passive-lstm-3 / checkpoint-76320 /trainer_state.json
abhinavp's picture
Training in progress, epoch 0, checkpoint
8410c42 verified
{
"best_metric": 4.75404691696167,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/lstm/3/checkpoints/checkpoint-76320",
"epoch": 0.025000606015738065,
"eval_steps": 10,
"global_step": 76320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.999998362119627e-05,
"loss": 10.8209,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 7.5568,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 7.0563,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 6.9949,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 6.9457,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 6.8667,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 6.7087,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 6.6099,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 6.5194,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 6.4426,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 6.3859,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 6.3178,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989936862987376e-05,
"loss": 6.2549,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989098268236324e-05,
"loss": 6.1917,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988259673485272e-05,
"loss": 6.14,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.98742107873422e-05,
"loss": 6.0751,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986582483983168e-05,
"loss": 6.0418,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985743889232116e-05,
"loss": 5.9947,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984905294481064e-05,
"loss": 5.9555,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984066699730012e-05,
"loss": 5.917,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.8862,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.982391148108281e-05,
"loss": 5.8472,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.981552553357229e-05,
"loss": 5.8198,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 5.778,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 5.7659,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790384069844466e-05,
"loss": 5.7291,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9781998122333946e-05,
"loss": 5.7097,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773612174823426e-05,
"loss": 5.6721,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.97652262273129e-05,
"loss": 5.6552,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.975684027980238e-05,
"loss": 5.6355,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.974845433229186e-05,
"loss": 5.6104,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.974008476358507e-05,
"loss": 5.6008,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.973169881607455e-05,
"loss": 5.5742,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.972331286856403e-05,
"loss": 5.5579,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714926921053515e-05,
"loss": 5.5499,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.9706540973542995e-05,
"loss": 5.5264,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.9698171404836204e-05,
"loss": 5.5173,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.9689785457325684e-05,
"loss": 5.471,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.9681399509815164e-05,
"loss": 5.477,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.9673013562304644e-05,
"loss": 5.4379,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.9664627614794124e-05,
"loss": 5.4437,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965625804608733e-05,
"loss": 5.4203,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964787209857681e-05,
"loss": 5.4179,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963948615106629e-05,
"loss": 5.3885,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.963110020355577e-05,
"loss": 5.3944,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962271425604525e-05,
"loss": 5.3831,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.961432830853473e-05,
"loss": 5.3749,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960595873982795e-05,
"loss": 5.3678,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959757279231743e-05,
"loss": 5.3383,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958918684480691e-05,
"loss": 5.3293,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958080089729639e-05,
"loss": 5.3294,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957241494978587e-05,
"loss": 5.3153,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956402900227535e-05,
"loss": 5.3102,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955564305476483e-05,
"loss": 5.2874,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954727348605804e-05,
"loss": 5.2838,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.953888753854752e-05,
"loss": 5.2638,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530501591037e-05,
"loss": 5.2849,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.952211564352648e-05,
"loss": 5.2387,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951372969601596e-05,
"loss": 5.2546,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950534374850543e-05,
"loss": 5.2441,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949695780099492e-05,
"loss": 5.2235,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948858823228813e-05,
"loss": 5.2221,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.948020228477761e-05,
"loss": 5.2082,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.9471816337267086e-05,
"loss": 5.1982,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.9463430389756566e-05,
"loss": 5.191,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945507719985351e-05,
"loss": 5.2049,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944669125234299e-05,
"loss": 5.1792,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943830530483247e-05,
"loss": 5.1701,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942991935732195e-05,
"loss": 5.1601,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942153340981143e-05,
"loss": 5.157,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.9413147462300904e-05,
"loss": 5.1632,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404761514790384e-05,
"loss": 5.1577,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.939637556727987e-05,
"loss": 5.1337,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.938798961976935e-05,
"loss": 5.1489,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379636429866296e-05,
"loss": 5.1414,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371250482355776e-05,
"loss": 5.1144,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362864534845256e-05,
"loss": 5.1127,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.935447858733473e-05,
"loss": 5.103,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.934609263982421e-05,
"loss": 5.0947,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.933770669231369e-05,
"loss": 5.0891,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.932932074480317e-05,
"loss": 5.0944,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932093479729265e-05,
"loss": 5.0902,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931254884978213e-05,
"loss": 5.0915,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930417928107534e-05,
"loss": 5.0665,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.9295793333564825e-05,
"loss": 5.0488,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.9287423764858034e-05,
"loss": 5.0552,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.9279037817347514e-05,
"loss": 5.0559,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.9270651869836994e-05,
"loss": 5.0544,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.9262265922326474e-05,
"loss": 5.0371,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.9253879974815954e-05,
"loss": 5.0355,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.9245494027305433e-05,
"loss": 5.0355,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.9237108079794913e-05,
"loss": 5.0281,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922872213228439e-05,
"loss": 5.0165,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922033618477387e-05,
"loss": 5.0128,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921195023726335e-05,
"loss": 5.0001,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920356428975283e-05,
"loss": 5.0129,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919519472104604e-05,
"loss": 5.0006,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918680877353552e-05,
"loss": 4.979,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917842282602501e-05,
"loss": 4.9847,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.917003687851449e-05,
"loss": 4.982,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.916165093100397e-05,
"loss": 4.9889,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915328136229718e-05,
"loss": 4.9709,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914489541478666e-05,
"loss": 4.9677,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.913650946727614e-05,
"loss": 4.9518,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.912812351976562e-05,
"loss": 4.9576,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911973757225509e-05,
"loss": 4.9371,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.911136800354831e-05,
"loss": 4.9483,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910298205603779e-05,
"loss": 4.9349,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.909459610852727e-05,
"loss": 4.9384,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908621016101675e-05,
"loss": 4.9274,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907782421350623e-05,
"loss": 4.9304,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906943826599571e-05,
"loss": 4.9166,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906105231848519e-05,
"loss": 4.9207,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905266637097467e-05,
"loss": 4.9128,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.9044296802267876e-05,
"loss": 4.9061,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.9035910854757356e-05,
"loss": 4.9177,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.9027524907246836e-05,
"loss": 4.9003,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019155338540045e-05,
"loss": 4.9002,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010769391029525e-05,
"loss": 4.8912,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002383443519005e-05,
"loss": 4.8944,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994013874812214e-05,
"loss": 4.8892,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.89856279273017e-05,
"loss": 4.8825,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.897724197979118e-05,
"loss": 4.8876,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.896885603228066e-05,
"loss": 4.874,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.896047008477014e-05,
"loss": 4.8753,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.895208413725962e-05,
"loss": 4.8537,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.89436981897491e-05,
"loss": 4.8603,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.893531224223858e-05,
"loss": 4.8618,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892694267353179e-05,
"loss": 4.8521,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891855672602127e-05,
"loss": 4.8429,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891017077851075e-05,
"loss": 4.8658,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890178483100023e-05,
"loss": 4.8481,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889339888348971e-05,
"loss": 4.8407,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888501293597919e-05,
"loss": 4.8387,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.88766433672724e-05,
"loss": 4.8334,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886825741976188e-05,
"loss": 4.8353,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.8859871472251365e-05,
"loss": 4.8324,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.8851485524740845e-05,
"loss": 4.8275,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.8843099577230325e-05,
"loss": 4.829,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.8834713629719805e-05,
"loss": 4.828,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882632768220928e-05,
"loss": 4.8157,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881794173469876e-05,
"loss": 4.8062,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.8809572165991974e-05,
"loss": 4.8178,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880120259728518e-05,
"loss": 4.8037,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879281664977466e-05,
"loss": 4.8074,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878443070226414e-05,
"loss": 4.7987,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877604475475362e-05,
"loss": 4.7962,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.87676588072431e-05,
"loss": 4.8014,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875927285973258e-05,
"loss": 4.7887,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.87509032910258e-05,
"loss": 4.7862,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.75404691696167,
"eval_runtime": 294.4935,
"eval_samples_per_second": 1295.754,
"eval_steps_per_second": 40.493,
"step": 76320
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 3.153557465701272e+16,
"trial_name": null,
"trial_params": null
}