minddrive / trainer_state.json
zhangthu's picture
Upload deepsight model checkpoint
d65b496 verified
Raw
History Blame Contribute Delete
139 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 21.85909712722298,
"eval_steps": 500,
"global_step": 8000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.027359781121751026,
"grad_norm": 108.924072265625,
"learning_rate": 7.705479452054794e-07,
"loss": 10.3397,
"step": 10
},
{
"epoch": 0.05471956224350205,
"grad_norm": 46.91159439086914,
"learning_rate": 1.6267123287671233e-06,
"loss": 9.8008,
"step": 20
},
{
"epoch": 0.08207934336525308,
"grad_norm": 39.599334716796875,
"learning_rate": 2.482876712328767e-06,
"loss": 8.6126,
"step": 30
},
{
"epoch": 0.1094391244870041,
"grad_norm": 9.677482604980469,
"learning_rate": 3.3390410958904114e-06,
"loss": 7.6908,
"step": 40
},
{
"epoch": 0.13679890560875513,
"grad_norm": 7.38622522354126,
"learning_rate": 4.195205479452055e-06,
"loss": 7.1743,
"step": 50
},
{
"epoch": 0.16415868673050615,
"grad_norm": 1.656615972518921,
"learning_rate": 5.051369863013699e-06,
"loss": 6.9899,
"step": 60
},
{
"epoch": 0.19151846785225718,
"grad_norm": 6.234030246734619,
"learning_rate": 5.907534246575343e-06,
"loss": 6.8517,
"step": 70
},
{
"epoch": 0.2188782489740082,
"grad_norm": 1.446900725364685,
"learning_rate": 6.763698630136987e-06,
"loss": 6.6603,
"step": 80
},
{
"epoch": 0.24623803009575923,
"grad_norm": 1.646415114402771,
"learning_rate": 7.61986301369863e-06,
"loss": 6.4274,
"step": 90
},
{
"epoch": 0.27359781121751026,
"grad_norm": 1.9842872619628906,
"learning_rate": 8.476027397260275e-06,
"loss": 6.1851,
"step": 100
},
{
"epoch": 0.3009575923392613,
"grad_norm": 2.7530670166015625,
"learning_rate": 9.332191780821919e-06,
"loss": 5.9863,
"step": 110
},
{
"epoch": 0.3283173734610123,
"grad_norm": 1.670802116394043,
"learning_rate": 1.0188356164383562e-05,
"loss": 5.786,
"step": 120
},
{
"epoch": 0.35567715458276333,
"grad_norm": 1.2471222877502441,
"learning_rate": 1.1044520547945206e-05,
"loss": 5.6355,
"step": 130
},
{
"epoch": 0.38303693570451436,
"grad_norm": 1.112801432609558,
"learning_rate": 1.1900684931506851e-05,
"loss": 5.5226,
"step": 140
},
{
"epoch": 0.4103967168262654,
"grad_norm": 1.029639720916748,
"learning_rate": 1.2756849315068494e-05,
"loss": 5.4719,
"step": 150
},
{
"epoch": 0.4377564979480164,
"grad_norm": 1.3304383754730225,
"learning_rate": 1.3613013698630136e-05,
"loss": 5.4087,
"step": 160
},
{
"epoch": 0.46511627906976744,
"grad_norm": 1.0809189081192017,
"learning_rate": 1.4469178082191781e-05,
"loss": 5.3731,
"step": 170
},
{
"epoch": 0.49247606019151846,
"grad_norm": 1.050671935081482,
"learning_rate": 1.5325342465753425e-05,
"loss": 5.3977,
"step": 180
},
{
"epoch": 0.5198358413132695,
"grad_norm": 0.914842963218689,
"learning_rate": 1.618150684931507e-05,
"loss": 5.3667,
"step": 190
},
{
"epoch": 0.5471956224350205,
"grad_norm": 1.194571614265442,
"learning_rate": 1.703767123287671e-05,
"loss": 5.3606,
"step": 200
},
{
"epoch": 0.5745554035567716,
"grad_norm": 0.9293230772018433,
"learning_rate": 1.7893835616438355e-05,
"loss": 5.3636,
"step": 210
},
{
"epoch": 0.6019151846785226,
"grad_norm": 0.8678876161575317,
"learning_rate": 1.8750000000000002e-05,
"loss": 5.3489,
"step": 220
},
{
"epoch": 0.6292749658002736,
"grad_norm": 0.8161562085151672,
"learning_rate": 1.9606164383561645e-05,
"loss": 5.3334,
"step": 230
},
{
"epoch": 0.6566347469220246,
"grad_norm": 0.923994779586792,
"learning_rate": 2.046232876712329e-05,
"loss": 5.3508,
"step": 240
},
{
"epoch": 0.6839945280437757,
"grad_norm": 0.8996877074241638,
"learning_rate": 2.1318493150684932e-05,
"loss": 5.337,
"step": 250
},
{
"epoch": 0.7113543091655267,
"grad_norm": 0.7687991857528687,
"learning_rate": 2.2174657534246575e-05,
"loss": 5.3483,
"step": 260
},
{
"epoch": 0.7387140902872777,
"grad_norm": 0.7895988821983337,
"learning_rate": 2.3030821917808222e-05,
"loss": 5.3308,
"step": 270
},
{
"epoch": 0.7660738714090287,
"grad_norm": 0.7869711518287659,
"learning_rate": 2.3886986301369866e-05,
"loss": 5.3397,
"step": 280
},
{
"epoch": 0.7934336525307798,
"grad_norm": 0.7213762998580933,
"learning_rate": 2.4743150684931506e-05,
"loss": 5.3171,
"step": 290
},
{
"epoch": 0.8207934336525308,
"grad_norm": 1.3395589590072632,
"learning_rate": 2.5599315068493153e-05,
"loss": 5.3219,
"step": 300
},
{
"epoch": 0.8481532147742818,
"grad_norm": 0.7428280711174011,
"learning_rate": 2.6455479452054793e-05,
"loss": 5.3191,
"step": 310
},
{
"epoch": 0.8755129958960328,
"grad_norm": 0.7201780080795288,
"learning_rate": 2.731164383561644e-05,
"loss": 5.3208,
"step": 320
},
{
"epoch": 0.9028727770177839,
"grad_norm": 0.7931995987892151,
"learning_rate": 2.8167808219178083e-05,
"loss": 5.3302,
"step": 330
},
{
"epoch": 0.9302325581395349,
"grad_norm": 0.7398663759231567,
"learning_rate": 2.902397260273973e-05,
"loss": 5.3177,
"step": 340
},
{
"epoch": 0.957592339261286,
"grad_norm": 0.7114181518554688,
"learning_rate": 2.988013698630137e-05,
"loss": 5.3008,
"step": 350
},
{
"epoch": 0.9849521203830369,
"grad_norm": 0.8410310745239258,
"learning_rate": 3.073630136986301e-05,
"loss": 5.2878,
"step": 360
},
{
"epoch": 1.0109439124487003,
"grad_norm": 0.7429641485214233,
"learning_rate": 3.1592465753424663e-05,
"loss": 5.0452,
"step": 370
},
{
"epoch": 1.0383036935704515,
"grad_norm": 0.7275601029396057,
"learning_rate": 3.2448630136986303e-05,
"loss": 5.2838,
"step": 380
},
{
"epoch": 1.0656634746922025,
"grad_norm": 0.6859568357467651,
"learning_rate": 3.330479452054795e-05,
"loss": 5.2884,
"step": 390
},
{
"epoch": 1.0930232558139534,
"grad_norm": 0.6706216335296631,
"learning_rate": 3.416095890410959e-05,
"loss": 5.2761,
"step": 400
},
{
"epoch": 1.1203830369357046,
"grad_norm": 0.7339576482772827,
"learning_rate": 3.501712328767123e-05,
"loss": 5.269,
"step": 410
},
{
"epoch": 1.1477428180574556,
"grad_norm": 0.6373031139373779,
"learning_rate": 3.587328767123288e-05,
"loss": 5.2589,
"step": 420
},
{
"epoch": 1.1751025991792066,
"grad_norm": 0.6594179272651672,
"learning_rate": 3.6729452054794524e-05,
"loss": 5.2607,
"step": 430
},
{
"epoch": 1.2024623803009575,
"grad_norm": 0.6628535389900208,
"learning_rate": 3.7585616438356164e-05,
"loss": 5.2669,
"step": 440
},
{
"epoch": 1.2298221614227085,
"grad_norm": 0.6285620927810669,
"learning_rate": 3.844178082191781e-05,
"loss": 5.2483,
"step": 450
},
{
"epoch": 1.2571819425444597,
"grad_norm": 0.6873233914375305,
"learning_rate": 3.929794520547945e-05,
"loss": 5.2478,
"step": 460
},
{
"epoch": 1.2845417236662107,
"grad_norm": 0.7597825527191162,
"learning_rate": 4.01541095890411e-05,
"loss": 5.2425,
"step": 470
},
{
"epoch": 1.3119015047879616,
"grad_norm": 0.7047654986381531,
"learning_rate": 4.1010273972602745e-05,
"loss": 5.2201,
"step": 480
},
{
"epoch": 1.3392612859097128,
"grad_norm": 0.7056351900100708,
"learning_rate": 4.1866438356164385e-05,
"loss": 5.2199,
"step": 490
},
{
"epoch": 1.3666210670314638,
"grad_norm": 0.8730869889259338,
"learning_rate": 4.272260273972603e-05,
"loss": 5.2438,
"step": 500
},
{
"epoch": 1.3939808481532148,
"grad_norm": 0.6341605186462402,
"learning_rate": 4.357876712328767e-05,
"loss": 5.228,
"step": 510
},
{
"epoch": 1.4213406292749657,
"grad_norm": 0.7111886143684387,
"learning_rate": 4.443493150684932e-05,
"loss": 5.1975,
"step": 520
},
{
"epoch": 1.4487004103967167,
"grad_norm": 1.011742353439331,
"learning_rate": 4.529109589041096e-05,
"loss": 5.1971,
"step": 530
},
{
"epoch": 1.476060191518468,
"grad_norm": 0.610575795173645,
"learning_rate": 4.6147260273972605e-05,
"loss": 5.1742,
"step": 540
},
{
"epoch": 1.5034199726402189,
"grad_norm": 0.7265263795852661,
"learning_rate": 4.700342465753425e-05,
"loss": 5.177,
"step": 550
},
{
"epoch": 1.53077975376197,
"grad_norm": 0.961155354976654,
"learning_rate": 4.785958904109589e-05,
"loss": 5.1744,
"step": 560
},
{
"epoch": 1.558139534883721,
"grad_norm": 0.7132388353347778,
"learning_rate": 4.871575342465753e-05,
"loss": 5.1624,
"step": 570
},
{
"epoch": 1.585499316005472,
"grad_norm": 0.8946133255958557,
"learning_rate": 4.957191780821918e-05,
"loss": 5.1641,
"step": 580
},
{
"epoch": 1.612859097127223,
"grad_norm": 0.6424533724784851,
"learning_rate": 5.0428082191780826e-05,
"loss": 5.1597,
"step": 590
},
{
"epoch": 1.640218878248974,
"grad_norm": 0.7116928696632385,
"learning_rate": 5.1284246575342466e-05,
"loss": 5.1431,
"step": 600
},
{
"epoch": 1.667578659370725,
"grad_norm": 0.5708991885185242,
"learning_rate": 5.2140410958904106e-05,
"loss": 5.1362,
"step": 610
},
{
"epoch": 1.694938440492476,
"grad_norm": 0.7077378034591675,
"learning_rate": 5.299657534246576e-05,
"loss": 5.129,
"step": 620
},
{
"epoch": 1.722298221614227,
"grad_norm": 0.660061240196228,
"learning_rate": 5.38527397260274e-05,
"loss": 5.0988,
"step": 630
},
{
"epoch": 1.7496580027359783,
"grad_norm": 0.643544614315033,
"learning_rate": 5.4708904109589046e-05,
"loss": 5.0985,
"step": 640
},
{
"epoch": 1.7770177838577292,
"grad_norm": 0.6883527040481567,
"learning_rate": 5.5565068493150686e-05,
"loss": 5.1039,
"step": 650
},
{
"epoch": 1.8043775649794802,
"grad_norm": 0.7061907052993774,
"learning_rate": 5.6421232876712326e-05,
"loss": 5.0852,
"step": 660
},
{
"epoch": 1.8317373461012312,
"grad_norm": 0.6760084629058838,
"learning_rate": 5.727739726027398e-05,
"loss": 5.0624,
"step": 670
},
{
"epoch": 1.8590971272229821,
"grad_norm": 0.8989177346229553,
"learning_rate": 5.813356164383562e-05,
"loss": 5.0514,
"step": 680
},
{
"epoch": 1.8864569083447331,
"grad_norm": 0.6468309760093689,
"learning_rate": 5.898972602739726e-05,
"loss": 5.0351,
"step": 690
},
{
"epoch": 1.9138166894664843,
"grad_norm": 0.6348468661308289,
"learning_rate": 5.984589041095891e-05,
"loss": 5.0217,
"step": 700
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.6308233737945557,
"learning_rate": 6.070205479452055e-05,
"loss": 5.0042,
"step": 710
},
{
"epoch": 1.9685362517099865,
"grad_norm": 0.6770797967910767,
"learning_rate": 6.15582191780822e-05,
"loss": 4.967,
"step": 720
},
{
"epoch": 1.9958960328317374,
"grad_norm": 1.0092562437057495,
"learning_rate": 6.241438356164385e-05,
"loss": 4.9914,
"step": 730
},
{
"epoch": 2.0218878248974006,
"grad_norm": 1.0360336303710938,
"learning_rate": 6.327054794520548e-05,
"loss": 4.6945,
"step": 740
},
{
"epoch": 2.049247606019152,
"grad_norm": 0.8279546499252319,
"learning_rate": 6.412671232876713e-05,
"loss": 4.8816,
"step": 750
},
{
"epoch": 2.076607387140903,
"grad_norm": 0.7100883722305298,
"learning_rate": 6.498287671232876e-05,
"loss": 4.899,
"step": 760
},
{
"epoch": 2.103967168262654,
"grad_norm": 0.729179859161377,
"learning_rate": 6.583904109589042e-05,
"loss": 4.888,
"step": 770
},
{
"epoch": 2.131326949384405,
"grad_norm": 0.7334495782852173,
"learning_rate": 6.669520547945205e-05,
"loss": 4.9031,
"step": 780
},
{
"epoch": 2.158686730506156,
"grad_norm": 0.7759724855422974,
"learning_rate": 6.75513698630137e-05,
"loss": 4.8494,
"step": 790
},
{
"epoch": 2.186046511627907,
"grad_norm": 0.7172778248786926,
"learning_rate": 6.840753424657535e-05,
"loss": 4.8329,
"step": 800
},
{
"epoch": 2.213406292749658,
"grad_norm": 0.6919230818748474,
"learning_rate": 6.926369863013698e-05,
"loss": 4.833,
"step": 810
},
{
"epoch": 2.2407660738714092,
"grad_norm": 0.8211776614189148,
"learning_rate": 7.011986301369864e-05,
"loss": 4.808,
"step": 820
},
{
"epoch": 2.26812585499316,
"grad_norm": 0.6922405958175659,
"learning_rate": 7.097602739726028e-05,
"loss": 4.8219,
"step": 830
},
{
"epoch": 2.295485636114911,
"grad_norm": 0.631967306137085,
"learning_rate": 7.183219178082192e-05,
"loss": 4.7738,
"step": 840
},
{
"epoch": 2.322845417236662,
"grad_norm": 0.7965299487113953,
"learning_rate": 7.268835616438357e-05,
"loss": 4.7908,
"step": 850
},
{
"epoch": 2.350205198358413,
"grad_norm": 0.912044107913971,
"learning_rate": 7.35445205479452e-05,
"loss": 4.7849,
"step": 860
},
{
"epoch": 2.377564979480164,
"grad_norm": 0.7506100535392761,
"learning_rate": 7.440068493150685e-05,
"loss": 4.6735,
"step": 870
},
{
"epoch": 2.404924760601915,
"grad_norm": 0.7374458909034729,
"learning_rate": 7.52568493150685e-05,
"loss": 4.7273,
"step": 880
},
{
"epoch": 2.432284541723666,
"grad_norm": 0.8312308192253113,
"learning_rate": 7.611301369863014e-05,
"loss": 4.6863,
"step": 890
},
{
"epoch": 2.459644322845417,
"grad_norm": 6.82420015335083,
"learning_rate": 7.696917808219179e-05,
"loss": 4.7248,
"step": 900
},
{
"epoch": 2.4870041039671684,
"grad_norm": 0.7571892738342285,
"learning_rate": 7.782534246575342e-05,
"loss": 4.6727,
"step": 910
},
{
"epoch": 2.5143638850889194,
"grad_norm": 0.7730613946914673,
"learning_rate": 7.868150684931507e-05,
"loss": 4.6484,
"step": 920
},
{
"epoch": 2.5417236662106704,
"grad_norm": 0.9539445638656616,
"learning_rate": 7.953767123287672e-05,
"loss": 4.6307,
"step": 930
},
{
"epoch": 2.5690834473324213,
"grad_norm": 0.8009161949157715,
"learning_rate": 8.039383561643836e-05,
"loss": 4.7371,
"step": 940
},
{
"epoch": 2.5964432284541723,
"grad_norm": 1.1688209772109985,
"learning_rate": 8.125000000000001e-05,
"loss": 4.688,
"step": 950
},
{
"epoch": 2.6238030095759233,
"grad_norm": 0.911170482635498,
"learning_rate": 8.210616438356164e-05,
"loss": 4.6532,
"step": 960
},
{
"epoch": 2.6511627906976747,
"grad_norm": 0.8589779138565063,
"learning_rate": 8.296232876712329e-05,
"loss": 4.7151,
"step": 970
},
{
"epoch": 2.6785225718194257,
"grad_norm": 0.8912016749382019,
"learning_rate": 8.381849315068494e-05,
"loss": 4.584,
"step": 980
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.6658539175987244,
"learning_rate": 8.467465753424658e-05,
"loss": 4.5391,
"step": 990
},
{
"epoch": 2.7332421340629276,
"grad_norm": 0.7096680402755737,
"learning_rate": 8.553082191780823e-05,
"loss": 4.6146,
"step": 1000
},
{
"epoch": 2.7606019151846786,
"grad_norm": 0.9317521452903748,
"learning_rate": 8.638698630136986e-05,
"loss": 4.6209,
"step": 1010
},
{
"epoch": 2.7879616963064295,
"grad_norm": 0.7910155653953552,
"learning_rate": 8.724315068493151e-05,
"loss": 4.5929,
"step": 1020
},
{
"epoch": 2.8153214774281805,
"grad_norm": 0.8584729433059692,
"learning_rate": 8.809931506849316e-05,
"loss": 4.5904,
"step": 1030
},
{
"epoch": 2.8426812585499315,
"grad_norm": 0.8470874428749084,
"learning_rate": 8.89554794520548e-05,
"loss": 4.5757,
"step": 1040
},
{
"epoch": 2.8700410396716824,
"grad_norm": 0.7197965383529663,
"learning_rate": 8.981164383561644e-05,
"loss": 4.5535,
"step": 1050
},
{
"epoch": 2.8974008207934334,
"grad_norm": 0.8799428939819336,
"learning_rate": 9.066780821917808e-05,
"loss": 4.5422,
"step": 1060
},
{
"epoch": 2.924760601915185,
"grad_norm": 0.9093697667121887,
"learning_rate": 9.152397260273973e-05,
"loss": 4.5226,
"step": 1070
},
{
"epoch": 2.952120383036936,
"grad_norm": 0.8678073883056641,
"learning_rate": 9.238013698630136e-05,
"loss": 4.5103,
"step": 1080
},
{
"epoch": 2.9794801641586868,
"grad_norm": 0.8454898595809937,
"learning_rate": 9.323630136986302e-05,
"loss": 4.4193,
"step": 1090
},
{
"epoch": 3.0054719562243504,
"grad_norm": 1.0316234827041626,
"learning_rate": 9.409246575342466e-05,
"loss": 4.1728,
"step": 1100
},
{
"epoch": 3.0328317373461013,
"grad_norm": 1.0127418041229248,
"learning_rate": 9.49486301369863e-05,
"loss": 4.3864,
"step": 1110
},
{
"epoch": 3.0601915184678523,
"grad_norm": 1.0852205753326416,
"learning_rate": 9.580479452054795e-05,
"loss": 4.2815,
"step": 1120
},
{
"epoch": 3.0875512995896033,
"grad_norm": 0.9563126564025879,
"learning_rate": 9.666095890410958e-05,
"loss": 4.3006,
"step": 1130
},
{
"epoch": 3.1149110807113543,
"grad_norm": 0.8504606485366821,
"learning_rate": 9.751712328767125e-05,
"loss": 4.3326,
"step": 1140
},
{
"epoch": 3.1422708618331052,
"grad_norm": 0.90372633934021,
"learning_rate": 9.837328767123288e-05,
"loss": 4.2369,
"step": 1150
},
{
"epoch": 3.169630642954856,
"grad_norm": 0.9138619899749756,
"learning_rate": 9.922945205479453e-05,
"loss": 4.2533,
"step": 1160
},
{
"epoch": 3.196990424076607,
"grad_norm": 0.8798762559890747,
"learning_rate": 9.999999776710118e-05,
"loss": 4.3048,
"step": 1170
},
{
"epoch": 3.2243502051983586,
"grad_norm": 0.9538683295249939,
"learning_rate": 9.999972981948274e-05,
"loss": 4.2993,
"step": 1180
},
{
"epoch": 3.2517099863201095,
"grad_norm": 0.8241022825241089,
"learning_rate": 9.999901529484025e-05,
"loss": 4.269,
"step": 1190
},
{
"epoch": 3.2790697674418605,
"grad_norm": 0.6876862645149231,
"learning_rate": 9.999785419955556e-05,
"loss": 4.31,
"step": 1200
},
{
"epoch": 3.3064295485636115,
"grad_norm": 0.7811799645423889,
"learning_rate": 9.999624654399908e-05,
"loss": 4.2854,
"step": 1210
},
{
"epoch": 3.3337893296853625,
"grad_norm": 0.8641256093978882,
"learning_rate": 9.999419234252974e-05,
"loss": 4.2598,
"step": 1220
},
{
"epoch": 3.3611491108071134,
"grad_norm": 0.8587515354156494,
"learning_rate": 9.999169161349483e-05,
"loss": 4.288,
"step": 1230
},
{
"epoch": 3.3885088919288644,
"grad_norm": 0.7273943424224854,
"learning_rate": 9.99887443792298e-05,
"loss": 4.2348,
"step": 1240
},
{
"epoch": 3.415868673050616,
"grad_norm": 0.9900055527687073,
"learning_rate": 9.998535066605816e-05,
"loss": 4.2331,
"step": 1250
},
{
"epoch": 3.443228454172367,
"grad_norm": 0.7297905087471008,
"learning_rate": 9.998151050429116e-05,
"loss": 4.2203,
"step": 1260
},
{
"epoch": 3.4705882352941178,
"grad_norm": 0.814732015132904,
"learning_rate": 9.997722392822755e-05,
"loss": 4.2054,
"step": 1270
},
{
"epoch": 3.4979480164158687,
"grad_norm": 1.163109302520752,
"learning_rate": 9.997249097615325e-05,
"loss": 4.1995,
"step": 1280
},
{
"epoch": 3.5253077975376197,
"grad_norm": 0.9196097254753113,
"learning_rate": 9.996731169034106e-05,
"loss": 4.1926,
"step": 1290
},
{
"epoch": 3.5526675786593707,
"grad_norm": 1.0712329149246216,
"learning_rate": 9.99616861170502e-05,
"loss": 4.1976,
"step": 1300
},
{
"epoch": 3.5800273597811216,
"grad_norm": 1.0013926029205322,
"learning_rate": 9.9955614306526e-05,
"loss": 4.1654,
"step": 1310
},
{
"epoch": 3.6073871409028726,
"grad_norm": 1.137804627418518,
"learning_rate": 9.994909631299938e-05,
"loss": 4.135,
"step": 1320
},
{
"epoch": 3.6347469220246236,
"grad_norm": 0.9330766201019287,
"learning_rate": 9.994213219468638e-05,
"loss": 4.2052,
"step": 1330
},
{
"epoch": 3.662106703146375,
"grad_norm": 0.9428395628929138,
"learning_rate": 9.993472201378759e-05,
"loss": 4.2319,
"step": 1340
},
{
"epoch": 3.689466484268126,
"grad_norm": 0.9099085330963135,
"learning_rate": 9.992686583648778e-05,
"loss": 4.0931,
"step": 1350
},
{
"epoch": 3.716826265389877,
"grad_norm": 0.8021682500839233,
"learning_rate": 9.991856373295503e-05,
"loss": 4.0936,
"step": 1360
},
{
"epoch": 3.744186046511628,
"grad_norm": 1.0988266468048096,
"learning_rate": 9.990981577734034e-05,
"loss": 4.1154,
"step": 1370
},
{
"epoch": 3.771545827633379,
"grad_norm": 0.8525907397270203,
"learning_rate": 9.990062204777684e-05,
"loss": 4.141,
"step": 1380
},
{
"epoch": 3.79890560875513,
"grad_norm": 1.0354300737380981,
"learning_rate": 9.989098262637917e-05,
"loss": 4.095,
"step": 1390
},
{
"epoch": 3.8262653898768813,
"grad_norm": 0.9018966555595398,
"learning_rate": 9.988089759924264e-05,
"loss": 4.089,
"step": 1400
},
{
"epoch": 3.853625170998632,
"grad_norm": 1.0135756731033325,
"learning_rate": 9.987036705644259e-05,
"loss": 4.0547,
"step": 1410
},
{
"epoch": 3.880984952120383,
"grad_norm": 1.1471967697143555,
"learning_rate": 9.98593910920335e-05,
"loss": 4.0537,
"step": 1420
},
{
"epoch": 3.908344733242134,
"grad_norm": 1.05537748336792,
"learning_rate": 9.984796980404815e-05,
"loss": 4.0033,
"step": 1430
},
{
"epoch": 3.935704514363885,
"grad_norm": 0.770301103591919,
"learning_rate": 9.98361032944968e-05,
"loss": 4.0572,
"step": 1440
},
{
"epoch": 3.963064295485636,
"grad_norm": 1.1304564476013184,
"learning_rate": 9.982379166936622e-05,
"loss": 3.9502,
"step": 1450
},
{
"epoch": 3.990424076607387,
"grad_norm": 0.8373954892158508,
"learning_rate": 9.98110350386188e-05,
"loss": 4.0933,
"step": 1460
},
{
"epoch": 4.01641586867305,
"grad_norm": 1.0730388164520264,
"learning_rate": 9.979783351619151e-05,
"loss": 3.6957,
"step": 1470
},
{
"epoch": 4.043775649794801,
"grad_norm": 1.027195692062378,
"learning_rate": 9.978418721999492e-05,
"loss": 3.8039,
"step": 1480
},
{
"epoch": 4.071135430916553,
"grad_norm": 1.1432992219924927,
"learning_rate": 9.977009627191216e-05,
"loss": 3.7143,
"step": 1490
},
{
"epoch": 4.098495212038304,
"grad_norm": 1.5777379274368286,
"learning_rate": 9.975556079779774e-05,
"loss": 3.9107,
"step": 1500
},
{
"epoch": 4.125854993160055,
"grad_norm": 1.5348365306854248,
"learning_rate": 9.974058092747655e-05,
"loss": 3.7954,
"step": 1510
},
{
"epoch": 4.153214774281806,
"grad_norm": 1.3011326789855957,
"learning_rate": 9.972515679474266e-05,
"loss": 3.7641,
"step": 1520
},
{
"epoch": 4.180574555403557,
"grad_norm": 1.031332015991211,
"learning_rate": 9.970928853735803e-05,
"loss": 3.7675,
"step": 1530
},
{
"epoch": 4.207934336525308,
"grad_norm": 1.0175678730010986,
"learning_rate": 9.969297629705148e-05,
"loss": 3.7914,
"step": 1540
},
{
"epoch": 4.235294117647059,
"grad_norm": 1.224719524383545,
"learning_rate": 9.967622021951715e-05,
"loss": 3.7265,
"step": 1550
},
{
"epoch": 4.26265389876881,
"grad_norm": 1.3663736581802368,
"learning_rate": 9.965902045441349e-05,
"loss": 3.8017,
"step": 1560
},
{
"epoch": 4.290013679890561,
"grad_norm": 1.4514472484588623,
"learning_rate": 9.964137715536171e-05,
"loss": 3.8975,
"step": 1570
},
{
"epoch": 4.317373461012312,
"grad_norm": 1.1808091402053833,
"learning_rate": 9.96232904799445e-05,
"loss": 3.8818,
"step": 1580
},
{
"epoch": 4.344733242134063,
"grad_norm": 0.9281204342842102,
"learning_rate": 9.960476058970463e-05,
"loss": 3.7538,
"step": 1590
},
{
"epoch": 4.372093023255814,
"grad_norm": 1.257411241531372,
"learning_rate": 9.958578765014342e-05,
"loss": 3.7657,
"step": 1600
},
{
"epoch": 4.399452804377565,
"grad_norm": 1.1760926246643066,
"learning_rate": 9.956637183071937e-05,
"loss": 3.6769,
"step": 1610
},
{
"epoch": 4.426812585499316,
"grad_norm": 0.9812909960746765,
"learning_rate": 9.954651330484664e-05,
"loss": 3.7848,
"step": 1620
},
{
"epoch": 4.454172366621067,
"grad_norm": 0.9935783743858337,
"learning_rate": 9.952621224989336e-05,
"loss": 3.8005,
"step": 1630
},
{
"epoch": 4.4815321477428185,
"grad_norm": 1.1927580833435059,
"learning_rate": 9.950546884718023e-05,
"loss": 3.774,
"step": 1640
},
{
"epoch": 4.5088919288645695,
"grad_norm": 1.4293582439422607,
"learning_rate": 9.948428328197879e-05,
"loss": 3.84,
"step": 1650
},
{
"epoch": 4.53625170998632,
"grad_norm": 1.0753121376037598,
"learning_rate": 9.946265574350979e-05,
"loss": 3.8591,
"step": 1660
},
{
"epoch": 4.563611491108071,
"grad_norm": 1.0058224201202393,
"learning_rate": 9.944058642494152e-05,
"loss": 3.7587,
"step": 1670
},
{
"epoch": 4.590971272229822,
"grad_norm": 1.0442452430725098,
"learning_rate": 9.941807552338804e-05,
"loss": 3.7696,
"step": 1680
},
{
"epoch": 4.618331053351573,
"grad_norm": 1.4075640439987183,
"learning_rate": 9.939512323990749e-05,
"loss": 3.8431,
"step": 1690
},
{
"epoch": 4.645690834473324,
"grad_norm": 1.1082020998001099,
"learning_rate": 9.93717297795002e-05,
"loss": 3.7459,
"step": 1700
},
{
"epoch": 4.673050615595075,
"grad_norm": 1.102745532989502,
"learning_rate": 9.934789535110696e-05,
"loss": 3.7705,
"step": 1710
},
{
"epoch": 4.700410396716826,
"grad_norm": 1.1125171184539795,
"learning_rate": 9.932362016760706e-05,
"loss": 3.805,
"step": 1720
},
{
"epoch": 4.727770177838577,
"grad_norm": 0.9724751710891724,
"learning_rate": 9.929890444581648e-05,
"loss": 3.7315,
"step": 1730
},
{
"epoch": 4.755129958960328,
"grad_norm": 1.0714983940124512,
"learning_rate": 9.927374840648587e-05,
"loss": 3.7966,
"step": 1740
},
{
"epoch": 4.782489740082079,
"grad_norm": 0.9851176738739014,
"learning_rate": 9.924815227429862e-05,
"loss": 3.7883,
"step": 1750
},
{
"epoch": 4.80984952120383,
"grad_norm": 0.8671346306800842,
"learning_rate": 9.922211627786886e-05,
"loss": 3.7598,
"step": 1760
},
{
"epoch": 4.837209302325581,
"grad_norm": 1.0850796699523926,
"learning_rate": 9.919564064973942e-05,
"loss": 3.6914,
"step": 1770
},
{
"epoch": 4.864569083447332,
"grad_norm": 0.9974954128265381,
"learning_rate": 9.916872562637971e-05,
"loss": 3.7613,
"step": 1780
},
{
"epoch": 4.891928864569083,
"grad_norm": 1.1243762969970703,
"learning_rate": 9.914137144818364e-05,
"loss": 3.7619,
"step": 1790
},
{
"epoch": 4.919288645690834,
"grad_norm": 1.035919427871704,
"learning_rate": 9.91135783594675e-05,
"loss": 3.7513,
"step": 1800
},
{
"epoch": 4.946648426812586,
"grad_norm": 1.2536144256591797,
"learning_rate": 9.908534660846771e-05,
"loss": 3.7849,
"step": 1810
},
{
"epoch": 4.974008207934337,
"grad_norm": 1.2192177772521973,
"learning_rate": 9.905667644733868e-05,
"loss": 3.7436,
"step": 1820
},
{
"epoch": 5.0,
"grad_norm": 1.2116731405258179,
"learning_rate": 9.902756813215045e-05,
"loss": 3.5384,
"step": 1830
},
{
"epoch": 5.027359781121751,
"grad_norm": 1.4592894315719604,
"learning_rate": 9.899802192288659e-05,
"loss": 3.3977,
"step": 1840
},
{
"epoch": 5.054719562243502,
"grad_norm": 1.2525421380996704,
"learning_rate": 9.896803808344164e-05,
"loss": 3.4254,
"step": 1850
},
{
"epoch": 5.082079343365253,
"grad_norm": 1.446140170097351,
"learning_rate": 9.893761688161894e-05,
"loss": 3.3259,
"step": 1860
},
{
"epoch": 5.109439124487004,
"grad_norm": 1.4736069440841675,
"learning_rate": 9.890675858912815e-05,
"loss": 3.3374,
"step": 1870
},
{
"epoch": 5.136798905608755,
"grad_norm": 1.5665637254714966,
"learning_rate": 9.887546348158286e-05,
"loss": 3.3545,
"step": 1880
},
{
"epoch": 5.164158686730506,
"grad_norm": 1.2610946893692017,
"learning_rate": 9.884373183849809e-05,
"loss": 3.2986,
"step": 1890
},
{
"epoch": 5.191518467852257,
"grad_norm": 1.2117750644683838,
"learning_rate": 9.881156394328782e-05,
"loss": 3.3309,
"step": 1900
},
{
"epoch": 5.218878248974008,
"grad_norm": 1.4232985973358154,
"learning_rate": 9.877896008326247e-05,
"loss": 3.355,
"step": 1910
},
{
"epoch": 5.24623803009576,
"grad_norm": 1.1137495040893555,
"learning_rate": 9.87459205496263e-05,
"loss": 3.3061,
"step": 1920
},
{
"epoch": 5.273597811217511,
"grad_norm": 1.3471006155014038,
"learning_rate": 9.871244563747487e-05,
"loss": 3.4957,
"step": 1930
},
{
"epoch": 5.300957592339262,
"grad_norm": 1.1785922050476074,
"learning_rate": 9.867853564579225e-05,
"loss": 3.3743,
"step": 1940
},
{
"epoch": 5.3283173734610125,
"grad_norm": 1.392377257347107,
"learning_rate": 9.864419087744864e-05,
"loss": 3.4454,
"step": 1950
},
{
"epoch": 5.3556771545827635,
"grad_norm": 1.149776577949524,
"learning_rate": 9.860941163919733e-05,
"loss": 3.3712,
"step": 1960
},
{
"epoch": 5.3830369357045145,
"grad_norm": 1.3407548666000366,
"learning_rate": 9.857419824167216e-05,
"loss": 3.4483,
"step": 1970
},
{
"epoch": 5.410396716826265,
"grad_norm": 1.110605239868164,
"learning_rate": 9.853855099938476e-05,
"loss": 3.3094,
"step": 1980
},
{
"epoch": 5.437756497948016,
"grad_norm": 1.1149815320968628,
"learning_rate": 9.85024702307216e-05,
"loss": 3.4469,
"step": 1990
},
{
"epoch": 5.465116279069767,
"grad_norm": 1.6999688148498535,
"learning_rate": 9.846595625794128e-05,
"loss": 3.3631,
"step": 2000
},
{
"epoch": 5.492476060191518,
"grad_norm": 1.4271352291107178,
"learning_rate": 9.842900940717159e-05,
"loss": 3.3652,
"step": 2010
},
{
"epoch": 5.519835841313269,
"grad_norm": 1.1635977029800415,
"learning_rate": 9.839163000840661e-05,
"loss": 3.447,
"step": 2020
},
{
"epoch": 5.54719562243502,
"grad_norm": 1.150688648223877,
"learning_rate": 9.835381839550376e-05,
"loss": 3.462,
"step": 2030
},
{
"epoch": 5.574555403556771,
"grad_norm": 1.157528042793274,
"learning_rate": 9.831557490618079e-05,
"loss": 3.472,
"step": 2040
},
{
"epoch": 5.601915184678522,
"grad_norm": 1.1749804019927979,
"learning_rate": 9.827689988201288e-05,
"loss": 3.3191,
"step": 2050
},
{
"epoch": 5.629274965800274,
"grad_norm": 1.2330487966537476,
"learning_rate": 9.823779366842936e-05,
"loss": 3.4122,
"step": 2060
},
{
"epoch": 5.656634746922025,
"grad_norm": 1.4110299348831177,
"learning_rate": 9.81982566147109e-05,
"loss": 3.3948,
"step": 2070
},
{
"epoch": 5.683994528043776,
"grad_norm": 1.3113088607788086,
"learning_rate": 9.815828907398619e-05,
"loss": 3.3534,
"step": 2080
},
{
"epoch": 5.711354309165527,
"grad_norm": 1.182449460029602,
"learning_rate": 9.811789140322884e-05,
"loss": 3.4261,
"step": 2090
},
{
"epoch": 5.738714090287278,
"grad_norm": 1.3011144399642944,
"learning_rate": 9.807706396325429e-05,
"loss": 3.4271,
"step": 2100
},
{
"epoch": 5.766073871409029,
"grad_norm": 1.1780519485473633,
"learning_rate": 9.80358071187164e-05,
"loss": 3.4282,
"step": 2110
},
{
"epoch": 5.79343365253078,
"grad_norm": 1.1470130681991577,
"learning_rate": 9.799412123810436e-05,
"loss": 3.3772,
"step": 2120
},
{
"epoch": 5.820793433652531,
"grad_norm": 1.223630666732788,
"learning_rate": 9.79520066937393e-05,
"loss": 3.4185,
"step": 2130
},
{
"epoch": 5.848153214774282,
"grad_norm": 1.1980472803115845,
"learning_rate": 9.7909463861771e-05,
"loss": 3.3719,
"step": 2140
},
{
"epoch": 5.875512995896033,
"grad_norm": 1.2304013967514038,
"learning_rate": 9.786649312217455e-05,
"loss": 3.4083,
"step": 2150
},
{
"epoch": 5.902872777017784,
"grad_norm": 1.0939279794692993,
"learning_rate": 9.782309485874694e-05,
"loss": 3.3645,
"step": 2160
},
{
"epoch": 5.930232558139535,
"grad_norm": 1.0186386108398438,
"learning_rate": 9.777926945910359e-05,
"loss": 3.3867,
"step": 2170
},
{
"epoch": 5.957592339261286,
"grad_norm": 1.1139832735061646,
"learning_rate": 9.773501731467495e-05,
"loss": 3.4506,
"step": 2180
},
{
"epoch": 5.984952120383037,
"grad_norm": 1.3426393270492554,
"learning_rate": 9.769033882070297e-05,
"loss": 3.4173,
"step": 2190
},
{
"epoch": 6.010943912448701,
"grad_norm": 2.4351305961608887,
"learning_rate": 9.76452343762376e-05,
"loss": 3.0589,
"step": 2200
},
{
"epoch": 6.038303693570452,
"grad_norm": 1.7416296005249023,
"learning_rate": 9.75997043841332e-05,
"loss": 2.9526,
"step": 2210
},
{
"epoch": 6.065663474692203,
"grad_norm": 1.677554726600647,
"learning_rate": 9.755374925104488e-05,
"loss": 2.8457,
"step": 2220
},
{
"epoch": 6.093023255813954,
"grad_norm": 2.02108097076416,
"learning_rate": 9.750736938742504e-05,
"loss": 2.7604,
"step": 2230
},
{
"epoch": 6.120383036935705,
"grad_norm": 1.8085135221481323,
"learning_rate": 9.746056520751953e-05,
"loss": 2.8228,
"step": 2240
},
{
"epoch": 6.147742818057456,
"grad_norm": 1.8664541244506836,
"learning_rate": 9.741333712936404e-05,
"loss": 2.8169,
"step": 2250
},
{
"epoch": 6.175102599179207,
"grad_norm": 1.7058764696121216,
"learning_rate": 9.736568557478035e-05,
"loss": 2.8312,
"step": 2260
},
{
"epoch": 6.2024623803009575,
"grad_norm": 1.8776473999023438,
"learning_rate": 9.731761096937252e-05,
"loss": 2.8457,
"step": 2270
},
{
"epoch": 6.2298221614227085,
"grad_norm": 1.5233421325683594,
"learning_rate": 9.726911374252319e-05,
"loss": 2.8071,
"step": 2280
},
{
"epoch": 6.2571819425444595,
"grad_norm": 1.7072609663009644,
"learning_rate": 9.722019432738962e-05,
"loss": 2.8671,
"step": 2290
},
{
"epoch": 6.2845417236662104,
"grad_norm": 1.4505999088287354,
"learning_rate": 9.717085316089993e-05,
"loss": 2.8291,
"step": 2300
},
{
"epoch": 6.311901504787961,
"grad_norm": 1.4433178901672363,
"learning_rate": 9.712109068374909e-05,
"loss": 2.8472,
"step": 2310
},
{
"epoch": 6.339261285909712,
"grad_norm": 1.3486769199371338,
"learning_rate": 9.707090734039512e-05,
"loss": 2.9171,
"step": 2320
},
{
"epoch": 6.366621067031463,
"grad_norm": 1.486006259918213,
"learning_rate": 9.702030357905498e-05,
"loss": 2.8284,
"step": 2330
},
{
"epoch": 6.393980848153214,
"grad_norm": 1.5222647190093994,
"learning_rate": 9.696927985170065e-05,
"loss": 2.8851,
"step": 2340
},
{
"epoch": 6.421340629274966,
"grad_norm": 1.7620928287506104,
"learning_rate": 9.691783661405513e-05,
"loss": 2.965,
"step": 2350
},
{
"epoch": 6.448700410396717,
"grad_norm": 1.424559473991394,
"learning_rate": 9.686597432558819e-05,
"loss": 2.8121,
"step": 2360
},
{
"epoch": 6.476060191518468,
"grad_norm": 1.5835916996002197,
"learning_rate": 9.681369344951252e-05,
"loss": 2.9462,
"step": 2370
},
{
"epoch": 6.503419972640219,
"grad_norm": 1.7365058660507202,
"learning_rate": 9.676099445277938e-05,
"loss": 2.8682,
"step": 2380
},
{
"epoch": 6.53077975376197,
"grad_norm": 1.477047324180603,
"learning_rate": 9.670787780607453e-05,
"loss": 2.9708,
"step": 2390
},
{
"epoch": 6.558139534883721,
"grad_norm": 1.5638231039047241,
"learning_rate": 9.665434398381401e-05,
"loss": 2.8585,
"step": 2400
},
{
"epoch": 6.585499316005472,
"grad_norm": 1.6344341039657593,
"learning_rate": 9.660039346413994e-05,
"loss": 2.8746,
"step": 2410
},
{
"epoch": 6.612859097127223,
"grad_norm": 1.4233092069625854,
"learning_rate": 9.654602672891613e-05,
"loss": 2.9764,
"step": 2420
},
{
"epoch": 6.640218878248974,
"grad_norm": 1.3977011442184448,
"learning_rate": 9.649124426372392e-05,
"loss": 2.8692,
"step": 2430
},
{
"epoch": 6.667578659370725,
"grad_norm": 1.542202353477478,
"learning_rate": 9.643604655785776e-05,
"loss": 2.8571,
"step": 2440
},
{
"epoch": 6.694938440492476,
"grad_norm": 1.7587088346481323,
"learning_rate": 9.638043410432087e-05,
"loss": 2.8904,
"step": 2450
},
{
"epoch": 6.722298221614227,
"grad_norm": 1.6545684337615967,
"learning_rate": 9.63244073998208e-05,
"loss": 2.9859,
"step": 2460
},
{
"epoch": 6.749658002735978,
"grad_norm": 1.4840025901794434,
"learning_rate": 9.626796694476503e-05,
"loss": 3.0061,
"step": 2470
},
{
"epoch": 6.777017783857729,
"grad_norm": 1.3737751245498657,
"learning_rate": 9.621111324325649e-05,
"loss": 2.9635,
"step": 2480
},
{
"epoch": 6.804377564979481,
"grad_norm": 1.4819165468215942,
"learning_rate": 9.615384680308907e-05,
"loss": 2.9223,
"step": 2490
},
{
"epoch": 6.831737346101232,
"grad_norm": 1.5234416723251343,
"learning_rate": 9.609616813574306e-05,
"loss": 3.0179,
"step": 2500
},
{
"epoch": 6.859097127222983,
"grad_norm": 1.4282548427581787,
"learning_rate": 9.603807775638062e-05,
"loss": 2.9568,
"step": 2510
},
{
"epoch": 6.886456908344734,
"grad_norm": 1.4413877725601196,
"learning_rate": 9.597957618384109e-05,
"loss": 2.99,
"step": 2520
},
{
"epoch": 6.9138166894664845,
"grad_norm": 1.513663649559021,
"learning_rate": 9.592066394063648e-05,
"loss": 2.9621,
"step": 2530
},
{
"epoch": 6.9411764705882355,
"grad_norm": 1.3762191534042358,
"learning_rate": 9.58613415529467e-05,
"loss": 2.9866,
"step": 2540
},
{
"epoch": 6.9685362517099865,
"grad_norm": 1.4097365140914917,
"learning_rate": 9.580160955061493e-05,
"loss": 2.9241,
"step": 2550
},
{
"epoch": 6.995896032831737,
"grad_norm": 1.5960757732391357,
"learning_rate": 9.574146846714285e-05,
"loss": 3.0212,
"step": 2560
},
{
"epoch": 7.021887824897401,
"grad_norm": 2.500211238861084,
"learning_rate": 9.568091883968589e-05,
"loss": 2.2289,
"step": 2570
},
{
"epoch": 7.049247606019152,
"grad_norm": 2.6547000408172607,
"learning_rate": 9.561996120904841e-05,
"loss": 2.2243,
"step": 2580
},
{
"epoch": 7.0766073871409025,
"grad_norm": 2.052117347717285,
"learning_rate": 9.555859611967891e-05,
"loss": 2.2437,
"step": 2590
},
{
"epoch": 7.1039671682626535,
"grad_norm": 2.249936103820801,
"learning_rate": 9.549682411966512e-05,
"loss": 2.2265,
"step": 2600
},
{
"epoch": 7.131326949384405,
"grad_norm": 2.7790284156799316,
"learning_rate": 9.543464576072916e-05,
"loss": 2.291,
"step": 2610
},
{
"epoch": 7.158686730506156,
"grad_norm": 2.4017059803009033,
"learning_rate": 9.537206159822252e-05,
"loss": 2.1677,
"step": 2620
},
{
"epoch": 7.186046511627907,
"grad_norm": 2.120255708694458,
"learning_rate": 9.530907219112124e-05,
"loss": 2.1673,
"step": 2630
},
{
"epoch": 7.213406292749658,
"grad_norm": 2.6208271980285645,
"learning_rate": 9.524567810202078e-05,
"loss": 2.2499,
"step": 2640
},
{
"epoch": 7.240766073871409,
"grad_norm": 2.2075304985046387,
"learning_rate": 9.51818798971311e-05,
"loss": 2.2051,
"step": 2650
},
{
"epoch": 7.26812585499316,
"grad_norm": 2.1943113803863525,
"learning_rate": 9.511767814627148e-05,
"loss": 2.2611,
"step": 2660
},
{
"epoch": 7.295485636114911,
"grad_norm": 2.2060434818267822,
"learning_rate": 9.50530734228656e-05,
"loss": 2.1784,
"step": 2670
},
{
"epoch": 7.322845417236662,
"grad_norm": 1.8865910768508911,
"learning_rate": 9.498806630393627e-05,
"loss": 2.2186,
"step": 2680
},
{
"epoch": 7.350205198358413,
"grad_norm": 1.9956541061401367,
"learning_rate": 9.492265737010031e-05,
"loss": 2.1978,
"step": 2690
},
{
"epoch": 7.377564979480164,
"grad_norm": 2.1337575912475586,
"learning_rate": 9.485684720556345e-05,
"loss": 2.2586,
"step": 2700
},
{
"epoch": 7.404924760601915,
"grad_norm": 1.8486758470535278,
"learning_rate": 9.479063639811501e-05,
"loss": 2.2797,
"step": 2710
},
{
"epoch": 7.432284541723666,
"grad_norm": 1.852810263633728,
"learning_rate": 9.472402553912269e-05,
"loss": 2.3055,
"step": 2720
},
{
"epoch": 7.459644322845417,
"grad_norm": 2.111374855041504,
"learning_rate": 9.46570152235273e-05,
"loss": 2.3293,
"step": 2730
},
{
"epoch": 7.487004103967168,
"grad_norm": 2.0791540145874023,
"learning_rate": 9.45896060498374e-05,
"loss": 2.2887,
"step": 2740
},
{
"epoch": 7.514363885088919,
"grad_norm": 2.3228745460510254,
"learning_rate": 9.4521798620124e-05,
"loss": 2.3107,
"step": 2750
},
{
"epoch": 7.54172366621067,
"grad_norm": 1.8341909646987915,
"learning_rate": 9.445359354001522e-05,
"loss": 2.2522,
"step": 2760
},
{
"epoch": 7.569083447332421,
"grad_norm": 2.034987211227417,
"learning_rate": 9.438499141869072e-05,
"loss": 2.3179,
"step": 2770
},
{
"epoch": 7.596443228454173,
"grad_norm": 1.9935376644134521,
"learning_rate": 9.43159928688765e-05,
"loss": 2.3077,
"step": 2780
},
{
"epoch": 7.623803009575924,
"grad_norm": 1.9922071695327759,
"learning_rate": 9.424659850683919e-05,
"loss": 2.4192,
"step": 2790
},
{
"epoch": 7.651162790697675,
"grad_norm": 1.9138284921646118,
"learning_rate": 9.417680895238072e-05,
"loss": 2.3329,
"step": 2800
},
{
"epoch": 7.678522571819426,
"grad_norm": 2.0336709022521973,
"learning_rate": 9.410662482883266e-05,
"loss": 2.3158,
"step": 2810
},
{
"epoch": 7.705882352941177,
"grad_norm": 1.9535430669784546,
"learning_rate": 9.403604676305075e-05,
"loss": 2.3931,
"step": 2820
},
{
"epoch": 7.733242134062928,
"grad_norm": 1.7450408935546875,
"learning_rate": 9.396507538540926e-05,
"loss": 2.3268,
"step": 2830
},
{
"epoch": 7.760601915184679,
"grad_norm": 2.0969526767730713,
"learning_rate": 9.389371132979534e-05,
"loss": 2.4411,
"step": 2840
},
{
"epoch": 7.7879616963064295,
"grad_norm": 2.2872345447540283,
"learning_rate": 9.38219552336034e-05,
"loss": 2.3672,
"step": 2850
},
{
"epoch": 7.8153214774281805,
"grad_norm": 2.089909791946411,
"learning_rate": 9.374980773772932e-05,
"loss": 2.436,
"step": 2860
},
{
"epoch": 7.8426812585499315,
"grad_norm": 2.183398485183716,
"learning_rate": 9.367726948656494e-05,
"loss": 2.2927,
"step": 2870
},
{
"epoch": 7.8700410396716824,
"grad_norm": 1.9796818494796753,
"learning_rate": 9.360434112799204e-05,
"loss": 2.359,
"step": 2880
},
{
"epoch": 7.897400820793433,
"grad_norm": 1.9431896209716797,
"learning_rate": 9.353102331337675e-05,
"loss": 2.3863,
"step": 2890
},
{
"epoch": 7.924760601915184,
"grad_norm": 1.8616565465927124,
"learning_rate": 9.345731669756362e-05,
"loss": 2.391,
"step": 2900
},
{
"epoch": 7.952120383036935,
"grad_norm": 1.7693086862564087,
"learning_rate": 9.338322193886985e-05,
"loss": 2.3664,
"step": 2910
},
{
"epoch": 7.979480164158687,
"grad_norm": 1.8138867616653442,
"learning_rate": 9.330873969907934e-05,
"loss": 2.3916,
"step": 2920
},
{
"epoch": 8.00547195622435,
"grad_norm": 2.7118518352508545,
"learning_rate": 9.323387064343681e-05,
"loss": 2.1808,
"step": 2930
},
{
"epoch": 8.0328317373461,
"grad_norm": 2.688577651977539,
"learning_rate": 9.315861544064191e-05,
"loss": 1.5727,
"step": 2940
},
{
"epoch": 8.060191518467851,
"grad_norm": 2.6679704189300537,
"learning_rate": 9.308297476284313e-05,
"loss": 1.5628,
"step": 2950
},
{
"epoch": 8.087551299589602,
"grad_norm": 2.6832151412963867,
"learning_rate": 9.300694928563192e-05,
"loss": 1.5101,
"step": 2960
},
{
"epoch": 8.114911080711355,
"grad_norm": 2.402343273162842,
"learning_rate": 9.293053968803655e-05,
"loss": 1.4595,
"step": 2970
},
{
"epoch": 8.142270861833106,
"grad_norm": 2.364119291305542,
"learning_rate": 9.285374665251616e-05,
"loss": 1.4641,
"step": 2980
},
{
"epoch": 8.169630642954857,
"grad_norm": 2.2149391174316406,
"learning_rate": 9.277657086495453e-05,
"loss": 1.4611,
"step": 2990
},
{
"epoch": 8.196990424076608,
"grad_norm": 2.76865553855896,
"learning_rate": 9.269901301465404e-05,
"loss": 1.5176,
"step": 3000
},
{
"epoch": 8.224350205198359,
"grad_norm": 2.479785919189453,
"learning_rate": 9.262107379432957e-05,
"loss": 1.4396,
"step": 3010
},
{
"epoch": 8.25170998632011,
"grad_norm": 4.035619258880615,
"learning_rate": 9.254275390010215e-05,
"loss": 1.5584,
"step": 3020
},
{
"epoch": 8.279069767441861,
"grad_norm": 2.317028045654297,
"learning_rate": 9.246405403149284e-05,
"loss": 1.5888,
"step": 3030
},
{
"epoch": 8.306429548563612,
"grad_norm": 2.6506307125091553,
"learning_rate": 9.238497489141655e-05,
"loss": 1.5278,
"step": 3040
},
{
"epoch": 8.333789329685363,
"grad_norm": 2.743687152862549,
"learning_rate": 9.23055171861756e-05,
"loss": 1.5378,
"step": 3050
},
{
"epoch": 8.361149110807114,
"grad_norm": 2.770232915878296,
"learning_rate": 9.222568162545355e-05,
"loss": 1.4971,
"step": 3060
},
{
"epoch": 8.388508891928865,
"grad_norm": 2.4142062664031982,
"learning_rate": 9.214546892230881e-05,
"loss": 1.5929,
"step": 3070
},
{
"epoch": 8.415868673050616,
"grad_norm": 2.8501765727996826,
"learning_rate": 9.206487979316824e-05,
"loss": 1.5939,
"step": 3080
},
{
"epoch": 8.443228454172367,
"grad_norm": 2.7158608436584473,
"learning_rate": 9.198391495782079e-05,
"loss": 1.5819,
"step": 3090
},
{
"epoch": 8.470588235294118,
"grad_norm": 2.708183765411377,
"learning_rate": 9.190257513941109e-05,
"loss": 1.6113,
"step": 3100
},
{
"epoch": 8.497948016415869,
"grad_norm": 2.5786938667297363,
"learning_rate": 9.182086106443293e-05,
"loss": 1.5864,
"step": 3110
},
{
"epoch": 8.52530779753762,
"grad_norm": 2.567384958267212,
"learning_rate": 9.173877346272283e-05,
"loss": 1.6151,
"step": 3120
},
{
"epoch": 8.55266757865937,
"grad_norm": 2.715834140777588,
"learning_rate": 9.16563130674535e-05,
"loss": 1.6565,
"step": 3130
},
{
"epoch": 8.580027359781122,
"grad_norm": 2.477998733520508,
"learning_rate": 9.157348061512727e-05,
"loss": 1.5792,
"step": 3140
},
{
"epoch": 8.607387140902873,
"grad_norm": 2.6714956760406494,
"learning_rate": 9.149027684556953e-05,
"loss": 1.6927,
"step": 3150
},
{
"epoch": 8.634746922024624,
"grad_norm": 2.620237350463867,
"learning_rate": 9.140670250192214e-05,
"loss": 1.6285,
"step": 3160
},
{
"epoch": 8.662106703146375,
"grad_norm": 2.4811065196990967,
"learning_rate": 9.132275833063677e-05,
"loss": 1.6466,
"step": 3170
},
{
"epoch": 8.689466484268126,
"grad_norm": 2.1317250728607178,
"learning_rate": 9.123844508146823e-05,
"loss": 1.6476,
"step": 3180
},
{
"epoch": 8.716826265389876,
"grad_norm": 2.234564781188965,
"learning_rate": 9.11537635074678e-05,
"loss": 1.6552,
"step": 3190
},
{
"epoch": 8.744186046511627,
"grad_norm": 2.135258197784424,
"learning_rate": 9.106871436497645e-05,
"loss": 1.6676,
"step": 3200
},
{
"epoch": 8.771545827633378,
"grad_norm": 2.0397415161132812,
"learning_rate": 9.098329841361816e-05,
"loss": 1.657,
"step": 3210
},
{
"epoch": 8.79890560875513,
"grad_norm": 2.4961605072021484,
"learning_rate": 9.089751641629308e-05,
"loss": 1.7213,
"step": 3220
},
{
"epoch": 8.82626538987688,
"grad_norm": 2.8516130447387695,
"learning_rate": 9.081136913917072e-05,
"loss": 1.7606,
"step": 3230
},
{
"epoch": 8.853625170998631,
"grad_norm": 2.5965471267700195,
"learning_rate": 9.072485735168313e-05,
"loss": 1.6631,
"step": 3240
},
{
"epoch": 8.880984952120382,
"grad_norm": 2.5616679191589355,
"learning_rate": 9.063798182651804e-05,
"loss": 1.664,
"step": 3250
},
{
"epoch": 8.908344733242133,
"grad_norm": 2.4064130783081055,
"learning_rate": 9.055074333961189e-05,
"loss": 1.7392,
"step": 3260
},
{
"epoch": 8.935704514363884,
"grad_norm": 2.3244357109069824,
"learning_rate": 9.046314267014297e-05,
"loss": 1.7403,
"step": 3270
},
{
"epoch": 8.963064295485637,
"grad_norm": 2.722261905670166,
"learning_rate": 9.037518060052446e-05,
"loss": 1.6781,
"step": 3280
},
{
"epoch": 8.990424076607388,
"grad_norm": 2.5195419788360596,
"learning_rate": 9.028685791639736e-05,
"loss": 1.7001,
"step": 3290
},
{
"epoch": 9.01641586867305,
"grad_norm": 3.1865060329437256,
"learning_rate": 9.019817540662356e-05,
"loss": 1.199,
"step": 3300
},
{
"epoch": 9.043775649794801,
"grad_norm": 2.6850881576538086,
"learning_rate": 9.010913386327879e-05,
"loss": 0.8667,
"step": 3310
},
{
"epoch": 9.071135430916552,
"grad_norm": 2.4270949363708496,
"learning_rate": 9.001973408164548e-05,
"loss": 0.8219,
"step": 3320
},
{
"epoch": 9.098495212038303,
"grad_norm": 2.442011594772339,
"learning_rate": 8.992997686020572e-05,
"loss": 0.8133,
"step": 3330
},
{
"epoch": 9.125854993160054,
"grad_norm": 2.6411468982696533,
"learning_rate": 8.98398630006341e-05,
"loss": 0.8892,
"step": 3340
},
{
"epoch": 9.153214774281805,
"grad_norm": 2.8654303550720215,
"learning_rate": 8.974939330779054e-05,
"loss": 0.8011,
"step": 3350
},
{
"epoch": 9.180574555403556,
"grad_norm": 3.015768527984619,
"learning_rate": 8.965856858971313e-05,
"loss": 0.8513,
"step": 3360
},
{
"epoch": 9.207934336525307,
"grad_norm": 2.670635461807251,
"learning_rate": 8.956738965761094e-05,
"loss": 0.8854,
"step": 3370
},
{
"epoch": 9.235294117647058,
"grad_norm": 2.8876519203186035,
"learning_rate": 8.947585732585663e-05,
"loss": 0.8651,
"step": 3380
},
{
"epoch": 9.26265389876881,
"grad_norm": 2.6166300773620605,
"learning_rate": 8.93839724119794e-05,
"loss": 0.886,
"step": 3390
},
{
"epoch": 9.290013679890562,
"grad_norm": 2.2493016719818115,
"learning_rate": 8.929173573665748e-05,
"loss": 0.9145,
"step": 3400
},
{
"epoch": 9.317373461012313,
"grad_norm": 2.4153900146484375,
"learning_rate": 8.919914812371093e-05,
"loss": 0.9705,
"step": 3410
},
{
"epoch": 9.344733242134064,
"grad_norm": 2.472249746322632,
"learning_rate": 8.910621040009424e-05,
"loss": 0.8995,
"step": 3420
},
{
"epoch": 9.372093023255815,
"grad_norm": 2.34696888923645,
"learning_rate": 8.901292339588891e-05,
"loss": 0.9154,
"step": 3430
},
{
"epoch": 9.399452804377566,
"grad_norm": 2.495896100997925,
"learning_rate": 8.891928794429613e-05,
"loss": 0.9082,
"step": 3440
},
{
"epoch": 9.426812585499317,
"grad_norm": 2.6187784671783447,
"learning_rate": 8.882530488162924e-05,
"loss": 0.9176,
"step": 3450
},
{
"epoch": 9.454172366621068,
"grad_norm": 2.736407518386841,
"learning_rate": 8.873097504730628e-05,
"loss": 0.9351,
"step": 3460
},
{
"epoch": 9.481532147742818,
"grad_norm": 2.8737196922302246,
"learning_rate": 8.863629928384254e-05,
"loss": 0.9054,
"step": 3470
},
{
"epoch": 9.50889192886457,
"grad_norm": 2.50784969329834,
"learning_rate": 8.854127843684303e-05,
"loss": 0.9352,
"step": 3480
},
{
"epoch": 9.53625170998632,
"grad_norm": 2.5840797424316406,
"learning_rate": 8.844591335499486e-05,
"loss": 0.9575,
"step": 3490
},
{
"epoch": 9.563611491108071,
"grad_norm": 2.456085205078125,
"learning_rate": 8.835020489005973e-05,
"loss": 0.9363,
"step": 3500
},
{
"epoch": 9.590971272229822,
"grad_norm": 2.376094341278076,
"learning_rate": 8.825415389686628e-05,
"loss": 0.9441,
"step": 3510
},
{
"epoch": 9.618331053351573,
"grad_norm": 2.7783279418945312,
"learning_rate": 8.815776123330247e-05,
"loss": 0.975,
"step": 3520
},
{
"epoch": 9.645690834473324,
"grad_norm": 2.5872397422790527,
"learning_rate": 8.806102776030795e-05,
"loss": 1.0137,
"step": 3530
},
{
"epoch": 9.673050615595075,
"grad_norm": 2.8060836791992188,
"learning_rate": 8.79639543418663e-05,
"loss": 0.9747,
"step": 3540
},
{
"epoch": 9.700410396716826,
"grad_norm": 2.501214027404785,
"learning_rate": 8.78665418449974e-05,
"loss": 1.0105,
"step": 3550
},
{
"epoch": 9.727770177838577,
"grad_norm": 2.874805450439453,
"learning_rate": 8.776879113974958e-05,
"loss": 1.0462,
"step": 3560
},
{
"epoch": 9.755129958960328,
"grad_norm": 2.649526357650757,
"learning_rate": 8.767070309919194e-05,
"loss": 1.0486,
"step": 3570
},
{
"epoch": 9.78248974008208,
"grad_norm": 2.8548691272735596,
"learning_rate": 8.757227859940655e-05,
"loss": 1.0118,
"step": 3580
},
{
"epoch": 9.80984952120383,
"grad_norm": 3.102064847946167,
"learning_rate": 8.747351851948052e-05,
"loss": 1.0325,
"step": 3590
},
{
"epoch": 9.837209302325581,
"grad_norm": 2.5407538414001465,
"learning_rate": 8.737442374149832e-05,
"loss": 1.0167,
"step": 3600
},
{
"epoch": 9.864569083447332,
"grad_norm": 3.0104925632476807,
"learning_rate": 8.72749951505337e-05,
"loss": 1.013,
"step": 3610
},
{
"epoch": 9.891928864569083,
"grad_norm": 2.5798354148864746,
"learning_rate": 8.717523363464198e-05,
"loss": 1.0123,
"step": 3620
},
{
"epoch": 9.919288645690834,
"grad_norm": 2.271873950958252,
"learning_rate": 8.707514008485196e-05,
"loss": 1.0461,
"step": 3630
},
{
"epoch": 9.946648426812585,
"grad_norm": 2.6398229598999023,
"learning_rate": 8.697471539515811e-05,
"loss": 1.0745,
"step": 3640
},
{
"epoch": 9.974008207934336,
"grad_norm": 2.401278257369995,
"learning_rate": 8.687396046251241e-05,
"loss": 1.0296,
"step": 3650
},
{
"epoch": 10.0,
"grad_norm": 3.144273519515991,
"learning_rate": 8.677287618681653e-05,
"loss": 1.055,
"step": 3660
},
{
"epoch": 10.027359781121751,
"grad_norm": 2.081324577331543,
"learning_rate": 8.667146347091362e-05,
"loss": 0.4022,
"step": 3670
},
{
"epoch": 10.054719562243502,
"grad_norm": 1.9966222047805786,
"learning_rate": 8.656972322058032e-05,
"loss": 0.4213,
"step": 3680
},
{
"epoch": 10.082079343365253,
"grad_norm": 2.4245169162750244,
"learning_rate": 8.646765634451877e-05,
"loss": 0.3918,
"step": 3690
},
{
"epoch": 10.109439124487004,
"grad_norm": 2.3648147583007812,
"learning_rate": 8.636526375434831e-05,
"loss": 0.4305,
"step": 3700
},
{
"epoch": 10.136798905608755,
"grad_norm": 1.9869210720062256,
"learning_rate": 8.626254636459741e-05,
"loss": 0.4146,
"step": 3710
},
{
"epoch": 10.164158686730506,
"grad_norm": 2.2358055114746094,
"learning_rate": 8.615950509269559e-05,
"loss": 0.4144,
"step": 3720
},
{
"epoch": 10.191518467852257,
"grad_norm": 2.2317137718200684,
"learning_rate": 8.605614085896511e-05,
"loss": 0.4567,
"step": 3730
},
{
"epoch": 10.218878248974008,
"grad_norm": 1.986158013343811,
"learning_rate": 8.595245458661278e-05,
"loss": 0.4487,
"step": 3740
},
{
"epoch": 10.246238030095759,
"grad_norm": 2.0380935668945312,
"learning_rate": 8.584844720172176e-05,
"loss": 0.4422,
"step": 3750
},
{
"epoch": 10.27359781121751,
"grad_norm": 2.418177843093872,
"learning_rate": 8.574411963324322e-05,
"loss": 0.4791,
"step": 3760
},
{
"epoch": 10.30095759233926,
"grad_norm": 2.607536792755127,
"learning_rate": 8.563947281298811e-05,
"loss": 0.4536,
"step": 3770
},
{
"epoch": 10.328317373461012,
"grad_norm": 2.239516258239746,
"learning_rate": 8.55345076756188e-05,
"loss": 0.4936,
"step": 3780
},
{
"epoch": 10.355677154582763,
"grad_norm": 2.1065726280212402,
"learning_rate": 8.542922515864071e-05,
"loss": 0.4848,
"step": 3790
},
{
"epoch": 10.383036935704514,
"grad_norm": 2.186018466949463,
"learning_rate": 8.5323626202394e-05,
"loss": 0.4796,
"step": 3800
},
{
"epoch": 10.410396716826265,
"grad_norm": 2.4769375324249268,
"learning_rate": 8.52177117500451e-05,
"loss": 0.4857,
"step": 3810
},
{
"epoch": 10.437756497948016,
"grad_norm": 2.442934989929199,
"learning_rate": 8.511148274757837e-05,
"loss": 0.4784,
"step": 3820
},
{
"epoch": 10.465116279069768,
"grad_norm": 2.5560991764068604,
"learning_rate": 8.500494014378753e-05,
"loss": 0.5105,
"step": 3830
},
{
"epoch": 10.49247606019152,
"grad_norm": 2.310737371444702,
"learning_rate": 8.489808489026734e-05,
"loss": 0.4889,
"step": 3840
},
{
"epoch": 10.51983584131327,
"grad_norm": 2.6425797939300537,
"learning_rate": 8.479091794140494e-05,
"loss": 0.5134,
"step": 3850
},
{
"epoch": 10.547195622435021,
"grad_norm": 2.3795275688171387,
"learning_rate": 8.46834402543715e-05,
"loss": 0.5012,
"step": 3860
},
{
"epoch": 10.574555403556772,
"grad_norm": 2.646310567855835,
"learning_rate": 8.457565278911348e-05,
"loss": 0.5165,
"step": 3870
},
{
"epoch": 10.601915184678523,
"grad_norm": 2.9664266109466553,
"learning_rate": 8.44675565083442e-05,
"loss": 0.5107,
"step": 3880
},
{
"epoch": 10.629274965800274,
"grad_norm": 2.160407543182373,
"learning_rate": 8.435915237753516e-05,
"loss": 0.5206,
"step": 3890
},
{
"epoch": 10.656634746922025,
"grad_norm": 2.6070010662078857,
"learning_rate": 8.425044136490751e-05,
"loss": 0.5406,
"step": 3900
},
{
"epoch": 10.683994528043776,
"grad_norm": 2.4874308109283447,
"learning_rate": 8.414142444142333e-05,
"loss": 0.5311,
"step": 3910
},
{
"epoch": 10.711354309165527,
"grad_norm": 2.687506675720215,
"learning_rate": 8.403210258077688e-05,
"loss": 0.5588,
"step": 3920
},
{
"epoch": 10.738714090287278,
"grad_norm": 2.764559268951416,
"learning_rate": 8.39224767593861e-05,
"loss": 0.5494,
"step": 3930
},
{
"epoch": 10.766073871409029,
"grad_norm": 2.135972261428833,
"learning_rate": 8.381254795638372e-05,
"loss": 0.546,
"step": 3940
},
{
"epoch": 10.79343365253078,
"grad_norm": 2.353956699371338,
"learning_rate": 8.370231715360864e-05,
"loss": 0.5378,
"step": 3950
},
{
"epoch": 10.82079343365253,
"grad_norm": 2.407003402709961,
"learning_rate": 8.3591785335597e-05,
"loss": 0.5636,
"step": 3960
},
{
"epoch": 10.848153214774282,
"grad_norm": 2.409064531326294,
"learning_rate": 8.34809534895736e-05,
"loss": 0.5452,
"step": 3970
},
{
"epoch": 10.875512995896033,
"grad_norm": 2.189385175704956,
"learning_rate": 8.336982260544286e-05,
"loss": 0.5651,
"step": 3980
},
{
"epoch": 10.902872777017784,
"grad_norm": 2.276256799697876,
"learning_rate": 8.325839367578016e-05,
"loss": 0.5891,
"step": 3990
},
{
"epoch": 10.930232558139535,
"grad_norm": 2.5703511238098145,
"learning_rate": 8.314666769582289e-05,
"loss": 0.5699,
"step": 4000
},
{
"epoch": 10.957592339261286,
"grad_norm": 2.272239923477173,
"learning_rate": 8.303464566346152e-05,
"loss": 0.5777,
"step": 4010
},
{
"epoch": 10.984952120383037,
"grad_norm": 2.359806537628174,
"learning_rate": 8.292232857923079e-05,
"loss": 0.5891,
"step": 4020
},
{
"epoch": 11.0109439124487,
"grad_norm": 1.5687215328216553,
"learning_rate": 8.280971744630069e-05,
"loss": 0.385,
"step": 4030
},
{
"epoch": 11.038303693570452,
"grad_norm": 1.485263466835022,
"learning_rate": 8.269681327046757e-05,
"loss": 0.187,
"step": 4040
},
{
"epoch": 11.065663474692203,
"grad_norm": 1.5766488313674927,
"learning_rate": 8.258361706014509e-05,
"loss": 0.1894,
"step": 4050
},
{
"epoch": 11.093023255813954,
"grad_norm": 1.679360032081604,
"learning_rate": 8.247012982635522e-05,
"loss": 0.1962,
"step": 4060
},
{
"epoch": 11.120383036935705,
"grad_norm": 1.7835336923599243,
"learning_rate": 8.235635258271932e-05,
"loss": 0.1997,
"step": 4070
},
{
"epoch": 11.147742818057456,
"grad_norm": 1.9064172506332397,
"learning_rate": 8.224228634544889e-05,
"loss": 0.2199,
"step": 4080
},
{
"epoch": 11.175102599179207,
"grad_norm": 1.7779500484466553,
"learning_rate": 8.212793213333664e-05,
"loss": 0.2143,
"step": 4090
},
{
"epoch": 11.202462380300958,
"grad_norm": 1.6907753944396973,
"learning_rate": 8.201329096774742e-05,
"loss": 0.2363,
"step": 4100
},
{
"epoch": 11.229822161422709,
"grad_norm": 1.8031010627746582,
"learning_rate": 8.189836387260892e-05,
"loss": 0.2326,
"step": 4110
},
{
"epoch": 11.25718194254446,
"grad_norm": 2.0248312950134277,
"learning_rate": 8.17831518744027e-05,
"loss": 0.2413,
"step": 4120
},
{
"epoch": 11.28454172366621,
"grad_norm": 1.9498226642608643,
"learning_rate": 8.166765600215498e-05,
"loss": 0.2438,
"step": 4130
},
{
"epoch": 11.311901504787961,
"grad_norm": 1.8161996603012085,
"learning_rate": 8.155187728742736e-05,
"loss": 0.2432,
"step": 4140
},
{
"epoch": 11.339261285909712,
"grad_norm": 1.8568273782730103,
"learning_rate": 8.143581676430771e-05,
"loss": 0.255,
"step": 4150
},
{
"epoch": 11.366621067031463,
"grad_norm": 2.202374219894409,
"learning_rate": 8.131947546940091e-05,
"loss": 0.2546,
"step": 4160
},
{
"epoch": 11.393980848153214,
"grad_norm": 2.066796064376831,
"learning_rate": 8.120285444181952e-05,
"loss": 0.2551,
"step": 4170
},
{
"epoch": 11.421340629274965,
"grad_norm": 2.0504369735717773,
"learning_rate": 8.108595472317465e-05,
"loss": 0.2523,
"step": 4180
},
{
"epoch": 11.448700410396716,
"grad_norm": 2.011749029159546,
"learning_rate": 8.096877735756647e-05,
"loss": 0.2657,
"step": 4190
},
{
"epoch": 11.476060191518467,
"grad_norm": 1.6771564483642578,
"learning_rate": 8.085132339157504e-05,
"loss": 0.2584,
"step": 4200
},
{
"epoch": 11.503419972640218,
"grad_norm": 2.031188488006592,
"learning_rate": 8.073359387425088e-05,
"loss": 0.2738,
"step": 4210
},
{
"epoch": 11.53077975376197,
"grad_norm": 1.7527649402618408,
"learning_rate": 8.06155898571056e-05,
"loss": 0.272,
"step": 4220
},
{
"epoch": 11.55813953488372,
"grad_norm": 1.8100972175598145,
"learning_rate": 8.049731239410258e-05,
"loss": 0.284,
"step": 4230
},
{
"epoch": 11.585499316005471,
"grad_norm": 1.7095541954040527,
"learning_rate": 8.037876254164747e-05,
"loss": 0.2779,
"step": 4240
},
{
"epoch": 11.612859097127224,
"grad_norm": 2.1026771068573,
"learning_rate": 8.025994135857877e-05,
"loss": 0.2753,
"step": 4250
},
{
"epoch": 11.640218878248975,
"grad_norm": 2.4466755390167236,
"learning_rate": 8.014084990615846e-05,
"loss": 0.2831,
"step": 4260
},
{
"epoch": 11.667578659370726,
"grad_norm": 2.0461161136627197,
"learning_rate": 8.00214892480624e-05,
"loss": 0.2852,
"step": 4270
},
{
"epoch": 11.694938440492477,
"grad_norm": 2.0975372791290283,
"learning_rate": 7.990186045037091e-05,
"loss": 0.2884,
"step": 4280
},
{
"epoch": 11.722298221614228,
"grad_norm": 1.9395639896392822,
"learning_rate": 7.97819645815592e-05,
"loss": 0.2901,
"step": 4290
},
{
"epoch": 11.749658002735979,
"grad_norm": 1.8611444234848022,
"learning_rate": 7.966180271248784e-05,
"loss": 0.2981,
"step": 4300
},
{
"epoch": 11.77701778385773,
"grad_norm": 2.0806777477264404,
"learning_rate": 7.954137591639329e-05,
"loss": 0.2975,
"step": 4310
},
{
"epoch": 11.80437756497948,
"grad_norm": 2.0935049057006836,
"learning_rate": 7.942068526887811e-05,
"loss": 0.305,
"step": 4320
},
{
"epoch": 11.831737346101232,
"grad_norm": 2.0660412311553955,
"learning_rate": 7.929973184790154e-05,
"loss": 0.3079,
"step": 4330
},
{
"epoch": 11.859097127222983,
"grad_norm": 1.9550343751907349,
"learning_rate": 7.917851673376982e-05,
"loss": 0.3067,
"step": 4340
},
{
"epoch": 11.886456908344734,
"grad_norm": 1.9984105825424194,
"learning_rate": 7.905704100912646e-05,
"loss": 0.3072,
"step": 4350
},
{
"epoch": 11.913816689466485,
"grad_norm": 2.224447011947632,
"learning_rate": 7.89353057589427e-05,
"loss": 0.3092,
"step": 4360
},
{
"epoch": 11.941176470588236,
"grad_norm": 1.90882408618927,
"learning_rate": 7.881331207050773e-05,
"loss": 0.3244,
"step": 4370
},
{
"epoch": 11.968536251709986,
"grad_norm": 2.045947551727295,
"learning_rate": 7.8691061033419e-05,
"loss": 0.3224,
"step": 4380
},
{
"epoch": 11.995896032831737,
"grad_norm": 2.139143466949463,
"learning_rate": 7.85685537395725e-05,
"loss": 0.3321,
"step": 4390
},
{
"epoch": 12.021887824897401,
"grad_norm": 1.2766697406768799,
"learning_rate": 7.844579128315299e-05,
"loss": 0.1295,
"step": 4400
},
{
"epoch": 12.049247606019152,
"grad_norm": 1.2042920589447021,
"learning_rate": 7.832277476062424e-05,
"loss": 0.1073,
"step": 4410
},
{
"epoch": 12.076607387140903,
"grad_norm": 1.2516529560089111,
"learning_rate": 7.819950527071927e-05,
"loss": 0.12,
"step": 4420
},
{
"epoch": 12.103967168262654,
"grad_norm": 1.3125766515731812,
"learning_rate": 7.807598391443041e-05,
"loss": 0.129,
"step": 4430
},
{
"epoch": 12.131326949384405,
"grad_norm": 1.2805269956588745,
"learning_rate": 7.795221179499967e-05,
"loss": 0.1288,
"step": 4440
},
{
"epoch": 12.158686730506156,
"grad_norm": 1.3272207975387573,
"learning_rate": 7.782819001790869e-05,
"loss": 0.1412,
"step": 4450
},
{
"epoch": 12.186046511627907,
"grad_norm": 1.330671787261963,
"learning_rate": 7.770391969086899e-05,
"loss": 0.1396,
"step": 4460
},
{
"epoch": 12.213406292749658,
"grad_norm": 1.4824010133743286,
"learning_rate": 7.7579401923812e-05,
"loss": 0.143,
"step": 4470
},
{
"epoch": 12.24076607387141,
"grad_norm": 1.4328900575637817,
"learning_rate": 7.745463782887928e-05,
"loss": 0.1562,
"step": 4480
},
{
"epoch": 12.26812585499316,
"grad_norm": 1.4229873418807983,
"learning_rate": 7.732962852041232e-05,
"loss": 0.1529,
"step": 4490
},
{
"epoch": 12.295485636114911,
"grad_norm": 1.422752857208252,
"learning_rate": 7.72043751149429e-05,
"loss": 0.1502,
"step": 4500
},
{
"epoch": 12.322845417236662,
"grad_norm": 1.5239604711532593,
"learning_rate": 7.707887873118291e-05,
"loss": 0.1624,
"step": 4510
},
{
"epoch": 12.350205198358413,
"grad_norm": 1.5180604457855225,
"learning_rate": 7.695314049001447e-05,
"loss": 0.1621,
"step": 4520
},
{
"epoch": 12.377564979480164,
"grad_norm": 1.5136972665786743,
"learning_rate": 7.682716151447982e-05,
"loss": 0.1668,
"step": 4530
},
{
"epoch": 12.404924760601915,
"grad_norm": 1.422076940536499,
"learning_rate": 7.670094292977134e-05,
"loss": 0.1661,
"step": 4540
},
{
"epoch": 12.432284541723666,
"grad_norm": 1.5080851316452026,
"learning_rate": 7.657448586322158e-05,
"loss": 0.1745,
"step": 4550
},
{
"epoch": 12.459644322845417,
"grad_norm": 1.5718903541564941,
"learning_rate": 7.644779144429298e-05,
"loss": 0.1738,
"step": 4560
},
{
"epoch": 12.487004103967168,
"grad_norm": 1.5404165983200073,
"learning_rate": 7.632086080456803e-05,
"loss": 0.1725,
"step": 4570
},
{
"epoch": 12.514363885088919,
"grad_norm": 1.5033949613571167,
"learning_rate": 7.619369507773898e-05,
"loss": 0.1732,
"step": 4580
},
{
"epoch": 12.54172366621067,
"grad_norm": 1.5548062324523926,
"learning_rate": 7.606629539959782e-05,
"loss": 0.1754,
"step": 4590
},
{
"epoch": 12.569083447332421,
"grad_norm": 1.5595543384552002,
"learning_rate": 7.593866290802608e-05,
"loss": 0.1799,
"step": 4600
},
{
"epoch": 12.596443228454172,
"grad_norm": 1.6693882942199707,
"learning_rate": 7.581079874298468e-05,
"loss": 0.1807,
"step": 4610
},
{
"epoch": 12.623803009575923,
"grad_norm": 1.6664587259292603,
"learning_rate": 7.568270404650375e-05,
"loss": 0.1847,
"step": 4620
},
{
"epoch": 12.651162790697674,
"grad_norm": 1.5869115591049194,
"learning_rate": 7.555437996267247e-05,
"loss": 0.1848,
"step": 4630
},
{
"epoch": 12.678522571819425,
"grad_norm": 1.747053861618042,
"learning_rate": 7.542582763762875e-05,
"loss": 0.1899,
"step": 4640
},
{
"epoch": 12.705882352941176,
"grad_norm": 1.5815576314926147,
"learning_rate": 7.529704821954909e-05,
"loss": 0.1874,
"step": 4650
},
{
"epoch": 12.733242134062927,
"grad_norm": 1.5363188982009888,
"learning_rate": 7.51680428586383e-05,
"loss": 0.1848,
"step": 4660
},
{
"epoch": 12.76060191518468,
"grad_norm": 1.4765170812606812,
"learning_rate": 7.503881270711924e-05,
"loss": 0.1898,
"step": 4670
},
{
"epoch": 12.787961696306429,
"grad_norm": 1.5592812299728394,
"learning_rate": 7.490935891922241e-05,
"loss": 0.1892,
"step": 4680
},
{
"epoch": 12.815321477428181,
"grad_norm": 1.7455358505249023,
"learning_rate": 7.477968265117586e-05,
"loss": 0.1958,
"step": 4690
},
{
"epoch": 12.842681258549932,
"grad_norm": 1.5827257633209229,
"learning_rate": 7.464978506119466e-05,
"loss": 0.1956,
"step": 4700
},
{
"epoch": 12.870041039671683,
"grad_norm": 1.8754746913909912,
"learning_rate": 7.451966730947068e-05,
"loss": 0.2007,
"step": 4710
},
{
"epoch": 12.897400820793434,
"grad_norm": 1.6817800998687744,
"learning_rate": 7.438933055816216e-05,
"loss": 0.2003,
"step": 4720
},
{
"epoch": 12.924760601915185,
"grad_norm": 1.7609623670578003,
"learning_rate": 7.425877597138336e-05,
"loss": 0.1942,
"step": 4730
},
{
"epoch": 12.952120383036936,
"grad_norm": 1.5804462432861328,
"learning_rate": 7.412800471519414e-05,
"loss": 0.1945,
"step": 4740
},
{
"epoch": 12.979480164158687,
"grad_norm": 1.572085976600647,
"learning_rate": 7.399701795758962e-05,
"loss": 0.193,
"step": 4750
},
{
"epoch": 13.00547195622435,
"grad_norm": 0.9358009099960327,
"learning_rate": 7.38658168684896e-05,
"loss": 0.1632,
"step": 4760
},
{
"epoch": 13.0328317373461,
"grad_norm": 0.9048413038253784,
"learning_rate": 7.373440261972828e-05,
"loss": 0.0715,
"step": 4770
},
{
"epoch": 13.060191518467851,
"grad_norm": 1.0731923580169678,
"learning_rate": 7.360277638504369e-05,
"loss": 0.0781,
"step": 4780
},
{
"epoch": 13.087551299589602,
"grad_norm": 1.0774123668670654,
"learning_rate": 7.347093934006721e-05,
"loss": 0.0857,
"step": 4790
},
{
"epoch": 13.114911080711355,
"grad_norm": 1.0120024681091309,
"learning_rate": 7.333889266231314e-05,
"loss": 0.0925,
"step": 4800
},
{
"epoch": 13.142270861833106,
"grad_norm": 1.0840078592300415,
"learning_rate": 7.320663753116804e-05,
"loss": 0.0967,
"step": 4810
},
{
"epoch": 13.169630642954857,
"grad_norm": 1.089255452156067,
"learning_rate": 7.307417512788039e-05,
"loss": 0.1003,
"step": 4820
},
{
"epoch": 13.196990424076608,
"grad_norm": 1.1200820207595825,
"learning_rate": 7.294150663554991e-05,
"loss": 0.1062,
"step": 4830
},
{
"epoch": 13.224350205198359,
"grad_norm": 1.0958836078643799,
"learning_rate": 7.2808633239117e-05,
"loss": 0.106,
"step": 4840
},
{
"epoch": 13.25170998632011,
"grad_norm": 1.1710079908370972,
"learning_rate": 7.26755561253522e-05,
"loss": 0.1094,
"step": 4850
},
{
"epoch": 13.279069767441861,
"grad_norm": 1.1231863498687744,
"learning_rate": 7.254227648284551e-05,
"loss": 0.111,
"step": 4860
},
{
"epoch": 13.306429548563612,
"grad_norm": 1.185207486152649,
"learning_rate": 7.240879550199596e-05,
"loss": 0.1128,
"step": 4870
},
{
"epoch": 13.333789329685363,
"grad_norm": 1.10126793384552,
"learning_rate": 7.227511437500074e-05,
"loss": 0.1175,
"step": 4880
},
{
"epoch": 13.361149110807114,
"grad_norm": 1.154030203819275,
"learning_rate": 7.214123429584469e-05,
"loss": 0.1178,
"step": 4890
},
{
"epoch": 13.388508891928865,
"grad_norm": 1.2188189029693604,
"learning_rate": 7.200715646028963e-05,
"loss": 0.1217,
"step": 4900
},
{
"epoch": 13.415868673050616,
"grad_norm": 1.2287869453430176,
"learning_rate": 7.187288206586366e-05,
"loss": 0.1205,
"step": 4910
},
{
"epoch": 13.443228454172367,
"grad_norm": 1.2189929485321045,
"learning_rate": 7.173841231185043e-05,
"loss": 0.1186,
"step": 4920
},
{
"epoch": 13.470588235294118,
"grad_norm": 1.209676742553711,
"learning_rate": 7.160374839927847e-05,
"loss": 0.1227,
"step": 4930
},
{
"epoch": 13.497948016415869,
"grad_norm": 1.2361178398132324,
"learning_rate": 7.146889153091051e-05,
"loss": 0.1255,
"step": 4940
},
{
"epoch": 13.52530779753762,
"grad_norm": 1.2039564847946167,
"learning_rate": 7.133384291123261e-05,
"loss": 0.126,
"step": 4950
},
{
"epoch": 13.55266757865937,
"grad_norm": 1.1753249168395996,
"learning_rate": 7.11986037464435e-05,
"loss": 0.1264,
"step": 4960
},
{
"epoch": 13.580027359781122,
"grad_norm": 1.3375935554504395,
"learning_rate": 7.106317524444377e-05,
"loss": 0.1292,
"step": 4970
},
{
"epoch": 13.607387140902873,
"grad_norm": 1.2629966735839844,
"learning_rate": 7.092755861482511e-05,
"loss": 0.1263,
"step": 4980
},
{
"epoch": 13.634746922024624,
"grad_norm": 1.3769235610961914,
"learning_rate": 7.079175506885951e-05,
"loss": 0.1316,
"step": 4990
},
{
"epoch": 13.662106703146375,
"grad_norm": 1.264193058013916,
"learning_rate": 7.065576581948837e-05,
"loss": 0.1347,
"step": 5000
},
{
"epoch": 13.689466484268126,
"grad_norm": 1.2223986387252808,
"learning_rate": 7.051959208131175e-05,
"loss": 0.1325,
"step": 5010
},
{
"epoch": 13.716826265389876,
"grad_norm": 1.2866407632827759,
"learning_rate": 7.038323507057746e-05,
"loss": 0.1337,
"step": 5020
},
{
"epoch": 13.744186046511627,
"grad_norm": 1.3503751754760742,
"learning_rate": 7.024669600517024e-05,
"loss": 0.1353,
"step": 5030
},
{
"epoch": 13.771545827633378,
"grad_norm": 1.2799268960952759,
"learning_rate": 7.01099761046009e-05,
"loss": 0.133,
"step": 5040
},
{
"epoch": 13.79890560875513,
"grad_norm": 1.282147765159607,
"learning_rate": 6.997307658999535e-05,
"loss": 0.134,
"step": 5050
},
{
"epoch": 13.82626538987688,
"grad_norm": 1.2931870222091675,
"learning_rate": 6.983599868408373e-05,
"loss": 0.1362,
"step": 5060
},
{
"epoch": 13.853625170998631,
"grad_norm": 1.2335541248321533,
"learning_rate": 6.969874361118956e-05,
"loss": 0.1355,
"step": 5070
},
{
"epoch": 13.880984952120382,
"grad_norm": 1.3108147382736206,
"learning_rate": 6.956131259721868e-05,
"loss": 0.1369,
"step": 5080
},
{
"epoch": 13.908344733242133,
"grad_norm": 1.309849739074707,
"learning_rate": 6.942370686964839e-05,
"loss": 0.1385,
"step": 5090
},
{
"epoch": 13.935704514363884,
"grad_norm": 1.2410293817520142,
"learning_rate": 6.928592765751645e-05,
"loss": 0.1365,
"step": 5100
},
{
"epoch": 13.963064295485637,
"grad_norm": 1.379789113998413,
"learning_rate": 6.914797619141013e-05,
"loss": 0.1362,
"step": 5110
},
{
"epoch": 13.990424076607388,
"grad_norm": 1.3390743732452393,
"learning_rate": 6.900985370345522e-05,
"loss": 0.1393,
"step": 5120
},
{
"epoch": 14.01641586867305,
"grad_norm": 0.8094517588615417,
"learning_rate": 6.887156142730493e-05,
"loss": 0.0812,
"step": 5130
},
{
"epoch": 14.043775649794801,
"grad_norm": 0.7880135178565979,
"learning_rate": 6.873310059812902e-05,
"loss": 0.0558,
"step": 5140
},
{
"epoch": 14.071135430916552,
"grad_norm": 0.7849863171577454,
"learning_rate": 6.859447245260268e-05,
"loss": 0.0567,
"step": 5150
},
{
"epoch": 14.098495212038303,
"grad_norm": 0.8471216559410095,
"learning_rate": 6.845567822889547e-05,
"loss": 0.0589,
"step": 5160
},
{
"epoch": 14.125854993160054,
"grad_norm": 0.8714724779129028,
"learning_rate": 6.831671916666036e-05,
"loss": 0.0599,
"step": 5170
},
{
"epoch": 14.153214774281805,
"grad_norm": 0.9662114977836609,
"learning_rate": 6.81775965070225e-05,
"loss": 0.0597,
"step": 5180
},
{
"epoch": 14.180574555403556,
"grad_norm": 0.8515005707740784,
"learning_rate": 6.803831149256827e-05,
"loss": 0.0621,
"step": 5190
},
{
"epoch": 14.207934336525307,
"grad_norm": 0.8311191201210022,
"learning_rate": 6.789886536733418e-05,
"loss": 0.0628,
"step": 5200
},
{
"epoch": 14.235294117647058,
"grad_norm": 0.8105677962303162,
"learning_rate": 6.775925937679564e-05,
"loss": 0.065,
"step": 5210
},
{
"epoch": 14.26265389876881,
"grad_norm": 0.8579919934272766,
"learning_rate": 6.761949476785596e-05,
"loss": 0.0656,
"step": 5220
},
{
"epoch": 14.290013679890562,
"grad_norm": 0.8626765012741089,
"learning_rate": 6.747957278883513e-05,
"loss": 0.0681,
"step": 5230
},
{
"epoch": 14.317373461012313,
"grad_norm": 0.8655377626419067,
"learning_rate": 6.733949468945874e-05,
"loss": 0.0679,
"step": 5240
},
{
"epoch": 14.344733242134064,
"grad_norm": 0.8607134819030762,
"learning_rate": 6.719926172084677e-05,
"loss": 0.0703,
"step": 5250
},
{
"epoch": 14.372093023255815,
"grad_norm": 0.934761643409729,
"learning_rate": 6.705887513550239e-05,
"loss": 0.0714,
"step": 5260
},
{
"epoch": 14.399452804377566,
"grad_norm": 0.8964748978614807,
"learning_rate": 6.691833618730088e-05,
"loss": 0.074,
"step": 5270
},
{
"epoch": 14.426812585499317,
"grad_norm": 0.9311953186988831,
"learning_rate": 6.677764613147832e-05,
"loss": 0.0769,
"step": 5280
},
{
"epoch": 14.454172366621068,
"grad_norm": 0.9542168974876404,
"learning_rate": 6.663680622462041e-05,
"loss": 0.078,
"step": 5290
},
{
"epoch": 14.481532147742818,
"grad_norm": 0.954710841178894,
"learning_rate": 6.64958177246513e-05,
"loss": 0.0797,
"step": 5300
},
{
"epoch": 14.50889192886457,
"grad_norm": 0.9762632846832275,
"learning_rate": 6.635468189082229e-05,
"loss": 0.0811,
"step": 5310
},
{
"epoch": 14.53625170998632,
"grad_norm": 0.9707111120223999,
"learning_rate": 6.621339998370059e-05,
"loss": 0.0838,
"step": 5320
},
{
"epoch": 14.563611491108071,
"grad_norm": 1.0272313356399536,
"learning_rate": 6.607197326515808e-05,
"loss": 0.0844,
"step": 5330
},
{
"epoch": 14.590971272229822,
"grad_norm": 1.0671648979187012,
"learning_rate": 6.593040299836008e-05,
"loss": 0.0864,
"step": 5340
},
{
"epoch": 14.618331053351573,
"grad_norm": 1.0127677917480469,
"learning_rate": 6.578869044775397e-05,
"loss": 0.0874,
"step": 5350
},
{
"epoch": 14.645690834473324,
"grad_norm": 1.1917623281478882,
"learning_rate": 6.5646836879058e-05,
"loss": 0.0876,
"step": 5360
},
{
"epoch": 14.673050615595075,
"grad_norm": 1.0231889486312866,
"learning_rate": 6.550484355924987e-05,
"loss": 0.0889,
"step": 5370
},
{
"epoch": 14.700410396716826,
"grad_norm": 1.0932745933532715,
"learning_rate": 6.536271175655551e-05,
"loss": 0.0913,
"step": 5380
},
{
"epoch": 14.727770177838577,
"grad_norm": 1.0013642311096191,
"learning_rate": 6.522044274043777e-05,
"loss": 0.091,
"step": 5390
},
{
"epoch": 14.755129958960328,
"grad_norm": 1.049109935760498,
"learning_rate": 6.507803778158498e-05,
"loss": 0.094,
"step": 5400
},
{
"epoch": 14.78248974008208,
"grad_norm": 1.0501512289047241,
"learning_rate": 6.493549815189965e-05,
"loss": 0.0938,
"step": 5410
},
{
"epoch": 14.80984952120383,
"grad_norm": 1.0627617835998535,
"learning_rate": 6.479282512448715e-05,
"loss": 0.0943,
"step": 5420
},
{
"epoch": 14.837209302325581,
"grad_norm": 1.1133873462677002,
"learning_rate": 6.465001997364428e-05,
"loss": 0.0959,
"step": 5430
},
{
"epoch": 14.864569083447332,
"grad_norm": 1.1022565364837646,
"learning_rate": 6.450708397484792e-05,
"loss": 0.0957,
"step": 5440
},
{
"epoch": 14.891928864569083,
"grad_norm": 1.09109628200531,
"learning_rate": 6.436401840474363e-05,
"loss": 0.0959,
"step": 5450
},
{
"epoch": 14.919288645690834,
"grad_norm": 1.0429644584655762,
"learning_rate": 6.422082454113421e-05,
"loss": 0.0981,
"step": 5460
},
{
"epoch": 14.946648426812585,
"grad_norm": 1.0722014904022217,
"learning_rate": 6.407750366296842e-05,
"loss": 0.0968,
"step": 5470
},
{
"epoch": 14.974008207934336,
"grad_norm": 1.1190056800842285,
"learning_rate": 6.393405705032936e-05,
"loss": 0.0983,
"step": 5480
},
{
"epoch": 15.0,
"grad_norm": 1.1476560831069946,
"learning_rate": 6.37904859844232e-05,
"loss": 0.0934,
"step": 5490
},
{
"epoch": 15.027359781121751,
"grad_norm": 0.66266268491745,
"learning_rate": 6.364679174756766e-05,
"loss": 0.0432,
"step": 5500
},
{
"epoch": 15.054719562243502,
"grad_norm": 0.7139880657196045,
"learning_rate": 6.350297562318055e-05,
"loss": 0.044,
"step": 5510
},
{
"epoch": 15.082079343365253,
"grad_norm": 0.6576858758926392,
"learning_rate": 6.335903889576835e-05,
"loss": 0.0428,
"step": 5520
},
{
"epoch": 15.109439124487004,
"grad_norm": 0.6100230813026428,
"learning_rate": 6.321498285091472e-05,
"loss": 0.0408,
"step": 5530
},
{
"epoch": 15.136798905608755,
"grad_norm": 0.6284418702125549,
"learning_rate": 6.307080877526902e-05,
"loss": 0.0392,
"step": 5540
},
{
"epoch": 15.164158686730506,
"grad_norm": 0.6035795211791992,
"learning_rate": 6.292651795653479e-05,
"loss": 0.0381,
"step": 5550
},
{
"epoch": 15.191518467852257,
"grad_norm": 0.5752121210098267,
"learning_rate": 6.278211168345827e-05,
"loss": 0.0378,
"step": 5560
},
{
"epoch": 15.218878248974008,
"grad_norm": 0.6118842363357544,
"learning_rate": 6.263759124581694e-05,
"loss": 0.0377,
"step": 5570
},
{
"epoch": 15.246238030095759,
"grad_norm": 0.5554078221321106,
"learning_rate": 6.249295793440786e-05,
"loss": 0.0358,
"step": 5580
},
{
"epoch": 15.27359781121751,
"grad_norm": 0.6214984655380249,
"learning_rate": 6.234821304103633e-05,
"loss": 0.0366,
"step": 5590
},
{
"epoch": 15.30095759233926,
"grad_norm": 0.6126103401184082,
"learning_rate": 6.220335785850421e-05,
"loss": 0.0369,
"step": 5600
},
{
"epoch": 15.328317373461012,
"grad_norm": 0.6230517625808716,
"learning_rate": 6.205839368059838e-05,
"loss": 0.0362,
"step": 5610
},
{
"epoch": 15.355677154582763,
"grad_norm": 0.6359356045722961,
"learning_rate": 6.19133218020793e-05,
"loss": 0.0371,
"step": 5620
},
{
"epoch": 15.383036935704514,
"grad_norm": 0.6141124963760376,
"learning_rate": 6.17681435186693e-05,
"loss": 0.0372,
"step": 5630
},
{
"epoch": 15.410396716826265,
"grad_norm": 0.6061652302742004,
"learning_rate": 6.162286012704113e-05,
"loss": 0.0368,
"step": 5640
},
{
"epoch": 15.437756497948016,
"grad_norm": 0.5912964940071106,
"learning_rate": 6.147747292480628e-05,
"loss": 0.0383,
"step": 5650
},
{
"epoch": 15.465116279069768,
"grad_norm": 0.5969256162643433,
"learning_rate": 6.133198321050342e-05,
"loss": 0.0369,
"step": 5660
},
{
"epoch": 15.49247606019152,
"grad_norm": 0.5808072090148926,
"learning_rate": 6.11863922835869e-05,
"loss": 0.038,
"step": 5670
},
{
"epoch": 15.51983584131327,
"grad_norm": 0.6238718628883362,
"learning_rate": 6.104070144441497e-05,
"loss": 0.0379,
"step": 5680
},
{
"epoch": 15.547195622435021,
"grad_norm": 0.6226553320884705,
"learning_rate": 6.089491199423828e-05,
"loss": 0.0377,
"step": 5690
},
{
"epoch": 15.574555403556772,
"grad_norm": 0.632961630821228,
"learning_rate": 6.0749025235188285e-05,
"loss": 0.0367,
"step": 5700
},
{
"epoch": 15.601915184678523,
"grad_norm": 0.6193811893463135,
"learning_rate": 6.060304247026546e-05,
"loss": 0.038,
"step": 5710
},
{
"epoch": 15.629274965800274,
"grad_norm": 0.5914502739906311,
"learning_rate": 6.045696500332789e-05,
"loss": 0.0374,
"step": 5720
},
{
"epoch": 15.656634746922025,
"grad_norm": 0.6544764041900635,
"learning_rate": 6.0310794139079383e-05,
"loss": 0.0383,
"step": 5730
},
{
"epoch": 15.683994528043776,
"grad_norm": 0.6221988797187805,
"learning_rate": 6.016453118305802e-05,
"loss": 0.0386,
"step": 5740
},
{
"epoch": 15.711354309165527,
"grad_norm": 0.5998134613037109,
"learning_rate": 6.0018177441624335e-05,
"loss": 0.0392,
"step": 5750
},
{
"epoch": 15.738714090287278,
"grad_norm": 0.656602144241333,
"learning_rate": 5.987173422194978e-05,
"loss": 0.0389,
"step": 5760
},
{
"epoch": 15.766073871409029,
"grad_norm": 0.6018205285072327,
"learning_rate": 5.972520283200498e-05,
"loss": 0.0393,
"step": 5770
},
{
"epoch": 15.79343365253078,
"grad_norm": 0.6513863205909729,
"learning_rate": 5.9578584580548036e-05,
"loss": 0.0395,
"step": 5780
},
{
"epoch": 15.82079343365253,
"grad_norm": 0.6165125370025635,
"learning_rate": 5.943188077711286e-05,
"loss": 0.039,
"step": 5790
},
{
"epoch": 15.848153214774282,
"grad_norm": 0.6451523303985596,
"learning_rate": 5.92850927319975e-05,
"loss": 0.0399,
"step": 5800
},
{
"epoch": 15.875512995896033,
"grad_norm": 0.6117334961891174,
"learning_rate": 5.913822175625242e-05,
"loss": 0.0398,
"step": 5810
},
{
"epoch": 15.902872777017784,
"grad_norm": 0.6188967227935791,
"learning_rate": 5.899126916166876e-05,
"loss": 0.0406,
"step": 5820
},
{
"epoch": 15.930232558139535,
"grad_norm": 0.6454140543937683,
"learning_rate": 5.884423626076667e-05,
"loss": 0.0407,
"step": 5830
},
{
"epoch": 15.957592339261286,
"grad_norm": 0.6732134222984314,
"learning_rate": 5.869712436678354e-05,
"loss": 0.0408,
"step": 5840
},
{
"epoch": 15.984952120383037,
"grad_norm": 0.6668757200241089,
"learning_rate": 5.854993479366231e-05,
"loss": 0.042,
"step": 5850
},
{
"epoch": 16.0109439124487,
"grad_norm": 0.37282848358154297,
"learning_rate": 5.840266885603972e-05,
"loss": 0.03,
"step": 5860
},
{
"epoch": 16.03830369357045,
"grad_norm": 0.37796393036842346,
"learning_rate": 5.825532786923455e-05,
"loss": 0.0182,
"step": 5870
},
{
"epoch": 16.0656634746922,
"grad_norm": 0.3875877261161804,
"learning_rate": 5.810791314923592e-05,
"loss": 0.0182,
"step": 5880
},
{
"epoch": 16.093023255813954,
"grad_norm": 0.3626669645309448,
"learning_rate": 5.796042601269146e-05,
"loss": 0.0167,
"step": 5890
},
{
"epoch": 16.120383036935703,
"grad_norm": 0.3603843152523041,
"learning_rate": 5.781286777689563e-05,
"loss": 0.0152,
"step": 5900
},
{
"epoch": 16.147742818057456,
"grad_norm": 0.3385431468486786,
"learning_rate": 5.766523975977791e-05,
"loss": 0.0154,
"step": 5910
},
{
"epoch": 16.175102599179205,
"grad_norm": 0.3303395211696625,
"learning_rate": 5.7517543279891037e-05,
"loss": 0.0153,
"step": 5920
},
{
"epoch": 16.202462380300958,
"grad_norm": 0.35541489720344543,
"learning_rate": 5.736977965639923e-05,
"loss": 0.0148,
"step": 5930
},
{
"epoch": 16.22982216142271,
"grad_norm": 0.3332382142543793,
"learning_rate": 5.722195020906639e-05,
"loss": 0.0154,
"step": 5940
},
{
"epoch": 16.25718194254446,
"grad_norm": 0.3340345621109009,
"learning_rate": 5.707405625824435e-05,
"loss": 0.0151,
"step": 5950
},
{
"epoch": 16.284541723666212,
"grad_norm": 0.34100693464279175,
"learning_rate": 5.692609912486107e-05,
"loss": 0.0149,
"step": 5960
},
{
"epoch": 16.31190150478796,
"grad_norm": 0.3536655604839325,
"learning_rate": 5.677808013040881e-05,
"loss": 0.0145,
"step": 5970
},
{
"epoch": 16.339261285909714,
"grad_norm": 0.3327285945415497,
"learning_rate": 5.663000059693233e-05,
"loss": 0.0145,
"step": 5980
},
{
"epoch": 16.366621067031463,
"grad_norm": 0.3213931620121002,
"learning_rate": 5.648186184701717e-05,
"loss": 0.0141,
"step": 5990
},
{
"epoch": 16.393980848153216,
"grad_norm": 0.32875871658325195,
"learning_rate": 5.63336652037777e-05,
"loss": 0.0143,
"step": 6000
},
{
"epoch": 16.421340629274965,
"grad_norm": 0.3572709262371063,
"learning_rate": 5.618541199084537e-05,
"loss": 0.0145,
"step": 6010
},
{
"epoch": 16.448700410396718,
"grad_norm": 0.2989596426486969,
"learning_rate": 5.603710353235695e-05,
"loss": 0.0139,
"step": 6020
},
{
"epoch": 16.476060191518467,
"grad_norm": 0.33748066425323486,
"learning_rate": 5.588874115294256e-05,
"loss": 0.0146,
"step": 6030
},
{
"epoch": 16.50341997264022,
"grad_norm": 0.329308420419693,
"learning_rate": 5.574032617771399e-05,
"loss": 0.0137,
"step": 6040
},
{
"epoch": 16.53077975376197,
"grad_norm": 0.35348793864250183,
"learning_rate": 5.559185993225272e-05,
"loss": 0.0145,
"step": 6050
},
{
"epoch": 16.558139534883722,
"grad_norm": 0.32704129815101624,
"learning_rate": 5.544334374259823e-05,
"loss": 0.0144,
"step": 6060
},
{
"epoch": 16.58549931600547,
"grad_norm": 0.33737605810165405,
"learning_rate": 5.529477893523606e-05,
"loss": 0.0145,
"step": 6070
},
{
"epoch": 16.612859097127224,
"grad_norm": 0.3684633672237396,
"learning_rate": 5.5146166837085925e-05,
"loss": 0.015,
"step": 6080
},
{
"epoch": 16.640218878248973,
"grad_norm": 0.3475631773471832,
"learning_rate": 5.4997508775490015e-05,
"loss": 0.0146,
"step": 6090
},
{
"epoch": 16.667578659370726,
"grad_norm": 0.34317895770072937,
"learning_rate": 5.484880607820097e-05,
"loss": 0.0146,
"step": 6100
},
{
"epoch": 16.694938440492475,
"grad_norm": 0.3417471647262573,
"learning_rate": 5.470006007337012e-05,
"loss": 0.0148,
"step": 6110
},
{
"epoch": 16.722298221614228,
"grad_norm": 0.36149823665618896,
"learning_rate": 5.455127208953564e-05,
"loss": 0.015,
"step": 6120
},
{
"epoch": 16.749658002735977,
"grad_norm": 0.3797590732574463,
"learning_rate": 5.4402443455610596e-05,
"loss": 0.0149,
"step": 6130
},
{
"epoch": 16.77701778385773,
"grad_norm": 0.3667081594467163,
"learning_rate": 5.4253575500871144e-05,
"loss": 0.0145,
"step": 6140
},
{
"epoch": 16.80437756497948,
"grad_norm": 0.32740288972854614,
"learning_rate": 5.410466955494462e-05,
"loss": 0.0149,
"step": 6150
},
{
"epoch": 16.83173734610123,
"grad_norm": 0.31508126854896545,
"learning_rate": 5.3955726947797695e-05,
"loss": 0.0145,
"step": 6160
},
{
"epoch": 16.85909712722298,
"grad_norm": 0.35782861709594727,
"learning_rate": 5.38067490097245e-05,
"loss": 0.0153,
"step": 6170
},
{
"epoch": 16.886456908344734,
"grad_norm": 0.33792731165885925,
"learning_rate": 5.36577370713347e-05,
"loss": 0.0145,
"step": 6180
},
{
"epoch": 16.913816689466483,
"grad_norm": 0.36968305706977844,
"learning_rate": 5.350869246354164e-05,
"loss": 0.0164,
"step": 6190
},
{
"epoch": 16.941176470588236,
"grad_norm": 0.3754122853279114,
"learning_rate": 5.3359616517550474e-05,
"loss": 0.0153,
"step": 6200
},
{
"epoch": 16.968536251709985,
"grad_norm": 0.3644407093524933,
"learning_rate": 5.321051056484623e-05,
"loss": 0.0158,
"step": 6210
},
{
"epoch": 16.995896032831737,
"grad_norm": 0.34643101692199707,
"learning_rate": 5.306137593718196e-05,
"loss": 0.0158,
"step": 6220
},
{
"epoch": 17.0218878248974,
"grad_norm": 0.2399960607290268,
"learning_rate": 5.291221396656684e-05,
"loss": 0.0081,
"step": 6230
},
{
"epoch": 17.04924760601915,
"grad_norm": 0.21385091543197632,
"learning_rate": 5.276302598525422e-05,
"loss": 0.0073,
"step": 6240
},
{
"epoch": 17.076607387140903,
"grad_norm": 0.22016951441764832,
"learning_rate": 5.261381332572982e-05,
"loss": 0.0073,
"step": 6250
},
{
"epoch": 17.103967168262653,
"grad_norm": 0.20825804769992828,
"learning_rate": 5.246457732069975e-05,
"loss": 0.007,
"step": 6260
},
{
"epoch": 17.131326949384405,
"grad_norm": 0.21500270068645477,
"learning_rate": 5.2315319303078616e-05,
"loss": 0.0065,
"step": 6270
},
{
"epoch": 17.158686730506155,
"grad_norm": 0.21949522197246552,
"learning_rate": 5.216604060597766e-05,
"loss": 0.0069,
"step": 6280
},
{
"epoch": 17.186046511627907,
"grad_norm": 0.22590532898902893,
"learning_rate": 5.20167425626928e-05,
"loss": 0.0067,
"step": 6290
},
{
"epoch": 17.213406292749657,
"grad_norm": 0.22894498705863953,
"learning_rate": 5.186742650669277e-05,
"loss": 0.0064,
"step": 6300
},
{
"epoch": 17.24076607387141,
"grad_norm": 0.205558642745018,
"learning_rate": 5.171809377160719e-05,
"loss": 0.0061,
"step": 6310
},
{
"epoch": 17.26812585499316,
"grad_norm": 0.24590414762496948,
"learning_rate": 5.1568745691214606e-05,
"loss": 0.0066,
"step": 6320
},
{
"epoch": 17.29548563611491,
"grad_norm": 0.20685908198356628,
"learning_rate": 5.141938359943066e-05,
"loss": 0.0062,
"step": 6330
},
{
"epoch": 17.32284541723666,
"grad_norm": 0.23585104942321777,
"learning_rate": 5.1270008830296156e-05,
"loss": 0.0064,
"step": 6340
},
{
"epoch": 17.350205198358413,
"grad_norm": 0.22263722121715546,
"learning_rate": 5.112062271796505e-05,
"loss": 0.006,
"step": 6350
},
{
"epoch": 17.377564979480166,
"grad_norm": 0.2212512344121933,
"learning_rate": 5.09712265966927e-05,
"loss": 0.0063,
"step": 6360
},
{
"epoch": 17.404924760601915,
"grad_norm": 0.22064076364040375,
"learning_rate": 5.082182180082382e-05,
"loss": 0.0061,
"step": 6370
},
{
"epoch": 17.432284541723668,
"grad_norm": 0.20909123122692108,
"learning_rate": 5.0672409664780586e-05,
"loss": 0.0062,
"step": 6380
},
{
"epoch": 17.459644322845417,
"grad_norm": 0.205373615026474,
"learning_rate": 5.0522991523050766e-05,
"loss": 0.0061,
"step": 6390
},
{
"epoch": 17.48700410396717,
"grad_norm": 0.20341329276561737,
"learning_rate": 5.0373568710175734e-05,
"loss": 0.0064,
"step": 6400
},
{
"epoch": 17.51436388508892,
"grad_norm": 0.24754370748996735,
"learning_rate": 5.022414256073862e-05,
"loss": 0.0063,
"step": 6410
},
{
"epoch": 17.54172366621067,
"grad_norm": 0.2462504655122757,
"learning_rate": 5.007471440935234e-05,
"loss": 0.0063,
"step": 6420
},
{
"epoch": 17.56908344733242,
"grad_norm": 0.2234465628862381,
"learning_rate": 4.992528559064768e-05,
"loss": 0.0064,
"step": 6430
},
{
"epoch": 17.596443228454174,
"grad_norm": 0.240615114569664,
"learning_rate": 4.9775857439261394e-05,
"loss": 0.0066,
"step": 6440
},
{
"epoch": 17.623803009575923,
"grad_norm": 0.2524745762348175,
"learning_rate": 4.962643128982427e-05,
"loss": 0.0065,
"step": 6450
},
{
"epoch": 17.651162790697676,
"grad_norm": 0.2284671664237976,
"learning_rate": 4.947700847694925e-05,
"loss": 0.0064,
"step": 6460
},
{
"epoch": 17.678522571819425,
"grad_norm": 0.24384962022304535,
"learning_rate": 4.932759033521943e-05,
"loss": 0.0069,
"step": 6470
},
{
"epoch": 17.705882352941178,
"grad_norm": 0.22949935495853424,
"learning_rate": 4.917817819917619e-05,
"loss": 0.0064,
"step": 6480
},
{
"epoch": 17.733242134062927,
"grad_norm": 0.23356693983078003,
"learning_rate": 4.902877340330731e-05,
"loss": 0.0074,
"step": 6490
},
{
"epoch": 17.76060191518468,
"grad_norm": 0.23319274187088013,
"learning_rate": 4.887937728203496e-05,
"loss": 0.007,
"step": 6500
},
{
"epoch": 17.78796169630643,
"grad_norm": 0.25012335181236267,
"learning_rate": 4.872999116970386e-05,
"loss": 0.0072,
"step": 6510
},
{
"epoch": 17.81532147742818,
"grad_norm": 0.23877964913845062,
"learning_rate": 4.858061640056934e-05,
"loss": 0.0066,
"step": 6520
},
{
"epoch": 17.84268125854993,
"grad_norm": 0.28133559226989746,
"learning_rate": 4.8431254308785406e-05,
"loss": 0.007,
"step": 6530
},
{
"epoch": 17.870041039671683,
"grad_norm": 0.22307847440242767,
"learning_rate": 4.8281906228392825e-05,
"loss": 0.0066,
"step": 6540
},
{
"epoch": 17.897400820793433,
"grad_norm": 0.231056809425354,
"learning_rate": 4.813257349330722e-05,
"loss": 0.0071,
"step": 6550
},
{
"epoch": 17.924760601915185,
"grad_norm": 0.2446180284023285,
"learning_rate": 4.798325743730721e-05,
"loss": 0.0066,
"step": 6560
},
{
"epoch": 17.952120383036934,
"grad_norm": 0.2306022197008133,
"learning_rate": 4.783395939402236e-05,
"loss": 0.0073,
"step": 6570
},
{
"epoch": 17.979480164158687,
"grad_norm": 0.2445317804813385,
"learning_rate": 4.768468069692139e-05,
"loss": 0.0071,
"step": 6580
},
{
"epoch": 18.00547195622435,
"grad_norm": 0.16725143790245056,
"learning_rate": 4.7535422679300263e-05,
"loss": 0.0059,
"step": 6590
},
{
"epoch": 18.0328317373461,
"grad_norm": 0.16170556843280792,
"learning_rate": 4.738618667427019e-05,
"loss": 0.0041,
"step": 6600
},
{
"epoch": 18.060191518467853,
"grad_norm": 0.1556808352470398,
"learning_rate": 4.723697401474579e-05,
"loss": 0.0038,
"step": 6610
},
{
"epoch": 18.087551299589602,
"grad_norm": 0.179331973195076,
"learning_rate": 4.708778603343319e-05,
"loss": 0.004,
"step": 6620
},
{
"epoch": 18.114911080711355,
"grad_norm": 0.1519869565963745,
"learning_rate": 4.6938624062818054e-05,
"loss": 0.0035,
"step": 6630
},
{
"epoch": 18.142270861833104,
"grad_norm": 0.17615194618701935,
"learning_rate": 4.678948943515379e-05,
"loss": 0.0038,
"step": 6640
},
{
"epoch": 18.169630642954857,
"grad_norm": 0.16421492397785187,
"learning_rate": 4.664038348244956e-05,
"loss": 0.0038,
"step": 6650
},
{
"epoch": 18.196990424076606,
"grad_norm": 0.1540912538766861,
"learning_rate": 4.6491307536458375e-05,
"loss": 0.0035,
"step": 6660
},
{
"epoch": 18.22435020519836,
"grad_norm": 0.1508628875017166,
"learning_rate": 4.6342262928665315e-05,
"loss": 0.0035,
"step": 6670
},
{
"epoch": 18.251709986320108,
"grad_norm": 0.18183846771717072,
"learning_rate": 4.61932509902755e-05,
"loss": 0.0034,
"step": 6680
},
{
"epoch": 18.27906976744186,
"grad_norm": 0.16749843955039978,
"learning_rate": 4.6044273052202324e-05,
"loss": 0.0033,
"step": 6690
},
{
"epoch": 18.30642954856361,
"grad_norm": 0.1775486320257187,
"learning_rate": 4.589533044505539e-05,
"loss": 0.0037,
"step": 6700
},
{
"epoch": 18.333789329685363,
"grad_norm": 0.16683386266231537,
"learning_rate": 4.574642449912887e-05,
"loss": 0.0035,
"step": 6710
},
{
"epoch": 18.361149110807112,
"grad_norm": 0.16168291866779327,
"learning_rate": 4.559755654438942e-05,
"loss": 0.0036,
"step": 6720
},
{
"epoch": 18.388508891928865,
"grad_norm": 0.1892244964838028,
"learning_rate": 4.544872791046436e-05,
"loss": 0.0036,
"step": 6730
},
{
"epoch": 18.415868673050614,
"grad_norm": 0.19447311758995056,
"learning_rate": 4.529993992662988e-05,
"loss": 0.0035,
"step": 6740
},
{
"epoch": 18.443228454172367,
"grad_norm": 0.17521780729293823,
"learning_rate": 4.515119392179906e-05,
"loss": 0.0034,
"step": 6750
},
{
"epoch": 18.470588235294116,
"grad_norm": 0.16449834406375885,
"learning_rate": 4.500249122451e-05,
"loss": 0.0037,
"step": 6760
},
{
"epoch": 18.49794801641587,
"grad_norm": 0.17732852697372437,
"learning_rate": 4.485383316291408e-05,
"loss": 0.0036,
"step": 6770
},
{
"epoch": 18.52530779753762,
"grad_norm": 0.18374419212341309,
"learning_rate": 4.470522106476397e-05,
"loss": 0.0039,
"step": 6780
},
{
"epoch": 18.55266757865937,
"grad_norm": 0.19527047872543335,
"learning_rate": 4.4556656257401786e-05,
"loss": 0.0037,
"step": 6790
},
{
"epoch": 18.580027359781123,
"grad_norm": 0.1950490027666092,
"learning_rate": 4.440814006774728e-05,
"loss": 0.0038,
"step": 6800
},
{
"epoch": 18.607387140902873,
"grad_norm": 0.17217150330543518,
"learning_rate": 4.425967382228604e-05,
"loss": 0.0034,
"step": 6810
},
{
"epoch": 18.634746922024625,
"grad_norm": 0.18268582224845886,
"learning_rate": 4.4111258847057454e-05,
"loss": 0.0039,
"step": 6820
},
{
"epoch": 18.662106703146375,
"grad_norm": 0.17226670682430267,
"learning_rate": 4.396289646764306e-05,
"loss": 0.0039,
"step": 6830
},
{
"epoch": 18.689466484268127,
"grad_norm": 0.1835619956254959,
"learning_rate": 4.381458800915463e-05,
"loss": 0.0038,
"step": 6840
},
{
"epoch": 18.716826265389876,
"grad_norm": 0.17223717272281647,
"learning_rate": 4.366633479622232e-05,
"loss": 0.0038,
"step": 6850
},
{
"epoch": 18.74418604651163,
"grad_norm": 0.18623270094394684,
"learning_rate": 4.351813815298284e-05,
"loss": 0.0036,
"step": 6860
},
{
"epoch": 18.77154582763338,
"grad_norm": 0.17093777656555176,
"learning_rate": 4.3369999403067666e-05,
"loss": 0.004,
"step": 6870
},
{
"epoch": 18.79890560875513,
"grad_norm": 0.20035335421562195,
"learning_rate": 4.322191986959121e-05,
"loss": 0.004,
"step": 6880
},
{
"epoch": 18.82626538987688,
"grad_norm": 0.18462000787258148,
"learning_rate": 4.3073900875138945e-05,
"loss": 0.0043,
"step": 6890
},
{
"epoch": 18.853625170998633,
"grad_norm": 0.1928115040063858,
"learning_rate": 4.292594374175565e-05,
"loss": 0.0041,
"step": 6900
},
{
"epoch": 18.880984952120382,
"grad_norm": 0.2051820009946823,
"learning_rate": 4.2778049790933627e-05,
"loss": 0.0041,
"step": 6910
},
{
"epoch": 18.908344733242135,
"grad_norm": 0.17575626075267792,
"learning_rate": 4.263022034360079e-05,
"loss": 0.0041,
"step": 6920
},
{
"epoch": 18.935704514363884,
"grad_norm": 0.1800376921892166,
"learning_rate": 4.248245672010896e-05,
"loss": 0.0042,
"step": 6930
},
{
"epoch": 18.963064295485637,
"grad_norm": 0.1879468858242035,
"learning_rate": 4.2334760240222105e-05,
"loss": 0.0042,
"step": 6940
},
{
"epoch": 18.990424076607386,
"grad_norm": 0.18099690973758698,
"learning_rate": 4.2187132223104374e-05,
"loss": 0.0042,
"step": 6950
},
{
"epoch": 19.01641586867305,
"grad_norm": 0.14218460023403168,
"learning_rate": 4.2039573987308545e-05,
"loss": 0.0032,
"step": 6960
},
{
"epoch": 19.043775649794803,
"grad_norm": 0.16449818015098572,
"learning_rate": 4.18920868507641e-05,
"loss": 0.0029,
"step": 6970
},
{
"epoch": 19.071135430916552,
"grad_norm": 0.13452163338661194,
"learning_rate": 4.174467213076546e-05,
"loss": 0.0024,
"step": 6980
},
{
"epoch": 19.098495212038305,
"grad_norm": 0.1514378935098648,
"learning_rate": 4.15973311439603e-05,
"loss": 0.0026,
"step": 6990
},
{
"epoch": 19.125854993160054,
"grad_norm": 0.15224389731884003,
"learning_rate": 4.145006520633769e-05,
"loss": 0.0024,
"step": 7000
},
{
"epoch": 19.153214774281807,
"grad_norm": 0.1427348554134369,
"learning_rate": 4.1302875633216473e-05,
"loss": 0.0023,
"step": 7010
},
{
"epoch": 19.180574555403556,
"grad_norm": 0.17503510415554047,
"learning_rate": 4.115576373923334e-05,
"loss": 0.0024,
"step": 7020
},
{
"epoch": 19.20793433652531,
"grad_norm": 0.1383901685476303,
"learning_rate": 4.100873083833123e-05,
"loss": 0.0022,
"step": 7030
},
{
"epoch": 19.235294117647058,
"grad_norm": 0.1591738760471344,
"learning_rate": 4.08617782437476e-05,
"loss": 0.0022,
"step": 7040
},
{
"epoch": 19.26265389876881,
"grad_norm": 0.16079847514629364,
"learning_rate": 4.07149072680025e-05,
"loss": 0.0025,
"step": 7050
},
{
"epoch": 19.29001367989056,
"grad_norm": 0.15590646862983704,
"learning_rate": 4.0568119222887146e-05,
"loss": 0.0025,
"step": 7060
},
{
"epoch": 19.317373461012313,
"grad_norm": 0.14387153089046478,
"learning_rate": 4.042141541945199e-05,
"loss": 0.0023,
"step": 7070
},
{
"epoch": 19.344733242134062,
"grad_norm": 0.14333535730838776,
"learning_rate": 4.027479716799503e-05,
"loss": 0.0021,
"step": 7080
},
{
"epoch": 19.372093023255815,
"grad_norm": 0.14790193736553192,
"learning_rate": 4.012826577805022e-05,
"loss": 0.0025,
"step": 7090
},
{
"epoch": 19.399452804377564,
"grad_norm": 0.15470997989177704,
"learning_rate": 3.998182255837568e-05,
"loss": 0.0024,
"step": 7100
},
{
"epoch": 19.426812585499317,
"grad_norm": 0.6740047931671143,
"learning_rate": 3.9835468816942e-05,
"loss": 0.0025,
"step": 7110
},
{
"epoch": 19.454172366621066,
"grad_norm": 0.1887519359588623,
"learning_rate": 3.968920586092063e-05,
"loss": 0.0031,
"step": 7120
},
{
"epoch": 19.48153214774282,
"grad_norm": 0.19010402262210846,
"learning_rate": 3.954303499667212e-05,
"loss": 0.0035,
"step": 7130
},
{
"epoch": 19.508891928864568,
"grad_norm": 0.18507176637649536,
"learning_rate": 3.9396957529734546e-05,
"loss": 0.0032,
"step": 7140
},
{
"epoch": 19.53625170998632,
"grad_norm": 0.16173076629638672,
"learning_rate": 3.925097476481173e-05,
"loss": 0.0032,
"step": 7150
},
{
"epoch": 19.56361149110807,
"grad_norm": 0.17077334225177765,
"learning_rate": 3.910508800576172e-05,
"loss": 0.0031,
"step": 7160
},
{
"epoch": 19.590971272229822,
"grad_norm": 0.16841229796409607,
"learning_rate": 3.895929855558505e-05,
"loss": 0.0031,
"step": 7170
},
{
"epoch": 19.61833105335157,
"grad_norm": 0.1734660565853119,
"learning_rate": 3.881360771641311e-05,
"loss": 0.0032,
"step": 7180
},
{
"epoch": 19.645690834473324,
"grad_norm": 0.1757352352142334,
"learning_rate": 3.8668016789496574e-05,
"loss": 0.0032,
"step": 7190
},
{
"epoch": 19.673050615595074,
"grad_norm": 0.1816571205854416,
"learning_rate": 3.852252707519375e-05,
"loss": 0.0031,
"step": 7200
},
{
"epoch": 19.700410396716826,
"grad_norm": 0.167251855134964,
"learning_rate": 3.8377139872958887e-05,
"loss": 0.0029,
"step": 7210
},
{
"epoch": 19.72777017783858,
"grad_norm": 0.14826416969299316,
"learning_rate": 3.8231856481330696e-05,
"loss": 0.003,
"step": 7220
},
{
"epoch": 19.755129958960328,
"grad_norm": 0.18200914561748505,
"learning_rate": 3.808667819792071e-05,
"loss": 0.0032,
"step": 7230
},
{
"epoch": 19.78248974008208,
"grad_norm": 0.14513128995895386,
"learning_rate": 3.794160631940163e-05,
"loss": 0.0031,
"step": 7240
},
{
"epoch": 19.80984952120383,
"grad_norm": 0.18026135861873627,
"learning_rate": 3.77966421414958e-05,
"loss": 0.003,
"step": 7250
},
{
"epoch": 19.837209302325583,
"grad_norm": 0.17554031312465668,
"learning_rate": 3.765178695896368e-05,
"loss": 0.0031,
"step": 7260
},
{
"epoch": 19.864569083447332,
"grad_norm": 0.1713663786649704,
"learning_rate": 3.750704206559215e-05,
"loss": 0.0034,
"step": 7270
},
{
"epoch": 19.891928864569085,
"grad_norm": 0.1802830547094345,
"learning_rate": 3.7362408754183074e-05,
"loss": 0.0033,
"step": 7280
},
{
"epoch": 19.919288645690834,
"grad_norm": 0.17740604281425476,
"learning_rate": 3.7217888316541724e-05,
"loss": 0.0035,
"step": 7290
},
{
"epoch": 19.946648426812587,
"grad_norm": 0.18996040523052216,
"learning_rate": 3.707348204346523e-05,
"loss": 0.0036,
"step": 7300
},
{
"epoch": 19.974008207934336,
"grad_norm": 0.15936149656772614,
"learning_rate": 3.6929191224730994e-05,
"loss": 0.0033,
"step": 7310
},
{
"epoch": 20.0,
"grad_norm": 0.19856102764606476,
"learning_rate": 3.6785017149085274e-05,
"loss": 0.0031,
"step": 7320
},
{
"epoch": 20.027359781121753,
"grad_norm": 0.12226132303476334,
"learning_rate": 3.6640961104231663e-05,
"loss": 0.0022,
"step": 7330
},
{
"epoch": 20.054719562243502,
"grad_norm": 0.14228872954845428,
"learning_rate": 3.6497024376819465e-05,
"loss": 0.0021,
"step": 7340
},
{
"epoch": 20.082079343365255,
"grad_norm": 0.14044199883937836,
"learning_rate": 3.6353208252432344e-05,
"loss": 0.0022,
"step": 7350
},
{
"epoch": 20.109439124487004,
"grad_norm": 0.10883612185716629,
"learning_rate": 3.6209514015576805e-05,
"loss": 0.0019,
"step": 7360
},
{
"epoch": 20.136798905608757,
"grad_norm": 0.15866024792194366,
"learning_rate": 3.6065942949670645e-05,
"loss": 0.002,
"step": 7370
},
{
"epoch": 20.164158686730506,
"grad_norm": 0.12476954609155655,
"learning_rate": 3.592249633703159e-05,
"loss": 0.0018,
"step": 7380
},
{
"epoch": 20.19151846785226,
"grad_norm": 0.13289691507816315,
"learning_rate": 3.57791754588658e-05,
"loss": 0.0019,
"step": 7390
},
{
"epoch": 20.218878248974008,
"grad_norm": 0.11743517220020294,
"learning_rate": 3.56359815952564e-05,
"loss": 0.0018,
"step": 7400
},
{
"epoch": 20.24623803009576,
"grad_norm": 0.1258155256509781,
"learning_rate": 3.549291602515209e-05,
"loss": 0.0018,
"step": 7410
},
{
"epoch": 20.27359781121751,
"grad_norm": 0.12627840042114258,
"learning_rate": 3.534998002635574e-05,
"loss": 0.0017,
"step": 7420
},
{
"epoch": 20.300957592339262,
"grad_norm": 0.1615668535232544,
"learning_rate": 3.520717487551286e-05,
"loss": 0.0019,
"step": 7430
},
{
"epoch": 20.32831737346101,
"grad_norm": 0.1235918328166008,
"learning_rate": 3.506450184810036e-05,
"loss": 0.0017,
"step": 7440
},
{
"epoch": 20.355677154582764,
"grad_norm": 0.12315589189529419,
"learning_rate": 3.492196221841502e-05,
"loss": 0.0018,
"step": 7450
},
{
"epoch": 20.383036935704514,
"grad_norm": 0.11821737140417099,
"learning_rate": 3.4779557259562246e-05,
"loss": 0.0018,
"step": 7460
},
{
"epoch": 20.410396716826266,
"grad_norm": 0.13333961367607117,
"learning_rate": 3.4637288243444494e-05,
"loss": 0.0018,
"step": 7470
},
{
"epoch": 20.437756497948016,
"grad_norm": 0.12324558943510056,
"learning_rate": 3.449515644075015e-05,
"loss": 0.002,
"step": 7480
},
{
"epoch": 20.46511627906977,
"grad_norm": 0.15187454223632812,
"learning_rate": 3.4353163120942035e-05,
"loss": 0.0018,
"step": 7490
},
{
"epoch": 20.492476060191517,
"grad_norm": 0.14215128123760223,
"learning_rate": 3.421130955224603e-05,
"loss": 0.002,
"step": 7500
},
{
"epoch": 20.51983584131327,
"grad_norm": 0.12410500645637512,
"learning_rate": 3.4069597001639925e-05,
"loss": 0.0018,
"step": 7510
},
{
"epoch": 20.54719562243502,
"grad_norm": 0.1700277477502823,
"learning_rate": 3.392802673484193e-05,
"loss": 0.002,
"step": 7520
},
{
"epoch": 20.574555403556772,
"grad_norm": 0.12269216775894165,
"learning_rate": 3.378660001629943e-05,
"loss": 0.002,
"step": 7530
},
{
"epoch": 20.60191518467852,
"grad_norm": 0.1367766410112381,
"learning_rate": 3.3645318109177726e-05,
"loss": 0.0022,
"step": 7540
},
{
"epoch": 20.629274965800274,
"grad_norm": 0.1393338292837143,
"learning_rate": 3.3504182275348715e-05,
"loss": 0.002,
"step": 7550
},
{
"epoch": 20.656634746922023,
"grad_norm": 0.1607707440853119,
"learning_rate": 3.3363193775379605e-05,
"loss": 0.0021,
"step": 7560
},
{
"epoch": 20.683994528043776,
"grad_norm": 0.14136134088039398,
"learning_rate": 3.322235386852169e-05,
"loss": 0.002,
"step": 7570
},
{
"epoch": 20.711354309165525,
"grad_norm": 0.12284570932388306,
"learning_rate": 3.3081663812699135e-05,
"loss": 0.0019,
"step": 7580
},
{
"epoch": 20.738714090287278,
"grad_norm": 0.14074423909187317,
"learning_rate": 3.294112486449762e-05,
"loss": 0.002,
"step": 7590
},
{
"epoch": 20.766073871409027,
"grad_norm": 0.1331290304660797,
"learning_rate": 3.280073827915324e-05,
"loss": 0.0019,
"step": 7600
},
{
"epoch": 20.79343365253078,
"grad_norm": 0.1495884507894516,
"learning_rate": 3.2660505310541256e-05,
"loss": 0.0021,
"step": 7610
},
{
"epoch": 20.82079343365253,
"grad_norm": 0.1477614939212799,
"learning_rate": 3.2520427211164875e-05,
"loss": 0.0021,
"step": 7620
},
{
"epoch": 20.848153214774282,
"grad_norm": 0.15515491366386414,
"learning_rate": 3.238050523214404e-05,
"loss": 0.0021,
"step": 7630
},
{
"epoch": 20.87551299589603,
"grad_norm": 0.1545867621898651,
"learning_rate": 3.2240740623204356e-05,
"loss": 0.0022,
"step": 7640
},
{
"epoch": 20.902872777017784,
"grad_norm": 0.16105583310127258,
"learning_rate": 3.210113463266583e-05,
"loss": 0.0022,
"step": 7650
},
{
"epoch": 20.930232558139537,
"grad_norm": 0.13224723935127258,
"learning_rate": 3.196168850743174e-05,
"loss": 0.0022,
"step": 7660
},
{
"epoch": 20.957592339261286,
"grad_norm": 0.14821544289588928,
"learning_rate": 3.1822403492977516e-05,
"loss": 0.0022,
"step": 7670
},
{
"epoch": 20.98495212038304,
"grad_norm": 0.15173478424549103,
"learning_rate": 3.1683280833339664e-05,
"loss": 0.0022,
"step": 7680
},
{
"epoch": 21.0109439124487,
"grad_norm": 0.1002042144536972,
"learning_rate": 3.1544321771104534e-05,
"loss": 0.0018,
"step": 7690
},
{
"epoch": 21.03830369357045,
"grad_norm": 0.11058831214904785,
"learning_rate": 3.140552754739732e-05,
"loss": 0.0014,
"step": 7700
},
{
"epoch": 21.0656634746922,
"grad_norm": 0.09244994819164276,
"learning_rate": 3.126689940187099e-05,
"loss": 0.0012,
"step": 7710
},
{
"epoch": 21.093023255813954,
"grad_norm": 0.09339479357004166,
"learning_rate": 3.1128438572695085e-05,
"loss": 0.0013,
"step": 7720
},
{
"epoch": 21.120383036935703,
"grad_norm": 0.11077740788459778,
"learning_rate": 3.09901462965448e-05,
"loss": 0.0014,
"step": 7730
},
{
"epoch": 21.147742818057456,
"grad_norm": 0.08118260651826859,
"learning_rate": 3.085202380858988e-05,
"loss": 0.0013,
"step": 7740
},
{
"epoch": 21.175102599179205,
"grad_norm": 0.12861959636211395,
"learning_rate": 3.071407234248356e-05,
"loss": 0.0014,
"step": 7750
},
{
"epoch": 21.202462380300958,
"grad_norm": 0.12473811209201813,
"learning_rate": 3.057629313035163e-05,
"loss": 0.0013,
"step": 7760
},
{
"epoch": 21.22982216142271,
"grad_norm": 0.09730850160121918,
"learning_rate": 3.043868740278133e-05,
"loss": 0.0011,
"step": 7770
},
{
"epoch": 21.25718194254446,
"grad_norm": 0.0872691422700882,
"learning_rate": 3.0301256388810458e-05,
"loss": 0.0013,
"step": 7780
},
{
"epoch": 21.284541723666212,
"grad_norm": 0.08004292100667953,
"learning_rate": 3.016400131591628e-05,
"loss": 0.0013,
"step": 7790
},
{
"epoch": 21.31190150478796,
"grad_norm": 0.11337770521640778,
"learning_rate": 3.002692341000466e-05,
"loss": 0.0015,
"step": 7800
},
{
"epoch": 21.339261285909714,
"grad_norm": 0.08427674323320389,
"learning_rate": 2.989002389539912e-05,
"loss": 0.0012,
"step": 7810
},
{
"epoch": 21.366621067031463,
"grad_norm": 0.10581520199775696,
"learning_rate": 2.975330399482976e-05,
"loss": 0.0014,
"step": 7820
},
{
"epoch": 21.393980848153216,
"grad_norm": 0.12079137563705444,
"learning_rate": 2.9616764929422557e-05,
"loss": 0.0014,
"step": 7830
},
{
"epoch": 21.421340629274965,
"grad_norm": 0.11981716752052307,
"learning_rate": 2.9480407918688274e-05,
"loss": 0.0014,
"step": 7840
},
{
"epoch": 21.448700410396718,
"grad_norm": 0.08894160389900208,
"learning_rate": 2.9344234180511638e-05,
"loss": 0.0014,
"step": 7850
},
{
"epoch": 21.476060191518467,
"grad_norm": 0.0949430912733078,
"learning_rate": 2.9208244931140493e-05,
"loss": 0.0015,
"step": 7860
},
{
"epoch": 21.50341997264022,
"grad_norm": 0.10954757034778595,
"learning_rate": 2.9072441385174892e-05,
"loss": 0.0014,
"step": 7870
},
{
"epoch": 21.53077975376197,
"grad_norm": 0.15860018134117126,
"learning_rate": 2.8936824755556258e-05,
"loss": 0.0015,
"step": 7880
},
{
"epoch": 21.558139534883722,
"grad_norm": 0.10013443976640701,
"learning_rate": 2.8801396253556522e-05,
"loss": 0.0013,
"step": 7890
},
{
"epoch": 21.58549931600547,
"grad_norm": 0.10873597115278244,
"learning_rate": 2.8666157088767416e-05,
"loss": 0.0013,
"step": 7900
},
{
"epoch": 21.612859097127224,
"grad_norm": 0.13066299259662628,
"learning_rate": 2.8531108469089496e-05,
"loss": 0.0014,
"step": 7910
},
{
"epoch": 21.640218878248973,
"grad_norm": 0.10153175890445709,
"learning_rate": 2.8396251600721537e-05,
"loss": 0.0014,
"step": 7920
},
{
"epoch": 21.667578659370726,
"grad_norm": 0.11935560405254364,
"learning_rate": 2.8261587688149582e-05,
"loss": 0.0015,
"step": 7930
},
{
"epoch": 21.694938440492475,
"grad_norm": 0.10965074598789215,
"learning_rate": 2.812711793413636e-05,
"loss": 0.0014,
"step": 7940
},
{
"epoch": 21.722298221614228,
"grad_norm": 0.10216519236564636,
"learning_rate": 2.7992843539710368e-05,
"loss": 0.0013,
"step": 7950
},
{
"epoch": 21.749658002735977,
"grad_norm": 0.11993634700775146,
"learning_rate": 2.7858765704155304e-05,
"loss": 0.0016,
"step": 7960
},
{
"epoch": 21.77701778385773,
"grad_norm": 0.10926508158445358,
"learning_rate": 2.772488562499927e-05,
"loss": 0.0017,
"step": 7970
},
{
"epoch": 21.80437756497948,
"grad_norm": 0.11312492936849594,
"learning_rate": 2.759120449800405e-05,
"loss": 0.0016,
"step": 7980
},
{
"epoch": 21.83173734610123,
"grad_norm": 0.11366380751132965,
"learning_rate": 2.7457723517154488e-05,
"loss": 0.0015,
"step": 7990
},
{
"epoch": 21.85909712722298,
"grad_norm": 0.11987926065921783,
"learning_rate": 2.732444387464783e-05,
"loss": 0.0015,
"step": 8000
}
],
"logging_steps": 10,
"max_steps": 11680,
"num_input_tokens_seen": 0,
"num_train_epochs": 32,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.151464205353222e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}