{ "best_metric": 4.2855424880981445, "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/4/checkpoints/checkpoint-305280", "epoch": 0.025000606015738065, "eval_steps": 10, "global_step": 305280, "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": 10.82, "step": 1 }, { "epoch": 0.0, "learning_rate": 4.999161405248948e-05, "loss": 7.5473, "step": 512 }, { "epoch": 0.0, "learning_rate": 4.998322810497896e-05, "loss": 7.0522, "step": 1024 }, { "epoch": 0.0, "learning_rate": 4.997484215746844e-05, "loss": 6.9852, "step": 1536 }, { "epoch": 0.0, "learning_rate": 4.996645620995792e-05, "loss": 6.9385, "step": 2048 }, { "epoch": 0.0, "learning_rate": 4.99580702624474e-05, "loss": 6.8823, "step": 2560 }, { "epoch": 0.0, "learning_rate": 4.994968431493688e-05, "loss": 6.7148, "step": 3072 }, { "epoch": 0.0, "learning_rate": 4.994131474623009e-05, "loss": 6.6152, "step": 3584 }, { "epoch": 0.0, "learning_rate": 4.993292879871958e-05, "loss": 6.5142, "step": 4096 }, { "epoch": 0.0, "learning_rate": 4.992454285120906e-05, "loss": 6.4434, "step": 4608 }, { "epoch": 0.0, "learning_rate": 4.991615690369854e-05, "loss": 6.3732, "step": 5120 }, { "epoch": 0.0, "learning_rate": 4.990777095618801e-05, "loss": 6.3151, "step": 5632 }, { "epoch": 0.0, "learning_rate": 4.989938500867749e-05, "loss": 6.2482, "step": 6144 }, { "epoch": 0.0, "learning_rate": 4.989099906116697e-05, "loss": 6.1796, "step": 6656 }, { "epoch": 0.0, "learning_rate": 4.988261311365645e-05, "loss": 6.1237, "step": 7168 }, { "epoch": 0.0, "learning_rate": 4.987422716614593e-05, "loss": 6.0564, "step": 7680 }, { "epoch": 0.0, "learning_rate": 4.986584121863541e-05, "loss": 6.0163, "step": 8192 }, { "epoch": 0.0, "learning_rate": 4.985745527112489e-05, "loss": 5.975, "step": 8704 }, { "epoch": 0.0, "learning_rate": 4.984906932361437e-05, "loss": 5.9339, "step": 9216 }, { "epoch": 0.0, "learning_rate": 4.984068337610385e-05, "loss": 5.8988, "step": 9728 }, { "epoch": 0.0, "learning_rate": 4.983229742859333e-05, "loss": 5.8686, "step": 10240 }, { "epoch": 0.0, "learning_rate": 4.982391148108281e-05, "loss": 5.8293, "step": 10752 }, { "epoch": 0.0, "learning_rate": 4.981552553357229e-05, "loss": 5.805, "step": 11264 }, { "epoch": 0.0, "learning_rate": 4.980713958606178e-05, "loss": 5.7631, "step": 11776 }, { "epoch": 0.0, "learning_rate": 4.9798770017354986e-05, "loss": 5.7504, "step": 12288 }, { "epoch": 0.0, "learning_rate": 4.9790384069844466e-05, "loss": 5.7189, "step": 12800 }, { "epoch": 0.0, "learning_rate": 4.9781998122333946e-05, "loss": 5.7002, "step": 13312 }, { "epoch": 0.0, "learning_rate": 4.9773612174823426e-05, "loss": 5.6721, "step": 13824 }, { "epoch": 0.0, "learning_rate": 4.9765242606116635e-05, "loss": 5.6404, "step": 14336 }, { "epoch": 0.0, "learning_rate": 4.9756856658606115e-05, "loss": 5.6248, "step": 14848 }, { "epoch": 0.01, "learning_rate": 4.9748470711095595e-05, "loss": 5.6, "step": 15360 }, { "epoch": 0.01, "learning_rate": 4.974008476358507e-05, "loss": 5.5899, "step": 15872 }, { "epoch": 0.01, "learning_rate": 4.973169881607455e-05, "loss": 5.5641, "step": 16384 }, { "epoch": 0.01, "learning_rate": 4.9723329247367764e-05, "loss": 5.5525, "step": 16896 }, { "epoch": 0.01, "learning_rate": 4.9714943299857244e-05, "loss": 5.542, "step": 17408 }, { "epoch": 0.01, "learning_rate": 4.970657373115046e-05, "loss": 5.5196, "step": 17920 }, { "epoch": 0.01, "learning_rate": 4.969818778363994e-05, "loss": 5.5083, "step": 18432 }, { "epoch": 0.01, "learning_rate": 4.968980183612942e-05, "loss": 5.4655, "step": 18944 }, { "epoch": 0.01, "learning_rate": 4.96814158886189e-05, "loss": 5.462, "step": 19456 }, { "epoch": 0.01, "learning_rate": 4.967302994110837e-05, "loss": 5.4422, "step": 19968 }, { "epoch": 0.01, "learning_rate": 4.966464399359785e-05, "loss": 5.4265, "step": 20480 }, { "epoch": 0.01, "learning_rate": 4.965625804608733e-05, "loss": 5.4222, "step": 20992 }, { "epoch": 0.01, "learning_rate": 4.964787209857681e-05, "loss": 5.4158, "step": 21504 }, { "epoch": 0.01, "learning_rate": 4.963950252987002e-05, "loss": 5.3867, "step": 22016 }, { "epoch": 0.01, "learning_rate": 4.96311165823595e-05, "loss": 5.3863, "step": 22528 }, { "epoch": 0.01, "learning_rate": 4.962273063484898e-05, "loss": 5.3819, "step": 23040 }, { "epoch": 0.01, "learning_rate": 4.961434468733847e-05, "loss": 5.3691, "step": 23552 }, { "epoch": 0.01, "learning_rate": 4.960595873982795e-05, "loss": 5.3632, "step": 24064 }, { "epoch": 0.01, "learning_rate": 4.959757279231743e-05, "loss": 5.3471, "step": 24576 }, { "epoch": 0.01, "learning_rate": 4.958918684480691e-05, "loss": 5.3251, "step": 25088 }, { "epoch": 0.01, "learning_rate": 4.958080089729639e-05, "loss": 5.3285, "step": 25600 }, { "epoch": 0.01, "learning_rate": 4.957241494978587e-05, "loss": 5.3084, "step": 26112 }, { "epoch": 0.01, "learning_rate": 4.956402900227535e-05, "loss": 5.3114, "step": 26624 }, { "epoch": 0.01, "learning_rate": 4.955564305476483e-05, "loss": 5.2796, "step": 27136 }, { "epoch": 0.01, "learning_rate": 4.954725710725431e-05, "loss": 5.2933, "step": 27648 }, { "epoch": 0.01, "learning_rate": 4.953888753854752e-05, "loss": 5.2666, "step": 28160 }, { "epoch": 0.01, "learning_rate": 4.9530517969840727e-05, "loss": 5.2844, "step": 28672 }, { "epoch": 0.01, "learning_rate": 4.9522132022330207e-05, "loss": 5.256, "step": 29184 }, { "epoch": 0.01, "learning_rate": 4.9513746074819686e-05, "loss": 5.2434, "step": 29696 }, { "epoch": 0.01, "learning_rate": 4.9505360127309166e-05, "loss": 5.2383, "step": 30208 }, { "epoch": 0.01, "learning_rate": 4.949699055860238e-05, "loss": 5.2334, "step": 30720 }, { "epoch": 0.01, "learning_rate": 4.948860461109186e-05, "loss": 5.2186, "step": 31232 }, { "epoch": 0.01, "learning_rate": 4.948021866358134e-05, "loss": 5.2216, "step": 31744 }, { "epoch": 0.01, "learning_rate": 4.947183271607082e-05, "loss": 5.1928, "step": 32256 }, { "epoch": 0.01, "learning_rate": 4.94634467685603e-05, "loss": 5.1905, "step": 32768 }, { "epoch": 0.01, "learning_rate": 4.945507719985351e-05, "loss": 5.2017, "step": 33280 }, { "epoch": 0.01, "learning_rate": 4.944669125234299e-05, "loss": 5.1932, "step": 33792 }, { "epoch": 0.01, "learning_rate": 4.943830530483247e-05, "loss": 5.1767, "step": 34304 }, { "epoch": 0.01, "learning_rate": 4.942991935732195e-05, "loss": 5.1594, "step": 34816 }, { "epoch": 0.01, "learning_rate": 4.942153340981143e-05, "loss": 5.1598, "step": 35328 }, { "epoch": 0.01, "learning_rate": 4.9413147462300904e-05, "loss": 5.1606, "step": 35840 }, { "epoch": 0.01, "learning_rate": 4.9404761514790384e-05, "loss": 5.153, "step": 36352 }, { "epoch": 0.01, "learning_rate": 4.939637556727987e-05, "loss": 5.1494, "step": 36864 }, { "epoch": 0.01, "learning_rate": 4.938800599857309e-05, "loss": 5.1453, "step": 37376 }, { "epoch": 0.01, "learning_rate": 4.937962005106256e-05, "loss": 5.145, "step": 37888 }, { "epoch": 0.01, "learning_rate": 4.937123410355204e-05, "loss": 5.1365, "step": 38400 }, { "epoch": 0.01, "learning_rate": 4.9362864534845256e-05, "loss": 5.1112, "step": 38912 }, { "epoch": 0.01, "learning_rate": 4.935447858733473e-05, "loss": 5.1063, "step": 39424 }, { "epoch": 0.01, "learning_rate": 4.934609263982421e-05, "loss": 5.104, "step": 39936 }, { "epoch": 0.01, "learning_rate": 4.933770669231369e-05, "loss": 5.0925, "step": 40448 }, { "epoch": 0.01, "learning_rate": 4.932932074480317e-05, "loss": 5.1014, "step": 40960 }, { "epoch": 0.01, "learning_rate": 4.932093479729265e-05, "loss": 5.0837, "step": 41472 }, { "epoch": 0.01, "learning_rate": 4.931256522858586e-05, "loss": 5.0995, "step": 41984 }, { "epoch": 0.01, "learning_rate": 4.930417928107534e-05, "loss": 5.0745, "step": 42496 }, { "epoch": 0.01, "learning_rate": 4.9295793333564825e-05, "loss": 5.0608, "step": 43008 }, { "epoch": 0.01, "learning_rate": 4.9287407386054305e-05, "loss": 5.0643, "step": 43520 }, { "epoch": 0.01, "learning_rate": 4.9279021438543785e-05, "loss": 5.0646, "step": 44032 }, { "epoch": 0.01, "learning_rate": 4.9270635491033264e-05, "loss": 5.0604, "step": 44544 }, { "epoch": 0.01, "learning_rate": 4.9262265922326474e-05, "loss": 5.0487, "step": 45056 }, { "epoch": 0.01, "learning_rate": 4.9253879974815954e-05, "loss": 5.0434, "step": 45568 }, { "epoch": 0.02, "learning_rate": 4.9245494027305433e-05, "loss": 5.0357, "step": 46080 }, { "epoch": 0.02, "learning_rate": 4.9237108079794913e-05, "loss": 5.0342, "step": 46592 }, { "epoch": 0.02, "learning_rate": 4.922872213228439e-05, "loss": 5.0238, "step": 47104 }, { "epoch": 0.02, "learning_rate": 4.92203525635776e-05, "loss": 5.0192, "step": 47616 }, { "epoch": 0.02, "learning_rate": 4.921196661606708e-05, "loss": 5.0186, "step": 48128 }, { "epoch": 0.02, "learning_rate": 4.920358066855656e-05, "loss": 5.0238, "step": 48640 }, { "epoch": 0.02, "learning_rate": 4.919519472104604e-05, "loss": 5.0035, "step": 49152 }, { "epoch": 0.02, "learning_rate": 4.918680877353552e-05, "loss": 4.9966, "step": 49664 }, { "epoch": 0.02, "learning_rate": 4.917842282602501e-05, "loss": 4.9899, "step": 50176 }, { "epoch": 0.02, "learning_rate": 4.917003687851449e-05, "loss": 4.988, "step": 50688 }, { "epoch": 0.02, "learning_rate": 4.91616673098077e-05, "loss": 4.9917, "step": 51200 }, { "epoch": 0.02, "learning_rate": 4.915328136229718e-05, "loss": 4.9787, "step": 51712 }, { "epoch": 0.02, "learning_rate": 4.914489541478666e-05, "loss": 4.9719, "step": 52224 }, { "epoch": 0.02, "learning_rate": 4.913650946727614e-05, "loss": 4.9635, "step": 52736 }, { "epoch": 0.02, "learning_rate": 4.912812351976562e-05, "loss": 4.9665, "step": 53248 }, { "epoch": 0.02, "learning_rate": 4.911975395105883e-05, "loss": 4.9394, "step": 53760 }, { "epoch": 0.02, "learning_rate": 4.911136800354831e-05, "loss": 4.9543, "step": 54272 }, { "epoch": 0.02, "learning_rate": 4.910298205603779e-05, "loss": 4.941, "step": 54784 }, { "epoch": 0.02, "learning_rate": 4.909459610852727e-05, "loss": 4.9473, "step": 55296 }, { "epoch": 0.02, "learning_rate": 4.908621016101675e-05, "loss": 4.9354, "step": 55808 }, { "epoch": 0.02, "learning_rate": 4.907782421350623e-05, "loss": 4.9431, "step": 56320 }, { "epoch": 0.02, "learning_rate": 4.906945464479944e-05, "loss": 4.9294, "step": 56832 }, { "epoch": 0.02, "learning_rate": 4.906106869728892e-05, "loss": 4.9222, "step": 57344 }, { "epoch": 0.02, "learning_rate": 4.9052682749778396e-05, "loss": 4.9307, "step": 57856 }, { "epoch": 0.02, "learning_rate": 4.9044296802267876e-05, "loss": 4.9087, "step": 58368 }, { "epoch": 0.02, "learning_rate": 4.9035910854757356e-05, "loss": 4.925, "step": 58880 }, { "epoch": 0.02, "learning_rate": 4.9027541286050565e-05, "loss": 4.9147, "step": 59392 }, { "epoch": 0.02, "learning_rate": 4.901917171734378e-05, "loss": 4.9125, "step": 59904 }, { "epoch": 0.02, "learning_rate": 4.901078576983326e-05, "loss": 4.9, "step": 60416 }, { "epoch": 0.02, "learning_rate": 4.900239982232274e-05, "loss": 4.8989, "step": 60928 }, { "epoch": 0.02, "learning_rate": 4.8994013874812214e-05, "loss": 4.9007, "step": 61440 }, { "epoch": 0.02, "learning_rate": 4.898564430610543e-05, "loss": 4.8932, "step": 61952 }, { "epoch": 0.02, "learning_rate": 4.897725835859491e-05, "loss": 4.892, "step": 62464 }, { "epoch": 0.02, "learning_rate": 4.8968872411084396e-05, "loss": 4.881, "step": 62976 }, { "epoch": 0.02, "learning_rate": 4.896048646357387e-05, "loss": 4.8836, "step": 63488 }, { "epoch": 0.02, "learning_rate": 4.895210051606335e-05, "loss": 4.8642, "step": 64000 }, { "epoch": 0.02, "learning_rate": 4.894371456855283e-05, "loss": 4.8687, "step": 64512 }, { "epoch": 0.02, "learning_rate": 4.893532862104231e-05, "loss": 4.8521, "step": 65024 }, { "epoch": 0.02, "learning_rate": 4.892694267353179e-05, "loss": 4.8586, "step": 65536 }, { "epoch": 0.02, "learning_rate": 4.8918573104825e-05, "loss": 4.8453, "step": 66048 }, { "epoch": 0.02, "learning_rate": 4.891018715731448e-05, "loss": 4.8663, "step": 66560 }, { "epoch": 0.02, "learning_rate": 4.890180120980396e-05, "loss": 4.8714, "step": 67072 }, { "epoch": 0.02, "learning_rate": 4.889341526229344e-05, "loss": 4.8554, "step": 67584 }, { "epoch": 0.02, "learning_rate": 4.888502931478292e-05, "loss": 4.8347, "step": 68096 }, { "epoch": 0.02, "learning_rate": 4.8876659746076134e-05, "loss": 4.8453, "step": 68608 }, { "epoch": 0.02, "learning_rate": 4.8868273798565614e-05, "loss": 4.8437, "step": 69120 }, { "epoch": 0.02, "learning_rate": 4.8859887851055094e-05, "loss": 4.8376, "step": 69632 }, { "epoch": 0.02, "learning_rate": 4.8851501903544574e-05, "loss": 4.8323, "step": 70144 }, { "epoch": 0.02, "learning_rate": 4.8843115956034054e-05, "loss": 4.8346, "step": 70656 }, { "epoch": 0.02, "learning_rate": 4.8834730008523534e-05, "loss": 4.8301, "step": 71168 }, { "epoch": 0.02, "learning_rate": 4.882637681862047e-05, "loss": 4.8193, "step": 71680 }, { "epoch": 0.02, "learning_rate": 4.881799087110995e-05, "loss": 4.8205, "step": 72192 }, { "epoch": 0.02, "learning_rate": 4.880960492359943e-05, "loss": 4.8198, "step": 72704 }, { "epoch": 0.02, "learning_rate": 4.880121897608891e-05, "loss": 4.8089, "step": 73216 }, { "epoch": 0.02, "learning_rate": 4.879283302857839e-05, "loss": 4.8085, "step": 73728 }, { "epoch": 0.02, "learning_rate": 4.878444708106787e-05, "loss": 4.817, "step": 74240 }, { "epoch": 0.02, "learning_rate": 4.877606113355735e-05, "loss": 4.804, "step": 74752 }, { "epoch": 0.02, "learning_rate": 4.876767518604683e-05, "loss": 4.8023, "step": 75264 }, { "epoch": 0.02, "learning_rate": 4.875928923853632e-05, "loss": 4.8159, "step": 75776 }, { "epoch": 0.02, "learning_rate": 4.87509032910258e-05, "loss": 4.7916, "step": 76288 }, { "epoch": 0.03, "eval_loss": 4.78051233291626, "eval_runtime": 327.7198, "eval_samples_per_second": 1164.382, "eval_steps_per_second": 36.388, "step": 76320 }, { "epoch": 0.0, "learning_rate": 4.874251734351528e-05, "loss": 4.7776, "step": 76800 }, { "epoch": 0.0, "learning_rate": 4.873413139600475e-05, "loss": 4.7857, "step": 77312 }, { "epoch": 0.0, "learning_rate": 4.872574544849423e-05, "loss": 4.7992, "step": 77824 }, { "epoch": 0.0, "learning_rate": 4.871735950098371e-05, "loss": 4.7871, "step": 78336 }, { "epoch": 0.0, "learning_rate": 4.870897355347319e-05, "loss": 4.7927, "step": 78848 }, { "epoch": 0.0, "learning_rate": 4.870058760596267e-05, "loss": 4.768, "step": 79360 }, { "epoch": 0.0, "learning_rate": 4.869220165845215e-05, "loss": 4.7849, "step": 79872 }, { "epoch": 0.0, "learning_rate": 4.868381571094163e-05, "loss": 4.7561, "step": 80384 }, { "epoch": 0.0, "learning_rate": 4.867542976343111e-05, "loss": 4.7715, "step": 80896 }, { "epoch": 0.0, "learning_rate": 4.866704381592059e-05, "loss": 4.7592, "step": 81408 }, { "epoch": 0.0, "learning_rate": 4.865865786841007e-05, "loss": 4.7752, "step": 81920 }, { "epoch": 0.0, "learning_rate": 4.865027192089955e-05, "loss": 4.7705, "step": 82432 }, { "epoch": 0.0, "learning_rate": 4.864190235219277e-05, "loss": 4.757, "step": 82944 }, { "epoch": 0.0, "learning_rate": 4.863351640468225e-05, "loss": 4.7516, "step": 83456 }, { "epoch": 0.0, "learning_rate": 4.862513045717173e-05, "loss": 4.7382, "step": 83968 }, { "epoch": 0.0, "learning_rate": 4.861674450966121e-05, "loss": 4.7402, "step": 84480 }, { "epoch": 0.0, "learning_rate": 4.860835856215069e-05, "loss": 4.7511, "step": 84992 }, { "epoch": 0.0, "learning_rate": 4.8599988993443896e-05, "loss": 4.7458, "step": 85504 }, { "epoch": 0.0, "learning_rate": 4.8591603045933376e-05, "loss": 4.7392, "step": 86016 }, { "epoch": 0.0, "learning_rate": 4.8583217098422856e-05, "loss": 4.7553, "step": 86528 }, { "epoch": 0.0, "learning_rate": 4.8574831150912336e-05, "loss": 4.7487, "step": 87040 }, { "epoch": 0.0, "learning_rate": 4.8566445203401816e-05, "loss": 4.7406, "step": 87552 }, { "epoch": 0.0, "learning_rate": 4.855805925589129e-05, "loss": 4.7285, "step": 88064 }, { "epoch": 0.0, "learning_rate": 4.854967330838077e-05, "loss": 4.7484, "step": 88576 }, { "epoch": 0.0, "learning_rate": 4.8541287360870256e-05, "loss": 4.7236, "step": 89088 }, { "epoch": 0.0, "learning_rate": 4.8532901413359736e-05, "loss": 4.7327, "step": 89600 }, { "epoch": 0.0, "learning_rate": 4.8524515465849216e-05, "loss": 4.7289, "step": 90112 }, { "epoch": 0.0, "learning_rate": 4.8516129518338696e-05, "loss": 4.723, "step": 90624 }, { "epoch": 0.0, "learning_rate": 4.8507759949631905e-05, "loss": 4.7116, "step": 91136 }, { "epoch": 0.01, "learning_rate": 4.8499374002121385e-05, "loss": 4.7129, "step": 91648 }, { "epoch": 0.01, "learning_rate": 4.8490988054610865e-05, "loss": 4.7155, "step": 92160 }, { "epoch": 0.01, "learning_rate": 4.8482602107100345e-05, "loss": 4.7109, "step": 92672 }, { "epoch": 0.01, "learning_rate": 4.8474216159589824e-05, "loss": 4.7111, "step": 93184 }, { "epoch": 0.01, "learning_rate": 4.8465846590883034e-05, "loss": 4.7176, "step": 93696 }, { "epoch": 0.01, "learning_rate": 4.8457460643372514e-05, "loss": 4.7082, "step": 94208 }, { "epoch": 0.01, "learning_rate": 4.8449074695861993e-05, "loss": 4.707, "step": 94720 }, { "epoch": 0.01, "learning_rate": 4.8440688748351473e-05, "loss": 4.6852, "step": 95232 }, { "epoch": 0.01, "learning_rate": 4.843230280084095e-05, "loss": 4.6927, "step": 95744 }, { "epoch": 0.01, "learning_rate": 4.842391685333043e-05, "loss": 4.683, "step": 96256 }, { "epoch": 0.01, "learning_rate": 4.841554728462365e-05, "loss": 4.6794, "step": 96768 }, { "epoch": 0.01, "learning_rate": 4.840716133711313e-05, "loss": 4.6897, "step": 97280 }, { "epoch": 0.01, "learning_rate": 4.839877538960261e-05, "loss": 4.692, "step": 97792 }, { "epoch": 0.01, "learning_rate": 4.839038944209209e-05, "loss": 4.6743, "step": 98304 }, { "epoch": 0.01, "learning_rate": 4.838200349458157e-05, "loss": 4.6823, "step": 98816 }, { "epoch": 0.01, "learning_rate": 4.837361754707105e-05, "loss": 4.6783, "step": 99328 }, { "epoch": 0.01, "learning_rate": 4.836523159956053e-05, "loss": 4.6822, "step": 99840 }, { "epoch": 0.01, "learning_rate": 4.835687840965747e-05, "loss": 4.6852, "step": 100352 }, { "epoch": 0.01, "learning_rate": 4.834849246214695e-05, "loss": 4.676, "step": 100864 }, { "epoch": 0.01, "learning_rate": 4.834010651463643e-05, "loss": 4.6651, "step": 101376 }, { "epoch": 0.01, "learning_rate": 4.833172056712591e-05, "loss": 4.6721, "step": 101888 }, { "epoch": 0.01, "learning_rate": 4.832333461961539e-05, "loss": 4.6605, "step": 102400 }, { "epoch": 0.01, "learning_rate": 4.8314948672104874e-05, "loss": 4.6692, "step": 102912 }, { "epoch": 0.01, "learning_rate": 4.8306562724594354e-05, "loss": 4.6561, "step": 103424 }, { "epoch": 0.01, "learning_rate": 4.829819315588756e-05, "loss": 4.6656, "step": 103936 }, { "epoch": 0.01, "learning_rate": 4.828980720837704e-05, "loss": 4.6456, "step": 104448 }, { "epoch": 0.01, "learning_rate": 4.828142126086652e-05, "loss": 4.6701, "step": 104960 }, { "epoch": 0.01, "learning_rate": 4.8273035313356e-05, "loss": 4.6512, "step": 105472 }, { "epoch": 0.01, "learning_rate": 4.8264649365845476e-05, "loss": 4.6475, "step": 105984 }, { "epoch": 0.01, "learning_rate": 4.8256263418334956e-05, "loss": 4.6545, "step": 106496 }, { "epoch": 0.01, "learning_rate": 4.824789384962817e-05, "loss": 4.6456, "step": 107008 }, { "epoch": 0.01, "learning_rate": 4.823950790211765e-05, "loss": 4.6386, "step": 107520 }, { "epoch": 0.01, "learning_rate": 4.8231121954607125e-05, "loss": 4.6526, "step": 108032 }, { "epoch": 0.01, "learning_rate": 4.822273600709661e-05, "loss": 4.6237, "step": 108544 }, { "epoch": 0.01, "learning_rate": 4.821435005958609e-05, "loss": 4.6294, "step": 109056 }, { "epoch": 0.01, "learning_rate": 4.820596411207557e-05, "loss": 4.6415, "step": 109568 }, { "epoch": 0.01, "learning_rate": 4.819757816456505e-05, "loss": 4.641, "step": 110080 }, { "epoch": 0.01, "learning_rate": 4.818919221705453e-05, "loss": 4.6324, "step": 110592 }, { "epoch": 0.01, "learning_rate": 4.818080626954401e-05, "loss": 4.6229, "step": 111104 }, { "epoch": 0.01, "learning_rate": 4.817242032203349e-05, "loss": 4.6226, "step": 111616 }, { "epoch": 0.01, "learning_rate": 4.816403437452297e-05, "loss": 4.6271, "step": 112128 }, { "epoch": 0.01, "learning_rate": 4.815566480581618e-05, "loss": 4.6328, "step": 112640 }, { "epoch": 0.01, "learning_rate": 4.814727885830566e-05, "loss": 4.6267, "step": 113152 }, { "epoch": 0.01, "learning_rate": 4.813889291079514e-05, "loss": 4.627, "step": 113664 }, { "epoch": 0.01, "learning_rate": 4.813052334208835e-05, "loss": 4.6331, "step": 114176 }, { "epoch": 0.01, "learning_rate": 4.812213739457783e-05, "loss": 4.63, "step": 114688 }, { "epoch": 0.01, "learning_rate": 4.811375144706731e-05, "loss": 4.613, "step": 115200 }, { "epoch": 0.01, "learning_rate": 4.8105365499556796e-05, "loss": 4.6108, "step": 115712 }, { "epoch": 0.01, "learning_rate": 4.8096979552046276e-05, "loss": 4.6192, "step": 116224 }, { "epoch": 0.01, "learning_rate": 4.8088609983339485e-05, "loss": 4.6036, "step": 116736 }, { "epoch": 0.01, "learning_rate": 4.8080224035828965e-05, "loss": 4.616, "step": 117248 }, { "epoch": 0.01, "learning_rate": 4.8071838088318445e-05, "loss": 4.6031, "step": 117760 }, { "epoch": 0.01, "learning_rate": 4.8063452140807925e-05, "loss": 4.6259, "step": 118272 }, { "epoch": 0.01, "learning_rate": 4.8055066193297405e-05, "loss": 4.6062, "step": 118784 }, { "epoch": 0.01, "learning_rate": 4.8046680245786885e-05, "loss": 4.591, "step": 119296 }, { "epoch": 0.01, "learning_rate": 4.8038294298276365e-05, "loss": 4.6016, "step": 119808 }, { "epoch": 0.01, "learning_rate": 4.802990835076584e-05, "loss": 4.6053, "step": 120320 }, { "epoch": 0.01, "learning_rate": 4.802152240325532e-05, "loss": 4.6077, "step": 120832 }, { "epoch": 0.01, "learning_rate": 4.8013152834548534e-05, "loss": 4.5984, "step": 121344 }, { "epoch": 0.01, "learning_rate": 4.8004766887038014e-05, "loss": 4.5972, "step": 121856 }, { "epoch": 0.02, "learning_rate": 4.7996380939527494e-05, "loss": 4.6018, "step": 122368 }, { "epoch": 0.02, "learning_rate": 4.7987994992016974e-05, "loss": 4.5901, "step": 122880 }, { "epoch": 0.02, "learning_rate": 4.797962542331019e-05, "loss": 4.5895, "step": 123392 }, { "epoch": 0.02, "learning_rate": 4.797123947579966e-05, "loss": 4.5898, "step": 123904 }, { "epoch": 0.02, "learning_rate": 4.796285352828914e-05, "loss": 4.5948, "step": 124416 }, { "epoch": 0.02, "learning_rate": 4.795446758077862e-05, "loss": 4.5942, "step": 124928 }, { "epoch": 0.02, "learning_rate": 4.79460816332681e-05, "loss": 4.5902, "step": 125440 }, { "epoch": 0.02, "learning_rate": 4.793769568575758e-05, "loss": 4.576, "step": 125952 }, { "epoch": 0.02, "learning_rate": 4.792930973824706e-05, "loss": 4.578, "step": 126464 }, { "epoch": 0.02, "learning_rate": 4.792092379073654e-05, "loss": 4.5817, "step": 126976 }, { "epoch": 0.02, "learning_rate": 4.791255422202975e-05, "loss": 4.5812, "step": 127488 }, { "epoch": 0.02, "learning_rate": 4.790416827451923e-05, "loss": 4.5773, "step": 128000 }, { "epoch": 0.02, "learning_rate": 4.789578232700871e-05, "loss": 4.5739, "step": 128512 }, { "epoch": 0.02, "learning_rate": 4.788741275830193e-05, "loss": 4.5662, "step": 129024 }, { "epoch": 0.02, "learning_rate": 4.787902681079141e-05, "loss": 4.5773, "step": 129536 }, { "epoch": 0.02, "learning_rate": 4.787064086328089e-05, "loss": 4.5465, "step": 130048 }, { "epoch": 0.02, "learning_rate": 4.786225491577037e-05, "loss": 4.5706, "step": 130560 }, { "epoch": 0.02, "learning_rate": 4.785386896825985e-05, "loss": 4.5619, "step": 131072 }, { "epoch": 0.02, "learning_rate": 4.784548302074933e-05, "loss": 4.5695, "step": 131584 }, { "epoch": 0.02, "learning_rate": 4.7837113452042536e-05, "loss": 4.5577, "step": 132096 }, { "epoch": 0.02, "learning_rate": 4.7828727504532016e-05, "loss": 4.5702, "step": 132608 }, { "epoch": 0.02, "learning_rate": 4.7820341557021496e-05, "loss": 4.5579, "step": 133120 }, { "epoch": 0.02, "learning_rate": 4.7811955609510976e-05, "loss": 4.5548, "step": 133632 }, { "epoch": 0.02, "learning_rate": 4.7803569662000456e-05, "loss": 4.5673, "step": 134144 }, { "epoch": 0.02, "learning_rate": 4.7795183714489936e-05, "loss": 4.5496, "step": 134656 }, { "epoch": 0.02, "learning_rate": 4.7786797766979416e-05, "loss": 4.5632, "step": 135168 }, { "epoch": 0.02, "learning_rate": 4.7778411819468896e-05, "loss": 4.5632, "step": 135680 }, { "epoch": 0.02, "learning_rate": 4.777005862956584e-05, "loss": 4.557, "step": 136192 }, { "epoch": 0.02, "learning_rate": 4.776167268205532e-05, "loss": 4.5481, "step": 136704 }, { "epoch": 0.02, "learning_rate": 4.77532867345448e-05, "loss": 4.5524, "step": 137216 }, { "epoch": 0.02, "learning_rate": 4.774490078703428e-05, "loss": 4.5562, "step": 137728 }, { "epoch": 0.02, "learning_rate": 4.773651483952376e-05, "loss": 4.5412, "step": 138240 }, { "epoch": 0.02, "learning_rate": 4.772814527081697e-05, "loss": 4.5578, "step": 138752 }, { "epoch": 0.02, "learning_rate": 4.771975932330645e-05, "loss": 4.5408, "step": 139264 }, { "epoch": 0.02, "learning_rate": 4.771137337579593e-05, "loss": 4.548, "step": 139776 }, { "epoch": 0.02, "learning_rate": 4.770298742828541e-05, "loss": 4.5406, "step": 140288 }, { "epoch": 0.02, "learning_rate": 4.769461785957862e-05, "loss": 4.5389, "step": 140800 }, { "epoch": 0.02, "learning_rate": 4.7686231912068105e-05, "loss": 4.5208, "step": 141312 }, { "epoch": 0.02, "learning_rate": 4.7677845964557585e-05, "loss": 4.5354, "step": 141824 }, { "epoch": 0.02, "learning_rate": 4.7669460017047065e-05, "loss": 4.5219, "step": 142336 }, { "epoch": 0.02, "learning_rate": 4.7661090448340274e-05, "loss": 4.5445, "step": 142848 }, { "epoch": 0.02, "learning_rate": 4.7652704500829754e-05, "loss": 4.5557, "step": 143360 }, { "epoch": 0.02, "learning_rate": 4.7644318553319234e-05, "loss": 4.5394, "step": 143872 }, { "epoch": 0.02, "learning_rate": 4.7635932605808714e-05, "loss": 4.5206, "step": 144384 }, { "epoch": 0.02, "learning_rate": 4.7627546658298194e-05, "loss": 4.535, "step": 144896 }, { "epoch": 0.02, "learning_rate": 4.7619160710787674e-05, "loss": 4.5288, "step": 145408 }, { "epoch": 0.02, "learning_rate": 4.761077476327715e-05, "loss": 4.532, "step": 145920 }, { "epoch": 0.02, "learning_rate": 4.760240519457036e-05, "loss": 4.5285, "step": 146432 }, { "epoch": 0.02, "learning_rate": 4.759401924705984e-05, "loss": 4.5286, "step": 146944 }, { "epoch": 0.02, "learning_rate": 4.758563329954932e-05, "loss": 4.5242, "step": 147456 }, { "epoch": 0.02, "learning_rate": 4.75772473520388e-05, "loss": 4.5249, "step": 147968 }, { "epoch": 0.02, "learning_rate": 4.756887778333202e-05, "loss": 4.5242, "step": 148480 }, { "epoch": 0.02, "learning_rate": 4.75604918358215e-05, "loss": 4.5268, "step": 148992 }, { "epoch": 0.02, "learning_rate": 4.755210588831097e-05, "loss": 4.5184, "step": 149504 }, { "epoch": 0.02, "learning_rate": 4.754371994080045e-05, "loss": 4.5138, "step": 150016 }, { "epoch": 0.02, "learning_rate": 4.753535037209367e-05, "loss": 4.528, "step": 150528 }, { "epoch": 0.02, "learning_rate": 4.752696442458315e-05, "loss": 4.5219, "step": 151040 }, { "epoch": 0.02, "learning_rate": 4.751857847707262e-05, "loss": 4.5113, "step": 151552 }, { "epoch": 0.02, "learning_rate": 4.751020890836584e-05, "loss": 4.5309, "step": 152064 }, { "epoch": 0.02, "learning_rate": 4.750182296085532e-05, "loss": 4.5106, "step": 152576 }, { "epoch": 0.03, "eval_loss": 4.504174709320068, "eval_runtime": 293.4935, "eval_samples_per_second": 1300.168, "eval_steps_per_second": 40.631, "step": 152640 }, { "epoch": 1.0, "learning_rate": 4.74934370133448e-05, "loss": 4.5019, "step": 153088 }, { "epoch": 1.0, "learning_rate": 4.748505106583428e-05, "loss": 4.504, "step": 153600 }, { "epoch": 1.0, "learning_rate": 4.747666511832376e-05, "loss": 4.5213, "step": 154112 }, { "epoch": 1.0, "learning_rate": 4.746829554961697e-05, "loss": 4.5127, "step": 154624 }, { "epoch": 1.0, "learning_rate": 4.7459909602106446e-05, "loss": 4.5167, "step": 155136 }, { "epoch": 1.0, "learning_rate": 4.7451523654595926e-05, "loss": 4.4965, "step": 155648 }, { "epoch": 1.0, "learning_rate": 4.7443137707085406e-05, "loss": 4.515, "step": 156160 }, { "epoch": 1.0, "learning_rate": 4.7434751759574886e-05, "loss": 4.4864, "step": 156672 }, { "epoch": 1.0, "learning_rate": 4.7426365812064366e-05, "loss": 4.5106, "step": 157184 }, { "epoch": 1.0, "learning_rate": 4.7417996243357575e-05, "loss": 4.494, "step": 157696 }, { "epoch": 1.0, "learning_rate": 4.7409610295847055e-05, "loss": 4.5109, "step": 158208 }, { "epoch": 1.0, "learning_rate": 4.7401224348336535e-05, "loss": 4.5093, "step": 158720 }, { "epoch": 1.0, "learning_rate": 4.7392838400826015e-05, "loss": 4.4959, "step": 159232 }, { "epoch": 1.0, "learning_rate": 4.7384452453315495e-05, "loss": 4.4926, "step": 159744 }, { "epoch": 1.0, "learning_rate": 4.7376066505804975e-05, "loss": 4.4861, "step": 160256 }, { "epoch": 1.0, "learning_rate": 4.736768055829446e-05, "loss": 4.4848, "step": 160768 }, { "epoch": 1.0, "learning_rate": 4.735929461078394e-05, "loss": 4.4913, "step": 161280 }, { "epoch": 1.0, "learning_rate": 4.735090866327342e-05, "loss": 4.4923, "step": 161792 }, { "epoch": 1.0, "learning_rate": 4.734253909456663e-05, "loss": 4.4909, "step": 162304 }, { "epoch": 1.0, "learning_rate": 4.733415314705611e-05, "loss": 4.502, "step": 162816 }, { "epoch": 1.0, "learning_rate": 4.732576719954559e-05, "loss": 4.5011, "step": 163328 }, { "epoch": 1.0, "learning_rate": 4.731738125203507e-05, "loss": 4.4945, "step": 163840 }, { "epoch": 1.0, "learning_rate": 4.730899530452455e-05, "loss": 4.4859, "step": 164352 }, { "epoch": 1.0, "learning_rate": 4.730062573581776e-05, "loss": 4.4972, "step": 164864 }, { "epoch": 1.0, "learning_rate": 4.729223978830724e-05, "loss": 4.4831, "step": 165376 }, { "epoch": 1.0, "learning_rate": 4.728385384079672e-05, "loss": 4.4908, "step": 165888 }, { "epoch": 1.0, "learning_rate": 4.72754678932862e-05, "loss": 4.4874, "step": 166400 }, { "epoch": 1.0, "learning_rate": 4.726708194577568e-05, "loss": 4.4826, "step": 166912 }, { "epoch": 1.0, "learning_rate": 4.725869599826516e-05, "loss": 4.4736, "step": 167424 }, { "epoch": 1.01, "learning_rate": 4.725031005075464e-05, "loss": 4.4761, "step": 167936 }, { "epoch": 1.01, "learning_rate": 4.7241940482047855e-05, "loss": 4.4784, "step": 168448 }, { "epoch": 1.01, "learning_rate": 4.7233554534537335e-05, "loss": 4.4804, "step": 168960 }, { "epoch": 1.01, "learning_rate": 4.722516858702681e-05, "loss": 4.4839, "step": 169472 }, { "epoch": 1.01, "learning_rate": 4.721678263951629e-05, "loss": 4.4796, "step": 169984 }, { "epoch": 1.01, "learning_rate": 4.720839669200577e-05, "loss": 4.4771, "step": 170496 }, { "epoch": 1.01, "learning_rate": 4.7200027123298984e-05, "loss": 4.4758, "step": 171008 }, { "epoch": 1.01, "learning_rate": 4.719164117578846e-05, "loss": 4.4578, "step": 171520 }, { "epoch": 1.01, "learning_rate": 4.718325522827794e-05, "loss": 4.4687, "step": 172032 }, { "epoch": 1.01, "learning_rate": 4.717486928076742e-05, "loss": 4.4602, "step": 172544 }, { "epoch": 1.01, "learning_rate": 4.71664833332569e-05, "loss": 4.4589, "step": 173056 }, { "epoch": 1.01, "learning_rate": 4.7158097385746384e-05, "loss": 4.464, "step": 173568 }, { "epoch": 1.01, "learning_rate": 4.714972781703959e-05, "loss": 4.4711, "step": 174080 }, { "epoch": 1.01, "learning_rate": 4.714134186952907e-05, "loss": 4.4581, "step": 174592 }, { "epoch": 1.01, "learning_rate": 4.713295592201855e-05, "loss": 4.4585, "step": 175104 }, { "epoch": 1.01, "learning_rate": 4.712456997450803e-05, "loss": 4.4662, "step": 175616 }, { "epoch": 1.01, "learning_rate": 4.711618402699751e-05, "loss": 4.46, "step": 176128 }, { "epoch": 1.01, "learning_rate": 4.710781445829072e-05, "loss": 4.4736, "step": 176640 }, { "epoch": 1.01, "learning_rate": 4.70994285107802e-05, "loss": 4.4581, "step": 177152 }, { "epoch": 1.01, "learning_rate": 4.709104256326968e-05, "loss": 4.4478, "step": 177664 }, { "epoch": 1.01, "learning_rate": 4.708265661575916e-05, "loss": 4.4604, "step": 178176 }, { "epoch": 1.01, "learning_rate": 4.707427066824864e-05, "loss": 4.4505, "step": 178688 }, { "epoch": 1.01, "learning_rate": 4.706588472073812e-05, "loss": 4.4595, "step": 179200 }, { "epoch": 1.01, "learning_rate": 4.70574987732276e-05, "loss": 4.4471, "step": 179712 }, { "epoch": 1.01, "learning_rate": 4.704911282571708e-05, "loss": 4.4516, "step": 180224 }, { "epoch": 1.01, "learning_rate": 4.70407432570103e-05, "loss": 4.4429, "step": 180736 }, { "epoch": 1.01, "learning_rate": 4.703235730949978e-05, "loss": 4.4565, "step": 181248 }, { "epoch": 1.01, "learning_rate": 4.702397136198926e-05, "loss": 4.4457, "step": 181760 }, { "epoch": 1.01, "learning_rate": 4.701558541447874e-05, "loss": 4.4454, "step": 182272 }, { "epoch": 1.01, "learning_rate": 4.700719946696822e-05, "loss": 4.4522, "step": 182784 }, { "epoch": 1.01, "learning_rate": 4.69988135194577e-05, "loss": 4.4395, "step": 183296 }, { "epoch": 1.01, "learning_rate": 4.6990443950750906e-05, "loss": 4.4345, "step": 183808 }, { "epoch": 1.01, "learning_rate": 4.6982058003240386e-05, "loss": 4.4566, "step": 184320 }, { "epoch": 1.01, "learning_rate": 4.6973672055729866e-05, "loss": 4.4232, "step": 184832 }, { "epoch": 1.01, "learning_rate": 4.6965302487023075e-05, "loss": 4.4286, "step": 185344 }, { "epoch": 1.01, "learning_rate": 4.6956916539512555e-05, "loss": 4.4376, "step": 185856 }, { "epoch": 1.01, "learning_rate": 4.6948530592002035e-05, "loss": 4.449, "step": 186368 }, { "epoch": 1.01, "learning_rate": 4.6940144644491515e-05, "loss": 4.4353, "step": 186880 }, { "epoch": 1.01, "learning_rate": 4.6931758696980995e-05, "loss": 4.4264, "step": 187392 }, { "epoch": 1.01, "learning_rate": 4.6923372749470475e-05, "loss": 4.4313, "step": 187904 }, { "epoch": 1.01, "learning_rate": 4.6914986801959955e-05, "loss": 4.4303, "step": 188416 }, { "epoch": 1.01, "learning_rate": 4.6906600854449435e-05, "loss": 4.4409, "step": 188928 }, { "epoch": 1.01, "learning_rate": 4.6898231285742644e-05, "loss": 4.4369, "step": 189440 }, { "epoch": 1.01, "learning_rate": 4.688986171703586e-05, "loss": 4.4348, "step": 189952 }, { "epoch": 1.01, "learning_rate": 4.688147576952534e-05, "loss": 4.4396, "step": 190464 }, { "epoch": 1.01, "learning_rate": 4.687308982201481e-05, "loss": 4.442, "step": 190976 }, { "epoch": 1.01, "learning_rate": 4.686470387450429e-05, "loss": 4.4226, "step": 191488 }, { "epoch": 1.01, "learning_rate": 4.685631792699377e-05, "loss": 4.4227, "step": 192000 }, { "epoch": 1.01, "learning_rate": 4.684793197948325e-05, "loss": 4.4332, "step": 192512 }, { "epoch": 1.01, "learning_rate": 4.683954603197274e-05, "loss": 4.4152, "step": 193024 }, { "epoch": 1.01, "learning_rate": 4.683116008446222e-05, "loss": 4.4318, "step": 193536 }, { "epoch": 1.01, "learning_rate": 4.68227741369517e-05, "loss": 4.4173, "step": 194048 }, { "epoch": 1.01, "learning_rate": 4.681438818944118e-05, "loss": 4.4375, "step": 194560 }, { "epoch": 1.01, "learning_rate": 4.680600224193066e-05, "loss": 4.4238, "step": 195072 }, { "epoch": 1.01, "learning_rate": 4.679763267322387e-05, "loss": 4.4116, "step": 195584 }, { "epoch": 1.01, "learning_rate": 4.678924672571335e-05, "loss": 4.416, "step": 196096 }, { "epoch": 1.01, "learning_rate": 4.678086077820283e-05, "loss": 4.4232, "step": 196608 }, { "epoch": 1.01, "learning_rate": 4.677247483069231e-05, "loss": 4.431, "step": 197120 }, { "epoch": 1.01, "learning_rate": 4.676408888318179e-05, "loss": 4.4172, "step": 197632 }, { "epoch": 1.01, "learning_rate": 4.6755719314475e-05, "loss": 4.4225, "step": 198144 }, { "epoch": 1.02, "learning_rate": 4.674733336696448e-05, "loss": 4.4125, "step": 198656 }, { "epoch": 1.02, "learning_rate": 4.673894741945396e-05, "loss": 4.4215, "step": 199168 }, { "epoch": 1.02, "learning_rate": 4.673057785074717e-05, "loss": 4.4132, "step": 199680 }, { "epoch": 1.02, "learning_rate": 4.672219190323665e-05, "loss": 4.4096, "step": 200192 }, { "epoch": 1.02, "learning_rate": 4.671380595572613e-05, "loss": 4.4203, "step": 200704 }, { "epoch": 1.02, "learning_rate": 4.670542000821561e-05, "loss": 4.4184, "step": 201216 }, { "epoch": 1.02, "learning_rate": 4.669703406070509e-05, "loss": 4.4177, "step": 201728 }, { "epoch": 1.02, "learning_rate": 4.668864811319457e-05, "loss": 4.4027, "step": 202240 }, { "epoch": 1.02, "learning_rate": 4.668026216568405e-05, "loss": 4.4058, "step": 202752 }, { "epoch": 1.02, "learning_rate": 4.667187621817353e-05, "loss": 4.4103, "step": 203264 }, { "epoch": 1.02, "learning_rate": 4.6663490270663006e-05, "loss": 4.4091, "step": 203776 }, { "epoch": 1.02, "learning_rate": 4.6655104323152486e-05, "loss": 4.4089, "step": 204288 }, { "epoch": 1.02, "learning_rate": 4.66467347544457e-05, "loss": 4.4003, "step": 204800 }, { "epoch": 1.02, "learning_rate": 4.6638348806935175e-05, "loss": 4.3996, "step": 205312 }, { "epoch": 1.02, "learning_rate": 4.662996285942466e-05, "loss": 4.4099, "step": 205824 }, { "epoch": 1.02, "learning_rate": 4.662157691191414e-05, "loss": 4.3817, "step": 206336 }, { "epoch": 1.02, "learning_rate": 4.661320734320736e-05, "loss": 4.4004, "step": 206848 }, { "epoch": 1.02, "learning_rate": 4.660482139569683e-05, "loss": 4.3916, "step": 207360 }, { "epoch": 1.02, "learning_rate": 4.659643544818631e-05, "loss": 4.4071, "step": 207872 }, { "epoch": 1.02, "learning_rate": 4.658804950067579e-05, "loss": 4.3957, "step": 208384 }, { "epoch": 1.02, "learning_rate": 4.6579679931969006e-05, "loss": 4.4037, "step": 208896 }, { "epoch": 1.02, "learning_rate": 4.657129398445848e-05, "loss": 4.3946, "step": 209408 }, { "epoch": 1.02, "learning_rate": 4.656290803694796e-05, "loss": 4.3927, "step": 209920 }, { "epoch": 1.02, "learning_rate": 4.655452208943744e-05, "loss": 4.4037, "step": 210432 }, { "epoch": 1.02, "learning_rate": 4.654613614192692e-05, "loss": 4.3914, "step": 210944 }, { "epoch": 1.02, "learning_rate": 4.65377501944164e-05, "loss": 4.404, "step": 211456 }, { "epoch": 1.02, "learning_rate": 4.652936424690588e-05, "loss": 4.3992, "step": 211968 }, { "epoch": 1.02, "learning_rate": 4.652097829939536e-05, "loss": 4.3985, "step": 212480 }, { "epoch": 1.02, "learning_rate": 4.6512608730688575e-05, "loss": 4.39, "step": 212992 }, { "epoch": 1.02, "learning_rate": 4.6504222783178055e-05, "loss": 4.3856, "step": 213504 }, { "epoch": 1.02, "learning_rate": 4.6495836835667535e-05, "loss": 4.4009, "step": 214016 }, { "epoch": 1.02, "learning_rate": 4.6487450888157015e-05, "loss": 4.3843, "step": 214528 }, { "epoch": 1.02, "learning_rate": 4.6479064940646495e-05, "loss": 4.3996, "step": 215040 }, { "epoch": 1.02, "learning_rate": 4.6470695371939704e-05, "loss": 4.3843, "step": 215552 }, { "epoch": 1.02, "learning_rate": 4.6462309424429184e-05, "loss": 4.3894, "step": 216064 }, { "epoch": 1.02, "learning_rate": 4.6453923476918664e-05, "loss": 4.389, "step": 216576 }, { "epoch": 1.02, "learning_rate": 4.6445537529408144e-05, "loss": 4.3852, "step": 217088 }, { "epoch": 1.02, "learning_rate": 4.6437151581897624e-05, "loss": 4.3666, "step": 217600 }, { "epoch": 1.02, "learning_rate": 4.6428765634387104e-05, "loss": 4.3846, "step": 218112 }, { "epoch": 1.02, "learning_rate": 4.6420379686876584e-05, "loss": 4.3722, "step": 218624 }, { "epoch": 1.02, "learning_rate": 4.64120101181698e-05, "loss": 4.3871, "step": 219136 }, { "epoch": 1.02, "learning_rate": 4.640362417065928e-05, "loss": 4.4007, "step": 219648 }, { "epoch": 1.02, "learning_rate": 4.639523822314876e-05, "loss": 4.3949, "step": 220160 }, { "epoch": 1.02, "learning_rate": 4.638685227563824e-05, "loss": 4.3641, "step": 220672 }, { "epoch": 1.02, "learning_rate": 4.637848270693145e-05, "loss": 4.3882, "step": 221184 }, { "epoch": 1.02, "learning_rate": 4.637009675942093e-05, "loss": 4.3827, "step": 221696 }, { "epoch": 1.02, "learning_rate": 4.636171081191041e-05, "loss": 4.3809, "step": 222208 }, { "epoch": 1.02, "learning_rate": 4.635332486439989e-05, "loss": 4.3706, "step": 222720 }, { "epoch": 1.02, "learning_rate": 4.634493891688936e-05, "loss": 4.3836, "step": 223232 }, { "epoch": 1.02, "learning_rate": 4.633655296937884e-05, "loss": 4.3772, "step": 223744 }, { "epoch": 1.02, "learning_rate": 4.632816702186832e-05, "loss": 4.3758, "step": 224256 }, { "epoch": 1.02, "learning_rate": 4.63197810743578e-05, "loss": 4.3782, "step": 224768 }, { "epoch": 1.02, "learning_rate": 4.631141150565102e-05, "loss": 4.3797, "step": 225280 }, { "epoch": 1.02, "learning_rate": 4.63030255581405e-05, "loss": 4.3755, "step": 225792 }, { "epoch": 1.02, "learning_rate": 4.629463961062998e-05, "loss": 4.3728, "step": 226304 }, { "epoch": 1.02, "learning_rate": 4.628625366311946e-05, "loss": 4.3783, "step": 226816 }, { "epoch": 1.02, "learning_rate": 4.627786771560894e-05, "loss": 4.3778, "step": 227328 }, { "epoch": 1.02, "learning_rate": 4.626948176809842e-05, "loss": 4.3668, "step": 227840 }, { "epoch": 1.02, "learning_rate": 4.62610958205879e-05, "loss": 4.3842, "step": 228352 }, { "epoch": 1.02, "learning_rate": 4.625270987307738e-05, "loss": 4.3687, "step": 228864 }, { "epoch": 1.03, "eval_loss": 4.369020462036133, "eval_runtime": 293.4838, "eval_samples_per_second": 1300.211, "eval_steps_per_second": 40.633, "step": 228960 }, { "epoch": 0.0, "learning_rate": 4.6244340304370586e-05, "loss": 4.3576, "step": 229376 }, { "epoch": 0.0, "learning_rate": 4.6235954356860066e-05, "loss": 4.3609, "step": 229888 }, { "epoch": 0.0, "learning_rate": 4.6227568409349546e-05, "loss": 4.3805, "step": 230400 }, { "epoch": 0.0, "learning_rate": 4.6219182461839026e-05, "loss": 4.3682, "step": 230912 }, { "epoch": 0.0, "learning_rate": 4.6210796514328506e-05, "loss": 4.3807, "step": 231424 }, { "epoch": 0.0, "learning_rate": 4.6202410566817986e-05, "loss": 4.3553, "step": 231936 }, { "epoch": 0.0, "learning_rate": 4.6194024619307466e-05, "loss": 4.3746, "step": 232448 }, { "epoch": 0.0, "learning_rate": 4.6185638671796946e-05, "loss": 4.3446, "step": 232960 }, { "epoch": 0.0, "learning_rate": 4.617725272428643e-05, "loss": 4.3762, "step": 233472 }, { "epoch": 0.0, "learning_rate": 4.6168866776775906e-05, "loss": 4.3568, "step": 233984 }, { "epoch": 0.0, "learning_rate": 4.6160480829265386e-05, "loss": 4.3721, "step": 234496 }, { "epoch": 0.0, "learning_rate": 4.6152094881754866e-05, "loss": 4.3707, "step": 235008 }, { "epoch": 0.0, "learning_rate": 4.614372531304808e-05, "loss": 4.3542, "step": 235520 }, { "epoch": 0.0, "learning_rate": 4.6135339365537555e-05, "loss": 4.3602, "step": 236032 }, { "epoch": 0.0, "learning_rate": 4.6126953418027035e-05, "loss": 4.3531, "step": 236544 }, { "epoch": 0.0, "learning_rate": 4.6118567470516515e-05, "loss": 4.3465, "step": 237056 }, { "epoch": 0.0, "learning_rate": 4.6110181523005995e-05, "loss": 4.3617, "step": 237568 }, { "epoch": 0.0, "learning_rate": 4.6101795575495475e-05, "loss": 4.3574, "step": 238080 }, { "epoch": 0.0, "learning_rate": 4.6093409627984955e-05, "loss": 4.3582, "step": 238592 }, { "epoch": 0.0, "learning_rate": 4.6085023680474435e-05, "loss": 4.366, "step": 239104 }, { "epoch": 0.0, "learning_rate": 4.607665411176765e-05, "loss": 4.3698, "step": 239616 }, { "epoch": 0.0, "learning_rate": 4.606826816425713e-05, "loss": 4.3607, "step": 240128 }, { "epoch": 0.0, "learning_rate": 4.605988221674661e-05, "loss": 4.3538, "step": 240640 }, { "epoch": 0.0, "learning_rate": 4.605151264803982e-05, "loss": 4.3653, "step": 241152 }, { "epoch": 0.0, "learning_rate": 4.60431267005293e-05, "loss": 4.3514, "step": 241664 }, { "epoch": 0.0, "learning_rate": 4.603474075301878e-05, "loss": 4.3636, "step": 242176 }, { "epoch": 0.0, "learning_rate": 4.602635480550826e-05, "loss": 4.3564, "step": 242688 }, { "epoch": 0.0, "learning_rate": 4.601796885799774e-05, "loss": 4.349, "step": 243200 }, { "epoch": 0.0, "learning_rate": 4.600958291048722e-05, "loss": 4.3438, "step": 243712 }, { "epoch": 0.01, "learning_rate": 4.60011969629767e-05, "loss": 4.3505, "step": 244224 }, { "epoch": 0.01, "learning_rate": 4.599281101546618e-05, "loss": 4.3466, "step": 244736 }, { "epoch": 0.01, "learning_rate": 4.598444144675939e-05, "loss": 4.3545, "step": 245248 }, { "epoch": 0.01, "learning_rate": 4.597605549924887e-05, "loss": 4.3544, "step": 245760 }, { "epoch": 0.01, "learning_rate": 4.596766955173835e-05, "loss": 4.3518, "step": 246272 }, { "epoch": 0.01, "learning_rate": 4.5959283604227835e-05, "loss": 4.3519, "step": 246784 }, { "epoch": 0.01, "learning_rate": 4.5950914035521044e-05, "loss": 4.3489, "step": 247296 }, { "epoch": 0.01, "learning_rate": 4.5942528088010524e-05, "loss": 4.3332, "step": 247808 }, { "epoch": 0.01, "learning_rate": 4.5934142140500004e-05, "loss": 4.3418, "step": 248320 }, { "epoch": 0.01, "learning_rate": 4.5925756192989484e-05, "loss": 4.3342, "step": 248832 }, { "epoch": 0.01, "learning_rate": 4.5917370245478964e-05, "loss": 4.3396, "step": 249344 }, { "epoch": 0.01, "learning_rate": 4.590898429796844e-05, "loss": 4.3418, "step": 249856 }, { "epoch": 0.01, "learning_rate": 4.590059835045792e-05, "loss": 4.3453, "step": 250368 }, { "epoch": 0.01, "learning_rate": 4.589222878175113e-05, "loss": 4.3412, "step": 250880 }, { "epoch": 0.01, "learning_rate": 4.588384283424061e-05, "loss": 4.3325, "step": 251392 }, { "epoch": 0.01, "learning_rate": 4.587545688673009e-05, "loss": 4.345, "step": 251904 }, { "epoch": 0.01, "learning_rate": 4.586707093921957e-05, "loss": 4.3431, "step": 252416 }, { "epoch": 0.01, "learning_rate": 4.585868499170905e-05, "loss": 4.3468, "step": 252928 }, { "epoch": 0.01, "learning_rate": 4.585029904419853e-05, "loss": 4.3382, "step": 253440 }, { "epoch": 0.01, "learning_rate": 4.584192947549174e-05, "loss": 4.326, "step": 253952 }, { "epoch": 0.01, "learning_rate": 4.583354352798122e-05, "loss": 4.3409, "step": 254464 }, { "epoch": 0.01, "learning_rate": 4.582517395927444e-05, "loss": 4.3303, "step": 254976 }, { "epoch": 0.01, "learning_rate": 4.581678801176391e-05, "loss": 4.3389, "step": 255488 }, { "epoch": 0.01, "learning_rate": 4.580840206425339e-05, "loss": 4.3287, "step": 256000 }, { "epoch": 0.01, "learning_rate": 4.580001611674287e-05, "loss": 4.3275, "step": 256512 }, { "epoch": 0.01, "learning_rate": 4.579163016923235e-05, "loss": 4.3279, "step": 257024 }, { "epoch": 0.01, "learning_rate": 4.578324422172183e-05, "loss": 4.3339, "step": 257536 }, { "epoch": 0.01, "learning_rate": 4.577485827421131e-05, "loss": 4.3302, "step": 258048 }, { "epoch": 0.01, "learning_rate": 4.576647232670079e-05, "loss": 4.3315, "step": 258560 }, { "epoch": 0.01, "learning_rate": 4.575808637919027e-05, "loss": 4.3302, "step": 259072 }, { "epoch": 0.01, "learning_rate": 4.574970043167976e-05, "loss": 4.322, "step": 259584 }, { "epoch": 0.01, "learning_rate": 4.5741330862972966e-05, "loss": 4.321, "step": 260096 }, { "epoch": 0.01, "learning_rate": 4.5732944915462446e-05, "loss": 4.3418, "step": 260608 }, { "epoch": 0.01, "learning_rate": 4.5724558967951926e-05, "loss": 4.3103, "step": 261120 }, { "epoch": 0.01, "learning_rate": 4.5716173020441406e-05, "loss": 4.3121, "step": 261632 }, { "epoch": 0.01, "learning_rate": 4.5707787072930886e-05, "loss": 4.3189, "step": 262144 }, { "epoch": 0.01, "learning_rate": 4.5699401125420366e-05, "loss": 4.3357, "step": 262656 }, { "epoch": 0.01, "learning_rate": 4.5691015177909846e-05, "loss": 4.3204, "step": 263168 }, { "epoch": 0.01, "learning_rate": 4.5682629230399326e-05, "loss": 4.3188, "step": 263680 }, { "epoch": 0.01, "learning_rate": 4.5674259661692535e-05, "loss": 4.3111, "step": 264192 }, { "epoch": 0.01, "learning_rate": 4.5665873714182015e-05, "loss": 4.3174, "step": 264704 }, { "epoch": 0.01, "learning_rate": 4.565752052427896e-05, "loss": 4.3283, "step": 265216 }, { "epoch": 0.01, "learning_rate": 4.564913457676844e-05, "loss": 4.3217, "step": 265728 }, { "epoch": 0.01, "learning_rate": 4.564074862925792e-05, "loss": 4.3201, "step": 266240 }, { "epoch": 0.01, "learning_rate": 4.56323626817474e-05, "loss": 4.3314, "step": 266752 }, { "epoch": 0.01, "learning_rate": 4.562397673423688e-05, "loss": 4.3293, "step": 267264 }, { "epoch": 0.01, "learning_rate": 4.561559078672636e-05, "loss": 4.3104, "step": 267776 }, { "epoch": 0.01, "learning_rate": 4.560720483921584e-05, "loss": 4.3133, "step": 268288 }, { "epoch": 0.01, "learning_rate": 4.559881889170532e-05, "loss": 4.3262, "step": 268800 }, { "epoch": 0.01, "learning_rate": 4.55904329441948e-05, "loss": 4.3008, "step": 269312 }, { "epoch": 0.01, "learning_rate": 4.558204699668427e-05, "loss": 4.3211, "step": 269824 }, { "epoch": 0.01, "learning_rate": 4.557367742797749e-05, "loss": 4.3068, "step": 270336 }, { "epoch": 0.01, "learning_rate": 4.556529148046697e-05, "loss": 4.3291, "step": 270848 }, { "epoch": 0.01, "learning_rate": 4.555690553295645e-05, "loss": 4.3139, "step": 271360 }, { "epoch": 0.01, "learning_rate": 4.554851958544593e-05, "loss": 4.3049, "step": 271872 }, { "epoch": 0.01, "learning_rate": 4.554013363793541e-05, "loss": 4.3034, "step": 272384 }, { "epoch": 0.01, "learning_rate": 4.553174769042489e-05, "loss": 4.3183, "step": 272896 }, { "epoch": 0.01, "learning_rate": 4.552336174291437e-05, "loss": 4.3232, "step": 273408 }, { "epoch": 0.01, "learning_rate": 4.551499217420758e-05, "loss": 4.3123, "step": 273920 }, { "epoch": 0.01, "learning_rate": 4.550660622669706e-05, "loss": 4.3112, "step": 274432 }, { "epoch": 0.02, "learning_rate": 4.549822027918654e-05, "loss": 4.3046, "step": 274944 }, { "epoch": 0.02, "learning_rate": 4.548983433167602e-05, "loss": 4.3163, "step": 275456 }, { "epoch": 0.02, "learning_rate": 4.5481464762969226e-05, "loss": 4.3062, "step": 275968 }, { "epoch": 0.02, "learning_rate": 4.547309519426244e-05, "loss": 4.3058, "step": 276480 }, { "epoch": 0.02, "learning_rate": 4.546470924675192e-05, "loss": 4.3148, "step": 276992 }, { "epoch": 0.02, "learning_rate": 4.54563232992414e-05, "loss": 4.3134, "step": 277504 }, { "epoch": 0.02, "learning_rate": 4.544793735173088e-05, "loss": 4.3097, "step": 278016 }, { "epoch": 0.02, "learning_rate": 4.543955140422036e-05, "loss": 4.2963, "step": 278528 }, { "epoch": 0.02, "learning_rate": 4.543116545670984e-05, "loss": 4.3048, "step": 279040 }, { "epoch": 0.02, "learning_rate": 4.542277950919932e-05, "loss": 4.3037, "step": 279552 }, { "epoch": 0.02, "learning_rate": 4.54143935616888e-05, "loss": 4.3065, "step": 280064 }, { "epoch": 0.02, "learning_rate": 4.540602399298201e-05, "loss": 4.3049, "step": 280576 }, { "epoch": 0.02, "learning_rate": 4.539763804547149e-05, "loss": 4.2962, "step": 281088 }, { "epoch": 0.02, "learning_rate": 4.538925209796097e-05, "loss": 4.3005, "step": 281600 }, { "epoch": 0.02, "learning_rate": 4.538086615045045e-05, "loss": 4.3052, "step": 282112 }, { "epoch": 0.02, "learning_rate": 4.537248020293993e-05, "loss": 4.2811, "step": 282624 }, { "epoch": 0.02, "learning_rate": 4.536409425542941e-05, "loss": 4.3005, "step": 283136 }, { "epoch": 0.02, "learning_rate": 4.535572468672262e-05, "loss": 4.2857, "step": 283648 }, { "epoch": 0.02, "learning_rate": 4.53473387392121e-05, "loss": 4.3038, "step": 284160 }, { "epoch": 0.02, "learning_rate": 4.533895279170158e-05, "loss": 4.2959, "step": 284672 }, { "epoch": 0.02, "learning_rate": 4.533056684419107e-05, "loss": 4.2995, "step": 285184 }, { "epoch": 0.02, "learning_rate": 4.5322197275484276e-05, "loss": 4.2916, "step": 285696 }, { "epoch": 0.02, "learning_rate": 4.5313811327973756e-05, "loss": 4.2954, "step": 286208 }, { "epoch": 0.02, "learning_rate": 4.5305425380463236e-05, "loss": 4.3044, "step": 286720 }, { "epoch": 0.02, "learning_rate": 4.5297039432952716e-05, "loss": 4.288, "step": 287232 }, { "epoch": 0.02, "learning_rate": 4.5288669864245925e-05, "loss": 4.3001, "step": 287744 }, { "epoch": 0.02, "learning_rate": 4.5280283916735405e-05, "loss": 4.3, "step": 288256 }, { "epoch": 0.02, "learning_rate": 4.5271914348028614e-05, "loss": 4.2982, "step": 288768 }, { "epoch": 0.02, "learning_rate": 4.5263528400518094e-05, "loss": 4.2998, "step": 289280 }, { "epoch": 0.02, "learning_rate": 4.5255142453007574e-05, "loss": 4.283, "step": 289792 }, { "epoch": 0.02, "learning_rate": 4.5246756505497054e-05, "loss": 4.3028, "step": 290304 }, { "epoch": 0.02, "learning_rate": 4.5238370557986534e-05, "loss": 4.2891, "step": 290816 }, { "epoch": 0.02, "learning_rate": 4.522998461047602e-05, "loss": 4.2976, "step": 291328 }, { "epoch": 0.02, "learning_rate": 4.52215986629655e-05, "loss": 4.2863, "step": 291840 }, { "epoch": 0.02, "learning_rate": 4.521321271545498e-05, "loss": 4.2921, "step": 292352 }, { "epoch": 0.02, "learning_rate": 4.520484314674819e-05, "loss": 4.2905, "step": 292864 }, { "epoch": 0.02, "learning_rate": 4.519645719923767e-05, "loss": 4.2857, "step": 293376 }, { "epoch": 0.02, "learning_rate": 4.518807125172715e-05, "loss": 4.2732, "step": 293888 }, { "epoch": 0.02, "learning_rate": 4.517968530421663e-05, "loss": 4.2849, "step": 294400 }, { "epoch": 0.02, "learning_rate": 4.517129935670611e-05, "loss": 4.2796, "step": 294912 }, { "epoch": 0.02, "learning_rate": 4.516291340919558e-05, "loss": 4.2903, "step": 295424 }, { "epoch": 0.02, "learning_rate": 4.515452746168506e-05, "loss": 4.3007, "step": 295936 }, { "epoch": 0.02, "learning_rate": 4.514615789297828e-05, "loss": 4.2982, "step": 296448 }, { "epoch": 0.02, "learning_rate": 4.513777194546776e-05, "loss": 4.27, "step": 296960 }, { "epoch": 0.02, "learning_rate": 4.512938599795724e-05, "loss": 4.2952, "step": 297472 }, { "epoch": 0.02, "learning_rate": 4.512100005044672e-05, "loss": 4.286, "step": 297984 }, { "epoch": 0.02, "learning_rate": 4.51126141029362e-05, "loss": 4.2854, "step": 298496 }, { "epoch": 0.02, "learning_rate": 4.510422815542568e-05, "loss": 4.2795, "step": 299008 }, { "epoch": 0.02, "learning_rate": 4.509584220791516e-05, "loss": 4.2886, "step": 299520 }, { "epoch": 0.02, "learning_rate": 4.508745626040464e-05, "loss": 4.2796, "step": 300032 }, { "epoch": 0.02, "learning_rate": 4.507908669169785e-05, "loss": 4.2879, "step": 300544 }, { "epoch": 0.02, "learning_rate": 4.507070074418733e-05, "loss": 4.2809, "step": 301056 }, { "epoch": 0.02, "learning_rate": 4.5062331175480536e-05, "loss": 4.2842, "step": 301568 }, { "epoch": 0.02, "learning_rate": 4.5053945227970016e-05, "loss": 4.2868, "step": 302080 }, { "epoch": 0.02, "learning_rate": 4.5045559280459496e-05, "loss": 4.2788, "step": 302592 }, { "epoch": 0.02, "learning_rate": 4.5037173332948976e-05, "loss": 4.2823, "step": 303104 }, { "epoch": 0.02, "learning_rate": 4.502880376424219e-05, "loss": 4.2913, "step": 303616 }, { "epoch": 0.02, "learning_rate": 4.502041781673167e-05, "loss": 4.2731, "step": 304128 }, { "epoch": 0.02, "learning_rate": 4.501203186922115e-05, "loss": 4.2894, "step": 304640 }, { "epoch": 0.02, "learning_rate": 4.500364592171063e-05, "loss": 4.282, "step": 305152 }, { "epoch": 0.03, "eval_loss": 4.2855424880981445, "eval_runtime": 283.2935, "eval_samples_per_second": 1346.981, "eval_steps_per_second": 42.094, "step": 305280 } ], "logging_steps": 512, "max_steps": 3052726, "num_train_epochs": 9223372036854775807, "save_steps": 10, "total_flos": 1.2223819719679363e+17, "trial_name": null, "trial_params": null }