546206261394498e-05,
"loss": 0.0191,
"step": 4920
},
{
"grad_norm": 0.0999777540564537,
"learning_rate": 5.529767562334437e-05,
"loss": 0.0168,
"step": 4930
},
{
"grad_norm": 0.12223142385482788,
"learning_rate": 5.5133230698211926e-05,
"loss": 0.0155,
"step": 4940
},
{
"grad_norm": 0.09065050631761551,
"learning_rate": 5.496872963689096e-05,
"loss": 0.0132,
"step": 4950
},
{
"grad_norm": 0.146714985370636,
"learning_rate": 5.4804174238338756e-05,
"loss": 0.0209,
"step": 4960
},
{
"grad_norm": 0.11266348510980606,
"learning_rate": 5.463956630210678e-05,
"loss": 0.0139,
"step": 4970
},
{
"grad_norm": 0.07989142090082169,
"learning_rate": 5.4474907628321046e-05,
"loss": 0.011,
"step": 4980
},
{
"grad_norm": 0.14550931751728058,
"learning_rate": 5.431020001766244e-05,
"loss": 0.0177,
"step": 4990
},
{
"grad_norm": 0.1117406040430069,
"learning_rate": 5.4145445271346986e-05,
"loss": 0.0178,
"step": 5000
},
{
"grad_norm": 0.12149982154369354,
"learning_rate": 5.398064519110622e-05,
"loss": 0.0175,
"step": 5010
},
{
"grad_norm": 0.21780666708946228,
"learning_rate": 5.3815801579167394e-05,
"loss": 0.0167,
"step": 5020
},
{
"grad_norm": 0.17102351784706116,
"learning_rate": 5.365091623823382e-05,
"loss": 0.0175,
"step": 5030
},
{
"grad_norm": 0.11842578649520874,
"learning_rate": 5.348599097146521e-05,
"loss": 0.0128,
"step": 5040
},
{
"grad_norm": 0.1212945505976677,
"learning_rate": 5.3321027582457836e-05,
"loss": 0.0156,
"step": 5050
},
{
"grad_norm": 0.09221485257148743,
"learning_rate": 5.315602787522491e-05,
"loss": 0.012,
"step": 5060
},
{
"grad_norm": 0.094135582447052,
"learning_rate": 5.299099365417678e-05,
"loss": 0.0155,
"step": 5070
},
{
"grad_norm": 0.11543729156255722,
"learning_rate": 5.2825926724101236e-05,
"loss": 0.0182,
"step": 5080
},
{
"grad_norm": 0.17128892242908478,
"learning_rate": 5.26608288901438e-05,
"loss": 0.018,
"step": 5090
},
{
"grad_norm": 0.12547394633293152,
"learning_rate": 5.24957019577879e-05,
"loss": 0.0215,
"step": 5100
},
{
"grad_norm": 0.21015821397304535,
"learning_rate": 5.2330547732835266e-05,
"loss": 0.0166,
"step": 5110
},
{
"grad_norm": 0.34002071619033813,
"learning_rate": 5.2165368021385996e-05,
"loss": 0.0179,
"step": 5120
},
{
"grad_norm": 0.14867143332958221,
"learning_rate": 5.200016462981897e-05,
"loss": 0.0145,
"step": 5130
},
{
"grad_norm": 0.09806505590677261,
"learning_rate": 5.1834939364772015e-05,
"loss": 0.0155,
"step": 5140
},
{
"grad_norm": 0.14923235774040222,
"learning_rate": 5.166969403312214e-05,
"loss": 0.018,
"step": 5150
},
{
"grad_norm": 0.1007002517580986,
"learning_rate": 5.1504430441965844e-05,
"loss": 0.015,
"step": 5160
},
{
"grad_norm": 0.19337940216064453,
"learning_rate": 5.133915039859923e-05,
"loss": 0.0164,
"step": 5170
},
{
"grad_norm": 0.0948144942522049,
"learning_rate": 5.1173855710498444e-05,
"loss": 0.0177,
"step": 5180
},
{
"grad_norm": 0.09900026768445969,
"learning_rate": 5.100854818529967e-05,
"loss": 0.0146,
"step": 5190
},
{
"grad_norm": 0.10444221645593643,
"learning_rate": 5.084322963077951e-05,
"loss": 0.0155,
"step": 5200
},
{
"grad_norm": 0.0999322310090065,
"learning_rate": 5.067790185483522e-05,
"loss": 0.0172,
"step": 5210
},
{
"grad_norm": 0.13305051624774933,
"learning_rate": 5.0512566665464844e-05,
"loss": 0.0128,
"step": 5220
},
{
"grad_norm": 0.21085679531097412,
"learning_rate": 5.034722587074755e-05,
"loss": 0.0192,
"step": 5230
},
{
"grad_norm": 0.10392367094755173,
"learning_rate": 5.018188127882375e-05,
"loss": 0.0139,
"step": 5240
},
{
"grad_norm": 0.119481660425663,
"learning_rate": 5.0016534697875417e-05,
"loss": 0.0125,
"step": 5250
},
{
"grad_norm": 0.11169359087944031,
"learning_rate": 4.9851187936106294e-05,
"loss": 0.0159,
"step": 5260
},
{
"grad_norm": 0.0979238748550415,
"learning_rate": 4.968584280172206e-05,
"loss": 0.0155,
"step": 5270
},
{
"grad_norm": 0.1328539103269577,
"learning_rate": 4.95205011029106e-05,
"loss": 0.0108,
"step": 5280
},
{
"grad_norm": 0.11174710094928741,
"learning_rate": 4.935516464782227e-05,
"loss": 0.0142,
"step": 5290
},
{
"grad_norm": 0.1004805713891983,
"learning_rate": 4.918983524455003e-05,
"loss": 0.0194,
"step": 5300
},
{
"grad_norm": 0.13593004643917084,
"learning_rate": 4.9024514701109766e-05,
"loss": 0.028,
"step": 5310
},
{
"grad_norm": 0.12582026422023773,
"learning_rate": 4.885920482542043e-05,
"loss": 0.0166,
"step": 5320
},
{
"grad_norm": 0.11316690593957901,
"learning_rate": 4.869390742528438e-05,
"loss": 0.0195,
"step": 5330
},
{
"grad_norm": 0.07854901999235153,
"learning_rate": 4.852862430836744e-05,
"loss": 0.0137,
"step": 5340
},
{
"grad_norm": 0.1230713501572609,
"learning_rate": 4.836335728217933e-05,
"loss": 0.0207,
"step": 5350
},
{
"grad_norm": 0.11357402056455612,
"learning_rate": 4.819810815405379e-05,
"loss": 0.0168,
"step": 5360
},
{
"grad_norm": 0.1529230773448944,
"learning_rate": 4.803287873112877e-05,
"loss": 0.0116,
"step": 5370
},
{
"grad_norm": 0.08224724233150482,
"learning_rate": 4.786767082032681e-05,
"loss": 0.0189,
"step": 5380
},
{
"grad_norm": 0.17682884633541107,
"learning_rate": 4.77024862283351e-05,
"loss": 0.0183,
"step": 5390
},
{
"grad_norm": 0.09716563671827316,
"learning_rate": 4.753732676158593e-05,
"loss": 0.0174,
"step": 5400
},
{
"grad_norm": 0.08129530400037766,
"learning_rate": 4.737219422623672e-05,
"loss": 0.0139,
"step": 5410
},
{
"grad_norm": 0.12539704144001007,
"learning_rate": 4.720709042815044e-05,
"loss": 0.0182,
"step": 5420
},
{
"grad_norm": 0.09867323935031891,
"learning_rate": 4.704201717287578e-05,
"loss": 0.0122,
"step": 5430
},
{
"grad_norm": 0.10833761841058731,
"learning_rate": 4.6876976265627404e-05,
"loss": 0.0164,
"step": 5440
},
{
"grad_norm": 0.11536791175603867,
"learning_rate": 4.671196951126626e-05,
"loss": 0.0153,
"step": 5450
},
{
"grad_norm": 0.09528806805610657,
"learning_rate": 4.654699871427971e-05,
"loss": 0.0132,
"step": 5460
},
{
"grad_norm": 0.15388181805610657,
"learning_rate": 4.6382065678762034e-05,
"loss": 0.0198,
"step": 5470
},
{
"grad_norm": 0.08998192846775055,
"learning_rate": 4.6217172208394424e-05,
"loss": 0.0109,
"step": 5480
},
{
"grad_norm": 0.07558619976043701,
"learning_rate": 4.605232010642549e-05,
"loss": 0.0119,
"step": 5490
},
{
"grad_norm": 0.10081010311841965,
"learning_rate": 4.588751117565142e-05,
"loss": 0.0131,
"step": 5500
},
{
"grad_norm": 0.12107884138822556,
"learning_rate": 4.5722747218396214e-05,
"loss": 0.0146,
"step": 5510
},
{
"grad_norm": 0.11925487220287323,
"learning_rate": 4.5558030036492194e-05,
"loss": 0.0135,
"step": 5520
},
{
"grad_norm": 0.12009993940591812,
"learning_rate": 4.539336143125999e-05,
"loss": 0.0142,
"step": 5530
},
{
"grad_norm": 0.06643348187208176,
"learning_rate": 4.522874320348916e-05,
"loss": 0.0134,
"step": 5540
},
{
"grad_norm": 0.1483471393585205,
"learning_rate": 4.506417715341821e-05,
"loss": 0.0134,
"step": 5550
},
{
"grad_norm": 0.14901196956634521,
"learning_rate": 4.489966508071511e-05,
"loss": 0.0149,
"step": 5560
},
{
"grad_norm": 0.1339421421289444,
"learning_rate": 4.4735208784457575e-05,
"loss": 0.0154,
"step": 5570
},
{
"grad_norm": 0.09905123710632324,
"learning_rate": 4.457081006311325e-05,
"loss": 0.0111,
"step": 5580
},
{
"grad_norm": 0.08451845496892929,
"learning_rate": 4.440647071452027e-05,
"loss": 0.0125,
"step": 5590
},
{
"grad_norm": 0.10544228553771973,
"learning_rate": 4.424219253586737e-05,
"loss": 0.0168,
"step": 5600
},
{
"grad_norm": 0.10120836645364761,
"learning_rate": 4.407797732367443e-05,
"loss": 0.0145,
"step": 5610
},
{
"grad_norm": 0.10984590649604797,
"learning_rate": 4.391382687377268e-05,
"loss": 0.0164,
"step": 5620
},
{
"grad_norm": 0.06651951372623444,
"learning_rate": 4.374974298128512e-05,
"loss": 0.0107,
"step": 5630
},
{
"grad_norm": 0.14574792981147766,
"learning_rate": 4.358572744060699e-05,
"loss": 0.0166,
"step": 5640
},
{
"grad_norm": 0.09139572083950043,
"learning_rate": 4.342178204538588e-05,
"loss": 0.014,
"step": 5650
},
{
"grad_norm": 0.17951051890850067,
"learning_rate": 4.325790858850241e-05,
"loss": 0.0136,
"step": 5660
},
{
"grad_norm": 0.10618460178375244,
"learning_rate": 4.309410886205043e-05,
"loss": 0.0109,
"step": 5670
},
{
"grad_norm": 0.10198267549276352,
"learning_rate": 4.293038465731752e-05,
"loss": 0.0187,
"step": 5680
},
{
"grad_norm": 0.12924885749816895,
"learning_rate": 4.276673776476533e-05,
"loss": 0.0107,
"step": 5690
},
{
"grad_norm": 0.13997332751750946,
"learning_rate": 4.260316997401007e-05,
"loss": 0.0169,
"step": 5700
},
{
"grad_norm": 0.07969796657562256,
"learning_rate": 4.243968307380293e-05,
"loss": 0.0174,
"step": 5710
},
{
"grad_norm": 0.10636241734027863,
"learning_rate": 4.22762788520104e-05,
"loss": 0.011,
"step": 5720
},
{
"grad_norm": 0.08587957918643951,
"learning_rate": 4.211295909559491e-05,
"loss": 0.0146,
"step": 5730
},
{
"grad_norm": 0.12792466580867767,
"learning_rate": 4.194972559059511e-05,
"loss": 0.0104,
"step": 5740
},
{
"grad_norm": 0.07166793942451477,
"learning_rate": 4.178658012210651e-05,
"loss": 0.0139,
"step": 5750
},
{
"grad_norm": 0.06277091056108475,
"learning_rate": 4.162352447426177e-05,
"loss": 0.0108,
"step": 5760
},
{
"grad_norm": 0.10091009736061096,
"learning_rate": 4.146056043021135e-05,
"loss": 0.0106,
"step": 5770
},
{
"grad_norm": 0.0645943358540535,
"learning_rate": 4.1297689772103944e-05,
"loss": 0.0134,
"step": 5780
},
{
"grad_norm": 0.10704907774925232,
"learning_rate": 4.113491428106694e-05,
"loss": 0.0128,
"step": 5790
},
{
"grad_norm": 0.10993106663227081,
"learning_rate": 4.0972235737187055e-05,
"loss": 0.013,
"step": 5800
},
{
"grad_norm": 0.1654406189918518,
"learning_rate": 4.080965591949076e-05,
"loss": 0.0236,
"step": 5810
},
{
"grad_norm": 0.12182041257619858,
"learning_rate": 4.0647176605924924e-05,
"loss": 0.0211,
"step": 5820
},
{
"grad_norm": 0.07427161931991577,
"learning_rate": 4.0484799573337255e-05,
"loss": 0.0154,
"step": 5830
},
{
"grad_norm": 0.0983806699514389,
"learning_rate": 4.032252659745699e-05,
"loss": 0.0128,
"step": 5840
},
{
"grad_norm": 0.10781994462013245,
"learning_rate": 4.016035945287539e-05,
"loss": 0.0129,
"step": 5850
},
{
"grad_norm": 0.10114669799804688,
"learning_rate": 3.999829991302635e-05,
"loss": 0.0145,
"step": 5860
},
{
"grad_norm": 0.1134527251124382,
"learning_rate": 3.983634975016707e-05,
"loss": 0.0125,
"step": 5870
},
{
"grad_norm": 0.08050957322120667,
"learning_rate": 3.967451073535854e-05,
"loss": 0.0166,
"step": 5880
},
{
"grad_norm": 0.11550536006689072,
"learning_rate": 3.951278463844633e-05,
"loss": 0.0126,
"step": 5890
},
{
"grad_norm": 0.14811114966869354,
"learning_rate": 3.935117322804111e-05,
"loss": 0.0147,
"step": 5900
},
{
"grad_norm": 0.09880905598402023,
"learning_rate": 3.918967827149938e-05,
"loss": 0.0161,
"step": 5910
},
{
"grad_norm": 0.15333198010921478,
"learning_rate": 3.9028301534904094e-05,
"loss": 0.0173,
"step": 5920
},
{
"grad_norm": 0.12640492618083954,
"learning_rate": 3.88670447830454e-05,
"loss": 0.0147,
"step": 5930
},
{
"grad_norm": 0.163604736328125,
"learning_rate": 3.870590977940132e-05,
"loss": 0.0171,
"step": 5940
},
{
"grad_norm": 0.11071155965328217,
"learning_rate": 3.8544898286118404e-05,
"loss": 0.0121,
"step": 5950
},
{
"grad_norm": 0.07924096286296844,
"learning_rate": 3.838401206399257e-05,
"loss": 0.0127,
"step": 5960
},
{
"grad_norm": 0.07424289733171463,
"learning_rate": 3.822325287244975e-05,
"loss": 0.0125,
"step": 5970
},
{
"grad_norm": 0.08219657093286514,
"learning_rate": 3.8062622469526725e-05,
"loss": 0.0094,
"step": 5980
},
{
"grad_norm": 0.1099315956234932,
"learning_rate": 3.790212261185183e-05,
"loss": 0.0144,
"step": 5990
},
{
"grad_norm": 0.09500293433666229,
"learning_rate": 3.7741755054625794e-05,
"loss": 0.0167,
"step": 6000
},
{
"grad_norm": 0.1208152174949646,
"learning_rate": 3.758152155160255e-05,
"loss": 0.0141,
"step": 6010
},
{
"grad_norm": 0.14845283329486847,
"learning_rate": 3.742142385506999e-05,
"loss": 0.0132,
"step": 6020
},
{
"grad_norm": 0.062457937747240067,
"learning_rate": 3.72614637158309e-05,
"loss": 0.0171,
"step": 6030
},
{
"grad_norm": 0.16369330883026123,
"learning_rate": 3.710164288318371e-05,
"loss": 0.0196,
"step": 6040
},
{
"grad_norm": 0.1389397531747818,
"learning_rate": 3.694196310490345e-05,
"loss": 0.0146,
"step": 6050
},
{
"grad_norm": 0.16105593740940094,
"learning_rate": 3.678242612722259e-05,
"loss": 0.0126,
"step": 6060
},
{
"grad_norm": 0.07928600907325745,
"learning_rate": 3.6623033694811953e-05,
"loss": 0.0111,
"step": 6070
},
{
"grad_norm": 0.0712757557630539,
"learning_rate": 3.6463787550761665e-05,
"loss": 0.0116,
"step": 6080
},
{
"grad_norm": 0.06778154522180557,
"learning_rate": 3.630468943656202e-05,
"loss": 0.0144,
"step": 6090
},
{
"grad_norm": 0.06639878451824188,
"learning_rate": 3.6145741092084523e-05,
"loss": 0.0144,
"step": 6100
},
{
"grad_norm": 0.12484387308359146,
"learning_rate": 3.598694425556278e-05,
"loss": 0.017,
"step": 6110
},
{
"grad_norm": 0.0718843936920166,
"learning_rate": 3.58283006635736e-05,
"loss": 0.0119,
"step": 6120
},
{
"grad_norm": 0.1509476900100708,
"learning_rate": 3.566981205101781e-05,
"loss": 0.0196,
"step": 6130
},
{
"grad_norm": 0.09593743085861206,
"learning_rate": 3.5511480151101556e-05,
"loss": 0.0126,
"step": 6140
},
{
"grad_norm": 0.11232764273881912,
"learning_rate": 3.5353306695317104e-05,
"loss": 0.0118,
"step": 6150
},
{
"grad_norm": 0.11152388900518417,
"learning_rate": 3.519529341342402e-05,
"loss": 0.0153,
"step": 6160
},
{
"grad_norm": 0.10718723386526108,
"learning_rate": 3.503744203343026e-05,
"loss": 0.0137,
"step": 6170
},
{
"grad_norm": 0.09171156585216522,
"learning_rate": 3.487975428157318e-05,
"loss": 0.0138,
"step": 6180
},
{
"grad_norm": 0.07840190827846527,
"learning_rate": 3.472223188230083e-05,
"loss": 0.0108,
"step": 6190
},
{
"grad_norm": 0.10643889755010605,
"learning_rate": 3.4564876558252866e-05,
"loss": 0.0157,
"step": 6200
},
{
"grad_norm": 0.10364333540201187,
"learning_rate": 3.440769003024195e-05,
"loss": 0.013,
"step": 6210
},
{
"grad_norm": 0.1575847715139389,
"learning_rate": 3.425067401723477e-05,
"loss": 0.0183,
"step": 6220
},
{
"grad_norm": 0.16690440475940704,
"learning_rate": 3.409383023633325e-05,
"loss": 0.0142,
"step": 6230
},
{
"grad_norm": 0.0983443409204483,
"learning_rate": 3.3937160402755894e-05,
"loss": 0.0111,
"step": 6240
},
{
"grad_norm": 0.09937821328639984,
"learning_rate": 3.378066622981885e-05,
"loss": 0.0122,
"step": 6250
},
{
"grad_norm": 0.1097458079457283,
"learning_rate": 3.362434942891738e-05,
"loss": 0.0154,
"step": 6260
},
{
"grad_norm": 0.10768865793943405,
"learning_rate": 3.346821170950693e-05,
"loss": 0.0154,
"step": 6270
},
{
"grad_norm": 0.16550977528095245,
"learning_rate": 3.3312254779084585e-05,
"loss": 0.0121,
"step": 6280
},
{
"grad_norm": 0.09241650253534317,
"learning_rate": 3.315648034317039e-05,
"loss": 0.0115,
"step": 6290
},
{
"grad_norm": 0.1035611629486084,
"learning_rate": 3.3000890105288564e-05,
"loss": 0.0141,
"step": 6300
},
{
"grad_norm": 0.06898240745067596,
"learning_rate": 3.284548576694908e-05,
"loss": 0.0091,
"step": 6310
},
{
"grad_norm": 0.14829108119010925,
"learning_rate": 3.2690269027628815e-05,
"loss": 0.0146,
"step": 6320
},
{
"grad_norm": 0.1735328733921051,
"learning_rate": 3.253524158475324e-05,
"loss": 0.0158,
"step": 6330
},
{
"grad_norm": 0.07043437659740448,
"learning_rate": 3.238040513367757e-05,
"loss": 0.0127,
"step": 6340
},
{
"grad_norm": 0.10460063815116882,
"learning_rate": 3.222576136766843e-05,
"loss": 0.0131,
"step": 6350
},
{
"grad_norm": 0.11066054552793503,
"learning_rate": 3.2071311977885324e-05,
"loss": 0.0119,
"step": 6360
},
{
"grad_norm": 0.09701026231050491,
"learning_rate": 3.191705865336197e-05,
"loss": 0.0122,
"step": 6370
},
{
"grad_norm": 0.15069259703159332,
"learning_rate": 3.1763003080988075e-05,
"loss": 0.0149,
"step": 6380
},
{
"grad_norm": 0.11791059374809265,
"learning_rate": 3.160914694549063e-05,
"loss": 0.0163,
"step": 6390
},
{
"grad_norm": 0.08860627561807632,
"learning_rate": 3.145549192941573e-05,
"loss": 0.0139,
"step": 6400
},
{
"grad_norm": 0.07669701427221298,
"learning_rate": 3.130203971310999e-05,
"loss": 0.014,
"step": 6410
},
{
"grad_norm": 0.07614731043577194,
"learning_rate": 3.114879197470225e-05,
"loss": 0.0122,
"step": 6420
},
{
"grad_norm": 0.08702623844146729,
"learning_rate": 3.0995750390085285e-05,
"loss": 0.013,
"step": 6430
},
{
"grad_norm": 0.0708293616771698,
"learning_rate": 3.084291663289728e-05,
"loss": 0.0097,
"step": 6440
},
{
"grad_norm": 0.1287541687488556,
"learning_rate": 3.069029237450375e-05,
"loss": 0.0162,
"step": 6450
},
{
"grad_norm": 0.08659978210926056,
"learning_rate": 3.053787928397911e-05,
"loss": 0.0098,
"step": 6460
},
{
"grad_norm": 0.08566563576459885,
"learning_rate": 3.0385679028088526e-05,
"loss": 0.0116,
"step": 6470
},
{
"grad_norm": 0.08606121689081192,
"learning_rate": 3.023369327126959e-05,
"loss": 0.0082,
"step": 6480
},
{
"grad_norm": 0.06804236024618149,
"learning_rate": 3.0081923675614198e-05,
"loss": 0.0087,
"step": 6490
},
{
"grad_norm": 0.10266627371311188,
"learning_rate": 2.993037190085034e-05,
"loss": 0.0111,
"step": 6500
},
{
"grad_norm": 0.14793899655342102,
"learning_rate": 2.977903960432392e-05,
"loss": 0.0216,
"step": 6510
},
{
"grad_norm": 0.11101231724023819,
"learning_rate": 2.9627928440980722e-05,
"loss": 0.015,
"step": 6520
},
{
"grad_norm": 0.11928931623697281,
"learning_rate": 2.9477040063348183e-05,
"loss": 0.0098,
"step": 6530
},
{
"grad_norm": 0.13965009152889252,
"learning_rate": 2.9326376121517456e-05,
"loss": 0.0152,
"step": 6540
},
{
"grad_norm": 0.08400373160839081,
"learning_rate": 2.9175938263125236e-05,
"loss": 0.0126,
"step": 6550
},
{
"grad_norm": 0.053806230425834656,
"learning_rate": 2.9025728133335873e-05,
"loss": 0.0084,
"step": 6560
},
{
"grad_norm": 0.07494649291038513,
"learning_rate": 2.8875747374823288e-05,
"loss": 0.0149,
"step": 6570
},
{
"grad_norm": 0.10612501204013824,
"learning_rate": 2.872599762775298e-05,
"loss": 0.0099,
"step": 6580
},
{
"grad_norm": 0.05629488080739975,
"learning_rate": 2.857648052976425e-05,
"loss": 0.0113,
"step": 6590
},
{
"grad_norm": 0.102164126932621,
"learning_rate": 2.8427197715952047e-05,
"loss": 0.0109,
"step": 6600
},
{
"grad_norm": 0.16999095678329468,
"learning_rate": 2.8278150818849393e-05,
"loss": 0.0079,
"step": 6610
},
{
"grad_norm": 0.08116182684898376,
"learning_rate": 2.812934146840922e-05,
"loss": 0.0136,
"step": 6620
},
{
"grad_norm": 0.08186657726764679,
"learning_rate": 2.7980771291986764e-05,
"loss": 0.0105,
"step": 6630
},
{
"grad_norm": 0.06212032958865166,
"learning_rate": 2.783244191432167e-05,
"loss": 0.0095,
"step": 6640
},
{
"grad_norm": 0.07709891349077225,
"learning_rate": 2.768435495752022e-05,
"loss": 0.0105,
"step": 6650
},
{
"grad_norm": 0.13410453498363495,
"learning_rate": 2.753651204103771e-05,
"loss": 0.016,
"step": 6660
},
{
"grad_norm": 0.10239517688751221,
"learning_rate": 2.7388914781660523e-05,
"loss": 0.0111,
"step": 6670
},
{
"grad_norm": 0.0703016147017479,
"learning_rate": 2.7241564793488693e-05,
"loss": 0.0105,
"step": 6680
},
{
"grad_norm": 0.1461678445339203,
"learning_rate": 2.7094463687918037e-05,
"loss": 0.0152,
"step": 6690
},
{
"grad_norm": 0.07751160860061646,
"learning_rate": 2.694761307362268e-05,
"loss": 0.015,
"step": 6700
},
{
"grad_norm": 0.08921916782855988,
"learning_rate": 2.6801014556537467e-05,
"loss": 0.0124,
"step": 6710
},
{
"grad_norm": 0.10372833907604218,
"learning_rate": 2.6654669739840243e-05,
"loss": 0.0106,
"step": 6720
},
{
"grad_norm": 0.07931507378816605,
"learning_rate": 2.650858022393451e-05,
"loss": 0.0157,
"step": 6730
},
{
"grad_norm": 0.1153360977768898,
"learning_rate": 2.6362747606431747e-05,
"loss": 0.0174,
"step": 6740
},
{
"grad_norm": 0.06669391691684723,
"learning_rate": 2.6217173482134172e-05,
"loss": 0.0108,
"step": 6750
},
{
"grad_norm": 0.07651269435882568,
"learning_rate": 2.6071859443017044e-05,
"loss": 0.0174,
"step": 6760
},
{
"grad_norm": 0.10819702595472336,
"learning_rate": 2.5926807078211414e-05,
"loss": 0.0153,
"step": 6770
},
{
"grad_norm": 0.07689729332923889,
"learning_rate": 2.5782017973986728e-05,
"loss": 0.0128,
"step": 6780
},
{
"grad_norm": 0.1622001677751541,
"learning_rate": 2.5637493713733374e-05,
"loss": 0.0115,
"step": 6790
},
{
"grad_norm": 0.11402573436498642,
"learning_rate": 2.549323587794559e-05,
"loss": 0.019,
"step": 6800
},
{
"grad_norm": 0.1884138435125351,
"learning_rate": 2.5349246044203895e-05,
"loss": 0.0138,
"step": 6810
},
{
"grad_norm": 0.08098918199539185,
"learning_rate": 2.520552578715808e-05,
"loss": 0.0104,
"step": 6820
},
{
"grad_norm": 0.10553104430437088,
"learning_rate": 2.506207667850981e-05,
"loss": 0.0141,
"step": 6830
},
{
"grad_norm": 0.06774193048477173,
"learning_rate": 2.4918900286995555e-05,
"loss": 0.0097,
"step": 6840
},
{
"grad_norm": 0.10253611207008362,
"learning_rate": 2.4775998178369458e-05,
"loss": 0.012,
"step": 6850
},
{
"grad_norm": 0.1957240253686905,
"learning_rate": 2.4633371915386017e-05,
"loss": 0.0109,
"step": 6860
},
{
"grad_norm": 0.08355480432510376,
"learning_rate": 2.4491023057783235e-05,
"loss": 0.0087,
"step": 6870
},
{
"grad_norm": 0.09997954219579697,
"learning_rate": 2.4348953162265375e-05,
"loss": 0.0166,
"step": 6880
},
{
"grad_norm": 0.08258730173110962,
"learning_rate": 2.420716378248607e-05,
"loss": 0.0095,
"step": 6890
},
{
"grad_norm": 0.1287717968225479,
"learning_rate": 2.4065656469031266e-05,
"loss": 0.0109,
"step": 6900
},
{
"grad_norm": 0.10503353178501129,
"learning_rate": 2.3924432769402268e-05,
"loss": 0.016,
"step": 6910
},
{
"grad_norm": 0.07276376336812973,
"learning_rate": 2.3783494227998844e-05,
"loss": 0.0136,
"step": 6920
},
{
"grad_norm": 0.0894055888056755,
"learning_rate": 2.3642842386102264e-05,
"loss": 0.0101,
"step": 6930
},
{
"grad_norm": 0.06587056815624237,
"learning_rate": 2.3502478781858567e-05,
"loss": 0.0112,
"step": 6940
},
{
"grad_norm": 0.07652408629655838,
"learning_rate": 2.3362404950261628e-05,
"loss": 0.0127,
"step": 6950
},
{
"grad_norm": 0.14036282896995544,
"learning_rate": 2.3222622423136458e-05,
"loss": 0.0141,
"step": 6960
},
{
"grad_norm": 0.07171545922756195,
"learning_rate": 2.3083132729122332e-05,
"loss": 0.012,
"step": 6970
},
{
"grad_norm": 0.12350364029407501,
"learning_rate": 2.294393739365621e-05,
"loss": 0.012,
"step": 6980
},
{
"grad_norm": 0.12356127798557281,
"learning_rate": 2.2805037938956e-05,
"loss": 0.0212,
"step": 6990
},
{
"grad_norm": 0.10291753709316254,
"learning_rate": 2.266643588400386e-05,
"loss": 0.0118,
"step": 7000
},
{
"grad_norm": 0.09324789047241211,
"learning_rate": 2.252813274452969e-05,
"loss": 0.0141,
"step": 7010
},
{
"grad_norm": 0.12701277434825897,
"learning_rate": 2.2390130032994427e-05,
"loss": 0.0108,
"step": 7020
},
{
"grad_norm": 0.06723277270793915,
"learning_rate": 2.2252429258573633e-05,
"loss": 0.0096,
"step": 7030
},
{
"grad_norm": 0.10593508929014206,
"learning_rate": 2.2115031927140904e-05,
"loss": 0.0122,
"step": 7040
},
{
"grad_norm": 0.10403140634298325,
"learning_rate": 2.1977939541251463e-05,
"loss": 0.0121,
"step": 7050
},
{
"grad_norm": 0.07704979181289673,
"learning_rate": 2.1841153600125684e-05,
"loss": 0.0107,
"step": 7060
},
{
"grad_norm": 0.09538368880748749,
"learning_rate": 2.170467559963267e-05,
"loss": 0.0147,
"step": 7070
},
{
"grad_norm": 0.13659921288490295,
"learning_rate": 2.1568507032273982e-05,
"loss": 0.0163,
"step": 7080
},
{
"grad_norm": 0.09706605225801468,
"learning_rate": 2.1432649387167264e-05,
"loss": 0.0114,
"step": 7090
},
{
"grad_norm": 0.11672662198543549,
"learning_rate": 2.1297104150029973e-05,
"loss": 0.0135,
"step": 7100
},
{
"grad_norm": 0.0744546577334404,
"learning_rate": 2.116187280316307e-05,
"loss": 0.0132,
"step": 7110
},
{
"grad_norm": 0.07638552784919739,
"learning_rate": 2.1026956825434908e-05,
"loss": 0.0105,
"step": 7120
},
{
"grad_norm": 0.08116145431995392,
"learning_rate": 2.0892357692265017e-05,
"loss": 0.0128,
"step": 7130
},
{
"grad_norm": 0.08376798033714294,
"learning_rate": 2.0758076875607947e-05,
"loss": 0.0132,
"step": 7140
},
{
"grad_norm": 0.06961897760629654,
"learning_rate": 2.0624115843937207e-05,
"loss": 0.0114,
"step": 7150
},
{
"grad_norm": 0.04945152997970581,
"learning_rate": 2.0490476062229157e-05,
"loss": 0.0165,
"step": 7160
},
{
"grad_norm": 0.08555549383163452,
"learning_rate": 2.035715899194704e-05,
"loss": 0.0114,
"step": 7170
},
{
"grad_norm": 0.12474948912858963,
"learning_rate": 2.022416609102499e-05,
"loss": 0.0102,
"step": 7180
},
{
"grad_norm": 0.05101489648222923,
"learning_rate": 2.009149881385205e-05,
"loss": 0.0105,
"step": 7190
},
{
"grad_norm": 0.16036275029182434,
"learning_rate": 1.995915861125634e-05,
"loss": 0.015,
"step": 7200
},
{
"grad_norm": 0.11775407940149307,
"learning_rate": 1.9827146930489065e-05,
"loss": 0.0102,
"step": 7210
},
{
"grad_norm": 0.09199395030736923,
"learning_rate": 1.9695465215208848e-05,
"loss": 0.0098,
"step": 7220
},
{
"grad_norm": 0.10746108740568161,
"learning_rate": 1.9564114905465813e-05,
"loss": 0.0109,
"step": 7230
},
{
"grad_norm": 0.07385744154453278,
"learning_rate": 1.9433097437685936e-05,
"loss": 0.0106,
"step": 7240
},
{
"grad_norm": 0.10727902501821518,
"learning_rate": 1.930241424465521e-05,
"loss": 0.0129,
"step": 7250
},
{
"grad_norm": 0.06855831295251846,
"learning_rate": 1.9172066755504115e-05,
"loss": 0.0094,
"step": 7260
},
{
"grad_norm": 0.09228988736867905,
"learning_rate": 1.9042056395691914e-05,
"loss": 0.0156,
"step": 7270
},
{
"grad_norm": 0.10431600362062454,
"learning_rate": 1.8912384586991066e-05,
"loss": 0.0094,
"step": 7280
},
{
"grad_norm": 0.11698517948389053,
"learning_rate": 1.8783052747471717e-05,
"loss": 0.0114,
"step": 7290
},
{
"grad_norm": 0.08065880835056305,
"learning_rate": 1.865406229148611e-05,
"loss": 0.015,
"step": 7300
},
{
"grad_norm": 0.0702187791466713,
"learning_rate": 1.8525414629653233e-05,
"loss": 0.0091,
"step": 7310
},
{
"grad_norm": 0.0950561910867691,
"learning_rate": 1.8397111168843255e-05,
"loss": 0.0098,
"step": 7320
},
{
"grad_norm": 0.0656697079539299,
"learning_rate": 1.8269153312162323e-05,
"loss": 0.0102,
"step": 7330
},
{
"grad_norm": 0.09684677422046661,
"learning_rate": 1.8141542458937054e-05,
"loss": 0.0127,
"step": 7340
},
{
"grad_norm": 0.1340114176273346,
"learning_rate": 1.8014280004699268e-05,
"loss": 0.0108,
"step": 7350
},
{
"grad_norm": 0.07838889211416245,
"learning_rate": 1.788736734117078e-05,
"loss": 0.0142,
"step": 7360
},
{
"grad_norm": 0.0720207467675209,
"learning_rate": 1.7760805856248152e-05,
"loss": 0.0132,
"step": 7370
},
{
"grad_norm": 0.06288359314203262,
"learning_rate": 1.7634596933987518e-05,
"loss": 0.0122,
"step": 7380
},
{
"grad_norm": 0.0680171549320221,
"learning_rate": 1.7508741954589404e-05,
"loss": 0.0106,
"step": 7390
},
{
"grad_norm": 0.15043655037879944,
"learning_rate": 1.7383242294383717e-05,
"loss": 0.0099,
"step": 7400
},
{
"grad_norm": 0.11125855892896652,
"learning_rate": 1.7258099325814632e-05,
"loss": 0.011,
"step": 7410
},
{
"grad_norm": 0.060009945183992386,
"learning_rate": 1.7133314417425594e-05,
"loss": 0.0178,
"step": 7420
},
{
"grad_norm": 0.10388922691345215,
"learning_rate": 1.7008888933844408e-05,
"loss": 0.0144,
"step": 7430
},
{
"grad_norm": 0.08130010217428207,
"learning_rate": 1.6884824235768172e-05,
"loss": 0.0109,
"step": 7440
},
{
"grad_norm": 0.10215283185243607,
"learning_rate": 1.6761121679948592e-05,
"loss": 0.0083,
"step": 7450
},
{
"grad_norm": 0.08346996456384659,
"learning_rate": 1.663778261917695e-05,
"loss": 0.0101,
"step": 7460
},
{
"grad_norm": 0.1447107046842575,
"learning_rate": 1.651480840226952e-05,
"loss": 0.012,
"step": 7470
},
{
"grad_norm": 0.2145552933216095,
"learning_rate": 1.639220037405258e-05,
"loss": 0.0143,
"step": 7480
},
{
"grad_norm": 0.06829545646905899,
"learning_rate": 1.6269959875347906e-05,
"loss": 0.0105,
"step": 7490
},
{
"grad_norm": 0.09343858808279037,
"learning_rate": 1.614808824295802e-05,
"loss": 0.0133,
"step": 7500
},
{
"grad_norm": 0.17119063436985016,
"learning_rate": 1.602658680965152e-05,
"loss": 0.0086,
"step": 7510
},
{
"grad_norm": 0.11219792068004608,
"learning_rate": 1.5905456904148686e-05,
"loss": 0.0136,
"step": 7520
},
{
"grad_norm": 0.1286395639181137,
"learning_rate": 1.57846998511067e-05,
"loss": 0.0174,
"step": 7530
},
{
"grad_norm": 0.0740140900015831,
"learning_rate": 1.566431697110538e-05,
"loss": 0.0116,
"step": 7540
},
{
"grad_norm": 0.10018734633922577,
"learning_rate": 1.554430958063259e-05,
"loss": 0.0128,
"step": 7550
},
{
"grad_norm": 0.10073462128639221,
"learning_rate": 1.5424678992069912e-05,
"loss": 0.0102,
"step": 7560
},
{
"grad_norm": 0.05530090630054474,
"learning_rate": 1.5305426513678362e-05,
"loss": 0.009,
"step": 7570
},
{
"grad_norm": 0.11774101853370667,
"learning_rate": 1.518655344958388e-05,
"loss": 0.0112,
"step": 7580
},
{
"grad_norm": 0.05505017563700676,
"learning_rate": 1.5068061099763275e-05,
"loss": 0.012,
"step": 7590
},
{
"grad_norm": 0.07832878082990646,
"learning_rate": 1.494995076002988e-05,
"loss": 0.0109,
"step": 7600
},
{
"grad_norm": 0.12978103756904602,
"learning_rate": 1.4832223722019456e-05,
"loss": 0.0117,
"step": 7610
},
{
"grad_norm": 0.06087205559015274,
"learning_rate": 1.4714881273176035e-05,
"loss": 0.0079,
"step": 7620
},
{
"grad_norm": 0.08805029094219208,
"learning_rate": 1.4597924696737835e-05,
"loss": 0.0172,
"step": 7630
},
{
"grad_norm": 0.19606037437915802,
"learning_rate": 1.4481355271723252e-05,
"loss": 0.0092,
"step": 7640
},
{
"grad_norm": 0.10848986357450485,
"learning_rate": 1.4365174272916809e-05,
"loss": 0.0125,
"step": 7650
},
{
"grad_norm": 0.12749142944812775,
"learning_rate": 1.4249382970855319e-05,
"loss": 0.0106,
"step": 7660
},
{
"grad_norm": 0.05358705669641495,
"learning_rate": 1.4133982631813903e-05,
"loss": 0.0091,
"step": 7670
},
{
"grad_norm": 0.08048982173204422,
"learning_rate": 1.4018974517792194e-05,
"loss": 0.0097,
"step": 7680
},
{
"grad_norm": 0.08677206188440323,
"learning_rate": 1.390435988650048e-05,
"loss": 0.0106,
"step": 7690
},
{
"grad_norm": 0.1321980506181717,
"learning_rate": 1.3790139991346006e-05,
"loss": 0.0117,
"step": 7700
},
{
"grad_norm": 0.06233120337128639,
"learning_rate": 1.367631608141926e-05,
"loss": 0.009,
"step": 7710
},
{
"grad_norm": 0.07186523824930191,
"learning_rate": 1.3562889401480278e-05,
"loss": 0.0151,
"step": 7720
},
{
"grad_norm": 0.08325783163309097,
"learning_rate": 1.3449861191945074e-05,
"loss": 0.0115,
"step": 7730
},
{
"grad_norm": 0.14051955938339233,
"learning_rate": 1.3337232688872009e-05,
"loss": 0.0098,
"step": 7740
},
{
"grad_norm": 0.13237300515174866,
"learning_rate": 1.3225005123948364e-05,
"loss": 0.0109,
"step": 7750
},
{
"grad_norm": 0.07133188098669052,
"learning_rate": 1.311317972447681e-05,
"loss": 0.0093,
"step": 7760
},
{
"grad_norm": 0.09805900603532791,
"learning_rate": 1.3001757713361996e-05,
"loss": 0.0081,
"step": 7770
},
{
"grad_norm": 0.09867582470178604,
"learning_rate": 1.2890740309097204e-05,
"loss": 0.0108,
"step": 7780
},
{
"grad_norm": 0.09767255932092667,
"learning_rate": 1.2780128725750944e-05,
"loss": 0.0132,
"step": 7790
},
{
"grad_norm": 0.09998299181461334,
"learning_rate": 1.266992417295379e-05,
"loss": 0.0092,
"step": 7800
},
{
"grad_norm": 0.09136645495891571,
"learning_rate": 1.2560127855885073e-05,
"loss": 0.0101,
"step": 7810
},
{
"grad_norm": 0.0668412446975708,
"learning_rate": 1.2450740975259745e-05,
"loss": 0.0102,
"step": 7820
},
{
"grad_norm": 0.15097054839134216,
"learning_rate": 1.234176472731517e-05,
"loss": 0.0125,
"step": 7830
},
{
"grad_norm": 0.07216800004243851,
"learning_rate": 1.2233200303798158e-05,
"loss": 0.008,
"step": 7840
},
{
"grad_norm": 0.1251133233308792,
"learning_rate": 1.2125048891951846e-05,
"loss": 0.0105,
"step": 7850
},
{
"grad_norm": 0.05678839236497879,
"learning_rate": 1.2017311674502745e-05,
"loss": 0.0098,
"step": 7860
},
{
"grad_norm": 0.07825085520744324,
"learning_rate": 1.1909989829647822e-05,
"loss": 0.0114,
"step": 7870
},
{
"grad_norm": 0.08573894202709198,
"learning_rate": 1.1803084531041553e-05,
"loss": 0.0115,
"step": 7880
},
{
"grad_norm": 0.18001435697078705,
"learning_rate": 1.1696596947783162e-05,
"loss": 0.0114,
"step": 7890
},
{
"grad_norm": 0.07100951671600342,
"learning_rate": 1.1590528244403803e-05,
"loss": 0.0141,
"step": 7900
},
{
"grad_norm": 0.06841184198856354,
"learning_rate": 1.148487958085382e-05,
"loss": 0.0095,
"step": 7910
},
{
"grad_norm": 0.043873380869627,
"learning_rate": 1.1379652112490086e-05,
"loss": 0.0089,
"step": 7920
},
{
"grad_norm": 0.09613429009914398,
"learning_rate": 1.1274846990063315e-05,
"loss": 0.011,
"step": 7930
},
{
"grad_norm": 0.11672299355268478,
"learning_rate": 1.117046535970554e-05,
"loss": 0.013,
"step": 7940
},
{
"grad_norm": 0.12638463079929352,
"learning_rate": 1.106650836291755e-05,
"loss": 0.0087,
"step": 7950
},
{
"grad_norm": 0.1008707657456398,
"learning_rate": 1.0962977136556418e-05,
"loss": 0.0105,
"step": 7960
},
{
"grad_norm": 0.07977188378572464,
"learning_rate": 1.0859872812823024e-05,
"loss": 0.0111,
"step": 7970
},
{
"grad_norm": 0.0645359456539154,
"learning_rate": 1.0757196519249747e-05,
"loss": 0.0079,
"step": 7980
},
{
"grad_norm": 0.09372685849666595,
"learning_rate": 1.0654949378688077e-05,
"loss": 0.0108,
"step": 7990
},
{
"grad_norm": 0.06485147029161453,
"learning_rate": 1.0553132509296376e-05,
"loss": 0.0124,
"step": 8000
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}