tomekkorbak's picture
Training in progress, step 1004
3a81cdc
raw
history blame
215 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.07975850015888147,
"global_step": 1004,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 7.936507936507937e-07,
"loss": 2.9309,
"theoretical_loss": 3.4868973533572363,
"tokens_seen": 1650130944
},
{
"epoch": 0.0,
"learning_rate": 1.5873015873015873e-06,
"loss": 3.0573,
"theoretical_loss": 3.4868733789973354,
"tokens_seen": 1650262016
},
{
"epoch": 0.0,
"learning_rate": 2.3809523809523808e-06,
"loss": 2.9048,
"theoretical_loss": 3.4868494070746388,
"tokens_seen": 1650393088
},
{
"epoch": 0.0,
"learning_rate": 3.1746031746031746e-06,
"loss": 3.1852,
"theoretical_loss": 3.4868254375887053,
"tokens_seen": 1650524160
},
{
"epoch": 0.0,
"learning_rate": 3.968253968253968e-06,
"loss": 2.9406,
"theoretical_loss": 3.4868014705390937,
"tokens_seen": 1650655232
},
{
"epoch": 0.0,
"learning_rate": 4.7619047619047615e-06,
"loss": 2.9301,
"theoretical_loss": 3.4867775059253625,
"tokens_seen": 1650786304
},
{
"epoch": 0.0,
"learning_rate": 5.555555555555556e-06,
"loss": 2.9694,
"theoretical_loss": 3.4867535437470716,
"tokens_seen": 1650917376
},
{
"epoch": 0.0,
"learning_rate": 6.349206349206349e-06,
"loss": 2.8537,
"theoretical_loss": 3.486729584003779,
"tokens_seen": 1651048448
},
{
"epoch": 0.0,
"learning_rate": 7.142857142857143e-06,
"loss": 3.1275,
"theoretical_loss": 3.4867056266950454,
"tokens_seen": 1651179520
},
{
"epoch": 0.0,
"learning_rate": 7.936507936507936e-06,
"loss": 2.8591,
"theoretical_loss": 3.4866816718204294,
"tokens_seen": 1651310592
},
{
"epoch": 0.0,
"learning_rate": 8.73015873015873e-06,
"loss": 2.9591,
"theoretical_loss": 3.4866577193794903,
"tokens_seen": 1651441664
},
{
"epoch": 0.0,
"learning_rate": 9.523809523809523e-06,
"loss": 2.9381,
"theoretical_loss": 3.486633769371788,
"tokens_seen": 1651572736
},
{
"epoch": 0.0,
"objective/train/docs_used": 911303,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.985459327697754,
"objective/train/theoretical_loss": 3.486621795280263,
"objective/train/tokens_used": 22097376,
"theoretical_loss": 3.486621795280263,
"tokens_seen": 1651638272
},
{
"epoch": 0.0,
"learning_rate": 1.0317460317460318e-05,
"loss": 3.0636,
"theoretical_loss": 3.4866098217968826,
"tokens_seen": 1651703808
},
{
"epoch": 0.0,
"learning_rate": 1.1111111111111112e-05,
"loss": 2.9173,
"theoretical_loss": 3.486585876654333,
"tokens_seen": 1651834880
},
{
"epoch": 0.0,
"learning_rate": 1.1904761904761905e-05,
"loss": 2.9235,
"theoretical_loss": 3.4865619339437,
"tokens_seen": 1651965952
},
{
"epoch": 0.0,
"learning_rate": 1.2698412698412699e-05,
"loss": 2.9036,
"theoretical_loss": 3.4865379936645438,
"tokens_seen": 1652097024
},
{
"epoch": 0.0,
"learning_rate": 1.3492063492063492e-05,
"loss": 2.8327,
"theoretical_loss": 3.486514055816424,
"tokens_seen": 1652228096
},
{
"epoch": 0.0,
"learning_rate": 1.4285714285714285e-05,
"loss": 2.8069,
"theoretical_loss": 3.4864901203989014,
"tokens_seen": 1652359168
},
{
"epoch": 0.0,
"learning_rate": 1.5079365079365079e-05,
"loss": 2.6994,
"theoretical_loss": 3.4864661874115357,
"tokens_seen": 1652490240
},
{
"epoch": 0.0,
"learning_rate": 1.5873015873015872e-05,
"loss": 2.8277,
"theoretical_loss": 3.486442256853888,
"tokens_seen": 1652621312
},
{
"epoch": 0.0,
"learning_rate": 1.6666666666666667e-05,
"loss": 2.8779,
"theoretical_loss": 3.4864183287255193,
"tokens_seen": 1652752384
},
{
"epoch": 0.0,
"learning_rate": 1.746031746031746e-05,
"loss": 2.929,
"theoretical_loss": 3.4863944030259897,
"tokens_seen": 1652883456
},
{
"epoch": 0.0,
"learning_rate": 1.8253968253968254e-05,
"loss": 2.7098,
"theoretical_loss": 3.4863704797548607,
"tokens_seen": 1653014528
},
{
"epoch": 0.0,
"learning_rate": 1.9047619047619046e-05,
"loss": 2.7617,
"theoretical_loss": 3.4863465589116927,
"tokens_seen": 1653145600
},
{
"epoch": 0.0,
"objective/train/docs_used": 912643,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 3.1410162448883057,
"objective/train/theoretical_loss": 3.486322640496047,
"objective/train/tokens_used": 23735776,
"theoretical_loss": 3.486322640496047,
"tokens_seen": 1653276672
},
{
"epoch": 0.0,
"learning_rate": 1.984126984126984e-05,
"loss": 2.8668,
"theoretical_loss": 3.486322640496047,
"tokens_seen": 1653276672
},
{
"epoch": 0.0,
"learning_rate": 2.0634920634920636e-05,
"loss": 2.7427,
"theoretical_loss": 3.486298724507485,
"tokens_seen": 1653407744
},
{
"epoch": 0.0,
"learning_rate": 2.1428571428571428e-05,
"loss": 2.9108,
"theoretical_loss": 3.4862748109455675,
"tokens_seen": 1653538816
},
{
"epoch": 0.0,
"learning_rate": 2.2222222222222223e-05,
"loss": 2.8479,
"theoretical_loss": 3.486250899809857,
"tokens_seen": 1653669888
},
{
"epoch": 0.0,
"learning_rate": 2.3015873015873015e-05,
"loss": 2.9489,
"theoretical_loss": 3.4862269910999135,
"tokens_seen": 1653800960
},
{
"epoch": 0.0,
"learning_rate": 2.380952380952381e-05,
"loss": 2.7904,
"theoretical_loss": 3.4862030848153003,
"tokens_seen": 1653932032
},
{
"epoch": 0.0,
"learning_rate": 2.4603174603174602e-05,
"loss": 2.6043,
"theoretical_loss": 3.4861791809555784,
"tokens_seen": 1654063104
},
{
"epoch": 0.0,
"learning_rate": 2.5396825396825397e-05,
"loss": 2.9128,
"theoretical_loss": 3.48615527952031,
"tokens_seen": 1654194176
},
{
"epoch": 0.0,
"learning_rate": 2.6190476190476192e-05,
"loss": 2.7843,
"theoretical_loss": 3.486131380509057,
"tokens_seen": 1654325248
},
{
"epoch": 0.0,
"learning_rate": 2.6984126984126984e-05,
"loss": 2.8452,
"theoretical_loss": 3.4861074839213813,
"tokens_seen": 1654456320
},
{
"epoch": 0.0,
"learning_rate": 2.777777777777778e-05,
"loss": 2.6435,
"theoretical_loss": 3.4860835897568454,
"tokens_seen": 1654587392
},
{
"epoch": 0.0,
"learning_rate": 2.857142857142857e-05,
"loss": 2.7978,
"theoretical_loss": 3.4860596980150116,
"tokens_seen": 1654718464
},
{
"epoch": 0.0,
"learning_rate": 2.9365079365079366e-05,
"loss": 2.9673,
"theoretical_loss": 3.4860358086954424,
"tokens_seen": 1654849536
},
{
"epoch": 0.0,
"objective/train/docs_used": 913892,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.85264253616333,
"objective/train/theoretical_loss": 3.48602386494387,
"objective/train/tokens_used": 25374176,
"theoretical_loss": 3.48602386494387,
"tokens_seen": 1654915072
},
{
"epoch": 0.0,
"learning_rate": 3.0158730158730158e-05,
"loss": 2.9861,
"theoretical_loss": 3.4860119217977004,
"tokens_seen": 1654980608
},
{
"epoch": 0.0,
"learning_rate": 3.095238095238095e-05,
"loss": 2.9108,
"theoretical_loss": 3.485988037321348,
"tokens_seen": 1655111680
},
{
"epoch": 0.0,
"learning_rate": 3.1746031746031745e-05,
"loss": 2.7628,
"theoretical_loss": 3.4859641552659486,
"tokens_seen": 1655242752
},
{
"epoch": 0.0,
"learning_rate": 3.253968253968254e-05,
"loss": 2.6715,
"theoretical_loss": 3.485940275631065,
"tokens_seen": 1655373824
},
{
"epoch": 0.0,
"learning_rate": 3.3333333333333335e-05,
"loss": 2.9045,
"theoretical_loss": 3.4859163984162596,
"tokens_seen": 1655504896
},
{
"epoch": 0.0,
"learning_rate": 3.412698412698413e-05,
"loss": 2.7567,
"theoretical_loss": 3.4858925236210965,
"tokens_seen": 1655635968
},
{
"epoch": 0.0,
"learning_rate": 3.492063492063492e-05,
"loss": 2.787,
"theoretical_loss": 3.4858686512451387,
"tokens_seen": 1655767040
},
{
"epoch": 0.0,
"learning_rate": 3.571428571428572e-05,
"loss": 2.6928,
"theoretical_loss": 3.4858447812879487,
"tokens_seen": 1655898112
},
{
"epoch": 0.0,
"learning_rate": 3.650793650793651e-05,
"loss": 2.7383,
"theoretical_loss": 3.485820913749091,
"tokens_seen": 1656029184
},
{
"epoch": 0.0,
"learning_rate": 3.730158730158731e-05,
"loss": 2.6014,
"theoretical_loss": 3.4857970486281293,
"tokens_seen": 1656160256
},
{
"epoch": 0.0,
"learning_rate": 3.809523809523809e-05,
"loss": 2.8116,
"theoretical_loss": 3.485773185924627,
"tokens_seen": 1656291328
},
{
"epoch": 0.0,
"learning_rate": 3.888888888888889e-05,
"loss": 2.6689,
"theoretical_loss": 3.4857493256381473,
"tokens_seen": 1656422400
},
{
"epoch": 0.0,
"objective/train/docs_used": 914561,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.774996280670166,
"objective/train/theoretical_loss": 3.4857254677682548,
"objective/train/tokens_used": 27012576,
"theoretical_loss": 3.4857254677682548,
"tokens_seen": 1656553472
},
{
"epoch": 0.0,
"learning_rate": 3.968253968253968e-05,
"loss": 2.6872,
"theoretical_loss": 3.4857254677682548,
"tokens_seen": 1656553472
},
{
"epoch": 0.0,
"learning_rate": 4.047619047619048e-05,
"loss": 2.5722,
"theoretical_loss": 3.485701612314514,
"tokens_seen": 1656684544
},
{
"epoch": 0.0,
"learning_rate": 4.126984126984127e-05,
"loss": 2.6782,
"theoretical_loss": 3.4856777592764883,
"tokens_seen": 1656815616
},
{
"epoch": 0.0,
"learning_rate": 4.2063492063492065e-05,
"loss": 2.6618,
"theoretical_loss": 3.4856539086537426,
"tokens_seen": 1656946688
},
{
"epoch": 0.0,
"learning_rate": 4.2857142857142856e-05,
"loss": 2.6818,
"theoretical_loss": 3.485630060445841,
"tokens_seen": 1657077760
},
{
"epoch": 0.0,
"learning_rate": 4.3650793650793655e-05,
"loss": 2.4444,
"theoretical_loss": 3.485606214652347,
"tokens_seen": 1657208832
},
{
"epoch": 0.0,
"learning_rate": 4.4444444444444447e-05,
"loss": 2.5045,
"theoretical_loss": 3.4855823712728276,
"tokens_seen": 1657339904
},
{
"epoch": 0.0,
"learning_rate": 4.523809523809524e-05,
"loss": 2.5876,
"theoretical_loss": 3.4855585303068453,
"tokens_seen": 1657470976
},
{
"epoch": 0.0,
"learning_rate": 4.603174603174603e-05,
"loss": 2.5061,
"theoretical_loss": 3.4855346917539665,
"tokens_seen": 1657602048
},
{
"epoch": 0.0,
"learning_rate": 4.682539682539683e-05,
"loss": 2.6817,
"theoretical_loss": 3.4855108556137546,
"tokens_seen": 1657733120
},
{
"epoch": 0.0,
"learning_rate": 4.761904761904762e-05,
"loss": 2.6159,
"theoretical_loss": 3.4854870218857763,
"tokens_seen": 1657864192
},
{
"epoch": 0.0,
"learning_rate": 4.841269841269841e-05,
"loss": 2.5539,
"theoretical_loss": 3.485463190569596,
"tokens_seen": 1657995264
},
{
"epoch": 0.0,
"learning_rate": 4.9206349206349204e-05,
"loss": 2.6999,
"theoretical_loss": 3.485439361664779,
"tokens_seen": 1658126336
},
{
"epoch": 0.0,
"objective/train/docs_used": 915678,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6932287216186523,
"objective/train/theoretical_loss": 3.4854274481164964,
"objective/train/tokens_used": 28650976,
"theoretical_loss": 3.4854274481164964,
"tokens_seen": 1658191872
},
{
"epoch": 0.01,
"learning_rate": 5e-05,
"loss": 2.5566,
"theoretical_loss": 3.4854155351708913,
"tokens_seen": 1658257408
},
{
"epoch": 0.01,
"learning_rate": 5.0793650793650794e-05,
"loss": 2.7198,
"theoretical_loss": 3.4853917110874972,
"tokens_seen": 1658388480
},
{
"epoch": 0.01,
"learning_rate": 5.158730158730159e-05,
"loss": 2.4993,
"theoretical_loss": 3.4853678894141638,
"tokens_seen": 1658519552
},
{
"epoch": 0.01,
"learning_rate": 5.2380952380952384e-05,
"loss": 2.5541,
"theoretical_loss": 3.4853440701504557,
"tokens_seen": 1658650624
},
{
"epoch": 0.01,
"learning_rate": 5.3174603174603176e-05,
"loss": 2.5189,
"theoretical_loss": 3.4853202532959395,
"tokens_seen": 1658781696
},
{
"epoch": 0.01,
"learning_rate": 5.396825396825397e-05,
"loss": 2.322,
"theoretical_loss": 3.4852964388501806,
"tokens_seen": 1658912768
},
{
"epoch": 0.01,
"learning_rate": 5.4761904761904766e-05,
"loss": 2.6172,
"theoretical_loss": 3.485272626812746,
"tokens_seen": 1659043840
},
{
"epoch": 0.01,
"learning_rate": 5.555555555555556e-05,
"loss": 2.5944,
"theoretical_loss": 3.4852488171832006,
"tokens_seen": 1659174912
},
{
"epoch": 0.01,
"learning_rate": 5.634920634920635e-05,
"loss": 2.5423,
"theoretical_loss": 3.4852250099611117,
"tokens_seen": 1659305984
},
{
"epoch": 0.01,
"learning_rate": 5.714285714285714e-05,
"loss": 2.5272,
"theoretical_loss": 3.485201205146046,
"tokens_seen": 1659437056
},
{
"epoch": 0.01,
"learning_rate": 5.793650793650795e-05,
"loss": 2.6078,
"theoretical_loss": 3.4851774027375693,
"tokens_seen": 1659568128
},
{
"epoch": 0.01,
"learning_rate": 5.873015873015873e-05,
"loss": 2.5119,
"theoretical_loss": 3.4851536027352483,
"tokens_seen": 1659699200
},
{
"epoch": 0.01,
"objective/train/docs_used": 916417,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.456707239151001,
"objective/train/theoretical_loss": 3.48512980513865,
"objective/train/tokens_used": 30289376,
"theoretical_loss": 3.48512980513865,
"tokens_seen": 1659830272
},
{
"epoch": 0.01,
"learning_rate": 5.9523809523809524e-05,
"loss": 2.6637,
"theoretical_loss": 3.48512980513865,
"tokens_seen": 1659830272
},
{
"epoch": 0.01,
"learning_rate": 6.0317460317460316e-05,
"loss": 2.3894,
"theoretical_loss": 3.4851060099473417,
"tokens_seen": 1659961344
},
{
"epoch": 0.01,
"learning_rate": 6.111111111111112e-05,
"loss": 2.5928,
"theoretical_loss": 3.4850822171608895,
"tokens_seen": 1660092416
},
{
"epoch": 0.01,
"learning_rate": 6.19047619047619e-05,
"loss": 2.4428,
"theoretical_loss": 3.4850584267788607,
"tokens_seen": 1660223488
},
{
"epoch": 0.01,
"learning_rate": 6.26984126984127e-05,
"loss": 2.7376,
"theoretical_loss": 3.4850346388008235,
"tokens_seen": 1660354560
},
{
"epoch": 0.01,
"learning_rate": 6.349206349206349e-05,
"loss": 2.4582,
"theoretical_loss": 3.4850108532263437,
"tokens_seen": 1660485632
},
{
"epoch": 0.01,
"learning_rate": 6.428571428571429e-05,
"loss": 2.409,
"theoretical_loss": 3.48498707005499,
"tokens_seen": 1660616704
},
{
"epoch": 0.01,
"learning_rate": 6.507936507936509e-05,
"loss": 2.5177,
"theoretical_loss": 3.4849632892863296,
"tokens_seen": 1660747776
},
{
"epoch": 0.01,
"learning_rate": 6.587301587301587e-05,
"loss": 2.6281,
"theoretical_loss": 3.48493951091993,
"tokens_seen": 1660878848
},
{
"epoch": 0.01,
"learning_rate": 6.666666666666667e-05,
"loss": 2.5806,
"theoretical_loss": 3.4849157349553592,
"tokens_seen": 1661009920
},
{
"epoch": 0.01,
"learning_rate": 6.746031746031747e-05,
"loss": 2.5608,
"theoretical_loss": 3.4848919613921847,
"tokens_seen": 1661140992
},
{
"epoch": 0.01,
"learning_rate": 6.825396825396825e-05,
"loss": 2.5385,
"theoretical_loss": 3.484868190229975,
"tokens_seen": 1661272064
},
{
"epoch": 0.01,
"learning_rate": 6.904761904761905e-05,
"loss": 2.5632,
"theoretical_loss": 3.484844421468298,
"tokens_seen": 1661403136
},
{
"epoch": 0.01,
"objective/train/docs_used": 917689,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.0480103492736816,
"objective/train/theoretical_loss": 3.484832537987524,
"objective/train/tokens_used": 31927776,
"theoretical_loss": 3.484832537987524,
"tokens_seen": 1661468672
},
{
"epoch": 0.01,
"learning_rate": 6.984126984126984e-05,
"loss": 2.3842,
"theoretical_loss": 3.4848206551067213,
"tokens_seen": 1661534208
},
{
"epoch": 0.01,
"learning_rate": 7.063492063492065e-05,
"loss": 2.4595,
"theoretical_loss": 3.484796891144814,
"tokens_seen": 1661665280
},
{
"epoch": 0.01,
"learning_rate": 7.142857142857143e-05,
"loss": 2.5882,
"theoretical_loss": 3.4847731295821447,
"tokens_seen": 1661796352
},
{
"epoch": 0.01,
"learning_rate": 7.222222222222222e-05,
"loss": 2.3931,
"theoretical_loss": 3.4847493704182817,
"tokens_seen": 1661927424
},
{
"epoch": 0.01,
"learning_rate": 7.301587301587302e-05,
"loss": 2.622,
"theoretical_loss": 3.4847256136527935,
"tokens_seen": 1662058496
},
{
"epoch": 0.01,
"learning_rate": 7.380952380952382e-05,
"loss": 2.5425,
"theoretical_loss": 3.4847018592852494,
"tokens_seen": 1662189568
},
{
"epoch": 0.01,
"learning_rate": 7.460317460317461e-05,
"loss": 2.601,
"theoretical_loss": 3.4846781073152173,
"tokens_seen": 1662320640
},
{
"epoch": 0.01,
"learning_rate": 7.53968253968254e-05,
"loss": 2.5712,
"theoretical_loss": 3.484654357742267,
"tokens_seen": 1662451712
},
{
"epoch": 0.01,
"learning_rate": 7.619047619047618e-05,
"loss": 2.2686,
"theoretical_loss": 3.4846306105659677,
"tokens_seen": 1662582784
},
{
"epoch": 0.01,
"learning_rate": 7.6984126984127e-05,
"loss": 2.3971,
"theoretical_loss": 3.4846068657858877,
"tokens_seen": 1662713856
},
{
"epoch": 0.01,
"learning_rate": 7.777777777777778e-05,
"loss": 2.5189,
"theoretical_loss": 3.4845831234015976,
"tokens_seen": 1662844928
},
{
"epoch": 0.01,
"learning_rate": 7.857142857142858e-05,
"loss": 2.3924,
"theoretical_loss": 3.484559383412666,
"tokens_seen": 1662976000
},
{
"epoch": 0.01,
"objective/train/docs_used": 918239,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.3232572078704834,
"objective/train/theoretical_loss": 3.4845356458186627,
"objective/train/tokens_used": 33566176,
"theoretical_loss": 3.4845356458186627,
"tokens_seen": 1663107072
},
{
"epoch": 0.01,
"learning_rate": 7.936507936507937e-05,
"loss": 2.5743,
"theoretical_loss": 3.4845356458186627,
"tokens_seen": 1663107072
},
{
"epoch": 0.01,
"learning_rate": 8.015873015873016e-05,
"loss": 2.4116,
"theoretical_loss": 3.4845119106191578,
"tokens_seen": 1663238144
},
{
"epoch": 0.01,
"learning_rate": 8.095238095238096e-05,
"loss": 2.5243,
"theoretical_loss": 3.48448817781372,
"tokens_seen": 1663369216
},
{
"epoch": 0.01,
"learning_rate": 8.174603174603175e-05,
"loss": 2.4798,
"theoretical_loss": 3.484464447401921,
"tokens_seen": 1663500288
},
{
"epoch": 0.01,
"learning_rate": 8.253968253968255e-05,
"loss": 2.3456,
"theoretical_loss": 3.4844407193833282,
"tokens_seen": 1663631360
},
{
"epoch": 0.01,
"learning_rate": 8.333333333333334e-05,
"loss": 2.4678,
"theoretical_loss": 3.484416993757514,
"tokens_seen": 1663762432
},
{
"epoch": 0.01,
"learning_rate": 8.412698412698413e-05,
"loss": 2.4106,
"theoretical_loss": 3.4843932705240483,
"tokens_seen": 1663893504
},
{
"epoch": 0.01,
"learning_rate": 8.492063492063493e-05,
"loss": 2.3669,
"theoretical_loss": 3.4843695496825005,
"tokens_seen": 1664024576
},
{
"epoch": 0.01,
"learning_rate": 8.571428571428571e-05,
"loss": 2.5092,
"theoretical_loss": 3.4843458312324413,
"tokens_seen": 1664155648
},
{
"epoch": 0.01,
"learning_rate": 8.650793650793651e-05,
"loss": 2.473,
"theoretical_loss": 3.484322115173442,
"tokens_seen": 1664286720
},
{
"epoch": 0.01,
"learning_rate": 8.730158730158731e-05,
"loss": 2.5359,
"theoretical_loss": 3.4842984015050726,
"tokens_seen": 1664417792
},
{
"epoch": 0.01,
"learning_rate": 8.80952380952381e-05,
"loss": 2.6096,
"theoretical_loss": 3.484274690226904,
"tokens_seen": 1664548864
},
{
"epoch": 0.01,
"learning_rate": 8.888888888888889e-05,
"loss": 2.4428,
"theoretical_loss": 3.484250981338507,
"tokens_seen": 1664679936
},
{
"epoch": 0.01,
"objective/train/docs_used": 919306,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.603062629699707,
"objective/train/theoretical_loss": 3.484239127790339,
"objective/train/tokens_used": 35204576,
"theoretical_loss": 3.484239127790339,
"tokens_seen": 1664745472
},
{
"epoch": 0.01,
"learning_rate": 8.968253968253969e-05,
"loss": 2.5448,
"theoretical_loss": 3.484227274839453,
"tokens_seen": 1664811008
},
{
"epoch": 0.01,
"learning_rate": 9.047619047619048e-05,
"loss": 2.5698,
"theoretical_loss": 3.484203570729313,
"tokens_seen": 1664942080
},
{
"epoch": 0.01,
"learning_rate": 9.126984126984128e-05,
"loss": 2.4996,
"theoretical_loss": 3.484179869007658,
"tokens_seen": 1665073152
},
{
"epoch": 0.01,
"learning_rate": 9.206349206349206e-05,
"loss": 2.5152,
"theoretical_loss": 3.4841561696740597,
"tokens_seen": 1665204224
},
{
"epoch": 0.01,
"learning_rate": 9.285714285714286e-05,
"loss": 2.4996,
"theoretical_loss": 3.4841324727280893,
"tokens_seen": 1665335296
},
{
"epoch": 0.01,
"learning_rate": 9.365079365079366e-05,
"loss": 2.47,
"theoretical_loss": 3.484108778169318,
"tokens_seen": 1665466368
},
{
"epoch": 0.01,
"learning_rate": 9.444444444444444e-05,
"loss": 2.3922,
"theoretical_loss": 3.484085085997318,
"tokens_seen": 1665597440
},
{
"epoch": 0.01,
"learning_rate": 9.523809523809524e-05,
"loss": 2.4767,
"theoretical_loss": 3.484061396211661,
"tokens_seen": 1665728512
},
{
"epoch": 0.01,
"learning_rate": 9.603174603174604e-05,
"loss": 2.5377,
"theoretical_loss": 3.4840377088119188,
"tokens_seen": 1665859584
},
{
"epoch": 0.01,
"learning_rate": 9.682539682539682e-05,
"loss": 2.61,
"theoretical_loss": 3.4840140237976636,
"tokens_seen": 1665990656
},
{
"epoch": 0.01,
"learning_rate": 9.761904761904762e-05,
"loss": 2.4506,
"theoretical_loss": 3.483990341168467,
"tokens_seen": 1666121728
},
{
"epoch": 0.01,
"learning_rate": 9.841269841269841e-05,
"loss": 2.4311,
"theoretical_loss": 3.483966660923902,
"tokens_seen": 1666252800
},
{
"epoch": 0.01,
"objective/train/docs_used": 919806,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.3898022174835205,
"objective/train/theoretical_loss": 3.4839429830635407,
"objective/train/tokens_used": 36842976,
"theoretical_loss": 3.4839429830635407,
"tokens_seen": 1666383872
},
{
"epoch": 0.01,
"learning_rate": 9.920634920634922e-05,
"loss": 2.543,
"theoretical_loss": 3.4839429830635407,
"tokens_seen": 1666383872
},
{
"epoch": 0.01,
"learning_rate": 0.0001,
"loss": 2.4395,
"theoretical_loss": 3.4839193075869543,
"tokens_seen": 1666514944
},
{
"epoch": 0.01,
"learning_rate": 9.999197560584176e-05,
"loss": 2.4916,
"theoretical_loss": 3.4838956344937175,
"tokens_seen": 1666646016
},
{
"epoch": 0.01,
"learning_rate": 9.998395121168352e-05,
"loss": 2.3888,
"theoretical_loss": 3.4838719637834012,
"tokens_seen": 1666777088
},
{
"epoch": 0.01,
"learning_rate": 9.997592681752528e-05,
"loss": 2.5063,
"theoretical_loss": 3.483848295455579,
"tokens_seen": 1666908160
},
{
"epoch": 0.01,
"learning_rate": 9.996790242336704e-05,
"loss": 2.4951,
"theoretical_loss": 3.483824629509824,
"tokens_seen": 1667039232
},
{
"epoch": 0.01,
"learning_rate": 9.99598780292088e-05,
"loss": 2.4226,
"theoretical_loss": 3.4838009659457088,
"tokens_seen": 1667170304
},
{
"epoch": 0.01,
"learning_rate": 9.995185363505056e-05,
"loss": 2.4431,
"theoretical_loss": 3.4837773047628064,
"tokens_seen": 1667301376
},
{
"epoch": 0.01,
"learning_rate": 9.994382924089232e-05,
"loss": 2.4434,
"theoretical_loss": 3.4837536459606904,
"tokens_seen": 1667432448
},
{
"epoch": 0.01,
"learning_rate": 9.993580484673407e-05,
"loss": 2.5021,
"theoretical_loss": 3.4837299895389333,
"tokens_seen": 1667563520
},
{
"epoch": 0.01,
"learning_rate": 9.992778045257584e-05,
"loss": 2.5622,
"theoretical_loss": 3.48370633549711,
"tokens_seen": 1667694592
},
{
"epoch": 0.01,
"learning_rate": 9.991975605841759e-05,
"loss": 2.524,
"theoretical_loss": 3.4836826838347923,
"tokens_seen": 1667825664
},
{
"epoch": 0.01,
"learning_rate": 9.991173166425936e-05,
"loss": 2.5055,
"theoretical_loss": 3.483659034551555,
"tokens_seen": 1667956736
},
{
"epoch": 0.01,
"objective/train/docs_used": 921005,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6571173667907715,
"objective/train/theoretical_loss": 3.4836472108019585,
"objective/train/tokens_used": 38481376,
"theoretical_loss": 3.4836472108019585,
"tokens_seen": 1668022272
},
{
"epoch": 0.01,
"learning_rate": 9.990370727010111e-05,
"loss": 2.5956,
"theoretical_loss": 3.483635387646972,
"tokens_seen": 1668087808
},
{
"epoch": 0.01,
"learning_rate": 9.989568287594288e-05,
"loss": 2.5279,
"theoretical_loss": 3.4836117431206164,
"tokens_seen": 1668218880
},
{
"epoch": 0.01,
"learning_rate": 9.988765848178463e-05,
"loss": 2.45,
"theoretical_loss": 3.483588100972063,
"tokens_seen": 1668349952
},
{
"epoch": 0.01,
"learning_rate": 9.987963408762638e-05,
"loss": 2.5056,
"theoretical_loss": 3.483564461200885,
"tokens_seen": 1668481024
},
{
"epoch": 0.01,
"learning_rate": 9.987160969346815e-05,
"loss": 2.56,
"theoretical_loss": 3.4835408238066567,
"tokens_seen": 1668612096
},
{
"epoch": 0.01,
"learning_rate": 9.98635852993099e-05,
"loss": 2.4673,
"theoretical_loss": 3.4835171887889533,
"tokens_seen": 1668743168
},
{
"epoch": 0.01,
"learning_rate": 9.985556090515167e-05,
"loss": 2.3887,
"theoretical_loss": 3.483493556147349,
"tokens_seen": 1668874240
},
{
"epoch": 0.01,
"learning_rate": 9.984753651099342e-05,
"loss": 2.478,
"theoretical_loss": 3.4834699258814172,
"tokens_seen": 1669005312
},
{
"epoch": 0.01,
"learning_rate": 9.983951211683519e-05,
"loss": 2.6136,
"theoretical_loss": 3.483446297990734,
"tokens_seen": 1669136384
},
{
"epoch": 0.01,
"learning_rate": 9.983148772267694e-05,
"loss": 2.5876,
"theoretical_loss": 3.4834226724748736,
"tokens_seen": 1669267456
},
{
"epoch": 0.01,
"learning_rate": 9.982346332851871e-05,
"loss": 2.5149,
"theoretical_loss": 3.4833990493334106,
"tokens_seen": 1669398528
},
{
"epoch": 0.01,
"learning_rate": 9.981543893436046e-05,
"loss": 2.5755,
"theoretical_loss": 3.48337542856592,
"tokens_seen": 1669529600
},
{
"epoch": 0.01,
"objective/train/docs_used": 921680,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.329754590988159,
"objective/train/theoretical_loss": 3.4833518101719774,
"objective/train/tokens_used": 40119776,
"theoretical_loss": 3.4833518101719774,
"tokens_seen": 1669660672
},
{
"epoch": 0.01,
"learning_rate": 9.980741454020222e-05,
"loss": 2.4671,
"theoretical_loss": 3.4833518101719774,
"tokens_seen": 1669660672
},
{
"epoch": 0.01,
"learning_rate": 9.979939014604398e-05,
"loss": 2.3942,
"theoretical_loss": 3.4833281941511576,
"tokens_seen": 1669791744
},
{
"epoch": 0.01,
"learning_rate": 9.979136575188574e-05,
"loss": 2.4857,
"theoretical_loss": 3.4833045805030363,
"tokens_seen": 1669922816
},
{
"epoch": 0.01,
"learning_rate": 9.97833413577275e-05,
"loss": 2.4502,
"theoretical_loss": 3.483280969227188,
"tokens_seen": 1670053888
},
{
"epoch": 0.01,
"learning_rate": 9.977531696356926e-05,
"loss": 2.5831,
"theoretical_loss": 3.4832573603231887,
"tokens_seen": 1670184960
},
{
"epoch": 0.01,
"learning_rate": 9.976729256941102e-05,
"loss": 2.561,
"theoretical_loss": 3.4832337537906146,
"tokens_seen": 1670316032
},
{
"epoch": 0.01,
"learning_rate": 9.975926817525277e-05,
"loss": 2.5869,
"theoretical_loss": 3.4832101496290404,
"tokens_seen": 1670447104
},
{
"epoch": 0.01,
"learning_rate": 9.975124378109453e-05,
"loss": 2.4206,
"theoretical_loss": 3.4831865478380433,
"tokens_seen": 1670578176
},
{
"epoch": 0.01,
"learning_rate": 9.97432193869363e-05,
"loss": 2.3991,
"theoretical_loss": 3.4831629484171978,
"tokens_seen": 1670709248
},
{
"epoch": 0.01,
"learning_rate": 9.973519499277805e-05,
"loss": 2.3821,
"theoretical_loss": 3.4831393513660807,
"tokens_seen": 1670840320
},
{
"epoch": 0.01,
"learning_rate": 9.972717059861981e-05,
"loss": 2.5258,
"theoretical_loss": 3.483115756684268,
"tokens_seen": 1670971392
},
{
"epoch": 0.01,
"learning_rate": 9.971914620446157e-05,
"loss": 2.4604,
"theoretical_loss": 3.4830921643713366,
"tokens_seen": 1671102464
},
{
"epoch": 0.01,
"learning_rate": 9.971112181030333e-05,
"loss": 2.5027,
"theoretical_loss": 3.483068574426862,
"tokens_seen": 1671233536
},
{
"epoch": 0.01,
"objective/train/docs_used": 922598,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.674171209335327,
"objective/train/theoretical_loss": 3.4830567803426637,
"objective/train/tokens_used": 41758176,
"theoretical_loss": 3.4830567803426637,
"tokens_seen": 1671299072
},
{
"epoch": 0.01,
"learning_rate": 9.970309741614509e-05,
"loss": 2.5194,
"theoretical_loss": 3.483044986850421,
"tokens_seen": 1671364608
},
{
"epoch": 0.01,
"learning_rate": 9.969507302198684e-05,
"loss": 2.4984,
"theoretical_loss": 3.48302140164159,
"tokens_seen": 1671495680
},
{
"epoch": 0.01,
"learning_rate": 9.968704862782861e-05,
"loss": 2.6113,
"theoretical_loss": 3.482997818799947,
"tokens_seen": 1671626752
},
{
"epoch": 0.01,
"learning_rate": 9.967902423367036e-05,
"loss": 2.6013,
"theoretical_loss": 3.4829742383250673,
"tokens_seen": 1671757824
},
{
"epoch": 0.01,
"learning_rate": 9.967099983951213e-05,
"loss": 2.4007,
"theoretical_loss": 3.4829506602165283,
"tokens_seen": 1671888896
},
{
"epoch": 0.01,
"learning_rate": 9.966297544535388e-05,
"loss": 2.4473,
"theoretical_loss": 3.482927084473907,
"tokens_seen": 1672019968
},
{
"epoch": 0.01,
"learning_rate": 9.965495105119565e-05,
"loss": 2.3941,
"theoretical_loss": 3.482903511096781,
"tokens_seen": 1672151040
},
{
"epoch": 0.01,
"learning_rate": 9.96469266570374e-05,
"loss": 2.5564,
"theoretical_loss": 3.482879940084727,
"tokens_seen": 1672282112
},
{
"epoch": 0.01,
"learning_rate": 9.963890226287915e-05,
"loss": 2.4979,
"theoretical_loss": 3.4828563714373226,
"tokens_seen": 1672413184
},
{
"epoch": 0.01,
"learning_rate": 9.963087786872092e-05,
"loss": 2.4506,
"theoretical_loss": 3.4828328051541453,
"tokens_seen": 1672544256
},
{
"epoch": 0.01,
"learning_rate": 9.962285347456267e-05,
"loss": 2.5131,
"theoretical_loss": 3.482809241234773,
"tokens_seen": 1672675328
},
{
"epoch": 0.01,
"learning_rate": 9.961482908040444e-05,
"loss": 2.4079,
"theoretical_loss": 3.482785679678783,
"tokens_seen": 1672806400
},
{
"epoch": 0.01,
"objective/train/docs_used": 923236,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.478933095932007,
"objective/train/theoretical_loss": 3.482762120485753,
"objective/train/tokens_used": 43396576,
"theoretical_loss": 3.482762120485753,
"tokens_seen": 1672937472
},
{
"epoch": 0.01,
"learning_rate": 9.960680468624619e-05,
"loss": 2.5005,
"theoretical_loss": 3.482762120485753,
"tokens_seen": 1672937472
},
{
"epoch": 0.01,
"learning_rate": 9.959878029208796e-05,
"loss": 2.669,
"theoretical_loss": 3.482738563655261,
"tokens_seen": 1673068544
},
{
"epoch": 0.01,
"learning_rate": 9.959075589792971e-05,
"loss": 2.5706,
"theoretical_loss": 3.4827150091868853,
"tokens_seen": 1673199616
},
{
"epoch": 0.01,
"learning_rate": 9.958273150377147e-05,
"loss": 2.5295,
"theoretical_loss": 3.482691457080204,
"tokens_seen": 1673330688
},
{
"epoch": 0.01,
"learning_rate": 9.957470710961323e-05,
"loss": 2.4358,
"theoretical_loss": 3.482667907334795,
"tokens_seen": 1673461760
},
{
"epoch": 0.01,
"learning_rate": 9.956668271545499e-05,
"loss": 2.4467,
"theoretical_loss": 3.482644359950237,
"tokens_seen": 1673592832
},
{
"epoch": 0.01,
"learning_rate": 9.955865832129675e-05,
"loss": 2.3988,
"theoretical_loss": 3.4826208149261078,
"tokens_seen": 1673723904
},
{
"epoch": 0.01,
"learning_rate": 9.95506339271385e-05,
"loss": 2.521,
"theoretical_loss": 3.482597272261987,
"tokens_seen": 1673854976
},
{
"epoch": 0.01,
"learning_rate": 9.954260953298027e-05,
"loss": 2.409,
"theoretical_loss": 3.4825737319574523,
"tokens_seen": 1673986048
},
{
"epoch": 0.01,
"learning_rate": 9.953458513882203e-05,
"loss": 2.5548,
"theoretical_loss": 3.4825501940120835,
"tokens_seen": 1674117120
},
{
"epoch": 0.01,
"learning_rate": 9.952656074466378e-05,
"loss": 2.4541,
"theoretical_loss": 3.4825266584254586,
"tokens_seen": 1674248192
},
{
"epoch": 0.01,
"learning_rate": 9.951853635050554e-05,
"loss": 2.5549,
"theoretical_loss": 3.482503125197157,
"tokens_seen": 1674379264
},
{
"epoch": 0.01,
"learning_rate": 9.95105119563473e-05,
"loss": 2.4687,
"theoretical_loss": 3.4824795943267577,
"tokens_seen": 1674510336
},
{
"epoch": 0.01,
"objective/train/docs_used": 924425,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.032245397567749,
"objective/train/theoretical_loss": 3.48246782977564,
"objective/train/tokens_used": 45034976,
"theoretical_loss": 3.48246782977564,
"tokens_seen": 1674575872
},
{
"epoch": 0.01,
"learning_rate": 9.950248756218906e-05,
"loss": 2.4781,
"theoretical_loss": 3.4824560658138397,
"tokens_seen": 1674641408
},
{
"epoch": 0.02,
"learning_rate": 9.949446316803082e-05,
"loss": 2.6007,
"theoretical_loss": 3.482432539657983,
"tokens_seen": 1674772480
},
{
"epoch": 0.02,
"learning_rate": 9.948643877387258e-05,
"loss": 2.4198,
"theoretical_loss": 3.4824090158587664,
"tokens_seen": 1674903552
},
{
"epoch": 0.02,
"learning_rate": 9.947841437971434e-05,
"loss": 2.49,
"theoretical_loss": 3.4823854944157695,
"tokens_seen": 1675034624
},
{
"epoch": 0.02,
"learning_rate": 9.94703899855561e-05,
"loss": 2.4733,
"theoretical_loss": 3.482361975328572,
"tokens_seen": 1675165696
},
{
"epoch": 0.02,
"learning_rate": 9.946236559139786e-05,
"loss": 2.6164,
"theoretical_loss": 3.482338458596754,
"tokens_seen": 1675296768
},
{
"epoch": 0.02,
"learning_rate": 9.945434119723961e-05,
"loss": 2.4729,
"theoretical_loss": 3.482314944219895,
"tokens_seen": 1675427840
},
{
"epoch": 0.02,
"learning_rate": 9.944631680308138e-05,
"loss": 2.5106,
"theoretical_loss": 3.482291432197575,
"tokens_seen": 1675558912
},
{
"epoch": 0.02,
"learning_rate": 9.943829240892313e-05,
"loss": 2.4255,
"theoretical_loss": 3.482267922529374,
"tokens_seen": 1675689984
},
{
"epoch": 0.02,
"learning_rate": 9.94302680147649e-05,
"loss": 2.4172,
"theoretical_loss": 3.482244415214873,
"tokens_seen": 1675821056
},
{
"epoch": 0.02,
"learning_rate": 9.942224362060665e-05,
"loss": 2.4485,
"theoretical_loss": 3.482220910253651,
"tokens_seen": 1675952128
},
{
"epoch": 0.02,
"learning_rate": 9.941421922644842e-05,
"loss": 2.5783,
"theoretical_loss": 3.482197407645289,
"tokens_seen": 1676083200
},
{
"epoch": 0.02,
"objective/train/docs_used": 924948,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.1431517601013184,
"objective/train/theoretical_loss": 3.4821739073893676,
"objective/train/tokens_used": 46673376,
"theoretical_loss": 3.4821739073893676,
"tokens_seen": 1676214272
},
{
"epoch": 0.02,
"learning_rate": 9.940619483229017e-05,
"loss": 2.4842,
"theoretical_loss": 3.4821739073893676,
"tokens_seen": 1676214272
},
{
"epoch": 0.02,
"learning_rate": 9.939817043813192e-05,
"loss": 2.5048,
"theoretical_loss": 3.482150409485467,
"tokens_seen": 1676345344
},
{
"epoch": 0.02,
"learning_rate": 9.939014604397369e-05,
"loss": 2.4242,
"theoretical_loss": 3.482126913933169,
"tokens_seen": 1676476416
},
{
"epoch": 0.02,
"learning_rate": 9.938212164981544e-05,
"loss": 2.4716,
"theoretical_loss": 3.482103420732053,
"tokens_seen": 1676607488
},
{
"epoch": 0.02,
"learning_rate": 9.937409725565721e-05,
"loss": 2.4006,
"theoretical_loss": 3.482079929881701,
"tokens_seen": 1676738560
},
{
"epoch": 0.02,
"learning_rate": 9.936607286149896e-05,
"loss": 2.5438,
"theoretical_loss": 3.482056441381694,
"tokens_seen": 1676869632
},
{
"epoch": 0.02,
"learning_rate": 9.935804846734073e-05,
"loss": 2.4948,
"theoretical_loss": 3.4820329552316123,
"tokens_seen": 1677000704
},
{
"epoch": 0.02,
"learning_rate": 9.935002407318248e-05,
"loss": 2.4135,
"theoretical_loss": 3.482009471431038,
"tokens_seen": 1677131776
},
{
"epoch": 0.02,
"learning_rate": 9.934199967902424e-05,
"loss": 2.4302,
"theoretical_loss": 3.4819859899795516,
"tokens_seen": 1677262848
},
{
"epoch": 0.02,
"learning_rate": 9.9333975284866e-05,
"loss": 2.4168,
"theoretical_loss": 3.481962510876736,
"tokens_seen": 1677393920
},
{
"epoch": 0.02,
"learning_rate": 9.932595089070776e-05,
"loss": 2.648,
"theoretical_loss": 3.481939034122171,
"tokens_seen": 1677524992
},
{
"epoch": 0.02,
"learning_rate": 9.931792649654952e-05,
"loss": 2.4332,
"theoretical_loss": 3.4819155597154396,
"tokens_seen": 1677656064
},
{
"epoch": 0.02,
"learning_rate": 9.930990210239128e-05,
"loss": 2.4625,
"theoretical_loss": 3.4818920876561235,
"tokens_seen": 1677787136
},
{
"epoch": 0.02,
"objective/train/docs_used": 926165,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.494450569152832,
"objective/train/theoretical_loss": 3.4818803525066153,
"objective/train/tokens_used": 48311776,
"theoretical_loss": 3.4818803525066153,
"tokens_seen": 1677852672
},
{
"epoch": 0.02,
"learning_rate": 9.930187770823304e-05,
"loss": 2.4562,
"theoretical_loss": 3.481868617943804,
"tokens_seen": 1677918208
},
{
"epoch": 0.02,
"learning_rate": 9.92938533140748e-05,
"loss": 2.4934,
"theoretical_loss": 3.481845150578063,
"tokens_seen": 1678049280
},
{
"epoch": 0.02,
"learning_rate": 9.928582891991655e-05,
"loss": 2.4625,
"theoretical_loss": 3.481821685558484,
"tokens_seen": 1678180352
},
{
"epoch": 0.02,
"learning_rate": 9.927780452575832e-05,
"loss": 2.3664,
"theoretical_loss": 3.4817982228846476,
"tokens_seen": 1678311424
},
{
"epoch": 0.02,
"learning_rate": 9.926978013160007e-05,
"loss": 2.4554,
"theoretical_loss": 3.481774762556137,
"tokens_seen": 1678442496
},
{
"epoch": 0.02,
"learning_rate": 9.926175573744183e-05,
"loss": 2.6398,
"theoretical_loss": 3.4817513045725343,
"tokens_seen": 1678573568
},
{
"epoch": 0.02,
"learning_rate": 9.925373134328359e-05,
"loss": 2.3632,
"theoretical_loss": 3.4817278489334225,
"tokens_seen": 1678704640
},
{
"epoch": 0.02,
"learning_rate": 9.924570694912535e-05,
"loss": 2.6002,
"theoretical_loss": 3.481704395638383,
"tokens_seen": 1678835712
},
{
"epoch": 0.02,
"learning_rate": 9.923768255496711e-05,
"loss": 2.5935,
"theoretical_loss": 3.4816809446870005,
"tokens_seen": 1678966784
},
{
"epoch": 0.02,
"learning_rate": 9.922965816080886e-05,
"loss": 2.6889,
"theoretical_loss": 3.4816574960788564,
"tokens_seen": 1679097856
},
{
"epoch": 0.02,
"learning_rate": 9.922163376665063e-05,
"loss": 2.2718,
"theoretical_loss": 3.4816340498135343,
"tokens_seen": 1679228928
},
{
"epoch": 0.02,
"learning_rate": 9.921360937249238e-05,
"loss": 2.4179,
"theoretical_loss": 3.4816106058906175,
"tokens_seen": 1679360000
},
{
"epoch": 0.02,
"objective/train/docs_used": 926923,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7133426666259766,
"objective/train/theoretical_loss": 3.481587164309688,
"objective/train/tokens_used": 49950176,
"theoretical_loss": 3.481587164309688,
"tokens_seen": 1679491072
},
{
"epoch": 0.02,
"learning_rate": 9.920558497833415e-05,
"loss": 2.4835,
"theoretical_loss": 3.481587164309688,
"tokens_seen": 1679491072
},
{
"epoch": 0.02,
"learning_rate": 9.91975605841759e-05,
"loss": 2.5025,
"theoretical_loss": 3.4815637250703304,
"tokens_seen": 1679622144
},
{
"epoch": 0.02,
"learning_rate": 9.918953619001767e-05,
"loss": 2.4519,
"theoretical_loss": 3.4815402881721274,
"tokens_seen": 1679753216
},
{
"epoch": 0.02,
"learning_rate": 9.918151179585942e-05,
"loss": 2.5021,
"theoretical_loss": 3.4815168536146626,
"tokens_seen": 1679884288
},
{
"epoch": 0.02,
"learning_rate": 9.917348740170119e-05,
"loss": 2.4785,
"theoretical_loss": 3.48149342139752,
"tokens_seen": 1680015360
},
{
"epoch": 0.02,
"learning_rate": 9.916546300754294e-05,
"loss": 2.5138,
"theoretical_loss": 3.481469991520283,
"tokens_seen": 1680146432
},
{
"epoch": 0.02,
"learning_rate": 9.91574386133847e-05,
"loss": 2.4019,
"theoretical_loss": 3.4814465639825354,
"tokens_seen": 1680277504
},
{
"epoch": 0.02,
"learning_rate": 9.914941421922646e-05,
"loss": 2.5178,
"theoretical_loss": 3.481423138783861,
"tokens_seen": 1680408576
},
{
"epoch": 0.02,
"learning_rate": 9.914138982506821e-05,
"loss": 2.4615,
"theoretical_loss": 3.4813997159238443,
"tokens_seen": 1680539648
},
{
"epoch": 0.02,
"learning_rate": 9.913336543090998e-05,
"loss": 2.533,
"theoretical_loss": 3.481376295402069,
"tokens_seen": 1680670720
},
{
"epoch": 0.02,
"learning_rate": 9.912534103675173e-05,
"loss": 2.5408,
"theoretical_loss": 3.4813528772181193,
"tokens_seen": 1680801792
},
{
"epoch": 0.02,
"learning_rate": 9.91173166425935e-05,
"loss": 2.4864,
"theoretical_loss": 3.4813294613715797,
"tokens_seen": 1680932864
},
{
"epoch": 0.02,
"learning_rate": 9.910929224843525e-05,
"loss": 2.567,
"theoretical_loss": 3.4813060478620352,
"tokens_seen": 1681063936
},
{
"epoch": 0.02,
"objective/train/docs_used": 927976,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.612856864929199,
"objective/train/theoretical_loss": 3.4812943419835065,
"objective/train/tokens_used": 51588576,
"theoretical_loss": 3.4812943419835065,
"tokens_seen": 1681129472
},
{
"epoch": 0.02,
"learning_rate": 9.9101267854277e-05,
"loss": 2.515,
"theoretical_loss": 3.48128263668907,
"tokens_seen": 1681195008
},
{
"epoch": 0.02,
"learning_rate": 9.909324346011877e-05,
"loss": 2.5095,
"theoretical_loss": 3.4812592278522687,
"tokens_seen": 1681326080
},
{
"epoch": 0.02,
"learning_rate": 9.908521906596053e-05,
"loss": 2.6043,
"theoretical_loss": 3.4812358213512162,
"tokens_seen": 1681457152
},
{
"epoch": 0.02,
"learning_rate": 9.907719467180229e-05,
"loss": 2.6276,
"theoretical_loss": 3.4812124171854966,
"tokens_seen": 1681588224
},
{
"epoch": 0.02,
"learning_rate": 9.906917027764405e-05,
"loss": 2.3657,
"theoretical_loss": 3.481189015354696,
"tokens_seen": 1681719296
},
{
"epoch": 0.02,
"learning_rate": 9.906114588348581e-05,
"loss": 2.4345,
"theoretical_loss": 3.481165615858399,
"tokens_seen": 1681850368
},
{
"epoch": 0.02,
"learning_rate": 9.905312148932757e-05,
"loss": 2.6024,
"theoretical_loss": 3.481142218696191,
"tokens_seen": 1681981440
},
{
"epoch": 0.02,
"learning_rate": 9.904509709516932e-05,
"loss": 2.5844,
"theoretical_loss": 3.481118823867657,
"tokens_seen": 1682112512
},
{
"epoch": 0.02,
"learning_rate": 9.903707270101109e-05,
"loss": 2.4985,
"theoretical_loss": 3.481095431372383,
"tokens_seen": 1682243584
},
{
"epoch": 0.02,
"learning_rate": 9.902904830685284e-05,
"loss": 2.4409,
"theoretical_loss": 3.481072041209954,
"tokens_seen": 1682374656
},
{
"epoch": 0.02,
"learning_rate": 9.90210239126946e-05,
"loss": 2.4786,
"theoretical_loss": 3.481048653379955,
"tokens_seen": 1682505728
},
{
"epoch": 0.02,
"learning_rate": 9.901299951853636e-05,
"loss": 2.4033,
"theoretical_loss": 3.4810252678819733,
"tokens_seen": 1682636800
},
{
"epoch": 0.02,
"objective/train/docs_used": 929273,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6835052967071533,
"objective/train/theoretical_loss": 3.481001884715594,
"objective/train/tokens_used": 53226976,
"theoretical_loss": 3.481001884715594,
"tokens_seen": 1682767872
},
{
"epoch": 0.02,
"learning_rate": 9.900497512437812e-05,
"loss": 2.5045,
"theoretical_loss": 3.481001884715594,
"tokens_seen": 1682767872
},
{
"epoch": 0.02,
"learning_rate": 9.899695073021988e-05,
"loss": 2.4493,
"theoretical_loss": 3.4809785038804026,
"tokens_seen": 1682898944
},
{
"epoch": 0.02,
"learning_rate": 9.898892633606163e-05,
"loss": 2.5062,
"theoretical_loss": 3.480955125375986,
"tokens_seen": 1683030016
},
{
"epoch": 0.02,
"learning_rate": 9.89809019419034e-05,
"loss": 2.5045,
"theoretical_loss": 3.48093174920193,
"tokens_seen": 1683161088
},
{
"epoch": 0.02,
"learning_rate": 9.897287754774515e-05,
"loss": 2.5248,
"theoretical_loss": 3.48090837535782,
"tokens_seen": 1683292160
},
{
"epoch": 0.02,
"learning_rate": 9.896485315358692e-05,
"loss": 2.5576,
"theoretical_loss": 3.4808850038432437,
"tokens_seen": 1683423232
},
{
"epoch": 0.02,
"learning_rate": 9.895682875942867e-05,
"loss": 2.398,
"theoretical_loss": 3.4808616346577868,
"tokens_seen": 1683554304
},
{
"epoch": 0.02,
"learning_rate": 9.894880436527044e-05,
"loss": 2.5147,
"theoretical_loss": 3.4808382678010363,
"tokens_seen": 1683685376
},
{
"epoch": 0.02,
"learning_rate": 9.894077997111219e-05,
"loss": 2.5106,
"theoretical_loss": 3.480814903272579,
"tokens_seen": 1683816448
},
{
"epoch": 0.02,
"learning_rate": 9.893275557695394e-05,
"loss": 2.4724,
"theoretical_loss": 3.4807915410720005,
"tokens_seen": 1683947520
},
{
"epoch": 0.02,
"learning_rate": 9.892473118279571e-05,
"loss": 2.5067,
"theoretical_loss": 3.4807681811988895,
"tokens_seen": 1684078592
},
{
"epoch": 0.02,
"learning_rate": 9.891670678863746e-05,
"loss": 2.5444,
"theoretical_loss": 3.4807448236528318,
"tokens_seen": 1684209664
},
{
"epoch": 0.02,
"learning_rate": 9.890868239447923e-05,
"loss": 2.5985,
"theoretical_loss": 3.4807214684334147,
"tokens_seen": 1684340736
},
{
"epoch": 0.02,
"objective/train/docs_used": 929850,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.568603992462158,
"objective/train/theoretical_loss": 3.480709791696068,
"objective/train/tokens_used": 54865376,
"theoretical_loss": 3.480709791696068,
"tokens_seen": 1684406272
},
{
"epoch": 0.02,
"learning_rate": 9.890065800032098e-05,
"loss": 2.4515,
"theoretical_loss": 3.4806981155402257,
"tokens_seen": 1684471808
},
{
"epoch": 0.02,
"learning_rate": 9.889263360616275e-05,
"loss": 2.4635,
"theoretical_loss": 3.4806747649728518,
"tokens_seen": 1684602880
},
{
"epoch": 0.02,
"learning_rate": 9.88846092120045e-05,
"loss": 2.5249,
"theoretical_loss": 3.480651416730881,
"tokens_seen": 1684733952
},
{
"epoch": 0.02,
"learning_rate": 9.887658481784626e-05,
"loss": 2.5362,
"theoretical_loss": 3.4806280708139,
"tokens_seen": 1684865024
},
{
"epoch": 0.02,
"learning_rate": 9.886856042368802e-05,
"loss": 2.4881,
"theoretical_loss": 3.480604727221497,
"tokens_seen": 1684996096
},
{
"epoch": 0.02,
"learning_rate": 9.886053602952978e-05,
"loss": 2.555,
"theoretical_loss": 3.4805813859532595,
"tokens_seen": 1685127168
},
{
"epoch": 0.02,
"learning_rate": 9.885251163537154e-05,
"loss": 2.6367,
"theoretical_loss": 3.480558047008776,
"tokens_seen": 1685258240
},
{
"epoch": 0.02,
"learning_rate": 9.88444872412133e-05,
"loss": 2.4256,
"theoretical_loss": 3.4805347103876327,
"tokens_seen": 1685389312
},
{
"epoch": 0.02,
"learning_rate": 9.883646284705506e-05,
"loss": 2.2893,
"theoretical_loss": 3.48051137608942,
"tokens_seen": 1685520384
},
{
"epoch": 0.02,
"learning_rate": 9.882843845289682e-05,
"loss": 2.5847,
"theoretical_loss": 3.4804880441137245,
"tokens_seen": 1685651456
},
{
"epoch": 0.02,
"learning_rate": 9.882041405873857e-05,
"loss": 2.6211,
"theoretical_loss": 3.4804647144601346,
"tokens_seen": 1685782528
},
{
"epoch": 0.02,
"learning_rate": 9.881238966458034e-05,
"loss": 2.4622,
"theoretical_loss": 3.4804413871282396,
"tokens_seen": 1685913600
},
{
"epoch": 0.02,
"objective/train/docs_used": 931063,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.578721523284912,
"objective/train/theoretical_loss": 3.480418062117627,
"objective/train/tokens_used": 56503776,
"theoretical_loss": 3.480418062117627,
"tokens_seen": 1686044672
},
{
"epoch": 0.02,
"learning_rate": 9.880436527042209e-05,
"loss": 2.5689,
"theoretical_loss": 3.480418062117627,
"tokens_seen": 1686044672
},
{
"epoch": 0.02,
"learning_rate": 9.879634087626386e-05,
"loss": 2.4969,
"theoretical_loss": 3.4803947394278856,
"tokens_seen": 1686175744
},
{
"epoch": 0.02,
"learning_rate": 9.878831648210561e-05,
"loss": 2.5164,
"theoretical_loss": 3.4803714190586037,
"tokens_seen": 1686306816
},
{
"epoch": 0.02,
"learning_rate": 9.878029208794736e-05,
"loss": 2.4945,
"theoretical_loss": 3.4803481010093718,
"tokens_seen": 1686437888
},
{
"epoch": 0.02,
"learning_rate": 9.877226769378913e-05,
"loss": 2.399,
"theoretical_loss": 3.4803247852797767,
"tokens_seen": 1686568960
},
{
"epoch": 0.02,
"learning_rate": 9.876424329963088e-05,
"loss": 2.4799,
"theoretical_loss": 3.4803014718694087,
"tokens_seen": 1686700032
},
{
"epoch": 0.02,
"learning_rate": 9.875621890547265e-05,
"loss": 2.35,
"theoretical_loss": 3.480278160777856,
"tokens_seen": 1686831104
},
{
"epoch": 0.02,
"learning_rate": 9.87481945113144e-05,
"loss": 2.4886,
"theoretical_loss": 3.480254852004709,
"tokens_seen": 1686962176
},
{
"epoch": 0.02,
"learning_rate": 9.874017011715617e-05,
"loss": 2.5754,
"theoretical_loss": 3.4802315455495565,
"tokens_seen": 1687093248
},
{
"epoch": 0.02,
"learning_rate": 9.873214572299792e-05,
"loss": 2.5741,
"theoretical_loss": 3.480208241411987,
"tokens_seen": 1687224320
},
{
"epoch": 0.02,
"learning_rate": 9.872412132883967e-05,
"loss": 2.5755,
"theoretical_loss": 3.480184939591591,
"tokens_seen": 1687355392
},
{
"epoch": 0.02,
"learning_rate": 9.871609693468144e-05,
"loss": 2.4047,
"theoretical_loss": 3.4801616400879585,
"tokens_seen": 1687486464
},
{
"epoch": 0.02,
"learning_rate": 9.87080725405232e-05,
"loss": 2.4871,
"theoretical_loss": 3.480138342900678,
"tokens_seen": 1687617536
},
{
"epoch": 0.02,
"objective/train/docs_used": 931367,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.5317652225494385,
"objective/train/theoretical_loss": 3.480126695175542,
"objective/train/tokens_used": 58142176,
"theoretical_loss": 3.480126695175542,
"tokens_seen": 1687683072
},
{
"epoch": 0.02,
"learning_rate": 9.870004814636496e-05,
"loss": 2.6203,
"theoretical_loss": 3.48011504802934,
"tokens_seen": 1687748608
},
{
"epoch": 0.02,
"learning_rate": 9.869202375220671e-05,
"loss": 2.4963,
"theoretical_loss": 3.4800917554735347,
"tokens_seen": 1687879680
},
{
"epoch": 0.02,
"learning_rate": 9.868399935804847e-05,
"loss": 2.5753,
"theoretical_loss": 3.480068465232852,
"tokens_seen": 1688010752
},
{
"epoch": 0.02,
"learning_rate": 9.867597496389023e-05,
"loss": 2.546,
"theoretical_loss": 3.4800451773068817,
"tokens_seen": 1688141824
},
{
"epoch": 0.02,
"learning_rate": 9.866795056973199e-05,
"loss": 2.3913,
"theoretical_loss": 3.4800218916952144,
"tokens_seen": 1688272896
},
{
"epoch": 0.02,
"learning_rate": 9.865992617557375e-05,
"loss": 2.5884,
"theoretical_loss": 3.47999860839744,
"tokens_seen": 1688403968
},
{
"epoch": 0.02,
"learning_rate": 9.86519017814155e-05,
"loss": 2.5369,
"theoretical_loss": 3.4799753274131495,
"tokens_seen": 1688535040
},
{
"epoch": 0.02,
"learning_rate": 9.864387738725727e-05,
"loss": 2.6298,
"theoretical_loss": 3.4799520487419335,
"tokens_seen": 1688666112
},
{
"epoch": 0.02,
"learning_rate": 9.863585299309903e-05,
"loss": 2.567,
"theoretical_loss": 3.4799287723833823,
"tokens_seen": 1688797184
},
{
"epoch": 0.02,
"learning_rate": 9.862782859894078e-05,
"loss": 2.4449,
"theoretical_loss": 3.4799054983370867,
"tokens_seen": 1688928256
},
{
"epoch": 0.02,
"learning_rate": 9.861980420478255e-05,
"loss": 2.5075,
"theoretical_loss": 3.4798822266026384,
"tokens_seen": 1689059328
},
{
"epoch": 0.02,
"learning_rate": 9.86117798106243e-05,
"loss": 2.5648,
"theoretical_loss": 3.4798589571796272,
"tokens_seen": 1689190400
},
{
"epoch": 0.02,
"objective/train/docs_used": 932516,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.2863895893096924,
"objective/train/theoretical_loss": 3.4798356900676444,
"objective/train/tokens_used": 59780576,
"theoretical_loss": 3.4798356900676444,
"tokens_seen": 1689321472
},
{
"epoch": 0.02,
"learning_rate": 9.860375541646607e-05,
"loss": 2.4437,
"theoretical_loss": 3.4798356900676444,
"tokens_seen": 1689321472
},
{
"epoch": 0.02,
"learning_rate": 9.859573102230782e-05,
"loss": 2.4915,
"theoretical_loss": 3.479812425266282,
"tokens_seen": 1689452544
},
{
"epoch": 0.02,
"learning_rate": 9.858770662814957e-05,
"loss": 2.5486,
"theoretical_loss": 3.479789162775131,
"tokens_seen": 1689583616
},
{
"epoch": 0.02,
"learning_rate": 9.857968223399134e-05,
"loss": 2.4756,
"theoretical_loss": 3.479765902593782,
"tokens_seen": 1689714688
},
{
"epoch": 0.02,
"learning_rate": 9.857165783983309e-05,
"loss": 2.4487,
"theoretical_loss": 3.4797426447218274,
"tokens_seen": 1689845760
},
{
"epoch": 0.02,
"learning_rate": 9.856363344567486e-05,
"loss": 2.5253,
"theoretical_loss": 3.4797193891588583,
"tokens_seen": 1689976832
},
{
"epoch": 0.02,
"learning_rate": 9.855560905151661e-05,
"loss": 2.4716,
"theoretical_loss": 3.479696135904467,
"tokens_seen": 1690107904
},
{
"epoch": 0.02,
"learning_rate": 9.854758465735838e-05,
"loss": 2.5138,
"theoretical_loss": 3.479672884958245,
"tokens_seen": 1690238976
},
{
"epoch": 0.02,
"learning_rate": 9.853956026320013e-05,
"loss": 2.4722,
"theoretical_loss": 3.4796496363197837,
"tokens_seen": 1690370048
},
{
"epoch": 0.02,
"learning_rate": 9.853153586904188e-05,
"loss": 2.4039,
"theoretical_loss": 3.4796263899886757,
"tokens_seen": 1690501120
},
{
"epoch": 0.02,
"learning_rate": 9.852351147488365e-05,
"loss": 2.2929,
"theoretical_loss": 3.4796031459645134,
"tokens_seen": 1690632192
},
{
"epoch": 0.02,
"learning_rate": 9.85154870807254e-05,
"loss": 2.5445,
"theoretical_loss": 3.4795799042468882,
"tokens_seen": 1690763264
},
{
"epoch": 0.02,
"learning_rate": 9.850746268656717e-05,
"loss": 2.4633,
"theoretical_loss": 3.4795566648353935,
"tokens_seen": 1690894336
},
{
"epoch": 0.02,
"objective/train/docs_used": 933032,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.596432685852051,
"objective/train/theoretical_loss": 3.4795450459943167,
"objective/train/tokens_used": 61418976,
"theoretical_loss": 3.4795450459943167,
"tokens_seen": 1690959872
},
{
"epoch": 0.02,
"learning_rate": 9.849943829240892e-05,
"loss": 2.5766,
"theoretical_loss": 3.4795334277296206,
"tokens_seen": 1691025408
},
{
"epoch": 0.02,
"learning_rate": 9.849141389825068e-05,
"loss": 2.4845,
"theoretical_loss": 3.4795101929291627,
"tokens_seen": 1691156480
},
{
"epoch": 0.03,
"learning_rate": 9.848338950409244e-05,
"loss": 2.4419,
"theoretical_loss": 3.4794869604336123,
"tokens_seen": 1691287552
},
{
"epoch": 0.03,
"learning_rate": 9.84753651099342e-05,
"loss": 2.6624,
"theoretical_loss": 3.4794637302425624,
"tokens_seen": 1691418624
},
{
"epoch": 0.03,
"learning_rate": 9.846734071577596e-05,
"loss": 2.4082,
"theoretical_loss": 3.4794405023556054,
"tokens_seen": 1691549696
},
{
"epoch": 0.03,
"learning_rate": 9.845931632161772e-05,
"loss": 2.5021,
"theoretical_loss": 3.4794172767723346,
"tokens_seen": 1691680768
},
{
"epoch": 0.03,
"learning_rate": 9.845129192745948e-05,
"loss": 2.6212,
"theoretical_loss": 3.479394053492343,
"tokens_seen": 1691811840
},
{
"epoch": 0.03,
"learning_rate": 9.844326753330124e-05,
"loss": 2.4537,
"theoretical_loss": 3.4793708325152237,
"tokens_seen": 1691942912
},
{
"epoch": 0.03,
"learning_rate": 9.843524313914299e-05,
"loss": 2.5129,
"theoretical_loss": 3.47934761384057,
"tokens_seen": 1692073984
},
{
"epoch": 0.03,
"learning_rate": 9.842721874498476e-05,
"loss": 2.5562,
"theoretical_loss": 3.4793243974679755,
"tokens_seen": 1692205056
},
{
"epoch": 0.03,
"learning_rate": 9.841919435082651e-05,
"loss": 2.409,
"theoretical_loss": 3.479301183397033,
"tokens_seen": 1692336128
},
{
"epoch": 0.03,
"learning_rate": 9.841116995666828e-05,
"loss": 2.5037,
"theoretical_loss": 3.4792779716273365,
"tokens_seen": 1692467200
},
{
"epoch": 0.03,
"objective/train/docs_used": 934026,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.466470241546631,
"objective/train/theoretical_loss": 3.4792547621584804,
"objective/train/tokens_used": 63057376,
"theoretical_loss": 3.4792547621584804,
"tokens_seen": 1692598272
},
{
"epoch": 0.03,
"learning_rate": 9.840314556251003e-05,
"loss": 2.4919,
"theoretical_loss": 3.4792547621584804,
"tokens_seen": 1692598272
},
{
"epoch": 0.03,
"learning_rate": 9.839512116835178e-05,
"loss": 2.5155,
"theoretical_loss": 3.479231554990057,
"tokens_seen": 1692729344
},
{
"epoch": 0.03,
"learning_rate": 9.838709677419355e-05,
"loss": 2.4349,
"theoretical_loss": 3.4792083501216613,
"tokens_seen": 1692860416
},
{
"epoch": 0.03,
"learning_rate": 9.83790723800353e-05,
"loss": 2.2746,
"theoretical_loss": 3.4791851475528874,
"tokens_seen": 1692991488
},
{
"epoch": 0.03,
"learning_rate": 9.837104798587707e-05,
"loss": 2.6368,
"theoretical_loss": 3.4791619472833286,
"tokens_seen": 1693122560
},
{
"epoch": 0.03,
"learning_rate": 9.836302359171882e-05,
"loss": 2.4889,
"theoretical_loss": 3.479138749312579,
"tokens_seen": 1693253632
},
{
"epoch": 0.03,
"learning_rate": 9.835499919756059e-05,
"loss": 2.5798,
"theoretical_loss": 3.4791155536402343,
"tokens_seen": 1693384704
},
{
"epoch": 0.03,
"learning_rate": 9.834697480340234e-05,
"loss": 2.6683,
"theoretical_loss": 3.479092360265887,
"tokens_seen": 1693515776
},
{
"epoch": 0.03,
"learning_rate": 9.83389504092441e-05,
"loss": 2.4851,
"theoretical_loss": 3.479069169189133,
"tokens_seen": 1693646848
},
{
"epoch": 0.03,
"learning_rate": 9.833092601508586e-05,
"loss": 2.3859,
"theoretical_loss": 3.4790459804095666,
"tokens_seen": 1693777920
},
{
"epoch": 0.03,
"learning_rate": 9.832290162092762e-05,
"loss": 2.5253,
"theoretical_loss": 3.4790227939267826,
"tokens_seen": 1693908992
},
{
"epoch": 0.03,
"learning_rate": 9.831487722676938e-05,
"loss": 2.2977,
"theoretical_loss": 3.478999609740375,
"tokens_seen": 1694040064
},
{
"epoch": 0.03,
"learning_rate": 9.830685283261113e-05,
"loss": 2.4333,
"theoretical_loss": 3.4789764278499398,
"tokens_seen": 1694171136
},
{
"epoch": 0.03,
"objective/train/docs_used": 934482,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.2957074642181396,
"objective/train/theoretical_loss": 3.478964837765585,
"objective/train/tokens_used": 64695776,
"theoretical_loss": 3.478964837765585,
"tokens_seen": 1694236672
},
{
"epoch": 0.03,
"learning_rate": 9.82988284384529e-05,
"loss": 2.5303,
"theoretical_loss": 3.4789532482550714,
"tokens_seen": 1694302208
},
{
"epoch": 0.03,
"learning_rate": 9.829080404429465e-05,
"loss": 2.4688,
"theoretical_loss": 3.478930070955365,
"tokens_seen": 1694433280
},
{
"epoch": 0.03,
"learning_rate": 9.828277965013641e-05,
"loss": 2.443,
"theoretical_loss": 3.478906895950416,
"tokens_seen": 1694564352
},
{
"epoch": 0.03,
"learning_rate": 9.827475525597817e-05,
"loss": 2.5702,
"theoretical_loss": 3.4788837232398198,
"tokens_seen": 1694695424
},
{
"epoch": 0.03,
"learning_rate": 9.826673086181993e-05,
"loss": 2.3373,
"theoretical_loss": 3.4788605528231713,
"tokens_seen": 1694826496
},
{
"epoch": 0.03,
"learning_rate": 9.82587064676617e-05,
"loss": 2.4548,
"theoretical_loss": 3.4788373847000664,
"tokens_seen": 1694957568
},
{
"epoch": 0.03,
"learning_rate": 9.825068207350345e-05,
"loss": 2.6429,
"theoretical_loss": 3.478814218870101,
"tokens_seen": 1695088640
},
{
"epoch": 0.03,
"learning_rate": 9.824265767934521e-05,
"loss": 2.3599,
"theoretical_loss": 3.4787910553328705,
"tokens_seen": 1695219712
},
{
"epoch": 0.03,
"learning_rate": 9.823463328518697e-05,
"loss": 2.3772,
"theoretical_loss": 3.4787678940879707,
"tokens_seen": 1695350784
},
{
"epoch": 0.03,
"learning_rate": 9.822660889102873e-05,
"loss": 2.5326,
"theoretical_loss": 3.478744735134998,
"tokens_seen": 1695481856
},
{
"epoch": 0.03,
"learning_rate": 9.821858449687049e-05,
"loss": 2.6442,
"theoretical_loss": 3.4787215784735475,
"tokens_seen": 1695612928
},
{
"epoch": 0.03,
"learning_rate": 9.821056010271224e-05,
"loss": 2.5572,
"theoretical_loss": 3.478698424103216,
"tokens_seen": 1695744000
},
{
"epoch": 0.03,
"objective/train/docs_used": 935631,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.5070106983184814,
"objective/train/theoretical_loss": 3.4786752720236005,
"objective/train/tokens_used": 66334176,
"theoretical_loss": 3.4786752720236005,
"tokens_seen": 1695875072
},
{
"epoch": 0.03,
"learning_rate": 9.820253570855401e-05,
"loss": 2.6289,
"theoretical_loss": 3.4786752720236005,
"tokens_seen": 1695875072
},
{
"epoch": 0.03,
"learning_rate": 9.819451131439576e-05,
"loss": 2.659,
"theoretical_loss": 3.478652122234296,
"tokens_seen": 1696006144
},
{
"epoch": 0.03,
"learning_rate": 9.818648692023753e-05,
"loss": 2.575,
"theoretical_loss": 3.4786289747348995,
"tokens_seen": 1696137216
},
{
"epoch": 0.03,
"learning_rate": 9.817846252607928e-05,
"loss": 2.637,
"theoretical_loss": 3.4786058295250077,
"tokens_seen": 1696268288
},
{
"epoch": 0.03,
"learning_rate": 9.817043813192105e-05,
"loss": 2.5304,
"theoretical_loss": 3.4785826866042173,
"tokens_seen": 1696399360
},
{
"epoch": 0.03,
"learning_rate": 9.81624137377628e-05,
"loss": 2.4708,
"theoretical_loss": 3.4785595459721246,
"tokens_seen": 1696530432
},
{
"epoch": 0.03,
"learning_rate": 9.815438934360455e-05,
"loss": 2.5821,
"theoretical_loss": 3.4785364076283267,
"tokens_seen": 1696661504
},
{
"epoch": 0.03,
"learning_rate": 9.814636494944632e-05,
"loss": 2.6878,
"theoretical_loss": 3.478513271572421,
"tokens_seen": 1696792576
},
{
"epoch": 0.03,
"learning_rate": 9.813834055528807e-05,
"loss": 2.5641,
"theoretical_loss": 3.4784901378040036,
"tokens_seen": 1696923648
},
{
"epoch": 0.03,
"learning_rate": 9.813031616112984e-05,
"loss": 2.5774,
"theoretical_loss": 3.4784670063226732,
"tokens_seen": 1697054720
},
{
"epoch": 0.03,
"learning_rate": 9.812229176697159e-05,
"loss": 2.5069,
"theoretical_loss": 3.4784438771280257,
"tokens_seen": 1697185792
},
{
"epoch": 0.03,
"learning_rate": 9.811426737281336e-05,
"loss": 2.4408,
"theoretical_loss": 3.4784207502196587,
"tokens_seen": 1697316864
},
{
"epoch": 0.03,
"learning_rate": 9.810624297865511e-05,
"loss": 2.6084,
"theoretical_loss": 3.47839762559717,
"tokens_seen": 1697447936
},
{
"epoch": 0.03,
"objective/train/docs_used": 936086,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.701629877090454,
"objective/train/theoretical_loss": 3.478386064143004,
"objective/train/tokens_used": 67972576,
"theoretical_loss": 3.478386064143004,
"tokens_seen": 1697513472
},
{
"epoch": 0.03,
"learning_rate": 9.809821858449687e-05,
"loss": 2.5616,
"theoretical_loss": 3.4783745032601567,
"tokens_seen": 1697579008
},
{
"epoch": 0.03,
"learning_rate": 9.809019419033863e-05,
"loss": 2.5896,
"theoretical_loss": 3.4783513832082176,
"tokens_seen": 1697710080
},
{
"epoch": 0.03,
"learning_rate": 9.808216979618039e-05,
"loss": 2.4651,
"theoretical_loss": 3.4783282654409495,
"tokens_seen": 1697841152
},
{
"epoch": 0.03,
"learning_rate": 9.807414540202215e-05,
"loss": 2.5471,
"theoretical_loss": 3.4783051499579507,
"tokens_seen": 1697972224
},
{
"epoch": 0.03,
"learning_rate": 9.80661210078639e-05,
"loss": 2.5858,
"theoretical_loss": 3.4782820367588183,
"tokens_seen": 1698103296
},
{
"epoch": 0.03,
"learning_rate": 9.805809661370567e-05,
"loss": 2.4588,
"theoretical_loss": 3.4782589258431518,
"tokens_seen": 1698234368
},
{
"epoch": 0.03,
"learning_rate": 9.805007221954742e-05,
"loss": 2.4407,
"theoretical_loss": 3.4782358172105483,
"tokens_seen": 1698365440
},
{
"epoch": 0.03,
"learning_rate": 9.804204782538918e-05,
"loss": 2.6123,
"theoretical_loss": 3.478212710860607,
"tokens_seen": 1698496512
},
{
"epoch": 0.03,
"learning_rate": 9.803402343123094e-05,
"loss": 2.3523,
"theoretical_loss": 3.4781896067929257,
"tokens_seen": 1698627584
},
{
"epoch": 0.03,
"learning_rate": 9.80259990370727e-05,
"loss": 2.5561,
"theoretical_loss": 3.4781665050071027,
"tokens_seen": 1698758656
},
{
"epoch": 0.03,
"learning_rate": 9.801797464291446e-05,
"loss": 2.5973,
"theoretical_loss": 3.4781434055027365,
"tokens_seen": 1698889728
},
{
"epoch": 0.03,
"learning_rate": 9.800995024875622e-05,
"loss": 2.3148,
"theoretical_loss": 3.4781203082794265,
"tokens_seen": 1699020800
},
{
"epoch": 0.03,
"objective/train/docs_used": 937360,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.2998149394989014,
"objective/train/theoretical_loss": 3.4780972133367714,
"objective/train/tokens_used": 69610976,
"theoretical_loss": 3.4780972133367714,
"tokens_seen": 1699151872
},
{
"epoch": 0.03,
"learning_rate": 9.800192585459798e-05,
"loss": 2.5517,
"theoretical_loss": 3.4780972133367714,
"tokens_seen": 1699151872
},
{
"epoch": 0.03,
"learning_rate": 9.799390146043974e-05,
"loss": 2.5891,
"theoretical_loss": 3.47807412067437,
"tokens_seen": 1699282944
},
{
"epoch": 0.03,
"learning_rate": 9.798587706628149e-05,
"loss": 2.5741,
"theoretical_loss": 3.478051030291821,
"tokens_seen": 1699414016
},
{
"epoch": 0.03,
"learning_rate": 9.797785267212326e-05,
"loss": 2.6164,
"theoretical_loss": 3.4780279421887235,
"tokens_seen": 1699545088
},
{
"epoch": 0.03,
"learning_rate": 9.796982827796501e-05,
"loss": 2.6742,
"theoretical_loss": 3.478004856364677,
"tokens_seen": 1699676160
},
{
"epoch": 0.03,
"learning_rate": 9.796180388380678e-05,
"loss": 2.5587,
"theoretical_loss": 3.4779817728192803,
"tokens_seen": 1699807232
},
{
"epoch": 0.03,
"learning_rate": 9.795377948964853e-05,
"loss": 2.41,
"theoretical_loss": 3.477958691552134,
"tokens_seen": 1699938304
},
{
"epoch": 0.03,
"learning_rate": 9.79457550954903e-05,
"loss": 2.4044,
"theoretical_loss": 3.4779356125628365,
"tokens_seen": 1700069376
},
{
"epoch": 0.03,
"learning_rate": 9.793773070133205e-05,
"loss": 2.6259,
"theoretical_loss": 3.4779125358509875,
"tokens_seen": 1700200448
},
{
"epoch": 0.03,
"learning_rate": 9.792970630717382e-05,
"loss": 2.54,
"theoretical_loss": 3.4778894614161873,
"tokens_seen": 1700331520
},
{
"epoch": 0.03,
"learning_rate": 9.792168191301557e-05,
"loss": 2.4259,
"theoretical_loss": 3.4778663892580353,
"tokens_seen": 1700462592
},
{
"epoch": 0.03,
"learning_rate": 9.791365751885732e-05,
"loss": 2.7077,
"theoretical_loss": 3.477843319376131,
"tokens_seen": 1700593664
},
{
"epoch": 0.03,
"learning_rate": 9.790563312469909e-05,
"loss": 2.6439,
"theoretical_loss": 3.4778202517700754,
"tokens_seen": 1700724736
},
{
"epoch": 0.03,
"objective/train/docs_used": 938428,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.8109610080718994,
"objective/train/theoretical_loss": 3.4778087188203655,
"objective/train/tokens_used": 71249376,
"theoretical_loss": 3.4778087188203655,
"tokens_seen": 1700790272
},
{
"epoch": 0.03,
"learning_rate": 9.789760873054084e-05,
"loss": 2.4903,
"theoretical_loss": 3.477797186439468,
"tokens_seen": 1700855808
},
{
"epoch": 0.03,
"learning_rate": 9.788958433638261e-05,
"loss": 2.6681,
"theoretical_loss": 3.477774123383909,
"tokens_seen": 1700986880
},
{
"epoch": 0.03,
"learning_rate": 9.788155994222436e-05,
"loss": 2.4544,
"theoretical_loss": 3.477751062602999,
"tokens_seen": 1701117952
},
{
"epoch": 0.03,
"learning_rate": 9.787353554806613e-05,
"loss": 2.627,
"theoretical_loss": 3.4777280040963383,
"tokens_seen": 1701249024
},
{
"epoch": 0.03,
"learning_rate": 9.786551115390788e-05,
"loss": 2.5838,
"theoretical_loss": 3.477704947863527,
"tokens_seen": 1701380096
},
{
"epoch": 0.03,
"learning_rate": 9.785748675974964e-05,
"loss": 2.5787,
"theoretical_loss": 3.4776818939041663,
"tokens_seen": 1701511168
},
{
"epoch": 0.03,
"learning_rate": 9.78494623655914e-05,
"loss": 2.6475,
"theoretical_loss": 3.477658842217857,
"tokens_seen": 1701642240
},
{
"epoch": 0.03,
"learning_rate": 9.784143797143316e-05,
"loss": 2.529,
"theoretical_loss": 3.4776357928041994,
"tokens_seen": 1701773312
},
{
"epoch": 0.03,
"learning_rate": 9.783341357727492e-05,
"loss": 2.4775,
"theoretical_loss": 3.4776127456627948,
"tokens_seen": 1701904384
},
{
"epoch": 0.03,
"learning_rate": 9.782538918311668e-05,
"loss": 2.6175,
"theoretical_loss": 3.477589700793244,
"tokens_seen": 1702035456
},
{
"epoch": 0.03,
"learning_rate": 9.781736478895844e-05,
"loss": 2.5555,
"theoretical_loss": 3.477566658195148,
"tokens_seen": 1702166528
},
{
"epoch": 0.03,
"learning_rate": 9.78093403948002e-05,
"loss": 2.5131,
"theoretical_loss": 3.4775436178681085,
"tokens_seen": 1702297600
},
{
"epoch": 0.03,
"objective/train/docs_used": 939017,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7329940795898438,
"objective/train/theoretical_loss": 3.477520579811727,
"objective/train/tokens_used": 72887776,
"theoretical_loss": 3.477520579811727,
"tokens_seen": 1702428672
},
{
"epoch": 0.03,
"learning_rate": 9.780131600064195e-05,
"loss": 2.5521,
"theoretical_loss": 3.477520579811727,
"tokens_seen": 1702428672
},
{
"epoch": 0.03,
"learning_rate": 9.779329160648371e-05,
"loss": 2.6362,
"theoretical_loss": 3.4774975440256037,
"tokens_seen": 1702559744
},
{
"epoch": 0.03,
"learning_rate": 9.778526721232547e-05,
"loss": 2.6066,
"theoretical_loss": 3.4774745105093414,
"tokens_seen": 1702690816
},
{
"epoch": 0.03,
"learning_rate": 9.777724281816723e-05,
"loss": 2.5312,
"theoretical_loss": 3.477451479262541,
"tokens_seen": 1702821888
},
{
"epoch": 0.03,
"learning_rate": 9.776921842400899e-05,
"loss": 2.5702,
"theoretical_loss": 3.477428450284805,
"tokens_seen": 1702952960
},
{
"epoch": 0.03,
"learning_rate": 9.776119402985075e-05,
"loss": 2.6024,
"theoretical_loss": 3.4774054235757346,
"tokens_seen": 1703084032
},
{
"epoch": 0.03,
"learning_rate": 9.775316963569251e-05,
"loss": 2.7145,
"theoretical_loss": 3.4773823991349317,
"tokens_seen": 1703215104
},
{
"epoch": 0.03,
"learning_rate": 9.774514524153426e-05,
"loss": 2.4479,
"theoretical_loss": 3.477359376961999,
"tokens_seen": 1703346176
},
{
"epoch": 0.03,
"learning_rate": 9.773712084737603e-05,
"loss": 2.5557,
"theoretical_loss": 3.4773363570565374,
"tokens_seen": 1703477248
},
{
"epoch": 0.03,
"learning_rate": 9.772909645321778e-05,
"loss": 2.514,
"theoretical_loss": 3.47731333941815,
"tokens_seen": 1703608320
},
{
"epoch": 0.03,
"learning_rate": 9.772107205905955e-05,
"loss": 2.4775,
"theoretical_loss": 3.4772903240464395,
"tokens_seen": 1703739392
},
{
"epoch": 0.03,
"learning_rate": 9.77130476649013e-05,
"loss": 2.5532,
"theoretical_loss": 3.4772673109410075,
"tokens_seen": 1703870464
},
{
"epoch": 0.03,
"learning_rate": 9.770502327074307e-05,
"loss": 2.5049,
"theoretical_loss": 3.4772443001014564,
"tokens_seen": 1704001536
},
{
"epoch": 0.03,
"objective/train/docs_used": 939656,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.5736076831817627,
"objective/train/theoretical_loss": 3.4772327955312625,
"objective/train/tokens_used": 74526176,
"theoretical_loss": 3.4772327955312625,
"tokens_seen": 1704067072
},
{
"epoch": 0.03,
"learning_rate": 9.769699887658482e-05,
"loss": 2.5509,
"theoretical_loss": 3.47722129152739,
"tokens_seen": 1704132608
},
{
"epoch": 0.03,
"learning_rate": 9.768897448242657e-05,
"loss": 2.5471,
"theoretical_loss": 3.4771982852184102,
"tokens_seen": 1704263680
},
{
"epoch": 0.03,
"learning_rate": 9.768095008826834e-05,
"loss": 2.4507,
"theoretical_loss": 3.4771752811741194,
"tokens_seen": 1704394752
},
{
"epoch": 0.03,
"learning_rate": 9.767292569411009e-05,
"loss": 2.5674,
"theoretical_loss": 3.477152279394122,
"tokens_seen": 1704525824
},
{
"epoch": 0.03,
"learning_rate": 9.766490129995186e-05,
"loss": 2.618,
"theoretical_loss": 3.4771292798780196,
"tokens_seen": 1704656896
},
{
"epoch": 0.03,
"learning_rate": 9.765687690579361e-05,
"loss": 2.5407,
"theoretical_loss": 3.4771062826254155,
"tokens_seen": 1704787968
},
{
"epoch": 0.03,
"learning_rate": 9.764885251163538e-05,
"loss": 2.6236,
"theoretical_loss": 3.477083287635914,
"tokens_seen": 1704919040
},
{
"epoch": 0.03,
"learning_rate": 9.764082811747713e-05,
"loss": 2.7451,
"theoretical_loss": 3.4770602949091174,
"tokens_seen": 1705050112
},
{
"epoch": 0.03,
"learning_rate": 9.763280372331889e-05,
"loss": 2.5501,
"theoretical_loss": 3.4770373044446297,
"tokens_seen": 1705181184
},
{
"epoch": 0.03,
"learning_rate": 9.762477932916065e-05,
"loss": 2.6444,
"theoretical_loss": 3.4770143162420544,
"tokens_seen": 1705312256
},
{
"epoch": 0.03,
"learning_rate": 9.76167549350024e-05,
"loss": 2.6054,
"theoretical_loss": 3.4769913303009945,
"tokens_seen": 1705443328
},
{
"epoch": 0.03,
"learning_rate": 9.760873054084417e-05,
"loss": 2.6033,
"theoretical_loss": 3.4769683466210544,
"tokens_seen": 1705574400
},
{
"epoch": 0.03,
"objective/train/docs_used": 940998,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.642338752746582,
"objective/train/theoretical_loss": 3.4769453652018374,
"objective/train/tokens_used": 76164576,
"theoretical_loss": 3.4769453652018374,
"tokens_seen": 1705705472
},
{
"epoch": 0.03,
"learning_rate": 9.760070614668593e-05,
"loss": 2.5956,
"theoretical_loss": 3.4769453652018374,
"tokens_seen": 1705705472
},
{
"epoch": 0.03,
"learning_rate": 9.759268175252769e-05,
"loss": 2.5025,
"theoretical_loss": 3.4769223860429483,
"tokens_seen": 1705836544
},
{
"epoch": 0.03,
"learning_rate": 9.758465735836945e-05,
"loss": 2.6917,
"theoretical_loss": 3.47689940914399,
"tokens_seen": 1705967616
},
{
"epoch": 0.03,
"learning_rate": 9.757663296421121e-05,
"loss": 2.5622,
"theoretical_loss": 3.4768764345045677,
"tokens_seen": 1706098688
},
{
"epoch": 0.03,
"learning_rate": 9.756860857005296e-05,
"loss": 2.5032,
"theoretical_loss": 3.4768534621242853,
"tokens_seen": 1706229760
},
{
"epoch": 0.03,
"learning_rate": 9.756058417589472e-05,
"loss": 2.6986,
"theoretical_loss": 3.4768304920027466,
"tokens_seen": 1706360832
},
{
"epoch": 0.03,
"learning_rate": 9.755255978173648e-05,
"loss": 2.5713,
"theoretical_loss": 3.4768075241395566,
"tokens_seen": 1706491904
},
{
"epoch": 0.03,
"learning_rate": 9.754453538757824e-05,
"loss": 2.6232,
"theoretical_loss": 3.4767845585343196,
"tokens_seen": 1706622976
},
{
"epoch": 0.03,
"learning_rate": 9.753651099342e-05,
"loss": 2.5417,
"theoretical_loss": 3.47676159518664,
"tokens_seen": 1706754048
},
{
"epoch": 0.03,
"learning_rate": 9.752848659926176e-05,
"loss": 2.5311,
"theoretical_loss": 3.4767386340961233,
"tokens_seen": 1706885120
},
{
"epoch": 0.03,
"learning_rate": 9.752046220510352e-05,
"loss": 2.5849,
"theoretical_loss": 3.4767156752623736,
"tokens_seen": 1707016192
},
{
"epoch": 0.03,
"learning_rate": 9.751243781094528e-05,
"loss": 2.6568,
"theoretical_loss": 3.476692718684996,
"tokens_seen": 1707147264
},
{
"epoch": 0.03,
"learning_rate": 9.750441341678703e-05,
"loss": 2.4312,
"theoretical_loss": 3.4766697643635958,
"tokens_seen": 1707278336
},
{
"epoch": 0.03,
"objective/train/docs_used": 942016,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.860006809234619,
"objective/train/theoretical_loss": 3.4766582880487635,
"objective/train/tokens_used": 77802976,
"theoretical_loss": 3.4766582880487635,
"tokens_seen": 1707343872
},
{
"epoch": 0.03,
"learning_rate": 9.74963890226288e-05,
"loss": 2.6738,
"theoretical_loss": 3.4766468122977776,
"tokens_seen": 1707409408
},
{
"epoch": 0.03,
"learning_rate": 9.748836462847055e-05,
"loss": 2.507,
"theoretical_loss": 3.4766238624871475,
"tokens_seen": 1707540480
},
{
"epoch": 0.03,
"learning_rate": 9.748034023431232e-05,
"loss": 2.6324,
"theoretical_loss": 3.4766009149313097,
"tokens_seen": 1707671552
},
{
"epoch": 0.04,
"learning_rate": 9.747231584015407e-05,
"loss": 2.5417,
"theoretical_loss": 3.4765779696298704,
"tokens_seen": 1707802624
},
{
"epoch": 0.04,
"learning_rate": 9.746429144599584e-05,
"loss": 2.568,
"theoretical_loss": 3.476555026582435,
"tokens_seen": 1707933696
},
{
"epoch": 0.04,
"learning_rate": 9.745626705183759e-05,
"loss": 2.4916,
"theoretical_loss": 3.4765320857886097,
"tokens_seen": 1708064768
},
{
"epoch": 0.04,
"learning_rate": 9.744824265767934e-05,
"loss": 2.4715,
"theoretical_loss": 3.476509147247999,
"tokens_seen": 1708195840
},
{
"epoch": 0.04,
"learning_rate": 9.744021826352111e-05,
"loss": 2.4469,
"theoretical_loss": 3.4764862109602097,
"tokens_seen": 1708326912
},
{
"epoch": 0.04,
"learning_rate": 9.743219386936286e-05,
"loss": 2.6352,
"theoretical_loss": 3.476463276924847,
"tokens_seen": 1708457984
},
{
"epoch": 0.04,
"learning_rate": 9.742416947520463e-05,
"loss": 2.6409,
"theoretical_loss": 3.4764403451415173,
"tokens_seen": 1708589056
},
{
"epoch": 0.04,
"learning_rate": 9.741614508104638e-05,
"loss": 2.6966,
"theoretical_loss": 3.476417415609827,
"tokens_seen": 1708720128
},
{
"epoch": 0.04,
"learning_rate": 9.740812068688815e-05,
"loss": 2.5196,
"theoretical_loss": 3.476394488329382,
"tokens_seen": 1708851200
},
{
"epoch": 0.04,
"objective/train/docs_used": 942685,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.43986439704895,
"objective/train/theoretical_loss": 3.4763715632997885,
"objective/train/tokens_used": 79441376,
"theoretical_loss": 3.4763715632997885,
"tokens_seen": 1708982272
},
{
"epoch": 0.04,
"learning_rate": 9.74000962927299e-05,
"loss": 2.5859,
"theoretical_loss": 3.4763715632997885,
"tokens_seen": 1708982272
},
{
"epoch": 0.04,
"learning_rate": 9.739207189857166e-05,
"loss": 2.669,
"theoretical_loss": 3.476348640520653,
"tokens_seen": 1709113344
},
{
"epoch": 0.04,
"learning_rate": 9.738404750441342e-05,
"loss": 2.6132,
"theoretical_loss": 3.476325719991582,
"tokens_seen": 1709244416
},
{
"epoch": 0.04,
"learning_rate": 9.737602311025518e-05,
"loss": 2.5786,
"theoretical_loss": 3.4763028017121824,
"tokens_seen": 1709375488
},
{
"epoch": 0.04,
"learning_rate": 9.736799871609694e-05,
"loss": 2.4671,
"theoretical_loss": 3.476279885682061,
"tokens_seen": 1709506560
},
{
"epoch": 0.04,
"learning_rate": 9.73599743219387e-05,
"loss": 2.8103,
"theoretical_loss": 3.476256971900824,
"tokens_seen": 1709637632
},
{
"epoch": 0.04,
"learning_rate": 9.735194992778046e-05,
"loss": 2.6216,
"theoretical_loss": 3.476234060368079,
"tokens_seen": 1709768704
},
{
"epoch": 0.04,
"learning_rate": 9.734392553362222e-05,
"loss": 2.5491,
"theoretical_loss": 3.476211151083432,
"tokens_seen": 1709899776
},
{
"epoch": 0.04,
"learning_rate": 9.733590113946397e-05,
"loss": 2.5661,
"theoretical_loss": 3.476188244046491,
"tokens_seen": 1710030848
},
{
"epoch": 0.04,
"learning_rate": 9.732787674530574e-05,
"loss": 2.7132,
"theoretical_loss": 3.4761653392568634,
"tokens_seen": 1710161920
},
{
"epoch": 0.04,
"learning_rate": 9.731985235114749e-05,
"loss": 2.5686,
"theoretical_loss": 3.476142436714156,
"tokens_seen": 1710292992
},
{
"epoch": 0.04,
"learning_rate": 9.731182795698925e-05,
"loss": 2.7215,
"theoretical_loss": 3.4761195364179764,
"tokens_seen": 1710424064
},
{
"epoch": 0.04,
"learning_rate": 9.730380356283101e-05,
"loss": 2.4812,
"theoretical_loss": 3.476096638367932,
"tokens_seen": 1710555136
},
{
"epoch": 0.04,
"objective/train/docs_used": 944008,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.545758008956909,
"objective/train/theoretical_loss": 3.4760851901850875,
"objective/train/tokens_used": 81079776,
"theoretical_loss": 3.4760851901850875,
"tokens_seen": 1710620672
},
{
"epoch": 0.04,
"learning_rate": 9.729577916867277e-05,
"loss": 2.5449,
"theoretical_loss": 3.4760737425636297,
"tokens_seen": 1710686208
},
{
"epoch": 0.04,
"learning_rate": 9.728775477451453e-05,
"loss": 2.7649,
"theoretical_loss": 3.476050849004679,
"tokens_seen": 1710817280
},
{
"epoch": 0.04,
"learning_rate": 9.72797303803563e-05,
"loss": 2.4776,
"theoretical_loss": 3.476027957690686,
"tokens_seen": 1710948352
},
{
"epoch": 0.04,
"learning_rate": 9.727170598619805e-05,
"loss": 2.4808,
"theoretical_loss": 3.4760050686212596,
"tokens_seen": 1711079424
},
{
"epoch": 0.04,
"learning_rate": 9.72636815920398e-05,
"loss": 2.5526,
"theoretical_loss": 3.475982181796007,
"tokens_seen": 1711210496
},
{
"epoch": 0.04,
"learning_rate": 9.725565719788157e-05,
"loss": 2.5859,
"theoretical_loss": 3.475959297214537,
"tokens_seen": 1711341568
},
{
"epoch": 0.04,
"learning_rate": 9.724763280372332e-05,
"loss": 2.6257,
"theoretical_loss": 3.4759364148764575,
"tokens_seen": 1711472640
},
{
"epoch": 0.04,
"learning_rate": 9.723960840956509e-05,
"loss": 2.7139,
"theoretical_loss": 3.4759135347813768,
"tokens_seen": 1711603712
},
{
"epoch": 0.04,
"learning_rate": 9.723158401540684e-05,
"loss": 2.6359,
"theoretical_loss": 3.4758906569289034,
"tokens_seen": 1711734784
},
{
"epoch": 0.04,
"learning_rate": 9.722355962124861e-05,
"loss": 2.6534,
"theoretical_loss": 3.475867781318646,
"tokens_seen": 1711865856
},
{
"epoch": 0.04,
"learning_rate": 9.721553522709036e-05,
"loss": 2.6935,
"theoretical_loss": 3.4758449079502123,
"tokens_seen": 1711996928
},
{
"epoch": 0.04,
"learning_rate": 9.720751083293211e-05,
"loss": 2.6434,
"theoretical_loss": 3.475822036823212,
"tokens_seen": 1712128000
},
{
"epoch": 0.04,
"objective/train/docs_used": 944514,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.2325708866119385,
"objective/train/theoretical_loss": 3.4757991679372533,
"objective/train/tokens_used": 82718176,
"theoretical_loss": 3.4757991679372533,
"tokens_seen": 1712259072
},
{
"epoch": 0.04,
"learning_rate": 9.719948643877388e-05,
"loss": 2.3,
"theoretical_loss": 3.4757991679372533,
"tokens_seen": 1712259072
},
{
"epoch": 0.04,
"learning_rate": 9.719146204461563e-05,
"loss": 2.7386,
"theoretical_loss": 3.4757763012919454,
"tokens_seen": 1712390144
},
{
"epoch": 0.04,
"learning_rate": 9.71834376504574e-05,
"loss": 2.6512,
"theoretical_loss": 3.4757534368868974,
"tokens_seen": 1712521216
},
{
"epoch": 0.04,
"learning_rate": 9.717541325629915e-05,
"loss": 2.5087,
"theoretical_loss": 3.4757305747217178,
"tokens_seen": 1712652288
},
{
"epoch": 0.04,
"learning_rate": 9.716738886214092e-05,
"loss": 2.5342,
"theoretical_loss": 3.475707714796016,
"tokens_seen": 1712783360
},
{
"epoch": 0.04,
"learning_rate": 9.715936446798267e-05,
"loss": 2.4806,
"theoretical_loss": 3.4756848571094014,
"tokens_seen": 1712914432
},
{
"epoch": 0.04,
"learning_rate": 9.715134007382443e-05,
"loss": 2.6971,
"theoretical_loss": 3.475662001661483,
"tokens_seen": 1713045504
},
{
"epoch": 0.04,
"learning_rate": 9.714331567966619e-05,
"loss": 2.6806,
"theoretical_loss": 3.4756391484518714,
"tokens_seen": 1713176576
},
{
"epoch": 0.04,
"learning_rate": 9.713529128550795e-05,
"loss": 2.554,
"theoretical_loss": 3.475616297480175,
"tokens_seen": 1713307648
},
{
"epoch": 0.04,
"learning_rate": 9.712726689134971e-05,
"loss": 2.579,
"theoretical_loss": 3.4755934487460047,
"tokens_seen": 1713438720
},
{
"epoch": 0.04,
"learning_rate": 9.711924249719147e-05,
"loss": 2.6336,
"theoretical_loss": 3.4755706022489687,
"tokens_seen": 1713569792
},
{
"epoch": 0.04,
"learning_rate": 9.711121810303323e-05,
"loss": 2.4975,
"theoretical_loss": 3.4755477579886778,
"tokens_seen": 1713700864
},
{
"epoch": 0.04,
"learning_rate": 9.710319370887499e-05,
"loss": 2.5121,
"theoretical_loss": 3.4755249159647414,
"tokens_seen": 1713831936
},
{
"epoch": 0.04,
"objective/train/docs_used": 945657,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.636183261871338,
"objective/train/theoretical_loss": 3.4755134957912848,
"objective/train/tokens_used": 84356576,
"theoretical_loss": 3.4755134957912848,
"tokens_seen": 1713897472
},
{
"epoch": 0.04,
"learning_rate": 9.709516931471674e-05,
"loss": 2.5273,
"theoretical_loss": 3.4755020761767703,
"tokens_seen": 1713963008
},
{
"epoch": 0.04,
"learning_rate": 9.70871449205585e-05,
"loss": 2.4317,
"theoretical_loss": 3.4754792386243745,
"tokens_seen": 1714094080
},
{
"epoch": 0.04,
"learning_rate": 9.707912052640026e-05,
"loss": 2.4492,
"theoretical_loss": 3.475456403307164,
"tokens_seen": 1714225152
},
{
"epoch": 0.04,
"learning_rate": 9.707109613224202e-05,
"loss": 2.6308,
"theoretical_loss": 3.475433570224749,
"tokens_seen": 1714356224
},
{
"epoch": 0.04,
"learning_rate": 9.706307173808378e-05,
"loss": 2.5065,
"theoretical_loss": 3.47541073937674,
"tokens_seen": 1714487296
},
{
"epoch": 0.04,
"learning_rate": 9.705504734392554e-05,
"loss": 2.4417,
"theoretical_loss": 3.4753879107627483,
"tokens_seen": 1714618368
},
{
"epoch": 0.04,
"learning_rate": 9.70470229497673e-05,
"loss": 2.5615,
"theoretical_loss": 3.4753650843823833,
"tokens_seen": 1714749440
},
{
"epoch": 0.04,
"learning_rate": 9.703899855560905e-05,
"loss": 2.5363,
"theoretical_loss": 3.475342260235257,
"tokens_seen": 1714880512
},
{
"epoch": 0.04,
"learning_rate": 9.703097416145082e-05,
"loss": 2.6402,
"theoretical_loss": 3.4753194383209793,
"tokens_seen": 1715011584
},
{
"epoch": 0.04,
"learning_rate": 9.702294976729257e-05,
"loss": 2.6409,
"theoretical_loss": 3.4752966186391614,
"tokens_seen": 1715142656
},
{
"epoch": 0.04,
"learning_rate": 9.701492537313434e-05,
"loss": 2.5318,
"theoretical_loss": 3.4752738011894144,
"tokens_seen": 1715273728
},
{
"epoch": 0.04,
"learning_rate": 9.700690097897609e-05,
"loss": 2.5512,
"theoretical_loss": 3.47525098597135,
"tokens_seen": 1715404800
},
{
"epoch": 0.04,
"objective/train/docs_used": 946326,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.3408730030059814,
"objective/train/theoretical_loss": 3.4752281729845786,
"objective/train/tokens_used": 85994976,
"theoretical_loss": 3.4752281729845786,
"tokens_seen": 1715535872
},
{
"epoch": 0.04,
"learning_rate": 9.699887658481786e-05,
"loss": 2.4938,
"theoretical_loss": 3.4752281729845786,
"tokens_seen": 1715535872
},
{
"epoch": 0.04,
"learning_rate": 9.699085219065961e-05,
"loss": 2.5903,
"theoretical_loss": 3.475205362228712,
"tokens_seen": 1715666944
},
{
"epoch": 0.04,
"learning_rate": 9.698282779650138e-05,
"loss": 2.6315,
"theoretical_loss": 3.4751825537033607,
"tokens_seen": 1715798016
},
{
"epoch": 0.04,
"learning_rate": 9.697480340234313e-05,
"loss": 2.5496,
"theoretical_loss": 3.4751597474081377,
"tokens_seen": 1715929088
},
{
"epoch": 0.04,
"learning_rate": 9.696677900818488e-05,
"loss": 2.5582,
"theoretical_loss": 3.4751369433426538,
"tokens_seen": 1716060160
},
{
"epoch": 0.04,
"learning_rate": 9.695875461402665e-05,
"loss": 2.4528,
"theoretical_loss": 3.4751141415065208,
"tokens_seen": 1716191232
},
{
"epoch": 0.04,
"learning_rate": 9.69507302198684e-05,
"loss": 2.4352,
"theoretical_loss": 3.4750913418993505,
"tokens_seen": 1716322304
},
{
"epoch": 0.04,
"learning_rate": 9.694270582571017e-05,
"loss": 2.4735,
"theoretical_loss": 3.475068544520755,
"tokens_seen": 1716453376
},
{
"epoch": 0.04,
"learning_rate": 9.693468143155192e-05,
"loss": 2.6383,
"theoretical_loss": 3.475045749370346,
"tokens_seen": 1716584448
},
{
"epoch": 0.04,
"learning_rate": 9.692665703739369e-05,
"loss": 2.6037,
"theoretical_loss": 3.475022956447736,
"tokens_seen": 1716715520
},
{
"epoch": 0.04,
"learning_rate": 9.691863264323544e-05,
"loss": 2.5211,
"theoretical_loss": 3.4750001657525367,
"tokens_seen": 1716846592
},
{
"epoch": 0.04,
"learning_rate": 9.69106082490772e-05,
"loss": 2.5279,
"theoretical_loss": 3.474977377284361,
"tokens_seen": 1716977664
},
{
"epoch": 0.04,
"learning_rate": 9.690258385491896e-05,
"loss": 2.471,
"theoretical_loss": 3.4749545910428212,
"tokens_seen": 1717108736
},
{
"epoch": 0.04,
"objective/train/docs_used": 947520,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7320938110351562,
"objective/train/theoretical_loss": 3.474943198756918,
"objective/train/tokens_used": 87633376,
"theoretical_loss": 3.474943198756918,
"tokens_seen": 1717174272
},
{
"epoch": 0.04,
"learning_rate": 9.689455946076072e-05,
"loss": 2.5489,
"theoretical_loss": 3.474931807027529,
"tokens_seen": 1717239808
},
{
"epoch": 0.04,
"learning_rate": 9.688653506660248e-05,
"loss": 2.3712,
"theoretical_loss": 3.474909025238098,
"tokens_seen": 1717370880
},
{
"epoch": 0.04,
"learning_rate": 9.687851067244424e-05,
"loss": 2.5661,
"theoretical_loss": 3.4748862456741403,
"tokens_seen": 1717501952
},
{
"epoch": 0.04,
"learning_rate": 9.6870486278286e-05,
"loss": 2.6223,
"theoretical_loss": 3.474863468335269,
"tokens_seen": 1717633024
},
{
"epoch": 0.04,
"learning_rate": 9.686246188412776e-05,
"loss": 2.6434,
"theoretical_loss": 3.474840693221096,
"tokens_seen": 1717764096
},
{
"epoch": 0.04,
"learning_rate": 9.685443748996951e-05,
"loss": 2.6146,
"theoretical_loss": 3.474817920331236,
"tokens_seen": 1717895168
},
{
"epoch": 0.04,
"learning_rate": 9.684641309581128e-05,
"loss": 2.5673,
"theoretical_loss": 3.474795149665301,
"tokens_seen": 1718026240
},
{
"epoch": 0.04,
"learning_rate": 9.683838870165303e-05,
"loss": 2.5127,
"theoretical_loss": 3.4747723812229045,
"tokens_seen": 1718157312
},
{
"epoch": 0.04,
"learning_rate": 9.68303643074948e-05,
"loss": 2.5503,
"theoretical_loss": 3.4747496150036596,
"tokens_seen": 1718288384
},
{
"epoch": 0.04,
"learning_rate": 9.682233991333655e-05,
"loss": 2.6373,
"theoretical_loss": 3.47472685100718,
"tokens_seen": 1718419456
},
{
"epoch": 0.04,
"learning_rate": 9.681431551917831e-05,
"loss": 2.5278,
"theoretical_loss": 3.4747040892330787,
"tokens_seen": 1718550528
},
{
"epoch": 0.04,
"learning_rate": 9.680629112502007e-05,
"loss": 2.5697,
"theoretical_loss": 3.474681329680969,
"tokens_seen": 1718681600
},
{
"epoch": 0.04,
"objective/train/docs_used": 948221,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.1624414920806885,
"objective/train/theoretical_loss": 3.4746585723504655,
"objective/train/tokens_used": 89271776,
"theoretical_loss": 3.4746585723504655,
"tokens_seen": 1718812672
},
{
"epoch": 0.04,
"learning_rate": 9.679826673086182e-05,
"loss": 2.4717,
"theoretical_loss": 3.4746585723504655,
"tokens_seen": 1718812672
},
{
"epoch": 0.04,
"learning_rate": 9.679024233670359e-05,
"loss": 2.5336,
"theoretical_loss": 3.4746358172411815,
"tokens_seen": 1718943744
},
{
"epoch": 0.04,
"learning_rate": 9.678221794254534e-05,
"loss": 2.5264,
"theoretical_loss": 3.4746130643527304,
"tokens_seen": 1719074816
},
{
"epoch": 0.04,
"learning_rate": 9.677419354838711e-05,
"loss": 2.598,
"theoretical_loss": 3.474590313684727,
"tokens_seen": 1719205888
},
{
"epoch": 0.04,
"learning_rate": 9.676616915422886e-05,
"loss": 2.492,
"theoretical_loss": 3.474567565236785,
"tokens_seen": 1719336960
},
{
"epoch": 0.04,
"learning_rate": 9.675814476007063e-05,
"loss": 2.4323,
"theoretical_loss": 3.4745448190085177,
"tokens_seen": 1719468032
},
{
"epoch": 0.04,
"learning_rate": 9.675012036591238e-05,
"loss": 2.5456,
"theoretical_loss": 3.4745220749995407,
"tokens_seen": 1719599104
},
{
"epoch": 0.04,
"learning_rate": 9.674209597175413e-05,
"loss": 2.4039,
"theoretical_loss": 3.474499333209468,
"tokens_seen": 1719730176
},
{
"epoch": 0.04,
"learning_rate": 9.67340715775959e-05,
"loss": 2.4198,
"theoretical_loss": 3.474476593637913,
"tokens_seen": 1719861248
},
{
"epoch": 0.04,
"learning_rate": 9.672604718343765e-05,
"loss": 2.6933,
"theoretical_loss": 3.4744538562844913,
"tokens_seen": 1719992320
},
{
"epoch": 0.04,
"learning_rate": 9.671802278927942e-05,
"loss": 2.5493,
"theoretical_loss": 3.4744311211488172,
"tokens_seen": 1720123392
},
{
"epoch": 0.04,
"learning_rate": 9.670999839512117e-05,
"loss": 2.6969,
"theoretical_loss": 3.4744083882305055,
"tokens_seen": 1720254464
},
{
"epoch": 0.04,
"learning_rate": 9.670197400096294e-05,
"loss": 2.4579,
"theoretical_loss": 3.474385657529171,
"tokens_seen": 1720385536
},
{
"epoch": 0.04,
"objective/train/docs_used": 949201,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.1461188793182373,
"objective/train/theoretical_loss": 3.47437429300975,
"objective/train/tokens_used": 90910176,
"theoretical_loss": 3.47437429300975,
"tokens_seen": 1720451072
},
{
"epoch": 0.04,
"learning_rate": 9.669394960680469e-05,
"loss": 2.4628,
"theoretical_loss": 3.4743629290444282,
"tokens_seen": 1720516608
},
{
"epoch": 0.04,
"learning_rate": 9.668592521264645e-05,
"loss": 2.4322,
"theoretical_loss": 3.474340202775893,
"tokens_seen": 1720647680
},
{
"epoch": 0.04,
"learning_rate": 9.667790081848821e-05,
"loss": 2.5498,
"theoretical_loss": 3.47431747872318,
"tokens_seen": 1720778752
},
{
"epoch": 0.04,
"learning_rate": 9.666987642432997e-05,
"loss": 2.5226,
"theoretical_loss": 3.474294756885904,
"tokens_seen": 1720909824
},
{
"epoch": 0.04,
"learning_rate": 9.666185203017173e-05,
"loss": 2.4954,
"theoretical_loss": 3.474272037263681,
"tokens_seen": 1721040896
},
{
"epoch": 0.04,
"learning_rate": 9.665382763601349e-05,
"loss": 2.6002,
"theoretical_loss": 3.474249319856126,
"tokens_seen": 1721171968
},
{
"epoch": 0.04,
"learning_rate": 9.664580324185525e-05,
"loss": 2.4917,
"theoretical_loss": 3.4742266046628543,
"tokens_seen": 1721303040
},
{
"epoch": 0.04,
"learning_rate": 9.6637778847697e-05,
"loss": 2.479,
"theoretical_loss": 3.474203891683482,
"tokens_seen": 1721434112
},
{
"epoch": 0.04,
"learning_rate": 9.662975445353877e-05,
"loss": 2.479,
"theoretical_loss": 3.474181180917624,
"tokens_seen": 1721565184
},
{
"epoch": 0.04,
"learning_rate": 9.662173005938053e-05,
"loss": 2.717,
"theoretical_loss": 3.4741584723648975,
"tokens_seen": 1721696256
},
{
"epoch": 0.04,
"learning_rate": 9.661370566522228e-05,
"loss": 2.6332,
"theoretical_loss": 3.474135766024917,
"tokens_seen": 1721827328
},
{
"epoch": 0.04,
"learning_rate": 9.660568127106405e-05,
"loss": 2.6836,
"theoretical_loss": 3.474113061897299,
"tokens_seen": 1721958400
},
{
"epoch": 0.04,
"objective/train/docs_used": 949909,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.497939348220825,
"objective/train/theoretical_loss": 3.474090359981659,
"objective/train/tokens_used": 92548576,
"theoretical_loss": 3.474090359981659,
"tokens_seen": 1722089472
},
{
"epoch": 0.04,
"learning_rate": 9.65976568769058e-05,
"loss": 2.5073,
"theoretical_loss": 3.474090359981659,
"tokens_seen": 1722089472
},
{
"epoch": 0.04,
"learning_rate": 9.658963248274757e-05,
"loss": 2.4187,
"theoretical_loss": 3.4740676602776146,
"tokens_seen": 1722220544
},
{
"epoch": 0.04,
"learning_rate": 9.658160808858932e-05,
"loss": 2.5354,
"theoretical_loss": 3.4740449627847805,
"tokens_seen": 1722351616
},
{
"epoch": 0.04,
"learning_rate": 9.657358369443108e-05,
"loss": 2.428,
"theoretical_loss": 3.4740222675027734,
"tokens_seen": 1722482688
},
{
"epoch": 0.04,
"learning_rate": 9.656555930027284e-05,
"loss": 2.5802,
"theoretical_loss": 3.4739995744312107,
"tokens_seen": 1722613760
},
{
"epoch": 0.04,
"learning_rate": 9.655753490611459e-05,
"loss": 2.6047,
"theoretical_loss": 3.473976883569708,
"tokens_seen": 1722744832
},
{
"epoch": 0.04,
"learning_rate": 9.654951051195636e-05,
"loss": 2.6242,
"theoretical_loss": 3.4739541949178827,
"tokens_seen": 1722875904
},
{
"epoch": 0.04,
"learning_rate": 9.654148611779811e-05,
"loss": 2.4371,
"theoretical_loss": 3.47393150847535,
"tokens_seen": 1723006976
},
{
"epoch": 0.04,
"learning_rate": 9.653346172363988e-05,
"loss": 2.3743,
"theoretical_loss": 3.473908824241729,
"tokens_seen": 1723138048
},
{
"epoch": 0.04,
"learning_rate": 9.652543732948163e-05,
"loss": 2.7518,
"theoretical_loss": 3.4738861422166343,
"tokens_seen": 1723269120
},
{
"epoch": 0.04,
"learning_rate": 9.65174129353234e-05,
"loss": 2.44,
"theoretical_loss": 3.4738634623996845,
"tokens_seen": 1723400192
},
{
"epoch": 0.04,
"learning_rate": 9.650938854116515e-05,
"loss": 2.5699,
"theoretical_loss": 3.4738407847904966,
"tokens_seen": 1723531264
},
{
"epoch": 0.04,
"learning_rate": 9.65013641470069e-05,
"loss": 2.4851,
"theoretical_loss": 3.473818109388687,
"tokens_seen": 1723662336
},
{
"epoch": 0.04,
"objective/train/docs_used": 951235,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.2583699226379395,
"objective/train/theoretical_loss": 3.47380677251543,
"objective/train/tokens_used": 94186976,
"theoretical_loss": 3.47380677251543,
"tokens_seen": 1723727872
},
{
"epoch": 0.04,
"learning_rate": 9.649333975284867e-05,
"loss": 2.5272,
"theoretical_loss": 3.473795436193874,
"tokens_seen": 1723793408
},
{
"epoch": 0.04,
"learning_rate": 9.648531535869042e-05,
"loss": 2.5135,
"theoretical_loss": 3.4737727652056742,
"tokens_seen": 1723924480
},
{
"epoch": 0.04,
"learning_rate": 9.647729096453219e-05,
"loss": 2.6295,
"theoretical_loss": 3.473750096423705,
"tokens_seen": 1724055552
},
{
"epoch": 0.04,
"learning_rate": 9.646926657037394e-05,
"loss": 2.6414,
"theoretical_loss": 3.4737274298475844,
"tokens_seen": 1724186624
},
{
"epoch": 0.05,
"learning_rate": 9.646124217621571e-05,
"loss": 2.5256,
"theoretical_loss": 3.4737047654769304,
"tokens_seen": 1724317696
},
{
"epoch": 0.05,
"learning_rate": 9.645321778205746e-05,
"loss": 2.664,
"theoretical_loss": 3.4736821033113605,
"tokens_seen": 1724448768
},
{
"epoch": 0.05,
"learning_rate": 9.644519338789922e-05,
"loss": 2.5327,
"theoretical_loss": 3.473659443350493,
"tokens_seen": 1724579840
},
{
"epoch": 0.05,
"learning_rate": 9.643716899374098e-05,
"loss": 2.6298,
"theoretical_loss": 3.4736367855939445,
"tokens_seen": 1724710912
},
{
"epoch": 0.05,
"learning_rate": 9.642914459958274e-05,
"loss": 2.5402,
"theoretical_loss": 3.4736141300413337,
"tokens_seen": 1724841984
},
{
"epoch": 0.05,
"learning_rate": 9.64211202054245e-05,
"loss": 2.4692,
"theoretical_loss": 3.47359147669228,
"tokens_seen": 1724973056
},
{
"epoch": 0.05,
"learning_rate": 9.641309581126626e-05,
"loss": 2.6825,
"theoretical_loss": 3.4735688255464003,
"tokens_seen": 1725104128
},
{
"epoch": 0.05,
"learning_rate": 9.640507141710802e-05,
"loss": 2.4039,
"theoretical_loss": 3.4735461766033136,
"tokens_seen": 1725235200
},
{
"epoch": 0.05,
"objective/train/docs_used": 951881,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6248931884765625,
"objective/train/theoretical_loss": 3.4735235298626375,
"objective/train/tokens_used": 95825376,
"theoretical_loss": 3.4735235298626375,
"tokens_seen": 1725366272
},
{
"epoch": 0.05,
"learning_rate": 9.639704702294978e-05,
"loss": 2.55,
"theoretical_loss": 3.4735235298626375,
"tokens_seen": 1725366272
},
{
"epoch": 0.05,
"learning_rate": 9.638902262879153e-05,
"loss": 2.3494,
"theoretical_loss": 3.473500885323992,
"tokens_seen": 1725497344
},
{
"epoch": 0.05,
"learning_rate": 9.63809982346333e-05,
"loss": 2.6021,
"theoretical_loss": 3.473478242986994,
"tokens_seen": 1725628416
},
{
"epoch": 0.05,
"learning_rate": 9.637297384047505e-05,
"loss": 2.4924,
"theoretical_loss": 3.473455602851264,
"tokens_seen": 1725759488
},
{
"epoch": 0.05,
"learning_rate": 9.636494944631682e-05,
"loss": 2.4347,
"theoretical_loss": 3.4734329649164195,
"tokens_seen": 1725890560
},
{
"epoch": 0.05,
"learning_rate": 9.635692505215857e-05,
"loss": 2.3637,
"theoretical_loss": 3.47341032918208,
"tokens_seen": 1726021632
},
{
"epoch": 0.05,
"learning_rate": 9.634890065800034e-05,
"loss": 2.5177,
"theoretical_loss": 3.4733876956478644,
"tokens_seen": 1726152704
},
{
"epoch": 0.05,
"learning_rate": 9.634087626384209e-05,
"loss": 2.5704,
"theoretical_loss": 3.473365064313392,
"tokens_seen": 1726283776
},
{
"epoch": 0.05,
"learning_rate": 9.633285186968385e-05,
"loss": 2.6099,
"theoretical_loss": 3.4733424351782816,
"tokens_seen": 1726414848
},
{
"epoch": 0.05,
"learning_rate": 9.632482747552561e-05,
"loss": 2.6702,
"theoretical_loss": 3.4733198082421533,
"tokens_seen": 1726545920
},
{
"epoch": 0.05,
"learning_rate": 9.631680308136736e-05,
"loss": 2.5298,
"theoretical_loss": 3.4732971835046254,
"tokens_seen": 1726676992
},
{
"epoch": 0.05,
"learning_rate": 9.630877868720913e-05,
"loss": 2.4978,
"theoretical_loss": 3.4732745609653177,
"tokens_seen": 1726808064
},
{
"epoch": 0.05,
"learning_rate": 9.630075429305088e-05,
"loss": 2.4702,
"theoretical_loss": 3.4732519406238507,
"tokens_seen": 1726939136
},
{
"epoch": 0.05,
"objective/train/docs_used": 953135,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 3.060582160949707,
"objective/train/theoretical_loss": 3.4732406312771884,
"objective/train/tokens_used": 97463776,
"theoretical_loss": 3.4732406312771884,
"tokens_seen": 1727004672
},
{
"epoch": 0.05,
"learning_rate": 9.629272989889265e-05,
"loss": 2.5876,
"theoretical_loss": 3.473229322479843,
"tokens_seen": 1727070208
},
{
"epoch": 0.05,
"learning_rate": 9.62847055047344e-05,
"loss": 2.6043,
"theoretical_loss": 3.473206706532915,
"tokens_seen": 1727201280
},
{
"epoch": 0.05,
"learning_rate": 9.627668111057617e-05,
"loss": 2.5915,
"theoretical_loss": 3.4731840927826867,
"tokens_seen": 1727332352
},
{
"epoch": 0.05,
"learning_rate": 9.626865671641792e-05,
"loss": 2.4664,
"theoretical_loss": 3.4731614812287774,
"tokens_seen": 1727463424
},
{
"epoch": 0.05,
"learning_rate": 9.626063232225967e-05,
"loss": 2.5555,
"theoretical_loss": 3.4731388718708076,
"tokens_seen": 1727594496
},
{
"epoch": 0.05,
"learning_rate": 9.625260792810144e-05,
"loss": 2.5768,
"theoretical_loss": 3.4731162647083975,
"tokens_seen": 1727725568
},
{
"epoch": 0.05,
"learning_rate": 9.62445835339432e-05,
"loss": 2.6285,
"theoretical_loss": 3.473093659741167,
"tokens_seen": 1727856640
},
{
"epoch": 0.05,
"learning_rate": 9.623655913978496e-05,
"loss": 2.6888,
"theoretical_loss": 3.473071056968737,
"tokens_seen": 1727987712
},
{
"epoch": 0.05,
"learning_rate": 9.622853474562671e-05,
"loss": 2.6184,
"theoretical_loss": 3.4730484563907273,
"tokens_seen": 1728118784
},
{
"epoch": 0.05,
"learning_rate": 9.622051035146848e-05,
"loss": 2.6123,
"theoretical_loss": 3.473025858006759,
"tokens_seen": 1728249856
},
{
"epoch": 0.05,
"learning_rate": 9.621248595731023e-05,
"loss": 2.4683,
"theoretical_loss": 3.4730032618164524,
"tokens_seen": 1728380928
},
{
"epoch": 0.05,
"learning_rate": 9.620446156315199e-05,
"loss": 2.4986,
"theoretical_loss": 3.472980667819428,
"tokens_seen": 1728512000
},
{
"epoch": 0.05,
"objective/train/docs_used": 953685,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.4338219165802,
"objective/train/theoretical_loss": 3.472958076015307,
"objective/train/tokens_used": 99102176,
"theoretical_loss": 3.472958076015307,
"tokens_seen": 1728643072
},
{
"epoch": 0.05,
"learning_rate": 9.619643716899375e-05,
"loss": 2.6793,
"theoretical_loss": 3.472958076015307,
"tokens_seen": 1728643072
},
{
"epoch": 0.05,
"learning_rate": 9.61884127748355e-05,
"loss": 2.3993,
"theoretical_loss": 3.4729354864037107,
"tokens_seen": 1728774144
},
{
"epoch": 0.05,
"learning_rate": 9.618038838067727e-05,
"loss": 2.5727,
"theoretical_loss": 3.4729128989842595,
"tokens_seen": 1728905216
},
{
"epoch": 0.05,
"learning_rate": 9.617236398651903e-05,
"loss": 2.5781,
"theoretical_loss": 3.4728903137565745,
"tokens_seen": 1729036288
},
{
"epoch": 0.05,
"learning_rate": 9.616433959236079e-05,
"loss": 2.5062,
"theoretical_loss": 3.472867730720277,
"tokens_seen": 1729167360
},
{
"epoch": 0.05,
"learning_rate": 9.615631519820255e-05,
"loss": 2.6331,
"theoretical_loss": 3.4728451498749884,
"tokens_seen": 1729298432
},
{
"epoch": 0.05,
"learning_rate": 9.61482908040443e-05,
"loss": 2.6771,
"theoretical_loss": 3.4728225712203304,
"tokens_seen": 1729429504
},
{
"epoch": 0.05,
"learning_rate": 9.614026640988607e-05,
"loss": 2.562,
"theoretical_loss": 3.472799994755924,
"tokens_seen": 1729560576
},
{
"epoch": 0.05,
"learning_rate": 9.613224201572782e-05,
"loss": 2.5274,
"theoretical_loss": 3.4727774204813904,
"tokens_seen": 1729691648
},
{
"epoch": 0.05,
"learning_rate": 9.612421762156959e-05,
"loss": 2.4391,
"theoretical_loss": 3.472754848396352,
"tokens_seen": 1729822720
},
{
"epoch": 0.05,
"learning_rate": 9.611619322741134e-05,
"loss": 2.5883,
"theoretical_loss": 3.4727322785004304,
"tokens_seen": 1729953792
},
{
"epoch": 0.05,
"learning_rate": 9.610816883325309e-05,
"loss": 2.7006,
"theoretical_loss": 3.4727097107932474,
"tokens_seen": 1730084864
},
{
"epoch": 0.05,
"learning_rate": 9.610014443909486e-05,
"loss": 2.5706,
"theoretical_loss": 3.472687145274425,
"tokens_seen": 1730215936
},
{
"epoch": 0.05,
"objective/train/docs_used": 954893,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.3295490741729736,
"objective/train/theoretical_loss": 3.472675863335531,
"objective/train/tokens_used": 100740576,
"theoretical_loss": 3.472675863335531,
"tokens_seen": 1730281472
},
{
"epoch": 0.05,
"learning_rate": 9.609212004493661e-05,
"loss": 2.5245,
"theoretical_loss": 3.472664581943585,
"tokens_seen": 1730347008
},
{
"epoch": 0.05,
"learning_rate": 9.608409565077838e-05,
"loss": 2.5552,
"theoretical_loss": 3.4726420208003494,
"tokens_seen": 1730478080
},
{
"epoch": 0.05,
"learning_rate": 9.607607125662013e-05,
"loss": 2.5185,
"theoretical_loss": 3.4726194618443413,
"tokens_seen": 1730609152
},
{
"epoch": 0.05,
"learning_rate": 9.60680468624619e-05,
"loss": 2.53,
"theoretical_loss": 3.472596905075182,
"tokens_seen": 1730740224
},
{
"epoch": 0.05,
"learning_rate": 9.606002246830365e-05,
"loss": 2.6368,
"theoretical_loss": 3.472574350492495,
"tokens_seen": 1730871296
},
{
"epoch": 0.05,
"learning_rate": 9.60519980741454e-05,
"loss": 2.4409,
"theoretical_loss": 3.4725517980959015,
"tokens_seen": 1731002368
},
{
"epoch": 0.05,
"learning_rate": 9.604397367998717e-05,
"loss": 2.6554,
"theoretical_loss": 3.472529247885025,
"tokens_seen": 1731133440
},
{
"epoch": 0.05,
"learning_rate": 9.603594928582892e-05,
"loss": 2.6305,
"theoretical_loss": 3.4725066998594882,
"tokens_seen": 1731264512
},
{
"epoch": 0.05,
"learning_rate": 9.602792489167069e-05,
"loss": 2.5356,
"theoretical_loss": 3.4724841540189137,
"tokens_seen": 1731395584
},
{
"epoch": 0.05,
"learning_rate": 9.601990049751244e-05,
"loss": 2.5259,
"theoretical_loss": 3.4724616103629242,
"tokens_seen": 1731526656
},
{
"epoch": 0.05,
"learning_rate": 9.60118761033542e-05,
"loss": 2.5755,
"theoretical_loss": 3.472439068891143,
"tokens_seen": 1731657728
},
{
"epoch": 0.05,
"learning_rate": 9.600385170919596e-05,
"loss": 2.6055,
"theoretical_loss": 3.4724165296031924,
"tokens_seen": 1731788800
},
{
"epoch": 0.05,
"objective/train/docs_used": 956184,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.5680079460144043,
"objective/train/theoretical_loss": 3.472393992498697,
"objective/train/tokens_used": 102378976,
"theoretical_loss": 3.472393992498697,
"tokens_seen": 1731919872
},
{
"epoch": 0.05,
"learning_rate": 9.599582731503772e-05,
"loss": 2.6725,
"theoretical_loss": 3.472393992498697,
"tokens_seen": 1731919872
},
{
"epoch": 0.05,
"learning_rate": 9.598780292087948e-05,
"loss": 2.5937,
"theoretical_loss": 3.472371457577279,
"tokens_seen": 1732050944
},
{
"epoch": 0.05,
"learning_rate": 9.597977852672124e-05,
"loss": 2.613,
"theoretical_loss": 3.472348924838562,
"tokens_seen": 1732182016
},
{
"epoch": 0.05,
"learning_rate": 9.5971754132563e-05,
"loss": 2.5003,
"theoretical_loss": 3.4723263942821694,
"tokens_seen": 1732313088
},
{
"epoch": 0.05,
"learning_rate": 9.596372973840476e-05,
"loss": 2.5422,
"theoretical_loss": 3.4723038659077248,
"tokens_seen": 1732444160
},
{
"epoch": 0.05,
"learning_rate": 9.595570534424651e-05,
"loss": 2.5334,
"theoretical_loss": 3.4722813397148515,
"tokens_seen": 1732575232
},
{
"epoch": 0.05,
"learning_rate": 9.594768095008828e-05,
"loss": 2.4992,
"theoretical_loss": 3.4722588157031744,
"tokens_seen": 1732706304
},
{
"epoch": 0.05,
"learning_rate": 9.593965655593003e-05,
"loss": 2.5993,
"theoretical_loss": 3.4722362938723155,
"tokens_seen": 1732837376
},
{
"epoch": 0.05,
"learning_rate": 9.59316321617718e-05,
"loss": 2.5522,
"theoretical_loss": 3.472213774221901,
"tokens_seen": 1732968448
},
{
"epoch": 0.05,
"learning_rate": 9.592360776761355e-05,
"loss": 2.5903,
"theoretical_loss": 3.472191256751552,
"tokens_seen": 1733099520
},
{
"epoch": 0.05,
"learning_rate": 9.59155833734553e-05,
"loss": 2.55,
"theoretical_loss": 3.4721687414608953,
"tokens_seen": 1733230592
},
{
"epoch": 0.05,
"learning_rate": 9.590755897929707e-05,
"loss": 2.6655,
"theoretical_loss": 3.472146228349554,
"tokens_seen": 1733361664
},
{
"epoch": 0.05,
"learning_rate": 9.589953458513882e-05,
"loss": 2.5992,
"theoretical_loss": 3.472123717417152,
"tokens_seen": 1733492736
},
{
"epoch": 0.05,
"objective/train/docs_used": 956719,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.863294839859009,
"objective/train/theoretical_loss": 3.472112462767936,
"objective/train/tokens_used": 104017376,
"theoretical_loss": 3.472112462767936,
"tokens_seen": 1733558272
},
{
"epoch": 0.05,
"learning_rate": 9.589151019098059e-05,
"loss": 2.592,
"theoretical_loss": 3.4721012086633145,
"tokens_seen": 1733623808
},
{
"epoch": 0.05,
"learning_rate": 9.588348579682234e-05,
"loss": 2.6002,
"theoretical_loss": 3.472078702087665,
"tokens_seen": 1733754880
},
{
"epoch": 0.05,
"learning_rate": 9.587546140266411e-05,
"loss": 2.396,
"theoretical_loss": 3.4720561976898288,
"tokens_seen": 1733885952
},
{
"epoch": 0.05,
"learning_rate": 9.586743700850586e-05,
"loss": 2.5257,
"theoretical_loss": 3.47203369546943,
"tokens_seen": 1734017024
},
{
"epoch": 0.05,
"learning_rate": 9.585941261434761e-05,
"loss": 2.2733,
"theoretical_loss": 3.472011195426095,
"tokens_seen": 1734148096
},
{
"epoch": 0.05,
"learning_rate": 9.585138822018938e-05,
"loss": 2.5829,
"theoretical_loss": 3.4719886975594463,
"tokens_seen": 1734279168
},
{
"epoch": 0.05,
"learning_rate": 9.584336382603113e-05,
"loss": 2.5333,
"theoretical_loss": 3.47196620186911,
"tokens_seen": 1734410240
},
{
"epoch": 0.05,
"learning_rate": 9.58353394318729e-05,
"loss": 2.6064,
"theoretical_loss": 3.4719437083547113,
"tokens_seen": 1734541312
},
{
"epoch": 0.05,
"learning_rate": 9.582731503771465e-05,
"loss": 2.5828,
"theoretical_loss": 3.4719212170158755,
"tokens_seen": 1734672384
},
{
"epoch": 0.05,
"learning_rate": 9.581929064355641e-05,
"loss": 2.5385,
"theoretical_loss": 3.471898727852227,
"tokens_seen": 1734803456
},
{
"epoch": 0.05,
"learning_rate": 9.581126624939817e-05,
"loss": 2.5934,
"theoretical_loss": 3.4718762408633914,
"tokens_seen": 1734934528
},
{
"epoch": 0.05,
"learning_rate": 9.580324185523993e-05,
"loss": 2.6143,
"theoretical_loss": 3.471853756048994,
"tokens_seen": 1735065600
},
{
"epoch": 0.05,
"objective/train/docs_used": 957813,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.456376314163208,
"objective/train/theoretical_loss": 3.471831273408661,
"objective/train/tokens_used": 105655776,
"theoretical_loss": 3.471831273408661,
"tokens_seen": 1735196672
},
{
"epoch": 0.05,
"learning_rate": 9.57952174610817e-05,
"loss": 2.5648,
"theoretical_loss": 3.471831273408661,
"tokens_seen": 1735196672
},
{
"epoch": 0.05,
"learning_rate": 9.578719306692345e-05,
"loss": 2.6123,
"theoretical_loss": 3.471808792942017,
"tokens_seen": 1735327744
},
{
"epoch": 0.05,
"learning_rate": 9.57791686727652e-05,
"loss": 2.5439,
"theoretical_loss": 3.4717863146486887,
"tokens_seen": 1735458816
},
{
"epoch": 0.05,
"learning_rate": 9.577114427860697e-05,
"loss": 2.5435,
"theoretical_loss": 3.471763838528301,
"tokens_seen": 1735589888
},
{
"epoch": 0.05,
"learning_rate": 9.576311988444872e-05,
"loss": 2.6144,
"theoretical_loss": 3.47174136458048,
"tokens_seen": 1735720960
},
{
"epoch": 0.05,
"learning_rate": 9.575509549029049e-05,
"loss": 2.6676,
"theoretical_loss": 3.4717188928048524,
"tokens_seen": 1735852032
},
{
"epoch": 0.05,
"learning_rate": 9.574707109613224e-05,
"loss": 2.6404,
"theoretical_loss": 3.471696423201043,
"tokens_seen": 1735983104
},
{
"epoch": 0.05,
"learning_rate": 9.5739046701974e-05,
"loss": 2.42,
"theoretical_loss": 3.4716739557686793,
"tokens_seen": 1736114176
},
{
"epoch": 0.05,
"learning_rate": 9.573102230781576e-05,
"loss": 2.5927,
"theoretical_loss": 3.4716514905073863,
"tokens_seen": 1736245248
},
{
"epoch": 0.05,
"learning_rate": 9.572299791365751e-05,
"loss": 2.5552,
"theoretical_loss": 3.4716290274167907,
"tokens_seen": 1736376320
},
{
"epoch": 0.05,
"learning_rate": 9.571497351949928e-05,
"loss": 2.5069,
"theoretical_loss": 3.4716065664965194,
"tokens_seen": 1736507392
},
{
"epoch": 0.05,
"learning_rate": 9.570694912534103e-05,
"loss": 2.6358,
"theoretical_loss": 3.4715841077461986,
"tokens_seen": 1736638464
},
{
"epoch": 0.05,
"learning_rate": 9.56989247311828e-05,
"loss": 2.5779,
"theoretical_loss": 3.4715616511654552,
"tokens_seen": 1736769536
},
{
"epoch": 0.05,
"objective/train/docs_used": 958321,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.462451934814453,
"objective/train/theoretical_loss": 3.4715504236885577,
"objective/train/tokens_used": 107294176,
"theoretical_loss": 3.4715504236885577,
"tokens_seen": 1736835072
},
{
"epoch": 0.05,
"learning_rate": 9.569090033702455e-05,
"loss": 2.5436,
"theoretical_loss": 3.4715391967539153,
"tokens_seen": 1736900608
},
{
"epoch": 0.05,
"learning_rate": 9.568287594286632e-05,
"loss": 2.7413,
"theoretical_loss": 3.471516744511206,
"tokens_seen": 1737031680
},
{
"epoch": 0.05,
"learning_rate": 9.567485154870807e-05,
"loss": 2.4977,
"theoretical_loss": 3.471494294436954,
"tokens_seen": 1737162752
},
{
"epoch": 0.05,
"learning_rate": 9.566682715454983e-05,
"loss": 2.4846,
"theoretical_loss": 3.471471846530787,
"tokens_seen": 1737293824
},
{
"epoch": 0.05,
"learning_rate": 9.565880276039159e-05,
"loss": 2.5388,
"theoretical_loss": 3.4714494007923307,
"tokens_seen": 1737424896
},
{
"epoch": 0.05,
"learning_rate": 9.565077836623335e-05,
"loss": 2.6694,
"theoretical_loss": 3.4714269572212135,
"tokens_seen": 1737555968
},
{
"epoch": 0.05,
"learning_rate": 9.564275397207511e-05,
"loss": 2.5805,
"theoretical_loss": 3.4714045158170626,
"tokens_seen": 1737687040
},
{
"epoch": 0.05,
"learning_rate": 9.563472957791687e-05,
"loss": 2.4512,
"theoretical_loss": 3.4713820765795047,
"tokens_seen": 1737818112
},
{
"epoch": 0.05,
"learning_rate": 9.562670518375863e-05,
"loss": 2.4375,
"theoretical_loss": 3.471359639508168,
"tokens_seen": 1737949184
},
{
"epoch": 0.05,
"learning_rate": 9.561868078960038e-05,
"loss": 2.5466,
"theoretical_loss": 3.471337204602679,
"tokens_seen": 1738080256
},
{
"epoch": 0.05,
"learning_rate": 9.561065639544214e-05,
"loss": 2.6156,
"theoretical_loss": 3.471314771862666,
"tokens_seen": 1738211328
},
{
"epoch": 0.05,
"learning_rate": 9.56026320012839e-05,
"loss": 2.7741,
"theoretical_loss": 3.4712923412877563,
"tokens_seen": 1738342400
},
{
"epoch": 0.05,
"objective/train/docs_used": 959398,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.5910706520080566,
"objective/train/theoretical_loss": 3.4712699128775784,
"objective/train/tokens_used": 108932576,
"theoretical_loss": 3.4712699128775784,
"tokens_seen": 1738473472
},
{
"epoch": 0.05,
"learning_rate": 9.559460760712566e-05,
"loss": 2.5787,
"theoretical_loss": 3.4712699128775784,
"tokens_seen": 1738473472
},
{
"epoch": 0.05,
"learning_rate": 9.558658321296742e-05,
"loss": 2.7085,
"theoretical_loss": 3.4712474866317597,
"tokens_seen": 1738604544
},
{
"epoch": 0.05,
"learning_rate": 9.557855881880918e-05,
"loss": 2.5791,
"theoretical_loss": 3.4712250625499284,
"tokens_seen": 1738735616
},
{
"epoch": 0.05,
"learning_rate": 9.557053442465094e-05,
"loss": 2.5154,
"theoretical_loss": 3.471202640631713,
"tokens_seen": 1738866688
},
{
"epoch": 0.05,
"learning_rate": 9.55625100304927e-05,
"loss": 2.6328,
"theoretical_loss": 3.4711802208767404,
"tokens_seen": 1738997760
},
{
"epoch": 0.05,
"learning_rate": 9.555448563633445e-05,
"loss": 2.5779,
"theoretical_loss": 3.4711578032846404,
"tokens_seen": 1739128832
},
{
"epoch": 0.05,
"learning_rate": 9.554646124217622e-05,
"loss": 2.4909,
"theoretical_loss": 3.47113538785504,
"tokens_seen": 1739259904
},
{
"epoch": 0.05,
"learning_rate": 9.553843684801797e-05,
"loss": 2.6613,
"theoretical_loss": 3.471112974587568,
"tokens_seen": 1739390976
},
{
"epoch": 0.05,
"learning_rate": 9.553041245385974e-05,
"loss": 2.647,
"theoretical_loss": 3.471090563481854,
"tokens_seen": 1739522048
},
{
"epoch": 0.05,
"learning_rate": 9.552238805970149e-05,
"loss": 2.5482,
"theoretical_loss": 3.4710681545375253,
"tokens_seen": 1739653120
},
{
"epoch": 0.05,
"learning_rate": 9.551436366554326e-05,
"loss": 2.6415,
"theoretical_loss": 3.471045747754211,
"tokens_seen": 1739784192
},
{
"epoch": 0.05,
"learning_rate": 9.550633927138501e-05,
"loss": 2.5234,
"theoretical_loss": 3.4710233431315407,
"tokens_seen": 1739915264
},
{
"epoch": 0.05,
"learning_rate": 9.549831487722676e-05,
"loss": 2.4932,
"theoretical_loss": 3.471000940669142,
"tokens_seen": 1740046336
},
{
"epoch": 0.05,
"objective/train/docs_used": 960697,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7677383422851562,
"objective/train/theoretical_loss": 3.470989740247929,
"objective/train/tokens_used": 110570976,
"theoretical_loss": 3.470989740247929,
"tokens_seen": 1740111872
},
{
"epoch": 0.05,
"learning_rate": 9.549029048306853e-05,
"loss": 2.5452,
"theoretical_loss": 3.4709785403666453,
"tokens_seen": 1740177408
},
{
"epoch": 0.05,
"learning_rate": 9.548226608891028e-05,
"loss": 2.5708,
"theoretical_loss": 3.4709561422236783,
"tokens_seen": 1740308480
},
{
"epoch": 0.05,
"learning_rate": 9.547424169475205e-05,
"loss": 2.7149,
"theoretical_loss": 3.4709337462398713,
"tokens_seen": 1740439552
},
{
"epoch": 0.05,
"learning_rate": 9.54662173005938e-05,
"loss": 2.5015,
"theoretical_loss": 3.470911352414853,
"tokens_seen": 1740570624
},
{
"epoch": 0.05,
"learning_rate": 9.545819290643557e-05,
"loss": 2.6489,
"theoretical_loss": 3.470888960748253,
"tokens_seen": 1740701696
},
{
"epoch": 0.06,
"learning_rate": 9.545016851227732e-05,
"loss": 2.6192,
"theoretical_loss": 3.4708665712397004,
"tokens_seen": 1740832768
},
{
"epoch": 0.06,
"learning_rate": 9.544214411811908e-05,
"loss": 2.6038,
"theoretical_loss": 3.4708441838888255,
"tokens_seen": 1740963840
},
{
"epoch": 0.06,
"learning_rate": 9.543411972396084e-05,
"loss": 2.6892,
"theoretical_loss": 3.4708217986952574,
"tokens_seen": 1741094912
},
{
"epoch": 0.06,
"learning_rate": 9.54260953298026e-05,
"loss": 2.5735,
"theoretical_loss": 3.4707994156586253,
"tokens_seen": 1741225984
},
{
"epoch": 0.06,
"learning_rate": 9.541807093564436e-05,
"loss": 2.6655,
"theoretical_loss": 3.47077703477856,
"tokens_seen": 1741357056
},
{
"epoch": 0.06,
"learning_rate": 9.541004654148612e-05,
"loss": 2.5846,
"theoretical_loss": 3.470754656054691,
"tokens_seen": 1741488128
},
{
"epoch": 0.06,
"learning_rate": 9.540202214732788e-05,
"loss": 2.539,
"theoretical_loss": 3.470732279486649,
"tokens_seen": 1741619200
},
{
"epoch": 0.06,
"objective/train/docs_used": 961372,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.166609525680542,
"objective/train/theoretical_loss": 3.470709905074062,
"objective/train/tokens_used": 112209376,
"theoretical_loss": 3.470709905074062,
"tokens_seen": 1741750272
},
{
"epoch": 0.06,
"learning_rate": 9.539399775316964e-05,
"loss": 2.6481,
"theoretical_loss": 3.470709905074062,
"tokens_seen": 1741750272
},
{
"epoch": 0.06,
"learning_rate": 9.53859733590114e-05,
"loss": 2.4939,
"theoretical_loss": 3.470687532816563,
"tokens_seen": 1741881344
},
{
"epoch": 0.06,
"learning_rate": 9.537794896485315e-05,
"loss": 2.488,
"theoretical_loss": 3.4706651627137806,
"tokens_seen": 1742012416
},
{
"epoch": 0.06,
"learning_rate": 9.536992457069491e-05,
"loss": 2.5908,
"theoretical_loss": 3.470642794765345,
"tokens_seen": 1742143488
},
{
"epoch": 0.06,
"learning_rate": 9.536190017653667e-05,
"loss": 2.5974,
"theoretical_loss": 3.4706204289708875,
"tokens_seen": 1742274560
},
{
"epoch": 0.06,
"learning_rate": 9.535387578237843e-05,
"loss": 2.5184,
"theoretical_loss": 3.4705980653300386,
"tokens_seen": 1742405632
},
{
"epoch": 0.06,
"learning_rate": 9.53458513882202e-05,
"loss": 2.4658,
"theoretical_loss": 3.4705757038424285,
"tokens_seen": 1742536704
},
{
"epoch": 0.06,
"learning_rate": 9.533782699406195e-05,
"loss": 2.6165,
"theoretical_loss": 3.470553344507688,
"tokens_seen": 1742667776
},
{
"epoch": 0.06,
"learning_rate": 9.532980259990371e-05,
"loss": 2.5697,
"theoretical_loss": 3.470530987325448,
"tokens_seen": 1742798848
},
{
"epoch": 0.06,
"learning_rate": 9.532177820574547e-05,
"loss": 2.7383,
"theoretical_loss": 3.4705086322953393,
"tokens_seen": 1742929920
},
{
"epoch": 0.06,
"learning_rate": 9.531375381158722e-05,
"loss": 2.5499,
"theoretical_loss": 3.4704862794169937,
"tokens_seen": 1743060992
},
{
"epoch": 0.06,
"learning_rate": 9.530572941742899e-05,
"loss": 2.5394,
"theoretical_loss": 3.470463928690041,
"tokens_seen": 1743192064
},
{
"epoch": 0.06,
"learning_rate": 9.529770502327074e-05,
"loss": 2.5673,
"theoretical_loss": 3.4704415801141133,
"tokens_seen": 1743323136
},
{
"epoch": 0.06,
"objective/train/docs_used": 962529,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.988215923309326,
"objective/train/theoretical_loss": 3.4704304066326683,
"objective/train/tokens_used": 113847776,
"theoretical_loss": 3.4704304066326683,
"tokens_seen": 1743388672
},
{
"epoch": 0.06,
"learning_rate": 9.528968062911251e-05,
"loss": 2.7066,
"theoretical_loss": 3.4704192336888413,
"tokens_seen": 1743454208
},
{
"epoch": 0.06,
"learning_rate": 9.528165623495426e-05,
"loss": 2.6361,
"theoretical_loss": 3.4703968894138573,
"tokens_seen": 1743585280
},
{
"epoch": 0.06,
"learning_rate": 9.527363184079603e-05,
"loss": 2.6597,
"theoretical_loss": 3.470374547288792,
"tokens_seen": 1743716352
},
{
"epoch": 0.06,
"learning_rate": 9.526560744663778e-05,
"loss": 2.7227,
"theoretical_loss": 3.4703522073132773,
"tokens_seen": 1743847424
},
{
"epoch": 0.06,
"learning_rate": 9.525758305247953e-05,
"loss": 2.4906,
"theoretical_loss": 3.470329869486944,
"tokens_seen": 1743978496
},
{
"epoch": 0.06,
"learning_rate": 9.52495586583213e-05,
"loss": 2.6045,
"theoretical_loss": 3.4703075338094256,
"tokens_seen": 1744109568
},
{
"epoch": 0.06,
"learning_rate": 9.524153426416305e-05,
"loss": 2.5417,
"theoretical_loss": 3.470285200280352,
"tokens_seen": 1744240640
},
{
"epoch": 0.06,
"learning_rate": 9.523350987000482e-05,
"loss": 2.6117,
"theoretical_loss": 3.470262868899357,
"tokens_seen": 1744371712
},
{
"epoch": 0.06,
"learning_rate": 9.522548547584657e-05,
"loss": 2.6588,
"theoretical_loss": 3.4702405396660705,
"tokens_seen": 1744502784
},
{
"epoch": 0.06,
"learning_rate": 9.521746108168834e-05,
"loss": 2.6923,
"theoretical_loss": 3.470218212580126,
"tokens_seen": 1744633856
},
{
"epoch": 0.06,
"learning_rate": 9.520943668753009e-05,
"loss": 2.7159,
"theoretical_loss": 3.4701958876411556,
"tokens_seen": 1744764928
},
{
"epoch": 0.06,
"learning_rate": 9.520141229337185e-05,
"loss": 2.5938,
"theoretical_loss": 3.4701735648487912,
"tokens_seen": 1744896000
},
{
"epoch": 0.06,
"objective/train/docs_used": 963086,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.850843667984009,
"objective/train/theoretical_loss": 3.4701512442026656,
"objective/train/tokens_used": 115486176,
"theoretical_loss": 3.4701512442026656,
"tokens_seen": 1745027072
},
{
"epoch": 0.06,
"learning_rate": 9.519338789921361e-05,
"loss": 2.575,
"theoretical_loss": 3.4701512442026656,
"tokens_seen": 1745027072
},
{
"epoch": 0.06,
"learning_rate": 9.518536350505537e-05,
"loss": 2.6056,
"theoretical_loss": 3.4701289257024106,
"tokens_seen": 1745158144
},
{
"epoch": 0.06,
"learning_rate": 9.517733911089713e-05,
"loss": 2.5319,
"theoretical_loss": 3.47010660934766,
"tokens_seen": 1745289216
},
{
"epoch": 0.06,
"learning_rate": 9.516931471673889e-05,
"loss": 2.5956,
"theoretical_loss": 3.4700842951380446,
"tokens_seen": 1745420288
},
{
"epoch": 0.06,
"learning_rate": 9.516129032258065e-05,
"loss": 2.5044,
"theoretical_loss": 3.4700619830731982,
"tokens_seen": 1745551360
},
{
"epoch": 0.06,
"learning_rate": 9.51532659284224e-05,
"loss": 2.499,
"theoretical_loss": 3.4700396731527543,
"tokens_seen": 1745682432
},
{
"epoch": 0.06,
"learning_rate": 9.514524153426416e-05,
"loss": 2.6185,
"theoretical_loss": 3.4700173653763446,
"tokens_seen": 1745813504
},
{
"epoch": 0.06,
"learning_rate": 9.513721714010593e-05,
"loss": 2.5509,
"theoretical_loss": 3.469995059743603,
"tokens_seen": 1745944576
},
{
"epoch": 0.06,
"learning_rate": 9.512919274594768e-05,
"loss": 2.5919,
"theoretical_loss": 3.469972756254162,
"tokens_seen": 1746075648
},
{
"epoch": 0.06,
"learning_rate": 9.512116835178944e-05,
"loss": 2.4592,
"theoretical_loss": 3.4699504549076545,
"tokens_seen": 1746206720
},
{
"epoch": 0.06,
"learning_rate": 9.51131439576312e-05,
"loss": 2.6279,
"theoretical_loss": 3.4699281557037147,
"tokens_seen": 1746337792
},
{
"epoch": 0.06,
"learning_rate": 9.510511956347296e-05,
"loss": 2.6709,
"theoretical_loss": 3.4699058586419755,
"tokens_seen": 1746468864
},
{
"epoch": 0.06,
"learning_rate": 9.509709516931472e-05,
"loss": 2.5483,
"theoretical_loss": 3.46988356372207,
"tokens_seen": 1746599936
},
{
"epoch": 0.06,
"objective/train/docs_used": 964151,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.168619394302368,
"objective/train/theoretical_loss": 3.469872417065191,
"objective/train/tokens_used": 117124576,
"theoretical_loss": 3.469872417065191,
"tokens_seen": 1746665472
},
{
"epoch": 0.06,
"learning_rate": 9.508907077515648e-05,
"loss": 2.4492,
"theoretical_loss": 3.4698612709436327,
"tokens_seen": 1746731008
},
{
"epoch": 0.06,
"learning_rate": 9.508104638099824e-05,
"loss": 2.6619,
"theoretical_loss": 3.469838980306297,
"tokens_seen": 1746862080
},
{
"epoch": 0.06,
"learning_rate": 9.507302198683999e-05,
"loss": 2.6721,
"theoretical_loss": 3.4698166918096955,
"tokens_seen": 1746993152
},
{
"epoch": 0.06,
"learning_rate": 9.506499759268176e-05,
"loss": 2.502,
"theoretical_loss": 3.469794405453463,
"tokens_seen": 1747124224
},
{
"epoch": 0.06,
"learning_rate": 9.505697319852351e-05,
"loss": 2.599,
"theoretical_loss": 3.4697721212372334,
"tokens_seen": 1747255296
},
{
"epoch": 0.06,
"learning_rate": 9.504894880436528e-05,
"loss": 2.6716,
"theoretical_loss": 3.4697498391606403,
"tokens_seen": 1747386368
},
{
"epoch": 0.06,
"learning_rate": 9.504092441020703e-05,
"loss": 2.5568,
"theoretical_loss": 3.469727559223318,
"tokens_seen": 1747517440
},
{
"epoch": 0.06,
"learning_rate": 9.50329000160488e-05,
"loss": 2.6531,
"theoretical_loss": 3.469705281424901,
"tokens_seen": 1747648512
},
{
"epoch": 0.06,
"learning_rate": 9.502487562189055e-05,
"loss": 2.653,
"theoretical_loss": 3.469683005765023,
"tokens_seen": 1747779584
},
{
"epoch": 0.06,
"learning_rate": 9.50168512277323e-05,
"loss": 2.547,
"theoretical_loss": 3.469660732243319,
"tokens_seen": 1747910656
},
{
"epoch": 0.06,
"learning_rate": 9.500882683357407e-05,
"loss": 2.5207,
"theoretical_loss": 3.469638460859423,
"tokens_seen": 1748041728
},
{
"epoch": 0.06,
"learning_rate": 9.500080243941582e-05,
"loss": 2.6592,
"theoretical_loss": 3.4696161916129693,
"tokens_seen": 1748172800
},
{
"epoch": 0.06,
"objective/train/docs_used": 964837,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.9922051429748535,
"objective/train/theoretical_loss": 3.469593924503593,
"objective/train/tokens_used": 118762976,
"theoretical_loss": 3.469593924503593,
"tokens_seen": 1748303872
},
{
"epoch": 0.06,
"learning_rate": 9.499277804525759e-05,
"loss": 2.5553,
"theoretical_loss": 3.469593924503593,
"tokens_seen": 1748303872
},
{
"epoch": 0.06,
"learning_rate": 9.498475365109934e-05,
"loss": 2.5655,
"theoretical_loss": 3.4695716595309287,
"tokens_seen": 1748434944
},
{
"epoch": 0.06,
"learning_rate": 9.497672925694111e-05,
"loss": 2.6482,
"theoretical_loss": 3.469549396694611,
"tokens_seen": 1748566016
},
{
"epoch": 0.06,
"learning_rate": 9.496870486278286e-05,
"loss": 2.5774,
"theoretical_loss": 3.469527135994275,
"tokens_seen": 1748697088
},
{
"epoch": 0.06,
"learning_rate": 9.496068046862462e-05,
"loss": 2.4727,
"theoretical_loss": 3.469504877429556,
"tokens_seen": 1748828160
},
{
"epoch": 0.06,
"learning_rate": 9.495265607446638e-05,
"loss": 2.5101,
"theoretical_loss": 3.4694826210000884,
"tokens_seen": 1748959232
},
{
"epoch": 0.06,
"learning_rate": 9.494463168030814e-05,
"loss": 2.6768,
"theoretical_loss": 3.4694603667055075,
"tokens_seen": 1749090304
},
{
"epoch": 0.06,
"learning_rate": 9.49366072861499e-05,
"loss": 2.553,
"theoretical_loss": 3.4694381145454494,
"tokens_seen": 1749221376
},
{
"epoch": 0.06,
"learning_rate": 9.492858289199166e-05,
"loss": 2.4986,
"theoretical_loss": 3.469415864519548,
"tokens_seen": 1749352448
},
{
"epoch": 0.06,
"learning_rate": 9.492055849783342e-05,
"loss": 2.66,
"theoretical_loss": 3.469393616627441,
"tokens_seen": 1749483520
},
{
"epoch": 0.06,
"learning_rate": 9.491253410367518e-05,
"loss": 2.5936,
"theoretical_loss": 3.4693713708687612,
"tokens_seen": 1749614592
},
{
"epoch": 0.06,
"learning_rate": 9.490450970951693e-05,
"loss": 2.5735,
"theoretical_loss": 3.4693491272431456,
"tokens_seen": 1749745664
},
{
"epoch": 0.06,
"learning_rate": 9.48964853153587e-05,
"loss": 2.6236,
"theoretical_loss": 3.46932688575023,
"tokens_seen": 1749876736
},
{
"epoch": 0.06,
"objective/train/docs_used": 965967,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.4741082191467285,
"objective/train/theoretical_loss": 3.4693157658034206,
"objective/train/tokens_used": 120401376,
"theoretical_loss": 3.4693157658034206,
"tokens_seen": 1749942272
},
{
"epoch": 0.06,
"learning_rate": 9.488846092120045e-05,
"loss": 2.6312,
"theoretical_loss": 3.46930464638965,
"tokens_seen": 1750007808
},
{
"epoch": 0.06,
"learning_rate": 9.488043652704221e-05,
"loss": 2.6914,
"theoretical_loss": 3.4692824091610412,
"tokens_seen": 1750138880
},
{
"epoch": 0.06,
"learning_rate": 9.487241213288397e-05,
"loss": 2.4969,
"theoretical_loss": 3.4692601740640403,
"tokens_seen": 1750269952
},
{
"epoch": 0.06,
"learning_rate": 9.486438773872573e-05,
"loss": 2.568,
"theoretical_loss": 3.4692379410982825,
"tokens_seen": 1750401024
},
{
"epoch": 0.06,
"learning_rate": 9.485636334456749e-05,
"loss": 2.5101,
"theoretical_loss": 3.4692157102634047,
"tokens_seen": 1750532096
},
{
"epoch": 0.06,
"learning_rate": 9.484833895040924e-05,
"loss": 2.6857,
"theoretical_loss": 3.4691934815590426,
"tokens_seen": 1750663168
},
{
"epoch": 0.06,
"learning_rate": 9.484031455625101e-05,
"loss": 2.4291,
"theoretical_loss": 3.4691712549848326,
"tokens_seen": 1750794240
},
{
"epoch": 0.06,
"learning_rate": 9.483229016209276e-05,
"loss": 2.6526,
"theoretical_loss": 3.469149030540412,
"tokens_seen": 1750925312
},
{
"epoch": 0.06,
"learning_rate": 9.482426576793453e-05,
"loss": 2.5627,
"theoretical_loss": 3.4691268082254156,
"tokens_seen": 1751056384
},
{
"epoch": 0.06,
"learning_rate": 9.481624137377628e-05,
"loss": 2.5353,
"theoretical_loss": 3.4691045880394817,
"tokens_seen": 1751187456
},
{
"epoch": 0.06,
"learning_rate": 9.480821697961805e-05,
"loss": 2.5783,
"theoretical_loss": 3.4690823699822455,
"tokens_seen": 1751318528
},
{
"epoch": 0.06,
"learning_rate": 9.48001925854598e-05,
"loss": 2.6184,
"theoretical_loss": 3.469060154053345,
"tokens_seen": 1751449600
},
{
"epoch": 0.06,
"objective/train/docs_used": 966884,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.789921760559082,
"objective/train/theoretical_loss": 3.4690379402524165,
"objective/train/tokens_used": 122039776,
"theoretical_loss": 3.4690379402524165,
"tokens_seen": 1751580672
},
{
"epoch": 0.06,
"learning_rate": 9.479216819130155e-05,
"loss": 2.5307,
"theoretical_loss": 3.4690379402524165,
"tokens_seen": 1751580672
},
{
"epoch": 0.06,
"learning_rate": 9.478414379714332e-05,
"loss": 2.5257,
"theoretical_loss": 3.4690157285790972,
"tokens_seen": 1751711744
},
{
"epoch": 0.06,
"learning_rate": 9.477611940298507e-05,
"loss": 2.561,
"theoretical_loss": 3.468993519033024,
"tokens_seen": 1751842816
},
{
"epoch": 0.06,
"learning_rate": 9.476809500882684e-05,
"loss": 2.3812,
"theoretical_loss": 3.468971311613834,
"tokens_seen": 1751973888
},
{
"epoch": 0.06,
"learning_rate": 9.476007061466859e-05,
"loss": 2.5491,
"theoretical_loss": 3.468949106321164,
"tokens_seen": 1752104960
},
{
"epoch": 0.06,
"learning_rate": 9.475204622051036e-05,
"loss": 2.5825,
"theoretical_loss": 3.468926903154652,
"tokens_seen": 1752236032
},
{
"epoch": 0.06,
"learning_rate": 9.474402182635211e-05,
"loss": 2.608,
"theoretical_loss": 3.4689047021139356,
"tokens_seen": 1752367104
},
{
"epoch": 0.06,
"learning_rate": 9.473599743219388e-05,
"loss": 2.6951,
"theoretical_loss": 3.4688825031986514,
"tokens_seen": 1752498176
},
{
"epoch": 0.06,
"learning_rate": 9.472797303803563e-05,
"loss": 2.5484,
"theoretical_loss": 3.4688603064084376,
"tokens_seen": 1752629248
},
{
"epoch": 0.06,
"learning_rate": 9.471994864387739e-05,
"loss": 2.7742,
"theoretical_loss": 3.4688381117429317,
"tokens_seen": 1752760320
},
{
"epoch": 0.06,
"learning_rate": 9.471192424971915e-05,
"loss": 2.5422,
"theoretical_loss": 3.4688159192017713,
"tokens_seen": 1752891392
},
{
"epoch": 0.06,
"learning_rate": 9.47038998555609e-05,
"loss": 2.637,
"theoretical_loss": 3.4687937287845947,
"tokens_seen": 1753022464
},
{
"epoch": 0.06,
"learning_rate": 9.469587546140267e-05,
"loss": 2.5073,
"theoretical_loss": 3.4687715404910398,
"tokens_seen": 1753153536
},
{
"epoch": 0.06,
"objective/train/docs_used": 967417,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.34035325050354,
"objective/train/theoretical_loss": 3.4687604471405065,
"objective/train/tokens_used": 123678176,
"theoretical_loss": 3.4687604471405065,
"tokens_seen": 1753219072
},
{
"epoch": 0.06,
"learning_rate": 9.468785106724443e-05,
"loss": 2.5376,
"theoretical_loss": 3.4687493543207433,
"tokens_seen": 1753284608
},
{
"epoch": 0.06,
"learning_rate": 9.467982667308619e-05,
"loss": 2.6251,
"theoretical_loss": 3.468727170273345,
"tokens_seen": 1753415680
},
{
"epoch": 0.06,
"learning_rate": 9.467180227892795e-05,
"loss": 2.8049,
"theoretical_loss": 3.468704988348483,
"tokens_seen": 1753546752
},
{
"epoch": 0.06,
"learning_rate": 9.46637778847697e-05,
"loss": 2.5002,
"theoretical_loss": 3.468682808545794,
"tokens_seen": 1753677824
},
{
"epoch": 0.06,
"learning_rate": 9.465575349061147e-05,
"loss": 2.5396,
"theoretical_loss": 3.4686606308649184,
"tokens_seen": 1753808896
},
{
"epoch": 0.06,
"learning_rate": 9.464772909645322e-05,
"loss": 2.5852,
"theoretical_loss": 3.468638455305493,
"tokens_seen": 1753939968
},
{
"epoch": 0.06,
"learning_rate": 9.463970470229499e-05,
"loss": 2.5664,
"theoretical_loss": 3.468616281867157,
"tokens_seen": 1754071040
},
{
"epoch": 0.06,
"learning_rate": 9.463168030813674e-05,
"loss": 2.6855,
"theoretical_loss": 3.46859411054955,
"tokens_seen": 1754202112
},
{
"epoch": 0.06,
"learning_rate": 9.46236559139785e-05,
"loss": 2.6069,
"theoretical_loss": 3.468571941352309,
"tokens_seen": 1754333184
},
{
"epoch": 0.06,
"learning_rate": 9.461563151982026e-05,
"loss": 2.5943,
"theoretical_loss": 3.4685497742750737,
"tokens_seen": 1754464256
},
{
"epoch": 0.06,
"learning_rate": 9.460760712566201e-05,
"loss": 2.5968,
"theoretical_loss": 3.4685276093174835,
"tokens_seen": 1754595328
},
{
"epoch": 0.06,
"learning_rate": 9.459958273150378e-05,
"loss": 2.4725,
"theoretical_loss": 3.4685054464791767,
"tokens_seen": 1754726400
},
{
"epoch": 0.06,
"objective/train/docs_used": 968410,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6490659713745117,
"objective/train/theoretical_loss": 3.4684832857597923,
"objective/train/tokens_used": 125316576,
"theoretical_loss": 3.4684832857597923,
"tokens_seen": 1754857472
},
{
"epoch": 0.06,
"learning_rate": 9.459155833734553e-05,
"loss": 2.5628,
"theoretical_loss": 3.4684832857597923,
"tokens_seen": 1754857472
},
{
"epoch": 0.06,
"learning_rate": 9.45835339431873e-05,
"loss": 2.5849,
"theoretical_loss": 3.46846112715897,
"tokens_seen": 1754988544
},
{
"epoch": 0.06,
"learning_rate": 9.457550954902905e-05,
"loss": 2.6675,
"theoretical_loss": 3.468438970676348,
"tokens_seen": 1755119616
},
{
"epoch": 0.06,
"learning_rate": 9.456748515487082e-05,
"loss": 2.5566,
"theoretical_loss": 3.468416816311567,
"tokens_seen": 1755250688
},
{
"epoch": 0.06,
"learning_rate": 9.455946076071257e-05,
"loss": 2.4097,
"theoretical_loss": 3.468394664064266,
"tokens_seen": 1755381760
},
{
"epoch": 0.06,
"learning_rate": 9.455143636655432e-05,
"loss": 2.6195,
"theoretical_loss": 3.468372513934084,
"tokens_seen": 1755512832
},
{
"epoch": 0.06,
"learning_rate": 9.454341197239609e-05,
"loss": 2.5105,
"theoretical_loss": 3.468350365920662,
"tokens_seen": 1755643904
},
{
"epoch": 0.06,
"learning_rate": 9.453538757823784e-05,
"loss": 2.5745,
"theoretical_loss": 3.468328220023638,
"tokens_seen": 1755774976
},
{
"epoch": 0.06,
"learning_rate": 9.452736318407961e-05,
"loss": 2.5794,
"theoretical_loss": 3.4683060762426523,
"tokens_seen": 1755906048
},
{
"epoch": 0.06,
"learning_rate": 9.451933878992136e-05,
"loss": 2.563,
"theoretical_loss": 3.4682839345773457,
"tokens_seen": 1756037120
},
{
"epoch": 0.06,
"learning_rate": 9.451131439576313e-05,
"loss": 2.5961,
"theoretical_loss": 3.468261795027357,
"tokens_seen": 1756168192
},
{
"epoch": 0.06,
"learning_rate": 9.450329000160488e-05,
"loss": 2.6463,
"theoretical_loss": 3.4682396575923264,
"tokens_seen": 1756299264
},
{
"epoch": 0.06,
"learning_rate": 9.449526560744664e-05,
"loss": 2.5798,
"theoretical_loss": 3.468217522271895,
"tokens_seen": 1756430336
},
{
"epoch": 0.06,
"objective/train/docs_used": 968954,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.701629638671875,
"objective/train/theoretical_loss": 3.468206455404541,
"objective/train/tokens_used": 126954976,
"theoretical_loss": 3.468206455404541,
"tokens_seen": 1756495872
},
{
"epoch": 0.06,
"learning_rate": 9.44872412132884e-05,
"loss": 2.4543,
"theoretical_loss": 3.468195389065702,
"tokens_seen": 1756561408
},
{
"epoch": 0.06,
"learning_rate": 9.447921681913016e-05,
"loss": 2.5349,
"theoretical_loss": 3.4681732579733886,
"tokens_seen": 1756692480
},
{
"epoch": 0.06,
"learning_rate": 9.447119242497192e-05,
"loss": 2.5532,
"theoretical_loss": 3.4681511289945943,
"tokens_seen": 1756823552
},
{
"epoch": 0.06,
"learning_rate": 9.446316803081368e-05,
"loss": 2.5709,
"theoretical_loss": 3.46812900212896,
"tokens_seen": 1756954624
},
{
"epoch": 0.06,
"learning_rate": 9.445514363665544e-05,
"loss": 2.5352,
"theoretical_loss": 3.468106877376126,
"tokens_seen": 1757085696
},
{
"epoch": 0.06,
"learning_rate": 9.44471192424972e-05,
"loss": 2.6435,
"theoretical_loss": 3.468084754735734,
"tokens_seen": 1757216768
},
{
"epoch": 0.07,
"learning_rate": 9.443909484833896e-05,
"loss": 2.6461,
"theoretical_loss": 3.468062634207424,
"tokens_seen": 1757347840
},
{
"epoch": 0.07,
"learning_rate": 9.443107045418072e-05,
"loss": 2.655,
"theoretical_loss": 3.4680405157908365,
"tokens_seen": 1757478912
},
{
"epoch": 0.07,
"learning_rate": 9.442304606002247e-05,
"loss": 2.6677,
"theoretical_loss": 3.4680183994856133,
"tokens_seen": 1757609984
},
{
"epoch": 0.07,
"learning_rate": 9.441502166586424e-05,
"loss": 2.6228,
"theoretical_loss": 3.4679962852913944,
"tokens_seen": 1757741056
},
{
"epoch": 0.07,
"learning_rate": 9.440699727170599e-05,
"loss": 2.6249,
"theoretical_loss": 3.4679741732078213,
"tokens_seen": 1757872128
},
{
"epoch": 0.07,
"learning_rate": 9.439897287754776e-05,
"loss": 2.5379,
"theoretical_loss": 3.467952063234536,
"tokens_seen": 1758003200
},
{
"epoch": 0.07,
"objective/train/docs_used": 970319,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6145591735839844,
"objective/train/theoretical_loss": 3.4679299553711793,
"objective/train/tokens_used": 128593376,
"theoretical_loss": 3.4679299553711793,
"tokens_seen": 1758134272
},
{
"epoch": 0.07,
"learning_rate": 9.439094848338951e-05,
"loss": 2.6071,
"theoretical_loss": 3.4679299553711793,
"tokens_seen": 1758134272
},
{
"epoch": 0.07,
"learning_rate": 9.438292408923127e-05,
"loss": 2.4915,
"theoretical_loss": 3.467907849617392,
"tokens_seen": 1758265344
},
{
"epoch": 0.07,
"learning_rate": 9.437489969507303e-05,
"loss": 2.4659,
"theoretical_loss": 3.467885745972816,
"tokens_seen": 1758396416
},
{
"epoch": 0.07,
"learning_rate": 9.436687530091478e-05,
"loss": 2.5481,
"theoretical_loss": 3.467863644437093,
"tokens_seen": 1758527488
},
{
"epoch": 0.07,
"learning_rate": 9.435885090675655e-05,
"loss": 2.6001,
"theoretical_loss": 3.4678415450098643,
"tokens_seen": 1758658560
},
{
"epoch": 0.07,
"learning_rate": 9.43508265125983e-05,
"loss": 2.5004,
"theoretical_loss": 3.467819447690772,
"tokens_seen": 1758789632
},
{
"epoch": 0.07,
"learning_rate": 9.434280211844007e-05,
"loss": 2.6864,
"theoretical_loss": 3.467797352479458,
"tokens_seen": 1758920704
},
{
"epoch": 0.07,
"learning_rate": 9.433477772428182e-05,
"loss": 2.6084,
"theoretical_loss": 3.4677752593755633,
"tokens_seen": 1759051776
},
{
"epoch": 0.07,
"learning_rate": 9.432675333012359e-05,
"loss": 2.553,
"theoretical_loss": 3.4677531683787306,
"tokens_seen": 1759182848
},
{
"epoch": 0.07,
"learning_rate": 9.431872893596534e-05,
"loss": 2.6378,
"theoretical_loss": 3.4677310794886025,
"tokens_seen": 1759313920
},
{
"epoch": 0.07,
"learning_rate": 9.43107045418071e-05,
"loss": 2.38,
"theoretical_loss": 3.46770899270482,
"tokens_seen": 1759444992
},
{
"epoch": 0.07,
"learning_rate": 9.430268014764886e-05,
"loss": 2.6261,
"theoretical_loss": 3.467686908027026,
"tokens_seen": 1759576064
},
{
"epoch": 0.07,
"learning_rate": 9.429465575349061e-05,
"loss": 2.6348,
"theoretical_loss": 3.4676648254548628,
"tokens_seen": 1759707136
},
{
"epoch": 0.07,
"objective/train/docs_used": 970997,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7837326526641846,
"objective/train/theoretical_loss": 3.467653784958281,
"objective/train/tokens_used": 130231776,
"theoretical_loss": 3.467653784958281,
"tokens_seen": 1759772672
},
{
"epoch": 0.07,
"learning_rate": 9.428663135933238e-05,
"loss": 2.6637,
"theoretical_loss": 3.467642744987973,
"tokens_seen": 1759838208
},
{
"epoch": 0.07,
"learning_rate": 9.427860696517413e-05,
"loss": 2.6002,
"theoretical_loss": 3.467620666625999,
"tokens_seen": 1759969280
},
{
"epoch": 0.07,
"learning_rate": 9.42705825710159e-05,
"loss": 2.3965,
"theoretical_loss": 3.4675985903685826,
"tokens_seen": 1760100352
},
{
"epoch": 0.07,
"learning_rate": 9.426255817685765e-05,
"loss": 2.4942,
"theoretical_loss": 3.4675765162153676,
"tokens_seen": 1760231424
},
{
"epoch": 0.07,
"learning_rate": 9.42545337826994e-05,
"loss": 2.6354,
"theoretical_loss": 3.4675544441659962,
"tokens_seen": 1760362496
},
{
"epoch": 0.07,
"learning_rate": 9.424650938854117e-05,
"loss": 2.5839,
"theoretical_loss": 3.467532374220112,
"tokens_seen": 1760493568
},
{
"epoch": 0.07,
"learning_rate": 9.423848499438293e-05,
"loss": 2.5664,
"theoretical_loss": 3.4675103063773567,
"tokens_seen": 1760624640
},
{
"epoch": 0.07,
"learning_rate": 9.423046060022469e-05,
"loss": 2.6749,
"theoretical_loss": 3.4674882406373744,
"tokens_seen": 1760755712
},
{
"epoch": 0.07,
"learning_rate": 9.422243620606645e-05,
"loss": 2.5527,
"theoretical_loss": 3.4674661769998076,
"tokens_seen": 1760886784
},
{
"epoch": 0.07,
"learning_rate": 9.421441181190821e-05,
"loss": 2.4963,
"theoretical_loss": 3.4674441154643,
"tokens_seen": 1761017856
},
{
"epoch": 0.07,
"learning_rate": 9.420638741774997e-05,
"loss": 2.5026,
"theoretical_loss": 3.467422056030494,
"tokens_seen": 1761148928
},
{
"epoch": 0.07,
"learning_rate": 9.419836302359172e-05,
"loss": 2.546,
"theoretical_loss": 3.467399998698034,
"tokens_seen": 1761280000
},
{
"epoch": 0.07,
"objective/train/docs_used": 972106,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7436439990997314,
"objective/train/theoretical_loss": 3.4673779434665635,
"objective/train/tokens_used": 131870176,
"theoretical_loss": 3.4673779434665635,
"tokens_seen": 1761411072
},
{
"epoch": 0.07,
"learning_rate": 9.419033862943349e-05,
"loss": 2.5274,
"theoretical_loss": 3.4673779434665635,
"tokens_seen": 1761411072
},
{
"epoch": 0.07,
"learning_rate": 9.418231423527524e-05,
"loss": 2.7645,
"theoretical_loss": 3.4673558903357247,
"tokens_seen": 1761542144
},
{
"epoch": 0.07,
"learning_rate": 9.4174289841117e-05,
"loss": 2.498,
"theoretical_loss": 3.467333839305163,
"tokens_seen": 1761673216
},
{
"epoch": 0.07,
"learning_rate": 9.416626544695876e-05,
"loss": 2.6431,
"theoretical_loss": 3.467311790374521,
"tokens_seen": 1761804288
},
{
"epoch": 0.07,
"learning_rate": 9.415824105280053e-05,
"loss": 2.6502,
"theoretical_loss": 3.467289743543443,
"tokens_seen": 1761935360
},
{
"epoch": 0.07,
"learning_rate": 9.415021665864228e-05,
"loss": 2.5922,
"theoretical_loss": 3.4672676988115727,
"tokens_seen": 1762066432
},
{
"epoch": 0.07,
"learning_rate": 9.414219226448404e-05,
"loss": 2.5547,
"theoretical_loss": 3.4672456561785543,
"tokens_seen": 1762197504
},
{
"epoch": 0.07,
"learning_rate": 9.41341678703258e-05,
"loss": 2.5285,
"theoretical_loss": 3.4672236156440315,
"tokens_seen": 1762328576
},
{
"epoch": 0.07,
"learning_rate": 9.412614347616755e-05,
"loss": 2.7133,
"theoretical_loss": 3.467201577207649,
"tokens_seen": 1762459648
},
{
"epoch": 0.07,
"learning_rate": 9.411811908200932e-05,
"loss": 2.516,
"theoretical_loss": 3.4671795408690507,
"tokens_seen": 1762590720
},
{
"epoch": 0.07,
"learning_rate": 9.411009468785107e-05,
"loss": 2.4072,
"theoretical_loss": 3.467157506627881,
"tokens_seen": 1762721792
},
{
"epoch": 0.07,
"learning_rate": 9.410207029369284e-05,
"loss": 2.3683,
"theoretical_loss": 3.4671354744837837,
"tokens_seen": 1762852864
},
{
"epoch": 0.07,
"learning_rate": 9.409404589953459e-05,
"loss": 2.6088,
"theoretical_loss": 3.4671134444364045,
"tokens_seen": 1762983936
},
{
"epoch": 0.07,
"objective/train/docs_used": 972675,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.4672791957855225,
"objective/train/theoretical_loss": 3.467102430198873,
"objective/train/tokens_used": 133508576,
"theoretical_loss": 3.467102430198873,
"tokens_seen": 1763049472
},
{
"epoch": 0.07,
"learning_rate": 9.408602150537636e-05,
"loss": 2.4902,
"theoretical_loss": 3.4670914164853874,
"tokens_seen": 1763115008
},
{
"epoch": 0.07,
"learning_rate": 9.407799711121811e-05,
"loss": 2.5745,
"theoretical_loss": 3.4670693906303773,
"tokens_seen": 1763246080
},
{
"epoch": 0.07,
"learning_rate": 9.406997271705986e-05,
"loss": 2.6415,
"theoretical_loss": 3.4670473668710184,
"tokens_seen": 1763377152
},
{
"epoch": 0.07,
"learning_rate": 9.406194832290163e-05,
"loss": 2.6236,
"theoretical_loss": 3.467025345206956,
"tokens_seen": 1763508224
},
{
"epoch": 0.07,
"learning_rate": 9.405392392874338e-05,
"loss": 2.5835,
"theoretical_loss": 3.4670033256378354,
"tokens_seen": 1763639296
},
{
"epoch": 0.07,
"learning_rate": 9.404589953458515e-05,
"loss": 2.5896,
"theoretical_loss": 3.4669813081633007,
"tokens_seen": 1763770368
},
{
"epoch": 0.07,
"learning_rate": 9.40378751404269e-05,
"loss": 2.4005,
"theoretical_loss": 3.466959292782998,
"tokens_seen": 1763901440
},
{
"epoch": 0.07,
"learning_rate": 9.402985074626867e-05,
"loss": 2.6698,
"theoretical_loss": 3.4669372794965723,
"tokens_seen": 1764032512
},
{
"epoch": 0.07,
"learning_rate": 9.402182635211042e-05,
"loss": 2.4925,
"theoretical_loss": 3.4669152683036684,
"tokens_seen": 1764163584
},
{
"epoch": 0.07,
"learning_rate": 9.401380195795218e-05,
"loss": 2.4732,
"theoretical_loss": 3.466893259203932,
"tokens_seen": 1764294656
},
{
"epoch": 0.07,
"learning_rate": 9.400577756379394e-05,
"loss": 2.4619,
"theoretical_loss": 3.4668712521970084,
"tokens_seen": 1764425728
},
{
"epoch": 0.07,
"learning_rate": 9.39977531696357e-05,
"loss": 2.4386,
"theoretical_loss": 3.466849247282543,
"tokens_seen": 1764556800
},
{
"epoch": 0.07,
"objective/train/docs_used": 973758,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.705296039581299,
"objective/train/theoretical_loss": 3.466827244460182,
"objective/train/tokens_used": 135146976,
"theoretical_loss": 3.466827244460182,
"tokens_seen": 1764687872
},
{
"epoch": 0.07,
"learning_rate": 9.398972877547746e-05,
"loss": 2.6076,
"theoretical_loss": 3.466827244460182,
"tokens_seen": 1764687872
},
{
"epoch": 0.07,
"learning_rate": 9.398170438131922e-05,
"loss": 2.5514,
"theoretical_loss": 3.466805243729571,
"tokens_seen": 1764818944
},
{
"epoch": 0.07,
"learning_rate": 9.397367998716098e-05,
"loss": 2.4123,
"theoretical_loss": 3.4667832450903555,
"tokens_seen": 1764950016
},
{
"epoch": 0.07,
"learning_rate": 9.396565559300274e-05,
"loss": 2.5775,
"theoretical_loss": 3.4667612485421815,
"tokens_seen": 1765081088
},
{
"epoch": 0.07,
"learning_rate": 9.395763119884449e-05,
"loss": 2.443,
"theoretical_loss": 3.466739254084695,
"tokens_seen": 1765212160
},
{
"epoch": 0.07,
"learning_rate": 9.394960680468626e-05,
"loss": 2.436,
"theoretical_loss": 3.4667172617175424,
"tokens_seen": 1765343232
},
{
"epoch": 0.07,
"learning_rate": 9.394158241052801e-05,
"loss": 2.5424,
"theoretical_loss": 3.4666952714403694,
"tokens_seen": 1765474304
},
{
"epoch": 0.07,
"learning_rate": 9.393355801636978e-05,
"loss": 2.4734,
"theoretical_loss": 3.466673283252822,
"tokens_seen": 1765605376
},
{
"epoch": 0.07,
"learning_rate": 9.392553362221153e-05,
"loss": 2.5327,
"theoretical_loss": 3.4666512971545473,
"tokens_seen": 1765736448
},
{
"epoch": 0.07,
"learning_rate": 9.39175092280533e-05,
"loss": 2.5953,
"theoretical_loss": 3.4666293131451917,
"tokens_seen": 1765867520
},
{
"epoch": 0.07,
"learning_rate": 9.390948483389505e-05,
"loss": 2.5874,
"theoretical_loss": 3.4666073312244006,
"tokens_seen": 1765998592
},
{
"epoch": 0.07,
"learning_rate": 9.39014604397368e-05,
"loss": 2.4698,
"theoretical_loss": 3.4665853513918217,
"tokens_seen": 1766129664
},
{
"epoch": 0.07,
"learning_rate": 9.389343604557857e-05,
"loss": 2.4093,
"theoretical_loss": 3.4665633736471015,
"tokens_seen": 1766260736
},
{
"epoch": 0.07,
"objective/train/docs_used": 974221,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.551424026489258,
"objective/train/theoretical_loss": 3.466552385557578,
"objective/train/tokens_used": 136785376,
"theoretical_loss": 3.466552385557578,
"tokens_seen": 1766326272
},
{
"epoch": 0.07,
"learning_rate": 9.388541165142032e-05,
"loss": 2.5791,
"theoretical_loss": 3.4665413979898863,
"tokens_seen": 1766391808
},
{
"epoch": 0.07,
"learning_rate": 9.387738725726209e-05,
"loss": 2.4563,
"theoretical_loss": 3.466519424419823,
"tokens_seen": 1766522880
},
{
"epoch": 0.07,
"learning_rate": 9.386936286310384e-05,
"loss": 2.4674,
"theoretical_loss": 3.4664974529365593,
"tokens_seen": 1766653952
},
{
"epoch": 0.07,
"learning_rate": 9.386133846894561e-05,
"loss": 2.6283,
"theoretical_loss": 3.4664754835397416,
"tokens_seen": 1766785024
},
{
"epoch": 0.07,
"learning_rate": 9.385331407478736e-05,
"loss": 2.4859,
"theoretical_loss": 3.4664535162290164,
"tokens_seen": 1766916096
},
{
"epoch": 0.07,
"learning_rate": 9.384528968062911e-05,
"loss": 2.4277,
"theoretical_loss": 3.466431551004032,
"tokens_seen": 1767047168
},
{
"epoch": 0.07,
"learning_rate": 9.383726528647088e-05,
"loss": 2.4934,
"theoretical_loss": 3.466409587864436,
"tokens_seen": 1767178240
},
{
"epoch": 0.07,
"learning_rate": 9.382924089231263e-05,
"loss": 2.476,
"theoretical_loss": 3.4663876268098734,
"tokens_seen": 1767309312
},
{
"epoch": 0.07,
"learning_rate": 9.38212164981544e-05,
"loss": 2.6492,
"theoretical_loss": 3.4663656678399946,
"tokens_seen": 1767440384
},
{
"epoch": 0.07,
"learning_rate": 9.381319210399615e-05,
"loss": 2.4337,
"theoretical_loss": 3.4663437109544453,
"tokens_seen": 1767571456
},
{
"epoch": 0.07,
"learning_rate": 9.380516770983792e-05,
"loss": 2.5528,
"theoretical_loss": 3.4663217561528734,
"tokens_seen": 1767702528
},
{
"epoch": 0.07,
"learning_rate": 9.379714331567967e-05,
"loss": 2.5765,
"theoretical_loss": 3.466299803434927,
"tokens_seen": 1767833600
},
{
"epoch": 0.07,
"objective/train/docs_used": 975253,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 3.134099006652832,
"objective/train/theoretical_loss": 3.4662778528002534,
"objective/train/tokens_used": 138423776,
"theoretical_loss": 3.4662778528002534,
"tokens_seen": 1767964672
},
{
"epoch": 0.07,
"learning_rate": 9.378911892152144e-05,
"loss": 2.5428,
"theoretical_loss": 3.4662778528002534,
"tokens_seen": 1767964672
},
{
"epoch": 0.07,
"learning_rate": 9.37810945273632e-05,
"loss": 2.4819,
"theoretical_loss": 3.4662559042485013,
"tokens_seen": 1768095744
},
{
"epoch": 0.07,
"learning_rate": 9.377307013320495e-05,
"loss": 2.6061,
"theoretical_loss": 3.466233957779318,
"tokens_seen": 1768226816
},
{
"epoch": 0.07,
"learning_rate": 9.376504573904671e-05,
"loss": 2.6773,
"theoretical_loss": 3.466212013392351,
"tokens_seen": 1768357888
},
{
"epoch": 0.07,
"learning_rate": 9.375702134488847e-05,
"loss": 2.5121,
"theoretical_loss": 3.46619007108725,
"tokens_seen": 1768488960
},
{
"epoch": 0.07,
"learning_rate": 9.374899695073023e-05,
"loss": 2.5764,
"theoretical_loss": 3.466168130863662,
"tokens_seen": 1768620032
},
{
"epoch": 0.07,
"learning_rate": 9.374097255657199e-05,
"loss": 2.594,
"theoretical_loss": 3.466146192721235,
"tokens_seen": 1768751104
},
{
"epoch": 0.07,
"learning_rate": 9.373294816241375e-05,
"loss": 2.6474,
"theoretical_loss": 3.4661242566596187,
"tokens_seen": 1768882176
},
{
"epoch": 0.07,
"learning_rate": 9.37249237682555e-05,
"loss": 2.6483,
"theoretical_loss": 3.46610232267846,
"tokens_seen": 1769013248
},
{
"epoch": 0.07,
"learning_rate": 9.371689937409726e-05,
"loss": 2.4411,
"theoretical_loss": 3.4660803907774094,
"tokens_seen": 1769144320
},
{
"epoch": 0.07,
"learning_rate": 9.370887497993903e-05,
"loss": 2.5386,
"theoretical_loss": 3.4660584609561145,
"tokens_seen": 1769275392
},
{
"epoch": 0.07,
"learning_rate": 9.370085058578078e-05,
"loss": 2.427,
"theoretical_loss": 3.4660365332142233,
"tokens_seen": 1769406464
},
{
"epoch": 0.07,
"learning_rate": 9.369282619162255e-05,
"loss": 2.6418,
"theoretical_loss": 3.4660146075513856,
"tokens_seen": 1769537536
},
{
"epoch": 0.07,
"objective/train/docs_used": 975853,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.6023433208465576,
"objective/train/theoretical_loss": 3.466003645499502,
"objective/train/tokens_used": 140062176,
"theoretical_loss": 3.466003645499502,
"tokens_seen": 1769603072
},
{
"epoch": 0.07,
"learning_rate": 9.36848017974643e-05,
"loss": 2.4855,
"theoretical_loss": 3.46599268396725,
"tokens_seen": 1769668608
},
{
"epoch": 0.07,
"learning_rate": 9.367677740330607e-05,
"loss": 2.5534,
"theoretical_loss": 3.465970762461466,
"tokens_seen": 1769799680
},
{
"epoch": 0.07,
"learning_rate": 9.366875300914782e-05,
"loss": 2.7021,
"theoretical_loss": 3.4659488430336816,
"tokens_seen": 1769930752
},
{
"epoch": 0.07,
"learning_rate": 9.366072861498957e-05,
"loss": 2.5897,
"theoretical_loss": 3.4659269256835468,
"tokens_seen": 1770061824
},
{
"epoch": 0.07,
"learning_rate": 9.365270422083134e-05,
"loss": 2.5387,
"theoretical_loss": 3.4659050104107103,
"tokens_seen": 1770192896
},
{
"epoch": 0.07,
"learning_rate": 9.364467982667309e-05,
"loss": 2.4854,
"theoretical_loss": 3.465883097214822,
"tokens_seen": 1770323968
},
{
"epoch": 0.07,
"learning_rate": 9.363665543251486e-05,
"loss": 2.6062,
"theoretical_loss": 3.465861186095531,
"tokens_seen": 1770455040
},
{
"epoch": 0.07,
"learning_rate": 9.362863103835661e-05,
"loss": 2.6616,
"theoretical_loss": 3.4658392770524866,
"tokens_seen": 1770586112
},
{
"epoch": 0.07,
"learning_rate": 9.362060664419838e-05,
"loss": 2.54,
"theoretical_loss": 3.4658173700853387,
"tokens_seen": 1770717184
},
{
"epoch": 0.07,
"learning_rate": 9.361258225004013e-05,
"loss": 2.4839,
"theoretical_loss": 3.465795465193737,
"tokens_seen": 1770848256
},
{
"epoch": 0.07,
"learning_rate": 9.360455785588188e-05,
"loss": 2.5301,
"theoretical_loss": 3.465773562377331,
"tokens_seen": 1770979328
},
{
"epoch": 0.07,
"learning_rate": 9.359653346172365e-05,
"loss": 2.4655,
"theoretical_loss": 3.4657516616357706,
"tokens_seen": 1771110400
},
{
"epoch": 0.07,
"objective/train/docs_used": 976931,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.991319179534912,
"objective/train/theoretical_loss": 3.4657297629687056,
"objective/train/tokens_used": 141700576,
"theoretical_loss": 3.4657297629687056,
"tokens_seen": 1771241472
},
{
"epoch": 0.07,
"learning_rate": 9.35885090675654e-05,
"loss": 2.6133,
"theoretical_loss": 3.4657297629687056,
"tokens_seen": 1771241472
},
{
"epoch": 0.07,
"learning_rate": 9.358048467340717e-05,
"loss": 2.5932,
"theoretical_loss": 3.4657078663757868,
"tokens_seen": 1771372544
},
{
"epoch": 0.07,
"learning_rate": 9.357246027924892e-05,
"loss": 2.5936,
"theoretical_loss": 3.4656859718566633,
"tokens_seen": 1771503616
},
{
"epoch": 0.07,
"learning_rate": 9.356443588509069e-05,
"loss": 2.5908,
"theoretical_loss": 3.4656640794109856,
"tokens_seen": 1771634688
},
{
"epoch": 0.07,
"learning_rate": 9.355641149093244e-05,
"loss": 2.5438,
"theoretical_loss": 3.4656421890384044,
"tokens_seen": 1771765760
},
{
"epoch": 0.07,
"learning_rate": 9.35483870967742e-05,
"loss": 2.4126,
"theoretical_loss": 3.465620300738569,
"tokens_seen": 1771896832
},
{
"epoch": 0.07,
"learning_rate": 9.354036270261596e-05,
"loss": 2.6821,
"theoretical_loss": 3.465598414511131,
"tokens_seen": 1772027904
},
{
"epoch": 0.07,
"learning_rate": 9.353233830845772e-05,
"loss": 2.5196,
"theoretical_loss": 3.46557653035574,
"tokens_seen": 1772158976
},
{
"epoch": 0.07,
"learning_rate": 9.352431391429948e-05,
"loss": 2.6704,
"theoretical_loss": 3.4655546482720476,
"tokens_seen": 1772290048
},
{
"epoch": 0.07,
"learning_rate": 9.351628952014124e-05,
"loss": 2.6822,
"theoretical_loss": 3.4655327682597035,
"tokens_seen": 1772421120
},
{
"epoch": 0.07,
"learning_rate": 9.3508265125983e-05,
"loss": 2.4813,
"theoretical_loss": 3.465510890318359,
"tokens_seen": 1772552192
},
{
"epoch": 0.07,
"learning_rate": 9.350024073182476e-05,
"loss": 2.5966,
"theoretical_loss": 3.465489014447665,
"tokens_seen": 1772683264
},
{
"epoch": 0.07,
"learning_rate": 9.349221633766652e-05,
"loss": 2.6041,
"theoretical_loss": 3.465467140647272,
"tokens_seen": 1772814336
},
{
"epoch": 0.07,
"objective/train/docs_used": 977871,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.325096368789673,
"objective/train/theoretical_loss": 3.4654562045233295,
"objective/train/tokens_used": 143338976,
"theoretical_loss": 3.4654562045233295,
"tokens_seen": 1772879872
},
{
"epoch": 0.07,
"learning_rate": 9.348419194350828e-05,
"loss": 2.5723,
"theoretical_loss": 3.465445268916832,
"tokens_seen": 1772945408
},
{
"epoch": 0.07,
"learning_rate": 9.347616754935003e-05,
"loss": 2.5643,
"theoretical_loss": 3.4654233992559944,
"tokens_seen": 1773076480
},
{
"epoch": 0.07,
"learning_rate": 9.34681431551918e-05,
"loss": 2.7092,
"theoretical_loss": 3.465401531664412,
"tokens_seen": 1773207552
},
{
"epoch": 0.07,
"learning_rate": 9.346011876103355e-05,
"loss": 2.5261,
"theoretical_loss": 3.465379666141735,
"tokens_seen": 1773338624
},
{
"epoch": 0.07,
"learning_rate": 9.345209436687532e-05,
"loss": 2.6486,
"theoretical_loss": 3.4653578026876155,
"tokens_seen": 1773469696
},
{
"epoch": 0.07,
"learning_rate": 9.344406997271707e-05,
"loss": 2.4702,
"theoretical_loss": 3.4653359413017046,
"tokens_seen": 1773600768
},
{
"epoch": 0.07,
"learning_rate": 9.343604557855882e-05,
"loss": 2.6074,
"theoretical_loss": 3.465314081983654,
"tokens_seen": 1773731840
},
{
"epoch": 0.08,
"learning_rate": 9.342802118440059e-05,
"loss": 2.7033,
"theoretical_loss": 3.4652922247331155,
"tokens_seen": 1773862912
},
{
"epoch": 0.08,
"learning_rate": 9.341999679024234e-05,
"loss": 2.6171,
"theoretical_loss": 3.4652703695497404,
"tokens_seen": 1773993984
},
{
"epoch": 0.08,
"learning_rate": 9.341197239608411e-05,
"loss": 2.6683,
"theoretical_loss": 3.4652485164331805,
"tokens_seen": 1774125056
},
{
"epoch": 0.08,
"learning_rate": 9.340394800192586e-05,
"loss": 2.5473,
"theoretical_loss": 3.4652266653830877,
"tokens_seen": 1774256128
},
{
"epoch": 0.08,
"learning_rate": 9.339592360776763e-05,
"loss": 2.503,
"theoretical_loss": 3.465204816399114,
"tokens_seen": 1774387200
},
{
"epoch": 0.08,
"objective/train/docs_used": 978597,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.45396089553833,
"objective/train/theoretical_loss": 3.4651829694809115,
"objective/train/tokens_used": 144977376,
"theoretical_loss": 3.4651829694809115,
"tokens_seen": 1774518272
},
{
"epoch": 0.08,
"learning_rate": 9.338789921360938e-05,
"loss": 2.3952,
"theoretical_loss": 3.4651829694809115,
"tokens_seen": 1774518272
},
{
"epoch": 0.08,
"learning_rate": 9.337987481945113e-05,
"loss": 2.6399,
"theoretical_loss": 3.4651611246281324,
"tokens_seen": 1774649344
},
{
"epoch": 0.08,
"learning_rate": 9.33718504252929e-05,
"loss": 2.53,
"theoretical_loss": 3.465139281840429,
"tokens_seen": 1774780416
},
{
"epoch": 0.08,
"learning_rate": 9.336382603113465e-05,
"loss": 2.4813,
"theoretical_loss": 3.465117441117453,
"tokens_seen": 1774911488
},
{
"epoch": 0.08,
"learning_rate": 9.335580163697642e-05,
"loss": 2.6267,
"theoretical_loss": 3.4650956024588577,
"tokens_seen": 1775042560
},
{
"epoch": 0.08,
"learning_rate": 9.334777724281817e-05,
"loss": 2.4972,
"theoretical_loss": 3.465073765864295,
"tokens_seen": 1775173632
},
{
"epoch": 0.08,
"learning_rate": 9.333975284865993e-05,
"loss": 2.6277,
"theoretical_loss": 3.4650519313334174,
"tokens_seen": 1775304704
},
{
"epoch": 0.08,
"learning_rate": 9.33317284545017e-05,
"loss": 2.5438,
"theoretical_loss": 3.465030098865877,
"tokens_seen": 1775435776
},
{
"epoch": 0.08,
"learning_rate": 9.332370406034345e-05,
"loss": 2.4372,
"theoretical_loss": 3.465008268461328,
"tokens_seen": 1775566848
},
{
"epoch": 0.08,
"learning_rate": 9.331567966618521e-05,
"loss": 2.3739,
"theoretical_loss": 3.4649864401194215,
"tokens_seen": 1775697920
},
{
"epoch": 0.08,
"learning_rate": 9.330765527202697e-05,
"loss": 2.4286,
"theoretical_loss": 3.464964613839812,
"tokens_seen": 1775828992
},
{
"epoch": 0.08,
"learning_rate": 9.329963087786872e-05,
"loss": 2.6974,
"theoretical_loss": 3.464942789622151,
"tokens_seen": 1775960064
},
{
"epoch": 0.08,
"learning_rate": 9.329160648371049e-05,
"loss": 2.4616,
"theoretical_loss": 3.4649209674660923,
"tokens_seen": 1776091136
},
{
"epoch": 0.08,
"objective/train/docs_used": 979583,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.8104419708251953,
"objective/train/theoretical_loss": 3.464910057161056,
"objective/train/tokens_used": 146615776,
"theoretical_loss": 3.464910057161056,
"tokens_seen": 1776156672
},
{
"epoch": 0.08,
"learning_rate": 9.328358208955224e-05,
"loss": 2.6414,
"theoretical_loss": 3.4648991473712893,
"tokens_seen": 1776222208
},
{
"epoch": 0.08,
"learning_rate": 9.3275557695394e-05,
"loss": 2.524,
"theoretical_loss": 3.464877329337395,
"tokens_seen": 1776353280
},
{
"epoch": 0.08,
"learning_rate": 9.326753330123576e-05,
"loss": 2.7874,
"theoretical_loss": 3.4648555133640624,
"tokens_seen": 1776484352
},
{
"epoch": 0.08,
"learning_rate": 9.325950890707753e-05,
"loss": 2.5007,
"theoretical_loss": 3.4648336994509448,
"tokens_seen": 1776615424
},
{
"epoch": 0.08,
"learning_rate": 9.325148451291928e-05,
"loss": 2.6567,
"theoretical_loss": 3.464811887597696,
"tokens_seen": 1776746496
},
{
"epoch": 0.08,
"learning_rate": 9.324346011876103e-05,
"loss": 2.5699,
"theoretical_loss": 3.4647900778039693,
"tokens_seen": 1776877568
},
{
"epoch": 0.08,
"learning_rate": 9.32354357246028e-05,
"loss": 2.5771,
"theoretical_loss": 3.4647682700694187,
"tokens_seen": 1777008640
},
{
"epoch": 0.08,
"learning_rate": 9.322741133044455e-05,
"loss": 2.6564,
"theoretical_loss": 3.4647464643936976,
"tokens_seen": 1777139712
},
{
"epoch": 0.08,
"learning_rate": 9.321938693628632e-05,
"loss": 2.5478,
"theoretical_loss": 3.4647246607764606,
"tokens_seen": 1777270784
},
{
"epoch": 0.08,
"learning_rate": 9.321136254212807e-05,
"loss": 2.5323,
"theoretical_loss": 3.4647028592173603,
"tokens_seen": 1777401856
},
{
"epoch": 0.08,
"learning_rate": 9.320333814796983e-05,
"loss": 2.6142,
"theoretical_loss": 3.4646810597160513,
"tokens_seen": 1777532928
},
{
"epoch": 0.08,
"learning_rate": 9.319531375381159e-05,
"loss": 2.6264,
"theoretical_loss": 3.4646592622721872,
"tokens_seen": 1777664000
},
{
"epoch": 0.08,
"objective/train/docs_used": 979848,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.569380044937134,
"objective/train/theoretical_loss": 3.4646374668854234,
"objective/train/tokens_used": 148254176,
"theoretical_loss": 3.4646374668854234,
"tokens_seen": 1777795072
},
{
"epoch": 0.08,
"learning_rate": 9.318728935965335e-05,
"loss": 2.6984,
"theoretical_loss": 3.4646374668854234,
"tokens_seen": 1777795072
},
{
"epoch": 0.08,
"learning_rate": 9.317926496549511e-05,
"loss": 2.4607,
"theoretical_loss": 3.464615673555413,
"tokens_seen": 1777926144
},
{
"epoch": 0.08,
"learning_rate": 9.317124057133686e-05,
"loss": 2.5602,
"theoretical_loss": 3.4645938822818105,
"tokens_seen": 1778057216
},
{
"epoch": 0.08,
"learning_rate": 9.316321617717863e-05,
"loss": 2.5766,
"theoretical_loss": 3.4645720930642705,
"tokens_seen": 1778188288
},
{
"epoch": 0.08,
"learning_rate": 9.315519178302038e-05,
"loss": 2.5064,
"theoretical_loss": 3.4645503059024474,
"tokens_seen": 1778319360
},
{
"epoch": 0.08,
"learning_rate": 9.314716738886214e-05,
"loss": 2.6483,
"theoretical_loss": 3.4645285207959953,
"tokens_seen": 1778450432
},
{
"epoch": 0.08,
"learning_rate": 9.31391429947039e-05,
"loss": 2.437,
"theoretical_loss": 3.4645067377445695,
"tokens_seen": 1778581504
},
{
"epoch": 0.08,
"learning_rate": 9.313111860054566e-05,
"loss": 2.7707,
"theoretical_loss": 3.4644849567478246,
"tokens_seen": 1778712576
},
{
"epoch": 0.08,
"learning_rate": 9.312309420638742e-05,
"loss": 2.6932,
"theoretical_loss": 3.464463177805415,
"tokens_seen": 1778843648
},
{
"epoch": 0.08,
"learning_rate": 9.311506981222918e-05,
"loss": 2.5334,
"theoretical_loss": 3.4644414009169964,
"tokens_seen": 1778974720
},
{
"epoch": 0.08,
"learning_rate": 9.310704541807093e-05,
"loss": 2.6573,
"theoretical_loss": 3.464419626082223,
"tokens_seen": 1779105792
},
{
"epoch": 0.08,
"learning_rate": 9.30990210239127e-05,
"loss": 2.5421,
"theoretical_loss": 3.46439785330075,
"tokens_seen": 1779236864
},
{
"epoch": 0.08,
"learning_rate": 9.309099662975445e-05,
"loss": 2.6231,
"theoretical_loss": 3.464376082572233,
"tokens_seen": 1779367936
},
{
"epoch": 0.08,
"objective/train/docs_used": 981185,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.7941625118255615,
"objective/train/theoretical_loss": 3.464365197977725,
"objective/train/tokens_used": 149892576,
"theoretical_loss": 3.464365197977725,
"tokens_seen": 1779433472
},
{
"epoch": 0.08,
"learning_rate": 9.308297223559622e-05,
"loss": 2.5164,
"theoretical_loss": 3.464354313896327,
"tokens_seen": 1779499008
},
{
"epoch": 0.08,
"learning_rate": 9.307494784143797e-05,
"loss": 2.5201,
"theoretical_loss": 3.4643325472726865,
"tokens_seen": 1779630080
},
{
"epoch": 0.08,
"learning_rate": 9.306692344727974e-05,
"loss": 2.4844,
"theoretical_loss": 3.4643107827009674,
"tokens_seen": 1779761152
},
{
"epoch": 0.08,
"learning_rate": 9.305889905312149e-05,
"loss": 2.6485,
"theoretical_loss": 3.464289020180826,
"tokens_seen": 1779892224
},
{
"epoch": 0.08,
"learning_rate": 9.305087465896324e-05,
"loss": 2.5563,
"theoretical_loss": 3.4642672597119173,
"tokens_seen": 1780023296
},
{
"epoch": 0.08,
"learning_rate": 9.304285026480501e-05,
"loss": 2.4439,
"theoretical_loss": 3.464245501293896,
"tokens_seen": 1780154368
},
{
"epoch": 0.08,
"learning_rate": 9.303482587064676e-05,
"loss": 2.5935,
"theoretical_loss": 3.4642237449264193,
"tokens_seen": 1780285440
},
{
"epoch": 0.08,
"learning_rate": 9.302680147648853e-05,
"loss": 2.6076,
"theoretical_loss": 3.4642019906091424,
"tokens_seen": 1780416512
},
{
"epoch": 0.08,
"learning_rate": 9.301877708233028e-05,
"loss": 2.6269,
"theoretical_loss": 3.464180238341721,
"tokens_seen": 1780547584
},
{
"epoch": 0.08,
"learning_rate": 9.301075268817204e-05,
"loss": 2.5585,
"theoretical_loss": 3.464158488123811,
"tokens_seen": 1780678656
},
{
"epoch": 0.08,
"learning_rate": 9.30027282940138e-05,
"loss": 2.6408,
"theoretical_loss": 3.464136739955069,
"tokens_seen": 1780809728
},
{
"epoch": 0.08,
"learning_rate": 9.299470389985556e-05,
"loss": 2.7265,
"theoretical_loss": 3.4641149938351505,
"tokens_seen": 1780940800
},
{
"epoch": 0.08,
"objective/train/docs_used": 981883,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 2.705595016479492,
"objective/train/theoretical_loss": 3.4640932497637125,
"objective/train/tokens_used": 151530976,
"theoretical_loss": 3.4640932497637125,
"tokens_seen": 1781071872
},
{
"epoch": 0.08,
"learning_rate": 9.298667950569732e-05,
"loss": 2.6347,
"theoretical_loss": 3.4640932497637125,
"tokens_seen": 1781071872
},
{
"epoch": 0.08,
"learning_rate": 9.297865511153908e-05,
"loss": 2.4974,
"theoretical_loss": 3.4640715077404103,
"tokens_seen": 1781202944
},
{
"epoch": 0.08,
"learning_rate": 9.297063071738084e-05,
"loss": 2.5395,
"theoretical_loss": 3.464049767764901,
"tokens_seen": 1781334016
},
{
"epoch": 0.08,
"learning_rate": 9.29626063232226e-05,
"loss": 2.5233,
"theoretical_loss": 3.4640280298368413,
"tokens_seen": 1781465088
},
{
"epoch": 0.08,
"learning_rate": 9.295458192906435e-05,
"loss": 2.5728,
"theoretical_loss": 3.4640062939558867,
"tokens_seen": 1781596160
}
],
"max_steps": 12588,
"num_train_epochs": 9223372036854775807,
"total_flos": 6.7158323232768e+16,
"trial_name": null,
"trial_params": null
}