9b-111 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
3cb768c verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1482,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002699055330634278,
"grad_norm": 123.96797180175781,
"learning_rate": 1.6666666666666668e-07,
"loss": 0.29786407947540283,
"step": 2
},
{
"epoch": 0.005398110661268556,
"grad_norm": 3.0589780807495117,
"learning_rate": 5.000000000000001e-07,
"loss": 0.9299415349960327,
"step": 4
},
{
"epoch": 0.008097165991902834,
"grad_norm": 2.4749207496643066,
"learning_rate": 8.333333333333333e-07,
"loss": 0.9372931718826294,
"step": 6
},
{
"epoch": 0.010796221322537112,
"grad_norm": 2.2769699096679688,
"learning_rate": 1.1666666666666668e-06,
"loss": 0.5104070901870728,
"step": 8
},
{
"epoch": 0.01349527665317139,
"grad_norm": 3.0426244735717773,
"learning_rate": 1.5e-06,
"loss": 0.5902462005615234,
"step": 10
},
{
"epoch": 0.016194331983805668,
"grad_norm": 3.86154842376709,
"learning_rate": 1.8333333333333335e-06,
"loss": 0.928583025932312,
"step": 12
},
{
"epoch": 0.018893387314439947,
"grad_norm": 1.8792176246643066,
"learning_rate": 2.166666666666667e-06,
"loss": 0.8758266568183899,
"step": 14
},
{
"epoch": 0.021592442645074223,
"grad_norm": 4.866174221038818,
"learning_rate": 2.5e-06,
"loss": 1.0566331148147583,
"step": 16
},
{
"epoch": 0.024291497975708502,
"grad_norm": 3.7274439334869385,
"learning_rate": 2.8333333333333335e-06,
"loss": 0.4785526394844055,
"step": 18
},
{
"epoch": 0.02699055330634278,
"grad_norm": 2.2945733070373535,
"learning_rate": 3.166666666666667e-06,
"loss": 0.4686698913574219,
"step": 20
},
{
"epoch": 0.029689608636977057,
"grad_norm": 2.5992486476898193,
"learning_rate": 3.5000000000000004e-06,
"loss": 0.41955965757369995,
"step": 22
},
{
"epoch": 0.032388663967611336,
"grad_norm": 2.801939010620117,
"learning_rate": 3.833333333333334e-06,
"loss": 0.871375322341919,
"step": 24
},
{
"epoch": 0.03508771929824561,
"grad_norm": 2.4493536949157715,
"learning_rate": 4.166666666666667e-06,
"loss": 0.4926496744155884,
"step": 26
},
{
"epoch": 0.037786774628879895,
"grad_norm": 4.693727970123291,
"learning_rate": 4.5e-06,
"loss": 1.0131165981292725,
"step": 28
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.3234457969665527,
"learning_rate": 4.833333333333333e-06,
"loss": 0.9670289158821106,
"step": 30
},
{
"epoch": 0.043184885290148446,
"grad_norm": 10.311660766601562,
"learning_rate": 5.166666666666667e-06,
"loss": 0.4570980668067932,
"step": 32
},
{
"epoch": 0.04588394062078273,
"grad_norm": 2.5042009353637695,
"learning_rate": 5.500000000000001e-06,
"loss": 1.0683726072311401,
"step": 34
},
{
"epoch": 0.048582995951417005,
"grad_norm": 1.1646201610565186,
"learning_rate": 5.833333333333334e-06,
"loss": 0.45573946833610535,
"step": 36
},
{
"epoch": 0.05128205128205128,
"grad_norm": 1.9673726558685303,
"learning_rate": 6.166666666666667e-06,
"loss": 0.600303590297699,
"step": 38
},
{
"epoch": 0.05398110661268556,
"grad_norm": 1.8087180852890015,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.9155614376068115,
"step": 40
},
{
"epoch": 0.05668016194331984,
"grad_norm": 2.9125490188598633,
"learning_rate": 6.833333333333333e-06,
"loss": 0.10254953056573868,
"step": 42
},
{
"epoch": 0.059379217273954114,
"grad_norm": 4.365234851837158,
"learning_rate": 7.166666666666667e-06,
"loss": 0.5865437984466553,
"step": 44
},
{
"epoch": 0.0620782726045884,
"grad_norm": 3.153306245803833,
"learning_rate": 7.5e-06,
"loss": 0.4570964276790619,
"step": 46
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.16092848777771,
"learning_rate": 7.833333333333333e-06,
"loss": 0.9249556064605713,
"step": 48
},
{
"epoch": 0.06747638326585695,
"grad_norm": 1.8864729404449463,
"learning_rate": 8.166666666666668e-06,
"loss": 0.9879755973815918,
"step": 50
},
{
"epoch": 0.07017543859649122,
"grad_norm": 2.141915798187256,
"learning_rate": 8.500000000000002e-06,
"loss": 0.9798436164855957,
"step": 52
},
{
"epoch": 0.0728744939271255,
"grad_norm": 1.9038501977920532,
"learning_rate": 8.833333333333334e-06,
"loss": 0.9050228595733643,
"step": 54
},
{
"epoch": 0.07557354925775979,
"grad_norm": 2.1748414039611816,
"learning_rate": 9.166666666666666e-06,
"loss": 0.8556973934173584,
"step": 56
},
{
"epoch": 0.07827260458839407,
"grad_norm": 1.771748423576355,
"learning_rate": 9.5e-06,
"loss": 0.97296142578125,
"step": 58
},
{
"epoch": 0.08097165991902834,
"grad_norm": 12.455157279968262,
"learning_rate": 9.833333333333333e-06,
"loss": 0.8485205769538879,
"step": 60
},
{
"epoch": 0.08367071524966262,
"grad_norm": 1.9719027280807495,
"learning_rate": 1.0166666666666667e-05,
"loss": 0.8762905597686768,
"step": 62
},
{
"epoch": 0.08636977058029689,
"grad_norm": 2.098322868347168,
"learning_rate": 1.05e-05,
"loss": 0.8045662641525269,
"step": 64
},
{
"epoch": 0.08906882591093117,
"grad_norm": 2.10783314704895,
"learning_rate": 1.0833333333333334e-05,
"loss": 0.9314976334571838,
"step": 66
},
{
"epoch": 0.09176788124156546,
"grad_norm": 1.4629249572753906,
"learning_rate": 1.1166666666666668e-05,
"loss": 1.1287637948989868,
"step": 68
},
{
"epoch": 0.09446693657219973,
"grad_norm": 2.19218111038208,
"learning_rate": 1.1500000000000002e-05,
"loss": 1.0018784999847412,
"step": 70
},
{
"epoch": 0.09716599190283401,
"grad_norm": 7.496009349822998,
"learning_rate": 1.1833333333333334e-05,
"loss": 0.6217132806777954,
"step": 72
},
{
"epoch": 0.09986504723346828,
"grad_norm": 1.461382269859314,
"learning_rate": 1.2166666666666668e-05,
"loss": 0.4434390664100647,
"step": 74
},
{
"epoch": 0.10256410256410256,
"grad_norm": 3.232837438583374,
"learning_rate": 1.25e-05,
"loss": 0.9331040382385254,
"step": 76
},
{
"epoch": 0.10526315789473684,
"grad_norm": 2.710008382797241,
"learning_rate": 1.2499943912852736e-05,
"loss": 0.7011754512786865,
"step": 78
},
{
"epoch": 0.10796221322537113,
"grad_norm": 2.020285129547119,
"learning_rate": 1.2499775652529443e-05,
"loss": 1.0530972480773926,
"step": 80
},
{
"epoch": 0.1106612685560054,
"grad_norm": 2.3170576095581055,
"learning_rate": 1.249949522238558e-05,
"loss": 0.697028398513794,
"step": 82
},
{
"epoch": 0.11336032388663968,
"grad_norm": 2.3631045818328857,
"learning_rate": 1.2499102628013513e-05,
"loss": 0.8030006885528564,
"step": 84
},
{
"epoch": 0.11605937921727395,
"grad_norm": 0.7326921820640564,
"learning_rate": 1.2498597877242402e-05,
"loss": 0.596689760684967,
"step": 86
},
{
"epoch": 0.11875843454790823,
"grad_norm": 2.40205454826355,
"learning_rate": 1.249798098013803e-05,
"loss": 0.7292033433914185,
"step": 88
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.8302048444747925,
"learning_rate": 1.2497251949002623e-05,
"loss": 0.5806452035903931,
"step": 90
},
{
"epoch": 0.1241565452091768,
"grad_norm": 1.6161532402038574,
"learning_rate": 1.2496410798374584e-05,
"loss": 1.0763652324676514,
"step": 92
},
{
"epoch": 0.12685560053981107,
"grad_norm": 1.9346797466278076,
"learning_rate": 1.2495457545028227e-05,
"loss": 0.9050993919372559,
"step": 94
},
{
"epoch": 0.12955465587044535,
"grad_norm": 2.1635751724243164,
"learning_rate": 1.2494392207973419e-05,
"loss": 1.0081508159637451,
"step": 96
},
{
"epoch": 0.13225371120107962,
"grad_norm": 1.9737285375595093,
"learning_rate": 1.2493214808455219e-05,
"loss": 0.8886507749557495,
"step": 98
},
{
"epoch": 0.1349527665317139,
"grad_norm": 2.099431276321411,
"learning_rate": 1.2491925369953435e-05,
"loss": 0.9926242828369141,
"step": 100
},
{
"epoch": 0.13765182186234817,
"grad_norm": 2.662266254425049,
"learning_rate": 1.249052391818218e-05,
"loss": 0.9976869821548462,
"step": 102
},
{
"epoch": 0.14035087719298245,
"grad_norm": 1.6853854656219482,
"learning_rate": 1.248901048108934e-05,
"loss": 0.9816886782646179,
"step": 104
},
{
"epoch": 0.14304993252361672,
"grad_norm": 6.228536605834961,
"learning_rate": 1.2487385088856024e-05,
"loss": 0.6037951707839966,
"step": 106
},
{
"epoch": 0.145748987854251,
"grad_norm": 7.536047458648682,
"learning_rate": 1.2485647773895957e-05,
"loss": 0.6610171794891357,
"step": 108
},
{
"epoch": 0.1484480431848853,
"grad_norm": 4.089591979980469,
"learning_rate": 1.2483798570854848e-05,
"loss": 0.36474066972732544,
"step": 110
},
{
"epoch": 0.15114709851551958,
"grad_norm": 1.2712372541427612,
"learning_rate": 1.248183751660968e-05,
"loss": 0.5292806625366211,
"step": 112
},
{
"epoch": 0.15384615384615385,
"grad_norm": 1.7450406551361084,
"learning_rate": 1.2479764650267988e-05,
"loss": 1.097339153289795,
"step": 114
},
{
"epoch": 0.15654520917678813,
"grad_norm": 2.237950563430786,
"learning_rate": 1.2477580013167072e-05,
"loss": 0.976284384727478,
"step": 116
},
{
"epoch": 0.1592442645074224,
"grad_norm": 2.6984260082244873,
"learning_rate": 1.2475283648873178e-05,
"loss": 0.8394799828529358,
"step": 118
},
{
"epoch": 0.16194331983805668,
"grad_norm": 2.801074743270874,
"learning_rate": 1.2472875603180622e-05,
"loss": 0.7012876272201538,
"step": 120
},
{
"epoch": 0.16464237516869096,
"grad_norm": 1.151301383972168,
"learning_rate": 1.2470355924110887e-05,
"loss": 1.0838816165924072,
"step": 122
},
{
"epoch": 0.16734143049932523,
"grad_norm": 1.6192210912704468,
"learning_rate": 1.2467724661911659e-05,
"loss": 0.9945567846298218,
"step": 124
},
{
"epoch": 0.1700404858299595,
"grad_norm": 1.7934941053390503,
"learning_rate": 1.2464981869055819e-05,
"loss": 0.9529829621315002,
"step": 126
},
{
"epoch": 0.17273954116059378,
"grad_norm": 18.378650665283203,
"learning_rate": 1.2462127600240409e-05,
"loss": 0.8152964115142822,
"step": 128
},
{
"epoch": 0.17543859649122806,
"grad_norm": 38.34910583496094,
"learning_rate": 1.2459161912385535e-05,
"loss": 0.6358547806739807,
"step": 130
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6077762842178345,
"learning_rate": 1.245608486463323e-05,
"loss": 0.16767336428165436,
"step": 132
},
{
"epoch": 0.18083670715249664,
"grad_norm": 3.260507106781006,
"learning_rate": 1.2452896518346279e-05,
"loss": 0.6223729848861694,
"step": 134
},
{
"epoch": 0.18353576248313092,
"grad_norm": 3.0357773303985596,
"learning_rate": 1.2449596937106993e-05,
"loss": 0.40284159779548645,
"step": 136
},
{
"epoch": 0.1862348178137652,
"grad_norm": 1.1678791046142578,
"learning_rate": 1.2446186186715941e-05,
"loss": 0.988430380821228,
"step": 138
},
{
"epoch": 0.18893387314439947,
"grad_norm": 1.1722887754440308,
"learning_rate": 1.2442664335190634e-05,
"loss": 0.9420180916786194,
"step": 140
},
{
"epoch": 0.19163292847503374,
"grad_norm": 1.7688356637954712,
"learning_rate": 1.243903145276418e-05,
"loss": 0.8568445444107056,
"step": 142
},
{
"epoch": 0.19433198380566802,
"grad_norm": 1.7096655368804932,
"learning_rate": 1.2435287611883869e-05,
"loss": 0.6841754913330078,
"step": 144
},
{
"epoch": 0.1970310391363023,
"grad_norm": 1.9151872396469116,
"learning_rate": 1.2431432887209741e-05,
"loss": 0.9281893968582153,
"step": 146
},
{
"epoch": 0.19973009446693657,
"grad_norm": 1.6319060325622559,
"learning_rate": 1.2427467355613087e-05,
"loss": 0.9721434116363525,
"step": 148
},
{
"epoch": 0.20242914979757085,
"grad_norm": 2.102926015853882,
"learning_rate": 1.2423391096174922e-05,
"loss": 0.630251407623291,
"step": 150
},
{
"epoch": 0.20512820512820512,
"grad_norm": 3.118698835372925,
"learning_rate": 1.2419204190184405e-05,
"loss": 1.116424322128296,
"step": 152
},
{
"epoch": 0.2078272604588394,
"grad_norm": 5.269786357879639,
"learning_rate": 1.2414906721137222e-05,
"loss": 0.7768229246139526,
"step": 154
},
{
"epoch": 0.21052631578947367,
"grad_norm": 1.7130775451660156,
"learning_rate": 1.2410498774733918e-05,
"loss": 0.7465885877609253,
"step": 156
},
{
"epoch": 0.21322537112010798,
"grad_norm": 1.2827574014663696,
"learning_rate": 1.2405980438878187e-05,
"loss": 0.9992589354515076,
"step": 158
},
{
"epoch": 0.21592442645074225,
"grad_norm": 1.3526352643966675,
"learning_rate": 1.2401351803675115e-05,
"loss": 1.1087861061096191,
"step": 160
},
{
"epoch": 0.21862348178137653,
"grad_norm": 5.795386791229248,
"learning_rate": 1.23966129614294e-05,
"loss": 0.8273687362670898,
"step": 162
},
{
"epoch": 0.2213225371120108,
"grad_norm": 1.4797230958938599,
"learning_rate": 1.239176400664349e-05,
"loss": 0.9844716787338257,
"step": 164
},
{
"epoch": 0.22402159244264508,
"grad_norm": 2.5438950061798096,
"learning_rate": 1.2386805036015709e-05,
"loss": 1.0064153671264648,
"step": 166
},
{
"epoch": 0.22672064777327935,
"grad_norm": 24.848285675048828,
"learning_rate": 1.2381736148438333e-05,
"loss": 0.8047782778739929,
"step": 168
},
{
"epoch": 0.22941970310391363,
"grad_norm": 2.069371223449707,
"learning_rate": 1.2376557444995605e-05,
"loss": 0.4937904477119446,
"step": 170
},
{
"epoch": 0.2321187584345479,
"grad_norm": 2.0151240825653076,
"learning_rate": 1.237126902896173e-05,
"loss": 1.0987937450408936,
"step": 172
},
{
"epoch": 0.23481781376518218,
"grad_norm": 1.6710108518600464,
"learning_rate": 1.2365871005798818e-05,
"loss": 0.6136937141418457,
"step": 174
},
{
"epoch": 0.23751686909581646,
"grad_norm": 19.61103630065918,
"learning_rate": 1.2360363483154764e-05,
"loss": 1.1899774074554443,
"step": 176
},
{
"epoch": 0.24021592442645073,
"grad_norm": 1.5459237098693848,
"learning_rate": 1.235474657086112e-05,
"loss": 0.9780693054199219,
"step": 178
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.487262487411499,
"learning_rate": 1.2349020380930897e-05,
"loss": 0.6330715417861938,
"step": 180
},
{
"epoch": 0.24561403508771928,
"grad_norm": 0.6218187212944031,
"learning_rate": 1.2343185027556323e-05,
"loss": 0.5066222548484802,
"step": 182
},
{
"epoch": 0.2483130904183536,
"grad_norm": 1.9460417032241821,
"learning_rate": 1.2337240627106584e-05,
"loss": 0.9712158441543579,
"step": 184
},
{
"epoch": 0.25101214574898784,
"grad_norm": 1.5485562086105347,
"learning_rate": 1.233118729812549e-05,
"loss": 1.0215809345245361,
"step": 186
},
{
"epoch": 0.25371120107962214,
"grad_norm": 1.5811684131622314,
"learning_rate": 1.2325025161329109e-05,
"loss": 0.6591629385948181,
"step": 188
},
{
"epoch": 0.2564102564102564,
"grad_norm": 2.3522331714630127,
"learning_rate": 1.2318754339603373e-05,
"loss": 0.8668273687362671,
"step": 190
},
{
"epoch": 0.2591093117408907,
"grad_norm": 7.786831378936768,
"learning_rate": 1.2312374958001619e-05,
"loss": 0.850255012512207,
"step": 192
},
{
"epoch": 0.26180836707152494,
"grad_norm": 7.128757476806641,
"learning_rate": 1.2305887143742084e-05,
"loss": 0.933728814125061,
"step": 194
},
{
"epoch": 0.26450742240215924,
"grad_norm": 1.2906512022018433,
"learning_rate": 1.2299291026205397e-05,
"loss": 1.0171899795532227,
"step": 196
},
{
"epoch": 0.26720647773279355,
"grad_norm": 9.613722801208496,
"learning_rate": 1.2292586736931966e-05,
"loss": 0.17835456132888794,
"step": 198
},
{
"epoch": 0.2699055330634278,
"grad_norm": 1.638985514640808,
"learning_rate": 1.228577440961938e-05,
"loss": 0.9216899871826172,
"step": 200
},
{
"epoch": 0.2726045883940621,
"grad_norm": 1.2578356266021729,
"learning_rate": 1.2278854180119726e-05,
"loss": 1.0369343757629395,
"step": 202
},
{
"epoch": 0.27530364372469635,
"grad_norm": 1.291988730430603,
"learning_rate": 1.2271826186436896e-05,
"loss": 0.9166511297225952,
"step": 204
},
{
"epoch": 0.27800269905533065,
"grad_norm": 1.1957826614379883,
"learning_rate": 1.2264690568723817e-05,
"loss": 0.764278769493103,
"step": 206
},
{
"epoch": 0.2807017543859649,
"grad_norm": 1.4660202264785767,
"learning_rate": 1.2257447469279669e-05,
"loss": 0.6684662699699402,
"step": 208
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.8246656656265259,
"learning_rate": 1.2250097032547043e-05,
"loss": 1.0045970678329468,
"step": 210
},
{
"epoch": 0.28609986504723345,
"grad_norm": 1.6462355852127075,
"learning_rate": 1.2242639405109057e-05,
"loss": 0.9771850109100342,
"step": 212
},
{
"epoch": 0.28879892037786775,
"grad_norm": 1.5680208206176758,
"learning_rate": 1.2235074735686443e-05,
"loss": 0.9396547079086304,
"step": 214
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.9619158506393433,
"learning_rate": 1.222740317513457e-05,
"loss": 1.0256106853485107,
"step": 216
},
{
"epoch": 0.2941970310391363,
"grad_norm": 0.649040699005127,
"learning_rate": 1.2219624876440437e-05,
"loss": 0.7259390354156494,
"step": 218
},
{
"epoch": 0.2968960863697706,
"grad_norm": 1.64500093460083,
"learning_rate": 1.2211739994719637e-05,
"loss": 0.9923442602157593,
"step": 220
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.3503235578536987,
"learning_rate": 1.2203748687213235e-05,
"loss": 1.0567125082015991,
"step": 222
},
{
"epoch": 0.30229419703103916,
"grad_norm": 1.9163061380386353,
"learning_rate": 1.219565111328467e-05,
"loss": 0.8977484703063965,
"step": 224
},
{
"epoch": 0.3049932523616734,
"grad_norm": 1.6293538808822632,
"learning_rate": 1.2187447434416537e-05,
"loss": 1.0445246696472168,
"step": 226
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.5400710105895996,
"learning_rate": 1.2179137814207398e-05,
"loss": 0.5192906856536865,
"step": 228
},
{
"epoch": 0.31039136302294196,
"grad_norm": 30.060245513916016,
"learning_rate": 1.2170722418368503e-05,
"loss": 0.9097606539726257,
"step": 230
},
{
"epoch": 0.31309041835357626,
"grad_norm": 4.271488189697266,
"learning_rate": 1.2162201414720487e-05,
"loss": 0.6040419340133667,
"step": 232
},
{
"epoch": 0.3157894736842105,
"grad_norm": 15.518794059753418,
"learning_rate": 1.2153574973190035e-05,
"loss": 0.3896315097808838,
"step": 234
},
{
"epoch": 0.3184885290148448,
"grad_norm": 3.9835827350616455,
"learning_rate": 1.2144843265806474e-05,
"loss": 0.6756294965744019,
"step": 236
},
{
"epoch": 0.32118758434547906,
"grad_norm": 1.1679621934890747,
"learning_rate": 1.2136006466698362e-05,
"loss": 1.0688567161560059,
"step": 238
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.6078269481658936,
"learning_rate": 1.2127064752090001e-05,
"loss": 0.6071195602416992,
"step": 240
},
{
"epoch": 0.3265856950067476,
"grad_norm": 4.92821741104126,
"learning_rate": 1.211801830029793e-05,
"loss": 0.757681131362915,
"step": 242
},
{
"epoch": 0.3292847503373819,
"grad_norm": 1.6710680723190308,
"learning_rate": 1.2108867291727363e-05,
"loss": 0.9659180045127869,
"step": 244
},
{
"epoch": 0.3319838056680162,
"grad_norm": 2.416759729385376,
"learning_rate": 1.2099611908868602e-05,
"loss": 0.9776782393455505,
"step": 246
},
{
"epoch": 0.33468286099865047,
"grad_norm": 1.6332911252975464,
"learning_rate": 1.2090252336293388e-05,
"loss": 0.9600896835327148,
"step": 248
},
{
"epoch": 0.33738191632928477,
"grad_norm": 5.349152088165283,
"learning_rate": 1.2080788760651221e-05,
"loss": 0.5158158540725708,
"step": 250
},
{
"epoch": 0.340080971659919,
"grad_norm": 31.22907066345215,
"learning_rate": 1.2071221370665644e-05,
"loss": 0.6257212162017822,
"step": 252
},
{
"epoch": 0.3427800269905533,
"grad_norm": 2.575179100036621,
"learning_rate": 1.2061550357130476e-05,
"loss": 1.0216875076293945,
"step": 254
},
{
"epoch": 0.34547908232118757,
"grad_norm": 2.1095643043518066,
"learning_rate": 1.2051775912906002e-05,
"loss": 0.6756740808486938,
"step": 256
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.6237268447875977,
"learning_rate": 1.2041898232915136e-05,
"loss": 0.7705059051513672,
"step": 258
},
{
"epoch": 0.3508771929824561,
"grad_norm": 1.3587663173675537,
"learning_rate": 1.203191751413953e-05,
"loss": 0.9567830562591553,
"step": 260
},
{
"epoch": 0.3535762483130904,
"grad_norm": 2.174368143081665,
"learning_rate": 1.2021833955615643e-05,
"loss": 0.5454485416412354,
"step": 262
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.7718565464019775,
"learning_rate": 1.2011647758430774e-05,
"loss": 0.971337080001831,
"step": 264
},
{
"epoch": 0.358974358974359,
"grad_norm": 1.3361896276474,
"learning_rate": 1.2001359125719057e-05,
"loss": 0.9904292821884155,
"step": 266
},
{
"epoch": 0.3616734143049933,
"grad_norm": 3.2935264110565186,
"learning_rate": 1.1990968262657397e-05,
"loss": 1.0136817693710327,
"step": 268
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.8393303155899048,
"learning_rate": 1.1980475376461393e-05,
"loss": 0.07791149616241455,
"step": 270
},
{
"epoch": 0.36707152496626183,
"grad_norm": 2.3149139881134033,
"learning_rate": 1.1969880676381196e-05,
"loss": 0.9124975800514221,
"step": 272
},
{
"epoch": 0.3697705802968961,
"grad_norm": 1.0495190620422363,
"learning_rate": 1.1959184373697345e-05,
"loss": 1.007908582687378,
"step": 274
},
{
"epoch": 0.3724696356275304,
"grad_norm": 4.054476261138916,
"learning_rate": 1.1948386681716538e-05,
"loss": 0.5559310913085938,
"step": 276
},
{
"epoch": 0.37516869095816463,
"grad_norm": 1.1862071752548218,
"learning_rate": 1.1937487815767397e-05,
"loss": 0.7214630842208862,
"step": 278
},
{
"epoch": 0.37786774628879893,
"grad_norm": 0.8050886988639832,
"learning_rate": 1.1926487993196163e-05,
"loss": 0.7162290215492249,
"step": 280
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.8456649780273438,
"learning_rate": 1.191538743336236e-05,
"loss": 0.8286449909210205,
"step": 282
},
{
"epoch": 0.3832658569500675,
"grad_norm": 12.547414779663086,
"learning_rate": 1.1904186357634428e-05,
"loss": 0.35176369547843933,
"step": 284
},
{
"epoch": 0.38596491228070173,
"grad_norm": 1.5508333444595337,
"learning_rate": 1.1892884989385298e-05,
"loss": 0.9995942711830139,
"step": 286
},
{
"epoch": 0.38866396761133604,
"grad_norm": 3.0489025115966797,
"learning_rate": 1.1881483553987962e-05,
"loss": 0.41472408175468445,
"step": 288
},
{
"epoch": 0.3913630229419703,
"grad_norm": 18.13665199279785,
"learning_rate": 1.1869982278810934e-05,
"loss": 0.7451214790344238,
"step": 290
},
{
"epoch": 0.3940620782726046,
"grad_norm": 1.6674484014511108,
"learning_rate": 1.185838139321377e-05,
"loss": 0.6856521964073181,
"step": 292
},
{
"epoch": 0.3967611336032389,
"grad_norm": 5.430974960327148,
"learning_rate": 1.1846681128542443e-05,
"loss": 0.8595588207244873,
"step": 294
},
{
"epoch": 0.39946018893387314,
"grad_norm": 1.4392977952957153,
"learning_rate": 1.183488171812477e-05,
"loss": 0.7703130841255188,
"step": 296
},
{
"epoch": 0.40215924426450744,
"grad_norm": 1.7315428256988525,
"learning_rate": 1.1822983397265734e-05,
"loss": 0.9766842126846313,
"step": 298
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.4848453998565674,
"learning_rate": 1.18109864032428e-05,
"loss": 0.904464840888977,
"step": 300
},
{
"epoch": 0.407557354925776,
"grad_norm": 1.4237099885940552,
"learning_rate": 1.1798890975301189e-05,
"loss": 1.0174388885498047,
"step": 302
},
{
"epoch": 0.41025641025641024,
"grad_norm": 1.5580227375030518,
"learning_rate": 1.178669735464909e-05,
"loss": 1.038877248764038,
"step": 304
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.6760661602020264,
"learning_rate": 1.177440578445287e-05,
"loss": 0.5212962031364441,
"step": 306
},
{
"epoch": 0.4156545209176788,
"grad_norm": 1.1963051557540894,
"learning_rate": 1.176201650983221e-05,
"loss": 0.7233787775039673,
"step": 308
},
{
"epoch": 0.4183535762483131,
"grad_norm": 1.6447457075119019,
"learning_rate": 1.1749529777855224e-05,
"loss": 1.0451695919036865,
"step": 310
},
{
"epoch": 0.42105263157894735,
"grad_norm": 307.018798828125,
"learning_rate": 1.1736945837533531e-05,
"loss": 0.7426882982254028,
"step": 312
},
{
"epoch": 0.42375168690958165,
"grad_norm": 1.5677286386489868,
"learning_rate": 1.1724264939817287e-05,
"loss": 0.9855917692184448,
"step": 314
},
{
"epoch": 0.42645074224021595,
"grad_norm": 1.4524308443069458,
"learning_rate": 1.171148733759018e-05,
"loss": 1.0825656652450562,
"step": 316
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.332364797592163,
"learning_rate": 1.1698613285664393e-05,
"loss": 0.21006017923355103,
"step": 318
},
{
"epoch": 0.4318488529014845,
"grad_norm": 1.1307263374328613,
"learning_rate": 1.1685643040775515e-05,
"loss": 0.4775347113609314,
"step": 320
},
{
"epoch": 0.43454790823211875,
"grad_norm": 5.774119853973389,
"learning_rate": 1.1672576861577423e-05,
"loss": 1.05266273021698,
"step": 322
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.3685437440872192,
"learning_rate": 1.165941500863713e-05,
"loss": 1.074377417564392,
"step": 324
},
{
"epoch": 0.4399460188933873,
"grad_norm": 1.943925142288208,
"learning_rate": 1.1646157744429583e-05,
"loss": 0.9657345414161682,
"step": 326
},
{
"epoch": 0.4426450742240216,
"grad_norm": 1.660365343093872,
"learning_rate": 1.1632805333332424e-05,
"loss": 0.9795082807540894,
"step": 328
},
{
"epoch": 0.44534412955465585,
"grad_norm": 7.672799587249756,
"learning_rate": 1.1619358041620728e-05,
"loss": 0.16447825729846954,
"step": 330
},
{
"epoch": 0.44804318488529016,
"grad_norm": 1.6035205125808716,
"learning_rate": 1.1605816137461693e-05,
"loss": 1.0562779903411865,
"step": 332
},
{
"epoch": 0.4507422402159244,
"grad_norm": 1.6066842079162598,
"learning_rate": 1.1592179890909276e-05,
"loss": 0.5840043425559998,
"step": 334
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.5669262409210205,
"learning_rate": 1.157844957389883e-05,
"loss": 0.9664932489395142,
"step": 336
},
{
"epoch": 0.45614035087719296,
"grad_norm": 1.8515318632125854,
"learning_rate": 1.1564625460241665e-05,
"loss": 0.9272164702415466,
"step": 338
},
{
"epoch": 0.45883940620782726,
"grad_norm": 1.9376684427261353,
"learning_rate": 1.1550707825619594e-05,
"loss": 0.9817652702331543,
"step": 340
},
{
"epoch": 0.46153846153846156,
"grad_norm": 5.883458137512207,
"learning_rate": 1.1536696947579434e-05,
"loss": 0.6517050266265869,
"step": 342
},
{
"epoch": 0.4642375168690958,
"grad_norm": 1.574130654335022,
"learning_rate": 1.1522593105527472e-05,
"loss": 0.7160367369651794,
"step": 344
},
{
"epoch": 0.4669365721997301,
"grad_norm": 1.5215942859649658,
"learning_rate": 1.1508396580723888e-05,
"loss": 0.9305225014686584,
"step": 346
},
{
"epoch": 0.46963562753036436,
"grad_norm": 10.74595832824707,
"learning_rate": 1.1494107656277158e-05,
"loss": 0.23292149603366852,
"step": 348
},
{
"epoch": 0.47233468286099867,
"grad_norm": 1.3209294080734253,
"learning_rate": 1.1479726617138396e-05,
"loss": 0.5377740859985352,
"step": 350
},
{
"epoch": 0.4750337381916329,
"grad_norm": 5.547281265258789,
"learning_rate": 1.1465253750095676e-05,
"loss": 0.5457713007926941,
"step": 352
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.200194239616394,
"learning_rate": 1.145068934376831e-05,
"loss": 1.032280445098877,
"step": 354
},
{
"epoch": 0.48043184885290147,
"grad_norm": 1.527847409248352,
"learning_rate": 1.1436033688601103e-05,
"loss": 0.7377746105194092,
"step": 356
},
{
"epoch": 0.48313090418353577,
"grad_norm": 2.019571304321289,
"learning_rate": 1.142128707685855e-05,
"loss": 1.0082802772521973,
"step": 358
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.2335315942764282,
"learning_rate": 1.1406449802619007e-05,
"loss": 1.0848407745361328,
"step": 360
},
{
"epoch": 0.4885290148448043,
"grad_norm": 4.238769054412842,
"learning_rate": 1.1391522161768835e-05,
"loss": 0.9359495043754578,
"step": 362
},
{
"epoch": 0.49122807017543857,
"grad_norm": 7.280695915222168,
"learning_rate": 1.137650445199649e-05,
"loss": 0.7889329791069031,
"step": 364
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.1879907846450806,
"learning_rate": 1.13613969727866e-05,
"loss": 0.9769068956375122,
"step": 366
},
{
"epoch": 0.4966261808367072,
"grad_norm": 2.0206735134124756,
"learning_rate": 1.1346200025413972e-05,
"loss": 0.8470125198364258,
"step": 368
},
{
"epoch": 0.4993252361673414,
"grad_norm": 1.0538994073867798,
"learning_rate": 1.1330913912937602e-05,
"loss": 1.0040283203125,
"step": 370
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.0824867486953735,
"learning_rate": 1.131553894019463e-05,
"loss": 1.0739049911499023,
"step": 372
},
{
"epoch": 0.50472334682861,
"grad_norm": 2.1606943607330322,
"learning_rate": 1.1300075413794247e-05,
"loss": 0.9793626070022583,
"step": 374
},
{
"epoch": 0.5074224021592443,
"grad_norm": 1.1736167669296265,
"learning_rate": 1.1284523642111592e-05,
"loss": 0.7736898064613342,
"step": 376
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.4421758651733398,
"learning_rate": 1.1268883935281603e-05,
"loss": 0.9196012020111084,
"step": 378
},
{
"epoch": 0.5128205128205128,
"grad_norm": 1.8012371063232422,
"learning_rate": 1.1253156605192829e-05,
"loss": 0.7796334028244019,
"step": 380
},
{
"epoch": 0.5155195681511471,
"grad_norm": 1.5929791927337646,
"learning_rate": 1.1237341965481207e-05,
"loss": 1.0640032291412354,
"step": 382
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.9453246593475342,
"learning_rate": 1.1221440331523815e-05,
"loss": 0.9769595861434937,
"step": 384
},
{
"epoch": 0.5209176788124157,
"grad_norm": 10.83289623260498,
"learning_rate": 1.1205452020432574e-05,
"loss": 0.5430967211723328,
"step": 386
},
{
"epoch": 0.5236167341430499,
"grad_norm": 4.249309062957764,
"learning_rate": 1.1189377351047935e-05,
"loss": 0.5993819832801819,
"step": 388
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.6194641590118408,
"learning_rate": 1.1173216643932509e-05,
"loss": 1.113552451133728,
"step": 390
},
{
"epoch": 0.5290148448043185,
"grad_norm": 1.8865735530853271,
"learning_rate": 1.1156970221364681e-05,
"loss": 0.9687294960021973,
"step": 392
},
{
"epoch": 0.5317139001349528,
"grad_norm": 1.3079495429992676,
"learning_rate": 1.1140638407332185e-05,
"loss": 0.5798075199127197,
"step": 394
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.5965051651000977,
"learning_rate": 1.1124221527525635e-05,
"loss": 1.0805583000183105,
"step": 396
},
{
"epoch": 0.5371120107962213,
"grad_norm": 1.7013156414031982,
"learning_rate": 1.1107719909332037e-05,
"loss": 1.0401415824890137,
"step": 398
},
{
"epoch": 0.5398110661268556,
"grad_norm": 2.132007360458374,
"learning_rate": 1.1091133881828261e-05,
"loss": 0.8972035646438599,
"step": 400
},
{
"epoch": 0.5425101214574899,
"grad_norm": 11.097935676574707,
"learning_rate": 1.1074463775774472e-05,
"loss": 0.5467524528503418,
"step": 402
},
{
"epoch": 0.5452091767881242,
"grad_norm": 1.6173795461654663,
"learning_rate": 1.1057709923607537e-05,
"loss": 0.955978274345398,
"step": 404
},
{
"epoch": 0.5479082321187584,
"grad_norm": 2.0655932426452637,
"learning_rate": 1.1040872659434402e-05,
"loss": 0.6293669939041138,
"step": 406
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.8929414749145508,
"learning_rate": 1.1023952319025414e-05,
"loss": 0.9048166871070862,
"step": 408
},
{
"epoch": 0.553306342780027,
"grad_norm": 1.0451692342758179,
"learning_rate": 1.1006949239807643e-05,
"loss": 1.1323614120483398,
"step": 410
},
{
"epoch": 0.5560053981106613,
"grad_norm": 6.36082649230957,
"learning_rate": 1.0989863760858138e-05,
"loss": 0.9985231757164001,
"step": 412
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.1424140930175781,
"learning_rate": 1.0972696222897176e-05,
"loss": 0.9446185827255249,
"step": 414
},
{
"epoch": 0.5614035087719298,
"grad_norm": 3.7635855674743652,
"learning_rate": 1.0955446968281463e-05,
"loss": 0.8581997156143188,
"step": 416
},
{
"epoch": 0.5641025641025641,
"grad_norm": 2.5408518314361572,
"learning_rate": 1.0938116340997299e-05,
"loss": 0.8878166675567627,
"step": 418
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.8976860046386719,
"learning_rate": 1.0920704686653735e-05,
"loss": 0.6632760167121887,
"step": 420
},
{
"epoch": 0.5695006747638327,
"grad_norm": 4.141200065612793,
"learning_rate": 1.0903212352475665e-05,
"loss": 0.7282555103302002,
"step": 422
},
{
"epoch": 0.5721997300944669,
"grad_norm": 1.912293553352356,
"learning_rate": 1.0885639687296914e-05,
"loss": 0.9195336699485779,
"step": 424
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.864601492881775,
"learning_rate": 1.0867987041553269e-05,
"loss": 0.9130004644393921,
"step": 426
},
{
"epoch": 0.5775978407557355,
"grad_norm": 4.861878395080566,
"learning_rate": 1.0850254767275501e-05,
"loss": 1.0303194522857666,
"step": 428
},
{
"epoch": 0.5802968960863698,
"grad_norm": 1.301428198814392,
"learning_rate": 1.083244321808234e-05,
"loss": 1.012520432472229,
"step": 430
},
{
"epoch": 0.582995951417004,
"grad_norm": 2.791686534881592,
"learning_rate": 1.0814552749173427e-05,
"loss": 0.86846923828125,
"step": 432
},
{
"epoch": 0.5856950067476383,
"grad_norm": 3.153118133544922,
"learning_rate": 1.0796583717322221e-05,
"loss": 0.7495797872543335,
"step": 434
},
{
"epoch": 0.5883940620782726,
"grad_norm": 1.4211339950561523,
"learning_rate": 1.0778536480868895e-05,
"loss": 0.4938827157020569,
"step": 436
},
{
"epoch": 0.5910931174089069,
"grad_norm": 2.191868782043457,
"learning_rate": 1.0760411399713189e-05,
"loss": 1.0031588077545166,
"step": 438
},
{
"epoch": 0.5937921727395412,
"grad_norm": 1.890841007232666,
"learning_rate": 1.0742208835307222e-05,
"loss": 0.7012419104576111,
"step": 440
},
{
"epoch": 0.5964912280701754,
"grad_norm": 5.149887561798096,
"learning_rate": 1.0723929150648296e-05,
"loss": 0.6346755027770996,
"step": 442
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.7442600727081299,
"learning_rate": 1.0705572710271656e-05,
"loss": 0.9421064853668213,
"step": 444
},
{
"epoch": 0.601889338731444,
"grad_norm": 1.8889329433441162,
"learning_rate": 1.0687139880243208e-05,
"loss": 0.8577780723571777,
"step": 446
},
{
"epoch": 0.6045883940620783,
"grad_norm": 1.529708743095398,
"learning_rate": 1.0668631028152236e-05,
"loss": 1.0774247646331787,
"step": 448
},
{
"epoch": 0.6072874493927125,
"grad_norm": 5.384032726287842,
"learning_rate": 1.0650046523104059e-05,
"loss": 0.6740553379058838,
"step": 450
},
{
"epoch": 0.6099865047233468,
"grad_norm": 1.94712233543396,
"learning_rate": 1.0631386735712675e-05,
"loss": 0.9021118879318237,
"step": 452
},
{
"epoch": 0.6126855600539811,
"grad_norm": 2.5946881771087646,
"learning_rate": 1.0612652038093372e-05,
"loss": 1.0916223526000977,
"step": 454
},
{
"epoch": 0.6153846153846154,
"grad_norm": 4.924251556396484,
"learning_rate": 1.0593842803855302e-05,
"loss": 0.6892099380493164,
"step": 456
},
{
"epoch": 0.6180836707152496,
"grad_norm": 1.7946263551712036,
"learning_rate": 1.0574959408094039e-05,
"loss": 0.9302884340286255,
"step": 458
},
{
"epoch": 0.6207827260458839,
"grad_norm": 3.0235397815704346,
"learning_rate": 1.0556002227384083e-05,
"loss": 0.9852180480957031,
"step": 460
},
{
"epoch": 0.6234817813765182,
"grad_norm": 2.6003530025482178,
"learning_rate": 1.0536971639771371e-05,
"loss": 1.0105475187301636,
"step": 462
},
{
"epoch": 0.6261808367071525,
"grad_norm": 1.9353368282318115,
"learning_rate": 1.0517868024765723e-05,
"loss": 0.53856360912323,
"step": 464
},
{
"epoch": 0.6288798920377868,
"grad_norm": 1.5394912958145142,
"learning_rate": 1.0498691763333278e-05,
"loss": 0.7180197238922119,
"step": 466
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.3707131147384644,
"learning_rate": 1.0479443237888898e-05,
"loss": 0.8788055181503296,
"step": 468
},
{
"epoch": 0.6342780026990553,
"grad_norm": 2.12168288230896,
"learning_rate": 1.0460122832288538e-05,
"loss": 0.9377783536911011,
"step": 470
},
{
"epoch": 0.6369770580296896,
"grad_norm": 1.5926860570907593,
"learning_rate": 1.0440730931821603e-05,
"loss": 0.9266502857208252,
"step": 472
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.3771506547927856,
"learning_rate": 1.0421267923203245e-05,
"loss": 0.8251262903213501,
"step": 474
},
{
"epoch": 0.6423751686909581,
"grad_norm": 1.7435166835784912,
"learning_rate": 1.0401734194566667e-05,
"loss": 1.051794171333313,
"step": 476
},
{
"epoch": 0.6450742240215924,
"grad_norm": 3.4853711128234863,
"learning_rate": 1.0382130135455377e-05,
"loss": 0.9569863080978394,
"step": 478
},
{
"epoch": 0.6477732793522267,
"grad_norm": 6.124602794647217,
"learning_rate": 1.0362456136815419e-05,
"loss": 0.6598426103591919,
"step": 480
},
{
"epoch": 0.650472334682861,
"grad_norm": 3.28501033782959,
"learning_rate": 1.0342712590987575e-05,
"loss": 0.9515646696090698,
"step": 482
},
{
"epoch": 0.6531713900134952,
"grad_norm": 1.8997509479522705,
"learning_rate": 1.0322899891699554e-05,
"loss": 0.7211359739303589,
"step": 484
},
{
"epoch": 0.6558704453441295,
"grad_norm": 2.0254898071289062,
"learning_rate": 1.030301843405812e-05,
"loss": 0.8057195544242859,
"step": 486
},
{
"epoch": 0.6585695006747638,
"grad_norm": 7.1337480545043945,
"learning_rate": 1.0283068614541222e-05,
"loss": 0.6165438890457153,
"step": 488
},
{
"epoch": 0.6612685560053981,
"grad_norm": 1.6425862312316895,
"learning_rate": 1.0263050830990097e-05,
"loss": 1.0961238145828247,
"step": 490
},
{
"epoch": 0.6639676113360324,
"grad_norm": 4.242962837219238,
"learning_rate": 1.0242965482601322e-05,
"loss": 0.21474683284759521,
"step": 492
},
{
"epoch": 0.6666666666666666,
"grad_norm": 1.5638395547866821,
"learning_rate": 1.0222812969918858e-05,
"loss": 0.9058488011360168,
"step": 494
},
{
"epoch": 0.6693657219973009,
"grad_norm": 6.208765506744385,
"learning_rate": 1.0202593694826065e-05,
"loss": 0.39533284306526184,
"step": 496
},
{
"epoch": 0.6720647773279352,
"grad_norm": 1.744771957397461,
"learning_rate": 1.0182308060537681e-05,
"loss": 1.0490394830703735,
"step": 498
},
{
"epoch": 0.6747638326585695,
"grad_norm": 1.165880560874939,
"learning_rate": 1.016195647159179e-05,
"loss": 1.1421074867248535,
"step": 500
},
{
"epoch": 0.6774628879892037,
"grad_norm": 3.3272783756256104,
"learning_rate": 1.0141539333841752e-05,
"loss": 0.9005064964294434,
"step": 502
},
{
"epoch": 0.680161943319838,
"grad_norm": 2.1412487030029297,
"learning_rate": 1.0121057054448099e-05,
"loss": 0.6484662294387817,
"step": 504
},
{
"epoch": 0.6828609986504723,
"grad_norm": 1.8351943492889404,
"learning_rate": 1.0100510041870434e-05,
"loss": 0.9145583510398865,
"step": 506
},
{
"epoch": 0.6855600539811066,
"grad_norm": 1.9452043771743774,
"learning_rate": 1.0079898705859268e-05,
"loss": 1.101776123046875,
"step": 508
},
{
"epoch": 0.6882591093117408,
"grad_norm": 1.2626560926437378,
"learning_rate": 1.0059223457447861e-05,
"loss": 1.0483142137527466,
"step": 510
},
{
"epoch": 0.6909581646423751,
"grad_norm": 0.7712414860725403,
"learning_rate": 1.0038484708944019e-05,
"loss": 0.545906662940979,
"step": 512
},
{
"epoch": 0.6936572199730094,
"grad_norm": 1.8449019193649292,
"learning_rate": 1.001768287392187e-05,
"loss": 0.5840798616409302,
"step": 514
},
{
"epoch": 0.6963562753036437,
"grad_norm": 1.3261218070983887,
"learning_rate": 9.996818367213623e-06,
"loss": 0.8815082311630249,
"step": 516
},
{
"epoch": 0.699055330634278,
"grad_norm": 2.2945189476013184,
"learning_rate": 9.97589160490129e-06,
"loss": 0.9285835027694702,
"step": 518
},
{
"epoch": 0.7017543859649122,
"grad_norm": 8.204078674316406,
"learning_rate": 9.95490300430839e-06,
"loss": 0.593896746635437,
"step": 520
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.5768344402313232,
"learning_rate": 9.933852983991631e-06,
"loss": 1.00489342212677,
"step": 522
},
{
"epoch": 0.7071524966261808,
"grad_norm": 3.006584405899048,
"learning_rate": 9.91274196373255e-06,
"loss": 1.0357744693756104,
"step": 524
},
{
"epoch": 0.7098515519568152,
"grad_norm": 6.199564456939697,
"learning_rate": 9.891570364529163e-06,
"loss": 0.642838716506958,
"step": 526
},
{
"epoch": 0.7125506072874493,
"grad_norm": 1.7646253108978271,
"learning_rate": 9.87033860858755e-06,
"loss": 0.9927930235862732,
"step": 528
},
{
"epoch": 0.7152496626180836,
"grad_norm": 4.5279221534729,
"learning_rate": 9.849047119313439e-06,
"loss": 0.2820289433002472,
"step": 530
},
{
"epoch": 0.717948717948718,
"grad_norm": 1.3871742486953735,
"learning_rate": 9.827696321303773e-06,
"loss": 0.9865570068359375,
"step": 532
},
{
"epoch": 0.7206477732793523,
"grad_norm": 5.380123615264893,
"learning_rate": 9.806286640338232e-06,
"loss": 0.23654060065746307,
"step": 534
},
{
"epoch": 0.7233468286099866,
"grad_norm": 1.4577468633651733,
"learning_rate": 9.78481850337075e-06,
"loss": 0.7747691869735718,
"step": 536
},
{
"epoch": 0.7260458839406208,
"grad_norm": 3.0036792755126953,
"learning_rate": 9.763292338520986e-06,
"loss": 0.9602346420288086,
"step": 538
},
{
"epoch": 0.728744939271255,
"grad_norm": 1.4889585971832275,
"learning_rate": 9.741708575065808e-06,
"loss": 0.5133143663406372,
"step": 540
},
{
"epoch": 0.7314439946018894,
"grad_norm": 7.9594268798828125,
"learning_rate": 9.720067643430716e-06,
"loss": 0.310360312461853,
"step": 542
},
{
"epoch": 0.7341430499325237,
"grad_norm": 2.530517578125,
"learning_rate": 9.698369975181262e-06,
"loss": 1.0847041606903076,
"step": 544
},
{
"epoch": 0.7368421052631579,
"grad_norm": 7.523575305938721,
"learning_rate": 9.676616003014447e-06,
"loss": 0.5520514249801636,
"step": 546
},
{
"epoch": 0.7395411605937922,
"grad_norm": 16.927391052246094,
"learning_rate": 9.65480616075009e-06,
"loss": 0.6531012058258057,
"step": 548
},
{
"epoch": 0.7422402159244265,
"grad_norm": 1.8876800537109375,
"learning_rate": 9.632940883322178e-06,
"loss": 0.9499740600585938,
"step": 550
},
{
"epoch": 0.7449392712550608,
"grad_norm": 1.4345674514770508,
"learning_rate": 9.611020606770188e-06,
"loss": 0.5219556093215942,
"step": 552
},
{
"epoch": 0.747638326585695,
"grad_norm": 5.749211311340332,
"learning_rate": 9.5890457682304e-06,
"loss": 0.6242225170135498,
"step": 554
},
{
"epoch": 0.7503373819163293,
"grad_norm": 2.7486813068389893,
"learning_rate": 9.567016805927166e-06,
"loss": 0.6217697858810425,
"step": 556
},
{
"epoch": 0.7530364372469636,
"grad_norm": 1.5207133293151855,
"learning_rate": 9.544934159164188e-06,
"loss": 1.0374534130096436,
"step": 558
},
{
"epoch": 0.7557354925775979,
"grad_norm": 6.610947132110596,
"learning_rate": 9.522798268315749e-06,
"loss": 0.6023949384689331,
"step": 560
},
{
"epoch": 0.7584345479082322,
"grad_norm": 1.1763848066329956,
"learning_rate": 9.500609574817922e-06,
"loss": 0.693153440952301,
"step": 562
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.46580708026885986,
"learning_rate": 9.478368521159782e-06,
"loss": 0.1448395848274231,
"step": 564
},
{
"epoch": 0.7638326585695007,
"grad_norm": 2.9862771034240723,
"learning_rate": 9.456075550874576e-06,
"loss": 0.6183888912200928,
"step": 566
},
{
"epoch": 0.766531713900135,
"grad_norm": 63.45586013793945,
"learning_rate": 9.433731108530874e-06,
"loss": 0.9780744314193726,
"step": 568
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.2516928911209106,
"learning_rate": 9.411335639723707e-06,
"loss": 1.024972677230835,
"step": 570
},
{
"epoch": 0.7719298245614035,
"grad_norm": 1.1070282459259033,
"learning_rate": 9.388889591065685e-06,
"loss": 0.9459818601608276,
"step": 572
},
{
"epoch": 0.7746288798920378,
"grad_norm": 1.320272445678711,
"learning_rate": 9.36639341017808e-06,
"loss": 1.0482739210128784,
"step": 574
},
{
"epoch": 0.7773279352226721,
"grad_norm": 1.0126550197601318,
"learning_rate": 9.343847545681912e-06,
"loss": 0.7617048025131226,
"step": 576
},
{
"epoch": 0.7800269905533064,
"grad_norm": 1.111310362815857,
"learning_rate": 9.321252447188992e-06,
"loss": 1.0313292741775513,
"step": 578
},
{
"epoch": 0.7827260458839406,
"grad_norm": 2.3641977310180664,
"learning_rate": 9.298608565292958e-06,
"loss": 0.8116118311882019,
"step": 580
},
{
"epoch": 0.7854251012145749,
"grad_norm": 3.368452548980713,
"learning_rate": 9.275916351560302e-06,
"loss": 0.85608971118927,
"step": 582
},
{
"epoch": 0.7881241565452092,
"grad_norm": 4.553686141967773,
"learning_rate": 9.25317625852134e-06,
"loss": 0.1353357583284378,
"step": 584
},
{
"epoch": 0.7908232118758435,
"grad_norm": 1.8922858238220215,
"learning_rate": 9.230388739661216e-06,
"loss": 1.093623399734497,
"step": 586
},
{
"epoch": 0.7935222672064778,
"grad_norm": 1.1269725561141968,
"learning_rate": 9.207554249410829e-06,
"loss": 0.7518768310546875,
"step": 588
},
{
"epoch": 0.796221322537112,
"grad_norm": 1.5451685190200806,
"learning_rate": 9.184673243137799e-06,
"loss": 0.9263067245483398,
"step": 590
},
{
"epoch": 0.7989203778677463,
"grad_norm": 1.5247893333435059,
"learning_rate": 9.161746177137366e-06,
"loss": 0.8519529104232788,
"step": 592
},
{
"epoch": 0.8016194331983806,
"grad_norm": 2.638950824737549,
"learning_rate": 9.138773508623304e-06,
"loss": 1.0825176239013672,
"step": 594
},
{
"epoch": 0.8043184885290149,
"grad_norm": 2.3280320167541504,
"learning_rate": 9.115755695718785e-06,
"loss": 1.038777470588684,
"step": 596
},
{
"epoch": 0.8070175438596491,
"grad_norm": 1.8069512844085693,
"learning_rate": 9.092693197447267e-06,
"loss": 0.9702070951461792,
"step": 598
},
{
"epoch": 0.8097165991902834,
"grad_norm": 5.8682379722595215,
"learning_rate": 9.069586473723325e-06,
"loss": 0.5837458372116089,
"step": 600
},
{
"epoch": 0.8124156545209177,
"grad_norm": 2.862755298614502,
"learning_rate": 9.046435985343476e-06,
"loss": 0.8376547694206238,
"step": 602
},
{
"epoch": 0.815114709851552,
"grad_norm": 1.6289156675338745,
"learning_rate": 9.023242193977e-06,
"loss": 0.8585331439971924,
"step": 604
},
{
"epoch": 0.8178137651821862,
"grad_norm": 2.9280383586883545,
"learning_rate": 9.000005562156733e-06,
"loss": 0.9037019610404968,
"step": 606
},
{
"epoch": 0.8205128205128205,
"grad_norm": 1.8420056104660034,
"learning_rate": 8.976726553269837e-06,
"loss": 0.9938711524009705,
"step": 608
},
{
"epoch": 0.8232118758434548,
"grad_norm": 1.1752334833145142,
"learning_rate": 8.953405631548557e-06,
"loss": 1.0280206203460693,
"step": 610
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.6784051656723022,
"learning_rate": 8.930043262060974e-06,
"loss": 0.8855659365653992,
"step": 612
},
{
"epoch": 0.8286099865047234,
"grad_norm": 2.5772430896759033,
"learning_rate": 8.906639910701719e-06,
"loss": 0.9628806710243225,
"step": 614
},
{
"epoch": 0.8313090418353576,
"grad_norm": 1.87814462184906,
"learning_rate": 8.883196044182693e-06,
"loss": 0.56829833984375,
"step": 616
},
{
"epoch": 0.8340080971659919,
"grad_norm": 2.191746950149536,
"learning_rate": 8.859712130023752e-06,
"loss": 0.9893323183059692,
"step": 618
},
{
"epoch": 0.8367071524966262,
"grad_norm": 1.060782790184021,
"learning_rate": 8.836188636543382e-06,
"loss": 0.7178077697753906,
"step": 620
},
{
"epoch": 0.8394062078272605,
"grad_norm": 1.4658266305923462,
"learning_rate": 8.812626032849373e-06,
"loss": 0.5496494770050049,
"step": 622
},
{
"epoch": 0.8421052631578947,
"grad_norm": 1.6255152225494385,
"learning_rate": 8.789024788829443e-06,
"loss": 0.6712936162948608,
"step": 624
},
{
"epoch": 0.844804318488529,
"grad_norm": 1.417540192604065,
"learning_rate": 8.765385375141892e-06,
"loss": 0.7042181491851807,
"step": 626
},
{
"epoch": 0.8475033738191633,
"grad_norm": 1.9119606018066406,
"learning_rate": 8.741708263206186e-06,
"loss": 1.062172532081604,
"step": 628
},
{
"epoch": 0.8502024291497976,
"grad_norm": 7.317932605743408,
"learning_rate": 8.717993925193591e-06,
"loss": 0.7059786915779114,
"step": 630
},
{
"epoch": 0.8529014844804319,
"grad_norm": 1.9727838039398193,
"learning_rate": 8.694242834017729e-06,
"loss": 1.1658766269683838,
"step": 632
},
{
"epoch": 0.8556005398110661,
"grad_norm": 2.5640006065368652,
"learning_rate": 8.670455463325162e-06,
"loss": 0.8735545873641968,
"step": 634
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.479243516921997,
"learning_rate": 8.646632287485932e-06,
"loss": 0.8225340843200684,
"step": 636
},
{
"epoch": 0.8609986504723347,
"grad_norm": 1.6788865327835083,
"learning_rate": 8.622773781584127e-06,
"loss": 0.9871320724487305,
"step": 638
},
{
"epoch": 0.863697705802969,
"grad_norm": 4.9331440925598145,
"learning_rate": 8.598880421408375e-06,
"loss": 0.24402402341365814,
"step": 640
},
{
"epoch": 0.8663967611336032,
"grad_norm": 1.409420132637024,
"learning_rate": 8.57495268344238e-06,
"loss": 1.0236334800720215,
"step": 642
},
{
"epoch": 0.8690958164642375,
"grad_norm": 2.8709068298339844,
"learning_rate": 8.550991044855408e-06,
"loss": 0.9956761598587036,
"step": 644
},
{
"epoch": 0.8717948717948718,
"grad_norm": 1.2389776706695557,
"learning_rate": 8.526995983492777e-06,
"loss": 0.9422023296356201,
"step": 646
},
{
"epoch": 0.8744939271255061,
"grad_norm": 2.6241936683654785,
"learning_rate": 8.50296797786632e-06,
"loss": 0.9618534445762634,
"step": 648
},
{
"epoch": 0.8771929824561403,
"grad_norm": 5.383096218109131,
"learning_rate": 8.478907507144855e-06,
"loss": 0.5624386072158813,
"step": 650
},
{
"epoch": 0.8798920377867746,
"grad_norm": 2.1661858558654785,
"learning_rate": 8.454815051144614e-06,
"loss": 0.709771454334259,
"step": 652
},
{
"epoch": 0.8825910931174089,
"grad_norm": 2.7425076961517334,
"learning_rate": 8.430691090319692e-06,
"loss": 1.019305944442749,
"step": 654
},
{
"epoch": 0.8852901484480432,
"grad_norm": 1.5749717950820923,
"learning_rate": 8.406536105752446e-06,
"loss": 0.991958498954773,
"step": 656
},
{
"epoch": 0.8879892037786775,
"grad_norm": 1.0196173191070557,
"learning_rate": 8.382350579143922e-06,
"loss": 1.085745096206665,
"step": 658
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.9763017296791077,
"learning_rate": 8.35813499280423e-06,
"loss": 1.089850664138794,
"step": 660
},
{
"epoch": 0.893387314439946,
"grad_norm": 14.465669631958008,
"learning_rate": 8.333889829642934e-06,
"loss": 0.7980184555053711,
"step": 662
},
{
"epoch": 0.8960863697705803,
"grad_norm": 2.9542925357818604,
"learning_rate": 8.309615573159432e-06,
"loss": 0.3161945343017578,
"step": 664
},
{
"epoch": 0.8987854251012146,
"grad_norm": 3.6325623989105225,
"learning_rate": 8.285312707433288e-06,
"loss": 0.7513540983200073,
"step": 666
},
{
"epoch": 0.9014844804318488,
"grad_norm": 1.150533676147461,
"learning_rate": 8.260981717114611e-06,
"loss": 0.9773034453392029,
"step": 668
},
{
"epoch": 0.9041835357624831,
"grad_norm": 1.3201379776000977,
"learning_rate": 8.23662308741436e-06,
"loss": 1.034916639328003,
"step": 670
},
{
"epoch": 0.9068825910931174,
"grad_norm": 7.727103233337402,
"learning_rate": 8.21223730409469e-06,
"loss": 0.7840131521224976,
"step": 672
},
{
"epoch": 0.9095816464237517,
"grad_norm": 1.597216010093689,
"learning_rate": 8.187824853459248e-06,
"loss": 0.9697774648666382,
"step": 674
},
{
"epoch": 0.9122807017543859,
"grad_norm": 1.5465730428695679,
"learning_rate": 8.163386222343491e-06,
"loss": 0.9404412508010864,
"step": 676
},
{
"epoch": 0.9149797570850202,
"grad_norm": 1.401080846786499,
"learning_rate": 8.138921898104967e-06,
"loss": 1.0558454990386963,
"step": 678
},
{
"epoch": 0.9176788124156545,
"grad_norm": 6.414635181427002,
"learning_rate": 8.1144323686136e-06,
"loss": 0.33971545100212097,
"step": 680
},
{
"epoch": 0.9203778677462888,
"grad_norm": 1.2682753801345825,
"learning_rate": 8.08991812224195e-06,
"loss": 1.0396260023117065,
"step": 682
},
{
"epoch": 0.9230769230769231,
"grad_norm": 10.276638984680176,
"learning_rate": 8.0653796478555e-06,
"loss": 0.9345718622207642,
"step": 684
},
{
"epoch": 0.9257759784075573,
"grad_norm": 2.0107717514038086,
"learning_rate": 8.040817434802875e-06,
"loss": 1.0054938793182373,
"step": 686
},
{
"epoch": 0.9284750337381916,
"grad_norm": 1.338994026184082,
"learning_rate": 8.01623197290611e-06,
"loss": 1.0551596879959106,
"step": 688
},
{
"epoch": 0.9311740890688259,
"grad_norm": 3.6777682304382324,
"learning_rate": 7.991623752450866e-06,
"loss": 0.8525782823562622,
"step": 690
},
{
"epoch": 0.9338731443994602,
"grad_norm": 2.8609213829040527,
"learning_rate": 7.966993264176657e-06,
"loss": 0.8466967344284058,
"step": 692
},
{
"epoch": 0.9365721997300944,
"grad_norm": 1.2463239431381226,
"learning_rate": 7.94234099926706e-06,
"loss": 0.936937153339386,
"step": 694
},
{
"epoch": 0.9392712550607287,
"grad_norm": 1.7466425895690918,
"learning_rate": 7.917667449339935e-06,
"loss": 0.9347761869430542,
"step": 696
},
{
"epoch": 0.941970310391363,
"grad_norm": 1.8468271493911743,
"learning_rate": 7.892973106437598e-06,
"loss": 0.8284968137741089,
"step": 698
},
{
"epoch": 0.9446693657219973,
"grad_norm": 1.7733769416809082,
"learning_rate": 7.868258463017027e-06,
"loss": 0.9093437194824219,
"step": 700
},
{
"epoch": 0.9473684210526315,
"grad_norm": 1.6979347467422485,
"learning_rate": 7.843524011940038e-06,
"loss": 0.5854538679122925,
"step": 702
},
{
"epoch": 0.9500674763832658,
"grad_norm": 1.380608081817627,
"learning_rate": 7.818770246463445e-06,
"loss": 0.7273447513580322,
"step": 704
},
{
"epoch": 0.9527665317139001,
"grad_norm": 1.1277148723602295,
"learning_rate": 7.793997660229233e-06,
"loss": 1.00750732421875,
"step": 706
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.6885129809379578,
"learning_rate": 7.769206747254719e-06,
"loss": 0.18471550941467285,
"step": 708
},
{
"epoch": 0.9581646423751687,
"grad_norm": 3.316200017929077,
"learning_rate": 7.744398001922686e-06,
"loss": 0.9240825176239014,
"step": 710
},
{
"epoch": 0.9608636977058029,
"grad_norm": 4.04154109954834,
"learning_rate": 7.719571918971539e-06,
"loss": 0.41226115822792053,
"step": 712
},
{
"epoch": 0.9635627530364372,
"grad_norm": 1.1670745611190796,
"learning_rate": 7.694728993485423e-06,
"loss": 0.8245844841003418,
"step": 714
},
{
"epoch": 0.9662618083670715,
"grad_norm": 2.671957492828369,
"learning_rate": 7.669869720884367e-06,
"loss": 0.9521523714065552,
"step": 716
},
{
"epoch": 0.9689608636977058,
"grad_norm": 1.659239649772644,
"learning_rate": 7.64499459691439e-06,
"loss": 0.9625018835067749,
"step": 718
},
{
"epoch": 0.97165991902834,
"grad_norm": 4.6280364990234375,
"learning_rate": 7.620104117637621e-06,
"loss": 0.8236821293830872,
"step": 720
},
{
"epoch": 0.9743589743589743,
"grad_norm": 1.4543287754058838,
"learning_rate": 7.595198779422408e-06,
"loss": 1.03853178024292,
"step": 722
},
{
"epoch": 0.9770580296896086,
"grad_norm": 9.59180736541748,
"learning_rate": 7.570279078933419e-06,
"loss": 0.2856726050376892,
"step": 724
},
{
"epoch": 0.979757085020243,
"grad_norm": 4.740160942077637,
"learning_rate": 7.54534551312173e-06,
"loss": 0.1729520559310913,
"step": 726
},
{
"epoch": 0.9824561403508771,
"grad_norm": 1.3076441287994385,
"learning_rate": 7.520398579214927e-06,
"loss": 0.958693265914917,
"step": 728
},
{
"epoch": 0.9851551956815114,
"grad_norm": 1.226672887802124,
"learning_rate": 7.495438774707179e-06,
"loss": 0.8376245498657227,
"step": 730
},
{
"epoch": 0.9878542510121457,
"grad_norm": 1.606249451637268,
"learning_rate": 7.4704665973493254e-06,
"loss": 0.9087053537368774,
"step": 732
},
{
"epoch": 0.99055330634278,
"grad_norm": 3.589003801345825,
"learning_rate": 7.445482545138942e-06,
"loss": 0.8934133052825928,
"step": 734
},
{
"epoch": 0.9932523616734144,
"grad_norm": 1.0932425260543823,
"learning_rate": 7.420487116310418e-06,
"loss": 1.093599796295166,
"step": 736
},
{
"epoch": 0.9959514170040485,
"grad_norm": 1.64720618724823,
"learning_rate": 7.3954808093250175e-06,
"loss": 1.0401995182037354,
"step": 738
},
{
"epoch": 0.9986504723346828,
"grad_norm": 1.071671724319458,
"learning_rate": 7.37046412286093e-06,
"loss": 0.6273572444915771,
"step": 740
},
{
"epoch": 1.001349527665317,
"grad_norm": 1.4724849462509155,
"learning_rate": 7.34543755580334e-06,
"loss": 0.885528564453125,
"step": 742
},
{
"epoch": 1.0040485829959513,
"grad_norm": 1.077073097229004,
"learning_rate": 7.320401607234473e-06,
"loss": 0.7546196579933167,
"step": 744
},
{
"epoch": 1.0067476383265856,
"grad_norm": 1.7151180505752563,
"learning_rate": 7.295356776423637e-06,
"loss": 0.6727256774902344,
"step": 746
},
{
"epoch": 1.00944669365722,
"grad_norm": 7.7168073654174805,
"learning_rate": 7.270303562817274e-06,
"loss": 0.44044023752212524,
"step": 748
},
{
"epoch": 1.0121457489878543,
"grad_norm": 2.8954923152923584,
"learning_rate": 7.245242466028994e-06,
"loss": 0.8065706491470337,
"step": 750
},
{
"epoch": 1.0148448043184886,
"grad_norm": 3.5301990509033203,
"learning_rate": 7.220173985829614e-06,
"loss": 0.5840882062911987,
"step": 752
},
{
"epoch": 1.0175438596491229,
"grad_norm": 5.2072062492370605,
"learning_rate": 7.195098622137196e-06,
"loss": 0.4359647035598755,
"step": 754
},
{
"epoch": 1.0202429149797572,
"grad_norm": 1.3144277334213257,
"learning_rate": 7.170016875007067e-06,
"loss": 0.7932149171829224,
"step": 756
},
{
"epoch": 1.0229419703103915,
"grad_norm": 2.169553756713867,
"learning_rate": 7.144929244621857e-06,
"loss": 0.6181255578994751,
"step": 758
},
{
"epoch": 1.0256410256410255,
"grad_norm": 3.7381906509399414,
"learning_rate": 7.1198362312815205e-06,
"loss": 0.20642229914665222,
"step": 760
},
{
"epoch": 1.0283400809716599,
"grad_norm": 3.731247663497925,
"learning_rate": 7.094738335393361e-06,
"loss": 0.7006998062133789,
"step": 762
},
{
"epoch": 1.0310391363022942,
"grad_norm": 1.382784366607666,
"learning_rate": 7.069636057462043e-06,
"loss": 0.6871730089187622,
"step": 764
},
{
"epoch": 1.0337381916329285,
"grad_norm": 1.5015952587127686,
"learning_rate": 7.044529898079626e-06,
"loss": 0.8277305364608765,
"step": 766
},
{
"epoch": 1.0364372469635628,
"grad_norm": 2.0695316791534424,
"learning_rate": 7.01942035791557e-06,
"loss": 0.4166184365749359,
"step": 768
},
{
"epoch": 1.039136302294197,
"grad_norm": 1.3989157676696777,
"learning_rate": 6.994307937706755e-06,
"loss": 0.90754234790802,
"step": 770
},
{
"epoch": 1.0418353576248314,
"grad_norm": 1.9147645235061646,
"learning_rate": 6.969193138247497e-06,
"loss": 0.7407764196395874,
"step": 772
},
{
"epoch": 1.0445344129554657,
"grad_norm": 3.022681951522827,
"learning_rate": 6.944076460379559e-06,
"loss": 0.46522819995880127,
"step": 774
},
{
"epoch": 1.0472334682861,
"grad_norm": 1.5226314067840576,
"learning_rate": 6.918958404982161e-06,
"loss": 0.7765253782272339,
"step": 776
},
{
"epoch": 1.049932523616734,
"grad_norm": 2.1499991416931152,
"learning_rate": 6.893839472961994e-06,
"loss": 0.7544516324996948,
"step": 778
},
{
"epoch": 1.0526315789473684,
"grad_norm": 3.095790386199951,
"learning_rate": 6.868720165243235e-06,
"loss": 0.6773780584335327,
"step": 780
},
{
"epoch": 1.0553306342780027,
"grad_norm": 1.459345817565918,
"learning_rate": 6.843600982757548e-06,
"loss": 0.7147480249404907,
"step": 782
},
{
"epoch": 1.058029689608637,
"grad_norm": 3.3086650371551514,
"learning_rate": 6.818482426434101e-06,
"loss": 0.7261034250259399,
"step": 784
},
{
"epoch": 1.0607287449392713,
"grad_norm": 1.651834487915039,
"learning_rate": 6.79336499718958e-06,
"loss": 0.8136019706726074,
"step": 786
},
{
"epoch": 1.0634278002699056,
"grad_norm": 7.757126331329346,
"learning_rate": 6.768249195918188e-06,
"loss": 0.5220491886138916,
"step": 788
},
{
"epoch": 1.0661268556005399,
"grad_norm": 4.931647777557373,
"learning_rate": 6.743135523481661e-06,
"loss": 0.7346121072769165,
"step": 790
},
{
"epoch": 1.0688259109311742,
"grad_norm": 42.31561279296875,
"learning_rate": 6.718024480699292e-06,
"loss": 0.4509182274341583,
"step": 792
},
{
"epoch": 1.0715249662618083,
"grad_norm": 6.802929878234863,
"learning_rate": 6.692916568337926e-06,
"loss": 0.20489805936813354,
"step": 794
},
{
"epoch": 1.0742240215924426,
"grad_norm": 2.2887001037597656,
"learning_rate": 6.667812287101978e-06,
"loss": 0.7869080305099487,
"step": 796
},
{
"epoch": 1.0769230769230769,
"grad_norm": 1.7047935724258423,
"learning_rate": 6.64271213762346e-06,
"loss": 0.7171746492385864,
"step": 798
},
{
"epoch": 1.0796221322537112,
"grad_norm": 1.3638522624969482,
"learning_rate": 6.617616620451979e-06,
"loss": 0.8181086778640747,
"step": 800
},
{
"epoch": 1.0823211875843455,
"grad_norm": 2.479830741882324,
"learning_rate": 6.592526236044769e-06,
"loss": 0.6260647177696228,
"step": 802
},
{
"epoch": 1.0850202429149798,
"grad_norm": 2.7747585773468018,
"learning_rate": 6.567441484756705e-06,
"loss": 0.6399790048599243,
"step": 804
},
{
"epoch": 1.087719298245614,
"grad_norm": 2.4006686210632324,
"learning_rate": 6.5423628668303235e-06,
"loss": 0.6539818644523621,
"step": 806
},
{
"epoch": 1.0904183535762484,
"grad_norm": 3.984959125518799,
"learning_rate": 6.517290882385854e-06,
"loss": 0.4466148614883423,
"step": 808
},
{
"epoch": 1.0931174089068827,
"grad_norm": 6.600959300994873,
"learning_rate": 6.492226031411236e-06,
"loss": 0.42955803871154785,
"step": 810
},
{
"epoch": 1.0958164642375168,
"grad_norm": 2.7055540084838867,
"learning_rate": 6.467168813752154e-06,
"loss": 0.6232951283454895,
"step": 812
},
{
"epoch": 1.098515519568151,
"grad_norm": 3.5650031566619873,
"learning_rate": 6.442119729102068e-06,
"loss": 0.6016315221786499,
"step": 814
},
{
"epoch": 1.1012145748987854,
"grad_norm": 3.2496516704559326,
"learning_rate": 6.41707927699225e-06,
"loss": 0.4725358486175537,
"step": 816
},
{
"epoch": 1.1039136302294197,
"grad_norm": 1.7780163288116455,
"learning_rate": 6.3920479567818155e-06,
"loss": 0.6818860173225403,
"step": 818
},
{
"epoch": 1.106612685560054,
"grad_norm": 3.279601812362671,
"learning_rate": 6.367026267647783e-06,
"loss": 0.7466439008712769,
"step": 820
},
{
"epoch": 1.1093117408906883,
"grad_norm": 2.5535480976104736,
"learning_rate": 6.342014708575093e-06,
"loss": 0.8270366787910461,
"step": 822
},
{
"epoch": 1.1120107962213226,
"grad_norm": 1.4318671226501465,
"learning_rate": 6.317013778346681e-06,
"loss": 0.4883289933204651,
"step": 824
},
{
"epoch": 1.114709851551957,
"grad_norm": 6.088655948638916,
"learning_rate": 6.2920239755335176e-06,
"loss": 0.4648905098438263,
"step": 826
},
{
"epoch": 1.117408906882591,
"grad_norm": 3.39751935005188,
"learning_rate": 6.2670457984846695e-06,
"loss": 0.79395592212677,
"step": 828
},
{
"epoch": 1.1201079622132253,
"grad_norm": 7.12752628326416,
"learning_rate": 6.2420797453173574e-06,
"loss": 0.22918568551540375,
"step": 830
},
{
"epoch": 1.1228070175438596,
"grad_norm": 1.8719902038574219,
"learning_rate": 6.217126313907036e-06,
"loss": 0.7311997413635254,
"step": 832
},
{
"epoch": 1.125506072874494,
"grad_norm": 9.373211860656738,
"learning_rate": 6.192186001877449e-06,
"loss": 0.487352192401886,
"step": 834
},
{
"epoch": 1.1282051282051282,
"grad_norm": 1.653995394706726,
"learning_rate": 6.167259306590711e-06,
"loss": 0.5706923007965088,
"step": 836
},
{
"epoch": 1.1309041835357625,
"grad_norm": 1.8983337879180908,
"learning_rate": 6.142346725137395e-06,
"loss": 0.7815992832183838,
"step": 838
},
{
"epoch": 1.1336032388663968,
"grad_norm": 2.8028581142425537,
"learning_rate": 6.117448754326609e-06,
"loss": 0.34738290309906006,
"step": 840
},
{
"epoch": 1.136302294197031,
"grad_norm": 3.3673501014709473,
"learning_rate": 6.092565890676098e-06,
"loss": 0.3341389000415802,
"step": 842
},
{
"epoch": 1.1390013495276654,
"grad_norm": 4.791003704071045,
"learning_rate": 6.06769863040234e-06,
"loss": 0.48775380849838257,
"step": 844
},
{
"epoch": 1.1417004048582995,
"grad_norm": 3.2612569332122803,
"learning_rate": 6.042847469410645e-06,
"loss": 0.7673020958900452,
"step": 846
},
{
"epoch": 1.1443994601889338,
"grad_norm": 2.966323137283325,
"learning_rate": 6.018012903285271e-06,
"loss": 0.7615069150924683,
"step": 848
},
{
"epoch": 1.147098515519568,
"grad_norm": 7.46718168258667,
"learning_rate": 5.993195427279541e-06,
"loss": 0.8174636960029602,
"step": 850
},
{
"epoch": 1.1497975708502024,
"grad_norm": 1.3792356252670288,
"learning_rate": 5.968395536305962e-06,
"loss": 0.4660506844520569,
"step": 852
},
{
"epoch": 1.1524966261808367,
"grad_norm": 1.9996525049209595,
"learning_rate": 5.943613724926361e-06,
"loss": 0.6199110746383667,
"step": 854
},
{
"epoch": 1.155195681511471,
"grad_norm": 5.937838554382324,
"learning_rate": 5.918850487342022e-06,
"loss": 0.6069660186767578,
"step": 856
},
{
"epoch": 1.1578947368421053,
"grad_norm": 1.8015501499176025,
"learning_rate": 5.894106317383823e-06,
"loss": 0.45053383708000183,
"step": 858
},
{
"epoch": 1.1605937921727396,
"grad_norm": 2.855788469314575,
"learning_rate": 5.869381708502398e-06,
"loss": 0.5916150808334351,
"step": 860
},
{
"epoch": 1.163292847503374,
"grad_norm": 1.6809040307998657,
"learning_rate": 5.844677153758292e-06,
"loss": 0.48865312337875366,
"step": 862
},
{
"epoch": 1.165991902834008,
"grad_norm": 2.5966737270355225,
"learning_rate": 5.819993145812122e-06,
"loss": 0.42146843671798706,
"step": 864
},
{
"epoch": 1.1686909581646423,
"grad_norm": 1.4814966917037964,
"learning_rate": 5.795330176914772e-06,
"loss": 0.6994541883468628,
"step": 866
},
{
"epoch": 1.1713900134952766,
"grad_norm": 2.2960119247436523,
"learning_rate": 5.770688738897551e-06,
"loss": 0.5806586146354675,
"step": 868
},
{
"epoch": 1.174089068825911,
"grad_norm": 3.2501308917999268,
"learning_rate": 5.7460693231624e-06,
"loss": 0.8271964192390442,
"step": 870
},
{
"epoch": 1.1767881241565452,
"grad_norm": 3.3459393978118896,
"learning_rate": 5.7214724206720915e-06,
"loss": 0.31470566987991333,
"step": 872
},
{
"epoch": 1.1794871794871795,
"grad_norm": 1.6450824737548828,
"learning_rate": 5.696898521940438e-06,
"loss": 0.7387674450874329,
"step": 874
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.4505995810031891,
"learning_rate": 5.6723481170225055e-06,
"loss": 0.3696012496948242,
"step": 876
},
{
"epoch": 1.1848852901484481,
"grad_norm": 0.2561063766479492,
"learning_rate": 5.647821695504846e-06,
"loss": 0.423138827085495,
"step": 878
},
{
"epoch": 1.1875843454790824,
"grad_norm": 1.6858747005462646,
"learning_rate": 5.623319746495732e-06,
"loss": 0.32702261209487915,
"step": 880
},
{
"epoch": 1.1902834008097165,
"grad_norm": 1.5090140104293823,
"learning_rate": 5.598842758615401e-06,
"loss": 0.6388052701950073,
"step": 882
},
{
"epoch": 1.1929824561403508,
"grad_norm": 2.3837759494781494,
"learning_rate": 5.574391219986317e-06,
"loss": 0.38176387548446655,
"step": 884
},
{
"epoch": 1.1956815114709851,
"grad_norm": 2.500481605529785,
"learning_rate": 5.549965618223428e-06,
"loss": 0.7306652069091797,
"step": 886
},
{
"epoch": 1.1983805668016194,
"grad_norm": 3.897916793823242,
"learning_rate": 5.525566440424449e-06,
"loss": 0.49203619360923767,
"step": 888
},
{
"epoch": 1.2010796221322537,
"grad_norm": 1.7262345552444458,
"learning_rate": 5.501194173160148e-06,
"loss": 0.37901103496551514,
"step": 890
},
{
"epoch": 1.203778677462888,
"grad_norm": 4.88112735748291,
"learning_rate": 5.476849302464638e-06,
"loss": 0.35398244857788086,
"step": 892
},
{
"epoch": 1.2064777327935223,
"grad_norm": 1.9734995365142822,
"learning_rate": 5.452532313825684e-06,
"loss": 0.7075091004371643,
"step": 894
},
{
"epoch": 1.2091767881241566,
"grad_norm": 5.391160011291504,
"learning_rate": 5.4282436921750305e-06,
"loss": 0.5342161655426025,
"step": 896
},
{
"epoch": 1.211875843454791,
"grad_norm": 1.6796256303787231,
"learning_rate": 5.403983921878722e-06,
"loss": 0.498091459274292,
"step": 898
},
{
"epoch": 1.214574898785425,
"grad_norm": 1.9090220928192139,
"learning_rate": 5.37975348672744e-06,
"loss": 0.6458582878112793,
"step": 900
},
{
"epoch": 1.2172739541160593,
"grad_norm": 2.673980712890625,
"learning_rate": 5.355552869926876e-06,
"loss": 0.6077509522438049,
"step": 902
},
{
"epoch": 1.2199730094466936,
"grad_norm": 2.16408371925354,
"learning_rate": 5.33138255408807e-06,
"loss": 0.8459935188293457,
"step": 904
},
{
"epoch": 1.222672064777328,
"grad_norm": 1.250825047492981,
"learning_rate": 5.3072430212177975e-06,
"loss": 0.5663684606552124,
"step": 906
},
{
"epoch": 1.2253711201079622,
"grad_norm": 2.2070729732513428,
"learning_rate": 5.283134752708964e-06,
"loss": 0.04916539788246155,
"step": 908
},
{
"epoch": 1.2280701754385965,
"grad_norm": 1.4894299507141113,
"learning_rate": 5.2590582293309935e-06,
"loss": 0.6701165437698364,
"step": 910
},
{
"epoch": 1.2307692307692308,
"grad_norm": 2.152111291885376,
"learning_rate": 5.235013931220241e-06,
"loss": 0.37132638692855835,
"step": 912
},
{
"epoch": 1.2334682860998651,
"grad_norm": 0.9825637340545654,
"learning_rate": 5.2110023378704315e-06,
"loss": 0.35279959440231323,
"step": 914
},
{
"epoch": 1.2361673414304994,
"grad_norm": 1.7362233400344849,
"learning_rate": 5.187023928123085e-06,
"loss": 0.727353572845459,
"step": 916
},
{
"epoch": 1.2388663967611335,
"grad_norm": 1.6653279066085815,
"learning_rate": 5.163079180157966e-06,
"loss": 0.7612659931182861,
"step": 918
},
{
"epoch": 1.2415654520917678,
"grad_norm": 2.676962375640869,
"learning_rate": 5.139168571483559e-06,
"loss": 0.47323521971702576,
"step": 920
},
{
"epoch": 1.2442645074224021,
"grad_norm": 2.4546144008636475,
"learning_rate": 5.115292578927536e-06,
"loss": 0.3562048077583313,
"step": 922
},
{
"epoch": 1.2469635627530364,
"grad_norm": 1.6883822679519653,
"learning_rate": 5.091451678627251e-06,
"loss": 0.6207482814788818,
"step": 924
},
{
"epoch": 1.2496626180836707,
"grad_norm": 1.189233660697937,
"learning_rate": 5.06764634602025e-06,
"loss": 0.5157153010368347,
"step": 926
},
{
"epoch": 1.252361673414305,
"grad_norm": 3.9678120613098145,
"learning_rate": 5.043877055834775e-06,
"loss": 0.44284674525260925,
"step": 928
},
{
"epoch": 1.2550607287449393,
"grad_norm": 1.0479458570480347,
"learning_rate": 5.020144282080314e-06,
"loss": 0.7401950359344482,
"step": 930
},
{
"epoch": 1.2577597840755734,
"grad_norm": 15.801717758178711,
"learning_rate": 4.996448498038139e-06,
"loss": 0.39344197511672974,
"step": 932
},
{
"epoch": 1.260458839406208,
"grad_norm": 1.5226280689239502,
"learning_rate": 4.972790176251868e-06,
"loss": 0.6036249399185181,
"step": 934
},
{
"epoch": 1.263157894736842,
"grad_norm": 1.5542678833007812,
"learning_rate": 4.949169788518046e-06,
"loss": 0.830246090888977,
"step": 936
},
{
"epoch": 1.2658569500674763,
"grad_norm": 1.36121666431427,
"learning_rate": 4.92558780587673e-06,
"loss": 0.4464884400367737,
"step": 938
},
{
"epoch": 1.2685560053981106,
"grad_norm": 1.4898165464401245,
"learning_rate": 4.9020446986021015e-06,
"loss": 0.3654748201370239,
"step": 940
},
{
"epoch": 1.271255060728745,
"grad_norm": 1.4254517555236816,
"learning_rate": 4.8785409361930865e-06,
"loss": 0.7567611336708069,
"step": 942
},
{
"epoch": 1.2739541160593792,
"grad_norm": 1.619158387184143,
"learning_rate": 4.855076987363988e-06,
"loss": 0.7386656999588013,
"step": 944
},
{
"epoch": 1.2766531713900136,
"grad_norm": 1.5548937320709229,
"learning_rate": 4.831653320035144e-06,
"loss": 0.9016460180282593,
"step": 946
},
{
"epoch": 1.2793522267206479,
"grad_norm": 1.3139351606369019,
"learning_rate": 4.808270401323599e-06,
"loss": 0.712449312210083,
"step": 948
},
{
"epoch": 1.282051282051282,
"grad_norm": 1.7884823083877563,
"learning_rate": 4.7849286975337815e-06,
"loss": 0.6212611198425293,
"step": 950
},
{
"epoch": 1.2847503373819165,
"grad_norm": 1.0487291812896729,
"learning_rate": 4.7616286741482036e-06,
"loss": 0.7121725082397461,
"step": 952
},
{
"epoch": 1.2874493927125505,
"grad_norm": 1.6497894525527954,
"learning_rate": 4.738370795818193e-06,
"loss": 0.6121946573257446,
"step": 954
},
{
"epoch": 1.2901484480431848,
"grad_norm": 2.2327287197113037,
"learning_rate": 4.715155526354606e-06,
"loss": 0.8555269241333008,
"step": 956
},
{
"epoch": 1.2928475033738192,
"grad_norm": 17.18166160583496,
"learning_rate": 4.6919833287185954e-06,
"loss": 0.7826257944107056,
"step": 958
},
{
"epoch": 1.2955465587044535,
"grad_norm": 1.892973780632019,
"learning_rate": 4.6688546650123706e-06,
"loss": 0.8690598011016846,
"step": 960
},
{
"epoch": 1.2982456140350878,
"grad_norm": 1.4111993312835693,
"learning_rate": 4.645769996469983e-06,
"loss": 0.6364820599555969,
"step": 962
},
{
"epoch": 1.300944669365722,
"grad_norm": 3.973740339279175,
"learning_rate": 4.622729783448124e-06,
"loss": 0.5858311653137207,
"step": 964
},
{
"epoch": 1.3036437246963564,
"grad_norm": 2.5259652137756348,
"learning_rate": 4.599734485416955e-06,
"loss": 0.7325659990310669,
"step": 966
},
{
"epoch": 1.3063427800269904,
"grad_norm": 13.180168151855469,
"learning_rate": 4.576784560950933e-06,
"loss": 0.34615111351013184,
"step": 968
},
{
"epoch": 1.3090418353576248,
"grad_norm": 1.4625717401504517,
"learning_rate": 4.553880467719675e-06,
"loss": 0.578520655632019,
"step": 970
},
{
"epoch": 1.311740890688259,
"grad_norm": 8.096935272216797,
"learning_rate": 4.531022662478826e-06,
"loss": 0.37505674362182617,
"step": 972
},
{
"epoch": 1.3144399460188934,
"grad_norm": 1.8905644416809082,
"learning_rate": 4.5082116010609495e-06,
"loss": 0.6498898267745972,
"step": 974
},
{
"epoch": 1.3171390013495277,
"grad_norm": 1.8634613752365112,
"learning_rate": 4.48544773836644e-06,
"loss": 0.6129083037376404,
"step": 976
},
{
"epoch": 1.319838056680162,
"grad_norm": 2.667171001434326,
"learning_rate": 4.462731528354454e-06,
"loss": 0.46260231733322144,
"step": 978
},
{
"epoch": 1.3225371120107963,
"grad_norm": 3.009416103363037,
"learning_rate": 4.440063424033848e-06,
"loss": 0.692058801651001,
"step": 980
},
{
"epoch": 1.3252361673414306,
"grad_norm": 1.555985689163208,
"learning_rate": 4.417443877454153e-06,
"loss": 0.7391829490661621,
"step": 982
},
{
"epoch": 1.3279352226720649,
"grad_norm": 2.0486674308776855,
"learning_rate": 4.3948733396965576e-06,
"loss": 0.5268170237541199,
"step": 984
},
{
"epoch": 1.330634278002699,
"grad_norm": 1.6344236135482788,
"learning_rate": 4.372352260864907e-06,
"loss": 0.6643505096435547,
"step": 986
},
{
"epoch": 1.3333333333333333,
"grad_norm": 2.9619522094726562,
"learning_rate": 4.3498810900767415e-06,
"loss": 0.5001938343048096,
"step": 988
},
{
"epoch": 1.3360323886639676,
"grad_norm": 1.5613045692443848,
"learning_rate": 4.327460275454321e-06,
"loss": 0.7717913389205933,
"step": 990
},
{
"epoch": 1.3387314439946019,
"grad_norm": 1.4568570852279663,
"learning_rate": 4.3050902641156995e-06,
"loss": 0.3589901328086853,
"step": 992
},
{
"epoch": 1.3414304993252362,
"grad_norm": 5.554232120513916,
"learning_rate": 4.282771502165812e-06,
"loss": 0.6063488721847534,
"step": 994
},
{
"epoch": 1.3441295546558705,
"grad_norm": 1.037520408630371,
"learning_rate": 4.260504434687568e-06,
"loss": 0.7692219018936157,
"step": 996
},
{
"epoch": 1.3468286099865048,
"grad_norm": 7.167163848876953,
"learning_rate": 4.238289505732982e-06,
"loss": 0.3994293808937073,
"step": 998
},
{
"epoch": 1.349527665317139,
"grad_norm": 1.551005482673645,
"learning_rate": 4.21612715831432e-06,
"loss": 0.33824437856674194,
"step": 1000
},
{
"epoch": 1.3522267206477734,
"grad_norm": 2.381816864013672,
"learning_rate": 4.1940178343952596e-06,
"loss": 0.7282351851463318,
"step": 1002
},
{
"epoch": 1.3549257759784075,
"grad_norm": 1.5229370594024658,
"learning_rate": 4.171961974882077e-06,
"loss": 0.8121545314788818,
"step": 1004
},
{
"epoch": 1.3576248313090418,
"grad_norm": 1.3013309240341187,
"learning_rate": 4.149960019614857e-06,
"loss": 0.6453372240066528,
"step": 1006
},
{
"epoch": 1.360323886639676,
"grad_norm": 1.6318532228469849,
"learning_rate": 4.1280124073587234e-06,
"loss": 0.6700369119644165,
"step": 1008
},
{
"epoch": 1.3630229419703104,
"grad_norm": 1.9664186239242554,
"learning_rate": 4.106119575795082e-06,
"loss": 0.6605565547943115,
"step": 1010
},
{
"epoch": 1.3657219973009447,
"grad_norm": 1.618058681488037,
"learning_rate": 4.084281961512902e-06,
"loss": 0.37986230850219727,
"step": 1012
},
{
"epoch": 1.368421052631579,
"grad_norm": 1.9911377429962158,
"learning_rate": 4.062500000000001e-06,
"loss": 0.6968834400177002,
"step": 1014
},
{
"epoch": 1.3711201079622133,
"grad_norm": 3.1421597003936768,
"learning_rate": 4.040774125634362e-06,
"loss": 0.8204485774040222,
"step": 1016
},
{
"epoch": 1.3738191632928474,
"grad_norm": 1.7936409711837769,
"learning_rate": 4.019104771675477e-06,
"loss": 0.4428718090057373,
"step": 1018
},
{
"epoch": 1.376518218623482,
"grad_norm": 2.101984977722168,
"learning_rate": 3.997492370255699e-06,
"loss": 0.8280647993087769,
"step": 1020
},
{
"epoch": 1.379217273954116,
"grad_norm": 1.8467909097671509,
"learning_rate": 3.975937352371629e-06,
"loss": 0.3088380992412567,
"step": 1022
},
{
"epoch": 1.3819163292847503,
"grad_norm": 2.341449737548828,
"learning_rate": 3.954440147875524e-06,
"loss": 0.39887773990631104,
"step": 1024
},
{
"epoch": 1.3846153846153846,
"grad_norm": 22.286470413208008,
"learning_rate": 3.9330011854667124e-06,
"loss": 0.5050676465034485,
"step": 1026
},
{
"epoch": 1.3873144399460189,
"grad_norm": 2.2894952297210693,
"learning_rate": 3.91162089268306e-06,
"loss": 0.7694636583328247,
"step": 1028
},
{
"epoch": 1.3900134952766532,
"grad_norm": 1.8550227880477905,
"learning_rate": 3.890299695892435e-06,
"loss": 0.6376620531082153,
"step": 1030
},
{
"epoch": 1.3927125506072875,
"grad_norm": 1.3034743070602417,
"learning_rate": 3.8690380202842075e-06,
"loss": 0.5672460794448853,
"step": 1032
},
{
"epoch": 1.3954116059379218,
"grad_norm": 14.647714614868164,
"learning_rate": 3.847836289860769e-06,
"loss": 0.3966209292411804,
"step": 1034
},
{
"epoch": 1.3981106612685559,
"grad_norm": 4.088876724243164,
"learning_rate": 3.826694927429083e-06,
"loss": 0.5219507217407227,
"step": 1036
},
{
"epoch": 1.4008097165991904,
"grad_norm": 1.5946576595306396,
"learning_rate": 3.805614354592246e-06,
"loss": 0.6981402635574341,
"step": 1038
},
{
"epoch": 1.4035087719298245,
"grad_norm": 1.6788384914398193,
"learning_rate": 3.784594991741079e-06,
"loss": 0.7897889018058777,
"step": 1040
},
{
"epoch": 1.4062078272604588,
"grad_norm": 3.879822015762329,
"learning_rate": 3.7636372580457534e-06,
"loss": 0.3763599395751953,
"step": 1042
},
{
"epoch": 1.408906882591093,
"grad_norm": 3.674297332763672,
"learning_rate": 3.742741571447424e-06,
"loss": 0.10553102195262909,
"step": 1044
},
{
"epoch": 1.4116059379217274,
"grad_norm": 2.0286495685577393,
"learning_rate": 3.7219083486498912e-06,
"loss": 0.3425998091697693,
"step": 1046
},
{
"epoch": 1.4143049932523617,
"grad_norm": 2.1781485080718994,
"learning_rate": 3.7011380051113058e-06,
"loss": 0.7947150468826294,
"step": 1048
},
{
"epoch": 1.417004048582996,
"grad_norm": 1.2689515352249146,
"learning_rate": 3.6804309550358677e-06,
"loss": 0.46531498432159424,
"step": 1050
},
{
"epoch": 1.4197031039136303,
"grad_norm": 1.434031367301941,
"learning_rate": 3.659787611365574e-06,
"loss": 0.6526649594306946,
"step": 1052
},
{
"epoch": 1.4224021592442644,
"grad_norm": 2.48591947555542,
"learning_rate": 3.6392083857719812e-06,
"loss": 0.5354222059249878,
"step": 1054
},
{
"epoch": 1.425101214574899,
"grad_norm": 4.407574653625488,
"learning_rate": 3.6186936886479986e-06,
"loss": 0.6338904500007629,
"step": 1056
},
{
"epoch": 1.427800269905533,
"grad_norm": 12.838454246520996,
"learning_rate": 3.598243929099704e-06,
"loss": 0.47556883096694946,
"step": 1058
},
{
"epoch": 1.4304993252361673,
"grad_norm": 1.669755220413208,
"learning_rate": 3.577859514938181e-06,
"loss": 0.5890032052993774,
"step": 1060
},
{
"epoch": 1.4331983805668016,
"grad_norm": 1.2008280754089355,
"learning_rate": 3.5575408526713886e-06,
"loss": 0.3775816559791565,
"step": 1062
},
{
"epoch": 1.435897435897436,
"grad_norm": 2.5516271591186523,
"learning_rate": 3.5372883474960595e-06,
"loss": 0.5098634362220764,
"step": 1064
},
{
"epoch": 1.4385964912280702,
"grad_norm": 1.3215935230255127,
"learning_rate": 3.5171024032896113e-06,
"loss": 0.726750910282135,
"step": 1066
},
{
"epoch": 1.4412955465587045,
"grad_norm": 3.994313955307007,
"learning_rate": 3.4969834226020947e-06,
"loss": 0.39119070768356323,
"step": 1068
},
{
"epoch": 1.4439946018893388,
"grad_norm": 1.602039098739624,
"learning_rate": 3.4769318066481747e-06,
"loss": 0.6564367413520813,
"step": 1070
},
{
"epoch": 1.446693657219973,
"grad_norm": 8.2996244430542,
"learning_rate": 3.4569479552991156e-06,
"loss": 0.35547560453414917,
"step": 1072
},
{
"epoch": 1.4493927125506074,
"grad_norm": 1.2243651151657104,
"learning_rate": 3.437032267074816e-06,
"loss": 0.8508844971656799,
"step": 1074
},
{
"epoch": 1.4520917678812415,
"grad_norm": 1.6840155124664307,
"learning_rate": 3.4171851391358574e-06,
"loss": 0.6419157385826111,
"step": 1076
},
{
"epoch": 1.4547908232118758,
"grad_norm": 1.2368465662002563,
"learning_rate": 3.3974069672755875e-06,
"loss": 0.10182754695415497,
"step": 1078
},
{
"epoch": 1.45748987854251,
"grad_norm": 13.594934463500977,
"learning_rate": 3.3776981459122213e-06,
"loss": 0.47692227363586426,
"step": 1080
},
{
"epoch": 1.4601889338731444,
"grad_norm": 1.5118765830993652,
"learning_rate": 3.3580590680809876e-06,
"loss": 0.7316707372665405,
"step": 1082
},
{
"epoch": 1.4628879892037787,
"grad_norm": 1.727602481842041,
"learning_rate": 3.338490125426275e-06,
"loss": 0.6650782823562622,
"step": 1084
},
{
"epoch": 1.465587044534413,
"grad_norm": 1.1840381622314453,
"learning_rate": 3.3189917081938286e-06,
"loss": 0.7887551188468933,
"step": 1086
},
{
"epoch": 1.4682860998650473,
"grad_norm": 1.3345993757247925,
"learning_rate": 3.2995642052229782e-06,
"loss": 0.4664420485496521,
"step": 1088
},
{
"epoch": 1.4709851551956814,
"grad_norm": 1.6331321001052856,
"learning_rate": 3.2802080039388673e-06,
"loss": 0.8616183400154114,
"step": 1090
},
{
"epoch": 1.4736842105263157,
"grad_norm": 1.847618818283081,
"learning_rate": 3.260923490344732e-06,
"loss": 0.5459949970245361,
"step": 1092
},
{
"epoch": 1.47638326585695,
"grad_norm": 3.2221004962921143,
"learning_rate": 3.2417110490142145e-06,
"loss": 0.6761457324028015,
"step": 1094
},
{
"epoch": 1.4790823211875843,
"grad_norm": 1.1452100276947021,
"learning_rate": 3.222571063083678e-06,
"loss": 0.7724871635437012,
"step": 1096
},
{
"epoch": 1.4817813765182186,
"grad_norm": 1.6045210361480713,
"learning_rate": 3.2035039142445718e-06,
"loss": 0.8433651924133301,
"step": 1098
},
{
"epoch": 1.484480431848853,
"grad_norm": 1.699333667755127,
"learning_rate": 3.184509982735827e-06,
"loss": 0.7413595914840698,
"step": 1100
},
{
"epoch": 1.4871794871794872,
"grad_norm": 1.9658582210540771,
"learning_rate": 3.1655896473362632e-06,
"loss": 0.5522735118865967,
"step": 1102
},
{
"epoch": 1.4898785425101215,
"grad_norm": 1.078214168548584,
"learning_rate": 3.14674328535704e-06,
"loss": 0.21913763880729675,
"step": 1104
},
{
"epoch": 1.4925775978407558,
"grad_norm": 1.8495062589645386,
"learning_rate": 3.1279712726341337e-06,
"loss": 0.6720937490463257,
"step": 1106
},
{
"epoch": 1.49527665317139,
"grad_norm": 1.9412720203399658,
"learning_rate": 3.109273983520838e-06,
"loss": 0.7124868631362915,
"step": 1108
},
{
"epoch": 1.4979757085020242,
"grad_norm": 1.5045949220657349,
"learning_rate": 3.090651790880304e-06,
"loss": 0.725304365158081,
"step": 1110
},
{
"epoch": 1.5006747638326585,
"grad_norm": 1.9396291971206665,
"learning_rate": 3.0721050660781e-06,
"loss": 0.7659965753555298,
"step": 1112
},
{
"epoch": 1.5033738191632928,
"grad_norm": 1.9754250049591064,
"learning_rate": 3.053634178974804e-06,
"loss": 0.511092483997345,
"step": 1114
},
{
"epoch": 1.5060728744939271,
"grad_norm": 1.1582471132278442,
"learning_rate": 3.0352394979186396e-06,
"loss": 0.6046227216720581,
"step": 1116
},
{
"epoch": 1.5087719298245614,
"grad_norm": 6.374830722808838,
"learning_rate": 3.0169213897381145e-06,
"loss": 0.46287164092063904,
"step": 1118
},
{
"epoch": 1.5114709851551957,
"grad_norm": 1.767095923423767,
"learning_rate": 2.9986802197347152e-06,
"loss": 0.900627613067627,
"step": 1120
},
{
"epoch": 1.5141700404858298,
"grad_norm": 1.4868059158325195,
"learning_rate": 2.9805163516756204e-06,
"loss": 0.769967257976532,
"step": 1122
},
{
"epoch": 1.5168690958164643,
"grad_norm": 2.632185697555542,
"learning_rate": 2.962430147786445e-06,
"loss": 0.6432470083236694,
"step": 1124
},
{
"epoch": 1.5195681511470984,
"grad_norm": 9.906206130981445,
"learning_rate": 2.9444219687440157e-06,
"loss": 0.5127274990081787,
"step": 1126
},
{
"epoch": 1.522267206477733,
"grad_norm": 2.0955259799957275,
"learning_rate": 2.9264921736691867e-06,
"loss": 0.42444682121276855,
"step": 1128
},
{
"epoch": 1.524966261808367,
"grad_norm": 2.149416208267212,
"learning_rate": 2.9086411201196646e-06,
"loss": 0.5323820114135742,
"step": 1130
},
{
"epoch": 1.5276653171390013,
"grad_norm": 1.5409173965454102,
"learning_rate": 2.8908691640828852e-06,
"loss": 0.6285885572433472,
"step": 1132
},
{
"epoch": 1.5303643724696356,
"grad_norm": 2.386502265930176,
"learning_rate": 2.873176659968921e-06,
"loss": 0.575620174407959,
"step": 1134
},
{
"epoch": 1.53306342780027,
"grad_norm": 2.00553297996521,
"learning_rate": 2.8555639606033987e-06,
"loss": 0.4259451627731323,
"step": 1136
},
{
"epoch": 1.5357624831309042,
"grad_norm": 7.872402667999268,
"learning_rate": 2.8380314172204726e-06,
"loss": 0.11073433607816696,
"step": 1138
},
{
"epoch": 1.5384615384615383,
"grad_norm": 5.999627590179443,
"learning_rate": 2.8205793794558236e-06,
"loss": 0.6745946407318115,
"step": 1140
},
{
"epoch": 1.5411605937921729,
"grad_norm": 2.0469369888305664,
"learning_rate": 2.8032081953396775e-06,
"loss": 0.5133538842201233,
"step": 1142
},
{
"epoch": 1.543859649122807,
"grad_norm": 0.9723069667816162,
"learning_rate": 2.785918211289868e-06,
"loss": 0.692731499671936,
"step": 1144
},
{
"epoch": 1.5465587044534415,
"grad_norm": 1.7804145812988281,
"learning_rate": 2.768709772104932e-06,
"loss": 0.4003831744194031,
"step": 1146
},
{
"epoch": 1.5492577597840755,
"grad_norm": 1.6913983821868896,
"learning_rate": 2.7515832209572283e-06,
"loss": 0.6252081990242004,
"step": 1148
},
{
"epoch": 1.5519568151147098,
"grad_norm": 3.494619369506836,
"learning_rate": 2.734538899386098e-06,
"loss": 0.7116725444793701,
"step": 1150
},
{
"epoch": 1.5546558704453441,
"grad_norm": 2.3654909133911133,
"learning_rate": 2.7175771472910535e-06,
"loss": 0.7219235301017761,
"step": 1152
},
{
"epoch": 1.5573549257759784,
"grad_norm": 1.9625266790390015,
"learning_rate": 2.7006983029249958e-06,
"loss": 0.7391921281814575,
"step": 1154
},
{
"epoch": 1.5600539811066128,
"grad_norm": 1.6443387269973755,
"learning_rate": 2.6839027028874717e-06,
"loss": 0.7720010280609131,
"step": 1156
},
{
"epoch": 1.5627530364372468,
"grad_norm": 3.267526388168335,
"learning_rate": 2.667190682117967e-06,
"loss": 0.6769675016403198,
"step": 1158
},
{
"epoch": 1.5654520917678814,
"grad_norm": 1.439118504524231,
"learning_rate": 2.6505625738892153e-06,
"loss": 0.7269835472106934,
"step": 1160
},
{
"epoch": 1.5681511470985154,
"grad_norm": 4.097200393676758,
"learning_rate": 2.634018709800558e-06,
"loss": 0.6407755613327026,
"step": 1162
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.7139143347740173,
"learning_rate": 2.61755941977134e-06,
"loss": 0.30610889196395874,
"step": 1164
},
{
"epoch": 1.573549257759784,
"grad_norm": 1.56275475025177,
"learning_rate": 2.601185032034313e-06,
"loss": 0.683242678642273,
"step": 1166
},
{
"epoch": 1.5762483130904184,
"grad_norm": 3.5380778312683105,
"learning_rate": 2.5848958731291024e-06,
"loss": 0.38974276185035706,
"step": 1168
},
{
"epoch": 1.5789473684210527,
"grad_norm": 4.699647903442383,
"learning_rate": 2.568692267895693e-06,
"loss": 0.7000415325164795,
"step": 1170
},
{
"epoch": 1.581646423751687,
"grad_norm": 1.5647377967834473,
"learning_rate": 2.5525745394679455e-06,
"loss": 0.7570105791091919,
"step": 1172
},
{
"epoch": 1.5843454790823213,
"grad_norm": 5.9123334884643555,
"learning_rate": 2.536543009267165e-06,
"loss": 0.20975083112716675,
"step": 1174
},
{
"epoch": 1.5870445344129553,
"grad_norm": 1.4215142726898193,
"learning_rate": 2.520597996995678e-06,
"loss": 0.7487507462501526,
"step": 1176
},
{
"epoch": 1.5897435897435899,
"grad_norm": 5.5586371421813965,
"learning_rate": 2.5047398206304597e-06,
"loss": 0.4733960032463074,
"step": 1178
},
{
"epoch": 1.592442645074224,
"grad_norm": 6.647611141204834,
"learning_rate": 2.4889687964168013e-06,
"loss": 0.5053025484085083,
"step": 1180
},
{
"epoch": 1.5951417004048583,
"grad_norm": 3.1487364768981934,
"learning_rate": 2.4732852388619934e-06,
"loss": 0.48154693841934204,
"step": 1182
},
{
"epoch": 1.5978407557354926,
"grad_norm": 1.7464215755462646,
"learning_rate": 2.4576894607290587e-06,
"loss": 0.4145604968070984,
"step": 1184
},
{
"epoch": 1.6005398110661269,
"grad_norm": 1.1593838930130005,
"learning_rate": 2.4421817730305156e-06,
"loss": 0.39196354150772095,
"step": 1186
},
{
"epoch": 1.6032388663967612,
"grad_norm": 1.4402377605438232,
"learning_rate": 2.4267624850221723e-06,
"loss": 0.5265844464302063,
"step": 1188
},
{
"epoch": 1.6059379217273952,
"grad_norm": 2.604595184326172,
"learning_rate": 2.4114319041969635e-06,
"loss": 0.5243338942527771,
"step": 1190
},
{
"epoch": 1.6086369770580298,
"grad_norm": 0.9398858547210693,
"learning_rate": 2.3961903362788154e-06,
"loss": 0.5175016522407532,
"step": 1192
},
{
"epoch": 1.6113360323886639,
"grad_norm": 1.7844507694244385,
"learning_rate": 2.3810380852165517e-06,
"loss": 0.6253082752227783,
"step": 1194
},
{
"epoch": 1.6140350877192984,
"grad_norm": 2.234337329864502,
"learning_rate": 2.365975453177828e-06,
"loss": 0.7877857685089111,
"step": 1196
},
{
"epoch": 1.6167341430499325,
"grad_norm": 1.4822181463241577,
"learning_rate": 2.351002740543113e-06,
"loss": 0.5377737283706665,
"step": 1198
},
{
"epoch": 1.6194331983805668,
"grad_norm": 1.9002721309661865,
"learning_rate": 2.336120245899691e-06,
"loss": 0.6274523735046387,
"step": 1200
},
{
"epoch": 1.622132253711201,
"grad_norm": 2.3867383003234863,
"learning_rate": 2.3213282660357083e-06,
"loss": 0.4820527136325836,
"step": 1202
},
{
"epoch": 1.6248313090418354,
"grad_norm": 0.8942919969558716,
"learning_rate": 2.3066270959342636e-06,
"loss": 0.3185853064060211,
"step": 1204
},
{
"epoch": 1.6275303643724697,
"grad_norm": 3.2428300380706787,
"learning_rate": 2.2920170287675116e-06,
"loss": 0.24991720914840698,
"step": 1206
},
{
"epoch": 1.6302294197031038,
"grad_norm": 1.798754096031189,
"learning_rate": 2.2774983558908253e-06,
"loss": 0.756623387336731,
"step": 1208
},
{
"epoch": 1.6329284750337383,
"grad_norm": 1.6678624153137207,
"learning_rate": 2.2630713668369872e-06,
"loss": 0.4156914949417114,
"step": 1210
},
{
"epoch": 1.6356275303643724,
"grad_norm": 1.567728877067566,
"learning_rate": 2.2487363493104073e-06,
"loss": 0.7142396569252014,
"step": 1212
},
{
"epoch": 1.6383265856950069,
"grad_norm": 5.427001953125,
"learning_rate": 2.2344935891813945e-06,
"loss": 0.14030721783638,
"step": 1214
},
{
"epoch": 1.641025641025641,
"grad_norm": 2.0503334999084473,
"learning_rate": 2.2203433704804495e-06,
"loss": 0.7471613883972168,
"step": 1216
},
{
"epoch": 1.6437246963562753,
"grad_norm": 1.1235144138336182,
"learning_rate": 2.206285975392603e-06,
"loss": 0.5582278966903687,
"step": 1218
},
{
"epoch": 1.6464237516869096,
"grad_norm": 1.2342617511749268,
"learning_rate": 2.1923216842517874e-06,
"loss": 0.6891179084777832,
"step": 1220
},
{
"epoch": 1.6491228070175439,
"grad_norm": 4.849971771240234,
"learning_rate": 2.178450775535252e-06,
"loss": 0.6573494076728821,
"step": 1222
},
{
"epoch": 1.6518218623481782,
"grad_norm": 33.77346420288086,
"learning_rate": 2.164673525857998e-06,
"loss": 0.4994286000728607,
"step": 1224
},
{
"epoch": 1.6545209176788123,
"grad_norm": 1.8156238794326782,
"learning_rate": 2.150990209967273e-06,
"loss": 0.746261715888977,
"step": 1226
},
{
"epoch": 1.6572199730094468,
"grad_norm": 3.0515799522399902,
"learning_rate": 2.137401100737088e-06,
"loss": 0.7395338416099548,
"step": 1228
},
{
"epoch": 1.6599190283400809,
"grad_norm": 2.6969211101531982,
"learning_rate": 2.123906469162774e-06,
"loss": 0.7482485175132751,
"step": 1230
},
{
"epoch": 1.6626180836707154,
"grad_norm": 3.7500598430633545,
"learning_rate": 2.1105065843555795e-06,
"loss": 0.4869375228881836,
"step": 1232
},
{
"epoch": 1.6653171390013495,
"grad_norm": 1.6825298070907593,
"learning_rate": 2.097201713537307e-06,
"loss": 0.6156188249588013,
"step": 1234
},
{
"epoch": 1.6680161943319838,
"grad_norm": 5.1931471824646,
"learning_rate": 2.083992122034976e-06,
"loss": 0.3456149101257324,
"step": 1236
},
{
"epoch": 1.670715249662618,
"grad_norm": 2.117143392562866,
"learning_rate": 2.07087807327554e-06,
"loss": 0.44808173179626465,
"step": 1238
},
{
"epoch": 1.6734143049932524,
"grad_norm": 1.7127982378005981,
"learning_rate": 2.0578598287806287e-06,
"loss": 0.36599722504615784,
"step": 1240
},
{
"epoch": 1.6761133603238867,
"grad_norm": 1.7033370733261108,
"learning_rate": 2.044937648161333e-06,
"loss": 0.5988764762878418,
"step": 1242
},
{
"epoch": 1.6788124156545208,
"grad_norm": 2.8658814430236816,
"learning_rate": 2.0321117891130313e-06,
"loss": 0.38747280836105347,
"step": 1244
},
{
"epoch": 1.6815114709851553,
"grad_norm": 4.310657024383545,
"learning_rate": 2.0193825074102454e-06,
"loss": 0.7432397603988647,
"step": 1246
},
{
"epoch": 1.6842105263157894,
"grad_norm": 4.345435619354248,
"learning_rate": 2.006750056901543e-06,
"loss": 0.5351995229721069,
"step": 1248
},
{
"epoch": 1.686909581646424,
"grad_norm": 1.7507013082504272,
"learning_rate": 1.9942146895044764e-06,
"loss": 0.7102921009063721,
"step": 1250
},
{
"epoch": 1.689608636977058,
"grad_norm": 1.8140771389007568,
"learning_rate": 1.9817766552005553e-06,
"loss": 0.5322377681732178,
"step": 1252
},
{
"epoch": 1.6923076923076923,
"grad_norm": 1.937525987625122,
"learning_rate": 1.969436202030262e-06,
"loss": 0.5792850852012634,
"step": 1254
},
{
"epoch": 1.6950067476383266,
"grad_norm": 1.4619972705841064,
"learning_rate": 1.957193576088111e-06,
"loss": 0.6791541576385498,
"step": 1256
},
{
"epoch": 1.697705802968961,
"grad_norm": 17.1634578704834,
"learning_rate": 1.9450490215177322e-06,
"loss": 0.46500182151794434,
"step": 1258
},
{
"epoch": 1.7004048582995952,
"grad_norm": 2.4044148921966553,
"learning_rate": 1.9330027805070092e-06,
"loss": 0.3937757611274719,
"step": 1260
},
{
"epoch": 1.7031039136302293,
"grad_norm": 2.509300708770752,
"learning_rate": 1.9210550932832463e-06,
"loss": 0.42199158668518066,
"step": 1262
},
{
"epoch": 1.7058029689608638,
"grad_norm": 3.170949935913086,
"learning_rate": 1.909206198108378e-06,
"loss": 0.6984668970108032,
"step": 1264
},
{
"epoch": 1.708502024291498,
"grad_norm": 1.8157047033309937,
"learning_rate": 1.8974563312742188e-06,
"loss": 0.47702908515930176,
"step": 1266
},
{
"epoch": 1.7112010796221324,
"grad_norm": 2.528952121734619,
"learning_rate": 1.8858057270977524e-06,
"loss": 0.555398166179657,
"step": 1268
},
{
"epoch": 1.7139001349527665,
"grad_norm": 2.8582353591918945,
"learning_rate": 1.874254617916454e-06,
"loss": 0.5454248189926147,
"step": 1270
},
{
"epoch": 1.7165991902834008,
"grad_norm": 1.347306728363037,
"learning_rate": 1.8628032340836596e-06,
"loss": 0.762856125831604,
"step": 1272
},
{
"epoch": 1.719298245614035,
"grad_norm": 4.5018229484558105,
"learning_rate": 1.8514518039639768e-06,
"loss": 0.3407154679298401,
"step": 1274
},
{
"epoch": 1.7219973009446694,
"grad_norm": 2.1817541122436523,
"learning_rate": 1.8402005539287227e-06,
"loss": 0.4855961799621582,
"step": 1276
},
{
"epoch": 1.7246963562753037,
"grad_norm": 2.419599771499634,
"learning_rate": 1.8290497083514133e-06,
"loss": 0.40242263674736023,
"step": 1278
},
{
"epoch": 1.7273954116059378,
"grad_norm": 2.2311151027679443,
"learning_rate": 1.817999489603293e-06,
"loss": 0.33770740032196045,
"step": 1280
},
{
"epoch": 1.7300944669365723,
"grad_norm": 1.1219701766967773,
"learning_rate": 1.8070501180488932e-06,
"loss": 0.2575661242008209,
"step": 1282
},
{
"epoch": 1.7327935222672064,
"grad_norm": 2.5874390602111816,
"learning_rate": 1.79620181204164e-06,
"loss": 0.435394287109375,
"step": 1284
},
{
"epoch": 1.7354925775978407,
"grad_norm": 1.4016201496124268,
"learning_rate": 1.7854547879195035e-06,
"loss": 0.38581591844558716,
"step": 1286
},
{
"epoch": 1.738191632928475,
"grad_norm": 1.244989275932312,
"learning_rate": 1.7748092600006785e-06,
"loss": 0.5393861532211304,
"step": 1288
},
{
"epoch": 1.7408906882591093,
"grad_norm": 3.02862286567688,
"learning_rate": 1.7642654405793119e-06,
"loss": 0.6616307497024536,
"step": 1290
},
{
"epoch": 1.7435897435897436,
"grad_norm": 2.827266216278076,
"learning_rate": 1.7538235399212737e-06,
"loss": 0.6251143217086792,
"step": 1292
},
{
"epoch": 1.746288798920378,
"grad_norm": 4.890537261962891,
"learning_rate": 1.7434837662599568e-06,
"loss": 0.3798263669013977,
"step": 1294
},
{
"epoch": 1.7489878542510122,
"grad_norm": 19.619352340698242,
"learning_rate": 1.7332463257921268e-06,
"loss": 0.2934688329696655,
"step": 1296
},
{
"epoch": 1.7516869095816463,
"grad_norm": 1.7478938102722168,
"learning_rate": 1.7231114226738156e-06,
"loss": 0.8473851680755615,
"step": 1298
},
{
"epoch": 1.7543859649122808,
"grad_norm": 1.7794421911239624,
"learning_rate": 1.7130792590162397e-06,
"loss": 0.5401483774185181,
"step": 1300
},
{
"epoch": 1.757085020242915,
"grad_norm": 1.8848437070846558,
"learning_rate": 1.7031500348817794e-06,
"loss": 0.877458930015564,
"step": 1302
},
{
"epoch": 1.7597840755735492,
"grad_norm": 1.747219443321228,
"learning_rate": 1.6933239482799831e-06,
"loss": 0.6049869656562805,
"step": 1304
},
{
"epoch": 1.7624831309041835,
"grad_norm": 2.7758688926696777,
"learning_rate": 1.6836011951636216e-06,
"loss": 0.46802860498428345,
"step": 1306
},
{
"epoch": 1.7651821862348178,
"grad_norm": 3.131430149078369,
"learning_rate": 1.6739819694247778e-06,
"loss": 0.7983341217041016,
"step": 1308
},
{
"epoch": 1.7678812415654521,
"grad_norm": 1.9045193195343018,
"learning_rate": 1.6644664628909834e-06,
"loss": 0.7016539573669434,
"step": 1310
},
{
"epoch": 1.7705802968960864,
"grad_norm": 2.361384630203247,
"learning_rate": 1.6550548653213892e-06,
"loss": 0.5561462640762329,
"step": 1312
},
{
"epoch": 1.7732793522267207,
"grad_norm": 3.259943962097168,
"learning_rate": 1.6457473644029872e-06,
"loss": 0.4066377282142639,
"step": 1314
},
{
"epoch": 1.7759784075573548,
"grad_norm": 2.0142383575439453,
"learning_rate": 1.636544145746861e-06,
"loss": 0.7119549512863159,
"step": 1316
},
{
"epoch": 1.7786774628879893,
"grad_norm": 2.7605931758880615,
"learning_rate": 1.6274453928844864e-06,
"loss": 0.516852617263794,
"step": 1318
},
{
"epoch": 1.7813765182186234,
"grad_norm": 1.3088191747665405,
"learning_rate": 1.6184512872640761e-06,
"loss": 0.37209126353263855,
"step": 1320
},
{
"epoch": 1.7840755735492577,
"grad_norm": 1.4621187448501587,
"learning_rate": 1.6095620082469544e-06,
"loss": 0.4754895269870758,
"step": 1322
},
{
"epoch": 1.786774628879892,
"grad_norm": 2.0048413276672363,
"learning_rate": 1.6007777331039826e-06,
"loss": 0.567115068435669,
"step": 1324
},
{
"epoch": 1.7894736842105263,
"grad_norm": 1.5625020265579224,
"learning_rate": 1.5920986370120274e-06,
"loss": 0.620823860168457,
"step": 1326
},
{
"epoch": 1.7921727395411606,
"grad_norm": 2.2350094318389893,
"learning_rate": 1.5835248930504632e-06,
"loss": 0.6646386384963989,
"step": 1328
},
{
"epoch": 1.7948717948717947,
"grad_norm": 1.4773541688919067,
"learning_rate": 1.575056672197722e-06,
"loss": 0.7824012041091919,
"step": 1330
},
{
"epoch": 1.7975708502024292,
"grad_norm": 2.12636661529541,
"learning_rate": 1.566694143327883e-06,
"loss": 0.6168233156204224,
"step": 1332
},
{
"epoch": 1.8002699055330633,
"grad_norm": 2.9663407802581787,
"learning_rate": 1.5584374732073067e-06,
"loss": 0.31019002199172974,
"step": 1334
},
{
"epoch": 1.8029689608636978,
"grad_norm": 1.7707102298736572,
"learning_rate": 1.5502868264913084e-06,
"loss": 0.5860764384269714,
"step": 1336
},
{
"epoch": 1.805668016194332,
"grad_norm": 1.6591790914535522,
"learning_rate": 1.5422423657208736e-06,
"loss": 0.728720486164093,
"step": 1338
},
{
"epoch": 1.8083670715249662,
"grad_norm": 20.809070587158203,
"learning_rate": 1.5343042513194182e-06,
"loss": 0.4329081177711487,
"step": 1340
},
{
"epoch": 1.8110661268556005,
"grad_norm": 1.4043638706207275,
"learning_rate": 1.5264726415895871e-06,
"loss": 0.690475583076477,
"step": 1342
},
{
"epoch": 1.8137651821862348,
"grad_norm": 1.4320697784423828,
"learning_rate": 1.5187476927101011e-06,
"loss": 0.3367083668708801,
"step": 1344
},
{
"epoch": 1.8164642375168691,
"grad_norm": 3.004023313522339,
"learning_rate": 1.5111295587326372e-06,
"loss": 0.8619272708892822,
"step": 1346
},
{
"epoch": 1.8191632928475032,
"grad_norm": 4.751201629638672,
"learning_rate": 1.5036183915787609e-06,
"loss": 0.20347274839878082,
"step": 1348
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.9431716799736023,
"learning_rate": 1.4962143410368955e-06,
"loss": 0.0988755151629448,
"step": 1350
},
{
"epoch": 1.8245614035087718,
"grad_norm": 0.4600701928138733,
"learning_rate": 1.4889175547593333e-06,
"loss": 0.4559515714645386,
"step": 1352
},
{
"epoch": 1.8272604588394064,
"grad_norm": 2.7570862770080566,
"learning_rate": 1.4817281782592927e-06,
"loss": 0.3706595301628113,
"step": 1354
},
{
"epoch": 1.8299595141700404,
"grad_norm": 1.6503052711486816,
"learning_rate": 1.4746463549080147e-06,
"loss": 0.3438398838043213,
"step": 1356
},
{
"epoch": 1.8326585695006747,
"grad_norm": 1.4355121850967407,
"learning_rate": 1.467672225931907e-06,
"loss": 0.6213847398757935,
"step": 1358
},
{
"epoch": 1.835357624831309,
"grad_norm": 1.1493998765945435,
"learning_rate": 1.4608059304097255e-06,
"loss": 0.8824546337127686,
"step": 1360
},
{
"epoch": 1.8380566801619433,
"grad_norm": 1.216939926147461,
"learning_rate": 1.4540476052697988e-06,
"loss": 0.7696167826652527,
"step": 1362
},
{
"epoch": 1.8407557354925776,
"grad_norm": 6.155215740203857,
"learning_rate": 1.4473973852873018e-06,
"loss": 0.6460918188095093,
"step": 1364
},
{
"epoch": 1.8434547908232117,
"grad_norm": 0.5168651342391968,
"learning_rate": 1.4408554030815654e-06,
"loss": 0.224877268075943,
"step": 1366
},
{
"epoch": 1.8461538461538463,
"grad_norm": 3.582855224609375,
"learning_rate": 1.434421789113432e-06,
"loss": 0.3557729721069336,
"step": 1368
},
{
"epoch": 1.8488529014844803,
"grad_norm": 9.36894702911377,
"learning_rate": 1.428096671682653e-06,
"loss": 0.8958494663238525,
"step": 1370
},
{
"epoch": 1.8515519568151149,
"grad_norm": 4.210360050201416,
"learning_rate": 1.4218801769253341e-06,
"loss": 0.7161762714385986,
"step": 1372
},
{
"epoch": 1.854251012145749,
"grad_norm": 2.4463114738464355,
"learning_rate": 1.415772428811415e-06,
"loss": 0.028064336627721786,
"step": 1374
},
{
"epoch": 1.8569500674763832,
"grad_norm": 2.058156967163086,
"learning_rate": 1.4097735491421994e-06,
"loss": 0.6387984752655029,
"step": 1376
},
{
"epoch": 1.8596491228070176,
"grad_norm": 1.5363142490386963,
"learning_rate": 1.403883657547927e-06,
"loss": 0.697824239730835,
"step": 1378
},
{
"epoch": 1.8623481781376519,
"grad_norm": 4.654923439025879,
"learning_rate": 1.3981028714853856e-06,
"loss": 0.5086572766304016,
"step": 1380
},
{
"epoch": 1.8650472334682862,
"grad_norm": 1.4995976686477661,
"learning_rate": 1.39243130623557e-06,
"loss": 0.6153897047042847,
"step": 1382
},
{
"epoch": 1.8677462887989202,
"grad_norm": 1.5821216106414795,
"learning_rate": 1.3868690749013852e-06,
"loss": 0.4210359454154968,
"step": 1384
},
{
"epoch": 1.8704453441295548,
"grad_norm": 1.385354995727539,
"learning_rate": 1.3814162884053855e-06,
"loss": 0.8175305128097534,
"step": 1386
},
{
"epoch": 1.8731443994601888,
"grad_norm": 1.201923131942749,
"learning_rate": 1.376073055487567e-06,
"loss": 0.4728902578353882,
"step": 1388
},
{
"epoch": 1.8758434547908234,
"grad_norm": 2.1456212997436523,
"learning_rate": 1.3708394827031984e-06,
"loss": 0.621558427810669,
"step": 1390
},
{
"epoch": 1.8785425101214575,
"grad_norm": 2.024261236190796,
"learning_rate": 1.365715674420695e-06,
"loss": 0.6893987655639648,
"step": 1392
},
{
"epoch": 1.8812415654520918,
"grad_norm": 3.368173837661743,
"learning_rate": 1.360701732819537e-06,
"loss": 0.4840161204338074,
"step": 1394
},
{
"epoch": 1.883940620782726,
"grad_norm": 2.0906198024749756,
"learning_rate": 1.3557977578882333e-06,
"loss": 0.360158234834671,
"step": 1396
},
{
"epoch": 1.8866396761133604,
"grad_norm": 2.162259817123413,
"learning_rate": 1.3510038474223268e-06,
"loss": 0.35252755880355835,
"step": 1398
},
{
"epoch": 1.8893387314439947,
"grad_norm": 2.556044340133667,
"learning_rate": 1.346320097022444e-06,
"loss": 0.7171377539634705,
"step": 1400
},
{
"epoch": 1.8920377867746287,
"grad_norm": 3.5639069080352783,
"learning_rate": 1.3417466000923887e-06,
"loss": 0.45300090312957764,
"step": 1402
},
{
"epoch": 1.8947368421052633,
"grad_norm": 4.419144630432129,
"learning_rate": 1.3372834478372803e-06,
"loss": 0.72588050365448,
"step": 1404
},
{
"epoch": 1.8974358974358974,
"grad_norm": 1.746862769126892,
"learning_rate": 1.3329307292617312e-06,
"loss": 0.4953262209892273,
"step": 1406
},
{
"epoch": 1.9001349527665317,
"grad_norm": 1.7869200706481934,
"learning_rate": 1.3286885311680783e-06,
"loss": 0.4988870322704315,
"step": 1408
},
{
"epoch": 1.902834008097166,
"grad_norm": 3.3469603061676025,
"learning_rate": 1.3245569381546465e-06,
"loss": 0.3246951699256897,
"step": 1410
},
{
"epoch": 1.9055330634278003,
"grad_norm": 2.1499931812286377,
"learning_rate": 1.320536032614063e-06,
"loss": 0.6533905267715454,
"step": 1412
},
{
"epoch": 1.9082321187584346,
"grad_norm": 2.3623106479644775,
"learning_rate": 1.3166258947316151e-06,
"loss": 0.5628916025161743,
"step": 1414
},
{
"epoch": 1.9109311740890689,
"grad_norm": 2.8255856037139893,
"learning_rate": 1.312826602483652e-06,
"loss": 0.09494273364543915,
"step": 1416
},
{
"epoch": 1.9136302294197032,
"grad_norm": 1.3409160375595093,
"learning_rate": 1.309138231636027e-06,
"loss": 0.8202773332595825,
"step": 1418
},
{
"epoch": 1.9163292847503373,
"grad_norm": 8.893857955932617,
"learning_rate": 1.3055608557425885e-06,
"loss": 0.464354932308197,
"step": 1420
},
{
"epoch": 1.9190283400809718,
"grad_norm": 3.1871230602264404,
"learning_rate": 1.3020945461437132e-06,
"loss": 0.5134360790252686,
"step": 1422
},
{
"epoch": 1.9217273954116059,
"grad_norm": 4.8007354736328125,
"learning_rate": 1.2987393719648823e-06,
"loss": 0.4733040928840637,
"step": 1424
},
{
"epoch": 1.9244264507422402,
"grad_norm": 1.6076239347457886,
"learning_rate": 1.2954954001153042e-06,
"loss": 0.4948502779006958,
"step": 1426
},
{
"epoch": 1.9271255060728745,
"grad_norm": 1.6794874668121338,
"learning_rate": 1.2923626952865786e-06,
"loss": 0.6524335741996765,
"step": 1428
},
{
"epoch": 1.9298245614035088,
"grad_norm": 4.740302085876465,
"learning_rate": 1.2893413199514102e-06,
"loss": 0.7467541694641113,
"step": 1430
},
{
"epoch": 1.932523616734143,
"grad_norm": 1.1458661556243896,
"learning_rate": 1.286431334362358e-06,
"loss": 0.47734931111335754,
"step": 1432
},
{
"epoch": 1.9352226720647774,
"grad_norm": 2.340743064880371,
"learning_rate": 1.283632796550635e-06,
"loss": 0.3793056607246399,
"step": 1434
},
{
"epoch": 1.9379217273954117,
"grad_norm": 2.9522817134857178,
"learning_rate": 1.2809457623249545e-06,
"loss": 0.7396039962768555,
"step": 1436
},
{
"epoch": 1.9406207827260458,
"grad_norm": 1.8434336185455322,
"learning_rate": 1.2783702852704132e-06,
"loss": 0.4388154149055481,
"step": 1438
},
{
"epoch": 1.9433198380566803,
"grad_norm": 2.278200626373291,
"learning_rate": 1.2759064167474226e-06,
"loss": 0.3468833863735199,
"step": 1440
},
{
"epoch": 1.9460188933873144,
"grad_norm": 1.7238056659698486,
"learning_rate": 1.2735542058906884e-06,
"loss": 0.6383070945739746,
"step": 1442
},
{
"epoch": 1.9487179487179487,
"grad_norm": 3.111173391342163,
"learning_rate": 1.2713136996082269e-06,
"loss": 0.6295847296714783,
"step": 1444
},
{
"epoch": 1.951417004048583,
"grad_norm": 1.7762092351913452,
"learning_rate": 1.2691849425804317e-06,
"loss": 0.6602956056594849,
"step": 1446
},
{
"epoch": 1.9541160593792173,
"grad_norm": 1.167322039604187,
"learning_rate": 1.2671679772591808e-06,
"loss": 0.46681874990463257,
"step": 1448
},
{
"epoch": 1.9568151147098516,
"grad_norm": 3.337437152862549,
"learning_rate": 1.265262843866992e-06,
"loss": 0.5883115530014038,
"step": 1450
},
{
"epoch": 1.9595141700404857,
"grad_norm": 2.6898999214172363,
"learning_rate": 1.2634695803962196e-06,
"loss": 0.600220263004303,
"step": 1452
},
{
"epoch": 1.9622132253711202,
"grad_norm": 1.3048210144042969,
"learning_rate": 1.2617882226082982e-06,
"loss": 0.8725317716598511,
"step": 1454
},
{
"epoch": 1.9649122807017543,
"grad_norm": 1.329509973526001,
"learning_rate": 1.2602188040330271e-06,
"loss": 0.7059814929962158,
"step": 1456
},
{
"epoch": 1.9676113360323888,
"grad_norm": 1.0768544673919678,
"learning_rate": 1.2587613559679029e-06,
"loss": 0.4005241394042969,
"step": 1458
},
{
"epoch": 1.9703103913630229,
"grad_norm": 1.7787357568740845,
"learning_rate": 1.2574159074774968e-06,
"loss": 0.5970302820205688,
"step": 1460
},
{
"epoch": 1.9730094466936572,
"grad_norm": 1.7534894943237305,
"learning_rate": 1.2561824853928736e-06,
"loss": 0.8058260679244995,
"step": 1462
},
{
"epoch": 1.9757085020242915,
"grad_norm": 1.4299421310424805,
"learning_rate": 1.255061114311054e-06,
"loss": 0.49718576669692993,
"step": 1464
},
{
"epoch": 1.9784075573549258,
"grad_norm": 1.9105091094970703,
"learning_rate": 1.2540518165945298e-06,
"loss": 0.486233115196228,
"step": 1466
},
{
"epoch": 1.98110661268556,
"grad_norm": 1.4206773042678833,
"learning_rate": 1.2531546123708135e-06,
"loss": 0.15221592783927917,
"step": 1468
},
{
"epoch": 1.9838056680161942,
"grad_norm": 2.441514730453491,
"learning_rate": 1.252369519532038e-06,
"loss": 0.655547022819519,
"step": 1470
},
{
"epoch": 1.9865047233468287,
"grad_norm": 1.990182876586914,
"learning_rate": 1.2516965537346026e-06,
"loss": 0.41459017992019653,
"step": 1472
},
{
"epoch": 1.9892037786774628,
"grad_norm": 5.279425144195557,
"learning_rate": 1.2511357283988542e-06,
"loss": 0.651479959487915,
"step": 1474
},
{
"epoch": 1.9919028340080973,
"grad_norm": 1.2038631439208984,
"learning_rate": 1.250687054708827e-06,
"loss": 0.7315380573272705,
"step": 1476
},
{
"epoch": 1.9946018893387314,
"grad_norm": 1.1363749504089355,
"learning_rate": 1.2503505416120152e-06,
"loss": 0.6686517596244812,
"step": 1478
},
{
"epoch": 1.9973009446693657,
"grad_norm": 3.964367628097534,
"learning_rate": 1.2501261958191954e-06,
"loss": 0.2313126027584076,
"step": 1480
},
{
"epoch": 2.0,
"grad_norm": 0.6272908449172974,
"learning_rate": 1.2500140218042928e-06,
"loss": 0.46454429626464844,
"step": 1482
},
{
"epoch": 2.0,
"step": 1482,
"total_flos": 3.038711167334744e+18,
"train_loss": 0.6924078768212223,
"train_runtime": 3699.7491,
"train_samples_per_second": 12.017,
"train_steps_per_second": 0.401
}
],
"logging_steps": 2,
"max_steps": 1482,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.038711167334744e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}