misultrachat400k8_4_2024 / trainer_state.json
RadAlienware's picture
checkpoint-12400-uploaded-manually
ff5a6b8 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9899409228804088,
"eval_steps": 500,
"global_step": 12400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 3.703125,
"learning_rate": 1.9999968642467102e-05,
"loss": 4.2386,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 1.8515625,
"learning_rate": 1.999987438156715e-05,
"loss": 3.1965,
"step": 40
},
{
"epoch": 0.0,
"grad_norm": 1.4921875,
"learning_rate": 1.9999717217822316e-05,
"loss": 2.7844,
"step": 60
},
{
"epoch": 0.01,
"grad_norm": 1.25,
"learning_rate": 1.999949715222121e-05,
"loss": 2.6013,
"step": 80
},
{
"epoch": 0.01,
"grad_norm": 2.28125,
"learning_rate": 1.9999214186148133e-05,
"loss": 2.5417,
"step": 100
},
{
"epoch": 0.01,
"grad_norm": 1.1953125,
"learning_rate": 1.9998868321383038e-05,
"loss": 2.4376,
"step": 120
},
{
"epoch": 0.01,
"grad_norm": 1.25,
"learning_rate": 1.9998459560101546e-05,
"loss": 2.3875,
"step": 140
},
{
"epoch": 0.01,
"grad_norm": 1.3359375,
"learning_rate": 1.9997987904874905e-05,
"loss": 2.3568,
"step": 160
},
{
"epoch": 0.01,
"grad_norm": 1.359375,
"learning_rate": 1.9997453358670004e-05,
"loss": 2.3034,
"step": 180
},
{
"epoch": 0.02,
"grad_norm": 1.9296875,
"learning_rate": 1.9996855924849337e-05,
"loss": 2.2779,
"step": 200
},
{
"epoch": 0.02,
"grad_norm": 1.6015625,
"learning_rate": 1.999619560717097e-05,
"loss": 2.2728,
"step": 220
},
{
"epoch": 0.02,
"grad_norm": 1.4921875,
"learning_rate": 1.9995472409788548e-05,
"loss": 2.2436,
"step": 240
},
{
"epoch": 0.02,
"grad_norm": 1.625,
"learning_rate": 1.999468633725125e-05,
"loss": 2.2062,
"step": 260
},
{
"epoch": 0.02,
"grad_norm": 1.6875,
"learning_rate": 1.9993837394503745e-05,
"loss": 2.1873,
"step": 280
},
{
"epoch": 0.02,
"grad_norm": 2.328125,
"learning_rate": 1.99929255868862e-05,
"loss": 2.1973,
"step": 300
},
{
"epoch": 0.03,
"grad_norm": 1.5546875,
"learning_rate": 1.999195092013422e-05,
"loss": 2.1891,
"step": 320
},
{
"epoch": 0.03,
"grad_norm": 2.8125,
"learning_rate": 1.99909134003788e-05,
"loss": 2.1813,
"step": 340
},
{
"epoch": 0.03,
"grad_norm": 2.1875,
"learning_rate": 1.998981303414633e-05,
"loss": 2.1609,
"step": 360
},
{
"epoch": 0.03,
"grad_norm": 2.234375,
"learning_rate": 1.9988649828358504e-05,
"loss": 2.1693,
"step": 380
},
{
"epoch": 0.03,
"grad_norm": 1.8359375,
"learning_rate": 1.9987423790332315e-05,
"loss": 2.1465,
"step": 400
},
{
"epoch": 0.03,
"grad_norm": 1.71875,
"learning_rate": 1.9986134927779986e-05,
"loss": 2.1387,
"step": 420
},
{
"epoch": 0.04,
"grad_norm": 1.734375,
"learning_rate": 1.998478324880893e-05,
"loss": 2.1236,
"step": 440
},
{
"epoch": 0.04,
"grad_norm": 1.7265625,
"learning_rate": 1.9983368761921703e-05,
"loss": 2.1144,
"step": 460
},
{
"epoch": 0.04,
"grad_norm": 1.640625,
"learning_rate": 1.9981891476015936e-05,
"loss": 2.1164,
"step": 480
},
{
"epoch": 0.04,
"grad_norm": 1.65625,
"learning_rate": 1.99803514003843e-05,
"loss": 2.1083,
"step": 500
},
{
"epoch": 0.04,
"grad_norm": 1.8515625,
"learning_rate": 1.9978748544714427e-05,
"loss": 2.0906,
"step": 520
},
{
"epoch": 0.04,
"grad_norm": 1.890625,
"learning_rate": 1.997708291908886e-05,
"loss": 2.1043,
"step": 540
},
{
"epoch": 0.04,
"grad_norm": 2.0,
"learning_rate": 1.9975354533984995e-05,
"loss": 2.1028,
"step": 560
},
{
"epoch": 0.05,
"grad_norm": 1.8984375,
"learning_rate": 1.9973563400274994e-05,
"loss": 2.082,
"step": 580
},
{
"epoch": 0.05,
"grad_norm": 2.1875,
"learning_rate": 1.9971709529225754e-05,
"loss": 2.0806,
"step": 600
},
{
"epoch": 0.05,
"grad_norm": 2.375,
"learning_rate": 1.9969792932498783e-05,
"loss": 2.0803,
"step": 620
},
{
"epoch": 0.05,
"grad_norm": 1.84375,
"learning_rate": 1.9967813622150177e-05,
"loss": 2.0731,
"step": 640
},
{
"epoch": 0.05,
"grad_norm": 2.015625,
"learning_rate": 1.996577161063052e-05,
"loss": 2.0662,
"step": 660
},
{
"epoch": 0.05,
"grad_norm": 1.859375,
"learning_rate": 1.99636669107848e-05,
"loss": 2.0472,
"step": 680
},
{
"epoch": 0.06,
"grad_norm": 1.9609375,
"learning_rate": 1.996149953585235e-05,
"loss": 2.0562,
"step": 700
},
{
"epoch": 0.06,
"grad_norm": 1.84375,
"learning_rate": 1.9959269499466746e-05,
"loss": 2.0587,
"step": 720
},
{
"epoch": 0.06,
"grad_norm": 1.890625,
"learning_rate": 1.9956976815655723e-05,
"loss": 2.0576,
"step": 740
},
{
"epoch": 0.06,
"grad_norm": 1.9609375,
"learning_rate": 1.99546214988411e-05,
"loss": 2.0508,
"step": 760
},
{
"epoch": 0.06,
"grad_norm": 1.953125,
"learning_rate": 1.9952203563838676e-05,
"loss": 2.034,
"step": 780
},
{
"epoch": 0.06,
"grad_norm": 1.8046875,
"learning_rate": 1.9949723025858136e-05,
"loss": 2.0259,
"step": 800
},
{
"epoch": 0.07,
"grad_norm": 2.078125,
"learning_rate": 1.994717990050297e-05,
"loss": 2.0439,
"step": 820
},
{
"epoch": 0.07,
"grad_norm": 2.875,
"learning_rate": 1.9944574203770365e-05,
"loss": 2.0371,
"step": 840
},
{
"epoch": 0.07,
"grad_norm": 2.515625,
"learning_rate": 1.994190595205109e-05,
"loss": 2.0375,
"step": 860
},
{
"epoch": 0.07,
"grad_norm": 2.0625,
"learning_rate": 1.9939175162129427e-05,
"loss": 2.0227,
"step": 880
},
{
"epoch": 0.07,
"grad_norm": 2.078125,
"learning_rate": 1.9936381851183032e-05,
"loss": 2.0182,
"step": 900
},
{
"epoch": 0.07,
"grad_norm": 1.8203125,
"learning_rate": 1.9933526036782852e-05,
"loss": 2.0208,
"step": 920
},
{
"epoch": 0.08,
"grad_norm": 1.9921875,
"learning_rate": 1.993060773689299e-05,
"loss": 2.0177,
"step": 940
},
{
"epoch": 0.08,
"grad_norm": 1.8125,
"learning_rate": 1.992762696987062e-05,
"loss": 2.0208,
"step": 960
},
{
"epoch": 0.08,
"grad_norm": 2.0,
"learning_rate": 1.9924583754465842e-05,
"loss": 1.9938,
"step": 980
},
{
"epoch": 0.08,
"grad_norm": 2.109375,
"learning_rate": 1.9921478109821598e-05,
"loss": 2.0132,
"step": 1000
},
{
"epoch": 0.08,
"grad_norm": 1.8046875,
"learning_rate": 1.9918310055473515e-05,
"loss": 2.0062,
"step": 1020
},
{
"epoch": 0.08,
"grad_norm": 1.8828125,
"learning_rate": 1.991507961134981e-05,
"loss": 2.0074,
"step": 1040
},
{
"epoch": 0.08,
"grad_norm": 2.046875,
"learning_rate": 1.9911786797771144e-05,
"loss": 2.0153,
"step": 1060
},
{
"epoch": 0.09,
"grad_norm": 2.078125,
"learning_rate": 1.990843163545052e-05,
"loss": 1.996,
"step": 1080
},
{
"epoch": 0.09,
"grad_norm": 2.21875,
"learning_rate": 1.990501414549312e-05,
"loss": 2.0067,
"step": 1100
},
{
"epoch": 0.09,
"grad_norm": 1.8359375,
"learning_rate": 1.9901534349396204e-05,
"loss": 1.9922,
"step": 1120
},
{
"epoch": 0.09,
"grad_norm": 2.25,
"learning_rate": 1.9897992269048953e-05,
"loss": 1.9953,
"step": 1140
},
{
"epoch": 0.09,
"grad_norm": 2.109375,
"learning_rate": 1.9894387926732342e-05,
"loss": 1.9968,
"step": 1160
},
{
"epoch": 0.09,
"grad_norm": 2.390625,
"learning_rate": 1.9890721345118987e-05,
"loss": 1.9851,
"step": 1180
},
{
"epoch": 0.1,
"grad_norm": 2.171875,
"learning_rate": 1.988699254727303e-05,
"loss": 1.9749,
"step": 1200
},
{
"epoch": 0.1,
"grad_norm": 2.25,
"learning_rate": 1.988320155664996e-05,
"loss": 2.003,
"step": 1220
},
{
"epoch": 0.1,
"grad_norm": 2.140625,
"learning_rate": 1.9879348397096482e-05,
"loss": 1.9779,
"step": 1240
},
{
"epoch": 0.1,
"grad_norm": 2.09375,
"learning_rate": 1.9875433092850376e-05,
"loss": 1.9633,
"step": 1260
},
{
"epoch": 0.1,
"grad_norm": 1.9921875,
"learning_rate": 1.9871455668540325e-05,
"loss": 1.9824,
"step": 1280
},
{
"epoch": 0.1,
"grad_norm": 2.3125,
"learning_rate": 1.9867416149185774e-05,
"loss": 1.9785,
"step": 1300
},
{
"epoch": 0.11,
"grad_norm": 2.265625,
"learning_rate": 1.9863314560196775e-05,
"loss": 1.9923,
"step": 1320
},
{
"epoch": 0.11,
"grad_norm": 2.140625,
"learning_rate": 1.9859150927373803e-05,
"loss": 1.9839,
"step": 1340
},
{
"epoch": 0.11,
"grad_norm": 2.4375,
"learning_rate": 1.9854925276907627e-05,
"loss": 1.985,
"step": 1360
},
{
"epoch": 0.11,
"grad_norm": 1.953125,
"learning_rate": 1.985063763537913e-05,
"loss": 1.974,
"step": 1380
},
{
"epoch": 0.11,
"grad_norm": 1.9453125,
"learning_rate": 1.9846288029759124e-05,
"loss": 1.9801,
"step": 1400
},
{
"epoch": 0.11,
"grad_norm": 2.03125,
"learning_rate": 1.984187648740822e-05,
"loss": 1.9733,
"step": 1420
},
{
"epoch": 0.11,
"grad_norm": 2.046875,
"learning_rate": 1.983740303607662e-05,
"loss": 1.9653,
"step": 1440
},
{
"epoch": 0.12,
"grad_norm": 2.4375,
"learning_rate": 1.9832867703903953e-05,
"loss": 1.9672,
"step": 1460
},
{
"epoch": 0.12,
"grad_norm": 2.15625,
"learning_rate": 1.9828270519419115e-05,
"loss": 1.9625,
"step": 1480
},
{
"epoch": 0.12,
"grad_norm": 2.0625,
"learning_rate": 1.9823611511540064e-05,
"loss": 1.9542,
"step": 1500
},
{
"epoch": 0.12,
"grad_norm": 2.328125,
"learning_rate": 1.9818890709573652e-05,
"loss": 1.9475,
"step": 1520
},
{
"epoch": 0.12,
"grad_norm": 2.234375,
"learning_rate": 1.9814108143215446e-05,
"loss": 1.9642,
"step": 1540
},
{
"epoch": 0.12,
"grad_norm": 2.015625,
"learning_rate": 1.9809263842549516e-05,
"loss": 1.9541,
"step": 1560
},
{
"epoch": 0.13,
"grad_norm": 1.9921875,
"learning_rate": 1.980435783804828e-05,
"loss": 1.956,
"step": 1580
},
{
"epoch": 0.13,
"grad_norm": 1.96875,
"learning_rate": 1.9799390160572295e-05,
"loss": 1.9812,
"step": 1600
},
{
"epoch": 0.13,
"grad_norm": 1.953125,
"learning_rate": 1.979436084137005e-05,
"loss": 1.9617,
"step": 1620
},
{
"epoch": 0.13,
"grad_norm": 2.125,
"learning_rate": 1.9789269912077792e-05,
"loss": 1.9534,
"step": 1640
},
{
"epoch": 0.13,
"grad_norm": 2.140625,
"learning_rate": 1.9784117404719324e-05,
"loss": 1.9519,
"step": 1660
},
{
"epoch": 0.13,
"grad_norm": 2.046875,
"learning_rate": 1.977890335170578e-05,
"loss": 1.9448,
"step": 1680
},
{
"epoch": 0.14,
"grad_norm": 2.21875,
"learning_rate": 1.9773627785835454e-05,
"loss": 1.9361,
"step": 1700
},
{
"epoch": 0.14,
"grad_norm": 2.265625,
"learning_rate": 1.9768290740293573e-05,
"loss": 1.9485,
"step": 1720
},
{
"epoch": 0.14,
"grad_norm": 2.046875,
"learning_rate": 1.9762892248652093e-05,
"loss": 1.9356,
"step": 1740
},
{
"epoch": 0.14,
"grad_norm": 2.453125,
"learning_rate": 1.975743234486949e-05,
"loss": 1.9484,
"step": 1760
},
{
"epoch": 0.14,
"grad_norm": 2.28125,
"learning_rate": 1.9751911063290542e-05,
"loss": 1.9358,
"step": 1780
},
{
"epoch": 0.14,
"grad_norm": 2.203125,
"learning_rate": 1.974632843864612e-05,
"loss": 1.9453,
"step": 1800
},
{
"epoch": 0.15,
"grad_norm": 2.359375,
"learning_rate": 1.9740684506052958e-05,
"loss": 1.9217,
"step": 1820
},
{
"epoch": 0.15,
"grad_norm": 2.171875,
"learning_rate": 1.9734979301013445e-05,
"loss": 1.9243,
"step": 1840
},
{
"epoch": 0.15,
"grad_norm": 2.09375,
"learning_rate": 1.9729212859415397e-05,
"loss": 1.9421,
"step": 1860
},
{
"epoch": 0.15,
"grad_norm": 2.625,
"learning_rate": 1.9723385217531824e-05,
"loss": 1.9311,
"step": 1880
},
{
"epoch": 0.15,
"grad_norm": 1.9609375,
"learning_rate": 1.9717496412020717e-05,
"loss": 1.9402,
"step": 1900
},
{
"epoch": 0.15,
"grad_norm": 2.140625,
"learning_rate": 1.9711546479924797e-05,
"loss": 1.9433,
"step": 1920
},
{
"epoch": 0.15,
"grad_norm": 2.203125,
"learning_rate": 1.9705535458671304e-05,
"loss": 1.9181,
"step": 1940
},
{
"epoch": 0.16,
"grad_norm": 2.375,
"learning_rate": 1.9699463386071748e-05,
"loss": 1.929,
"step": 1960
},
{
"epoch": 0.16,
"grad_norm": 2.25,
"learning_rate": 1.9693330300321666e-05,
"loss": 1.941,
"step": 1980
},
{
"epoch": 0.16,
"grad_norm": 2.3125,
"learning_rate": 1.96871362400004e-05,
"loss": 1.9172,
"step": 2000
},
{
"epoch": 0.16,
"grad_norm": 2.375,
"learning_rate": 1.9680881244070848e-05,
"loss": 1.9103,
"step": 2020
},
{
"epoch": 0.16,
"grad_norm": 1.9921875,
"learning_rate": 1.96745653518792e-05,
"loss": 1.9323,
"step": 2040
},
{
"epoch": 0.16,
"grad_norm": 2.015625,
"learning_rate": 1.9668188603154716e-05,
"loss": 1.9333,
"step": 2060
},
{
"epoch": 0.17,
"grad_norm": 2.640625,
"learning_rate": 1.9661751038009463e-05,
"loss": 1.9243,
"step": 2080
},
{
"epoch": 0.17,
"grad_norm": 2.03125,
"learning_rate": 1.965525269693807e-05,
"loss": 1.9386,
"step": 2100
},
{
"epoch": 0.17,
"grad_norm": 2.8125,
"learning_rate": 1.9648693620817455e-05,
"loss": 1.9293,
"step": 2120
},
{
"epoch": 0.17,
"grad_norm": 2.125,
"learning_rate": 1.96420738509066e-05,
"loss": 1.9175,
"step": 2140
},
{
"epoch": 0.17,
"grad_norm": 1.90625,
"learning_rate": 1.963539342884626e-05,
"loss": 1.9176,
"step": 2160
},
{
"epoch": 0.17,
"grad_norm": 2.15625,
"learning_rate": 1.9628652396658725e-05,
"loss": 1.9182,
"step": 2180
},
{
"epoch": 0.18,
"grad_norm": 2.015625,
"learning_rate": 1.9621850796747528e-05,
"loss": 1.9048,
"step": 2200
},
{
"epoch": 0.18,
"grad_norm": 2.4375,
"learning_rate": 1.9614988671897208e-05,
"loss": 1.9209,
"step": 2220
},
{
"epoch": 0.18,
"grad_norm": 2.1875,
"learning_rate": 1.960806606527303e-05,
"loss": 1.9064,
"step": 2240
},
{
"epoch": 0.18,
"grad_norm": 2.515625,
"learning_rate": 1.96010830204207e-05,
"loss": 1.9192,
"step": 2260
},
{
"epoch": 0.18,
"grad_norm": 2.203125,
"learning_rate": 1.9594039581266107e-05,
"loss": 1.9326,
"step": 2280
},
{
"epoch": 0.18,
"grad_norm": 2.0,
"learning_rate": 1.958693579211505e-05,
"loss": 1.9194,
"step": 2300
},
{
"epoch": 0.19,
"grad_norm": 2.0625,
"learning_rate": 1.957977169765294e-05,
"loss": 1.8903,
"step": 2320
},
{
"epoch": 0.19,
"grad_norm": 2.703125,
"learning_rate": 1.957254734294454e-05,
"loss": 1.9135,
"step": 2340
},
{
"epoch": 0.19,
"grad_norm": 2.125,
"learning_rate": 1.956526277343366e-05,
"loss": 1.9228,
"step": 2360
},
{
"epoch": 0.19,
"grad_norm": 2.28125,
"learning_rate": 1.95579180349429e-05,
"loss": 1.9094,
"step": 2380
},
{
"epoch": 0.19,
"grad_norm": 2.015625,
"learning_rate": 1.955051317367333e-05,
"loss": 1.9102,
"step": 2400
},
{
"epoch": 0.19,
"grad_norm": 2.109375,
"learning_rate": 1.9543048236204215e-05,
"loss": 1.8987,
"step": 2420
},
{
"epoch": 0.19,
"grad_norm": 2.328125,
"learning_rate": 1.9535523269492733e-05,
"loss": 1.9124,
"step": 2440
},
{
"epoch": 0.2,
"grad_norm": 2.171875,
"learning_rate": 1.9527938320873652e-05,
"loss": 1.9137,
"step": 2460
},
{
"epoch": 0.2,
"grad_norm": 2.078125,
"learning_rate": 1.9520293438059065e-05,
"loss": 1.9078,
"step": 2480
},
{
"epoch": 0.2,
"grad_norm": 1.9765625,
"learning_rate": 1.9512588669138055e-05,
"loss": 1.9092,
"step": 2500
},
{
"epoch": 0.2,
"grad_norm": 2.09375,
"learning_rate": 1.9504824062576425e-05,
"loss": 1.9114,
"step": 2520
},
{
"epoch": 0.2,
"grad_norm": 2.046875,
"learning_rate": 1.949699966721637e-05,
"loss": 1.9121,
"step": 2540
},
{
"epoch": 0.2,
"grad_norm": 2.171875,
"learning_rate": 1.9489115532276182e-05,
"loss": 1.9139,
"step": 2560
},
{
"epoch": 0.21,
"grad_norm": 2.0625,
"learning_rate": 1.9481171707349936e-05,
"loss": 1.8889,
"step": 2580
},
{
"epoch": 0.21,
"grad_norm": 2.03125,
"learning_rate": 1.9473168242407183e-05,
"loss": 1.9233,
"step": 2600
},
{
"epoch": 0.21,
"grad_norm": 2.203125,
"learning_rate": 1.9465105187792617e-05,
"loss": 1.8928,
"step": 2620
},
{
"epoch": 0.21,
"grad_norm": 2.15625,
"learning_rate": 1.9456982594225787e-05,
"loss": 1.9101,
"step": 2640
},
{
"epoch": 0.21,
"grad_norm": 2.234375,
"learning_rate": 1.9448800512800762e-05,
"loss": 1.8862,
"step": 2660
},
{
"epoch": 0.21,
"grad_norm": 2.203125,
"learning_rate": 1.9440558994985805e-05,
"loss": 1.8912,
"step": 2680
},
{
"epoch": 0.22,
"grad_norm": 2.3125,
"learning_rate": 1.943225809262306e-05,
"loss": 1.8983,
"step": 2700
},
{
"epoch": 0.22,
"grad_norm": 2.28125,
"learning_rate": 1.942389785792822e-05,
"loss": 1.9031,
"step": 2720
},
{
"epoch": 0.22,
"grad_norm": 2.0625,
"learning_rate": 1.94154783434902e-05,
"loss": 1.9023,
"step": 2740
},
{
"epoch": 0.22,
"grad_norm": 2.46875,
"learning_rate": 1.940699960227081e-05,
"loss": 1.8974,
"step": 2760
},
{
"epoch": 0.22,
"grad_norm": 2.0625,
"learning_rate": 1.939846168760441e-05,
"loss": 1.9007,
"step": 2780
},
{
"epoch": 0.22,
"grad_norm": 2.15625,
"learning_rate": 1.938986465319759e-05,
"loss": 1.8949,
"step": 2800
},
{
"epoch": 0.23,
"grad_norm": 2.375,
"learning_rate": 1.9381208553128813e-05,
"loss": 1.8864,
"step": 2820
},
{
"epoch": 0.23,
"grad_norm": 2.359375,
"learning_rate": 1.9372493441848105e-05,
"loss": 1.9024,
"step": 2840
},
{
"epoch": 0.23,
"grad_norm": 2.109375,
"learning_rate": 1.9363719374176683e-05,
"loss": 1.8891,
"step": 2860
},
{
"epoch": 0.23,
"grad_norm": 2.28125,
"learning_rate": 1.935488640530662e-05,
"loss": 1.8849,
"step": 2880
},
{
"epoch": 0.23,
"grad_norm": 2.234375,
"learning_rate": 1.9345994590800498e-05,
"loss": 1.8939,
"step": 2900
},
{
"epoch": 0.23,
"grad_norm": 2.453125,
"learning_rate": 1.9337043986591064e-05,
"loss": 1.8903,
"step": 2920
},
{
"epoch": 0.23,
"grad_norm": 2.265625,
"learning_rate": 1.9328034648980874e-05,
"loss": 1.8731,
"step": 2940
},
{
"epoch": 0.24,
"grad_norm": 2.203125,
"learning_rate": 1.9318966634641936e-05,
"loss": 1.8781,
"step": 2960
},
{
"epoch": 0.24,
"grad_norm": 2.109375,
"learning_rate": 1.9309840000615358e-05,
"loss": 1.8855,
"step": 2980
},
{
"epoch": 0.24,
"grad_norm": 2.03125,
"learning_rate": 1.930065480431098e-05,
"loss": 1.89,
"step": 3000
},
{
"epoch": 0.24,
"grad_norm": 2.53125,
"learning_rate": 1.9291411103507033e-05,
"loss": 1.878,
"step": 3020
},
{
"epoch": 0.24,
"grad_norm": 2.453125,
"learning_rate": 1.9282108956349754e-05,
"loss": 1.8896,
"step": 3040
},
{
"epoch": 0.24,
"grad_norm": 2.140625,
"learning_rate": 1.9272748421353023e-05,
"loss": 1.8763,
"step": 3060
},
{
"epoch": 0.25,
"grad_norm": 2.015625,
"learning_rate": 1.9263329557398012e-05,
"loss": 1.8741,
"step": 3080
},
{
"epoch": 0.25,
"grad_norm": 2.28125,
"learning_rate": 1.9253852423732803e-05,
"loss": 1.8664,
"step": 3100
},
{
"epoch": 0.25,
"grad_norm": 2.25,
"learning_rate": 1.9244317079972007e-05,
"loss": 1.8706,
"step": 3120
},
{
"epoch": 0.25,
"grad_norm": 2.3125,
"learning_rate": 1.92347235860964e-05,
"loss": 1.8791,
"step": 3140
},
{
"epoch": 0.25,
"grad_norm": 2.34375,
"learning_rate": 1.9225072002452557e-05,
"loss": 1.8834,
"step": 3160
},
{
"epoch": 0.25,
"grad_norm": 2.359375,
"learning_rate": 1.9215362389752434e-05,
"loss": 1.8849,
"step": 3180
},
{
"epoch": 0.26,
"grad_norm": 2.3125,
"learning_rate": 1.9205594809073035e-05,
"loss": 1.8804,
"step": 3200
},
{
"epoch": 0.26,
"grad_norm": 2.3125,
"learning_rate": 1.9195769321855984e-05,
"loss": 1.8717,
"step": 3220
},
{
"epoch": 0.26,
"grad_norm": 2.1875,
"learning_rate": 1.9185885989907173e-05,
"loss": 1.8701,
"step": 3240
},
{
"epoch": 0.26,
"grad_norm": 2.0625,
"learning_rate": 1.917594487539635e-05,
"loss": 1.8764,
"step": 3260
},
{
"epoch": 0.26,
"grad_norm": 2.234375,
"learning_rate": 1.9165946040856747e-05,
"loss": 1.8695,
"step": 3280
},
{
"epoch": 0.26,
"grad_norm": 2.59375,
"learning_rate": 1.9155889549184657e-05,
"loss": 1.8747,
"step": 3300
},
{
"epoch": 0.27,
"grad_norm": 3.3125,
"learning_rate": 1.9145775463639073e-05,
"loss": 1.858,
"step": 3320
},
{
"epoch": 0.27,
"grad_norm": 2.4375,
"learning_rate": 1.9135603847841266e-05,
"loss": 1.8668,
"step": 3340
},
{
"epoch": 0.27,
"grad_norm": 2.28125,
"learning_rate": 1.9125374765774404e-05,
"loss": 1.8479,
"step": 3360
},
{
"epoch": 0.27,
"grad_norm": 2.1875,
"learning_rate": 1.911508828178312e-05,
"loss": 1.8627,
"step": 3380
},
{
"epoch": 0.27,
"grad_norm": 2.40625,
"learning_rate": 1.9104744460573156e-05,
"loss": 1.8924,
"step": 3400
},
{
"epoch": 0.27,
"grad_norm": 2.15625,
"learning_rate": 1.909434336721089e-05,
"loss": 1.8739,
"step": 3420
},
{
"epoch": 0.27,
"grad_norm": 2.1875,
"learning_rate": 1.9083885067122985e-05,
"loss": 1.8762,
"step": 3440
},
{
"epoch": 0.28,
"grad_norm": 2.5,
"learning_rate": 1.9073369626095958e-05,
"loss": 1.8711,
"step": 3460
},
{
"epoch": 0.28,
"grad_norm": 2.046875,
"learning_rate": 1.9062797110275743e-05,
"loss": 1.8768,
"step": 3480
},
{
"epoch": 0.28,
"grad_norm": 2.65625,
"learning_rate": 1.9052167586167315e-05,
"loss": 1.8683,
"step": 3500
},
{
"epoch": 0.28,
"grad_norm": 2.234375,
"learning_rate": 1.9041481120634248e-05,
"loss": 1.8697,
"step": 3520
},
{
"epoch": 0.28,
"grad_norm": 2.375,
"learning_rate": 1.9030737780898284e-05,
"loss": 1.863,
"step": 3540
},
{
"epoch": 0.28,
"grad_norm": 2.25,
"learning_rate": 1.9019937634538946e-05,
"loss": 1.8664,
"step": 3560
},
{
"epoch": 0.29,
"grad_norm": 2.203125,
"learning_rate": 1.900908074949307e-05,
"loss": 1.8684,
"step": 3580
},
{
"epoch": 0.29,
"grad_norm": 2.1875,
"learning_rate": 1.8998167194054425e-05,
"loss": 1.8525,
"step": 3600
},
{
"epoch": 0.29,
"grad_norm": 2.5625,
"learning_rate": 1.8987197036873227e-05,
"loss": 1.8582,
"step": 3620
},
{
"epoch": 0.29,
"grad_norm": 2.796875,
"learning_rate": 1.897617034695576e-05,
"loss": 1.8664,
"step": 3640
},
{
"epoch": 0.29,
"grad_norm": 2.1875,
"learning_rate": 1.8965087193663906e-05,
"loss": 1.8692,
"step": 3660
},
{
"epoch": 0.29,
"grad_norm": 2.21875,
"learning_rate": 1.895394764671473e-05,
"loss": 1.8534,
"step": 3680
},
{
"epoch": 0.3,
"grad_norm": 2.359375,
"learning_rate": 1.894275177618004e-05,
"loss": 1.852,
"step": 3700
},
{
"epoch": 0.3,
"grad_norm": 2.4375,
"learning_rate": 1.893149965248592e-05,
"loss": 1.8699,
"step": 3720
},
{
"epoch": 0.3,
"grad_norm": 2.203125,
"learning_rate": 1.8920191346412326e-05,
"loss": 1.8649,
"step": 3740
},
{
"epoch": 0.3,
"grad_norm": 2.15625,
"learning_rate": 1.8908826929092607e-05,
"loss": 1.857,
"step": 3760
},
{
"epoch": 0.3,
"grad_norm": 2.28125,
"learning_rate": 1.8897406472013084e-05,
"loss": 1.8404,
"step": 3780
},
{
"epoch": 0.3,
"grad_norm": 2.65625,
"learning_rate": 1.8885930047012585e-05,
"loss": 1.864,
"step": 3800
},
{
"epoch": 0.3,
"grad_norm": 2.640625,
"learning_rate": 1.887439772628199e-05,
"loss": 1.8578,
"step": 3820
},
{
"epoch": 0.31,
"grad_norm": 1.9765625,
"learning_rate": 1.886280958236379e-05,
"loss": 1.8603,
"step": 3840
},
{
"epoch": 0.31,
"grad_norm": 2.484375,
"learning_rate": 1.8851165688151627e-05,
"loss": 1.8603,
"step": 3860
},
{
"epoch": 0.31,
"grad_norm": 2.65625,
"learning_rate": 1.8839466116889823e-05,
"loss": 1.8752,
"step": 3880
},
{
"epoch": 0.31,
"grad_norm": 2.140625,
"learning_rate": 1.882771094217293e-05,
"loss": 1.8628,
"step": 3900
},
{
"epoch": 0.31,
"grad_norm": 2.28125,
"learning_rate": 1.8815900237945284e-05,
"loss": 1.8575,
"step": 3920
},
{
"epoch": 0.31,
"grad_norm": 2.28125,
"learning_rate": 1.8804034078500497e-05,
"loss": 1.85,
"step": 3940
},
{
"epoch": 0.32,
"grad_norm": 2.140625,
"learning_rate": 1.8792112538481025e-05,
"loss": 1.8687,
"step": 3960
},
{
"epoch": 0.32,
"grad_norm": 2.21875,
"learning_rate": 1.8780135692877693e-05,
"loss": 1.8465,
"step": 3980
},
{
"epoch": 0.32,
"grad_norm": 2.3125,
"learning_rate": 1.8768103617029213e-05,
"loss": 1.8569,
"step": 4000
},
{
"epoch": 0.32,
"grad_norm": 2.328125,
"learning_rate": 1.8756016386621712e-05,
"loss": 1.8401,
"step": 4020
},
{
"epoch": 0.32,
"grad_norm": 2.21875,
"learning_rate": 1.874387407768827e-05,
"loss": 1.8356,
"step": 4040
},
{
"epoch": 0.32,
"grad_norm": 2.21875,
"learning_rate": 1.873167676660842e-05,
"loss": 1.8605,
"step": 4060
},
{
"epoch": 0.33,
"grad_norm": 2.125,
"learning_rate": 1.8719424530107674e-05,
"loss": 1.8598,
"step": 4080
},
{
"epoch": 0.33,
"grad_norm": 2.890625,
"learning_rate": 1.8707117445257067e-05,
"loss": 1.8512,
"step": 4100
},
{
"epoch": 0.33,
"grad_norm": 2.59375,
"learning_rate": 1.8694755589472633e-05,
"loss": 1.8482,
"step": 4120
},
{
"epoch": 0.33,
"grad_norm": 2.359375,
"learning_rate": 1.8682339040514933e-05,
"loss": 1.8479,
"step": 4140
},
{
"epoch": 0.33,
"grad_norm": 2.140625,
"learning_rate": 1.8669867876488578e-05,
"loss": 1.8397,
"step": 4160
},
{
"epoch": 0.33,
"grad_norm": 2.0625,
"learning_rate": 1.8657342175841722e-05,
"loss": 1.8579,
"step": 4180
},
{
"epoch": 0.34,
"grad_norm": 2.78125,
"learning_rate": 1.8644762017365576e-05,
"loss": 1.8508,
"step": 4200
},
{
"epoch": 0.34,
"grad_norm": 2.34375,
"learning_rate": 1.863212748019391e-05,
"loss": 1.8335,
"step": 4220
},
{
"epoch": 0.34,
"grad_norm": 3.03125,
"learning_rate": 1.861943864380255e-05,
"loss": 1.8415,
"step": 4240
},
{
"epoch": 0.34,
"grad_norm": 2.375,
"learning_rate": 1.86066955880089e-05,
"loss": 1.8543,
"step": 4260
},
{
"epoch": 0.34,
"grad_norm": 2.625,
"learning_rate": 1.85938983929714e-05,
"loss": 1.861,
"step": 4280
},
{
"epoch": 0.34,
"grad_norm": 2.4375,
"learning_rate": 1.858104713918907e-05,
"loss": 1.8387,
"step": 4300
},
{
"epoch": 0.34,
"grad_norm": 2.21875,
"learning_rate": 1.8568141907500964e-05,
"loss": 1.8561,
"step": 4320
},
{
"epoch": 0.35,
"grad_norm": 2.484375,
"learning_rate": 1.8555182779085678e-05,
"loss": 1.8442,
"step": 4340
},
{
"epoch": 0.35,
"grad_norm": 2.34375,
"learning_rate": 1.8542169835460846e-05,
"loss": 1.8582,
"step": 4360
},
{
"epoch": 0.35,
"grad_norm": 2.265625,
"learning_rate": 1.8529103158482605e-05,
"loss": 1.8319,
"step": 4380
},
{
"epoch": 0.35,
"grad_norm": 2.875,
"learning_rate": 1.8515982830345115e-05,
"loss": 1.8388,
"step": 4400
},
{
"epoch": 0.35,
"grad_norm": 2.28125,
"learning_rate": 1.850280893358e-05,
"loss": 1.8552,
"step": 4420
},
{
"epoch": 0.35,
"grad_norm": 2.296875,
"learning_rate": 1.848958155105586e-05,
"loss": 1.8317,
"step": 4440
},
{
"epoch": 0.36,
"grad_norm": 2.203125,
"learning_rate": 1.847630076597774e-05,
"loss": 1.8413,
"step": 4460
},
{
"epoch": 0.36,
"grad_norm": 2.5,
"learning_rate": 1.846296666188661e-05,
"loss": 1.8251,
"step": 4480
},
{
"epoch": 0.36,
"grad_norm": 2.15625,
"learning_rate": 1.8449579322658827e-05,
"loss": 1.8445,
"step": 4500
},
{
"epoch": 0.36,
"grad_norm": 2.40625,
"learning_rate": 1.8436138832505623e-05,
"loss": 1.8672,
"step": 4520
},
{
"epoch": 0.36,
"grad_norm": 2.109375,
"learning_rate": 1.842264527597257e-05,
"loss": 1.8343,
"step": 4540
},
{
"epoch": 0.36,
"grad_norm": 2.390625,
"learning_rate": 1.8409098737939038e-05,
"loss": 1.8272,
"step": 4560
},
{
"epoch": 0.37,
"grad_norm": 2.296875,
"learning_rate": 1.8395499303617677e-05,
"loss": 1.8448,
"step": 4580
},
{
"epoch": 0.37,
"grad_norm": 2.421875,
"learning_rate": 1.8381847058553872e-05,
"loss": 1.835,
"step": 4600
},
{
"epoch": 0.37,
"grad_norm": 2.171875,
"learning_rate": 1.8368142088625213e-05,
"loss": 1.8356,
"step": 4620
},
{
"epoch": 0.37,
"grad_norm": 2.40625,
"learning_rate": 1.8354384480040935e-05,
"loss": 1.8175,
"step": 4640
},
{
"epoch": 0.37,
"grad_norm": 2.703125,
"learning_rate": 1.83405743193414e-05,
"loss": 1.8218,
"step": 4660
},
{
"epoch": 0.37,
"grad_norm": 2.265625,
"learning_rate": 1.8326711693397537e-05,
"loss": 1.8409,
"step": 4680
},
{
"epoch": 0.38,
"grad_norm": 2.25,
"learning_rate": 1.831279668941031e-05,
"loss": 1.8471,
"step": 4700
},
{
"epoch": 0.38,
"grad_norm": 2.109375,
"learning_rate": 1.8298829394910146e-05,
"loss": 1.8708,
"step": 4720
},
{
"epoch": 0.38,
"grad_norm": 2.28125,
"learning_rate": 1.82848098977564e-05,
"loss": 1.8397,
"step": 4740
},
{
"epoch": 0.38,
"grad_norm": 2.25,
"learning_rate": 1.8270738286136815e-05,
"loss": 1.8166,
"step": 4760
},
{
"epoch": 0.38,
"grad_norm": 2.1875,
"learning_rate": 1.8256614648566937e-05,
"loss": 1.8257,
"step": 4780
},
{
"epoch": 0.38,
"grad_norm": 2.40625,
"learning_rate": 1.824243907388958e-05,
"loss": 1.8483,
"step": 4800
},
{
"epoch": 0.38,
"grad_norm": 2.59375,
"learning_rate": 1.8228211651274264e-05,
"loss": 1.8235,
"step": 4820
},
{
"epoch": 0.39,
"grad_norm": 2.640625,
"learning_rate": 1.8213932470216652e-05,
"loss": 1.8561,
"step": 4840
},
{
"epoch": 0.39,
"grad_norm": 2.828125,
"learning_rate": 1.8199601620537977e-05,
"loss": 1.8324,
"step": 4860
},
{
"epoch": 0.39,
"grad_norm": 2.171875,
"learning_rate": 1.81852191923845e-05,
"loss": 1.8389,
"step": 4880
},
{
"epoch": 0.39,
"grad_norm": 2.171875,
"learning_rate": 1.8170785276226915e-05,
"loss": 1.8372,
"step": 4900
},
{
"epoch": 0.39,
"grad_norm": 2.484375,
"learning_rate": 1.8156299962859805e-05,
"loss": 1.8367,
"step": 4920
},
{
"epoch": 0.39,
"grad_norm": 2.21875,
"learning_rate": 1.8141763343401057e-05,
"loss": 1.8078,
"step": 4940
},
{
"epoch": 0.4,
"grad_norm": 2.21875,
"learning_rate": 1.8127175509291292e-05,
"loss": 1.8181,
"step": 4960
},
{
"epoch": 0.4,
"grad_norm": 2.359375,
"learning_rate": 1.8112536552293286e-05,
"loss": 1.8273,
"step": 4980
},
{
"epoch": 0.4,
"grad_norm": 3.140625,
"learning_rate": 1.80978465644914e-05,
"loss": 1.8302,
"step": 5000
},
{
"epoch": 0.4,
"grad_norm": 2.125,
"learning_rate": 1.8083105638291e-05,
"loss": 1.8469,
"step": 5020
},
{
"epoch": 0.4,
"grad_norm": 2.46875,
"learning_rate": 1.8068313866417876e-05,
"loss": 1.8235,
"step": 5040
},
{
"epoch": 0.4,
"grad_norm": 2.328125,
"learning_rate": 1.8053471341917636e-05,
"loss": 1.8302,
"step": 5060
},
{
"epoch": 0.41,
"grad_norm": 2.625,
"learning_rate": 1.8038578158155163e-05,
"loss": 1.8218,
"step": 5080
},
{
"epoch": 0.41,
"grad_norm": 2.234375,
"learning_rate": 1.8023634408814e-05,
"loss": 1.8322,
"step": 5100
},
{
"epoch": 0.41,
"grad_norm": 2.4375,
"learning_rate": 1.8008640187895755e-05,
"loss": 1.8091,
"step": 5120
},
{
"epoch": 0.41,
"grad_norm": 2.53125,
"learning_rate": 1.7993595589719533e-05,
"loss": 1.828,
"step": 5140
},
{
"epoch": 0.41,
"grad_norm": 2.40625,
"learning_rate": 1.797850070892132e-05,
"loss": 1.8188,
"step": 5160
},
{
"epoch": 0.41,
"grad_norm": 2.84375,
"learning_rate": 1.7963355640453407e-05,
"loss": 1.8106,
"step": 5180
},
{
"epoch": 0.42,
"grad_norm": 2.15625,
"learning_rate": 1.7948160479583783e-05,
"loss": 1.8172,
"step": 5200
},
{
"epoch": 0.42,
"grad_norm": 2.375,
"learning_rate": 1.793291532189553e-05,
"loss": 1.8324,
"step": 5220
},
{
"epoch": 0.42,
"grad_norm": 2.359375,
"learning_rate": 1.791762026328623e-05,
"loss": 1.8202,
"step": 5240
},
{
"epoch": 0.42,
"grad_norm": 2.171875,
"learning_rate": 1.7902275399967363e-05,
"loss": 1.8183,
"step": 5260
},
{
"epoch": 0.42,
"grad_norm": 2.5625,
"learning_rate": 1.78868808284637e-05,
"loss": 1.8347,
"step": 5280
},
{
"epoch": 0.42,
"grad_norm": 2.5625,
"learning_rate": 1.7871436645612685e-05,
"loss": 1.831,
"step": 5300
},
{
"epoch": 0.42,
"grad_norm": 2.796875,
"learning_rate": 1.785594294856385e-05,
"loss": 1.8263,
"step": 5320
},
{
"epoch": 0.43,
"grad_norm": 2.265625,
"learning_rate": 1.7840399834778176e-05,
"loss": 1.847,
"step": 5340
},
{
"epoch": 0.43,
"grad_norm": 2.375,
"learning_rate": 1.7824807402027504e-05,
"loss": 1.8249,
"step": 5360
},
{
"epoch": 0.43,
"grad_norm": 2.34375,
"learning_rate": 1.78091657483939e-05,
"loss": 1.8206,
"step": 5380
},
{
"epoch": 0.43,
"grad_norm": 2.328125,
"learning_rate": 1.779347497226905e-05,
"loss": 1.8251,
"step": 5400
},
{
"epoch": 0.43,
"grad_norm": 2.5,
"learning_rate": 1.777773517235364e-05,
"loss": 1.8226,
"step": 5420
},
{
"epoch": 0.43,
"grad_norm": 2.1875,
"learning_rate": 1.7761946447656736e-05,
"loss": 1.8309,
"step": 5440
},
{
"epoch": 0.44,
"grad_norm": 2.234375,
"learning_rate": 1.7746108897495157e-05,
"loss": 1.8283,
"step": 5460
},
{
"epoch": 0.44,
"grad_norm": 2.796875,
"learning_rate": 1.7730222621492846e-05,
"loss": 1.8275,
"step": 5480
},
{
"epoch": 0.44,
"grad_norm": 2.390625,
"learning_rate": 1.7714287719580254e-05,
"loss": 1.8059,
"step": 5500
},
{
"epoch": 0.44,
"grad_norm": 2.28125,
"learning_rate": 1.769830429199371e-05,
"loss": 1.8235,
"step": 5520
},
{
"epoch": 0.44,
"grad_norm": 2.40625,
"learning_rate": 1.7682272439274778e-05,
"loss": 1.8104,
"step": 5540
},
{
"epoch": 0.44,
"grad_norm": 2.609375,
"learning_rate": 1.766619226226965e-05,
"loss": 1.8212,
"step": 5560
},
{
"epoch": 0.45,
"grad_norm": 2.140625,
"learning_rate": 1.765006386212847e-05,
"loss": 1.8269,
"step": 5580
},
{
"epoch": 0.45,
"grad_norm": 2.21875,
"learning_rate": 1.763388734030475e-05,
"loss": 1.8212,
"step": 5600
},
{
"epoch": 0.45,
"grad_norm": 2.390625,
"learning_rate": 1.7617662798554685e-05,
"loss": 1.8447,
"step": 5620
},
{
"epoch": 0.45,
"grad_norm": 2.53125,
"learning_rate": 1.7601390338936547e-05,
"loss": 1.8244,
"step": 5640
},
{
"epoch": 0.45,
"grad_norm": 2.375,
"learning_rate": 1.7585070063810014e-05,
"loss": 1.8125,
"step": 5660
},
{
"epoch": 0.45,
"grad_norm": 2.53125,
"learning_rate": 1.7568702075835557e-05,
"loss": 1.8114,
"step": 5680
},
{
"epoch": 0.46,
"grad_norm": 2.34375,
"learning_rate": 1.7552286477973766e-05,
"loss": 1.8136,
"step": 5700
},
{
"epoch": 0.46,
"grad_norm": 2.359375,
"learning_rate": 1.7535823373484716e-05,
"loss": 1.8261,
"step": 5720
},
{
"epoch": 0.46,
"grad_norm": 2.3125,
"learning_rate": 1.751931286592732e-05,
"loss": 1.8085,
"step": 5740
},
{
"epoch": 0.46,
"grad_norm": 2.34375,
"learning_rate": 1.7502755059158683e-05,
"loss": 1.8297,
"step": 5760
},
{
"epoch": 0.46,
"grad_norm": 2.296875,
"learning_rate": 1.7486150057333416e-05,
"loss": 1.7937,
"step": 5780
},
{
"epoch": 0.46,
"grad_norm": 2.15625,
"learning_rate": 1.7469497964903018e-05,
"loss": 1.8052,
"step": 5800
},
{
"epoch": 0.46,
"grad_norm": 2.328125,
"learning_rate": 1.7452798886615205e-05,
"loss": 1.8216,
"step": 5820
},
{
"epoch": 0.47,
"grad_norm": 2.171875,
"learning_rate": 1.7436052927513254e-05,
"loss": 1.8322,
"step": 5840
},
{
"epoch": 0.47,
"grad_norm": 2.484375,
"learning_rate": 1.741926019293533e-05,
"loss": 1.8182,
"step": 5860
},
{
"epoch": 0.47,
"grad_norm": 2.578125,
"learning_rate": 1.740242078851384e-05,
"loss": 1.8262,
"step": 5880
},
{
"epoch": 0.47,
"grad_norm": 2.734375,
"learning_rate": 1.7385534820174757e-05,
"loss": 1.7948,
"step": 5900
},
{
"epoch": 0.47,
"grad_norm": 3.0,
"learning_rate": 1.7368602394136964e-05,
"loss": 1.8332,
"step": 5920
},
{
"epoch": 0.47,
"grad_norm": 2.234375,
"learning_rate": 1.735162361691157e-05,
"loss": 1.8016,
"step": 5940
},
{
"epoch": 0.48,
"grad_norm": 2.25,
"learning_rate": 1.7334598595301257e-05,
"loss": 1.8103,
"step": 5960
},
{
"epoch": 0.48,
"grad_norm": 2.953125,
"learning_rate": 1.7317527436399603e-05,
"loss": 1.8014,
"step": 5980
},
{
"epoch": 0.48,
"grad_norm": 2.25,
"learning_rate": 1.7300410247590402e-05,
"loss": 1.8071,
"step": 6000
},
{
"epoch": 0.48,
"grad_norm": 2.375,
"learning_rate": 1.7283247136546996e-05,
"loss": 1.809,
"step": 6020
},
{
"epoch": 0.48,
"grad_norm": 2.46875,
"learning_rate": 1.7266038211231583e-05,
"loss": 1.8236,
"step": 6040
},
{
"epoch": 0.48,
"grad_norm": 2.234375,
"learning_rate": 1.724878357989457e-05,
"loss": 1.8306,
"step": 6060
},
{
"epoch": 0.49,
"grad_norm": 2.265625,
"learning_rate": 1.7231483351073858e-05,
"loss": 1.8165,
"step": 6080
},
{
"epoch": 0.49,
"grad_norm": 2.1875,
"learning_rate": 1.721413763359417e-05,
"loss": 1.8162,
"step": 6100
},
{
"epoch": 0.49,
"grad_norm": 2.25,
"learning_rate": 1.7196746536566376e-05,
"loss": 1.8346,
"step": 6120
},
{
"epoch": 0.49,
"grad_norm": 2.265625,
"learning_rate": 1.71793101693868e-05,
"loss": 1.8082,
"step": 6140
},
{
"epoch": 0.49,
"grad_norm": 2.15625,
"learning_rate": 1.7161828641736527e-05,
"loss": 1.8105,
"step": 6160
},
{
"epoch": 0.49,
"grad_norm": 2.4375,
"learning_rate": 1.7144302063580726e-05,
"loss": 1.8105,
"step": 6180
},
{
"epoch": 0.49,
"grad_norm": 2.46875,
"learning_rate": 1.712673054516794e-05,
"loss": 1.8232,
"step": 6200
},
{
"epoch": 0.5,
"grad_norm": 2.75,
"learning_rate": 1.7109114197029408e-05,
"loss": 1.8227,
"step": 6220
},
{
"epoch": 0.5,
"grad_norm": 2.265625,
"learning_rate": 1.7091453129978363e-05,
"loss": 1.8181,
"step": 6240
},
{
"epoch": 0.5,
"grad_norm": 2.59375,
"learning_rate": 1.7073747455109336e-05,
"loss": 1.8006,
"step": 6260
},
{
"epoch": 0.5,
"grad_norm": 2.71875,
"learning_rate": 1.7055997283797463e-05,
"loss": 1.7975,
"step": 6280
},
{
"epoch": 0.5,
"grad_norm": 2.28125,
"learning_rate": 1.7038202727697766e-05,
"loss": 1.8105,
"step": 6300
},
{
"epoch": 0.5,
"grad_norm": 2.59375,
"learning_rate": 1.7020363898744477e-05,
"loss": 1.7994,
"step": 6320
},
{
"epoch": 0.51,
"grad_norm": 2.59375,
"learning_rate": 1.7002480909150316e-05,
"loss": 1.8193,
"step": 6340
},
{
"epoch": 0.51,
"grad_norm": 2.40625,
"learning_rate": 1.6984553871405783e-05,
"loss": 1.8347,
"step": 6360
},
{
"epoch": 0.51,
"grad_norm": 2.234375,
"learning_rate": 1.6966582898278466e-05,
"loss": 1.8159,
"step": 6380
},
{
"epoch": 0.51,
"grad_norm": 2.3125,
"learning_rate": 1.694856810281232e-05,
"loss": 1.8053,
"step": 6400
},
{
"epoch": 0.51,
"grad_norm": 2.34375,
"learning_rate": 1.6930509598326948e-05,
"loss": 1.828,
"step": 6420
},
{
"epoch": 0.51,
"grad_norm": 2.1875,
"learning_rate": 1.6912407498416914e-05,
"loss": 1.8186,
"step": 6440
},
{
"epoch": 0.52,
"grad_norm": 2.296875,
"learning_rate": 1.689426191695101e-05,
"loss": 1.8027,
"step": 6460
},
{
"epoch": 0.52,
"grad_norm": 2.375,
"learning_rate": 1.6876072968071532e-05,
"loss": 1.8098,
"step": 6480
},
{
"epoch": 0.52,
"grad_norm": 2.46875,
"learning_rate": 1.6857840766193586e-05,
"loss": 1.8129,
"step": 6500
},
{
"epoch": 0.52,
"grad_norm": 2.328125,
"learning_rate": 1.6839565426004346e-05,
"loss": 1.8054,
"step": 6520
},
{
"epoch": 0.52,
"grad_norm": 2.390625,
"learning_rate": 1.6821247062462347e-05,
"loss": 1.8123,
"step": 6540
},
{
"epoch": 0.52,
"grad_norm": 2.328125,
"learning_rate": 1.6802885790796753e-05,
"loss": 1.8074,
"step": 6560
},
{
"epoch": 0.53,
"grad_norm": 2.65625,
"learning_rate": 1.678448172650664e-05,
"loss": 1.7996,
"step": 6580
},
{
"epoch": 0.53,
"grad_norm": 2.609375,
"learning_rate": 1.676603498536026e-05,
"loss": 1.8098,
"step": 6600
},
{
"epoch": 0.53,
"grad_norm": 2.265625,
"learning_rate": 1.6747545683394322e-05,
"loss": 1.8016,
"step": 6620
},
{
"epoch": 0.53,
"grad_norm": 2.6875,
"learning_rate": 1.672901393691325e-05,
"loss": 1.8093,
"step": 6640
},
{
"epoch": 0.53,
"grad_norm": 2.484375,
"learning_rate": 1.6710439862488478e-05,
"loss": 1.8023,
"step": 6660
},
{
"epoch": 0.53,
"grad_norm": 2.234375,
"learning_rate": 1.6691823576957676e-05,
"loss": 1.8075,
"step": 6680
},
{
"epoch": 0.53,
"grad_norm": 2.6875,
"learning_rate": 1.667316519742405e-05,
"loss": 1.8052,
"step": 6700
},
{
"epoch": 0.54,
"grad_norm": 2.34375,
"learning_rate": 1.6654464841255586e-05,
"loss": 1.8011,
"step": 6720
},
{
"epoch": 0.54,
"grad_norm": 2.25,
"learning_rate": 1.663572262608433e-05,
"loss": 1.8075,
"step": 6740
},
{
"epoch": 0.54,
"grad_norm": 2.296875,
"learning_rate": 1.6616938669805622e-05,
"loss": 1.7911,
"step": 6760
},
{
"epoch": 0.54,
"grad_norm": 2.203125,
"learning_rate": 1.659811309057738e-05,
"loss": 1.8026,
"step": 6780
},
{
"epoch": 0.54,
"grad_norm": 2.1875,
"learning_rate": 1.6579246006819335e-05,
"loss": 1.8088,
"step": 6800
},
{
"epoch": 0.54,
"grad_norm": 2.328125,
"learning_rate": 1.6560337537212306e-05,
"loss": 1.8155,
"step": 6820
},
{
"epoch": 0.55,
"grad_norm": 2.21875,
"learning_rate": 1.6541387800697438e-05,
"loss": 1.7997,
"step": 6840
},
{
"epoch": 0.55,
"grad_norm": 2.25,
"learning_rate": 1.6522396916475468e-05,
"loss": 1.8253,
"step": 6860
},
{
"epoch": 0.55,
"grad_norm": 2.265625,
"learning_rate": 1.650336500400595e-05,
"loss": 1.8037,
"step": 6880
},
{
"epoch": 0.55,
"grad_norm": 2.421875,
"learning_rate": 1.6484292183006542e-05,
"loss": 1.8154,
"step": 6900
},
{
"epoch": 0.55,
"grad_norm": 2.515625,
"learning_rate": 1.6465178573452214e-05,
"loss": 1.8169,
"step": 6920
},
{
"epoch": 0.55,
"grad_norm": 2.359375,
"learning_rate": 1.6446024295574522e-05,
"loss": 1.8002,
"step": 6940
},
{
"epoch": 0.56,
"grad_norm": 2.28125,
"learning_rate": 1.6426829469860837e-05,
"loss": 1.7999,
"step": 6960
},
{
"epoch": 0.56,
"grad_norm": 2.375,
"learning_rate": 1.6407594217053587e-05,
"loss": 1.7973,
"step": 6980
},
{
"epoch": 0.56,
"grad_norm": 2.296875,
"learning_rate": 1.638831865814951e-05,
"loss": 1.8073,
"step": 7000
},
{
"epoch": 0.56,
"grad_norm": 2.265625,
"learning_rate": 1.6369002914398874e-05,
"loss": 1.795,
"step": 7020
},
{
"epoch": 0.56,
"grad_norm": 2.34375,
"learning_rate": 1.6349647107304724e-05,
"loss": 1.7985,
"step": 7040
},
{
"epoch": 0.56,
"grad_norm": 2.125,
"learning_rate": 1.633025135862213e-05,
"loss": 1.7936,
"step": 7060
},
{
"epoch": 0.57,
"grad_norm": 2.890625,
"learning_rate": 1.6310815790357404e-05,
"loss": 1.8036,
"step": 7080
},
{
"epoch": 0.57,
"grad_norm": 2.4375,
"learning_rate": 1.6291340524767327e-05,
"loss": 1.8046,
"step": 7100
},
{
"epoch": 0.57,
"grad_norm": 2.6875,
"learning_rate": 1.6271825684358404e-05,
"loss": 1.8052,
"step": 7120
},
{
"epoch": 0.57,
"grad_norm": 2.484375,
"learning_rate": 1.625227139188607e-05,
"loss": 1.8105,
"step": 7140
},
{
"epoch": 0.57,
"grad_norm": 2.375,
"learning_rate": 1.6232677770353936e-05,
"loss": 1.7952,
"step": 7160
},
{
"epoch": 0.57,
"grad_norm": 2.625,
"learning_rate": 1.621304494301301e-05,
"loss": 1.8102,
"step": 7180
},
{
"epoch": 0.57,
"grad_norm": 2.609375,
"learning_rate": 1.6193373033360904e-05,
"loss": 1.7962,
"step": 7200
},
{
"epoch": 0.58,
"grad_norm": 2.328125,
"learning_rate": 1.6173662165141084e-05,
"loss": 1.8078,
"step": 7220
},
{
"epoch": 0.58,
"grad_norm": 2.5,
"learning_rate": 1.6153912462342073e-05,
"loss": 1.8051,
"step": 7240
},
{
"epoch": 0.58,
"grad_norm": 2.15625,
"learning_rate": 1.6134124049196688e-05,
"loss": 1.8057,
"step": 7260
},
{
"epoch": 0.58,
"grad_norm": 2.78125,
"learning_rate": 1.6114297050181235e-05,
"loss": 1.8153,
"step": 7280
},
{
"epoch": 0.58,
"grad_norm": 2.171875,
"learning_rate": 1.6094431590014746e-05,
"loss": 1.8047,
"step": 7300
},
{
"epoch": 0.58,
"grad_norm": 2.359375,
"learning_rate": 1.6074527793658186e-05,
"loss": 1.8069,
"step": 7320
},
{
"epoch": 0.59,
"grad_norm": 2.4375,
"learning_rate": 1.605458578631367e-05,
"loss": 1.7919,
"step": 7340
},
{
"epoch": 0.59,
"grad_norm": 2.390625,
"learning_rate": 1.6034605693423676e-05,
"loss": 1.8104,
"step": 7360
},
{
"epoch": 0.59,
"grad_norm": 2.703125,
"learning_rate": 1.6014587640670244e-05,
"loss": 1.7971,
"step": 7380
},
{
"epoch": 0.59,
"grad_norm": 2.703125,
"learning_rate": 1.599453175397421e-05,
"loss": 1.7987,
"step": 7400
},
{
"epoch": 0.59,
"grad_norm": 2.46875,
"learning_rate": 1.597443815949439e-05,
"loss": 1.8057,
"step": 7420
},
{
"epoch": 0.59,
"grad_norm": 2.359375,
"learning_rate": 1.59543069836268e-05,
"loss": 1.7817,
"step": 7440
},
{
"epoch": 0.6,
"grad_norm": 2.515625,
"learning_rate": 1.5934138353003845e-05,
"loss": 1.8009,
"step": 7460
},
{
"epoch": 0.6,
"grad_norm": 2.5,
"learning_rate": 1.5913932394493548e-05,
"loss": 1.7939,
"step": 7480
},
{
"epoch": 0.6,
"grad_norm": 2.40625,
"learning_rate": 1.589368923519874e-05,
"loss": 1.8014,
"step": 7500
},
{
"epoch": 0.6,
"grad_norm": 2.328125,
"learning_rate": 1.587340900245624e-05,
"loss": 1.7879,
"step": 7520
},
{
"epoch": 0.6,
"grad_norm": 2.25,
"learning_rate": 1.5853091823836087e-05,
"loss": 1.8,
"step": 7540
},
{
"epoch": 0.6,
"grad_norm": 2.140625,
"learning_rate": 1.5832737827140727e-05,
"loss": 1.7894,
"step": 7560
},
{
"epoch": 0.61,
"grad_norm": 2.6875,
"learning_rate": 1.581234714040419e-05,
"loss": 1.7845,
"step": 7580
},
{
"epoch": 0.61,
"grad_norm": 2.359375,
"learning_rate": 1.5791919891891313e-05,
"loss": 1.7841,
"step": 7600
},
{
"epoch": 0.61,
"grad_norm": 2.3125,
"learning_rate": 1.5771456210096913e-05,
"loss": 1.8057,
"step": 7620
},
{
"epoch": 0.61,
"grad_norm": 2.84375,
"learning_rate": 1.5750956223744985e-05,
"loss": 1.7961,
"step": 7640
},
{
"epoch": 0.61,
"grad_norm": 2.28125,
"learning_rate": 1.5730420061787898e-05,
"loss": 1.7908,
"step": 7660
},
{
"epoch": 0.61,
"grad_norm": 2.578125,
"learning_rate": 1.5709847853405574e-05,
"loss": 1.7888,
"step": 7680
},
{
"epoch": 0.61,
"grad_norm": 2.328125,
"learning_rate": 1.568923972800468e-05,
"loss": 1.7742,
"step": 7700
},
{
"epoch": 0.62,
"grad_norm": 2.484375,
"learning_rate": 1.566859581521782e-05,
"loss": 1.7902,
"step": 7720
},
{
"epoch": 0.62,
"grad_norm": 2.46875,
"learning_rate": 1.5647916244902707e-05,
"loss": 1.7918,
"step": 7740
},
{
"epoch": 0.62,
"grad_norm": 2.390625,
"learning_rate": 1.5627201147141357e-05,
"loss": 1.806,
"step": 7760
},
{
"epoch": 0.62,
"grad_norm": 2.25,
"learning_rate": 1.5606450652239263e-05,
"loss": 1.7925,
"step": 7780
},
{
"epoch": 0.62,
"grad_norm": 2.578125,
"learning_rate": 1.5585664890724584e-05,
"loss": 1.7921,
"step": 7800
},
{
"epoch": 0.62,
"grad_norm": 2.390625,
"learning_rate": 1.5564843993347313e-05,
"loss": 1.7901,
"step": 7820
},
{
"epoch": 0.63,
"grad_norm": 2.390625,
"learning_rate": 1.5543988091078467e-05,
"loss": 1.7881,
"step": 7840
},
{
"epoch": 0.63,
"grad_norm": 2.6875,
"learning_rate": 1.5523097315109245e-05,
"loss": 1.7948,
"step": 7860
},
{
"epoch": 0.63,
"grad_norm": 2.484375,
"learning_rate": 1.5502171796850226e-05,
"loss": 1.7958,
"step": 7880
},
{
"epoch": 0.63,
"grad_norm": 2.40625,
"learning_rate": 1.5481211667930528e-05,
"loss": 1.7911,
"step": 7900
},
{
"epoch": 0.63,
"grad_norm": 2.453125,
"learning_rate": 1.5460217060196986e-05,
"loss": 1.7709,
"step": 7920
},
{
"epoch": 0.63,
"grad_norm": 2.40625,
"learning_rate": 1.54391881057133e-05,
"loss": 1.7914,
"step": 7940
},
{
"epoch": 0.64,
"grad_norm": 2.328125,
"learning_rate": 1.541812493675925e-05,
"loss": 1.8062,
"step": 7960
},
{
"epoch": 0.64,
"grad_norm": 2.34375,
"learning_rate": 1.539702768582982e-05,
"loss": 1.8074,
"step": 7980
},
{
"epoch": 0.64,
"grad_norm": 2.75,
"learning_rate": 1.5375896485634386e-05,
"loss": 1.7788,
"step": 8000
},
{
"epoch": 0.64,
"grad_norm": 2.453125,
"learning_rate": 1.5354731469095884e-05,
"loss": 1.7814,
"step": 8020
},
{
"epoch": 0.64,
"grad_norm": 2.421875,
"learning_rate": 1.5333532769349955e-05,
"loss": 1.7854,
"step": 8040
},
{
"epoch": 0.64,
"grad_norm": 2.65625,
"learning_rate": 1.5312300519744135e-05,
"loss": 1.7869,
"step": 8060
},
{
"epoch": 0.65,
"grad_norm": 2.625,
"learning_rate": 1.529103485383699e-05,
"loss": 1.7736,
"step": 8080
},
{
"epoch": 0.65,
"grad_norm": 2.28125,
"learning_rate": 1.5269735905397278e-05,
"loss": 1.7966,
"step": 8100
},
{
"epoch": 0.65,
"grad_norm": 2.59375,
"learning_rate": 1.524840380840314e-05,
"loss": 1.7907,
"step": 8120
},
{
"epoch": 0.65,
"grad_norm": 2.671875,
"learning_rate": 1.5227038697041216e-05,
"loss": 1.7767,
"step": 8140
},
{
"epoch": 0.65,
"grad_norm": 2.234375,
"learning_rate": 1.520564070570582e-05,
"loss": 1.7963,
"step": 8160
},
{
"epoch": 0.65,
"grad_norm": 2.421875,
"learning_rate": 1.5184209968998098e-05,
"loss": 1.7822,
"step": 8180
},
{
"epoch": 0.65,
"grad_norm": 2.390625,
"learning_rate": 1.5162746621725176e-05,
"loss": 1.7806,
"step": 8200
},
{
"epoch": 0.66,
"grad_norm": 2.359375,
"learning_rate": 1.5141250798899307e-05,
"loss": 1.7836,
"step": 8220
},
{
"epoch": 0.66,
"grad_norm": 2.640625,
"learning_rate": 1.5119722635737035e-05,
"loss": 1.7825,
"step": 8240
},
{
"epoch": 0.66,
"grad_norm": 2.953125,
"learning_rate": 1.5098162267658323e-05,
"loss": 1.7877,
"step": 8260
},
{
"epoch": 0.66,
"grad_norm": 2.3125,
"learning_rate": 1.5076569830285736e-05,
"loss": 1.791,
"step": 8280
},
{
"epoch": 0.66,
"grad_norm": 2.4375,
"learning_rate": 1.5054945459443544e-05,
"loss": 1.781,
"step": 8300
},
{
"epoch": 0.66,
"grad_norm": 2.609375,
"learning_rate": 1.5033289291156905e-05,
"loss": 1.7873,
"step": 8320
},
{
"epoch": 0.67,
"grad_norm": 2.296875,
"learning_rate": 1.501160146165099e-05,
"loss": 1.7963,
"step": 8340
},
{
"epoch": 0.67,
"grad_norm": 2.515625,
"learning_rate": 1.498988210735013e-05,
"loss": 1.794,
"step": 8360
},
{
"epoch": 0.67,
"grad_norm": 2.265625,
"learning_rate": 1.4968131364876952e-05,
"loss": 1.8001,
"step": 8380
},
{
"epoch": 0.67,
"grad_norm": 2.46875,
"learning_rate": 1.4946349371051541e-05,
"loss": 1.7728,
"step": 8400
},
{
"epoch": 0.67,
"grad_norm": 2.59375,
"learning_rate": 1.4924536262890557e-05,
"loss": 1.7732,
"step": 8420
},
{
"epoch": 0.67,
"grad_norm": 2.671875,
"learning_rate": 1.4902692177606368e-05,
"loss": 1.7822,
"step": 8440
},
{
"epoch": 0.68,
"grad_norm": 2.609375,
"learning_rate": 1.4880817252606226e-05,
"loss": 1.7862,
"step": 8460
},
{
"epoch": 0.68,
"grad_norm": 2.421875,
"learning_rate": 1.4858911625491352e-05,
"loss": 1.801,
"step": 8480
},
{
"epoch": 0.68,
"grad_norm": 2.40625,
"learning_rate": 1.4836975434056102e-05,
"loss": 1.8229,
"step": 8500
},
{
"epoch": 0.68,
"grad_norm": 2.578125,
"learning_rate": 1.48150088162871e-05,
"loss": 1.7954,
"step": 8520
},
{
"epoch": 0.68,
"grad_norm": 2.46875,
"learning_rate": 1.4793011910362352e-05,
"loss": 1.7996,
"step": 8540
},
{
"epoch": 0.68,
"grad_norm": 2.40625,
"learning_rate": 1.4770984854650397e-05,
"loss": 1.8033,
"step": 8560
},
{
"epoch": 0.68,
"grad_norm": 2.140625,
"learning_rate": 1.4748927787709417e-05,
"loss": 1.7883,
"step": 8580
},
{
"epoch": 0.69,
"grad_norm": 2.265625,
"learning_rate": 1.4726840848286385e-05,
"loss": 1.7939,
"step": 8600
},
{
"epoch": 0.69,
"grad_norm": 2.421875,
"learning_rate": 1.4704724175316181e-05,
"loss": 1.7975,
"step": 8620
},
{
"epoch": 0.69,
"grad_norm": 2.46875,
"learning_rate": 1.4682577907920707e-05,
"loss": 1.8029,
"step": 8640
},
{
"epoch": 0.69,
"grad_norm": 2.359375,
"learning_rate": 1.4660402185408046e-05,
"loss": 1.7807,
"step": 8660
},
{
"epoch": 0.69,
"grad_norm": 2.5,
"learning_rate": 1.4638197147271548e-05,
"loss": 1.7953,
"step": 8680
},
{
"epoch": 0.69,
"grad_norm": 2.4375,
"learning_rate": 1.4615962933188981e-05,
"loss": 1.7902,
"step": 8700
},
{
"epoch": 0.7,
"grad_norm": 2.5625,
"learning_rate": 1.4593699683021625e-05,
"loss": 1.7849,
"step": 8720
},
{
"epoch": 0.7,
"grad_norm": 2.765625,
"learning_rate": 1.4571407536813422e-05,
"loss": 1.7814,
"step": 8740
},
{
"epoch": 0.7,
"grad_norm": 2.375,
"learning_rate": 1.4549086634790075e-05,
"loss": 1.7932,
"step": 8760
},
{
"epoch": 0.7,
"grad_norm": 2.4375,
"learning_rate": 1.4526737117358167e-05,
"loss": 1.789,
"step": 8780
},
{
"epoch": 0.7,
"grad_norm": 2.546875,
"learning_rate": 1.4504359125104292e-05,
"loss": 1.7828,
"step": 8800
},
{
"epoch": 0.7,
"grad_norm": 2.34375,
"learning_rate": 1.4481952798794152e-05,
"loss": 1.7876,
"step": 8820
},
{
"epoch": 0.71,
"grad_norm": 2.328125,
"learning_rate": 1.4459518279371692e-05,
"loss": 1.794,
"step": 8840
},
{
"epoch": 0.71,
"grad_norm": 2.734375,
"learning_rate": 1.4437055707958184e-05,
"loss": 1.7919,
"step": 8860
},
{
"epoch": 0.71,
"grad_norm": 2.421875,
"learning_rate": 1.4414565225851371e-05,
"loss": 1.7846,
"step": 8880
},
{
"epoch": 0.71,
"grad_norm": 2.453125,
"learning_rate": 1.4392046974524565e-05,
"loss": 1.7843,
"step": 8900
},
{
"epoch": 0.71,
"grad_norm": 2.40625,
"learning_rate": 1.4369501095625747e-05,
"loss": 1.7726,
"step": 8920
},
{
"epoch": 0.71,
"grad_norm": 2.390625,
"learning_rate": 1.4346927730976691e-05,
"loss": 1.7836,
"step": 8940
},
{
"epoch": 0.72,
"grad_norm": 2.8125,
"learning_rate": 1.4324327022572073e-05,
"loss": 1.776,
"step": 8960
},
{
"epoch": 0.72,
"grad_norm": 2.296875,
"learning_rate": 1.4301699112578557e-05,
"loss": 1.7903,
"step": 8980
},
{
"epoch": 0.72,
"grad_norm": 2.5,
"learning_rate": 1.4279044143333926e-05,
"loss": 1.7757,
"step": 9000
},
{
"epoch": 0.72,
"grad_norm": 2.3125,
"learning_rate": 1.425636225734617e-05,
"loss": 1.7705,
"step": 9020
},
{
"epoch": 0.72,
"grad_norm": 2.5,
"learning_rate": 1.42336535972926e-05,
"loss": 1.8011,
"step": 9040
},
{
"epoch": 0.72,
"grad_norm": 2.375,
"learning_rate": 1.4210918306018937e-05,
"loss": 1.7795,
"step": 9060
},
{
"epoch": 0.72,
"grad_norm": 2.78125,
"learning_rate": 1.4188156526538435e-05,
"loss": 1.7965,
"step": 9080
},
{
"epoch": 0.73,
"grad_norm": 2.796875,
"learning_rate": 1.4165368402030952e-05,
"loss": 1.7631,
"step": 9100
},
{
"epoch": 0.73,
"grad_norm": 2.453125,
"learning_rate": 1.4142554075842083e-05,
"loss": 1.7949,
"step": 9120
},
{
"epoch": 0.73,
"grad_norm": 2.625,
"learning_rate": 1.4119713691482228e-05,
"loss": 1.785,
"step": 9140
},
{
"epoch": 0.73,
"grad_norm": 2.46875,
"learning_rate": 1.4096847392625708e-05,
"loss": 1.777,
"step": 9160
},
{
"epoch": 0.73,
"grad_norm": 2.546875,
"learning_rate": 1.4073955323109859e-05,
"loss": 1.779,
"step": 9180
},
{
"epoch": 0.73,
"grad_norm": 2.5,
"learning_rate": 1.4051037626934112e-05,
"loss": 1.7815,
"step": 9200
},
{
"epoch": 0.74,
"grad_norm": 2.828125,
"learning_rate": 1.4028094448259113e-05,
"loss": 1.7852,
"step": 9220
},
{
"epoch": 0.74,
"grad_norm": 2.375,
"learning_rate": 1.4005125931405792e-05,
"loss": 1.7999,
"step": 9240
},
{
"epoch": 0.74,
"grad_norm": 2.390625,
"learning_rate": 1.3982132220854472e-05,
"loss": 1.791,
"step": 9260
},
{
"epoch": 0.74,
"grad_norm": 2.484375,
"learning_rate": 1.3959113461243952e-05,
"loss": 1.7836,
"step": 9280
},
{
"epoch": 0.74,
"grad_norm": 2.6875,
"learning_rate": 1.3936069797370591e-05,
"loss": 1.778,
"step": 9300
},
{
"epoch": 0.74,
"grad_norm": 2.234375,
"learning_rate": 1.3913001374187421e-05,
"loss": 1.8065,
"step": 9320
},
{
"epoch": 0.75,
"grad_norm": 2.671875,
"learning_rate": 1.3889908336803198e-05,
"loss": 1.8035,
"step": 9340
},
{
"epoch": 0.75,
"grad_norm": 2.4375,
"learning_rate": 1.3866790830481529e-05,
"loss": 1.7789,
"step": 9360
},
{
"epoch": 0.75,
"grad_norm": 2.65625,
"learning_rate": 1.3843649000639933e-05,
"loss": 1.7706,
"step": 9380
},
{
"epoch": 0.75,
"grad_norm": 2.546875,
"learning_rate": 1.3820482992848929e-05,
"loss": 1.7685,
"step": 9400
},
{
"epoch": 0.75,
"grad_norm": 2.46875,
"learning_rate": 1.3797292952831127e-05,
"loss": 1.7687,
"step": 9420
},
{
"epoch": 0.75,
"grad_norm": 2.375,
"learning_rate": 1.3774079026460308e-05,
"loss": 1.7768,
"step": 9440
},
{
"epoch": 0.76,
"grad_norm": 2.5,
"learning_rate": 1.3750841359760511e-05,
"loss": 1.7878,
"step": 9460
},
{
"epoch": 0.76,
"grad_norm": 2.40625,
"learning_rate": 1.37275800989051e-05,
"loss": 1.792,
"step": 9480
},
{
"epoch": 0.76,
"grad_norm": 2.75,
"learning_rate": 1.3704295390215868e-05,
"loss": 1.7822,
"step": 9500
},
{
"epoch": 0.76,
"grad_norm": 2.4375,
"learning_rate": 1.3680987380162095e-05,
"loss": 1.7831,
"step": 9520
},
{
"epoch": 0.76,
"grad_norm": 2.453125,
"learning_rate": 1.3657656215359634e-05,
"loss": 1.7819,
"step": 9540
},
{
"epoch": 0.76,
"grad_norm": 2.546875,
"learning_rate": 1.3634302042569995e-05,
"loss": 1.7839,
"step": 9560
},
{
"epoch": 0.76,
"grad_norm": 2.3125,
"learning_rate": 1.3610925008699413e-05,
"loss": 1.7905,
"step": 9580
},
{
"epoch": 0.77,
"grad_norm": 2.484375,
"learning_rate": 1.3587525260797934e-05,
"loss": 1.7785,
"step": 9600
},
{
"epoch": 0.77,
"grad_norm": 2.328125,
"learning_rate": 1.3564102946058468e-05,
"loss": 1.7846,
"step": 9620
},
{
"epoch": 0.77,
"grad_norm": 2.53125,
"learning_rate": 1.3540658211815898e-05,
"loss": 1.7841,
"step": 9640
},
{
"epoch": 0.77,
"grad_norm": 2.453125,
"learning_rate": 1.3517191205546121e-05,
"loss": 1.774,
"step": 9660
},
{
"epoch": 0.77,
"grad_norm": 2.578125,
"learning_rate": 1.3493702074865139e-05,
"loss": 1.7947,
"step": 9680
},
{
"epoch": 0.77,
"grad_norm": 2.390625,
"learning_rate": 1.3470190967528118e-05,
"loss": 1.7843,
"step": 9700
},
{
"epoch": 0.78,
"grad_norm": 2.296875,
"learning_rate": 1.3446658031428474e-05,
"loss": 1.7796,
"step": 9720
},
{
"epoch": 0.78,
"grad_norm": 2.5,
"learning_rate": 1.3423103414596929e-05,
"loss": 1.7713,
"step": 9740
},
{
"epoch": 0.78,
"grad_norm": 2.734375,
"learning_rate": 1.3399527265200581e-05,
"loss": 1.7769,
"step": 9760
},
{
"epoch": 0.78,
"grad_norm": 2.546875,
"learning_rate": 1.3375929731541986e-05,
"loss": 1.7823,
"step": 9780
},
{
"epoch": 0.78,
"grad_norm": 2.59375,
"learning_rate": 1.3352310962058202e-05,
"loss": 1.7642,
"step": 9800
},
{
"epoch": 0.78,
"grad_norm": 2.296875,
"learning_rate": 1.332867110531988e-05,
"loss": 1.7841,
"step": 9820
},
{
"epoch": 0.79,
"grad_norm": 2.609375,
"learning_rate": 1.3305010310030311e-05,
"loss": 1.7897,
"step": 9840
},
{
"epoch": 0.79,
"grad_norm": 2.421875,
"learning_rate": 1.3281328725024496e-05,
"loss": 1.7813,
"step": 9860
},
{
"epoch": 0.79,
"grad_norm": 2.359375,
"learning_rate": 1.3257626499268217e-05,
"loss": 1.7828,
"step": 9880
},
{
"epoch": 0.79,
"grad_norm": 2.390625,
"learning_rate": 1.3233903781857084e-05,
"loss": 1.7809,
"step": 9900
},
{
"epoch": 0.79,
"grad_norm": 2.453125,
"learning_rate": 1.3210160722015619e-05,
"loss": 1.7768,
"step": 9920
},
{
"epoch": 0.79,
"grad_norm": 2.46875,
"learning_rate": 1.3186397469096295e-05,
"loss": 1.7816,
"step": 9940
},
{
"epoch": 0.8,
"grad_norm": 2.703125,
"learning_rate": 1.3162614172578614e-05,
"loss": 1.7741,
"step": 9960
},
{
"epoch": 0.8,
"grad_norm": 2.484375,
"learning_rate": 1.3138810982068154e-05,
"loss": 1.7801,
"step": 9980
},
{
"epoch": 0.8,
"grad_norm": 2.46875,
"learning_rate": 1.3114988047295638e-05,
"loss": 1.7711,
"step": 10000
},
{
"epoch": 0.8,
"grad_norm": 2.3125,
"learning_rate": 1.3091145518115982e-05,
"loss": 1.7807,
"step": 10020
},
{
"epoch": 0.8,
"grad_norm": 2.578125,
"learning_rate": 1.3067283544507366e-05,
"loss": 1.7835,
"step": 10040
},
{
"epoch": 0.8,
"grad_norm": 2.609375,
"learning_rate": 1.3043402276570276e-05,
"loss": 1.7746,
"step": 10060
},
{
"epoch": 0.8,
"grad_norm": 2.53125,
"learning_rate": 1.3019501864526565e-05,
"loss": 1.7742,
"step": 10080
},
{
"epoch": 0.81,
"grad_norm": 2.25,
"learning_rate": 1.2995582458718518e-05,
"loss": 1.7811,
"step": 10100
},
{
"epoch": 0.81,
"grad_norm": 2.3125,
"learning_rate": 1.2971644209607893e-05,
"loss": 1.7684,
"step": 10120
},
{
"epoch": 0.81,
"grad_norm": 2.703125,
"learning_rate": 1.2947687267774973e-05,
"loss": 1.7778,
"step": 10140
},
{
"epoch": 0.81,
"grad_norm": 2.46875,
"learning_rate": 1.2923711783917637e-05,
"loss": 1.7587,
"step": 10160
},
{
"epoch": 0.81,
"grad_norm": 2.296875,
"learning_rate": 1.2899717908850385e-05,
"loss": 1.784,
"step": 10180
},
{
"epoch": 0.81,
"grad_norm": 2.515625,
"learning_rate": 1.2875705793503424e-05,
"loss": 1.773,
"step": 10200
},
{
"epoch": 0.82,
"grad_norm": 2.6875,
"learning_rate": 1.2851675588921677e-05,
"loss": 1.7721,
"step": 10220
},
{
"epoch": 0.82,
"grad_norm": 2.75,
"learning_rate": 1.2827627446263877e-05,
"loss": 1.7781,
"step": 10240
},
{
"epoch": 0.82,
"grad_norm": 2.734375,
"learning_rate": 1.2803561516801575e-05,
"loss": 1.7935,
"step": 10260
},
{
"epoch": 0.82,
"grad_norm": 2.34375,
"learning_rate": 1.2779477951918217e-05,
"loss": 1.7746,
"step": 10280
},
{
"epoch": 0.82,
"grad_norm": 2.453125,
"learning_rate": 1.2755376903108183e-05,
"loss": 1.7783,
"step": 10300
},
{
"epoch": 0.82,
"grad_norm": 2.421875,
"learning_rate": 1.2731258521975829e-05,
"loss": 1.7812,
"step": 10320
},
{
"epoch": 0.83,
"grad_norm": 2.46875,
"learning_rate": 1.2707122960234544e-05,
"loss": 1.7742,
"step": 10340
},
{
"epoch": 0.83,
"grad_norm": 2.671875,
"learning_rate": 1.2682970369705773e-05,
"loss": 1.7585,
"step": 10360
},
{
"epoch": 0.83,
"grad_norm": 2.671875,
"learning_rate": 1.2658800902318103e-05,
"loss": 1.7848,
"step": 10380
},
{
"epoch": 0.83,
"grad_norm": 2.609375,
"learning_rate": 1.2634614710106266e-05,
"loss": 1.7784,
"step": 10400
},
{
"epoch": 0.83,
"grad_norm": 2.421875,
"learning_rate": 1.2610411945210199e-05,
"loss": 1.7762,
"step": 10420
},
{
"epoch": 0.83,
"grad_norm": 2.359375,
"learning_rate": 1.2586192759874094e-05,
"loss": 1.7686,
"step": 10440
},
{
"epoch": 0.84,
"grad_norm": 2.734375,
"learning_rate": 1.2561957306445428e-05,
"loss": 1.7861,
"step": 10460
},
{
"epoch": 0.84,
"grad_norm": 2.53125,
"learning_rate": 1.253770573737402e-05,
"loss": 1.7744,
"step": 10480
},
{
"epoch": 0.84,
"grad_norm": 2.40625,
"learning_rate": 1.2513438205211048e-05,
"loss": 1.7703,
"step": 10500
},
{
"epoch": 0.84,
"grad_norm": 2.65625,
"learning_rate": 1.2489154862608111e-05,
"loss": 1.7785,
"step": 10520
},
{
"epoch": 0.84,
"grad_norm": 2.453125,
"learning_rate": 1.2464855862316263e-05,
"loss": 1.7789,
"step": 10540
},
{
"epoch": 0.84,
"grad_norm": 2.484375,
"learning_rate": 1.244054135718505e-05,
"loss": 1.7766,
"step": 10560
},
{
"epoch": 0.84,
"grad_norm": 2.484375,
"learning_rate": 1.2416211500161546e-05,
"loss": 1.7805,
"step": 10580
},
{
"epoch": 0.85,
"grad_norm": 2.5,
"learning_rate": 1.2391866444289394e-05,
"loss": 1.7769,
"step": 10600
},
{
"epoch": 0.85,
"grad_norm": 2.5,
"learning_rate": 1.2367506342707851e-05,
"loss": 1.7727,
"step": 10620
},
{
"epoch": 0.85,
"grad_norm": 2.640625,
"learning_rate": 1.2343131348650806e-05,
"loss": 1.7603,
"step": 10640
},
{
"epoch": 0.85,
"grad_norm": 2.375,
"learning_rate": 1.231874161544583e-05,
"loss": 1.7681,
"step": 10660
},
{
"epoch": 0.85,
"grad_norm": 2.515625,
"learning_rate": 1.2294337296513219e-05,
"loss": 1.7705,
"step": 10680
},
{
"epoch": 0.85,
"grad_norm": 2.375,
"learning_rate": 1.2269918545365e-05,
"loss": 1.7692,
"step": 10700
},
{
"epoch": 0.86,
"grad_norm": 2.53125,
"learning_rate": 1.2245485515604004e-05,
"loss": 1.7685,
"step": 10720
},
{
"epoch": 0.86,
"grad_norm": 2.84375,
"learning_rate": 1.2221038360922863e-05,
"loss": 1.7873,
"step": 10740
},
{
"epoch": 0.86,
"grad_norm": 2.5,
"learning_rate": 1.219657723510307e-05,
"loss": 1.779,
"step": 10760
},
{
"epoch": 0.86,
"grad_norm": 2.65625,
"learning_rate": 1.2172102292013994e-05,
"loss": 1.7963,
"step": 10780
},
{
"epoch": 0.86,
"grad_norm": 2.4375,
"learning_rate": 1.2147613685611928e-05,
"loss": 1.7737,
"step": 10800
},
{
"epoch": 0.86,
"grad_norm": 2.671875,
"learning_rate": 1.212311156993911e-05,
"loss": 1.7578,
"step": 10820
},
{
"epoch": 0.87,
"grad_norm": 2.875,
"learning_rate": 1.2098596099122745e-05,
"loss": 1.7649,
"step": 10840
},
{
"epoch": 0.87,
"grad_norm": 2.734375,
"learning_rate": 1.2074067427374068e-05,
"loss": 1.782,
"step": 10860
},
{
"epoch": 0.87,
"grad_norm": 2.640625,
"learning_rate": 1.2049525708987331e-05,
"loss": 1.7729,
"step": 10880
},
{
"epoch": 0.87,
"grad_norm": 2.546875,
"learning_rate": 1.2024971098338868e-05,
"loss": 1.7769,
"step": 10900
},
{
"epoch": 0.87,
"grad_norm": 2.40625,
"learning_rate": 1.2000403749886108e-05,
"loss": 1.7761,
"step": 10920
},
{
"epoch": 0.87,
"grad_norm": 2.640625,
"learning_rate": 1.1975823818166596e-05,
"loss": 1.7476,
"step": 10940
},
{
"epoch": 0.87,
"grad_norm": 2.609375,
"learning_rate": 1.1951231457797047e-05,
"loss": 1.7814,
"step": 10960
},
{
"epoch": 0.88,
"grad_norm": 2.703125,
"learning_rate": 1.1926626823472338e-05,
"loss": 1.7691,
"step": 10980
},
{
"epoch": 0.88,
"grad_norm": 2.65625,
"learning_rate": 1.1902010069964569e-05,
"loss": 1.7756,
"step": 11000
},
{
"epoch": 0.88,
"grad_norm": 2.5625,
"learning_rate": 1.1877381352122064e-05,
"loss": 1.7833,
"step": 11020
},
{
"epoch": 0.88,
"grad_norm": 2.4375,
"learning_rate": 1.1852740824868416e-05,
"loss": 1.7659,
"step": 11040
},
{
"epoch": 0.88,
"grad_norm": 2.46875,
"learning_rate": 1.1828088643201492e-05,
"loss": 1.772,
"step": 11060
},
{
"epoch": 0.88,
"grad_norm": 2.546875,
"learning_rate": 1.180342496219248e-05,
"loss": 1.7516,
"step": 11080
},
{
"epoch": 0.89,
"grad_norm": 2.609375,
"learning_rate": 1.17787499369849e-05,
"loss": 1.7647,
"step": 11100
},
{
"epoch": 0.89,
"grad_norm": 2.515625,
"learning_rate": 1.1754063722793624e-05,
"loss": 1.769,
"step": 11120
},
{
"epoch": 0.89,
"grad_norm": 2.671875,
"learning_rate": 1.1729366474903923e-05,
"loss": 1.7813,
"step": 11140
},
{
"epoch": 0.89,
"grad_norm": 2.734375,
"learning_rate": 1.1704658348670455e-05,
"loss": 1.7669,
"step": 11160
},
{
"epoch": 0.89,
"grad_norm": 2.328125,
"learning_rate": 1.1679939499516317e-05,
"loss": 1.7846,
"step": 11180
},
{
"epoch": 0.89,
"grad_norm": 2.375,
"learning_rate": 1.165521008293206e-05,
"loss": 1.7719,
"step": 11200
},
{
"epoch": 0.9,
"grad_norm": 2.65625,
"learning_rate": 1.1630470254474697e-05,
"loss": 1.7625,
"step": 11220
},
{
"epoch": 0.9,
"grad_norm": 2.578125,
"learning_rate": 1.1605720169766752e-05,
"loss": 1.7721,
"step": 11240
},
{
"epoch": 0.9,
"grad_norm": 2.53125,
"learning_rate": 1.1580959984495243e-05,
"loss": 1.7558,
"step": 11260
},
{
"epoch": 0.9,
"grad_norm": 2.609375,
"learning_rate": 1.1556189854410744e-05,
"loss": 1.7633,
"step": 11280
},
{
"epoch": 0.9,
"grad_norm": 2.4375,
"learning_rate": 1.1531409935326377e-05,
"loss": 1.7632,
"step": 11300
},
{
"epoch": 0.9,
"grad_norm": 2.5,
"learning_rate": 1.1506620383116835e-05,
"loss": 1.7925,
"step": 11320
},
{
"epoch": 0.91,
"grad_norm": 2.453125,
"learning_rate": 1.1481821353717418e-05,
"loss": 1.7667,
"step": 11340
},
{
"epoch": 0.91,
"grad_norm": 2.484375,
"learning_rate": 1.145701300312303e-05,
"loss": 1.7733,
"step": 11360
},
{
"epoch": 0.91,
"grad_norm": 2.703125,
"learning_rate": 1.1432195487387223e-05,
"loss": 1.7772,
"step": 11380
},
{
"epoch": 0.91,
"grad_norm": 2.640625,
"learning_rate": 1.1407368962621184e-05,
"loss": 1.7459,
"step": 11400
},
{
"epoch": 0.91,
"grad_norm": 2.59375,
"learning_rate": 1.1382533584992783e-05,
"loss": 1.7608,
"step": 11420
},
{
"epoch": 0.91,
"grad_norm": 2.4375,
"learning_rate": 1.1357689510725571e-05,
"loss": 1.749,
"step": 11440
},
{
"epoch": 0.91,
"grad_norm": 2.5625,
"learning_rate": 1.1332836896097808e-05,
"loss": 1.77,
"step": 11460
},
{
"epoch": 0.92,
"grad_norm": 2.5,
"learning_rate": 1.1307975897441473e-05,
"loss": 1.7676,
"step": 11480
},
{
"epoch": 0.92,
"grad_norm": 2.546875,
"learning_rate": 1.1283106671141282e-05,
"loss": 1.7755,
"step": 11500
},
{
"epoch": 0.92,
"grad_norm": 2.5625,
"learning_rate": 1.1258229373633713e-05,
"loss": 1.7742,
"step": 11520
},
{
"epoch": 0.92,
"grad_norm": 2.828125,
"learning_rate": 1.1233344161406008e-05,
"loss": 1.7606,
"step": 11540
},
{
"epoch": 0.92,
"grad_norm": 2.640625,
"learning_rate": 1.12084511909952e-05,
"loss": 1.7749,
"step": 11560
},
{
"epoch": 0.92,
"grad_norm": 2.640625,
"learning_rate": 1.1183550618987118e-05,
"loss": 1.7868,
"step": 11580
},
{
"epoch": 0.93,
"grad_norm": 2.640625,
"learning_rate": 1.1158642602015415e-05,
"loss": 1.7712,
"step": 11600
},
{
"epoch": 0.93,
"grad_norm": 2.46875,
"learning_rate": 1.1133727296760572e-05,
"loss": 1.7732,
"step": 11620
},
{
"epoch": 0.93,
"grad_norm": 2.546875,
"learning_rate": 1.110880485994891e-05,
"loss": 1.7672,
"step": 11640
},
{
"epoch": 0.93,
"grad_norm": 2.375,
"learning_rate": 1.1083875448351626e-05,
"loss": 1.7858,
"step": 11660
},
{
"epoch": 0.93,
"grad_norm": 2.40625,
"learning_rate": 1.1058939218783772e-05,
"loss": 1.7683,
"step": 11680
},
{
"epoch": 0.93,
"grad_norm": 2.59375,
"learning_rate": 1.10339963281033e-05,
"loss": 1.7813,
"step": 11700
},
{
"epoch": 0.94,
"grad_norm": 2.578125,
"learning_rate": 1.100904693321006e-05,
"loss": 1.7745,
"step": 11720
},
{
"epoch": 0.94,
"grad_norm": 2.484375,
"learning_rate": 1.0984091191044816e-05,
"loss": 1.7848,
"step": 11740
},
{
"epoch": 0.94,
"grad_norm": 2.5,
"learning_rate": 1.0959129258588257e-05,
"loss": 1.7518,
"step": 11760
},
{
"epoch": 0.94,
"grad_norm": 2.53125,
"learning_rate": 1.0934161292860008e-05,
"loss": 1.7768,
"step": 11780
},
{
"epoch": 0.94,
"grad_norm": 2.5625,
"learning_rate": 1.0909187450917656e-05,
"loss": 1.7602,
"step": 11800
},
{
"epoch": 0.94,
"grad_norm": 2.515625,
"learning_rate": 1.0884207889855735e-05,
"loss": 1.758,
"step": 11820
},
{
"epoch": 0.95,
"grad_norm": 2.5625,
"learning_rate": 1.0859222766804778e-05,
"loss": 1.7761,
"step": 11840
},
{
"epoch": 0.95,
"grad_norm": 2.609375,
"learning_rate": 1.0834232238930283e-05,
"loss": 1.7606,
"step": 11860
},
{
"epoch": 0.95,
"grad_norm": 2.59375,
"learning_rate": 1.0809236463431754e-05,
"loss": 1.779,
"step": 11880
},
{
"epoch": 0.95,
"grad_norm": 2.765625,
"learning_rate": 1.0784235597541708e-05,
"loss": 1.771,
"step": 11900
},
{
"epoch": 0.95,
"grad_norm": 2.5625,
"learning_rate": 1.075922979852468e-05,
"loss": 1.7654,
"step": 11920
},
{
"epoch": 0.95,
"grad_norm": 2.703125,
"learning_rate": 1.073421922367623e-05,
"loss": 1.7758,
"step": 11940
},
{
"epoch": 0.95,
"grad_norm": 2.453125,
"learning_rate": 1.0709204030321972e-05,
"loss": 1.7592,
"step": 11960
},
{
"epoch": 0.96,
"grad_norm": 2.546875,
"learning_rate": 1.068418437581656e-05,
"loss": 1.7741,
"step": 11980
},
{
"epoch": 0.96,
"grad_norm": 2.46875,
"learning_rate": 1.0659160417542721e-05,
"loss": 1.759,
"step": 12000
},
{
"epoch": 0.96,
"grad_norm": 2.6875,
"learning_rate": 1.0634132312910245e-05,
"loss": 1.7809,
"step": 12020
},
{
"epoch": 0.96,
"grad_norm": 2.65625,
"learning_rate": 1.060910021935501e-05,
"loss": 1.7811,
"step": 12040
},
{
"epoch": 0.96,
"grad_norm": 2.59375,
"learning_rate": 1.0584064294337983e-05,
"loss": 1.761,
"step": 12060
},
{
"epoch": 0.96,
"grad_norm": 2.40625,
"learning_rate": 1.0559024695344233e-05,
"loss": 1.7515,
"step": 12080
},
{
"epoch": 0.97,
"grad_norm": 2.359375,
"learning_rate": 1.0533981579881938e-05,
"loss": 1.7861,
"step": 12100
},
{
"epoch": 0.97,
"grad_norm": 2.546875,
"learning_rate": 1.0508935105481402e-05,
"loss": 1.7643,
"step": 12120
},
{
"epoch": 0.97,
"grad_norm": 2.546875,
"learning_rate": 1.0483885429694051e-05,
"loss": 1.7745,
"step": 12140
},
{
"epoch": 0.97,
"grad_norm": 2.46875,
"learning_rate": 1.0458832710091448e-05,
"loss": 1.7539,
"step": 12160
},
{
"epoch": 0.97,
"grad_norm": 2.421875,
"learning_rate": 1.0433777104264313e-05,
"loss": 1.7546,
"step": 12180
},
{
"epoch": 0.97,
"grad_norm": 2.53125,
"learning_rate": 1.0408718769821512e-05,
"loss": 1.7606,
"step": 12200
},
{
"epoch": 0.98,
"grad_norm": 2.765625,
"learning_rate": 1.0383657864389077e-05,
"loss": 1.7583,
"step": 12220
},
{
"epoch": 0.98,
"grad_norm": 2.828125,
"learning_rate": 1.0358594545609207e-05,
"loss": 1.7659,
"step": 12240
},
{
"epoch": 0.98,
"grad_norm": 2.53125,
"learning_rate": 1.0333528971139297e-05,
"loss": 1.7601,
"step": 12260
},
{
"epoch": 0.98,
"grad_norm": 2.421875,
"learning_rate": 1.0308461298650923e-05,
"loss": 1.7612,
"step": 12280
},
{
"epoch": 0.98,
"grad_norm": 2.484375,
"learning_rate": 1.0283391685828844e-05,
"loss": 1.7646,
"step": 12300
},
{
"epoch": 0.98,
"grad_norm": 2.765625,
"learning_rate": 1.0258320290370051e-05,
"loss": 1.7741,
"step": 12320
},
{
"epoch": 0.99,
"grad_norm": 2.359375,
"learning_rate": 1.0233247269982732e-05,
"loss": 1.7616,
"step": 12340
},
{
"epoch": 0.99,
"grad_norm": 2.625,
"learning_rate": 1.0208172782385295e-05,
"loss": 1.7502,
"step": 12360
},
{
"epoch": 0.99,
"grad_norm": 2.65625,
"learning_rate": 1.0183096985305385e-05,
"loss": 1.7806,
"step": 12380
},
{
"epoch": 0.99,
"grad_norm": 2.578125,
"learning_rate": 1.0158020036478881e-05,
"loss": 1.7728,
"step": 12400
}
],
"logging_steps": 20,
"max_steps": 25052,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 200,
"total_flos": 1.9252630386848563e+19,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}