15984 lines
372 KiB
JSON
15984 lines
372 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.175,
|
|
"eval_steps": 650,
|
|
"global_step": 2275,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 7.692307692307693e-05,
|
|
"grad_norm": 3.953125,
|
|
"learning_rate": 1e-07,
|
|
"loss": 2.715610980987549,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.00015384615384615385,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 1.8239655172413794e-06,
|
|
"loss": 2.5245614051818848,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.00023076923076923076,
|
|
"grad_norm": 1.953125,
|
|
"learning_rate": 3.547931034482759e-06,
|
|
"loss": 2.520277738571167,
|
|
"step": 3
|
|
},
|
|
{
|
|
"epoch": 0.0003076923076923077,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 5.271896551724138e-06,
|
|
"loss": 2.482571601867676,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.0003846153846153846,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 6.995862068965518e-06,
|
|
"loss": 2.4671521186828613,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.00046153846153846153,
|
|
"grad_norm": 3.59375,
|
|
"learning_rate": 8.719827586206896e-06,
|
|
"loss": 2.6088168621063232,
|
|
"step": 6
|
|
},
|
|
{
|
|
"epoch": 0.0005384615384615384,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 1.0443793103448277e-05,
|
|
"loss": 2.4543330669403076,
|
|
"step": 7
|
|
},
|
|
{
|
|
"epoch": 0.0006153846153846154,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 1.2167758620689656e-05,
|
|
"loss": 2.4998857975006104,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.0006923076923076923,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 1.3891724137931037e-05,
|
|
"loss": 2.4842681884765625,
|
|
"step": 9
|
|
},
|
|
{
|
|
"epoch": 0.0007692307692307692,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 1.5615689655172412e-05,
|
|
"loss": 2.485731363296509,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.0008461538461538462,
|
|
"grad_norm": 3.921875,
|
|
"learning_rate": 1.733965517241379e-05,
|
|
"loss": 2.615187883377075,
|
|
"step": 11
|
|
},
|
|
{
|
|
"epoch": 0.0009230769230769231,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 1.9063620689655174e-05,
|
|
"loss": 2.509403944015503,
|
|
"step": 12
|
|
},
|
|
{
|
|
"epoch": 0.001,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 2.078758620689655e-05,
|
|
"loss": 2.51432466506958,
|
|
"step": 13
|
|
},
|
|
{
|
|
"epoch": 0.0010769230769230769,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 2.2511551724137932e-05,
|
|
"loss": 2.5114333629608154,
|
|
"step": 14
|
|
},
|
|
{
|
|
"epoch": 0.001153846153846154,
|
|
"grad_norm": 2.328125,
|
|
"learning_rate": 2.423551724137931e-05,
|
|
"loss": 2.534618854522705,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.0012307692307692308,
|
|
"grad_norm": 3.875,
|
|
"learning_rate": 2.595948275862069e-05,
|
|
"loss": 2.7278454303741455,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.0013076923076923077,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 2.7683448275862068e-05,
|
|
"loss": 2.4931509494781494,
|
|
"step": 17
|
|
},
|
|
{
|
|
"epoch": 0.0013846153846153845,
|
|
"grad_norm": 1.953125,
|
|
"learning_rate": 2.940741379310345e-05,
|
|
"loss": 2.5176568031311035,
|
|
"step": 18
|
|
},
|
|
{
|
|
"epoch": 0.0014615384615384616,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 3.113137931034483e-05,
|
|
"loss": 2.5142030715942383,
|
|
"step": 19
|
|
},
|
|
{
|
|
"epoch": 0.0015384615384615385,
|
|
"grad_norm": 1.921875,
|
|
"learning_rate": 3.285534482758621e-05,
|
|
"loss": 2.4683785438537598,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.0016153846153846153,
|
|
"grad_norm": 3.5625,
|
|
"learning_rate": 3.4579310344827584e-05,
|
|
"loss": 2.6252732276916504,
|
|
"step": 21
|
|
},
|
|
{
|
|
"epoch": 0.0016923076923076924,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 3.630327586206897e-05,
|
|
"loss": 2.5508525371551514,
|
|
"step": 22
|
|
},
|
|
{
|
|
"epoch": 0.0017692307692307693,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 3.802724137931035e-05,
|
|
"loss": 2.4862871170043945,
|
|
"step": 23
|
|
},
|
|
{
|
|
"epoch": 0.0018461538461538461,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 3.975120689655173e-05,
|
|
"loss": 2.5072364807128906,
|
|
"step": 24
|
|
},
|
|
{
|
|
"epoch": 0.0019230769230769232,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 4.147517241379311e-05,
|
|
"loss": 2.457385540008545,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.002,
|
|
"grad_norm": 3.796875,
|
|
"learning_rate": 4.3199137931034484e-05,
|
|
"loss": 2.6437313556671143,
|
|
"step": 26
|
|
},
|
|
{
|
|
"epoch": 0.002076923076923077,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 4.492310344827587e-05,
|
|
"loss": 2.5256571769714355,
|
|
"step": 27
|
|
},
|
|
{
|
|
"epoch": 0.0021538461538461538,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 4.6647068965517246e-05,
|
|
"loss": 2.546295166015625,
|
|
"step": 28
|
|
},
|
|
{
|
|
"epoch": 0.0022307692307692306,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 4.837103448275862e-05,
|
|
"loss": 2.4317626953125,
|
|
"step": 29
|
|
},
|
|
{
|
|
"epoch": 0.002307692307692308,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 5.0095e-05,
|
|
"loss": 2.468461036682129,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.0023846153846153848,
|
|
"grad_norm": 3.25,
|
|
"learning_rate": 5.1818965517241385e-05,
|
|
"loss": 2.6549417972564697,
|
|
"step": 31
|
|
},
|
|
{
|
|
"epoch": 0.0024615384615384616,
|
|
"grad_norm": 2.859375,
|
|
"learning_rate": 5.354293103448277e-05,
|
|
"loss": 2.5670971870422363,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.0025384615384615385,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 5.5266896551724146e-05,
|
|
"loss": 2.5535831451416016,
|
|
"step": 33
|
|
},
|
|
{
|
|
"epoch": 0.0026153846153846153,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 5.699086206896552e-05,
|
|
"loss": 2.451829433441162,
|
|
"step": 34
|
|
},
|
|
{
|
|
"epoch": 0.002692307692307692,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 5.87148275862069e-05,
|
|
"loss": 2.4535882472991943,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.002769230769230769,
|
|
"grad_norm": 3.375,
|
|
"learning_rate": 6.043879310344828e-05,
|
|
"loss": 2.655374765396118,
|
|
"step": 36
|
|
},
|
|
{
|
|
"epoch": 0.0028461538461538463,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 6.216275862068966e-05,
|
|
"loss": 2.49287748336792,
|
|
"step": 37
|
|
},
|
|
{
|
|
"epoch": 0.002923076923076923,
|
|
"grad_norm": 2.234375,
|
|
"learning_rate": 6.388672413793103e-05,
|
|
"loss": 2.52790904045105,
|
|
"step": 38
|
|
},
|
|
{
|
|
"epoch": 0.003,
|
|
"grad_norm": 1.859375,
|
|
"learning_rate": 6.561068965517242e-05,
|
|
"loss": 2.4939723014831543,
|
|
"step": 39
|
|
},
|
|
{
|
|
"epoch": 0.003076923076923077,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 6.73346551724138e-05,
|
|
"loss": 2.5158371925354004,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.0031538461538461538,
|
|
"grad_norm": 3.15625,
|
|
"learning_rate": 6.905862068965517e-05,
|
|
"loss": 2.6229329109191895,
|
|
"step": 41
|
|
},
|
|
{
|
|
"epoch": 0.0032307692307692306,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 7.078258620689656e-05,
|
|
"loss": 2.5449061393737793,
|
|
"step": 42
|
|
},
|
|
{
|
|
"epoch": 0.0033076923076923075,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 7.250655172413793e-05,
|
|
"loss": 2.516963481903076,
|
|
"step": 43
|
|
},
|
|
{
|
|
"epoch": 0.003384615384615385,
|
|
"grad_norm": 1.828125,
|
|
"learning_rate": 7.423051724137931e-05,
|
|
"loss": 2.4559173583984375,
|
|
"step": 44
|
|
},
|
|
{
|
|
"epoch": 0.0034615384615384616,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 7.59544827586207e-05,
|
|
"loss": 2.427342176437378,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.0035384615384615385,
|
|
"grad_norm": 4.15625,
|
|
"learning_rate": 7.767844827586207e-05,
|
|
"loss": 2.6887357234954834,
|
|
"step": 46
|
|
},
|
|
{
|
|
"epoch": 0.0036153846153846154,
|
|
"grad_norm": 2.125,
|
|
"learning_rate": 7.940241379310346e-05,
|
|
"loss": 2.497612714767456,
|
|
"step": 47
|
|
},
|
|
{
|
|
"epoch": 0.0036923076923076922,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 8.112637931034483e-05,
|
|
"loss": 2.467332363128662,
|
|
"step": 48
|
|
},
|
|
{
|
|
"epoch": 0.003769230769230769,
|
|
"grad_norm": 1.8203125,
|
|
"learning_rate": 8.285034482758621e-05,
|
|
"loss": 2.472888708114624,
|
|
"step": 49
|
|
},
|
|
{
|
|
"epoch": 0.0038461538461538464,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 8.457431034482759e-05,
|
|
"loss": 2.4593591690063477,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.003923076923076923,
|
|
"grad_norm": 4.0,
|
|
"learning_rate": 8.629827586206897e-05,
|
|
"loss": 2.7112505435943604,
|
|
"step": 51
|
|
},
|
|
{
|
|
"epoch": 0.004,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 8.802224137931036e-05,
|
|
"loss": 2.5246660709381104,
|
|
"step": 52
|
|
},
|
|
{
|
|
"epoch": 0.004076923076923077,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 8.974620689655173e-05,
|
|
"loss": 2.4498183727264404,
|
|
"step": 53
|
|
},
|
|
{
|
|
"epoch": 0.004153846153846154,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 9.14701724137931e-05,
|
|
"loss": 2.503476619720459,
|
|
"step": 54
|
|
},
|
|
{
|
|
"epoch": 0.004230769230769231,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 9.319413793103449e-05,
|
|
"loss": 2.486161708831787,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.0043076923076923075,
|
|
"grad_norm": 3.578125,
|
|
"learning_rate": 9.491810344827587e-05,
|
|
"loss": 2.6522369384765625,
|
|
"step": 56
|
|
},
|
|
{
|
|
"epoch": 0.004384615384615384,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 9.664206896551724e-05,
|
|
"loss": 2.492124319076538,
|
|
"step": 57
|
|
},
|
|
{
|
|
"epoch": 0.004461538461538461,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 9.836603448275862e-05,
|
|
"loss": 2.5395843982696533,
|
|
"step": 58
|
|
},
|
|
{
|
|
"epoch": 0.004538461538461538,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 0.00010009,
|
|
"loss": 2.4535837173461914,
|
|
"step": 59
|
|
},
|
|
{
|
|
"epoch": 0.004615384615384616,
|
|
"grad_norm": 1.6640625,
|
|
"learning_rate": 0.00010181396551724139,
|
|
"loss": 2.4544646739959717,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.004692307692307693,
|
|
"grad_norm": 3.515625,
|
|
"learning_rate": 0.00010353793103448277,
|
|
"loss": 2.6017041206359863,
|
|
"step": 61
|
|
},
|
|
{
|
|
"epoch": 0.0047692307692307695,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 0.00010526189655172416,
|
|
"loss": 2.518475294113159,
|
|
"step": 62
|
|
},
|
|
{
|
|
"epoch": 0.004846153846153846,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 0.00010698586206896554,
|
|
"loss": 2.4987432956695557,
|
|
"step": 63
|
|
},
|
|
{
|
|
"epoch": 0.004923076923076923,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 0.0001087098275862069,
|
|
"loss": 2.4908711910247803,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.005,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.00011043379310344829,
|
|
"loss": 2.45800518989563,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.005076923076923077,
|
|
"grad_norm": 3.484375,
|
|
"learning_rate": 0.00011215775862068967,
|
|
"loss": 2.5989508628845215,
|
|
"step": 66
|
|
},
|
|
{
|
|
"epoch": 0.005153846153846154,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 0.00011388172413793103,
|
|
"loss": 2.5074410438537598,
|
|
"step": 67
|
|
},
|
|
{
|
|
"epoch": 0.005230769230769231,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 0.00011560568965517242,
|
|
"loss": 2.5043785572052,
|
|
"step": 68
|
|
},
|
|
{
|
|
"epoch": 0.0053076923076923075,
|
|
"grad_norm": 1.875,
|
|
"learning_rate": 0.0001173296551724138,
|
|
"loss": 2.4298713207244873,
|
|
"step": 69
|
|
},
|
|
{
|
|
"epoch": 0.005384615384615384,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 0.00011905362068965518,
|
|
"loss": 2.5199856758117676,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.005461538461538461,
|
|
"grad_norm": 3.5625,
|
|
"learning_rate": 0.00012077758620689655,
|
|
"loss": 2.613662004470825,
|
|
"step": 71
|
|
},
|
|
{
|
|
"epoch": 0.005538461538461538,
|
|
"grad_norm": 1.8203125,
|
|
"learning_rate": 0.00012250155172413793,
|
|
"loss": 2.488935708999634,
|
|
"step": 72
|
|
},
|
|
{
|
|
"epoch": 0.005615384615384615,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 0.0001242255172413793,
|
|
"loss": 2.478847026824951,
|
|
"step": 73
|
|
},
|
|
{
|
|
"epoch": 0.005692307692307693,
|
|
"grad_norm": 1.90625,
|
|
"learning_rate": 0.0001259494827586207,
|
|
"loss": 2.458312749862671,
|
|
"step": 74
|
|
},
|
|
{
|
|
"epoch": 0.0057692307692307696,
|
|
"grad_norm": 2.4375,
|
|
"learning_rate": 0.00012767344827586206,
|
|
"loss": 2.5617313385009766,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.005846153846153846,
|
|
"grad_norm": 3.15625,
|
|
"learning_rate": 0.00012939741379310344,
|
|
"loss": 2.559849977493286,
|
|
"step": 76
|
|
},
|
|
{
|
|
"epoch": 0.005923076923076923,
|
|
"grad_norm": 2.234375,
|
|
"learning_rate": 0.00013112137931034485,
|
|
"loss": 2.5504000186920166,
|
|
"step": 77
|
|
},
|
|
{
|
|
"epoch": 0.006,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 0.00013284534482758622,
|
|
"loss": 2.4511373043060303,
|
|
"step": 78
|
|
},
|
|
{
|
|
"epoch": 0.006076923076923077,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 0.0001345693103448276,
|
|
"loss": 2.5114150047302246,
|
|
"step": 79
|
|
},
|
|
{
|
|
"epoch": 0.006153846153846154,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 0.00013629327586206898,
|
|
"loss": 2.5414695739746094,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.006230769230769231,
|
|
"grad_norm": 3.625,
|
|
"learning_rate": 0.00013801724137931033,
|
|
"loss": 2.603720188140869,
|
|
"step": 81
|
|
},
|
|
{
|
|
"epoch": 0.0063076923076923076,
|
|
"grad_norm": 1.9921875,
|
|
"learning_rate": 0.00013974120689655173,
|
|
"loss": 2.4926371574401855,
|
|
"step": 82
|
|
},
|
|
{
|
|
"epoch": 0.006384615384615384,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 0.0001414651724137931,
|
|
"loss": 2.508887529373169,
|
|
"step": 83
|
|
},
|
|
{
|
|
"epoch": 0.006461538461538461,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 0.0001431891379310345,
|
|
"loss": 2.4919683933258057,
|
|
"step": 84
|
|
},
|
|
{
|
|
"epoch": 0.006538461538461538,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 0.00014491310344827586,
|
|
"loss": 2.5718719959259033,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.006615384615384615,
|
|
"grad_norm": 3.375,
|
|
"learning_rate": 0.00014663706896551724,
|
|
"loss": 2.6013636589050293,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 0.006692307692307693,
|
|
"grad_norm": 1.953125,
|
|
"learning_rate": 0.00014836103448275862,
|
|
"loss": 2.4935154914855957,
|
|
"step": 87
|
|
},
|
|
{
|
|
"epoch": 0.00676923076923077,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 0.00015008500000000002,
|
|
"loss": 2.4900732040405273,
|
|
"step": 88
|
|
},
|
|
{
|
|
"epoch": 0.006846153846153846,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 0.0001518089655172414,
|
|
"loss": 2.4658381938934326,
|
|
"step": 89
|
|
},
|
|
{
|
|
"epoch": 0.006923076923076923,
|
|
"grad_norm": 2.3125,
|
|
"learning_rate": 0.00015353293103448278,
|
|
"loss": 2.540864944458008,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.007,
|
|
"grad_norm": 3.578125,
|
|
"learning_rate": 0.00015525689655172413,
|
|
"loss": 2.602909803390503,
|
|
"step": 91
|
|
},
|
|
{
|
|
"epoch": 0.007076923076923077,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 0.00015698086206896553,
|
|
"loss": 2.50783634185791,
|
|
"step": 92
|
|
},
|
|
{
|
|
"epoch": 0.007153846153846154,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 0.0001587048275862069,
|
|
"loss": 2.445280075073242,
|
|
"step": 93
|
|
},
|
|
{
|
|
"epoch": 0.007230769230769231,
|
|
"grad_norm": 1.703125,
|
|
"learning_rate": 0.0001604287931034483,
|
|
"loss": 2.4400634765625,
|
|
"step": 94
|
|
},
|
|
{
|
|
"epoch": 0.007307692307692308,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 0.00016215275862068967,
|
|
"loss": 2.5592000484466553,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.0073846153846153844,
|
|
"grad_norm": 3.28125,
|
|
"learning_rate": 0.00016387672413793104,
|
|
"loss": 2.5681962966918945,
|
|
"step": 96
|
|
},
|
|
{
|
|
"epoch": 0.007461538461538461,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 0.00016560068965517242,
|
|
"loss": 2.5396571159362793,
|
|
"step": 97
|
|
},
|
|
{
|
|
"epoch": 0.007538461538461538,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 0.0001673246551724138,
|
|
"loss": 2.571528911590576,
|
|
"step": 98
|
|
},
|
|
{
|
|
"epoch": 0.007615384615384615,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 0.00016904862068965518,
|
|
"loss": 2.4412589073181152,
|
|
"step": 99
|
|
},
|
|
{
|
|
"epoch": 0.007692307692307693,
|
|
"grad_norm": 2.21875,
|
|
"learning_rate": 0.00017077258620689655,
|
|
"loss": 2.519843578338623,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.00776923076923077,
|
|
"grad_norm": 3.140625,
|
|
"learning_rate": 0.00017249655172413793,
|
|
"loss": 2.5585575103759766,
|
|
"step": 101
|
|
},
|
|
{
|
|
"epoch": 0.007846153846153846,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 0.00017422051724137933,
|
|
"loss": 2.496175527572632,
|
|
"step": 102
|
|
},
|
|
{
|
|
"epoch": 0.007923076923076923,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 0.0001759444827586207,
|
|
"loss": 2.48506760597229,
|
|
"step": 103
|
|
},
|
|
{
|
|
"epoch": 0.008,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0001776684482758621,
|
|
"loss": 2.4228076934814453,
|
|
"step": 104
|
|
},
|
|
{
|
|
"epoch": 0.008076923076923077,
|
|
"grad_norm": 2.40625,
|
|
"learning_rate": 0.00017939241379310347,
|
|
"loss": 2.5842766761779785,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.008153846153846154,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 0.0001811163793103448,
|
|
"loss": 2.528358221054077,
|
|
"step": 106
|
|
},
|
|
{
|
|
"epoch": 0.00823076923076923,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 0.0001828403448275862,
|
|
"loss": 2.5383832454681396,
|
|
"step": 107
|
|
},
|
|
{
|
|
"epoch": 0.008307692307692308,
|
|
"grad_norm": 1.765625,
|
|
"learning_rate": 0.0001845643103448276,
|
|
"loss": 2.471806764602661,
|
|
"step": 108
|
|
},
|
|
{
|
|
"epoch": 0.008384615384615384,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 0.00018628827586206898,
|
|
"loss": 2.5172243118286133,
|
|
"step": 109
|
|
},
|
|
{
|
|
"epoch": 0.008461538461538461,
|
|
"grad_norm": 3.21875,
|
|
"learning_rate": 0.00018801224137931035,
|
|
"loss": 2.596045732498169,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.008538461538461538,
|
|
"grad_norm": 2.796875,
|
|
"learning_rate": 0.00018973620689655173,
|
|
"loss": 2.5905539989471436,
|
|
"step": 111
|
|
},
|
|
{
|
|
"epoch": 0.008615384615384615,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 0.00019146017241379308,
|
|
"loss": 2.539374351501465,
|
|
"step": 112
|
|
},
|
|
{
|
|
"epoch": 0.008692307692307692,
|
|
"grad_norm": 1.90625,
|
|
"learning_rate": 0.00019318413793103449,
|
|
"loss": 2.4923410415649414,
|
|
"step": 113
|
|
},
|
|
{
|
|
"epoch": 0.008769230769230769,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 0.00019490810344827586,
|
|
"loss": 2.475188970565796,
|
|
"step": 114
|
|
},
|
|
{
|
|
"epoch": 0.008846153846153846,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 0.00019663206896551724,
|
|
"loss": 2.541165351867676,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.008923076923076922,
|
|
"grad_norm": 3.609375,
|
|
"learning_rate": 0.00019835603448275862,
|
|
"loss": 2.622483491897583,
|
|
"step": 116
|
|
},
|
|
{
|
|
"epoch": 0.009,
|
|
"grad_norm": 1.828125,
|
|
"learning_rate": 0.00020008,
|
|
"loss": 2.520427703857422,
|
|
"step": 117
|
|
},
|
|
{
|
|
"epoch": 0.009076923076923076,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 0.00020180396551724137,
|
|
"loss": 2.4928951263427734,
|
|
"step": 118
|
|
},
|
|
{
|
|
"epoch": 0.009153846153846153,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 0.00020352793103448275,
|
|
"loss": 2.495316505432129,
|
|
"step": 119
|
|
},
|
|
{
|
|
"epoch": 0.009230769230769232,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 0.00020525189655172413,
|
|
"loss": 2.478546142578125,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.009307692307692308,
|
|
"grad_norm": 3.46875,
|
|
"learning_rate": 0.00020697586206896553,
|
|
"loss": 2.6133782863616943,
|
|
"step": 121
|
|
},
|
|
{
|
|
"epoch": 0.009384615384615385,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 0.0002086998275862069,
|
|
"loss": 2.5370616912841797,
|
|
"step": 122
|
|
},
|
|
{
|
|
"epoch": 0.009461538461538462,
|
|
"grad_norm": 1.96875,
|
|
"learning_rate": 0.00021042379310344831,
|
|
"loss": 2.499112129211426,
|
|
"step": 123
|
|
},
|
|
{
|
|
"epoch": 0.009538461538461539,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.0002121477586206897,
|
|
"loss": 2.4469897747039795,
|
|
"step": 124
|
|
},
|
|
{
|
|
"epoch": 0.009615384615384616,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 0.00021387172413793107,
|
|
"loss": 2.457775831222534,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.009692307692307693,
|
|
"grad_norm": 3.0,
|
|
"learning_rate": 0.00021559568965517242,
|
|
"loss": 2.6406683921813965,
|
|
"step": 126
|
|
},
|
|
{
|
|
"epoch": 0.00976923076923077,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 0.0002173196551724138,
|
|
"loss": 2.474027156829834,
|
|
"step": 127
|
|
},
|
|
{
|
|
"epoch": 0.009846153846153846,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 0.0002190436206896552,
|
|
"loss": 2.484086275100708,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 0.009923076923076923,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 0.00022076758620689658,
|
|
"loss": 2.4875786304473877,
|
|
"step": 129
|
|
},
|
|
{
|
|
"epoch": 0.01,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 0.00022249155172413796,
|
|
"loss": 2.481260061264038,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.010076923076923077,
|
|
"grad_norm": 3.234375,
|
|
"learning_rate": 0.00022421551724137933,
|
|
"loss": 2.6462113857269287,
|
|
"step": 131
|
|
},
|
|
{
|
|
"epoch": 0.010153846153846154,
|
|
"grad_norm": 1.7578125,
|
|
"learning_rate": 0.00022593948275862068,
|
|
"loss": 2.4575135707855225,
|
|
"step": 132
|
|
},
|
|
{
|
|
"epoch": 0.01023076923076923,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 0.00022766344827586206,
|
|
"loss": 2.486748218536377,
|
|
"step": 133
|
|
},
|
|
{
|
|
"epoch": 0.010307692307692308,
|
|
"grad_norm": 1.8046875,
|
|
"learning_rate": 0.00022938741379310346,
|
|
"loss": 2.467606544494629,
|
|
"step": 134
|
|
},
|
|
{
|
|
"epoch": 0.010384615384615384,
|
|
"grad_norm": 1.8984375,
|
|
"learning_rate": 0.00023111137931034484,
|
|
"loss": 2.4958724975585938,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.010461538461538461,
|
|
"grad_norm": 3.265625,
|
|
"learning_rate": 0.00023283534482758622,
|
|
"loss": 2.6400978565216064,
|
|
"step": 136
|
|
},
|
|
{
|
|
"epoch": 0.010538461538461538,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 0.0002345593103448276,
|
|
"loss": 2.4838132858276367,
|
|
"step": 137
|
|
},
|
|
{
|
|
"epoch": 0.010615384615384615,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 0.00023628327586206895,
|
|
"loss": 2.521484613418579,
|
|
"step": 138
|
|
},
|
|
{
|
|
"epoch": 0.010692307692307692,
|
|
"grad_norm": 1.9140625,
|
|
"learning_rate": 0.00023800724137931035,
|
|
"loss": 2.4443867206573486,
|
|
"step": 139
|
|
},
|
|
{
|
|
"epoch": 0.010769230769230769,
|
|
"grad_norm": 1.7890625,
|
|
"learning_rate": 0.00023973120689655173,
|
|
"loss": 2.4744012355804443,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.010846153846153846,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 0.0002414551724137931,
|
|
"loss": 2.6226348876953125,
|
|
"step": 141
|
|
},
|
|
{
|
|
"epoch": 0.010923076923076923,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 0.00024317913793103448,
|
|
"loss": 2.476577043533325,
|
|
"step": 142
|
|
},
|
|
{
|
|
"epoch": 0.011,
|
|
"grad_norm": 1.796875,
|
|
"learning_rate": 0.00024490310344827586,
|
|
"loss": 2.4843080043792725,
|
|
"step": 143
|
|
},
|
|
{
|
|
"epoch": 0.011076923076923076,
|
|
"grad_norm": 1.7265625,
|
|
"learning_rate": 0.00024662706896551724,
|
|
"loss": 2.4499170780181885,
|
|
"step": 144
|
|
},
|
|
{
|
|
"epoch": 0.011153846153846153,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 0.0002483510344827586,
|
|
"loss": 2.495089530944824,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.01123076923076923,
|
|
"grad_norm": 2.921875,
|
|
"learning_rate": 0.000250075,
|
|
"loss": 2.5947680473327637,
|
|
"step": 146
|
|
},
|
|
{
|
|
"epoch": 0.011307692307692309,
|
|
"grad_norm": 2.140625,
|
|
"learning_rate": 0.00025179896551724137,
|
|
"loss": 2.524608612060547,
|
|
"step": 147
|
|
},
|
|
{
|
|
"epoch": 0.011384615384615385,
|
|
"grad_norm": 1.6015625,
|
|
"learning_rate": 0.0002535229310344828,
|
|
"loss": 2.4754514694213867,
|
|
"step": 148
|
|
},
|
|
{
|
|
"epoch": 0.011461538461538462,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 0.0002552468965517241,
|
|
"loss": 2.4597668647766113,
|
|
"step": 149
|
|
},
|
|
{
|
|
"epoch": 0.011538461538461539,
|
|
"grad_norm": 1.9375,
|
|
"learning_rate": 0.0002569708620689655,
|
|
"loss": 2.506704092025757,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.011615384615384616,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 0.0002586948275862069,
|
|
"loss": 2.546921968460083,
|
|
"step": 151
|
|
},
|
|
{
|
|
"epoch": 0.011692307692307693,
|
|
"grad_norm": 2.0,
|
|
"learning_rate": 0.0002604187931034483,
|
|
"loss": 2.4841909408569336,
|
|
"step": 152
|
|
},
|
|
{
|
|
"epoch": 0.01176923076923077,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 0.0002621427586206897,
|
|
"loss": 2.4821231365203857,
|
|
"step": 153
|
|
},
|
|
{
|
|
"epoch": 0.011846153846153847,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.00026386672413793107,
|
|
"loss": 2.4113540649414062,
|
|
"step": 154
|
|
},
|
|
{
|
|
"epoch": 0.011923076923076923,
|
|
"grad_norm": 1.7421875,
|
|
"learning_rate": 0.00026559068965517244,
|
|
"loss": 2.4521214962005615,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.012,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 0.0002673146551724138,
|
|
"loss": 2.628316640853882,
|
|
"step": 156
|
|
},
|
|
{
|
|
"epoch": 0.012076923076923077,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 0.0002690386206896552,
|
|
"loss": 2.494407892227173,
|
|
"step": 157
|
|
},
|
|
{
|
|
"epoch": 0.012153846153846154,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 0.0002707625862068966,
|
|
"loss": 2.4451377391815186,
|
|
"step": 158
|
|
},
|
|
{
|
|
"epoch": 0.01223076923076923,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 0.00027248655172413795,
|
|
"loss": 2.51568865776062,
|
|
"step": 159
|
|
},
|
|
{
|
|
"epoch": 0.012307692307692308,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 0.0002742105172413793,
|
|
"loss": 2.5070724487304688,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.012384615384615385,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 0.00027593448275862065,
|
|
"loss": 2.541414737701416,
|
|
"step": 161
|
|
},
|
|
{
|
|
"epoch": 0.012461538461538461,
|
|
"grad_norm": 2.546875,
|
|
"learning_rate": 0.0002776584482758621,
|
|
"loss": 2.5513970851898193,
|
|
"step": 162
|
|
},
|
|
{
|
|
"epoch": 0.012538461538461538,
|
|
"grad_norm": 2.015625,
|
|
"learning_rate": 0.00027938241379310346,
|
|
"loss": 2.4974682331085205,
|
|
"step": 163
|
|
},
|
|
{
|
|
"epoch": 0.012615384615384615,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.00028110637931034484,
|
|
"loss": 2.4881885051727295,
|
|
"step": 164
|
|
},
|
|
{
|
|
"epoch": 0.012692307692307692,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 0.0002828303448275862,
|
|
"loss": 2.46138072013855,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 0.012769230769230769,
|
|
"grad_norm": 2.5625,
|
|
"learning_rate": 0.0002845543103448276,
|
|
"loss": 2.552140712738037,
|
|
"step": 166
|
|
},
|
|
{
|
|
"epoch": 0.012846153846153846,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 0.00028627827586206897,
|
|
"loss": 2.5298194885253906,
|
|
"step": 167
|
|
},
|
|
{
|
|
"epoch": 0.012923076923076923,
|
|
"grad_norm": 1.515625,
|
|
"learning_rate": 0.00028800224137931035,
|
|
"loss": 2.4455153942108154,
|
|
"step": 168
|
|
},
|
|
{
|
|
"epoch": 0.013,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 0.00028972620689655173,
|
|
"loss": 2.4622244834899902,
|
|
"step": 169
|
|
},
|
|
{
|
|
"epoch": 0.013076923076923076,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 0.0002914501724137931,
|
|
"loss": 2.4402244091033936,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.013153846153846153,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 0.0002931741379310345,
|
|
"loss": 2.564606189727783,
|
|
"step": 171
|
|
},
|
|
{
|
|
"epoch": 0.01323076923076923,
|
|
"grad_norm": 2.9375,
|
|
"learning_rate": 0.00029489810344827586,
|
|
"loss": 2.5341174602508545,
|
|
"step": 172
|
|
},
|
|
{
|
|
"epoch": 0.013307692307692307,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.00029662206896551724,
|
|
"loss": 2.4565837383270264,
|
|
"step": 173
|
|
},
|
|
{
|
|
"epoch": 0.013384615384615385,
|
|
"grad_norm": 1.8671875,
|
|
"learning_rate": 0.0002983460344827586,
|
|
"loss": 2.468684196472168,
|
|
"step": 174
|
|
},
|
|
{
|
|
"epoch": 0.013461538461538462,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 0.00030007000000000005,
|
|
"loss": 2.454108238220215,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.01353846153846154,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 0.0003017939655172414,
|
|
"loss": 2.532437324523926,
|
|
"step": 176
|
|
},
|
|
{
|
|
"epoch": 0.013615384615384616,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 0.0003035179310344828,
|
|
"loss": 2.5242857933044434,
|
|
"step": 177
|
|
},
|
|
{
|
|
"epoch": 0.013692307692307693,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.0003052418965517242,
|
|
"loss": 2.446073532104492,
|
|
"step": 178
|
|
},
|
|
{
|
|
"epoch": 0.01376923076923077,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 0.00030696586206896556,
|
|
"loss": 2.468057870864868,
|
|
"step": 179
|
|
},
|
|
{
|
|
"epoch": 0.013846153846153847,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.0003086898275862069,
|
|
"loss": 2.435492753982544,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.013923076923076923,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 0.00031041379310344826,
|
|
"loss": 2.5175681114196777,
|
|
"step": 181
|
|
},
|
|
{
|
|
"epoch": 0.014,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 0.00031213775862068963,
|
|
"loss": 2.5015745162963867,
|
|
"step": 182
|
|
},
|
|
{
|
|
"epoch": 0.014076923076923077,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 0.00031386172413793106,
|
|
"loss": 2.5343284606933594,
|
|
"step": 183
|
|
},
|
|
{
|
|
"epoch": 0.014153846153846154,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 0.00031558568965517244,
|
|
"loss": 2.440843105316162,
|
|
"step": 184
|
|
},
|
|
{
|
|
"epoch": 0.01423076923076923,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 0.0003173096551724138,
|
|
"loss": 2.464001178741455,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 0.014307692307692308,
|
|
"grad_norm": 1.6953125,
|
|
"learning_rate": 0.0003190336206896552,
|
|
"loss": 2.490817070007324,
|
|
"step": 186
|
|
},
|
|
{
|
|
"epoch": 0.014384615384615385,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 0.0003207575862068966,
|
|
"loss": 2.5020222663879395,
|
|
"step": 187
|
|
},
|
|
{
|
|
"epoch": 0.014461538461538461,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.00032248155172413795,
|
|
"loss": 2.4789650440216064,
|
|
"step": 188
|
|
},
|
|
{
|
|
"epoch": 0.014538461538461538,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.00032420551724137933,
|
|
"loss": 2.4255244731903076,
|
|
"step": 189
|
|
},
|
|
{
|
|
"epoch": 0.014615384615384615,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.0003259294827586207,
|
|
"loss": 2.4255170822143555,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.014692307692307692,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 0.0003276534482758621,
|
|
"loss": 2.5182228088378906,
|
|
"step": 191
|
|
},
|
|
{
|
|
"epoch": 0.014769230769230769,
|
|
"grad_norm": 1.8203125,
|
|
"learning_rate": 0.0003293774137931034,
|
|
"loss": 2.506859540939331,
|
|
"step": 192
|
|
},
|
|
{
|
|
"epoch": 0.014846153846153846,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.00033110137931034484,
|
|
"loss": 2.4471051692962646,
|
|
"step": 193
|
|
},
|
|
{
|
|
"epoch": 0.014923076923076923,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0003328253448275862,
|
|
"loss": 2.4206173419952393,
|
|
"step": 194
|
|
},
|
|
{
|
|
"epoch": 0.015,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.0003345493103448276,
|
|
"loss": 2.4293782711029053,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 0.015076923076923076,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 0.00033627327586206897,
|
|
"loss": 2.5011918544769287,
|
|
"step": 196
|
|
},
|
|
{
|
|
"epoch": 0.015153846153846153,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 0.00033799724137931035,
|
|
"loss": 2.468625545501709,
|
|
"step": 197
|
|
},
|
|
{
|
|
"epoch": 0.01523076923076923,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.0003397212068965517,
|
|
"loss": 2.4359755516052246,
|
|
"step": 198
|
|
},
|
|
{
|
|
"epoch": 0.015307692307692307,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.0003414451724137931,
|
|
"loss": 2.4055593013763428,
|
|
"step": 199
|
|
},
|
|
{
|
|
"epoch": 0.015384615384615385,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.0003431691379310345,
|
|
"loss": 2.427013635635376,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.015461538461538462,
|
|
"grad_norm": 2.5625,
|
|
"learning_rate": 0.00034489310344827586,
|
|
"loss": 2.561696767807007,
|
|
"step": 201
|
|
},
|
|
{
|
|
"epoch": 0.01553846153846154,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 0.00034661706896551724,
|
|
"loss": 2.454472541809082,
|
|
"step": 202
|
|
},
|
|
{
|
|
"epoch": 0.015615384615384616,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00034834103448275867,
|
|
"loss": 2.4774181842803955,
|
|
"step": 203
|
|
},
|
|
{
|
|
"epoch": 0.015692307692307693,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.00035006500000000004,
|
|
"loss": 2.411586284637451,
|
|
"step": 204
|
|
},
|
|
{
|
|
"epoch": 0.015769230769230768,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.0003517889655172414,
|
|
"loss": 2.4032504558563232,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 0.015846153846153847,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 0.00035351293103448274,
|
|
"loss": 2.5443458557128906,
|
|
"step": 206
|
|
},
|
|
{
|
|
"epoch": 0.015923076923076922,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0003552368965517242,
|
|
"loss": 2.4499576091766357,
|
|
"step": 207
|
|
},
|
|
{
|
|
"epoch": 0.016,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0003569608620689655,
|
|
"loss": 2.4166741371154785,
|
|
"step": 208
|
|
},
|
|
{
|
|
"epoch": 0.016076923076923075,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00035868482758620693,
|
|
"loss": 2.420236110687256,
|
|
"step": 209
|
|
},
|
|
{
|
|
"epoch": 0.016153846153846154,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.00036040879310344825,
|
|
"loss": 2.426898241043091,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.01623076923076923,
|
|
"grad_norm": 2.484375,
|
|
"learning_rate": 0.00036213275862068963,
|
|
"loss": 2.5344972610473633,
|
|
"step": 211
|
|
},
|
|
{
|
|
"epoch": 0.016307692307692308,
|
|
"grad_norm": 1.46875,
|
|
"learning_rate": 0.000363856724137931,
|
|
"loss": 2.4651527404785156,
|
|
"step": 212
|
|
},
|
|
{
|
|
"epoch": 0.016384615384615383,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0003655806896551724,
|
|
"loss": 2.4288556575775146,
|
|
"step": 213
|
|
},
|
|
{
|
|
"epoch": 0.01646153846153846,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0003673046551724138,
|
|
"loss": 2.357262134552002,
|
|
"step": 214
|
|
},
|
|
{
|
|
"epoch": 0.01653846153846154,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0003690286206896552,
|
|
"loss": 2.388572931289673,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 0.016615384615384615,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 0.0003707525862068966,
|
|
"loss": 2.558690071105957,
|
|
"step": 216
|
|
},
|
|
{
|
|
"epoch": 0.016692307692307694,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00037247655172413795,
|
|
"loss": 2.4204773902893066,
|
|
"step": 217
|
|
},
|
|
{
|
|
"epoch": 0.01676923076923077,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.0003742005172413794,
|
|
"loss": 2.4480881690979004,
|
|
"step": 218
|
|
},
|
|
{
|
|
"epoch": 0.016846153846153848,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0003759244827586207,
|
|
"loss": 2.3939409255981445,
|
|
"step": 219
|
|
},
|
|
{
|
|
"epoch": 0.016923076923076923,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0003776484482758621,
|
|
"loss": 2.446916341781616,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.017,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 0.00037937241379310346,
|
|
"loss": 2.520115852355957,
|
|
"step": 221
|
|
},
|
|
{
|
|
"epoch": 0.017076923076923076,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00038109637931034484,
|
|
"loss": 2.4025721549987793,
|
|
"step": 222
|
|
},
|
|
{
|
|
"epoch": 0.017153846153846155,
|
|
"grad_norm": 0.90625,
|
|
"learning_rate": 0.00038282034482758616,
|
|
"loss": 2.3889217376708984,
|
|
"step": 223
|
|
},
|
|
{
|
|
"epoch": 0.01723076923076923,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00038454431034482765,
|
|
"loss": 2.405648708343506,
|
|
"step": 224
|
|
},
|
|
{
|
|
"epoch": 0.01730769230769231,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00038626827586206897,
|
|
"loss": 2.416717529296875,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.017384615384615384,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 0.0003879922413793104,
|
|
"loss": 2.5468807220458984,
|
|
"step": 226
|
|
},
|
|
{
|
|
"epoch": 0.017461538461538462,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.0003897162068965517,
|
|
"loss": 2.395967960357666,
|
|
"step": 227
|
|
},
|
|
{
|
|
"epoch": 0.017538461538461537,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00039144017241379316,
|
|
"loss": 2.4252467155456543,
|
|
"step": 228
|
|
},
|
|
{
|
|
"epoch": 0.017615384615384616,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0003931641379310345,
|
|
"loss": 2.406794786453247,
|
|
"step": 229
|
|
},
|
|
{
|
|
"epoch": 0.01769230769230769,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.0003948881034482759,
|
|
"loss": 2.4611074924468994,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.01776923076923077,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 0.00039661206896551723,
|
|
"loss": 2.4630239009857178,
|
|
"step": 231
|
|
},
|
|
{
|
|
"epoch": 0.017846153846153845,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0003983360344827586,
|
|
"loss": 2.4144043922424316,
|
|
"step": 232
|
|
},
|
|
{
|
|
"epoch": 0.017923076923076924,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00040006,
|
|
"loss": 2.460479974746704,
|
|
"step": 233
|
|
},
|
|
{
|
|
"epoch": 0.018,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0004017839655172414,
|
|
"loss": 2.400005340576172,
|
|
"step": 234
|
|
},
|
|
{
|
|
"epoch": 0.018076923076923077,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00040350793103448274,
|
|
"loss": 2.437112331390381,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 0.018153846153846152,
|
|
"grad_norm": 1.6796875,
|
|
"learning_rate": 0.0004052318965517242,
|
|
"loss": 2.4266703128814697,
|
|
"step": 236
|
|
},
|
|
{
|
|
"epoch": 0.01823076923076923,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0004069558620689655,
|
|
"loss": 2.4272959232330322,
|
|
"step": 237
|
|
},
|
|
{
|
|
"epoch": 0.018307692307692306,
|
|
"grad_norm": 0.94140625,
|
|
"learning_rate": 0.00040867982758620693,
|
|
"loss": 2.36838960647583,
|
|
"step": 238
|
|
},
|
|
{
|
|
"epoch": 0.018384615384615385,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.00041040379310344825,
|
|
"loss": 2.409811019897461,
|
|
"step": 239
|
|
},
|
|
{
|
|
"epoch": 0.018461538461538463,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.0004121277586206897,
|
|
"loss": 2.4161369800567627,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.01853846153846154,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 0.00041385172413793106,
|
|
"loss": 2.485699415206909,
|
|
"step": 241
|
|
},
|
|
{
|
|
"epoch": 0.018615384615384617,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.00041557568965517244,
|
|
"loss": 2.4011106491088867,
|
|
"step": 242
|
|
},
|
|
{
|
|
"epoch": 0.018692307692307692,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 0.0004172996551724138,
|
|
"loss": 2.396042823791504,
|
|
"step": 243
|
|
},
|
|
{
|
|
"epoch": 0.01876923076923077,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.00041902362068965514,
|
|
"loss": 2.3792614936828613,
|
|
"step": 244
|
|
},
|
|
{
|
|
"epoch": 0.018846153846153846,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0004207475862068966,
|
|
"loss": 2.4582300186157227,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 0.018923076923076924,
|
|
"grad_norm": 1.5234375,
|
|
"learning_rate": 0.00042247155172413795,
|
|
"loss": 2.4576497077941895,
|
|
"step": 246
|
|
},
|
|
{
|
|
"epoch": 0.019,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0004241955172413794,
|
|
"loss": 2.4750797748565674,
|
|
"step": 247
|
|
},
|
|
{
|
|
"epoch": 0.019076923076923078,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.0004259194827586207,
|
|
"loss": 2.4101741313934326,
|
|
"step": 248
|
|
},
|
|
{
|
|
"epoch": 0.019153846153846153,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00042764344827586213,
|
|
"loss": 2.379774808883667,
|
|
"step": 249
|
|
},
|
|
{
|
|
"epoch": 0.019230769230769232,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00042936741379310346,
|
|
"loss": 2.421823740005493,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.019307692307692307,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.00043109137931034484,
|
|
"loss": 2.41143536567688,
|
|
"step": 251
|
|
},
|
|
{
|
|
"epoch": 0.019384615384615386,
|
|
"grad_norm": 0.82421875,
|
|
"learning_rate": 0.0004328153448275862,
|
|
"loss": 2.3944203853607178,
|
|
"step": 252
|
|
},
|
|
{
|
|
"epoch": 0.01946153846153846,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 0.0004345393103448276,
|
|
"loss": 2.407639503479004,
|
|
"step": 253
|
|
},
|
|
{
|
|
"epoch": 0.01953846153846154,
|
|
"grad_norm": 0.80859375,
|
|
"learning_rate": 0.0004362632758620689,
|
|
"loss": 2.4225754737854004,
|
|
"step": 254
|
|
},
|
|
{
|
|
"epoch": 0.019615384615384614,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.0004379872413793104,
|
|
"loss": 2.443262815475464,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 0.019692307692307693,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0004397112068965517,
|
|
"loss": 2.467907190322876,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 0.019769230769230768,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 0.00044143517241379315,
|
|
"loss": 2.405221462249756,
|
|
"step": 257
|
|
},
|
|
{
|
|
"epoch": 0.019846153846153847,
|
|
"grad_norm": 0.734375,
|
|
"learning_rate": 0.0004431591379310345,
|
|
"loss": 2.386744737625122,
|
|
"step": 258
|
|
},
|
|
{
|
|
"epoch": 0.019923076923076922,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.0004448831034482759,
|
|
"loss": 2.3745787143707275,
|
|
"step": 259
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.00044660706896551723,
|
|
"loss": 2.4517741203308105,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.020076923076923076,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.00044833103448275866,
|
|
"loss": 2.423065423965454,
|
|
"step": 261
|
|
},
|
|
{
|
|
"epoch": 0.020153846153846154,
|
|
"grad_norm": 0.91796875,
|
|
"learning_rate": 0.000450055,
|
|
"loss": 2.4110941886901855,
|
|
"step": 262
|
|
},
|
|
{
|
|
"epoch": 0.02023076923076923,
|
|
"grad_norm": 0.65625,
|
|
"learning_rate": 0.00045177896551724136,
|
|
"loss": 2.360673666000366,
|
|
"step": 263
|
|
},
|
|
{
|
|
"epoch": 0.020307692307692308,
|
|
"grad_norm": 0.875,
|
|
"learning_rate": 0.00045350293103448274,
|
|
"loss": 2.397526264190674,
|
|
"step": 264
|
|
},
|
|
{
|
|
"epoch": 0.020384615384615383,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.0004552268965517241,
|
|
"loss": 2.4497227668762207,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 0.02046153846153846,
|
|
"grad_norm": 0.8203125,
|
|
"learning_rate": 0.0004569508620689655,
|
|
"loss": 2.4259109497070312,
|
|
"step": 266
|
|
},
|
|
{
|
|
"epoch": 0.02053846153846154,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.00045867482758620693,
|
|
"loss": 2.4793429374694824,
|
|
"step": 267
|
|
},
|
|
{
|
|
"epoch": 0.020615384615384615,
|
|
"grad_norm": 0.9140625,
|
|
"learning_rate": 0.00046039879310344836,
|
|
"loss": 2.4151601791381836,
|
|
"step": 268
|
|
},
|
|
{
|
|
"epoch": 0.020692307692307694,
|
|
"grad_norm": 0.98046875,
|
|
"learning_rate": 0.0004621227586206897,
|
|
"loss": 2.3758158683776855,
|
|
"step": 269
|
|
},
|
|
{
|
|
"epoch": 0.02076923076923077,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.0004638467241379311,
|
|
"loss": 2.4561290740966797,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.020846153846153848,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.00046557068965517244,
|
|
"loss": 2.393204927444458,
|
|
"step": 271
|
|
},
|
|
{
|
|
"epoch": 0.020923076923076923,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.0004672946551724138,
|
|
"loss": 2.4570980072021484,
|
|
"step": 272
|
|
},
|
|
{
|
|
"epoch": 0.021,
|
|
"grad_norm": 0.78515625,
|
|
"learning_rate": 0.0004690186206896552,
|
|
"loss": 2.3872392177581787,
|
|
"step": 273
|
|
},
|
|
{
|
|
"epoch": 0.021076923076923076,
|
|
"grad_norm": 0.703125,
|
|
"learning_rate": 0.00047074258620689657,
|
|
"loss": 2.390110969543457,
|
|
"step": 274
|
|
},
|
|
{
|
|
"epoch": 0.021153846153846155,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0004724665517241379,
|
|
"loss": 2.477142095565796,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.02123076923076923,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004741905172413794,
|
|
"loss": 2.4440720081329346,
|
|
"step": 276
|
|
},
|
|
{
|
|
"epoch": 0.02130769230769231,
|
|
"grad_norm": 0.828125,
|
|
"learning_rate": 0.0004759144827586207,
|
|
"loss": 2.4130265712738037,
|
|
"step": 277
|
|
},
|
|
{
|
|
"epoch": 0.021384615384615384,
|
|
"grad_norm": 0.73828125,
|
|
"learning_rate": 0.00047763844827586213,
|
|
"loss": 2.401531219482422,
|
|
"step": 278
|
|
},
|
|
{
|
|
"epoch": 0.021461538461538462,
|
|
"grad_norm": 0.76953125,
|
|
"learning_rate": 0.00047936241379310346,
|
|
"loss": 2.3781306743621826,
|
|
"step": 279
|
|
},
|
|
{
|
|
"epoch": 0.021538461538461538,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0004810863793103449,
|
|
"loss": 2.4172561168670654,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.021615384615384616,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.0004828103448275862,
|
|
"loss": 2.4107325077056885,
|
|
"step": 281
|
|
},
|
|
{
|
|
"epoch": 0.02169230769230769,
|
|
"grad_norm": 0.81640625,
|
|
"learning_rate": 0.0004845343103448276,
|
|
"loss": 2.4322867393493652,
|
|
"step": 282
|
|
},
|
|
{
|
|
"epoch": 0.02176923076923077,
|
|
"grad_norm": 0.765625,
|
|
"learning_rate": 0.00048625827586206897,
|
|
"loss": 2.4008307456970215,
|
|
"step": 283
|
|
},
|
|
{
|
|
"epoch": 0.021846153846153845,
|
|
"grad_norm": 0.77734375,
|
|
"learning_rate": 0.00048798224137931034,
|
|
"loss": 2.3974852561950684,
|
|
"step": 284
|
|
},
|
|
{
|
|
"epoch": 0.021923076923076924,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0004897062068965517,
|
|
"loss": 2.423407554626465,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 0.022,
|
|
"grad_norm": 0.8828125,
|
|
"learning_rate": 0.000491430172413793,
|
|
"loss": 2.38500714302063,
|
|
"step": 286
|
|
},
|
|
{
|
|
"epoch": 0.022076923076923077,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 0.0004931541379310344,
|
|
"loss": 2.383406639099121,
|
|
"step": 287
|
|
},
|
|
{
|
|
"epoch": 0.022153846153846152,
|
|
"grad_norm": 0.67578125,
|
|
"learning_rate": 0.0004948781034482758,
|
|
"loss": 2.405846357345581,
|
|
"step": 288
|
|
},
|
|
{
|
|
"epoch": 0.02223076923076923,
|
|
"grad_norm": 0.69921875,
|
|
"learning_rate": 0.0004966020689655172,
|
|
"loss": 2.380526542663574,
|
|
"step": 289
|
|
},
|
|
{
|
|
"epoch": 0.022307692307692306,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.0004983260344827586,
|
|
"loss": 2.4697611331939697,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.022384615384615385,
|
|
"grad_norm": 0.9453125,
|
|
"learning_rate": 0.0005000499999999999,
|
|
"loss": 2.390803575515747,
|
|
"step": 291
|
|
},
|
|
{
|
|
"epoch": 0.02246153846153846,
|
|
"grad_norm": 0.7578125,
|
|
"learning_rate": 0.0005017739655172414,
|
|
"loss": 2.3932652473449707,
|
|
"step": 292
|
|
},
|
|
{
|
|
"epoch": 0.02253846153846154,
|
|
"grad_norm": 0.61328125,
|
|
"learning_rate": 0.0005034979310344827,
|
|
"loss": 2.3618335723876953,
|
|
"step": 293
|
|
},
|
|
{
|
|
"epoch": 0.022615384615384617,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.0005052218965517242,
|
|
"loss": 2.413799285888672,
|
|
"step": 294
|
|
},
|
|
{
|
|
"epoch": 0.022692307692307692,
|
|
"grad_norm": 0.89453125,
|
|
"learning_rate": 0.0005069458620689655,
|
|
"loss": 2.371918201446533,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 0.02276923076923077,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0005086698275862069,
|
|
"loss": 2.4191057682037354,
|
|
"step": 296
|
|
},
|
|
{
|
|
"epoch": 0.022846153846153846,
|
|
"grad_norm": 0.48046875,
|
|
"learning_rate": 0.0005103937931034482,
|
|
"loss": 2.3885951042175293,
|
|
"step": 297
|
|
},
|
|
{
|
|
"epoch": 0.022923076923076924,
|
|
"grad_norm": 0.58984375,
|
|
"learning_rate": 0.0005121177586206897,
|
|
"loss": 2.3618862628936768,
|
|
"step": 298
|
|
},
|
|
{
|
|
"epoch": 0.023,
|
|
"grad_norm": 0.6328125,
|
|
"learning_rate": 0.000513841724137931,
|
|
"loss": 2.346761465072632,
|
|
"step": 299
|
|
},
|
|
{
|
|
"epoch": 0.023076923076923078,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.0005155656896551723,
|
|
"loss": 2.4326131343841553,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.023153846153846153,
|
|
"grad_norm": 0.953125,
|
|
"learning_rate": 0.0005172896551724137,
|
|
"loss": 2.4212794303894043,
|
|
"step": 301
|
|
},
|
|
{
|
|
"epoch": 0.023230769230769232,
|
|
"grad_norm": 0.59765625,
|
|
"learning_rate": 0.0005190136206896551,
|
|
"loss": 2.404571294784546,
|
|
"step": 302
|
|
},
|
|
{
|
|
"epoch": 0.023307692307692307,
|
|
"grad_norm": 0.46875,
|
|
"learning_rate": 0.0005207375862068966,
|
|
"loss": 2.3497302532196045,
|
|
"step": 303
|
|
},
|
|
{
|
|
"epoch": 0.023384615384615386,
|
|
"grad_norm": 0.6640625,
|
|
"learning_rate": 0.0005224615517241378,
|
|
"loss": 2.3841183185577393,
|
|
"step": 304
|
|
},
|
|
{
|
|
"epoch": 0.02346153846153846,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0005241855172413793,
|
|
"loss": 2.446011781692505,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 0.02353846153846154,
|
|
"grad_norm": 0.9375,
|
|
"learning_rate": 0.0005259094827586206,
|
|
"loss": 2.3934285640716553,
|
|
"step": 306
|
|
},
|
|
{
|
|
"epoch": 0.023615384615384614,
|
|
"grad_norm": 0.5859375,
|
|
"learning_rate": 0.0005276334482758621,
|
|
"loss": 2.413163185119629,
|
|
"step": 307
|
|
},
|
|
{
|
|
"epoch": 0.023692307692307693,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 0.0005293574137931033,
|
|
"loss": 2.374394655227661,
|
|
"step": 308
|
|
},
|
|
{
|
|
"epoch": 0.023769230769230768,
|
|
"grad_norm": 0.75390625,
|
|
"learning_rate": 0.0005310813793103448,
|
|
"loss": 2.3822245597839355,
|
|
"step": 309
|
|
},
|
|
{
|
|
"epoch": 0.023846153846153847,
|
|
"grad_norm": 0.68359375,
|
|
"learning_rate": 0.0005328053448275861,
|
|
"loss": 2.4050559997558594,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.023923076923076922,
|
|
"grad_norm": 0.7578125,
|
|
"learning_rate": 0.0005345293103448276,
|
|
"loss": 2.387981653213501,
|
|
"step": 311
|
|
},
|
|
{
|
|
"epoch": 0.024,
|
|
"grad_norm": 0.6640625,
|
|
"learning_rate": 0.000536253275862069,
|
|
"loss": 2.3852033615112305,
|
|
"step": 312
|
|
},
|
|
{
|
|
"epoch": 0.024076923076923076,
|
|
"grad_norm": 0.62890625,
|
|
"learning_rate": 0.0005379772413793103,
|
|
"loss": 2.3823788166046143,
|
|
"step": 313
|
|
},
|
|
{
|
|
"epoch": 0.024153846153846154,
|
|
"grad_norm": 0.5625,
|
|
"learning_rate": 0.0005397012068965517,
|
|
"loss": 2.360262870788574,
|
|
"step": 314
|
|
},
|
|
{
|
|
"epoch": 0.02423076923076923,
|
|
"grad_norm": 0.94921875,
|
|
"learning_rate": 0.0005414251724137931,
|
|
"loss": 2.4232230186462402,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 0.024307692307692308,
|
|
"grad_norm": 0.55859375,
|
|
"learning_rate": 0.0005431491379310345,
|
|
"loss": 2.398618459701538,
|
|
"step": 316
|
|
},
|
|
{
|
|
"epoch": 0.024384615384615383,
|
|
"grad_norm": 0.61328125,
|
|
"learning_rate": 0.0005448731034482759,
|
|
"loss": 2.402404308319092,
|
|
"step": 317
|
|
},
|
|
{
|
|
"epoch": 0.02446153846153846,
|
|
"grad_norm": 0.625,
|
|
"learning_rate": 0.0005465970689655172,
|
|
"loss": 2.3911709785461426,
|
|
"step": 318
|
|
},
|
|
{
|
|
"epoch": 0.024538461538461537,
|
|
"grad_norm": 0.46875,
|
|
"learning_rate": 0.0005483210344827585,
|
|
"loss": 2.3570752143859863,
|
|
"step": 319
|
|
},
|
|
{
|
|
"epoch": 0.024615384615384615,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000550045,
|
|
"loss": 2.498175859451294,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.024692307692307694,
|
|
"grad_norm": 0.6796875,
|
|
"learning_rate": 0.0005517689655172413,
|
|
"loss": 2.399850368499756,
|
|
"step": 321
|
|
},
|
|
{
|
|
"epoch": 0.02476923076923077,
|
|
"grad_norm": 0.380859375,
|
|
"learning_rate": 0.0005534929310344827,
|
|
"loss": 2.3354363441467285,
|
|
"step": 322
|
|
},
|
|
{
|
|
"epoch": 0.024846153846153848,
|
|
"grad_norm": 0.451171875,
|
|
"learning_rate": 0.0005552168965517241,
|
|
"loss": 2.397901773452759,
|
|
"step": 323
|
|
},
|
|
{
|
|
"epoch": 0.024923076923076923,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 0.0005569408620689654,
|
|
"loss": 2.3905577659606934,
|
|
"step": 324
|
|
},
|
|
{
|
|
"epoch": 0.025,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0005586648275862069,
|
|
"loss": 2.4413623809814453,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 0.025076923076923077,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 0.0005603887931034482,
|
|
"loss": 2.371628761291504,
|
|
"step": 326
|
|
},
|
|
{
|
|
"epoch": 0.025153846153846155,
|
|
"grad_norm": 0.380859375,
|
|
"learning_rate": 0.0005621127586206896,
|
|
"loss": 2.345428466796875,
|
|
"step": 327
|
|
},
|
|
{
|
|
"epoch": 0.02523076923076923,
|
|
"grad_norm": 0.48046875,
|
|
"learning_rate": 0.000563836724137931,
|
|
"loss": 2.341475486755371,
|
|
"step": 328
|
|
},
|
|
{
|
|
"epoch": 0.02530769230769231,
|
|
"grad_norm": 0.453125,
|
|
"learning_rate": 0.0005655606896551724,
|
|
"loss": 2.3648860454559326,
|
|
"step": 329
|
|
},
|
|
{
|
|
"epoch": 0.025384615384615384,
|
|
"grad_norm": 0.98828125,
|
|
"learning_rate": 0.0005672846551724138,
|
|
"loss": 2.4247031211853027,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.025461538461538463,
|
|
"grad_norm": 0.3984375,
|
|
"learning_rate": 0.0005690086206896551,
|
|
"loss": 2.3547112941741943,
|
|
"step": 331
|
|
},
|
|
{
|
|
"epoch": 0.025538461538461538,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 0.0005707325862068965,
|
|
"loss": 2.372105121612549,
|
|
"step": 332
|
|
},
|
|
{
|
|
"epoch": 0.025615384615384616,
|
|
"grad_norm": 0.6328125,
|
|
"learning_rate": 0.0005724565517241379,
|
|
"loss": 2.3514902591705322,
|
|
"step": 333
|
|
},
|
|
{
|
|
"epoch": 0.02569230769230769,
|
|
"grad_norm": 0.6015625,
|
|
"learning_rate": 0.0005741805172413795,
|
|
"loss": 2.368028402328491,
|
|
"step": 334
|
|
},
|
|
{
|
|
"epoch": 0.02576923076923077,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0005759044827586206,
|
|
"loss": 2.4327218532562256,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 0.025846153846153845,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 0.0005776284482758621,
|
|
"loss": 2.4009580612182617,
|
|
"step": 336
|
|
},
|
|
{
|
|
"epoch": 0.025923076923076924,
|
|
"grad_norm": 0.33984375,
|
|
"learning_rate": 0.0005793524137931034,
|
|
"loss": 2.3773748874664307,
|
|
"step": 337
|
|
},
|
|
{
|
|
"epoch": 0.026,
|
|
"grad_norm": 0.4296875,
|
|
"learning_rate": 0.0005810763793103449,
|
|
"loss": 2.3357174396514893,
|
|
"step": 338
|
|
},
|
|
{
|
|
"epoch": 0.026076923076923077,
|
|
"grad_norm": 0.3984375,
|
|
"learning_rate": 0.0005828003448275862,
|
|
"loss": 2.3467323780059814,
|
|
"step": 339
|
|
},
|
|
{
|
|
"epoch": 0.026153846153846153,
|
|
"grad_norm": 0.84765625,
|
|
"learning_rate": 0.0005845243103448275,
|
|
"loss": 2.422316789627075,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.02623076923076923,
|
|
"grad_norm": 0.388671875,
|
|
"learning_rate": 0.0005862482758620689,
|
|
"loss": 2.3779399394989014,
|
|
"step": 341
|
|
},
|
|
{
|
|
"epoch": 0.026307692307692306,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 0.0005879722413793103,
|
|
"loss": 2.40360164642334,
|
|
"step": 342
|
|
},
|
|
{
|
|
"epoch": 0.026384615384615385,
|
|
"grad_norm": 0.46484375,
|
|
"learning_rate": 0.0005896962068965517,
|
|
"loss": 2.371128797531128,
|
|
"step": 343
|
|
},
|
|
{
|
|
"epoch": 0.02646153846153846,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 0.000591420172413793,
|
|
"loss": 2.3675079345703125,
|
|
"step": 344
|
|
},
|
|
{
|
|
"epoch": 0.02653846153846154,
|
|
"grad_norm": 0.87109375,
|
|
"learning_rate": 0.0005931441379310344,
|
|
"loss": 2.402824878692627,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 0.026615384615384614,
|
|
"grad_norm": 0.384765625,
|
|
"learning_rate": 0.0005948681034482758,
|
|
"loss": 2.4022397994995117,
|
|
"step": 346
|
|
},
|
|
{
|
|
"epoch": 0.026692307692307692,
|
|
"grad_norm": 0.51953125,
|
|
"learning_rate": 0.0005965920689655172,
|
|
"loss": 2.410256862640381,
|
|
"step": 347
|
|
},
|
|
{
|
|
"epoch": 0.02676923076923077,
|
|
"grad_norm": 0.455078125,
|
|
"learning_rate": 0.0005983160344827585,
|
|
"loss": 2.363457202911377,
|
|
"step": 348
|
|
},
|
|
{
|
|
"epoch": 0.026846153846153846,
|
|
"grad_norm": 0.345703125,
|
|
"learning_rate": 0.00060004,
|
|
"loss": 2.338977575302124,
|
|
"step": 349
|
|
},
|
|
{
|
|
"epoch": 0.026923076923076925,
|
|
"grad_norm": 0.73046875,
|
|
"learning_rate": 0.0006017639655172413,
|
|
"loss": 2.3813695907592773,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.027,
|
|
"grad_norm": 0.703125,
|
|
"learning_rate": 0.0006034879310344828,
|
|
"loss": 2.387315034866333,
|
|
"step": 351
|
|
},
|
|
{
|
|
"epoch": 0.02707692307692308,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 0.0006052118965517241,
|
|
"loss": 2.3980159759521484,
|
|
"step": 352
|
|
},
|
|
{
|
|
"epoch": 0.027153846153846153,
|
|
"grad_norm": 0.4375,
|
|
"learning_rate": 0.0006069358620689655,
|
|
"loss": 2.412670612335205,
|
|
"step": 353
|
|
},
|
|
{
|
|
"epoch": 0.027230769230769232,
|
|
"grad_norm": 0.2333984375,
|
|
"learning_rate": 0.0006086598275862069,
|
|
"loss": 2.344902992248535,
|
|
"step": 354
|
|
},
|
|
{
|
|
"epoch": 0.027307692307692307,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 0.0006103837931034483,
|
|
"loss": 2.386260747909546,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 0.027384615384615386,
|
|
"grad_norm": 0.61328125,
|
|
"learning_rate": 0.0006121077586206897,
|
|
"loss": 2.3970417976379395,
|
|
"step": 356
|
|
},
|
|
{
|
|
"epoch": 0.02746153846153846,
|
|
"grad_norm": 0.486328125,
|
|
"learning_rate": 0.000613831724137931,
|
|
"loss": 2.3776445388793945,
|
|
"step": 357
|
|
},
|
|
{
|
|
"epoch": 0.02753846153846154,
|
|
"grad_norm": 0.4140625,
|
|
"learning_rate": 0.0006155556896551724,
|
|
"loss": 2.3585381507873535,
|
|
"step": 358
|
|
},
|
|
{
|
|
"epoch": 0.027615384615384615,
|
|
"grad_norm": 0.392578125,
|
|
"learning_rate": 0.0006172796551724137,
|
|
"loss": 2.354534864425659,
|
|
"step": 359
|
|
},
|
|
{
|
|
"epoch": 0.027692307692307693,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 0.0006190036206896552,
|
|
"loss": 2.375436544418335,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.02776923076923077,
|
|
"grad_norm": 0.8671875,
|
|
"learning_rate": 0.0006207275862068966,
|
|
"loss": 2.421811819076538,
|
|
"step": 361
|
|
},
|
|
{
|
|
"epoch": 0.027846153846153847,
|
|
"grad_norm": 0.380859375,
|
|
"learning_rate": 0.0006224515517241378,
|
|
"loss": 2.3739116191864014,
|
|
"step": 362
|
|
},
|
|
{
|
|
"epoch": 0.027923076923076922,
|
|
"grad_norm": 0.41796875,
|
|
"learning_rate": 0.0006241755172413792,
|
|
"loss": 2.3735122680664062,
|
|
"step": 363
|
|
},
|
|
{
|
|
"epoch": 0.028,
|
|
"grad_norm": 0.45703125,
|
|
"learning_rate": 0.0006258994827586206,
|
|
"loss": 2.332568407058716,
|
|
"step": 364
|
|
},
|
|
{
|
|
"epoch": 0.028076923076923076,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 0.0006276234482758621,
|
|
"loss": 2.3882017135620117,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 0.028153846153846154,
|
|
"grad_norm": 0.498046875,
|
|
"learning_rate": 0.0006293474137931033,
|
|
"loss": 2.407604932785034,
|
|
"step": 366
|
|
},
|
|
{
|
|
"epoch": 0.02823076923076923,
|
|
"grad_norm": 0.353515625,
|
|
"learning_rate": 0.0006310713793103448,
|
|
"loss": 2.3894269466400146,
|
|
"step": 367
|
|
},
|
|
{
|
|
"epoch": 0.028307692307692308,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 0.0006327953448275861,
|
|
"loss": 2.3749516010284424,
|
|
"step": 368
|
|
},
|
|
{
|
|
"epoch": 0.028384615384615383,
|
|
"grad_norm": 0.4375,
|
|
"learning_rate": 0.0006345193103448276,
|
|
"loss": 2.358680486679077,
|
|
"step": 369
|
|
},
|
|
{
|
|
"epoch": 0.02846153846153846,
|
|
"grad_norm": 0.294921875,
|
|
"learning_rate": 0.0006362432758620689,
|
|
"loss": 2.3393640518188477,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.028538461538461537,
|
|
"grad_norm": 0.984375,
|
|
"learning_rate": 0.0006379672413793104,
|
|
"loss": 2.4257075786590576,
|
|
"step": 371
|
|
},
|
|
{
|
|
"epoch": 0.028615384615384615,
|
|
"grad_norm": 0.2333984375,
|
|
"learning_rate": 0.0006396912068965516,
|
|
"loss": 2.320903778076172,
|
|
"step": 372
|
|
},
|
|
{
|
|
"epoch": 0.028692307692307694,
|
|
"grad_norm": 0.31640625,
|
|
"learning_rate": 0.0006414151724137931,
|
|
"loss": 2.350449562072754,
|
|
"step": 373
|
|
},
|
|
{
|
|
"epoch": 0.02876923076923077,
|
|
"grad_norm": 0.330078125,
|
|
"learning_rate": 0.0006431391379310346,
|
|
"loss": 2.3557543754577637,
|
|
"step": 374
|
|
},
|
|
{
|
|
"epoch": 0.028846153846153848,
|
|
"grad_norm": 0.451171875,
|
|
"learning_rate": 0.0006448631034482758,
|
|
"loss": 2.3453969955444336,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 0.028923076923076923,
|
|
"grad_norm": 0.65625,
|
|
"learning_rate": 0.0006465870689655173,
|
|
"loss": 2.4152159690856934,
|
|
"step": 376
|
|
},
|
|
{
|
|
"epoch": 0.029,
|
|
"grad_norm": 0.40625,
|
|
"learning_rate": 0.0006483110344827586,
|
|
"loss": 2.364250421524048,
|
|
"step": 377
|
|
},
|
|
{
|
|
"epoch": 0.029076923076923077,
|
|
"grad_norm": 0.29296875,
|
|
"learning_rate": 0.0006500350000000001,
|
|
"loss": 2.3461904525756836,
|
|
"step": 378
|
|
},
|
|
{
|
|
"epoch": 0.029153846153846155,
|
|
"grad_norm": 0.435546875,
|
|
"learning_rate": 0.0006517589655172414,
|
|
"loss": 2.3484559059143066,
|
|
"step": 379
|
|
},
|
|
{
|
|
"epoch": 0.02923076923076923,
|
|
"grad_norm": 0.3203125,
|
|
"learning_rate": 0.0006534829310344827,
|
|
"loss": 2.3288581371307373,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.02930769230769231,
|
|
"grad_norm": 0.671875,
|
|
"learning_rate": 0.0006552068965517241,
|
|
"loss": 2.3826522827148438,
|
|
"step": 381
|
|
},
|
|
{
|
|
"epoch": 0.029384615384615384,
|
|
"grad_norm": 0.36328125,
|
|
"learning_rate": 0.0006569308620689655,
|
|
"loss": 2.3861920833587646,
|
|
"step": 382
|
|
},
|
|
{
|
|
"epoch": 0.029461538461538463,
|
|
"grad_norm": 0.32421875,
|
|
"learning_rate": 0.0006586548275862068,
|
|
"loss": 2.3691675662994385,
|
|
"step": 383
|
|
},
|
|
{
|
|
"epoch": 0.029538461538461538,
|
|
"grad_norm": 0.44921875,
|
|
"learning_rate": 0.0006603787931034482,
|
|
"loss": 2.3411216735839844,
|
|
"step": 384
|
|
},
|
|
{
|
|
"epoch": 0.029615384615384616,
|
|
"grad_norm": 0.296875,
|
|
"learning_rate": 0.0006621027586206896,
|
|
"loss": 2.347240686416626,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 0.02969230769230769,
|
|
"grad_norm": 0.70703125,
|
|
"learning_rate": 0.000663826724137931,
|
|
"loss": 2.4139244556427,
|
|
"step": 386
|
|
},
|
|
{
|
|
"epoch": 0.02976923076923077,
|
|
"grad_norm": 0.400390625,
|
|
"learning_rate": 0.0006655506896551724,
|
|
"loss": 2.393260955810547,
|
|
"step": 387
|
|
},
|
|
{
|
|
"epoch": 0.029846153846153845,
|
|
"grad_norm": 0.333984375,
|
|
"learning_rate": 0.0006672746551724138,
|
|
"loss": 2.3918356895446777,
|
|
"step": 388
|
|
},
|
|
{
|
|
"epoch": 0.029923076923076924,
|
|
"grad_norm": 0.365234375,
|
|
"learning_rate": 0.0006689986206896551,
|
|
"loss": 2.3979084491729736,
|
|
"step": 389
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"grad_norm": 0.369140625,
|
|
"learning_rate": 0.0006707225862068965,
|
|
"loss": 2.3705573081970215,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.030076923076923077,
|
|
"grad_norm": 0.6484375,
|
|
"learning_rate": 0.0006724465517241379,
|
|
"loss": 2.3797779083251953,
|
|
"step": 391
|
|
},
|
|
{
|
|
"epoch": 0.030153846153846153,
|
|
"grad_norm": 0.38671875,
|
|
"learning_rate": 0.0006741705172413793,
|
|
"loss": 2.375246047973633,
|
|
"step": 392
|
|
},
|
|
{
|
|
"epoch": 0.03023076923076923,
|
|
"grad_norm": 0.3203125,
|
|
"learning_rate": 0.0006758944827586206,
|
|
"loss": 2.3799750804901123,
|
|
"step": 393
|
|
},
|
|
{
|
|
"epoch": 0.030307692307692306,
|
|
"grad_norm": 0.3046875,
|
|
"learning_rate": 0.000677618448275862,
|
|
"loss": 2.3571839332580566,
|
|
"step": 394
|
|
},
|
|
{
|
|
"epoch": 0.030384615384615385,
|
|
"grad_norm": 0.373046875,
|
|
"learning_rate": 0.0006793424137931034,
|
|
"loss": 2.3437602519989014,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 0.03046153846153846,
|
|
"grad_norm": 0.640625,
|
|
"learning_rate": 0.0006810663793103449,
|
|
"loss": 2.4107651710510254,
|
|
"step": 396
|
|
},
|
|
{
|
|
"epoch": 0.03053846153846154,
|
|
"grad_norm": 0.4921875,
|
|
"learning_rate": 0.0006827903448275861,
|
|
"loss": 2.362891912460327,
|
|
"step": 397
|
|
},
|
|
{
|
|
"epoch": 0.030615384615384614,
|
|
"grad_norm": 0.431640625,
|
|
"learning_rate": 0.0006845143103448276,
|
|
"loss": 2.399916887283325,
|
|
"step": 398
|
|
},
|
|
{
|
|
"epoch": 0.030692307692307692,
|
|
"grad_norm": 0.291015625,
|
|
"learning_rate": 0.0006862382758620689,
|
|
"loss": 2.3406789302825928,
|
|
"step": 399
|
|
},
|
|
{
|
|
"epoch": 0.03076923076923077,
|
|
"grad_norm": 0.3125,
|
|
"learning_rate": 0.0006879622413793104,
|
|
"loss": 2.3349015712738037,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.030846153846153846,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 0.0006896862068965517,
|
|
"loss": 2.3763208389282227,
|
|
"step": 401
|
|
},
|
|
{
|
|
"epoch": 0.030923076923076925,
|
|
"grad_norm": 0.333984375,
|
|
"learning_rate": 0.000691410172413793,
|
|
"loss": 2.380683183670044,
|
|
"step": 402
|
|
},
|
|
{
|
|
"epoch": 0.031,
|
|
"grad_norm": 0.22265625,
|
|
"learning_rate": 0.0006931341379310344,
|
|
"loss": 2.321301221847534,
|
|
"step": 403
|
|
},
|
|
{
|
|
"epoch": 0.03107692307692308,
|
|
"grad_norm": 0.28125,
|
|
"learning_rate": 0.0006948581034482758,
|
|
"loss": 2.3186697959899902,
|
|
"step": 404
|
|
},
|
|
{
|
|
"epoch": 0.031153846153846153,
|
|
"grad_norm": 0.3203125,
|
|
"learning_rate": 0.0006965820689655173,
|
|
"loss": 2.3403284549713135,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 0.031230769230769232,
|
|
"grad_norm": 0.55859375,
|
|
"learning_rate": 0.0006983060344827585,
|
|
"loss": 2.3800857067108154,
|
|
"step": 406
|
|
},
|
|
{
|
|
"epoch": 0.03130769230769231,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 0.00070003,
|
|
"loss": 2.402909755706787,
|
|
"step": 407
|
|
},
|
|
{
|
|
"epoch": 0.031384615384615386,
|
|
"grad_norm": 0.26171875,
|
|
"learning_rate": 0.0007017539655172414,
|
|
"loss": 2.3591811656951904,
|
|
"step": 408
|
|
},
|
|
{
|
|
"epoch": 0.031461538461538464,
|
|
"grad_norm": 0.2578125,
|
|
"learning_rate": 0.0007034779310344828,
|
|
"loss": 2.3548638820648193,
|
|
"step": 409
|
|
},
|
|
{
|
|
"epoch": 0.031538461538461536,
|
|
"grad_norm": 0.375,
|
|
"learning_rate": 0.0007052018965517242,
|
|
"loss": 2.375188112258911,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 0.031615384615384615,
|
|
"grad_norm": 0.56640625,
|
|
"learning_rate": 0.0007069258620689654,
|
|
"loss": 2.36104416847229,
|
|
"step": 411
|
|
},
|
|
{
|
|
"epoch": 0.03169230769230769,
|
|
"grad_norm": 0.478515625,
|
|
"learning_rate": 0.0007086498275862068,
|
|
"loss": 2.4026284217834473,
|
|
"step": 412
|
|
},
|
|
{
|
|
"epoch": 0.03176923076923077,
|
|
"grad_norm": 0.35546875,
|
|
"learning_rate": 0.0007103737931034483,
|
|
"loss": 2.377183437347412,
|
|
"step": 413
|
|
},
|
|
{
|
|
"epoch": 0.031846153846153843,
|
|
"grad_norm": 0.236328125,
|
|
"learning_rate": 0.0007120977586206897,
|
|
"loss": 2.346937417984009,
|
|
"step": 414
|
|
},
|
|
{
|
|
"epoch": 0.03192307692307692,
|
|
"grad_norm": 0.232421875,
|
|
"learning_rate": 0.0007138217241379309,
|
|
"loss": 2.333737373352051,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 0.032,
|
|
"grad_norm": 0.494140625,
|
|
"learning_rate": 0.0007155456896551724,
|
|
"loss": 2.388359546661377,
|
|
"step": 416
|
|
},
|
|
{
|
|
"epoch": 0.03207692307692308,
|
|
"grad_norm": 0.234375,
|
|
"learning_rate": 0.0007172696551724138,
|
|
"loss": 2.3490195274353027,
|
|
"step": 417
|
|
},
|
|
{
|
|
"epoch": 0.03215384615384615,
|
|
"grad_norm": 0.2890625,
|
|
"learning_rate": 0.0007189936206896552,
|
|
"loss": 2.4062962532043457,
|
|
"step": 418
|
|
},
|
|
{
|
|
"epoch": 0.03223076923076923,
|
|
"grad_norm": 0.34765625,
|
|
"learning_rate": 0.0007207175862068965,
|
|
"loss": 2.370055675506592,
|
|
"step": 419
|
|
},
|
|
{
|
|
"epoch": 0.03230769230769231,
|
|
"grad_norm": 0.244140625,
|
|
"learning_rate": 0.0007224415517241379,
|
|
"loss": 2.3560547828674316,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.03238461538461539,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 0.0007241655172413792,
|
|
"loss": 2.383244276046753,
|
|
"step": 421
|
|
},
|
|
{
|
|
"epoch": 0.03246153846153846,
|
|
"grad_norm": 0.3359375,
|
|
"learning_rate": 0.0007258894827586207,
|
|
"loss": 2.329597234725952,
|
|
"step": 422
|
|
},
|
|
{
|
|
"epoch": 0.03253846153846154,
|
|
"grad_norm": 0.232421875,
|
|
"learning_rate": 0.000727613448275862,
|
|
"loss": 2.360773801803589,
|
|
"step": 423
|
|
},
|
|
{
|
|
"epoch": 0.032615384615384616,
|
|
"grad_norm": 0.1640625,
|
|
"learning_rate": 0.0007293374137931034,
|
|
"loss": 2.3317320346832275,
|
|
"step": 424
|
|
},
|
|
{
|
|
"epoch": 0.032692307692307694,
|
|
"grad_norm": 0.1826171875,
|
|
"learning_rate": 0.0007310613793103447,
|
|
"loss": 2.2932634353637695,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 0.032769230769230766,
|
|
"grad_norm": 0.47265625,
|
|
"learning_rate": 0.0007327853448275862,
|
|
"loss": 2.3766391277313232,
|
|
"step": 426
|
|
},
|
|
{
|
|
"epoch": 0.032846153846153844,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 0.0007345093103448276,
|
|
"loss": 2.395951509475708,
|
|
"step": 427
|
|
},
|
|
{
|
|
"epoch": 0.03292307692307692,
|
|
"grad_norm": 0.28125,
|
|
"learning_rate": 0.0007362332758620688,
|
|
"loss": 2.3560235500335693,
|
|
"step": 428
|
|
},
|
|
{
|
|
"epoch": 0.033,
|
|
"grad_norm": 0.283203125,
|
|
"learning_rate": 0.0007379572413793103,
|
|
"loss": 2.365208148956299,
|
|
"step": 429
|
|
},
|
|
{
|
|
"epoch": 0.03307692307692308,
|
|
"grad_norm": 0.177734375,
|
|
"learning_rate": 0.0007396812068965517,
|
|
"loss": 2.332125425338745,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 0.03315384615384615,
|
|
"grad_norm": 0.29296875,
|
|
"learning_rate": 0.0007414051724137932,
|
|
"loss": 2.3273558616638184,
|
|
"step": 431
|
|
},
|
|
{
|
|
"epoch": 0.03323076923076923,
|
|
"grad_norm": 0.51953125,
|
|
"learning_rate": 0.0007431291379310344,
|
|
"loss": 2.4034628868103027,
|
|
"step": 432
|
|
},
|
|
{
|
|
"epoch": 0.03330769230769231,
|
|
"grad_norm": 0.240234375,
|
|
"learning_rate": 0.0007448531034482758,
|
|
"loss": 2.378164529800415,
|
|
"step": 433
|
|
},
|
|
{
|
|
"epoch": 0.03338461538461539,
|
|
"grad_norm": 0.23046875,
|
|
"learning_rate": 0.0007465770689655172,
|
|
"loss": 2.3748064041137695,
|
|
"step": 434
|
|
},
|
|
{
|
|
"epoch": 0.03346153846153846,
|
|
"grad_norm": 0.26171875,
|
|
"learning_rate": 0.0007483010344827587,
|
|
"loss": 2.3467514514923096,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 0.03353846153846154,
|
|
"grad_norm": 0.3125,
|
|
"learning_rate": 0.000750025,
|
|
"loss": 2.359844446182251,
|
|
"step": 436
|
|
},
|
|
{
|
|
"epoch": 0.033615384615384616,
|
|
"grad_norm": 0.466796875,
|
|
"learning_rate": 0.0007517489655172414,
|
|
"loss": 2.382267713546753,
|
|
"step": 437
|
|
},
|
|
{
|
|
"epoch": 0.033692307692307695,
|
|
"grad_norm": 0.1650390625,
|
|
"learning_rate": 0.0007534729310344828,
|
|
"loss": 2.3543038368225098,
|
|
"step": 438
|
|
},
|
|
{
|
|
"epoch": 0.03376923076923077,
|
|
"grad_norm": 0.2236328125,
|
|
"learning_rate": 0.0007551968965517241,
|
|
"loss": 2.332155227661133,
|
|
"step": 439
|
|
},
|
|
{
|
|
"epoch": 0.033846153846153845,
|
|
"grad_norm": 0.1953125,
|
|
"learning_rate": 0.0007569208620689655,
|
|
"loss": 2.3615312576293945,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.033923076923076924,
|
|
"grad_norm": 0.1669921875,
|
|
"learning_rate": 0.0007586448275862069,
|
|
"loss": 2.3229894638061523,
|
|
"step": 441
|
|
},
|
|
{
|
|
"epoch": 0.034,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 0.0007603687931034482,
|
|
"loss": 2.3574373722076416,
|
|
"step": 442
|
|
},
|
|
{
|
|
"epoch": 0.034076923076923074,
|
|
"grad_norm": 0.27734375,
|
|
"learning_rate": 0.0007620927586206896,
|
|
"loss": 2.3913779258728027,
|
|
"step": 443
|
|
},
|
|
{
|
|
"epoch": 0.03415384615384615,
|
|
"grad_norm": 0.2578125,
|
|
"learning_rate": 0.000763816724137931,
|
|
"loss": 2.358253240585327,
|
|
"step": 444
|
|
},
|
|
{
|
|
"epoch": 0.03423076923076923,
|
|
"grad_norm": 0.2099609375,
|
|
"learning_rate": 0.0007655406896551723,
|
|
"loss": 2.3381638526916504,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 0.03430769230769231,
|
|
"grad_norm": 0.13671875,
|
|
"learning_rate": 0.0007672646551724137,
|
|
"loss": 2.3228919506073,
|
|
"step": 446
|
|
},
|
|
{
|
|
"epoch": 0.03438461538461538,
|
|
"grad_norm": 0.33984375,
|
|
"learning_rate": 0.0007689886206896552,
|
|
"loss": 2.3871500492095947,
|
|
"step": 447
|
|
},
|
|
{
|
|
"epoch": 0.03446153846153846,
|
|
"grad_norm": 0.291015625,
|
|
"learning_rate": 0.0007707125862068964,
|
|
"loss": 2.3592708110809326,
|
|
"step": 448
|
|
},
|
|
{
|
|
"epoch": 0.03453846153846154,
|
|
"grad_norm": 0.1826171875,
|
|
"learning_rate": 0.0007724365517241379,
|
|
"loss": 2.3636491298675537,
|
|
"step": 449
|
|
},
|
|
{
|
|
"epoch": 0.03461538461538462,
|
|
"grad_norm": 0.21484375,
|
|
"learning_rate": 0.0007741605172413793,
|
|
"loss": 2.3158223628997803,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.03469230769230769,
|
|
"grad_norm": 0.14453125,
|
|
"learning_rate": 0.0007758844827586207,
|
|
"loss": 2.3360321521759033,
|
|
"step": 451
|
|
},
|
|
{
|
|
"epoch": 0.03476923076923077,
|
|
"grad_norm": 0.47265625,
|
|
"learning_rate": 0.0007776084482758619,
|
|
"loss": 2.4034264087677,
|
|
"step": 452
|
|
},
|
|
{
|
|
"epoch": 0.034846153846153846,
|
|
"grad_norm": 0.177734375,
|
|
"learning_rate": 0.0007793324137931034,
|
|
"loss": 2.3332104682922363,
|
|
"step": 453
|
|
},
|
|
{
|
|
"epoch": 0.034923076923076925,
|
|
"grad_norm": 0.1591796875,
|
|
"learning_rate": 0.0007810563793103448,
|
|
"loss": 2.3474137783050537,
|
|
"step": 454
|
|
},
|
|
{
|
|
"epoch": 0.035,
|
|
"grad_norm": 0.1533203125,
|
|
"learning_rate": 0.0007827803448275861,
|
|
"loss": 2.300544500350952,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 0.035076923076923075,
|
|
"grad_norm": 0.2109375,
|
|
"learning_rate": 0.0007845043103448277,
|
|
"loss": 2.3495965003967285,
|
|
"step": 456
|
|
},
|
|
{
|
|
"epoch": 0.035153846153846154,
|
|
"grad_norm": 0.294921875,
|
|
"learning_rate": 0.0007862282758620689,
|
|
"loss": 2.351247787475586,
|
|
"step": 457
|
|
},
|
|
{
|
|
"epoch": 0.03523076923076923,
|
|
"grad_norm": 0.14453125,
|
|
"learning_rate": 0.0007879522413793104,
|
|
"loss": 2.3424439430236816,
|
|
"step": 458
|
|
},
|
|
{
|
|
"epoch": 0.03530769230769231,
|
|
"grad_norm": 0.1259765625,
|
|
"learning_rate": 0.0007896762068965518,
|
|
"loss": 2.3269729614257812,
|
|
"step": 459
|
|
},
|
|
{
|
|
"epoch": 0.03538461538461538,
|
|
"grad_norm": 0.2158203125,
|
|
"learning_rate": 0.0007914001724137931,
|
|
"loss": 2.339643716812134,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.03546153846153846,
|
|
"grad_norm": 0.26171875,
|
|
"learning_rate": 0.0007931241379310344,
|
|
"loss": 2.351191759109497,
|
|
"step": 461
|
|
},
|
|
{
|
|
"epoch": 0.03553846153846154,
|
|
"grad_norm": 0.287109375,
|
|
"learning_rate": 0.0007948481034482759,
|
|
"loss": 2.38104248046875,
|
|
"step": 462
|
|
},
|
|
{
|
|
"epoch": 0.03561538461538462,
|
|
"grad_norm": 0.1337890625,
|
|
"learning_rate": 0.0007965720689655172,
|
|
"loss": 2.3408782482147217,
|
|
"step": 463
|
|
},
|
|
{
|
|
"epoch": 0.03569230769230769,
|
|
"grad_norm": 0.125,
|
|
"learning_rate": 0.0007982960344827587,
|
|
"loss": 2.3358283042907715,
|
|
"step": 464
|
|
},
|
|
{
|
|
"epoch": 0.03576923076923077,
|
|
"grad_norm": 0.1318359375,
|
|
"learning_rate": 0.0008000199999999999,
|
|
"loss": 2.3343422412872314,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 0.03584615384615385,
|
|
"grad_norm": 0.3984375,
|
|
"learning_rate": 0.0008017439655172413,
|
|
"loss": 2.407003402709961,
|
|
"step": 466
|
|
},
|
|
{
|
|
"epoch": 0.035923076923076926,
|
|
"grad_norm": 0.419921875,
|
|
"learning_rate": 0.0008034679310344828,
|
|
"loss": 2.3830714225769043,
|
|
"step": 467
|
|
},
|
|
{
|
|
"epoch": 0.036,
|
|
"grad_norm": 0.19140625,
|
|
"learning_rate": 0.0008051918965517242,
|
|
"loss": 2.3635828495025635,
|
|
"step": 468
|
|
},
|
|
{
|
|
"epoch": 0.036076923076923076,
|
|
"grad_norm": 0.2109375,
|
|
"learning_rate": 0.0008069158620689654,
|
|
"loss": 2.3600637912750244,
|
|
"step": 469
|
|
},
|
|
{
|
|
"epoch": 0.036153846153846154,
|
|
"grad_norm": 0.1728515625,
|
|
"learning_rate": 0.0008086398275862068,
|
|
"loss": 2.314394474029541,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 0.03623076923076923,
|
|
"grad_norm": 0.125,
|
|
"learning_rate": 0.0008103637931034483,
|
|
"loss": 2.3252623081207275,
|
|
"step": 471
|
|
},
|
|
{
|
|
"epoch": 0.036307692307692305,
|
|
"grad_norm": 0.365234375,
|
|
"learning_rate": 0.0008120877586206897,
|
|
"loss": 2.372300386428833,
|
|
"step": 472
|
|
},
|
|
{
|
|
"epoch": 0.03638461538461538,
|
|
"grad_norm": 0.16015625,
|
|
"learning_rate": 0.0008138117241379309,
|
|
"loss": 2.381964683532715,
|
|
"step": 473
|
|
},
|
|
{
|
|
"epoch": 0.03646153846153846,
|
|
"grad_norm": 0.1962890625,
|
|
"learning_rate": 0.0008155356896551723,
|
|
"loss": 2.3746445178985596,
|
|
"step": 474
|
|
},
|
|
{
|
|
"epoch": 0.03653846153846154,
|
|
"grad_norm": 0.154296875,
|
|
"learning_rate": 0.0008172596551724138,
|
|
"loss": 2.34956431388855,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 0.03661538461538461,
|
|
"grad_norm": 0.1630859375,
|
|
"learning_rate": 0.0008189836206896552,
|
|
"loss": 2.333387613296509,
|
|
"step": 476
|
|
},
|
|
{
|
|
"epoch": 0.03669230769230769,
|
|
"grad_norm": 0.326171875,
|
|
"learning_rate": 0.0008207075862068964,
|
|
"loss": 2.3585548400878906,
|
|
"step": 477
|
|
},
|
|
{
|
|
"epoch": 0.03676923076923077,
|
|
"grad_norm": 0.1767578125,
|
|
"learning_rate": 0.0008224315517241379,
|
|
"loss": 2.3274805545806885,
|
|
"step": 478
|
|
},
|
|
{
|
|
"epoch": 0.03684615384615385,
|
|
"grad_norm": 0.1337890625,
|
|
"learning_rate": 0.0008241555172413793,
|
|
"loss": 2.326779842376709,
|
|
"step": 479
|
|
},
|
|
{
|
|
"epoch": 0.036923076923076927,
|
|
"grad_norm": 0.1455078125,
|
|
"learning_rate": 0.0008258794827586208,
|
|
"loss": 2.368696451187134,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.037,
|
|
"grad_norm": 0.1923828125,
|
|
"learning_rate": 0.0008276034482758621,
|
|
"loss": 2.315467119216919,
|
|
"step": 481
|
|
},
|
|
{
|
|
"epoch": 0.03707692307692308,
|
|
"grad_norm": 0.287109375,
|
|
"learning_rate": 0.0008293274137931034,
|
|
"loss": 2.36686372756958,
|
|
"step": 482
|
|
},
|
|
{
|
|
"epoch": 0.037153846153846155,
|
|
"grad_norm": 0.1943359375,
|
|
"learning_rate": 0.0008310513793103448,
|
|
"loss": 2.382108211517334,
|
|
"step": 483
|
|
},
|
|
{
|
|
"epoch": 0.037230769230769234,
|
|
"grad_norm": 0.11181640625,
|
|
"learning_rate": 0.0008327753448275862,
|
|
"loss": 2.358457088470459,
|
|
"step": 484
|
|
},
|
|
{
|
|
"epoch": 0.037307692307692306,
|
|
"grad_norm": 0.1337890625,
|
|
"learning_rate": 0.0008344993103448276,
|
|
"loss": 2.3562204837799072,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 0.037384615384615384,
|
|
"grad_norm": 0.12109375,
|
|
"learning_rate": 0.000836223275862069,
|
|
"loss": 2.3102526664733887,
|
|
"step": 486
|
|
},
|
|
{
|
|
"epoch": 0.03746153846153846,
|
|
"grad_norm": 0.244140625,
|
|
"learning_rate": 0.0008379472413793102,
|
|
"loss": 2.3606557846069336,
|
|
"step": 487
|
|
},
|
|
{
|
|
"epoch": 0.03753846153846154,
|
|
"grad_norm": 0.1171875,
|
|
"learning_rate": 0.0008396712068965517,
|
|
"loss": 2.342296600341797,
|
|
"step": 488
|
|
},
|
|
{
|
|
"epoch": 0.03761538461538461,
|
|
"grad_norm": 0.1240234375,
|
|
"learning_rate": 0.0008413951724137932,
|
|
"loss": 2.320117473602295,
|
|
"step": 489
|
|
},
|
|
{
|
|
"epoch": 0.03769230769230769,
|
|
"grad_norm": 0.1494140625,
|
|
"learning_rate": 0.0008431191379310344,
|
|
"loss": 2.3178086280822754,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 0.03776923076923077,
|
|
"grad_norm": 0.10107421875,
|
|
"learning_rate": 0.0008448431034482758,
|
|
"loss": 2.3238868713378906,
|
|
"step": 491
|
|
},
|
|
{
|
|
"epoch": 0.03784615384615385,
|
|
"grad_norm": 0.240234375,
|
|
"learning_rate": 0.0008465670689655171,
|
|
"loss": 2.355005979537964,
|
|
"step": 492
|
|
},
|
|
{
|
|
"epoch": 0.03792307692307692,
|
|
"grad_norm": 0.126953125,
|
|
"learning_rate": 0.0008482910344827587,
|
|
"loss": 2.3518574237823486,
|
|
"step": 493
|
|
},
|
|
{
|
|
"epoch": 0.038,
|
|
"grad_norm": 0.1181640625,
|
|
"learning_rate": 0.0008500149999999999,
|
|
"loss": 2.3575267791748047,
|
|
"step": 494
|
|
},
|
|
{
|
|
"epoch": 0.03807692307692308,
|
|
"grad_norm": 0.12255859375,
|
|
"learning_rate": 0.0008517389655172413,
|
|
"loss": 2.324350357055664,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 0.038153846153846156,
|
|
"grad_norm": 0.244140625,
|
|
"learning_rate": 0.0008534629310344827,
|
|
"loss": 2.3231608867645264,
|
|
"step": 496
|
|
},
|
|
{
|
|
"epoch": 0.03823076923076923,
|
|
"grad_norm": 0.212890625,
|
|
"learning_rate": 0.0008551868965517242,
|
|
"loss": 2.3450562953948975,
|
|
"step": 497
|
|
},
|
|
{
|
|
"epoch": 0.038307692307692306,
|
|
"grad_norm": 0.11767578125,
|
|
"learning_rate": 0.0008569108620689654,
|
|
"loss": 2.3314554691314697,
|
|
"step": 498
|
|
},
|
|
{
|
|
"epoch": 0.038384615384615385,
|
|
"grad_norm": 0.14453125,
|
|
"learning_rate": 0.0008586348275862069,
|
|
"loss": 2.352429151535034,
|
|
"step": 499
|
|
},
|
|
{
|
|
"epoch": 0.038461538461538464,
|
|
"grad_norm": 0.13671875,
|
|
"learning_rate": 0.0008603587931034483,
|
|
"loss": 2.3274359703063965,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.038538461538461535,
|
|
"grad_norm": 0.29296875,
|
|
"learning_rate": 0.0008620827586206896,
|
|
"loss": 2.352755546569824,
|
|
"step": 501
|
|
},
|
|
{
|
|
"epoch": 0.038615384615384614,
|
|
"grad_norm": 0.126953125,
|
|
"learning_rate": 0.000863806724137931,
|
|
"loss": 2.362879991531372,
|
|
"step": 502
|
|
},
|
|
{
|
|
"epoch": 0.03869230769230769,
|
|
"grad_norm": 0.17578125,
|
|
"learning_rate": 0.0008655306896551724,
|
|
"loss": 2.3583664894104004,
|
|
"step": 503
|
|
},
|
|
{
|
|
"epoch": 0.03876923076923077,
|
|
"grad_norm": 0.08984375,
|
|
"learning_rate": 0.0008672546551724137,
|
|
"loss": 2.354548215866089,
|
|
"step": 504
|
|
},
|
|
{
|
|
"epoch": 0.03884615384615384,
|
|
"grad_norm": 0.1552734375,
|
|
"learning_rate": 0.0008689786206896551,
|
|
"loss": 2.321981430053711,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 0.03892307692307692,
|
|
"grad_norm": 0.263671875,
|
|
"learning_rate": 0.0008707025862068965,
|
|
"loss": 2.3475794792175293,
|
|
"step": 506
|
|
},
|
|
{
|
|
"epoch": 0.039,
|
|
"grad_norm": 0.15625,
|
|
"learning_rate": 0.0008724265517241378,
|
|
"loss": 2.35663104057312,
|
|
"step": 507
|
|
},
|
|
{
|
|
"epoch": 0.03907692307692308,
|
|
"grad_norm": 0.11083984375,
|
|
"learning_rate": 0.0008741505172413793,
|
|
"loss": 2.355034828186035,
|
|
"step": 508
|
|
},
|
|
{
|
|
"epoch": 0.03915384615384616,
|
|
"grad_norm": 0.09716796875,
|
|
"learning_rate": 0.0008758744827586207,
|
|
"loss": 2.3241186141967773,
|
|
"step": 509
|
|
},
|
|
{
|
|
"epoch": 0.03923076923076923,
|
|
"grad_norm": 0.0947265625,
|
|
"learning_rate": 0.0008775984482758621,
|
|
"loss": 2.295342445373535,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 0.03930769230769231,
|
|
"grad_norm": 0.22265625,
|
|
"learning_rate": 0.0008793224137931034,
|
|
"loss": 2.37380313873291,
|
|
"step": 511
|
|
},
|
|
{
|
|
"epoch": 0.039384615384615386,
|
|
"grad_norm": 0.134765625,
|
|
"learning_rate": 0.0008810463793103448,
|
|
"loss": 2.3448140621185303,
|
|
"step": 512
|
|
},
|
|
{
|
|
"epoch": 0.039461538461538465,
|
|
"grad_norm": 0.1103515625,
|
|
"learning_rate": 0.0008827703448275863,
|
|
"loss": 2.3571507930755615,
|
|
"step": 513
|
|
},
|
|
{
|
|
"epoch": 0.039538461538461536,
|
|
"grad_norm": 0.10498046875,
|
|
"learning_rate": 0.0008844943103448276,
|
|
"loss": 2.3463470935821533,
|
|
"step": 514
|
|
},
|
|
{
|
|
"epoch": 0.039615384615384615,
|
|
"grad_norm": 0.126953125,
|
|
"learning_rate": 0.0008862182758620689,
|
|
"loss": 2.3382456302642822,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 0.03969230769230769,
|
|
"grad_norm": 0.2080078125,
|
|
"learning_rate": 0.0008879422413793103,
|
|
"loss": 2.35414457321167,
|
|
"step": 516
|
|
},
|
|
{
|
|
"epoch": 0.03976923076923077,
|
|
"grad_norm": 0.1298828125,
|
|
"learning_rate": 0.0008896662068965518,
|
|
"loss": 2.3108327388763428,
|
|
"step": 517
|
|
},
|
|
{
|
|
"epoch": 0.039846153846153844,
|
|
"grad_norm": 0.09619140625,
|
|
"learning_rate": 0.0008913901724137931,
|
|
"loss": 2.346640110015869,
|
|
"step": 518
|
|
},
|
|
{
|
|
"epoch": 0.03992307692307692,
|
|
"grad_norm": 0.0859375,
|
|
"learning_rate": 0.0008931141379310344,
|
|
"loss": 2.3373820781707764,
|
|
"step": 519
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"grad_norm": 0.1376953125,
|
|
"learning_rate": 0.0008948381034482759,
|
|
"loss": 2.305022716522217,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.04007692307692308,
|
|
"grad_norm": 0.203125,
|
|
"learning_rate": 0.0008965620689655173,
|
|
"loss": 2.356292724609375,
|
|
"step": 521
|
|
},
|
|
{
|
|
"epoch": 0.04015384615384615,
|
|
"grad_norm": 0.119140625,
|
|
"learning_rate": 0.0008982860344827586,
|
|
"loss": 2.3204169273376465,
|
|
"step": 522
|
|
},
|
|
{
|
|
"epoch": 0.04023076923076923,
|
|
"grad_norm": 0.12890625,
|
|
"learning_rate": 0.0009000099999999999,
|
|
"loss": 2.352160692214966,
|
|
"step": 523
|
|
},
|
|
{
|
|
"epoch": 0.04030769230769231,
|
|
"grad_norm": 0.1513671875,
|
|
"learning_rate": 0.0009017339655172414,
|
|
"loss": 2.35078763961792,
|
|
"step": 524
|
|
},
|
|
{
|
|
"epoch": 0.04038461538461539,
|
|
"grad_norm": 0.09423828125,
|
|
"learning_rate": 0.0009034579310344827,
|
|
"loss": 2.309633255004883,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 0.04046153846153846,
|
|
"grad_norm": 0.25390625,
|
|
"learning_rate": 0.0009051818965517242,
|
|
"loss": 2.338869571685791,
|
|
"step": 526
|
|
},
|
|
{
|
|
"epoch": 0.04053846153846154,
|
|
"grad_norm": 0.1669921875,
|
|
"learning_rate": 0.0009069058620689654,
|
|
"loss": 2.3838884830474854,
|
|
"step": 527
|
|
},
|
|
{
|
|
"epoch": 0.040615384615384616,
|
|
"grad_norm": 0.099609375,
|
|
"learning_rate": 0.0009086298275862068,
|
|
"loss": 2.343787670135498,
|
|
"step": 528
|
|
},
|
|
{
|
|
"epoch": 0.040692307692307694,
|
|
"grad_norm": 0.08349609375,
|
|
"learning_rate": 0.0009103537931034483,
|
|
"loss": 2.316413164138794,
|
|
"step": 529
|
|
},
|
|
{
|
|
"epoch": 0.040769230769230766,
|
|
"grad_norm": 0.09130859375,
|
|
"learning_rate": 0.0009120777586206897,
|
|
"loss": 2.325157403945923,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 0.040846153846153845,
|
|
"grad_norm": 0.25390625,
|
|
"learning_rate": 0.0009138017241379309,
|
|
"loss": 2.3492913246154785,
|
|
"step": 531
|
|
},
|
|
{
|
|
"epoch": 0.04092307692307692,
|
|
"grad_norm": 0.158203125,
|
|
"learning_rate": 0.0009155256896551723,
|
|
"loss": 2.3442044258117676,
|
|
"step": 532
|
|
},
|
|
{
|
|
"epoch": 0.041,
|
|
"grad_norm": 0.10009765625,
|
|
"learning_rate": 0.0009172496551724138,
|
|
"loss": 2.345303773880005,
|
|
"step": 533
|
|
},
|
|
{
|
|
"epoch": 0.04107692307692308,
|
|
"grad_norm": 0.09326171875,
|
|
"learning_rate": 0.0009189736206896552,
|
|
"loss": 2.3094093799591064,
|
|
"step": 534
|
|
},
|
|
{
|
|
"epoch": 0.04115384615384615,
|
|
"grad_norm": 0.1962890625,
|
|
"learning_rate": 0.0009206975862068967,
|
|
"loss": 2.3518834114074707,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 0.04123076923076923,
|
|
"grad_norm": 0.1630859375,
|
|
"learning_rate": 0.0009224215517241378,
|
|
"loss": 2.3075547218322754,
|
|
"step": 536
|
|
},
|
|
{
|
|
"epoch": 0.04130769230769231,
|
|
"grad_norm": 0.1533203125,
|
|
"learning_rate": 0.0009241455172413793,
|
|
"loss": 2.3844878673553467,
|
|
"step": 537
|
|
},
|
|
{
|
|
"epoch": 0.04138461538461539,
|
|
"grad_norm": 0.08056640625,
|
|
"learning_rate": 0.0009258694827586207,
|
|
"loss": 2.3263814449310303,
|
|
"step": 538
|
|
},
|
|
{
|
|
"epoch": 0.04146153846153846,
|
|
"grad_norm": 0.103515625,
|
|
"learning_rate": 0.0009275934482758622,
|
|
"loss": 2.306187152862549,
|
|
"step": 539
|
|
},
|
|
{
|
|
"epoch": 0.04153846153846154,
|
|
"grad_norm": 0.09814453125,
|
|
"learning_rate": 0.0009293174137931034,
|
|
"loss": 2.294544219970703,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.04161538461538462,
|
|
"grad_norm": 0.1845703125,
|
|
"learning_rate": 0.0009310413793103448,
|
|
"loss": 2.339218854904175,
|
|
"step": 541
|
|
},
|
|
{
|
|
"epoch": 0.041692307692307695,
|
|
"grad_norm": 0.10986328125,
|
|
"learning_rate": 0.0009327653448275863,
|
|
"loss": 2.3683807849884033,
|
|
"step": 542
|
|
},
|
|
{
|
|
"epoch": 0.04176923076923077,
|
|
"grad_norm": 0.08349609375,
|
|
"learning_rate": 0.0009344893103448276,
|
|
"loss": 2.3212006092071533,
|
|
"step": 543
|
|
},
|
|
{
|
|
"epoch": 0.041846153846153845,
|
|
"grad_norm": 0.150390625,
|
|
"learning_rate": 0.000936213275862069,
|
|
"loss": 2.328866720199585,
|
|
"step": 544
|
|
},
|
|
{
|
|
"epoch": 0.041923076923076924,
|
|
"grad_norm": 0.095703125,
|
|
"learning_rate": 0.0009379372413793103,
|
|
"loss": 2.3314998149871826,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 0.042,
|
|
"grad_norm": 0.2001953125,
|
|
"learning_rate": 0.0009396612068965517,
|
|
"loss": 2.3379597663879395,
|
|
"step": 546
|
|
},
|
|
{
|
|
"epoch": 0.042076923076923074,
|
|
"grad_norm": 0.0869140625,
|
|
"learning_rate": 0.0009413851724137931,
|
|
"loss": 2.3489468097686768,
|
|
"step": 547
|
|
},
|
|
{
|
|
"epoch": 0.04215384615384615,
|
|
"grad_norm": 0.1357421875,
|
|
"learning_rate": 0.0009431091379310345,
|
|
"loss": 2.3574469089508057,
|
|
"step": 548
|
|
},
|
|
{
|
|
"epoch": 0.04223076923076923,
|
|
"grad_norm": 0.1279296875,
|
|
"learning_rate": 0.0009448331034482757,
|
|
"loss": 2.307446002960205,
|
|
"step": 549
|
|
},
|
|
{
|
|
"epoch": 0.04230769230769231,
|
|
"grad_norm": 0.10595703125,
|
|
"learning_rate": 0.0009465570689655172,
|
|
"loss": 2.298635244369507,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.04238461538461538,
|
|
"grad_norm": 0.11376953125,
|
|
"learning_rate": 0.0009482810344827587,
|
|
"loss": 2.3411130905151367,
|
|
"step": 551
|
|
},
|
|
{
|
|
"epoch": 0.04246153846153846,
|
|
"grad_norm": 0.09228515625,
|
|
"learning_rate": 0.0009500049999999999,
|
|
"loss": 2.3329997062683105,
|
|
"step": 552
|
|
},
|
|
{
|
|
"epoch": 0.04253846153846154,
|
|
"grad_norm": 0.087890625,
|
|
"learning_rate": 0.0009517289655172413,
|
|
"loss": 2.329939126968384,
|
|
"step": 553
|
|
},
|
|
{
|
|
"epoch": 0.04261538461538462,
|
|
"grad_norm": 0.091796875,
|
|
"learning_rate": 0.0009534529310344827,
|
|
"loss": 2.320272445678711,
|
|
"step": 554
|
|
},
|
|
{
|
|
"epoch": 0.04269230769230769,
|
|
"grad_norm": 0.115234375,
|
|
"learning_rate": 0.0009551768965517242,
|
|
"loss": 2.3082468509674072,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 0.04276923076923077,
|
|
"grad_norm": 0.236328125,
|
|
"learning_rate": 0.0009569008620689654,
|
|
"loss": 2.3427658081054688,
|
|
"step": 556
|
|
},
|
|
{
|
|
"epoch": 0.042846153846153846,
|
|
"grad_norm": 0.078125,
|
|
"learning_rate": 0.0009586248275862069,
|
|
"loss": 2.3370916843414307,
|
|
"step": 557
|
|
},
|
|
{
|
|
"epoch": 0.042923076923076925,
|
|
"grad_norm": 0.08642578125,
|
|
"learning_rate": 0.0009603487931034482,
|
|
"loss": 2.318396806716919,
|
|
"step": 558
|
|
},
|
|
{
|
|
"epoch": 0.043,
|
|
"grad_norm": 0.0888671875,
|
|
"learning_rate": 0.0009620727586206897,
|
|
"loss": 2.2979509830474854,
|
|
"step": 559
|
|
},
|
|
{
|
|
"epoch": 0.043076923076923075,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.0009637967241379309,
|
|
"loss": 2.315047025680542,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 0.043153846153846154,
|
|
"grad_norm": 0.18359375,
|
|
"learning_rate": 0.0009655206896551724,
|
|
"loss": 2.3645732402801514,
|
|
"step": 561
|
|
},
|
|
{
|
|
"epoch": 0.04323076923076923,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.0009672446551724139,
|
|
"loss": 2.343644380569458,
|
|
"step": 562
|
|
},
|
|
{
|
|
"epoch": 0.04330769230769231,
|
|
"grad_norm": 0.0986328125,
|
|
"learning_rate": 0.0009689686206896551,
|
|
"loss": 2.341093063354492,
|
|
"step": 563
|
|
},
|
|
{
|
|
"epoch": 0.04338461538461538,
|
|
"grad_norm": 0.07080078125,
|
|
"learning_rate": 0.0009706925862068966,
|
|
"loss": 2.311235189437866,
|
|
"step": 564
|
|
},
|
|
{
|
|
"epoch": 0.04346153846153846,
|
|
"grad_norm": 0.08203125,
|
|
"learning_rate": 0.0009724165517241379,
|
|
"loss": 2.29046368598938,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 0.04353846153846154,
|
|
"grad_norm": 0.140625,
|
|
"learning_rate": 0.0009741405172413793,
|
|
"loss": 2.3184750080108643,
|
|
"step": 566
|
|
},
|
|
{
|
|
"epoch": 0.04361538461538462,
|
|
"grad_norm": 0.07763671875,
|
|
"learning_rate": 0.0009758644827586206,
|
|
"loss": 2.3313910961151123,
|
|
"step": 567
|
|
},
|
|
{
|
|
"epoch": 0.04369230769230769,
|
|
"grad_norm": 0.0791015625,
|
|
"learning_rate": 0.0009775884482758621,
|
|
"loss": 2.336766481399536,
|
|
"step": 568
|
|
},
|
|
{
|
|
"epoch": 0.04376923076923077,
|
|
"grad_norm": 0.1083984375,
|
|
"learning_rate": 0.0009793124137931035,
|
|
"loss": 2.319382667541504,
|
|
"step": 569
|
|
},
|
|
{
|
|
"epoch": 0.04384615384615385,
|
|
"grad_norm": 0.1328125,
|
|
"learning_rate": 0.0009810363793103449,
|
|
"loss": 2.32136869430542,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 0.043923076923076926,
|
|
"grad_norm": 0.181640625,
|
|
"learning_rate": 0.0009827603448275862,
|
|
"loss": 2.3420627117156982,
|
|
"step": 571
|
|
},
|
|
{
|
|
"epoch": 0.044,
|
|
"grad_norm": 0.10595703125,
|
|
"learning_rate": 0.0009844843103448278,
|
|
"loss": 2.347048044204712,
|
|
"step": 572
|
|
},
|
|
{
|
|
"epoch": 0.044076923076923076,
|
|
"grad_norm": 0.076171875,
|
|
"learning_rate": 0.000986208275862069,
|
|
"loss": 2.3363776206970215,
|
|
"step": 573
|
|
},
|
|
{
|
|
"epoch": 0.044153846153846155,
|
|
"grad_norm": 0.0712890625,
|
|
"learning_rate": 0.0009879322413793104,
|
|
"loss": 2.3200719356536865,
|
|
"step": 574
|
|
},
|
|
{
|
|
"epoch": 0.04423076923076923,
|
|
"grad_norm": 0.0927734375,
|
|
"learning_rate": 0.0009896562068965518,
|
|
"loss": 2.310849666595459,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 0.044307692307692305,
|
|
"grad_norm": 0.2392578125,
|
|
"learning_rate": 0.0009913801724137934,
|
|
"loss": 2.3255860805511475,
|
|
"step": 576
|
|
},
|
|
{
|
|
"epoch": 0.044384615384615383,
|
|
"grad_norm": 0.1171875,
|
|
"learning_rate": 0.0009931041379310345,
|
|
"loss": 2.3304312229156494,
|
|
"step": 577
|
|
},
|
|
{
|
|
"epoch": 0.04446153846153846,
|
|
"grad_norm": 0.08349609375,
|
|
"learning_rate": 0.0009948281034482759,
|
|
"loss": 2.3498709201812744,
|
|
"step": 578
|
|
},
|
|
{
|
|
"epoch": 0.04453846153846154,
|
|
"grad_norm": 0.0830078125,
|
|
"learning_rate": 0.0009965520689655173,
|
|
"loss": 2.3063101768493652,
|
|
"step": 579
|
|
},
|
|
{
|
|
"epoch": 0.04461538461538461,
|
|
"grad_norm": 0.08447265625,
|
|
"learning_rate": 0.0009982760344827589,
|
|
"loss": 2.3217151165008545,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 0.04469230769230769,
|
|
"grad_norm": 0.08447265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3535659313201904,
|
|
"step": 581
|
|
},
|
|
{
|
|
"epoch": 0.04476923076923077,
|
|
"grad_norm": 0.07763671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.362304210662842,
|
|
"step": 582
|
|
},
|
|
{
|
|
"epoch": 0.04484615384615385,
|
|
"grad_norm": 0.087890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.311096668243408,
|
|
"step": 583
|
|
},
|
|
{
|
|
"epoch": 0.04492307692307692,
|
|
"grad_norm": 0.08349609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289670705795288,
|
|
"step": 584
|
|
},
|
|
{
|
|
"epoch": 0.045,
|
|
"grad_norm": 0.091796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.329411506652832,
|
|
"step": 585
|
|
},
|
|
{
|
|
"epoch": 0.04507692307692308,
|
|
"grad_norm": 0.185546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3375794887542725,
|
|
"step": 586
|
|
},
|
|
{
|
|
"epoch": 0.045153846153846156,
|
|
"grad_norm": 0.07470703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.342571258544922,
|
|
"step": 587
|
|
},
|
|
{
|
|
"epoch": 0.045230769230769234,
|
|
"grad_norm": 0.07177734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3473896980285645,
|
|
"step": 588
|
|
},
|
|
{
|
|
"epoch": 0.045307692307692306,
|
|
"grad_norm": 0.06689453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2865121364593506,
|
|
"step": 589
|
|
},
|
|
{
|
|
"epoch": 0.045384615384615384,
|
|
"grad_norm": 0.0771484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.311269760131836,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 0.04546153846153846,
|
|
"grad_norm": 0.150390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309694766998291,
|
|
"step": 591
|
|
},
|
|
{
|
|
"epoch": 0.04553846153846154,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.32133150100708,
|
|
"step": 592
|
|
},
|
|
{
|
|
"epoch": 0.04561538461538461,
|
|
"grad_norm": 0.06884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3544232845306396,
|
|
"step": 593
|
|
},
|
|
{
|
|
"epoch": 0.04569230769230769,
|
|
"grad_norm": 0.091796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2849857807159424,
|
|
"step": 594
|
|
},
|
|
{
|
|
"epoch": 0.04576923076923077,
|
|
"grad_norm": 0.07568359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.318098545074463,
|
|
"step": 595
|
|
},
|
|
{
|
|
"epoch": 0.04584615384615385,
|
|
"grad_norm": 0.1767578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.354386329650879,
|
|
"step": 596
|
|
},
|
|
{
|
|
"epoch": 0.04592307692307692,
|
|
"grad_norm": 0.1220703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3512113094329834,
|
|
"step": 597
|
|
},
|
|
{
|
|
"epoch": 0.046,
|
|
"grad_norm": 0.07861328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.350881814956665,
|
|
"step": 598
|
|
},
|
|
{
|
|
"epoch": 0.04607692307692308,
|
|
"grad_norm": 0.06591796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2994861602783203,
|
|
"step": 599
|
|
},
|
|
{
|
|
"epoch": 0.046153846153846156,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2979249954223633,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.04623076923076923,
|
|
"grad_norm": 0.162109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.342385768890381,
|
|
"step": 601
|
|
},
|
|
{
|
|
"epoch": 0.04630769230769231,
|
|
"grad_norm": 0.09716796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3670525550842285,
|
|
"step": 602
|
|
},
|
|
{
|
|
"epoch": 0.046384615384615385,
|
|
"grad_norm": 0.087890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313377618789673,
|
|
"step": 603
|
|
},
|
|
{
|
|
"epoch": 0.046461538461538464,
|
|
"grad_norm": 0.061767578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2833125591278076,
|
|
"step": 604
|
|
},
|
|
{
|
|
"epoch": 0.046538461538461535,
|
|
"grad_norm": 0.0810546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298144817352295,
|
|
"step": 605
|
|
},
|
|
{
|
|
"epoch": 0.046615384615384614,
|
|
"grad_norm": 0.095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3198747634887695,
|
|
"step": 606
|
|
},
|
|
{
|
|
"epoch": 0.04669230769230769,
|
|
"grad_norm": 0.06787109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.37497615814209,
|
|
"step": 607
|
|
},
|
|
{
|
|
"epoch": 0.04676923076923077,
|
|
"grad_norm": 0.0771484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3337717056274414,
|
|
"step": 608
|
|
},
|
|
{
|
|
"epoch": 0.04684615384615384,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3055007457733154,
|
|
"step": 609
|
|
},
|
|
{
|
|
"epoch": 0.04692307692307692,
|
|
"grad_norm": 0.07373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3105177879333496,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 0.047,
|
|
"grad_norm": 0.072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3223698139190674,
|
|
"step": 611
|
|
},
|
|
{
|
|
"epoch": 0.04707692307692308,
|
|
"grad_norm": 0.076171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.335233211517334,
|
|
"step": 612
|
|
},
|
|
{
|
|
"epoch": 0.04715384615384616,
|
|
"grad_norm": 0.0634765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3164377212524414,
|
|
"step": 613
|
|
},
|
|
{
|
|
"epoch": 0.04723076923076923,
|
|
"grad_norm": 0.08544921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2866997718811035,
|
|
"step": 614
|
|
},
|
|
{
|
|
"epoch": 0.04730769230769231,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2829737663269043,
|
|
"step": 615
|
|
},
|
|
{
|
|
"epoch": 0.047384615384615386,
|
|
"grad_norm": 0.1611328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3216052055358887,
|
|
"step": 616
|
|
},
|
|
{
|
|
"epoch": 0.047461538461538465,
|
|
"grad_norm": 0.064453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3419198989868164,
|
|
"step": 617
|
|
},
|
|
{
|
|
"epoch": 0.047538461538461536,
|
|
"grad_norm": 0.0615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3307266235351562,
|
|
"step": 618
|
|
},
|
|
{
|
|
"epoch": 0.047615384615384615,
|
|
"grad_norm": 0.08837890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.322772979736328,
|
|
"step": 619
|
|
},
|
|
{
|
|
"epoch": 0.047692307692307694,
|
|
"grad_norm": 0.0732421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3054792881011963,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 0.04776923076923077,
|
|
"grad_norm": 0.111328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3276708126068115,
|
|
"step": 621
|
|
},
|
|
{
|
|
"epoch": 0.047846153846153844,
|
|
"grad_norm": 0.0771484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313854217529297,
|
|
"step": 622
|
|
},
|
|
{
|
|
"epoch": 0.04792307692307692,
|
|
"grad_norm": 0.08984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3213045597076416,
|
|
"step": 623
|
|
},
|
|
{
|
|
"epoch": 0.048,
|
|
"grad_norm": 0.0869140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.344265937805176,
|
|
"step": 624
|
|
},
|
|
{
|
|
"epoch": 0.04807692307692308,
|
|
"grad_norm": 0.0712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3091607093811035,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 0.04815384615384615,
|
|
"grad_norm": 0.1845703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3512611389160156,
|
|
"step": 626
|
|
},
|
|
{
|
|
"epoch": 0.04823076923076923,
|
|
"grad_norm": 0.07958984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3566741943359375,
|
|
"step": 627
|
|
},
|
|
{
|
|
"epoch": 0.04830769230769231,
|
|
"grad_norm": 0.061279296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3344690799713135,
|
|
"step": 628
|
|
},
|
|
{
|
|
"epoch": 0.04838461538461539,
|
|
"grad_norm": 0.06689453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.32702898979187,
|
|
"step": 629
|
|
},
|
|
{
|
|
"epoch": 0.04846153846153846,
|
|
"grad_norm": 0.0615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304476261138916,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 0.04853846153846154,
|
|
"grad_norm": 0.1806640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3405394554138184,
|
|
"step": 631
|
|
},
|
|
{
|
|
"epoch": 0.048615384615384616,
|
|
"grad_norm": 0.06884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3125710487365723,
|
|
"step": 632
|
|
},
|
|
{
|
|
"epoch": 0.048692307692307694,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3211774826049805,
|
|
"step": 633
|
|
},
|
|
{
|
|
"epoch": 0.048769230769230766,
|
|
"grad_norm": 0.061279296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302459716796875,
|
|
"step": 634
|
|
},
|
|
{
|
|
"epoch": 0.048846153846153845,
|
|
"grad_norm": 0.061767578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298485517501831,
|
|
"step": 635
|
|
},
|
|
{
|
|
"epoch": 0.04892307692307692,
|
|
"grad_norm": 0.087890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.340817928314209,
|
|
"step": 636
|
|
},
|
|
{
|
|
"epoch": 0.049,
|
|
"grad_norm": 0.07373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.317852258682251,
|
|
"step": 637
|
|
},
|
|
{
|
|
"epoch": 0.049076923076923074,
|
|
"grad_norm": 0.0673828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.332689046859741,
|
|
"step": 638
|
|
},
|
|
{
|
|
"epoch": 0.04915384615384615,
|
|
"grad_norm": 0.061279296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293757200241089,
|
|
"step": 639
|
|
},
|
|
{
|
|
"epoch": 0.04923076923076923,
|
|
"grad_norm": 0.07958984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2918901443481445,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 0.04930769230769231,
|
|
"grad_norm": 0.0810546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3383333683013916,
|
|
"step": 641
|
|
},
|
|
{
|
|
"epoch": 0.04938461538461539,
|
|
"grad_norm": 0.0908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.327881336212158,
|
|
"step": 642
|
|
},
|
|
{
|
|
"epoch": 0.04946153846153846,
|
|
"grad_norm": 0.072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.277453899383545,
|
|
"step": 643
|
|
},
|
|
{
|
|
"epoch": 0.04953846153846154,
|
|
"grad_norm": 0.09326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3405628204345703,
|
|
"step": 644
|
|
},
|
|
{
|
|
"epoch": 0.04961538461538462,
|
|
"grad_norm": 0.09228515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3202908039093018,
|
|
"step": 645
|
|
},
|
|
{
|
|
"epoch": 0.049692307692307695,
|
|
"grad_norm": 0.07373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3298017978668213,
|
|
"step": 646
|
|
},
|
|
{
|
|
"epoch": 0.04976923076923077,
|
|
"grad_norm": 0.060791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3572258949279785,
|
|
"step": 647
|
|
},
|
|
{
|
|
"epoch": 0.049846153846153846,
|
|
"grad_norm": 0.07080078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.321516752243042,
|
|
"step": 648
|
|
},
|
|
{
|
|
"epoch": 0.049923076923076924,
|
|
"grad_norm": 0.0888671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293996810913086,
|
|
"step": 649
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"grad_norm": 0.0859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3441944122314453,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"eval_loss": 2.057891368865967,
|
|
"eval_runtime": 737.2152,
|
|
"eval_samples_per_second": 25.789,
|
|
"eval_steps_per_second": 6.447,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.050076923076923074,
|
|
"grad_norm": 0.1171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3347079753875732,
|
|
"step": 651
|
|
},
|
|
{
|
|
"epoch": 0.05015384615384615,
|
|
"grad_norm": 0.06787109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3095526695251465,
|
|
"step": 652
|
|
},
|
|
{
|
|
"epoch": 0.05023076923076923,
|
|
"grad_norm": 0.06103515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3205273151397705,
|
|
"step": 653
|
|
},
|
|
{
|
|
"epoch": 0.05030769230769231,
|
|
"grad_norm": 0.06103515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3023438453674316,
|
|
"step": 654
|
|
},
|
|
{
|
|
"epoch": 0.05038461538461538,
|
|
"grad_norm": 0.05810546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3185677528381348,
|
|
"step": 655
|
|
},
|
|
{
|
|
"epoch": 0.05046153846153846,
|
|
"grad_norm": 0.11572265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3112826347351074,
|
|
"step": 656
|
|
},
|
|
{
|
|
"epoch": 0.05053846153846154,
|
|
"grad_norm": 0.0634765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.310549259185791,
|
|
"step": 657
|
|
},
|
|
{
|
|
"epoch": 0.05061538461538462,
|
|
"grad_norm": 0.07080078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3319876194000244,
|
|
"step": 658
|
|
},
|
|
{
|
|
"epoch": 0.05069230769230769,
|
|
"grad_norm": 0.0615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3014066219329834,
|
|
"step": 659
|
|
},
|
|
{
|
|
"epoch": 0.05076923076923077,
|
|
"grad_norm": 0.057373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300014019012451,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 0.050846153846153846,
|
|
"grad_norm": 0.1630859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3478991985321045,
|
|
"step": 661
|
|
},
|
|
{
|
|
"epoch": 0.050923076923076925,
|
|
"grad_norm": 0.061279296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.333362340927124,
|
|
"step": 662
|
|
},
|
|
{
|
|
"epoch": 0.051,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.310438394546509,
|
|
"step": 663
|
|
},
|
|
{
|
|
"epoch": 0.051076923076923075,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287053346633911,
|
|
"step": 664
|
|
},
|
|
{
|
|
"epoch": 0.051153846153846154,
|
|
"grad_norm": 0.07421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3353095054626465,
|
|
"step": 665
|
|
},
|
|
{
|
|
"epoch": 0.05123076923076923,
|
|
"grad_norm": 0.06494140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.327256917953491,
|
|
"step": 666
|
|
},
|
|
{
|
|
"epoch": 0.05130769230769231,
|
|
"grad_norm": 0.061767578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3253142833709717,
|
|
"step": 667
|
|
},
|
|
{
|
|
"epoch": 0.05138461538461538,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3325653076171875,
|
|
"step": 668
|
|
},
|
|
{
|
|
"epoch": 0.05146153846153846,
|
|
"grad_norm": 0.0634765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2972283363342285,
|
|
"step": 669
|
|
},
|
|
{
|
|
"epoch": 0.05153846153846154,
|
|
"grad_norm": 0.1259765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.31846284866333,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 0.05161538461538462,
|
|
"grad_norm": 0.0859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3505542278289795,
|
|
"step": 671
|
|
},
|
|
{
|
|
"epoch": 0.05169230769230769,
|
|
"grad_norm": 0.058837890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3241498470306396,
|
|
"step": 672
|
|
},
|
|
{
|
|
"epoch": 0.05176923076923077,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3097870349884033,
|
|
"step": 673
|
|
},
|
|
{
|
|
"epoch": 0.05184615384615385,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3205947875976562,
|
|
"step": 674
|
|
},
|
|
{
|
|
"epoch": 0.051923076923076926,
|
|
"grad_norm": 0.08935546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3349335193634033,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 0.052,
|
|
"grad_norm": 0.06494140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.335045576095581,
|
|
"step": 676
|
|
},
|
|
{
|
|
"epoch": 0.052076923076923076,
|
|
"grad_norm": 0.06689453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.321113348007202,
|
|
"step": 677
|
|
},
|
|
{
|
|
"epoch": 0.052153846153846155,
|
|
"grad_norm": 0.07470703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3279402256011963,
|
|
"step": 678
|
|
},
|
|
{
|
|
"epoch": 0.05223076923076923,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301842212677002,
|
|
"step": 679
|
|
},
|
|
{
|
|
"epoch": 0.052307692307692305,
|
|
"grad_norm": 0.11865234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3382153511047363,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 0.052384615384615384,
|
|
"grad_norm": 0.06982421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3207309246063232,
|
|
"step": 681
|
|
},
|
|
{
|
|
"epoch": 0.05246153846153846,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.307994842529297,
|
|
"step": 682
|
|
},
|
|
{
|
|
"epoch": 0.05253846153846154,
|
|
"grad_norm": 0.056640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2960360050201416,
|
|
"step": 683
|
|
},
|
|
{
|
|
"epoch": 0.05261538461538461,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2915523052215576,
|
|
"step": 684
|
|
},
|
|
{
|
|
"epoch": 0.05269230769230769,
|
|
"grad_norm": 0.1328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3074569702148438,
|
|
"step": 685
|
|
},
|
|
{
|
|
"epoch": 0.05276923076923077,
|
|
"grad_norm": 0.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3074331283569336,
|
|
"step": 686
|
|
},
|
|
{
|
|
"epoch": 0.05284615384615385,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309718132019043,
|
|
"step": 687
|
|
},
|
|
{
|
|
"epoch": 0.05292307692307692,
|
|
"grad_norm": 0.08154296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285754442214966,
|
|
"step": 688
|
|
},
|
|
{
|
|
"epoch": 0.053,
|
|
"grad_norm": 0.0634765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3056955337524414,
|
|
"step": 689
|
|
},
|
|
{
|
|
"epoch": 0.05307692307692308,
|
|
"grad_norm": 0.11181640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3402061462402344,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 0.053153846153846156,
|
|
"grad_norm": 0.0830078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3174920082092285,
|
|
"step": 691
|
|
},
|
|
{
|
|
"epoch": 0.05323076923076923,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.328782081604004,
|
|
"step": 692
|
|
},
|
|
{
|
|
"epoch": 0.053307692307692306,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3112332820892334,
|
|
"step": 693
|
|
},
|
|
{
|
|
"epoch": 0.053384615384615385,
|
|
"grad_norm": 0.0615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.330681562423706,
|
|
"step": 694
|
|
},
|
|
{
|
|
"epoch": 0.05346153846153846,
|
|
"grad_norm": 0.060546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.319808006286621,
|
|
"step": 695
|
|
},
|
|
{
|
|
"epoch": 0.05353846153846154,
|
|
"grad_norm": 0.05859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2788195610046387,
|
|
"step": 696
|
|
},
|
|
{
|
|
"epoch": 0.05361538461538461,
|
|
"grad_norm": 0.058349609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3012642860412598,
|
|
"step": 697
|
|
},
|
|
{
|
|
"epoch": 0.05369230769230769,
|
|
"grad_norm": 0.05908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.295173406600952,
|
|
"step": 698
|
|
},
|
|
{
|
|
"epoch": 0.05376923076923077,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2661664485931396,
|
|
"step": 699
|
|
},
|
|
{
|
|
"epoch": 0.05384615384615385,
|
|
"grad_norm": 0.134765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3316609859466553,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.05392307692307692,
|
|
"grad_norm": 0.06396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3306350708007812,
|
|
"step": 701
|
|
},
|
|
{
|
|
"epoch": 0.054,
|
|
"grad_norm": 0.0556640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.295478343963623,
|
|
"step": 702
|
|
},
|
|
{
|
|
"epoch": 0.05407692307692308,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279919385910034,
|
|
"step": 703
|
|
},
|
|
{
|
|
"epoch": 0.05415384615384616,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.295625686645508,
|
|
"step": 704
|
|
},
|
|
{
|
|
"epoch": 0.05423076923076923,
|
|
"grad_norm": 0.1376953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3119618892669678,
|
|
"step": 705
|
|
},
|
|
{
|
|
"epoch": 0.05430769230769231,
|
|
"grad_norm": 0.060791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.322444200515747,
|
|
"step": 706
|
|
},
|
|
{
|
|
"epoch": 0.054384615384615385,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3012478351593018,
|
|
"step": 707
|
|
},
|
|
{
|
|
"epoch": 0.054461538461538464,
|
|
"grad_norm": 0.057373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2694203853607178,
|
|
"step": 708
|
|
},
|
|
{
|
|
"epoch": 0.054538461538461536,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2987098693847656,
|
|
"step": 709
|
|
},
|
|
{
|
|
"epoch": 0.054615384615384614,
|
|
"grad_norm": 0.078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.346726179122925,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 0.05469230769230769,
|
|
"grad_norm": 0.060546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.308034658432007,
|
|
"step": 711
|
|
},
|
|
{
|
|
"epoch": 0.05476923076923077,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3248188495635986,
|
|
"step": 712
|
|
},
|
|
{
|
|
"epoch": 0.05484615384615384,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300440549850464,
|
|
"step": 713
|
|
},
|
|
{
|
|
"epoch": 0.05492307692307692,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2859814167022705,
|
|
"step": 714
|
|
},
|
|
{
|
|
"epoch": 0.055,
|
|
"grad_norm": 0.10302734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3244669437408447,
|
|
"step": 715
|
|
},
|
|
{
|
|
"epoch": 0.05507692307692308,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.348273515701294,
|
|
"step": 716
|
|
},
|
|
{
|
|
"epoch": 0.05515384615384615,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3048415184020996,
|
|
"step": 717
|
|
},
|
|
{
|
|
"epoch": 0.05523076923076923,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2858529090881348,
|
|
"step": 718
|
|
},
|
|
{
|
|
"epoch": 0.05530769230769231,
|
|
"grad_norm": 0.05859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2916669845581055,
|
|
"step": 719
|
|
},
|
|
{
|
|
"epoch": 0.055384615384615386,
|
|
"grad_norm": 0.087890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.323831081390381,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 0.055461538461538465,
|
|
"grad_norm": 0.061279296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3186769485473633,
|
|
"step": 721
|
|
},
|
|
{
|
|
"epoch": 0.05553846153846154,
|
|
"grad_norm": 0.05810546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294464111328125,
|
|
"step": 722
|
|
},
|
|
{
|
|
"epoch": 0.055615384615384615,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2951059341430664,
|
|
"step": 723
|
|
},
|
|
{
|
|
"epoch": 0.055692307692307694,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3066041469573975,
|
|
"step": 724
|
|
},
|
|
{
|
|
"epoch": 0.05576923076923077,
|
|
"grad_norm": 0.06494140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.348342180252075,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 0.055846153846153844,
|
|
"grad_norm": 0.06103515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.344860553741455,
|
|
"step": 726
|
|
},
|
|
{
|
|
"epoch": 0.05592307692307692,
|
|
"grad_norm": 0.0654296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299546003341675,
|
|
"step": 727
|
|
},
|
|
{
|
|
"epoch": 0.056,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.321791648864746,
|
|
"step": 728
|
|
},
|
|
{
|
|
"epoch": 0.05607692307692308,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3141982555389404,
|
|
"step": 729
|
|
},
|
|
{
|
|
"epoch": 0.05615384615384615,
|
|
"grad_norm": 0.07080078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.326770782470703,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 0.05623076923076923,
|
|
"grad_norm": 0.060302734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3316993713378906,
|
|
"step": 731
|
|
},
|
|
{
|
|
"epoch": 0.05630769230769231,
|
|
"grad_norm": 0.06640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3425333499908447,
|
|
"step": 732
|
|
},
|
|
{
|
|
"epoch": 0.05638461538461539,
|
|
"grad_norm": 0.053955078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3052456378936768,
|
|
"step": 733
|
|
},
|
|
{
|
|
"epoch": 0.05646153846153846,
|
|
"grad_norm": 0.06982421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274475336074829,
|
|
"step": 734
|
|
},
|
|
{
|
|
"epoch": 0.05653846153846154,
|
|
"grad_norm": 0.10205078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3607637882232666,
|
|
"step": 735
|
|
},
|
|
{
|
|
"epoch": 0.056615384615384616,
|
|
"grad_norm": 0.07666015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313366651535034,
|
|
"step": 736
|
|
},
|
|
{
|
|
"epoch": 0.056692307692307695,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2943239212036133,
|
|
"step": 737
|
|
},
|
|
{
|
|
"epoch": 0.056769230769230766,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299917221069336,
|
|
"step": 738
|
|
},
|
|
{
|
|
"epoch": 0.056846153846153845,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290950298309326,
|
|
"step": 739
|
|
},
|
|
{
|
|
"epoch": 0.05692307692307692,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.330909013748169,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 0.057,
|
|
"grad_norm": 0.060302734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3106706142425537,
|
|
"step": 741
|
|
},
|
|
{
|
|
"epoch": 0.057076923076923074,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28202486038208,
|
|
"step": 742
|
|
},
|
|
{
|
|
"epoch": 0.05715384615384615,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28804874420166,
|
|
"step": 743
|
|
},
|
|
{
|
|
"epoch": 0.05723076923076923,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290073871612549,
|
|
"step": 744
|
|
},
|
|
{
|
|
"epoch": 0.05730769230769231,
|
|
"grad_norm": 0.07373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3489372730255127,
|
|
"step": 745
|
|
},
|
|
{
|
|
"epoch": 0.05738461538461539,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.330341100692749,
|
|
"step": 746
|
|
},
|
|
{
|
|
"epoch": 0.05746153846153846,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.316121816635132,
|
|
"step": 747
|
|
},
|
|
{
|
|
"epoch": 0.05753846153846154,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.284105062484741,
|
|
"step": 748
|
|
},
|
|
{
|
|
"epoch": 0.05761538461538462,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285276412963867,
|
|
"step": 749
|
|
},
|
|
{
|
|
"epoch": 0.057692307692307696,
|
|
"grad_norm": 0.080078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.322348117828369,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.05776923076923077,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3278934955596924,
|
|
"step": 751
|
|
},
|
|
{
|
|
"epoch": 0.057846153846153846,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2983362674713135,
|
|
"step": 752
|
|
},
|
|
{
|
|
"epoch": 0.057923076923076924,
|
|
"grad_norm": 0.08544921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3132517337799072,
|
|
"step": 753
|
|
},
|
|
{
|
|
"epoch": 0.058,
|
|
"grad_norm": 0.06689453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3043758869171143,
|
|
"step": 754
|
|
},
|
|
{
|
|
"epoch": 0.058076923076923075,
|
|
"grad_norm": 0.06298828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3408679962158203,
|
|
"step": 755
|
|
},
|
|
{
|
|
"epoch": 0.05815384615384615,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.306149959564209,
|
|
"step": 756
|
|
},
|
|
{
|
|
"epoch": 0.05823076923076923,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2903904914855957,
|
|
"step": 757
|
|
},
|
|
{
|
|
"epoch": 0.05830769230769231,
|
|
"grad_norm": 0.0654296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2846899032592773,
|
|
"step": 758
|
|
},
|
|
{
|
|
"epoch": 0.05838461538461538,
|
|
"grad_norm": 0.07763671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30981707572937,
|
|
"step": 759
|
|
},
|
|
{
|
|
"epoch": 0.05846153846153846,
|
|
"grad_norm": 0.1591796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2949326038360596,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 0.05853846153846154,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3338418006896973,
|
|
"step": 761
|
|
},
|
|
{
|
|
"epoch": 0.05861538461538462,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3034679889678955,
|
|
"step": 762
|
|
},
|
|
{
|
|
"epoch": 0.05869230769230769,
|
|
"grad_norm": 0.059814453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.288644790649414,
|
|
"step": 763
|
|
},
|
|
{
|
|
"epoch": 0.05876923076923077,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3021903038024902,
|
|
"step": 764
|
|
},
|
|
{
|
|
"epoch": 0.05884615384615385,
|
|
"grad_norm": 0.0927734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3286383152008057,
|
|
"step": 765
|
|
},
|
|
{
|
|
"epoch": 0.058923076923076925,
|
|
"grad_norm": 0.0556640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.340181827545166,
|
|
"step": 766
|
|
},
|
|
{
|
|
"epoch": 0.059,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3203792572021484,
|
|
"step": 767
|
|
},
|
|
{
|
|
"epoch": 0.059076923076923075,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3107333183288574,
|
|
"step": 768
|
|
},
|
|
{
|
|
"epoch": 0.059153846153846154,
|
|
"grad_norm": 0.0673828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2832186222076416,
|
|
"step": 769
|
|
},
|
|
{
|
|
"epoch": 0.05923076923076923,
|
|
"grad_norm": 0.076171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3454933166503906,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 0.059307692307692304,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3276398181915283,
|
|
"step": 771
|
|
},
|
|
{
|
|
"epoch": 0.05938461538461538,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3059475421905518,
|
|
"step": 772
|
|
},
|
|
{
|
|
"epoch": 0.05946153846153846,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281750440597534,
|
|
"step": 773
|
|
},
|
|
{
|
|
"epoch": 0.05953846153846154,
|
|
"grad_norm": 0.06396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3230948448181152,
|
|
"step": 774
|
|
},
|
|
{
|
|
"epoch": 0.05961538461538462,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3180582523345947,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 0.05969230769230769,
|
|
"grad_norm": 0.0771484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3201804161071777,
|
|
"step": 776
|
|
},
|
|
{
|
|
"epoch": 0.05976923076923077,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.323944091796875,
|
|
"step": 777
|
|
},
|
|
{
|
|
"epoch": 0.05984615384615385,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2945234775543213,
|
|
"step": 778
|
|
},
|
|
{
|
|
"epoch": 0.059923076923076926,
|
|
"grad_norm": 0.0673828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302614450454712,
|
|
"step": 779
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.327493906021118,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 0.060076923076923076,
|
|
"grad_norm": 0.0703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.303632974624634,
|
|
"step": 781
|
|
},
|
|
{
|
|
"epoch": 0.060153846153846155,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3138580322265625,
|
|
"step": 782
|
|
},
|
|
{
|
|
"epoch": 0.060230769230769234,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3245837688446045,
|
|
"step": 783
|
|
},
|
|
{
|
|
"epoch": 0.060307692307692305,
|
|
"grad_norm": 0.0703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2760727405548096,
|
|
"step": 784
|
|
},
|
|
{
|
|
"epoch": 0.060384615384615384,
|
|
"grad_norm": 0.08056640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3182547092437744,
|
|
"step": 785
|
|
},
|
|
{
|
|
"epoch": 0.06046153846153846,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3168816566467285,
|
|
"step": 786
|
|
},
|
|
{
|
|
"epoch": 0.06053846153846154,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.325174570083618,
|
|
"step": 787
|
|
},
|
|
{
|
|
"epoch": 0.06061538461538461,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2896981239318848,
|
|
"step": 788
|
|
},
|
|
{
|
|
"epoch": 0.06069230769230769,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301520347595215,
|
|
"step": 789
|
|
},
|
|
{
|
|
"epoch": 0.06076923076923077,
|
|
"grad_norm": 0.0703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.318538188934326,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 0.06084615384615385,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3031766414642334,
|
|
"step": 791
|
|
},
|
|
{
|
|
"epoch": 0.06092307692307692,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3114781379699707,
|
|
"step": 792
|
|
},
|
|
{
|
|
"epoch": 0.061,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2933480739593506,
|
|
"step": 793
|
|
},
|
|
{
|
|
"epoch": 0.06107692307692308,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269770622253418,
|
|
"step": 794
|
|
},
|
|
{
|
|
"epoch": 0.061153846153846156,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3052868843078613,
|
|
"step": 795
|
|
},
|
|
{
|
|
"epoch": 0.06123076923076923,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.324531316757202,
|
|
"step": 796
|
|
},
|
|
{
|
|
"epoch": 0.061307692307692306,
|
|
"grad_norm": 0.091796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3485219478607178,
|
|
"step": 797
|
|
},
|
|
{
|
|
"epoch": 0.061384615384615385,
|
|
"grad_norm": 0.056640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2947423458099365,
|
|
"step": 798
|
|
},
|
|
{
|
|
"epoch": 0.06146153846153846,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2920188903808594,
|
|
"step": 799
|
|
},
|
|
{
|
|
"epoch": 0.06153846153846154,
|
|
"grad_norm": 0.09765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.335989236831665,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.061615384615384614,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2920074462890625,
|
|
"step": 801
|
|
},
|
|
{
|
|
"epoch": 0.06169230769230769,
|
|
"grad_norm": 0.05859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.305457592010498,
|
|
"step": 802
|
|
},
|
|
{
|
|
"epoch": 0.06176923076923077,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2792134284973145,
|
|
"step": 803
|
|
},
|
|
{
|
|
"epoch": 0.06184615384615385,
|
|
"grad_norm": 0.059814453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292145252227783,
|
|
"step": 804
|
|
},
|
|
{
|
|
"epoch": 0.06192307692307692,
|
|
"grad_norm": 0.057373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.318772077560425,
|
|
"step": 805
|
|
},
|
|
{
|
|
"epoch": 0.062,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3019418716430664,
|
|
"step": 806
|
|
},
|
|
{
|
|
"epoch": 0.06207692307692308,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286505937576294,
|
|
"step": 807
|
|
},
|
|
{
|
|
"epoch": 0.06215384615384616,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287654399871826,
|
|
"step": 808
|
|
},
|
|
{
|
|
"epoch": 0.06223076923076923,
|
|
"grad_norm": 0.08154296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302781820297241,
|
|
"step": 809
|
|
},
|
|
{
|
|
"epoch": 0.06230769230769231,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.330990791320801,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 0.062384615384615386,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.343008041381836,
|
|
"step": 811
|
|
},
|
|
{
|
|
"epoch": 0.062461538461538464,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2964632511138916,
|
|
"step": 812
|
|
},
|
|
{
|
|
"epoch": 0.06253846153846154,
|
|
"grad_norm": 0.06396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2926793098449707,
|
|
"step": 813
|
|
},
|
|
{
|
|
"epoch": 0.06261538461538461,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2835094928741455,
|
|
"step": 814
|
|
},
|
|
{
|
|
"epoch": 0.06269230769230769,
|
|
"grad_norm": 0.06201171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3310914039611816,
|
|
"step": 815
|
|
},
|
|
{
|
|
"epoch": 0.06276923076923077,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3154735565185547,
|
|
"step": 816
|
|
},
|
|
{
|
|
"epoch": 0.06284615384615384,
|
|
"grad_norm": 0.059814453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3333277702331543,
|
|
"step": 817
|
|
},
|
|
{
|
|
"epoch": 0.06292307692307693,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3227367401123047,
|
|
"step": 818
|
|
},
|
|
{
|
|
"epoch": 0.063,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282726526260376,
|
|
"step": 819
|
|
},
|
|
{
|
|
"epoch": 0.06307692307692307,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.316619396209717,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 0.06315384615384616,
|
|
"grad_norm": 0.057861328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279437780380249,
|
|
"step": 821
|
|
},
|
|
{
|
|
"epoch": 0.06323076923076923,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292891025543213,
|
|
"step": 822
|
|
},
|
|
{
|
|
"epoch": 0.0633076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2755231857299805,
|
|
"step": 823
|
|
},
|
|
{
|
|
"epoch": 0.06338461538461539,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3167638778686523,
|
|
"step": 824
|
|
},
|
|
{
|
|
"epoch": 0.06346153846153846,
|
|
"grad_norm": 0.0888671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.330213785171509,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 0.06353846153846154,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3486156463623047,
|
|
"step": 826
|
|
},
|
|
{
|
|
"epoch": 0.06361538461538462,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285367727279663,
|
|
"step": 827
|
|
},
|
|
{
|
|
"epoch": 0.06369230769230769,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2871158123016357,
|
|
"step": 828
|
|
},
|
|
{
|
|
"epoch": 0.06376923076923077,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3146228790283203,
|
|
"step": 829
|
|
},
|
|
{
|
|
"epoch": 0.06384615384615384,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3339030742645264,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 0.06392307692307693,
|
|
"grad_norm": 0.058349609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.334737777709961,
|
|
"step": 831
|
|
},
|
|
{
|
|
"epoch": 0.064,
|
|
"grad_norm": 0.053955078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2767813205718994,
|
|
"step": 832
|
|
},
|
|
{
|
|
"epoch": 0.06407692307692307,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292248010635376,
|
|
"step": 833
|
|
},
|
|
{
|
|
"epoch": 0.06415384615384616,
|
|
"grad_norm": 0.05859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.307952404022217,
|
|
"step": 834
|
|
},
|
|
{
|
|
"epoch": 0.06423076923076923,
|
|
"grad_norm": 0.054443359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3395590782165527,
|
|
"step": 835
|
|
},
|
|
{
|
|
"epoch": 0.0643076923076923,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.295976161956787,
|
|
"step": 836
|
|
},
|
|
{
|
|
"epoch": 0.06438461538461539,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2714438438415527,
|
|
"step": 837
|
|
},
|
|
{
|
|
"epoch": 0.06446153846153846,
|
|
"grad_norm": 0.0751953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2810120582580566,
|
|
"step": 838
|
|
},
|
|
{
|
|
"epoch": 0.06453846153846154,
|
|
"grad_norm": 0.0869140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2733139991760254,
|
|
"step": 839
|
|
},
|
|
{
|
|
"epoch": 0.06461538461538462,
|
|
"grad_norm": 0.06396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.332858085632324,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 0.06469230769230769,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29681658744812,
|
|
"step": 841
|
|
},
|
|
{
|
|
"epoch": 0.06476923076923077,
|
|
"grad_norm": 0.0556640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2928733825683594,
|
|
"step": 842
|
|
},
|
|
{
|
|
"epoch": 0.06484615384615385,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2736690044403076,
|
|
"step": 843
|
|
},
|
|
{
|
|
"epoch": 0.06492307692307692,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29630708694458,
|
|
"step": 844
|
|
},
|
|
{
|
|
"epoch": 0.065,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3279128074645996,
|
|
"step": 845
|
|
},
|
|
{
|
|
"epoch": 0.06507692307692307,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304863452911377,
|
|
"step": 846
|
|
},
|
|
{
|
|
"epoch": 0.06515384615384616,
|
|
"grad_norm": 0.06787109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2948756217956543,
|
|
"step": 847
|
|
},
|
|
{
|
|
"epoch": 0.06523076923076923,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2934534549713135,
|
|
"step": 848
|
|
},
|
|
{
|
|
"epoch": 0.0653076923076923,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3094704151153564,
|
|
"step": 849
|
|
},
|
|
{
|
|
"epoch": 0.06538461538461539,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3199169635772705,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.06546153846153846,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.316528558731079,
|
|
"step": 851
|
|
},
|
|
{
|
|
"epoch": 0.06553846153846153,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.314857244491577,
|
|
"step": 852
|
|
},
|
|
{
|
|
"epoch": 0.06561538461538462,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296419382095337,
|
|
"step": 853
|
|
},
|
|
{
|
|
"epoch": 0.06569230769230769,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299081563949585,
|
|
"step": 854
|
|
},
|
|
{
|
|
"epoch": 0.06576923076923077,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3324499130249023,
|
|
"step": 855
|
|
},
|
|
{
|
|
"epoch": 0.06584615384615385,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3335134983062744,
|
|
"step": 856
|
|
},
|
|
{
|
|
"epoch": 0.06592307692307692,
|
|
"grad_norm": 0.095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.317650079727173,
|
|
"step": 857
|
|
},
|
|
{
|
|
"epoch": 0.066,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2703654766082764,
|
|
"step": 858
|
|
},
|
|
{
|
|
"epoch": 0.06607692307692307,
|
|
"grad_norm": 0.06640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3029470443725586,
|
|
"step": 859
|
|
},
|
|
{
|
|
"epoch": 0.06615384615384616,
|
|
"grad_norm": 0.06494140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289628744125366,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 0.06623076923076923,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309727907180786,
|
|
"step": 861
|
|
},
|
|
{
|
|
"epoch": 0.0663076923076923,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3196537494659424,
|
|
"step": 862
|
|
},
|
|
{
|
|
"epoch": 0.06638461538461539,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2857706546783447,
|
|
"step": 863
|
|
},
|
|
{
|
|
"epoch": 0.06646153846153846,
|
|
"grad_norm": 0.0615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2825918197631836,
|
|
"step": 864
|
|
},
|
|
{
|
|
"epoch": 0.06653846153846153,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3174006938934326,
|
|
"step": 865
|
|
},
|
|
{
|
|
"epoch": 0.06661538461538462,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3245511054992676,
|
|
"step": 866
|
|
},
|
|
{
|
|
"epoch": 0.06669230769230769,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.332166910171509,
|
|
"step": 867
|
|
},
|
|
{
|
|
"epoch": 0.06676923076923078,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3035104274749756,
|
|
"step": 868
|
|
},
|
|
{
|
|
"epoch": 0.06684615384615385,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3174326419830322,
|
|
"step": 869
|
|
},
|
|
{
|
|
"epoch": 0.06692307692307692,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3395447731018066,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 0.067,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.324138641357422,
|
|
"step": 871
|
|
},
|
|
{
|
|
"epoch": 0.06707692307692308,
|
|
"grad_norm": 0.053955078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.326456069946289,
|
|
"step": 872
|
|
},
|
|
{
|
|
"epoch": 0.06715384615384615,
|
|
"grad_norm": 0.07080078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2988696098327637,
|
|
"step": 873
|
|
},
|
|
{
|
|
"epoch": 0.06723076923076923,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.259901285171509,
|
|
"step": 874
|
|
},
|
|
{
|
|
"epoch": 0.0673076923076923,
|
|
"grad_norm": 0.057861328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279069185256958,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 0.06738461538461539,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2892115116119385,
|
|
"step": 876
|
|
},
|
|
{
|
|
"epoch": 0.06746153846153846,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278977155685425,
|
|
"step": 877
|
|
},
|
|
{
|
|
"epoch": 0.06753846153846153,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2916462421417236,
|
|
"step": 878
|
|
},
|
|
{
|
|
"epoch": 0.06761538461538462,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29710054397583,
|
|
"step": 879
|
|
},
|
|
{
|
|
"epoch": 0.06769230769230769,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3256068229675293,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 0.06776923076923078,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3011467456817627,
|
|
"step": 881
|
|
},
|
|
{
|
|
"epoch": 0.06784615384615385,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2740840911865234,
|
|
"step": 882
|
|
},
|
|
{
|
|
"epoch": 0.06792307692307692,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26505708694458,
|
|
"step": 883
|
|
},
|
|
{
|
|
"epoch": 0.068,
|
|
"grad_norm": 0.06884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3060576915740967,
|
|
"step": 884
|
|
},
|
|
{
|
|
"epoch": 0.06807692307692308,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.303182363510132,
|
|
"step": 885
|
|
},
|
|
{
|
|
"epoch": 0.06815384615384615,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300964832305908,
|
|
"step": 886
|
|
},
|
|
{
|
|
"epoch": 0.06823076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290008068084717,
|
|
"step": 887
|
|
},
|
|
{
|
|
"epoch": 0.0683076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271760940551758,
|
|
"step": 888
|
|
},
|
|
{
|
|
"epoch": 0.06838461538461539,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3053879737854004,
|
|
"step": 889
|
|
},
|
|
{
|
|
"epoch": 0.06846153846153846,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309889554977417,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 0.06853846153846153,
|
|
"grad_norm": 0.059814453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3188300132751465,
|
|
"step": 891
|
|
},
|
|
{
|
|
"epoch": 0.06861538461538462,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2871434688568115,
|
|
"step": 892
|
|
},
|
|
{
|
|
"epoch": 0.06869230769230769,
|
|
"grad_norm": 0.057861328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28722882270813,
|
|
"step": 893
|
|
},
|
|
{
|
|
"epoch": 0.06876923076923076,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3059065341949463,
|
|
"step": 894
|
|
},
|
|
{
|
|
"epoch": 0.06884615384615385,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.316551685333252,
|
|
"step": 895
|
|
},
|
|
{
|
|
"epoch": 0.06892307692307692,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3162333965301514,
|
|
"step": 896
|
|
},
|
|
{
|
|
"epoch": 0.069,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287226676940918,
|
|
"step": 897
|
|
},
|
|
{
|
|
"epoch": 0.06907692307692308,
|
|
"grad_norm": 0.064453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3021342754364014,
|
|
"step": 898
|
|
},
|
|
{
|
|
"epoch": 0.06915384615384615,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.316467046737671,
|
|
"step": 899
|
|
},
|
|
{
|
|
"epoch": 0.06923076923076923,
|
|
"grad_norm": 0.072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300720453262329,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.0693076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3137147426605225,
|
|
"step": 901
|
|
},
|
|
{
|
|
"epoch": 0.06938461538461538,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2972307205200195,
|
|
"step": 902
|
|
},
|
|
{
|
|
"epoch": 0.06946153846153846,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290743350982666,
|
|
"step": 903
|
|
},
|
|
{
|
|
"epoch": 0.06953846153846154,
|
|
"grad_norm": 0.06884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.327735662460327,
|
|
"step": 904
|
|
},
|
|
{
|
|
"epoch": 0.06961538461538462,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.325681447982788,
|
|
"step": 905
|
|
},
|
|
{
|
|
"epoch": 0.06969230769230769,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3194191455841064,
|
|
"step": 906
|
|
},
|
|
{
|
|
"epoch": 0.06976923076923076,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.291700601577759,
|
|
"step": 907
|
|
},
|
|
{
|
|
"epoch": 0.06984615384615385,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3248202800750732,
|
|
"step": 908
|
|
},
|
|
{
|
|
"epoch": 0.06992307692307692,
|
|
"grad_norm": 0.062255859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3016631603240967,
|
|
"step": 909
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294468641281128,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 0.07007692307692308,
|
|
"grad_norm": 0.056640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.318389892578125,
|
|
"step": 911
|
|
},
|
|
{
|
|
"epoch": 0.07015384615384615,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2747318744659424,
|
|
"step": 912
|
|
},
|
|
{
|
|
"epoch": 0.07023076923076924,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2888641357421875,
|
|
"step": 913
|
|
},
|
|
{
|
|
"epoch": 0.07030769230769231,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2877848148345947,
|
|
"step": 914
|
|
},
|
|
{
|
|
"epoch": 0.07038461538461538,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2950026988983154,
|
|
"step": 915
|
|
},
|
|
{
|
|
"epoch": 0.07046153846153846,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3021254539489746,
|
|
"step": 916
|
|
},
|
|
{
|
|
"epoch": 0.07053846153846154,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2522494792938232,
|
|
"step": 917
|
|
},
|
|
{
|
|
"epoch": 0.07061538461538462,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267174005508423,
|
|
"step": 918
|
|
},
|
|
{
|
|
"epoch": 0.0706923076923077,
|
|
"grad_norm": 0.08740234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3412327766418457,
|
|
"step": 919
|
|
},
|
|
{
|
|
"epoch": 0.07076923076923076,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302635431289673,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 0.07084615384615385,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3009374141693115,
|
|
"step": 921
|
|
},
|
|
{
|
|
"epoch": 0.07092307692307692,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2577121257781982,
|
|
"step": 922
|
|
},
|
|
{
|
|
"epoch": 0.071,
|
|
"grad_norm": 0.056640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280217170715332,
|
|
"step": 923
|
|
},
|
|
{
|
|
"epoch": 0.07107692307692308,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.338879108428955,
|
|
"step": 924
|
|
},
|
|
{
|
|
"epoch": 0.07115384615384615,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3136560916900635,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 0.07123076923076924,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.318369150161743,
|
|
"step": 926
|
|
},
|
|
{
|
|
"epoch": 0.07130769230769231,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281522035598755,
|
|
"step": 927
|
|
},
|
|
{
|
|
"epoch": 0.07138461538461538,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269068956375122,
|
|
"step": 928
|
|
},
|
|
{
|
|
"epoch": 0.07146153846153847,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301034450531006,
|
|
"step": 929
|
|
},
|
|
{
|
|
"epoch": 0.07153846153846154,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304685115814209,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 0.07161538461538461,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272529125213623,
|
|
"step": 931
|
|
},
|
|
{
|
|
"epoch": 0.0716923076923077,
|
|
"grad_norm": 0.054443359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3107452392578125,
|
|
"step": 932
|
|
},
|
|
{
|
|
"epoch": 0.07176923076923077,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283876419067383,
|
|
"step": 933
|
|
},
|
|
{
|
|
"epoch": 0.07184615384615385,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3353426456451416,
|
|
"step": 934
|
|
},
|
|
{
|
|
"epoch": 0.07192307692307692,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3079395294189453,
|
|
"step": 935
|
|
},
|
|
{
|
|
"epoch": 0.072,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2812912464141846,
|
|
"step": 936
|
|
},
|
|
{
|
|
"epoch": 0.07207692307692308,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301910638809204,
|
|
"step": 937
|
|
},
|
|
{
|
|
"epoch": 0.07215384615384615,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2904052734375,
|
|
"step": 938
|
|
},
|
|
{
|
|
"epoch": 0.07223076923076924,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3040075302124023,
|
|
"step": 939
|
|
},
|
|
{
|
|
"epoch": 0.07230769230769231,
|
|
"grad_norm": 0.062255859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3262479305267334,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 0.07238461538461538,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298692464828491,
|
|
"step": 941
|
|
},
|
|
{
|
|
"epoch": 0.07246153846153847,
|
|
"grad_norm": 0.08837890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2825262546539307,
|
|
"step": 942
|
|
},
|
|
{
|
|
"epoch": 0.07253846153846154,
|
|
"grad_norm": 0.062255859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.310617685317993,
|
|
"step": 943
|
|
},
|
|
{
|
|
"epoch": 0.07261538461538461,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302220106124878,
|
|
"step": 944
|
|
},
|
|
{
|
|
"epoch": 0.0726923076923077,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.314629077911377,
|
|
"step": 945
|
|
},
|
|
{
|
|
"epoch": 0.07276923076923077,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2959587574005127,
|
|
"step": 946
|
|
},
|
|
{
|
|
"epoch": 0.07284615384615385,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279834270477295,
|
|
"step": 947
|
|
},
|
|
{
|
|
"epoch": 0.07292307692307692,
|
|
"grad_norm": 0.09326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.311004161834717,
|
|
"step": 948
|
|
},
|
|
{
|
|
"epoch": 0.073,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.322566032409668,
|
|
"step": 949
|
|
},
|
|
{
|
|
"epoch": 0.07307692307692308,
|
|
"grad_norm": 0.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3385963439941406,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.07315384615384615,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3293559551239014,
|
|
"step": 951
|
|
},
|
|
{
|
|
"epoch": 0.07323076923076922,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2633073329925537,
|
|
"step": 952
|
|
},
|
|
{
|
|
"epoch": 0.07330769230769231,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2970666885375977,
|
|
"step": 953
|
|
},
|
|
{
|
|
"epoch": 0.07338461538461538,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3071272373199463,
|
|
"step": 954
|
|
},
|
|
{
|
|
"epoch": 0.07346153846153847,
|
|
"grad_norm": 0.06494140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.332554340362549,
|
|
"step": 955
|
|
},
|
|
{
|
|
"epoch": 0.07353846153846154,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296982765197754,
|
|
"step": 956
|
|
},
|
|
{
|
|
"epoch": 0.07361538461538461,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2971534729003906,
|
|
"step": 957
|
|
},
|
|
{
|
|
"epoch": 0.0736923076923077,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.291956901550293,
|
|
"step": 958
|
|
},
|
|
{
|
|
"epoch": 0.07376923076923077,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285183906555176,
|
|
"step": 959
|
|
},
|
|
{
|
|
"epoch": 0.07384615384615385,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3221824169158936,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 0.07392307692307692,
|
|
"grad_norm": 0.0673828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2732675075531006,
|
|
"step": 961
|
|
},
|
|
{
|
|
"epoch": 0.074,
|
|
"grad_norm": 0.060546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3073689937591553,
|
|
"step": 962
|
|
},
|
|
{
|
|
"epoch": 0.07407692307692308,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270061731338501,
|
|
"step": 963
|
|
},
|
|
{
|
|
"epoch": 0.07415384615384615,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296302556991577,
|
|
"step": 964
|
|
},
|
|
{
|
|
"epoch": 0.07423076923076923,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.323195457458496,
|
|
"step": 965
|
|
},
|
|
{
|
|
"epoch": 0.07430769230769231,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.312408685684204,
|
|
"step": 966
|
|
},
|
|
{
|
|
"epoch": 0.07438461538461538,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.253072500228882,
|
|
"step": 967
|
|
},
|
|
{
|
|
"epoch": 0.07446153846153847,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280721664428711,
|
|
"step": 968
|
|
},
|
|
{
|
|
"epoch": 0.07453846153846154,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.291304349899292,
|
|
"step": 969
|
|
},
|
|
{
|
|
"epoch": 0.07461538461538461,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304917812347412,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 0.0746923076923077,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2871367931365967,
|
|
"step": 971
|
|
},
|
|
{
|
|
"epoch": 0.07476923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2903988361358643,
|
|
"step": 972
|
|
},
|
|
{
|
|
"epoch": 0.07484615384615384,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.297746419906616,
|
|
"step": 973
|
|
},
|
|
{
|
|
"epoch": 0.07492307692307693,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301907539367676,
|
|
"step": 974
|
|
},
|
|
{
|
|
"epoch": 0.075,
|
|
"grad_norm": 0.0791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.316561222076416,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 0.07507692307692308,
|
|
"grad_norm": 0.068359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281508445739746,
|
|
"step": 976
|
|
},
|
|
{
|
|
"epoch": 0.07515384615384615,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3163793087005615,
|
|
"step": 977
|
|
},
|
|
{
|
|
"epoch": 0.07523076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2741358280181885,
|
|
"step": 978
|
|
},
|
|
{
|
|
"epoch": 0.07530769230769231,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2786741256713867,
|
|
"step": 979
|
|
},
|
|
{
|
|
"epoch": 0.07538461538461538,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.310197353363037,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 0.07546153846153845,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3116824626922607,
|
|
"step": 981
|
|
},
|
|
{
|
|
"epoch": 0.07553846153846154,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2891335487365723,
|
|
"step": 982
|
|
},
|
|
{
|
|
"epoch": 0.07561538461538461,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3003146648406982,
|
|
"step": 983
|
|
},
|
|
{
|
|
"epoch": 0.0756923076923077,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3153605461120605,
|
|
"step": 984
|
|
},
|
|
{
|
|
"epoch": 0.07576923076923077,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3112599849700928,
|
|
"step": 985
|
|
},
|
|
{
|
|
"epoch": 0.07584615384615384,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2927730083465576,
|
|
"step": 986
|
|
},
|
|
{
|
|
"epoch": 0.07592307692307693,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287766218185425,
|
|
"step": 987
|
|
},
|
|
{
|
|
"epoch": 0.076,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301114082336426,
|
|
"step": 988
|
|
},
|
|
{
|
|
"epoch": 0.07607692307692308,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3123831748962402,
|
|
"step": 989
|
|
},
|
|
{
|
|
"epoch": 0.07615384615384616,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3312795162200928,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 0.07623076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260502815246582,
|
|
"step": 991
|
|
},
|
|
{
|
|
"epoch": 0.07630769230769231,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2809510231018066,
|
|
"step": 992
|
|
},
|
|
{
|
|
"epoch": 0.07638461538461538,
|
|
"grad_norm": 0.062255859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2654502391815186,
|
|
"step": 993
|
|
},
|
|
{
|
|
"epoch": 0.07646153846153846,
|
|
"grad_norm": 0.059326171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.310722827911377,
|
|
"step": 994
|
|
},
|
|
{
|
|
"epoch": 0.07653846153846154,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2980918884277344,
|
|
"step": 995
|
|
},
|
|
{
|
|
"epoch": 0.07661538461538461,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2813501358032227,
|
|
"step": 996
|
|
},
|
|
{
|
|
"epoch": 0.0766923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272542953491211,
|
|
"step": 997
|
|
},
|
|
{
|
|
"epoch": 0.07676923076923077,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302832841873169,
|
|
"step": 998
|
|
},
|
|
{
|
|
"epoch": 0.07684615384615384,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290461540222168,
|
|
"step": 999
|
|
},
|
|
{
|
|
"epoch": 0.07692307692307693,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298267364501953,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3101072311401367,
|
|
"step": 1001
|
|
},
|
|
{
|
|
"epoch": 0.07707692307692307,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2810518741607666,
|
|
"step": 1002
|
|
},
|
|
{
|
|
"epoch": 0.07715384615384616,
|
|
"grad_norm": 0.053955078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.322326898574829,
|
|
"step": 1003
|
|
},
|
|
{
|
|
"epoch": 0.07723076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2773804664611816,
|
|
"step": 1004
|
|
},
|
|
{
|
|
"epoch": 0.07730769230769231,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.312965154647827,
|
|
"step": 1005
|
|
},
|
|
{
|
|
"epoch": 0.07738461538461539,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281033515930176,
|
|
"step": 1006
|
|
},
|
|
{
|
|
"epoch": 0.07746153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2645020484924316,
|
|
"step": 1007
|
|
},
|
|
{
|
|
"epoch": 0.07753846153846154,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2907135486602783,
|
|
"step": 1008
|
|
},
|
|
{
|
|
"epoch": 0.07761538461538461,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3120787143707275,
|
|
"step": 1009
|
|
},
|
|
{
|
|
"epoch": 0.07769230769230769,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2961792945861816,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 0.07776923076923077,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2996582984924316,
|
|
"step": 1011
|
|
},
|
|
{
|
|
"epoch": 0.07784615384615384,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271503210067749,
|
|
"step": 1012
|
|
},
|
|
{
|
|
"epoch": 0.07792307692307693,
|
|
"grad_norm": 0.05810546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3171017169952393,
|
|
"step": 1013
|
|
},
|
|
{
|
|
"epoch": 0.078,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3152401447296143,
|
|
"step": 1014
|
|
},
|
|
{
|
|
"epoch": 0.07807692307692307,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2928383350372314,
|
|
"step": 1015
|
|
},
|
|
{
|
|
"epoch": 0.07815384615384616,
|
|
"grad_norm": 0.07275390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301924228668213,
|
|
"step": 1016
|
|
},
|
|
{
|
|
"epoch": 0.07823076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2766737937927246,
|
|
"step": 1017
|
|
},
|
|
{
|
|
"epoch": 0.07830769230769231,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2728281021118164,
|
|
"step": 1018
|
|
},
|
|
{
|
|
"epoch": 0.07838461538461539,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3059580326080322,
|
|
"step": 1019
|
|
},
|
|
{
|
|
"epoch": 0.07846153846153846,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2678041458129883,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 0.07853846153846154,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2974562644958496,
|
|
"step": 1021
|
|
},
|
|
{
|
|
"epoch": 0.07861538461538461,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2603352069854736,
|
|
"step": 1022
|
|
},
|
|
{
|
|
"epoch": 0.07869230769230769,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3144993782043457,
|
|
"step": 1023
|
|
},
|
|
{
|
|
"epoch": 0.07876923076923077,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304680109024048,
|
|
"step": 1024
|
|
},
|
|
{
|
|
"epoch": 0.07884615384615384,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272590398788452,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 0.07892307692307693,
|
|
"grad_norm": 0.0556640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2839744091033936,
|
|
"step": 1026
|
|
},
|
|
{
|
|
"epoch": 0.079,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289086103439331,
|
|
"step": 1027
|
|
},
|
|
{
|
|
"epoch": 0.07907692307692307,
|
|
"grad_norm": 0.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3061249256134033,
|
|
"step": 1028
|
|
},
|
|
{
|
|
"epoch": 0.07915384615384616,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309112310409546,
|
|
"step": 1029
|
|
},
|
|
{
|
|
"epoch": 0.07923076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298206329345703,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 0.0793076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3083884716033936,
|
|
"step": 1031
|
|
},
|
|
{
|
|
"epoch": 0.07938461538461539,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290154218673706,
|
|
"step": 1032
|
|
},
|
|
{
|
|
"epoch": 0.07946153846153846,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3301620483398438,
|
|
"step": 1033
|
|
},
|
|
{
|
|
"epoch": 0.07953846153846154,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.305955410003662,
|
|
"step": 1034
|
|
},
|
|
{
|
|
"epoch": 0.07961538461538462,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3127284049987793,
|
|
"step": 1035
|
|
},
|
|
{
|
|
"epoch": 0.07969230769230769,
|
|
"grad_norm": 0.054443359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2667782306671143,
|
|
"step": 1036
|
|
},
|
|
{
|
|
"epoch": 0.07976923076923077,
|
|
"grad_norm": 0.054443359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2877755165100098,
|
|
"step": 1037
|
|
},
|
|
{
|
|
"epoch": 0.07984615384615384,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300727367401123,
|
|
"step": 1038
|
|
},
|
|
{
|
|
"epoch": 0.07992307692307693,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.319269895553589,
|
|
"step": 1039
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"grad_norm": 0.057861328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2641561031341553,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 0.08007692307692307,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2780354022979736,
|
|
"step": 1041
|
|
},
|
|
{
|
|
"epoch": 0.08015384615384616,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2633230686187744,
|
|
"step": 1042
|
|
},
|
|
{
|
|
"epoch": 0.08023076923076923,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293245553970337,
|
|
"step": 1043
|
|
},
|
|
{
|
|
"epoch": 0.0803076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3000288009643555,
|
|
"step": 1044
|
|
},
|
|
{
|
|
"epoch": 0.08038461538461539,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2614498138427734,
|
|
"step": 1045
|
|
},
|
|
{
|
|
"epoch": 0.08046153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2788732051849365,
|
|
"step": 1046
|
|
},
|
|
{
|
|
"epoch": 0.08053846153846154,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267451047897339,
|
|
"step": 1047
|
|
},
|
|
{
|
|
"epoch": 0.08061538461538462,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.323251724243164,
|
|
"step": 1048
|
|
},
|
|
{
|
|
"epoch": 0.08069230769230769,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3116002082824707,
|
|
"step": 1049
|
|
},
|
|
{
|
|
"epoch": 0.08076923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27949595451355,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.08084615384615385,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285147190093994,
|
|
"step": 1051
|
|
},
|
|
{
|
|
"epoch": 0.08092307692307692,
|
|
"grad_norm": 0.059814453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3041603565216064,
|
|
"step": 1052
|
|
},
|
|
{
|
|
"epoch": 0.081,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309488534927368,
|
|
"step": 1053
|
|
},
|
|
{
|
|
"epoch": 0.08107692307692307,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3022758960723877,
|
|
"step": 1054
|
|
},
|
|
{
|
|
"epoch": 0.08115384615384616,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29506516456604,
|
|
"step": 1055
|
|
},
|
|
{
|
|
"epoch": 0.08123076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2633209228515625,
|
|
"step": 1056
|
|
},
|
|
{
|
|
"epoch": 0.0813076923076923,
|
|
"grad_norm": 0.057373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3223156929016113,
|
|
"step": 1057
|
|
},
|
|
{
|
|
"epoch": 0.08138461538461539,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3068323135375977,
|
|
"step": 1058
|
|
},
|
|
{
|
|
"epoch": 0.08146153846153846,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296396255493164,
|
|
"step": 1059
|
|
},
|
|
{
|
|
"epoch": 0.08153846153846153,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.262054204940796,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 0.08161538461538462,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2597947120666504,
|
|
"step": 1061
|
|
},
|
|
{
|
|
"epoch": 0.08169230769230769,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3057541847229004,
|
|
"step": 1062
|
|
},
|
|
{
|
|
"epoch": 0.08176923076923077,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299349546432495,
|
|
"step": 1063
|
|
},
|
|
{
|
|
"epoch": 0.08184615384615385,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3280625343322754,
|
|
"step": 1064
|
|
},
|
|
{
|
|
"epoch": 0.08192307692307692,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2907445430755615,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 0.082,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269679546356201,
|
|
"step": 1066
|
|
},
|
|
{
|
|
"epoch": 0.08207692307692308,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2925684452056885,
|
|
"step": 1067
|
|
},
|
|
{
|
|
"epoch": 0.08215384615384616,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304046869277954,
|
|
"step": 1068
|
|
},
|
|
{
|
|
"epoch": 0.08223076923076923,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296741008758545,
|
|
"step": 1069
|
|
},
|
|
{
|
|
"epoch": 0.0823076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2638139724731445,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 0.08238461538461539,
|
|
"grad_norm": 0.05615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281353235244751,
|
|
"step": 1071
|
|
},
|
|
{
|
|
"epoch": 0.08246153846153846,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2957816123962402,
|
|
"step": 1072
|
|
},
|
|
{
|
|
"epoch": 0.08253846153846153,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3074984550476074,
|
|
"step": 1073
|
|
},
|
|
{
|
|
"epoch": 0.08261538461538462,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2877724170684814,
|
|
"step": 1074
|
|
},
|
|
{
|
|
"epoch": 0.08269230769230769,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3082876205444336,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 0.08276923076923078,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2747561931610107,
|
|
"step": 1076
|
|
},
|
|
{
|
|
"epoch": 0.08284615384615385,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2904655933380127,
|
|
"step": 1077
|
|
},
|
|
{
|
|
"epoch": 0.08292307692307692,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298311710357666,
|
|
"step": 1078
|
|
},
|
|
{
|
|
"epoch": 0.083,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3248517513275146,
|
|
"step": 1079
|
|
},
|
|
{
|
|
"epoch": 0.08307692307692308,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2707276344299316,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 0.08315384615384615,
|
|
"grad_norm": 0.053955078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304731607437134,
|
|
"step": 1081
|
|
},
|
|
{
|
|
"epoch": 0.08323076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278836488723755,
|
|
"step": 1082
|
|
},
|
|
{
|
|
"epoch": 0.0833076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3161325454711914,
|
|
"step": 1083
|
|
},
|
|
{
|
|
"epoch": 0.08338461538461539,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2808938026428223,
|
|
"step": 1084
|
|
},
|
|
{
|
|
"epoch": 0.08346153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2673373222351074,
|
|
"step": 1085
|
|
},
|
|
{
|
|
"epoch": 0.08353846153846153,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2976694107055664,
|
|
"step": 1086
|
|
},
|
|
{
|
|
"epoch": 0.08361538461538462,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.312603712081909,
|
|
"step": 1087
|
|
},
|
|
{
|
|
"epoch": 0.08369230769230769,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.305006980895996,
|
|
"step": 1088
|
|
},
|
|
{
|
|
"epoch": 0.08376923076923076,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2964980602264404,
|
|
"step": 1089
|
|
},
|
|
{
|
|
"epoch": 0.08384615384615385,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2556395530700684,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 0.08392307692307692,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.312662363052368,
|
|
"step": 1091
|
|
},
|
|
{
|
|
"epoch": 0.084,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2630302906036377,
|
|
"step": 1092
|
|
},
|
|
{
|
|
"epoch": 0.08407692307692308,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3013787269592285,
|
|
"step": 1093
|
|
},
|
|
{
|
|
"epoch": 0.08415384615384615,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.277423620223999,
|
|
"step": 1094
|
|
},
|
|
{
|
|
"epoch": 0.08423076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274897336959839,
|
|
"step": 1095
|
|
},
|
|
{
|
|
"epoch": 0.0843076923076923,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2931325435638428,
|
|
"step": 1096
|
|
},
|
|
{
|
|
"epoch": 0.08438461538461539,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3233582973480225,
|
|
"step": 1097
|
|
},
|
|
{
|
|
"epoch": 0.08446153846153846,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286414623260498,
|
|
"step": 1098
|
|
},
|
|
{
|
|
"epoch": 0.08453846153846153,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2803916931152344,
|
|
"step": 1099
|
|
},
|
|
{
|
|
"epoch": 0.08461538461538462,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282362699508667,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.08469230769230769,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289262533187866,
|
|
"step": 1101
|
|
},
|
|
{
|
|
"epoch": 0.08476923076923076,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3200526237487793,
|
|
"step": 1102
|
|
},
|
|
{
|
|
"epoch": 0.08484615384615385,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280045986175537,
|
|
"step": 1103
|
|
},
|
|
{
|
|
"epoch": 0.08492307692307692,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2825357913970947,
|
|
"step": 1104
|
|
},
|
|
{
|
|
"epoch": 0.085,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2580947875976562,
|
|
"step": 1105
|
|
},
|
|
{
|
|
"epoch": 0.08507692307692308,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.318635940551758,
|
|
"step": 1106
|
|
},
|
|
{
|
|
"epoch": 0.08515384615384615,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270728826522827,
|
|
"step": 1107
|
|
},
|
|
{
|
|
"epoch": 0.08523076923076923,
|
|
"grad_norm": 0.07177734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2848944664001465,
|
|
"step": 1108
|
|
},
|
|
{
|
|
"epoch": 0.0853076923076923,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2848424911499023,
|
|
"step": 1109
|
|
},
|
|
{
|
|
"epoch": 0.08538461538461538,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2506837844848633,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 0.08546153846153846,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2977240085601807,
|
|
"step": 1111
|
|
},
|
|
{
|
|
"epoch": 0.08553846153846154,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2983500957489014,
|
|
"step": 1112
|
|
},
|
|
{
|
|
"epoch": 0.08561538461538462,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.262930154800415,
|
|
"step": 1113
|
|
},
|
|
{
|
|
"epoch": 0.08569230769230769,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27022385597229,
|
|
"step": 1114
|
|
},
|
|
{
|
|
"epoch": 0.08576923076923076,
|
|
"grad_norm": 0.064453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2647488117218018,
|
|
"step": 1115
|
|
},
|
|
{
|
|
"epoch": 0.08584615384615385,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26766300201416,
|
|
"step": 1116
|
|
},
|
|
{
|
|
"epoch": 0.08592307692307692,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2883410453796387,
|
|
"step": 1117
|
|
},
|
|
{
|
|
"epoch": 0.086,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2884602546691895,
|
|
"step": 1118
|
|
},
|
|
{
|
|
"epoch": 0.08607692307692308,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273092746734619,
|
|
"step": 1119
|
|
},
|
|
{
|
|
"epoch": 0.08615384615384615,
|
|
"grad_norm": 0.09375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267637014389038,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 0.08623076923076924,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3263449668884277,
|
|
"step": 1121
|
|
},
|
|
{
|
|
"epoch": 0.08630769230769231,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.315425395965576,
|
|
"step": 1122
|
|
},
|
|
{
|
|
"epoch": 0.08638461538461538,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269979476928711,
|
|
"step": 1123
|
|
},
|
|
{
|
|
"epoch": 0.08646153846153846,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280012369155884,
|
|
"step": 1124
|
|
},
|
|
{
|
|
"epoch": 0.08653846153846154,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276994228363037,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 0.08661538461538462,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3179986476898193,
|
|
"step": 1126
|
|
},
|
|
{
|
|
"epoch": 0.0866923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279191017150879,
|
|
"step": 1127
|
|
},
|
|
{
|
|
"epoch": 0.08676923076923077,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2809793949127197,
|
|
"step": 1128
|
|
},
|
|
{
|
|
"epoch": 0.08684615384615385,
|
|
"grad_norm": 0.060302734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2735235691070557,
|
|
"step": 1129
|
|
},
|
|
{
|
|
"epoch": 0.08692307692307692,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.288597345352173,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 0.087,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2967991828918457,
|
|
"step": 1131
|
|
},
|
|
{
|
|
"epoch": 0.08707692307692308,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3090128898620605,
|
|
"step": 1132
|
|
},
|
|
{
|
|
"epoch": 0.08715384615384615,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264988422393799,
|
|
"step": 1133
|
|
},
|
|
{
|
|
"epoch": 0.08723076923076924,
|
|
"grad_norm": 0.06640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2928619384765625,
|
|
"step": 1134
|
|
},
|
|
{
|
|
"epoch": 0.08730769230769231,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2968597412109375,
|
|
"step": 1135
|
|
},
|
|
{
|
|
"epoch": 0.08738461538461538,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.325543165206909,
|
|
"step": 1136
|
|
},
|
|
{
|
|
"epoch": 0.08746153846153847,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2998580932617188,
|
|
"step": 1137
|
|
},
|
|
{
|
|
"epoch": 0.08753846153846154,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296109914779663,
|
|
"step": 1138
|
|
},
|
|
{
|
|
"epoch": 0.08761538461538461,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2908480167388916,
|
|
"step": 1139
|
|
},
|
|
{
|
|
"epoch": 0.0876923076923077,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287315845489502,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 0.08776923076923077,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313964366912842,
|
|
"step": 1141
|
|
},
|
|
{
|
|
"epoch": 0.08784615384615385,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.305881977081299,
|
|
"step": 1142
|
|
},
|
|
{
|
|
"epoch": 0.08792307692307692,
|
|
"grad_norm": 0.06396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286377429962158,
|
|
"step": 1143
|
|
},
|
|
{
|
|
"epoch": 0.088,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260868787765503,
|
|
"step": 1144
|
|
},
|
|
{
|
|
"epoch": 0.08807692307692308,
|
|
"grad_norm": 0.08251953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278489828109741,
|
|
"step": 1145
|
|
},
|
|
{
|
|
"epoch": 0.08815384615384615,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279378652572632,
|
|
"step": 1146
|
|
},
|
|
{
|
|
"epoch": 0.08823076923076924,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2857887744903564,
|
|
"step": 1147
|
|
},
|
|
{
|
|
"epoch": 0.08830769230769231,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2911343574523926,
|
|
"step": 1148
|
|
},
|
|
{
|
|
"epoch": 0.08838461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.241732120513916,
|
|
"step": 1149
|
|
},
|
|
{
|
|
"epoch": 0.08846153846153847,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2709946632385254,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.08853846153846154,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.321845054626465,
|
|
"step": 1151
|
|
},
|
|
{
|
|
"epoch": 0.08861538461538461,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301917552947998,
|
|
"step": 1152
|
|
},
|
|
{
|
|
"epoch": 0.0886923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.288893699645996,
|
|
"step": 1153
|
|
},
|
|
{
|
|
"epoch": 0.08876923076923077,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2965023517608643,
|
|
"step": 1154
|
|
},
|
|
{
|
|
"epoch": 0.08884615384615385,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27234148979187,
|
|
"step": 1155
|
|
},
|
|
{
|
|
"epoch": 0.08892307692307692,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3088207244873047,
|
|
"step": 1156
|
|
},
|
|
{
|
|
"epoch": 0.089,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2988831996917725,
|
|
"step": 1157
|
|
},
|
|
{
|
|
"epoch": 0.08907692307692308,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2818374633789062,
|
|
"step": 1158
|
|
},
|
|
{
|
|
"epoch": 0.08915384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2730796337127686,
|
|
"step": 1159
|
|
},
|
|
{
|
|
"epoch": 0.08923076923076922,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2946746349334717,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 0.08930769230769231,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.328730583190918,
|
|
"step": 1161
|
|
},
|
|
{
|
|
"epoch": 0.08938461538461538,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283555746078491,
|
|
"step": 1162
|
|
},
|
|
{
|
|
"epoch": 0.08946153846153847,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2920055389404297,
|
|
"step": 1163
|
|
},
|
|
{
|
|
"epoch": 0.08953846153846154,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266439437866211,
|
|
"step": 1164
|
|
},
|
|
{
|
|
"epoch": 0.08961538461538461,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29899263381958,
|
|
"step": 1165
|
|
},
|
|
{
|
|
"epoch": 0.0896923076923077,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3058907985687256,
|
|
"step": 1166
|
|
},
|
|
{
|
|
"epoch": 0.08976923076923077,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304705858230591,
|
|
"step": 1167
|
|
},
|
|
{
|
|
"epoch": 0.08984615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276823043823242,
|
|
"step": 1168
|
|
},
|
|
{
|
|
"epoch": 0.08992307692307693,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265761137008667,
|
|
"step": 1169
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2932844161987305,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 0.09007692307692308,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2975382804870605,
|
|
"step": 1171
|
|
},
|
|
{
|
|
"epoch": 0.09015384615384615,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293301582336426,
|
|
"step": 1172
|
|
},
|
|
{
|
|
"epoch": 0.09023076923076923,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.307666301727295,
|
|
"step": 1173
|
|
},
|
|
{
|
|
"epoch": 0.09030769230769231,
|
|
"grad_norm": 0.057373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26143217086792,
|
|
"step": 1174
|
|
},
|
|
{
|
|
"epoch": 0.09038461538461538,
|
|
"grad_norm": 0.0693359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3002991676330566,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 0.09046153846153847,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289588212966919,
|
|
"step": 1176
|
|
},
|
|
{
|
|
"epoch": 0.09053846153846154,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2892026901245117,
|
|
"step": 1177
|
|
},
|
|
{
|
|
"epoch": 0.09061538461538461,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3153703212738037,
|
|
"step": 1178
|
|
},
|
|
{
|
|
"epoch": 0.0906923076923077,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2453489303588867,
|
|
"step": 1179
|
|
},
|
|
{
|
|
"epoch": 0.09076923076923077,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2859535217285156,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 0.09084615384615384,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3124125003814697,
|
|
"step": 1181
|
|
},
|
|
{
|
|
"epoch": 0.09092307692307693,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309913396835327,
|
|
"step": 1182
|
|
},
|
|
{
|
|
"epoch": 0.091,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2879085540771484,
|
|
"step": 1183
|
|
},
|
|
{
|
|
"epoch": 0.09107692307692308,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273616075515747,
|
|
"step": 1184
|
|
},
|
|
{
|
|
"epoch": 0.09115384615384615,
|
|
"grad_norm": 0.060791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279472827911377,
|
|
"step": 1185
|
|
},
|
|
{
|
|
"epoch": 0.09123076923076923,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289283514022827,
|
|
"step": 1186
|
|
},
|
|
{
|
|
"epoch": 0.09130769230769231,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3052752017974854,
|
|
"step": 1187
|
|
},
|
|
{
|
|
"epoch": 0.09138461538461538,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.321653366088867,
|
|
"step": 1188
|
|
},
|
|
{
|
|
"epoch": 0.09146153846153846,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2671780586242676,
|
|
"step": 1189
|
|
},
|
|
{
|
|
"epoch": 0.09153846153846154,
|
|
"grad_norm": 0.083984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2692716121673584,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 0.09161538461538461,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285646677017212,
|
|
"step": 1191
|
|
},
|
|
{
|
|
"epoch": 0.0916923076923077,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2806739807128906,
|
|
"step": 1192
|
|
},
|
|
{
|
|
"epoch": 0.09176923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313441753387451,
|
|
"step": 1193
|
|
},
|
|
{
|
|
"epoch": 0.09184615384615384,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.249051570892334,
|
|
"step": 1194
|
|
},
|
|
{
|
|
"epoch": 0.09192307692307693,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2691502571105957,
|
|
"step": 1195
|
|
},
|
|
{
|
|
"epoch": 0.092,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2777552604675293,
|
|
"step": 1196
|
|
},
|
|
{
|
|
"epoch": 0.09207692307692307,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2808332443237305,
|
|
"step": 1197
|
|
},
|
|
{
|
|
"epoch": 0.09215384615384616,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2748212814331055,
|
|
"step": 1198
|
|
},
|
|
{
|
|
"epoch": 0.09223076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3020434379577637,
|
|
"step": 1199
|
|
},
|
|
{
|
|
"epoch": 0.09230769230769231,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299642324447632,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.09238461538461538,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3027184009552,
|
|
"step": 1201
|
|
},
|
|
{
|
|
"epoch": 0.09246153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2934000492095947,
|
|
"step": 1202
|
|
},
|
|
{
|
|
"epoch": 0.09253846153846154,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3042309284210205,
|
|
"step": 1203
|
|
},
|
|
{
|
|
"epoch": 0.09261538461538461,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2785143852233887,
|
|
"step": 1204
|
|
},
|
|
{
|
|
"epoch": 0.0926923076923077,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.297877311706543,
|
|
"step": 1205
|
|
},
|
|
{
|
|
"epoch": 0.09276923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267688035964966,
|
|
"step": 1206
|
|
},
|
|
{
|
|
"epoch": 0.09284615384615384,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299767255783081,
|
|
"step": 1207
|
|
},
|
|
{
|
|
"epoch": 0.09292307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258415460586548,
|
|
"step": 1208
|
|
},
|
|
{
|
|
"epoch": 0.093,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2686636447906494,
|
|
"step": 1209
|
|
},
|
|
{
|
|
"epoch": 0.09307692307692307,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2934834957122803,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 0.09315384615384616,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2880213260650635,
|
|
"step": 1211
|
|
},
|
|
{
|
|
"epoch": 0.09323076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2660391330718994,
|
|
"step": 1212
|
|
},
|
|
{
|
|
"epoch": 0.09330769230769231,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27244234085083,
|
|
"step": 1213
|
|
},
|
|
{
|
|
"epoch": 0.09338461538461539,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.268784999847412,
|
|
"step": 1214
|
|
},
|
|
{
|
|
"epoch": 0.09346153846153846,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.288210153579712,
|
|
"step": 1215
|
|
},
|
|
{
|
|
"epoch": 0.09353846153846154,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2903499603271484,
|
|
"step": 1216
|
|
},
|
|
{
|
|
"epoch": 0.09361538461538461,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2972047328948975,
|
|
"step": 1217
|
|
},
|
|
{
|
|
"epoch": 0.09369230769230769,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2878780364990234,
|
|
"step": 1218
|
|
},
|
|
{
|
|
"epoch": 0.09376923076923077,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.288731098175049,
|
|
"step": 1219
|
|
},
|
|
{
|
|
"epoch": 0.09384615384615384,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.324207067489624,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 0.09392307692307693,
|
|
"grad_norm": 0.0693359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.309567928314209,
|
|
"step": 1221
|
|
},
|
|
{
|
|
"epoch": 0.094,
|
|
"grad_norm": 0.05908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2663397789001465,
|
|
"step": 1222
|
|
},
|
|
{
|
|
"epoch": 0.09407692307692307,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273252010345459,
|
|
"step": 1223
|
|
},
|
|
{
|
|
"epoch": 0.09415384615384616,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266460418701172,
|
|
"step": 1224
|
|
},
|
|
{
|
|
"epoch": 0.09423076923076923,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2710120677948,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 0.09430769230769231,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2897017002105713,
|
|
"step": 1226
|
|
},
|
|
{
|
|
"epoch": 0.09438461538461539,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.315492630004883,
|
|
"step": 1227
|
|
},
|
|
{
|
|
"epoch": 0.09446153846153846,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2808477878570557,
|
|
"step": 1228
|
|
},
|
|
{
|
|
"epoch": 0.09453846153846154,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2841110229492188,
|
|
"step": 1229
|
|
},
|
|
{
|
|
"epoch": 0.09461538461538462,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278153657913208,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 0.09469230769230769,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2819409370422363,
|
|
"step": 1231
|
|
},
|
|
{
|
|
"epoch": 0.09476923076923077,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3058526515960693,
|
|
"step": 1232
|
|
},
|
|
{
|
|
"epoch": 0.09484615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2852118015289307,
|
|
"step": 1233
|
|
},
|
|
{
|
|
"epoch": 0.09492307692307693,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2704591751098633,
|
|
"step": 1234
|
|
},
|
|
{
|
|
"epoch": 0.095,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2657039165496826,
|
|
"step": 1235
|
|
},
|
|
{
|
|
"epoch": 0.09507692307692307,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3157958984375,
|
|
"step": 1236
|
|
},
|
|
{
|
|
"epoch": 0.09515384615384616,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2732717990875244,
|
|
"step": 1237
|
|
},
|
|
{
|
|
"epoch": 0.09523076923076923,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269211769104004,
|
|
"step": 1238
|
|
},
|
|
{
|
|
"epoch": 0.0953076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265678644180298,
|
|
"step": 1239
|
|
},
|
|
{
|
|
"epoch": 0.09538461538461539,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300659656524658,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 0.09546153846153846,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2837026119232178,
|
|
"step": 1241
|
|
},
|
|
{
|
|
"epoch": 0.09553846153846154,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3083789348602295,
|
|
"step": 1242
|
|
},
|
|
{
|
|
"epoch": 0.09561538461538462,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301541566848755,
|
|
"step": 1243
|
|
},
|
|
{
|
|
"epoch": 0.09569230769230769,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292228937149048,
|
|
"step": 1244
|
|
},
|
|
{
|
|
"epoch": 0.09576923076923077,
|
|
"grad_norm": 0.06494140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304903030395508,
|
|
"step": 1245
|
|
},
|
|
{
|
|
"epoch": 0.09584615384615384,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2788262367248535,
|
|
"step": 1246
|
|
},
|
|
{
|
|
"epoch": 0.09592307692307692,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3141837120056152,
|
|
"step": 1247
|
|
},
|
|
{
|
|
"epoch": 0.096,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2924115657806396,
|
|
"step": 1248
|
|
},
|
|
{
|
|
"epoch": 0.09607692307692307,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290032386779785,
|
|
"step": 1249
|
|
},
|
|
{
|
|
"epoch": 0.09615384615384616,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28208327293396,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.09623076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.295142650604248,
|
|
"step": 1251
|
|
},
|
|
{
|
|
"epoch": 0.0963076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.308838367462158,
|
|
"step": 1252
|
|
},
|
|
{
|
|
"epoch": 0.09638461538461539,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2511932849884033,
|
|
"step": 1253
|
|
},
|
|
{
|
|
"epoch": 0.09646153846153846,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2687454223632812,
|
|
"step": 1254
|
|
},
|
|
{
|
|
"epoch": 0.09653846153846155,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3206777572631836,
|
|
"step": 1255
|
|
},
|
|
{
|
|
"epoch": 0.09661538461538462,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3087830543518066,
|
|
"step": 1256
|
|
},
|
|
{
|
|
"epoch": 0.09669230769230769,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2966227531433105,
|
|
"step": 1257
|
|
},
|
|
{
|
|
"epoch": 0.09676923076923077,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2704806327819824,
|
|
"step": 1258
|
|
},
|
|
{
|
|
"epoch": 0.09684615384615385,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2696475982666016,
|
|
"step": 1259
|
|
},
|
|
{
|
|
"epoch": 0.09692307692307692,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286139965057373,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 0.097,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2985122203826904,
|
|
"step": 1261
|
|
},
|
|
{
|
|
"epoch": 0.09707692307692307,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.284963607788086,
|
|
"step": 1262
|
|
},
|
|
{
|
|
"epoch": 0.09715384615384616,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2590343952178955,
|
|
"step": 1263
|
|
},
|
|
{
|
|
"epoch": 0.09723076923076923,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.249835729598999,
|
|
"step": 1264
|
|
},
|
|
{
|
|
"epoch": 0.0973076923076923,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.275681972503662,
|
|
"step": 1265
|
|
},
|
|
{
|
|
"epoch": 0.09738461538461539,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2853381633758545,
|
|
"step": 1266
|
|
},
|
|
{
|
|
"epoch": 0.09746153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2849013805389404,
|
|
"step": 1267
|
|
},
|
|
{
|
|
"epoch": 0.09753846153846153,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298964262008667,
|
|
"step": 1268
|
|
},
|
|
{
|
|
"epoch": 0.09761538461538462,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2910454273223877,
|
|
"step": 1269
|
|
},
|
|
{
|
|
"epoch": 0.09769230769230769,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2650763988494873,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 0.09776923076923077,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273305892944336,
|
|
"step": 1271
|
|
},
|
|
{
|
|
"epoch": 0.09784615384615385,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294053316116333,
|
|
"step": 1272
|
|
},
|
|
{
|
|
"epoch": 0.09792307692307692,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2761731147766113,
|
|
"step": 1273
|
|
},
|
|
{
|
|
"epoch": 0.098,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2782843112945557,
|
|
"step": 1274
|
|
},
|
|
{
|
|
"epoch": 0.09807692307692308,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300074815750122,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 0.09815384615384615,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285184860229492,
|
|
"step": 1276
|
|
},
|
|
{
|
|
"epoch": 0.09823076923076923,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.31616473197937,
|
|
"step": 1277
|
|
},
|
|
{
|
|
"epoch": 0.0983076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2815208435058594,
|
|
"step": 1278
|
|
},
|
|
{
|
|
"epoch": 0.09838461538461539,
|
|
"grad_norm": 0.056884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287889242172241,
|
|
"step": 1279
|
|
},
|
|
{
|
|
"epoch": 0.09846153846153846,
|
|
"grad_norm": 0.06884765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3188366889953613,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 0.09853846153846153,
|
|
"grad_norm": 0.059814453125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2763428688049316,
|
|
"step": 1281
|
|
},
|
|
{
|
|
"epoch": 0.09861538461538462,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2720305919647217,
|
|
"step": 1282
|
|
},
|
|
{
|
|
"epoch": 0.09869230769230769,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2689857482910156,
|
|
"step": 1283
|
|
},
|
|
{
|
|
"epoch": 0.09876923076923078,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2800867557525635,
|
|
"step": 1284
|
|
},
|
|
{
|
|
"epoch": 0.09884615384615385,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2887043952941895,
|
|
"step": 1285
|
|
},
|
|
{
|
|
"epoch": 0.09892307692307692,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270484209060669,
|
|
"step": 1286
|
|
},
|
|
{
|
|
"epoch": 0.099,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.248594284057617,
|
|
"step": 1287
|
|
},
|
|
{
|
|
"epoch": 0.09907692307692308,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265702247619629,
|
|
"step": 1288
|
|
},
|
|
{
|
|
"epoch": 0.09915384615384615,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260300636291504,
|
|
"step": 1289
|
|
},
|
|
{
|
|
"epoch": 0.09923076923076923,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2642955780029297,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 0.0993076923076923,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3012382984161377,
|
|
"step": 1291
|
|
},
|
|
{
|
|
"epoch": 0.09938461538461539,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2843809127807617,
|
|
"step": 1292
|
|
},
|
|
{
|
|
"epoch": 0.09946153846153846,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2684121131896973,
|
|
"step": 1293
|
|
},
|
|
{
|
|
"epoch": 0.09953846153846153,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2633073329925537,
|
|
"step": 1294
|
|
},
|
|
{
|
|
"epoch": 0.09961538461538462,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2599241733551025,
|
|
"step": 1295
|
|
},
|
|
{
|
|
"epoch": 0.09969230769230769,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.325059413909912,
|
|
"step": 1296
|
|
},
|
|
{
|
|
"epoch": 0.09976923076923076,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3176729679107666,
|
|
"step": 1297
|
|
},
|
|
{
|
|
"epoch": 0.09984615384615385,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2927942276000977,
|
|
"step": 1298
|
|
},
|
|
{
|
|
"epoch": 0.09992307692307692,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280928611755371,
|
|
"step": 1299
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3054182529449463,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"eval_loss": 1.946685552597046,
|
|
"eval_runtime": 762.7599,
|
|
"eval_samples_per_second": 24.925,
|
|
"eval_steps_per_second": 6.231,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.10007692307692308,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30716609954834,
|
|
"step": 1301
|
|
},
|
|
{
|
|
"epoch": 0.10015384615384615,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270382881164551,
|
|
"step": 1302
|
|
},
|
|
{
|
|
"epoch": 0.10023076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2984094619750977,
|
|
"step": 1303
|
|
},
|
|
{
|
|
"epoch": 0.1003076923076923,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3063979148864746,
|
|
"step": 1304
|
|
},
|
|
{
|
|
"epoch": 0.10038461538461538,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283358335494995,
|
|
"step": 1305
|
|
},
|
|
{
|
|
"epoch": 0.10046153846153846,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2748124599456787,
|
|
"step": 1306
|
|
},
|
|
{
|
|
"epoch": 0.10053846153846153,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2953109741210938,
|
|
"step": 1307
|
|
},
|
|
{
|
|
"epoch": 0.10061538461538462,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2948238849639893,
|
|
"step": 1308
|
|
},
|
|
{
|
|
"epoch": 0.10069230769230769,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28383207321167,
|
|
"step": 1309
|
|
},
|
|
{
|
|
"epoch": 0.10076923076923076,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2949390411376953,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 0.10084615384615385,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2710659503936768,
|
|
"step": 1311
|
|
},
|
|
{
|
|
"epoch": 0.10092307692307692,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2791402339935303,
|
|
"step": 1312
|
|
},
|
|
{
|
|
"epoch": 0.101,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2703442573547363,
|
|
"step": 1313
|
|
},
|
|
{
|
|
"epoch": 0.10107692307692308,
|
|
"grad_norm": 0.05908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2775673866271973,
|
|
"step": 1314
|
|
},
|
|
{
|
|
"epoch": 0.10115384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.247141122817993,
|
|
"step": 1315
|
|
},
|
|
{
|
|
"epoch": 0.10123076923076924,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.247110605239868,
|
|
"step": 1316
|
|
},
|
|
{
|
|
"epoch": 0.1013076923076923,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2949914932250977,
|
|
"step": 1317
|
|
},
|
|
{
|
|
"epoch": 0.10138461538461538,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2976455688476562,
|
|
"step": 1318
|
|
},
|
|
{
|
|
"epoch": 0.10146153846153846,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2871017456054688,
|
|
"step": 1319
|
|
},
|
|
{
|
|
"epoch": 0.10153846153846154,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2629425525665283,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 0.10161538461538462,
|
|
"grad_norm": 0.060791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.308692455291748,
|
|
"step": 1321
|
|
},
|
|
{
|
|
"epoch": 0.10169230769230769,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285892963409424,
|
|
"step": 1322
|
|
},
|
|
{
|
|
"epoch": 0.10176923076923076,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.311063528060913,
|
|
"step": 1323
|
|
},
|
|
{
|
|
"epoch": 0.10184615384615385,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2664830684661865,
|
|
"step": 1324
|
|
},
|
|
{
|
|
"epoch": 0.10192307692307692,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2596077919006348,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 0.102,
|
|
"grad_norm": 0.060546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.297745943069458,
|
|
"step": 1326
|
|
},
|
|
{
|
|
"epoch": 0.10207692307692308,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3122897148132324,
|
|
"step": 1327
|
|
},
|
|
{
|
|
"epoch": 0.10215384615384615,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304755449295044,
|
|
"step": 1328
|
|
},
|
|
{
|
|
"epoch": 0.10223076923076924,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276322364807129,
|
|
"step": 1329
|
|
},
|
|
{
|
|
"epoch": 0.10230769230769231,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2992680072784424,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 0.10238461538461538,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2786922454833984,
|
|
"step": 1331
|
|
},
|
|
{
|
|
"epoch": 0.10246153846153847,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2944834232330322,
|
|
"step": 1332
|
|
},
|
|
{
|
|
"epoch": 0.10253846153846154,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.291032314300537,
|
|
"step": 1333
|
|
},
|
|
{
|
|
"epoch": 0.10261538461538462,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289757013320923,
|
|
"step": 1334
|
|
},
|
|
{
|
|
"epoch": 0.1026923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299837589263916,
|
|
"step": 1335
|
|
},
|
|
{
|
|
"epoch": 0.10276923076923077,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313868522644043,
|
|
"step": 1336
|
|
},
|
|
{
|
|
"epoch": 0.10284615384615385,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274547815322876,
|
|
"step": 1337
|
|
},
|
|
{
|
|
"epoch": 0.10292307692307692,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3073570728302,
|
|
"step": 1338
|
|
},
|
|
{
|
|
"epoch": 0.103,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.284780740737915,
|
|
"step": 1339
|
|
},
|
|
{
|
|
"epoch": 0.10307692307692308,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.261195421218872,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 0.10315384615384615,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2715954780578613,
|
|
"step": 1341
|
|
},
|
|
{
|
|
"epoch": 0.10323076923076924,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2876977920532227,
|
|
"step": 1342
|
|
},
|
|
{
|
|
"epoch": 0.10330769230769231,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2772912979125977,
|
|
"step": 1343
|
|
},
|
|
{
|
|
"epoch": 0.10338461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287768602371216,
|
|
"step": 1344
|
|
},
|
|
{
|
|
"epoch": 0.10346153846153847,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.222832441329956,
|
|
"step": 1345
|
|
},
|
|
{
|
|
"epoch": 0.10353846153846154,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2838244438171387,
|
|
"step": 1346
|
|
},
|
|
{
|
|
"epoch": 0.10361538461538461,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2957444190979004,
|
|
"step": 1347
|
|
},
|
|
{
|
|
"epoch": 0.1036923076923077,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.305013418197632,
|
|
"step": 1348
|
|
},
|
|
{
|
|
"epoch": 0.10376923076923077,
|
|
"grad_norm": 0.060791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293652057647705,
|
|
"step": 1349
|
|
},
|
|
{
|
|
"epoch": 0.10384615384615385,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264406204223633,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.10392307692307692,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2855851650238037,
|
|
"step": 1351
|
|
},
|
|
{
|
|
"epoch": 0.104,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296412467956543,
|
|
"step": 1352
|
|
},
|
|
{
|
|
"epoch": 0.10407692307692308,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3006083965301514,
|
|
"step": 1353
|
|
},
|
|
{
|
|
"epoch": 0.10415384615384615,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2769317626953125,
|
|
"step": 1354
|
|
},
|
|
{
|
|
"epoch": 0.10423076923076922,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2655787467956543,
|
|
"step": 1355
|
|
},
|
|
{
|
|
"epoch": 0.10430769230769231,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287201166152954,
|
|
"step": 1356
|
|
},
|
|
{
|
|
"epoch": 0.10438461538461538,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2740657329559326,
|
|
"step": 1357
|
|
},
|
|
{
|
|
"epoch": 0.10446153846153847,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2906572818756104,
|
|
"step": 1358
|
|
},
|
|
{
|
|
"epoch": 0.10453846153846154,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2772610187530518,
|
|
"step": 1359
|
|
},
|
|
{
|
|
"epoch": 0.10461538461538461,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2385993003845215,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 0.1046923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281891345977783,
|
|
"step": 1361
|
|
},
|
|
{
|
|
"epoch": 0.10476923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.32733154296875,
|
|
"step": 1362
|
|
},
|
|
{
|
|
"epoch": 0.10484615384615385,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2746388912200928,
|
|
"step": 1363
|
|
},
|
|
{
|
|
"epoch": 0.10492307692307692,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3054635524749756,
|
|
"step": 1364
|
|
},
|
|
{
|
|
"epoch": 0.105,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290755033493042,
|
|
"step": 1365
|
|
},
|
|
{
|
|
"epoch": 0.10507692307692308,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274430990219116,
|
|
"step": 1366
|
|
},
|
|
{
|
|
"epoch": 0.10515384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2832305431365967,
|
|
"step": 1367
|
|
},
|
|
{
|
|
"epoch": 0.10523076923076922,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2990829944610596,
|
|
"step": 1368
|
|
},
|
|
{
|
|
"epoch": 0.10530769230769231,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2982945442199707,
|
|
"step": 1369
|
|
},
|
|
{
|
|
"epoch": 0.10538461538461538,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2940561771392822,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 0.10546153846153847,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266427993774414,
|
|
"step": 1371
|
|
},
|
|
{
|
|
"epoch": 0.10553846153846154,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2973361015319824,
|
|
"step": 1372
|
|
},
|
|
{
|
|
"epoch": 0.10561538461538461,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2848243713378906,
|
|
"step": 1373
|
|
},
|
|
{
|
|
"epoch": 0.1056923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28057599067688,
|
|
"step": 1374
|
|
},
|
|
{
|
|
"epoch": 0.10576923076923077,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2632758617401123,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 0.10584615384615384,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30037260055542,
|
|
"step": 1376
|
|
},
|
|
{
|
|
"epoch": 0.10592307692307693,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269190549850464,
|
|
"step": 1377
|
|
},
|
|
{
|
|
"epoch": 0.106,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2729926109313965,
|
|
"step": 1378
|
|
},
|
|
{
|
|
"epoch": 0.10607692307692308,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30131459236145,
|
|
"step": 1379
|
|
},
|
|
{
|
|
"epoch": 0.10615384615384615,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2575595378875732,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 0.10623076923076923,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2911674976348877,
|
|
"step": 1381
|
|
},
|
|
{
|
|
"epoch": 0.10630769230769231,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2789673805236816,
|
|
"step": 1382
|
|
},
|
|
{
|
|
"epoch": 0.10638461538461538,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3095202445983887,
|
|
"step": 1383
|
|
},
|
|
{
|
|
"epoch": 0.10646153846153845,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2416727542877197,
|
|
"step": 1384
|
|
},
|
|
{
|
|
"epoch": 0.10653846153846154,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2462611198425293,
|
|
"step": 1385
|
|
},
|
|
{
|
|
"epoch": 0.10661538461538461,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.314155101776123,
|
|
"step": 1386
|
|
},
|
|
{
|
|
"epoch": 0.1066923076923077,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.306645393371582,
|
|
"step": 1387
|
|
},
|
|
{
|
|
"epoch": 0.10676923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2926292419433594,
|
|
"step": 1388
|
|
},
|
|
{
|
|
"epoch": 0.10684615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2843070030212402,
|
|
"step": 1389
|
|
},
|
|
{
|
|
"epoch": 0.10692307692307693,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289640188217163,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 0.107,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276277780532837,
|
|
"step": 1391
|
|
},
|
|
{
|
|
"epoch": 0.10707692307692308,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3010096549987793,
|
|
"step": 1392
|
|
},
|
|
{
|
|
"epoch": 0.10715384615384616,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267055034637451,
|
|
"step": 1393
|
|
},
|
|
{
|
|
"epoch": 0.10723076923076923,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.314847230911255,
|
|
"step": 1394
|
|
},
|
|
{
|
|
"epoch": 0.10730769230769231,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2732081413269043,
|
|
"step": 1395
|
|
},
|
|
{
|
|
"epoch": 0.10738461538461538,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264951705932617,
|
|
"step": 1396
|
|
},
|
|
{
|
|
"epoch": 0.10746153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2896957397460938,
|
|
"step": 1397
|
|
},
|
|
{
|
|
"epoch": 0.10753846153846154,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2945103645324707,
|
|
"step": 1398
|
|
},
|
|
{
|
|
"epoch": 0.10761538461538461,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266144037246704,
|
|
"step": 1399
|
|
},
|
|
{
|
|
"epoch": 0.1076923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264249801635742,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.10776923076923077,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264883041381836,
|
|
"step": 1401
|
|
},
|
|
{
|
|
"epoch": 0.10784615384615384,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290191411972046,
|
|
"step": 1402
|
|
},
|
|
{
|
|
"epoch": 0.10792307692307693,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269263982772827,
|
|
"step": 1403
|
|
},
|
|
{
|
|
"epoch": 0.108,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2436695098876953,
|
|
"step": 1404
|
|
},
|
|
{
|
|
"epoch": 0.10807692307692307,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2692906856536865,
|
|
"step": 1405
|
|
},
|
|
{
|
|
"epoch": 0.10815384615384616,
|
|
"grad_norm": 0.054443359375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2629220485687256,
|
|
"step": 1406
|
|
},
|
|
{
|
|
"epoch": 0.10823076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3091468811035156,
|
|
"step": 1407
|
|
},
|
|
{
|
|
"epoch": 0.10830769230769231,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2767493724823,
|
|
"step": 1408
|
|
},
|
|
{
|
|
"epoch": 0.10838461538461538,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272688627243042,
|
|
"step": 1409
|
|
},
|
|
{
|
|
"epoch": 0.10846153846153846,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280780076980591,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 0.10853846153846154,
|
|
"grad_norm": 0.0595703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283330202102661,
|
|
"step": 1411
|
|
},
|
|
{
|
|
"epoch": 0.10861538461538461,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30291485786438,
|
|
"step": 1412
|
|
},
|
|
{
|
|
"epoch": 0.1086923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30545973777771,
|
|
"step": 1413
|
|
},
|
|
{
|
|
"epoch": 0.10876923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.245992422103882,
|
|
"step": 1414
|
|
},
|
|
{
|
|
"epoch": 0.10884615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.22948956489563,
|
|
"step": 1415
|
|
},
|
|
{
|
|
"epoch": 0.10892307692307693,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.303201675415039,
|
|
"step": 1416
|
|
},
|
|
{
|
|
"epoch": 0.109,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2701869010925293,
|
|
"step": 1417
|
|
},
|
|
{
|
|
"epoch": 0.10907692307692307,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2969796657562256,
|
|
"step": 1418
|
|
},
|
|
{
|
|
"epoch": 0.10915384615384616,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280252695083618,
|
|
"step": 1419
|
|
},
|
|
{
|
|
"epoch": 0.10923076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265605926513672,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 0.10930769230769231,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2656404972076416,
|
|
"step": 1421
|
|
},
|
|
{
|
|
"epoch": 0.10938461538461539,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290250778198242,
|
|
"step": 1422
|
|
},
|
|
{
|
|
"epoch": 0.10946153846153846,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294367551803589,
|
|
"step": 1423
|
|
},
|
|
{
|
|
"epoch": 0.10953846153846154,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.257459878921509,
|
|
"step": 1424
|
|
},
|
|
{
|
|
"epoch": 0.10961538461538461,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269968271255493,
|
|
"step": 1425
|
|
},
|
|
{
|
|
"epoch": 0.10969230769230769,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3199453353881836,
|
|
"step": 1426
|
|
},
|
|
{
|
|
"epoch": 0.10976923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3053741455078125,
|
|
"step": 1427
|
|
},
|
|
{
|
|
"epoch": 0.10984615384615384,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2965688705444336,
|
|
"step": 1428
|
|
},
|
|
{
|
|
"epoch": 0.10992307692307693,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2699058055877686,
|
|
"step": 1429
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2566442489624023,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 0.11007692307692307,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286675453186035,
|
|
"step": 1431
|
|
},
|
|
{
|
|
"epoch": 0.11015384615384616,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304194688796997,
|
|
"step": 1432
|
|
},
|
|
{
|
|
"epoch": 0.11023076923076923,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300928831100464,
|
|
"step": 1433
|
|
},
|
|
{
|
|
"epoch": 0.1103076923076923,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298708438873291,
|
|
"step": 1434
|
|
},
|
|
{
|
|
"epoch": 0.11038461538461539,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2713160514831543,
|
|
"step": 1435
|
|
},
|
|
{
|
|
"epoch": 0.11046153846153846,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273848533630371,
|
|
"step": 1436
|
|
},
|
|
{
|
|
"epoch": 0.11053846153846154,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2776575088500977,
|
|
"step": 1437
|
|
},
|
|
{
|
|
"epoch": 0.11061538461538462,
|
|
"grad_norm": 0.06982421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2810027599334717,
|
|
"step": 1438
|
|
},
|
|
{
|
|
"epoch": 0.11069230769230769,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2893192768096924,
|
|
"step": 1439
|
|
},
|
|
{
|
|
"epoch": 0.11076923076923077,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2713875770568848,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 0.11084615384615384,
|
|
"grad_norm": 0.057373046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271865129470825,
|
|
"step": 1441
|
|
},
|
|
{
|
|
"epoch": 0.11092307692307693,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281644105911255,
|
|
"step": 1442
|
|
},
|
|
{
|
|
"epoch": 0.111,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3070154190063477,
|
|
"step": 1443
|
|
},
|
|
{
|
|
"epoch": 0.11107692307692307,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281196117401123,
|
|
"step": 1444
|
|
},
|
|
{
|
|
"epoch": 0.11115384615384616,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2554454803466797,
|
|
"step": 1445
|
|
},
|
|
{
|
|
"epoch": 0.11123076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283013105392456,
|
|
"step": 1446
|
|
},
|
|
{
|
|
"epoch": 0.1113076923076923,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302604913711548,
|
|
"step": 1447
|
|
},
|
|
{
|
|
"epoch": 0.11138461538461539,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3334763050079346,
|
|
"step": 1448
|
|
},
|
|
{
|
|
"epoch": 0.11146153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2651431560516357,
|
|
"step": 1449
|
|
},
|
|
{
|
|
"epoch": 0.11153846153846154,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269610643386841,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.11161538461538462,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274177074432373,
|
|
"step": 1451
|
|
},
|
|
{
|
|
"epoch": 0.11169230769230769,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2737534046173096,
|
|
"step": 1452
|
|
},
|
|
{
|
|
"epoch": 0.11176923076923077,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2951645851135254,
|
|
"step": 1453
|
|
},
|
|
{
|
|
"epoch": 0.11184615384615385,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2854413986206055,
|
|
"step": 1454
|
|
},
|
|
{
|
|
"epoch": 0.11192307692307692,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2611136436462402,
|
|
"step": 1455
|
|
},
|
|
{
|
|
"epoch": 0.112,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3067190647125244,
|
|
"step": 1456
|
|
},
|
|
{
|
|
"epoch": 0.11207692307692307,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2861433029174805,
|
|
"step": 1457
|
|
},
|
|
{
|
|
"epoch": 0.11215384615384616,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2706847190856934,
|
|
"step": 1458
|
|
},
|
|
{
|
|
"epoch": 0.11223076923076923,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292670249938965,
|
|
"step": 1459
|
|
},
|
|
{
|
|
"epoch": 0.1123076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2997994422912598,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 0.11238461538461539,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3197712898254395,
|
|
"step": 1461
|
|
},
|
|
{
|
|
"epoch": 0.11246153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2923693656921387,
|
|
"step": 1462
|
|
},
|
|
{
|
|
"epoch": 0.11253846153846153,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2927136421203613,
|
|
"step": 1463
|
|
},
|
|
{
|
|
"epoch": 0.11261538461538462,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2628865242004395,
|
|
"step": 1464
|
|
},
|
|
{
|
|
"epoch": 0.11269230769230769,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2685630321502686,
|
|
"step": 1465
|
|
},
|
|
{
|
|
"epoch": 0.11276923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2840230464935303,
|
|
"step": 1466
|
|
},
|
|
{
|
|
"epoch": 0.11284615384615385,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2555501461029053,
|
|
"step": 1467
|
|
},
|
|
{
|
|
"epoch": 0.11292307692307692,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.262644052505493,
|
|
"step": 1468
|
|
},
|
|
{
|
|
"epoch": 0.113,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270575523376465,
|
|
"step": 1469
|
|
},
|
|
{
|
|
"epoch": 0.11307692307692307,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27679181098938,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 0.11315384615384616,
|
|
"grad_norm": 0.0546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283435106277466,
|
|
"step": 1471
|
|
},
|
|
{
|
|
"epoch": 0.11323076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269765853881836,
|
|
"step": 1472
|
|
},
|
|
{
|
|
"epoch": 0.1133076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2577569484710693,
|
|
"step": 1473
|
|
},
|
|
{
|
|
"epoch": 0.11338461538461539,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304492235183716,
|
|
"step": 1474
|
|
},
|
|
{
|
|
"epoch": 0.11346153846153846,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2700109481811523,
|
|
"step": 1475
|
|
},
|
|
{
|
|
"epoch": 0.11353846153846153,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.297964572906494,
|
|
"step": 1476
|
|
},
|
|
{
|
|
"epoch": 0.11361538461538462,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282289743423462,
|
|
"step": 1477
|
|
},
|
|
{
|
|
"epoch": 0.11369230769230769,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3232014179229736,
|
|
"step": 1478
|
|
},
|
|
{
|
|
"epoch": 0.11376923076923078,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278773784637451,
|
|
"step": 1479
|
|
},
|
|
{
|
|
"epoch": 0.11384615384615385,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2720518112182617,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 0.11392307692307692,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2650418281555176,
|
|
"step": 1481
|
|
},
|
|
{
|
|
"epoch": 0.114,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.291524648666382,
|
|
"step": 1482
|
|
},
|
|
{
|
|
"epoch": 0.11407692307692308,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2733776569366455,
|
|
"step": 1483
|
|
},
|
|
{
|
|
"epoch": 0.11415384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2583673000335693,
|
|
"step": 1484
|
|
},
|
|
{
|
|
"epoch": 0.11423076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263597011566162,
|
|
"step": 1485
|
|
},
|
|
{
|
|
"epoch": 0.1143076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2798824310302734,
|
|
"step": 1486
|
|
},
|
|
{
|
|
"epoch": 0.11438461538461539,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3151471614837646,
|
|
"step": 1487
|
|
},
|
|
{
|
|
"epoch": 0.11446153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30068039894104,
|
|
"step": 1488
|
|
},
|
|
{
|
|
"epoch": 0.11453846153846153,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3293325901031494,
|
|
"step": 1489
|
|
},
|
|
{
|
|
"epoch": 0.11461538461538462,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2550833225250244,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 0.11469230769230769,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2640793323516846,
|
|
"step": 1491
|
|
},
|
|
{
|
|
"epoch": 0.11476923076923078,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265991687774658,
|
|
"step": 1492
|
|
},
|
|
{
|
|
"epoch": 0.11484615384615385,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2890121936798096,
|
|
"step": 1493
|
|
},
|
|
{
|
|
"epoch": 0.11492307692307692,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2937066555023193,
|
|
"step": 1494
|
|
},
|
|
{
|
|
"epoch": 0.115,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2715957164764404,
|
|
"step": 1495
|
|
},
|
|
{
|
|
"epoch": 0.11507692307692308,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28102707862854,
|
|
"step": 1496
|
|
},
|
|
{
|
|
"epoch": 0.11515384615384615,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2754833698272705,
|
|
"step": 1497
|
|
},
|
|
{
|
|
"epoch": 0.11523076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2781403064727783,
|
|
"step": 1498
|
|
},
|
|
{
|
|
"epoch": 0.1153076923076923,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2614293098449707,
|
|
"step": 1499
|
|
},
|
|
{
|
|
"epoch": 0.11538461538461539,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260953664779663,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.11546153846153846,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3188295364379883,
|
|
"step": 1501
|
|
},
|
|
{
|
|
"epoch": 0.11553846153846153,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2778615951538086,
|
|
"step": 1502
|
|
},
|
|
{
|
|
"epoch": 0.11561538461538462,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.312237024307251,
|
|
"step": 1503
|
|
},
|
|
{
|
|
"epoch": 0.11569230769230769,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282717227935791,
|
|
"step": 1504
|
|
},
|
|
{
|
|
"epoch": 0.11576923076923076,
|
|
"grad_norm": 0.061279296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.243504524230957,
|
|
"step": 1505
|
|
},
|
|
{
|
|
"epoch": 0.11584615384615385,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271209478378296,
|
|
"step": 1506
|
|
},
|
|
{
|
|
"epoch": 0.11592307692307692,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302647590637207,
|
|
"step": 1507
|
|
},
|
|
{
|
|
"epoch": 0.116,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.253922462463379,
|
|
"step": 1508
|
|
},
|
|
{
|
|
"epoch": 0.11607692307692308,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2622570991516113,
|
|
"step": 1509
|
|
},
|
|
{
|
|
"epoch": 0.11615384615384615,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2678589820861816,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 0.11623076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2822067737579346,
|
|
"step": 1511
|
|
},
|
|
{
|
|
"epoch": 0.1163076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.284210681915283,
|
|
"step": 1512
|
|
},
|
|
{
|
|
"epoch": 0.11638461538461538,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.288612127304077,
|
|
"step": 1513
|
|
},
|
|
{
|
|
"epoch": 0.11646153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296241044998169,
|
|
"step": 1514
|
|
},
|
|
{
|
|
"epoch": 0.11653846153846154,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.268479347229004,
|
|
"step": 1515
|
|
},
|
|
{
|
|
"epoch": 0.11661538461538462,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280991554260254,
|
|
"step": 1516
|
|
},
|
|
{
|
|
"epoch": 0.11669230769230769,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2837626934051514,
|
|
"step": 1517
|
|
},
|
|
{
|
|
"epoch": 0.11676923076923076,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2816648483276367,
|
|
"step": 1518
|
|
},
|
|
{
|
|
"epoch": 0.11684615384615385,
|
|
"grad_norm": 0.04638671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2672369480133057,
|
|
"step": 1519
|
|
},
|
|
{
|
|
"epoch": 0.11692307692307692,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.256504774093628,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 0.117,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2876079082489014,
|
|
"step": 1521
|
|
},
|
|
{
|
|
"epoch": 0.11707692307692308,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269368886947632,
|
|
"step": 1522
|
|
},
|
|
{
|
|
"epoch": 0.11715384615384615,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2736992835998535,
|
|
"step": 1523
|
|
},
|
|
{
|
|
"epoch": 0.11723076923076924,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2677578926086426,
|
|
"step": 1524
|
|
},
|
|
{
|
|
"epoch": 0.11730769230769231,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281013250350952,
|
|
"step": 1525
|
|
},
|
|
{
|
|
"epoch": 0.11738461538461538,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26878023147583,
|
|
"step": 1526
|
|
},
|
|
{
|
|
"epoch": 0.11746153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3150718212127686,
|
|
"step": 1527
|
|
},
|
|
{
|
|
"epoch": 0.11753846153846154,
|
|
"grad_norm": 0.0615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2946715354919434,
|
|
"step": 1528
|
|
},
|
|
{
|
|
"epoch": 0.11761538461538462,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26357364654541,
|
|
"step": 1529
|
|
},
|
|
{
|
|
"epoch": 0.1176923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2756173610687256,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 0.11776923076923076,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2683825492858887,
|
|
"step": 1531
|
|
},
|
|
{
|
|
"epoch": 0.11784615384615385,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3082659244537354,
|
|
"step": 1532
|
|
},
|
|
{
|
|
"epoch": 0.11792307692307692,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3077597618103027,
|
|
"step": 1533
|
|
},
|
|
{
|
|
"epoch": 0.118,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3121414184570312,
|
|
"step": 1534
|
|
},
|
|
{
|
|
"epoch": 0.11807692307692308,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290111541748047,
|
|
"step": 1535
|
|
},
|
|
{
|
|
"epoch": 0.11815384615384615,
|
|
"grad_norm": 0.05908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2705531120300293,
|
|
"step": 1536
|
|
},
|
|
{
|
|
"epoch": 0.11823076923076924,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2799899578094482,
|
|
"step": 1537
|
|
},
|
|
{
|
|
"epoch": 0.11830769230769231,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264899969100952,
|
|
"step": 1538
|
|
},
|
|
{
|
|
"epoch": 0.11838461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2515881061553955,
|
|
"step": 1539
|
|
},
|
|
{
|
|
"epoch": 0.11846153846153847,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2475528717041016,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 0.11853846153846154,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2652933597564697,
|
|
"step": 1541
|
|
},
|
|
{
|
|
"epoch": 0.11861538461538461,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.303346633911133,
|
|
"step": 1542
|
|
},
|
|
{
|
|
"epoch": 0.1186923076923077,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3145480155944824,
|
|
"step": 1543
|
|
},
|
|
{
|
|
"epoch": 0.11876923076923077,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2786316871643066,
|
|
"step": 1544
|
|
},
|
|
{
|
|
"epoch": 0.11884615384615385,
|
|
"grad_norm": 0.0634765625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2939202785491943,
|
|
"step": 1545
|
|
},
|
|
{
|
|
"epoch": 0.11892307692307692,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2987499237060547,
|
|
"step": 1546
|
|
},
|
|
{
|
|
"epoch": 0.119,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293464422225952,
|
|
"step": 1547
|
|
},
|
|
{
|
|
"epoch": 0.11907692307692308,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2967309951782227,
|
|
"step": 1548
|
|
},
|
|
{
|
|
"epoch": 0.11915384615384615,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274343490600586,
|
|
"step": 1549
|
|
},
|
|
{
|
|
"epoch": 0.11923076923076924,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2968032360076904,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.11930769230769231,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2990834712982178,
|
|
"step": 1551
|
|
},
|
|
{
|
|
"epoch": 0.11938461538461538,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292356014251709,
|
|
"step": 1552
|
|
},
|
|
{
|
|
"epoch": 0.11946153846153847,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2782278060913086,
|
|
"step": 1553
|
|
},
|
|
{
|
|
"epoch": 0.11953846153846154,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2601068019866943,
|
|
"step": 1554
|
|
},
|
|
{
|
|
"epoch": 0.11961538461538461,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2538583278656006,
|
|
"step": 1555
|
|
},
|
|
{
|
|
"epoch": 0.1196923076923077,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2632017135620117,
|
|
"step": 1556
|
|
},
|
|
{
|
|
"epoch": 0.11976923076923077,
|
|
"grad_norm": 0.05908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2795090675354004,
|
|
"step": 1557
|
|
},
|
|
{
|
|
"epoch": 0.11984615384615385,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2706642150878906,
|
|
"step": 1558
|
|
},
|
|
{
|
|
"epoch": 0.11992307692307692,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.253105640411377,
|
|
"step": 1559
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2669293880462646,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 0.12007692307692308,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.25414776802063,
|
|
"step": 1561
|
|
},
|
|
{
|
|
"epoch": 0.12015384615384615,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2825629711151123,
|
|
"step": 1562
|
|
},
|
|
{
|
|
"epoch": 0.12023076923076922,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252047538757324,
|
|
"step": 1563
|
|
},
|
|
{
|
|
"epoch": 0.12030769230769231,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266549587249756,
|
|
"step": 1564
|
|
},
|
|
{
|
|
"epoch": 0.12038461538461538,
|
|
"grad_norm": 0.060546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2919762134552,
|
|
"step": 1565
|
|
},
|
|
{
|
|
"epoch": 0.12046153846153847,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.313251256942749,
|
|
"step": 1566
|
|
},
|
|
{
|
|
"epoch": 0.12053846153846154,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273660659790039,
|
|
"step": 1567
|
|
},
|
|
{
|
|
"epoch": 0.12061538461538461,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.259809970855713,
|
|
"step": 1568
|
|
},
|
|
{
|
|
"epoch": 0.1206923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2640018463134766,
|
|
"step": 1569
|
|
},
|
|
{
|
|
"epoch": 0.12076923076923077,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2716188430786133,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 0.12084615384615384,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28609037399292,
|
|
"step": 1571
|
|
},
|
|
{
|
|
"epoch": 0.12092307692307692,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285327196121216,
|
|
"step": 1572
|
|
},
|
|
{
|
|
"epoch": 0.121,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3015406131744385,
|
|
"step": 1573
|
|
},
|
|
{
|
|
"epoch": 0.12107692307692308,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2647156715393066,
|
|
"step": 1574
|
|
},
|
|
{
|
|
"epoch": 0.12115384615384615,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2799525260925293,
|
|
"step": 1575
|
|
},
|
|
{
|
|
"epoch": 0.12123076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289839029312134,
|
|
"step": 1576
|
|
},
|
|
{
|
|
"epoch": 0.12130769230769231,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289036512374878,
|
|
"step": 1577
|
|
},
|
|
{
|
|
"epoch": 0.12138461538461538,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279052972793579,
|
|
"step": 1578
|
|
},
|
|
{
|
|
"epoch": 0.12146153846153847,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2502918243408203,
|
|
"step": 1579
|
|
},
|
|
{
|
|
"epoch": 0.12153846153846154,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2726213932037354,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 0.12161538461538461,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.298384189605713,
|
|
"step": 1581
|
|
},
|
|
{
|
|
"epoch": 0.1216923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279877185821533,
|
|
"step": 1582
|
|
},
|
|
{
|
|
"epoch": 0.12176923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2787983417510986,
|
|
"step": 1583
|
|
},
|
|
{
|
|
"epoch": 0.12184615384615384,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2900795936584473,
|
|
"step": 1584
|
|
},
|
|
{
|
|
"epoch": 0.12192307692307693,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278425455093384,
|
|
"step": 1585
|
|
},
|
|
{
|
|
"epoch": 0.122,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2718749046325684,
|
|
"step": 1586
|
|
},
|
|
{
|
|
"epoch": 0.12207692307692308,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.261878490447998,
|
|
"step": 1587
|
|
},
|
|
{
|
|
"epoch": 0.12215384615384615,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.255105972290039,
|
|
"step": 1588
|
|
},
|
|
{
|
|
"epoch": 0.12223076923076923,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2378487586975098,
|
|
"step": 1589
|
|
},
|
|
{
|
|
"epoch": 0.12230769230769231,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2835609912872314,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 0.12238461538461538,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2899961471557617,
|
|
"step": 1591
|
|
},
|
|
{
|
|
"epoch": 0.12246153846153846,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265063762664795,
|
|
"step": 1592
|
|
},
|
|
{
|
|
"epoch": 0.12253846153846154,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2753584384918213,
|
|
"step": 1593
|
|
},
|
|
{
|
|
"epoch": 0.12261538461538461,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27531361579895,
|
|
"step": 1594
|
|
},
|
|
{
|
|
"epoch": 0.1226923076923077,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2811598777770996,
|
|
"step": 1595
|
|
},
|
|
{
|
|
"epoch": 0.12276923076923077,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2755682468414307,
|
|
"step": 1596
|
|
},
|
|
{
|
|
"epoch": 0.12284615384615384,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2643797397613525,
|
|
"step": 1597
|
|
},
|
|
{
|
|
"epoch": 0.12292307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2604963779449463,
|
|
"step": 1598
|
|
},
|
|
{
|
|
"epoch": 0.123,
|
|
"grad_norm": 0.05859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2881381511688232,
|
|
"step": 1599
|
|
},
|
|
{
|
|
"epoch": 0.12307692307692308,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276512861251831,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.12315384615384616,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2865192890167236,
|
|
"step": 1601
|
|
},
|
|
{
|
|
"epoch": 0.12323076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296142578125,
|
|
"step": 1602
|
|
},
|
|
{
|
|
"epoch": 0.12330769230769231,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286663055419922,
|
|
"step": 1603
|
|
},
|
|
{
|
|
"epoch": 0.12338461538461538,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2854809761047363,
|
|
"step": 1604
|
|
},
|
|
{
|
|
"epoch": 0.12346153846153846,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272406816482544,
|
|
"step": 1605
|
|
},
|
|
{
|
|
"epoch": 0.12353846153846154,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29083514213562,
|
|
"step": 1606
|
|
},
|
|
{
|
|
"epoch": 0.12361538461538461,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2662692070007324,
|
|
"step": 1607
|
|
},
|
|
{
|
|
"epoch": 0.1236923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280508041381836,
|
|
"step": 1608
|
|
},
|
|
{
|
|
"epoch": 0.12376923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276205062866211,
|
|
"step": 1609
|
|
},
|
|
{
|
|
"epoch": 0.12384615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2868616580963135,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"epoch": 0.12392307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2714624404907227,
|
|
"step": 1611
|
|
},
|
|
{
|
|
"epoch": 0.124,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2588162422180176,
|
|
"step": 1612
|
|
},
|
|
{
|
|
"epoch": 0.12407692307692307,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302276372909546,
|
|
"step": 1613
|
|
},
|
|
{
|
|
"epoch": 0.12415384615384616,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3011579513549805,
|
|
"step": 1614
|
|
},
|
|
{
|
|
"epoch": 0.12423076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2835466861724854,
|
|
"step": 1615
|
|
},
|
|
{
|
|
"epoch": 0.12430769230769231,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.291175127029419,
|
|
"step": 1616
|
|
},
|
|
{
|
|
"epoch": 0.12438461538461539,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2790634632110596,
|
|
"step": 1617
|
|
},
|
|
{
|
|
"epoch": 0.12446153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2873899936676025,
|
|
"step": 1618
|
|
},
|
|
{
|
|
"epoch": 0.12453846153846154,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2850944995880127,
|
|
"step": 1619
|
|
},
|
|
{
|
|
"epoch": 0.12461538461538461,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29046630859375,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 0.12469230769230769,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266103506088257,
|
|
"step": 1621
|
|
},
|
|
{
|
|
"epoch": 0.12476923076923077,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3207457065582275,
|
|
"step": 1622
|
|
},
|
|
{
|
|
"epoch": 0.12484615384615384,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2807531356811523,
|
|
"step": 1623
|
|
},
|
|
{
|
|
"epoch": 0.12492307692307693,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282443046569824,
|
|
"step": 1624
|
|
},
|
|
{
|
|
"epoch": 0.125,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3139641284942627,
|
|
"step": 1625
|
|
},
|
|
{
|
|
"epoch": 0.12507692307692309,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.25911545753479,
|
|
"step": 1626
|
|
},
|
|
{
|
|
"epoch": 0.12515384615384614,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254673719406128,
|
|
"step": 1627
|
|
},
|
|
{
|
|
"epoch": 0.12523076923076923,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2908549308776855,
|
|
"step": 1628
|
|
},
|
|
{
|
|
"epoch": 0.12530769230769231,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2897024154663086,
|
|
"step": 1629
|
|
},
|
|
{
|
|
"epoch": 0.12538461538461537,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2716333866119385,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"epoch": 0.12546153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2682442665100098,
|
|
"step": 1631
|
|
},
|
|
{
|
|
"epoch": 0.12553846153846154,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2735748291015625,
|
|
"step": 1632
|
|
},
|
|
{
|
|
"epoch": 0.12561538461538463,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3005504608154297,
|
|
"step": 1633
|
|
},
|
|
{
|
|
"epoch": 0.1256923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3074841499328613,
|
|
"step": 1634
|
|
},
|
|
{
|
|
"epoch": 0.12576923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2942190170288086,
|
|
"step": 1635
|
|
},
|
|
{
|
|
"epoch": 0.12584615384615386,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2715466022491455,
|
|
"step": 1636
|
|
},
|
|
{
|
|
"epoch": 0.12592307692307692,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.25559139251709,
|
|
"step": 1637
|
|
},
|
|
{
|
|
"epoch": 0.126,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.268742084503174,
|
|
"step": 1638
|
|
},
|
|
{
|
|
"epoch": 0.1260769230769231,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279475212097168,
|
|
"step": 1639
|
|
},
|
|
{
|
|
"epoch": 0.12615384615384614,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2608654499053955,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 0.12623076923076923,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280597686767578,
|
|
"step": 1641
|
|
},
|
|
{
|
|
"epoch": 0.12630769230769232,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.243931293487549,
|
|
"step": 1642
|
|
},
|
|
{
|
|
"epoch": 0.12638461538461537,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279604196548462,
|
|
"step": 1643
|
|
},
|
|
{
|
|
"epoch": 0.12646153846153846,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2895431518554688,
|
|
"step": 1644
|
|
},
|
|
{
|
|
"epoch": 0.12653846153846154,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271253824234009,
|
|
"step": 1645
|
|
},
|
|
{
|
|
"epoch": 0.1266153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2561371326446533,
|
|
"step": 1646
|
|
},
|
|
{
|
|
"epoch": 0.1266923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2595255374908447,
|
|
"step": 1647
|
|
},
|
|
{
|
|
"epoch": 0.12676923076923077,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2872138023376465,
|
|
"step": 1648
|
|
},
|
|
{
|
|
"epoch": 0.12684615384615386,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289501667022705,
|
|
"step": 1649
|
|
},
|
|
{
|
|
"epoch": 0.12692307692307692,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265887975692749,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 0.127,
|
|
"grad_norm": 0.060791015625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2636659145355225,
|
|
"step": 1651
|
|
},
|
|
{
|
|
"epoch": 0.1270769230769231,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274228572845459,
|
|
"step": 1652
|
|
},
|
|
{
|
|
"epoch": 0.12715384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289851427078247,
|
|
"step": 1653
|
|
},
|
|
{
|
|
"epoch": 0.12723076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300541877746582,
|
|
"step": 1654
|
|
},
|
|
{
|
|
"epoch": 0.12730769230769232,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2791554927825928,
|
|
"step": 1655
|
|
},
|
|
{
|
|
"epoch": 0.12738461538461537,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290310859680176,
|
|
"step": 1656
|
|
},
|
|
{
|
|
"epoch": 0.12746153846153846,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2482283115386963,
|
|
"step": 1657
|
|
},
|
|
{
|
|
"epoch": 0.12753846153846154,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27512526512146,
|
|
"step": 1658
|
|
},
|
|
{
|
|
"epoch": 0.1276153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294180393218994,
|
|
"step": 1659
|
|
},
|
|
{
|
|
"epoch": 0.1276923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2733566761016846,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"epoch": 0.12776923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.261453151702881,
|
|
"step": 1661
|
|
},
|
|
{
|
|
"epoch": 0.12784615384615386,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2973203659057617,
|
|
"step": 1662
|
|
},
|
|
{
|
|
"epoch": 0.12792307692307692,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2817189693450928,
|
|
"step": 1663
|
|
},
|
|
{
|
|
"epoch": 0.128,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2538843154907227,
|
|
"step": 1664
|
|
},
|
|
{
|
|
"epoch": 0.1280769230769231,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3179283142089844,
|
|
"step": 1665
|
|
},
|
|
{
|
|
"epoch": 0.12815384615384615,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2670273780822754,
|
|
"step": 1666
|
|
},
|
|
{
|
|
"epoch": 0.12823076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2587456703186035,
|
|
"step": 1667
|
|
},
|
|
{
|
|
"epoch": 0.12830769230769232,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.297819137573242,
|
|
"step": 1668
|
|
},
|
|
{
|
|
"epoch": 0.12838461538461537,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2929530143737793,
|
|
"step": 1669
|
|
},
|
|
{
|
|
"epoch": 0.12846153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2958037853240967,
|
|
"step": 1670
|
|
},
|
|
{
|
|
"epoch": 0.12853846153846155,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272658586502075,
|
|
"step": 1671
|
|
},
|
|
{
|
|
"epoch": 0.1286153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2899296283721924,
|
|
"step": 1672
|
|
},
|
|
{
|
|
"epoch": 0.1286923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2703144550323486,
|
|
"step": 1673
|
|
},
|
|
{
|
|
"epoch": 0.12876923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3071935176849365,
|
|
"step": 1674
|
|
},
|
|
{
|
|
"epoch": 0.12884615384615383,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2602996826171875,
|
|
"step": 1675
|
|
},
|
|
{
|
|
"epoch": 0.12892307692307692,
|
|
"grad_norm": 0.06787109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2406485080718994,
|
|
"step": 1676
|
|
},
|
|
{
|
|
"epoch": 0.129,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.302485942840576,
|
|
"step": 1677
|
|
},
|
|
{
|
|
"epoch": 0.1290769230769231,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274357318878174,
|
|
"step": 1678
|
|
},
|
|
{
|
|
"epoch": 0.12915384615384615,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2737183570861816,
|
|
"step": 1679
|
|
},
|
|
{
|
|
"epoch": 0.12923076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2558298110961914,
|
|
"step": 1680
|
|
},
|
|
{
|
|
"epoch": 0.12930769230769232,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2567391395568848,
|
|
"step": 1681
|
|
},
|
|
{
|
|
"epoch": 0.12938461538461538,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.284274101257324,
|
|
"step": 1682
|
|
},
|
|
{
|
|
"epoch": 0.12946153846153846,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260139226913452,
|
|
"step": 1683
|
|
},
|
|
{
|
|
"epoch": 0.12953846153846155,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2660398483276367,
|
|
"step": 1684
|
|
},
|
|
{
|
|
"epoch": 0.1296153846153846,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270083427429199,
|
|
"step": 1685
|
|
},
|
|
{
|
|
"epoch": 0.1296923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2539010047912598,
|
|
"step": 1686
|
|
},
|
|
{
|
|
"epoch": 0.12976923076923078,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2528584003448486,
|
|
"step": 1687
|
|
},
|
|
{
|
|
"epoch": 0.12984615384615383,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2838165760040283,
|
|
"step": 1688
|
|
},
|
|
{
|
|
"epoch": 0.12992307692307692,
|
|
"grad_norm": 0.05615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2880709171295166,
|
|
"step": 1689
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299452543258667,
|
|
"step": 1690
|
|
},
|
|
{
|
|
"epoch": 0.1300769230769231,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.256288528442383,
|
|
"step": 1691
|
|
},
|
|
{
|
|
"epoch": 0.13015384615384615,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.253983974456787,
|
|
"step": 1692
|
|
},
|
|
{
|
|
"epoch": 0.13023076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267174005508423,
|
|
"step": 1693
|
|
},
|
|
{
|
|
"epoch": 0.13030769230769232,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2843220233917236,
|
|
"step": 1694
|
|
},
|
|
{
|
|
"epoch": 0.13038461538461538,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.250291347503662,
|
|
"step": 1695
|
|
},
|
|
{
|
|
"epoch": 0.13046153846153846,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2494959831237793,
|
|
"step": 1696
|
|
},
|
|
{
|
|
"epoch": 0.13053846153846155,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2667765617370605,
|
|
"step": 1697
|
|
},
|
|
{
|
|
"epoch": 0.1306153846153846,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252580404281616,
|
|
"step": 1698
|
|
},
|
|
{
|
|
"epoch": 0.1306923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286985397338867,
|
|
"step": 1699
|
|
},
|
|
{
|
|
"epoch": 0.13076923076923078,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2579896450042725,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 0.13084615384615383,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252689838409424,
|
|
"step": 1701
|
|
},
|
|
{
|
|
"epoch": 0.13092307692307692,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2586567401885986,
|
|
"step": 1702
|
|
},
|
|
{
|
|
"epoch": 0.131,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2927322387695312,
|
|
"step": 1703
|
|
},
|
|
{
|
|
"epoch": 0.13107692307692306,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30063533782959,
|
|
"step": 1704
|
|
},
|
|
{
|
|
"epoch": 0.13115384615384615,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2398111820220947,
|
|
"step": 1705
|
|
},
|
|
{
|
|
"epoch": 0.13123076923076923,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2353551387786865,
|
|
"step": 1706
|
|
},
|
|
{
|
|
"epoch": 0.13130769230769232,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2904791831970215,
|
|
"step": 1707
|
|
},
|
|
{
|
|
"epoch": 0.13138461538461538,
|
|
"grad_norm": 0.061767578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2558600902557373,
|
|
"step": 1708
|
|
},
|
|
{
|
|
"epoch": 0.13146153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304398775100708,
|
|
"step": 1709
|
|
},
|
|
{
|
|
"epoch": 0.13153846153846155,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2797787189483643,
|
|
"step": 1710
|
|
},
|
|
{
|
|
"epoch": 0.1316153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252304792404175,
|
|
"step": 1711
|
|
},
|
|
{
|
|
"epoch": 0.1316923076923077,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.277336359024048,
|
|
"step": 1712
|
|
},
|
|
{
|
|
"epoch": 0.13176923076923078,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3006503582000732,
|
|
"step": 1713
|
|
},
|
|
{
|
|
"epoch": 0.13184615384615384,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270364761352539,
|
|
"step": 1714
|
|
},
|
|
{
|
|
"epoch": 0.13192307692307692,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2639524936676025,
|
|
"step": 1715
|
|
},
|
|
{
|
|
"epoch": 0.132,
|
|
"grad_norm": 0.046630859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2485153675079346,
|
|
"step": 1716
|
|
},
|
|
{
|
|
"epoch": 0.13207692307692306,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290842294692993,
|
|
"step": 1717
|
|
},
|
|
{
|
|
"epoch": 0.13215384615384615,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2795801162719727,
|
|
"step": 1718
|
|
},
|
|
{
|
|
"epoch": 0.13223076923076924,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274472236633301,
|
|
"step": 1719
|
|
},
|
|
{
|
|
"epoch": 0.13230769230769232,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2750890254974365,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 0.13238461538461538,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267507553100586,
|
|
"step": 1721
|
|
},
|
|
{
|
|
"epoch": 0.13246153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2987735271453857,
|
|
"step": 1722
|
|
},
|
|
{
|
|
"epoch": 0.13253846153846155,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285726308822632,
|
|
"step": 1723
|
|
},
|
|
{
|
|
"epoch": 0.1326153846153846,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252462387084961,
|
|
"step": 1724
|
|
},
|
|
{
|
|
"epoch": 0.1326923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2626593112945557,
|
|
"step": 1725
|
|
},
|
|
{
|
|
"epoch": 0.13276923076923078,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2379705905914307,
|
|
"step": 1726
|
|
},
|
|
{
|
|
"epoch": 0.13284615384615384,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2965657711029053,
|
|
"step": 1727
|
|
},
|
|
{
|
|
"epoch": 0.13292307692307692,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270869016647339,
|
|
"step": 1728
|
|
},
|
|
{
|
|
"epoch": 0.133,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2948246002197266,
|
|
"step": 1729
|
|
},
|
|
{
|
|
"epoch": 0.13307692307692306,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2661702632904053,
|
|
"step": 1730
|
|
},
|
|
{
|
|
"epoch": 0.13315384615384615,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2255053520202637,
|
|
"step": 1731
|
|
},
|
|
{
|
|
"epoch": 0.13323076923076924,
|
|
"grad_norm": 0.05615234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2641711235046387,
|
|
"step": 1732
|
|
},
|
|
{
|
|
"epoch": 0.13330769230769232,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3161938190460205,
|
|
"step": 1733
|
|
},
|
|
{
|
|
"epoch": 0.13338461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.300751209259033,
|
|
"step": 1734
|
|
},
|
|
{
|
|
"epoch": 0.13346153846153846,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283698797225952,
|
|
"step": 1735
|
|
},
|
|
{
|
|
"epoch": 0.13353846153846155,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2718961238861084,
|
|
"step": 1736
|
|
},
|
|
{
|
|
"epoch": 0.1336153846153846,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2875566482543945,
|
|
"step": 1737
|
|
},
|
|
{
|
|
"epoch": 0.1336923076923077,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2726240158081055,
|
|
"step": 1738
|
|
},
|
|
{
|
|
"epoch": 0.13376923076923078,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2681424617767334,
|
|
"step": 1739
|
|
},
|
|
{
|
|
"epoch": 0.13384615384615384,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258366584777832,
|
|
"step": 1740
|
|
},
|
|
{
|
|
"epoch": 0.13392307692307692,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2528839111328125,
|
|
"step": 1741
|
|
},
|
|
{
|
|
"epoch": 0.134,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271008253097534,
|
|
"step": 1742
|
|
},
|
|
{
|
|
"epoch": 0.13407692307692307,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285386562347412,
|
|
"step": 1743
|
|
},
|
|
{
|
|
"epoch": 0.13415384615384615,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260132312774658,
|
|
"step": 1744
|
|
},
|
|
{
|
|
"epoch": 0.13423076923076924,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2513625621795654,
|
|
"step": 1745
|
|
},
|
|
{
|
|
"epoch": 0.1343076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3058507442474365,
|
|
"step": 1746
|
|
},
|
|
{
|
|
"epoch": 0.13438461538461538,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2620270252227783,
|
|
"step": 1747
|
|
},
|
|
{
|
|
"epoch": 0.13446153846153847,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2994725704193115,
|
|
"step": 1748
|
|
},
|
|
{
|
|
"epoch": 0.13453846153846155,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2942874431610107,
|
|
"step": 1749
|
|
},
|
|
{
|
|
"epoch": 0.1346153846153846,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2606778144836426,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 0.1346923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278118371963501,
|
|
"step": 1751
|
|
},
|
|
{
|
|
"epoch": 0.13476923076923078,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2976388931274414,
|
|
"step": 1752
|
|
},
|
|
{
|
|
"epoch": 0.13484615384615384,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2609193325042725,
|
|
"step": 1753
|
|
},
|
|
{
|
|
"epoch": 0.13492307692307692,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280362606048584,
|
|
"step": 1754
|
|
},
|
|
{
|
|
"epoch": 0.135,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2776880264282227,
|
|
"step": 1755
|
|
},
|
|
{
|
|
"epoch": 0.13507692307692307,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.256744861602783,
|
|
"step": 1756
|
|
},
|
|
{
|
|
"epoch": 0.13515384615384615,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2928833961486816,
|
|
"step": 1757
|
|
},
|
|
{
|
|
"epoch": 0.13523076923076924,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.299929618835449,
|
|
"step": 1758
|
|
},
|
|
{
|
|
"epoch": 0.1353076923076923,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2778468132019043,
|
|
"step": 1759
|
|
},
|
|
{
|
|
"epoch": 0.13538461538461538,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2509305477142334,
|
|
"step": 1760
|
|
},
|
|
{
|
|
"epoch": 0.13546153846153847,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30161190032959,
|
|
"step": 1761
|
|
},
|
|
{
|
|
"epoch": 0.13553846153846155,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.262180805206299,
|
|
"step": 1762
|
|
},
|
|
{
|
|
"epoch": 0.1356153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281710624694824,
|
|
"step": 1763
|
|
},
|
|
{
|
|
"epoch": 0.1356923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.259505033493042,
|
|
"step": 1764
|
|
},
|
|
{
|
|
"epoch": 0.13576923076923078,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2671782970428467,
|
|
"step": 1765
|
|
},
|
|
{
|
|
"epoch": 0.13584615384615384,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2654378414154053,
|
|
"step": 1766
|
|
},
|
|
{
|
|
"epoch": 0.13592307692307692,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2817962169647217,
|
|
"step": 1767
|
|
},
|
|
{
|
|
"epoch": 0.136,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267108678817749,
|
|
"step": 1768
|
|
},
|
|
{
|
|
"epoch": 0.13607692307692307,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.268860101699829,
|
|
"step": 1769
|
|
},
|
|
{
|
|
"epoch": 0.13615384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2692039012908936,
|
|
"step": 1770
|
|
},
|
|
{
|
|
"epoch": 0.13623076923076924,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2923927307128906,
|
|
"step": 1771
|
|
},
|
|
{
|
|
"epoch": 0.1363076923076923,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.312532901763916,
|
|
"step": 1772
|
|
},
|
|
{
|
|
"epoch": 0.13638461538461538,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2805674076080322,
|
|
"step": 1773
|
|
},
|
|
{
|
|
"epoch": 0.13646153846153847,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2458784580230713,
|
|
"step": 1774
|
|
},
|
|
{
|
|
"epoch": 0.13653846153846153,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2249228954315186,
|
|
"step": 1775
|
|
},
|
|
{
|
|
"epoch": 0.1366153846153846,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2917540073394775,
|
|
"step": 1776
|
|
},
|
|
{
|
|
"epoch": 0.1366923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2848899364471436,
|
|
"step": 1777
|
|
},
|
|
{
|
|
"epoch": 0.13676923076923078,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296670913696289,
|
|
"step": 1778
|
|
},
|
|
{
|
|
"epoch": 0.13684615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2725589275360107,
|
|
"step": 1779
|
|
},
|
|
{
|
|
"epoch": 0.13692307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2266669273376465,
|
|
"step": 1780
|
|
},
|
|
{
|
|
"epoch": 0.137,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2854397296905518,
|
|
"step": 1781
|
|
},
|
|
{
|
|
"epoch": 0.13707692307692307,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28179669380188,
|
|
"step": 1782
|
|
},
|
|
{
|
|
"epoch": 0.13715384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27282977104187,
|
|
"step": 1783
|
|
},
|
|
{
|
|
"epoch": 0.13723076923076924,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2888853549957275,
|
|
"step": 1784
|
|
},
|
|
{
|
|
"epoch": 0.1373076923076923,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2816598415374756,
|
|
"step": 1785
|
|
},
|
|
{
|
|
"epoch": 0.13738461538461538,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2662999629974365,
|
|
"step": 1786
|
|
},
|
|
{
|
|
"epoch": 0.13746153846153847,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2673592567443848,
|
|
"step": 1787
|
|
},
|
|
{
|
|
"epoch": 0.13753846153846153,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2976105213165283,
|
|
"step": 1788
|
|
},
|
|
{
|
|
"epoch": 0.1376153846153846,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.255948543548584,
|
|
"step": 1789
|
|
},
|
|
{
|
|
"epoch": 0.1376923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.240166425704956,
|
|
"step": 1790
|
|
},
|
|
{
|
|
"epoch": 0.13776923076923078,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2813053131103516,
|
|
"step": 1791
|
|
},
|
|
{
|
|
"epoch": 0.13784615384615384,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2759456634521484,
|
|
"step": 1792
|
|
},
|
|
{
|
|
"epoch": 0.13792307692307693,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285245418548584,
|
|
"step": 1793
|
|
},
|
|
{
|
|
"epoch": 0.138,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2666609287261963,
|
|
"step": 1794
|
|
},
|
|
{
|
|
"epoch": 0.13807692307692307,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.255413055419922,
|
|
"step": 1795
|
|
},
|
|
{
|
|
"epoch": 0.13815384615384615,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27553653717041,
|
|
"step": 1796
|
|
},
|
|
{
|
|
"epoch": 0.13823076923076924,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266387701034546,
|
|
"step": 1797
|
|
},
|
|
{
|
|
"epoch": 0.1383076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265894889831543,
|
|
"step": 1798
|
|
},
|
|
{
|
|
"epoch": 0.13838461538461538,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2743723392486572,
|
|
"step": 1799
|
|
},
|
|
{
|
|
"epoch": 0.13846153846153847,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.235647678375244,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 0.13853846153846153,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2631096839904785,
|
|
"step": 1801
|
|
},
|
|
{
|
|
"epoch": 0.1386153846153846,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289768934249878,
|
|
"step": 1802
|
|
},
|
|
{
|
|
"epoch": 0.1386923076923077,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2674362659454346,
|
|
"step": 1803
|
|
},
|
|
{
|
|
"epoch": 0.13876923076923076,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26521635055542,
|
|
"step": 1804
|
|
},
|
|
{
|
|
"epoch": 0.13884615384615384,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286775588989258,
|
|
"step": 1805
|
|
},
|
|
{
|
|
"epoch": 0.13892307692307693,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2861900329589844,
|
|
"step": 1806
|
|
},
|
|
{
|
|
"epoch": 0.139,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283564567565918,
|
|
"step": 1807
|
|
},
|
|
{
|
|
"epoch": 0.13907692307692307,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2869975566864014,
|
|
"step": 1808
|
|
},
|
|
{
|
|
"epoch": 0.13915384615384616,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2593445777893066,
|
|
"step": 1809
|
|
},
|
|
{
|
|
"epoch": 0.13923076923076924,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263089656829834,
|
|
"step": 1810
|
|
},
|
|
{
|
|
"epoch": 0.1393076923076923,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2862918376922607,
|
|
"step": 1811
|
|
},
|
|
{
|
|
"epoch": 0.13938461538461538,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.314255475997925,
|
|
"step": 1812
|
|
},
|
|
{
|
|
"epoch": 0.13946153846153847,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2668073177337646,
|
|
"step": 1813
|
|
},
|
|
{
|
|
"epoch": 0.13953846153846153,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.255507230758667,
|
|
"step": 1814
|
|
},
|
|
{
|
|
"epoch": 0.1396153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2894444465637207,
|
|
"step": 1815
|
|
},
|
|
{
|
|
"epoch": 0.1396923076923077,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2553839683532715,
|
|
"step": 1816
|
|
},
|
|
{
|
|
"epoch": 0.13976923076923076,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29831862449646,
|
|
"step": 1817
|
|
},
|
|
{
|
|
"epoch": 0.13984615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2692787647247314,
|
|
"step": 1818
|
|
},
|
|
{
|
|
"epoch": 0.13992307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.24756121635437,
|
|
"step": 1819
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252690553665161,
|
|
"step": 1820
|
|
},
|
|
{
|
|
"epoch": 0.14007692307692307,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293144941329956,
|
|
"step": 1821
|
|
},
|
|
{
|
|
"epoch": 0.14015384615384616,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278878688812256,
|
|
"step": 1822
|
|
},
|
|
{
|
|
"epoch": 0.14023076923076924,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2806026935577393,
|
|
"step": 1823
|
|
},
|
|
{
|
|
"epoch": 0.1403076923076923,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280390739440918,
|
|
"step": 1824
|
|
},
|
|
{
|
|
"epoch": 0.14038461538461539,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.235408306121826,
|
|
"step": 1825
|
|
},
|
|
{
|
|
"epoch": 0.14046153846153847,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2696902751922607,
|
|
"step": 1826
|
|
},
|
|
{
|
|
"epoch": 0.14053846153846153,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2772860527038574,
|
|
"step": 1827
|
|
},
|
|
{
|
|
"epoch": 0.14061538461538461,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2649617195129395,
|
|
"step": 1828
|
|
},
|
|
{
|
|
"epoch": 0.1406923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2517142295837402,
|
|
"step": 1829
|
|
},
|
|
{
|
|
"epoch": 0.14076923076923076,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2717506885528564,
|
|
"step": 1830
|
|
},
|
|
{
|
|
"epoch": 0.14084615384615384,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294996738433838,
|
|
"step": 1831
|
|
},
|
|
{
|
|
"epoch": 0.14092307692307693,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2639358043670654,
|
|
"step": 1832
|
|
},
|
|
{
|
|
"epoch": 0.141,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2608883380889893,
|
|
"step": 1833
|
|
},
|
|
{
|
|
"epoch": 0.14107692307692307,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.284914493560791,
|
|
"step": 1834
|
|
},
|
|
{
|
|
"epoch": 0.14115384615384616,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3178961277008057,
|
|
"step": 1835
|
|
},
|
|
{
|
|
"epoch": 0.14123076923076924,
|
|
"grad_norm": 0.0576171875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274600028991699,
|
|
"step": 1836
|
|
},
|
|
{
|
|
"epoch": 0.1413076923076923,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283984422683716,
|
|
"step": 1837
|
|
},
|
|
{
|
|
"epoch": 0.1413846153846154,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.275743007659912,
|
|
"step": 1838
|
|
},
|
|
{
|
|
"epoch": 0.14146153846153847,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266361713409424,
|
|
"step": 1839
|
|
},
|
|
{
|
|
"epoch": 0.14153846153846153,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2657713890075684,
|
|
"step": 1840
|
|
},
|
|
{
|
|
"epoch": 0.14161538461538462,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285396099090576,
|
|
"step": 1841
|
|
},
|
|
{
|
|
"epoch": 0.1416923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2641124725341797,
|
|
"step": 1842
|
|
},
|
|
{
|
|
"epoch": 0.14176923076923076,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2677161693573,
|
|
"step": 1843
|
|
},
|
|
{
|
|
"epoch": 0.14184615384615384,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2699713706970215,
|
|
"step": 1844
|
|
},
|
|
{
|
|
"epoch": 0.14192307692307693,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283129930496216,
|
|
"step": 1845
|
|
},
|
|
{
|
|
"epoch": 0.142,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2646987438201904,
|
|
"step": 1846
|
|
},
|
|
{
|
|
"epoch": 0.14207692307692307,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.296809196472168,
|
|
"step": 1847
|
|
},
|
|
{
|
|
"epoch": 0.14215384615384616,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2794697284698486,
|
|
"step": 1848
|
|
},
|
|
{
|
|
"epoch": 0.14223076923076924,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.256743907928467,
|
|
"step": 1849
|
|
},
|
|
{
|
|
"epoch": 0.1423076923076923,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269493818283081,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 0.1423846153846154,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2767274379730225,
|
|
"step": 1851
|
|
},
|
|
{
|
|
"epoch": 0.14246153846153847,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2685062885284424,
|
|
"step": 1852
|
|
},
|
|
{
|
|
"epoch": 0.14253846153846153,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.259917974472046,
|
|
"step": 1853
|
|
},
|
|
{
|
|
"epoch": 0.14261538461538462,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.240743637084961,
|
|
"step": 1854
|
|
},
|
|
{
|
|
"epoch": 0.1426923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2373547554016113,
|
|
"step": 1855
|
|
},
|
|
{
|
|
"epoch": 0.14276923076923076,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26086163520813,
|
|
"step": 1856
|
|
},
|
|
{
|
|
"epoch": 0.14284615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.275695323944092,
|
|
"step": 1857
|
|
},
|
|
{
|
|
"epoch": 0.14292307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2782771587371826,
|
|
"step": 1858
|
|
},
|
|
{
|
|
"epoch": 0.143,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.247370719909668,
|
|
"step": 1859
|
|
},
|
|
{
|
|
"epoch": 0.14307692307692307,
|
|
"grad_norm": 0.1298828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2914998531341553,
|
|
"step": 1860
|
|
},
|
|
{
|
|
"epoch": 0.14315384615384616,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3075144290924072,
|
|
"step": 1861
|
|
},
|
|
{
|
|
"epoch": 0.14323076923076922,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2480781078338623,
|
|
"step": 1862
|
|
},
|
|
{
|
|
"epoch": 0.1433076923076923,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258535146713257,
|
|
"step": 1863
|
|
},
|
|
{
|
|
"epoch": 0.1433846153846154,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254106044769287,
|
|
"step": 1864
|
|
},
|
|
{
|
|
"epoch": 0.14346153846153847,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274900436401367,
|
|
"step": 1865
|
|
},
|
|
{
|
|
"epoch": 0.14353846153846153,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2761924266815186,
|
|
"step": 1866
|
|
},
|
|
{
|
|
"epoch": 0.14361538461538462,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2860562801361084,
|
|
"step": 1867
|
|
},
|
|
{
|
|
"epoch": 0.1436923076923077,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.257594585418701,
|
|
"step": 1868
|
|
},
|
|
{
|
|
"epoch": 0.14376923076923076,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.236194133758545,
|
|
"step": 1869
|
|
},
|
|
{
|
|
"epoch": 0.14384615384615385,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2434449195861816,
|
|
"step": 1870
|
|
},
|
|
{
|
|
"epoch": 0.14392307692307693,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3009512424468994,
|
|
"step": 1871
|
|
},
|
|
{
|
|
"epoch": 0.144,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2913801670074463,
|
|
"step": 1872
|
|
},
|
|
{
|
|
"epoch": 0.14407692307692307,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2732880115509033,
|
|
"step": 1873
|
|
},
|
|
{
|
|
"epoch": 0.14415384615384616,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258152723312378,
|
|
"step": 1874
|
|
},
|
|
{
|
|
"epoch": 0.14423076923076922,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2879819869995117,
|
|
"step": 1875
|
|
},
|
|
{
|
|
"epoch": 0.1443076923076923,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.297663927078247,
|
|
"step": 1876
|
|
},
|
|
{
|
|
"epoch": 0.1443846153846154,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2909865379333496,
|
|
"step": 1877
|
|
},
|
|
{
|
|
"epoch": 0.14446153846153847,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2683980464935303,
|
|
"step": 1878
|
|
},
|
|
{
|
|
"epoch": 0.14453846153846153,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.24198055267334,
|
|
"step": 1879
|
|
},
|
|
{
|
|
"epoch": 0.14461538461538462,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2691781520843506,
|
|
"step": 1880
|
|
},
|
|
{
|
|
"epoch": 0.1446923076923077,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2805166244506836,
|
|
"step": 1881
|
|
},
|
|
{
|
|
"epoch": 0.14476923076923076,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2926366329193115,
|
|
"step": 1882
|
|
},
|
|
{
|
|
"epoch": 0.14484615384615385,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2829792499542236,
|
|
"step": 1883
|
|
},
|
|
{
|
|
"epoch": 0.14492307692307693,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.241101026535034,
|
|
"step": 1884
|
|
},
|
|
{
|
|
"epoch": 0.145,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2767767906188965,
|
|
"step": 1885
|
|
},
|
|
{
|
|
"epoch": 0.14507692307692308,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2572944164276123,
|
|
"step": 1886
|
|
},
|
|
{
|
|
"epoch": 0.14515384615384616,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2542853355407715,
|
|
"step": 1887
|
|
},
|
|
{
|
|
"epoch": 0.14523076923076922,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260773181915283,
|
|
"step": 1888
|
|
},
|
|
{
|
|
"epoch": 0.1453076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2678158283233643,
|
|
"step": 1889
|
|
},
|
|
{
|
|
"epoch": 0.1453846153846154,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.290088176727295,
|
|
"step": 1890
|
|
},
|
|
{
|
|
"epoch": 0.14546153846153845,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289501190185547,
|
|
"step": 1891
|
|
},
|
|
{
|
|
"epoch": 0.14553846153846153,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2840030193328857,
|
|
"step": 1892
|
|
},
|
|
{
|
|
"epoch": 0.14561538461538462,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2592079639434814,
|
|
"step": 1893
|
|
},
|
|
{
|
|
"epoch": 0.1456923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.22300386428833,
|
|
"step": 1894
|
|
},
|
|
{
|
|
"epoch": 0.14576923076923076,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2545740604400635,
|
|
"step": 1895
|
|
},
|
|
{
|
|
"epoch": 0.14584615384615385,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.244967222213745,
|
|
"step": 1896
|
|
},
|
|
{
|
|
"epoch": 0.14592307692307693,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286064863204956,
|
|
"step": 1897
|
|
},
|
|
{
|
|
"epoch": 0.146,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2727558612823486,
|
|
"step": 1898
|
|
},
|
|
{
|
|
"epoch": 0.14607692307692308,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2697718143463135,
|
|
"step": 1899
|
|
},
|
|
{
|
|
"epoch": 0.14615384615384616,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.295034408569336,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 0.14623076923076922,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2822105884552,
|
|
"step": 1901
|
|
},
|
|
{
|
|
"epoch": 0.1463076923076923,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2918193340301514,
|
|
"step": 1902
|
|
},
|
|
{
|
|
"epoch": 0.1463846153846154,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2655107975006104,
|
|
"step": 1903
|
|
},
|
|
{
|
|
"epoch": 0.14646153846153845,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260927677154541,
|
|
"step": 1904
|
|
},
|
|
{
|
|
"epoch": 0.14653846153846153,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2689876556396484,
|
|
"step": 1905
|
|
},
|
|
{
|
|
"epoch": 0.14661538461538462,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2911629676818848,
|
|
"step": 1906
|
|
},
|
|
{
|
|
"epoch": 0.1466923076923077,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282381296157837,
|
|
"step": 1907
|
|
},
|
|
{
|
|
"epoch": 0.14676923076923076,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289015531539917,
|
|
"step": 1908
|
|
},
|
|
{
|
|
"epoch": 0.14684615384615385,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2481818199157715,
|
|
"step": 1909
|
|
},
|
|
{
|
|
"epoch": 0.14692307692307693,
|
|
"grad_norm": 0.054931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2678375244140625,
|
|
"step": 1910
|
|
},
|
|
{
|
|
"epoch": 0.147,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2477240562438965,
|
|
"step": 1911
|
|
},
|
|
{
|
|
"epoch": 0.14707692307692308,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2969248294830322,
|
|
"step": 1912
|
|
},
|
|
{
|
|
"epoch": 0.14715384615384616,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263333559036255,
|
|
"step": 1913
|
|
},
|
|
{
|
|
"epoch": 0.14723076923076922,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2737507820129395,
|
|
"step": 1914
|
|
},
|
|
{
|
|
"epoch": 0.1473076923076923,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2942962646484375,
|
|
"step": 1915
|
|
},
|
|
{
|
|
"epoch": 0.1473846153846154,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272123336791992,
|
|
"step": 1916
|
|
},
|
|
{
|
|
"epoch": 0.14746153846153845,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3128662109375,
|
|
"step": 1917
|
|
},
|
|
{
|
|
"epoch": 0.14753846153846153,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28910493850708,
|
|
"step": 1918
|
|
},
|
|
{
|
|
"epoch": 0.14761538461538462,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2599358558654785,
|
|
"step": 1919
|
|
},
|
|
{
|
|
"epoch": 0.1476923076923077,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3085591793060303,
|
|
"step": 1920
|
|
},
|
|
{
|
|
"epoch": 0.14776923076923076,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269791603088379,
|
|
"step": 1921
|
|
},
|
|
{
|
|
"epoch": 0.14784615384615385,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281803846359253,
|
|
"step": 1922
|
|
},
|
|
{
|
|
"epoch": 0.14792307692307693,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2487127780914307,
|
|
"step": 1923
|
|
},
|
|
{
|
|
"epoch": 0.148,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2631592750549316,
|
|
"step": 1924
|
|
},
|
|
{
|
|
"epoch": 0.14807692307692308,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274519205093384,
|
|
"step": 1925
|
|
},
|
|
{
|
|
"epoch": 0.14815384615384616,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287407636642456,
|
|
"step": 1926
|
|
},
|
|
{
|
|
"epoch": 0.14823076923076922,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3052995204925537,
|
|
"step": 1927
|
|
},
|
|
{
|
|
"epoch": 0.1483076923076923,
|
|
"grad_norm": 0.046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.251197099685669,
|
|
"step": 1928
|
|
},
|
|
{
|
|
"epoch": 0.1483846153846154,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2611496448516846,
|
|
"step": 1929
|
|
},
|
|
{
|
|
"epoch": 0.14846153846153845,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2983431816101074,
|
|
"step": 1930
|
|
},
|
|
{
|
|
"epoch": 0.14853846153846154,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30291485786438,
|
|
"step": 1931
|
|
},
|
|
{
|
|
"epoch": 0.14861538461538462,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2789859771728516,
|
|
"step": 1932
|
|
},
|
|
{
|
|
"epoch": 0.14869230769230768,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269357204437256,
|
|
"step": 1933
|
|
},
|
|
{
|
|
"epoch": 0.14876923076923076,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26228928565979,
|
|
"step": 1934
|
|
},
|
|
{
|
|
"epoch": 0.14884615384615385,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285062551498413,
|
|
"step": 1935
|
|
},
|
|
{
|
|
"epoch": 0.14892307692307694,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2513651847839355,
|
|
"step": 1936
|
|
},
|
|
{
|
|
"epoch": 0.149,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269437074661255,
|
|
"step": 1937
|
|
},
|
|
{
|
|
"epoch": 0.14907692307692308,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2932040691375732,
|
|
"step": 1938
|
|
},
|
|
{
|
|
"epoch": 0.14915384615384616,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3026485443115234,
|
|
"step": 1939
|
|
},
|
|
{
|
|
"epoch": 0.14923076923076922,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2636752128601074,
|
|
"step": 1940
|
|
},
|
|
{
|
|
"epoch": 0.1493076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2750332355499268,
|
|
"step": 1941
|
|
},
|
|
{
|
|
"epoch": 0.1493846153846154,
|
|
"grad_norm": 0.05419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2903923988342285,
|
|
"step": 1942
|
|
},
|
|
{
|
|
"epoch": 0.14946153846153845,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2591378688812256,
|
|
"step": 1943
|
|
},
|
|
{
|
|
"epoch": 0.14953846153846154,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287057399749756,
|
|
"step": 1944
|
|
},
|
|
{
|
|
"epoch": 0.14961538461538462,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2497751712799072,
|
|
"step": 1945
|
|
},
|
|
{
|
|
"epoch": 0.14969230769230768,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2650794982910156,
|
|
"step": 1946
|
|
},
|
|
{
|
|
"epoch": 0.14976923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2551515102386475,
|
|
"step": 1947
|
|
},
|
|
{
|
|
"epoch": 0.14984615384615385,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.242255210876465,
|
|
"step": 1948
|
|
},
|
|
{
|
|
"epoch": 0.14992307692307694,
|
|
"grad_norm": 0.0537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2728545665740967,
|
|
"step": 1949
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2574617862701416,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"eval_loss": 1.927430272102356,
|
|
"eval_runtime": 734.6554,
|
|
"eval_samples_per_second": 25.879,
|
|
"eval_steps_per_second": 6.47,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.15007692307692308,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2611591815948486,
|
|
"step": 1951
|
|
},
|
|
{
|
|
"epoch": 0.15015384615384617,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3004865646362305,
|
|
"step": 1952
|
|
},
|
|
{
|
|
"epoch": 0.15023076923076922,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283640146255493,
|
|
"step": 1953
|
|
},
|
|
{
|
|
"epoch": 0.1503076923076923,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2714145183563232,
|
|
"step": 1954
|
|
},
|
|
{
|
|
"epoch": 0.1503846153846154,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.278994083404541,
|
|
"step": 1955
|
|
},
|
|
{
|
|
"epoch": 0.15046153846153845,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2710835933685303,
|
|
"step": 1956
|
|
},
|
|
{
|
|
"epoch": 0.15053846153846154,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258436679840088,
|
|
"step": 1957
|
|
},
|
|
{
|
|
"epoch": 0.15061538461538462,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2482452392578125,
|
|
"step": 1958
|
|
},
|
|
{
|
|
"epoch": 0.15069230769230768,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2680411338806152,
|
|
"step": 1959
|
|
},
|
|
{
|
|
"epoch": 0.15076923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.261289596557617,
|
|
"step": 1960
|
|
},
|
|
{
|
|
"epoch": 0.15084615384615385,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282461643218994,
|
|
"step": 1961
|
|
},
|
|
{
|
|
"epoch": 0.1509230769230769,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2777230739593506,
|
|
"step": 1962
|
|
},
|
|
{
|
|
"epoch": 0.151,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2788429260253906,
|
|
"step": 1963
|
|
},
|
|
{
|
|
"epoch": 0.15107692307692308,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2746407985687256,
|
|
"step": 1964
|
|
},
|
|
{
|
|
"epoch": 0.15115384615384617,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2464892864227295,
|
|
"step": 1965
|
|
},
|
|
{
|
|
"epoch": 0.15123076923076922,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26655650138855,
|
|
"step": 1966
|
|
},
|
|
{
|
|
"epoch": 0.1513076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2945399284362793,
|
|
"step": 1967
|
|
},
|
|
{
|
|
"epoch": 0.1513846153846154,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.25119948387146,
|
|
"step": 1968
|
|
},
|
|
{
|
|
"epoch": 0.15146153846153845,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27756667137146,
|
|
"step": 1969
|
|
},
|
|
{
|
|
"epoch": 0.15153846153846154,
|
|
"grad_norm": 0.056396484375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3048620223999023,
|
|
"step": 1970
|
|
},
|
|
{
|
|
"epoch": 0.15161538461538462,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2826743125915527,
|
|
"step": 1971
|
|
},
|
|
{
|
|
"epoch": 0.15169230769230768,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272800922393799,
|
|
"step": 1972
|
|
},
|
|
{
|
|
"epoch": 0.15176923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2565670013427734,
|
|
"step": 1973
|
|
},
|
|
{
|
|
"epoch": 0.15184615384615385,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.233677625656128,
|
|
"step": 1974
|
|
},
|
|
{
|
|
"epoch": 0.1519230769230769,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3020784854888916,
|
|
"step": 1975
|
|
},
|
|
{
|
|
"epoch": 0.152,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2980756759643555,
|
|
"step": 1976
|
|
},
|
|
{
|
|
"epoch": 0.15207692307692308,
|
|
"grad_norm": 0.06005859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.314485788345337,
|
|
"step": 1977
|
|
},
|
|
{
|
|
"epoch": 0.15215384615384617,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.29050874710083,
|
|
"step": 1978
|
|
},
|
|
{
|
|
"epoch": 0.15223076923076923,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2333693504333496,
|
|
"step": 1979
|
|
},
|
|
{
|
|
"epoch": 0.1523076923076923,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2860894203186035,
|
|
"step": 1980
|
|
},
|
|
{
|
|
"epoch": 0.1523846153846154,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2806735038757324,
|
|
"step": 1981
|
|
},
|
|
{
|
|
"epoch": 0.15246153846153845,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2486259937286377,
|
|
"step": 1982
|
|
},
|
|
{
|
|
"epoch": 0.15253846153846154,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2419447898864746,
|
|
"step": 1983
|
|
},
|
|
{
|
|
"epoch": 0.15261538461538462,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269508123397827,
|
|
"step": 1984
|
|
},
|
|
{
|
|
"epoch": 0.15269230769230768,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273178815841675,
|
|
"step": 1985
|
|
},
|
|
{
|
|
"epoch": 0.15276923076923077,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.277156114578247,
|
|
"step": 1986
|
|
},
|
|
{
|
|
"epoch": 0.15284615384615385,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.246988296508789,
|
|
"step": 1987
|
|
},
|
|
{
|
|
"epoch": 0.1529230769230769,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.236790657043457,
|
|
"step": 1988
|
|
},
|
|
{
|
|
"epoch": 0.153,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2617545127868652,
|
|
"step": 1989
|
|
},
|
|
{
|
|
"epoch": 0.15307692307692308,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3075740337371826,
|
|
"step": 1990
|
|
},
|
|
{
|
|
"epoch": 0.15315384615384614,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.256378173828125,
|
|
"step": 1991
|
|
},
|
|
{
|
|
"epoch": 0.15323076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2648181915283203,
|
|
"step": 1992
|
|
},
|
|
{
|
|
"epoch": 0.1533076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279500961303711,
|
|
"step": 1993
|
|
},
|
|
{
|
|
"epoch": 0.1533846153846154,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283208131790161,
|
|
"step": 1994
|
|
},
|
|
{
|
|
"epoch": 0.15346153846153845,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2594995498657227,
|
|
"step": 1995
|
|
},
|
|
{
|
|
"epoch": 0.15353846153846154,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2644295692443848,
|
|
"step": 1996
|
|
},
|
|
{
|
|
"epoch": 0.15361538461538463,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2948427200317383,
|
|
"step": 1997
|
|
},
|
|
{
|
|
"epoch": 0.15369230769230768,
|
|
"grad_norm": 0.04638671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265317440032959,
|
|
"step": 1998
|
|
},
|
|
{
|
|
"epoch": 0.15376923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.261195182800293,
|
|
"step": 1999
|
|
},
|
|
{
|
|
"epoch": 0.15384615384615385,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2672688961029053,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.1539230769230769,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.289055109024048,
|
|
"step": 2001
|
|
},
|
|
{
|
|
"epoch": 0.154,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.268690347671509,
|
|
"step": 2002
|
|
},
|
|
{
|
|
"epoch": 0.15407692307692308,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2621490955352783,
|
|
"step": 2003
|
|
},
|
|
{
|
|
"epoch": 0.15415384615384614,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274840831756592,
|
|
"step": 2004
|
|
},
|
|
{
|
|
"epoch": 0.15423076923076923,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2896721363067627,
|
|
"step": 2005
|
|
},
|
|
{
|
|
"epoch": 0.1543076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269467353820801,
|
|
"step": 2006
|
|
},
|
|
{
|
|
"epoch": 0.1543846153846154,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2868263721466064,
|
|
"step": 2007
|
|
},
|
|
{
|
|
"epoch": 0.15446153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2784883975982666,
|
|
"step": 2008
|
|
},
|
|
{
|
|
"epoch": 0.15453846153846154,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2453572750091553,
|
|
"step": 2009
|
|
},
|
|
{
|
|
"epoch": 0.15461538461538463,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2872321605682373,
|
|
"step": 2010
|
|
},
|
|
{
|
|
"epoch": 0.15469230769230768,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271270990371704,
|
|
"step": 2011
|
|
},
|
|
{
|
|
"epoch": 0.15476923076923077,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2491512298583984,
|
|
"step": 2012
|
|
},
|
|
{
|
|
"epoch": 0.15484615384615386,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2896981239318848,
|
|
"step": 2013
|
|
},
|
|
{
|
|
"epoch": 0.1549230769230769,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2706031799316406,
|
|
"step": 2014
|
|
},
|
|
{
|
|
"epoch": 0.155,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2481513023376465,
|
|
"step": 2015
|
|
},
|
|
{
|
|
"epoch": 0.15507692307692308,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267051935195923,
|
|
"step": 2016
|
|
},
|
|
{
|
|
"epoch": 0.15515384615384614,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2470664978027344,
|
|
"step": 2017
|
|
},
|
|
{
|
|
"epoch": 0.15523076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2602787017822266,
|
|
"step": 2018
|
|
},
|
|
{
|
|
"epoch": 0.1553076923076923,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2681219577789307,
|
|
"step": 2019
|
|
},
|
|
{
|
|
"epoch": 0.15538461538461537,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2746639251708984,
|
|
"step": 2020
|
|
},
|
|
{
|
|
"epoch": 0.15546153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2627737522125244,
|
|
"step": 2021
|
|
},
|
|
{
|
|
"epoch": 0.15553846153846154,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264252185821533,
|
|
"step": 2022
|
|
},
|
|
{
|
|
"epoch": 0.15561538461538463,
|
|
"grad_norm": 0.046630859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2676703929901123,
|
|
"step": 2023
|
|
},
|
|
{
|
|
"epoch": 0.15569230769230769,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254221200942993,
|
|
"step": 2024
|
|
},
|
|
{
|
|
"epoch": 0.15576923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2634100914001465,
|
|
"step": 2025
|
|
},
|
|
{
|
|
"epoch": 0.15584615384615386,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2772440910339355,
|
|
"step": 2026
|
|
},
|
|
{
|
|
"epoch": 0.15592307692307691,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27101993560791,
|
|
"step": 2027
|
|
},
|
|
{
|
|
"epoch": 0.156,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2323484420776367,
|
|
"step": 2028
|
|
},
|
|
{
|
|
"epoch": 0.15607692307692309,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.217000722885132,
|
|
"step": 2029
|
|
},
|
|
{
|
|
"epoch": 0.15615384615384614,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2774040699005127,
|
|
"step": 2030
|
|
},
|
|
{
|
|
"epoch": 0.15623076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2604198455810547,
|
|
"step": 2031
|
|
},
|
|
{
|
|
"epoch": 0.15630769230769231,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2874064445495605,
|
|
"step": 2032
|
|
},
|
|
{
|
|
"epoch": 0.15638461538461537,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2462234497070312,
|
|
"step": 2033
|
|
},
|
|
{
|
|
"epoch": 0.15646153846153846,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2650601863861084,
|
|
"step": 2034
|
|
},
|
|
{
|
|
"epoch": 0.15653846153846154,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2701189517974854,
|
|
"step": 2035
|
|
},
|
|
{
|
|
"epoch": 0.15661538461538463,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3030612468719482,
|
|
"step": 2036
|
|
},
|
|
{
|
|
"epoch": 0.1566923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2683217525482178,
|
|
"step": 2037
|
|
},
|
|
{
|
|
"epoch": 0.15676923076923077,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.230518102645874,
|
|
"step": 2038
|
|
},
|
|
{
|
|
"epoch": 0.15684615384615386,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2454538345336914,
|
|
"step": 2039
|
|
},
|
|
{
|
|
"epoch": 0.15692307692307692,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2565059661865234,
|
|
"step": 2040
|
|
},
|
|
{
|
|
"epoch": 0.157,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.249058246612549,
|
|
"step": 2041
|
|
},
|
|
{
|
|
"epoch": 0.1570769230769231,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.286410331726074,
|
|
"step": 2042
|
|
},
|
|
{
|
|
"epoch": 0.15715384615384614,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.237051010131836,
|
|
"step": 2043
|
|
},
|
|
{
|
|
"epoch": 0.15723076923076923,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264281749725342,
|
|
"step": 2044
|
|
},
|
|
{
|
|
"epoch": 0.15730769230769232,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2564070224761963,
|
|
"step": 2045
|
|
},
|
|
{
|
|
"epoch": 0.15738461538461537,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.270683526992798,
|
|
"step": 2046
|
|
},
|
|
{
|
|
"epoch": 0.15746153846153846,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.292073965072632,
|
|
"step": 2047
|
|
},
|
|
{
|
|
"epoch": 0.15753846153846154,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263036012649536,
|
|
"step": 2048
|
|
},
|
|
{
|
|
"epoch": 0.1576153846153846,
|
|
"grad_norm": 0.046630859375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2399768829345703,
|
|
"step": 2049
|
|
},
|
|
{
|
|
"epoch": 0.1576923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254948377609253,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 0.15776923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.275545597076416,
|
|
"step": 2051
|
|
},
|
|
{
|
|
"epoch": 0.15784615384615386,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.268822193145752,
|
|
"step": 2052
|
|
},
|
|
{
|
|
"epoch": 0.15792307692307692,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.274296760559082,
|
|
"step": 2053
|
|
},
|
|
{
|
|
"epoch": 0.158,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.229806423187256,
|
|
"step": 2054
|
|
},
|
|
{
|
|
"epoch": 0.1580769230769231,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2744927406311035,
|
|
"step": 2055
|
|
},
|
|
{
|
|
"epoch": 0.15815384615384614,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2844972610473633,
|
|
"step": 2056
|
|
},
|
|
{
|
|
"epoch": 0.15823076923076923,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2578840255737305,
|
|
"step": 2057
|
|
},
|
|
{
|
|
"epoch": 0.15830769230769232,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254976511001587,
|
|
"step": 2058
|
|
},
|
|
{
|
|
"epoch": 0.15838461538461537,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2789008617401123,
|
|
"step": 2059
|
|
},
|
|
{
|
|
"epoch": 0.15846153846153846,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.225954532623291,
|
|
"step": 2060
|
|
},
|
|
{
|
|
"epoch": 0.15853846153846154,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.25264048576355,
|
|
"step": 2061
|
|
},
|
|
{
|
|
"epoch": 0.1586153846153846,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.253605365753174,
|
|
"step": 2062
|
|
},
|
|
{
|
|
"epoch": 0.1586923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260505199432373,
|
|
"step": 2063
|
|
},
|
|
{
|
|
"epoch": 0.15876923076923077,
|
|
"grad_norm": 0.04638671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2449944019317627,
|
|
"step": 2064
|
|
},
|
|
{
|
|
"epoch": 0.15884615384615386,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.285703182220459,
|
|
"step": 2065
|
|
},
|
|
{
|
|
"epoch": 0.15892307692307692,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283221960067749,
|
|
"step": 2066
|
|
},
|
|
{
|
|
"epoch": 0.159,
|
|
"grad_norm": 0.05322265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2681379318237305,
|
|
"step": 2067
|
|
},
|
|
{
|
|
"epoch": 0.1590769230769231,
|
|
"grad_norm": 0.055419921875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280773878097534,
|
|
"step": 2068
|
|
},
|
|
{
|
|
"epoch": 0.15915384615384615,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265123128890991,
|
|
"step": 2069
|
|
},
|
|
{
|
|
"epoch": 0.15923076923076923,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.262887954711914,
|
|
"step": 2070
|
|
},
|
|
{
|
|
"epoch": 0.15930769230769232,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.236272096633911,
|
|
"step": 2071
|
|
},
|
|
{
|
|
"epoch": 0.15938461538461537,
|
|
"grad_norm": 0.052001953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258493185043335,
|
|
"step": 2072
|
|
},
|
|
{
|
|
"epoch": 0.15946153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252685546875,
|
|
"step": 2073
|
|
},
|
|
{
|
|
"epoch": 0.15953846153846155,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.273353338241577,
|
|
"step": 2074
|
|
},
|
|
{
|
|
"epoch": 0.1596153846153846,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.249328851699829,
|
|
"step": 2075
|
|
},
|
|
{
|
|
"epoch": 0.1596923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.277662754058838,
|
|
"step": 2076
|
|
},
|
|
{
|
|
"epoch": 0.15976923076923077,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.294999599456787,
|
|
"step": 2077
|
|
},
|
|
{
|
|
"epoch": 0.15984615384615386,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2736849784851074,
|
|
"step": 2078
|
|
},
|
|
{
|
|
"epoch": 0.15992307692307692,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.238288640975952,
|
|
"step": 2079
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282778739929199,
|
|
"step": 2080
|
|
},
|
|
{
|
|
"epoch": 0.1600769230769231,
|
|
"grad_norm": 0.060302734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264030694961548,
|
|
"step": 2081
|
|
},
|
|
{
|
|
"epoch": 0.16015384615384615,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254776954650879,
|
|
"step": 2082
|
|
},
|
|
{
|
|
"epoch": 0.16023076923076923,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267033815383911,
|
|
"step": 2083
|
|
},
|
|
{
|
|
"epoch": 0.16030769230769232,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2629544734954834,
|
|
"step": 2084
|
|
},
|
|
{
|
|
"epoch": 0.16038461538461538,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2406206130981445,
|
|
"step": 2085
|
|
},
|
|
{
|
|
"epoch": 0.16046153846153846,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.255542039871216,
|
|
"step": 2086
|
|
},
|
|
{
|
|
"epoch": 0.16053846153846155,
|
|
"grad_norm": 0.053466796875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2696950435638428,
|
|
"step": 2087
|
|
},
|
|
{
|
|
"epoch": 0.1606153846153846,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2823824882507324,
|
|
"step": 2088
|
|
},
|
|
{
|
|
"epoch": 0.1606923076923077,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.261863946914673,
|
|
"step": 2089
|
|
},
|
|
{
|
|
"epoch": 0.16076923076923078,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276240348815918,
|
|
"step": 2090
|
|
},
|
|
{
|
|
"epoch": 0.16084615384615383,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.301727771759033,
|
|
"step": 2091
|
|
},
|
|
{
|
|
"epoch": 0.16092307692307692,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264112949371338,
|
|
"step": 2092
|
|
},
|
|
{
|
|
"epoch": 0.161,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2914299964904785,
|
|
"step": 2093
|
|
},
|
|
{
|
|
"epoch": 0.1610769230769231,
|
|
"grad_norm": 0.04638671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2537949085235596,
|
|
"step": 2094
|
|
},
|
|
{
|
|
"epoch": 0.16115384615384615,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2659800052642822,
|
|
"step": 2095
|
|
},
|
|
{
|
|
"epoch": 0.16123076923076923,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2859396934509277,
|
|
"step": 2096
|
|
},
|
|
{
|
|
"epoch": 0.16130769230769232,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2577476501464844,
|
|
"step": 2097
|
|
},
|
|
{
|
|
"epoch": 0.16138461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2899820804595947,
|
|
"step": 2098
|
|
},
|
|
{
|
|
"epoch": 0.16146153846153846,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.25826358795166,
|
|
"step": 2099
|
|
},
|
|
{
|
|
"epoch": 0.16153846153846155,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2655086517333984,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 0.1616153846153846,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.269667148590088,
|
|
"step": 2101
|
|
},
|
|
{
|
|
"epoch": 0.1616923076923077,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2950732707977295,
|
|
"step": 2102
|
|
},
|
|
{
|
|
"epoch": 0.16176923076923078,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2657103538513184,
|
|
"step": 2103
|
|
},
|
|
{
|
|
"epoch": 0.16184615384615383,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2467617988586426,
|
|
"step": 2104
|
|
},
|
|
{
|
|
"epoch": 0.16192307692307692,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2122344970703125,
|
|
"step": 2105
|
|
},
|
|
{
|
|
"epoch": 0.162,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2707016468048096,
|
|
"step": 2106
|
|
},
|
|
{
|
|
"epoch": 0.1620769230769231,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2976884841918945,
|
|
"step": 2107
|
|
},
|
|
{
|
|
"epoch": 0.16215384615384615,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287379741668701,
|
|
"step": 2108
|
|
},
|
|
{
|
|
"epoch": 0.16223076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2386512756347656,
|
|
"step": 2109
|
|
},
|
|
{
|
|
"epoch": 0.16230769230769232,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2467823028564453,
|
|
"step": 2110
|
|
},
|
|
{
|
|
"epoch": 0.16238461538461538,
|
|
"grad_norm": 0.052978515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2786967754364014,
|
|
"step": 2111
|
|
},
|
|
{
|
|
"epoch": 0.16246153846153846,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.247628927230835,
|
|
"step": 2112
|
|
},
|
|
{
|
|
"epoch": 0.16253846153846155,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2741966247558594,
|
|
"step": 2113
|
|
},
|
|
{
|
|
"epoch": 0.1626153846153846,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2584421634674072,
|
|
"step": 2114
|
|
},
|
|
{
|
|
"epoch": 0.1626923076923077,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.252009391784668,
|
|
"step": 2115
|
|
},
|
|
{
|
|
"epoch": 0.16276923076923078,
|
|
"grad_norm": 0.051025390625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2879483699798584,
|
|
"step": 2116
|
|
},
|
|
{
|
|
"epoch": 0.16284615384615383,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.24045991897583,
|
|
"step": 2117
|
|
},
|
|
{
|
|
"epoch": 0.16292307692307692,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2728869915008545,
|
|
"step": 2118
|
|
},
|
|
{
|
|
"epoch": 0.163,
|
|
"grad_norm": 0.05517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2588205337524414,
|
|
"step": 2119
|
|
},
|
|
{
|
|
"epoch": 0.16307692307692306,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2482388019561768,
|
|
"step": 2120
|
|
},
|
|
{
|
|
"epoch": 0.16315384615384615,
|
|
"grad_norm": 0.052490234375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.279491901397705,
|
|
"step": 2121
|
|
},
|
|
{
|
|
"epoch": 0.16323076923076923,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27345871925354,
|
|
"step": 2122
|
|
},
|
|
{
|
|
"epoch": 0.16330769230769232,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.28446626663208,
|
|
"step": 2123
|
|
},
|
|
{
|
|
"epoch": 0.16338461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2640860080718994,
|
|
"step": 2124
|
|
},
|
|
{
|
|
"epoch": 0.16346153846153846,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2952218055725098,
|
|
"step": 2125
|
|
},
|
|
{
|
|
"epoch": 0.16353846153846155,
|
|
"grad_norm": 0.05029296875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.281426191329956,
|
|
"step": 2126
|
|
},
|
|
{
|
|
"epoch": 0.1636153846153846,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2729663848876953,
|
|
"step": 2127
|
|
},
|
|
{
|
|
"epoch": 0.1636923076923077,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2768783569335938,
|
|
"step": 2128
|
|
},
|
|
{
|
|
"epoch": 0.16376923076923078,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276730537414551,
|
|
"step": 2129
|
|
},
|
|
{
|
|
"epoch": 0.16384615384615384,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.254751443862915,
|
|
"step": 2130
|
|
},
|
|
{
|
|
"epoch": 0.16392307692307692,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2671515941619873,
|
|
"step": 2131
|
|
},
|
|
{
|
|
"epoch": 0.164,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2619895935058594,
|
|
"step": 2132
|
|
},
|
|
{
|
|
"epoch": 0.16407692307692306,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2611937522888184,
|
|
"step": 2133
|
|
},
|
|
{
|
|
"epoch": 0.16415384615384615,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2856714725494385,
|
|
"step": 2134
|
|
},
|
|
{
|
|
"epoch": 0.16423076923076924,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2483632564544678,
|
|
"step": 2135
|
|
},
|
|
{
|
|
"epoch": 0.16430769230769232,
|
|
"grad_norm": 0.050537109375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2388885021209717,
|
|
"step": 2136
|
|
},
|
|
{
|
|
"epoch": 0.16438461538461538,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2851388454437256,
|
|
"step": 2137
|
|
},
|
|
{
|
|
"epoch": 0.16446153846153846,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.3262550830841064,
|
|
"step": 2138
|
|
},
|
|
{
|
|
"epoch": 0.16453846153846155,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.266754388809204,
|
|
"step": 2139
|
|
},
|
|
{
|
|
"epoch": 0.1646153846153846,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.227847099304199,
|
|
"step": 2140
|
|
},
|
|
{
|
|
"epoch": 0.1646923076923077,
|
|
"grad_norm": 0.052734375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.275742769241333,
|
|
"step": 2141
|
|
},
|
|
{
|
|
"epoch": 0.16476923076923078,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2844882011413574,
|
|
"step": 2142
|
|
},
|
|
{
|
|
"epoch": 0.16484615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.30678391456604,
|
|
"step": 2143
|
|
},
|
|
{
|
|
"epoch": 0.16492307692307692,
|
|
"grad_norm": 0.04638671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2605345249176025,
|
|
"step": 2144
|
|
},
|
|
{
|
|
"epoch": 0.165,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.237731456756592,
|
|
"step": 2145
|
|
},
|
|
{
|
|
"epoch": 0.16507692307692307,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2550652027130127,
|
|
"step": 2146
|
|
},
|
|
{
|
|
"epoch": 0.16515384615384615,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.258378505706787,
|
|
"step": 2147
|
|
},
|
|
{
|
|
"epoch": 0.16523076923076924,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.293205738067627,
|
|
"step": 2148
|
|
},
|
|
{
|
|
"epoch": 0.1653076923076923,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280923366546631,
|
|
"step": 2149
|
|
},
|
|
{
|
|
"epoch": 0.16538461538461538,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.264120578765869,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 0.16546153846153847,
|
|
"grad_norm": 0.051513671875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.267935276031494,
|
|
"step": 2151
|
|
},
|
|
{
|
|
"epoch": 0.16553846153846155,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2810113430023193,
|
|
"step": 2152
|
|
},
|
|
{
|
|
"epoch": 0.1656153846153846,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.257805824279785,
|
|
"step": 2153
|
|
},
|
|
{
|
|
"epoch": 0.1656923076923077,
|
|
"grad_norm": 0.049560546875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271151065826416,
|
|
"step": 2154
|
|
},
|
|
{
|
|
"epoch": 0.16576923076923078,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2524173259735107,
|
|
"step": 2155
|
|
},
|
|
{
|
|
"epoch": 0.16584615384615384,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2825124263763428,
|
|
"step": 2156
|
|
},
|
|
{
|
|
"epoch": 0.16592307692307692,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2860419750213623,
|
|
"step": 2157
|
|
},
|
|
{
|
|
"epoch": 0.166,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263126850128174,
|
|
"step": 2158
|
|
},
|
|
{
|
|
"epoch": 0.16607692307692307,
|
|
"grad_norm": 0.04736328125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2514429092407227,
|
|
"step": 2159
|
|
},
|
|
{
|
|
"epoch": 0.16615384615384615,
|
|
"grad_norm": 0.048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2560830116271973,
|
|
"step": 2160
|
|
},
|
|
{
|
|
"epoch": 0.16623076923076924,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.260608196258545,
|
|
"step": 2161
|
|
},
|
|
{
|
|
"epoch": 0.1663076923076923,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2771072387695312,
|
|
"step": 2162
|
|
},
|
|
{
|
|
"epoch": 0.16638461538461538,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2711105346679688,
|
|
"step": 2163
|
|
},
|
|
{
|
|
"epoch": 0.16646153846153847,
|
|
"grad_norm": 0.05224609375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2795956134796143,
|
|
"step": 2164
|
|
},
|
|
{
|
|
"epoch": 0.16653846153846155,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.250424861907959,
|
|
"step": 2165
|
|
},
|
|
{
|
|
"epoch": 0.1666153846153846,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.287231683731079,
|
|
"step": 2166
|
|
},
|
|
{
|
|
"epoch": 0.1666923076923077,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2814106941223145,
|
|
"step": 2167
|
|
},
|
|
{
|
|
"epoch": 0.16676923076923078,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2628066539764404,
|
|
"step": 2168
|
|
},
|
|
{
|
|
"epoch": 0.16684615384615384,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276034116744995,
|
|
"step": 2169
|
|
},
|
|
{
|
|
"epoch": 0.16692307692307692,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2442615032196045,
|
|
"step": 2170
|
|
},
|
|
{
|
|
"epoch": 0.167,
|
|
"grad_norm": 0.0517578125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304964065551758,
|
|
"step": 2171
|
|
},
|
|
{
|
|
"epoch": 0.16707692307692307,
|
|
"grad_norm": 0.047119140625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2779390811920166,
|
|
"step": 2172
|
|
},
|
|
{
|
|
"epoch": 0.16715384615384615,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.276973247528076,
|
|
"step": 2173
|
|
},
|
|
{
|
|
"epoch": 0.16723076923076924,
|
|
"grad_norm": 0.048583984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2743124961853027,
|
|
"step": 2174
|
|
},
|
|
{
|
|
"epoch": 0.1673076923076923,
|
|
"grad_norm": 0.050048828125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2344045639038086,
|
|
"step": 2175
|
|
},
|
|
{
|
|
"epoch": 0.16738461538461538,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2733066082000732,
|
|
"step": 2176
|
|
},
|
|
{
|
|
"epoch": 0.16746153846153847,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2659289836883545,
|
|
"step": 2177
|
|
},
|
|
{
|
|
"epoch": 0.16753846153846153,
|
|
"grad_norm": 0.047607421875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2574241161346436,
|
|
"step": 2178
|
|
},
|
|
{
|
|
"epoch": 0.1676153846153846,
|
|
"grad_norm": 0.0478515625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2506704330444336,
|
|
"step": 2179
|
|
},
|
|
{
|
|
"epoch": 0.1676923076923077,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263676643371582,
|
|
"step": 2180
|
|
},
|
|
{
|
|
"epoch": 0.16776923076923078,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.272334337234497,
|
|
"step": 2181
|
|
},
|
|
{
|
|
"epoch": 0.16784615384615384,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.304509162902832,
|
|
"step": 2182
|
|
},
|
|
{
|
|
"epoch": 0.16792307692307692,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.283642292022705,
|
|
"step": 2183
|
|
},
|
|
{
|
|
"epoch": 0.168,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.244910955429077,
|
|
"step": 2184
|
|
},
|
|
{
|
|
"epoch": 0.16807692307692307,
|
|
"grad_norm": 0.055908203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2699849605560303,
|
|
"step": 2185
|
|
},
|
|
{
|
|
"epoch": 0.16815384615384615,
|
|
"grad_norm": 0.05712890625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2659099102020264,
|
|
"step": 2186
|
|
},
|
|
{
|
|
"epoch": 0.16823076923076924,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.263134241104126,
|
|
"step": 2187
|
|
},
|
|
{
|
|
"epoch": 0.1683076923076923,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.280508279800415,
|
|
"step": 2188
|
|
},
|
|
{
|
|
"epoch": 0.16838461538461538,
|
|
"grad_norm": 0.04833984375,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.282924175262451,
|
|
"step": 2189
|
|
},
|
|
{
|
|
"epoch": 0.16846153846153847,
|
|
"grad_norm": 0.049072265625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.26344633102417,
|
|
"step": 2190
|
|
},
|
|
{
|
|
"epoch": 0.16853846153846153,
|
|
"grad_norm": 0.05078125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.277963876724243,
|
|
"step": 2191
|
|
},
|
|
{
|
|
"epoch": 0.1686153846153846,
|
|
"grad_norm": 0.0498046875,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.265225410461426,
|
|
"step": 2192
|
|
},
|
|
{
|
|
"epoch": 0.1686923076923077,
|
|
"grad_norm": 0.05126953125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.271087408065796,
|
|
"step": 2193
|
|
},
|
|
{
|
|
"epoch": 0.16876923076923078,
|
|
"grad_norm": 0.048095703125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.27451491355896,
|
|
"step": 2194
|
|
},
|
|
{
|
|
"epoch": 0.16884615384615384,
|
|
"grad_norm": 0.04931640625,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.2593960762023926,
|
|
"step": 2195
|
|
},
|
|
{
|
|
"epoch": 0.16892307692307693,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 0.001,
|
|
"loss": 2.7150745391845703,
|
|
"step": 2196
|
|
},
|
|
{
|
|
"epoch": 0.169,
|
|
"grad_norm": 18.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.667558193206787,
|
|
"step": 2197
|
|
},
|
|
{
|
|
"epoch": 0.16907692307692307,
|
|
"grad_norm": 20.75,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.2370219230651855,
|
|
"step": 2198
|
|
},
|
|
{
|
|
"epoch": 0.16915384615384615,
|
|
"grad_norm": 23.875,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.5040693283081055,
|
|
"step": 2199
|
|
},
|
|
{
|
|
"epoch": 0.16923076923076924,
|
|
"grad_norm": 23.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.672176361083984,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 0.1693076923076923,
|
|
"grad_norm": 22.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.429105281829834,
|
|
"step": 2201
|
|
},
|
|
{
|
|
"epoch": 0.16938461538461538,
|
|
"grad_norm": 23.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.442510604858398,
|
|
"step": 2202
|
|
},
|
|
{
|
|
"epoch": 0.16946153846153847,
|
|
"grad_norm": 20.875,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.38027811050415,
|
|
"step": 2203
|
|
},
|
|
{
|
|
"epoch": 0.16953846153846153,
|
|
"grad_norm": 22.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.262911319732666,
|
|
"step": 2204
|
|
},
|
|
{
|
|
"epoch": 0.1696153846153846,
|
|
"grad_norm": 20.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.336555004119873,
|
|
"step": 2205
|
|
},
|
|
{
|
|
"epoch": 0.1696923076923077,
|
|
"grad_norm": 20.75,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.151665687561035,
|
|
"step": 2206
|
|
},
|
|
{
|
|
"epoch": 0.16976923076923076,
|
|
"grad_norm": 20.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.23635196685791,
|
|
"step": 2207
|
|
},
|
|
{
|
|
"epoch": 0.16984615384615384,
|
|
"grad_norm": 19.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.0901055335998535,
|
|
"step": 2208
|
|
},
|
|
{
|
|
"epoch": 0.16992307692307693,
|
|
"grad_norm": 20.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 6.039085865020752,
|
|
"step": 2209
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"grad_norm": 18.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.953739166259766,
|
|
"step": 2210
|
|
},
|
|
{
|
|
"epoch": 0.17007692307692307,
|
|
"grad_norm": 19.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.840315341949463,
|
|
"step": 2211
|
|
},
|
|
{
|
|
"epoch": 0.17015384615384616,
|
|
"grad_norm": 18.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.894529342651367,
|
|
"step": 2212
|
|
},
|
|
{
|
|
"epoch": 0.17023076923076924,
|
|
"grad_norm": 18.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.711856842041016,
|
|
"step": 2213
|
|
},
|
|
{
|
|
"epoch": 0.1703076923076923,
|
|
"grad_norm": 18.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.764537334442139,
|
|
"step": 2214
|
|
},
|
|
{
|
|
"epoch": 0.17038461538461538,
|
|
"grad_norm": 17.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.683034420013428,
|
|
"step": 2215
|
|
},
|
|
{
|
|
"epoch": 0.17046153846153847,
|
|
"grad_norm": 18.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.540406227111816,
|
|
"step": 2216
|
|
},
|
|
{
|
|
"epoch": 0.17053846153846153,
|
|
"grad_norm": 16.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.602969169616699,
|
|
"step": 2217
|
|
},
|
|
{
|
|
"epoch": 0.1706153846153846,
|
|
"grad_norm": 17.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.365934371948242,
|
|
"step": 2218
|
|
},
|
|
{
|
|
"epoch": 0.1706923076923077,
|
|
"grad_norm": 16.875,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.528835773468018,
|
|
"step": 2219
|
|
},
|
|
{
|
|
"epoch": 0.17076923076923076,
|
|
"grad_norm": 16.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.339080333709717,
|
|
"step": 2220
|
|
},
|
|
{
|
|
"epoch": 0.17084615384615384,
|
|
"grad_norm": 16.75,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.271631240844727,
|
|
"step": 2221
|
|
},
|
|
{
|
|
"epoch": 0.17092307692307693,
|
|
"grad_norm": 15.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.343720436096191,
|
|
"step": 2222
|
|
},
|
|
{
|
|
"epoch": 0.171,
|
|
"grad_norm": 16.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.032788276672363,
|
|
"step": 2223
|
|
},
|
|
{
|
|
"epoch": 0.17107692307692307,
|
|
"grad_norm": 15.5625,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.2195658683776855,
|
|
"step": 2224
|
|
},
|
|
{
|
|
"epoch": 0.17115384615384616,
|
|
"grad_norm": 15.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.088586330413818,
|
|
"step": 2225
|
|
},
|
|
{
|
|
"epoch": 0.17123076923076924,
|
|
"grad_norm": 15.6875,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.063693523406982,
|
|
"step": 2226
|
|
},
|
|
{
|
|
"epoch": 0.1713076923076923,
|
|
"grad_norm": 14.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.074687480926514,
|
|
"step": 2227
|
|
},
|
|
{
|
|
"epoch": 0.17138461538461539,
|
|
"grad_norm": 15.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.895683765411377,
|
|
"step": 2228
|
|
},
|
|
{
|
|
"epoch": 0.17146153846153847,
|
|
"grad_norm": 14.1875,
|
|
"learning_rate": 0.001,
|
|
"loss": 5.015110969543457,
|
|
"step": 2229
|
|
},
|
|
{
|
|
"epoch": 0.17153846153846153,
|
|
"grad_norm": 14.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.795098781585693,
|
|
"step": 2230
|
|
},
|
|
{
|
|
"epoch": 0.17161538461538461,
|
|
"grad_norm": 14.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.907761096954346,
|
|
"step": 2231
|
|
},
|
|
{
|
|
"epoch": 0.1716923076923077,
|
|
"grad_norm": 13.8125,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.786104679107666,
|
|
"step": 2232
|
|
},
|
|
{
|
|
"epoch": 0.17176923076923076,
|
|
"grad_norm": 14.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.677491188049316,
|
|
"step": 2233
|
|
},
|
|
{
|
|
"epoch": 0.17184615384615384,
|
|
"grad_norm": 13.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.765591144561768,
|
|
"step": 2234
|
|
},
|
|
{
|
|
"epoch": 0.17192307692307693,
|
|
"grad_norm": 13.9375,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.557199001312256,
|
|
"step": 2235
|
|
},
|
|
{
|
|
"epoch": 0.172,
|
|
"grad_norm": 13.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.724630355834961,
|
|
"step": 2236
|
|
},
|
|
{
|
|
"epoch": 0.17207692307692307,
|
|
"grad_norm": 13.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.509766578674316,
|
|
"step": 2237
|
|
},
|
|
{
|
|
"epoch": 0.17215384615384616,
|
|
"grad_norm": 13.1875,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.545156002044678,
|
|
"step": 2238
|
|
},
|
|
{
|
|
"epoch": 0.17223076923076924,
|
|
"grad_norm": 12.5625,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.523886203765869,
|
|
"step": 2239
|
|
},
|
|
{
|
|
"epoch": 0.1723076923076923,
|
|
"grad_norm": 13.3125,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.378265857696533,
|
|
"step": 2240
|
|
},
|
|
{
|
|
"epoch": 0.1723846153846154,
|
|
"grad_norm": 12.1875,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.571813583374023,
|
|
"step": 2241
|
|
},
|
|
{
|
|
"epoch": 0.17246153846153847,
|
|
"grad_norm": 12.75,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.290168762207031,
|
|
"step": 2242
|
|
},
|
|
{
|
|
"epoch": 0.17253846153846153,
|
|
"grad_norm": 12.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.4424309730529785,
|
|
"step": 2243
|
|
},
|
|
{
|
|
"epoch": 0.17261538461538461,
|
|
"grad_norm": 12.0,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.301406383514404,
|
|
"step": 2244
|
|
},
|
|
{
|
|
"epoch": 0.1726923076923077,
|
|
"grad_norm": 12.3125,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.285731315612793,
|
|
"step": 2245
|
|
},
|
|
{
|
|
"epoch": 0.17276923076923076,
|
|
"grad_norm": 11.375,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.296684265136719,
|
|
"step": 2246
|
|
},
|
|
{
|
|
"epoch": 0.17284615384615384,
|
|
"grad_norm": 11.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.111746311187744,
|
|
"step": 2247
|
|
},
|
|
{
|
|
"epoch": 0.17292307692307693,
|
|
"grad_norm": 11.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.252358913421631,
|
|
"step": 2248
|
|
},
|
|
{
|
|
"epoch": 0.173,
|
|
"grad_norm": 11.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.083554267883301,
|
|
"step": 2249
|
|
},
|
|
{
|
|
"epoch": 0.17307692307692307,
|
|
"grad_norm": 11.1875,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.205334663391113,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 0.17315384615384616,
|
|
"grad_norm": 10.5625,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.154065132141113,
|
|
"step": 2251
|
|
},
|
|
{
|
|
"epoch": 0.17323076923076924,
|
|
"grad_norm": 10.8125,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.016529083251953,
|
|
"step": 2252
|
|
},
|
|
{
|
|
"epoch": 0.1733076923076923,
|
|
"grad_norm": 10.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.161767959594727,
|
|
"step": 2253
|
|
},
|
|
{
|
|
"epoch": 0.1733846153846154,
|
|
"grad_norm": 10.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.8964240550994873,
|
|
"step": 2254
|
|
},
|
|
{
|
|
"epoch": 0.17346153846153847,
|
|
"grad_norm": 10.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 4.09371280670166,
|
|
"step": 2255
|
|
},
|
|
{
|
|
"epoch": 0.17353846153846153,
|
|
"grad_norm": 9.6875,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.9454965591430664,
|
|
"step": 2256
|
|
},
|
|
{
|
|
"epoch": 0.17361538461538462,
|
|
"grad_norm": 9.875,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.9787862300872803,
|
|
"step": 2257
|
|
},
|
|
{
|
|
"epoch": 0.1736923076923077,
|
|
"grad_norm": 9.4375,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.974217414855957,
|
|
"step": 2258
|
|
},
|
|
{
|
|
"epoch": 0.17376923076923076,
|
|
"grad_norm": 9.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.7777390480041504,
|
|
"step": 2259
|
|
},
|
|
{
|
|
"epoch": 0.17384615384615384,
|
|
"grad_norm": 9.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.9780449867248535,
|
|
"step": 2260
|
|
},
|
|
{
|
|
"epoch": 0.17392307692307693,
|
|
"grad_norm": 9.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.726745128631592,
|
|
"step": 2261
|
|
},
|
|
{
|
|
"epoch": 0.174,
|
|
"grad_norm": 9.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.945422649383545,
|
|
"step": 2262
|
|
},
|
|
{
|
|
"epoch": 0.17407692307692307,
|
|
"grad_norm": 8.5,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.761659860610962,
|
|
"step": 2263
|
|
},
|
|
{
|
|
"epoch": 0.17415384615384616,
|
|
"grad_norm": 8.75,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.8028013706207275,
|
|
"step": 2264
|
|
},
|
|
{
|
|
"epoch": 0.17423076923076924,
|
|
"grad_norm": 8.25,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.8051154613494873,
|
|
"step": 2265
|
|
},
|
|
{
|
|
"epoch": 0.1743076923076923,
|
|
"grad_norm": 8.0625,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6156961917877197,
|
|
"step": 2266
|
|
},
|
|
{
|
|
"epoch": 0.1743846153846154,
|
|
"grad_norm": 8.125,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.8753151893615723,
|
|
"step": 2267
|
|
},
|
|
{
|
|
"epoch": 0.17446153846153847,
|
|
"grad_norm": 7.8125,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6171770095825195,
|
|
"step": 2268
|
|
},
|
|
{
|
|
"epoch": 0.17453846153846153,
|
|
"grad_norm": 7.90625,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.7530620098114014,
|
|
"step": 2269
|
|
},
|
|
{
|
|
"epoch": 0.17461538461538462,
|
|
"grad_norm": 7.53125,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.679172992706299,
|
|
"step": 2270
|
|
},
|
|
{
|
|
"epoch": 0.1746923076923077,
|
|
"grad_norm": 7.3125,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.523942232131958,
|
|
"step": 2271
|
|
},
|
|
{
|
|
"epoch": 0.17476923076923076,
|
|
"grad_norm": 7.1875,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.699639081954956,
|
|
"step": 2272
|
|
},
|
|
{
|
|
"epoch": 0.17484615384615385,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5431265830993652,
|
|
"step": 2273
|
|
},
|
|
{
|
|
"epoch": 0.17492307692307693,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.7048654556274414,
|
|
"step": 2274
|
|
},
|
|
{
|
|
"epoch": 0.175,
|
|
"grad_norm": 6.625,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.541247844696045,
|
|
"step": 2275
|
|
}
|
|
],
|
|
"logging_steps": 1,
|
|
"max_steps": 13000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 9223372036854775807,
|
|
"save_steps": 325,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 6.598890270847716e+18,
|
|
"train_batch_size": 4,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|