diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,65844 @@ +{ + "best_global_step": 31600, + "best_metric": 0.31366169452667236, + "best_model_checkpoint": "saves/prompt-tuning/mistral-7b-instruct/train_multirc_1745950268/checkpoint-31600", + "epoch": 6.525328330206379, + "eval_steps": 200, + "global_step": 40000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0008157272208173587, + "grad_norm": 0.5166723132133484, + "learning_rate": 0.29999999259779675, + "loss": 3.0874, + "num_input_tokens_seen": 8784, + "step": 5 + }, + { + "epoch": 0.0016314544416347174, + "grad_norm": 0.03360513597726822, + "learning_rate": 0.29999996252634736, + "loss": 0.6102, + "num_input_tokens_seen": 18528, + "step": 10 + }, + { + "epoch": 0.002447181662452076, + "grad_norm": 0.014941602945327759, + "learning_rate": 0.2999999093230187, + "loss": 0.384, + "num_input_tokens_seen": 29168, + "step": 15 + }, + { + "epoch": 0.0032629088832694347, + "grad_norm": 0.025340773165225983, + "learning_rate": 0.299999832987819, + "loss": 0.3615, + "num_input_tokens_seen": 40624, + "step": 20 + }, + { + "epoch": 0.004078636104086793, + "grad_norm": 0.00806356966495514, + "learning_rate": 0.29999973352076004, + "loss": 0.4024, + "num_input_tokens_seen": 49968, + "step": 25 + }, + { + "epoch": 0.004894363324904152, + "grad_norm": 0.00269055413082242, + "learning_rate": 0.2999996109218572, + "loss": 0.3496, + "num_input_tokens_seen": 60672, + "step": 30 + }, + { + "epoch": 0.005710090545721511, + "grad_norm": 0.013894874602556229, + "learning_rate": 0.2999994651911293, + "loss": 0.3796, + "num_input_tokens_seen": 71408, + "step": 35 + }, + { + "epoch": 0.0065258177665388694, + "grad_norm": 0.0029685748741030693, + "learning_rate": 0.2999992963285989, + "loss": 0.3484, + "num_input_tokens_seen": 82752, + "step": 40 + }, + { + "epoch": 0.007341544987356228, + "grad_norm": 0.008195328526198864, + "learning_rate": 0.29999910433429194, + "loss": 0.3045, + "num_input_tokens_seen": 93296, + "step": 45 + }, + { + "epoch": 0.008157272208173586, + "grad_norm": 0.004941473249346018, + "learning_rate": 0.29999888920823814, + "loss": 0.3677, + "num_input_tokens_seen": 104592, + "step": 50 + }, + { + "epoch": 0.008972999428990946, + "grad_norm": 0.008505582809448242, + "learning_rate": 0.29999865095047057, + "loss": 0.4096, + "num_input_tokens_seen": 114976, + "step": 55 + }, + { + "epoch": 0.009788726649808304, + "grad_norm": 0.0014015829656273127, + "learning_rate": 0.29999838956102604, + "loss": 0.3382, + "num_input_tokens_seen": 124944, + "step": 60 + }, + { + "epoch": 0.010604453870625663, + "grad_norm": 0.010865691117942333, + "learning_rate": 0.29999810503994484, + "loss": 0.3542, + "num_input_tokens_seen": 134560, + "step": 65 + }, + { + "epoch": 0.011420181091443021, + "grad_norm": 0.010377690196037292, + "learning_rate": 0.29999779738727084, + "loss": 0.3557, + "num_input_tokens_seen": 144528, + "step": 70 + }, + { + "epoch": 0.012235908312260381, + "grad_norm": 0.005796477198600769, + "learning_rate": 0.29999746660305154, + "loss": 0.3552, + "num_input_tokens_seen": 155056, + "step": 75 + }, + { + "epoch": 0.013051635533077739, + "grad_norm": 0.001450508483685553, + "learning_rate": 0.2999971126873379, + "loss": 0.3471, + "num_input_tokens_seen": 164464, + "step": 80 + }, + { + "epoch": 0.013867362753895097, + "grad_norm": 0.004668163135647774, + "learning_rate": 0.2999967356401845, + "loss": 0.3535, + "num_input_tokens_seen": 174608, + "step": 85 + }, + { + "epoch": 0.014683089974712456, + "grad_norm": 0.003919953014701605, + "learning_rate": 0.29999633546164944, + "loss": 0.3549, + "num_input_tokens_seen": 185616, + "step": 90 + }, + { + "epoch": 0.015498817195529814, + "grad_norm": 0.0023572074715048075, + "learning_rate": 0.29999591215179444, + "loss": 0.3324, + "num_input_tokens_seen": 195360, + "step": 95 + }, + { + "epoch": 0.016314544416347172, + "grad_norm": 0.002270733704790473, + "learning_rate": 0.2999954657106849, + "loss": 0.3662, + "num_input_tokens_seen": 207152, + "step": 100 + }, + { + "epoch": 0.017130271637164532, + "grad_norm": 0.003650523256510496, + "learning_rate": 0.2999949961383896, + "loss": 0.3417, + "num_input_tokens_seen": 218016, + "step": 105 + }, + { + "epoch": 0.01794599885798189, + "grad_norm": 0.006146328989416361, + "learning_rate": 0.2999945034349809, + "loss": 0.3507, + "num_input_tokens_seen": 229680, + "step": 110 + }, + { + "epoch": 0.01876172607879925, + "grad_norm": 0.0061322348192334175, + "learning_rate": 0.2999939876005348, + "loss": 0.3454, + "num_input_tokens_seen": 238624, + "step": 115 + }, + { + "epoch": 0.019577453299616607, + "grad_norm": 0.0018613430438563228, + "learning_rate": 0.29999344863513094, + "loss": 0.3474, + "num_input_tokens_seen": 248592, + "step": 120 + }, + { + "epoch": 0.020393180520433967, + "grad_norm": 0.0018143381457775831, + "learning_rate": 0.2999928865388523, + "loss": 0.3555, + "num_input_tokens_seen": 259392, + "step": 125 + }, + { + "epoch": 0.021208907741251327, + "grad_norm": 0.0040846229530870914, + "learning_rate": 0.29999230131178567, + "loss": 0.3349, + "num_input_tokens_seen": 269664, + "step": 130 + }, + { + "epoch": 0.022024634962068683, + "grad_norm": 0.0020152025390416384, + "learning_rate": 0.2999916929540212, + "loss": 0.347, + "num_input_tokens_seen": 280800, + "step": 135 + }, + { + "epoch": 0.022840362182886043, + "grad_norm": 0.012577286921441555, + "learning_rate": 0.29999106146565285, + "loss": 0.3788, + "num_input_tokens_seen": 290736, + "step": 140 + }, + { + "epoch": 0.023656089403703402, + "grad_norm": 0.004239683039486408, + "learning_rate": 0.29999040684677786, + "loss": 0.3515, + "num_input_tokens_seen": 301152, + "step": 145 + }, + { + "epoch": 0.024471816624520762, + "grad_norm": 0.004605093039572239, + "learning_rate": 0.2999897290974972, + "loss": 0.3513, + "num_input_tokens_seen": 311040, + "step": 150 + }, + { + "epoch": 0.025287543845338118, + "grad_norm": 0.0015230309218168259, + "learning_rate": 0.2999890282179155, + "loss": 0.3515, + "num_input_tokens_seen": 322512, + "step": 155 + }, + { + "epoch": 0.026103271066155478, + "grad_norm": 0.008101596496999264, + "learning_rate": 0.29998830420814077, + "loss": 0.3459, + "num_input_tokens_seen": 333664, + "step": 160 + }, + { + "epoch": 0.026918998286972837, + "grad_norm": 0.00374515145085752, + "learning_rate": 0.2999875570682846, + "loss": 0.3384, + "num_input_tokens_seen": 345104, + "step": 165 + }, + { + "epoch": 0.027734725507790194, + "grad_norm": 0.0019926894456148148, + "learning_rate": 0.2999867867984623, + "loss": 0.3698, + "num_input_tokens_seen": 356288, + "step": 170 + }, + { + "epoch": 0.028550452728607553, + "grad_norm": 0.008918768726289272, + "learning_rate": 0.29998599339879267, + "loss": 0.358, + "num_input_tokens_seen": 365744, + "step": 175 + }, + { + "epoch": 0.029366179949424913, + "grad_norm": 0.007957767695188522, + "learning_rate": 0.29998517686939796, + "loss": 0.3428, + "num_input_tokens_seen": 376080, + "step": 180 + }, + { + "epoch": 0.030181907170242273, + "grad_norm": 0.0026493584737181664, + "learning_rate": 0.29998433721040413, + "loss": 0.338, + "num_input_tokens_seen": 386784, + "step": 185 + }, + { + "epoch": 0.03099763439105963, + "grad_norm": 0.0036443392746150494, + "learning_rate": 0.29998347442194073, + "loss": 0.3547, + "num_input_tokens_seen": 396944, + "step": 190 + }, + { + "epoch": 0.031813361611876985, + "grad_norm": 0.006455554161220789, + "learning_rate": 0.2999825885041407, + "loss": 0.304, + "num_input_tokens_seen": 407952, + "step": 195 + }, + { + "epoch": 0.032629088832694345, + "grad_norm": 0.007252972107380629, + "learning_rate": 0.29998167945714077, + "loss": 0.3351, + "num_input_tokens_seen": 418192, + "step": 200 + }, + { + "epoch": 0.032629088832694345, + "eval_loss": 0.34499409794807434, + "eval_runtime": 154.412, + "eval_samples_per_second": 17.648, + "eval_steps_per_second": 8.827, + "num_input_tokens_seen": 418192, + "step": 200 + }, + { + "epoch": 0.033444816053511704, + "grad_norm": 0.002480786992236972, + "learning_rate": 0.2999807472810811, + "loss": 0.3672, + "num_input_tokens_seen": 429056, + "step": 205 + }, + { + "epoch": 0.034260543274329064, + "grad_norm": 0.0018027732148766518, + "learning_rate": 0.29997979197610536, + "loss": 0.3329, + "num_input_tokens_seen": 440768, + "step": 210 + }, + { + "epoch": 0.035076270495146424, + "grad_norm": 0.0031286764424294233, + "learning_rate": 0.299978813542361, + "loss": 0.3204, + "num_input_tokens_seen": 451120, + "step": 215 + }, + { + "epoch": 0.03589199771596378, + "grad_norm": 0.0019084375817328691, + "learning_rate": 0.2999778119799988, + "loss": 0.3945, + "num_input_tokens_seen": 463056, + "step": 220 + }, + { + "epoch": 0.03670772493678114, + "grad_norm": 0.0015596814919263124, + "learning_rate": 0.29997678728917326, + "loss": 0.3569, + "num_input_tokens_seen": 474048, + "step": 225 + }, + { + "epoch": 0.0375234521575985, + "grad_norm": 0.008686738088726997, + "learning_rate": 0.2999757394700424, + "loss": 0.3642, + "num_input_tokens_seen": 483968, + "step": 230 + }, + { + "epoch": 0.038339179378415855, + "grad_norm": 0.005526532884687185, + "learning_rate": 0.29997466852276783, + "loss": 0.3606, + "num_input_tokens_seen": 494032, + "step": 235 + }, + { + "epoch": 0.039154906599233215, + "grad_norm": 0.0019534719176590443, + "learning_rate": 0.29997357444751466, + "loss": 0.3581, + "num_input_tokens_seen": 505168, + "step": 240 + }, + { + "epoch": 0.039970633820050575, + "grad_norm": 0.001520688645541668, + "learning_rate": 0.2999724572444516, + "loss": 0.376, + "num_input_tokens_seen": 514768, + "step": 245 + }, + { + "epoch": 0.040786361040867934, + "grad_norm": 0.006032690405845642, + "learning_rate": 0.29997131691375095, + "loss": 0.3424, + "num_input_tokens_seen": 525776, + "step": 250 + }, + { + "epoch": 0.041602088261685294, + "grad_norm": 0.005076623056083918, + "learning_rate": 0.2999701534555886, + "loss": 0.3729, + "num_input_tokens_seen": 535216, + "step": 255 + }, + { + "epoch": 0.042417815482502653, + "grad_norm": 0.005316622089594603, + "learning_rate": 0.2999689668701439, + "loss": 0.313, + "num_input_tokens_seen": 544864, + "step": 260 + }, + { + "epoch": 0.04323354270332001, + "grad_norm": 0.0019301895517855883, + "learning_rate": 0.29996775715759993, + "loss": 0.3478, + "num_input_tokens_seen": 556624, + "step": 265 + }, + { + "epoch": 0.044049269924137366, + "grad_norm": 0.0012865029275417328, + "learning_rate": 0.2999665243181432, + "loss": 0.3922, + "num_input_tokens_seen": 566320, + "step": 270 + }, + { + "epoch": 0.044864997144954726, + "grad_norm": 0.003332223743200302, + "learning_rate": 0.2999652683519638, + "loss": 0.3535, + "num_input_tokens_seen": 577312, + "step": 275 + }, + { + "epoch": 0.045680724365772085, + "grad_norm": 0.0012981403851881623, + "learning_rate": 0.29996398925925544, + "loss": 0.3459, + "num_input_tokens_seen": 587840, + "step": 280 + }, + { + "epoch": 0.046496451586589445, + "grad_norm": 0.004099343903362751, + "learning_rate": 0.2999626870402154, + "loss": 0.3752, + "num_input_tokens_seen": 598224, + "step": 285 + }, + { + "epoch": 0.047312178807406804, + "grad_norm": 0.0025638167280703783, + "learning_rate": 0.29996136169504445, + "loss": 0.3337, + "num_input_tokens_seen": 607856, + "step": 290 + }, + { + "epoch": 0.048127906028224164, + "grad_norm": 0.00537355849519372, + "learning_rate": 0.29996001322394694, + "loss": 0.3513, + "num_input_tokens_seen": 618656, + "step": 295 + }, + { + "epoch": 0.048943633249041524, + "grad_norm": 0.003202238352969289, + "learning_rate": 0.29995864162713093, + "loss": 0.3458, + "num_input_tokens_seen": 629056, + "step": 300 + }, + { + "epoch": 0.049759360469858877, + "grad_norm": 0.002908246126025915, + "learning_rate": 0.2999572469048079, + "loss": 0.347, + "num_input_tokens_seen": 641280, + "step": 305 + }, + { + "epoch": 0.050575087690676236, + "grad_norm": 0.003173286095261574, + "learning_rate": 0.29995582905719287, + "loss": 0.3544, + "num_input_tokens_seen": 650992, + "step": 310 + }, + { + "epoch": 0.051390814911493596, + "grad_norm": 0.002905767410993576, + "learning_rate": 0.2999543880845046, + "loss": 0.3482, + "num_input_tokens_seen": 662112, + "step": 315 + }, + { + "epoch": 0.052206542132310955, + "grad_norm": 0.0025589221622794867, + "learning_rate": 0.2999529239869652, + "loss": 0.3347, + "num_input_tokens_seen": 672912, + "step": 320 + }, + { + "epoch": 0.053022269353128315, + "grad_norm": 0.0014326254604384303, + "learning_rate": 0.2999514367648005, + "loss": 0.35, + "num_input_tokens_seen": 682864, + "step": 325 + }, + { + "epoch": 0.053837996573945675, + "grad_norm": 0.00595319177955389, + "learning_rate": 0.29994992641823987, + "loss": 0.319, + "num_input_tokens_seen": 692480, + "step": 330 + }, + { + "epoch": 0.054653723794763034, + "grad_norm": 0.0018536720890551805, + "learning_rate": 0.29994839294751613, + "loss": 0.3729, + "num_input_tokens_seen": 703312, + "step": 335 + }, + { + "epoch": 0.05546945101558039, + "grad_norm": 0.001526648411527276, + "learning_rate": 0.29994683635286584, + "loss": 0.3598, + "num_input_tokens_seen": 713744, + "step": 340 + }, + { + "epoch": 0.05628517823639775, + "grad_norm": 0.002793884137645364, + "learning_rate": 0.2999452566345291, + "loss": 0.3379, + "num_input_tokens_seen": 722880, + "step": 345 + }, + { + "epoch": 0.057100905457215106, + "grad_norm": 0.0052211349830031395, + "learning_rate": 0.2999436537927494, + "loss": 0.3276, + "num_input_tokens_seen": 732912, + "step": 350 + }, + { + "epoch": 0.057916632678032466, + "grad_norm": 0.0016644001007080078, + "learning_rate": 0.299942027827774, + "loss": 0.3163, + "num_input_tokens_seen": 744640, + "step": 355 + }, + { + "epoch": 0.058732359898849826, + "grad_norm": 0.004528427962213755, + "learning_rate": 0.29994037873985363, + "loss": 0.4466, + "num_input_tokens_seen": 754432, + "step": 360 + }, + { + "epoch": 0.059548087119667185, + "grad_norm": 0.0013441117480397224, + "learning_rate": 0.29993870652924254, + "loss": 0.344, + "num_input_tokens_seen": 766304, + "step": 365 + }, + { + "epoch": 0.060363814340484545, + "grad_norm": 0.002886069007217884, + "learning_rate": 0.29993701119619876, + "loss": 0.3498, + "num_input_tokens_seen": 777072, + "step": 370 + }, + { + "epoch": 0.0611795415613019, + "grad_norm": 0.005336853675544262, + "learning_rate": 0.2999352927409835, + "loss": 0.3518, + "num_input_tokens_seen": 787808, + "step": 375 + }, + { + "epoch": 0.06199526878211926, + "grad_norm": 0.0022375225089490414, + "learning_rate": 0.29993355116386194, + "loss": 0.3372, + "num_input_tokens_seen": 798016, + "step": 380 + }, + { + "epoch": 0.06281099600293662, + "grad_norm": 0.0018488684436306357, + "learning_rate": 0.29993178646510266, + "loss": 0.3274, + "num_input_tokens_seen": 807632, + "step": 385 + }, + { + "epoch": 0.06362672322375397, + "grad_norm": 0.003834428731352091, + "learning_rate": 0.2999299986449777, + "loss": 0.3583, + "num_input_tokens_seen": 816608, + "step": 390 + }, + { + "epoch": 0.06444245044457134, + "grad_norm": 0.0045864456333220005, + "learning_rate": 0.29992818770376284, + "loss": 0.3957, + "num_input_tokens_seen": 826208, + "step": 395 + }, + { + "epoch": 0.06525817766538869, + "grad_norm": 0.001193505828268826, + "learning_rate": 0.29992635364173725, + "loss": 0.342, + "num_input_tokens_seen": 836224, + "step": 400 + }, + { + "epoch": 0.06525817766538869, + "eval_loss": 0.3405257761478424, + "eval_runtime": 154.2469, + "eval_samples_per_second": 17.666, + "eval_steps_per_second": 8.836, + "num_input_tokens_seen": 836224, + "step": 400 + }, + { + "epoch": 0.06607390488620606, + "grad_norm": 0.005433475133031607, + "learning_rate": 0.2999244964591839, + "loss": 0.3574, + "num_input_tokens_seen": 846160, + "step": 405 + }, + { + "epoch": 0.06688963210702341, + "grad_norm": 0.0040536620654165745, + "learning_rate": 0.2999226161563891, + "loss": 0.3558, + "num_input_tokens_seen": 856608, + "step": 410 + }, + { + "epoch": 0.06770535932784078, + "grad_norm": 0.0035918112844228745, + "learning_rate": 0.2999207127336429, + "loss": 0.3706, + "num_input_tokens_seen": 867568, + "step": 415 + }, + { + "epoch": 0.06852108654865813, + "grad_norm": 0.0010855558793991804, + "learning_rate": 0.2999187861912387, + "loss": 0.344, + "num_input_tokens_seen": 878496, + "step": 420 + }, + { + "epoch": 0.06933681376947548, + "grad_norm": 0.0011223707115277648, + "learning_rate": 0.2999168365294737, + "loss": 0.3449, + "num_input_tokens_seen": 888784, + "step": 425 + }, + { + "epoch": 0.07015254099029285, + "grad_norm": 0.0015563002089038491, + "learning_rate": 0.29991486374864856, + "loss": 0.3238, + "num_input_tokens_seen": 899392, + "step": 430 + }, + { + "epoch": 0.0709682682111102, + "grad_norm": 0.0014149852795526385, + "learning_rate": 0.29991286784906745, + "loss": 0.3601, + "num_input_tokens_seen": 910112, + "step": 435 + }, + { + "epoch": 0.07178399543192757, + "grad_norm": 0.09900538623332977, + "learning_rate": 0.2999108488310382, + "loss": 0.7327, + "num_input_tokens_seen": 920000, + "step": 440 + }, + { + "epoch": 0.07259972265274492, + "grad_norm": 0.040844470262527466, + "learning_rate": 0.29990880669487213, + "loss": 0.3943, + "num_input_tokens_seen": 930064, + "step": 445 + }, + { + "epoch": 0.07341544987356229, + "grad_norm": 0.01378296222537756, + "learning_rate": 0.29990674144088425, + "loss": 0.3827, + "num_input_tokens_seen": 940512, + "step": 450 + }, + { + "epoch": 0.07423117709437964, + "grad_norm": 0.007743575610220432, + "learning_rate": 0.299904653069393, + "loss": 0.3837, + "num_input_tokens_seen": 951472, + "step": 455 + }, + { + "epoch": 0.075046904315197, + "grad_norm": 0.00873705092817545, + "learning_rate": 0.29990254158072044, + "loss": 0.415, + "num_input_tokens_seen": 961456, + "step": 460 + }, + { + "epoch": 0.07586263153601436, + "grad_norm": 0.00367817934602499, + "learning_rate": 0.2999004069751921, + "loss": 0.3487, + "num_input_tokens_seen": 972896, + "step": 465 + }, + { + "epoch": 0.07667835875683171, + "grad_norm": 0.005216001998633146, + "learning_rate": 0.2998982492531373, + "loss": 0.3538, + "num_input_tokens_seen": 984560, + "step": 470 + }, + { + "epoch": 0.07749408597764908, + "grad_norm": 0.007017599418759346, + "learning_rate": 0.2998960684148887, + "loss": 0.3418, + "num_input_tokens_seen": 994912, + "step": 475 + }, + { + "epoch": 0.07830981319846643, + "grad_norm": 0.003889421233907342, + "learning_rate": 0.29989386446078264, + "loss": 0.3245, + "num_input_tokens_seen": 1004656, + "step": 480 + }, + { + "epoch": 0.0791255404192838, + "grad_norm": 0.0031199154909700155, + "learning_rate": 0.299891637391159, + "loss": 0.4069, + "num_input_tokens_seen": 1016000, + "step": 485 + }, + { + "epoch": 0.07994126764010115, + "grad_norm": 0.005574753973633051, + "learning_rate": 0.2998893872063612, + "loss": 0.4094, + "num_input_tokens_seen": 1027104, + "step": 490 + }, + { + "epoch": 0.08075699486091852, + "grad_norm": 0.004014242440462112, + "learning_rate": 0.2998871139067363, + "loss": 0.3426, + "num_input_tokens_seen": 1035696, + "step": 495 + }, + { + "epoch": 0.08157272208173587, + "grad_norm": 0.003082490060478449, + "learning_rate": 0.2998848174926348, + "loss": 0.3381, + "num_input_tokens_seen": 1046704, + "step": 500 + }, + { + "epoch": 0.08238844930255322, + "grad_norm": 0.0038369374815374613, + "learning_rate": 0.2998824979644109, + "loss": 0.3371, + "num_input_tokens_seen": 1056976, + "step": 505 + }, + { + "epoch": 0.08320417652337059, + "grad_norm": 0.0036174741107970476, + "learning_rate": 0.29988015532242224, + "loss": 0.3322, + "num_input_tokens_seen": 1066640, + "step": 510 + }, + { + "epoch": 0.08401990374418794, + "grad_norm": 0.006264283787459135, + "learning_rate": 0.29987778956703015, + "loss": 0.368, + "num_input_tokens_seen": 1077552, + "step": 515 + }, + { + "epoch": 0.08483563096500531, + "grad_norm": 0.0016577474307268858, + "learning_rate": 0.2998754006985994, + "loss": 0.3703, + "num_input_tokens_seen": 1087552, + "step": 520 + }, + { + "epoch": 0.08565135818582266, + "grad_norm": 0.0063669029623270035, + "learning_rate": 0.29987298871749846, + "loss": 0.3458, + "num_input_tokens_seen": 1097904, + "step": 525 + }, + { + "epoch": 0.08646708540664003, + "grad_norm": 0.003886701539158821, + "learning_rate": 0.2998705536240992, + "loss": 0.3811, + "num_input_tokens_seen": 1108416, + "step": 530 + }, + { + "epoch": 0.08728281262745738, + "grad_norm": 0.004176350776106119, + "learning_rate": 0.2998680954187772, + "loss": 0.3506, + "num_input_tokens_seen": 1119776, + "step": 535 + }, + { + "epoch": 0.08809853984827473, + "grad_norm": 0.0049131340347230434, + "learning_rate": 0.2998656141019115, + "loss": 0.3598, + "num_input_tokens_seen": 1129920, + "step": 540 + }, + { + "epoch": 0.0889142670690921, + "grad_norm": 0.004840197507292032, + "learning_rate": 0.2998631096738848, + "loss": 0.3544, + "num_input_tokens_seen": 1140944, + "step": 545 + }, + { + "epoch": 0.08972999428990945, + "grad_norm": 0.0016731440555304289, + "learning_rate": 0.29986058213508326, + "loss": 0.3535, + "num_input_tokens_seen": 1152800, + "step": 550 + }, + { + "epoch": 0.09054572151072682, + "grad_norm": 0.007126728538423777, + "learning_rate": 0.29985803148589674, + "loss": 0.328, + "num_input_tokens_seen": 1164656, + "step": 555 + }, + { + "epoch": 0.09136144873154417, + "grad_norm": 0.0023244440089911222, + "learning_rate": 0.2998554577267185, + "loss": 0.3626, + "num_input_tokens_seen": 1175328, + "step": 560 + }, + { + "epoch": 0.09217717595236154, + "grad_norm": 0.0027704602107405663, + "learning_rate": 0.2998528608579455, + "loss": 0.3202, + "num_input_tokens_seen": 1185264, + "step": 565 + }, + { + "epoch": 0.09299290317317889, + "grad_norm": 0.0035634597297757864, + "learning_rate": 0.2998502408799781, + "loss": 0.3622, + "num_input_tokens_seen": 1194912, + "step": 570 + }, + { + "epoch": 0.09380863039399624, + "grad_norm": 0.005389500875025988, + "learning_rate": 0.2998475977932205, + "loss": 0.3423, + "num_input_tokens_seen": 1205792, + "step": 575 + }, + { + "epoch": 0.09462435761481361, + "grad_norm": 0.002999930176883936, + "learning_rate": 0.29984493159808023, + "loss": 0.3511, + "num_input_tokens_seen": 1216352, + "step": 580 + }, + { + "epoch": 0.09544008483563096, + "grad_norm": 0.0017796044703572989, + "learning_rate": 0.29984224229496836, + "loss": 0.3604, + "num_input_tokens_seen": 1225664, + "step": 585 + }, + { + "epoch": 0.09625581205644833, + "grad_norm": 0.00244000181555748, + "learning_rate": 0.2998395298842998, + "loss": 0.3198, + "num_input_tokens_seen": 1236912, + "step": 590 + }, + { + "epoch": 0.09707153927726568, + "grad_norm": 0.002502809278666973, + "learning_rate": 0.29983679436649263, + "loss": 0.3479, + "num_input_tokens_seen": 1248064, + "step": 595 + }, + { + "epoch": 0.09788726649808305, + "grad_norm": 0.0027176383882761, + "learning_rate": 0.2998340357419689, + "loss": 0.3445, + "num_input_tokens_seen": 1258320, + "step": 600 + }, + { + "epoch": 0.09788726649808305, + "eval_loss": 0.3395461142063141, + "eval_runtime": 154.881, + "eval_samples_per_second": 17.594, + "eval_steps_per_second": 8.8, + "num_input_tokens_seen": 1258320, + "step": 600 + }, + { + "epoch": 0.0987029937189004, + "grad_norm": 0.005158630199730396, + "learning_rate": 0.29983125401115385, + "loss": 0.3542, + "num_input_tokens_seen": 1268864, + "step": 605 + }, + { + "epoch": 0.09951872093971775, + "grad_norm": 0.004146914929151535, + "learning_rate": 0.29982844917447654, + "loss": 0.3362, + "num_input_tokens_seen": 1279616, + "step": 610 + }, + { + "epoch": 0.10033444816053512, + "grad_norm": 0.003671339713037014, + "learning_rate": 0.2998256212323695, + "loss": 0.3408, + "num_input_tokens_seen": 1290688, + "step": 615 + }, + { + "epoch": 0.10115017538135247, + "grad_norm": 0.001939187292009592, + "learning_rate": 0.29982277018526887, + "loss": 0.368, + "num_input_tokens_seen": 1300656, + "step": 620 + }, + { + "epoch": 0.10196590260216984, + "grad_norm": 0.005686121992766857, + "learning_rate": 0.2998198960336143, + "loss": 0.3337, + "num_input_tokens_seen": 1311056, + "step": 625 + }, + { + "epoch": 0.10278162982298719, + "grad_norm": 0.0021077822893857956, + "learning_rate": 0.299816998777849, + "loss": 0.3403, + "num_input_tokens_seen": 1321024, + "step": 630 + }, + { + "epoch": 0.10359735704380456, + "grad_norm": 0.0020597530528903008, + "learning_rate": 0.2998140784184197, + "loss": 0.3445, + "num_input_tokens_seen": 1330912, + "step": 635 + }, + { + "epoch": 0.10441308426462191, + "grad_norm": 0.006153590977191925, + "learning_rate": 0.2998111349557769, + "loss": 0.2993, + "num_input_tokens_seen": 1339840, + "step": 640 + }, + { + "epoch": 0.10522881148543926, + "grad_norm": 0.003954227548092604, + "learning_rate": 0.29980816839037444, + "loss": 0.3519, + "num_input_tokens_seen": 1350208, + "step": 645 + }, + { + "epoch": 0.10604453870625663, + "grad_norm": 0.007463100366294384, + "learning_rate": 0.2998051787226698, + "loss": 0.3275, + "num_input_tokens_seen": 1360720, + "step": 650 + }, + { + "epoch": 0.10686026592707398, + "grad_norm": 0.006658853497356176, + "learning_rate": 0.29980216595312403, + "loss": 0.3493, + "num_input_tokens_seen": 1371792, + "step": 655 + }, + { + "epoch": 0.10767599314789135, + "grad_norm": 0.003564865794032812, + "learning_rate": 0.29979913008220177, + "loss": 0.3527, + "num_input_tokens_seen": 1381456, + "step": 660 + }, + { + "epoch": 0.1084917203687087, + "grad_norm": 0.00762197794392705, + "learning_rate": 0.2997960711103711, + "loss": 0.3497, + "num_input_tokens_seen": 1390240, + "step": 665 + }, + { + "epoch": 0.10930744758952607, + "grad_norm": 0.0030213005375117064, + "learning_rate": 0.29979298903810386, + "loss": 0.3523, + "num_input_tokens_seen": 1401488, + "step": 670 + }, + { + "epoch": 0.11012317481034342, + "grad_norm": 0.005132665857672691, + "learning_rate": 0.29978988386587524, + "loss": 0.3596, + "num_input_tokens_seen": 1412208, + "step": 675 + }, + { + "epoch": 0.11093890203116077, + "grad_norm": 0.0030210576951503754, + "learning_rate": 0.2997867555941642, + "loss": 0.3399, + "num_input_tokens_seen": 1423232, + "step": 680 + }, + { + "epoch": 0.11175462925197814, + "grad_norm": 0.003074389649555087, + "learning_rate": 0.299783604223453, + "loss": 0.3399, + "num_input_tokens_seen": 1432896, + "step": 685 + }, + { + "epoch": 0.1125703564727955, + "grad_norm": 0.001969139324501157, + "learning_rate": 0.29978042975422786, + "loss": 0.3362, + "num_input_tokens_seen": 1443216, + "step": 690 + }, + { + "epoch": 0.11338608369361286, + "grad_norm": 0.009268972091376781, + "learning_rate": 0.29977723218697816, + "loss": 0.3947, + "num_input_tokens_seen": 1454352, + "step": 695 + }, + { + "epoch": 0.11420181091443021, + "grad_norm": 0.0017823850503191352, + "learning_rate": 0.299774011522197, + "loss": 0.3612, + "num_input_tokens_seen": 1464448, + "step": 700 + }, + { + "epoch": 0.11501753813524758, + "grad_norm": 0.004547430202364922, + "learning_rate": 0.29977076776038114, + "loss": 0.3602, + "num_input_tokens_seen": 1474704, + "step": 705 + }, + { + "epoch": 0.11583326535606493, + "grad_norm": 0.004583375062793493, + "learning_rate": 0.2997675009020307, + "loss": 0.3761, + "num_input_tokens_seen": 1486400, + "step": 710 + }, + { + "epoch": 0.11664899257688228, + "grad_norm": 0.0036545023322105408, + "learning_rate": 0.2997642109476496, + "loss": 0.3532, + "num_input_tokens_seen": 1497488, + "step": 715 + }, + { + "epoch": 0.11746471979769965, + "grad_norm": 0.004706677049398422, + "learning_rate": 0.299760897897745, + "loss": 0.3224, + "num_input_tokens_seen": 1506928, + "step": 720 + }, + { + "epoch": 0.118280447018517, + "grad_norm": 0.0030511573422700167, + "learning_rate": 0.29975756175282803, + "loss": 0.2521, + "num_input_tokens_seen": 1517280, + "step": 725 + }, + { + "epoch": 0.11909617423933437, + "grad_norm": 0.004239110741764307, + "learning_rate": 0.29975420251341306, + "loss": 0.5556, + "num_input_tokens_seen": 1526864, + "step": 730 + }, + { + "epoch": 0.11991190146015172, + "grad_norm": 0.005122567526996136, + "learning_rate": 0.29975082018001814, + "loss": 0.3684, + "num_input_tokens_seen": 1537712, + "step": 735 + }, + { + "epoch": 0.12072762868096909, + "grad_norm": 0.0009733770275488496, + "learning_rate": 0.2997474147531648, + "loss": 0.3712, + "num_input_tokens_seen": 1548752, + "step": 740 + }, + { + "epoch": 0.12154335590178644, + "grad_norm": 0.001045173848979175, + "learning_rate": 0.29974398623337833, + "loss": 0.3522, + "num_input_tokens_seen": 1560400, + "step": 745 + }, + { + "epoch": 0.1223590831226038, + "grad_norm": 0.0035189541522413492, + "learning_rate": 0.2997405346211873, + "loss": 0.351, + "num_input_tokens_seen": 1571696, + "step": 750 + }, + { + "epoch": 0.12317481034342116, + "grad_norm": 0.0012393012875691056, + "learning_rate": 0.2997370599171241, + "loss": 0.3514, + "num_input_tokens_seen": 1582720, + "step": 755 + }, + { + "epoch": 0.12399053756423851, + "grad_norm": 0.0027983086183667183, + "learning_rate": 0.2997335621217246, + "loss": 0.3518, + "num_input_tokens_seen": 1593936, + "step": 760 + }, + { + "epoch": 0.12480626478505588, + "grad_norm": 0.0022470999974757433, + "learning_rate": 0.29973004123552816, + "loss": 0.346, + "num_input_tokens_seen": 1604384, + "step": 765 + }, + { + "epoch": 0.12562199200587323, + "grad_norm": 0.0037341152783483267, + "learning_rate": 0.2997264972590777, + "loss": 0.3821, + "num_input_tokens_seen": 1614704, + "step": 770 + }, + { + "epoch": 0.1264377192266906, + "grad_norm": 0.003368651494383812, + "learning_rate": 0.29972293019291973, + "loss": 0.3253, + "num_input_tokens_seen": 1623680, + "step": 775 + }, + { + "epoch": 0.12725344644750794, + "grad_norm": 0.0018822546117007732, + "learning_rate": 0.2997193400376045, + "loss": 0.3301, + "num_input_tokens_seen": 1633664, + "step": 780 + }, + { + "epoch": 0.12806917366832532, + "grad_norm": 0.0014694606652483344, + "learning_rate": 0.2997157267936854, + "loss": 0.3537, + "num_input_tokens_seen": 1643552, + "step": 785 + }, + { + "epoch": 0.12888490088914267, + "grad_norm": 0.0011447025462985039, + "learning_rate": 0.2997120904617199, + "loss": 0.3836, + "num_input_tokens_seen": 1654544, + "step": 790 + }, + { + "epoch": 0.12970062810996003, + "grad_norm": 0.000939045159611851, + "learning_rate": 0.29970843104226863, + "loss": 0.3518, + "num_input_tokens_seen": 1663728, + "step": 795 + }, + { + "epoch": 0.13051635533077738, + "grad_norm": 0.0016299857525154948, + "learning_rate": 0.2997047485358959, + "loss": 0.3385, + "num_input_tokens_seen": 1673984, + "step": 800 + }, + { + "epoch": 0.13051635533077738, + "eval_loss": 0.38190150260925293, + "eval_runtime": 155.0881, + "eval_samples_per_second": 17.571, + "eval_steps_per_second": 8.789, + "num_input_tokens_seen": 1673984, + "step": 800 + }, + { + "epoch": 0.13133208255159476, + "grad_norm": 0.0015476613771170378, + "learning_rate": 0.2997010429431697, + "loss": 0.3408, + "num_input_tokens_seen": 1684640, + "step": 805 + }, + { + "epoch": 0.1321478097724121, + "grad_norm": 0.006398939061909914, + "learning_rate": 0.29969731426466134, + "loss": 0.4043, + "num_input_tokens_seen": 1694576, + "step": 810 + }, + { + "epoch": 0.13296353699322946, + "grad_norm": 0.0023466432467103004, + "learning_rate": 0.299693562500946, + "loss": 0.3862, + "num_input_tokens_seen": 1705856, + "step": 815 + }, + { + "epoch": 0.13377926421404682, + "grad_norm": 0.0018243200611323118, + "learning_rate": 0.29968978765260207, + "loss": 0.3411, + "num_input_tokens_seen": 1716432, + "step": 820 + }, + { + "epoch": 0.13459499143486417, + "grad_norm": 0.0036380398087203503, + "learning_rate": 0.2996859897202118, + "loss": 0.3213, + "num_input_tokens_seen": 1725936, + "step": 825 + }, + { + "epoch": 0.13541071865568155, + "grad_norm": 0.003462836379185319, + "learning_rate": 0.2996821687043609, + "loss": 0.3726, + "num_input_tokens_seen": 1736256, + "step": 830 + }, + { + "epoch": 0.1362264458764989, + "grad_norm": 0.0015389558393508196, + "learning_rate": 0.2996783246056384, + "loss": 0.3477, + "num_input_tokens_seen": 1746512, + "step": 835 + }, + { + "epoch": 0.13704217309731626, + "grad_norm": 0.0015814187936484814, + "learning_rate": 0.29967445742463744, + "loss": 0.3019, + "num_input_tokens_seen": 1756448, + "step": 840 + }, + { + "epoch": 0.1378579003181336, + "grad_norm": 0.001445625675842166, + "learning_rate": 0.29967056716195417, + "loss": 0.3024, + "num_input_tokens_seen": 1766528, + "step": 845 + }, + { + "epoch": 0.13867362753895096, + "grad_norm": 0.0017033072654157877, + "learning_rate": 0.2996666538181885, + "loss": 0.3905, + "num_input_tokens_seen": 1778784, + "step": 850 + }, + { + "epoch": 0.13948935475976834, + "grad_norm": 0.003498355159536004, + "learning_rate": 0.29966271739394407, + "loss": 0.3616, + "num_input_tokens_seen": 1790016, + "step": 855 + }, + { + "epoch": 0.1403050819805857, + "grad_norm": 0.00107050861697644, + "learning_rate": 0.29965875788982776, + "loss": 0.3365, + "num_input_tokens_seen": 1800608, + "step": 860 + }, + { + "epoch": 0.14112080920140305, + "grad_norm": 0.0011037947842851281, + "learning_rate": 0.2996547753064503, + "loss": 0.3264, + "num_input_tokens_seen": 1810608, + "step": 865 + }, + { + "epoch": 0.1419365364222204, + "grad_norm": 0.003449543146416545, + "learning_rate": 0.29965076964442583, + "loss": 0.3537, + "num_input_tokens_seen": 1822896, + "step": 870 + }, + { + "epoch": 0.14275226364303778, + "grad_norm": 0.003992166370153427, + "learning_rate": 0.299646740904372, + "loss": 0.3304, + "num_input_tokens_seen": 1835712, + "step": 875 + }, + { + "epoch": 0.14356799086385513, + "grad_norm": 0.0013193360064178705, + "learning_rate": 0.29964268908691016, + "loss": 0.3406, + "num_input_tokens_seen": 1845856, + "step": 880 + }, + { + "epoch": 0.14438371808467249, + "grad_norm": 0.003425218164920807, + "learning_rate": 0.29963861419266513, + "loss": 0.3261, + "num_input_tokens_seen": 1855360, + "step": 885 + }, + { + "epoch": 0.14519944530548984, + "grad_norm": 0.006420031655579805, + "learning_rate": 0.29963451622226533, + "loss": 0.3829, + "num_input_tokens_seen": 1865264, + "step": 890 + }, + { + "epoch": 0.1460151725263072, + "grad_norm": 0.0008779715280979872, + "learning_rate": 0.29963039517634277, + "loss": 0.3507, + "num_input_tokens_seen": 1874608, + "step": 895 + }, + { + "epoch": 0.14683089974712457, + "grad_norm": 0.0013013690477237105, + "learning_rate": 0.2996262510555328, + "loss": 0.3461, + "num_input_tokens_seen": 1884736, + "step": 900 + }, + { + "epoch": 0.14764662696794192, + "grad_norm": 0.0021814159117639065, + "learning_rate": 0.2996220838604746, + "loss": 0.3342, + "num_input_tokens_seen": 1896112, + "step": 905 + }, + { + "epoch": 0.14846235418875928, + "grad_norm": 0.002533081453293562, + "learning_rate": 0.29961789359181085, + "loss": 0.3361, + "num_input_tokens_seen": 1908128, + "step": 910 + }, + { + "epoch": 0.14927808140957663, + "grad_norm": 0.004384899511933327, + "learning_rate": 0.29961368025018764, + "loss": 0.3566, + "num_input_tokens_seen": 1918528, + "step": 915 + }, + { + "epoch": 0.150093808630394, + "grad_norm": 0.003769607050344348, + "learning_rate": 0.2996094438362548, + "loss": 0.355, + "num_input_tokens_seen": 1928496, + "step": 920 + }, + { + "epoch": 0.15090953585121136, + "grad_norm": 0.00226087705232203, + "learning_rate": 0.2996051843506657, + "loss": 0.3493, + "num_input_tokens_seen": 1939504, + "step": 925 + }, + { + "epoch": 0.15172526307202872, + "grad_norm": 0.0022942216601222754, + "learning_rate": 0.299600901794077, + "loss": 0.3496, + "num_input_tokens_seen": 1950464, + "step": 930 + }, + { + "epoch": 0.15254099029284607, + "grad_norm": 0.007707653567194939, + "learning_rate": 0.29959659616714923, + "loss": 0.338, + "num_input_tokens_seen": 1962032, + "step": 935 + }, + { + "epoch": 0.15335671751366342, + "grad_norm": 0.004171842243522406, + "learning_rate": 0.2995922674705464, + "loss": 0.3381, + "num_input_tokens_seen": 1972272, + "step": 940 + }, + { + "epoch": 0.1541724447344808, + "grad_norm": 0.0021839311812072992, + "learning_rate": 0.2995879157049361, + "loss": 0.3252, + "num_input_tokens_seen": 1982240, + "step": 945 + }, + { + "epoch": 0.15498817195529815, + "grad_norm": 0.0017911504255607724, + "learning_rate": 0.2995835408709893, + "loss": 0.3663, + "num_input_tokens_seen": 1993472, + "step": 950 + }, + { + "epoch": 0.1558038991761155, + "grad_norm": 0.002295766957104206, + "learning_rate": 0.29957914296938076, + "loss": 0.3466, + "num_input_tokens_seen": 2003360, + "step": 955 + }, + { + "epoch": 0.15661962639693286, + "grad_norm": 0.0038212991785258055, + "learning_rate": 0.2995747220007886, + "loss": 0.3478, + "num_input_tokens_seen": 2015504, + "step": 960 + }, + { + "epoch": 0.1574353536177502, + "grad_norm": 0.0018186277011409402, + "learning_rate": 0.2995702779658947, + "loss": 0.3458, + "num_input_tokens_seen": 2025280, + "step": 965 + }, + { + "epoch": 0.1582510808385676, + "grad_norm": 0.0024929745122790337, + "learning_rate": 0.29956581086538425, + "loss": 0.3123, + "num_input_tokens_seen": 2035136, + "step": 970 + }, + { + "epoch": 0.15906680805938495, + "grad_norm": 0.006560133304446936, + "learning_rate": 0.2995613206999462, + "loss": 0.3606, + "num_input_tokens_seen": 2045664, + "step": 975 + }, + { + "epoch": 0.1598825352802023, + "grad_norm": 0.002434995025396347, + "learning_rate": 0.29955680747027297, + "loss": 0.2959, + "num_input_tokens_seen": 2056400, + "step": 980 + }, + { + "epoch": 0.16069826250101965, + "grad_norm": 0.00487171346321702, + "learning_rate": 0.2995522711770607, + "loss": 0.3705, + "num_input_tokens_seen": 2066560, + "step": 985 + }, + { + "epoch": 0.16151398972183703, + "grad_norm": 0.004345785826444626, + "learning_rate": 0.2995477118210087, + "loss": 0.3619, + "num_input_tokens_seen": 2077728, + "step": 990 + }, + { + "epoch": 0.16232971694265438, + "grad_norm": 0.0038173154462128878, + "learning_rate": 0.29954312940282024, + "loss": 0.345, + "num_input_tokens_seen": 2087728, + "step": 995 + }, + { + "epoch": 0.16314544416347174, + "grad_norm": 0.016025803983211517, + "learning_rate": 0.29953852392320196, + "loss": 0.3856, + "num_input_tokens_seen": 2097344, + "step": 1000 + }, + { + "epoch": 0.16314544416347174, + "eval_loss": 0.3446209728717804, + "eval_runtime": 155.7072, + "eval_samples_per_second": 17.501, + "eval_steps_per_second": 8.754, + "num_input_tokens_seen": 2097344, + "step": 1000 + }, + { + "epoch": 0.1639611713842891, + "grad_norm": 0.004915114026516676, + "learning_rate": 0.2995338953828641, + "loss": 0.3711, + "num_input_tokens_seen": 2108880, + "step": 1005 + }, + { + "epoch": 0.16477689860510644, + "grad_norm": 0.008009597659111023, + "learning_rate": 0.2995292437825204, + "loss": 0.3684, + "num_input_tokens_seen": 2117952, + "step": 1010 + }, + { + "epoch": 0.16559262582592382, + "grad_norm": 0.0015510657103732228, + "learning_rate": 0.29952456912288816, + "loss": 0.3374, + "num_input_tokens_seen": 2129728, + "step": 1015 + }, + { + "epoch": 0.16640835304674118, + "grad_norm": 0.0033708109986037016, + "learning_rate": 0.2995198714046884, + "loss": 0.3386, + "num_input_tokens_seen": 2140256, + "step": 1020 + }, + { + "epoch": 0.16722408026755853, + "grad_norm": 0.0019502558279782534, + "learning_rate": 0.2995151506286454, + "loss": 0.3289, + "num_input_tokens_seen": 2151088, + "step": 1025 + }, + { + "epoch": 0.16803980748837588, + "grad_norm": 0.004487877711653709, + "learning_rate": 0.2995104067954873, + "loss": 0.3398, + "num_input_tokens_seen": 2161184, + "step": 1030 + }, + { + "epoch": 0.16885553470919323, + "grad_norm": 0.0021790775936096907, + "learning_rate": 0.2995056399059456, + "loss": 0.3429, + "num_input_tokens_seen": 2171520, + "step": 1035 + }, + { + "epoch": 0.16967126193001061, + "grad_norm": 0.004260600078850985, + "learning_rate": 0.2995008499607554, + "loss": 0.3281, + "num_input_tokens_seen": 2181968, + "step": 1040 + }, + { + "epoch": 0.17048698915082797, + "grad_norm": 0.002489289967343211, + "learning_rate": 0.2994960369606554, + "loss": 0.3185, + "num_input_tokens_seen": 2192800, + "step": 1045 + }, + { + "epoch": 0.17130271637164532, + "grad_norm": 0.0018319665687158704, + "learning_rate": 0.2994912009063878, + "loss": 0.352, + "num_input_tokens_seen": 2203744, + "step": 1050 + }, + { + "epoch": 0.17211844359246267, + "grad_norm": 0.004588298965245485, + "learning_rate": 0.29948634179869843, + "loss": 0.3322, + "num_input_tokens_seen": 2212240, + "step": 1055 + }, + { + "epoch": 0.17293417081328005, + "grad_norm": 0.0026091658510267735, + "learning_rate": 0.29948145963833656, + "loss": 0.3245, + "num_input_tokens_seen": 2223072, + "step": 1060 + }, + { + "epoch": 0.1737498980340974, + "grad_norm": 0.004484395030885935, + "learning_rate": 0.29947655442605514, + "loss": 0.2786, + "num_input_tokens_seen": 2233664, + "step": 1065 + }, + { + "epoch": 0.17456562525491476, + "grad_norm": 0.006189978215843439, + "learning_rate": 0.2994716261626106, + "loss": 0.3356, + "num_input_tokens_seen": 2243664, + "step": 1070 + }, + { + "epoch": 0.1753813524757321, + "grad_norm": 0.013744615018367767, + "learning_rate": 0.2994666748487629, + "loss": 0.3381, + "num_input_tokens_seen": 2254288, + "step": 1075 + }, + { + "epoch": 0.17619707969654946, + "grad_norm": 0.0046493117697536945, + "learning_rate": 0.2994617004852756, + "loss": 0.3466, + "num_input_tokens_seen": 2264576, + "step": 1080 + }, + { + "epoch": 0.17701280691736684, + "grad_norm": 0.004611166659742594, + "learning_rate": 0.2994567030729159, + "loss": 0.3586, + "num_input_tokens_seen": 2274832, + "step": 1085 + }, + { + "epoch": 0.1778285341381842, + "grad_norm": 0.0046843732707202435, + "learning_rate": 0.29945168261245436, + "loss": 0.3626, + "num_input_tokens_seen": 2285200, + "step": 1090 + }, + { + "epoch": 0.17864426135900155, + "grad_norm": 0.003826683387160301, + "learning_rate": 0.29944663910466524, + "loss": 0.3784, + "num_input_tokens_seen": 2298528, + "step": 1095 + }, + { + "epoch": 0.1794599885798189, + "grad_norm": 0.002508563920855522, + "learning_rate": 0.2994415725503263, + "loss": 0.3417, + "num_input_tokens_seen": 2309008, + "step": 1100 + }, + { + "epoch": 0.18027571580063625, + "grad_norm": 0.002256205305457115, + "learning_rate": 0.29943648295021885, + "loss": 0.3437, + "num_input_tokens_seen": 2319232, + "step": 1105 + }, + { + "epoch": 0.18109144302145364, + "grad_norm": 0.001988241681829095, + "learning_rate": 0.2994313703051278, + "loss": 0.3627, + "num_input_tokens_seen": 2329920, + "step": 1110 + }, + { + "epoch": 0.181907170242271, + "grad_norm": 0.002289233263581991, + "learning_rate": 0.29942623461584156, + "loss": 0.364, + "num_input_tokens_seen": 2341984, + "step": 1115 + }, + { + "epoch": 0.18272289746308834, + "grad_norm": 0.0059751663357019424, + "learning_rate": 0.29942107588315214, + "loss": 0.3321, + "num_input_tokens_seen": 2352416, + "step": 1120 + }, + { + "epoch": 0.1835386246839057, + "grad_norm": 0.0033521910663694143, + "learning_rate": 0.29941589410785513, + "loss": 0.3386, + "num_input_tokens_seen": 2362768, + "step": 1125 + }, + { + "epoch": 0.18435435190472307, + "grad_norm": 0.001786261098459363, + "learning_rate": 0.29941068929074954, + "loss": 0.3189, + "num_input_tokens_seen": 2373312, + "step": 1130 + }, + { + "epoch": 0.18517007912554043, + "grad_norm": 0.0027704753447324038, + "learning_rate": 0.2994054614326381, + "loss": 0.3394, + "num_input_tokens_seen": 2382832, + "step": 1135 + }, + { + "epoch": 0.18598580634635778, + "grad_norm": 0.003295681905001402, + "learning_rate": 0.29940021053432686, + "loss": 0.3287, + "num_input_tokens_seen": 2392368, + "step": 1140 + }, + { + "epoch": 0.18680153356717513, + "grad_norm": 0.0023016456980258226, + "learning_rate": 0.29939493659662575, + "loss": 0.3443, + "num_input_tokens_seen": 2404096, + "step": 1145 + }, + { + "epoch": 0.18761726078799248, + "grad_norm": 0.004299890249967575, + "learning_rate": 0.299389639620348, + "loss": 0.3278, + "num_input_tokens_seen": 2412688, + "step": 1150 + }, + { + "epoch": 0.18843298800880987, + "grad_norm": 0.003644842654466629, + "learning_rate": 0.29938431960631046, + "loss": 0.3184, + "num_input_tokens_seen": 2422624, + "step": 1155 + }, + { + "epoch": 0.18924871522962722, + "grad_norm": 0.0021060791332274675, + "learning_rate": 0.2993789765553335, + "loss": 0.309, + "num_input_tokens_seen": 2432320, + "step": 1160 + }, + { + "epoch": 0.19006444245044457, + "grad_norm": 0.00615809066221118, + "learning_rate": 0.2993736104682412, + "loss": 0.3797, + "num_input_tokens_seen": 2443248, + "step": 1165 + }, + { + "epoch": 0.19088016967126192, + "grad_norm": 0.0023389325942844152, + "learning_rate": 0.299368221345861, + "loss": 0.3228, + "num_input_tokens_seen": 2454384, + "step": 1170 + }, + { + "epoch": 0.19169589689207928, + "grad_norm": 0.012258321978151798, + "learning_rate": 0.29936280918902397, + "loss": 0.3372, + "num_input_tokens_seen": 2465456, + "step": 1175 + }, + { + "epoch": 0.19251162411289666, + "grad_norm": 0.006760898046195507, + "learning_rate": 0.2993573739985648, + "loss": 0.347, + "num_input_tokens_seen": 2476256, + "step": 1180 + }, + { + "epoch": 0.193327351333714, + "grad_norm": 0.0024150481913238764, + "learning_rate": 0.2993519157753216, + "loss": 0.3141, + "num_input_tokens_seen": 2487296, + "step": 1185 + }, + { + "epoch": 0.19414307855453136, + "grad_norm": 0.004783710464835167, + "learning_rate": 0.2993464345201361, + "loss": 0.339, + "num_input_tokens_seen": 2496448, + "step": 1190 + }, + { + "epoch": 0.19495880577534871, + "grad_norm": 0.003948025871068239, + "learning_rate": 0.2993409302338536, + "loss": 0.2638, + "num_input_tokens_seen": 2505856, + "step": 1195 + }, + { + "epoch": 0.1957745329961661, + "grad_norm": 0.006589039228856564, + "learning_rate": 0.2993354029173229, + "loss": 0.3948, + "num_input_tokens_seen": 2515056, + "step": 1200 + }, + { + "epoch": 0.1957745329961661, + "eval_loss": 0.3347358703613281, + "eval_runtime": 155.5363, + "eval_samples_per_second": 17.52, + "eval_steps_per_second": 8.763, + "num_input_tokens_seen": 2515056, + "step": 1200 + }, + { + "epoch": 0.19659026021698345, + "grad_norm": 0.004891451448202133, + "learning_rate": 0.2993298525713965, + "loss": 0.3079, + "num_input_tokens_seen": 2525712, + "step": 1205 + }, + { + "epoch": 0.1974059874378008, + "grad_norm": 0.006376969162374735, + "learning_rate": 0.29932427919693017, + "loss": 0.3701, + "num_input_tokens_seen": 2535872, + "step": 1210 + }, + { + "epoch": 0.19822171465861815, + "grad_norm": 0.0022574570029973984, + "learning_rate": 0.2993186827947834, + "loss": 0.297, + "num_input_tokens_seen": 2547248, + "step": 1215 + }, + { + "epoch": 0.1990374418794355, + "grad_norm": 0.0075279888696968555, + "learning_rate": 0.2993130633658194, + "loss": 0.3187, + "num_input_tokens_seen": 2558224, + "step": 1220 + }, + { + "epoch": 0.1998531691002529, + "grad_norm": 0.005779363214969635, + "learning_rate": 0.29930742091090456, + "loss": 0.3182, + "num_input_tokens_seen": 2568192, + "step": 1225 + }, + { + "epoch": 0.20066889632107024, + "grad_norm": 0.002334220800548792, + "learning_rate": 0.29930175543090914, + "loss": 0.3684, + "num_input_tokens_seen": 2580112, + "step": 1230 + }, + { + "epoch": 0.2014846235418876, + "grad_norm": 0.0023699982557445765, + "learning_rate": 0.2992960669267068, + "loss": 0.4031, + "num_input_tokens_seen": 2591808, + "step": 1235 + }, + { + "epoch": 0.20230035076270494, + "grad_norm": 0.0024309610016644, + "learning_rate": 0.29929035539917476, + "loss": 0.3378, + "num_input_tokens_seen": 2602336, + "step": 1240 + }, + { + "epoch": 0.2031160779835223, + "grad_norm": 0.0062577747739851475, + "learning_rate": 0.2992846208491938, + "loss": 0.369, + "num_input_tokens_seen": 2612096, + "step": 1245 + }, + { + "epoch": 0.20393180520433968, + "grad_norm": 0.0019762124866247177, + "learning_rate": 0.2992788632776483, + "loss": 0.3413, + "num_input_tokens_seen": 2622736, + "step": 1250 + }, + { + "epoch": 0.20474753242515703, + "grad_norm": 0.003143586218357086, + "learning_rate": 0.29927308268542613, + "loss": 0.3773, + "num_input_tokens_seen": 2632320, + "step": 1255 + }, + { + "epoch": 0.20556325964597438, + "grad_norm": 0.001771715353243053, + "learning_rate": 0.2992672790734187, + "loss": 0.3361, + "num_input_tokens_seen": 2644160, + "step": 1260 + }, + { + "epoch": 0.20637898686679174, + "grad_norm": 0.0033435882069170475, + "learning_rate": 0.299261452442521, + "loss": 0.3761, + "num_input_tokens_seen": 2655024, + "step": 1265 + }, + { + "epoch": 0.20719471408760912, + "grad_norm": 0.0013368253130465746, + "learning_rate": 0.29925560279363167, + "loss": 0.3501, + "num_input_tokens_seen": 2664848, + "step": 1270 + }, + { + "epoch": 0.20801044130842647, + "grad_norm": 0.002246229210868478, + "learning_rate": 0.29924973012765266, + "loss": 0.351, + "num_input_tokens_seen": 2675680, + "step": 1275 + }, + { + "epoch": 0.20882616852924382, + "grad_norm": 0.001761930761858821, + "learning_rate": 0.29924383444548974, + "loss": 0.3244, + "num_input_tokens_seen": 2686384, + "step": 1280 + }, + { + "epoch": 0.20964189575006117, + "grad_norm": 0.003817657707259059, + "learning_rate": 0.299237915748052, + "loss": 0.3459, + "num_input_tokens_seen": 2696576, + "step": 1285 + }, + { + "epoch": 0.21045762297087853, + "grad_norm": 0.004242570139467716, + "learning_rate": 0.2992319740362522, + "loss": 0.2932, + "num_input_tokens_seen": 2707376, + "step": 1290 + }, + { + "epoch": 0.2112733501916959, + "grad_norm": 0.02277778647840023, + "learning_rate": 0.2992260093110066, + "loss": 0.3684, + "num_input_tokens_seen": 2718768, + "step": 1295 + }, + { + "epoch": 0.21208907741251326, + "grad_norm": 0.00398236932232976, + "learning_rate": 0.2992200215732352, + "loss": 0.3846, + "num_input_tokens_seen": 2727728, + "step": 1300 + }, + { + "epoch": 0.2129048046333306, + "grad_norm": 0.003953870385885239, + "learning_rate": 0.2992140108238611, + "loss": 0.355, + "num_input_tokens_seen": 2739264, + "step": 1305 + }, + { + "epoch": 0.21372053185414797, + "grad_norm": 0.001860627206042409, + "learning_rate": 0.2992079770638115, + "loss": 0.3559, + "num_input_tokens_seen": 2749920, + "step": 1310 + }, + { + "epoch": 0.21453625907496532, + "grad_norm": 0.0017721714684739709, + "learning_rate": 0.29920192029401677, + "loss": 0.3471, + "num_input_tokens_seen": 2759440, + "step": 1315 + }, + { + "epoch": 0.2153519862957827, + "grad_norm": 0.0018087525386363268, + "learning_rate": 0.2991958405154109, + "loss": 0.3248, + "num_input_tokens_seen": 2770720, + "step": 1320 + }, + { + "epoch": 0.21616771351660005, + "grad_norm": 0.00143712037242949, + "learning_rate": 0.29918973772893154, + "loss": 0.3202, + "num_input_tokens_seen": 2781760, + "step": 1325 + }, + { + "epoch": 0.2169834407374174, + "grad_norm": 0.001632988452911377, + "learning_rate": 0.29918361193551973, + "loss": 0.3125, + "num_input_tokens_seen": 2791872, + "step": 1330 + }, + { + "epoch": 0.21779916795823476, + "grad_norm": 0.002183354226872325, + "learning_rate": 0.29917746313612026, + "loss": 0.384, + "num_input_tokens_seen": 2803104, + "step": 1335 + }, + { + "epoch": 0.21861489517905214, + "grad_norm": 0.004468036349862814, + "learning_rate": 0.29917129133168124, + "loss": 0.3338, + "num_input_tokens_seen": 2813296, + "step": 1340 + }, + { + "epoch": 0.2194306223998695, + "grad_norm": 0.0017906628781929612, + "learning_rate": 0.2991650965231546, + "loss": 0.3402, + "num_input_tokens_seen": 2824960, + "step": 1345 + }, + { + "epoch": 0.22024634962068684, + "grad_norm": 0.0031935549341142178, + "learning_rate": 0.29915887871149544, + "loss": 0.3377, + "num_input_tokens_seen": 2835984, + "step": 1350 + }, + { + "epoch": 0.2210620768415042, + "grad_norm": 0.0016794634284451604, + "learning_rate": 0.2991526378976628, + "loss": 0.3275, + "num_input_tokens_seen": 2845744, + "step": 1355 + }, + { + "epoch": 0.22187780406232155, + "grad_norm": 0.003599076997488737, + "learning_rate": 0.29914637408261896, + "loss": 0.2989, + "num_input_tokens_seen": 2857344, + "step": 1360 + }, + { + "epoch": 0.22269353128313893, + "grad_norm": 0.0019797568675130606, + "learning_rate": 0.29914008726733, + "loss": 0.3619, + "num_input_tokens_seen": 2867776, + "step": 1365 + }, + { + "epoch": 0.22350925850395628, + "grad_norm": 0.001570652355439961, + "learning_rate": 0.2991337774527653, + "loss": 0.2783, + "num_input_tokens_seen": 2878960, + "step": 1370 + }, + { + "epoch": 0.22432498572477363, + "grad_norm": 0.0011758863693103194, + "learning_rate": 0.2991274446398981, + "loss": 0.2998, + "num_input_tokens_seen": 2891024, + "step": 1375 + }, + { + "epoch": 0.225140712945591, + "grad_norm": 0.003272431902587414, + "learning_rate": 0.29912108882970484, + "loss": 0.2991, + "num_input_tokens_seen": 2901456, + "step": 1380 + }, + { + "epoch": 0.22595644016640834, + "grad_norm": 0.0024484898895025253, + "learning_rate": 0.2991147100231657, + "loss": 0.3043, + "num_input_tokens_seen": 2912016, + "step": 1385 + }, + { + "epoch": 0.22677216738722572, + "grad_norm": 0.00442514568567276, + "learning_rate": 0.2991083082212644, + "loss": 0.3343, + "num_input_tokens_seen": 2922192, + "step": 1390 + }, + { + "epoch": 0.22758789460804307, + "grad_norm": 0.011631554923951626, + "learning_rate": 0.2991018834249881, + "loss": 0.311, + "num_input_tokens_seen": 2933488, + "step": 1395 + }, + { + "epoch": 0.22840362182886043, + "grad_norm": 0.003801258746534586, + "learning_rate": 0.29909543563532764, + "loss": 0.3097, + "num_input_tokens_seen": 2943280, + "step": 1400 + }, + { + "epoch": 0.22840362182886043, + "eval_loss": 0.32623228430747986, + "eval_runtime": 155.6827, + "eval_samples_per_second": 17.504, + "eval_steps_per_second": 8.755, + "num_input_tokens_seen": 2943280, + "step": 1400 + }, + { + "epoch": 0.22921934904967778, + "grad_norm": 0.008248308673501015, + "learning_rate": 0.29908896485327746, + "loss": 0.3735, + "num_input_tokens_seen": 2955792, + "step": 1405 + }, + { + "epoch": 0.23003507627049516, + "grad_norm": 0.0042039272375404835, + "learning_rate": 0.29908247107983527, + "loss": 0.3709, + "num_input_tokens_seen": 2965968, + "step": 1410 + }, + { + "epoch": 0.2308508034913125, + "grad_norm": 0.004846850410103798, + "learning_rate": 0.29907595431600253, + "loss": 0.3478, + "num_input_tokens_seen": 2975968, + "step": 1415 + }, + { + "epoch": 0.23166653071212986, + "grad_norm": 0.0019089032430201769, + "learning_rate": 0.29906941456278424, + "loss": 0.371, + "num_input_tokens_seen": 2986016, + "step": 1420 + }, + { + "epoch": 0.23248225793294722, + "grad_norm": 0.005443074274808168, + "learning_rate": 0.2990628518211889, + "loss": 0.3154, + "num_input_tokens_seen": 2994768, + "step": 1425 + }, + { + "epoch": 0.23329798515376457, + "grad_norm": 0.0029183190781623125, + "learning_rate": 0.2990562660922286, + "loss": 0.3484, + "num_input_tokens_seen": 3003264, + "step": 1430 + }, + { + "epoch": 0.23411371237458195, + "grad_norm": 0.0014200868317857385, + "learning_rate": 0.2990496573769189, + "loss": 0.3632, + "num_input_tokens_seen": 3013904, + "step": 1435 + }, + { + "epoch": 0.2349294395953993, + "grad_norm": 0.0029175307136029005, + "learning_rate": 0.29904302567627894, + "loss": 0.335, + "num_input_tokens_seen": 3024432, + "step": 1440 + }, + { + "epoch": 0.23574516681621666, + "grad_norm": 0.007318297866731882, + "learning_rate": 0.2990363709913314, + "loss": 0.3737, + "num_input_tokens_seen": 3034624, + "step": 1445 + }, + { + "epoch": 0.236560894037034, + "grad_norm": 0.0016497453907504678, + "learning_rate": 0.29902969332310264, + "loss": 0.3355, + "num_input_tokens_seen": 3044288, + "step": 1450 + }, + { + "epoch": 0.23737662125785136, + "grad_norm": 0.0018152876291424036, + "learning_rate": 0.2990229926726223, + "loss": 0.343, + "num_input_tokens_seen": 3054384, + "step": 1455 + }, + { + "epoch": 0.23819234847866874, + "grad_norm": 0.0018590528052300215, + "learning_rate": 0.29901626904092365, + "loss": 0.2575, + "num_input_tokens_seen": 3065136, + "step": 1460 + }, + { + "epoch": 0.2390080756994861, + "grad_norm": 0.0018419299740344286, + "learning_rate": 0.2990095224290438, + "loss": 0.2977, + "num_input_tokens_seen": 3076384, + "step": 1465 + }, + { + "epoch": 0.23982380292030345, + "grad_norm": 0.0020858747884631157, + "learning_rate": 0.29900275283802297, + "loss": 0.3225, + "num_input_tokens_seen": 3088800, + "step": 1470 + }, + { + "epoch": 0.2406395301411208, + "grad_norm": 0.0028587800916284323, + "learning_rate": 0.2989959602689051, + "loss": 0.3407, + "num_input_tokens_seen": 3099328, + "step": 1475 + }, + { + "epoch": 0.24145525736193818, + "grad_norm": 0.0034131850115954876, + "learning_rate": 0.2989891447227379, + "loss": 0.3906, + "num_input_tokens_seen": 3109920, + "step": 1480 + }, + { + "epoch": 0.24227098458275553, + "grad_norm": 0.0033531251829117537, + "learning_rate": 0.29898230620057215, + "loss": 0.3296, + "num_input_tokens_seen": 3119360, + "step": 1485 + }, + { + "epoch": 0.24308671180357289, + "grad_norm": 0.005957566201686859, + "learning_rate": 0.2989754447034626, + "loss": 0.3761, + "num_input_tokens_seen": 3128896, + "step": 1490 + }, + { + "epoch": 0.24390243902439024, + "grad_norm": 0.0023242230527102947, + "learning_rate": 0.2989685602324673, + "loss": 0.3075, + "num_input_tokens_seen": 3140512, + "step": 1495 + }, + { + "epoch": 0.2447181662452076, + "grad_norm": 0.0018890727078542113, + "learning_rate": 0.298961652788648, + "loss": 0.329, + "num_input_tokens_seen": 3150480, + "step": 1500 + }, + { + "epoch": 0.24553389346602497, + "grad_norm": 0.002041296334937215, + "learning_rate": 0.29895472237306986, + "loss": 0.278, + "num_input_tokens_seen": 3161408, + "step": 1505 + }, + { + "epoch": 0.24634962068684232, + "grad_norm": 0.004123319871723652, + "learning_rate": 0.29894776898680164, + "loss": 0.3224, + "num_input_tokens_seen": 3171472, + "step": 1510 + }, + { + "epoch": 0.24716534790765968, + "grad_norm": 0.0018634287407621741, + "learning_rate": 0.29894079263091566, + "loss": 0.3643, + "num_input_tokens_seen": 3182176, + "step": 1515 + }, + { + "epoch": 0.24798107512847703, + "grad_norm": 0.001997335348278284, + "learning_rate": 0.2989337933064877, + "loss": 0.2779, + "num_input_tokens_seen": 3193104, + "step": 1520 + }, + { + "epoch": 0.24879680234929438, + "grad_norm": 0.003978698514401913, + "learning_rate": 0.29892677101459725, + "loss": 0.3316, + "num_input_tokens_seen": 3203872, + "step": 1525 + }, + { + "epoch": 0.24961252957011176, + "grad_norm": 0.0023778730537742376, + "learning_rate": 0.2989197257563272, + "loss": 0.3183, + "num_input_tokens_seen": 3213440, + "step": 1530 + }, + { + "epoch": 0.2504282567909291, + "grad_norm": 0.0013240821426734328, + "learning_rate": 0.2989126575327639, + "loss": 0.3357, + "num_input_tokens_seen": 3224720, + "step": 1535 + }, + { + "epoch": 0.25124398401174647, + "grad_norm": 0.0016874123830348253, + "learning_rate": 0.29890556634499754, + "loss": 0.2902, + "num_input_tokens_seen": 3236320, + "step": 1540 + }, + { + "epoch": 0.2520597112325638, + "grad_norm": 0.004500208888202906, + "learning_rate": 0.2988984521941216, + "loss": 0.4089, + "num_input_tokens_seen": 3246928, + "step": 1545 + }, + { + "epoch": 0.2528754384533812, + "grad_norm": 0.0012168819084763527, + "learning_rate": 0.29889131508123307, + "loss": 0.3055, + "num_input_tokens_seen": 3256672, + "step": 1550 + }, + { + "epoch": 0.2536911656741985, + "grad_norm": 0.005062466021627188, + "learning_rate": 0.2988841550074327, + "loss": 0.3499, + "num_input_tokens_seen": 3267200, + "step": 1555 + }, + { + "epoch": 0.2545068928950159, + "grad_norm": 0.0045015146024525166, + "learning_rate": 0.2988769719738246, + "loss": 0.3082, + "num_input_tokens_seen": 3277168, + "step": 1560 + }, + { + "epoch": 0.2553226201158333, + "grad_norm": 0.005249973852187395, + "learning_rate": 0.29886976598151666, + "loss": 0.3538, + "num_input_tokens_seen": 3287488, + "step": 1565 + }, + { + "epoch": 0.25613834733665064, + "grad_norm": 0.0027722730301320553, + "learning_rate": 0.29886253703161986, + "loss": 0.3329, + "num_input_tokens_seen": 3297952, + "step": 1570 + }, + { + "epoch": 0.256954074557468, + "grad_norm": 0.001710803247988224, + "learning_rate": 0.29885528512524917, + "loss": 0.3117, + "num_input_tokens_seen": 3306912, + "step": 1575 + }, + { + "epoch": 0.25776980177828535, + "grad_norm": 0.002775187138468027, + "learning_rate": 0.29884801026352287, + "loss": 0.3255, + "num_input_tokens_seen": 3315712, + "step": 1580 + }, + { + "epoch": 0.2585855289991027, + "grad_norm": 0.00623695831745863, + "learning_rate": 0.2988407124475629, + "loss": 0.28, + "num_input_tokens_seen": 3327312, + "step": 1585 + }, + { + "epoch": 0.25940125621992005, + "grad_norm": 0.005534789524972439, + "learning_rate": 0.2988333916784945, + "loss": 0.3602, + "num_input_tokens_seen": 3338992, + "step": 1590 + }, + { + "epoch": 0.2602169834407374, + "grad_norm": 0.0019673218484967947, + "learning_rate": 0.2988260479574468, + "loss": 0.4383, + "num_input_tokens_seen": 3350720, + "step": 1595 + }, + { + "epoch": 0.26103271066155476, + "grad_norm": 0.005785458721220493, + "learning_rate": 0.2988186812855523, + "loss": 0.2916, + "num_input_tokens_seen": 3360448, + "step": 1600 + }, + { + "epoch": 0.26103271066155476, + "eval_loss": 0.32416218519210815, + "eval_runtime": 155.7299, + "eval_samples_per_second": 17.498, + "eval_steps_per_second": 8.752, + "num_input_tokens_seen": 3360448, + "step": 1600 + }, + { + "epoch": 0.2618484378823721, + "grad_norm": 0.0028418435249477625, + "learning_rate": 0.29881129166394693, + "loss": 0.273, + "num_input_tokens_seen": 3371184, + "step": 1605 + }, + { + "epoch": 0.2626641651031895, + "grad_norm": 0.0019118610071018338, + "learning_rate": 0.29880387909377026, + "loss": 0.377, + "num_input_tokens_seen": 3383232, + "step": 1610 + }, + { + "epoch": 0.26347989232400687, + "grad_norm": 0.004118203651160002, + "learning_rate": 0.2987964435761655, + "loss": 0.2887, + "num_input_tokens_seen": 3394384, + "step": 1615 + }, + { + "epoch": 0.2642956195448242, + "grad_norm": 0.010001979768276215, + "learning_rate": 0.29878898511227925, + "loss": 0.4644, + "num_input_tokens_seen": 3404288, + "step": 1620 + }, + { + "epoch": 0.2651113467656416, + "grad_norm": 0.007667034864425659, + "learning_rate": 0.2987815037032617, + "loss": 0.3427, + "num_input_tokens_seen": 3415440, + "step": 1625 + }, + { + "epoch": 0.26592707398645893, + "grad_norm": 0.009596778079867363, + "learning_rate": 0.29877399935026655, + "loss": 0.3197, + "num_input_tokens_seen": 3426672, + "step": 1630 + }, + { + "epoch": 0.2667428012072763, + "grad_norm": 0.004166428931057453, + "learning_rate": 0.2987664720544511, + "loss": 0.3711, + "num_input_tokens_seen": 3436304, + "step": 1635 + }, + { + "epoch": 0.26755852842809363, + "grad_norm": 0.004024288151413202, + "learning_rate": 0.2987589218169761, + "loss": 0.3339, + "num_input_tokens_seen": 3447408, + "step": 1640 + }, + { + "epoch": 0.268374255648911, + "grad_norm": 0.00294339656829834, + "learning_rate": 0.29875134863900604, + "loss": 0.3254, + "num_input_tokens_seen": 3458144, + "step": 1645 + }, + { + "epoch": 0.26918998286972834, + "grad_norm": 0.0030152020044624805, + "learning_rate": 0.29874375252170865, + "loss": 0.3541, + "num_input_tokens_seen": 3469776, + "step": 1650 + }, + { + "epoch": 0.27000571009054575, + "grad_norm": 0.003197286743670702, + "learning_rate": 0.2987361334662553, + "loss": 0.3663, + "num_input_tokens_seen": 3480560, + "step": 1655 + }, + { + "epoch": 0.2708214373113631, + "grad_norm": 0.0026175675448030233, + "learning_rate": 0.29872849147382113, + "loss": 0.3406, + "num_input_tokens_seen": 3491264, + "step": 1660 + }, + { + "epoch": 0.27163716453218045, + "grad_norm": 0.0031507962848991156, + "learning_rate": 0.2987208265455845, + "loss": 0.3393, + "num_input_tokens_seen": 3501920, + "step": 1665 + }, + { + "epoch": 0.2724528917529978, + "grad_norm": 0.0031249332241714, + "learning_rate": 0.29871313868272753, + "loss": 0.3489, + "num_input_tokens_seen": 3511776, + "step": 1670 + }, + { + "epoch": 0.27326861897381516, + "grad_norm": 0.003907470498234034, + "learning_rate": 0.29870542788643567, + "loss": 0.3679, + "num_input_tokens_seen": 3522896, + "step": 1675 + }, + { + "epoch": 0.2740843461946325, + "grad_norm": 0.004410870838910341, + "learning_rate": 0.2986976941578981, + "loss": 0.3291, + "num_input_tokens_seen": 3533808, + "step": 1680 + }, + { + "epoch": 0.27490007341544986, + "grad_norm": 0.0026346067897975445, + "learning_rate": 0.29868993749830747, + "loss": 0.3335, + "num_input_tokens_seen": 3544960, + "step": 1685 + }, + { + "epoch": 0.2757158006362672, + "grad_norm": 0.0030530740041285753, + "learning_rate": 0.2986821579088598, + "loss": 0.37, + "num_input_tokens_seen": 3554672, + "step": 1690 + }, + { + "epoch": 0.27653152785708457, + "grad_norm": 0.001812435220927, + "learning_rate": 0.29867435539075504, + "loss": 0.3005, + "num_input_tokens_seen": 3565664, + "step": 1695 + }, + { + "epoch": 0.2773472550779019, + "grad_norm": 0.0014057470252737403, + "learning_rate": 0.2986665299451963, + "loss": 0.3179, + "num_input_tokens_seen": 3575744, + "step": 1700 + }, + { + "epoch": 0.27816298229871933, + "grad_norm": 0.0020887881983071566, + "learning_rate": 0.29865868157339037, + "loss": 0.2822, + "num_input_tokens_seen": 3585840, + "step": 1705 + }, + { + "epoch": 0.2789787095195367, + "grad_norm": 0.0034263127017766237, + "learning_rate": 0.2986508102765476, + "loss": 0.3523, + "num_input_tokens_seen": 3596816, + "step": 1710 + }, + { + "epoch": 0.27979443674035404, + "grad_norm": 0.0028459702152758837, + "learning_rate": 0.2986429160558818, + "loss": 0.2851, + "num_input_tokens_seen": 3607584, + "step": 1715 + }, + { + "epoch": 0.2806101639611714, + "grad_norm": 0.002237144624814391, + "learning_rate": 0.2986349989126104, + "loss": 0.4152, + "num_input_tokens_seen": 3617504, + "step": 1720 + }, + { + "epoch": 0.28142589118198874, + "grad_norm": 0.0027793985791504383, + "learning_rate": 0.29862705884795426, + "loss": 0.3229, + "num_input_tokens_seen": 3628496, + "step": 1725 + }, + { + "epoch": 0.2822416184028061, + "grad_norm": 0.0019755444955080748, + "learning_rate": 0.2986190958631379, + "loss": 0.3208, + "num_input_tokens_seen": 3638848, + "step": 1730 + }, + { + "epoch": 0.28305734562362345, + "grad_norm": 0.0030268202535808086, + "learning_rate": 0.29861110995938933, + "loss": 0.3469, + "num_input_tokens_seen": 3648784, + "step": 1735 + }, + { + "epoch": 0.2838730728444408, + "grad_norm": 0.0042704008519649506, + "learning_rate": 0.29860310113794, + "loss": 0.3194, + "num_input_tokens_seen": 3657760, + "step": 1740 + }, + { + "epoch": 0.28468880006525815, + "grad_norm": 0.004336914047598839, + "learning_rate": 0.29859506940002506, + "loss": 0.3633, + "num_input_tokens_seen": 3667456, + "step": 1745 + }, + { + "epoch": 0.28550452728607556, + "grad_norm": 0.0012928375508636236, + "learning_rate": 0.298587014746883, + "loss": 0.3509, + "num_input_tokens_seen": 3679296, + "step": 1750 + }, + { + "epoch": 0.2863202545068929, + "grad_norm": 0.0029745223000645638, + "learning_rate": 0.298578937179756, + "loss": 0.2666, + "num_input_tokens_seen": 3690176, + "step": 1755 + }, + { + "epoch": 0.28713598172771027, + "grad_norm": 0.0013210566248744726, + "learning_rate": 0.29857083669988976, + "loss": 0.3675, + "num_input_tokens_seen": 3699744, + "step": 1760 + }, + { + "epoch": 0.2879517089485276, + "grad_norm": 0.0029689688235521317, + "learning_rate": 0.29856271330853346, + "loss": 0.3145, + "num_input_tokens_seen": 3710016, + "step": 1765 + }, + { + "epoch": 0.28876743616934497, + "grad_norm": 0.0013818895677104592, + "learning_rate": 0.2985545670069398, + "loss": 0.3508, + "num_input_tokens_seen": 3720224, + "step": 1770 + }, + { + "epoch": 0.2895831633901623, + "grad_norm": 0.0018888642080128193, + "learning_rate": 0.29854639779636505, + "loss": 0.3319, + "num_input_tokens_seen": 3731280, + "step": 1775 + }, + { + "epoch": 0.2903988906109797, + "grad_norm": 0.0039171199314296246, + "learning_rate": 0.298538205678069, + "loss": 0.3495, + "num_input_tokens_seen": 3742352, + "step": 1780 + }, + { + "epoch": 0.29121461783179703, + "grad_norm": 0.0022311650682240725, + "learning_rate": 0.298529990653315, + "loss": 0.338, + "num_input_tokens_seen": 3752144, + "step": 1785 + }, + { + "epoch": 0.2920303450526144, + "grad_norm": 0.0019547520205378532, + "learning_rate": 0.29852175272336984, + "loss": 0.3199, + "num_input_tokens_seen": 3761760, + "step": 1790 + }, + { + "epoch": 0.2928460722734318, + "grad_norm": 0.0016626928700134158, + "learning_rate": 0.29851349188950405, + "loss": 0.3019, + "num_input_tokens_seen": 3771952, + "step": 1795 + }, + { + "epoch": 0.29366179949424914, + "grad_norm": 0.003795742057263851, + "learning_rate": 0.2985052081529914, + "loss": 0.2844, + "num_input_tokens_seen": 3782768, + "step": 1800 + }, + { + "epoch": 0.29366179949424914, + "eval_loss": 0.32318371534347534, + "eval_runtime": 156.1429, + "eval_samples_per_second": 17.452, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 3782768, + "step": 1800 + }, + { + "epoch": 0.2944775267150665, + "grad_norm": 0.0023093039635568857, + "learning_rate": 0.29849690151510944, + "loss": 0.3472, + "num_input_tokens_seen": 3794032, + "step": 1805 + }, + { + "epoch": 0.29529325393588385, + "grad_norm": 0.0017709826352074742, + "learning_rate": 0.2984885719771392, + "loss": 0.3286, + "num_input_tokens_seen": 3803696, + "step": 1810 + }, + { + "epoch": 0.2961089811567012, + "grad_norm": 0.00221143732778728, + "learning_rate": 0.2984802195403651, + "loss": 0.3069, + "num_input_tokens_seen": 3815184, + "step": 1815 + }, + { + "epoch": 0.29692470837751855, + "grad_norm": 0.002768902340903878, + "learning_rate": 0.2984718442060752, + "loss": 0.3158, + "num_input_tokens_seen": 3823280, + "step": 1820 + }, + { + "epoch": 0.2977404355983359, + "grad_norm": 0.004052346106618643, + "learning_rate": 0.2984634459755611, + "loss": 0.3156, + "num_input_tokens_seen": 3834880, + "step": 1825 + }, + { + "epoch": 0.29855616281915326, + "grad_norm": 0.0014681221218779683, + "learning_rate": 0.29845502485011793, + "loss": 0.3426, + "num_input_tokens_seen": 3846688, + "step": 1830 + }, + { + "epoch": 0.2993718900399706, + "grad_norm": 0.001786438631825149, + "learning_rate": 0.2984465808310444, + "loss": 0.3192, + "num_input_tokens_seen": 3856672, + "step": 1835 + }, + { + "epoch": 0.300187617260788, + "grad_norm": 0.0018404334550723433, + "learning_rate": 0.29843811391964253, + "loss": 0.3313, + "num_input_tokens_seen": 3867008, + "step": 1840 + }, + { + "epoch": 0.3010033444816054, + "grad_norm": 0.0021180950570851564, + "learning_rate": 0.2984296241172182, + "loss": 0.2999, + "num_input_tokens_seen": 3876096, + "step": 1845 + }, + { + "epoch": 0.3018190717024227, + "grad_norm": 0.001824252656660974, + "learning_rate": 0.29842111142508043, + "loss": 0.3665, + "num_input_tokens_seen": 3886288, + "step": 1850 + }, + { + "epoch": 0.3026347989232401, + "grad_norm": 0.003941659815609455, + "learning_rate": 0.29841257584454217, + "loss": 0.3259, + "num_input_tokens_seen": 3896096, + "step": 1855 + }, + { + "epoch": 0.30345052614405743, + "grad_norm": 0.004542078357189894, + "learning_rate": 0.29840401737691963, + "loss": 0.4292, + "num_input_tokens_seen": 3905632, + "step": 1860 + }, + { + "epoch": 0.3042662533648748, + "grad_norm": 0.002017942490056157, + "learning_rate": 0.29839543602353263, + "loss": 0.3132, + "num_input_tokens_seen": 3915856, + "step": 1865 + }, + { + "epoch": 0.30508198058569214, + "grad_norm": 0.0014085484435781837, + "learning_rate": 0.2983868317857046, + "loss": 0.3345, + "num_input_tokens_seen": 3926592, + "step": 1870 + }, + { + "epoch": 0.3058977078065095, + "grad_norm": 0.0014124109875410795, + "learning_rate": 0.2983782046647623, + "loss": 0.3091, + "num_input_tokens_seen": 3937344, + "step": 1875 + }, + { + "epoch": 0.30671343502732684, + "grad_norm": 0.003402130678296089, + "learning_rate": 0.2983695546620362, + "loss": 0.3478, + "num_input_tokens_seen": 3948224, + "step": 1880 + }, + { + "epoch": 0.3075291622481442, + "grad_norm": 0.0016870603431016207, + "learning_rate": 0.2983608817788603, + "loss": 0.3468, + "num_input_tokens_seen": 3958368, + "step": 1885 + }, + { + "epoch": 0.3083448894689616, + "grad_norm": 0.0013088658452033997, + "learning_rate": 0.29835218601657193, + "loss": 0.3362, + "num_input_tokens_seen": 3968960, + "step": 1890 + }, + { + "epoch": 0.30916061668977896, + "grad_norm": 0.002409295178949833, + "learning_rate": 0.29834346737651224, + "loss": 0.2638, + "num_input_tokens_seen": 3978656, + "step": 1895 + }, + { + "epoch": 0.3099763439105963, + "grad_norm": 0.0029921522364020348, + "learning_rate": 0.29833472586002563, + "loss": 0.4181, + "num_input_tokens_seen": 3989376, + "step": 1900 + }, + { + "epoch": 0.31079207113141366, + "grad_norm": 0.0015562610933557153, + "learning_rate": 0.29832596146846024, + "loss": 0.3526, + "num_input_tokens_seen": 4001264, + "step": 1905 + }, + { + "epoch": 0.311607798352231, + "grad_norm": 0.0025768429040908813, + "learning_rate": 0.2983171742031676, + "loss": 0.3254, + "num_input_tokens_seen": 4012576, + "step": 1910 + }, + { + "epoch": 0.31242352557304837, + "grad_norm": 0.0014259126037359238, + "learning_rate": 0.2983083640655028, + "loss": 0.3392, + "num_input_tokens_seen": 4022976, + "step": 1915 + }, + { + "epoch": 0.3132392527938657, + "grad_norm": 0.0012798506068065763, + "learning_rate": 0.29829953105682455, + "loss": 0.3196, + "num_input_tokens_seen": 4033664, + "step": 1920 + }, + { + "epoch": 0.31405498001468307, + "grad_norm": 0.001339602516964078, + "learning_rate": 0.29829067517849495, + "loss": 0.3131, + "num_input_tokens_seen": 4045264, + "step": 1925 + }, + { + "epoch": 0.3148707072355004, + "grad_norm": 0.003267345018684864, + "learning_rate": 0.2982817964318797, + "loss": 0.3366, + "num_input_tokens_seen": 4056288, + "step": 1930 + }, + { + "epoch": 0.31568643445631783, + "grad_norm": 0.0017089551547542214, + "learning_rate": 0.298272894818348, + "loss": 0.2745, + "num_input_tokens_seen": 4067472, + "step": 1935 + }, + { + "epoch": 0.3165021616771352, + "grad_norm": 0.002588986186310649, + "learning_rate": 0.2982639703392726, + "loss": 0.2654, + "num_input_tokens_seen": 4078976, + "step": 1940 + }, + { + "epoch": 0.31731788889795254, + "grad_norm": 0.0019725896418094635, + "learning_rate": 0.29825502299602974, + "loss": 0.3221, + "num_input_tokens_seen": 4088256, + "step": 1945 + }, + { + "epoch": 0.3181336161187699, + "grad_norm": 0.016597682610154152, + "learning_rate": 0.2982460527899993, + "loss": 0.4099, + "num_input_tokens_seen": 4098800, + "step": 1950 + }, + { + "epoch": 0.31894934333958724, + "grad_norm": 0.0036676793824881315, + "learning_rate": 0.29823705972256453, + "loss": 0.2985, + "num_input_tokens_seen": 4110048, + "step": 1955 + }, + { + "epoch": 0.3197650705604046, + "grad_norm": 0.0027454623486846685, + "learning_rate": 0.2982280437951123, + "loss": 0.328, + "num_input_tokens_seen": 4120000, + "step": 1960 + }, + { + "epoch": 0.32058079778122195, + "grad_norm": 0.002317698672413826, + "learning_rate": 0.298219005009033, + "loss": 0.334, + "num_input_tokens_seen": 4129840, + "step": 1965 + }, + { + "epoch": 0.3213965250020393, + "grad_norm": 0.002174587454646826, + "learning_rate": 0.29820994336572043, + "loss": 0.3102, + "num_input_tokens_seen": 4140224, + "step": 1970 + }, + { + "epoch": 0.32221225222285665, + "grad_norm": 0.002501687966287136, + "learning_rate": 0.2982008588665721, + "loss": 0.3316, + "num_input_tokens_seen": 4150832, + "step": 1975 + }, + { + "epoch": 0.32302797944367406, + "grad_norm": 0.0025143614038825035, + "learning_rate": 0.2981917515129889, + "loss": 0.3218, + "num_input_tokens_seen": 4161136, + "step": 1980 + }, + { + "epoch": 0.3238437066644914, + "grad_norm": 0.002193773863837123, + "learning_rate": 0.2981826213063753, + "loss": 0.3053, + "num_input_tokens_seen": 4171392, + "step": 1985 + }, + { + "epoch": 0.32465943388530877, + "grad_norm": 0.0014806127874180675, + "learning_rate": 0.2981734682481394, + "loss": 0.3414, + "num_input_tokens_seen": 4181984, + "step": 1990 + }, + { + "epoch": 0.3254751611061261, + "grad_norm": 0.0015751017490401864, + "learning_rate": 0.29816429233969255, + "loss": 0.2918, + "num_input_tokens_seen": 4193472, + "step": 1995 + }, + { + "epoch": 0.3262908883269435, + "grad_norm": 0.0023902757093310356, + "learning_rate": 0.2981550935824499, + "loss": 0.3879, + "num_input_tokens_seen": 4205680, + "step": 2000 + }, + { + "epoch": 0.3262908883269435, + "eval_loss": 0.3215888738632202, + "eval_runtime": 155.9429, + "eval_samples_per_second": 17.474, + "eval_steps_per_second": 8.74, + "num_input_tokens_seen": 4205680, + "step": 2000 + }, + { + "epoch": 0.3271066155477608, + "grad_norm": 0.0040078191086649895, + "learning_rate": 0.29814587197783, + "loss": 0.3611, + "num_input_tokens_seen": 4216496, + "step": 2005 + }, + { + "epoch": 0.3279223427685782, + "grad_norm": 0.001636652508750558, + "learning_rate": 0.29813662752725495, + "loss": 0.3561, + "num_input_tokens_seen": 4227232, + "step": 2010 + }, + { + "epoch": 0.32873806998939553, + "grad_norm": 0.004348531365394592, + "learning_rate": 0.29812736023215025, + "loss": 0.3632, + "num_input_tokens_seen": 4238512, + "step": 2015 + }, + { + "epoch": 0.3295537972102129, + "grad_norm": 0.0015615056036040187, + "learning_rate": 0.29811807009394514, + "loss": 0.3254, + "num_input_tokens_seen": 4250048, + "step": 2020 + }, + { + "epoch": 0.33036952443103024, + "grad_norm": 0.004085992928594351, + "learning_rate": 0.2981087571140723, + "loss": 0.3067, + "num_input_tokens_seen": 4260640, + "step": 2025 + }, + { + "epoch": 0.33118525165184765, + "grad_norm": 0.0011221643071621656, + "learning_rate": 0.2980994212939678, + "loss": 0.359, + "num_input_tokens_seen": 4271248, + "step": 2030 + }, + { + "epoch": 0.332000978872665, + "grad_norm": 0.0013492887374013662, + "learning_rate": 0.2980900626350715, + "loss": 0.3232, + "num_input_tokens_seen": 4281040, + "step": 2035 + }, + { + "epoch": 0.33281670609348235, + "grad_norm": 0.0018209065310657024, + "learning_rate": 0.29808068113882646, + "loss": 0.3188, + "num_input_tokens_seen": 4291520, + "step": 2040 + }, + { + "epoch": 0.3336324333142997, + "grad_norm": 0.0014607356861233711, + "learning_rate": 0.2980712768066795, + "loss": 0.3463, + "num_input_tokens_seen": 4302240, + "step": 2045 + }, + { + "epoch": 0.33444816053511706, + "grad_norm": 0.010053323581814766, + "learning_rate": 0.2980618496400809, + "loss": 0.383, + "num_input_tokens_seen": 4312128, + "step": 2050 + }, + { + "epoch": 0.3352638877559344, + "grad_norm": 0.002587942173704505, + "learning_rate": 0.2980523996404844, + "loss": 0.352, + "num_input_tokens_seen": 4322720, + "step": 2055 + }, + { + "epoch": 0.33607961497675176, + "grad_norm": 0.002561272121965885, + "learning_rate": 0.2980429268093473, + "loss": 0.3153, + "num_input_tokens_seen": 4332400, + "step": 2060 + }, + { + "epoch": 0.3368953421975691, + "grad_norm": 0.0017682205652818084, + "learning_rate": 0.29803343114813047, + "loss": 0.348, + "num_input_tokens_seen": 4342928, + "step": 2065 + }, + { + "epoch": 0.33771106941838647, + "grad_norm": 0.0024688798002898693, + "learning_rate": 0.2980239126582983, + "loss": 0.3264, + "num_input_tokens_seen": 4354400, + "step": 2070 + }, + { + "epoch": 0.3385267966392039, + "grad_norm": 0.0023402951192110777, + "learning_rate": 0.2980143713413186, + "loss": 0.3173, + "num_input_tokens_seen": 4365328, + "step": 2075 + }, + { + "epoch": 0.33934252386002123, + "grad_norm": 0.004305868409574032, + "learning_rate": 0.29800480719866274, + "loss": 0.3927, + "num_input_tokens_seen": 4374512, + "step": 2080 + }, + { + "epoch": 0.3401582510808386, + "grad_norm": 0.0022514367010444403, + "learning_rate": 0.2979952202318057, + "loss": 0.4114, + "num_input_tokens_seen": 4384480, + "step": 2085 + }, + { + "epoch": 0.34097397830165593, + "grad_norm": 0.0017198947025462985, + "learning_rate": 0.2979856104422259, + "loss": 0.3488, + "num_input_tokens_seen": 4396096, + "step": 2090 + }, + { + "epoch": 0.3417897055224733, + "grad_norm": 0.0010998568031936884, + "learning_rate": 0.2979759778314052, + "loss": 0.3088, + "num_input_tokens_seen": 4406496, + "step": 2095 + }, + { + "epoch": 0.34260543274329064, + "grad_norm": 0.0009473544778302312, + "learning_rate": 0.2979663224008292, + "loss": 0.3479, + "num_input_tokens_seen": 4416368, + "step": 2100 + }, + { + "epoch": 0.343421159964108, + "grad_norm": 0.0012090642703697085, + "learning_rate": 0.2979566441519868, + "loss": 0.3438, + "num_input_tokens_seen": 4427360, + "step": 2105 + }, + { + "epoch": 0.34423688718492534, + "grad_norm": 0.001029488048516214, + "learning_rate": 0.29794694308637054, + "loss": 0.378, + "num_input_tokens_seen": 4437280, + "step": 2110 + }, + { + "epoch": 0.3450526144057427, + "grad_norm": 0.0010899255285039544, + "learning_rate": 0.2979372192054764, + "loss": 0.3503, + "num_input_tokens_seen": 4447872, + "step": 2115 + }, + { + "epoch": 0.3458683416265601, + "grad_norm": 0.0014430626761168242, + "learning_rate": 0.297927472510804, + "loss": 0.3367, + "num_input_tokens_seen": 4458784, + "step": 2120 + }, + { + "epoch": 0.34668406884737746, + "grad_norm": 0.0011328174732625484, + "learning_rate": 0.29791770300385634, + "loss": 0.3276, + "num_input_tokens_seen": 4469728, + "step": 2125 + }, + { + "epoch": 0.3474997960681948, + "grad_norm": 0.0010235938243567944, + "learning_rate": 0.29790791068614003, + "loss": 0.3147, + "num_input_tokens_seen": 4479872, + "step": 2130 + }, + { + "epoch": 0.34831552328901216, + "grad_norm": 0.0021684609819203615, + "learning_rate": 0.2978980955591652, + "loss": 0.3844, + "num_input_tokens_seen": 4489920, + "step": 2135 + }, + { + "epoch": 0.3491312505098295, + "grad_norm": 0.0009185347589664161, + "learning_rate": 0.2978882576244454, + "loss": 0.3234, + "num_input_tokens_seen": 4499232, + "step": 2140 + }, + { + "epoch": 0.34994697773064687, + "grad_norm": 0.0012863995507359505, + "learning_rate": 0.2978783968834978, + "loss": 0.3349, + "num_input_tokens_seen": 4508528, + "step": 2145 + }, + { + "epoch": 0.3507627049514642, + "grad_norm": 0.0015510775847360492, + "learning_rate": 0.29786851333784303, + "loss": 0.308, + "num_input_tokens_seen": 4518880, + "step": 2150 + }, + { + "epoch": 0.3515784321722816, + "grad_norm": 0.0013991697924211621, + "learning_rate": 0.2978586069890053, + "loss": 0.3229, + "num_input_tokens_seen": 4530176, + "step": 2155 + }, + { + "epoch": 0.3523941593930989, + "grad_norm": 0.0024597763549536467, + "learning_rate": 0.29784867783851227, + "loss": 0.393, + "num_input_tokens_seen": 4540048, + "step": 2160 + }, + { + "epoch": 0.3532098866139163, + "grad_norm": 0.0015577560989186168, + "learning_rate": 0.2978387258878951, + "loss": 0.3347, + "num_input_tokens_seen": 4549744, + "step": 2165 + }, + { + "epoch": 0.3540256138347337, + "grad_norm": 0.0012287257704883814, + "learning_rate": 0.29782875113868856, + "loss": 0.3632, + "num_input_tokens_seen": 4561280, + "step": 2170 + }, + { + "epoch": 0.35484134105555104, + "grad_norm": 0.0030440078116953373, + "learning_rate": 0.2978187535924309, + "loss": 0.3517, + "num_input_tokens_seen": 4571232, + "step": 2175 + }, + { + "epoch": 0.3556570682763684, + "grad_norm": 0.0009547201334498823, + "learning_rate": 0.29780873325066376, + "loss": 0.3296, + "num_input_tokens_seen": 4580960, + "step": 2180 + }, + { + "epoch": 0.35647279549718575, + "grad_norm": 0.002222132869064808, + "learning_rate": 0.2977986901149325, + "loss": 0.3499, + "num_input_tokens_seen": 4590560, + "step": 2185 + }, + { + "epoch": 0.3572885227180031, + "grad_norm": 0.0014770691050216556, + "learning_rate": 0.29778862418678587, + "loss": 0.339, + "num_input_tokens_seen": 4600048, + "step": 2190 + }, + { + "epoch": 0.35810424993882045, + "grad_norm": 0.0009311647154390812, + "learning_rate": 0.29777853546777616, + "loss": 0.3121, + "num_input_tokens_seen": 4610688, + "step": 2195 + }, + { + "epoch": 0.3589199771596378, + "grad_norm": 0.0020446092821657658, + "learning_rate": 0.2977684239594592, + "loss": 0.3318, + "num_input_tokens_seen": 4620944, + "step": 2200 + }, + { + "epoch": 0.3589199771596378, + "eval_loss": 0.33026573061943054, + "eval_runtime": 155.8735, + "eval_samples_per_second": 17.482, + "eval_steps_per_second": 8.744, + "num_input_tokens_seen": 4620944, + "step": 2200 + }, + { + "epoch": 0.35973570438045516, + "grad_norm": 0.001399877481162548, + "learning_rate": 0.29775828966339424, + "loss": 0.3419, + "num_input_tokens_seen": 4630656, + "step": 2205 + }, + { + "epoch": 0.3605514316012725, + "grad_norm": 0.001528484863229096, + "learning_rate": 0.29774813258114424, + "loss": 0.3541, + "num_input_tokens_seen": 4642112, + "step": 2210 + }, + { + "epoch": 0.3613671588220899, + "grad_norm": 0.003570972941815853, + "learning_rate": 0.29773795271427544, + "loss": 0.3542, + "num_input_tokens_seen": 4652688, + "step": 2215 + }, + { + "epoch": 0.36218288604290727, + "grad_norm": 0.0037128690164536238, + "learning_rate": 0.2977277500643577, + "loss": 0.3677, + "num_input_tokens_seen": 4663840, + "step": 2220 + }, + { + "epoch": 0.3629986132637246, + "grad_norm": 0.0013342403108254075, + "learning_rate": 0.29771752463296447, + "loss": 0.3775, + "num_input_tokens_seen": 4674384, + "step": 2225 + }, + { + "epoch": 0.363814340484542, + "grad_norm": 0.0016134585021063685, + "learning_rate": 0.29770727642167266, + "loss": 0.3364, + "num_input_tokens_seen": 4685248, + "step": 2230 + }, + { + "epoch": 0.36463006770535933, + "grad_norm": 0.0012097612489014864, + "learning_rate": 0.29769700543206257, + "loss": 0.3488, + "num_input_tokens_seen": 4696576, + "step": 2235 + }, + { + "epoch": 0.3654457949261767, + "grad_norm": 0.0020931290928274393, + "learning_rate": 0.2976867116657182, + "loss": 0.3486, + "num_input_tokens_seen": 4706848, + "step": 2240 + }, + { + "epoch": 0.36626152214699403, + "grad_norm": 0.0024296753108501434, + "learning_rate": 0.2976763951242269, + "loss": 0.3182, + "num_input_tokens_seen": 4718224, + "step": 2245 + }, + { + "epoch": 0.3670772493678114, + "grad_norm": 0.0010482901707291603, + "learning_rate": 0.29766605580917965, + "loss": 0.3421, + "num_input_tokens_seen": 4728336, + "step": 2250 + }, + { + "epoch": 0.36789297658862874, + "grad_norm": 0.0008800657815299928, + "learning_rate": 0.29765569372217093, + "loss": 0.3322, + "num_input_tokens_seen": 4739264, + "step": 2255 + }, + { + "epoch": 0.36870870380944615, + "grad_norm": 0.0011697524460032582, + "learning_rate": 0.2976453088647987, + "loss": 0.3171, + "num_input_tokens_seen": 4746736, + "step": 2260 + }, + { + "epoch": 0.3695244310302635, + "grad_norm": 0.0011649366933852434, + "learning_rate": 0.2976349012386644, + "loss": 0.3507, + "num_input_tokens_seen": 4756976, + "step": 2265 + }, + { + "epoch": 0.37034015825108085, + "grad_norm": 0.002116328338161111, + "learning_rate": 0.29762447084537297, + "loss": 0.3275, + "num_input_tokens_seen": 4766448, + "step": 2270 + }, + { + "epoch": 0.3711558854718982, + "grad_norm": 0.0015562402550131083, + "learning_rate": 0.29761401768653306, + "loss": 0.3184, + "num_input_tokens_seen": 4777584, + "step": 2275 + }, + { + "epoch": 0.37197161269271556, + "grad_norm": 0.0008633274119347334, + "learning_rate": 0.29760354176375653, + "loss": 0.3335, + "num_input_tokens_seen": 4788064, + "step": 2280 + }, + { + "epoch": 0.3727873399135329, + "grad_norm": 0.0018718598876148462, + "learning_rate": 0.29759304307865897, + "loss": 0.3005, + "num_input_tokens_seen": 4799568, + "step": 2285 + }, + { + "epoch": 0.37360306713435026, + "grad_norm": 0.0013679268304258585, + "learning_rate": 0.2975825216328594, + "loss": 0.3379, + "num_input_tokens_seen": 4809200, + "step": 2290 + }, + { + "epoch": 0.3744187943551676, + "grad_norm": 0.0014313142746686935, + "learning_rate": 0.2975719774279804, + "loss": 0.3038, + "num_input_tokens_seen": 4818640, + "step": 2295 + }, + { + "epoch": 0.37523452157598497, + "grad_norm": 0.0033253678120672703, + "learning_rate": 0.29756141046564794, + "loss": 0.2812, + "num_input_tokens_seen": 4828816, + "step": 2300 + }, + { + "epoch": 0.3760502487968023, + "grad_norm": 0.0046171084977686405, + "learning_rate": 0.2975508207474916, + "loss": 0.3124, + "num_input_tokens_seen": 4838864, + "step": 2305 + }, + { + "epoch": 0.37686597601761973, + "grad_norm": 0.0015020931605249643, + "learning_rate": 0.2975402082751445, + "loss": 0.3098, + "num_input_tokens_seen": 4849728, + "step": 2310 + }, + { + "epoch": 0.3776817032384371, + "grad_norm": 0.002404296537861228, + "learning_rate": 0.29752957305024313, + "loss": 0.3522, + "num_input_tokens_seen": 4860432, + "step": 2315 + }, + { + "epoch": 0.37849743045925444, + "grad_norm": 0.0011109282495453954, + "learning_rate": 0.2975189150744277, + "loss": 0.3725, + "num_input_tokens_seen": 4870784, + "step": 2320 + }, + { + "epoch": 0.3793131576800718, + "grad_norm": 0.0035257500130683184, + "learning_rate": 0.29750823434934165, + "loss": 0.3747, + "num_input_tokens_seen": 4880768, + "step": 2325 + }, + { + "epoch": 0.38012888490088914, + "grad_norm": 0.001008775201626122, + "learning_rate": 0.29749753087663217, + "loss": 0.2734, + "num_input_tokens_seen": 4891536, + "step": 2330 + }, + { + "epoch": 0.3809446121217065, + "grad_norm": 0.0012855614768341184, + "learning_rate": 0.29748680465794985, + "loss": 0.3192, + "num_input_tokens_seen": 4901680, + "step": 2335 + }, + { + "epoch": 0.38176033934252385, + "grad_norm": 0.0014938790118321776, + "learning_rate": 0.29747605569494884, + "loss": 0.2687, + "num_input_tokens_seen": 4913824, + "step": 2340 + }, + { + "epoch": 0.3825760665633412, + "grad_norm": 0.003242469159886241, + "learning_rate": 0.29746528398928673, + "loss": 0.329, + "num_input_tokens_seen": 4924528, + "step": 2345 + }, + { + "epoch": 0.38339179378415855, + "grad_norm": 0.002823294373229146, + "learning_rate": 0.2974544895426247, + "loss": 0.3152, + "num_input_tokens_seen": 4934784, + "step": 2350 + }, + { + "epoch": 0.38420752100497596, + "grad_norm": 0.0017129918560385704, + "learning_rate": 0.29744367235662733, + "loss": 0.3222, + "num_input_tokens_seen": 4944768, + "step": 2355 + }, + { + "epoch": 0.3850232482257933, + "grad_norm": 0.0039434474892914295, + "learning_rate": 0.29743283243296276, + "loss": 0.2933, + "num_input_tokens_seen": 4954368, + "step": 2360 + }, + { + "epoch": 0.38583897544661067, + "grad_norm": 0.0012018640991300344, + "learning_rate": 0.29742196977330276, + "loss": 0.2893, + "num_input_tokens_seen": 4963616, + "step": 2365 + }, + { + "epoch": 0.386654702667428, + "grad_norm": 0.002270828001201153, + "learning_rate": 0.2974110843793223, + "loss": 0.3004, + "num_input_tokens_seen": 4973808, + "step": 2370 + }, + { + "epoch": 0.38747042988824537, + "grad_norm": 0.0021790112368762493, + "learning_rate": 0.2974001762527002, + "loss": 0.2697, + "num_input_tokens_seen": 4984320, + "step": 2375 + }, + { + "epoch": 0.3882861571090627, + "grad_norm": 0.002713443711400032, + "learning_rate": 0.2973892453951186, + "loss": 0.2364, + "num_input_tokens_seen": 4995488, + "step": 2380 + }, + { + "epoch": 0.3891018843298801, + "grad_norm": 0.0029887326527386904, + "learning_rate": 0.2973782918082631, + "loss": 0.3511, + "num_input_tokens_seen": 5006080, + "step": 2385 + }, + { + "epoch": 0.38991761155069743, + "grad_norm": 0.004999150987714529, + "learning_rate": 0.29736731549382295, + "loss": 0.2856, + "num_input_tokens_seen": 5016464, + "step": 2390 + }, + { + "epoch": 0.3907333387715148, + "grad_norm": 0.002697778632864356, + "learning_rate": 0.2973563164534908, + "loss": 0.3172, + "num_input_tokens_seen": 5026944, + "step": 2395 + }, + { + "epoch": 0.3915490659923322, + "grad_norm": 0.001116662286221981, + "learning_rate": 0.29734529468896287, + "loss": 0.3499, + "num_input_tokens_seen": 5037232, + "step": 2400 + }, + { + "epoch": 0.3915490659923322, + "eval_loss": 0.3219336271286011, + "eval_runtime": 155.8677, + "eval_samples_per_second": 17.483, + "eval_steps_per_second": 8.745, + "num_input_tokens_seen": 5037232, + "step": 2400 + }, + { + "epoch": 0.39236479321314954, + "grad_norm": 0.0026675297413021326, + "learning_rate": 0.2973342502019388, + "loss": 0.3338, + "num_input_tokens_seen": 5047536, + "step": 2405 + }, + { + "epoch": 0.3931805204339669, + "grad_norm": 0.003743786597624421, + "learning_rate": 0.2973231829941219, + "loss": 0.2988, + "num_input_tokens_seen": 5057408, + "step": 2410 + }, + { + "epoch": 0.39399624765478425, + "grad_norm": 0.0029051820747554302, + "learning_rate": 0.2973120930672188, + "loss": 0.3527, + "num_input_tokens_seen": 5068416, + "step": 2415 + }, + { + "epoch": 0.3948119748756016, + "grad_norm": 0.0016571574378758669, + "learning_rate": 0.2973009804229397, + "loss": 0.3539, + "num_input_tokens_seen": 5079936, + "step": 2420 + }, + { + "epoch": 0.39562770209641895, + "grad_norm": 0.004280771128833294, + "learning_rate": 0.29728984506299827, + "loss": 0.3699, + "num_input_tokens_seen": 5090080, + "step": 2425 + }, + { + "epoch": 0.3964434293172363, + "grad_norm": 0.0011250959942117333, + "learning_rate": 0.2972786869891118, + "loss": 0.3582, + "num_input_tokens_seen": 5101120, + "step": 2430 + }, + { + "epoch": 0.39725915653805366, + "grad_norm": 0.001081493217498064, + "learning_rate": 0.29726750620300096, + "loss": 0.3154, + "num_input_tokens_seen": 5110976, + "step": 2435 + }, + { + "epoch": 0.398074883758871, + "grad_norm": 0.0014019380323588848, + "learning_rate": 0.29725630270639003, + "loss": 0.4068, + "num_input_tokens_seen": 5121840, + "step": 2440 + }, + { + "epoch": 0.39889061097968836, + "grad_norm": 0.0018859247211366892, + "learning_rate": 0.2972450765010067, + "loss": 0.3301, + "num_input_tokens_seen": 5133120, + "step": 2445 + }, + { + "epoch": 0.3997063382005058, + "grad_norm": 0.00295349326916039, + "learning_rate": 0.29723382758858213, + "loss": 0.3422, + "num_input_tokens_seen": 5144096, + "step": 2450 + }, + { + "epoch": 0.4005220654213231, + "grad_norm": 0.0014913721242919564, + "learning_rate": 0.29722255597085107, + "loss": 0.3413, + "num_input_tokens_seen": 5154848, + "step": 2455 + }, + { + "epoch": 0.4013377926421405, + "grad_norm": 0.0013514095917344093, + "learning_rate": 0.2972112616495518, + "loss": 0.3271, + "num_input_tokens_seen": 5166112, + "step": 2460 + }, + { + "epoch": 0.40215351986295783, + "grad_norm": 0.0014050801983103156, + "learning_rate": 0.297199944626426, + "loss": 0.372, + "num_input_tokens_seen": 5177760, + "step": 2465 + }, + { + "epoch": 0.4029692470837752, + "grad_norm": 0.003792761592194438, + "learning_rate": 0.2971886049032189, + "loss": 0.3307, + "num_input_tokens_seen": 5188016, + "step": 2470 + }, + { + "epoch": 0.40378497430459254, + "grad_norm": 0.004798308480530977, + "learning_rate": 0.29717724248167926, + "loss": 0.3503, + "num_input_tokens_seen": 5199744, + "step": 2475 + }, + { + "epoch": 0.4046007015254099, + "grad_norm": 0.005721183028072119, + "learning_rate": 0.29716585736355927, + "loss": 0.3507, + "num_input_tokens_seen": 5210624, + "step": 2480 + }, + { + "epoch": 0.40541642874622724, + "grad_norm": 0.001982618821784854, + "learning_rate": 0.2971544495506147, + "loss": 0.3135, + "num_input_tokens_seen": 5219344, + "step": 2485 + }, + { + "epoch": 0.4062321559670446, + "grad_norm": 0.0010639833053573966, + "learning_rate": 0.2971430190446048, + "loss": 0.3324, + "num_input_tokens_seen": 5228400, + "step": 2490 + }, + { + "epoch": 0.407047883187862, + "grad_norm": 0.0026543112471699715, + "learning_rate": 0.2971315658472921, + "loss": 0.3761, + "num_input_tokens_seen": 5239056, + "step": 2495 + }, + { + "epoch": 0.40786361040867936, + "grad_norm": 0.0012394465738907456, + "learning_rate": 0.2971200899604431, + "loss": 0.3084, + "num_input_tokens_seen": 5248640, + "step": 2500 + }, + { + "epoch": 0.4086793376294967, + "grad_norm": 0.0017050763126462698, + "learning_rate": 0.29710859138582735, + "loss": 0.3297, + "num_input_tokens_seen": 5259312, + "step": 2505 + }, + { + "epoch": 0.40949506485031406, + "grad_norm": 0.0027209147810935974, + "learning_rate": 0.29709707012521813, + "loss": 0.2785, + "num_input_tokens_seen": 5269360, + "step": 2510 + }, + { + "epoch": 0.4103107920711314, + "grad_norm": 0.0006285311537794769, + "learning_rate": 0.29708552618039213, + "loss": 0.3981, + "num_input_tokens_seen": 5278528, + "step": 2515 + }, + { + "epoch": 0.41112651929194877, + "grad_norm": 0.002274584723636508, + "learning_rate": 0.2970739595531296, + "loss": 0.2977, + "num_input_tokens_seen": 5288864, + "step": 2520 + }, + { + "epoch": 0.4119422465127661, + "grad_norm": 0.0023584633599966764, + "learning_rate": 0.2970623702452143, + "loss": 0.3481, + "num_input_tokens_seen": 5298864, + "step": 2525 + }, + { + "epoch": 0.41275797373358347, + "grad_norm": 0.0034251066390424967, + "learning_rate": 0.2970507582584334, + "loss": 0.3046, + "num_input_tokens_seen": 5309824, + "step": 2530 + }, + { + "epoch": 0.4135737009544008, + "grad_norm": 0.001320572686381638, + "learning_rate": 0.2970391235945776, + "loss": 0.3242, + "num_input_tokens_seen": 5317872, + "step": 2535 + }, + { + "epoch": 0.41438942817521823, + "grad_norm": 0.0010635900544002652, + "learning_rate": 0.2970274662554412, + "loss": 0.3141, + "num_input_tokens_seen": 5327584, + "step": 2540 + }, + { + "epoch": 0.4152051553960356, + "grad_norm": 0.0018948122160509229, + "learning_rate": 0.2970157862428218, + "loss": 0.354, + "num_input_tokens_seen": 5337824, + "step": 2545 + }, + { + "epoch": 0.41602088261685294, + "grad_norm": 0.0008229347877204418, + "learning_rate": 0.2970040835585206, + "loss": 0.2951, + "num_input_tokens_seen": 5348480, + "step": 2550 + }, + { + "epoch": 0.4168366098376703, + "grad_norm": 0.0015554087003692985, + "learning_rate": 0.2969923582043424, + "loss": 0.3841, + "num_input_tokens_seen": 5358768, + "step": 2555 + }, + { + "epoch": 0.41765233705848764, + "grad_norm": 0.0012760718818753958, + "learning_rate": 0.2969806101820953, + "loss": 0.3205, + "num_input_tokens_seen": 5369472, + "step": 2560 + }, + { + "epoch": 0.418468064279305, + "grad_norm": 0.000977406743913889, + "learning_rate": 0.2969688394935911, + "loss": 0.3082, + "num_input_tokens_seen": 5381344, + "step": 2565 + }, + { + "epoch": 0.41928379150012235, + "grad_norm": 0.0008083070861175656, + "learning_rate": 0.2969570461406449, + "loss": 0.338, + "num_input_tokens_seen": 5390688, + "step": 2570 + }, + { + "epoch": 0.4200995187209397, + "grad_norm": 0.0011711996048688889, + "learning_rate": 0.29694523012507534, + "loss": 0.3536, + "num_input_tokens_seen": 5401984, + "step": 2575 + }, + { + "epoch": 0.42091524594175705, + "grad_norm": 0.0024995990097522736, + "learning_rate": 0.2969333914487048, + "loss": 0.3269, + "num_input_tokens_seen": 5411792, + "step": 2580 + }, + { + "epoch": 0.4217309731625744, + "grad_norm": 0.0010608760640025139, + "learning_rate": 0.2969215301133587, + "loss": 0.3059, + "num_input_tokens_seen": 5421616, + "step": 2585 + }, + { + "epoch": 0.4225467003833918, + "grad_norm": 0.0010715938406065106, + "learning_rate": 0.29690964612086634, + "loss": 0.3511, + "num_input_tokens_seen": 5430976, + "step": 2590 + }, + { + "epoch": 0.42336242760420917, + "grad_norm": 0.00189229438547045, + "learning_rate": 0.2968977394730604, + "loss": 0.3281, + "num_input_tokens_seen": 5442128, + "step": 2595 + }, + { + "epoch": 0.4241781548250265, + "grad_norm": 0.0013617717195302248, + "learning_rate": 0.296885810171777, + "loss": 0.354, + "num_input_tokens_seen": 5452736, + "step": 2600 + }, + { + "epoch": 0.4241781548250265, + "eval_loss": 0.32056233286857605, + "eval_runtime": 156.0898, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 5452736, + "step": 2600 + }, + { + "epoch": 0.4249938820458439, + "grad_norm": 0.001659345580264926, + "learning_rate": 0.2968738582188558, + "loss": 0.3647, + "num_input_tokens_seen": 5463808, + "step": 2605 + }, + { + "epoch": 0.4258096092666612, + "grad_norm": 0.002086037304252386, + "learning_rate": 0.2968618836161399, + "loss": 0.3241, + "num_input_tokens_seen": 5473520, + "step": 2610 + }, + { + "epoch": 0.4266253364874786, + "grad_norm": 0.001845355611294508, + "learning_rate": 0.296849886365476, + "loss": 0.3967, + "num_input_tokens_seen": 5483216, + "step": 2615 + }, + { + "epoch": 0.42744106370829593, + "grad_norm": 0.0019158029463142157, + "learning_rate": 0.2968378664687142, + "loss": 0.3432, + "num_input_tokens_seen": 5493136, + "step": 2620 + }, + { + "epoch": 0.4282567909291133, + "grad_norm": 0.0019027363741770387, + "learning_rate": 0.296825823927708, + "loss": 0.3284, + "num_input_tokens_seen": 5504016, + "step": 2625 + }, + { + "epoch": 0.42907251814993064, + "grad_norm": 0.0012797857634723186, + "learning_rate": 0.29681375874431476, + "loss": 0.3207, + "num_input_tokens_seen": 5514800, + "step": 2630 + }, + { + "epoch": 0.42988824537074805, + "grad_norm": 0.0026355506852269173, + "learning_rate": 0.29680167092039483, + "loss": 0.2721, + "num_input_tokens_seen": 5524208, + "step": 2635 + }, + { + "epoch": 0.4307039725915654, + "grad_norm": 0.0012351491022855043, + "learning_rate": 0.2967895604578125, + "loss": 0.3763, + "num_input_tokens_seen": 5535232, + "step": 2640 + }, + { + "epoch": 0.43151969981238275, + "grad_norm": 0.001741337007842958, + "learning_rate": 0.2967774273584352, + "loss": 0.2667, + "num_input_tokens_seen": 5546480, + "step": 2645 + }, + { + "epoch": 0.4323354270332001, + "grad_norm": 0.001070113736204803, + "learning_rate": 0.2967652716241342, + "loss": 0.3647, + "num_input_tokens_seen": 5556960, + "step": 2650 + }, + { + "epoch": 0.43315115425401746, + "grad_norm": 0.0013474292354658246, + "learning_rate": 0.29675309325678384, + "loss": 0.2777, + "num_input_tokens_seen": 5566976, + "step": 2655 + }, + { + "epoch": 0.4339668814748348, + "grad_norm": 0.002570071490481496, + "learning_rate": 0.29674089225826233, + "loss": 0.3658, + "num_input_tokens_seen": 5577440, + "step": 2660 + }, + { + "epoch": 0.43478260869565216, + "grad_norm": 0.001778421108610928, + "learning_rate": 0.29672866863045116, + "loss": 0.287, + "num_input_tokens_seen": 5587824, + "step": 2665 + }, + { + "epoch": 0.4355983359164695, + "grad_norm": 0.001508875167928636, + "learning_rate": 0.2967164223752354, + "loss": 0.3113, + "num_input_tokens_seen": 5598656, + "step": 2670 + }, + { + "epoch": 0.43641406313728687, + "grad_norm": 0.0017600301653146744, + "learning_rate": 0.2967041534945035, + "loss": 0.3558, + "num_input_tokens_seen": 5608656, + "step": 2675 + }, + { + "epoch": 0.4372297903581043, + "grad_norm": 0.0023241122253239155, + "learning_rate": 0.2966918619901476, + "loss": 0.3262, + "num_input_tokens_seen": 5618912, + "step": 2680 + }, + { + "epoch": 0.43804551757892163, + "grad_norm": 0.0010172269539907575, + "learning_rate": 0.2966795478640631, + "loss": 0.3475, + "num_input_tokens_seen": 5630256, + "step": 2685 + }, + { + "epoch": 0.438861244799739, + "grad_norm": 0.004137392155826092, + "learning_rate": 0.29666721111814903, + "loss": 0.3478, + "num_input_tokens_seen": 5640192, + "step": 2690 + }, + { + "epoch": 0.43967697202055633, + "grad_norm": 0.0011011360911652446, + "learning_rate": 0.2966548517543079, + "loss": 0.3527, + "num_input_tokens_seen": 5649744, + "step": 2695 + }, + { + "epoch": 0.4404926992413737, + "grad_norm": 0.0010245278244838119, + "learning_rate": 0.29664246977444564, + "loss": 0.3396, + "num_input_tokens_seen": 5658464, + "step": 2700 + }, + { + "epoch": 0.44130842646219104, + "grad_norm": 0.0011520982952788472, + "learning_rate": 0.2966300651804717, + "loss": 0.3025, + "num_input_tokens_seen": 5669632, + "step": 2705 + }, + { + "epoch": 0.4421241536830084, + "grad_norm": 0.0006628388073295355, + "learning_rate": 0.296617637974299, + "loss": 0.358, + "num_input_tokens_seen": 5679056, + "step": 2710 + }, + { + "epoch": 0.44293988090382574, + "grad_norm": 0.0012031752848997712, + "learning_rate": 0.2966051881578441, + "loss": 0.3052, + "num_input_tokens_seen": 5689376, + "step": 2715 + }, + { + "epoch": 0.4437556081246431, + "grad_norm": 0.003128199838101864, + "learning_rate": 0.29659271573302676, + "loss": 0.384, + "num_input_tokens_seen": 5700960, + "step": 2720 + }, + { + "epoch": 0.4445713353454605, + "grad_norm": 0.0019012097036466002, + "learning_rate": 0.2965802207017705, + "loss": 0.3203, + "num_input_tokens_seen": 5711680, + "step": 2725 + }, + { + "epoch": 0.44538706256627786, + "grad_norm": 0.0008627412607893348, + "learning_rate": 0.2965677030660021, + "loss": 0.3678, + "num_input_tokens_seen": 5721328, + "step": 2730 + }, + { + "epoch": 0.4462027897870952, + "grad_norm": 0.0009213874582201242, + "learning_rate": 0.2965551628276521, + "loss": 0.326, + "num_input_tokens_seen": 5732480, + "step": 2735 + }, + { + "epoch": 0.44701851700791256, + "grad_norm": 0.0007076473557390273, + "learning_rate": 0.29654259998865423, + "loss": 0.3343, + "num_input_tokens_seen": 5742896, + "step": 2740 + }, + { + "epoch": 0.4478342442287299, + "grad_norm": 0.0025438861921429634, + "learning_rate": 0.2965300145509458, + "loss": 0.3014, + "num_input_tokens_seen": 5752304, + "step": 2745 + }, + { + "epoch": 0.44864997144954727, + "grad_norm": 0.0019465407822281122, + "learning_rate": 0.2965174065164678, + "loss": 0.3314, + "num_input_tokens_seen": 5764032, + "step": 2750 + }, + { + "epoch": 0.4494656986703646, + "grad_norm": 0.00177590677049011, + "learning_rate": 0.2965047758871644, + "loss": 0.2726, + "num_input_tokens_seen": 5774960, + "step": 2755 + }, + { + "epoch": 0.450281425891182, + "grad_norm": 0.002632517833262682, + "learning_rate": 0.2964921226649835, + "loss": 0.4329, + "num_input_tokens_seen": 5786256, + "step": 2760 + }, + { + "epoch": 0.4510971531119993, + "grad_norm": 0.0010765629122033715, + "learning_rate": 0.2964794468518763, + "loss": 0.3054, + "num_input_tokens_seen": 5797536, + "step": 2765 + }, + { + "epoch": 0.4519128803328167, + "grad_norm": 0.0024050006177276373, + "learning_rate": 0.2964667484497977, + "loss": 0.2621, + "num_input_tokens_seen": 5807824, + "step": 2770 + }, + { + "epoch": 0.4527286075536341, + "grad_norm": 0.0025597205385565758, + "learning_rate": 0.29645402746070587, + "loss": 0.3276, + "num_input_tokens_seen": 5819072, + "step": 2775 + }, + { + "epoch": 0.45354433477445144, + "grad_norm": 0.044100966304540634, + "learning_rate": 0.2964412838865625, + "loss": 0.4165, + "num_input_tokens_seen": 5830272, + "step": 2780 + }, + { + "epoch": 0.4543600619952688, + "grad_norm": 0.03486764803528786, + "learning_rate": 0.29642851772933293, + "loss": 0.4173, + "num_input_tokens_seen": 5841280, + "step": 2785 + }, + { + "epoch": 0.45517578921608615, + "grad_norm": 0.01169503666460514, + "learning_rate": 0.29641572899098567, + "loss": 0.4328, + "num_input_tokens_seen": 5850640, + "step": 2790 + }, + { + "epoch": 0.4559915164369035, + "grad_norm": 0.16116808354854584, + "learning_rate": 0.29640291767349314, + "loss": 0.703, + "num_input_tokens_seen": 5862704, + "step": 2795 + }, + { + "epoch": 0.45680724365772085, + "grad_norm": 0.01505367737263441, + "learning_rate": 0.2963900837788308, + "loss": 0.5992, + "num_input_tokens_seen": 5872752, + "step": 2800 + }, + { + "epoch": 0.45680724365772085, + "eval_loss": 0.3763403594493866, + "eval_runtime": 155.5528, + "eval_samples_per_second": 17.518, + "eval_steps_per_second": 8.762, + "num_input_tokens_seen": 5872752, + "step": 2800 + }, + { + "epoch": 0.4576229708785382, + "grad_norm": 0.004922375548630953, + "learning_rate": 0.2963772273089779, + "loss": 0.549, + "num_input_tokens_seen": 5882464, + "step": 2805 + }, + { + "epoch": 0.45843869809935556, + "grad_norm": 0.005202272906899452, + "learning_rate": 0.2963643482659171, + "loss": 0.4174, + "num_input_tokens_seen": 5892048, + "step": 2810 + }, + { + "epoch": 0.4592544253201729, + "grad_norm": 0.0089210644364357, + "learning_rate": 0.2963514466516345, + "loss": 0.3998, + "num_input_tokens_seen": 5901504, + "step": 2815 + }, + { + "epoch": 0.4600701525409903, + "grad_norm": 0.0014745867811143398, + "learning_rate": 0.2963385224681196, + "loss": 0.447, + "num_input_tokens_seen": 5912832, + "step": 2820 + }, + { + "epoch": 0.46088587976180767, + "grad_norm": 0.0037267878651618958, + "learning_rate": 0.29632557571736556, + "loss": 0.3485, + "num_input_tokens_seen": 5923392, + "step": 2825 + }, + { + "epoch": 0.461701606982625, + "grad_norm": 0.0035348478704690933, + "learning_rate": 0.2963126064013689, + "loss": 0.3342, + "num_input_tokens_seen": 5933968, + "step": 2830 + }, + { + "epoch": 0.4625173342034424, + "grad_norm": 0.002370834583416581, + "learning_rate": 0.29629961452212966, + "loss": 0.4533, + "num_input_tokens_seen": 5943872, + "step": 2835 + }, + { + "epoch": 0.46333306142425973, + "grad_norm": 0.0032887067645788193, + "learning_rate": 0.2962866000816513, + "loss": 0.3461, + "num_input_tokens_seen": 5954752, + "step": 2840 + }, + { + "epoch": 0.4641487886450771, + "grad_norm": 0.005084976553916931, + "learning_rate": 0.2962735630819409, + "loss": 0.338, + "num_input_tokens_seen": 5965280, + "step": 2845 + }, + { + "epoch": 0.46496451586589443, + "grad_norm": 0.0033660195767879486, + "learning_rate": 0.2962605035250089, + "loss": 0.3568, + "num_input_tokens_seen": 5975792, + "step": 2850 + }, + { + "epoch": 0.4657802430867118, + "grad_norm": 0.0066399648785591125, + "learning_rate": 0.29624742141286914, + "loss": 0.3718, + "num_input_tokens_seen": 5986480, + "step": 2855 + }, + { + "epoch": 0.46659597030752914, + "grad_norm": 0.0018381003756076097, + "learning_rate": 0.29623431674753925, + "loss": 0.3262, + "num_input_tokens_seen": 5997264, + "step": 2860 + }, + { + "epoch": 0.46741169752834655, + "grad_norm": 0.0011168667115271091, + "learning_rate": 0.29622118953103993, + "loss": 0.3976, + "num_input_tokens_seen": 6006976, + "step": 2865 + }, + { + "epoch": 0.4682274247491639, + "grad_norm": 0.0018533595139160752, + "learning_rate": 0.2962080397653957, + "loss": 0.3552, + "num_input_tokens_seen": 6016912, + "step": 2870 + }, + { + "epoch": 0.46904315196998125, + "grad_norm": 0.001554404036141932, + "learning_rate": 0.29619486745263435, + "loss": 0.3356, + "num_input_tokens_seen": 6027248, + "step": 2875 + }, + { + "epoch": 0.4698588791907986, + "grad_norm": 0.0016410815296694636, + "learning_rate": 0.2961816725947873, + "loss": 0.3577, + "num_input_tokens_seen": 6037616, + "step": 2880 + }, + { + "epoch": 0.47067460641161596, + "grad_norm": 0.004019053652882576, + "learning_rate": 0.29616845519388924, + "loss": 0.4175, + "num_input_tokens_seen": 6047600, + "step": 2885 + }, + { + "epoch": 0.4714903336324333, + "grad_norm": 0.0011450763558968902, + "learning_rate": 0.2961552152519785, + "loss": 0.3645, + "num_input_tokens_seen": 6056960, + "step": 2890 + }, + { + "epoch": 0.47230606085325066, + "grad_norm": 0.0015798121457919478, + "learning_rate": 0.29614195277109695, + "loss": 0.389, + "num_input_tokens_seen": 6067408, + "step": 2895 + }, + { + "epoch": 0.473121788074068, + "grad_norm": 0.0009438001434318721, + "learning_rate": 0.2961286677532897, + "loss": 0.3549, + "num_input_tokens_seen": 6078240, + "step": 2900 + }, + { + "epoch": 0.47393751529488537, + "grad_norm": 0.0038902137894183397, + "learning_rate": 0.2961153602006055, + "loss": 0.3424, + "num_input_tokens_seen": 6087760, + "step": 2905 + }, + { + "epoch": 0.4747532425157027, + "grad_norm": 0.00355653278529644, + "learning_rate": 0.29610203011509656, + "loss": 0.3556, + "num_input_tokens_seen": 6098512, + "step": 2910 + }, + { + "epoch": 0.47556896973652013, + "grad_norm": 0.0011501912958920002, + "learning_rate": 0.29608867749881856, + "loss": 0.2696, + "num_input_tokens_seen": 6108592, + "step": 2915 + }, + { + "epoch": 0.4763846969573375, + "grad_norm": 0.0011065282160416245, + "learning_rate": 0.29607530235383067, + "loss": 0.4512, + "num_input_tokens_seen": 6118816, + "step": 2920 + }, + { + "epoch": 0.47720042417815484, + "grad_norm": 0.001950986566953361, + "learning_rate": 0.2960619046821954, + "loss": 0.3944, + "num_input_tokens_seen": 6129136, + "step": 2925 + }, + { + "epoch": 0.4780161513989722, + "grad_norm": 0.002957879099994898, + "learning_rate": 0.2960484844859789, + "loss": 0.377, + "num_input_tokens_seen": 6138768, + "step": 2930 + }, + { + "epoch": 0.47883187861978954, + "grad_norm": 0.00087858916958794, + "learning_rate": 0.29603504176725076, + "loss": 0.4014, + "num_input_tokens_seen": 6149824, + "step": 2935 + }, + { + "epoch": 0.4796476058406069, + "grad_norm": 0.0008996732067316771, + "learning_rate": 0.296021576528084, + "loss": 0.3487, + "num_input_tokens_seen": 6159184, + "step": 2940 + }, + { + "epoch": 0.48046333306142425, + "grad_norm": 0.0006052746321074665, + "learning_rate": 0.29600808877055507, + "loss": 0.3588, + "num_input_tokens_seen": 6168000, + "step": 2945 + }, + { + "epoch": 0.4812790602822416, + "grad_norm": 0.001752154086716473, + "learning_rate": 0.29599457849674404, + "loss": 0.3412, + "num_input_tokens_seen": 6178512, + "step": 2950 + }, + { + "epoch": 0.48209478750305895, + "grad_norm": 0.0017100481782108545, + "learning_rate": 0.2959810457087343, + "loss": 0.3574, + "num_input_tokens_seen": 6189360, + "step": 2955 + }, + { + "epoch": 0.48291051472387636, + "grad_norm": 0.0016837437869980931, + "learning_rate": 0.2959674904086128, + "loss": 0.3382, + "num_input_tokens_seen": 6199632, + "step": 2960 + }, + { + "epoch": 0.4837262419446937, + "grad_norm": 0.0014841118827462196, + "learning_rate": 0.2959539125984699, + "loss": 0.3218, + "num_input_tokens_seen": 6211280, + "step": 2965 + }, + { + "epoch": 0.48454196916551107, + "grad_norm": 0.0008986750617623329, + "learning_rate": 0.2959403122803996, + "loss": 0.3261, + "num_input_tokens_seen": 6221328, + "step": 2970 + }, + { + "epoch": 0.4853576963863284, + "grad_norm": 0.0009979759342968464, + "learning_rate": 0.2959266894564991, + "loss": 0.3536, + "num_input_tokens_seen": 6231648, + "step": 2975 + }, + { + "epoch": 0.48617342360714577, + "grad_norm": 0.0015320158563554287, + "learning_rate": 0.2959130441288692, + "loss": 0.3147, + "num_input_tokens_seen": 6243296, + "step": 2980 + }, + { + "epoch": 0.4869891508279631, + "grad_norm": 0.0012124666245654225, + "learning_rate": 0.2958993762996143, + "loss": 0.3412, + "num_input_tokens_seen": 6254080, + "step": 2985 + }, + { + "epoch": 0.4878048780487805, + "grad_norm": 0.0012792290654033422, + "learning_rate": 0.2958856859708421, + "loss": 0.3388, + "num_input_tokens_seen": 6263680, + "step": 2990 + }, + { + "epoch": 0.48862060526959783, + "grad_norm": 0.002522127004340291, + "learning_rate": 0.2958719731446638, + "loss": 0.3735, + "num_input_tokens_seen": 6274560, + "step": 2995 + }, + { + "epoch": 0.4894363324904152, + "grad_norm": 0.0006484489422291517, + "learning_rate": 0.29585823782319404, + "loss": 0.3572, + "num_input_tokens_seen": 6285232, + "step": 3000 + }, + { + "epoch": 0.4894363324904152, + "eval_loss": 0.36080917716026306, + "eval_runtime": 155.5498, + "eval_samples_per_second": 17.519, + "eval_steps_per_second": 8.762, + "num_input_tokens_seen": 6285232, + "step": 3000 + }, + { + "epoch": 0.4902520597112326, + "grad_norm": 0.0007740185828879476, + "learning_rate": 0.2958444800085511, + "loss": 0.3744, + "num_input_tokens_seen": 6296272, + "step": 3005 + }, + { + "epoch": 0.49106778693204994, + "grad_norm": 0.003669434692710638, + "learning_rate": 0.2958306997028565, + "loss": 0.3556, + "num_input_tokens_seen": 6306720, + "step": 3010 + }, + { + "epoch": 0.4918835141528673, + "grad_norm": 0.0008320682100020349, + "learning_rate": 0.2958168969082354, + "loss": 0.3301, + "num_input_tokens_seen": 6316480, + "step": 3015 + }, + { + "epoch": 0.49269924137368465, + "grad_norm": 0.0027711817529052496, + "learning_rate": 0.2958030716268164, + "loss": 0.3609, + "num_input_tokens_seen": 6326048, + "step": 3020 + }, + { + "epoch": 0.493514968594502, + "grad_norm": 0.0010422742925584316, + "learning_rate": 0.2957892238607314, + "loss": 0.3515, + "num_input_tokens_seen": 6335760, + "step": 3025 + }, + { + "epoch": 0.49433069581531935, + "grad_norm": 0.000790141464676708, + "learning_rate": 0.2957753536121161, + "loss": 0.3653, + "num_input_tokens_seen": 6345984, + "step": 3030 + }, + { + "epoch": 0.4951464230361367, + "grad_norm": 0.00197411235421896, + "learning_rate": 0.29576146088310923, + "loss": 0.3494, + "num_input_tokens_seen": 6356448, + "step": 3035 + }, + { + "epoch": 0.49596215025695406, + "grad_norm": 0.0018207589164376259, + "learning_rate": 0.2957475456758533, + "loss": 0.3615, + "num_input_tokens_seen": 6367440, + "step": 3040 + }, + { + "epoch": 0.4967778774777714, + "grad_norm": 0.0020789443515241146, + "learning_rate": 0.2957336079924944, + "loss": 0.352, + "num_input_tokens_seen": 6378768, + "step": 3045 + }, + { + "epoch": 0.49759360469858877, + "grad_norm": 0.0019111091969534755, + "learning_rate": 0.2957196478351816, + "loss": 0.351, + "num_input_tokens_seen": 6390032, + "step": 3050 + }, + { + "epoch": 0.4984093319194062, + "grad_norm": 0.0007412563427351415, + "learning_rate": 0.295705665206068, + "loss": 0.3398, + "num_input_tokens_seen": 6400784, + "step": 3055 + }, + { + "epoch": 0.4992250591402235, + "grad_norm": 0.002972503425553441, + "learning_rate": 0.2956916601073097, + "loss": 0.3125, + "num_input_tokens_seen": 6410320, + "step": 3060 + }, + { + "epoch": 0.5000407863610409, + "grad_norm": 0.0013289054622873664, + "learning_rate": 0.29567763254106655, + "loss": 0.333, + "num_input_tokens_seen": 6420368, + "step": 3065 + }, + { + "epoch": 0.5008565135818582, + "grad_norm": 0.0017550404882058501, + "learning_rate": 0.29566358250950175, + "loss": 0.396, + "num_input_tokens_seen": 6430912, + "step": 3070 + }, + { + "epoch": 0.5016722408026756, + "grad_norm": 0.0025683206040412188, + "learning_rate": 0.295649510014782, + "loss": 0.341, + "num_input_tokens_seen": 6440720, + "step": 3075 + }, + { + "epoch": 0.5024879680234929, + "grad_norm": 0.0009079358424060047, + "learning_rate": 0.2956354150590775, + "loss": 0.3921, + "num_input_tokens_seen": 6450800, + "step": 3080 + }, + { + "epoch": 0.5033036952443103, + "grad_norm": 0.0007847478846088052, + "learning_rate": 0.2956212976445618, + "loss": 0.3481, + "num_input_tokens_seen": 6460752, + "step": 3085 + }, + { + "epoch": 0.5041194224651276, + "grad_norm": 0.002690050285309553, + "learning_rate": 0.295607157773412, + "loss": 0.3336, + "num_input_tokens_seen": 6470208, + "step": 3090 + }, + { + "epoch": 0.504935149685945, + "grad_norm": 0.003966998308897018, + "learning_rate": 0.2955929954478087, + "loss": 0.3706, + "num_input_tokens_seen": 6480768, + "step": 3095 + }, + { + "epoch": 0.5057508769067623, + "grad_norm": 0.0023496695794165134, + "learning_rate": 0.29557881066993585, + "loss": 0.3767, + "num_input_tokens_seen": 6490928, + "step": 3100 + }, + { + "epoch": 0.5065666041275797, + "grad_norm": 0.0033786895219236612, + "learning_rate": 0.29556460344198093, + "loss": 0.3617, + "num_input_tokens_seen": 6501856, + "step": 3105 + }, + { + "epoch": 0.507382331348397, + "grad_norm": 0.0007717047119513154, + "learning_rate": 0.29555037376613486, + "loss": 0.3539, + "num_input_tokens_seen": 6513232, + "step": 3110 + }, + { + "epoch": 0.5081980585692144, + "grad_norm": 0.0009552224655635655, + "learning_rate": 0.29553612164459203, + "loss": 0.3242, + "num_input_tokens_seen": 6523632, + "step": 3115 + }, + { + "epoch": 0.5090137857900318, + "grad_norm": 0.003004846628755331, + "learning_rate": 0.29552184707955037, + "loss": 0.367, + "num_input_tokens_seen": 6534176, + "step": 3120 + }, + { + "epoch": 0.5098295130108492, + "grad_norm": 0.0013739587739109993, + "learning_rate": 0.29550755007321117, + "loss": 0.323, + "num_input_tokens_seen": 6544912, + "step": 3125 + }, + { + "epoch": 0.5106452402316666, + "grad_norm": 0.002363866427913308, + "learning_rate": 0.29549323062777916, + "loss": 0.34, + "num_input_tokens_seen": 6554656, + "step": 3130 + }, + { + "epoch": 0.5114609674524839, + "grad_norm": 0.0023303425405174494, + "learning_rate": 0.29547888874546263, + "loss": 0.3164, + "num_input_tokens_seen": 6564032, + "step": 3135 + }, + { + "epoch": 0.5122766946733013, + "grad_norm": 0.0010347734205424786, + "learning_rate": 0.2954645244284732, + "loss": 0.3491, + "num_input_tokens_seen": 6574544, + "step": 3140 + }, + { + "epoch": 0.5130924218941186, + "grad_norm": 0.0009075491689145565, + "learning_rate": 0.2954501376790261, + "loss": 0.3293, + "num_input_tokens_seen": 6584448, + "step": 3145 + }, + { + "epoch": 0.513908149114936, + "grad_norm": 0.0011109380284324288, + "learning_rate": 0.29543572849933997, + "loss": 0.3527, + "num_input_tokens_seen": 6593552, + "step": 3150 + }, + { + "epoch": 0.5147238763357533, + "grad_norm": 0.0009571441914886236, + "learning_rate": 0.2954212968916368, + "loss": 0.3327, + "num_input_tokens_seen": 6602464, + "step": 3155 + }, + { + "epoch": 0.5155396035565707, + "grad_norm": 0.0009727685246616602, + "learning_rate": 0.29540684285814217, + "loss": 0.338, + "num_input_tokens_seen": 6612000, + "step": 3160 + }, + { + "epoch": 0.516355330777388, + "grad_norm": 0.0010991772869601846, + "learning_rate": 0.2953923664010851, + "loss": 0.3812, + "num_input_tokens_seen": 6623184, + "step": 3165 + }, + { + "epoch": 0.5171710579982054, + "grad_norm": 0.001220783917233348, + "learning_rate": 0.295377867522698, + "loss": 0.3398, + "num_input_tokens_seen": 6634384, + "step": 3170 + }, + { + "epoch": 0.5179867852190227, + "grad_norm": 0.003890488063916564, + "learning_rate": 0.2953633462252168, + "loss": 0.3508, + "num_input_tokens_seen": 6644624, + "step": 3175 + }, + { + "epoch": 0.5188025124398401, + "grad_norm": 0.0018833045614883304, + "learning_rate": 0.2953488025108809, + "loss": 0.352, + "num_input_tokens_seen": 6654848, + "step": 3180 + }, + { + "epoch": 0.5196182396606575, + "grad_norm": 0.001407943549565971, + "learning_rate": 0.295334236381933, + "loss": 0.348, + "num_input_tokens_seen": 6665904, + "step": 3185 + }, + { + "epoch": 0.5204339668814748, + "grad_norm": 0.0018224777886644006, + "learning_rate": 0.29531964784061954, + "loss": 0.3451, + "num_input_tokens_seen": 6677872, + "step": 3190 + }, + { + "epoch": 0.5212496941022922, + "grad_norm": 0.0018023482989519835, + "learning_rate": 0.2953050368891902, + "loss": 0.3526, + "num_input_tokens_seen": 6688640, + "step": 3195 + }, + { + "epoch": 0.5220654213231095, + "grad_norm": 0.0006180281052365899, + "learning_rate": 0.29529040352989805, + "loss": 0.3438, + "num_input_tokens_seen": 6699264, + "step": 3200 + }, + { + "epoch": 0.5220654213231095, + "eval_loss": 0.35246187448501587, + "eval_runtime": 155.5893, + "eval_samples_per_second": 17.514, + "eval_steps_per_second": 8.76, + "num_input_tokens_seen": 6699264, + "step": 3200 + }, + { + "epoch": 0.5228811485439269, + "grad_norm": 0.0008175882394425571, + "learning_rate": 0.29527574776499993, + "loss": 0.3528, + "num_input_tokens_seen": 6710016, + "step": 3205 + }, + { + "epoch": 0.5236968757647442, + "grad_norm": 0.0015119300223886967, + "learning_rate": 0.2952610695967558, + "loss": 0.351, + "num_input_tokens_seen": 6720304, + "step": 3210 + }, + { + "epoch": 0.5245126029855617, + "grad_norm": 0.002173058921471238, + "learning_rate": 0.29524636902742935, + "loss": 0.3656, + "num_input_tokens_seen": 6730784, + "step": 3215 + }, + { + "epoch": 0.525328330206379, + "grad_norm": 0.0013735227985307574, + "learning_rate": 0.2952316460592875, + "loss": 0.3481, + "num_input_tokens_seen": 6740208, + "step": 3220 + }, + { + "epoch": 0.5261440574271964, + "grad_norm": 0.000541451561730355, + "learning_rate": 0.29521690069460066, + "loss": 0.3445, + "num_input_tokens_seen": 6750864, + "step": 3225 + }, + { + "epoch": 0.5269597846480137, + "grad_norm": 0.001269191619940102, + "learning_rate": 0.29520213293564285, + "loss": 0.3319, + "num_input_tokens_seen": 6761072, + "step": 3230 + }, + { + "epoch": 0.5277755118688311, + "grad_norm": 0.0025236839428544044, + "learning_rate": 0.29518734278469144, + "loss": 0.3111, + "num_input_tokens_seen": 6770928, + "step": 3235 + }, + { + "epoch": 0.5285912390896484, + "grad_norm": 0.0010894235456362367, + "learning_rate": 0.29517253024402723, + "loss": 0.3331, + "num_input_tokens_seen": 6781936, + "step": 3240 + }, + { + "epoch": 0.5294069663104658, + "grad_norm": 0.002050969284027815, + "learning_rate": 0.2951576953159345, + "loss": 0.2916, + "num_input_tokens_seen": 6791632, + "step": 3245 + }, + { + "epoch": 0.5302226935312832, + "grad_norm": 0.0020102905109524727, + "learning_rate": 0.29514283800270097, + "loss": 0.2874, + "num_input_tokens_seen": 6802000, + "step": 3250 + }, + { + "epoch": 0.5310384207521005, + "grad_norm": 0.0007512721349485219, + "learning_rate": 0.2951279583066179, + "loss": 0.4673, + "num_input_tokens_seen": 6812976, + "step": 3255 + }, + { + "epoch": 0.5318541479729179, + "grad_norm": 0.0020206605549901724, + "learning_rate": 0.2951130562299798, + "loss": 0.4041, + "num_input_tokens_seen": 6822992, + "step": 3260 + }, + { + "epoch": 0.5326698751937352, + "grad_norm": 0.0016776270931586623, + "learning_rate": 0.29509813177508487, + "loss": 0.3422, + "num_input_tokens_seen": 6833680, + "step": 3265 + }, + { + "epoch": 0.5334856024145526, + "grad_norm": 0.0005982061265967786, + "learning_rate": 0.2950831849442346, + "loss": 0.3533, + "num_input_tokens_seen": 6843968, + "step": 3270 + }, + { + "epoch": 0.5343013296353699, + "grad_norm": 0.0012573677813634276, + "learning_rate": 0.2950682157397339, + "loss": 0.373, + "num_input_tokens_seen": 6855040, + "step": 3275 + }, + { + "epoch": 0.5351170568561873, + "grad_norm": 0.0006232629530131817, + "learning_rate": 0.2950532241638914, + "loss": 0.3527, + "num_input_tokens_seen": 6866688, + "step": 3280 + }, + { + "epoch": 0.5359327840770046, + "grad_norm": 0.0014188137138262391, + "learning_rate": 0.2950382102190188, + "loss": 0.3561, + "num_input_tokens_seen": 6877808, + "step": 3285 + }, + { + "epoch": 0.536748511297822, + "grad_norm": 0.0024665119126439095, + "learning_rate": 0.2950231739074316, + "loss": 0.3342, + "num_input_tokens_seen": 6887824, + "step": 3290 + }, + { + "epoch": 0.5375642385186393, + "grad_norm": 0.0007493851589970291, + "learning_rate": 0.29500811523144843, + "loss": 0.332, + "num_input_tokens_seen": 6897936, + "step": 3295 + }, + { + "epoch": 0.5383799657394567, + "grad_norm": 0.0008521305862814188, + "learning_rate": 0.2949930341933917, + "loss": 0.3521, + "num_input_tokens_seen": 6908448, + "step": 3300 + }, + { + "epoch": 0.539195692960274, + "grad_norm": 0.000808502605650574, + "learning_rate": 0.29497793079558693, + "loss": 0.3247, + "num_input_tokens_seen": 6919680, + "step": 3305 + }, + { + "epoch": 0.5400114201810915, + "grad_norm": 0.0007616779766976833, + "learning_rate": 0.2949628050403633, + "loss": 0.2775, + "num_input_tokens_seen": 6930528, + "step": 3310 + }, + { + "epoch": 0.5408271474019088, + "grad_norm": 0.0015597775345668197, + "learning_rate": 0.2949476569300535, + "loss": 0.2243, + "num_input_tokens_seen": 6941360, + "step": 3315 + }, + { + "epoch": 0.5416428746227262, + "grad_norm": 0.0015994628192856908, + "learning_rate": 0.29493248646699344, + "loss": 0.3807, + "num_input_tokens_seen": 6951440, + "step": 3320 + }, + { + "epoch": 0.5424586018435436, + "grad_norm": 0.0014743269421160221, + "learning_rate": 0.29491729365352265, + "loss": 0.4059, + "num_input_tokens_seen": 6961856, + "step": 3325 + }, + { + "epoch": 0.5432743290643609, + "grad_norm": 0.0006350327748805285, + "learning_rate": 0.29490207849198397, + "loss": 0.3666, + "num_input_tokens_seen": 6971808, + "step": 3330 + }, + { + "epoch": 0.5440900562851783, + "grad_norm": 0.0007747460040263832, + "learning_rate": 0.29488684098472384, + "loss": 0.3403, + "num_input_tokens_seen": 6981968, + "step": 3335 + }, + { + "epoch": 0.5449057835059956, + "grad_norm": 0.000987814157269895, + "learning_rate": 0.2948715811340921, + "loss": 0.326, + "num_input_tokens_seen": 6992224, + "step": 3340 + }, + { + "epoch": 0.545721510726813, + "grad_norm": 0.0005675693973898888, + "learning_rate": 0.294856298942442, + "loss": 0.3602, + "num_input_tokens_seen": 7002768, + "step": 3345 + }, + { + "epoch": 0.5465372379476303, + "grad_norm": 0.0015274516772478819, + "learning_rate": 0.2948409944121302, + "loss": 0.3557, + "num_input_tokens_seen": 7012656, + "step": 3350 + }, + { + "epoch": 0.5473529651684477, + "grad_norm": 0.0015756093198433518, + "learning_rate": 0.29482566754551687, + "loss": 0.3545, + "num_input_tokens_seen": 7022368, + "step": 3355 + }, + { + "epoch": 0.548168692389265, + "grad_norm": 0.0005804491811431944, + "learning_rate": 0.2948103183449656, + "loss": 0.3514, + "num_input_tokens_seen": 7033024, + "step": 3360 + }, + { + "epoch": 0.5489844196100824, + "grad_norm": 0.0014394688187167048, + "learning_rate": 0.2947949468128435, + "loss": 0.3403, + "num_input_tokens_seen": 7043536, + "step": 3365 + }, + { + "epoch": 0.5498001468308997, + "grad_norm": 0.0012758499942719936, + "learning_rate": 0.2947795529515209, + "loss": 0.3592, + "num_input_tokens_seen": 7053824, + "step": 3370 + }, + { + "epoch": 0.5506158740517171, + "grad_norm": 0.0017407946288585663, + "learning_rate": 0.29476413676337193, + "loss": 0.3515, + "num_input_tokens_seen": 7064464, + "step": 3375 + }, + { + "epoch": 0.5514316012725344, + "grad_norm": 0.001590934582054615, + "learning_rate": 0.2947486982507738, + "loss": 0.3432, + "num_input_tokens_seen": 7075408, + "step": 3380 + }, + { + "epoch": 0.5522473284933518, + "grad_norm": 0.0005859872326254845, + "learning_rate": 0.29473323741610735, + "loss": 0.3381, + "num_input_tokens_seen": 7087296, + "step": 3385 + }, + { + "epoch": 0.5530630557141691, + "grad_norm": 0.0011597155826166272, + "learning_rate": 0.2947177542617569, + "loss": 0.3444, + "num_input_tokens_seen": 7098240, + "step": 3390 + }, + { + "epoch": 0.5538787829349865, + "grad_norm": 0.0011653074761852622, + "learning_rate": 0.2947022487901101, + "loss": 0.3399, + "num_input_tokens_seen": 7108736, + "step": 3395 + }, + { + "epoch": 0.5546945101558038, + "grad_norm": 0.0019088761182501912, + "learning_rate": 0.2946867210035581, + "loss": 0.3398, + "num_input_tokens_seen": 7118336, + "step": 3400 + }, + { + "epoch": 0.5546945101558038, + "eval_loss": 0.3392236828804016, + "eval_runtime": 155.6304, + "eval_samples_per_second": 17.509, + "eval_steps_per_second": 8.758, + "num_input_tokens_seen": 7118336, + "step": 3400 + }, + { + "epoch": 0.5555102373766213, + "grad_norm": 0.0006772142369300127, + "learning_rate": 0.2946711709044954, + "loss": 0.3599, + "num_input_tokens_seen": 7128304, + "step": 3405 + }, + { + "epoch": 0.5563259645974387, + "grad_norm": 0.0009665507241152227, + "learning_rate": 0.2946555984953202, + "loss": 0.3391, + "num_input_tokens_seen": 7137456, + "step": 3410 + }, + { + "epoch": 0.557141691818256, + "grad_norm": 0.0007525041582994163, + "learning_rate": 0.2946400037784338, + "loss": 0.3624, + "num_input_tokens_seen": 7147712, + "step": 3415 + }, + { + "epoch": 0.5579574190390734, + "grad_norm": 0.0016252625500783324, + "learning_rate": 0.29462438675624114, + "loss": 0.3362, + "num_input_tokens_seen": 7158608, + "step": 3420 + }, + { + "epoch": 0.5587731462598907, + "grad_norm": 0.001487727160565555, + "learning_rate": 0.2946087474311506, + "loss": 0.3547, + "num_input_tokens_seen": 7168880, + "step": 3425 + }, + { + "epoch": 0.5595888734807081, + "grad_norm": 0.0006462738383561373, + "learning_rate": 0.294593085805574, + "loss": 0.3528, + "num_input_tokens_seen": 7178576, + "step": 3430 + }, + { + "epoch": 0.5604046007015254, + "grad_norm": 0.0012036421103402972, + "learning_rate": 0.2945774018819264, + "loss": 0.3539, + "num_input_tokens_seen": 7189488, + "step": 3435 + }, + { + "epoch": 0.5612203279223428, + "grad_norm": 0.0011934980284422636, + "learning_rate": 0.2945616956626266, + "loss": 0.357, + "num_input_tokens_seen": 7199584, + "step": 3440 + }, + { + "epoch": 0.5620360551431601, + "grad_norm": 0.000627143366727978, + "learning_rate": 0.2945459671500966, + "loss": 0.3441, + "num_input_tokens_seen": 7208400, + "step": 3445 + }, + { + "epoch": 0.5628517823639775, + "grad_norm": 0.0013885938096791506, + "learning_rate": 0.2945302163467621, + "loss": 0.3496, + "num_input_tokens_seen": 7218704, + "step": 3450 + }, + { + "epoch": 0.5636675095847948, + "grad_norm": 0.002843694295734167, + "learning_rate": 0.2945144432550519, + "loss": 0.3569, + "num_input_tokens_seen": 7230304, + "step": 3455 + }, + { + "epoch": 0.5644832368056122, + "grad_norm": 0.0015580805484205484, + "learning_rate": 0.29449864787739843, + "loss": 0.3535, + "num_input_tokens_seen": 7240912, + "step": 3460 + }, + { + "epoch": 0.5652989640264295, + "grad_norm": 0.0026018237695097923, + "learning_rate": 0.2944828302162376, + "loss": 0.3625, + "num_input_tokens_seen": 7251872, + "step": 3465 + }, + { + "epoch": 0.5661146912472469, + "grad_norm": 0.0005846542771905661, + "learning_rate": 0.2944669902740087, + "loss": 0.3521, + "num_input_tokens_seen": 7262336, + "step": 3470 + }, + { + "epoch": 0.5669304184680642, + "grad_norm": 0.0017351104179397225, + "learning_rate": 0.2944511280531544, + "loss": 0.3514, + "num_input_tokens_seen": 7273344, + "step": 3475 + }, + { + "epoch": 0.5677461456888816, + "grad_norm": 0.0006473526591435075, + "learning_rate": 0.29443524355612083, + "loss": 0.3444, + "num_input_tokens_seen": 7282848, + "step": 3480 + }, + { + "epoch": 0.568561872909699, + "grad_norm": 0.0005010878667235374, + "learning_rate": 0.29441933678535764, + "loss": 0.3594, + "num_input_tokens_seen": 7293424, + "step": 3485 + }, + { + "epoch": 0.5693776001305163, + "grad_norm": 0.0008196360431611538, + "learning_rate": 0.29440340774331786, + "loss": 0.3526, + "num_input_tokens_seen": 7304192, + "step": 3490 + }, + { + "epoch": 0.5701933273513338, + "grad_norm": 0.0012767212465405464, + "learning_rate": 0.2943874564324579, + "loss": 0.3221, + "num_input_tokens_seen": 7315488, + "step": 3495 + }, + { + "epoch": 0.5710090545721511, + "grad_norm": 0.0006293316837400198, + "learning_rate": 0.2943714828552376, + "loss": 0.3366, + "num_input_tokens_seen": 7325840, + "step": 3500 + }, + { + "epoch": 0.5718247817929685, + "grad_norm": 0.00097968231420964, + "learning_rate": 0.29435548701412045, + "loss": 0.3359, + "num_input_tokens_seen": 7334864, + "step": 3505 + }, + { + "epoch": 0.5726405090137858, + "grad_norm": 0.002128956839442253, + "learning_rate": 0.2943394689115731, + "loss": 0.3343, + "num_input_tokens_seen": 7345472, + "step": 3510 + }, + { + "epoch": 0.5734562362346032, + "grad_norm": 0.0037135735619813204, + "learning_rate": 0.29432342855006577, + "loss": 0.3525, + "num_input_tokens_seen": 7356752, + "step": 3515 + }, + { + "epoch": 0.5742719634554205, + "grad_norm": 0.0009089869563467801, + "learning_rate": 0.294307365932072, + "loss": 0.3698, + "num_input_tokens_seen": 7366736, + "step": 3520 + }, + { + "epoch": 0.5750876906762379, + "grad_norm": 0.00240536080673337, + "learning_rate": 0.294291281060069, + "loss": 0.3418, + "num_input_tokens_seen": 7376416, + "step": 3525 + }, + { + "epoch": 0.5759034178970552, + "grad_norm": 0.001395562314428389, + "learning_rate": 0.29427517393653724, + "loss": 0.3436, + "num_input_tokens_seen": 7387456, + "step": 3530 + }, + { + "epoch": 0.5767191451178726, + "grad_norm": 0.0017167842015624046, + "learning_rate": 0.29425904456396046, + "loss": 0.3426, + "num_input_tokens_seen": 7398704, + "step": 3535 + }, + { + "epoch": 0.5775348723386899, + "grad_norm": 0.0012523906771093607, + "learning_rate": 0.2942428929448262, + "loss": 0.3547, + "num_input_tokens_seen": 7408832, + "step": 3540 + }, + { + "epoch": 0.5783505995595073, + "grad_norm": 0.0011418481590226293, + "learning_rate": 0.2942267190816252, + "loss": 0.3259, + "num_input_tokens_seen": 7419120, + "step": 3545 + }, + { + "epoch": 0.5791663267803246, + "grad_norm": 0.0007797508151270449, + "learning_rate": 0.2942105229768516, + "loss": 0.3541, + "num_input_tokens_seen": 7429712, + "step": 3550 + }, + { + "epoch": 0.579982054001142, + "grad_norm": 0.0008592383819632232, + "learning_rate": 0.29419430463300306, + "loss": 0.314, + "num_input_tokens_seen": 7440480, + "step": 3555 + }, + { + "epoch": 0.5807977812219594, + "grad_norm": 0.0018498505232855678, + "learning_rate": 0.2941780640525808, + "loss": 0.3394, + "num_input_tokens_seen": 7451840, + "step": 3560 + }, + { + "epoch": 0.5816135084427767, + "grad_norm": 0.0010221642442047596, + "learning_rate": 0.2941618012380891, + "loss": 0.3492, + "num_input_tokens_seen": 7461696, + "step": 3565 + }, + { + "epoch": 0.5824292356635941, + "grad_norm": 0.0006964880740270019, + "learning_rate": 0.29414551619203605, + "loss": 0.3391, + "num_input_tokens_seen": 7471568, + "step": 3570 + }, + { + "epoch": 0.5832449628844114, + "grad_norm": 0.0005947811296209693, + "learning_rate": 0.29412920891693295, + "loss": 0.3607, + "num_input_tokens_seen": 7482512, + "step": 3575 + }, + { + "epoch": 0.5840606901052288, + "grad_norm": 0.0005742863286286592, + "learning_rate": 0.2941128794152946, + "loss": 0.3361, + "num_input_tokens_seen": 7492016, + "step": 3580 + }, + { + "epoch": 0.5848764173260461, + "grad_norm": 0.0016339692519977689, + "learning_rate": 0.2940965276896392, + "loss": 0.35, + "num_input_tokens_seen": 7501328, + "step": 3585 + }, + { + "epoch": 0.5856921445468636, + "grad_norm": 0.000675518880598247, + "learning_rate": 0.2940801537424884, + "loss": 0.3475, + "num_input_tokens_seen": 7511952, + "step": 3590 + }, + { + "epoch": 0.5865078717676809, + "grad_norm": 0.002551097422838211, + "learning_rate": 0.2940637575763673, + "loss": 0.3425, + "num_input_tokens_seen": 7523520, + "step": 3595 + }, + { + "epoch": 0.5873235989884983, + "grad_norm": 0.0024090514052659273, + "learning_rate": 0.2940473391938043, + "loss": 0.3375, + "num_input_tokens_seen": 7533408, + "step": 3600 + }, + { + "epoch": 0.5873235989884983, + "eval_loss": 0.34082770347595215, + "eval_runtime": 155.6304, + "eval_samples_per_second": 17.509, + "eval_steps_per_second": 8.758, + "num_input_tokens_seen": 7533408, + "step": 3600 + }, + { + "epoch": 0.5881393262093156, + "grad_norm": 0.0006290565361268818, + "learning_rate": 0.29403089859733145, + "loss": 0.373, + "num_input_tokens_seen": 7543760, + "step": 3605 + }, + { + "epoch": 0.588955053430133, + "grad_norm": 0.001127478783018887, + "learning_rate": 0.294014435789484, + "loss": 0.3451, + "num_input_tokens_seen": 7554320, + "step": 3610 + }, + { + "epoch": 0.5897707806509503, + "grad_norm": 0.0011502604465931654, + "learning_rate": 0.2939979507728007, + "loss": 0.3359, + "num_input_tokens_seen": 7564672, + "step": 3615 + }, + { + "epoch": 0.5905865078717677, + "grad_norm": 0.0010107973357662559, + "learning_rate": 0.2939814435498239, + "loss": 0.3351, + "num_input_tokens_seen": 7574608, + "step": 3620 + }, + { + "epoch": 0.591402235092585, + "grad_norm": 0.0010768078500404954, + "learning_rate": 0.29396491412309905, + "loss": 0.3162, + "num_input_tokens_seen": 7586704, + "step": 3625 + }, + { + "epoch": 0.5922179623134024, + "grad_norm": 0.0009834383381530643, + "learning_rate": 0.2939483624951753, + "loss": 0.3572, + "num_input_tokens_seen": 7596992, + "step": 3630 + }, + { + "epoch": 0.5930336895342198, + "grad_norm": 0.0009693442261777818, + "learning_rate": 0.2939317886686051, + "loss": 0.3073, + "num_input_tokens_seen": 7607376, + "step": 3635 + }, + { + "epoch": 0.5938494167550371, + "grad_norm": 0.0010234774090349674, + "learning_rate": 0.2939151926459443, + "loss": 0.3235, + "num_input_tokens_seen": 7616608, + "step": 3640 + }, + { + "epoch": 0.5946651439758545, + "grad_norm": 0.002642586827278137, + "learning_rate": 0.2938985744297522, + "loss": 0.3925, + "num_input_tokens_seen": 7626416, + "step": 3645 + }, + { + "epoch": 0.5954808711966718, + "grad_norm": 0.0021748223807662725, + "learning_rate": 0.29388193402259166, + "loss": 0.3579, + "num_input_tokens_seen": 7637104, + "step": 3650 + }, + { + "epoch": 0.5962965984174892, + "grad_norm": 0.0007574478513561189, + "learning_rate": 0.29386527142702873, + "loss": 0.3447, + "num_input_tokens_seen": 7648816, + "step": 3655 + }, + { + "epoch": 0.5971123256383065, + "grad_norm": 0.0005829168949276209, + "learning_rate": 0.293848586645633, + "loss": 0.3594, + "num_input_tokens_seen": 7660336, + "step": 3660 + }, + { + "epoch": 0.5979280528591239, + "grad_norm": 0.0014376264298334718, + "learning_rate": 0.2938318796809775, + "loss": 0.3572, + "num_input_tokens_seen": 7670336, + "step": 3665 + }, + { + "epoch": 0.5987437800799412, + "grad_norm": 0.0006063358741812408, + "learning_rate": 0.29381515053563867, + "loss": 0.3455, + "num_input_tokens_seen": 7680000, + "step": 3670 + }, + { + "epoch": 0.5995595073007586, + "grad_norm": 0.0012612176360562444, + "learning_rate": 0.29379839921219636, + "loss": 0.3456, + "num_input_tokens_seen": 7690912, + "step": 3675 + }, + { + "epoch": 0.600375234521576, + "grad_norm": 0.0006609547417610884, + "learning_rate": 0.2937816257132338, + "loss": 0.3508, + "num_input_tokens_seen": 7703616, + "step": 3680 + }, + { + "epoch": 0.6011909617423934, + "grad_norm": 0.0018144131172448397, + "learning_rate": 0.2937648300413376, + "loss": 0.3477, + "num_input_tokens_seen": 7711952, + "step": 3685 + }, + { + "epoch": 0.6020066889632107, + "grad_norm": 0.0011835353216156363, + "learning_rate": 0.293748012199098, + "loss": 0.3561, + "num_input_tokens_seen": 7723216, + "step": 3690 + }, + { + "epoch": 0.6028224161840281, + "grad_norm": 0.0008448885637335479, + "learning_rate": 0.29373117218910844, + "loss": 0.3467, + "num_input_tokens_seen": 7732192, + "step": 3695 + }, + { + "epoch": 0.6036381434048455, + "grad_norm": 0.0017567477189004421, + "learning_rate": 0.2937143100139659, + "loss": 0.362, + "num_input_tokens_seen": 7742144, + "step": 3700 + }, + { + "epoch": 0.6044538706256628, + "grad_norm": 0.0015230164863169193, + "learning_rate": 0.29369742567627083, + "loss": 0.3561, + "num_input_tokens_seen": 7753600, + "step": 3705 + }, + { + "epoch": 0.6052695978464802, + "grad_norm": 0.0015012891963124275, + "learning_rate": 0.29368051917862675, + "loss": 0.3572, + "num_input_tokens_seen": 7763248, + "step": 3710 + }, + { + "epoch": 0.6060853250672975, + "grad_norm": 0.0014523961581289768, + "learning_rate": 0.2936635905236411, + "loss": 0.3431, + "num_input_tokens_seen": 7773424, + "step": 3715 + }, + { + "epoch": 0.6069010522881149, + "grad_norm": 0.0007238282705657184, + "learning_rate": 0.2936466397139244, + "loss": 0.3479, + "num_input_tokens_seen": 7783248, + "step": 3720 + }, + { + "epoch": 0.6077167795089322, + "grad_norm": 0.0006236144690774381, + "learning_rate": 0.2936296667520907, + "loss": 0.3557, + "num_input_tokens_seen": 7794048, + "step": 3725 + }, + { + "epoch": 0.6085325067297496, + "grad_norm": 0.000576700025703758, + "learning_rate": 0.2936126716407574, + "loss": 0.337, + "num_input_tokens_seen": 7803920, + "step": 3730 + }, + { + "epoch": 0.6093482339505669, + "grad_norm": 0.002927131485193968, + "learning_rate": 0.29359565438254537, + "loss": 0.3726, + "num_input_tokens_seen": 7816048, + "step": 3735 + }, + { + "epoch": 0.6101639611713843, + "grad_norm": 0.002276123035699129, + "learning_rate": 0.29357861498007887, + "loss": 0.3393, + "num_input_tokens_seen": 7827056, + "step": 3740 + }, + { + "epoch": 0.6109796883922016, + "grad_norm": 0.0006624266388826072, + "learning_rate": 0.29356155343598567, + "loss": 0.3628, + "num_input_tokens_seen": 7837392, + "step": 3745 + }, + { + "epoch": 0.611795415613019, + "grad_norm": 0.000532898644451052, + "learning_rate": 0.2935444697528968, + "loss": 0.3549, + "num_input_tokens_seen": 7847888, + "step": 3750 + }, + { + "epoch": 0.6126111428338363, + "grad_norm": 0.0011791781289502978, + "learning_rate": 0.2935273639334468, + "loss": 0.3485, + "num_input_tokens_seen": 7858064, + "step": 3755 + }, + { + "epoch": 0.6134268700546537, + "grad_norm": 0.0006830209749750793, + "learning_rate": 0.29351023598027365, + "loss": 0.3371, + "num_input_tokens_seen": 7867648, + "step": 3760 + }, + { + "epoch": 0.614242597275471, + "grad_norm": 0.001046227291226387, + "learning_rate": 0.2934930858960186, + "loss": 0.3426, + "num_input_tokens_seen": 7878496, + "step": 3765 + }, + { + "epoch": 0.6150583244962884, + "grad_norm": 0.002221597358584404, + "learning_rate": 0.29347591368332643, + "loss": 0.3106, + "num_input_tokens_seen": 7890016, + "step": 3770 + }, + { + "epoch": 0.6158740517171059, + "grad_norm": 0.000855901394970715, + "learning_rate": 0.2934587193448454, + "loss": 0.3474, + "num_input_tokens_seen": 7901056, + "step": 3775 + }, + { + "epoch": 0.6166897789379232, + "grad_norm": 0.0008113554795272648, + "learning_rate": 0.29344150288322696, + "loss": 0.3485, + "num_input_tokens_seen": 7911296, + "step": 3780 + }, + { + "epoch": 0.6175055061587406, + "grad_norm": 0.0007325623882934451, + "learning_rate": 0.2934242643011263, + "loss": 0.4063, + "num_input_tokens_seen": 7920896, + "step": 3785 + }, + { + "epoch": 0.6183212333795579, + "grad_norm": 0.0006423689774237573, + "learning_rate": 0.2934070036012016, + "loss": 0.3541, + "num_input_tokens_seen": 7931168, + "step": 3790 + }, + { + "epoch": 0.6191369606003753, + "grad_norm": 0.002420038217678666, + "learning_rate": 0.29338972078611475, + "loss": 0.3352, + "num_input_tokens_seen": 7941328, + "step": 3795 + }, + { + "epoch": 0.6199526878211926, + "grad_norm": 0.0011116871610283852, + "learning_rate": 0.2933724158585311, + "loss": 0.3466, + "num_input_tokens_seen": 7950560, + "step": 3800 + }, + { + "epoch": 0.6199526878211926, + "eval_loss": 0.34035706520080566, + "eval_runtime": 155.9066, + "eval_samples_per_second": 17.478, + "eval_steps_per_second": 8.742, + "num_input_tokens_seen": 7950560, + "step": 3800 + }, + { + "epoch": 0.62076841504201, + "grad_norm": 0.0006437241099774837, + "learning_rate": 0.29335508882111916, + "loss": 0.3331, + "num_input_tokens_seen": 7960672, + "step": 3805 + }, + { + "epoch": 0.6215841422628273, + "grad_norm": 0.0009861525613814592, + "learning_rate": 0.29333773967655097, + "loss": 0.315, + "num_input_tokens_seen": 7970848, + "step": 3810 + }, + { + "epoch": 0.6223998694836447, + "grad_norm": 0.0007758465362712741, + "learning_rate": 0.2933203684275021, + "loss": 0.3138, + "num_input_tokens_seen": 7983216, + "step": 3815 + }, + { + "epoch": 0.623215596704462, + "grad_norm": 0.0006971616530790925, + "learning_rate": 0.2933029750766513, + "loss": 0.3305, + "num_input_tokens_seen": 7994064, + "step": 3820 + }, + { + "epoch": 0.6240313239252794, + "grad_norm": 0.0034114676527678967, + "learning_rate": 0.2932855596266809, + "loss": 0.3623, + "num_input_tokens_seen": 8004656, + "step": 3825 + }, + { + "epoch": 0.6248470511460967, + "grad_norm": 0.000731572974473238, + "learning_rate": 0.2932681220802765, + "loss": 0.3146, + "num_input_tokens_seen": 8015360, + "step": 3830 + }, + { + "epoch": 0.6256627783669141, + "grad_norm": 0.0008502693963237107, + "learning_rate": 0.2932506624401274, + "loss": 0.39, + "num_input_tokens_seen": 8025040, + "step": 3835 + }, + { + "epoch": 0.6264785055877314, + "grad_norm": 0.0019493624567985535, + "learning_rate": 0.29323318070892584, + "loss": 0.3568, + "num_input_tokens_seen": 8036576, + "step": 3840 + }, + { + "epoch": 0.6272942328085488, + "grad_norm": 0.000976203999016434, + "learning_rate": 0.29321567688936784, + "loss": 0.3417, + "num_input_tokens_seen": 8045728, + "step": 3845 + }, + { + "epoch": 0.6281099600293661, + "grad_norm": 0.001254623755812645, + "learning_rate": 0.29319815098415275, + "loss": 0.3486, + "num_input_tokens_seen": 8056992, + "step": 3850 + }, + { + "epoch": 0.6289256872501835, + "grad_norm": 0.00259178108535707, + "learning_rate": 0.2931806029959832, + "loss": 0.3444, + "num_input_tokens_seen": 8065504, + "step": 3855 + }, + { + "epoch": 0.6297414144710008, + "grad_norm": 0.0010226838057860732, + "learning_rate": 0.29316303292756535, + "loss": 0.3339, + "num_input_tokens_seen": 8076384, + "step": 3860 + }, + { + "epoch": 0.6305571416918182, + "grad_norm": 0.00204673083499074, + "learning_rate": 0.29314544078160876, + "loss": 0.3252, + "num_input_tokens_seen": 8087296, + "step": 3865 + }, + { + "epoch": 0.6313728689126357, + "grad_norm": 0.0008041769033297896, + "learning_rate": 0.2931278265608263, + "loss": 0.3328, + "num_input_tokens_seen": 8098368, + "step": 3870 + }, + { + "epoch": 0.632188596133453, + "grad_norm": 0.0018854743102565408, + "learning_rate": 0.29311019026793433, + "loss": 0.3463, + "num_input_tokens_seen": 8109232, + "step": 3875 + }, + { + "epoch": 0.6330043233542704, + "grad_norm": 0.0009272661409340799, + "learning_rate": 0.29309253190565254, + "loss": 0.3317, + "num_input_tokens_seen": 8119808, + "step": 3880 + }, + { + "epoch": 0.6338200505750877, + "grad_norm": 0.0008428451837971807, + "learning_rate": 0.2930748514767042, + "loss": 0.3382, + "num_input_tokens_seen": 8129376, + "step": 3885 + }, + { + "epoch": 0.6346357777959051, + "grad_norm": 0.0007760237203910947, + "learning_rate": 0.29305714898381574, + "loss": 0.3799, + "num_input_tokens_seen": 8139168, + "step": 3890 + }, + { + "epoch": 0.6354515050167224, + "grad_norm": 0.0017368971602991223, + "learning_rate": 0.29303942442971714, + "loss": 0.3566, + "num_input_tokens_seen": 8149264, + "step": 3895 + }, + { + "epoch": 0.6362672322375398, + "grad_norm": 0.001624696422368288, + "learning_rate": 0.2930216778171417, + "loss": 0.3374, + "num_input_tokens_seen": 8161248, + "step": 3900 + }, + { + "epoch": 0.6370829594583571, + "grad_norm": 0.0005093521904200315, + "learning_rate": 0.2930039091488263, + "loss": 0.3539, + "num_input_tokens_seen": 8171248, + "step": 3905 + }, + { + "epoch": 0.6378986866791745, + "grad_norm": 0.001131294178776443, + "learning_rate": 0.29298611842751093, + "loss": 0.3328, + "num_input_tokens_seen": 8182976, + "step": 3910 + }, + { + "epoch": 0.6387144138999918, + "grad_norm": 0.0014804151142016053, + "learning_rate": 0.29296830565593923, + "loss": 0.3425, + "num_input_tokens_seen": 8192832, + "step": 3915 + }, + { + "epoch": 0.6395301411208092, + "grad_norm": 0.0014141841093078256, + "learning_rate": 0.2929504708368582, + "loss": 0.3424, + "num_input_tokens_seen": 8202992, + "step": 3920 + }, + { + "epoch": 0.6403458683416265, + "grad_norm": 0.001320424722507596, + "learning_rate": 0.29293261397301806, + "loss": 0.3456, + "num_input_tokens_seen": 8214176, + "step": 3925 + }, + { + "epoch": 0.6411615955624439, + "grad_norm": 0.0023495852947235107, + "learning_rate": 0.29291473506717275, + "loss": 0.3451, + "num_input_tokens_seen": 8225216, + "step": 3930 + }, + { + "epoch": 0.6419773227832613, + "grad_norm": 0.000542253430467099, + "learning_rate": 0.29289683412207923, + "loss": 0.3513, + "num_input_tokens_seen": 8235680, + "step": 3935 + }, + { + "epoch": 0.6427930500040786, + "grad_norm": 0.0016157758655026555, + "learning_rate": 0.29287891114049813, + "loss": 0.3594, + "num_input_tokens_seen": 8246480, + "step": 3940 + }, + { + "epoch": 0.643608777224896, + "grad_norm": 0.000730989733710885, + "learning_rate": 0.29286096612519347, + "loss": 0.3466, + "num_input_tokens_seen": 8256928, + "step": 3945 + }, + { + "epoch": 0.6444245044457133, + "grad_norm": 0.001220886711962521, + "learning_rate": 0.2928429990789325, + "loss": 0.3468, + "num_input_tokens_seen": 8268096, + "step": 3950 + }, + { + "epoch": 0.6452402316665307, + "grad_norm": 0.0007379367598332465, + "learning_rate": 0.29282501000448596, + "loss": 0.3186, + "num_input_tokens_seen": 8279440, + "step": 3955 + }, + { + "epoch": 0.6460559588873481, + "grad_norm": 0.000755560991819948, + "learning_rate": 0.2928069989046281, + "loss": 0.3095, + "num_input_tokens_seen": 8290144, + "step": 3960 + }, + { + "epoch": 0.6468716861081655, + "grad_norm": 0.003492318792268634, + "learning_rate": 0.2927889657821363, + "loss": 0.3976, + "num_input_tokens_seen": 8300240, + "step": 3965 + }, + { + "epoch": 0.6476874133289828, + "grad_norm": 0.0011360944481566548, + "learning_rate": 0.2927709106397916, + "loss": 0.371, + "num_input_tokens_seen": 8309632, + "step": 3970 + }, + { + "epoch": 0.6485031405498002, + "grad_norm": 0.0018521938472986221, + "learning_rate": 0.29275283348037834, + "loss": 0.3417, + "num_input_tokens_seen": 8320752, + "step": 3975 + }, + { + "epoch": 0.6493188677706175, + "grad_norm": 0.0006954994169063866, + "learning_rate": 0.29273473430668423, + "loss": 0.3508, + "num_input_tokens_seen": 8331136, + "step": 3980 + }, + { + "epoch": 0.6501345949914349, + "grad_norm": 0.0013802520697936416, + "learning_rate": 0.2927166131215003, + "loss": 0.3312, + "num_input_tokens_seen": 8341552, + "step": 3985 + }, + { + "epoch": 0.6509503222122522, + "grad_norm": 0.0008327278774231672, + "learning_rate": 0.2926984699276212, + "loss": 0.3513, + "num_input_tokens_seen": 8352176, + "step": 3990 + }, + { + "epoch": 0.6517660494330696, + "grad_norm": 0.0009512333199381828, + "learning_rate": 0.29268030472784473, + "loss": 0.3662, + "num_input_tokens_seen": 8361952, + "step": 3995 + }, + { + "epoch": 0.652581776653887, + "grad_norm": 0.0011538391700014472, + "learning_rate": 0.2926621175249723, + "loss": 0.3448, + "num_input_tokens_seen": 8372672, + "step": 4000 + }, + { + "epoch": 0.652581776653887, + "eval_loss": 0.3364817202091217, + "eval_runtime": 155.9871, + "eval_samples_per_second": 17.469, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 8372672, + "step": 4000 + }, + { + "epoch": 0.6533975038747043, + "grad_norm": 0.0010531063890084624, + "learning_rate": 0.29264390832180853, + "loss": 0.3276, + "num_input_tokens_seen": 8382752, + "step": 4005 + }, + { + "epoch": 0.6542132310955217, + "grad_norm": 0.0007207767921499908, + "learning_rate": 0.29262567712116144, + "loss": 0.3515, + "num_input_tokens_seen": 8394112, + "step": 4010 + }, + { + "epoch": 0.655028958316339, + "grad_norm": 0.0022957841865718365, + "learning_rate": 0.29260742392584266, + "loss": 0.3555, + "num_input_tokens_seen": 8404576, + "step": 4015 + }, + { + "epoch": 0.6558446855371564, + "grad_norm": 0.0006454138783738017, + "learning_rate": 0.292589148738667, + "loss": 0.3505, + "num_input_tokens_seen": 8414592, + "step": 4020 + }, + { + "epoch": 0.6566604127579737, + "grad_norm": 0.001161114196293056, + "learning_rate": 0.2925708515624527, + "loss": 0.3454, + "num_input_tokens_seen": 8425168, + "step": 4025 + }, + { + "epoch": 0.6574761399787911, + "grad_norm": 0.0014844425022602081, + "learning_rate": 0.29255253240002144, + "loss": 0.3324, + "num_input_tokens_seen": 8436768, + "step": 4030 + }, + { + "epoch": 0.6582918671996084, + "grad_norm": 0.002551761455833912, + "learning_rate": 0.2925341912541983, + "loss": 0.3587, + "num_input_tokens_seen": 8448176, + "step": 4035 + }, + { + "epoch": 0.6591075944204258, + "grad_norm": 0.0007941091316752136, + "learning_rate": 0.2925158281278116, + "loss": 0.3455, + "num_input_tokens_seen": 8460480, + "step": 4040 + }, + { + "epoch": 0.6599233216412431, + "grad_norm": 0.0016014787834137678, + "learning_rate": 0.29249744302369324, + "loss": 0.3435, + "num_input_tokens_seen": 8470400, + "step": 4045 + }, + { + "epoch": 0.6607390488620605, + "grad_norm": 0.0021273696329444647, + "learning_rate": 0.29247903594467844, + "loss": 0.3236, + "num_input_tokens_seen": 8480352, + "step": 4050 + }, + { + "epoch": 0.6615547760828779, + "grad_norm": 0.0006997511954978108, + "learning_rate": 0.2924606068936058, + "loss": 0.3548, + "num_input_tokens_seen": 8491344, + "step": 4055 + }, + { + "epoch": 0.6623705033036953, + "grad_norm": 0.0011304132640361786, + "learning_rate": 0.2924421558733173, + "loss": 0.3532, + "num_input_tokens_seen": 8502480, + "step": 4060 + }, + { + "epoch": 0.6631862305245126, + "grad_norm": 0.002333104144781828, + "learning_rate": 0.2924236828866583, + "loss": 0.3366, + "num_input_tokens_seen": 8512672, + "step": 4065 + }, + { + "epoch": 0.66400195774533, + "grad_norm": 0.00144060084130615, + "learning_rate": 0.29240518793647763, + "loss": 0.3432, + "num_input_tokens_seen": 8523856, + "step": 4070 + }, + { + "epoch": 0.6648176849661473, + "grad_norm": 0.0012930165976285934, + "learning_rate": 0.29238667102562743, + "loss": 0.3087, + "num_input_tokens_seen": 8536192, + "step": 4075 + }, + { + "epoch": 0.6656334121869647, + "grad_norm": 0.002154681133106351, + "learning_rate": 0.29236813215696317, + "loss": 0.3774, + "num_input_tokens_seen": 8545136, + "step": 4080 + }, + { + "epoch": 0.666449139407782, + "grad_norm": 0.0020003095269203186, + "learning_rate": 0.2923495713333439, + "loss": 0.3283, + "num_input_tokens_seen": 8555456, + "step": 4085 + }, + { + "epoch": 0.6672648666285994, + "grad_norm": 0.0010210952023044229, + "learning_rate": 0.29233098855763173, + "loss": 0.3844, + "num_input_tokens_seen": 8564304, + "step": 4090 + }, + { + "epoch": 0.6680805938494168, + "grad_norm": 0.0026901508681476116, + "learning_rate": 0.29231238383269254, + "loss": 0.3235, + "num_input_tokens_seen": 8575536, + "step": 4095 + }, + { + "epoch": 0.6688963210702341, + "grad_norm": 0.0013598148943856359, + "learning_rate": 0.2922937571613954, + "loss": 0.3275, + "num_input_tokens_seen": 8587056, + "step": 4100 + }, + { + "epoch": 0.6697120482910515, + "grad_norm": 0.0014149290509521961, + "learning_rate": 0.29227510854661265, + "loss": 0.3535, + "num_input_tokens_seen": 8597808, + "step": 4105 + }, + { + "epoch": 0.6705277755118688, + "grad_norm": 0.00163622060790658, + "learning_rate": 0.29225643799122025, + "loss": 0.3627, + "num_input_tokens_seen": 8608592, + "step": 4110 + }, + { + "epoch": 0.6713435027326862, + "grad_norm": 0.0022505319211632013, + "learning_rate": 0.2922377454980974, + "loss": 0.3165, + "num_input_tokens_seen": 8617824, + "step": 4115 + }, + { + "epoch": 0.6721592299535035, + "grad_norm": 0.0014938770327717066, + "learning_rate": 0.29221903107012676, + "loss": 0.3273, + "num_input_tokens_seen": 8628464, + "step": 4120 + }, + { + "epoch": 0.6729749571743209, + "grad_norm": 0.002423269907012582, + "learning_rate": 0.29220029471019426, + "loss": 0.3448, + "num_input_tokens_seen": 8639168, + "step": 4125 + }, + { + "epoch": 0.6737906843951382, + "grad_norm": 0.0020969994366168976, + "learning_rate": 0.2921815364211893, + "loss": 0.3492, + "num_input_tokens_seen": 8649680, + "step": 4130 + }, + { + "epoch": 0.6746064116159556, + "grad_norm": 0.002322229091078043, + "learning_rate": 0.29216275620600474, + "loss": 0.2697, + "num_input_tokens_seen": 8660400, + "step": 4135 + }, + { + "epoch": 0.6754221388367729, + "grad_norm": 0.0021362595725804567, + "learning_rate": 0.29214395406753657, + "loss": 0.2488, + "num_input_tokens_seen": 8670080, + "step": 4140 + }, + { + "epoch": 0.6762378660575903, + "grad_norm": 0.003578752279281616, + "learning_rate": 0.2921251300086844, + "loss": 0.3168, + "num_input_tokens_seen": 8679952, + "step": 4145 + }, + { + "epoch": 0.6770535932784078, + "grad_norm": 0.0023385859094560146, + "learning_rate": 0.2921062840323511, + "loss": 0.3775, + "num_input_tokens_seen": 8689920, + "step": 4150 + }, + { + "epoch": 0.6778693204992251, + "grad_norm": 0.0011437603970989585, + "learning_rate": 0.29208741614144307, + "loss": 0.3435, + "num_input_tokens_seen": 8700432, + "step": 4155 + }, + { + "epoch": 0.6786850477200425, + "grad_norm": 0.0014995950041338801, + "learning_rate": 0.2920685263388698, + "loss": 0.3135, + "num_input_tokens_seen": 8711104, + "step": 4160 + }, + { + "epoch": 0.6795007749408598, + "grad_norm": 0.0010980385122820735, + "learning_rate": 0.2920496146275445, + "loss": 0.2574, + "num_input_tokens_seen": 8721264, + "step": 4165 + }, + { + "epoch": 0.6803165021616772, + "grad_norm": 0.00512612285092473, + "learning_rate": 0.29203068101038343, + "loss": 0.455, + "num_input_tokens_seen": 8732112, + "step": 4170 + }, + { + "epoch": 0.6811322293824945, + "grad_norm": 0.0014178334968164563, + "learning_rate": 0.2920117254903065, + "loss": 0.3416, + "num_input_tokens_seen": 8742464, + "step": 4175 + }, + { + "epoch": 0.6819479566033119, + "grad_norm": 0.002347205299884081, + "learning_rate": 0.29199274807023695, + "loss": 0.3306, + "num_input_tokens_seen": 8754128, + "step": 4180 + }, + { + "epoch": 0.6827636838241292, + "grad_norm": 0.0013803837355226278, + "learning_rate": 0.29197374875310117, + "loss": 0.3158, + "num_input_tokens_seen": 8765280, + "step": 4185 + }, + { + "epoch": 0.6835794110449466, + "grad_norm": 0.004274320788681507, + "learning_rate": 0.2919547275418292, + "loss": 0.4152, + "num_input_tokens_seen": 8775712, + "step": 4190 + }, + { + "epoch": 0.6843951382657639, + "grad_norm": 0.004113332834094763, + "learning_rate": 0.29193568443935436, + "loss": 0.3474, + "num_input_tokens_seen": 8785376, + "step": 4195 + }, + { + "epoch": 0.6852108654865813, + "grad_norm": 0.001684392336755991, + "learning_rate": 0.2919166194486133, + "loss": 0.3478, + "num_input_tokens_seen": 8796048, + "step": 4200 + }, + { + "epoch": 0.6852108654865813, + "eval_loss": 0.3409584164619446, + "eval_runtime": 155.8414, + "eval_samples_per_second": 17.486, + "eval_steps_per_second": 8.746, + "num_input_tokens_seen": 8796048, + "step": 4200 + }, + { + "epoch": 0.6860265927073986, + "grad_norm": 0.0007555097690783441, + "learning_rate": 0.2918975325725461, + "loss": 0.3346, + "num_input_tokens_seen": 8807040, + "step": 4205 + }, + { + "epoch": 0.686842319928216, + "grad_norm": 0.0006923229084350169, + "learning_rate": 0.29187842381409607, + "loss": 0.3325, + "num_input_tokens_seen": 8817664, + "step": 4210 + }, + { + "epoch": 0.6876580471490333, + "grad_norm": 0.0017565051093697548, + "learning_rate": 0.29185929317621023, + "loss": 0.3523, + "num_input_tokens_seen": 8828880, + "step": 4215 + }, + { + "epoch": 0.6884737743698507, + "grad_norm": 0.0017085663275793195, + "learning_rate": 0.29184014066183867, + "loss": 0.3409, + "num_input_tokens_seen": 8839840, + "step": 4220 + }, + { + "epoch": 0.689289501590668, + "grad_norm": 0.0009496496641077101, + "learning_rate": 0.2918209662739349, + "loss": 0.3356, + "num_input_tokens_seen": 8850224, + "step": 4225 + }, + { + "epoch": 0.6901052288114854, + "grad_norm": 0.0010532630840316415, + "learning_rate": 0.29180177001545593, + "loss": 0.3477, + "num_input_tokens_seen": 8862592, + "step": 4230 + }, + { + "epoch": 0.6909209560323027, + "grad_norm": 0.0019435316789895296, + "learning_rate": 0.29178255188936203, + "loss": 0.333, + "num_input_tokens_seen": 8872656, + "step": 4235 + }, + { + "epoch": 0.6917366832531202, + "grad_norm": 0.0016798194264993072, + "learning_rate": 0.2917633118986169, + "loss": 0.3049, + "num_input_tokens_seen": 8882016, + "step": 4240 + }, + { + "epoch": 0.6925524104739376, + "grad_norm": 0.0027838638052344322, + "learning_rate": 0.2917440500461875, + "loss": 0.2757, + "num_input_tokens_seen": 8891712, + "step": 4245 + }, + { + "epoch": 0.6933681376947549, + "grad_norm": 0.003288568463176489, + "learning_rate": 0.29172476633504435, + "loss": 0.3093, + "num_input_tokens_seen": 8902000, + "step": 4250 + }, + { + "epoch": 0.6941838649155723, + "grad_norm": 0.0022371294908225536, + "learning_rate": 0.2917054607681612, + "loss": 0.3309, + "num_input_tokens_seen": 8911040, + "step": 4255 + }, + { + "epoch": 0.6949995921363896, + "grad_norm": 0.004496986512094736, + "learning_rate": 0.29168613334851523, + "loss": 0.3438, + "num_input_tokens_seen": 8920912, + "step": 4260 + }, + { + "epoch": 0.695815319357207, + "grad_norm": 0.0014287206577137113, + "learning_rate": 0.2916667840790869, + "loss": 0.3029, + "num_input_tokens_seen": 8931728, + "step": 4265 + }, + { + "epoch": 0.6966310465780243, + "grad_norm": 0.0015545543283224106, + "learning_rate": 0.2916474129628603, + "loss": 0.2957, + "num_input_tokens_seen": 8941616, + "step": 4270 + }, + { + "epoch": 0.6974467737988417, + "grad_norm": 0.0016169364098459482, + "learning_rate": 0.29162802000282245, + "loss": 0.2777, + "num_input_tokens_seen": 8951008, + "step": 4275 + }, + { + "epoch": 0.698262501019659, + "grad_norm": 0.003309482941403985, + "learning_rate": 0.2916086052019642, + "loss": 0.3001, + "num_input_tokens_seen": 8961856, + "step": 4280 + }, + { + "epoch": 0.6990782282404764, + "grad_norm": 0.002902928041294217, + "learning_rate": 0.2915891685632794, + "loss": 0.3183, + "num_input_tokens_seen": 8971792, + "step": 4285 + }, + { + "epoch": 0.6998939554612937, + "grad_norm": 0.004890352487564087, + "learning_rate": 0.29156971008976545, + "loss": 0.3711, + "num_input_tokens_seen": 8982000, + "step": 4290 + }, + { + "epoch": 0.7007096826821111, + "grad_norm": 0.002797371707856655, + "learning_rate": 0.2915502297844232, + "loss": 0.4048, + "num_input_tokens_seen": 8990832, + "step": 4295 + }, + { + "epoch": 0.7015254099029284, + "grad_norm": 0.0018097743159160018, + "learning_rate": 0.2915307276502566, + "loss": 0.299, + "num_input_tokens_seen": 8999904, + "step": 4300 + }, + { + "epoch": 0.7023411371237458, + "grad_norm": 0.0010221375850960612, + "learning_rate": 0.29151120369027334, + "loss": 0.4037, + "num_input_tokens_seen": 9009888, + "step": 4305 + }, + { + "epoch": 0.7031568643445631, + "grad_norm": 0.001642312970943749, + "learning_rate": 0.29149165790748405, + "loss": 0.3688, + "num_input_tokens_seen": 9020224, + "step": 4310 + }, + { + "epoch": 0.7039725915653805, + "grad_norm": 0.0012493684189394116, + "learning_rate": 0.291472090304903, + "loss": 0.3539, + "num_input_tokens_seen": 9030176, + "step": 4315 + }, + { + "epoch": 0.7047883187861979, + "grad_norm": 0.0006043706089258194, + "learning_rate": 0.2914525008855478, + "loss": 0.3359, + "num_input_tokens_seen": 9041312, + "step": 4320 + }, + { + "epoch": 0.7056040460070152, + "grad_norm": 0.001513864379376173, + "learning_rate": 0.2914328896524394, + "loss": 0.3183, + "num_input_tokens_seen": 9052368, + "step": 4325 + }, + { + "epoch": 0.7064197732278326, + "grad_norm": 0.0009209021809510887, + "learning_rate": 0.291413256608602, + "loss": 0.3459, + "num_input_tokens_seen": 9064096, + "step": 4330 + }, + { + "epoch": 0.70723550044865, + "grad_norm": 0.002728527644649148, + "learning_rate": 0.29139360175706336, + "loss": 0.3311, + "num_input_tokens_seen": 9074064, + "step": 4335 + }, + { + "epoch": 0.7080512276694674, + "grad_norm": 0.0015675330068916082, + "learning_rate": 0.2913739251008544, + "loss": 0.351, + "num_input_tokens_seen": 9085728, + "step": 4340 + }, + { + "epoch": 0.7088669548902847, + "grad_norm": 0.0013548055430874228, + "learning_rate": 0.29135422664300964, + "loss": 0.3422, + "num_input_tokens_seen": 9097440, + "step": 4345 + }, + { + "epoch": 0.7096826821111021, + "grad_norm": 0.0011119613191112876, + "learning_rate": 0.29133450638656677, + "loss": 0.3434, + "num_input_tokens_seen": 9108320, + "step": 4350 + }, + { + "epoch": 0.7104984093319194, + "grad_norm": 0.0017901345854625106, + "learning_rate": 0.2913147643345669, + "loss": 0.3588, + "num_input_tokens_seen": 9119232, + "step": 4355 + }, + { + "epoch": 0.7113141365527368, + "grad_norm": 0.0011172191007062793, + "learning_rate": 0.29129500049005447, + "loss": 0.3043, + "num_input_tokens_seen": 9128816, + "step": 4360 + }, + { + "epoch": 0.7121298637735541, + "grad_norm": 0.0012433392694219947, + "learning_rate": 0.2912752148560773, + "loss": 0.2841, + "num_input_tokens_seen": 9137840, + "step": 4365 + }, + { + "epoch": 0.7129455909943715, + "grad_norm": 0.0021135681308805943, + "learning_rate": 0.2912554074356866, + "loss": 0.366, + "num_input_tokens_seen": 9148672, + "step": 4370 + }, + { + "epoch": 0.7137613182151888, + "grad_norm": 0.002253643935546279, + "learning_rate": 0.2912355782319371, + "loss": 0.3721, + "num_input_tokens_seen": 9158720, + "step": 4375 + }, + { + "epoch": 0.7145770454360062, + "grad_norm": 0.001087624579668045, + "learning_rate": 0.2912157272478864, + "loss": 0.2902, + "num_input_tokens_seen": 9168240, + "step": 4380 + }, + { + "epoch": 0.7153927726568236, + "grad_norm": 0.0010568676516413689, + "learning_rate": 0.291195854486596, + "loss": 0.2904, + "num_input_tokens_seen": 9178704, + "step": 4385 + }, + { + "epoch": 0.7162084998776409, + "grad_norm": 0.0009495731210336089, + "learning_rate": 0.2911759599511305, + "loss": 0.3425, + "num_input_tokens_seen": 9189648, + "step": 4390 + }, + { + "epoch": 0.7170242270984583, + "grad_norm": 0.0015707132406532764, + "learning_rate": 0.29115604364455777, + "loss": 0.2944, + "num_input_tokens_seen": 9199696, + "step": 4395 + }, + { + "epoch": 0.7178399543192756, + "grad_norm": 0.0017782608047127724, + "learning_rate": 0.2911361055699493, + "loss": 0.3384, + "num_input_tokens_seen": 9210416, + "step": 4400 + }, + { + "epoch": 0.7178399543192756, + "eval_loss": 0.3222092390060425, + "eval_runtime": 156.1736, + "eval_samples_per_second": 17.449, + "eval_steps_per_second": 8.727, + "num_input_tokens_seen": 9210416, + "step": 4400 + }, + { + "epoch": 0.718655681540093, + "grad_norm": 0.0029527146834880114, + "learning_rate": 0.2911161457303797, + "loss": 0.3289, + "num_input_tokens_seen": 9220016, + "step": 4405 + }, + { + "epoch": 0.7194714087609103, + "grad_norm": 0.002007452305406332, + "learning_rate": 0.291096164128927, + "loss": 0.3331, + "num_input_tokens_seen": 9230944, + "step": 4410 + }, + { + "epoch": 0.7202871359817277, + "grad_norm": 0.005001790821552277, + "learning_rate": 0.2910761607686727, + "loss": 0.3737, + "num_input_tokens_seen": 9240528, + "step": 4415 + }, + { + "epoch": 0.721102863202545, + "grad_norm": 0.002044194145128131, + "learning_rate": 0.2910561356527016, + "loss": 0.3667, + "num_input_tokens_seen": 9250496, + "step": 4420 + }, + { + "epoch": 0.7219185904233625, + "grad_norm": 0.002278428990393877, + "learning_rate": 0.2910360887841017, + "loss": 0.3435, + "num_input_tokens_seen": 9260832, + "step": 4425 + }, + { + "epoch": 0.7227343176441798, + "grad_norm": 0.0016001742333173752, + "learning_rate": 0.2910160201659645, + "loss": 0.2821, + "num_input_tokens_seen": 9270352, + "step": 4430 + }, + { + "epoch": 0.7235500448649972, + "grad_norm": 0.0009348717285320163, + "learning_rate": 0.29099592980138494, + "loss": 0.323, + "num_input_tokens_seen": 9281120, + "step": 4435 + }, + { + "epoch": 0.7243657720858145, + "grad_norm": 0.0007745995535515249, + "learning_rate": 0.29097581769346115, + "loss": 0.3276, + "num_input_tokens_seen": 9292000, + "step": 4440 + }, + { + "epoch": 0.7251814993066319, + "grad_norm": 0.0010657825041562319, + "learning_rate": 0.29095568384529463, + "loss": 0.3479, + "num_input_tokens_seen": 9302480, + "step": 4445 + }, + { + "epoch": 0.7259972265274492, + "grad_norm": 0.00181719905231148, + "learning_rate": 0.2909355282599903, + "loss": 0.3369, + "num_input_tokens_seen": 9312464, + "step": 4450 + }, + { + "epoch": 0.7268129537482666, + "grad_norm": 0.0018904076423496008, + "learning_rate": 0.29091535094065635, + "loss": 0.3304, + "num_input_tokens_seen": 9322224, + "step": 4455 + }, + { + "epoch": 0.727628680969084, + "grad_norm": 0.001580826472491026, + "learning_rate": 0.2908951518904045, + "loss": 0.3293, + "num_input_tokens_seen": 9332176, + "step": 4460 + }, + { + "epoch": 0.7284444081899013, + "grad_norm": 0.0007734224200248718, + "learning_rate": 0.29087493111234963, + "loss": 0.3121, + "num_input_tokens_seen": 9344176, + "step": 4465 + }, + { + "epoch": 0.7292601354107187, + "grad_norm": 0.0024113974068313837, + "learning_rate": 0.29085468860961, + "loss": 0.3313, + "num_input_tokens_seen": 9355904, + "step": 4470 + }, + { + "epoch": 0.730075862631536, + "grad_norm": 0.0008996828109957278, + "learning_rate": 0.2908344243853073, + "loss": 0.3034, + "num_input_tokens_seen": 9366320, + "step": 4475 + }, + { + "epoch": 0.7308915898523534, + "grad_norm": 0.0021405008155852556, + "learning_rate": 0.2908141384425666, + "loss": 0.3551, + "num_input_tokens_seen": 9377504, + "step": 4480 + }, + { + "epoch": 0.7317073170731707, + "grad_norm": 0.001397660351358354, + "learning_rate": 0.2907938307845161, + "loss": 0.3228, + "num_input_tokens_seen": 9387936, + "step": 4485 + }, + { + "epoch": 0.7325230442939881, + "grad_norm": 0.0016000826144590974, + "learning_rate": 0.2907735014142876, + "loss": 0.3205, + "num_input_tokens_seen": 9398496, + "step": 4490 + }, + { + "epoch": 0.7333387715148054, + "grad_norm": 0.0021532459650188684, + "learning_rate": 0.2907531503350161, + "loss": 0.37, + "num_input_tokens_seen": 9408848, + "step": 4495 + }, + { + "epoch": 0.7341544987356228, + "grad_norm": 0.0017964029684662819, + "learning_rate": 0.29073277754983995, + "loss": 0.3094, + "num_input_tokens_seen": 9418224, + "step": 4500 + }, + { + "epoch": 0.7349702259564401, + "grad_norm": 0.0022990326397120953, + "learning_rate": 0.290712383061901, + "loss": 0.3498, + "num_input_tokens_seen": 9428464, + "step": 4505 + }, + { + "epoch": 0.7357859531772575, + "grad_norm": 0.0010057432809844613, + "learning_rate": 0.2906919668743443, + "loss": 0.3159, + "num_input_tokens_seen": 9437792, + "step": 4510 + }, + { + "epoch": 0.7366016803980748, + "grad_norm": 0.0011442821705713868, + "learning_rate": 0.29067152899031823, + "loss": 0.3482, + "num_input_tokens_seen": 9448144, + "step": 4515 + }, + { + "epoch": 0.7374174076188923, + "grad_norm": 0.0009812419302761555, + "learning_rate": 0.2906510694129746, + "loss": 0.3127, + "num_input_tokens_seen": 9457312, + "step": 4520 + }, + { + "epoch": 0.7382331348397096, + "grad_norm": 0.003093400038778782, + "learning_rate": 0.2906305881454685, + "loss": 0.3103, + "num_input_tokens_seen": 9469024, + "step": 4525 + }, + { + "epoch": 0.739048862060527, + "grad_norm": 0.0032218755222857, + "learning_rate": 0.2906100851909585, + "loss": 0.3405, + "num_input_tokens_seen": 9480000, + "step": 4530 + }, + { + "epoch": 0.7398645892813444, + "grad_norm": 0.0010040431516245008, + "learning_rate": 0.29058956055260626, + "loss": 0.2878, + "num_input_tokens_seen": 9491696, + "step": 4535 + }, + { + "epoch": 0.7406803165021617, + "grad_norm": 0.002233027946203947, + "learning_rate": 0.2905690142335771, + "loss": 0.3358, + "num_input_tokens_seen": 9502528, + "step": 4540 + }, + { + "epoch": 0.7414960437229791, + "grad_norm": 0.002588427159935236, + "learning_rate": 0.29054844623703946, + "loss": 0.3989, + "num_input_tokens_seen": 9512112, + "step": 4545 + }, + { + "epoch": 0.7423117709437964, + "grad_norm": 0.002240970963612199, + "learning_rate": 0.2905278565661651, + "loss": 0.3196, + "num_input_tokens_seen": 9523568, + "step": 4550 + }, + { + "epoch": 0.7431274981646138, + "grad_norm": 0.0013793676625937223, + "learning_rate": 0.2905072452241293, + "loss": 0.267, + "num_input_tokens_seen": 9533392, + "step": 4555 + }, + { + "epoch": 0.7439432253854311, + "grad_norm": 0.003387782257050276, + "learning_rate": 0.2904866122141106, + "loss": 0.3503, + "num_input_tokens_seen": 9544560, + "step": 4560 + }, + { + "epoch": 0.7447589526062485, + "grad_norm": 0.0032624464947730303, + "learning_rate": 0.2904659575392908, + "loss": 0.2686, + "num_input_tokens_seen": 9553984, + "step": 4565 + }, + { + "epoch": 0.7455746798270658, + "grad_norm": 0.0019009544048458338, + "learning_rate": 0.2904452812028551, + "loss": 0.2813, + "num_input_tokens_seen": 9563664, + "step": 4570 + }, + { + "epoch": 0.7463904070478832, + "grad_norm": 0.0010681245476007462, + "learning_rate": 0.2904245832079922, + "loss": 0.3016, + "num_input_tokens_seen": 9574256, + "step": 4575 + }, + { + "epoch": 0.7472061342687005, + "grad_norm": 0.002466883510351181, + "learning_rate": 0.29040386355789377, + "loss": 0.3308, + "num_input_tokens_seen": 9584800, + "step": 4580 + }, + { + "epoch": 0.7480218614895179, + "grad_norm": 0.0012558788293972611, + "learning_rate": 0.29038312225575524, + "loss": 0.373, + "num_input_tokens_seen": 9595120, + "step": 4585 + }, + { + "epoch": 0.7488375887103352, + "grad_norm": 0.0033187123481184244, + "learning_rate": 0.29036235930477505, + "loss": 0.3472, + "num_input_tokens_seen": 9605024, + "step": 4590 + }, + { + "epoch": 0.7496533159311526, + "grad_norm": 0.003067445708438754, + "learning_rate": 0.29034157470815514, + "loss": 0.2932, + "num_input_tokens_seen": 9617520, + "step": 4595 + }, + { + "epoch": 0.7504690431519699, + "grad_norm": 0.0013371068052947521, + "learning_rate": 0.2903207684691008, + "loss": 0.3026, + "num_input_tokens_seen": 9628832, + "step": 4600 + }, + { + "epoch": 0.7504690431519699, + "eval_loss": 0.3228461742401123, + "eval_runtime": 156.1072, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 9628832, + "step": 4600 + }, + { + "epoch": 0.7512847703727873, + "grad_norm": 0.0024242105428129435, + "learning_rate": 0.29029994059082054, + "loss": 0.3015, + "num_input_tokens_seen": 9639408, + "step": 4605 + }, + { + "epoch": 0.7521004975936046, + "grad_norm": 0.0014778063632547855, + "learning_rate": 0.2902790910765264, + "loss": 0.2952, + "num_input_tokens_seen": 9650752, + "step": 4610 + }, + { + "epoch": 0.7529162248144221, + "grad_norm": 0.0017629133071750402, + "learning_rate": 0.29025821992943346, + "loss": 0.2944, + "num_input_tokens_seen": 9660672, + "step": 4615 + }, + { + "epoch": 0.7537319520352395, + "grad_norm": 0.0017022376414388418, + "learning_rate": 0.29023732715276046, + "loss": 0.2697, + "num_input_tokens_seen": 9670256, + "step": 4620 + }, + { + "epoch": 0.7545476792560568, + "grad_norm": 0.0022384338080883026, + "learning_rate": 0.2902164127497293, + "loss": 0.4289, + "num_input_tokens_seen": 9680592, + "step": 4625 + }, + { + "epoch": 0.7553634064768742, + "grad_norm": 0.0028220440726727247, + "learning_rate": 0.2901954767235652, + "loss": 0.3425, + "num_input_tokens_seen": 9689904, + "step": 4630 + }, + { + "epoch": 0.7561791336976915, + "grad_norm": 0.0017985920421779156, + "learning_rate": 0.2901745190774968, + "loss": 0.3242, + "num_input_tokens_seen": 9699888, + "step": 4635 + }, + { + "epoch": 0.7569948609185089, + "grad_norm": 0.0024364041164517403, + "learning_rate": 0.290153539814756, + "loss": 0.3398, + "num_input_tokens_seen": 9710592, + "step": 4640 + }, + { + "epoch": 0.7578105881393262, + "grad_norm": 0.0008935498190112412, + "learning_rate": 0.2901325389385781, + "loss": 0.327, + "num_input_tokens_seen": 9721184, + "step": 4645 + }, + { + "epoch": 0.7586263153601436, + "grad_norm": 0.0037395760882645845, + "learning_rate": 0.2901115164522016, + "loss": 0.3548, + "num_input_tokens_seen": 9732384, + "step": 4650 + }, + { + "epoch": 0.7594420425809609, + "grad_norm": 0.0017213895916938782, + "learning_rate": 0.29009047235886865, + "loss": 0.3004, + "num_input_tokens_seen": 9742544, + "step": 4655 + }, + { + "epoch": 0.7602577698017783, + "grad_norm": 0.002058829413726926, + "learning_rate": 0.2900694066618243, + "loss": 0.3428, + "num_input_tokens_seen": 9752736, + "step": 4660 + }, + { + "epoch": 0.7610734970225956, + "grad_norm": 0.0011990817729383707, + "learning_rate": 0.2900483193643172, + "loss": 0.2938, + "num_input_tokens_seen": 9763744, + "step": 4665 + }, + { + "epoch": 0.761889224243413, + "grad_norm": 0.002634989097714424, + "learning_rate": 0.29002721046959934, + "loss": 0.2862, + "num_input_tokens_seen": 9775040, + "step": 4670 + }, + { + "epoch": 0.7627049514642303, + "grad_norm": 0.0014796295436099172, + "learning_rate": 0.29000607998092587, + "loss": 0.2992, + "num_input_tokens_seen": 9785328, + "step": 4675 + }, + { + "epoch": 0.7635206786850477, + "grad_norm": 0.002976446645334363, + "learning_rate": 0.2899849279015555, + "loss": 0.308, + "num_input_tokens_seen": 9795328, + "step": 4680 + }, + { + "epoch": 0.764336405905865, + "grad_norm": 0.0021798224188387394, + "learning_rate": 0.28996375423475007, + "loss": 0.3475, + "num_input_tokens_seen": 9805328, + "step": 4685 + }, + { + "epoch": 0.7651521331266824, + "grad_norm": 0.002107886364683509, + "learning_rate": 0.28994255898377486, + "loss": 0.3467, + "num_input_tokens_seen": 9814688, + "step": 4690 + }, + { + "epoch": 0.7659678603474998, + "grad_norm": 0.0006881779991090298, + "learning_rate": 0.2899213421518984, + "loss": 0.2545, + "num_input_tokens_seen": 9825664, + "step": 4695 + }, + { + "epoch": 0.7667835875683171, + "grad_norm": 0.0030328012071549892, + "learning_rate": 0.2899001037423926, + "loss": 0.2838, + "num_input_tokens_seen": 9837760, + "step": 4700 + }, + { + "epoch": 0.7675993147891346, + "grad_norm": 0.0013619129313156009, + "learning_rate": 0.28987884375853273, + "loss": 0.3208, + "num_input_tokens_seen": 9848512, + "step": 4705 + }, + { + "epoch": 0.7684150420099519, + "grad_norm": 0.0011368038831278682, + "learning_rate": 0.2898575622035974, + "loss": 0.3595, + "num_input_tokens_seen": 9858912, + "step": 4710 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.0016957747284322977, + "learning_rate": 0.2898362590808683, + "loss": 0.2448, + "num_input_tokens_seen": 9869024, + "step": 4715 + }, + { + "epoch": 0.7700464964515866, + "grad_norm": 0.0008419152582064271, + "learning_rate": 0.2898149343936308, + "loss": 0.2574, + "num_input_tokens_seen": 9878464, + "step": 4720 + }, + { + "epoch": 0.770862223672404, + "grad_norm": 0.0009039540309458971, + "learning_rate": 0.2897935881451734, + "loss": 0.3365, + "num_input_tokens_seen": 9888864, + "step": 4725 + }, + { + "epoch": 0.7716779508932213, + "grad_norm": 0.001688003190793097, + "learning_rate": 0.28977222033878797, + "loss": 0.2461, + "num_input_tokens_seen": 9899696, + "step": 4730 + }, + { + "epoch": 0.7724936781140387, + "grad_norm": 0.003010700223967433, + "learning_rate": 0.28975083097776966, + "loss": 0.2868, + "num_input_tokens_seen": 9911392, + "step": 4735 + }, + { + "epoch": 0.773309405334856, + "grad_norm": 0.006916660815477371, + "learning_rate": 0.28972942006541696, + "loss": 0.3218, + "num_input_tokens_seen": 9920304, + "step": 4740 + }, + { + "epoch": 0.7741251325556734, + "grad_norm": 0.0034634494222700596, + "learning_rate": 0.2897079876050318, + "loss": 0.2956, + "num_input_tokens_seen": 9930816, + "step": 4745 + }, + { + "epoch": 0.7749408597764907, + "grad_norm": 0.005117450375109911, + "learning_rate": 0.2896865335999192, + "loss": 0.2712, + "num_input_tokens_seen": 9940848, + "step": 4750 + }, + { + "epoch": 0.7757565869973081, + "grad_norm": 0.0012985988287255168, + "learning_rate": 0.28966505805338777, + "loss": 0.3736, + "num_input_tokens_seen": 9951920, + "step": 4755 + }, + { + "epoch": 0.7765723142181254, + "grad_norm": 0.0017857500351965427, + "learning_rate": 0.2896435609687492, + "loss": 0.3214, + "num_input_tokens_seen": 9962176, + "step": 4760 + }, + { + "epoch": 0.7773880414389428, + "grad_norm": 0.0014033698244020343, + "learning_rate": 0.2896220423493187, + "loss": 0.281, + "num_input_tokens_seen": 9973600, + "step": 4765 + }, + { + "epoch": 0.7782037686597602, + "grad_norm": 0.0008578443666920066, + "learning_rate": 0.28960050219841466, + "loss": 0.2729, + "num_input_tokens_seen": 9984528, + "step": 4770 + }, + { + "epoch": 0.7790194958805775, + "grad_norm": 0.0014563340228050947, + "learning_rate": 0.28957894051935884, + "loss": 0.3053, + "num_input_tokens_seen": 9993632, + "step": 4775 + }, + { + "epoch": 0.7798352231013949, + "grad_norm": 0.0020109142642468214, + "learning_rate": 0.2895573573154764, + "loss": 0.3477, + "num_input_tokens_seen": 10004448, + "step": 4780 + }, + { + "epoch": 0.7806509503222122, + "grad_norm": 0.0021665741223841906, + "learning_rate": 0.28953575259009556, + "loss": 0.2726, + "num_input_tokens_seen": 10015664, + "step": 4785 + }, + { + "epoch": 0.7814666775430296, + "grad_norm": 0.0020958222448825836, + "learning_rate": 0.2895141263465482, + "loss": 0.3162, + "num_input_tokens_seen": 10026576, + "step": 4790 + }, + { + "epoch": 0.7822824047638469, + "grad_norm": 0.0026109202299267054, + "learning_rate": 0.28949247858816934, + "loss": 0.3376, + "num_input_tokens_seen": 10037168, + "step": 4795 + }, + { + "epoch": 0.7830981319846644, + "grad_norm": 0.001513017574325204, + "learning_rate": 0.2894708093182973, + "loss": 0.3646, + "num_input_tokens_seen": 10048144, + "step": 4800 + }, + { + "epoch": 0.7830981319846644, + "eval_loss": 0.3192906677722931, + "eval_runtime": 156.1359, + "eval_samples_per_second": 17.453, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 10048144, + "step": 4800 + }, + { + "epoch": 0.7839138592054817, + "grad_norm": 0.0014610973885282874, + "learning_rate": 0.2894491185402737, + "loss": 0.3798, + "num_input_tokens_seen": 10057472, + "step": 4805 + }, + { + "epoch": 0.7847295864262991, + "grad_norm": 0.0011254171840846539, + "learning_rate": 0.2894274062574437, + "loss": 0.3936, + "num_input_tokens_seen": 10068720, + "step": 4810 + }, + { + "epoch": 0.7855453136471164, + "grad_norm": 0.0019502994837239385, + "learning_rate": 0.2894056724731554, + "loss": 0.3042, + "num_input_tokens_seen": 10078336, + "step": 4815 + }, + { + "epoch": 0.7863610408679338, + "grad_norm": 0.0007281806319952011, + "learning_rate": 0.28938391719076056, + "loss": 0.2963, + "num_input_tokens_seen": 10088064, + "step": 4820 + }, + { + "epoch": 0.7871767680887511, + "grad_norm": 0.0007417232263833284, + "learning_rate": 0.28936214041361413, + "loss": 0.3003, + "num_input_tokens_seen": 10098928, + "step": 4825 + }, + { + "epoch": 0.7879924953095685, + "grad_norm": 0.0032694493420422077, + "learning_rate": 0.2893403421450743, + "loss": 0.326, + "num_input_tokens_seen": 10108608, + "step": 4830 + }, + { + "epoch": 0.7888082225303858, + "grad_norm": 0.0005632973625324667, + "learning_rate": 0.2893185223885026, + "loss": 0.3093, + "num_input_tokens_seen": 10119168, + "step": 4835 + }, + { + "epoch": 0.7896239497512032, + "grad_norm": 0.002530734520405531, + "learning_rate": 0.289296681147264, + "loss": 0.3392, + "num_input_tokens_seen": 10129472, + "step": 4840 + }, + { + "epoch": 0.7904396769720206, + "grad_norm": 0.002390877576544881, + "learning_rate": 0.28927481842472663, + "loss": 0.2283, + "num_input_tokens_seen": 10138944, + "step": 4845 + }, + { + "epoch": 0.7912554041928379, + "grad_norm": 0.00384637713432312, + "learning_rate": 0.28925293422426207, + "loss": 0.3864, + "num_input_tokens_seen": 10150096, + "step": 4850 + }, + { + "epoch": 0.7920711314136553, + "grad_norm": 0.001627269433811307, + "learning_rate": 0.28923102854924504, + "loss": 0.2749, + "num_input_tokens_seen": 10158512, + "step": 4855 + }, + { + "epoch": 0.7928868586344726, + "grad_norm": 0.002952941693365574, + "learning_rate": 0.2892091014030537, + "loss": 0.3126, + "num_input_tokens_seen": 10167536, + "step": 4860 + }, + { + "epoch": 0.79370258585529, + "grad_norm": 0.0010198444360867143, + "learning_rate": 0.2891871527890696, + "loss": 0.3859, + "num_input_tokens_seen": 10178400, + "step": 4865 + }, + { + "epoch": 0.7945183130761073, + "grad_norm": 0.0013080495409667492, + "learning_rate": 0.2891651827106773, + "loss": 0.3209, + "num_input_tokens_seen": 10188240, + "step": 4870 + }, + { + "epoch": 0.7953340402969247, + "grad_norm": 0.0018730687443166971, + "learning_rate": 0.2891431911712651, + "loss": 0.2564, + "num_input_tokens_seen": 10198672, + "step": 4875 + }, + { + "epoch": 0.796149767517742, + "grad_norm": 0.0013093010056763887, + "learning_rate": 0.2891211781742241, + "loss": 0.2565, + "num_input_tokens_seen": 10208576, + "step": 4880 + }, + { + "epoch": 0.7969654947385594, + "grad_norm": 0.001371023477986455, + "learning_rate": 0.2890991437229492, + "loss": 0.4062, + "num_input_tokens_seen": 10219472, + "step": 4885 + }, + { + "epoch": 0.7977812219593767, + "grad_norm": 0.002174901310354471, + "learning_rate": 0.2890770878208383, + "loss": 0.3882, + "num_input_tokens_seen": 10230112, + "step": 4890 + }, + { + "epoch": 0.7985969491801942, + "grad_norm": 0.0013537859776988626, + "learning_rate": 0.28905501047129273, + "loss": 0.3244, + "num_input_tokens_seen": 10240992, + "step": 4895 + }, + { + "epoch": 0.7994126764010115, + "grad_norm": 0.0016178968362510204, + "learning_rate": 0.289032911677717, + "loss": 0.3336, + "num_input_tokens_seen": 10250336, + "step": 4900 + }, + { + "epoch": 0.8002284036218289, + "grad_norm": 0.001010962761938572, + "learning_rate": 0.28901079144351915, + "loss": 0.3397, + "num_input_tokens_seen": 10261600, + "step": 4905 + }, + { + "epoch": 0.8010441308426463, + "grad_norm": 0.0008778467890806496, + "learning_rate": 0.2889886497721103, + "loss": 0.3388, + "num_input_tokens_seen": 10271040, + "step": 4910 + }, + { + "epoch": 0.8018598580634636, + "grad_norm": 0.00146520568523556, + "learning_rate": 0.28896648666690505, + "loss": 0.366, + "num_input_tokens_seen": 10281568, + "step": 4915 + }, + { + "epoch": 0.802675585284281, + "grad_norm": 0.0007209221948869526, + "learning_rate": 0.2889443021313212, + "loss": 0.34, + "num_input_tokens_seen": 10289952, + "step": 4920 + }, + { + "epoch": 0.8034913125050983, + "grad_norm": 0.0018284699181094766, + "learning_rate": 0.28892209616877984, + "loss": 0.3239, + "num_input_tokens_seen": 10299536, + "step": 4925 + }, + { + "epoch": 0.8043070397259157, + "grad_norm": 0.0015225071692839265, + "learning_rate": 0.28889986878270546, + "loss": 0.3049, + "num_input_tokens_seen": 10309936, + "step": 4930 + }, + { + "epoch": 0.805122766946733, + "grad_norm": 0.0008819542126730084, + "learning_rate": 0.28887761997652583, + "loss": 0.3171, + "num_input_tokens_seen": 10319776, + "step": 4935 + }, + { + "epoch": 0.8059384941675504, + "grad_norm": 0.001785361673682928, + "learning_rate": 0.2888553497536719, + "loss": 0.3505, + "num_input_tokens_seen": 10330432, + "step": 4940 + }, + { + "epoch": 0.8067542213883677, + "grad_norm": 0.001377649838104844, + "learning_rate": 0.2888330581175781, + "loss": 0.2989, + "num_input_tokens_seen": 10341936, + "step": 4945 + }, + { + "epoch": 0.8075699486091851, + "grad_norm": 0.0009166896343231201, + "learning_rate": 0.28881074507168203, + "loss": 0.3616, + "num_input_tokens_seen": 10353488, + "step": 4950 + }, + { + "epoch": 0.8083856758300024, + "grad_norm": 0.0008294417639262974, + "learning_rate": 0.2887884106194247, + "loss": 0.2853, + "num_input_tokens_seen": 10364688, + "step": 4955 + }, + { + "epoch": 0.8092014030508198, + "grad_norm": 0.0024823874700814486, + "learning_rate": 0.28876605476425027, + "loss": 0.3349, + "num_input_tokens_seen": 10373104, + "step": 4960 + }, + { + "epoch": 0.8100171302716371, + "grad_norm": 0.001341267372481525, + "learning_rate": 0.2887436775096064, + "loss": 0.2775, + "num_input_tokens_seen": 10383776, + "step": 4965 + }, + { + "epoch": 0.8108328574924545, + "grad_norm": 0.0013850379036739469, + "learning_rate": 0.2887212788589439, + "loss": 0.3233, + "num_input_tokens_seen": 10395072, + "step": 4970 + }, + { + "epoch": 0.8116485847132718, + "grad_norm": 0.0014325660886242986, + "learning_rate": 0.2886988588157169, + "loss": 0.38, + "num_input_tokens_seen": 10407264, + "step": 4975 + }, + { + "epoch": 0.8124643119340892, + "grad_norm": 0.003903658129274845, + "learning_rate": 0.28867641738338284, + "loss": 0.3307, + "num_input_tokens_seen": 10417776, + "step": 4980 + }, + { + "epoch": 0.8132800391549067, + "grad_norm": 0.000594331999309361, + "learning_rate": 0.2886539545654026, + "loss": 0.2842, + "num_input_tokens_seen": 10428560, + "step": 4985 + }, + { + "epoch": 0.814095766375724, + "grad_norm": 0.0007164172711782157, + "learning_rate": 0.28863147036524006, + "loss": 0.3338, + "num_input_tokens_seen": 10439904, + "step": 4990 + }, + { + "epoch": 0.8149114935965414, + "grad_norm": 0.0012872953666374087, + "learning_rate": 0.2886089647863626, + "loss": 0.3218, + "num_input_tokens_seen": 10449136, + "step": 4995 + }, + { + "epoch": 0.8157272208173587, + "grad_norm": 0.0005984824383631349, + "learning_rate": 0.288586437832241, + "loss": 0.3149, + "num_input_tokens_seen": 10460144, + "step": 5000 + }, + { + "epoch": 0.8157272208173587, + "eval_loss": 0.32470157742500305, + "eval_runtime": 156.0585, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 10460144, + "step": 5000 + }, + { + "epoch": 0.8165429480381761, + "grad_norm": 0.0011932719498872757, + "learning_rate": 0.28856388950634904, + "loss": 0.2748, + "num_input_tokens_seen": 10470816, + "step": 5005 + }, + { + "epoch": 0.8173586752589934, + "grad_norm": 0.0020568605978041887, + "learning_rate": 0.288541319812164, + "loss": 0.3318, + "num_input_tokens_seen": 10482064, + "step": 5010 + }, + { + "epoch": 0.8181744024798108, + "grad_norm": 0.0011680201860144734, + "learning_rate": 0.2885187287531665, + "loss": 0.3196, + "num_input_tokens_seen": 10491408, + "step": 5015 + }, + { + "epoch": 0.8189901297006281, + "grad_norm": 0.006892974954098463, + "learning_rate": 0.2884961163328402, + "loss": 0.4044, + "num_input_tokens_seen": 10502896, + "step": 5020 + }, + { + "epoch": 0.8198058569214455, + "grad_norm": 0.0009961104951798916, + "learning_rate": 0.28847348255467237, + "loss": 0.3277, + "num_input_tokens_seen": 10513040, + "step": 5025 + }, + { + "epoch": 0.8206215841422628, + "grad_norm": 0.001897697220556438, + "learning_rate": 0.28845082742215333, + "loss": 0.3827, + "num_input_tokens_seen": 10523392, + "step": 5030 + }, + { + "epoch": 0.8214373113630802, + "grad_norm": 0.0007372269756160676, + "learning_rate": 0.2884281509387769, + "loss": 0.3408, + "num_input_tokens_seen": 10534592, + "step": 5035 + }, + { + "epoch": 0.8222530385838975, + "grad_norm": 0.0014414206380024552, + "learning_rate": 0.2884054531080399, + "loss": 0.3303, + "num_input_tokens_seen": 10544656, + "step": 5040 + }, + { + "epoch": 0.8230687658047149, + "grad_norm": 0.001617669709958136, + "learning_rate": 0.28838273393344277, + "loss": 0.3198, + "num_input_tokens_seen": 10554448, + "step": 5045 + }, + { + "epoch": 0.8238844930255322, + "grad_norm": 0.0017543075373396277, + "learning_rate": 0.288359993418489, + "loss": 0.3212, + "num_input_tokens_seen": 10564864, + "step": 5050 + }, + { + "epoch": 0.8247002202463496, + "grad_norm": 0.000666820676997304, + "learning_rate": 0.28833723156668556, + "loss": 0.3194, + "num_input_tokens_seen": 10575712, + "step": 5055 + }, + { + "epoch": 0.8255159474671669, + "grad_norm": 0.0005862181424163282, + "learning_rate": 0.2883144483815425, + "loss": 0.3461, + "num_input_tokens_seen": 10585792, + "step": 5060 + }, + { + "epoch": 0.8263316746879843, + "grad_norm": 0.001456849160604179, + "learning_rate": 0.28829164386657335, + "loss": 0.3331, + "num_input_tokens_seen": 10594688, + "step": 5065 + }, + { + "epoch": 0.8271474019088016, + "grad_norm": 0.0006418775301426649, + "learning_rate": 0.28826881802529486, + "loss": 0.3664, + "num_input_tokens_seen": 10603968, + "step": 5070 + }, + { + "epoch": 0.827963129129619, + "grad_norm": 0.0019763042218983173, + "learning_rate": 0.28824597086122705, + "loss": 0.3735, + "num_input_tokens_seen": 10614432, + "step": 5075 + }, + { + "epoch": 0.8287788563504365, + "grad_norm": 0.001020178315229714, + "learning_rate": 0.28822310237789317, + "loss": 0.3247, + "num_input_tokens_seen": 10624176, + "step": 5080 + }, + { + "epoch": 0.8295945835712538, + "grad_norm": 0.0005671300459653139, + "learning_rate": 0.2882002125788199, + "loss": 0.3703, + "num_input_tokens_seen": 10635168, + "step": 5085 + }, + { + "epoch": 0.8304103107920712, + "grad_norm": 0.001512977760285139, + "learning_rate": 0.2881773014675371, + "loss": 0.3669, + "num_input_tokens_seen": 10644336, + "step": 5090 + }, + { + "epoch": 0.8312260380128885, + "grad_norm": 0.0012904312461614609, + "learning_rate": 0.288154369047578, + "loss": 0.3469, + "num_input_tokens_seen": 10654656, + "step": 5095 + }, + { + "epoch": 0.8320417652337059, + "grad_norm": 0.001797419274225831, + "learning_rate": 0.28813141532247905, + "loss": 0.2888, + "num_input_tokens_seen": 10665008, + "step": 5100 + }, + { + "epoch": 0.8328574924545232, + "grad_norm": 0.0014407372800633311, + "learning_rate": 0.28810844029578, + "loss": 0.3485, + "num_input_tokens_seen": 10675616, + "step": 5105 + }, + { + "epoch": 0.8336732196753406, + "grad_norm": 0.0020855243783444166, + "learning_rate": 0.2880854439710238, + "loss": 0.3292, + "num_input_tokens_seen": 10686416, + "step": 5110 + }, + { + "epoch": 0.8344889468961579, + "grad_norm": 0.0013369523221626878, + "learning_rate": 0.28806242635175694, + "loss": 0.3207, + "num_input_tokens_seen": 10697056, + "step": 5115 + }, + { + "epoch": 0.8353046741169753, + "grad_norm": 0.001311844796873629, + "learning_rate": 0.2880393874415289, + "loss": 0.3149, + "num_input_tokens_seen": 10706544, + "step": 5120 + }, + { + "epoch": 0.8361204013377926, + "grad_norm": 0.0024499192368239164, + "learning_rate": 0.2880163272438926, + "loss": 0.3322, + "num_input_tokens_seen": 10716752, + "step": 5125 + }, + { + "epoch": 0.83693612855861, + "grad_norm": 0.0018589065875858068, + "learning_rate": 0.2879932457624042, + "loss": 0.3013, + "num_input_tokens_seen": 10727536, + "step": 5130 + }, + { + "epoch": 0.8377518557794273, + "grad_norm": 0.0012407248141244054, + "learning_rate": 0.2879701430006232, + "loss": 0.2597, + "num_input_tokens_seen": 10738000, + "step": 5135 + }, + { + "epoch": 0.8385675830002447, + "grad_norm": 0.0005454940837807953, + "learning_rate": 0.28794701896211233, + "loss": 0.3149, + "num_input_tokens_seen": 10748320, + "step": 5140 + }, + { + "epoch": 0.839383310221062, + "grad_norm": 0.0013937605544924736, + "learning_rate": 0.28792387365043753, + "loss": 0.3548, + "num_input_tokens_seen": 10757120, + "step": 5145 + }, + { + "epoch": 0.8401990374418794, + "grad_norm": 0.0021843560971319675, + "learning_rate": 0.28790070706916815, + "loss": 0.2856, + "num_input_tokens_seen": 10766192, + "step": 5150 + }, + { + "epoch": 0.8410147646626968, + "grad_norm": 0.001635857974179089, + "learning_rate": 0.2878775192218768, + "loss": 0.3334, + "num_input_tokens_seen": 10775088, + "step": 5155 + }, + { + "epoch": 0.8418304918835141, + "grad_norm": 0.005787878297269344, + "learning_rate": 0.2878543101121393, + "loss": 0.418, + "num_input_tokens_seen": 10786624, + "step": 5160 + }, + { + "epoch": 0.8426462191043315, + "grad_norm": 0.0014382967492565513, + "learning_rate": 0.28783107974353483, + "loss": 0.3328, + "num_input_tokens_seen": 10797376, + "step": 5165 + }, + { + "epoch": 0.8434619463251488, + "grad_norm": 0.0013609710149466991, + "learning_rate": 0.2878078281196457, + "loss": 0.3349, + "num_input_tokens_seen": 10806832, + "step": 5170 + }, + { + "epoch": 0.8442776735459663, + "grad_norm": 0.0011538314865902066, + "learning_rate": 0.28778455524405777, + "loss": 0.31, + "num_input_tokens_seen": 10817888, + "step": 5175 + }, + { + "epoch": 0.8450934007667836, + "grad_norm": 0.0007284412276931107, + "learning_rate": 0.2877612611203598, + "loss": 0.3123, + "num_input_tokens_seen": 10829120, + "step": 5180 + }, + { + "epoch": 0.845909127987601, + "grad_norm": 0.0011036060750484467, + "learning_rate": 0.28773794575214423, + "loss": 0.3549, + "num_input_tokens_seen": 10839248, + "step": 5185 + }, + { + "epoch": 0.8467248552084183, + "grad_norm": 0.0007364131161011755, + "learning_rate": 0.28771460914300645, + "loss": 0.325, + "num_input_tokens_seen": 10850624, + "step": 5190 + }, + { + "epoch": 0.8475405824292357, + "grad_norm": 0.0046900114975869656, + "learning_rate": 0.2876912512965454, + "loss": 0.3499, + "num_input_tokens_seen": 10861744, + "step": 5195 + }, + { + "epoch": 0.848356309650053, + "grad_norm": 0.0005929804756306112, + "learning_rate": 0.287667872216363, + "loss": 0.3029, + "num_input_tokens_seen": 10871296, + "step": 5200 + }, + { + "epoch": 0.848356309650053, + "eval_loss": 0.31762900948524475, + "eval_runtime": 156.1425, + "eval_samples_per_second": 17.452, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 10871296, + "step": 5200 + }, + { + "epoch": 0.8491720368708704, + "grad_norm": 0.001145743764936924, + "learning_rate": 0.2876444719060647, + "loss": 0.3098, + "num_input_tokens_seen": 10881776, + "step": 5205 + }, + { + "epoch": 0.8499877640916877, + "grad_norm": 0.001235578558407724, + "learning_rate": 0.287621050369259, + "loss": 0.2913, + "num_input_tokens_seen": 10890720, + "step": 5210 + }, + { + "epoch": 0.8508034913125051, + "grad_norm": 0.0009163934155367315, + "learning_rate": 0.28759760760955794, + "loss": 0.2811, + "num_input_tokens_seen": 10900528, + "step": 5215 + }, + { + "epoch": 0.8516192185333225, + "grad_norm": 0.002471217419952154, + "learning_rate": 0.2875741436305766, + "loss": 0.3667, + "num_input_tokens_seen": 10912048, + "step": 5220 + }, + { + "epoch": 0.8524349457541398, + "grad_norm": 0.0008888957090675831, + "learning_rate": 0.28755065843593347, + "loss": 0.3214, + "num_input_tokens_seen": 10921584, + "step": 5225 + }, + { + "epoch": 0.8532506729749572, + "grad_norm": 0.0013366041239351034, + "learning_rate": 0.2875271520292502, + "loss": 0.3289, + "num_input_tokens_seen": 10932208, + "step": 5230 + }, + { + "epoch": 0.8540664001957745, + "grad_norm": 0.0010787544306367636, + "learning_rate": 0.28750362441415184, + "loss": 0.3161, + "num_input_tokens_seen": 10941984, + "step": 5235 + }, + { + "epoch": 0.8548821274165919, + "grad_norm": 0.0011696713045239449, + "learning_rate": 0.28748007559426664, + "loss": 0.2998, + "num_input_tokens_seen": 10951584, + "step": 5240 + }, + { + "epoch": 0.8556978546374092, + "grad_norm": 0.0009502312168478966, + "learning_rate": 0.2874565055732261, + "loss": 0.3049, + "num_input_tokens_seen": 10961280, + "step": 5245 + }, + { + "epoch": 0.8565135818582266, + "grad_norm": 0.0019330558134242892, + "learning_rate": 0.28743291435466495, + "loss": 0.343, + "num_input_tokens_seen": 10972720, + "step": 5250 + }, + { + "epoch": 0.8573293090790439, + "grad_norm": 0.0008539496921002865, + "learning_rate": 0.2874093019422214, + "loss": 0.3384, + "num_input_tokens_seen": 10981936, + "step": 5255 + }, + { + "epoch": 0.8581450362998613, + "grad_norm": 0.0013879843754693866, + "learning_rate": 0.28738566833953666, + "loss": 0.3453, + "num_input_tokens_seen": 10992080, + "step": 5260 + }, + { + "epoch": 0.8589607635206787, + "grad_norm": 0.0010850467951968312, + "learning_rate": 0.28736201355025537, + "loss": 0.2647, + "num_input_tokens_seen": 11002832, + "step": 5265 + }, + { + "epoch": 0.8597764907414961, + "grad_norm": 0.0013848968083038926, + "learning_rate": 0.28733833757802535, + "loss": 0.3719, + "num_input_tokens_seen": 11013344, + "step": 5270 + }, + { + "epoch": 0.8605922179623134, + "grad_norm": 0.0008867940050549805, + "learning_rate": 0.28731464042649785, + "loss": 0.3256, + "num_input_tokens_seen": 11022304, + "step": 5275 + }, + { + "epoch": 0.8614079451831308, + "grad_norm": 0.0016384918708354235, + "learning_rate": 0.2872909220993271, + "loss": 0.3718, + "num_input_tokens_seen": 11033424, + "step": 5280 + }, + { + "epoch": 0.8622236724039481, + "grad_norm": 0.001569784595631063, + "learning_rate": 0.287267182600171, + "loss": 0.3453, + "num_input_tokens_seen": 11044112, + "step": 5285 + }, + { + "epoch": 0.8630393996247655, + "grad_norm": 0.000696200761012733, + "learning_rate": 0.2872434219326902, + "loss": 0.282, + "num_input_tokens_seen": 11055280, + "step": 5290 + }, + { + "epoch": 0.8638551268455829, + "grad_norm": 0.0007566320709884167, + "learning_rate": 0.28721964010054907, + "loss": 0.3181, + "num_input_tokens_seen": 11066416, + "step": 5295 + }, + { + "epoch": 0.8646708540664002, + "grad_norm": 0.0007043866207823157, + "learning_rate": 0.28719583710741503, + "loss": 0.3214, + "num_input_tokens_seen": 11075216, + "step": 5300 + }, + { + "epoch": 0.8654865812872176, + "grad_norm": 0.001919167465530336, + "learning_rate": 0.28717201295695877, + "loss": 0.3732, + "num_input_tokens_seen": 11085264, + "step": 5305 + }, + { + "epoch": 0.8663023085080349, + "grad_norm": 0.0005369902355596423, + "learning_rate": 0.28714816765285434, + "loss": 0.3213, + "num_input_tokens_seen": 11095536, + "step": 5310 + }, + { + "epoch": 0.8671180357288523, + "grad_norm": 0.0011691368417814374, + "learning_rate": 0.28712430119877896, + "loss": 0.3515, + "num_input_tokens_seen": 11107360, + "step": 5315 + }, + { + "epoch": 0.8679337629496696, + "grad_norm": 0.0012880684807896614, + "learning_rate": 0.28710041359841304, + "loss": 0.3312, + "num_input_tokens_seen": 11118336, + "step": 5320 + }, + { + "epoch": 0.868749490170487, + "grad_norm": 0.0006569003453478217, + "learning_rate": 0.28707650485544056, + "loss": 0.3265, + "num_input_tokens_seen": 11130064, + "step": 5325 + }, + { + "epoch": 0.8695652173913043, + "grad_norm": 0.0019254529615864158, + "learning_rate": 0.28705257497354836, + "loss": 0.3062, + "num_input_tokens_seen": 11139520, + "step": 5330 + }, + { + "epoch": 0.8703809446121217, + "grad_norm": 0.0010469005210325122, + "learning_rate": 0.28702862395642675, + "loss": 0.3261, + "num_input_tokens_seen": 11150224, + "step": 5335 + }, + { + "epoch": 0.871196671832939, + "grad_norm": 0.0006879929569549859, + "learning_rate": 0.28700465180776935, + "loss": 0.3626, + "num_input_tokens_seen": 11160848, + "step": 5340 + }, + { + "epoch": 0.8720123990537564, + "grad_norm": 0.0013986540725454688, + "learning_rate": 0.2869806585312729, + "loss": 0.3076, + "num_input_tokens_seen": 11171824, + "step": 5345 + }, + { + "epoch": 0.8728281262745737, + "grad_norm": 0.0026324521750211716, + "learning_rate": 0.28695664413063754, + "loss": 0.311, + "num_input_tokens_seen": 11182448, + "step": 5350 + }, + { + "epoch": 0.8736438534953911, + "grad_norm": 0.0013818513834849, + "learning_rate": 0.28693260860956654, + "loss": 0.3583, + "num_input_tokens_seen": 11192736, + "step": 5355 + }, + { + "epoch": 0.8744595807162086, + "grad_norm": 0.0011328537948429585, + "learning_rate": 0.2869085519717665, + "loss": 0.2669, + "num_input_tokens_seen": 11204752, + "step": 5360 + }, + { + "epoch": 0.8752753079370259, + "grad_norm": 0.0006440019351430237, + "learning_rate": 0.28688447422094726, + "loss": 0.3163, + "num_input_tokens_seen": 11216160, + "step": 5365 + }, + { + "epoch": 0.8760910351578433, + "grad_norm": 0.0014829334104433656, + "learning_rate": 0.2868603753608219, + "loss": 0.2733, + "num_input_tokens_seen": 11227664, + "step": 5370 + }, + { + "epoch": 0.8769067623786606, + "grad_norm": 0.0049444884061813354, + "learning_rate": 0.28683625539510665, + "loss": 0.3489, + "num_input_tokens_seen": 11238160, + "step": 5375 + }, + { + "epoch": 0.877722489599478, + "grad_norm": 0.0016436531441286206, + "learning_rate": 0.28681211432752135, + "loss": 0.3503, + "num_input_tokens_seen": 11248240, + "step": 5380 + }, + { + "epoch": 0.8785382168202953, + "grad_norm": 0.002831164514645934, + "learning_rate": 0.2867879521617887, + "loss": 0.3645, + "num_input_tokens_seen": 11257760, + "step": 5385 + }, + { + "epoch": 0.8793539440411127, + "grad_norm": 0.0010441725607961416, + "learning_rate": 0.28676376890163485, + "loss": 0.3014, + "num_input_tokens_seen": 11268608, + "step": 5390 + }, + { + "epoch": 0.88016967126193, + "grad_norm": 0.001201434526592493, + "learning_rate": 0.2867395645507891, + "loss": 0.2615, + "num_input_tokens_seen": 11278176, + "step": 5395 + }, + { + "epoch": 0.8809853984827474, + "grad_norm": 0.0014655718114227057, + "learning_rate": 0.2867153391129842, + "loss": 0.2841, + "num_input_tokens_seen": 11287600, + "step": 5400 + }, + { + "epoch": 0.8809853984827474, + "eval_loss": 0.3502575159072876, + "eval_runtime": 156.0518, + "eval_samples_per_second": 17.462, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 11287600, + "step": 5400 + }, + { + "epoch": 0.8818011257035647, + "grad_norm": 0.0018372762715443969, + "learning_rate": 0.28669109259195585, + "loss": 0.4421, + "num_input_tokens_seen": 11298624, + "step": 5405 + }, + { + "epoch": 0.8826168529243821, + "grad_norm": 0.0008314222795888782, + "learning_rate": 0.2866668249914433, + "loss": 0.3348, + "num_input_tokens_seen": 11307440, + "step": 5410 + }, + { + "epoch": 0.8834325801451994, + "grad_norm": 0.0007386492216028273, + "learning_rate": 0.2866425363151889, + "loss": 0.2795, + "num_input_tokens_seen": 11318288, + "step": 5415 + }, + { + "epoch": 0.8842483073660168, + "grad_norm": 0.0014771047281101346, + "learning_rate": 0.2866182265669382, + "loss": 0.3587, + "num_input_tokens_seen": 11328048, + "step": 5420 + }, + { + "epoch": 0.8850640345868341, + "grad_norm": 0.0017925486899912357, + "learning_rate": 0.28659389575044014, + "loss": 0.3276, + "num_input_tokens_seen": 11338048, + "step": 5425 + }, + { + "epoch": 0.8858797618076515, + "grad_norm": 0.0007746818591840565, + "learning_rate": 0.28656954386944683, + "loss": 0.3136, + "num_input_tokens_seen": 11349600, + "step": 5430 + }, + { + "epoch": 0.8866954890284688, + "grad_norm": 0.0006518039735965431, + "learning_rate": 0.28654517092771353, + "loss": 0.3048, + "num_input_tokens_seen": 11360000, + "step": 5435 + }, + { + "epoch": 0.8875112162492862, + "grad_norm": 0.0012950107920914888, + "learning_rate": 0.286520776928999, + "loss": 0.3422, + "num_input_tokens_seen": 11370624, + "step": 5440 + }, + { + "epoch": 0.8883269434701035, + "grad_norm": 0.001369382138364017, + "learning_rate": 0.286496361877065, + "loss": 0.3169, + "num_input_tokens_seen": 11381104, + "step": 5445 + }, + { + "epoch": 0.889142670690921, + "grad_norm": 0.0009714423795230687, + "learning_rate": 0.28647192577567676, + "loss": 0.3148, + "num_input_tokens_seen": 11392144, + "step": 5450 + }, + { + "epoch": 0.8899583979117384, + "grad_norm": 0.0014955688966438174, + "learning_rate": 0.28644746862860254, + "loss": 0.3356, + "num_input_tokens_seen": 11401760, + "step": 5455 + }, + { + "epoch": 0.8907741251325557, + "grad_norm": 0.0008027845760807395, + "learning_rate": 0.2864229904396139, + "loss": 0.3077, + "num_input_tokens_seen": 11412352, + "step": 5460 + }, + { + "epoch": 0.8915898523533731, + "grad_norm": 0.0011640962911769748, + "learning_rate": 0.28639849121248573, + "loss": 0.326, + "num_input_tokens_seen": 11421424, + "step": 5465 + }, + { + "epoch": 0.8924055795741904, + "grad_norm": 0.0009041184675879776, + "learning_rate": 0.28637397095099615, + "loss": 0.2842, + "num_input_tokens_seen": 11431680, + "step": 5470 + }, + { + "epoch": 0.8932213067950078, + "grad_norm": 0.0007662560092285275, + "learning_rate": 0.28634942965892646, + "loss": 0.3233, + "num_input_tokens_seen": 11442672, + "step": 5475 + }, + { + "epoch": 0.8940370340158251, + "grad_norm": 0.0013222425477579236, + "learning_rate": 0.28632486734006124, + "loss": 0.3478, + "num_input_tokens_seen": 11453712, + "step": 5480 + }, + { + "epoch": 0.8948527612366425, + "grad_norm": 0.0013737301342189312, + "learning_rate": 0.28630028399818835, + "loss": 0.2962, + "num_input_tokens_seen": 11464128, + "step": 5485 + }, + { + "epoch": 0.8956684884574598, + "grad_norm": 0.0016861003823578358, + "learning_rate": 0.2862756796370987, + "loss": 0.3318, + "num_input_tokens_seen": 11474208, + "step": 5490 + }, + { + "epoch": 0.8964842156782772, + "grad_norm": 0.0011057081865146756, + "learning_rate": 0.2862510542605868, + "loss": 0.2839, + "num_input_tokens_seen": 11485968, + "step": 5495 + }, + { + "epoch": 0.8972999428990945, + "grad_norm": 0.0016647791489958763, + "learning_rate": 0.2862264078724501, + "loss": 0.2857, + "num_input_tokens_seen": 11497104, + "step": 5500 + }, + { + "epoch": 0.8981156701199119, + "grad_norm": 0.0012007035547867417, + "learning_rate": 0.28620174047648933, + "loss": 0.3372, + "num_input_tokens_seen": 11507440, + "step": 5505 + }, + { + "epoch": 0.8989313973407292, + "grad_norm": 0.001037610461935401, + "learning_rate": 0.2861770520765086, + "loss": 0.2362, + "num_input_tokens_seen": 11518000, + "step": 5510 + }, + { + "epoch": 0.8997471245615466, + "grad_norm": 0.0022717721294611692, + "learning_rate": 0.2861523426763151, + "loss": 0.3536, + "num_input_tokens_seen": 11528592, + "step": 5515 + }, + { + "epoch": 0.900562851782364, + "grad_norm": 0.0006367280147969723, + "learning_rate": 0.2861276122797194, + "loss": 0.3193, + "num_input_tokens_seen": 11538432, + "step": 5520 + }, + { + "epoch": 0.9013785790031813, + "grad_norm": 0.0008934168145060539, + "learning_rate": 0.28610286089053516, + "loss": 0.3854, + "num_input_tokens_seen": 11548352, + "step": 5525 + }, + { + "epoch": 0.9021943062239987, + "grad_norm": 0.0017619330901652575, + "learning_rate": 0.28607808851257943, + "loss": 0.3215, + "num_input_tokens_seen": 11558896, + "step": 5530 + }, + { + "epoch": 0.903010033444816, + "grad_norm": 0.0005712034180760384, + "learning_rate": 0.28605329514967237, + "loss": 0.3587, + "num_input_tokens_seen": 11568304, + "step": 5535 + }, + { + "epoch": 0.9038257606656334, + "grad_norm": 0.0014995408710092306, + "learning_rate": 0.2860284808056374, + "loss": 0.2916, + "num_input_tokens_seen": 11579392, + "step": 5540 + }, + { + "epoch": 0.9046414878864508, + "grad_norm": 0.002275485312566161, + "learning_rate": 0.28600364548430135, + "loss": 0.3063, + "num_input_tokens_seen": 11590048, + "step": 5545 + }, + { + "epoch": 0.9054572151072682, + "grad_norm": 0.0009052076493389904, + "learning_rate": 0.28597878918949393, + "loss": 0.4083, + "num_input_tokens_seen": 11601216, + "step": 5550 + }, + { + "epoch": 0.9062729423280855, + "grad_norm": 0.0009083361364901066, + "learning_rate": 0.2859539119250485, + "loss": 0.2958, + "num_input_tokens_seen": 11611808, + "step": 5555 + }, + { + "epoch": 0.9070886695489029, + "grad_norm": 0.00129537598695606, + "learning_rate": 0.2859290136948013, + "loss": 0.4107, + "num_input_tokens_seen": 11623408, + "step": 5560 + }, + { + "epoch": 0.9079043967697202, + "grad_norm": 0.0011404891265556216, + "learning_rate": 0.28590409450259197, + "loss": 0.3294, + "num_input_tokens_seen": 11632416, + "step": 5565 + }, + { + "epoch": 0.9087201239905376, + "grad_norm": 0.000812943559139967, + "learning_rate": 0.28587915435226346, + "loss": 0.3427, + "num_input_tokens_seen": 11643248, + "step": 5570 + }, + { + "epoch": 0.9095358512113549, + "grad_norm": 0.0007024811347946525, + "learning_rate": 0.2858541932476617, + "loss": 0.3245, + "num_input_tokens_seen": 11654832, + "step": 5575 + }, + { + "epoch": 0.9103515784321723, + "grad_norm": 0.0005740235792472959, + "learning_rate": 0.2858292111926361, + "loss": 0.3335, + "num_input_tokens_seen": 11665696, + "step": 5580 + }, + { + "epoch": 0.9111673056529896, + "grad_norm": 0.0013332575326785445, + "learning_rate": 0.28580420819103924, + "loss": 0.3154, + "num_input_tokens_seen": 11675472, + "step": 5585 + }, + { + "epoch": 0.911983032873807, + "grad_norm": 0.0010334005346521735, + "learning_rate": 0.2857791842467269, + "loss": 0.3327, + "num_input_tokens_seen": 11685872, + "step": 5590 + }, + { + "epoch": 0.9127987600946244, + "grad_norm": 0.0006338933017104864, + "learning_rate": 0.2857541393635579, + "loss": 0.3765, + "num_input_tokens_seen": 11696304, + "step": 5595 + }, + { + "epoch": 0.9136144873154417, + "grad_norm": 0.0005694155697710812, + "learning_rate": 0.2857290735453948, + "loss": 0.3657, + "num_input_tokens_seen": 11707328, + "step": 5600 + }, + { + "epoch": 0.9136144873154417, + "eval_loss": 0.32543787360191345, + "eval_runtime": 155.9407, + "eval_samples_per_second": 17.475, + "eval_steps_per_second": 8.741, + "num_input_tokens_seen": 11707328, + "step": 5600 + }, + { + "epoch": 0.9144302145362591, + "grad_norm": 0.0004524579562712461, + "learning_rate": 0.28570398679610276, + "loss": 0.3271, + "num_input_tokens_seen": 11717936, + "step": 5605 + }, + { + "epoch": 0.9152459417570764, + "grad_norm": 0.0007683888543397188, + "learning_rate": 0.2856788791195506, + "loss": 0.3061, + "num_input_tokens_seen": 11728688, + "step": 5610 + }, + { + "epoch": 0.9160616689778938, + "grad_norm": 0.0018091164529323578, + "learning_rate": 0.28565375051961023, + "loss": 0.3246, + "num_input_tokens_seen": 11738016, + "step": 5615 + }, + { + "epoch": 0.9168773961987111, + "grad_norm": 0.0013068396365270019, + "learning_rate": 0.28562860100015686, + "loss": 0.3257, + "num_input_tokens_seen": 11747520, + "step": 5620 + }, + { + "epoch": 0.9176931234195285, + "grad_norm": 0.0007536066696047783, + "learning_rate": 0.2856034305650687, + "loss": 0.2887, + "num_input_tokens_seen": 11757344, + "step": 5625 + }, + { + "epoch": 0.9185088506403458, + "grad_norm": 0.002164301695302129, + "learning_rate": 0.28557823921822756, + "loss": 0.3245, + "num_input_tokens_seen": 11768352, + "step": 5630 + }, + { + "epoch": 0.9193245778611632, + "grad_norm": 0.001682449714280665, + "learning_rate": 0.2855530269635181, + "loss": 0.3103, + "num_input_tokens_seen": 11777824, + "step": 5635 + }, + { + "epoch": 0.9201403050819806, + "grad_norm": 0.0012137346202507615, + "learning_rate": 0.2855277938048284, + "loss": 0.2735, + "num_input_tokens_seen": 11788432, + "step": 5640 + }, + { + "epoch": 0.920956032302798, + "grad_norm": 0.0016050759004428983, + "learning_rate": 0.2855025397460498, + "loss": 0.364, + "num_input_tokens_seen": 11798352, + "step": 5645 + }, + { + "epoch": 0.9217717595236153, + "grad_norm": 0.0006402663420885801, + "learning_rate": 0.28547726479107666, + "loss": 0.2857, + "num_input_tokens_seen": 11807984, + "step": 5650 + }, + { + "epoch": 0.9225874867444327, + "grad_norm": 0.001102645881474018, + "learning_rate": 0.2854519689438068, + "loss": 0.3078, + "num_input_tokens_seen": 11818448, + "step": 5655 + }, + { + "epoch": 0.92340321396525, + "grad_norm": 0.001690243138000369, + "learning_rate": 0.2854266522081412, + "loss": 0.336, + "num_input_tokens_seen": 11828128, + "step": 5660 + }, + { + "epoch": 0.9242189411860674, + "grad_norm": 0.0014804549282416701, + "learning_rate": 0.28540131458798385, + "loss": 0.3512, + "num_input_tokens_seen": 11839344, + "step": 5665 + }, + { + "epoch": 0.9250346684068848, + "grad_norm": 0.0006690378650091588, + "learning_rate": 0.28537595608724226, + "loss": 0.3304, + "num_input_tokens_seen": 11849280, + "step": 5670 + }, + { + "epoch": 0.9258503956277021, + "grad_norm": 0.0018938019638881087, + "learning_rate": 0.28535057670982705, + "loss": 0.2627, + "num_input_tokens_seen": 11859296, + "step": 5675 + }, + { + "epoch": 0.9266661228485195, + "grad_norm": 0.0015777499647811055, + "learning_rate": 0.285325176459652, + "loss": 0.3838, + "num_input_tokens_seen": 11870304, + "step": 5680 + }, + { + "epoch": 0.9274818500693368, + "grad_norm": 0.0009405934251844883, + "learning_rate": 0.28529975534063406, + "loss": 0.3193, + "num_input_tokens_seen": 11880736, + "step": 5685 + }, + { + "epoch": 0.9282975772901542, + "grad_norm": 0.00306174554862082, + "learning_rate": 0.2852743133566936, + "loss": 0.4176, + "num_input_tokens_seen": 11891648, + "step": 5690 + }, + { + "epoch": 0.9291133045109715, + "grad_norm": 0.0014471039175987244, + "learning_rate": 0.2852488505117541, + "loss": 0.3176, + "num_input_tokens_seen": 11902480, + "step": 5695 + }, + { + "epoch": 0.9299290317317889, + "grad_norm": 0.001254696981050074, + "learning_rate": 0.28522336680974214, + "loss": 0.3562, + "num_input_tokens_seen": 11912592, + "step": 5700 + }, + { + "epoch": 0.9307447589526062, + "grad_norm": 0.0009726692805998027, + "learning_rate": 0.2851978622545877, + "loss": 0.3055, + "num_input_tokens_seen": 11923024, + "step": 5705 + }, + { + "epoch": 0.9315604861734236, + "grad_norm": 0.0009301457903347909, + "learning_rate": 0.285172336850224, + "loss": 0.2886, + "num_input_tokens_seen": 11933984, + "step": 5710 + }, + { + "epoch": 0.9323762133942409, + "grad_norm": 0.0005287854583002627, + "learning_rate": 0.2851467906005871, + "loss": 0.3024, + "num_input_tokens_seen": 11945744, + "step": 5715 + }, + { + "epoch": 0.9331919406150583, + "grad_norm": 0.0009527819347567856, + "learning_rate": 0.28512122350961683, + "loss": 0.3127, + "num_input_tokens_seen": 11955456, + "step": 5720 + }, + { + "epoch": 0.9340076678358756, + "grad_norm": 0.001318770693615079, + "learning_rate": 0.2850956355812559, + "loss": 0.3778, + "num_input_tokens_seen": 11965920, + "step": 5725 + }, + { + "epoch": 0.9348233950566931, + "grad_norm": 0.0006999018369242549, + "learning_rate": 0.28507002681945015, + "loss": 0.3549, + "num_input_tokens_seen": 11975232, + "step": 5730 + }, + { + "epoch": 0.9356391222775104, + "grad_norm": 0.0007602136465720832, + "learning_rate": 0.28504439722814895, + "loss": 0.3835, + "num_input_tokens_seen": 11985664, + "step": 5735 + }, + { + "epoch": 0.9364548494983278, + "grad_norm": 0.001212690374813974, + "learning_rate": 0.28501874681130457, + "loss": 0.2824, + "num_input_tokens_seen": 11996208, + "step": 5740 + }, + { + "epoch": 0.9372705767191452, + "grad_norm": 0.0006211533909663558, + "learning_rate": 0.2849930755728727, + "loss": 0.3186, + "num_input_tokens_seen": 12006992, + "step": 5745 + }, + { + "epoch": 0.9380863039399625, + "grad_norm": 0.002847872208803892, + "learning_rate": 0.28496738351681217, + "loss": 0.2958, + "num_input_tokens_seen": 12018064, + "step": 5750 + }, + { + "epoch": 0.9389020311607799, + "grad_norm": 0.0014025744749233127, + "learning_rate": 0.284941670647085, + "loss": 0.3144, + "num_input_tokens_seen": 12028128, + "step": 5755 + }, + { + "epoch": 0.9397177583815972, + "grad_norm": 0.0004352271498646587, + "learning_rate": 0.2849159369676563, + "loss": 0.2597, + "num_input_tokens_seen": 12040016, + "step": 5760 + }, + { + "epoch": 0.9405334856024146, + "grad_norm": 0.0012067031348124146, + "learning_rate": 0.2848901824824948, + "loss": 0.2873, + "num_input_tokens_seen": 12050608, + "step": 5765 + }, + { + "epoch": 0.9413492128232319, + "grad_norm": 0.0011298684403300285, + "learning_rate": 0.284864407195572, + "loss": 0.2937, + "num_input_tokens_seen": 12061296, + "step": 5770 + }, + { + "epoch": 0.9421649400440493, + "grad_norm": 0.000931062619201839, + "learning_rate": 0.28483861111086284, + "loss": 0.279, + "num_input_tokens_seen": 12072000, + "step": 5775 + }, + { + "epoch": 0.9429806672648666, + "grad_norm": 0.0006251083686947823, + "learning_rate": 0.2848127942323453, + "loss": 0.3197, + "num_input_tokens_seen": 12080992, + "step": 5780 + }, + { + "epoch": 0.943796394485684, + "grad_norm": 0.0007409377139993012, + "learning_rate": 0.2847869565640007, + "loss": 0.2882, + "num_input_tokens_seen": 12090272, + "step": 5785 + }, + { + "epoch": 0.9446121217065013, + "grad_norm": 0.0010686024324968457, + "learning_rate": 0.2847610981098136, + "loss": 0.3523, + "num_input_tokens_seen": 12100528, + "step": 5790 + }, + { + "epoch": 0.9454278489273187, + "grad_norm": 0.0005395633634179831, + "learning_rate": 0.2847352188737716, + "loss": 0.3195, + "num_input_tokens_seen": 12110352, + "step": 5795 + }, + { + "epoch": 0.946243576148136, + "grad_norm": 0.0010825146455317736, + "learning_rate": 0.2847093188598658, + "loss": 0.3327, + "num_input_tokens_seen": 12120624, + "step": 5800 + }, + { + "epoch": 0.946243576148136, + "eval_loss": 0.3238305151462555, + "eval_runtime": 156.0465, + "eval_samples_per_second": 17.463, + "eval_steps_per_second": 8.735, + "num_input_tokens_seen": 12120624, + "step": 5800 + }, + { + "epoch": 0.9470593033689534, + "grad_norm": 0.0012870727805420756, + "learning_rate": 0.28468339807209003, + "loss": 0.3397, + "num_input_tokens_seen": 12131440, + "step": 5805 + }, + { + "epoch": 0.9478750305897707, + "grad_norm": 0.0009699321235530078, + "learning_rate": 0.2846574565144418, + "loss": 0.305, + "num_input_tokens_seen": 12141856, + "step": 5810 + }, + { + "epoch": 0.9486907578105881, + "grad_norm": 0.0005382618983276188, + "learning_rate": 0.28463149419092154, + "loss": 0.3379, + "num_input_tokens_seen": 12152512, + "step": 5815 + }, + { + "epoch": 0.9495064850314054, + "grad_norm": 0.0010805901838466525, + "learning_rate": 0.284605511105533, + "loss": 0.2855, + "num_input_tokens_seen": 12163872, + "step": 5820 + }, + { + "epoch": 0.9503222122522229, + "grad_norm": 0.0009586413507349789, + "learning_rate": 0.28457950726228315, + "loss": 0.2605, + "num_input_tokens_seen": 12175536, + "step": 5825 + }, + { + "epoch": 0.9511379394730403, + "grad_norm": 0.0007782725151628256, + "learning_rate": 0.28455348266518193, + "loss": 0.3965, + "num_input_tokens_seen": 12186624, + "step": 5830 + }, + { + "epoch": 0.9519536666938576, + "grad_norm": 0.0008276398293673992, + "learning_rate": 0.28452743731824287, + "loss": 0.2854, + "num_input_tokens_seen": 12196976, + "step": 5835 + }, + { + "epoch": 0.952769393914675, + "grad_norm": 0.000515516847372055, + "learning_rate": 0.28450137122548236, + "loss": 0.3478, + "num_input_tokens_seen": 12206784, + "step": 5840 + }, + { + "epoch": 0.9535851211354923, + "grad_norm": 0.001388662843964994, + "learning_rate": 0.2844752843909201, + "loss": 0.3152, + "num_input_tokens_seen": 12217504, + "step": 5845 + }, + { + "epoch": 0.9544008483563097, + "grad_norm": 0.0004493473097681999, + "learning_rate": 0.28444917681857923, + "loss": 0.2557, + "num_input_tokens_seen": 12227360, + "step": 5850 + }, + { + "epoch": 0.955216575577127, + "grad_norm": 0.0004518165369518101, + "learning_rate": 0.28442304851248557, + "loss": 0.3301, + "num_input_tokens_seen": 12237344, + "step": 5855 + }, + { + "epoch": 0.9560323027979444, + "grad_norm": 0.0006291952449828386, + "learning_rate": 0.2843968994766686, + "loss": 0.3319, + "num_input_tokens_seen": 12248400, + "step": 5860 + }, + { + "epoch": 0.9568480300187617, + "grad_norm": 0.0015211799181997776, + "learning_rate": 0.28437072971516075, + "loss": 0.3148, + "num_input_tokens_seen": 12258224, + "step": 5865 + }, + { + "epoch": 0.9576637572395791, + "grad_norm": 0.001664430950768292, + "learning_rate": 0.2843445392319979, + "loss": 0.3104, + "num_input_tokens_seen": 12268736, + "step": 5870 + }, + { + "epoch": 0.9584794844603964, + "grad_norm": 0.0007471125572919846, + "learning_rate": 0.28431832803121865, + "loss": 0.3116, + "num_input_tokens_seen": 12278496, + "step": 5875 + }, + { + "epoch": 0.9592952116812138, + "grad_norm": 0.0010008630342781544, + "learning_rate": 0.28429209611686534, + "loss": 0.3223, + "num_input_tokens_seen": 12289088, + "step": 5880 + }, + { + "epoch": 0.9601109389020311, + "grad_norm": 0.000916619086638093, + "learning_rate": 0.28426584349298323, + "loss": 0.3164, + "num_input_tokens_seen": 12299728, + "step": 5885 + }, + { + "epoch": 0.9609266661228485, + "grad_norm": 0.0021812603808939457, + "learning_rate": 0.2842395701636207, + "loss": 0.3217, + "num_input_tokens_seen": 12309520, + "step": 5890 + }, + { + "epoch": 0.9617423933436658, + "grad_norm": 0.002276579150930047, + "learning_rate": 0.28421327613282954, + "loss": 0.3189, + "num_input_tokens_seen": 12319776, + "step": 5895 + }, + { + "epoch": 0.9625581205644832, + "grad_norm": 0.0006399665726348758, + "learning_rate": 0.28418696140466454, + "loss": 0.2944, + "num_input_tokens_seen": 12329168, + "step": 5900 + }, + { + "epoch": 0.9633738477853006, + "grad_norm": 0.0007726626354269683, + "learning_rate": 0.2841606259831838, + "loss": 0.3326, + "num_input_tokens_seen": 12339696, + "step": 5905 + }, + { + "epoch": 0.9641895750061179, + "grad_norm": 0.000858964049257338, + "learning_rate": 0.2841342698724486, + "loss": 0.3843, + "num_input_tokens_seen": 12349840, + "step": 5910 + }, + { + "epoch": 0.9650053022269353, + "grad_norm": 0.004363725893199444, + "learning_rate": 0.28410789307652334, + "loss": 0.3196, + "num_input_tokens_seen": 12361088, + "step": 5915 + }, + { + "epoch": 0.9658210294477527, + "grad_norm": 0.0006909170770086348, + "learning_rate": 0.2840814955994756, + "loss": 0.3038, + "num_input_tokens_seen": 12372080, + "step": 5920 + }, + { + "epoch": 0.9666367566685701, + "grad_norm": 0.0005849237786605954, + "learning_rate": 0.2840550774453763, + "loss": 0.3588, + "num_input_tokens_seen": 12382752, + "step": 5925 + }, + { + "epoch": 0.9674524838893874, + "grad_norm": 0.0008777655893936753, + "learning_rate": 0.28402863861829947, + "loss": 0.3427, + "num_input_tokens_seen": 12392704, + "step": 5930 + }, + { + "epoch": 0.9682682111102048, + "grad_norm": 0.0006631019059568644, + "learning_rate": 0.2840021791223222, + "loss": 0.3088, + "num_input_tokens_seen": 12404128, + "step": 5935 + }, + { + "epoch": 0.9690839383310221, + "grad_norm": 0.0007130609592422843, + "learning_rate": 0.2839756989615249, + "loss": 0.2753, + "num_input_tokens_seen": 12415936, + "step": 5940 + }, + { + "epoch": 0.9698996655518395, + "grad_norm": 0.000906288914848119, + "learning_rate": 0.28394919813999125, + "loss": 0.3154, + "num_input_tokens_seen": 12427104, + "step": 5945 + }, + { + "epoch": 0.9707153927726568, + "grad_norm": 0.0021283570677042007, + "learning_rate": 0.28392267666180787, + "loss": 0.3476, + "num_input_tokens_seen": 12437728, + "step": 5950 + }, + { + "epoch": 0.9715311199934742, + "grad_norm": 0.0014746139058843255, + "learning_rate": 0.2838961345310648, + "loss": 0.3627, + "num_input_tokens_seen": 12447968, + "step": 5955 + }, + { + "epoch": 0.9723468472142915, + "grad_norm": 0.001598145579919219, + "learning_rate": 0.2838695717518552, + "loss": 0.3168, + "num_input_tokens_seen": 12459072, + "step": 5960 + }, + { + "epoch": 0.9731625744351089, + "grad_norm": 0.0011322154896333814, + "learning_rate": 0.28384298832827526, + "loss": 0.3087, + "num_input_tokens_seen": 12470592, + "step": 5965 + }, + { + "epoch": 0.9739783016559262, + "grad_norm": 0.0016571765299886465, + "learning_rate": 0.28381638426442457, + "loss": 0.2903, + "num_input_tokens_seen": 12480864, + "step": 5970 + }, + { + "epoch": 0.9747940288767436, + "grad_norm": 0.00394893903285265, + "learning_rate": 0.2837897595644057, + "loss": 0.4229, + "num_input_tokens_seen": 12490672, + "step": 5975 + }, + { + "epoch": 0.975609756097561, + "grad_norm": 0.001482925727032125, + "learning_rate": 0.28376311423232475, + "loss": 0.3152, + "num_input_tokens_seen": 12501840, + "step": 5980 + }, + { + "epoch": 0.9764254833183783, + "grad_norm": 0.0012108649825677276, + "learning_rate": 0.2837364482722905, + "loss": 0.2897, + "num_input_tokens_seen": 12511216, + "step": 5985 + }, + { + "epoch": 0.9772412105391957, + "grad_norm": 0.0008006284479051828, + "learning_rate": 0.28370976168841533, + "loss": 0.3638, + "num_input_tokens_seen": 12521920, + "step": 5990 + }, + { + "epoch": 0.978056937760013, + "grad_norm": 0.0011215269332751632, + "learning_rate": 0.2836830544848146, + "loss": 0.3367, + "num_input_tokens_seen": 12533200, + "step": 5995 + }, + { + "epoch": 0.9788726649808304, + "grad_norm": 0.0020503622945398092, + "learning_rate": 0.2836563266656069, + "loss": 0.3498, + "num_input_tokens_seen": 12542416, + "step": 6000 + }, + { + "epoch": 0.9788726649808304, + "eval_loss": 0.32139912247657776, + "eval_runtime": 156.1589, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 12542416, + "step": 6000 + }, + { + "epoch": 0.9796883922016477, + "grad_norm": 0.0006524223135784268, + "learning_rate": 0.283629578234914, + "loss": 0.2504, + "num_input_tokens_seen": 12553152, + "step": 6005 + }, + { + "epoch": 0.9805041194224652, + "grad_norm": 0.0037967583630234003, + "learning_rate": 0.2836028091968608, + "loss": 0.3994, + "num_input_tokens_seen": 12562960, + "step": 6010 + }, + { + "epoch": 0.9813198466432825, + "grad_norm": 0.0009635387104935944, + "learning_rate": 0.28357601955557554, + "loss": 0.2913, + "num_input_tokens_seen": 12572144, + "step": 6015 + }, + { + "epoch": 0.9821355738640999, + "grad_norm": 0.0014124168083071709, + "learning_rate": 0.2835492093151894, + "loss": 0.3837, + "num_input_tokens_seen": 12583120, + "step": 6020 + }, + { + "epoch": 0.9829513010849172, + "grad_norm": 0.00071518070762977, + "learning_rate": 0.2835223784798369, + "loss": 0.3237, + "num_input_tokens_seen": 12594208, + "step": 6025 + }, + { + "epoch": 0.9837670283057346, + "grad_norm": 0.0015479176072403789, + "learning_rate": 0.2834955270536557, + "loss": 0.3994, + "num_input_tokens_seen": 12605456, + "step": 6030 + }, + { + "epoch": 0.984582755526552, + "grad_norm": 0.001667194184847176, + "learning_rate": 0.2834686550407866, + "loss": 0.3216, + "num_input_tokens_seen": 12615872, + "step": 6035 + }, + { + "epoch": 0.9853984827473693, + "grad_norm": 0.0007302198791876435, + "learning_rate": 0.28344176244537367, + "loss": 0.3123, + "num_input_tokens_seen": 12626144, + "step": 6040 + }, + { + "epoch": 0.9862142099681867, + "grad_norm": 0.0005407308926805854, + "learning_rate": 0.28341484927156396, + "loss": 0.3076, + "num_input_tokens_seen": 12636112, + "step": 6045 + }, + { + "epoch": 0.987029937189004, + "grad_norm": 0.0009110743412747979, + "learning_rate": 0.28338791552350795, + "loss": 0.342, + "num_input_tokens_seen": 12645168, + "step": 6050 + }, + { + "epoch": 0.9878456644098214, + "grad_norm": 0.0013451521517708898, + "learning_rate": 0.28336096120535914, + "loss": 0.3104, + "num_input_tokens_seen": 12656736, + "step": 6055 + }, + { + "epoch": 0.9886613916306387, + "grad_norm": 0.0013610690366476774, + "learning_rate": 0.2833339863212741, + "loss": 0.2941, + "num_input_tokens_seen": 12667472, + "step": 6060 + }, + { + "epoch": 0.9894771188514561, + "grad_norm": 0.0018596533918753266, + "learning_rate": 0.28330699087541283, + "loss": 0.2868, + "num_input_tokens_seen": 12678384, + "step": 6065 + }, + { + "epoch": 0.9902928460722734, + "grad_norm": 0.001612895866855979, + "learning_rate": 0.2832799748719384, + "loss": 0.3017, + "num_input_tokens_seen": 12689200, + "step": 6070 + }, + { + "epoch": 0.9911085732930908, + "grad_norm": 0.0007270342903211713, + "learning_rate": 0.28325293831501686, + "loss": 0.3354, + "num_input_tokens_seen": 12699568, + "step": 6075 + }, + { + "epoch": 0.9919243005139081, + "grad_norm": 0.0012539800954982638, + "learning_rate": 0.2832258812088177, + "loss": 0.3928, + "num_input_tokens_seen": 12710336, + "step": 6080 + }, + { + "epoch": 0.9927400277347255, + "grad_norm": 0.0011996247339993715, + "learning_rate": 0.2831988035575134, + "loss": 0.3026, + "num_input_tokens_seen": 12719392, + "step": 6085 + }, + { + "epoch": 0.9935557549555428, + "grad_norm": 0.0023828265257179737, + "learning_rate": 0.28317170536527975, + "loss": 0.3482, + "num_input_tokens_seen": 12729424, + "step": 6090 + }, + { + "epoch": 0.9943714821763602, + "grad_norm": 0.0010698542464524508, + "learning_rate": 0.2831445866362956, + "loss": 0.276, + "num_input_tokens_seen": 12741248, + "step": 6095 + }, + { + "epoch": 0.9951872093971775, + "grad_norm": 0.0016728847986087203, + "learning_rate": 0.2831174473747429, + "loss": 0.2902, + "num_input_tokens_seen": 12751184, + "step": 6100 + }, + { + "epoch": 0.996002936617995, + "grad_norm": 0.001149362069554627, + "learning_rate": 0.2830902875848071, + "loss": 0.3717, + "num_input_tokens_seen": 12760464, + "step": 6105 + }, + { + "epoch": 0.9968186638388123, + "grad_norm": 0.0015766486758366227, + "learning_rate": 0.28306310727067635, + "loss": 0.3598, + "num_input_tokens_seen": 12770672, + "step": 6110 + }, + { + "epoch": 0.9976343910596297, + "grad_norm": 0.0012423882726579905, + "learning_rate": 0.2830359064365423, + "loss": 0.3428, + "num_input_tokens_seen": 12781456, + "step": 6115 + }, + { + "epoch": 0.998450118280447, + "grad_norm": 0.000980466604232788, + "learning_rate": 0.28300868508659965, + "loss": 0.3471, + "num_input_tokens_seen": 12792592, + "step": 6120 + }, + { + "epoch": 0.9992658455012644, + "grad_norm": 0.000668411492370069, + "learning_rate": 0.28298144322504626, + "loss": 0.3206, + "num_input_tokens_seen": 12803840, + "step": 6125 + }, + { + "epoch": 1.0, + "grad_norm": 0.0021434943191707134, + "learning_rate": 0.2829541808560832, + "loss": 0.3229, + "num_input_tokens_seen": 12813920, + "step": 6130 + }, + { + "epoch": 1.0008157272208174, + "grad_norm": 0.0011109599145129323, + "learning_rate": 0.2829268979839146, + "loss": 0.3401, + "num_input_tokens_seen": 12824960, + "step": 6135 + }, + { + "epoch": 1.0016314544416347, + "grad_norm": 0.0012162828352302313, + "learning_rate": 0.2828995946127479, + "loss": 0.3114, + "num_input_tokens_seen": 12836112, + "step": 6140 + }, + { + "epoch": 1.002447181662452, + "grad_norm": 0.0007075215689837933, + "learning_rate": 0.2828722707467936, + "loss": 0.3003, + "num_input_tokens_seen": 12846560, + "step": 6145 + }, + { + "epoch": 1.0032629088832694, + "grad_norm": 0.0008012124453671277, + "learning_rate": 0.2828449263902653, + "loss": 0.3038, + "num_input_tokens_seen": 12858208, + "step": 6150 + }, + { + "epoch": 1.0040786361040868, + "grad_norm": 0.00239468808285892, + "learning_rate": 0.28281756154738, + "loss": 0.3299, + "num_input_tokens_seen": 12868608, + "step": 6155 + }, + { + "epoch": 1.0048943633249041, + "grad_norm": 0.0014215511037036777, + "learning_rate": 0.28279017622235764, + "loss": 0.3099, + "num_input_tokens_seen": 12878320, + "step": 6160 + }, + { + "epoch": 1.0057100905457215, + "grad_norm": 0.001335846376605332, + "learning_rate": 0.28276277041942127, + "loss": 0.3234, + "num_input_tokens_seen": 12889248, + "step": 6165 + }, + { + "epoch": 1.0065258177665388, + "grad_norm": 0.001983682857826352, + "learning_rate": 0.2827353441427974, + "loss": 0.2666, + "num_input_tokens_seen": 12900432, + "step": 6170 + }, + { + "epoch": 1.0073415449873562, + "grad_norm": 0.0019285471644252539, + "learning_rate": 0.2827078973967153, + "loss": 0.2783, + "num_input_tokens_seen": 12910800, + "step": 6175 + }, + { + "epoch": 1.0081572722081735, + "grad_norm": 0.0010410039685666561, + "learning_rate": 0.2826804301854078, + "loss": 0.3049, + "num_input_tokens_seen": 12921872, + "step": 6180 + }, + { + "epoch": 1.0089729994289909, + "grad_norm": 0.0013251702766865492, + "learning_rate": 0.2826529425131105, + "loss": 0.3653, + "num_input_tokens_seen": 12933024, + "step": 6185 + }, + { + "epoch": 1.0097887266498082, + "grad_norm": 0.0016101034125313163, + "learning_rate": 0.2826254343840625, + "loss": 0.3059, + "num_input_tokens_seen": 12943536, + "step": 6190 + }, + { + "epoch": 1.0106044538706256, + "grad_norm": 0.0009760434040799737, + "learning_rate": 0.2825979058025059, + "loss": 0.4049, + "num_input_tokens_seen": 12952080, + "step": 6195 + }, + { + "epoch": 1.011420181091443, + "grad_norm": 0.0010067711118608713, + "learning_rate": 0.2825703567726858, + "loss": 0.3557, + "num_input_tokens_seen": 12963008, + "step": 6200 + }, + { + "epoch": 1.011420181091443, + "eval_loss": 0.3178694546222687, + "eval_runtime": 156.0615, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 12963008, + "step": 6200 + }, + { + "epoch": 1.0122359083122603, + "grad_norm": 0.0012209475971758366, + "learning_rate": 0.2825427872988508, + "loss": 0.352, + "num_input_tokens_seen": 12974864, + "step": 6205 + }, + { + "epoch": 1.0130516355330776, + "grad_norm": 0.0008208458311855793, + "learning_rate": 0.28251519738525227, + "loss": 0.3087, + "num_input_tokens_seen": 12985216, + "step": 6210 + }, + { + "epoch": 1.013867362753895, + "grad_norm": 0.0008828988065943122, + "learning_rate": 0.28248758703614507, + "loss": 0.3163, + "num_input_tokens_seen": 12996976, + "step": 6215 + }, + { + "epoch": 1.0146830899747123, + "grad_norm": 0.0006987036322243512, + "learning_rate": 0.28245995625578696, + "loss": 0.2701, + "num_input_tokens_seen": 13007952, + "step": 6220 + }, + { + "epoch": 1.01549881719553, + "grad_norm": 0.001616298919543624, + "learning_rate": 0.282432305048439, + "loss": 0.2975, + "num_input_tokens_seen": 13016672, + "step": 6225 + }, + { + "epoch": 1.0163145444163473, + "grad_norm": 0.0009549586102366447, + "learning_rate": 0.28240463341836536, + "loss": 0.2982, + "num_input_tokens_seen": 13027520, + "step": 6230 + }, + { + "epoch": 1.0171302716371646, + "grad_norm": 0.0009515340207144618, + "learning_rate": 0.2823769413698334, + "loss": 0.3263, + "num_input_tokens_seen": 13037552, + "step": 6235 + }, + { + "epoch": 1.017945998857982, + "grad_norm": 0.0010316044790670276, + "learning_rate": 0.2823492289071135, + "loss": 0.3828, + "num_input_tokens_seen": 13047728, + "step": 6240 + }, + { + "epoch": 1.0187617260787993, + "grad_norm": 0.0006653384189121425, + "learning_rate": 0.2823214960344793, + "loss": 0.3355, + "num_input_tokens_seen": 13058880, + "step": 6245 + }, + { + "epoch": 1.0195774532996167, + "grad_norm": 0.000625389744527638, + "learning_rate": 0.28229374275620756, + "loss": 0.2766, + "num_input_tokens_seen": 13069856, + "step": 6250 + }, + { + "epoch": 1.020393180520434, + "grad_norm": 0.0009720442467369139, + "learning_rate": 0.28226596907657814, + "loss": 0.3457, + "num_input_tokens_seen": 13080160, + "step": 6255 + }, + { + "epoch": 1.0212089077412514, + "grad_norm": 0.0006199234048835933, + "learning_rate": 0.28223817499987414, + "loss": 0.37, + "num_input_tokens_seen": 13090832, + "step": 6260 + }, + { + "epoch": 1.0220246349620687, + "grad_norm": 0.000984571292065084, + "learning_rate": 0.2822103605303818, + "loss": 0.3192, + "num_input_tokens_seen": 13100976, + "step": 6265 + }, + { + "epoch": 1.022840362182886, + "grad_norm": 0.0007145425188355148, + "learning_rate": 0.2821825256723903, + "loss": 0.2807, + "num_input_tokens_seen": 13109904, + "step": 6270 + }, + { + "epoch": 1.0236560894037035, + "grad_norm": 0.001315491390414536, + "learning_rate": 0.2821546704301923, + "loss": 0.3814, + "num_input_tokens_seen": 13120032, + "step": 6275 + }, + { + "epoch": 1.0244718166245208, + "grad_norm": 0.000623669009655714, + "learning_rate": 0.2821267948080834, + "loss": 0.3255, + "num_input_tokens_seen": 13130448, + "step": 6280 + }, + { + "epoch": 1.0252875438453382, + "grad_norm": 0.0021557549480348825, + "learning_rate": 0.28209889881036226, + "loss": 0.3345, + "num_input_tokens_seen": 13142448, + "step": 6285 + }, + { + "epoch": 1.0261032710661555, + "grad_norm": 0.0013103581732138991, + "learning_rate": 0.28207098244133094, + "loss": 0.3118, + "num_input_tokens_seen": 13152592, + "step": 6290 + }, + { + "epoch": 1.0269189982869729, + "grad_norm": 0.0007934362511150539, + "learning_rate": 0.2820430457052943, + "loss": 0.3349, + "num_input_tokens_seen": 13163904, + "step": 6295 + }, + { + "epoch": 1.0277347255077902, + "grad_norm": 0.0010917377658188343, + "learning_rate": 0.28201508860656077, + "loss": 0.3308, + "num_input_tokens_seen": 13174064, + "step": 6300 + }, + { + "epoch": 1.0285504527286076, + "grad_norm": 0.000505866773892194, + "learning_rate": 0.2819871111494415, + "loss": 0.3689, + "num_input_tokens_seen": 13184528, + "step": 6305 + }, + { + "epoch": 1.029366179949425, + "grad_norm": 0.0007383420597761869, + "learning_rate": 0.28195911333825113, + "loss": 0.3281, + "num_input_tokens_seen": 13194464, + "step": 6310 + }, + { + "epoch": 1.0301819071702423, + "grad_norm": 0.000541556510142982, + "learning_rate": 0.28193109517730713, + "loss": 0.3006, + "num_input_tokens_seen": 13206272, + "step": 6315 + }, + { + "epoch": 1.0309976343910596, + "grad_norm": 0.0008423721301369369, + "learning_rate": 0.2819030566709303, + "loss": 0.3239, + "num_input_tokens_seen": 13216944, + "step": 6320 + }, + { + "epoch": 1.031813361611877, + "grad_norm": 0.0010244582081213593, + "learning_rate": 0.2818749978234445, + "loss": 0.2864, + "num_input_tokens_seen": 13227664, + "step": 6325 + }, + { + "epoch": 1.0326290888326943, + "grad_norm": 0.0010489802807569504, + "learning_rate": 0.2818469186391768, + "loss": 0.2573, + "num_input_tokens_seen": 13238832, + "step": 6330 + }, + { + "epoch": 1.0334448160535117, + "grad_norm": 0.0017450801096856594, + "learning_rate": 0.28181881912245743, + "loss": 0.3741, + "num_input_tokens_seen": 13248880, + "step": 6335 + }, + { + "epoch": 1.034260543274329, + "grad_norm": 0.0016866539372131228, + "learning_rate": 0.2817906992776195, + "loss": 0.3704, + "num_input_tokens_seen": 13259408, + "step": 6340 + }, + { + "epoch": 1.0350762704951464, + "grad_norm": 0.0010808075312525034, + "learning_rate": 0.28176255910899967, + "loss": 0.3036, + "num_input_tokens_seen": 13270432, + "step": 6345 + }, + { + "epoch": 1.0358919977159637, + "grad_norm": 0.0011952748754993081, + "learning_rate": 0.2817343986209373, + "loss": 0.3067, + "num_input_tokens_seen": 13282112, + "step": 6350 + }, + { + "epoch": 1.036707724936781, + "grad_norm": 0.0020363281946629286, + "learning_rate": 0.2817062178177753, + "loss": 0.3556, + "num_input_tokens_seen": 13293248, + "step": 6355 + }, + { + "epoch": 1.0375234521575984, + "grad_norm": 0.001097036642022431, + "learning_rate": 0.2816780167038593, + "loss": 0.2979, + "num_input_tokens_seen": 13303488, + "step": 6360 + }, + { + "epoch": 1.0383391793784158, + "grad_norm": 0.0016502422513440251, + "learning_rate": 0.28164979528353834, + "loss": 0.3673, + "num_input_tokens_seen": 13314848, + "step": 6365 + }, + { + "epoch": 1.0391549065992332, + "grad_norm": 0.0016043295618146658, + "learning_rate": 0.28162155356116453, + "loss": 0.3265, + "num_input_tokens_seen": 13325712, + "step": 6370 + }, + { + "epoch": 1.0399706338200505, + "grad_norm": 0.0004424923681654036, + "learning_rate": 0.28159329154109314, + "loss": 0.2809, + "num_input_tokens_seen": 13335664, + "step": 6375 + }, + { + "epoch": 1.0407863610408679, + "grad_norm": 0.00046899114386178553, + "learning_rate": 0.28156500922768246, + "loss": 0.5231, + "num_input_tokens_seen": 13345760, + "step": 6380 + }, + { + "epoch": 1.0416020882616852, + "grad_norm": 0.0012437719851732254, + "learning_rate": 0.28153670662529406, + "loss": 0.3386, + "num_input_tokens_seen": 13356144, + "step": 6385 + }, + { + "epoch": 1.0424178154825026, + "grad_norm": 0.0005771912401542068, + "learning_rate": 0.28150838373829246, + "loss": 0.3346, + "num_input_tokens_seen": 13366432, + "step": 6390 + }, + { + "epoch": 1.04323354270332, + "grad_norm": 0.0005077545647509396, + "learning_rate": 0.2814800405710455, + "loss": 0.2732, + "num_input_tokens_seen": 13378800, + "step": 6395 + }, + { + "epoch": 1.0440492699241373, + "grad_norm": 0.0008700948092155159, + "learning_rate": 0.2814516771279239, + "loss": 0.3641, + "num_input_tokens_seen": 13388336, + "step": 6400 + }, + { + "epoch": 1.0440492699241373, + "eval_loss": 0.32295671105384827, + "eval_runtime": 155.8139, + "eval_samples_per_second": 17.489, + "eval_steps_per_second": 8.748, + "num_input_tokens_seen": 13388336, + "step": 6400 + }, + { + "epoch": 1.0448649971449546, + "grad_norm": 0.0008301034686155617, + "learning_rate": 0.28142329341330186, + "loss": 0.3459, + "num_input_tokens_seen": 13399088, + "step": 6405 + }, + { + "epoch": 1.0456807243657722, + "grad_norm": 0.000826602685265243, + "learning_rate": 0.2813948894315564, + "loss": 0.3064, + "num_input_tokens_seen": 13410384, + "step": 6410 + }, + { + "epoch": 1.0464964515865895, + "grad_norm": 0.0016018039314076304, + "learning_rate": 0.2813664651870677, + "loss": 0.3026, + "num_input_tokens_seen": 13419216, + "step": 6415 + }, + { + "epoch": 1.047312178807407, + "grad_norm": 0.0010516223264858127, + "learning_rate": 0.28133802068421926, + "loss": 0.3964, + "num_input_tokens_seen": 13430224, + "step": 6420 + }, + { + "epoch": 1.0481279060282243, + "grad_norm": 0.0013267643516883254, + "learning_rate": 0.28130955592739754, + "loss": 0.3245, + "num_input_tokens_seen": 13441424, + "step": 6425 + }, + { + "epoch": 1.0489436332490416, + "grad_norm": 0.000644216313958168, + "learning_rate": 0.2812810709209922, + "loss": 0.2929, + "num_input_tokens_seen": 13451760, + "step": 6430 + }, + { + "epoch": 1.049759360469859, + "grad_norm": 0.0006156767485663295, + "learning_rate": 0.2812525656693959, + "loss": 0.3323, + "num_input_tokens_seen": 13462464, + "step": 6435 + }, + { + "epoch": 1.0505750876906763, + "grad_norm": 0.0007289440254680812, + "learning_rate": 0.28122404017700453, + "loss": 0.3104, + "num_input_tokens_seen": 13471968, + "step": 6440 + }, + { + "epoch": 1.0513908149114937, + "grad_norm": 0.0005940821720287204, + "learning_rate": 0.2811954944482171, + "loss": 0.2831, + "num_input_tokens_seen": 13483856, + "step": 6445 + }, + { + "epoch": 1.052206542132311, + "grad_norm": 0.0006517529836855829, + "learning_rate": 0.2811669284874358, + "loss": 0.2869, + "num_input_tokens_seen": 13494464, + "step": 6450 + }, + { + "epoch": 1.0530222693531284, + "grad_norm": 0.00036562365130521357, + "learning_rate": 0.2811383422990657, + "loss": 0.317, + "num_input_tokens_seen": 13503664, + "step": 6455 + }, + { + "epoch": 1.0538379965739457, + "grad_norm": 0.0013021518243476748, + "learning_rate": 0.2811097358875152, + "loss": 0.2392, + "num_input_tokens_seen": 13515824, + "step": 6460 + }, + { + "epoch": 1.054653723794763, + "grad_norm": 0.0006697159842588007, + "learning_rate": 0.2810811092571959, + "loss": 0.3422, + "num_input_tokens_seen": 13525712, + "step": 6465 + }, + { + "epoch": 1.0554694510155804, + "grad_norm": 0.0012632940197363496, + "learning_rate": 0.28105246241252224, + "loss": 0.2967, + "num_input_tokens_seen": 13536176, + "step": 6470 + }, + { + "epoch": 1.0562851782363978, + "grad_norm": 0.0010647661983966827, + "learning_rate": 0.28102379535791194, + "loss": 0.3197, + "num_input_tokens_seen": 13546704, + "step": 6475 + }, + { + "epoch": 1.0571009054572151, + "grad_norm": 0.001008500810712576, + "learning_rate": 0.2809951080977859, + "loss": 0.3245, + "num_input_tokens_seen": 13557520, + "step": 6480 + }, + { + "epoch": 1.0579166326780325, + "grad_norm": 0.0008449119632132351, + "learning_rate": 0.28096640063656797, + "loss": 0.3153, + "num_input_tokens_seen": 13568640, + "step": 6485 + }, + { + "epoch": 1.0587323598988498, + "grad_norm": 0.001287192921154201, + "learning_rate": 0.2809376729786852, + "loss": 0.3381, + "num_input_tokens_seen": 13580016, + "step": 6490 + }, + { + "epoch": 1.0595480871196672, + "grad_norm": 0.00214596395380795, + "learning_rate": 0.28090892512856785, + "loss": 0.2806, + "num_input_tokens_seen": 13589744, + "step": 6495 + }, + { + "epoch": 1.0603638143404845, + "grad_norm": 0.0004634445649571717, + "learning_rate": 0.2808801570906491, + "loss": 0.3216, + "num_input_tokens_seen": 13598896, + "step": 6500 + }, + { + "epoch": 1.061179541561302, + "grad_norm": 0.0018425980815663934, + "learning_rate": 0.2808513688693654, + "loss": 0.3323, + "num_input_tokens_seen": 13610352, + "step": 6505 + }, + { + "epoch": 1.0619952687821193, + "grad_norm": 0.0014948031166568398, + "learning_rate": 0.28082256046915627, + "loss": 0.3003, + "num_input_tokens_seen": 13621536, + "step": 6510 + }, + { + "epoch": 1.0628109960029366, + "grad_norm": 0.000860951142385602, + "learning_rate": 0.28079373189446427, + "loss": 0.3332, + "num_input_tokens_seen": 13632704, + "step": 6515 + }, + { + "epoch": 1.063626723223754, + "grad_norm": 0.0006001461879350245, + "learning_rate": 0.28076488314973513, + "loss": 0.3679, + "num_input_tokens_seen": 13642656, + "step": 6520 + }, + { + "epoch": 1.0644424504445713, + "grad_norm": 0.0029002202209085226, + "learning_rate": 0.28073601423941774, + "loss": 0.3948, + "num_input_tokens_seen": 13652800, + "step": 6525 + }, + { + "epoch": 1.0652581776653887, + "grad_norm": 0.0009980060858651996, + "learning_rate": 0.28070712516796403, + "loss": 0.3099, + "num_input_tokens_seen": 13664160, + "step": 6530 + }, + { + "epoch": 1.066073904886206, + "grad_norm": 0.0013570758746936917, + "learning_rate": 0.28067821593982906, + "loss": 0.3427, + "num_input_tokens_seen": 13675824, + "step": 6535 + }, + { + "epoch": 1.0668896321070234, + "grad_norm": 0.0005855279159732163, + "learning_rate": 0.28064928655947097, + "loss": 0.3354, + "num_input_tokens_seen": 13684960, + "step": 6540 + }, + { + "epoch": 1.0677053593278407, + "grad_norm": 0.0012672500452026725, + "learning_rate": 0.28062033703135103, + "loss": 0.3345, + "num_input_tokens_seen": 13696688, + "step": 6545 + }, + { + "epoch": 1.068521086548658, + "grad_norm": 0.001007338403724134, + "learning_rate": 0.2805913673599337, + "loss": 0.2908, + "num_input_tokens_seen": 13707440, + "step": 6550 + }, + { + "epoch": 1.0693368137694754, + "grad_norm": 0.0005731582059524953, + "learning_rate": 0.2805623775496864, + "loss": 0.3229, + "num_input_tokens_seen": 13718368, + "step": 6555 + }, + { + "epoch": 1.0701525409902928, + "grad_norm": 0.0014444963308051229, + "learning_rate": 0.2805333676050797, + "loss": 0.3255, + "num_input_tokens_seen": 13729152, + "step": 6560 + }, + { + "epoch": 1.0709682682111101, + "grad_norm": 0.0008368454873561859, + "learning_rate": 0.2805043375305873, + "loss": 0.2958, + "num_input_tokens_seen": 13740672, + "step": 6565 + }, + { + "epoch": 1.0717839954319275, + "grad_norm": 0.0008362674852833152, + "learning_rate": 0.2804752873306861, + "loss": 0.2675, + "num_input_tokens_seen": 13751360, + "step": 6570 + }, + { + "epoch": 1.0725997226527448, + "grad_norm": 0.001911698142066598, + "learning_rate": 0.2804462170098559, + "loss": 0.4189, + "num_input_tokens_seen": 13762896, + "step": 6575 + }, + { + "epoch": 1.0734154498735622, + "grad_norm": 0.0013350126100704074, + "learning_rate": 0.2804171265725797, + "loss": 0.3457, + "num_input_tokens_seen": 13774192, + "step": 6580 + }, + { + "epoch": 1.0742311770943795, + "grad_norm": 0.002430194756016135, + "learning_rate": 0.28038801602334373, + "loss": 0.3106, + "num_input_tokens_seen": 13784304, + "step": 6585 + }, + { + "epoch": 1.075046904315197, + "grad_norm": 0.0007345867343246937, + "learning_rate": 0.28035888536663717, + "loss": 0.3577, + "num_input_tokens_seen": 13794432, + "step": 6590 + }, + { + "epoch": 1.0758626315360145, + "grad_norm": 0.00085720099741593, + "learning_rate": 0.2803297346069522, + "loss": 0.2925, + "num_input_tokens_seen": 13805872, + "step": 6595 + }, + { + "epoch": 1.0766783587568316, + "grad_norm": 0.0007475372403860092, + "learning_rate": 0.28030056374878437, + "loss": 0.3471, + "num_input_tokens_seen": 13816224, + "step": 6600 + }, + { + "epoch": 1.0766783587568316, + "eval_loss": 0.330051988363266, + "eval_runtime": 156.0021, + "eval_samples_per_second": 17.468, + "eval_steps_per_second": 8.737, + "num_input_tokens_seen": 13816224, + "step": 6600 + }, + { + "epoch": 1.0774940859776492, + "grad_norm": 0.0006545928190462291, + "learning_rate": 0.2802713727966321, + "loss": 0.3066, + "num_input_tokens_seen": 13827264, + "step": 6605 + }, + { + "epoch": 1.0783098131984665, + "grad_norm": 0.0006637960905209184, + "learning_rate": 0.28024216175499717, + "loss": 0.3295, + "num_input_tokens_seen": 13837344, + "step": 6610 + }, + { + "epoch": 1.0791255404192839, + "grad_norm": 0.0013442556373775005, + "learning_rate": 0.2802129306283841, + "loss": 0.3752, + "num_input_tokens_seen": 13848368, + "step": 6615 + }, + { + "epoch": 1.0799412676401012, + "grad_norm": 0.0011590039357542992, + "learning_rate": 0.28018367942130074, + "loss": 0.3372, + "num_input_tokens_seen": 13858688, + "step": 6620 + }, + { + "epoch": 1.0807569948609186, + "grad_norm": 0.0006346975569613278, + "learning_rate": 0.28015440813825804, + "loss": 0.3425, + "num_input_tokens_seen": 13867728, + "step": 6625 + }, + { + "epoch": 1.081572722081736, + "grad_norm": 0.0010818393202498555, + "learning_rate": 0.28012511678377006, + "loss": 0.2975, + "num_input_tokens_seen": 13879216, + "step": 6630 + }, + { + "epoch": 1.0823884493025533, + "grad_norm": 0.0005228330264799297, + "learning_rate": 0.28009580536235373, + "loss": 0.3159, + "num_input_tokens_seen": 13888976, + "step": 6635 + }, + { + "epoch": 1.0832041765233706, + "grad_norm": 0.001181809464469552, + "learning_rate": 0.28006647387852934, + "loss": 0.333, + "num_input_tokens_seen": 13898064, + "step": 6640 + }, + { + "epoch": 1.084019903744188, + "grad_norm": 0.0006140198092907667, + "learning_rate": 0.28003712233682015, + "loss": 0.3205, + "num_input_tokens_seen": 13909136, + "step": 6645 + }, + { + "epoch": 1.0848356309650053, + "grad_norm": 0.0011640178272500634, + "learning_rate": 0.2800077507417526, + "loss": 0.2808, + "num_input_tokens_seen": 13919632, + "step": 6650 + }, + { + "epoch": 1.0856513581858227, + "grad_norm": 0.0008698559831827879, + "learning_rate": 0.2799783590978561, + "loss": 0.3728, + "num_input_tokens_seen": 13929552, + "step": 6655 + }, + { + "epoch": 1.08646708540664, + "grad_norm": 0.0009199125343002379, + "learning_rate": 0.2799489474096632, + "loss": 0.297, + "num_input_tokens_seen": 13941232, + "step": 6660 + }, + { + "epoch": 1.0872828126274574, + "grad_norm": 0.0005775021854788065, + "learning_rate": 0.27991951568170953, + "loss": 0.3499, + "num_input_tokens_seen": 13953264, + "step": 6665 + }, + { + "epoch": 1.0880985398482748, + "grad_norm": 0.0011679226299747825, + "learning_rate": 0.2798900639185339, + "loss": 0.3477, + "num_input_tokens_seen": 13963248, + "step": 6670 + }, + { + "epoch": 1.0889142670690921, + "grad_norm": 0.0006021509761922061, + "learning_rate": 0.2798605921246781, + "loss": 0.3313, + "num_input_tokens_seen": 13973680, + "step": 6675 + }, + { + "epoch": 1.0897299942899095, + "grad_norm": 0.0007308109197765589, + "learning_rate": 0.2798311003046871, + "loss": 0.3129, + "num_input_tokens_seen": 13984528, + "step": 6680 + }, + { + "epoch": 1.0905457215107268, + "grad_norm": 0.0006267097196541727, + "learning_rate": 0.2798015884631089, + "loss": 0.4088, + "num_input_tokens_seen": 13994032, + "step": 6685 + }, + { + "epoch": 1.0913614487315442, + "grad_norm": 0.002127984771504998, + "learning_rate": 0.27977205660449445, + "loss": 0.3285, + "num_input_tokens_seen": 14004976, + "step": 6690 + }, + { + "epoch": 1.0921771759523615, + "grad_norm": 0.0008664081688039005, + "learning_rate": 0.2797425047333981, + "loss": 0.3498, + "num_input_tokens_seen": 14014240, + "step": 6695 + }, + { + "epoch": 1.0929929031731789, + "grad_norm": 0.0005149489734321833, + "learning_rate": 0.27971293285437715, + "loss": 0.3392, + "num_input_tokens_seen": 14025152, + "step": 6700 + }, + { + "epoch": 1.0938086303939962, + "grad_norm": 0.0004840959154535085, + "learning_rate": 0.2796833409719918, + "loss": 0.3341, + "num_input_tokens_seen": 14036528, + "step": 6705 + }, + { + "epoch": 1.0946243576148136, + "grad_norm": 0.0010165638523176312, + "learning_rate": 0.27965372909080566, + "loss": 0.3553, + "num_input_tokens_seen": 14047568, + "step": 6710 + }, + { + "epoch": 1.095440084835631, + "grad_norm": 0.001095811021514237, + "learning_rate": 0.27962409721538506, + "loss": 0.2979, + "num_input_tokens_seen": 14057760, + "step": 6715 + }, + { + "epoch": 1.0962558120564483, + "grad_norm": 0.0005994185921736062, + "learning_rate": 0.27959444535029976, + "loss": 0.3437, + "num_input_tokens_seen": 14067936, + "step": 6720 + }, + { + "epoch": 1.0970715392772656, + "grad_norm": 0.0005757702165283263, + "learning_rate": 0.27956477350012243, + "loss": 0.315, + "num_input_tokens_seen": 14077616, + "step": 6725 + }, + { + "epoch": 1.097887266498083, + "grad_norm": 0.000491929764393717, + "learning_rate": 0.27953508166942875, + "loss": 0.3235, + "num_input_tokens_seen": 14087440, + "step": 6730 + }, + { + "epoch": 1.0987029937189003, + "grad_norm": 0.0007192486664280295, + "learning_rate": 0.27950536986279767, + "loss": 0.2966, + "num_input_tokens_seen": 14096880, + "step": 6735 + }, + { + "epoch": 1.0995187209397177, + "grad_norm": 0.0008505085716024041, + "learning_rate": 0.2794756380848111, + "loss": 0.3408, + "num_input_tokens_seen": 14108320, + "step": 6740 + }, + { + "epoch": 1.100334448160535, + "grad_norm": 0.000866222195327282, + "learning_rate": 0.279445886340054, + "loss": 0.3109, + "num_input_tokens_seen": 14116368, + "step": 6745 + }, + { + "epoch": 1.1011501753813524, + "grad_norm": 0.0016647030133754015, + "learning_rate": 0.27941611463311455, + "loss": 0.3621, + "num_input_tokens_seen": 14126352, + "step": 6750 + }, + { + "epoch": 1.1019659026021698, + "grad_norm": 0.0009757074294611812, + "learning_rate": 0.2793863229685839, + "loss": 0.3108, + "num_input_tokens_seen": 14137920, + "step": 6755 + }, + { + "epoch": 1.102781629822987, + "grad_norm": 0.0022787493653595448, + "learning_rate": 0.27935651135105627, + "loss": 0.3768, + "num_input_tokens_seen": 14148448, + "step": 6760 + }, + { + "epoch": 1.1035973570438045, + "grad_norm": 0.000652264105156064, + "learning_rate": 0.279326679785129, + "loss": 0.3265, + "num_input_tokens_seen": 14157248, + "step": 6765 + }, + { + "epoch": 1.1044130842646218, + "grad_norm": 0.0009993617422878742, + "learning_rate": 0.2792968282754024, + "loss": 0.343, + "num_input_tokens_seen": 14167216, + "step": 6770 + }, + { + "epoch": 1.1052288114854392, + "grad_norm": 0.0006668021087534726, + "learning_rate": 0.2792669568264801, + "loss": 0.3541, + "num_input_tokens_seen": 14176352, + "step": 6775 + }, + { + "epoch": 1.1060445387062567, + "grad_norm": 0.0007467989344149828, + "learning_rate": 0.27923706544296856, + "loss": 0.3467, + "num_input_tokens_seen": 14185840, + "step": 6780 + }, + { + "epoch": 1.1068602659270739, + "grad_norm": 0.0006195122259669006, + "learning_rate": 0.2792071541294775, + "loss": 0.3195, + "num_input_tokens_seen": 14196064, + "step": 6785 + }, + { + "epoch": 1.1076759931478914, + "grad_norm": 0.000709568674210459, + "learning_rate": 0.27917722289061947, + "loss": 0.3069, + "num_input_tokens_seen": 14207152, + "step": 6790 + }, + { + "epoch": 1.1084917203687088, + "grad_norm": 0.0004604713467415422, + "learning_rate": 0.27914727173101034, + "loss": 0.2844, + "num_input_tokens_seen": 14217392, + "step": 6795 + }, + { + "epoch": 1.1093074475895262, + "grad_norm": 0.000499785237479955, + "learning_rate": 0.279117300655269, + "loss": 0.3218, + "num_input_tokens_seen": 14228240, + "step": 6800 + }, + { + "epoch": 1.1093074475895262, + "eval_loss": 0.3166750967502594, + "eval_runtime": 156.0115, + "eval_samples_per_second": 17.467, + "eval_steps_per_second": 8.737, + "num_input_tokens_seen": 14228240, + "step": 6800 + }, + { + "epoch": 1.1101231748103435, + "grad_norm": 0.0006324832793325186, + "learning_rate": 0.2790873096680173, + "loss": 0.32, + "num_input_tokens_seen": 14237536, + "step": 6805 + }, + { + "epoch": 1.1109389020311609, + "grad_norm": 0.001243319595232606, + "learning_rate": 0.2790572987738802, + "loss": 0.3023, + "num_input_tokens_seen": 14249440, + "step": 6810 + }, + { + "epoch": 1.1117546292519782, + "grad_norm": 0.0008118283003568649, + "learning_rate": 0.27902726797748584, + "loss": 0.2675, + "num_input_tokens_seen": 14259952, + "step": 6815 + }, + { + "epoch": 1.1125703564727956, + "grad_norm": 0.00033010041806846857, + "learning_rate": 0.2789972172834652, + "loss": 0.2948, + "num_input_tokens_seen": 14270032, + "step": 6820 + }, + { + "epoch": 1.113386083693613, + "grad_norm": 0.0009974065469577909, + "learning_rate": 0.2789671466964527, + "loss": 0.3322, + "num_input_tokens_seen": 14279776, + "step": 6825 + }, + { + "epoch": 1.1142018109144303, + "grad_norm": 0.001084831077605486, + "learning_rate": 0.2789370562210854, + "loss": 0.3501, + "num_input_tokens_seen": 14289776, + "step": 6830 + }, + { + "epoch": 1.1150175381352476, + "grad_norm": 0.0008278153836727142, + "learning_rate": 0.27890694586200376, + "loss": 0.335, + "num_input_tokens_seen": 14298544, + "step": 6835 + }, + { + "epoch": 1.115833265356065, + "grad_norm": 0.0005646215286105871, + "learning_rate": 0.2788768156238511, + "loss": 0.2986, + "num_input_tokens_seen": 14308624, + "step": 6840 + }, + { + "epoch": 1.1166489925768823, + "grad_norm": 0.000733309134375304, + "learning_rate": 0.27884666551127385, + "loss": 0.3034, + "num_input_tokens_seen": 14318512, + "step": 6845 + }, + { + "epoch": 1.1174647197976997, + "grad_norm": 0.0011963247088715434, + "learning_rate": 0.2788164955289217, + "loss": 0.3252, + "num_input_tokens_seen": 14329520, + "step": 6850 + }, + { + "epoch": 1.118280447018517, + "grad_norm": 0.0006740109529346228, + "learning_rate": 0.27878630568144697, + "loss": 0.321, + "num_input_tokens_seen": 14339536, + "step": 6855 + }, + { + "epoch": 1.1190961742393344, + "grad_norm": 0.0008055256330408156, + "learning_rate": 0.2787560959735056, + "loss": 0.3744, + "num_input_tokens_seen": 14350128, + "step": 6860 + }, + { + "epoch": 1.1199119014601517, + "grad_norm": 0.0010715160751715302, + "learning_rate": 0.27872586640975616, + "loss": 0.2274, + "num_input_tokens_seen": 14360832, + "step": 6865 + }, + { + "epoch": 1.120727628680969, + "grad_norm": 0.00088644860079512, + "learning_rate": 0.27869561699486045, + "loss": 0.3479, + "num_input_tokens_seen": 14372192, + "step": 6870 + }, + { + "epoch": 1.1215433559017864, + "grad_norm": 0.0009118651505559683, + "learning_rate": 0.2786653477334833, + "loss": 0.3286, + "num_input_tokens_seen": 14382064, + "step": 6875 + }, + { + "epoch": 1.1223590831226038, + "grad_norm": 0.0009887823835015297, + "learning_rate": 0.2786350586302926, + "loss": 0.2889, + "num_input_tokens_seen": 14391312, + "step": 6880 + }, + { + "epoch": 1.1231748103434211, + "grad_norm": 0.0008154059178195894, + "learning_rate": 0.27860474968995935, + "loss": 0.3867, + "num_input_tokens_seen": 14401424, + "step": 6885 + }, + { + "epoch": 1.1239905375642385, + "grad_norm": 0.0006295953062362969, + "learning_rate": 0.27857442091715756, + "loss": 0.3345, + "num_input_tokens_seen": 14410832, + "step": 6890 + }, + { + "epoch": 1.1248062647850559, + "grad_norm": 0.0006650531431660056, + "learning_rate": 0.27854407231656425, + "loss": 0.3265, + "num_input_tokens_seen": 14421888, + "step": 6895 + }, + { + "epoch": 1.1256219920058732, + "grad_norm": 0.0007918119081296027, + "learning_rate": 0.2785137038928596, + "loss": 0.3735, + "num_input_tokens_seen": 14432848, + "step": 6900 + }, + { + "epoch": 1.1264377192266906, + "grad_norm": 0.0010489593259990215, + "learning_rate": 0.27848331565072687, + "loss": 0.3454, + "num_input_tokens_seen": 14443728, + "step": 6905 + }, + { + "epoch": 1.127253446447508, + "grad_norm": 0.0009881331352517009, + "learning_rate": 0.27845290759485225, + "loss": 0.2838, + "num_input_tokens_seen": 14455296, + "step": 6910 + }, + { + "epoch": 1.1280691736683253, + "grad_norm": 0.000554270634893328, + "learning_rate": 0.278422479729925, + "loss": 0.3679, + "num_input_tokens_seen": 14466016, + "step": 6915 + }, + { + "epoch": 1.1288849008891426, + "grad_norm": 0.001033591921441257, + "learning_rate": 0.2783920320606375, + "loss": 0.3421, + "num_input_tokens_seen": 14475296, + "step": 6920 + }, + { + "epoch": 1.12970062810996, + "grad_norm": 0.0009852772345766425, + "learning_rate": 0.2783615645916852, + "loss": 0.3386, + "num_input_tokens_seen": 14485280, + "step": 6925 + }, + { + "epoch": 1.1305163553307773, + "grad_norm": 0.0014642465393990278, + "learning_rate": 0.2783310773277666, + "loss": 0.35, + "num_input_tokens_seen": 14494960, + "step": 6930 + }, + { + "epoch": 1.1313320825515947, + "grad_norm": 0.0009581036283634603, + "learning_rate": 0.2783005702735831, + "loss": 0.2961, + "num_input_tokens_seen": 14506048, + "step": 6935 + }, + { + "epoch": 1.132147809772412, + "grad_norm": 0.0005946195451542735, + "learning_rate": 0.2782700434338394, + "loss": 0.3461, + "num_input_tokens_seen": 14515264, + "step": 6940 + }, + { + "epoch": 1.1329635369932294, + "grad_norm": 0.0003114934079349041, + "learning_rate": 0.278239496813243, + "loss": 0.3502, + "num_input_tokens_seen": 14524304, + "step": 6945 + }, + { + "epoch": 1.1337792642140467, + "grad_norm": 0.00039780489169061184, + "learning_rate": 0.27820893041650463, + "loss": 0.2967, + "num_input_tokens_seen": 14535152, + "step": 6950 + }, + { + "epoch": 1.134594991434864, + "grad_norm": 0.00041039413190446794, + "learning_rate": 0.27817834424833804, + "loss": 0.3365, + "num_input_tokens_seen": 14545488, + "step": 6955 + }, + { + "epoch": 1.1354107186556814, + "grad_norm": 0.0005173016106709838, + "learning_rate": 0.27814773831345996, + "loss": 0.3169, + "num_input_tokens_seen": 14556304, + "step": 6960 + }, + { + "epoch": 1.136226445876499, + "grad_norm": 0.0009043932659551501, + "learning_rate": 0.2781171126165902, + "loss": 0.3695, + "num_input_tokens_seen": 14567424, + "step": 6965 + }, + { + "epoch": 1.1370421730973161, + "grad_norm": 0.0004862657224293798, + "learning_rate": 0.2780864671624517, + "loss": 0.2787, + "num_input_tokens_seen": 14576928, + "step": 6970 + }, + { + "epoch": 1.1378579003181337, + "grad_norm": 0.0011931447079405189, + "learning_rate": 0.27805580195577034, + "loss": 0.2993, + "num_input_tokens_seen": 14585392, + "step": 6975 + }, + { + "epoch": 1.1386736275389508, + "grad_norm": 0.0010432382114231586, + "learning_rate": 0.2780251170012751, + "loss": 0.2945, + "num_input_tokens_seen": 14595792, + "step": 6980 + }, + { + "epoch": 1.1394893547597684, + "grad_norm": 0.0006243549287319183, + "learning_rate": 0.27799441230369787, + "loss": 0.3259, + "num_input_tokens_seen": 14606464, + "step": 6985 + }, + { + "epoch": 1.1403050819805858, + "grad_norm": 0.001123800058849156, + "learning_rate": 0.27796368786777387, + "loss": 0.3111, + "num_input_tokens_seen": 14616576, + "step": 6990 + }, + { + "epoch": 1.1411208092014031, + "grad_norm": 0.0009382382850162685, + "learning_rate": 0.277932943698241, + "loss": 0.3255, + "num_input_tokens_seen": 14626944, + "step": 6995 + }, + { + "epoch": 1.1419365364222205, + "grad_norm": 0.0004672382783610374, + "learning_rate": 0.2779021797998406, + "loss": 0.3316, + "num_input_tokens_seen": 14637984, + "step": 7000 + }, + { + "epoch": 1.1419365364222205, + "eval_loss": 0.3165501654148102, + "eval_runtime": 156.0094, + "eval_samples_per_second": 17.467, + "eval_steps_per_second": 8.737, + "num_input_tokens_seen": 14637984, + "step": 7000 + }, + { + "epoch": 1.1427522636430378, + "grad_norm": 0.0017864840338006616, + "learning_rate": 0.2778713961773167, + "loss": 0.338, + "num_input_tokens_seen": 14648064, + "step": 7005 + }, + { + "epoch": 1.1435679908638552, + "grad_norm": 0.0008107745088636875, + "learning_rate": 0.2778405928354166, + "loss": 0.3166, + "num_input_tokens_seen": 14658048, + "step": 7010 + }, + { + "epoch": 1.1443837180846725, + "grad_norm": 0.0004942925879731774, + "learning_rate": 0.27780976977889055, + "loss": 0.2969, + "num_input_tokens_seen": 14668256, + "step": 7015 + }, + { + "epoch": 1.14519944530549, + "grad_norm": 0.0005943296127952635, + "learning_rate": 0.27777892701249185, + "loss": 0.2997, + "num_input_tokens_seen": 14679280, + "step": 7020 + }, + { + "epoch": 1.1460151725263072, + "grad_norm": 0.0006757128867320716, + "learning_rate": 0.2777480645409768, + "loss": 0.3425, + "num_input_tokens_seen": 14688992, + "step": 7025 + }, + { + "epoch": 1.1468308997471246, + "grad_norm": 0.0003920726594515145, + "learning_rate": 0.27771718236910486, + "loss": 0.3489, + "num_input_tokens_seen": 14699328, + "step": 7030 + }, + { + "epoch": 1.147646626967942, + "grad_norm": 0.0011811456643044949, + "learning_rate": 0.27768628050163835, + "loss": 0.3272, + "num_input_tokens_seen": 14711264, + "step": 7035 + }, + { + "epoch": 1.1484623541887593, + "grad_norm": 0.0007965599652379751, + "learning_rate": 0.2776553589433428, + "loss": 0.3411, + "num_input_tokens_seen": 14721504, + "step": 7040 + }, + { + "epoch": 1.1492780814095767, + "grad_norm": 0.0010781108867377043, + "learning_rate": 0.27762441769898666, + "loss": 0.3555, + "num_input_tokens_seen": 14730896, + "step": 7045 + }, + { + "epoch": 1.150093808630394, + "grad_norm": 0.0006070437957532704, + "learning_rate": 0.2775934567733415, + "loss": 0.3, + "num_input_tokens_seen": 14740448, + "step": 7050 + }, + { + "epoch": 1.1509095358512114, + "grad_norm": 0.001150371041148901, + "learning_rate": 0.2775624761711819, + "loss": 0.3628, + "num_input_tokens_seen": 14751216, + "step": 7055 + }, + { + "epoch": 1.1517252630720287, + "grad_norm": 0.0009390509803779423, + "learning_rate": 0.2775314758972854, + "loss": 0.3318, + "num_input_tokens_seen": 14760720, + "step": 7060 + }, + { + "epoch": 1.152540990292846, + "grad_norm": 0.0008495066431351006, + "learning_rate": 0.2775004559564327, + "loss": 0.3062, + "num_input_tokens_seen": 14772848, + "step": 7065 + }, + { + "epoch": 1.1533567175136634, + "grad_norm": 0.0012819591211155057, + "learning_rate": 0.2774694163534073, + "loss": 0.3639, + "num_input_tokens_seen": 14782528, + "step": 7070 + }, + { + "epoch": 1.1541724447344808, + "grad_norm": 0.0010149044683203101, + "learning_rate": 0.27743835709299614, + "loss": 0.322, + "num_input_tokens_seen": 14791968, + "step": 7075 + }, + { + "epoch": 1.1549881719552981, + "grad_norm": 0.0006113193812780082, + "learning_rate": 0.2774072781799888, + "loss": 0.3239, + "num_input_tokens_seen": 14801616, + "step": 7080 + }, + { + "epoch": 1.1558038991761155, + "grad_norm": 0.0006507239886559546, + "learning_rate": 0.27737617961917804, + "loss": 0.3484, + "num_input_tokens_seen": 14812272, + "step": 7085 + }, + { + "epoch": 1.1566196263969328, + "grad_norm": 0.0007504763780161738, + "learning_rate": 0.27734506141535964, + "loss": 0.2955, + "num_input_tokens_seen": 14823552, + "step": 7090 + }, + { + "epoch": 1.1574353536177502, + "grad_norm": 0.00055135996080935, + "learning_rate": 0.2773139235733325, + "loss": 0.3004, + "num_input_tokens_seen": 14833920, + "step": 7095 + }, + { + "epoch": 1.1582510808385675, + "grad_norm": 0.0009696233319118619, + "learning_rate": 0.2772827660978984, + "loss": 0.3224, + "num_input_tokens_seen": 14842816, + "step": 7100 + }, + { + "epoch": 1.159066808059385, + "grad_norm": 0.0010645416332408786, + "learning_rate": 0.27725158899386226, + "loss": 0.2822, + "num_input_tokens_seen": 14853104, + "step": 7105 + }, + { + "epoch": 1.1598825352802022, + "grad_norm": 0.0011569594498723745, + "learning_rate": 0.27722039226603196, + "loss": 0.3567, + "num_input_tokens_seen": 14862176, + "step": 7110 + }, + { + "epoch": 1.1606982625010196, + "grad_norm": 0.0005182621534913778, + "learning_rate": 0.2771891759192184, + "loss": 0.3107, + "num_input_tokens_seen": 14872160, + "step": 7115 + }, + { + "epoch": 1.161513989721837, + "grad_norm": 0.0007239593542180955, + "learning_rate": 0.2771579399582355, + "loss": 0.3554, + "num_input_tokens_seen": 14882576, + "step": 7120 + }, + { + "epoch": 1.1623297169426543, + "grad_norm": 0.0007827209774404764, + "learning_rate": 0.2771266843879004, + "loss": 0.3005, + "num_input_tokens_seen": 14893040, + "step": 7125 + }, + { + "epoch": 1.1631454441634717, + "grad_norm": 0.0005349952261894941, + "learning_rate": 0.2770954092130329, + "loss": 0.3156, + "num_input_tokens_seen": 14904336, + "step": 7130 + }, + { + "epoch": 1.163961171384289, + "grad_norm": 0.0009000332211144269, + "learning_rate": 0.27706411443845613, + "loss": 0.2847, + "num_input_tokens_seen": 14915840, + "step": 7135 + }, + { + "epoch": 1.1647768986051064, + "grad_norm": 0.0006063259206712246, + "learning_rate": 0.27703280006899617, + "loss": 0.3227, + "num_input_tokens_seen": 14925776, + "step": 7140 + }, + { + "epoch": 1.1655926258259237, + "grad_norm": 0.0006671081646345556, + "learning_rate": 0.277001466109482, + "loss": 0.2779, + "num_input_tokens_seen": 14936208, + "step": 7145 + }, + { + "epoch": 1.1664083530467413, + "grad_norm": 0.0006749749300070107, + "learning_rate": 0.2769701125647458, + "loss": 0.3199, + "num_input_tokens_seen": 14947456, + "step": 7150 + }, + { + "epoch": 1.1672240802675584, + "grad_norm": 0.001933797262609005, + "learning_rate": 0.27693873943962266, + "loss": 0.3403, + "num_input_tokens_seen": 14956112, + "step": 7155 + }, + { + "epoch": 1.168039807488376, + "grad_norm": 0.0014118188992142677, + "learning_rate": 0.2769073467389506, + "loss": 0.4132, + "num_input_tokens_seen": 14965344, + "step": 7160 + }, + { + "epoch": 1.1688555347091931, + "grad_norm": 0.0016412826953455806, + "learning_rate": 0.2768759344675709, + "loss": 0.2842, + "num_input_tokens_seen": 14975888, + "step": 7165 + }, + { + "epoch": 1.1696712619300107, + "grad_norm": 0.0004257354012224823, + "learning_rate": 0.27684450263032767, + "loss": 0.3264, + "num_input_tokens_seen": 14986160, + "step": 7170 + }, + { + "epoch": 1.170486989150828, + "grad_norm": 0.001428481424227357, + "learning_rate": 0.2768130512320682, + "loss": 0.3386, + "num_input_tokens_seen": 14996544, + "step": 7175 + }, + { + "epoch": 1.1713027163716454, + "grad_norm": 0.0005097856046631932, + "learning_rate": 0.27678158027764244, + "loss": 0.3344, + "num_input_tokens_seen": 15007712, + "step": 7180 + }, + { + "epoch": 1.1721184435924628, + "grad_norm": 0.0005588092608377337, + "learning_rate": 0.27675008977190385, + "loss": 0.3202, + "num_input_tokens_seen": 15017056, + "step": 7185 + }, + { + "epoch": 1.17293417081328, + "grad_norm": 0.0009421606082469225, + "learning_rate": 0.2767185797197086, + "loss": 0.3131, + "num_input_tokens_seen": 15028048, + "step": 7190 + }, + { + "epoch": 1.1737498980340975, + "grad_norm": 0.0019289972260594368, + "learning_rate": 0.2766870501259159, + "loss": 0.3313, + "num_input_tokens_seen": 15038096, + "step": 7195 + }, + { + "epoch": 1.1745656252549148, + "grad_norm": 0.0012004096060991287, + "learning_rate": 0.276655500995388, + "loss": 0.2862, + "num_input_tokens_seen": 15049216, + "step": 7200 + }, + { + "epoch": 1.1745656252549148, + "eval_loss": 0.31657156348228455, + "eval_runtime": 156.1144, + "eval_samples_per_second": 17.455, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 15049216, + "step": 7200 + }, + { + "epoch": 1.1753813524757322, + "grad_norm": 0.0005434803897514939, + "learning_rate": 0.27662393233299015, + "loss": 0.3302, + "num_input_tokens_seen": 15060080, + "step": 7205 + }, + { + "epoch": 1.1761970796965495, + "grad_norm": 0.0008126180036924779, + "learning_rate": 0.27659234414359074, + "loss": 0.3116, + "num_input_tokens_seen": 15071408, + "step": 7210 + }, + { + "epoch": 1.1770128069173669, + "grad_norm": 0.0014177251141518354, + "learning_rate": 0.27656073643206097, + "loss": 0.3226, + "num_input_tokens_seen": 15080496, + "step": 7215 + }, + { + "epoch": 1.1778285341381842, + "grad_norm": 0.0013334930408746004, + "learning_rate": 0.27652910920327517, + "loss": 0.322, + "num_input_tokens_seen": 15090592, + "step": 7220 + }, + { + "epoch": 1.1786442613590016, + "grad_norm": 0.0008444556151516736, + "learning_rate": 0.2764974624621107, + "loss": 0.3748, + "num_input_tokens_seen": 15100832, + "step": 7225 + }, + { + "epoch": 1.179459988579819, + "grad_norm": 0.0010920284548774362, + "learning_rate": 0.2764657962134479, + "loss": 0.3545, + "num_input_tokens_seen": 15111104, + "step": 7230 + }, + { + "epoch": 1.1802757158006363, + "grad_norm": 0.0005661523318849504, + "learning_rate": 0.27643411046217, + "loss": 0.3458, + "num_input_tokens_seen": 15121152, + "step": 7235 + }, + { + "epoch": 1.1810914430214536, + "grad_norm": 0.0005858746590092778, + "learning_rate": 0.27640240521316334, + "loss": 0.3348, + "num_input_tokens_seen": 15132800, + "step": 7240 + }, + { + "epoch": 1.181907170242271, + "grad_norm": 0.0003644682001322508, + "learning_rate": 0.2763706804713174, + "loss": 0.314, + "num_input_tokens_seen": 15143632, + "step": 7245 + }, + { + "epoch": 1.1827228974630883, + "grad_norm": 0.0005006123683415353, + "learning_rate": 0.2763389362415245, + "loss": 0.3167, + "num_input_tokens_seen": 15154688, + "step": 7250 + }, + { + "epoch": 1.1835386246839057, + "grad_norm": 0.0014588994672521949, + "learning_rate": 0.27630717252867987, + "loss": 0.3314, + "num_input_tokens_seen": 15165136, + "step": 7255 + }, + { + "epoch": 1.184354351904723, + "grad_norm": 0.0004008500254712999, + "learning_rate": 0.276275389337682, + "loss": 0.3451, + "num_input_tokens_seen": 15173968, + "step": 7260 + }, + { + "epoch": 1.1851700791255404, + "grad_norm": 0.0014999104896560311, + "learning_rate": 0.2762435866734322, + "loss": 0.3336, + "num_input_tokens_seen": 15184720, + "step": 7265 + }, + { + "epoch": 1.1859858063463578, + "grad_norm": 0.0006607582326978445, + "learning_rate": 0.27621176454083485, + "loss": 0.3247, + "num_input_tokens_seen": 15195024, + "step": 7270 + }, + { + "epoch": 1.186801533567175, + "grad_norm": 0.0007454791339114308, + "learning_rate": 0.2761799229447973, + "loss": 0.2853, + "num_input_tokens_seen": 15206256, + "step": 7275 + }, + { + "epoch": 1.1876172607879925, + "grad_norm": 0.0007972342427819967, + "learning_rate": 0.27614806189023006, + "loss": 0.3186, + "num_input_tokens_seen": 15218832, + "step": 7280 + }, + { + "epoch": 1.1884329880088098, + "grad_norm": 0.0006535719148814678, + "learning_rate": 0.27611618138204636, + "loss": 0.2906, + "num_input_tokens_seen": 15229536, + "step": 7285 + }, + { + "epoch": 1.1892487152296272, + "grad_norm": 0.0014041749527677894, + "learning_rate": 0.2760842814251626, + "loss": 0.3946, + "num_input_tokens_seen": 15240512, + "step": 7290 + }, + { + "epoch": 1.1900644424504445, + "grad_norm": 0.0012399714905768633, + "learning_rate": 0.2760523620244982, + "loss": 0.3652, + "num_input_tokens_seen": 15250080, + "step": 7295 + }, + { + "epoch": 1.1908801696712619, + "grad_norm": 0.0058816587552428246, + "learning_rate": 0.27602042318497544, + "loss": 0.3273, + "num_input_tokens_seen": 15260704, + "step": 7300 + }, + { + "epoch": 1.1916958968920792, + "grad_norm": 0.0018553159898146987, + "learning_rate": 0.2759884649115198, + "loss": 0.3724, + "num_input_tokens_seen": 15269504, + "step": 7305 + }, + { + "epoch": 1.1925116241128966, + "grad_norm": 0.0005208657239563763, + "learning_rate": 0.2759564872090596, + "loss": 0.2916, + "num_input_tokens_seen": 15280064, + "step": 7310 + }, + { + "epoch": 1.193327351333714, + "grad_norm": 0.0011318603064864874, + "learning_rate": 0.2759244900825262, + "loss": 0.31, + "num_input_tokens_seen": 15291456, + "step": 7315 + }, + { + "epoch": 1.1941430785545313, + "grad_norm": 0.0016729385824874043, + "learning_rate": 0.2758924735368539, + "loss": 0.3115, + "num_input_tokens_seen": 15303968, + "step": 7320 + }, + { + "epoch": 1.1949588057753486, + "grad_norm": 0.0007479885243810713, + "learning_rate": 0.27586043757698014, + "loss": 0.4932, + "num_input_tokens_seen": 15313760, + "step": 7325 + }, + { + "epoch": 1.195774532996166, + "grad_norm": 0.007419213652610779, + "learning_rate": 0.27582838220784534, + "loss": 0.5014, + "num_input_tokens_seen": 15324192, + "step": 7330 + }, + { + "epoch": 1.1965902602169836, + "grad_norm": 0.006174216978251934, + "learning_rate": 0.27579630743439265, + "loss": 0.3826, + "num_input_tokens_seen": 15335456, + "step": 7335 + }, + { + "epoch": 1.1974059874378007, + "grad_norm": 0.0017082620179280639, + "learning_rate": 0.2757642132615686, + "loss": 0.3701, + "num_input_tokens_seen": 15345600, + "step": 7340 + }, + { + "epoch": 1.1982217146586183, + "grad_norm": 0.001243835547938943, + "learning_rate": 0.2757320996943223, + "loss": 0.3194, + "num_input_tokens_seen": 15356544, + "step": 7345 + }, + { + "epoch": 1.1990374418794354, + "grad_norm": 0.0006160566117614508, + "learning_rate": 0.2756999667376062, + "loss": 0.335, + "num_input_tokens_seen": 15367232, + "step": 7350 + }, + { + "epoch": 1.199853169100253, + "grad_norm": 0.0009901289595291018, + "learning_rate": 0.2756678143963756, + "loss": 0.3719, + "num_input_tokens_seen": 15377168, + "step": 7355 + }, + { + "epoch": 1.2006688963210703, + "grad_norm": 0.002064318163320422, + "learning_rate": 0.2756356426755888, + "loss": 0.3214, + "num_input_tokens_seen": 15387088, + "step": 7360 + }, + { + "epoch": 1.2014846235418877, + "grad_norm": 0.0022350528743118048, + "learning_rate": 0.27560345158020705, + "loss": 0.3363, + "num_input_tokens_seen": 15397584, + "step": 7365 + }, + { + "epoch": 1.202300350762705, + "grad_norm": 0.0016319799469783902, + "learning_rate": 0.27557124111519465, + "loss": 0.3993, + "num_input_tokens_seen": 15407264, + "step": 7370 + }, + { + "epoch": 1.2031160779835224, + "grad_norm": 0.002384593477472663, + "learning_rate": 0.27553901128551883, + "loss": 0.3499, + "num_input_tokens_seen": 15418080, + "step": 7375 + }, + { + "epoch": 1.2039318052043397, + "grad_norm": 0.000399889366235584, + "learning_rate": 0.2755067620961498, + "loss": 0.3508, + "num_input_tokens_seen": 15428464, + "step": 7380 + }, + { + "epoch": 1.204747532425157, + "grad_norm": 0.001461653271690011, + "learning_rate": 0.27547449355206094, + "loss": 0.3298, + "num_input_tokens_seen": 15438320, + "step": 7385 + }, + { + "epoch": 1.2055632596459744, + "grad_norm": 0.0007666203891858459, + "learning_rate": 0.2754422056582283, + "loss": 0.3488, + "num_input_tokens_seen": 15449360, + "step": 7390 + }, + { + "epoch": 1.2063789868667918, + "grad_norm": 0.0006161309429444373, + "learning_rate": 0.27540989841963115, + "loss": 0.3602, + "num_input_tokens_seen": 15460496, + "step": 7395 + }, + { + "epoch": 1.2071947140876091, + "grad_norm": 0.0005947009194642305, + "learning_rate": 0.27537757184125167, + "loss": 0.2954, + "num_input_tokens_seen": 15471984, + "step": 7400 + }, + { + "epoch": 1.2071947140876091, + "eval_loss": 0.3433842062950134, + "eval_runtime": 155.695, + "eval_samples_per_second": 17.502, + "eval_steps_per_second": 8.754, + "num_input_tokens_seen": 15471984, + "step": 7400 + }, + { + "epoch": 1.2080104413084265, + "grad_norm": 0.0026144422590732574, + "learning_rate": 0.275345225928075, + "loss": 0.407, + "num_input_tokens_seen": 15481472, + "step": 7405 + }, + { + "epoch": 1.2088261685292438, + "grad_norm": 0.0012338814558461308, + "learning_rate": 0.2753128606850893, + "loss": 0.366, + "num_input_tokens_seen": 15492544, + "step": 7410 + }, + { + "epoch": 1.2096418957500612, + "grad_norm": 0.002255290513858199, + "learning_rate": 0.2752804761172858, + "loss": 0.3734, + "num_input_tokens_seen": 15503408, + "step": 7415 + }, + { + "epoch": 1.2104576229708786, + "grad_norm": 0.0006136256270110607, + "learning_rate": 0.27524807222965836, + "loss": 0.3753, + "num_input_tokens_seen": 15514256, + "step": 7420 + }, + { + "epoch": 1.211273350191696, + "grad_norm": 0.0017258509760722518, + "learning_rate": 0.27521564902720436, + "loss": 0.3849, + "num_input_tokens_seen": 15525456, + "step": 7425 + }, + { + "epoch": 1.2120890774125133, + "grad_norm": 0.00040799571434035897, + "learning_rate": 0.2751832065149236, + "loss": 0.3466, + "num_input_tokens_seen": 15535600, + "step": 7430 + }, + { + "epoch": 1.2129048046333306, + "grad_norm": 0.0023727375082671642, + "learning_rate": 0.2751507446978193, + "loss": 0.3447, + "num_input_tokens_seen": 15546976, + "step": 7435 + }, + { + "epoch": 1.213720531854148, + "grad_norm": 0.001061539282090962, + "learning_rate": 0.2751182635808974, + "loss": 0.3064, + "num_input_tokens_seen": 15556928, + "step": 7440 + }, + { + "epoch": 1.2145362590749653, + "grad_norm": 0.0007028838736005127, + "learning_rate": 0.27508576316916694, + "loss": 0.369, + "num_input_tokens_seen": 15568192, + "step": 7445 + }, + { + "epoch": 1.2153519862957827, + "grad_norm": 0.000554053345695138, + "learning_rate": 0.2750532434676399, + "loss": 0.3405, + "num_input_tokens_seen": 15579360, + "step": 7450 + }, + { + "epoch": 1.2161677135166, + "grad_norm": 0.0004182024276815355, + "learning_rate": 0.27502070448133115, + "loss": 0.3058, + "num_input_tokens_seen": 15589680, + "step": 7455 + }, + { + "epoch": 1.2169834407374174, + "grad_norm": 0.0006424047169275582, + "learning_rate": 0.2749881462152587, + "loss": 0.2806, + "num_input_tokens_seen": 15600704, + "step": 7460 + }, + { + "epoch": 1.2177991679582347, + "grad_norm": 0.002104311715811491, + "learning_rate": 0.2749555686744434, + "loss": 0.4084, + "num_input_tokens_seen": 15611024, + "step": 7465 + }, + { + "epoch": 1.218614895179052, + "grad_norm": 0.001111306599341333, + "learning_rate": 0.2749229718639091, + "loss": 0.35, + "num_input_tokens_seen": 15621232, + "step": 7470 + }, + { + "epoch": 1.2194306223998694, + "grad_norm": 0.0008619750733487308, + "learning_rate": 0.27489035578868265, + "loss": 0.3652, + "num_input_tokens_seen": 15630976, + "step": 7475 + }, + { + "epoch": 1.2202463496206868, + "grad_norm": 0.0008383870590478182, + "learning_rate": 0.2748577204537939, + "loss": 0.3564, + "num_input_tokens_seen": 15640880, + "step": 7480 + }, + { + "epoch": 1.2210620768415041, + "grad_norm": 0.0007965630502440035, + "learning_rate": 0.2748250658642756, + "loss": 0.3528, + "num_input_tokens_seen": 15651040, + "step": 7485 + }, + { + "epoch": 1.2218778040623215, + "grad_norm": 0.0010911200661212206, + "learning_rate": 0.2747923920251634, + "loss": 0.3585, + "num_input_tokens_seen": 15661248, + "step": 7490 + }, + { + "epoch": 1.2226935312831388, + "grad_norm": 0.0002003743138629943, + "learning_rate": 0.27475969894149627, + "loss": 0.3461, + "num_input_tokens_seen": 15670560, + "step": 7495 + }, + { + "epoch": 1.2235092585039562, + "grad_norm": 0.00032642463338561356, + "learning_rate": 0.2747269866183156, + "loss": 0.3449, + "num_input_tokens_seen": 15681296, + "step": 7500 + }, + { + "epoch": 1.2243249857247736, + "grad_norm": 0.0005888153682462871, + "learning_rate": 0.27469425506066625, + "loss": 0.3364, + "num_input_tokens_seen": 15692880, + "step": 7505 + }, + { + "epoch": 1.225140712945591, + "grad_norm": 0.0010691258357837796, + "learning_rate": 0.27466150427359576, + "loss": 0.3395, + "num_input_tokens_seen": 15703168, + "step": 7510 + }, + { + "epoch": 1.2259564401664083, + "grad_norm": 0.0005221906467340887, + "learning_rate": 0.2746287342621547, + "loss": 0.3584, + "num_input_tokens_seen": 15713568, + "step": 7515 + }, + { + "epoch": 1.2267721673872258, + "grad_norm": 0.0005201178719289601, + "learning_rate": 0.2745959450313966, + "loss": 0.3195, + "num_input_tokens_seen": 15723872, + "step": 7520 + }, + { + "epoch": 1.227587894608043, + "grad_norm": 0.0011251755058765411, + "learning_rate": 0.27456313658637804, + "loss": 0.3526, + "num_input_tokens_seen": 15733472, + "step": 7525 + }, + { + "epoch": 1.2284036218288605, + "grad_norm": 0.00047092809109017253, + "learning_rate": 0.27453030893215846, + "loss": 0.3153, + "num_input_tokens_seen": 15743920, + "step": 7530 + }, + { + "epoch": 1.2292193490496777, + "grad_norm": 0.0004373456758912653, + "learning_rate": 0.2744974620738003, + "loss": 0.4123, + "num_input_tokens_seen": 15753648, + "step": 7535 + }, + { + "epoch": 1.2300350762704952, + "grad_norm": 0.000282588618574664, + "learning_rate": 0.27446459601636897, + "loss": 0.3654, + "num_input_tokens_seen": 15764160, + "step": 7540 + }, + { + "epoch": 1.2308508034913126, + "grad_norm": 0.000246456969762221, + "learning_rate": 0.2744317107649328, + "loss": 0.3522, + "num_input_tokens_seen": 15774512, + "step": 7545 + }, + { + "epoch": 1.23166653071213, + "grad_norm": 0.0006277721840888262, + "learning_rate": 0.2743988063245631, + "loss": 0.3703, + "num_input_tokens_seen": 15784944, + "step": 7550 + }, + { + "epoch": 1.2324822579329473, + "grad_norm": 0.0008805602556094527, + "learning_rate": 0.2743658827003342, + "loss": 0.3574, + "num_input_tokens_seen": 15795776, + "step": 7555 + }, + { + "epoch": 1.2332979851537647, + "grad_norm": 0.0007793106487952173, + "learning_rate": 0.27433293989732327, + "loss": 0.3574, + "num_input_tokens_seen": 15806304, + "step": 7560 + }, + { + "epoch": 1.234113712374582, + "grad_norm": 0.0002660374157130718, + "learning_rate": 0.27429997792061056, + "loss": 0.3433, + "num_input_tokens_seen": 15817376, + "step": 7565 + }, + { + "epoch": 1.2349294395953994, + "grad_norm": 0.00023994105868041515, + "learning_rate": 0.27426699677527927, + "loss": 0.3399, + "num_input_tokens_seen": 15827072, + "step": 7570 + }, + { + "epoch": 1.2357451668162167, + "grad_norm": 0.00028267400921322405, + "learning_rate": 0.2742339964664154, + "loss": 0.352, + "num_input_tokens_seen": 15839040, + "step": 7575 + }, + { + "epoch": 1.236560894037034, + "grad_norm": 0.0011355975875630975, + "learning_rate": 0.274200976999108, + "loss": 0.3227, + "num_input_tokens_seen": 15849296, + "step": 7580 + }, + { + "epoch": 1.2373766212578514, + "grad_norm": 0.0008732696296647191, + "learning_rate": 0.27416793837844916, + "loss": 0.3887, + "num_input_tokens_seen": 15859424, + "step": 7585 + }, + { + "epoch": 1.2381923484786688, + "grad_norm": 0.0014326974051073194, + "learning_rate": 0.27413488060953384, + "loss": 0.3515, + "num_input_tokens_seen": 15870864, + "step": 7590 + }, + { + "epoch": 1.2390080756994861, + "grad_norm": 0.0006931500392965972, + "learning_rate": 0.27410180369745996, + "loss": 0.3431, + "num_input_tokens_seen": 15881056, + "step": 7595 + }, + { + "epoch": 1.2398238029203035, + "grad_norm": 0.0006422604201361537, + "learning_rate": 0.27406870764732844, + "loss": 0.3416, + "num_input_tokens_seen": 15891152, + "step": 7600 + }, + { + "epoch": 1.2398238029203035, + "eval_loss": 0.3422994315624237, + "eval_runtime": 155.6771, + "eval_samples_per_second": 17.504, + "eval_steps_per_second": 8.755, + "num_input_tokens_seen": 15891152, + "step": 7600 + }, + { + "epoch": 1.2406395301411208, + "grad_norm": 0.0005827855202369392, + "learning_rate": 0.27403559246424297, + "loss": 0.3381, + "num_input_tokens_seen": 15901296, + "step": 7605 + }, + { + "epoch": 1.2414552573619382, + "grad_norm": 0.00032174590160138905, + "learning_rate": 0.2740024581533105, + "loss": 0.3395, + "num_input_tokens_seen": 15911280, + "step": 7610 + }, + { + "epoch": 1.2422709845827555, + "grad_norm": 0.0009327312000095844, + "learning_rate": 0.2739693047196406, + "loss": 0.3764, + "num_input_tokens_seen": 15920128, + "step": 7615 + }, + { + "epoch": 1.2430867118035729, + "grad_norm": 0.0005198067519813776, + "learning_rate": 0.27393613216834606, + "loss": 0.3417, + "num_input_tokens_seen": 15930768, + "step": 7620 + }, + { + "epoch": 1.2439024390243902, + "grad_norm": 0.0005925625446252525, + "learning_rate": 0.2739029405045424, + "loss": 0.3466, + "num_input_tokens_seen": 15940592, + "step": 7625 + }, + { + "epoch": 1.2447181662452076, + "grad_norm": 0.0008532751235179603, + "learning_rate": 0.2738697297333483, + "loss": 0.3413, + "num_input_tokens_seen": 15950160, + "step": 7630 + }, + { + "epoch": 1.245533893466025, + "grad_norm": 0.0006014998070895672, + "learning_rate": 0.2738364998598852, + "loss": 0.3528, + "num_input_tokens_seen": 15960880, + "step": 7635 + }, + { + "epoch": 1.2463496206868423, + "grad_norm": 0.0012540887109935284, + "learning_rate": 0.27380325088927765, + "loss": 0.347, + "num_input_tokens_seen": 15972064, + "step": 7640 + }, + { + "epoch": 1.2471653479076596, + "grad_norm": 0.0013087834231555462, + "learning_rate": 0.27376998282665294, + "loss": 0.3461, + "num_input_tokens_seen": 15983184, + "step": 7645 + }, + { + "epoch": 1.247981075128477, + "grad_norm": 0.00028711732011288404, + "learning_rate": 0.27373669567714154, + "loss": 0.3447, + "num_input_tokens_seen": 15994624, + "step": 7650 + }, + { + "epoch": 1.2487968023492944, + "grad_norm": 0.0008724182844161987, + "learning_rate": 0.27370338944587663, + "loss": 0.3742, + "num_input_tokens_seen": 16004336, + "step": 7655 + }, + { + "epoch": 1.2496125295701117, + "grad_norm": 0.0005514242220669985, + "learning_rate": 0.27367006413799455, + "loss": 0.34, + "num_input_tokens_seen": 16014400, + "step": 7660 + }, + { + "epoch": 1.250428256790929, + "grad_norm": 0.0004951076116412878, + "learning_rate": 0.2736367197586345, + "loss": 0.3202, + "num_input_tokens_seen": 16024720, + "step": 7665 + }, + { + "epoch": 1.2512439840117464, + "grad_norm": 0.0005530879134312272, + "learning_rate": 0.2736033563129385, + "loss": 0.3496, + "num_input_tokens_seen": 16036624, + "step": 7670 + }, + { + "epoch": 1.2520597112325638, + "grad_norm": 0.00044917193008586764, + "learning_rate": 0.27356997380605164, + "loss": 0.3262, + "num_input_tokens_seen": 16046240, + "step": 7675 + }, + { + "epoch": 1.2528754384533811, + "grad_norm": 0.00041269956273026764, + "learning_rate": 0.27353657224312194, + "loss": 0.3215, + "num_input_tokens_seen": 16057232, + "step": 7680 + }, + { + "epoch": 1.2536911656741985, + "grad_norm": 0.0009986162185668945, + "learning_rate": 0.2735031516293004, + "loss": 0.3333, + "num_input_tokens_seen": 16067296, + "step": 7685 + }, + { + "epoch": 1.2545068928950158, + "grad_norm": 0.0004358515434432775, + "learning_rate": 0.2734697119697408, + "loss": 0.3262, + "num_input_tokens_seen": 16079312, + "step": 7690 + }, + { + "epoch": 1.2553226201158334, + "grad_norm": 0.0017700174357742071, + "learning_rate": 0.27343625326959997, + "loss": 0.3836, + "num_input_tokens_seen": 16089776, + "step": 7695 + }, + { + "epoch": 1.2561383473366505, + "grad_norm": 0.000480107992189005, + "learning_rate": 0.27340277553403775, + "loss": 0.3408, + "num_input_tokens_seen": 16100544, + "step": 7700 + }, + { + "epoch": 1.256954074557468, + "grad_norm": 0.0002914616488851607, + "learning_rate": 0.2733692787682167, + "loss": 0.3418, + "num_input_tokens_seen": 16111424, + "step": 7705 + }, + { + "epoch": 1.2577698017782852, + "grad_norm": 0.0010857428424060345, + "learning_rate": 0.27333576297730255, + "loss": 0.307, + "num_input_tokens_seen": 16121056, + "step": 7710 + }, + { + "epoch": 1.2585855289991028, + "grad_norm": 0.0009951952379196882, + "learning_rate": 0.2733022281664638, + "loss": 0.3389, + "num_input_tokens_seen": 16130304, + "step": 7715 + }, + { + "epoch": 1.25940125621992, + "grad_norm": 0.00047141406685113907, + "learning_rate": 0.273268674340872, + "loss": 0.3213, + "num_input_tokens_seen": 16141936, + "step": 7720 + }, + { + "epoch": 1.2602169834407375, + "grad_norm": 0.0003814203373622149, + "learning_rate": 0.27323510150570146, + "loss": 0.3347, + "num_input_tokens_seen": 16151664, + "step": 7725 + }, + { + "epoch": 1.2610327106615546, + "grad_norm": 0.00037911938852630556, + "learning_rate": 0.27320150966612966, + "loss": 0.3727, + "num_input_tokens_seen": 16161936, + "step": 7730 + }, + { + "epoch": 1.2618484378823722, + "grad_norm": 0.00033001185511238873, + "learning_rate": 0.2731678988273368, + "loss": 0.3248, + "num_input_tokens_seen": 16172944, + "step": 7735 + }, + { + "epoch": 1.2626641651031896, + "grad_norm": 0.0008686059154570103, + "learning_rate": 0.27313426899450605, + "loss": 0.3646, + "num_input_tokens_seen": 16183136, + "step": 7740 + }, + { + "epoch": 1.263479892324007, + "grad_norm": 0.00027395537472330034, + "learning_rate": 0.27310062017282366, + "loss": 0.3456, + "num_input_tokens_seen": 16194592, + "step": 7745 + }, + { + "epoch": 1.2642956195448243, + "grad_norm": 0.00025680073304101825, + "learning_rate": 0.2730669523674787, + "loss": 0.3513, + "num_input_tokens_seen": 16203296, + "step": 7750 + }, + { + "epoch": 1.2651113467656416, + "grad_norm": 0.0012241012882441282, + "learning_rate": 0.2730332655836631, + "loss": 0.3413, + "num_input_tokens_seen": 16214096, + "step": 7755 + }, + { + "epoch": 1.265927073986459, + "grad_norm": 0.0006824050797149539, + "learning_rate": 0.2729995598265718, + "loss": 0.3432, + "num_input_tokens_seen": 16224400, + "step": 7760 + }, + { + "epoch": 1.2667428012072763, + "grad_norm": 0.00028741269488818944, + "learning_rate": 0.2729658351014027, + "loss": 0.3543, + "num_input_tokens_seen": 16235152, + "step": 7765 + }, + { + "epoch": 1.2675585284280937, + "grad_norm": 0.00048420947859995067, + "learning_rate": 0.27293209141335656, + "loss": 0.3383, + "num_input_tokens_seen": 16245552, + "step": 7770 + }, + { + "epoch": 1.268374255648911, + "grad_norm": 0.0007975480402819812, + "learning_rate": 0.27289832876763703, + "loss": 0.3672, + "num_input_tokens_seen": 16258272, + "step": 7775 + }, + { + "epoch": 1.2691899828697284, + "grad_norm": 0.00024961851886473596, + "learning_rate": 0.27286454716945074, + "loss": 0.3503, + "num_input_tokens_seen": 16268096, + "step": 7780 + }, + { + "epoch": 1.2700057100905457, + "grad_norm": 0.0006218688795343041, + "learning_rate": 0.27283074662400725, + "loss": 0.3524, + "num_input_tokens_seen": 16278752, + "step": 7785 + }, + { + "epoch": 1.270821437311363, + "grad_norm": 0.0011463436530902982, + "learning_rate": 0.2727969271365191, + "loss": 0.3375, + "num_input_tokens_seen": 16289328, + "step": 7790 + }, + { + "epoch": 1.2716371645321805, + "grad_norm": 0.0011193787213414907, + "learning_rate": 0.2727630887122016, + "loss": 0.3502, + "num_input_tokens_seen": 16299248, + "step": 7795 + }, + { + "epoch": 1.2724528917529978, + "grad_norm": 0.00029947044095024467, + "learning_rate": 0.27272923135627314, + "loss": 0.3428, + "num_input_tokens_seen": 16309376, + "step": 7800 + }, + { + "epoch": 1.2724528917529978, + "eval_loss": 0.33882591128349304, + "eval_runtime": 155.7589, + "eval_samples_per_second": 17.495, + "eval_steps_per_second": 8.751, + "num_input_tokens_seen": 16309376, + "step": 7800 + }, + { + "epoch": 1.2732686189738152, + "grad_norm": 0.0009026144980452955, + "learning_rate": 0.2726953550739548, + "loss": 0.3418, + "num_input_tokens_seen": 16319664, + "step": 7805 + }, + { + "epoch": 1.2740843461946325, + "grad_norm": 0.0006411803187802434, + "learning_rate": 0.27266145987047086, + "loss": 0.3368, + "num_input_tokens_seen": 16329776, + "step": 7810 + }, + { + "epoch": 1.2749000734154499, + "grad_norm": 0.0003389654739294201, + "learning_rate": 0.27262754575104836, + "loss": 0.3338, + "num_input_tokens_seen": 16339520, + "step": 7815 + }, + { + "epoch": 1.2757158006362672, + "grad_norm": 0.00044611821067519486, + "learning_rate": 0.27259361272091726, + "loss": 0.3419, + "num_input_tokens_seen": 16351040, + "step": 7820 + }, + { + "epoch": 1.2765315278570846, + "grad_norm": 0.0009084423654712737, + "learning_rate": 0.27255966078531046, + "loss": 0.3795, + "num_input_tokens_seen": 16360848, + "step": 7825 + }, + { + "epoch": 1.277347255077902, + "grad_norm": 0.00032267768983729184, + "learning_rate": 0.2725256899494638, + "loss": 0.3377, + "num_input_tokens_seen": 16372256, + "step": 7830 + }, + { + "epoch": 1.2781629822987193, + "grad_norm": 0.0005781434010714293, + "learning_rate": 0.272491700218616, + "loss": 0.3398, + "num_input_tokens_seen": 16384016, + "step": 7835 + }, + { + "epoch": 1.2789787095195366, + "grad_norm": 0.0008074213401414454, + "learning_rate": 0.27245769159800876, + "loss": 0.3347, + "num_input_tokens_seen": 16395104, + "step": 7840 + }, + { + "epoch": 1.279794436740354, + "grad_norm": 0.0008375453180633485, + "learning_rate": 0.2724236640928865, + "loss": 0.3239, + "num_input_tokens_seen": 16405280, + "step": 7845 + }, + { + "epoch": 1.2806101639611713, + "grad_norm": 0.00044305797200649977, + "learning_rate": 0.27238961770849673, + "loss": 0.3582, + "num_input_tokens_seen": 16415264, + "step": 7850 + }, + { + "epoch": 1.2814258911819887, + "grad_norm": 0.0008507455931976438, + "learning_rate": 0.27235555245008997, + "loss": 0.3314, + "num_input_tokens_seen": 16424928, + "step": 7855 + }, + { + "epoch": 1.282241618402806, + "grad_norm": 0.0009655548492446542, + "learning_rate": 0.2723214683229193, + "loss": 0.3478, + "num_input_tokens_seen": 16434864, + "step": 7860 + }, + { + "epoch": 1.2830573456236234, + "grad_norm": 0.0005146359326317906, + "learning_rate": 0.27228736533224107, + "loss": 0.353, + "num_input_tokens_seen": 16445584, + "step": 7865 + }, + { + "epoch": 1.2838730728444407, + "grad_norm": 0.0007352399989031255, + "learning_rate": 0.27225324348331437, + "loss": 0.3658, + "num_input_tokens_seen": 16455440, + "step": 7870 + }, + { + "epoch": 1.284688800065258, + "grad_norm": 0.0007306681363843381, + "learning_rate": 0.27221910278140116, + "loss": 0.3446, + "num_input_tokens_seen": 16465024, + "step": 7875 + }, + { + "epoch": 1.2855045272860757, + "grad_norm": 0.0008076491649262607, + "learning_rate": 0.2721849432317664, + "loss": 0.3516, + "num_input_tokens_seen": 16476128, + "step": 7880 + }, + { + "epoch": 1.2863202545068928, + "grad_norm": 0.001092958147637546, + "learning_rate": 0.2721507648396779, + "loss": 0.3549, + "num_input_tokens_seen": 16486704, + "step": 7885 + }, + { + "epoch": 1.2871359817277104, + "grad_norm": 0.0015915234107524157, + "learning_rate": 0.27211656761040653, + "loss": 0.3452, + "num_input_tokens_seen": 16497984, + "step": 7890 + }, + { + "epoch": 1.2879517089485275, + "grad_norm": 0.0008848176803439856, + "learning_rate": 0.2720823515492257, + "loss": 0.3516, + "num_input_tokens_seen": 16508848, + "step": 7895 + }, + { + "epoch": 1.288767436169345, + "grad_norm": 0.005277757532894611, + "learning_rate": 0.27204811666141215, + "loss": 0.3428, + "num_input_tokens_seen": 16519008, + "step": 7900 + }, + { + "epoch": 1.2895831633901622, + "grad_norm": 0.00042640409083105624, + "learning_rate": 0.2720138629522452, + "loss": 0.3454, + "num_input_tokens_seen": 16529120, + "step": 7905 + }, + { + "epoch": 1.2903988906109798, + "grad_norm": 0.20454703271389008, + "learning_rate": 0.2719795904270073, + "loss": 0.9576, + "num_input_tokens_seen": 16540032, + "step": 7910 + }, + { + "epoch": 1.291214617831797, + "grad_norm": 0.0012600607005879283, + "learning_rate": 0.2719452990909837, + "loss": 0.4672, + "num_input_tokens_seen": 16551664, + "step": 7915 + }, + { + "epoch": 1.2920303450526145, + "grad_norm": 0.0008644591434858739, + "learning_rate": 0.2719109889494625, + "loss": 0.338, + "num_input_tokens_seen": 16561856, + "step": 7920 + }, + { + "epoch": 1.2928460722734318, + "grad_norm": 0.0010027489624917507, + "learning_rate": 0.27187666000773475, + "loss": 0.3407, + "num_input_tokens_seen": 16572688, + "step": 7925 + }, + { + "epoch": 1.2936617994942492, + "grad_norm": 0.0005785388639196754, + "learning_rate": 0.2718423122710944, + "loss": 0.342, + "num_input_tokens_seen": 16583024, + "step": 7930 + }, + { + "epoch": 1.2944775267150666, + "grad_norm": 0.0014703483320772648, + "learning_rate": 0.2718079457448384, + "loss": 0.3202, + "num_input_tokens_seen": 16593936, + "step": 7935 + }, + { + "epoch": 1.295293253935884, + "grad_norm": 0.0015400657430291176, + "learning_rate": 0.27177356043426637, + "loss": 0.3513, + "num_input_tokens_seen": 16604016, + "step": 7940 + }, + { + "epoch": 1.2961089811567013, + "grad_norm": 0.000558488885872066, + "learning_rate": 0.27173915634468104, + "loss": 0.3489, + "num_input_tokens_seen": 16613744, + "step": 7945 + }, + { + "epoch": 1.2969247083775186, + "grad_norm": 0.0007427738164551556, + "learning_rate": 0.27170473348138796, + "loss": 0.3433, + "num_input_tokens_seen": 16625168, + "step": 7950 + }, + { + "epoch": 1.297740435598336, + "grad_norm": 0.0014876312343403697, + "learning_rate": 0.27167029184969554, + "loss": 0.333, + "num_input_tokens_seen": 16635664, + "step": 7955 + }, + { + "epoch": 1.2985561628191533, + "grad_norm": 0.001121046021580696, + "learning_rate": 0.27163583145491504, + "loss": 0.3526, + "num_input_tokens_seen": 16646736, + "step": 7960 + }, + { + "epoch": 1.2993718900399707, + "grad_norm": 0.0013750421348959208, + "learning_rate": 0.2716013523023608, + "loss": 0.3807, + "num_input_tokens_seen": 16658432, + "step": 7965 + }, + { + "epoch": 1.300187617260788, + "grad_norm": 0.0009318109368905425, + "learning_rate": 0.27156685439734995, + "loss": 0.3438, + "num_input_tokens_seen": 16668608, + "step": 7970 + }, + { + "epoch": 1.3010033444816054, + "grad_norm": 0.0011291721602901816, + "learning_rate": 0.2715323377452024, + "loss": 0.3336, + "num_input_tokens_seen": 16678224, + "step": 7975 + }, + { + "epoch": 1.3018190717024227, + "grad_norm": 0.0004831278638448566, + "learning_rate": 0.2714978023512411, + "loss": 0.3315, + "num_input_tokens_seen": 16688960, + "step": 7980 + }, + { + "epoch": 1.30263479892324, + "grad_norm": 0.0012109113158658147, + "learning_rate": 0.2714632482207918, + "loss": 0.3339, + "num_input_tokens_seen": 16699552, + "step": 7985 + }, + { + "epoch": 1.3034505261440574, + "grad_norm": 0.0011960617266595364, + "learning_rate": 0.2714286753591833, + "loss": 0.3314, + "num_input_tokens_seen": 16710720, + "step": 7990 + }, + { + "epoch": 1.3042662533648748, + "grad_norm": 0.000532337639015168, + "learning_rate": 0.27139408377174706, + "loss": 0.3501, + "num_input_tokens_seen": 16719968, + "step": 7995 + }, + { + "epoch": 1.3050819805856921, + "grad_norm": 0.0013678708346560597, + "learning_rate": 0.27135947346381756, + "loss": 0.3515, + "num_input_tokens_seen": 16729632, + "step": 8000 + }, + { + "epoch": 1.3050819805856921, + "eval_loss": 0.33501431345939636, + "eval_runtime": 155.5812, + "eval_samples_per_second": 17.515, + "eval_steps_per_second": 8.761, + "num_input_tokens_seen": 16729632, + "step": 8000 + }, + { + "epoch": 1.3058977078065095, + "grad_norm": 0.0005150751676410437, + "learning_rate": 0.2713248444407322, + "loss": 0.3471, + "num_input_tokens_seen": 16740000, + "step": 8005 + }, + { + "epoch": 1.3067134350273268, + "grad_norm": 0.0011403763201087713, + "learning_rate": 0.27129019670783106, + "loss": 0.3303, + "num_input_tokens_seen": 16748928, + "step": 8010 + }, + { + "epoch": 1.3075291622481442, + "grad_norm": 0.0006483474862761796, + "learning_rate": 0.27125553027045746, + "loss": 0.3387, + "num_input_tokens_seen": 16759728, + "step": 8015 + }, + { + "epoch": 1.3083448894689615, + "grad_norm": 0.0008542697760276496, + "learning_rate": 0.2712208451339572, + "loss": 0.3239, + "num_input_tokens_seen": 16770672, + "step": 8020 + }, + { + "epoch": 1.309160616689779, + "grad_norm": 0.0016876126173883677, + "learning_rate": 0.27118614130367935, + "loss": 0.3681, + "num_input_tokens_seen": 16781008, + "step": 8025 + }, + { + "epoch": 1.3099763439105963, + "grad_norm": 0.0027847937308251858, + "learning_rate": 0.2711514187849756, + "loss": 0.3353, + "num_input_tokens_seen": 16791008, + "step": 8030 + }, + { + "epoch": 1.3107920711314136, + "grad_norm": 0.001964820083230734, + "learning_rate": 0.27111667758320057, + "loss": 0.3413, + "num_input_tokens_seen": 16800896, + "step": 8035 + }, + { + "epoch": 1.311607798352231, + "grad_norm": 0.0013208871241658926, + "learning_rate": 0.27108191770371176, + "loss": 0.3329, + "num_input_tokens_seen": 16809072, + "step": 8040 + }, + { + "epoch": 1.3124235255730483, + "grad_norm": 0.0012912011006847024, + "learning_rate": 0.2710471391518697, + "loss": 0.3443, + "num_input_tokens_seen": 16818976, + "step": 8045 + }, + { + "epoch": 1.3132392527938657, + "grad_norm": 0.0008539034170098603, + "learning_rate": 0.2710123419330375, + "loss": 0.3019, + "num_input_tokens_seen": 16828176, + "step": 8050 + }, + { + "epoch": 1.314054980014683, + "grad_norm": 0.0007449094555340707, + "learning_rate": 0.2709775260525816, + "loss": 0.3786, + "num_input_tokens_seen": 16837680, + "step": 8055 + }, + { + "epoch": 1.3148707072355004, + "grad_norm": 0.0007740448927506804, + "learning_rate": 0.27094269151587075, + "loss": 0.3674, + "num_input_tokens_seen": 16847344, + "step": 8060 + }, + { + "epoch": 1.315686434456318, + "grad_norm": 0.00037260071258060634, + "learning_rate": 0.27090783832827703, + "loss": 0.3155, + "num_input_tokens_seen": 16858832, + "step": 8065 + }, + { + "epoch": 1.316502161677135, + "grad_norm": 0.0006035536644048989, + "learning_rate": 0.2708729664951753, + "loss": 0.3304, + "num_input_tokens_seen": 16868672, + "step": 8070 + }, + { + "epoch": 1.3173178888979526, + "grad_norm": 0.000582944368943572, + "learning_rate": 0.27083807602194304, + "loss": 0.293, + "num_input_tokens_seen": 16879392, + "step": 8075 + }, + { + "epoch": 1.3181336161187698, + "grad_norm": 0.0009736604406498373, + "learning_rate": 0.270803166913961, + "loss": 0.3543, + "num_input_tokens_seen": 16889568, + "step": 8080 + }, + { + "epoch": 1.3189493433395874, + "grad_norm": 0.0011609174543991685, + "learning_rate": 0.27076823917661247, + "loss": 0.3149, + "num_input_tokens_seen": 16899392, + "step": 8085 + }, + { + "epoch": 1.3197650705604045, + "grad_norm": 0.0013106701662763953, + "learning_rate": 0.2707332928152838, + "loss": 0.344, + "num_input_tokens_seen": 16910448, + "step": 8090 + }, + { + "epoch": 1.320580797781222, + "grad_norm": 0.0012587460223585367, + "learning_rate": 0.2706983278353641, + "loss": 0.3517, + "num_input_tokens_seen": 16919808, + "step": 8095 + }, + { + "epoch": 1.3213965250020392, + "grad_norm": 0.000753775704652071, + "learning_rate": 0.27066334424224553, + "loss": 0.3669, + "num_input_tokens_seen": 16930288, + "step": 8100 + }, + { + "epoch": 1.3222122522228568, + "grad_norm": 0.0009419381385669112, + "learning_rate": 0.27062834204132297, + "loss": 0.3306, + "num_input_tokens_seen": 16941200, + "step": 8105 + }, + { + "epoch": 1.3230279794436741, + "grad_norm": 0.0005944916047155857, + "learning_rate": 0.27059332123799407, + "loss": 0.3355, + "num_input_tokens_seen": 16950624, + "step": 8110 + }, + { + "epoch": 1.3238437066644915, + "grad_norm": 0.0014406164409592748, + "learning_rate": 0.27055828183765956, + "loss": 0.3612, + "num_input_tokens_seen": 16960960, + "step": 8115 + }, + { + "epoch": 1.3246594338853088, + "grad_norm": 0.0011270293034613132, + "learning_rate": 0.270523223845723, + "loss": 0.3332, + "num_input_tokens_seen": 16970224, + "step": 8120 + }, + { + "epoch": 1.3254751611061262, + "grad_norm": 0.000620954146143049, + "learning_rate": 0.2704881472675907, + "loss": 0.3953, + "num_input_tokens_seen": 16981600, + "step": 8125 + }, + { + "epoch": 1.3262908883269435, + "grad_norm": 0.0006132387788966298, + "learning_rate": 0.270453052108672, + "loss": 0.328, + "num_input_tokens_seen": 16991264, + "step": 8130 + }, + { + "epoch": 1.3271066155477609, + "grad_norm": 0.001666879397816956, + "learning_rate": 0.2704179383743789, + "loss": 0.3232, + "num_input_tokens_seen": 17003248, + "step": 8135 + }, + { + "epoch": 1.3279223427685782, + "grad_norm": 0.00140784727409482, + "learning_rate": 0.27038280607012644, + "loss": 0.3668, + "num_input_tokens_seen": 17014016, + "step": 8140 + }, + { + "epoch": 1.3287380699893956, + "grad_norm": 0.0007356796413660049, + "learning_rate": 0.27034765520133247, + "loss": 0.3458, + "num_input_tokens_seen": 17023840, + "step": 8145 + }, + { + "epoch": 1.329553797210213, + "grad_norm": 0.0011370661668479443, + "learning_rate": 0.2703124857734177, + "loss": 0.3548, + "num_input_tokens_seen": 17034176, + "step": 8150 + }, + { + "epoch": 1.3303695244310303, + "grad_norm": 0.001472663483582437, + "learning_rate": 0.27027729779180565, + "loss": 0.3219, + "num_input_tokens_seen": 17045424, + "step": 8155 + }, + { + "epoch": 1.3311852516518476, + "grad_norm": 0.0011588030029088259, + "learning_rate": 0.27024209126192283, + "loss": 0.3322, + "num_input_tokens_seen": 17056080, + "step": 8160 + }, + { + "epoch": 1.332000978872665, + "grad_norm": 0.0007446422823704779, + "learning_rate": 0.2702068661891984, + "loss": 0.325, + "num_input_tokens_seen": 17065776, + "step": 8165 + }, + { + "epoch": 1.3328167060934824, + "grad_norm": 0.0008699473110027611, + "learning_rate": 0.2701716225790647, + "loss": 0.3692, + "num_input_tokens_seen": 17076320, + "step": 8170 + }, + { + "epoch": 1.3336324333142997, + "grad_norm": 0.0005989583441987634, + "learning_rate": 0.27013636043695655, + "loss": 0.3391, + "num_input_tokens_seen": 17086304, + "step": 8175 + }, + { + "epoch": 1.334448160535117, + "grad_norm": 0.00176260678563267, + "learning_rate": 0.27010107976831194, + "loss": 0.3446, + "num_input_tokens_seen": 17096976, + "step": 8180 + }, + { + "epoch": 1.3352638877559344, + "grad_norm": 0.001982561545446515, + "learning_rate": 0.2700657805785715, + "loss": 0.3355, + "num_input_tokens_seen": 17108064, + "step": 8185 + }, + { + "epoch": 1.3360796149767518, + "grad_norm": 0.0020071184262633324, + "learning_rate": 0.2700304628731789, + "loss": 0.322, + "num_input_tokens_seen": 17118912, + "step": 8190 + }, + { + "epoch": 1.3368953421975691, + "grad_norm": 0.001459736842662096, + "learning_rate": 0.26999512665758046, + "loss": 0.3366, + "num_input_tokens_seen": 17129168, + "step": 8195 + }, + { + "epoch": 1.3377110694183865, + "grad_norm": 0.25722870230674744, + "learning_rate": 0.2699597719372256, + "loss": 0.4186, + "num_input_tokens_seen": 17139952, + "step": 8200 + }, + { + "epoch": 1.3377110694183865, + "eval_loss": 0.3455325663089752, + "eval_runtime": 156.0357, + "eval_samples_per_second": 17.464, + "eval_steps_per_second": 8.735, + "num_input_tokens_seen": 17139952, + "step": 8200 + }, + { + "epoch": 1.3385267966392038, + "grad_norm": 0.003148572752252221, + "learning_rate": 0.26992439871756635, + "loss": 0.3529, + "num_input_tokens_seen": 17149664, + "step": 8205 + }, + { + "epoch": 1.3393425238600212, + "grad_norm": 0.003440799657255411, + "learning_rate": 0.2698890070040578, + "loss": 0.346, + "num_input_tokens_seen": 17159040, + "step": 8210 + }, + { + "epoch": 1.3401582510808385, + "grad_norm": 0.002062548417598009, + "learning_rate": 0.2698535968021577, + "loss": 0.4074, + "num_input_tokens_seen": 17169360, + "step": 8215 + }, + { + "epoch": 1.3409739783016559, + "grad_norm": 0.0014075727667659521, + "learning_rate": 0.26981816811732684, + "loss": 0.3186, + "num_input_tokens_seen": 17181920, + "step": 8220 + }, + { + "epoch": 1.3417897055224732, + "grad_norm": 0.0016093463636934757, + "learning_rate": 0.26978272095502875, + "loss": 0.3938, + "num_input_tokens_seen": 17191600, + "step": 8225 + }, + { + "epoch": 1.3426054327432906, + "grad_norm": 0.0016663532005622983, + "learning_rate": 0.26974725532072974, + "loss": 0.3267, + "num_input_tokens_seen": 17201776, + "step": 8230 + }, + { + "epoch": 1.343421159964108, + "grad_norm": 0.0015454341191798449, + "learning_rate": 0.26971177121989914, + "loss": 0.3052, + "num_input_tokens_seen": 17212864, + "step": 8235 + }, + { + "epoch": 1.3442368871849253, + "grad_norm": 0.0006132564158178866, + "learning_rate": 0.2696762686580091, + "loss": 0.3779, + "num_input_tokens_seen": 17223888, + "step": 8240 + }, + { + "epoch": 1.3450526144057426, + "grad_norm": 0.0012949283700436354, + "learning_rate": 0.26964074764053436, + "loss": 0.3962, + "num_input_tokens_seen": 17234064, + "step": 8245 + }, + { + "epoch": 1.3458683416265602, + "grad_norm": 0.001418858882971108, + "learning_rate": 0.2696052081729529, + "loss": 0.3455, + "num_input_tokens_seen": 17243296, + "step": 8250 + }, + { + "epoch": 1.3466840688473773, + "grad_norm": 0.0011550962226465344, + "learning_rate": 0.2695696502607453, + "loss": 0.353, + "num_input_tokens_seen": 17253520, + "step": 8255 + }, + { + "epoch": 1.347499796068195, + "grad_norm": 0.0006165150552988052, + "learning_rate": 0.26953407390939504, + "loss": 0.3298, + "num_input_tokens_seen": 17264208, + "step": 8260 + }, + { + "epoch": 1.348315523289012, + "grad_norm": 0.001485359505750239, + "learning_rate": 0.26949847912438835, + "loss": 0.3694, + "num_input_tokens_seen": 17273296, + "step": 8265 + }, + { + "epoch": 1.3491312505098296, + "grad_norm": 0.0006427565240301192, + "learning_rate": 0.26946286591121454, + "loss": 0.3449, + "num_input_tokens_seen": 17284848, + "step": 8270 + }, + { + "epoch": 1.3499469777306468, + "grad_norm": 0.001587472390383482, + "learning_rate": 0.2694272342753655, + "loss": 0.3526, + "num_input_tokens_seen": 17296592, + "step": 8275 + }, + { + "epoch": 1.3507627049514643, + "grad_norm": 0.000961434852797538, + "learning_rate": 0.26939158422233617, + "loss": 0.3147, + "num_input_tokens_seen": 17307264, + "step": 8280 + }, + { + "epoch": 1.3515784321722815, + "grad_norm": 0.001044528093189001, + "learning_rate": 0.26935591575762413, + "loss": 0.3274, + "num_input_tokens_seen": 17317040, + "step": 8285 + }, + { + "epoch": 1.352394159393099, + "grad_norm": 0.001773353200405836, + "learning_rate": 0.26932022888672996, + "loss": 0.3831, + "num_input_tokens_seen": 17327072, + "step": 8290 + }, + { + "epoch": 1.3532098866139162, + "grad_norm": 0.0015599518083035946, + "learning_rate": 0.26928452361515703, + "loss": 0.3585, + "num_input_tokens_seen": 17337360, + "step": 8295 + }, + { + "epoch": 1.3540256138347337, + "grad_norm": 0.0013695117086172104, + "learning_rate": 0.26924879994841155, + "loss": 0.3192, + "num_input_tokens_seen": 17347232, + "step": 8300 + }, + { + "epoch": 1.354841341055551, + "grad_norm": 0.0009407741599716246, + "learning_rate": 0.2692130578920025, + "loss": 0.3336, + "num_input_tokens_seen": 17357424, + "step": 8305 + }, + { + "epoch": 1.3556570682763684, + "grad_norm": 0.004019458312541246, + "learning_rate": 0.26917729745144187, + "loss": 0.3003, + "num_input_tokens_seen": 17367408, + "step": 8310 + }, + { + "epoch": 1.3564727954971858, + "grad_norm": 0.0017520523397251964, + "learning_rate": 0.2691415186322443, + "loss": 0.3558, + "num_input_tokens_seen": 17378480, + "step": 8315 + }, + { + "epoch": 1.3572885227180032, + "grad_norm": 0.0009585568914189935, + "learning_rate": 0.2691057214399273, + "loss": 0.3235, + "num_input_tokens_seen": 17388784, + "step": 8320 + }, + { + "epoch": 1.3581042499388205, + "grad_norm": 0.000550182769075036, + "learning_rate": 0.2690699058800113, + "loss": 0.3426, + "num_input_tokens_seen": 17399552, + "step": 8325 + }, + { + "epoch": 1.3589199771596379, + "grad_norm": 0.0008135483367368579, + "learning_rate": 0.2690340719580194, + "loss": 0.3267, + "num_input_tokens_seen": 17411168, + "step": 8330 + }, + { + "epoch": 1.3597357043804552, + "grad_norm": 0.0014754504663869739, + "learning_rate": 0.2689982196794778, + "loss": 0.3371, + "num_input_tokens_seen": 17422224, + "step": 8335 + }, + { + "epoch": 1.3605514316012726, + "grad_norm": 0.0005276502924971282, + "learning_rate": 0.2689623490499153, + "loss": 0.3612, + "num_input_tokens_seen": 17433792, + "step": 8340 + }, + { + "epoch": 1.36136715882209, + "grad_norm": 0.0004006499657407403, + "learning_rate": 0.2689264600748636, + "loss": 0.36, + "num_input_tokens_seen": 17444688, + "step": 8345 + }, + { + "epoch": 1.3621828860429073, + "grad_norm": 0.0008935976657085121, + "learning_rate": 0.26889055275985724, + "loss": 0.3082, + "num_input_tokens_seen": 17454720, + "step": 8350 + }, + { + "epoch": 1.3629986132637246, + "grad_norm": 0.001325985067524016, + "learning_rate": 0.2688546271104335, + "loss": 0.3452, + "num_input_tokens_seen": 17463424, + "step": 8355 + }, + { + "epoch": 1.363814340484542, + "grad_norm": 0.0006125373183749616, + "learning_rate": 0.26881868313213275, + "loss": 0.3586, + "num_input_tokens_seen": 17474416, + "step": 8360 + }, + { + "epoch": 1.3646300677053593, + "grad_norm": 0.0006132800481282175, + "learning_rate": 0.2687827208304978, + "loss": 0.3474, + "num_input_tokens_seen": 17484496, + "step": 8365 + }, + { + "epoch": 1.3654457949261767, + "grad_norm": 0.0007210352341644466, + "learning_rate": 0.26874674021107464, + "loss": 0.3243, + "num_input_tokens_seen": 17495136, + "step": 8370 + }, + { + "epoch": 1.366261522146994, + "grad_norm": 0.0011377526680007577, + "learning_rate": 0.2687107412794118, + "loss": 0.3567, + "num_input_tokens_seen": 17504848, + "step": 8375 + }, + { + "epoch": 1.3670772493678114, + "grad_norm": 0.0006316872895695269, + "learning_rate": 0.26867472404106096, + "loss": 0.3494, + "num_input_tokens_seen": 17515072, + "step": 8380 + }, + { + "epoch": 1.3678929765886287, + "grad_norm": 0.00048323321971111, + "learning_rate": 0.26863868850157624, + "loss": 0.3095, + "num_input_tokens_seen": 17524672, + "step": 8385 + }, + { + "epoch": 1.368708703809446, + "grad_norm": 0.0012682478409260511, + "learning_rate": 0.26860263466651485, + "loss": 0.3211, + "num_input_tokens_seen": 17534976, + "step": 8390 + }, + { + "epoch": 1.3695244310302634, + "grad_norm": 0.001040998031385243, + "learning_rate": 0.26856656254143674, + "loss": 0.3494, + "num_input_tokens_seen": 17545248, + "step": 8395 + }, + { + "epoch": 1.3703401582510808, + "grad_norm": 0.0007067966507747769, + "learning_rate": 0.2685304721319047, + "loss": 0.3117, + "num_input_tokens_seen": 17557136, + "step": 8400 + }, + { + "epoch": 1.3703401582510808, + "eval_loss": 0.3229477107524872, + "eval_runtime": 155.9864, + "eval_samples_per_second": 17.469, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 17557136, + "step": 8400 + }, + { + "epoch": 1.3711558854718982, + "grad_norm": 0.0007836545701138675, + "learning_rate": 0.2684943634434843, + "loss": 0.2633, + "num_input_tokens_seen": 17567696, + "step": 8405 + }, + { + "epoch": 1.3719716126927155, + "grad_norm": 0.0006719935336150229, + "learning_rate": 0.268458236481744, + "loss": 0.4189, + "num_input_tokens_seen": 17577008, + "step": 8410 + }, + { + "epoch": 1.3727873399135329, + "grad_norm": 0.0010625491850078106, + "learning_rate": 0.2684220912522549, + "loss": 0.3879, + "num_input_tokens_seen": 17587344, + "step": 8415 + }, + { + "epoch": 1.3736030671343502, + "grad_norm": 0.0006186693790368736, + "learning_rate": 0.2683859277605913, + "loss": 0.3249, + "num_input_tokens_seen": 17598096, + "step": 8420 + }, + { + "epoch": 1.3744187943551676, + "grad_norm": 0.0010036516468971968, + "learning_rate": 0.2683497460123298, + "loss": 0.3517, + "num_input_tokens_seen": 17609488, + "step": 8425 + }, + { + "epoch": 1.375234521575985, + "grad_norm": 0.0010888660326600075, + "learning_rate": 0.26831354601305013, + "loss": 0.3306, + "num_input_tokens_seen": 17619920, + "step": 8430 + }, + { + "epoch": 1.3760502487968023, + "grad_norm": 0.0004128953441977501, + "learning_rate": 0.26827732776833496, + "loss": 0.3432, + "num_input_tokens_seen": 17629600, + "step": 8435 + }, + { + "epoch": 1.3768659760176196, + "grad_norm": 0.0008956619421951473, + "learning_rate": 0.26824109128376944, + "loss": 0.3179, + "num_input_tokens_seen": 17640752, + "step": 8440 + }, + { + "epoch": 1.3776817032384372, + "grad_norm": 0.0014901654794812202, + "learning_rate": 0.2682048365649417, + "loss": 0.3593, + "num_input_tokens_seen": 17650432, + "step": 8445 + }, + { + "epoch": 1.3784974304592543, + "grad_norm": 0.0011219236766919494, + "learning_rate": 0.2681685636174428, + "loss": 0.3415, + "num_input_tokens_seen": 17661600, + "step": 8450 + }, + { + "epoch": 1.379313157680072, + "grad_norm": 0.0010627409210428596, + "learning_rate": 0.2681322724468663, + "loss": 0.3781, + "num_input_tokens_seen": 17671792, + "step": 8455 + }, + { + "epoch": 1.380128884900889, + "grad_norm": 0.0012287518475204706, + "learning_rate": 0.2680959630588089, + "loss": 0.3018, + "num_input_tokens_seen": 17682128, + "step": 8460 + }, + { + "epoch": 1.3809446121217066, + "grad_norm": 0.0015729517908766866, + "learning_rate": 0.26805963545886985, + "loss": 0.3805, + "num_input_tokens_seen": 17694128, + "step": 8465 + }, + { + "epoch": 1.3817603393425237, + "grad_norm": 0.0019416973227635026, + "learning_rate": 0.26802328965265143, + "loss": 0.3239, + "num_input_tokens_seen": 17704128, + "step": 8470 + }, + { + "epoch": 1.3825760665633413, + "grad_norm": 0.0010382848558947444, + "learning_rate": 0.26798692564575854, + "loss": 0.2933, + "num_input_tokens_seen": 17712960, + "step": 8475 + }, + { + "epoch": 1.3833917937841584, + "grad_norm": 0.0010573812760412693, + "learning_rate": 0.26795054344379904, + "loss": 0.3477, + "num_input_tokens_seen": 17724096, + "step": 8480 + }, + { + "epoch": 1.384207521004976, + "grad_norm": 0.0003847660555038601, + "learning_rate": 0.2679141430523835, + "loss": 0.3097, + "num_input_tokens_seen": 17735728, + "step": 8485 + }, + { + "epoch": 1.3850232482257934, + "grad_norm": 0.0010953174205496907, + "learning_rate": 0.2678777244771252, + "loss": 0.3028, + "num_input_tokens_seen": 17745968, + "step": 8490 + }, + { + "epoch": 1.3858389754466107, + "grad_norm": 0.0008264125208370388, + "learning_rate": 0.2678412877236405, + "loss": 0.2833, + "num_input_tokens_seen": 17757744, + "step": 8495 + }, + { + "epoch": 1.386654702667428, + "grad_norm": 0.0012298582587391138, + "learning_rate": 0.2678048327975484, + "loss": 0.2573, + "num_input_tokens_seen": 17767824, + "step": 8500 + }, + { + "epoch": 1.3874704298882454, + "grad_norm": 0.0012136587174609303, + "learning_rate": 0.2677683597044706, + "loss": 0.3307, + "num_input_tokens_seen": 17778128, + "step": 8505 + }, + { + "epoch": 1.3882861571090628, + "grad_norm": 0.0019383929902687669, + "learning_rate": 0.2677318684500318, + "loss": 0.3682, + "num_input_tokens_seen": 17787888, + "step": 8510 + }, + { + "epoch": 1.3891018843298801, + "grad_norm": 0.002528421813622117, + "learning_rate": 0.2676953590398593, + "loss": 0.2798, + "num_input_tokens_seen": 17799664, + "step": 8515 + }, + { + "epoch": 1.3899176115506975, + "grad_norm": 0.0013523711822926998, + "learning_rate": 0.2676588314795834, + "loss": 0.3597, + "num_input_tokens_seen": 17809360, + "step": 8520 + }, + { + "epoch": 1.3907333387715148, + "grad_norm": 0.0010268025798723102, + "learning_rate": 0.26762228577483715, + "loss": 0.3116, + "num_input_tokens_seen": 17819184, + "step": 8525 + }, + { + "epoch": 1.3915490659923322, + "grad_norm": 0.0007249803165905178, + "learning_rate": 0.2675857219312563, + "loss": 0.352, + "num_input_tokens_seen": 17830320, + "step": 8530 + }, + { + "epoch": 1.3923647932131495, + "grad_norm": 0.0006790331681258976, + "learning_rate": 0.2675491399544794, + "loss": 0.3235, + "num_input_tokens_seen": 17838880, + "step": 8535 + }, + { + "epoch": 1.393180520433967, + "grad_norm": 0.000982416793704033, + "learning_rate": 0.2675125398501479, + "loss": 0.3202, + "num_input_tokens_seen": 17848848, + "step": 8540 + }, + { + "epoch": 1.3939962476547842, + "grad_norm": 0.0009415604872629046, + "learning_rate": 0.26747592162390604, + "loss": 0.3436, + "num_input_tokens_seen": 17859792, + "step": 8545 + }, + { + "epoch": 1.3948119748756016, + "grad_norm": 0.0006833195802755654, + "learning_rate": 0.26743928528140076, + "loss": 0.3224, + "num_input_tokens_seen": 17869872, + "step": 8550 + }, + { + "epoch": 1.395627702096419, + "grad_norm": 0.0005036804359406233, + "learning_rate": 0.26740263082828186, + "loss": 0.3124, + "num_input_tokens_seen": 17881392, + "step": 8555 + }, + { + "epoch": 1.3964434293172363, + "grad_norm": 0.0007697822875343263, + "learning_rate": 0.2673659582702019, + "loss": 0.3545, + "num_input_tokens_seen": 17892064, + "step": 8560 + }, + { + "epoch": 1.3972591565380537, + "grad_norm": 0.0007358399452641606, + "learning_rate": 0.2673292676128163, + "loss": 0.377, + "num_input_tokens_seen": 17903280, + "step": 8565 + }, + { + "epoch": 1.398074883758871, + "grad_norm": 0.0011454892810434103, + "learning_rate": 0.2672925588617831, + "loss": 0.3542, + "num_input_tokens_seen": 17913584, + "step": 8570 + }, + { + "epoch": 1.3988906109796884, + "grad_norm": 0.0007568441214971244, + "learning_rate": 0.2672558320227634, + "loss": 0.3212, + "num_input_tokens_seen": 17923328, + "step": 8575 + }, + { + "epoch": 1.3997063382005057, + "grad_norm": 0.0005132111255079508, + "learning_rate": 0.2672190871014209, + "loss": 0.3845, + "num_input_tokens_seen": 17933728, + "step": 8580 + }, + { + "epoch": 1.400522065421323, + "grad_norm": 0.0011786780087277293, + "learning_rate": 0.267182324103422, + "loss": 0.3454, + "num_input_tokens_seen": 17943376, + "step": 8585 + }, + { + "epoch": 1.4013377926421404, + "grad_norm": 0.0016412176191806793, + "learning_rate": 0.2671455430344362, + "loss": 0.3238, + "num_input_tokens_seen": 17954160, + "step": 8590 + }, + { + "epoch": 1.4021535198629578, + "grad_norm": 0.0011654424015432596, + "learning_rate": 0.2671087439001355, + "loss": 0.3542, + "num_input_tokens_seen": 17964384, + "step": 8595 + }, + { + "epoch": 1.4029692470837751, + "grad_norm": 0.0009036353440023959, + "learning_rate": 0.2670719267061948, + "loss": 0.3217, + "num_input_tokens_seen": 17974864, + "step": 8600 + }, + { + "epoch": 1.4029692470837751, + "eval_loss": 0.34172576665878296, + "eval_runtime": 155.5563, + "eval_samples_per_second": 17.518, + "eval_steps_per_second": 8.762, + "num_input_tokens_seen": 17974864, + "step": 8600 + }, + { + "epoch": 1.4037849743045925, + "grad_norm": 0.0010702534345909953, + "learning_rate": 0.2670350914582918, + "loss": 0.3509, + "num_input_tokens_seen": 17983792, + "step": 8605 + }, + { + "epoch": 1.4046007015254098, + "grad_norm": 0.0009645780664868653, + "learning_rate": 0.26699823816210694, + "loss": 0.3391, + "num_input_tokens_seen": 17994688, + "step": 8610 + }, + { + "epoch": 1.4054164287462272, + "grad_norm": 0.00047198316315189004, + "learning_rate": 0.26696136682332344, + "loss": 0.3084, + "num_input_tokens_seen": 18005024, + "step": 8615 + }, + { + "epoch": 1.4062321559670445, + "grad_norm": 0.0005465932190418243, + "learning_rate": 0.2669244774476274, + "loss": 0.3582, + "num_input_tokens_seen": 18015104, + "step": 8620 + }, + { + "epoch": 1.407047883187862, + "grad_norm": 0.0006257069762796164, + "learning_rate": 0.2668875700407075, + "loss": 0.3188, + "num_input_tokens_seen": 18024976, + "step": 8625 + }, + { + "epoch": 1.4078636104086795, + "grad_norm": 0.0015107266372069716, + "learning_rate": 0.26685064460825547, + "loss": 0.3335, + "num_input_tokens_seen": 18035520, + "step": 8630 + }, + { + "epoch": 1.4086793376294966, + "grad_norm": 0.0005655224085785449, + "learning_rate": 0.26681370115596553, + "loss": 0.3694, + "num_input_tokens_seen": 18045312, + "step": 8635 + }, + { + "epoch": 1.4094950648503142, + "grad_norm": 0.0013172278413549066, + "learning_rate": 0.26677673968953497, + "loss": 0.3029, + "num_input_tokens_seen": 18055872, + "step": 8640 + }, + { + "epoch": 1.4103107920711313, + "grad_norm": 0.000520397792570293, + "learning_rate": 0.2667397602146636, + "loss": 0.3103, + "num_input_tokens_seen": 18066208, + "step": 8645 + }, + { + "epoch": 1.4111265192919489, + "grad_norm": 0.001505224034190178, + "learning_rate": 0.2667027627370542, + "loss": 0.3398, + "num_input_tokens_seen": 18075968, + "step": 8650 + }, + { + "epoch": 1.411942246512766, + "grad_norm": 0.0009627080871723592, + "learning_rate": 0.26666574726241216, + "loss": 0.3355, + "num_input_tokens_seen": 18085728, + "step": 8655 + }, + { + "epoch": 1.4127579737335836, + "grad_norm": 0.00044872486614622176, + "learning_rate": 0.2666287137964458, + "loss": 0.3283, + "num_input_tokens_seen": 18096144, + "step": 8660 + }, + { + "epoch": 1.4135737009544007, + "grad_norm": 0.0009563640342094004, + "learning_rate": 0.26659166234486614, + "loss": 0.3266, + "num_input_tokens_seen": 18107840, + "step": 8665 + }, + { + "epoch": 1.4143894281752183, + "grad_norm": 0.00044019773486070335, + "learning_rate": 0.2665545929133869, + "loss": 0.2841, + "num_input_tokens_seen": 18118064, + "step": 8670 + }, + { + "epoch": 1.4152051553960356, + "grad_norm": 0.0012144878273829818, + "learning_rate": 0.2665175055077248, + "loss": 0.3885, + "num_input_tokens_seen": 18128656, + "step": 8675 + }, + { + "epoch": 1.416020882616853, + "grad_norm": 0.000491540355142206, + "learning_rate": 0.2664804001335991, + "loss": 0.2345, + "num_input_tokens_seen": 18140560, + "step": 8680 + }, + { + "epoch": 1.4168366098376703, + "grad_norm": 0.0005054217181168497, + "learning_rate": 0.26644327679673185, + "loss": 0.3037, + "num_input_tokens_seen": 18152048, + "step": 8685 + }, + { + "epoch": 1.4176523370584877, + "grad_norm": 0.002557041822001338, + "learning_rate": 0.26640613550284803, + "loss": 0.3446, + "num_input_tokens_seen": 18161808, + "step": 8690 + }, + { + "epoch": 1.418468064279305, + "grad_norm": 0.001157795893959701, + "learning_rate": 0.26636897625767525, + "loss": 0.3022, + "num_input_tokens_seen": 18173488, + "step": 8695 + }, + { + "epoch": 1.4192837915001224, + "grad_norm": 0.0008764455560594797, + "learning_rate": 0.266331799066944, + "loss": 0.247, + "num_input_tokens_seen": 18182448, + "step": 8700 + }, + { + "epoch": 1.4200995187209398, + "grad_norm": 0.0007525365217588842, + "learning_rate": 0.2662946039363874, + "loss": 0.284, + "num_input_tokens_seen": 18191296, + "step": 8705 + }, + { + "epoch": 1.420915245941757, + "grad_norm": 0.001262719975784421, + "learning_rate": 0.2662573908717414, + "loss": 0.3165, + "num_input_tokens_seen": 18202128, + "step": 8710 + }, + { + "epoch": 1.4217309731625745, + "grad_norm": 0.0012581974733620882, + "learning_rate": 0.2662201598787447, + "loss": 0.3745, + "num_input_tokens_seen": 18213280, + "step": 8715 + }, + { + "epoch": 1.4225467003833918, + "grad_norm": 0.002120323944836855, + "learning_rate": 0.2661829109631389, + "loss": 0.2874, + "num_input_tokens_seen": 18223936, + "step": 8720 + }, + { + "epoch": 1.4233624276042092, + "grad_norm": 0.001978317741304636, + "learning_rate": 0.26614564413066816, + "loss": 0.4007, + "num_input_tokens_seen": 18234800, + "step": 8725 + }, + { + "epoch": 1.4241781548250265, + "grad_norm": 0.0011171478545293212, + "learning_rate": 0.2661083593870795, + "loss": 0.3611, + "num_input_tokens_seen": 18244240, + "step": 8730 + }, + { + "epoch": 1.4249938820458439, + "grad_norm": 0.0010595156345516443, + "learning_rate": 0.26607105673812276, + "loss": 0.2976, + "num_input_tokens_seen": 18256240, + "step": 8735 + }, + { + "epoch": 1.4258096092666612, + "grad_norm": 0.0006357814418151975, + "learning_rate": 0.2660337361895504, + "loss": 0.3769, + "num_input_tokens_seen": 18268608, + "step": 8740 + }, + { + "epoch": 1.4266253364874786, + "grad_norm": 0.0015081280143931508, + "learning_rate": 0.26599639774711775, + "loss": 0.2639, + "num_input_tokens_seen": 18279936, + "step": 8745 + }, + { + "epoch": 1.427441063708296, + "grad_norm": 0.0003134124563075602, + "learning_rate": 0.2659590414165829, + "loss": 0.3188, + "num_input_tokens_seen": 18289552, + "step": 8750 + }, + { + "epoch": 1.4282567909291133, + "grad_norm": 0.0006064129993319511, + "learning_rate": 0.2659216672037066, + "loss": 0.2863, + "num_input_tokens_seen": 18301600, + "step": 8755 + }, + { + "epoch": 1.4290725181499306, + "grad_norm": 0.0006486538331955671, + "learning_rate": 0.26588427511425244, + "loss": 0.3107, + "num_input_tokens_seen": 18311328, + "step": 8760 + }, + { + "epoch": 1.429888245370748, + "grad_norm": 0.0003411301877349615, + "learning_rate": 0.26584686515398676, + "loss": 0.3139, + "num_input_tokens_seen": 18321408, + "step": 8765 + }, + { + "epoch": 1.4307039725915653, + "grad_norm": 0.0003961423353757709, + "learning_rate": 0.2658094373286787, + "loss": 0.286, + "num_input_tokens_seen": 18332720, + "step": 8770 + }, + { + "epoch": 1.4315196998123827, + "grad_norm": 0.0015173861756920815, + "learning_rate": 0.2657719916441, + "loss": 0.3215, + "num_input_tokens_seen": 18342576, + "step": 8775 + }, + { + "epoch": 1.4323354270332, + "grad_norm": 0.0008866150164976716, + "learning_rate": 0.2657345281060253, + "loss": 0.3155, + "num_input_tokens_seen": 18353360, + "step": 8780 + }, + { + "epoch": 1.4331511542540174, + "grad_norm": 0.0015520218294113874, + "learning_rate": 0.26569704672023203, + "loss": 0.3877, + "num_input_tokens_seen": 18363200, + "step": 8785 + }, + { + "epoch": 1.4339668814748348, + "grad_norm": 0.0017010357696563005, + "learning_rate": 0.26565954749250015, + "loss": 0.3255, + "num_input_tokens_seen": 18373456, + "step": 8790 + }, + { + "epoch": 1.434782608695652, + "grad_norm": 0.00166057120077312, + "learning_rate": 0.2656220304286126, + "loss": 0.3185, + "num_input_tokens_seen": 18384432, + "step": 8795 + }, + { + "epoch": 1.4355983359164695, + "grad_norm": 0.001079820329323411, + "learning_rate": 0.265584495534355, + "loss": 0.289, + "num_input_tokens_seen": 18394320, + "step": 8800 + }, + { + "epoch": 1.4355983359164695, + "eval_loss": 0.3266211450099945, + "eval_runtime": 155.5752, + "eval_samples_per_second": 17.516, + "eval_steps_per_second": 8.761, + "num_input_tokens_seen": 18394320, + "step": 8800 + }, + { + "epoch": 1.4364140631372868, + "grad_norm": 0.0005356412730179727, + "learning_rate": 0.2655469428155156, + "loss": 0.2979, + "num_input_tokens_seen": 18404992, + "step": 8805 + }, + { + "epoch": 1.4372297903581042, + "grad_norm": 0.0005098797264508903, + "learning_rate": 0.2655093722778856, + "loss": 0.255, + "num_input_tokens_seen": 18416096, + "step": 8810 + }, + { + "epoch": 1.4380455175789217, + "grad_norm": 0.0009012409136630595, + "learning_rate": 0.2654717839272588, + "loss": 0.321, + "num_input_tokens_seen": 18426832, + "step": 8815 + }, + { + "epoch": 1.4388612447997389, + "grad_norm": 0.0009344351128675044, + "learning_rate": 0.2654341777694318, + "loss": 0.3852, + "num_input_tokens_seen": 18437584, + "step": 8820 + }, + { + "epoch": 1.4396769720205564, + "grad_norm": 0.0009955504210665822, + "learning_rate": 0.265396553810204, + "loss": 0.3445, + "num_input_tokens_seen": 18447024, + "step": 8825 + }, + { + "epoch": 1.4404926992413736, + "grad_norm": 0.0021760137751698494, + "learning_rate": 0.26535891205537737, + "loss": 0.357, + "num_input_tokens_seen": 18458512, + "step": 8830 + }, + { + "epoch": 1.4413084264621912, + "grad_norm": 0.0008438961813226342, + "learning_rate": 0.26532125251075683, + "loss": 0.3497, + "num_input_tokens_seen": 18469824, + "step": 8835 + }, + { + "epoch": 1.4421241536830083, + "grad_norm": 0.0004877269093412906, + "learning_rate": 0.26528357518214996, + "loss": 0.3824, + "num_input_tokens_seen": 18479936, + "step": 8840 + }, + { + "epoch": 1.4429398809038259, + "grad_norm": 0.0010831572581082582, + "learning_rate": 0.26524588007536704, + "loss": 0.3639, + "num_input_tokens_seen": 18489072, + "step": 8845 + }, + { + "epoch": 1.443755608124643, + "grad_norm": 0.0015521417371928692, + "learning_rate": 0.26520816719622115, + "loss": 0.2905, + "num_input_tokens_seen": 18499200, + "step": 8850 + }, + { + "epoch": 1.4445713353454606, + "grad_norm": 0.0010062626097351313, + "learning_rate": 0.2651704365505281, + "loss": 0.3547, + "num_input_tokens_seen": 18510080, + "step": 8855 + }, + { + "epoch": 1.445387062566278, + "grad_norm": 0.0004975254996679723, + "learning_rate": 0.26513268814410634, + "loss": 0.3545, + "num_input_tokens_seen": 18517264, + "step": 8860 + }, + { + "epoch": 1.4462027897870953, + "grad_norm": 0.0004290191864129156, + "learning_rate": 0.2650949219827773, + "loss": 0.3363, + "num_input_tokens_seen": 18528544, + "step": 8865 + }, + { + "epoch": 1.4470185170079126, + "grad_norm": 0.0007064397213980556, + "learning_rate": 0.26505713807236486, + "loss": 0.3163, + "num_input_tokens_seen": 18540048, + "step": 8870 + }, + { + "epoch": 1.44783424422873, + "grad_norm": 0.0006041998276486993, + "learning_rate": 0.26501933641869585, + "loss": 0.3286, + "num_input_tokens_seen": 18550688, + "step": 8875 + }, + { + "epoch": 1.4486499714495473, + "grad_norm": 0.0005561452126130462, + "learning_rate": 0.26498151702759976, + "loss": 0.335, + "num_input_tokens_seen": 18560960, + "step": 8880 + }, + { + "epoch": 1.4494656986703647, + "grad_norm": 0.0004848630342166871, + "learning_rate": 0.2649436799049088, + "loss": 0.3112, + "num_input_tokens_seen": 18572224, + "step": 8885 + }, + { + "epoch": 1.450281425891182, + "grad_norm": 0.001222575199790299, + "learning_rate": 0.2649058250564579, + "loss": 0.3205, + "num_input_tokens_seen": 18581872, + "step": 8890 + }, + { + "epoch": 1.4510971531119994, + "grad_norm": 0.0010504914680495858, + "learning_rate": 0.26486795248808476, + "loss": 0.3843, + "num_input_tokens_seen": 18592736, + "step": 8895 + }, + { + "epoch": 1.4519128803328167, + "grad_norm": 0.0005310230772010982, + "learning_rate": 0.2648300622056298, + "loss": 0.3043, + "num_input_tokens_seen": 18603232, + "step": 8900 + }, + { + "epoch": 1.452728607553634, + "grad_norm": 0.031833354383707047, + "learning_rate": 0.2647921542149363, + "loss": 0.3676, + "num_input_tokens_seen": 18614256, + "step": 8905 + }, + { + "epoch": 1.4535443347744514, + "grad_norm": 0.00195384887047112, + "learning_rate": 0.26475422852185, + "loss": 0.3453, + "num_input_tokens_seen": 18625968, + "step": 8910 + }, + { + "epoch": 1.4543600619952688, + "grad_norm": 0.0006763515411876142, + "learning_rate": 0.2647162851322196, + "loss": 0.3328, + "num_input_tokens_seen": 18636048, + "step": 8915 + }, + { + "epoch": 1.4551757892160861, + "grad_norm": 0.000545258226338774, + "learning_rate": 0.2646783240518964, + "loss": 0.3241, + "num_input_tokens_seen": 18646144, + "step": 8920 + }, + { + "epoch": 1.4559915164369035, + "grad_norm": 0.0010180602548643947, + "learning_rate": 0.26464034528673447, + "loss": 0.3273, + "num_input_tokens_seen": 18657312, + "step": 8925 + }, + { + "epoch": 1.4568072436577209, + "grad_norm": 0.0004903950612060726, + "learning_rate": 0.26460234884259065, + "loss": 0.3616, + "num_input_tokens_seen": 18668976, + "step": 8930 + }, + { + "epoch": 1.4576229708785382, + "grad_norm": 0.0012348182499408722, + "learning_rate": 0.2645643347253245, + "loss": 0.3155, + "num_input_tokens_seen": 18679728, + "step": 8935 + }, + { + "epoch": 1.4584386980993556, + "grad_norm": 0.0009611008572392166, + "learning_rate": 0.2645263029407982, + "loss": 0.3867, + "num_input_tokens_seen": 18691296, + "step": 8940 + }, + { + "epoch": 1.459254425320173, + "grad_norm": 0.0010527314152568579, + "learning_rate": 0.2644882534948767, + "loss": 0.3573, + "num_input_tokens_seen": 18702080, + "step": 8945 + }, + { + "epoch": 1.4600701525409903, + "grad_norm": 0.001170526840724051, + "learning_rate": 0.2644501863934278, + "loss": 0.2996, + "num_input_tokens_seen": 18713120, + "step": 8950 + }, + { + "epoch": 1.4608858797618076, + "grad_norm": 0.0005134872044436634, + "learning_rate": 0.26441210164232193, + "loss": 0.3704, + "num_input_tokens_seen": 18723488, + "step": 8955 + }, + { + "epoch": 1.461701606982625, + "grad_norm": 0.0014453437179327011, + "learning_rate": 0.26437399924743216, + "loss": 0.3617, + "num_input_tokens_seen": 18733968, + "step": 8960 + }, + { + "epoch": 1.4625173342034423, + "grad_norm": 0.0004026776587124914, + "learning_rate": 0.26433587921463436, + "loss": 0.3131, + "num_input_tokens_seen": 18745664, + "step": 8965 + }, + { + "epoch": 1.4633330614242597, + "grad_norm": 0.0005141075816936791, + "learning_rate": 0.2642977415498072, + "loss": 0.328, + "num_input_tokens_seen": 18755680, + "step": 8970 + }, + { + "epoch": 1.464148788645077, + "grad_norm": 0.0009161701891571283, + "learning_rate": 0.26425958625883195, + "loss": 0.2822, + "num_input_tokens_seen": 18765920, + "step": 8975 + }, + { + "epoch": 1.4649645158658944, + "grad_norm": 0.0010506808757781982, + "learning_rate": 0.2642214133475926, + "loss": 0.2964, + "num_input_tokens_seen": 18776480, + "step": 8980 + }, + { + "epoch": 1.4657802430867117, + "grad_norm": 0.0006917060818523169, + "learning_rate": 0.26418322282197587, + "loss": 0.29, + "num_input_tokens_seen": 18787632, + "step": 8985 + }, + { + "epoch": 1.466595970307529, + "grad_norm": 0.0030167843215167522, + "learning_rate": 0.2641450146878714, + "loss": 0.3416, + "num_input_tokens_seen": 18797488, + "step": 8990 + }, + { + "epoch": 1.4674116975283464, + "grad_norm": 0.0007440244080498815, + "learning_rate": 0.26410678895117107, + "loss": 0.2885, + "num_input_tokens_seen": 18808880, + "step": 8995 + }, + { + "epoch": 1.468227424749164, + "grad_norm": 0.002238346030935645, + "learning_rate": 0.26406854561777, + "loss": 0.3238, + "num_input_tokens_seen": 18820208, + "step": 9000 + }, + { + "epoch": 1.468227424749164, + "eval_loss": 0.32159891724586487, + "eval_runtime": 155.9379, + "eval_samples_per_second": 17.475, + "eval_steps_per_second": 8.741, + "num_input_tokens_seen": 18820208, + "step": 9000 + }, + { + "epoch": 1.4690431519699811, + "grad_norm": 0.0003218413330614567, + "learning_rate": 0.26403028469356576, + "loss": 0.2849, + "num_input_tokens_seen": 18830160, + "step": 9005 + }, + { + "epoch": 1.4698588791907987, + "grad_norm": 0.0009661619551479816, + "learning_rate": 0.2639920061844585, + "loss": 0.3176, + "num_input_tokens_seen": 18839408, + "step": 9010 + }, + { + "epoch": 1.4706746064116158, + "grad_norm": 0.0007308617350645363, + "learning_rate": 0.2639537100963515, + "loss": 0.3919, + "num_input_tokens_seen": 18847520, + "step": 9015 + }, + { + "epoch": 1.4714903336324334, + "grad_norm": 0.00043160474160686135, + "learning_rate": 0.26391539643515033, + "loss": 0.3038, + "num_input_tokens_seen": 18858608, + "step": 9020 + }, + { + "epoch": 1.4723060608532506, + "grad_norm": 0.0007173013291321695, + "learning_rate": 0.26387706520676346, + "loss": 0.2431, + "num_input_tokens_seen": 18868416, + "step": 9025 + }, + { + "epoch": 1.4731217880740681, + "grad_norm": 0.0011075400980189443, + "learning_rate": 0.26383871641710205, + "loss": 0.3558, + "num_input_tokens_seen": 18877408, + "step": 9030 + }, + { + "epoch": 1.4739375152948853, + "grad_norm": 0.0015375574585050344, + "learning_rate": 0.26380035007208, + "loss": 0.3491, + "num_input_tokens_seen": 18887552, + "step": 9035 + }, + { + "epoch": 1.4747532425157028, + "grad_norm": 0.0011686085490509868, + "learning_rate": 0.26376196617761394, + "loss": 0.3258, + "num_input_tokens_seen": 18896832, + "step": 9040 + }, + { + "epoch": 1.4755689697365202, + "grad_norm": 0.001676783780567348, + "learning_rate": 0.263723564739623, + "loss": 0.3425, + "num_input_tokens_seen": 18907856, + "step": 9045 + }, + { + "epoch": 1.4763846969573375, + "grad_norm": 0.0006948191439732909, + "learning_rate": 0.2636851457640293, + "loss": 0.2836, + "num_input_tokens_seen": 18919008, + "step": 9050 + }, + { + "epoch": 1.477200424178155, + "grad_norm": 0.0009685778641141951, + "learning_rate": 0.26364670925675737, + "loss": 0.3293, + "num_input_tokens_seen": 18931344, + "step": 9055 + }, + { + "epoch": 1.4780161513989722, + "grad_norm": 0.0010284746531397104, + "learning_rate": 0.2636082552237347, + "loss": 0.3367, + "num_input_tokens_seen": 18940528, + "step": 9060 + }, + { + "epoch": 1.4788318786197896, + "grad_norm": 0.0004848528769798577, + "learning_rate": 0.26356978367089146, + "loss": 0.3197, + "num_input_tokens_seen": 18950784, + "step": 9065 + }, + { + "epoch": 1.479647605840607, + "grad_norm": 0.0016161822713911533, + "learning_rate": 0.26353129460416036, + "loss": 0.3326, + "num_input_tokens_seen": 18961616, + "step": 9070 + }, + { + "epoch": 1.4804633330614243, + "grad_norm": 0.0005054870271123946, + "learning_rate": 0.2634927880294769, + "loss": 0.3193, + "num_input_tokens_seen": 18972880, + "step": 9075 + }, + { + "epoch": 1.4812790602822417, + "grad_norm": 0.0009414623491466045, + "learning_rate": 0.26345426395277927, + "loss": 0.2991, + "num_input_tokens_seen": 18984704, + "step": 9080 + }, + { + "epoch": 1.482094787503059, + "grad_norm": 0.001064323354512453, + "learning_rate": 0.2634157223800084, + "loss": 0.2723, + "num_input_tokens_seen": 18996560, + "step": 9085 + }, + { + "epoch": 1.4829105147238764, + "grad_norm": 0.0012042855378240347, + "learning_rate": 0.26337716331710787, + "loss": 0.3739, + "num_input_tokens_seen": 19006176, + "step": 9090 + }, + { + "epoch": 1.4837262419446937, + "grad_norm": 0.0013092756271362305, + "learning_rate": 0.2633385867700239, + "loss": 0.2633, + "num_input_tokens_seen": 19017120, + "step": 9095 + }, + { + "epoch": 1.484541969165511, + "grad_norm": 0.0031640916131436825, + "learning_rate": 0.2632999927447056, + "loss": 0.3365, + "num_input_tokens_seen": 19028512, + "step": 9100 + }, + { + "epoch": 1.4853576963863284, + "grad_norm": 0.002919908380135894, + "learning_rate": 0.2632613812471046, + "loss": 0.2843, + "num_input_tokens_seen": 19038992, + "step": 9105 + }, + { + "epoch": 1.4861734236071458, + "grad_norm": 0.0021958055440336466, + "learning_rate": 0.2632227522831753, + "loss": 0.2803, + "num_input_tokens_seen": 19049728, + "step": 9110 + }, + { + "epoch": 1.4869891508279631, + "grad_norm": 0.0004656772071029991, + "learning_rate": 0.26318410585887475, + "loss": 0.3496, + "num_input_tokens_seen": 19059520, + "step": 9115 + }, + { + "epoch": 1.4878048780487805, + "grad_norm": 0.0022581482771784067, + "learning_rate": 0.2631454419801627, + "loss": 0.275, + "num_input_tokens_seen": 19070944, + "step": 9120 + }, + { + "epoch": 1.4886206052695978, + "grad_norm": 0.0006262877723202109, + "learning_rate": 0.2631067606530016, + "loss": 0.3466, + "num_input_tokens_seen": 19082672, + "step": 9125 + }, + { + "epoch": 1.4894363324904152, + "grad_norm": 0.002379129407927394, + "learning_rate": 0.2630680618833567, + "loss": 0.3025, + "num_input_tokens_seen": 19093536, + "step": 9130 + }, + { + "epoch": 1.4902520597112325, + "grad_norm": 0.0006862705340608954, + "learning_rate": 0.26302934567719566, + "loss": 0.3513, + "num_input_tokens_seen": 19104848, + "step": 9135 + }, + { + "epoch": 1.4910677869320499, + "grad_norm": 0.0009277355275116861, + "learning_rate": 0.2629906120404892, + "loss": 0.3891, + "num_input_tokens_seen": 19116704, + "step": 9140 + }, + { + "epoch": 1.4918835141528672, + "grad_norm": 0.0023360424675047398, + "learning_rate": 0.26295186097921036, + "loss": 0.3853, + "num_input_tokens_seen": 19128688, + "step": 9145 + }, + { + "epoch": 1.4926992413736846, + "grad_norm": 0.001803129562176764, + "learning_rate": 0.2629130924993351, + "loss": 0.3167, + "num_input_tokens_seen": 19141664, + "step": 9150 + }, + { + "epoch": 1.493514968594502, + "grad_norm": 0.001102640526369214, + "learning_rate": 0.2628743066068421, + "loss": 0.3878, + "num_input_tokens_seen": 19151408, + "step": 9155 + }, + { + "epoch": 1.4943306958153193, + "grad_norm": 0.0005829165456816554, + "learning_rate": 0.26283550330771244, + "loss": 0.3015, + "num_input_tokens_seen": 19162016, + "step": 9160 + }, + { + "epoch": 1.4951464230361367, + "grad_norm": 0.0005638590082526207, + "learning_rate": 0.2627966826079303, + "loss": 0.3276, + "num_input_tokens_seen": 19172064, + "step": 9165 + }, + { + "epoch": 1.495962150256954, + "grad_norm": 0.0006338460952974856, + "learning_rate": 0.26275784451348216, + "loss": 0.3127, + "num_input_tokens_seen": 19181456, + "step": 9170 + }, + { + "epoch": 1.4967778774777714, + "grad_norm": 0.00036910618655383587, + "learning_rate": 0.2627189890303574, + "loss": 0.3055, + "num_input_tokens_seen": 19191440, + "step": 9175 + }, + { + "epoch": 1.4975936046985887, + "grad_norm": 0.0008892737096175551, + "learning_rate": 0.262680116164548, + "loss": 0.3138, + "num_input_tokens_seen": 19203504, + "step": 9180 + }, + { + "epoch": 1.4984093319194063, + "grad_norm": 0.0006483875913545489, + "learning_rate": 0.2626412259220487, + "loss": 0.2345, + "num_input_tokens_seen": 19214064, + "step": 9185 + }, + { + "epoch": 1.4992250591402234, + "grad_norm": 0.00041485263500362635, + "learning_rate": 0.2626023183088568, + "loss": 0.3325, + "num_input_tokens_seen": 19223792, + "step": 9190 + }, + { + "epoch": 1.500040786361041, + "grad_norm": 0.0006280962261371315, + "learning_rate": 0.26256339333097234, + "loss": 0.3321, + "num_input_tokens_seen": 19233312, + "step": 9195 + }, + { + "epoch": 1.5008565135818581, + "grad_norm": 0.0008895988576114178, + "learning_rate": 0.2625244509943981, + "loss": 0.4018, + "num_input_tokens_seen": 19244192, + "step": 9200 + }, + { + "epoch": 1.5008565135818581, + "eval_loss": 0.3190626800060272, + "eval_runtime": 155.5896, + "eval_samples_per_second": 17.514, + "eval_steps_per_second": 8.76, + "num_input_tokens_seen": 19244192, + "step": 9200 + }, + { + "epoch": 1.5016722408026757, + "grad_norm": 0.0010743618477135897, + "learning_rate": 0.2624854913051395, + "loss": 0.3094, + "num_input_tokens_seen": 19254640, + "step": 9205 + }, + { + "epoch": 1.5024879680234928, + "grad_norm": 0.0004733838140964508, + "learning_rate": 0.26244651426920446, + "loss": 0.3264, + "num_input_tokens_seen": 19265568, + "step": 9210 + }, + { + "epoch": 1.5033036952443104, + "grad_norm": 0.0003500799066387117, + "learning_rate": 0.26240751989260386, + "loss": 0.2716, + "num_input_tokens_seen": 19275360, + "step": 9215 + }, + { + "epoch": 1.5041194224651275, + "grad_norm": 0.0020939658861607313, + "learning_rate": 0.2623685081813511, + "loss": 0.3223, + "num_input_tokens_seen": 19286800, + "step": 9220 + }, + { + "epoch": 1.504935149685945, + "grad_norm": 0.000393404858186841, + "learning_rate": 0.2623294791414623, + "loss": 0.2785, + "num_input_tokens_seen": 19296192, + "step": 9225 + }, + { + "epoch": 1.5057508769067622, + "grad_norm": 0.003156145801767707, + "learning_rate": 0.26229043277895614, + "loss": 0.3378, + "num_input_tokens_seen": 19306864, + "step": 9230 + }, + { + "epoch": 1.5065666041275798, + "grad_norm": 0.0010758355492725968, + "learning_rate": 0.2622513690998542, + "loss": 0.2841, + "num_input_tokens_seen": 19318288, + "step": 9235 + }, + { + "epoch": 1.507382331348397, + "grad_norm": 0.0006210255087353289, + "learning_rate": 0.26221228811018044, + "loss": 0.3184, + "num_input_tokens_seen": 19328112, + "step": 9240 + }, + { + "epoch": 1.5081980585692145, + "grad_norm": 0.0006295893690548837, + "learning_rate": 0.2621731898159617, + "loss": 0.3504, + "num_input_tokens_seen": 19339872, + "step": 9245 + }, + { + "epoch": 1.5090137857900316, + "grad_norm": 0.0006645281100645661, + "learning_rate": 0.26213407422322743, + "loss": 0.2922, + "num_input_tokens_seen": 19349424, + "step": 9250 + }, + { + "epoch": 1.5098295130108492, + "grad_norm": 0.0019557687919586897, + "learning_rate": 0.2620949413380098, + "loss": 0.3793, + "num_input_tokens_seen": 19360560, + "step": 9255 + }, + { + "epoch": 1.5106452402316666, + "grad_norm": 0.0009617840405553579, + "learning_rate": 0.26205579116634353, + "loss": 0.3869, + "num_input_tokens_seen": 19369968, + "step": 9260 + }, + { + "epoch": 1.511460967452484, + "grad_norm": 0.0009482190362177789, + "learning_rate": 0.26201662371426604, + "loss": 0.3467, + "num_input_tokens_seen": 19378624, + "step": 9265 + }, + { + "epoch": 1.5122766946733013, + "grad_norm": 0.0008507429738529027, + "learning_rate": 0.2619774389878175, + "loss": 0.3323, + "num_input_tokens_seen": 19388224, + "step": 9270 + }, + { + "epoch": 1.5130924218941186, + "grad_norm": 0.0023688911460340023, + "learning_rate": 0.2619382369930407, + "loss": 0.3275, + "num_input_tokens_seen": 19399120, + "step": 9275 + }, + { + "epoch": 1.513908149114936, + "grad_norm": 0.0008160388679243624, + "learning_rate": 0.261899017735981, + "loss": 0.2947, + "num_input_tokens_seen": 19409584, + "step": 9280 + }, + { + "epoch": 1.5147238763357533, + "grad_norm": 0.0008573423838242888, + "learning_rate": 0.2618597812226866, + "loss": 0.3443, + "num_input_tokens_seen": 19419360, + "step": 9285 + }, + { + "epoch": 1.5155396035565707, + "grad_norm": 0.0008822022355161607, + "learning_rate": 0.2618205274592082, + "loss": 0.3509, + "num_input_tokens_seen": 19428368, + "step": 9290 + }, + { + "epoch": 1.516355330777388, + "grad_norm": 0.0009909544605761766, + "learning_rate": 0.2617812564515992, + "loss": 0.3555, + "num_input_tokens_seen": 19438240, + "step": 9295 + }, + { + "epoch": 1.5171710579982054, + "grad_norm": 0.0013537841150537133, + "learning_rate": 0.2617419682059158, + "loss": 0.3168, + "num_input_tokens_seen": 19448784, + "step": 9300 + }, + { + "epoch": 1.5179867852190227, + "grad_norm": 0.002487300196662545, + "learning_rate": 0.26170266272821663, + "loss": 0.3181, + "num_input_tokens_seen": 19458752, + "step": 9305 + }, + { + "epoch": 1.51880251243984, + "grad_norm": 0.002225360367447138, + "learning_rate": 0.26166334002456315, + "loss": 0.7583, + "num_input_tokens_seen": 19469888, + "step": 9310 + }, + { + "epoch": 1.5196182396606575, + "grad_norm": 0.0011106771416962147, + "learning_rate": 0.2616240001010194, + "loss": 0.3508, + "num_input_tokens_seen": 19480848, + "step": 9315 + }, + { + "epoch": 1.5204339668814748, + "grad_norm": 0.0008614849648438394, + "learning_rate": 0.26158464296365197, + "loss": 0.3151, + "num_input_tokens_seen": 19491280, + "step": 9320 + }, + { + "epoch": 1.5212496941022922, + "grad_norm": 0.0010851698461920023, + "learning_rate": 0.2615452686185304, + "loss": 0.3102, + "num_input_tokens_seen": 19501904, + "step": 9325 + }, + { + "epoch": 1.5220654213231095, + "grad_norm": 0.0007264797459356487, + "learning_rate": 0.26150587707172673, + "loss": 0.3681, + "num_input_tokens_seen": 19511984, + "step": 9330 + }, + { + "epoch": 1.5228811485439269, + "grad_norm": 0.0009147971286438406, + "learning_rate": 0.2614664683293154, + "loss": 0.3118, + "num_input_tokens_seen": 19521952, + "step": 9335 + }, + { + "epoch": 1.5236968757647442, + "grad_norm": 0.0020913875196129084, + "learning_rate": 0.26142704239737397, + "loss": 0.3813, + "num_input_tokens_seen": 19531648, + "step": 9340 + }, + { + "epoch": 1.5245126029855616, + "grad_norm": 0.0013163305120542645, + "learning_rate": 0.26138759928198235, + "loss": 0.3527, + "num_input_tokens_seen": 19540672, + "step": 9345 + }, + { + "epoch": 1.5253283302063791, + "grad_norm": 0.0008750211563892663, + "learning_rate": 0.26134813898922304, + "loss": 0.3368, + "num_input_tokens_seen": 19549328, + "step": 9350 + }, + { + "epoch": 1.5261440574271963, + "grad_norm": 0.0009804537985473871, + "learning_rate": 0.26130866152518145, + "loss": 0.3108, + "num_input_tokens_seen": 19559872, + "step": 9355 + }, + { + "epoch": 1.5269597846480139, + "grad_norm": 0.0020631018560379744, + "learning_rate": 0.2612691668959455, + "loss": 0.3753, + "num_input_tokens_seen": 19570880, + "step": 9360 + }, + { + "epoch": 1.527775511868831, + "grad_norm": 0.0006905777263455093, + "learning_rate": 0.2612296551076057, + "loss": 0.3643, + "num_input_tokens_seen": 19581536, + "step": 9365 + }, + { + "epoch": 1.5285912390896486, + "grad_norm": 0.0012567484518513083, + "learning_rate": 0.26119012616625525, + "loss": 0.3396, + "num_input_tokens_seen": 19591344, + "step": 9370 + }, + { + "epoch": 1.5294069663104657, + "grad_norm": 0.0010946065885946155, + "learning_rate": 0.26115058007799, + "loss": 0.3523, + "num_input_tokens_seen": 19601664, + "step": 9375 + }, + { + "epoch": 1.5302226935312833, + "grad_norm": 0.002065102569758892, + "learning_rate": 0.26111101684890864, + "loss": 0.3383, + "num_input_tokens_seen": 19612192, + "step": 9380 + }, + { + "epoch": 1.5310384207521004, + "grad_norm": 0.0007725143223069608, + "learning_rate": 0.26107143648511205, + "loss": 0.3335, + "num_input_tokens_seen": 19622256, + "step": 9385 + }, + { + "epoch": 1.531854147972918, + "grad_norm": 0.0011493548518046737, + "learning_rate": 0.2610318389927042, + "loss": 0.3426, + "num_input_tokens_seen": 19633424, + "step": 9390 + }, + { + "epoch": 1.532669875193735, + "grad_norm": 0.0017142462311312556, + "learning_rate": 0.26099222437779146, + "loss": 0.3563, + "num_input_tokens_seen": 19643584, + "step": 9395 + }, + { + "epoch": 1.5334856024145527, + "grad_norm": 0.0012018659617751837, + "learning_rate": 0.26095259264648285, + "loss": 0.3232, + "num_input_tokens_seen": 19654192, + "step": 9400 + }, + { + "epoch": 1.5334856024145527, + "eval_loss": 0.3440909683704376, + "eval_runtime": 156.0057, + "eval_samples_per_second": 17.467, + "eval_steps_per_second": 8.737, + "num_input_tokens_seen": 19654192, + "step": 9400 + }, + { + "epoch": 1.5343013296353698, + "grad_norm": 0.0006925382767803967, + "learning_rate": 0.2609129438048902, + "loss": 0.3532, + "num_input_tokens_seen": 19665248, + "step": 9405 + }, + { + "epoch": 1.5351170568561874, + "grad_norm": 0.0015724935801699758, + "learning_rate": 0.2608732778591278, + "loss": 0.3126, + "num_input_tokens_seen": 19674784, + "step": 9410 + }, + { + "epoch": 1.5359327840770045, + "grad_norm": 0.0012054104590788484, + "learning_rate": 0.2608335948153126, + "loss": 0.319, + "num_input_tokens_seen": 19685632, + "step": 9415 + }, + { + "epoch": 1.536748511297822, + "grad_norm": 0.0006357654347084463, + "learning_rate": 0.26079389467956426, + "loss": 0.309, + "num_input_tokens_seen": 19697184, + "step": 9420 + }, + { + "epoch": 1.5375642385186392, + "grad_norm": 0.000698055955581367, + "learning_rate": 0.26075417745800505, + "loss": 0.3657, + "num_input_tokens_seen": 19708064, + "step": 9425 + }, + { + "epoch": 1.5383799657394568, + "grad_norm": 0.0007435831939801574, + "learning_rate": 0.26071444315675985, + "loss": 0.3336, + "num_input_tokens_seen": 19717664, + "step": 9430 + }, + { + "epoch": 1.539195692960274, + "grad_norm": 0.0006400939892046154, + "learning_rate": 0.2606746917819562, + "loss": 0.3129, + "num_input_tokens_seen": 19728736, + "step": 9435 + }, + { + "epoch": 1.5400114201810915, + "grad_norm": 0.0005352523876354098, + "learning_rate": 0.2606349233397242, + "loss": 0.3044, + "num_input_tokens_seen": 19740080, + "step": 9440 + }, + { + "epoch": 1.5408271474019088, + "grad_norm": 0.0015907324850559235, + "learning_rate": 0.26059513783619676, + "loss": 0.3379, + "num_input_tokens_seen": 19750688, + "step": 9445 + }, + { + "epoch": 1.5416428746227262, + "grad_norm": 0.0013727439800277352, + "learning_rate": 0.26055533527750924, + "loss": 0.3449, + "num_input_tokens_seen": 19762176, + "step": 9450 + }, + { + "epoch": 1.5424586018435436, + "grad_norm": 0.000896521145477891, + "learning_rate": 0.26051551566979964, + "loss": 0.2914, + "num_input_tokens_seen": 19771920, + "step": 9455 + }, + { + "epoch": 1.543274329064361, + "grad_norm": 0.0019606302957981825, + "learning_rate": 0.26047567901920876, + "loss": 0.2986, + "num_input_tokens_seen": 19782784, + "step": 9460 + }, + { + "epoch": 1.5440900562851783, + "grad_norm": 0.0010864522773772478, + "learning_rate": 0.2604358253318798, + "loss": 0.3067, + "num_input_tokens_seen": 19794256, + "step": 9465 + }, + { + "epoch": 1.5449057835059956, + "grad_norm": 0.001113697187975049, + "learning_rate": 0.26039595461395876, + "loss": 0.3298, + "num_input_tokens_seen": 19805168, + "step": 9470 + }, + { + "epoch": 1.545721510726813, + "grad_norm": 0.001437017577700317, + "learning_rate": 0.26035606687159424, + "loss": 0.3068, + "num_input_tokens_seen": 19816288, + "step": 9475 + }, + { + "epoch": 1.5465372379476303, + "grad_norm": 0.0013500561472028494, + "learning_rate": 0.26031616211093733, + "loss": 0.2746, + "num_input_tokens_seen": 19826144, + "step": 9480 + }, + { + "epoch": 1.5473529651684477, + "grad_norm": 0.0019449819810688496, + "learning_rate": 0.26027624033814195, + "loss": 0.3616, + "num_input_tokens_seen": 19835840, + "step": 9485 + }, + { + "epoch": 1.548168692389265, + "grad_norm": 0.004462305922061205, + "learning_rate": 0.2602363015593645, + "loss": 0.4126, + "num_input_tokens_seen": 19848000, + "step": 9490 + }, + { + "epoch": 1.5489844196100824, + "grad_norm": 0.0006106537184678018, + "learning_rate": 0.26019634578076395, + "loss": 0.3569, + "num_input_tokens_seen": 19858496, + "step": 9495 + }, + { + "epoch": 1.5498001468308997, + "grad_norm": 0.0010476280003786087, + "learning_rate": 0.26015637300850214, + "loss": 0.367, + "num_input_tokens_seen": 19870304, + "step": 9500 + }, + { + "epoch": 1.550615874051717, + "grad_norm": 0.0006308449665084481, + "learning_rate": 0.26011638324874325, + "loss": 0.3251, + "num_input_tokens_seen": 19881904, + "step": 9505 + }, + { + "epoch": 1.5514316012725344, + "grad_norm": 0.001185330911539495, + "learning_rate": 0.2600763765076543, + "loss": 0.3308, + "num_input_tokens_seen": 19892720, + "step": 9510 + }, + { + "epoch": 1.5522473284933518, + "grad_norm": 0.0010499792406335473, + "learning_rate": 0.2600363527914048, + "loss": 0.3416, + "num_input_tokens_seen": 19902800, + "step": 9515 + }, + { + "epoch": 1.5530630557141691, + "grad_norm": 0.0004784755001310259, + "learning_rate": 0.25999631210616686, + "loss": 0.3506, + "num_input_tokens_seen": 19913424, + "step": 9520 + }, + { + "epoch": 1.5538787829349865, + "grad_norm": 0.0007673536892980337, + "learning_rate": 0.25995625445811527, + "loss": 0.3365, + "num_input_tokens_seen": 19922976, + "step": 9525 + }, + { + "epoch": 1.5546945101558038, + "grad_norm": 0.0019712084904313087, + "learning_rate": 0.2599161798534275, + "loss": 0.3352, + "num_input_tokens_seen": 19932992, + "step": 9530 + }, + { + "epoch": 1.5555102373766214, + "grad_norm": 0.0011069567408412695, + "learning_rate": 0.25987608829828346, + "loss": 0.33, + "num_input_tokens_seen": 19944160, + "step": 9535 + }, + { + "epoch": 1.5563259645974385, + "grad_norm": 0.0008037110092118382, + "learning_rate": 0.25983597979886586, + "loss": 0.3152, + "num_input_tokens_seen": 19955216, + "step": 9540 + }, + { + "epoch": 1.5571416918182561, + "grad_norm": 0.0013685263693332672, + "learning_rate": 0.2597958543613599, + "loss": 0.3796, + "num_input_tokens_seen": 19965440, + "step": 9545 + }, + { + "epoch": 1.5579574190390733, + "grad_norm": 0.0015826961025595665, + "learning_rate": 0.25975571199195335, + "loss": 0.3773, + "num_input_tokens_seen": 19973936, + "step": 9550 + }, + { + "epoch": 1.5587731462598908, + "grad_norm": 0.0007500678184442222, + "learning_rate": 0.25971555269683677, + "loss": 0.3484, + "num_input_tokens_seen": 19984752, + "step": 9555 + }, + { + "epoch": 1.559588873480708, + "grad_norm": 0.000582374632358551, + "learning_rate": 0.25967537648220324, + "loss": 0.3457, + "num_input_tokens_seen": 19995056, + "step": 9560 + }, + { + "epoch": 1.5604046007015255, + "grad_norm": 0.0012609573313966393, + "learning_rate": 0.2596351833542483, + "loss": 0.3454, + "num_input_tokens_seen": 20005888, + "step": 9565 + }, + { + "epoch": 1.5612203279223427, + "grad_norm": 0.0010249987244606018, + "learning_rate": 0.25959497331917036, + "loss": 0.3125, + "num_input_tokens_seen": 20016688, + "step": 9570 + }, + { + "epoch": 1.5620360551431602, + "grad_norm": 0.001272124471142888, + "learning_rate": 0.2595547463831703, + "loss": 0.3666, + "num_input_tokens_seen": 20027808, + "step": 9575 + }, + { + "epoch": 1.5628517823639774, + "grad_norm": 0.0007100733346305788, + "learning_rate": 0.25951450255245156, + "loss": 0.3147, + "num_input_tokens_seen": 20038144, + "step": 9580 + }, + { + "epoch": 1.563667509584795, + "grad_norm": 0.0007304640021175146, + "learning_rate": 0.2594742418332203, + "loss": 0.3311, + "num_input_tokens_seen": 20049552, + "step": 9585 + }, + { + "epoch": 1.564483236805612, + "grad_norm": 0.0007717667031101882, + "learning_rate": 0.2594339642316852, + "loss": 0.3423, + "num_input_tokens_seen": 20057744, + "step": 9590 + }, + { + "epoch": 1.5652989640264297, + "grad_norm": 0.0011554567608982325, + "learning_rate": 0.2593936697540576, + "loss": 0.324, + "num_input_tokens_seen": 20067536, + "step": 9595 + }, + { + "epoch": 1.5661146912472468, + "grad_norm": 0.0008503863937221467, + "learning_rate": 0.2593533584065514, + "loss": 0.2839, + "num_input_tokens_seen": 20077520, + "step": 9600 + }, + { + "epoch": 1.5661146912472468, + "eval_loss": 0.3350810408592224, + "eval_runtime": 155.4153, + "eval_samples_per_second": 17.534, + "eval_steps_per_second": 8.77, + "num_input_tokens_seen": 20077520, + "step": 9600 + }, + { + "epoch": 1.5669304184680644, + "grad_norm": 0.001914996886625886, + "learning_rate": 0.2593130301953831, + "loss": 0.294, + "num_input_tokens_seen": 20086640, + "step": 9605 + }, + { + "epoch": 1.5677461456888815, + "grad_norm": 0.0017211042577400804, + "learning_rate": 0.2592726851267718, + "loss": 0.3746, + "num_input_tokens_seen": 20097472, + "step": 9610 + }, + { + "epoch": 1.568561872909699, + "grad_norm": 0.0006970216054469347, + "learning_rate": 0.2592323232069393, + "loss": 0.3768, + "num_input_tokens_seen": 20107104, + "step": 9615 + }, + { + "epoch": 1.5693776001305162, + "grad_norm": 0.0010165777057409286, + "learning_rate": 0.25919194444210986, + "loss": 0.3088, + "num_input_tokens_seen": 20117280, + "step": 9620 + }, + { + "epoch": 1.5701933273513338, + "grad_norm": 0.00039550932706333697, + "learning_rate": 0.2591515488385103, + "loss": 0.3299, + "num_input_tokens_seen": 20129120, + "step": 9625 + }, + { + "epoch": 1.5710090545721511, + "grad_norm": 0.0005806655972264707, + "learning_rate": 0.2591111364023704, + "loss": 0.3018, + "num_input_tokens_seen": 20138672, + "step": 9630 + }, + { + "epoch": 1.5718247817929685, + "grad_norm": 0.0009855846874415874, + "learning_rate": 0.259070707139922, + "loss": 0.3388, + "num_input_tokens_seen": 20149888, + "step": 9635 + }, + { + "epoch": 1.5726405090137858, + "grad_norm": 0.0013741590082645416, + "learning_rate": 0.25903026105739985, + "loss": 0.3055, + "num_input_tokens_seen": 20160192, + "step": 9640 + }, + { + "epoch": 1.5734562362346032, + "grad_norm": 0.0027232342399656773, + "learning_rate": 0.2589897981610413, + "loss": 0.344, + "num_input_tokens_seen": 20170608, + "step": 9645 + }, + { + "epoch": 1.5742719634554205, + "grad_norm": 0.00052365887677297, + "learning_rate": 0.2589493184570863, + "loss": 0.3257, + "num_input_tokens_seen": 20181936, + "step": 9650 + }, + { + "epoch": 1.5750876906762379, + "grad_norm": 0.0012822115095332265, + "learning_rate": 0.25890882195177717, + "loss": 0.2985, + "num_input_tokens_seen": 20191008, + "step": 9655 + }, + { + "epoch": 1.5759034178970552, + "grad_norm": 0.0007645913283340633, + "learning_rate": 0.25886830865135907, + "loss": 0.3645, + "num_input_tokens_seen": 20199840, + "step": 9660 + }, + { + "epoch": 1.5767191451178726, + "grad_norm": 0.001034095068462193, + "learning_rate": 0.25882777856207967, + "loss": 0.3571, + "num_input_tokens_seen": 20211152, + "step": 9665 + }, + { + "epoch": 1.57753487233869, + "grad_norm": 0.001197558711282909, + "learning_rate": 0.2587872316901892, + "loss": 0.3378, + "num_input_tokens_seen": 20220304, + "step": 9670 + }, + { + "epoch": 1.5783505995595073, + "grad_norm": 0.000923741317819804, + "learning_rate": 0.25874666804194046, + "loss": 0.3257, + "num_input_tokens_seen": 20230432, + "step": 9675 + }, + { + "epoch": 1.5791663267803246, + "grad_norm": 0.0009878658456727862, + "learning_rate": 0.258706087623589, + "loss": 0.3614, + "num_input_tokens_seen": 20240896, + "step": 9680 + }, + { + "epoch": 1.579982054001142, + "grad_norm": 0.0004994118353351951, + "learning_rate": 0.25866549044139264, + "loss": 0.3106, + "num_input_tokens_seen": 20252448, + "step": 9685 + }, + { + "epoch": 1.5807977812219594, + "grad_norm": 0.0011404890101402998, + "learning_rate": 0.25862487650161214, + "loss": 0.2969, + "num_input_tokens_seen": 20262768, + "step": 9690 + }, + { + "epoch": 1.5816135084427767, + "grad_norm": 0.0007052095024846494, + "learning_rate": 0.2585842458105106, + "loss": 0.3169, + "num_input_tokens_seen": 20273584, + "step": 9695 + }, + { + "epoch": 1.582429235663594, + "grad_norm": 0.0014938695821911097, + "learning_rate": 0.2585435983743538, + "loss": 0.3334, + "num_input_tokens_seen": 20285056, + "step": 9700 + }, + { + "epoch": 1.5832449628844114, + "grad_norm": 0.0011644480982795358, + "learning_rate": 0.2585029341994101, + "loss": 0.2976, + "num_input_tokens_seen": 20296000, + "step": 9705 + }, + { + "epoch": 1.5840606901052288, + "grad_norm": 0.0010557821951806545, + "learning_rate": 0.2584622532919504, + "loss": 0.3587, + "num_input_tokens_seen": 20306464, + "step": 9710 + }, + { + "epoch": 1.5848764173260461, + "grad_norm": 0.0011259742313995957, + "learning_rate": 0.2584215556582482, + "loss": 0.3816, + "num_input_tokens_seen": 20316320, + "step": 9715 + }, + { + "epoch": 1.5856921445468637, + "grad_norm": 0.002235927851870656, + "learning_rate": 0.25838084130457967, + "loss": 0.3149, + "num_input_tokens_seen": 20327488, + "step": 9720 + }, + { + "epoch": 1.5865078717676808, + "grad_norm": 0.001361423870548606, + "learning_rate": 0.2583401102372234, + "loss": 0.3503, + "num_input_tokens_seen": 20336832, + "step": 9725 + }, + { + "epoch": 1.5873235989884984, + "grad_norm": 0.001009644358418882, + "learning_rate": 0.2582993624624606, + "loss": 0.3165, + "num_input_tokens_seen": 20346288, + "step": 9730 + }, + { + "epoch": 1.5881393262093155, + "grad_norm": 0.0011860538506880403, + "learning_rate": 0.25825859798657513, + "loss": 0.3197, + "num_input_tokens_seen": 20356016, + "step": 9735 + }, + { + "epoch": 1.588955053430133, + "grad_norm": 0.0007378022419288754, + "learning_rate": 0.25821781681585343, + "loss": 0.3248, + "num_input_tokens_seen": 20368208, + "step": 9740 + }, + { + "epoch": 1.5897707806509502, + "grad_norm": 0.001620305934920907, + "learning_rate": 0.2581770189565844, + "loss": 0.3518, + "num_input_tokens_seen": 20379408, + "step": 9745 + }, + { + "epoch": 1.5905865078717678, + "grad_norm": 0.0006666569970548153, + "learning_rate": 0.25813620441505963, + "loss": 0.3591, + "num_input_tokens_seen": 20388864, + "step": 9750 + }, + { + "epoch": 1.591402235092585, + "grad_norm": 0.002633653813973069, + "learning_rate": 0.2580953731975732, + "loss": 0.3372, + "num_input_tokens_seen": 20399120, + "step": 9755 + }, + { + "epoch": 1.5922179623134025, + "grad_norm": 0.0007657923852093518, + "learning_rate": 0.2580545253104218, + "loss": 0.3373, + "num_input_tokens_seen": 20410464, + "step": 9760 + }, + { + "epoch": 1.5930336895342196, + "grad_norm": 0.0012323064729571342, + "learning_rate": 0.2580136607599047, + "loss": 0.3341, + "num_input_tokens_seen": 20420400, + "step": 9765 + }, + { + "epoch": 1.5938494167550372, + "grad_norm": 0.0007053759763948619, + "learning_rate": 0.2579727795523238, + "loss": 0.3853, + "num_input_tokens_seen": 20430512, + "step": 9770 + }, + { + "epoch": 1.5946651439758543, + "grad_norm": 0.0009026051266118884, + "learning_rate": 0.25793188169398334, + "loss": 0.3424, + "num_input_tokens_seen": 20440848, + "step": 9775 + }, + { + "epoch": 1.595480871196672, + "grad_norm": 0.00043024655315093696, + "learning_rate": 0.25789096719119037, + "loss": 0.3277, + "num_input_tokens_seen": 20451520, + "step": 9780 + }, + { + "epoch": 1.596296598417489, + "grad_norm": 0.0008795445319265127, + "learning_rate": 0.2578500360502544, + "loss": 0.3059, + "num_input_tokens_seen": 20461856, + "step": 9785 + }, + { + "epoch": 1.5971123256383066, + "grad_norm": 0.0005049572209827602, + "learning_rate": 0.2578090882774876, + "loss": 0.2888, + "num_input_tokens_seen": 20473648, + "step": 9790 + }, + { + "epoch": 1.5979280528591238, + "grad_norm": 0.0006494129193015397, + "learning_rate": 0.25776812387920456, + "loss": 0.2975, + "num_input_tokens_seen": 20482752, + "step": 9795 + }, + { + "epoch": 1.5987437800799413, + "grad_norm": 0.0007256320677697659, + "learning_rate": 0.2577271428617225, + "loss": 0.2611, + "num_input_tokens_seen": 20493344, + "step": 9800 + }, + { + "epoch": 1.5987437800799413, + "eval_loss": 0.3175891935825348, + "eval_runtime": 155.622, + "eval_samples_per_second": 17.51, + "eval_steps_per_second": 8.758, + "num_input_tokens_seen": 20493344, + "step": 9800 + }, + { + "epoch": 1.5995595073007585, + "grad_norm": 0.0004673349321819842, + "learning_rate": 0.25768614523136124, + "loss": 0.2769, + "num_input_tokens_seen": 20505264, + "step": 9805 + }, + { + "epoch": 1.600375234521576, + "grad_norm": 0.0012637930922210217, + "learning_rate": 0.25764513099444314, + "loss": 0.3081, + "num_input_tokens_seen": 20515632, + "step": 9810 + }, + { + "epoch": 1.6011909617423934, + "grad_norm": 0.0007955404580570757, + "learning_rate": 0.25760410015729307, + "loss": 0.3174, + "num_input_tokens_seen": 20526848, + "step": 9815 + }, + { + "epoch": 1.6020066889632107, + "grad_norm": 0.002286785515025258, + "learning_rate": 0.2575630527262385, + "loss": 0.3249, + "num_input_tokens_seen": 20536416, + "step": 9820 + }, + { + "epoch": 1.602822416184028, + "grad_norm": 0.0021824983414262533, + "learning_rate": 0.25752198870760945, + "loss": 0.4187, + "num_input_tokens_seen": 20547008, + "step": 9825 + }, + { + "epoch": 1.6036381434048455, + "grad_norm": 0.0006743657868355513, + "learning_rate": 0.2574809081077386, + "loss": 0.2875, + "num_input_tokens_seen": 20558064, + "step": 9830 + }, + { + "epoch": 1.6044538706256628, + "grad_norm": 0.00050344382179901, + "learning_rate": 0.257439810932961, + "loss": 0.3024, + "num_input_tokens_seen": 20568512, + "step": 9835 + }, + { + "epoch": 1.6052695978464802, + "grad_norm": 0.0005890533211641014, + "learning_rate": 0.2573986971896144, + "loss": 0.3262, + "num_input_tokens_seen": 20578256, + "step": 9840 + }, + { + "epoch": 1.6060853250672975, + "grad_norm": 0.0012423120206221938, + "learning_rate": 0.257357566884039, + "loss": 0.3496, + "num_input_tokens_seen": 20589344, + "step": 9845 + }, + { + "epoch": 1.6069010522881149, + "grad_norm": 0.0010050638811662793, + "learning_rate": 0.25731642002257765, + "loss": 0.3542, + "num_input_tokens_seen": 20600960, + "step": 9850 + }, + { + "epoch": 1.6077167795089322, + "grad_norm": 0.0010786553611978889, + "learning_rate": 0.25727525661157574, + "loss": 0.2673, + "num_input_tokens_seen": 20611424, + "step": 9855 + }, + { + "epoch": 1.6085325067297496, + "grad_norm": 0.0006611315184272826, + "learning_rate": 0.2572340766573811, + "loss": 0.2752, + "num_input_tokens_seen": 20621824, + "step": 9860 + }, + { + "epoch": 1.609348233950567, + "grad_norm": 0.0009005812462419271, + "learning_rate": 0.25719288016634434, + "loss": 0.2928, + "num_input_tokens_seen": 20631264, + "step": 9865 + }, + { + "epoch": 1.6101639611713843, + "grad_norm": 0.0009114118292927742, + "learning_rate": 0.25715166714481835, + "loss": 0.2762, + "num_input_tokens_seen": 20642096, + "step": 9870 + }, + { + "epoch": 1.6109796883922016, + "grad_norm": 0.0007952269515953958, + "learning_rate": 0.2571104375991587, + "loss": 0.2744, + "num_input_tokens_seen": 20652976, + "step": 9875 + }, + { + "epoch": 1.611795415613019, + "grad_norm": 0.001350030885078013, + "learning_rate": 0.2570691915357236, + "loss": 0.2382, + "num_input_tokens_seen": 20661920, + "step": 9880 + }, + { + "epoch": 1.6126111428338363, + "grad_norm": 0.002883619163185358, + "learning_rate": 0.2570279289608736, + "loss": 0.4609, + "num_input_tokens_seen": 20672336, + "step": 9885 + }, + { + "epoch": 1.6134268700546537, + "grad_norm": 0.0009797693928703666, + "learning_rate": 0.256986649880972, + "loss": 0.2722, + "num_input_tokens_seen": 20684448, + "step": 9890 + }, + { + "epoch": 1.614242597275471, + "grad_norm": 0.001720051746815443, + "learning_rate": 0.25694535430238447, + "loss": 0.4245, + "num_input_tokens_seen": 20695248, + "step": 9895 + }, + { + "epoch": 1.6150583244962884, + "grad_norm": 0.0005445706774480641, + "learning_rate": 0.25690404223147933, + "loss": 0.2967, + "num_input_tokens_seen": 20705488, + "step": 9900 + }, + { + "epoch": 1.615874051717106, + "grad_norm": 0.00106249435339123, + "learning_rate": 0.2568627136746275, + "loss": 0.2564, + "num_input_tokens_seen": 20715120, + "step": 9905 + }, + { + "epoch": 1.616689778937923, + "grad_norm": 0.0017738831229507923, + "learning_rate": 0.25682136863820226, + "loss": 0.3249, + "num_input_tokens_seen": 20724624, + "step": 9910 + }, + { + "epoch": 1.6175055061587407, + "grad_norm": 0.0006415793322958052, + "learning_rate": 0.25678000712857957, + "loss": 0.3035, + "num_input_tokens_seen": 20735664, + "step": 9915 + }, + { + "epoch": 1.6183212333795578, + "grad_norm": 0.001987265655770898, + "learning_rate": 0.2567386291521379, + "loss": 0.4157, + "num_input_tokens_seen": 20745104, + "step": 9920 + }, + { + "epoch": 1.6191369606003754, + "grad_norm": 0.0006328831659629941, + "learning_rate": 0.2566972347152583, + "loss": 0.4121, + "num_input_tokens_seen": 20755632, + "step": 9925 + }, + { + "epoch": 1.6199526878211925, + "grad_norm": 0.0004994075861759484, + "learning_rate": 0.2566558238243242, + "loss": 0.3485, + "num_input_tokens_seen": 20766304, + "step": 9930 + }, + { + "epoch": 1.62076841504201, + "grad_norm": 0.0006246473058126867, + "learning_rate": 0.25661439648572176, + "loss": 0.3452, + "num_input_tokens_seen": 20777712, + "step": 9935 + }, + { + "epoch": 1.6215841422628272, + "grad_norm": 0.0021843318827450275, + "learning_rate": 0.25657295270583963, + "loss": 0.2958, + "num_input_tokens_seen": 20788000, + "step": 9940 + }, + { + "epoch": 1.6223998694836448, + "grad_norm": 0.0011959514813497663, + "learning_rate": 0.25653149249106894, + "loss": 0.3152, + "num_input_tokens_seen": 20798720, + "step": 9945 + }, + { + "epoch": 1.623215596704462, + "grad_norm": 0.00057521479902789, + "learning_rate": 0.25649001584780323, + "loss": 0.3289, + "num_input_tokens_seen": 20808704, + "step": 9950 + }, + { + "epoch": 1.6240313239252795, + "grad_norm": 0.0011088292812928557, + "learning_rate": 0.2564485227824389, + "loss": 0.3463, + "num_input_tokens_seen": 20818848, + "step": 9955 + }, + { + "epoch": 1.6248470511460966, + "grad_norm": 0.0014528913889080286, + "learning_rate": 0.25640701330137466, + "loss": 0.3166, + "num_input_tokens_seen": 20828304, + "step": 9960 + }, + { + "epoch": 1.6256627783669142, + "grad_norm": 0.000800382171291858, + "learning_rate": 0.2563654874110117, + "loss": 0.2914, + "num_input_tokens_seen": 20837840, + "step": 9965 + }, + { + "epoch": 1.6264785055877313, + "grad_norm": 0.0008823578828014433, + "learning_rate": 0.256323945117754, + "loss": 0.3497, + "num_input_tokens_seen": 20848608, + "step": 9970 + }, + { + "epoch": 1.627294232808549, + "grad_norm": 0.0010887474054470658, + "learning_rate": 0.2562823864280078, + "loss": 0.2929, + "num_input_tokens_seen": 20859616, + "step": 9975 + }, + { + "epoch": 1.628109960029366, + "grad_norm": 0.0008019572705961764, + "learning_rate": 0.25624081134818194, + "loss": 0.3537, + "num_input_tokens_seen": 20869984, + "step": 9980 + }, + { + "epoch": 1.6289256872501836, + "grad_norm": 0.0015942577738314867, + "learning_rate": 0.2561992198846879, + "loss": 0.3382, + "num_input_tokens_seen": 20880448, + "step": 9985 + }, + { + "epoch": 1.6297414144710007, + "grad_norm": 0.0006546929944306612, + "learning_rate": 0.25615761204393955, + "loss": 0.3369, + "num_input_tokens_seen": 20890480, + "step": 9990 + }, + { + "epoch": 1.6305571416918183, + "grad_norm": 0.0007179654203355312, + "learning_rate": 0.2561159878323534, + "loss": 0.3557, + "num_input_tokens_seen": 20902416, + "step": 9995 + }, + { + "epoch": 1.6313728689126357, + "grad_norm": 0.0006927962531335652, + "learning_rate": 0.2560743472563483, + "loss": 0.3499, + "num_input_tokens_seen": 20912896, + "step": 10000 + }, + { + "epoch": 1.6313728689126357, + "eval_loss": 0.32187598943710327, + "eval_runtime": 155.9809, + "eval_samples_per_second": 17.47, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 20912896, + "step": 10000 + }, + { + "epoch": 1.632188596133453, + "grad_norm": 0.000853546429425478, + "learning_rate": 0.25603269032234593, + "loss": 0.3223, + "num_input_tokens_seen": 20923056, + "step": 10005 + }, + { + "epoch": 1.6330043233542704, + "grad_norm": 0.0014838615898042917, + "learning_rate": 0.2559910170367702, + "loss": 0.3532, + "num_input_tokens_seen": 20934336, + "step": 10010 + }, + { + "epoch": 1.6338200505750877, + "grad_norm": 0.00048771800356917083, + "learning_rate": 0.2559493274060477, + "loss": 0.3023, + "num_input_tokens_seen": 20945440, + "step": 10015 + }, + { + "epoch": 1.634635777795905, + "grad_norm": 0.000578461040277034, + "learning_rate": 0.2559076214366074, + "loss": 0.3123, + "num_input_tokens_seen": 20956544, + "step": 10020 + }, + { + "epoch": 1.6354515050167224, + "grad_norm": 0.0005923123680986464, + "learning_rate": 0.25586589913488106, + "loss": 0.324, + "num_input_tokens_seen": 20966560, + "step": 10025 + }, + { + "epoch": 1.6362672322375398, + "grad_norm": 0.0017068808665499091, + "learning_rate": 0.2558241605073026, + "loss": 0.3523, + "num_input_tokens_seen": 20976784, + "step": 10030 + }, + { + "epoch": 1.6370829594583571, + "grad_norm": 0.0005362624069675803, + "learning_rate": 0.25578240556030873, + "loss": 0.3086, + "num_input_tokens_seen": 20988256, + "step": 10035 + }, + { + "epoch": 1.6378986866791745, + "grad_norm": 0.0006033041281625628, + "learning_rate": 0.2557406343003386, + "loss": 0.3832, + "num_input_tokens_seen": 20997648, + "step": 10040 + }, + { + "epoch": 1.6387144138999918, + "grad_norm": 0.0006861377623863518, + "learning_rate": 0.25569884673383375, + "loss": 0.3221, + "num_input_tokens_seen": 21008480, + "step": 10045 + }, + { + "epoch": 1.6395301411208092, + "grad_norm": 0.0009015821269713342, + "learning_rate": 0.25565704286723856, + "loss": 0.3149, + "num_input_tokens_seen": 21018384, + "step": 10050 + }, + { + "epoch": 1.6403458683416265, + "grad_norm": 0.0007272171787917614, + "learning_rate": 0.25561522270699955, + "loss": 0.3005, + "num_input_tokens_seen": 21029424, + "step": 10055 + }, + { + "epoch": 1.641161595562444, + "grad_norm": 0.00026144200819544494, + "learning_rate": 0.25557338625956594, + "loss": 0.3313, + "num_input_tokens_seen": 21039936, + "step": 10060 + }, + { + "epoch": 1.6419773227832613, + "grad_norm": 0.0005230159731581807, + "learning_rate": 0.25553153353138947, + "loss": 0.3067, + "num_input_tokens_seen": 21049904, + "step": 10065 + }, + { + "epoch": 1.6427930500040786, + "grad_norm": 0.0010142180835828185, + "learning_rate": 0.2554896645289243, + "loss": 0.3455, + "num_input_tokens_seen": 21059312, + "step": 10070 + }, + { + "epoch": 1.643608777224896, + "grad_norm": 0.0006453785463236272, + "learning_rate": 0.2554477792586272, + "loss": 0.3165, + "num_input_tokens_seen": 21068816, + "step": 10075 + }, + { + "epoch": 1.6444245044457133, + "grad_norm": 0.0010698041878640652, + "learning_rate": 0.25540587772695744, + "loss": 0.3195, + "num_input_tokens_seen": 21080128, + "step": 10080 + }, + { + "epoch": 1.6452402316665307, + "grad_norm": 0.0005176096456125379, + "learning_rate": 0.2553639599403767, + "loss": 0.3696, + "num_input_tokens_seen": 21090800, + "step": 10085 + }, + { + "epoch": 1.6460559588873482, + "grad_norm": 0.0005302694044075906, + "learning_rate": 0.2553220259053493, + "loss": 0.3496, + "num_input_tokens_seen": 21101744, + "step": 10090 + }, + { + "epoch": 1.6468716861081654, + "grad_norm": 0.00044411036651581526, + "learning_rate": 0.2552800756283419, + "loss": 0.2707, + "num_input_tokens_seen": 21112320, + "step": 10095 + }, + { + "epoch": 1.647687413328983, + "grad_norm": 0.0007831835537217557, + "learning_rate": 0.25523810911582373, + "loss": 0.3173, + "num_input_tokens_seen": 21123376, + "step": 10100 + }, + { + "epoch": 1.6485031405498, + "grad_norm": 0.0012543441262096167, + "learning_rate": 0.25519612637426675, + "loss": 0.3108, + "num_input_tokens_seen": 21135104, + "step": 10105 + }, + { + "epoch": 1.6493188677706176, + "grad_norm": 0.0005784293753094971, + "learning_rate": 0.25515412741014504, + "loss": 0.3366, + "num_input_tokens_seen": 21145744, + "step": 10110 + }, + { + "epoch": 1.6501345949914348, + "grad_norm": 0.0006368635804392397, + "learning_rate": 0.2551121122299355, + "loss": 0.2852, + "num_input_tokens_seen": 21157488, + "step": 10115 + }, + { + "epoch": 1.6509503222122524, + "grad_norm": 0.0007263730512931943, + "learning_rate": 0.2550700808401173, + "loss": 0.3329, + "num_input_tokens_seen": 21167120, + "step": 10120 + }, + { + "epoch": 1.6517660494330695, + "grad_norm": 0.0009371183114126325, + "learning_rate": 0.2550280332471722, + "loss": 0.3558, + "num_input_tokens_seen": 21178704, + "step": 10125 + }, + { + "epoch": 1.652581776653887, + "grad_norm": 0.0005666097276844084, + "learning_rate": 0.2549859694575845, + "loss": 0.3105, + "num_input_tokens_seen": 21188640, + "step": 10130 + }, + { + "epoch": 1.6533975038747042, + "grad_norm": 0.0005598050192929804, + "learning_rate": 0.254943889477841, + "loss": 0.297, + "num_input_tokens_seen": 21199312, + "step": 10135 + }, + { + "epoch": 1.6542132310955218, + "grad_norm": 0.0013801170280203223, + "learning_rate": 0.25490179331443097, + "loss": 0.3223, + "num_input_tokens_seen": 21209824, + "step": 10140 + }, + { + "epoch": 1.655028958316339, + "grad_norm": 0.0022578793577849865, + "learning_rate": 0.25485968097384615, + "loss": 0.315, + "num_input_tokens_seen": 21220832, + "step": 10145 + }, + { + "epoch": 1.6558446855371565, + "grad_norm": 0.001153341494500637, + "learning_rate": 0.25481755246258075, + "loss": 0.3787, + "num_input_tokens_seen": 21230928, + "step": 10150 + }, + { + "epoch": 1.6566604127579736, + "grad_norm": 0.0014163325540721416, + "learning_rate": 0.2547754077871315, + "loss": 0.347, + "num_input_tokens_seen": 21240928, + "step": 10155 + }, + { + "epoch": 1.6574761399787912, + "grad_norm": 0.0008230687235482037, + "learning_rate": 0.25473324695399774, + "loss": 0.2863, + "num_input_tokens_seen": 21251280, + "step": 10160 + }, + { + "epoch": 1.6582918671996083, + "grad_norm": 0.0005906233564019203, + "learning_rate": 0.25469106996968105, + "loss": 0.3543, + "num_input_tokens_seen": 21261072, + "step": 10165 + }, + { + "epoch": 1.6591075944204259, + "grad_norm": 0.00046589208068326116, + "learning_rate": 0.2546488768406858, + "loss": 0.3259, + "num_input_tokens_seen": 21270416, + "step": 10170 + }, + { + "epoch": 1.659923321641243, + "grad_norm": 0.000994556350633502, + "learning_rate": 0.25460666757351863, + "loss": 0.3461, + "num_input_tokens_seen": 21280976, + "step": 10175 + }, + { + "epoch": 1.6607390488620606, + "grad_norm": 0.000574135803617537, + "learning_rate": 0.25456444217468877, + "loss": 0.2666, + "num_input_tokens_seen": 21290304, + "step": 10180 + }, + { + "epoch": 1.661554776082878, + "grad_norm": 0.0007760073058307171, + "learning_rate": 0.25452220065070785, + "loss": 0.3232, + "num_input_tokens_seen": 21299504, + "step": 10185 + }, + { + "epoch": 1.6623705033036953, + "grad_norm": 0.0015180453192442656, + "learning_rate": 0.2544799430080901, + "loss": 0.3046, + "num_input_tokens_seen": 21308288, + "step": 10190 + }, + { + "epoch": 1.6631862305245126, + "grad_norm": 0.001063089119270444, + "learning_rate": 0.2544376692533522, + "loss": 0.2888, + "num_input_tokens_seen": 21318704, + "step": 10195 + }, + { + "epoch": 1.66400195774533, + "grad_norm": 0.0004754812107421458, + "learning_rate": 0.2543953793930132, + "loss": 0.3144, + "num_input_tokens_seen": 21328976, + "step": 10200 + }, + { + "epoch": 1.66400195774533, + "eval_loss": 0.32003873586654663, + "eval_runtime": 155.6412, + "eval_samples_per_second": 17.508, + "eval_steps_per_second": 8.757, + "num_input_tokens_seen": 21328976, + "step": 10200 + }, + { + "epoch": 1.6648176849661473, + "grad_norm": 0.0022679278627038, + "learning_rate": 0.2543530734335948, + "loss": 0.3197, + "num_input_tokens_seen": 21339840, + "step": 10205 + }, + { + "epoch": 1.6656334121869647, + "grad_norm": 0.0018980265595018864, + "learning_rate": 0.2543107513816211, + "loss": 0.3544, + "num_input_tokens_seen": 21350368, + "step": 10210 + }, + { + "epoch": 1.666449139407782, + "grad_norm": 0.000265380454948172, + "learning_rate": 0.25426841324361865, + "loss": 0.3225, + "num_input_tokens_seen": 21360736, + "step": 10215 + }, + { + "epoch": 1.6672648666285994, + "grad_norm": 0.0012989231618121266, + "learning_rate": 0.2542260590261166, + "loss": 0.3917, + "num_input_tokens_seen": 21372096, + "step": 10220 + }, + { + "epoch": 1.6680805938494168, + "grad_norm": 0.0004311405646149069, + "learning_rate": 0.2541836887356465, + "loss": 0.3065, + "num_input_tokens_seen": 21382240, + "step": 10225 + }, + { + "epoch": 1.6688963210702341, + "grad_norm": 0.0008375413017347455, + "learning_rate": 0.2541413023787423, + "loss": 0.3224, + "num_input_tokens_seen": 21392864, + "step": 10230 + }, + { + "epoch": 1.6697120482910515, + "grad_norm": 0.0006203499506227672, + "learning_rate": 0.2540988999619405, + "loss": 0.3589, + "num_input_tokens_seen": 21403008, + "step": 10235 + }, + { + "epoch": 1.6705277755118688, + "grad_norm": 0.0005162979941815138, + "learning_rate": 0.25405648149178023, + "loss": 0.3172, + "num_input_tokens_seen": 21414048, + "step": 10240 + }, + { + "epoch": 1.6713435027326862, + "grad_norm": 0.0004551886231638491, + "learning_rate": 0.2540140469748028, + "loss": 0.3189, + "num_input_tokens_seen": 21425792, + "step": 10245 + }, + { + "epoch": 1.6721592299535035, + "grad_norm": 0.0013125762343406677, + "learning_rate": 0.25397159641755224, + "loss": 0.3337, + "num_input_tokens_seen": 21434736, + "step": 10250 + }, + { + "epoch": 1.6729749571743209, + "grad_norm": 0.0007216216763481498, + "learning_rate": 0.2539291298265749, + "loss": 0.3375, + "num_input_tokens_seen": 21445984, + "step": 10255 + }, + { + "epoch": 1.6737906843951382, + "grad_norm": 0.0006825675955042243, + "learning_rate": 0.2538866472084197, + "loss": 0.3544, + "num_input_tokens_seen": 21457440, + "step": 10260 + }, + { + "epoch": 1.6746064116159556, + "grad_norm": 0.0003526957880239934, + "learning_rate": 0.25384414856963794, + "loss": 0.3636, + "num_input_tokens_seen": 21468112, + "step": 10265 + }, + { + "epoch": 1.675422138836773, + "grad_norm": 0.0005670436657965183, + "learning_rate": 0.25380163391678356, + "loss": 0.3178, + "num_input_tokens_seen": 21478912, + "step": 10270 + }, + { + "epoch": 1.6762378660575903, + "grad_norm": 0.0007987121935002506, + "learning_rate": 0.2537591032564127, + "loss": 0.292, + "num_input_tokens_seen": 21489264, + "step": 10275 + }, + { + "epoch": 1.6770535932784076, + "grad_norm": 0.0004865903756581247, + "learning_rate": 0.25371655659508424, + "loss": 0.3156, + "num_input_tokens_seen": 21500944, + "step": 10280 + }, + { + "epoch": 1.6778693204992252, + "grad_norm": 0.001422533649019897, + "learning_rate": 0.25367399393935935, + "loss": 0.3356, + "num_input_tokens_seen": 21510848, + "step": 10285 + }, + { + "epoch": 1.6786850477200423, + "grad_norm": 0.0005163936875760555, + "learning_rate": 0.25363141529580174, + "loss": 0.3192, + "num_input_tokens_seen": 21520384, + "step": 10290 + }, + { + "epoch": 1.67950077494086, + "grad_norm": 0.0005504917935468256, + "learning_rate": 0.2535888206709776, + "loss": 0.3319, + "num_input_tokens_seen": 21531600, + "step": 10295 + }, + { + "epoch": 1.680316502161677, + "grad_norm": 0.00043824620661325753, + "learning_rate": 0.2535462100714555, + "loss": 0.306, + "num_input_tokens_seen": 21541760, + "step": 10300 + }, + { + "epoch": 1.6811322293824946, + "grad_norm": 0.0005514112999662757, + "learning_rate": 0.2535035835038066, + "loss": 0.3924, + "num_input_tokens_seen": 21552000, + "step": 10305 + }, + { + "epoch": 1.6819479566033118, + "grad_norm": 0.0008602272137068212, + "learning_rate": 0.2534609409746044, + "loss": 0.3431, + "num_input_tokens_seen": 21562416, + "step": 10310 + }, + { + "epoch": 1.6827636838241293, + "grad_norm": 0.002067946596071124, + "learning_rate": 0.253418282490425, + "loss": 0.3066, + "num_input_tokens_seen": 21572656, + "step": 10315 + }, + { + "epoch": 1.6835794110449465, + "grad_norm": 0.001091286656446755, + "learning_rate": 0.2533756080578467, + "loss": 0.2759, + "num_input_tokens_seen": 21583456, + "step": 10320 + }, + { + "epoch": 1.684395138265764, + "grad_norm": 0.0008767139515839517, + "learning_rate": 0.25333291768345056, + "loss": 0.408, + "num_input_tokens_seen": 21593696, + "step": 10325 + }, + { + "epoch": 1.6852108654865812, + "grad_norm": 0.0004360186867415905, + "learning_rate": 0.25329021137381996, + "loss": 0.3422, + "num_input_tokens_seen": 21602832, + "step": 10330 + }, + { + "epoch": 1.6860265927073987, + "grad_norm": 0.0006096746656112373, + "learning_rate": 0.25324748913554074, + "loss": 0.3167, + "num_input_tokens_seen": 21613760, + "step": 10335 + }, + { + "epoch": 1.6868423199282159, + "grad_norm": 0.0009129969985224307, + "learning_rate": 0.2532047509752013, + "loss": 0.3677, + "num_input_tokens_seen": 21622576, + "step": 10340 + }, + { + "epoch": 1.6876580471490334, + "grad_norm": 0.00047754525439813733, + "learning_rate": 0.25316199689939217, + "loss": 0.324, + "num_input_tokens_seen": 21634032, + "step": 10345 + }, + { + "epoch": 1.6884737743698506, + "grad_norm": 0.00047278444981202483, + "learning_rate": 0.2531192269147068, + "loss": 0.3429, + "num_input_tokens_seen": 21644592, + "step": 10350 + }, + { + "epoch": 1.6892895015906682, + "grad_norm": 0.0006230353028513491, + "learning_rate": 0.2530764410277407, + "loss": 0.2754, + "num_input_tokens_seen": 21655552, + "step": 10355 + }, + { + "epoch": 1.6901052288114853, + "grad_norm": 0.0011450143065303564, + "learning_rate": 0.25303363924509203, + "loss": 0.2805, + "num_input_tokens_seen": 21666960, + "step": 10360 + }, + { + "epoch": 1.6909209560323029, + "grad_norm": 0.00094487244496122, + "learning_rate": 0.25299082157336145, + "loss": 0.3711, + "num_input_tokens_seen": 21676608, + "step": 10365 + }, + { + "epoch": 1.6917366832531202, + "grad_norm": 0.0008095004013739526, + "learning_rate": 0.2529479880191519, + "loss": 0.3522, + "num_input_tokens_seen": 21686576, + "step": 10370 + }, + { + "epoch": 1.6925524104739376, + "grad_norm": 0.0009259566431865096, + "learning_rate": 0.2529051385890689, + "loss": 0.3502, + "num_input_tokens_seen": 21697104, + "step": 10375 + }, + { + "epoch": 1.693368137694755, + "grad_norm": 0.0016488739056512713, + "learning_rate": 0.2528622732897203, + "loss": 0.3254, + "num_input_tokens_seen": 21707376, + "step": 10380 + }, + { + "epoch": 1.6941838649155723, + "grad_norm": 0.00045551115181297064, + "learning_rate": 0.25281939212771654, + "loss": 0.3198, + "num_input_tokens_seen": 21718848, + "step": 10385 + }, + { + "epoch": 1.6949995921363896, + "grad_norm": 0.00040812708903104067, + "learning_rate": 0.2527764951096704, + "loss": 0.3254, + "num_input_tokens_seen": 21729968, + "step": 10390 + }, + { + "epoch": 1.695815319357207, + "grad_norm": 0.0007316325209103525, + "learning_rate": 0.2527335822421971, + "loss": 0.3234, + "num_input_tokens_seen": 21740832, + "step": 10395 + }, + { + "epoch": 1.6966310465780243, + "grad_norm": 0.0008116631070151925, + "learning_rate": 0.25269065353191444, + "loss": 0.3449, + "num_input_tokens_seen": 21752192, + "step": 10400 + }, + { + "epoch": 1.6966310465780243, + "eval_loss": 0.31743690371513367, + "eval_runtime": 155.6959, + "eval_samples_per_second": 17.502, + "eval_steps_per_second": 8.754, + "num_input_tokens_seen": 21752192, + "step": 10400 + }, + { + "epoch": 1.6974467737988417, + "grad_norm": 0.0015592441195622087, + "learning_rate": 0.2526477089854425, + "loss": 0.3362, + "num_input_tokens_seen": 21761344, + "step": 10405 + }, + { + "epoch": 1.698262501019659, + "grad_norm": 0.0005099293775856495, + "learning_rate": 0.25260474860940385, + "loss": 0.3238, + "num_input_tokens_seen": 21770720, + "step": 10410 + }, + { + "epoch": 1.6990782282404764, + "grad_norm": 0.0006699920631945133, + "learning_rate": 0.2525617724104236, + "loss": 0.3162, + "num_input_tokens_seen": 21780208, + "step": 10415 + }, + { + "epoch": 1.6998939554612937, + "grad_norm": 0.0018320067320019007, + "learning_rate": 0.25251878039512915, + "loss": 0.3651, + "num_input_tokens_seen": 21790624, + "step": 10420 + }, + { + "epoch": 1.700709682682111, + "grad_norm": 0.0010041018249467015, + "learning_rate": 0.25247577257015047, + "loss": 0.3732, + "num_input_tokens_seen": 21801008, + "step": 10425 + }, + { + "epoch": 1.7015254099029284, + "grad_norm": 0.000622111838310957, + "learning_rate": 0.2524327489421198, + "loss": 0.2854, + "num_input_tokens_seen": 21812656, + "step": 10430 + }, + { + "epoch": 1.7023411371237458, + "grad_norm": 0.0009403816075064242, + "learning_rate": 0.25238970951767203, + "loss": 0.3389, + "num_input_tokens_seen": 21824000, + "step": 10435 + }, + { + "epoch": 1.7031568643445631, + "grad_norm": 0.000692906673066318, + "learning_rate": 0.25234665430344433, + "loss": 0.3402, + "num_input_tokens_seen": 21834976, + "step": 10440 + }, + { + "epoch": 1.7039725915653805, + "grad_norm": 0.0007921420037746429, + "learning_rate": 0.2523035833060764, + "loss": 0.3472, + "num_input_tokens_seen": 21845344, + "step": 10445 + }, + { + "epoch": 1.7047883187861979, + "grad_norm": 0.0005106896278448403, + "learning_rate": 0.2522604965322103, + "loss": 0.3197, + "num_input_tokens_seen": 21856944, + "step": 10450 + }, + { + "epoch": 1.7056040460070152, + "grad_norm": 0.0007674300577491522, + "learning_rate": 0.25221739398849047, + "loss": 0.287, + "num_input_tokens_seen": 21867856, + "step": 10455 + }, + { + "epoch": 1.7064197732278326, + "grad_norm": 0.0006124316714704037, + "learning_rate": 0.252174275681564, + "loss": 0.3662, + "num_input_tokens_seen": 21877504, + "step": 10460 + }, + { + "epoch": 1.70723550044865, + "grad_norm": 0.0014153228839859366, + "learning_rate": 0.2521311416180802, + "loss": 0.3491, + "num_input_tokens_seen": 21886352, + "step": 10465 + }, + { + "epoch": 1.7080512276694675, + "grad_norm": 0.0002911556512117386, + "learning_rate": 0.25208799180469094, + "loss": 0.3845, + "num_input_tokens_seen": 21896976, + "step": 10470 + }, + { + "epoch": 1.7088669548902846, + "grad_norm": 0.0005009690648876131, + "learning_rate": 0.2520448262480504, + "loss": 0.2878, + "num_input_tokens_seen": 21906160, + "step": 10475 + }, + { + "epoch": 1.7096826821111022, + "grad_norm": 0.0005089633632451296, + "learning_rate": 0.25200164495481525, + "loss": 0.3299, + "num_input_tokens_seen": 21915792, + "step": 10480 + }, + { + "epoch": 1.7104984093319193, + "grad_norm": 0.00038709526415914297, + "learning_rate": 0.25195844793164474, + "loss": 0.3317, + "num_input_tokens_seen": 21926080, + "step": 10485 + }, + { + "epoch": 1.711314136552737, + "grad_norm": 0.0007596178911626339, + "learning_rate": 0.2519152351852001, + "loss": 0.336, + "num_input_tokens_seen": 21936656, + "step": 10490 + }, + { + "epoch": 1.712129863773554, + "grad_norm": 0.0003471379750408232, + "learning_rate": 0.25187200672214555, + "loss": 0.3252, + "num_input_tokens_seen": 21947488, + "step": 10495 + }, + { + "epoch": 1.7129455909943716, + "grad_norm": 0.0005008327425457537, + "learning_rate": 0.2518287625491473, + "loss": 0.3424, + "num_input_tokens_seen": 21957760, + "step": 10500 + }, + { + "epoch": 1.7137613182151887, + "grad_norm": 0.000405204453272745, + "learning_rate": 0.25178550267287425, + "loss": 0.3889, + "num_input_tokens_seen": 21967472, + "step": 10505 + }, + { + "epoch": 1.7145770454360063, + "grad_norm": 0.0009730899473652244, + "learning_rate": 0.2517422270999976, + "loss": 0.3314, + "num_input_tokens_seen": 21977568, + "step": 10510 + }, + { + "epoch": 1.7153927726568234, + "grad_norm": 0.0009918417781591415, + "learning_rate": 0.2516989358371909, + "loss": 0.3261, + "num_input_tokens_seen": 21988224, + "step": 10515 + }, + { + "epoch": 1.716208499877641, + "grad_norm": 0.00046357393148355186, + "learning_rate": 0.25165562889113025, + "loss": 0.3622, + "num_input_tokens_seen": 21998032, + "step": 10520 + }, + { + "epoch": 1.7170242270984581, + "grad_norm": 0.000349667010596022, + "learning_rate": 0.2516123062684942, + "loss": 0.3063, + "num_input_tokens_seen": 22008800, + "step": 10525 + }, + { + "epoch": 1.7178399543192757, + "grad_norm": 0.0011746804229915142, + "learning_rate": 0.25156896797596356, + "loss": 0.2939, + "num_input_tokens_seen": 22020000, + "step": 10530 + }, + { + "epoch": 1.7186556815400928, + "grad_norm": 0.0005617734859697521, + "learning_rate": 0.2515256140202216, + "loss": 0.2903, + "num_input_tokens_seen": 22031712, + "step": 10535 + }, + { + "epoch": 1.7194714087609104, + "grad_norm": 0.00047782392357476056, + "learning_rate": 0.25148224440795425, + "loss": 0.3586, + "num_input_tokens_seen": 22042352, + "step": 10540 + }, + { + "epoch": 1.7202871359817276, + "grad_norm": 0.0010298186680302024, + "learning_rate": 0.2514388591458494, + "loss": 0.3573, + "num_input_tokens_seen": 22054064, + "step": 10545 + }, + { + "epoch": 1.7211028632025451, + "grad_norm": 0.0010423911735415459, + "learning_rate": 0.2513954582405977, + "loss": 0.3446, + "num_input_tokens_seen": 22064368, + "step": 10550 + }, + { + "epoch": 1.7219185904233625, + "grad_norm": 0.00082802539691329, + "learning_rate": 0.2513520416988922, + "loss": 0.3241, + "num_input_tokens_seen": 22074064, + "step": 10555 + }, + { + "epoch": 1.7227343176441798, + "grad_norm": 0.0008749370463192463, + "learning_rate": 0.2513086095274281, + "loss": 0.2756, + "num_input_tokens_seen": 22084864, + "step": 10560 + }, + { + "epoch": 1.7235500448649972, + "grad_norm": 0.00043557476601563394, + "learning_rate": 0.25126516173290336, + "loss": 0.3358, + "num_input_tokens_seen": 22094544, + "step": 10565 + }, + { + "epoch": 1.7243657720858145, + "grad_norm": 0.0010163617553189397, + "learning_rate": 0.2512216983220181, + "loss": 0.2909, + "num_input_tokens_seen": 22102656, + "step": 10570 + }, + { + "epoch": 1.725181499306632, + "grad_norm": 0.0011209632502868772, + "learning_rate": 0.25117821930147494, + "loss": 0.3115, + "num_input_tokens_seen": 22112976, + "step": 10575 + }, + { + "epoch": 1.7259972265274492, + "grad_norm": 0.00042157541611231863, + "learning_rate": 0.2511347246779788, + "loss": 0.34, + "num_input_tokens_seen": 22123936, + "step": 10580 + }, + { + "epoch": 1.7268129537482666, + "grad_norm": 0.0003993313293904066, + "learning_rate": 0.25109121445823723, + "loss": 0.2837, + "num_input_tokens_seen": 22134736, + "step": 10585 + }, + { + "epoch": 1.727628680969084, + "grad_norm": 0.0010007658274844289, + "learning_rate": 0.25104768864896004, + "loss": 0.33, + "num_input_tokens_seen": 22144736, + "step": 10590 + }, + { + "epoch": 1.7284444081899013, + "grad_norm": 0.0009941380703821778, + "learning_rate": 0.2510041472568594, + "loss": 0.3126, + "num_input_tokens_seen": 22154656, + "step": 10595 + }, + { + "epoch": 1.7292601354107187, + "grad_norm": 0.0004776909772772342, + "learning_rate": 0.25096059028864987, + "loss": 0.2893, + "num_input_tokens_seen": 22164912, + "step": 10600 + }, + { + "epoch": 1.7292601354107187, + "eval_loss": 0.3173571228981018, + "eval_runtime": 155.7381, + "eval_samples_per_second": 17.497, + "eval_steps_per_second": 8.752, + "num_input_tokens_seen": 22164912, + "step": 10600 + }, + { + "epoch": 1.730075862631536, + "grad_norm": 0.0005753715522587299, + "learning_rate": 0.25091701775104863, + "loss": 0.4398, + "num_input_tokens_seen": 22174560, + "step": 10605 + }, + { + "epoch": 1.7308915898523534, + "grad_norm": 0.001011297688819468, + "learning_rate": 0.250873429650775, + "loss": 0.262, + "num_input_tokens_seen": 22185552, + "step": 10610 + }, + { + "epoch": 1.7317073170731707, + "grad_norm": 0.0006373569485731423, + "learning_rate": 0.25082982599455095, + "loss": 0.3869, + "num_input_tokens_seen": 22194352, + "step": 10615 + }, + { + "epoch": 1.732523044293988, + "grad_norm": 0.0012121282052248716, + "learning_rate": 0.2507862067891006, + "loss": 0.341, + "num_input_tokens_seen": 22203600, + "step": 10620 + }, + { + "epoch": 1.7333387715148054, + "grad_norm": 0.0014601523289456964, + "learning_rate": 0.25074257204115064, + "loss": 0.3708, + "num_input_tokens_seen": 22214720, + "step": 10625 + }, + { + "epoch": 1.7341544987356228, + "grad_norm": 0.0005050289328210056, + "learning_rate": 0.25069892175742997, + "loss": 0.3215, + "num_input_tokens_seen": 22227568, + "step": 10630 + }, + { + "epoch": 1.7349702259564401, + "grad_norm": 0.0005897438386455178, + "learning_rate": 0.25065525594467014, + "loss": 0.2939, + "num_input_tokens_seen": 22239392, + "step": 10635 + }, + { + "epoch": 1.7357859531772575, + "grad_norm": 0.0005142324371263385, + "learning_rate": 0.2506115746096049, + "loss": 0.2893, + "num_input_tokens_seen": 22249808, + "step": 10640 + }, + { + "epoch": 1.7366016803980748, + "grad_norm": 0.000497011817060411, + "learning_rate": 0.25056787775897055, + "loss": 0.3188, + "num_input_tokens_seen": 22260512, + "step": 10645 + }, + { + "epoch": 1.7374174076188922, + "grad_norm": 0.00030881879501976073, + "learning_rate": 0.2505241653995056, + "loss": 0.3168, + "num_input_tokens_seen": 22270144, + "step": 10650 + }, + { + "epoch": 1.7382331348397098, + "grad_norm": 0.0005574266542680562, + "learning_rate": 0.25048043753795113, + "loss": 0.3593, + "num_input_tokens_seen": 22280688, + "step": 10655 + }, + { + "epoch": 1.739048862060527, + "grad_norm": 0.0005572537775151432, + "learning_rate": 0.2504366941810504, + "loss": 0.3458, + "num_input_tokens_seen": 22291568, + "step": 10660 + }, + { + "epoch": 1.7398645892813445, + "grad_norm": 0.0008557679248042405, + "learning_rate": 0.2503929353355493, + "loss": 0.2988, + "num_input_tokens_seen": 22300912, + "step": 10665 + }, + { + "epoch": 1.7406803165021616, + "grad_norm": 0.0008545452728867531, + "learning_rate": 0.250349161008196, + "loss": 0.2669, + "num_input_tokens_seen": 22310688, + "step": 10670 + }, + { + "epoch": 1.7414960437229792, + "grad_norm": 0.0008194029214791954, + "learning_rate": 0.2503053712057409, + "loss": 0.3041, + "num_input_tokens_seen": 22320720, + "step": 10675 + }, + { + "epoch": 1.7423117709437963, + "grad_norm": 0.0009724179399199784, + "learning_rate": 0.25026156593493715, + "loss": 0.3048, + "num_input_tokens_seen": 22331072, + "step": 10680 + }, + { + "epoch": 1.7431274981646139, + "grad_norm": 0.0006240713410079479, + "learning_rate": 0.2502177452025399, + "loss": 0.3533, + "num_input_tokens_seen": 22341072, + "step": 10685 + }, + { + "epoch": 1.743943225385431, + "grad_norm": 0.00106913223862648, + "learning_rate": 0.25017390901530695, + "loss": 0.3336, + "num_input_tokens_seen": 22350624, + "step": 10690 + }, + { + "epoch": 1.7447589526062486, + "grad_norm": 0.0005017641233280301, + "learning_rate": 0.2501300573799984, + "loss": 0.3375, + "num_input_tokens_seen": 22361200, + "step": 10695 + }, + { + "epoch": 1.7455746798270657, + "grad_norm": 0.0007071220898069441, + "learning_rate": 0.2500861903033766, + "loss": 0.3053, + "num_input_tokens_seen": 22372464, + "step": 10700 + }, + { + "epoch": 1.7463904070478833, + "grad_norm": 0.000448412261903286, + "learning_rate": 0.25004230779220654, + "loss": 0.2712, + "num_input_tokens_seen": 22383936, + "step": 10705 + }, + { + "epoch": 1.7472061342687004, + "grad_norm": 0.0006312751211225986, + "learning_rate": 0.24999840985325542, + "loss": 0.3197, + "num_input_tokens_seen": 22394144, + "step": 10710 + }, + { + "epoch": 1.748021861489518, + "grad_norm": 0.0006126820808276534, + "learning_rate": 0.24995449649329285, + "loss": 0.3007, + "num_input_tokens_seen": 22404000, + "step": 10715 + }, + { + "epoch": 1.7488375887103351, + "grad_norm": 0.0010865862714126706, + "learning_rate": 0.2499105677190908, + "loss": 0.3056, + "num_input_tokens_seen": 22415232, + "step": 10720 + }, + { + "epoch": 1.7496533159311527, + "grad_norm": 0.0009889364009723067, + "learning_rate": 0.24986662353742364, + "loss": 0.3006, + "num_input_tokens_seen": 22424672, + "step": 10725 + }, + { + "epoch": 1.7504690431519698, + "grad_norm": 0.0009979249443858862, + "learning_rate": 0.24982266395506814, + "loss": 0.255, + "num_input_tokens_seen": 22434256, + "step": 10730 + }, + { + "epoch": 1.7512847703727874, + "grad_norm": 0.0011382090160623193, + "learning_rate": 0.2497786889788034, + "loss": 0.3165, + "num_input_tokens_seen": 22444288, + "step": 10735 + }, + { + "epoch": 1.7521004975936045, + "grad_norm": 0.0006959455204196274, + "learning_rate": 0.24973469861541095, + "loss": 0.3192, + "num_input_tokens_seen": 22454880, + "step": 10740 + }, + { + "epoch": 1.752916224814422, + "grad_norm": 0.0007442320929840207, + "learning_rate": 0.24969069287167456, + "loss": 0.2839, + "num_input_tokens_seen": 22466000, + "step": 10745 + }, + { + "epoch": 1.7537319520352395, + "grad_norm": 0.00033567563514225185, + "learning_rate": 0.2496466717543806, + "loss": 0.2661, + "num_input_tokens_seen": 22476240, + "step": 10750 + }, + { + "epoch": 1.7545476792560568, + "grad_norm": 0.0008930526091717184, + "learning_rate": 0.24960263527031762, + "loss": 0.3695, + "num_input_tokens_seen": 22485968, + "step": 10755 + }, + { + "epoch": 1.7553634064768742, + "grad_norm": 0.0003305840364191681, + "learning_rate": 0.24955858342627657, + "loss": 0.3397, + "num_input_tokens_seen": 22497824, + "step": 10760 + }, + { + "epoch": 1.7561791336976915, + "grad_norm": 0.0007587596192024648, + "learning_rate": 0.24951451622905083, + "loss": 0.3036, + "num_input_tokens_seen": 22509344, + "step": 10765 + }, + { + "epoch": 1.7569948609185089, + "grad_norm": 0.0010997520294040442, + "learning_rate": 0.24947043368543612, + "loss": 0.3671, + "num_input_tokens_seen": 22521760, + "step": 10770 + }, + { + "epoch": 1.7578105881393262, + "grad_norm": 0.0010637372033670545, + "learning_rate": 0.2494263358022305, + "loss": 0.3176, + "num_input_tokens_seen": 22532000, + "step": 10775 + }, + { + "epoch": 1.7586263153601436, + "grad_norm": 0.0009449108620174229, + "learning_rate": 0.24938222258623444, + "loss": 0.3486, + "num_input_tokens_seen": 22542112, + "step": 10780 + }, + { + "epoch": 1.759442042580961, + "grad_norm": 0.0004993940237909555, + "learning_rate": 0.24933809404425075, + "loss": 0.2839, + "num_input_tokens_seen": 22553504, + "step": 10785 + }, + { + "epoch": 1.7602577698017783, + "grad_norm": 0.000776585191488266, + "learning_rate": 0.24929395018308453, + "loss": 0.3261, + "num_input_tokens_seen": 22563872, + "step": 10790 + }, + { + "epoch": 1.7610734970225956, + "grad_norm": 0.00048603207687847316, + "learning_rate": 0.24924979100954348, + "loss": 0.3393, + "num_input_tokens_seen": 22575024, + "step": 10795 + }, + { + "epoch": 1.761889224243413, + "grad_norm": 0.00031024764757603407, + "learning_rate": 0.24920561653043735, + "loss": 0.2869, + "num_input_tokens_seen": 22585216, + "step": 10800 + }, + { + "epoch": 1.761889224243413, + "eval_loss": 0.31815019249916077, + "eval_runtime": 155.9058, + "eval_samples_per_second": 17.479, + "eval_steps_per_second": 8.742, + "num_input_tokens_seen": 22585216, + "step": 10800 + }, + { + "epoch": 1.7627049514642303, + "grad_norm": 0.0004149778396822512, + "learning_rate": 0.24916142675257846, + "loss": 0.2838, + "num_input_tokens_seen": 22595088, + "step": 10805 + }, + { + "epoch": 1.7635206786850477, + "grad_norm": 0.0008786368416622281, + "learning_rate": 0.24911722168278144, + "loss": 0.3961, + "num_input_tokens_seen": 22604784, + "step": 10810 + }, + { + "epoch": 1.764336405905865, + "grad_norm": 0.00060770328855142, + "learning_rate": 0.24907300132786328, + "loss": 0.3233, + "num_input_tokens_seen": 22615040, + "step": 10815 + }, + { + "epoch": 1.7651521331266824, + "grad_norm": 0.0007108795107342303, + "learning_rate": 0.24902876569464322, + "loss": 0.3117, + "num_input_tokens_seen": 22626928, + "step": 10820 + }, + { + "epoch": 1.7659678603474998, + "grad_norm": 0.0005148422787897289, + "learning_rate": 0.24898451478994305, + "loss": 0.3255, + "num_input_tokens_seen": 22636896, + "step": 10825 + }, + { + "epoch": 1.766783587568317, + "grad_norm": 0.0005242954357527196, + "learning_rate": 0.2489402486205868, + "loss": 0.244, + "num_input_tokens_seen": 22649296, + "step": 10830 + }, + { + "epoch": 1.7675993147891345, + "grad_norm": 0.00047725532203912735, + "learning_rate": 0.24889596719340085, + "loss": 0.3442, + "num_input_tokens_seen": 22658160, + "step": 10835 + }, + { + "epoch": 1.768415042009952, + "grad_norm": 0.0038429799024015665, + "learning_rate": 0.24885167051521392, + "loss": 0.387, + "num_input_tokens_seen": 22666944, + "step": 10840 + }, + { + "epoch": 1.7692307692307692, + "grad_norm": 0.0003431045915931463, + "learning_rate": 0.24880735859285716, + "loss": 0.3492, + "num_input_tokens_seen": 22678032, + "step": 10845 + }, + { + "epoch": 1.7700464964515867, + "grad_norm": 0.0011386690894141793, + "learning_rate": 0.24876303143316406, + "loss": 0.3754, + "num_input_tokens_seen": 22687536, + "step": 10850 + }, + { + "epoch": 1.7708622236724039, + "grad_norm": 0.0007080395589582622, + "learning_rate": 0.24871868904297031, + "loss": 0.341, + "num_input_tokens_seen": 22698784, + "step": 10855 + }, + { + "epoch": 1.7716779508932214, + "grad_norm": 0.0005101084243506193, + "learning_rate": 0.24867433142911416, + "loss": 0.2958, + "num_input_tokens_seen": 22709536, + "step": 10860 + }, + { + "epoch": 1.7724936781140386, + "grad_norm": 0.00083680369425565, + "learning_rate": 0.24862995859843612, + "loss": 0.3346, + "num_input_tokens_seen": 22718960, + "step": 10865 + }, + { + "epoch": 1.7733094053348561, + "grad_norm": 0.0009946634527295828, + "learning_rate": 0.24858557055777897, + "loss": 0.3128, + "num_input_tokens_seen": 22729696, + "step": 10870 + }, + { + "epoch": 1.7741251325556733, + "grad_norm": 0.00036680069752037525, + "learning_rate": 0.24854116731398793, + "loss": 0.339, + "num_input_tokens_seen": 22739808, + "step": 10875 + }, + { + "epoch": 1.7749408597764909, + "grad_norm": 0.0009156455635093153, + "learning_rate": 0.24849674887391052, + "loss": 0.3391, + "num_input_tokens_seen": 22752112, + "step": 10880 + }, + { + "epoch": 1.775756586997308, + "grad_norm": 0.0009991655824705958, + "learning_rate": 0.2484523152443967, + "loss": 0.3248, + "num_input_tokens_seen": 22763392, + "step": 10885 + }, + { + "epoch": 1.7765723142181256, + "grad_norm": 0.000792142644058913, + "learning_rate": 0.24840786643229862, + "loss": 0.3844, + "num_input_tokens_seen": 22773408, + "step": 10890 + }, + { + "epoch": 1.7773880414389427, + "grad_norm": 0.0007611254113726318, + "learning_rate": 0.2483634024444709, + "loss": 0.3119, + "num_input_tokens_seen": 22783936, + "step": 10895 + }, + { + "epoch": 1.7782037686597603, + "grad_norm": 0.0008845360134728253, + "learning_rate": 0.24831892328777033, + "loss": 0.3585, + "num_input_tokens_seen": 22795152, + "step": 10900 + }, + { + "epoch": 1.7790194958805774, + "grad_norm": 0.0005960068665444851, + "learning_rate": 0.2482744289690563, + "loss": 0.3404, + "num_input_tokens_seen": 22805104, + "step": 10905 + }, + { + "epoch": 1.779835223101395, + "grad_norm": 0.0007231996860355139, + "learning_rate": 0.2482299194951903, + "loss": 0.3165, + "num_input_tokens_seen": 22815520, + "step": 10910 + }, + { + "epoch": 1.780650950322212, + "grad_norm": 0.00034061228507198393, + "learning_rate": 0.2481853948730363, + "loss": 0.3356, + "num_input_tokens_seen": 22825424, + "step": 10915 + }, + { + "epoch": 1.7814666775430297, + "grad_norm": 0.0005643180338665843, + "learning_rate": 0.24814085510946052, + "loss": 0.3102, + "num_input_tokens_seen": 22835968, + "step": 10920 + }, + { + "epoch": 1.7822824047638468, + "grad_norm": 0.0007392407860606909, + "learning_rate": 0.24809630021133158, + "loss": 0.3452, + "num_input_tokens_seen": 22846432, + "step": 10925 + }, + { + "epoch": 1.7830981319846644, + "grad_norm": 0.00082031125202775, + "learning_rate": 0.24805173018552037, + "loss": 0.3032, + "num_input_tokens_seen": 22856160, + "step": 10930 + }, + { + "epoch": 1.7839138592054817, + "grad_norm": 0.0004354343400336802, + "learning_rate": 0.2480071450389002, + "loss": 0.3143, + "num_input_tokens_seen": 22866512, + "step": 10935 + }, + { + "epoch": 1.784729586426299, + "grad_norm": 0.0006922395550645888, + "learning_rate": 0.24796254477834662, + "loss": 0.3269, + "num_input_tokens_seen": 22876272, + "step": 10940 + }, + { + "epoch": 1.7855453136471164, + "grad_norm": 0.0005060981493443251, + "learning_rate": 0.24791792941073754, + "loss": 0.3013, + "num_input_tokens_seen": 22886592, + "step": 10945 + }, + { + "epoch": 1.7863610408679338, + "grad_norm": 0.0006246320554055274, + "learning_rate": 0.2478732989429533, + "loss": 0.3479, + "num_input_tokens_seen": 22897968, + "step": 10950 + }, + { + "epoch": 1.7871767680887511, + "grad_norm": 0.0005736483726650476, + "learning_rate": 0.24782865338187632, + "loss": 0.3074, + "num_input_tokens_seen": 22907792, + "step": 10955 + }, + { + "epoch": 1.7879924953095685, + "grad_norm": 0.0005252953851595521, + "learning_rate": 0.2477839927343916, + "loss": 0.3734, + "num_input_tokens_seen": 22917264, + "step": 10960 + }, + { + "epoch": 1.7888082225303858, + "grad_norm": 0.0005695904255844653, + "learning_rate": 0.2477393170073864, + "loss": 0.2952, + "num_input_tokens_seen": 22928608, + "step": 10965 + }, + { + "epoch": 1.7896239497512032, + "grad_norm": 0.0005634942790493369, + "learning_rate": 0.2476946262077503, + "loss": 0.3424, + "num_input_tokens_seen": 22939888, + "step": 10970 + }, + { + "epoch": 1.7904396769720206, + "grad_norm": 0.0015008801128715277, + "learning_rate": 0.24764992034237507, + "loss": 0.3559, + "num_input_tokens_seen": 22950864, + "step": 10975 + }, + { + "epoch": 1.791255404192838, + "grad_norm": 0.00044309248914942145, + "learning_rate": 0.24760519941815498, + "loss": 0.3066, + "num_input_tokens_seen": 22962656, + "step": 10980 + }, + { + "epoch": 1.7920711314136553, + "grad_norm": 0.0006128817331045866, + "learning_rate": 0.2475604634419866, + "loss": 0.293, + "num_input_tokens_seen": 22974112, + "step": 10985 + }, + { + "epoch": 1.7928868586344726, + "grad_norm": 0.0007779362495057285, + "learning_rate": 0.24751571242076872, + "loss": 0.2972, + "num_input_tokens_seen": 22985232, + "step": 10990 + }, + { + "epoch": 1.79370258585529, + "grad_norm": 0.0005935783265158534, + "learning_rate": 0.2474709463614025, + "loss": 0.3436, + "num_input_tokens_seen": 22995312, + "step": 10995 + }, + { + "epoch": 1.7945183130761073, + "grad_norm": 0.0013239224208518863, + "learning_rate": 0.24742616527079145, + "loss": 0.3774, + "num_input_tokens_seen": 23005600, + "step": 11000 + }, + { + "epoch": 1.7945183130761073, + "eval_loss": 0.33231860399246216, + "eval_runtime": 155.7721, + "eval_samples_per_second": 17.494, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 23005600, + "step": 11000 + }, + { + "epoch": 1.7953340402969247, + "grad_norm": 0.0006585658993571997, + "learning_rate": 0.24738136915584139, + "loss": 0.3322, + "num_input_tokens_seen": 23014720, + "step": 11005 + }, + { + "epoch": 1.796149767517742, + "grad_norm": 0.0005775223835371435, + "learning_rate": 0.24733655802346047, + "loss": 0.3323, + "num_input_tokens_seen": 23024944, + "step": 11010 + }, + { + "epoch": 1.7969654947385594, + "grad_norm": 0.00030387178412638605, + "learning_rate": 0.24729173188055906, + "loss": 0.3018, + "num_input_tokens_seen": 23035856, + "step": 11015 + }, + { + "epoch": 1.7977812219593767, + "grad_norm": 0.0005973592051304877, + "learning_rate": 0.24724689073404996, + "loss": 0.3148, + "num_input_tokens_seen": 23045360, + "step": 11020 + }, + { + "epoch": 1.7985969491801943, + "grad_norm": 0.0008502819109708071, + "learning_rate": 0.24720203459084822, + "loss": 0.309, + "num_input_tokens_seen": 23056368, + "step": 11025 + }, + { + "epoch": 1.7994126764010114, + "grad_norm": 0.0010154196061193943, + "learning_rate": 0.24715716345787123, + "loss": 0.3805, + "num_input_tokens_seen": 23066736, + "step": 11030 + }, + { + "epoch": 1.800228403621829, + "grad_norm": 0.001293670036830008, + "learning_rate": 0.2471122773420387, + "loss": 0.3326, + "num_input_tokens_seen": 23078992, + "step": 11035 + }, + { + "epoch": 1.8010441308426461, + "grad_norm": 0.0005162745364941657, + "learning_rate": 0.24706737625027259, + "loss": 0.3186, + "num_input_tokens_seen": 23089984, + "step": 11040 + }, + { + "epoch": 1.8018598580634637, + "grad_norm": 0.0004343291511759162, + "learning_rate": 0.24702246018949725, + "loss": 0.2806, + "num_input_tokens_seen": 23099904, + "step": 11045 + }, + { + "epoch": 1.8026755852842808, + "grad_norm": 0.000996316084638238, + "learning_rate": 0.2469775291666393, + "loss": 0.3889, + "num_input_tokens_seen": 23110208, + "step": 11050 + }, + { + "epoch": 1.8034913125050984, + "grad_norm": 0.0014544714940711856, + "learning_rate": 0.24693258318862765, + "loss": 0.3115, + "num_input_tokens_seen": 23120544, + "step": 11055 + }, + { + "epoch": 1.8043070397259156, + "grad_norm": 0.0007592611946165562, + "learning_rate": 0.2468876222623935, + "loss": 0.2975, + "num_input_tokens_seen": 23131472, + "step": 11060 + }, + { + "epoch": 1.8051227669467331, + "grad_norm": 0.00039643660420551896, + "learning_rate": 0.2468426463948705, + "loss": 0.3103, + "num_input_tokens_seen": 23140640, + "step": 11065 + }, + { + "epoch": 1.8059384941675503, + "grad_norm": 0.0008628187351860106, + "learning_rate": 0.24679765559299438, + "loss": 0.2791, + "num_input_tokens_seen": 23150800, + "step": 11070 + }, + { + "epoch": 1.8067542213883678, + "grad_norm": 0.0011154800886288285, + "learning_rate": 0.24675264986370332, + "loss": 0.3556, + "num_input_tokens_seen": 23161360, + "step": 11075 + }, + { + "epoch": 1.807569948609185, + "grad_norm": 0.0007615709910169244, + "learning_rate": 0.2467076292139378, + "loss": 0.2899, + "num_input_tokens_seen": 23171776, + "step": 11080 + }, + { + "epoch": 1.8083856758300025, + "grad_norm": 0.000513669743668288, + "learning_rate": 0.24666259365064055, + "loss": 0.2832, + "num_input_tokens_seen": 23181344, + "step": 11085 + }, + { + "epoch": 1.8092014030508197, + "grad_norm": 0.0005002831458114088, + "learning_rate": 0.24661754318075663, + "loss": 0.2996, + "num_input_tokens_seen": 23191808, + "step": 11090 + }, + { + "epoch": 1.8100171302716372, + "grad_norm": 0.000616150617133826, + "learning_rate": 0.2465724778112334, + "loss": 0.3759, + "num_input_tokens_seen": 23201312, + "step": 11095 + }, + { + "epoch": 1.8108328574924544, + "grad_norm": 0.0010417718440294266, + "learning_rate": 0.24652739754902042, + "loss": 0.3113, + "num_input_tokens_seen": 23212416, + "step": 11100 + }, + { + "epoch": 1.811648584713272, + "grad_norm": 0.0009069519001059234, + "learning_rate": 0.24648230240106975, + "loss": 0.3512, + "num_input_tokens_seen": 23221024, + "step": 11105 + }, + { + "epoch": 1.812464311934089, + "grad_norm": 0.0011855501215904951, + "learning_rate": 0.2464371923743356, + "loss": 0.3651, + "num_input_tokens_seen": 23230416, + "step": 11110 + }, + { + "epoch": 1.8132800391549067, + "grad_norm": 0.00041570886969566345, + "learning_rate": 0.24639206747577444, + "loss": 0.2968, + "num_input_tokens_seen": 23240192, + "step": 11115 + }, + { + "epoch": 1.814095766375724, + "grad_norm": 0.0005923124263063073, + "learning_rate": 0.24634692771234515, + "loss": 0.317, + "num_input_tokens_seen": 23248992, + "step": 11120 + }, + { + "epoch": 1.8149114935965414, + "grad_norm": 0.0004941904335282743, + "learning_rate": 0.2463017730910088, + "loss": 0.2984, + "num_input_tokens_seen": 23258688, + "step": 11125 + }, + { + "epoch": 1.8157272208173587, + "grad_norm": 0.0007221192354336381, + "learning_rate": 0.2462566036187289, + "loss": 0.302, + "num_input_tokens_seen": 23268448, + "step": 11130 + }, + { + "epoch": 1.816542948038176, + "grad_norm": 0.0009902867022901773, + "learning_rate": 0.24621141930247106, + "loss": 0.3018, + "num_input_tokens_seen": 23277616, + "step": 11135 + }, + { + "epoch": 1.8173586752589934, + "grad_norm": 0.0010118106147274375, + "learning_rate": 0.2461662201492033, + "loss": 0.3125, + "num_input_tokens_seen": 23287136, + "step": 11140 + }, + { + "epoch": 1.8181744024798108, + "grad_norm": 0.0008751368732191622, + "learning_rate": 0.24612100616589586, + "loss": 0.2833, + "num_input_tokens_seen": 23296880, + "step": 11145 + }, + { + "epoch": 1.8189901297006281, + "grad_norm": 0.0004997876822017133, + "learning_rate": 0.24607577735952135, + "loss": 0.3065, + "num_input_tokens_seen": 23305856, + "step": 11150 + }, + { + "epoch": 1.8198058569214455, + "grad_norm": 0.0004215153749100864, + "learning_rate": 0.24603053373705464, + "loss": 0.23, + "num_input_tokens_seen": 23315936, + "step": 11155 + }, + { + "epoch": 1.8206215841422628, + "grad_norm": 0.0014938621316105127, + "learning_rate": 0.2459852753054728, + "loss": 0.3209, + "num_input_tokens_seen": 23326752, + "step": 11160 + }, + { + "epoch": 1.8214373113630802, + "grad_norm": 0.0007267557666637003, + "learning_rate": 0.24594000207175526, + "loss": 0.2941, + "num_input_tokens_seen": 23338544, + "step": 11165 + }, + { + "epoch": 1.8222530385838975, + "grad_norm": 0.0011683502234518528, + "learning_rate": 0.2458947140428838, + "loss": 0.3636, + "num_input_tokens_seen": 23348800, + "step": 11170 + }, + { + "epoch": 1.8230687658047149, + "grad_norm": 0.0009711086750030518, + "learning_rate": 0.24584941122584233, + "loss": 0.4084, + "num_input_tokens_seen": 23359824, + "step": 11175 + }, + { + "epoch": 1.8238844930255322, + "grad_norm": 0.0004544725234154612, + "learning_rate": 0.24580409362761713, + "loss": 0.3915, + "num_input_tokens_seen": 23369920, + "step": 11180 + }, + { + "epoch": 1.8247002202463496, + "grad_norm": 0.001243923557922244, + "learning_rate": 0.2457587612551967, + "loss": 0.377, + "num_input_tokens_seen": 23380864, + "step": 11185 + }, + { + "epoch": 1.825515947467167, + "grad_norm": 0.0010294276289641857, + "learning_rate": 0.24571341411557193, + "loss": 0.2926, + "num_input_tokens_seen": 23391888, + "step": 11190 + }, + { + "epoch": 1.8263316746879843, + "grad_norm": 0.0005596186965703964, + "learning_rate": 0.2456680522157359, + "loss": 0.3329, + "num_input_tokens_seen": 23403312, + "step": 11195 + }, + { + "epoch": 1.8271474019088016, + "grad_norm": 0.00047175551299005747, + "learning_rate": 0.245622675562684, + "loss": 0.3332, + "num_input_tokens_seen": 23413712, + "step": 11200 + }, + { + "epoch": 1.8271474019088016, + "eval_loss": 0.32120412588119507, + "eval_runtime": 155.6111, + "eval_samples_per_second": 17.512, + "eval_steps_per_second": 8.759, + "num_input_tokens_seen": 23413712, + "step": 11200 + }, + { + "epoch": 1.827963129129619, + "grad_norm": 0.0008009933517314494, + "learning_rate": 0.24557728416341384, + "loss": 0.3261, + "num_input_tokens_seen": 23424544, + "step": 11205 + }, + { + "epoch": 1.8287788563504366, + "grad_norm": 0.0008438238292001188, + "learning_rate": 0.24553187802492538, + "loss": 0.3643, + "num_input_tokens_seen": 23435200, + "step": 11210 + }, + { + "epoch": 1.8295945835712537, + "grad_norm": 0.0003607615362852812, + "learning_rate": 0.24548645715422074, + "loss": 0.3498, + "num_input_tokens_seen": 23444192, + "step": 11215 + }, + { + "epoch": 1.8304103107920713, + "grad_norm": 0.00038504134863615036, + "learning_rate": 0.2454410215583045, + "loss": 0.29, + "num_input_tokens_seen": 23454144, + "step": 11220 + }, + { + "epoch": 1.8312260380128884, + "grad_norm": 0.0011182742891833186, + "learning_rate": 0.24539557124418332, + "loss": 0.355, + "num_input_tokens_seen": 23464656, + "step": 11225 + }, + { + "epoch": 1.832041765233706, + "grad_norm": 0.0006652456359006464, + "learning_rate": 0.24535010621886624, + "loss": 0.3279, + "num_input_tokens_seen": 23474112, + "step": 11230 + }, + { + "epoch": 1.8328574924545231, + "grad_norm": 0.0009225752437487245, + "learning_rate": 0.2453046264893646, + "loss": 0.3567, + "num_input_tokens_seen": 23485456, + "step": 11235 + }, + { + "epoch": 1.8336732196753407, + "grad_norm": 0.0010428056120872498, + "learning_rate": 0.24525913206269184, + "loss": 0.3235, + "num_input_tokens_seen": 23496576, + "step": 11240 + }, + { + "epoch": 1.8344889468961578, + "grad_norm": 0.0008310190169140697, + "learning_rate": 0.2452136229458638, + "loss": 0.3348, + "num_input_tokens_seen": 23506976, + "step": 11245 + }, + { + "epoch": 1.8353046741169754, + "grad_norm": 0.0004676055395975709, + "learning_rate": 0.24516809914589857, + "loss": 0.2961, + "num_input_tokens_seen": 23517776, + "step": 11250 + }, + { + "epoch": 1.8361204013377925, + "grad_norm": 0.0008899369859136641, + "learning_rate": 0.2451225606698165, + "loss": 0.3287, + "num_input_tokens_seen": 23528096, + "step": 11255 + }, + { + "epoch": 1.83693612855861, + "grad_norm": 0.0005893537309020758, + "learning_rate": 0.2450770075246402, + "loss": 0.3149, + "num_input_tokens_seen": 23538816, + "step": 11260 + }, + { + "epoch": 1.8377518557794272, + "grad_norm": 0.0005418189684860408, + "learning_rate": 0.24503143971739455, + "loss": 0.2775, + "num_input_tokens_seen": 23548688, + "step": 11265 + }, + { + "epoch": 1.8385675830002448, + "grad_norm": 0.0005224101478233933, + "learning_rate": 0.24498585725510663, + "loss": 0.3263, + "num_input_tokens_seen": 23559440, + "step": 11270 + }, + { + "epoch": 1.839383310221062, + "grad_norm": 0.0006388858309946954, + "learning_rate": 0.24494026014480583, + "loss": 0.3841, + "num_input_tokens_seen": 23569488, + "step": 11275 + }, + { + "epoch": 1.8401990374418795, + "grad_norm": 0.0006547382799908519, + "learning_rate": 0.24489464839352387, + "loss": 0.3538, + "num_input_tokens_seen": 23581088, + "step": 11280 + }, + { + "epoch": 1.8410147646626966, + "grad_norm": 0.0009281898965127766, + "learning_rate": 0.2448490220082946, + "loss": 0.3374, + "num_input_tokens_seen": 23590048, + "step": 11285 + }, + { + "epoch": 1.8418304918835142, + "grad_norm": 0.001320741605013609, + "learning_rate": 0.24480338099615415, + "loss": 0.3478, + "num_input_tokens_seen": 23600544, + "step": 11290 + }, + { + "epoch": 1.8426462191043314, + "grad_norm": 0.0005085073644295335, + "learning_rate": 0.244757725364141, + "loss": 0.319, + "num_input_tokens_seen": 23609984, + "step": 11295 + }, + { + "epoch": 1.843461946325149, + "grad_norm": 0.0003414139209780842, + "learning_rate": 0.24471205511929583, + "loss": 0.3127, + "num_input_tokens_seen": 23620400, + "step": 11300 + }, + { + "epoch": 1.8442776735459663, + "grad_norm": 0.0003844956518150866, + "learning_rate": 0.24466637026866145, + "loss": 0.3089, + "num_input_tokens_seen": 23629840, + "step": 11305 + }, + { + "epoch": 1.8450934007667836, + "grad_norm": 0.0004366193024907261, + "learning_rate": 0.2446206708192832, + "loss": 0.3286, + "num_input_tokens_seen": 23641696, + "step": 11310 + }, + { + "epoch": 1.845909127987601, + "grad_norm": 0.00044044890091754496, + "learning_rate": 0.2445749567782084, + "loss": 0.3331, + "num_input_tokens_seen": 23651680, + "step": 11315 + }, + { + "epoch": 1.8467248552084183, + "grad_norm": 0.00045015374780632555, + "learning_rate": 0.2445292281524868, + "loss": 0.3185, + "num_input_tokens_seen": 23662528, + "step": 11320 + }, + { + "epoch": 1.8475405824292357, + "grad_norm": 0.0004831848491448909, + "learning_rate": 0.24448348494917022, + "loss": 0.2728, + "num_input_tokens_seen": 23673328, + "step": 11325 + }, + { + "epoch": 1.848356309650053, + "grad_norm": 0.0012197286123409867, + "learning_rate": 0.24443772717531295, + "loss": 0.3577, + "num_input_tokens_seen": 23683872, + "step": 11330 + }, + { + "epoch": 1.8491720368708704, + "grad_norm": 0.000542963098268956, + "learning_rate": 0.24439195483797138, + "loss": 0.323, + "num_input_tokens_seen": 23693360, + "step": 11335 + }, + { + "epoch": 1.8499877640916877, + "grad_norm": 0.0006279461667872965, + "learning_rate": 0.24434616794420416, + "loss": 0.3451, + "num_input_tokens_seen": 23704112, + "step": 11340 + }, + { + "epoch": 1.850803491312505, + "grad_norm": 0.0009895446710288525, + "learning_rate": 0.24430036650107223, + "loss": 0.3384, + "num_input_tokens_seen": 23713984, + "step": 11345 + }, + { + "epoch": 1.8516192185333225, + "grad_norm": 0.0010170981986448169, + "learning_rate": 0.2442545505156387, + "loss": 0.3372, + "num_input_tokens_seen": 23724608, + "step": 11350 + }, + { + "epoch": 1.8524349457541398, + "grad_norm": 0.0003457177372183651, + "learning_rate": 0.24420871999496904, + "loss": 0.2962, + "num_input_tokens_seen": 23735472, + "step": 11355 + }, + { + "epoch": 1.8532506729749572, + "grad_norm": 0.0011163741583004594, + "learning_rate": 0.24416287494613084, + "loss": 0.3304, + "num_input_tokens_seen": 23746592, + "step": 11360 + }, + { + "epoch": 1.8540664001957745, + "grad_norm": 0.00039675403968431056, + "learning_rate": 0.24411701537619399, + "loss": 0.2547, + "num_input_tokens_seen": 23757536, + "step": 11365 + }, + { + "epoch": 1.8548821274165919, + "grad_norm": 0.001330796629190445, + "learning_rate": 0.24407114129223062, + "loss": 0.3131, + "num_input_tokens_seen": 23767584, + "step": 11370 + }, + { + "epoch": 1.8556978546374092, + "grad_norm": 0.00046234484761953354, + "learning_rate": 0.2440252527013151, + "loss": 0.3802, + "num_input_tokens_seen": 23777136, + "step": 11375 + }, + { + "epoch": 1.8565135818582266, + "grad_norm": 0.0005748730036430061, + "learning_rate": 0.24397934961052403, + "loss": 0.355, + "num_input_tokens_seen": 23787872, + "step": 11380 + }, + { + "epoch": 1.857329309079044, + "grad_norm": 0.0008035072241909802, + "learning_rate": 0.24393343202693618, + "loss": 0.2877, + "num_input_tokens_seen": 23796464, + "step": 11385 + }, + { + "epoch": 1.8581450362998613, + "grad_norm": 0.001027230522595346, + "learning_rate": 0.2438874999576327, + "loss": 0.3335, + "num_input_tokens_seen": 23807360, + "step": 11390 + }, + { + "epoch": 1.8589607635206788, + "grad_norm": 0.0007038271287456155, + "learning_rate": 0.24384155340969688, + "loss": 0.361, + "num_input_tokens_seen": 23816768, + "step": 11395 + }, + { + "epoch": 1.859776490741496, + "grad_norm": 0.0007810801034793258, + "learning_rate": 0.24379559239021423, + "loss": 0.3697, + "num_input_tokens_seen": 23827536, + "step": 11400 + }, + { + "epoch": 1.859776490741496, + "eval_loss": 0.3198401629924774, + "eval_runtime": 155.7513, + "eval_samples_per_second": 17.496, + "eval_steps_per_second": 8.751, + "num_input_tokens_seen": 23827536, + "step": 11400 + }, + { + "epoch": 1.8605922179623136, + "grad_norm": 0.001022770651616156, + "learning_rate": 0.2437496169062725, + "loss": 0.3626, + "num_input_tokens_seen": 23837824, + "step": 11405 + }, + { + "epoch": 1.8614079451831307, + "grad_norm": 0.0008251392282545567, + "learning_rate": 0.24370362696496176, + "loss": 0.3383, + "num_input_tokens_seen": 23847904, + "step": 11410 + }, + { + "epoch": 1.8622236724039483, + "grad_norm": 0.0004406454099807888, + "learning_rate": 0.24365762257337417, + "loss": 0.344, + "num_input_tokens_seen": 23856816, + "step": 11415 + }, + { + "epoch": 1.8630393996247654, + "grad_norm": 0.0007890978013165295, + "learning_rate": 0.2436116037386042, + "loss": 0.3221, + "num_input_tokens_seen": 23865392, + "step": 11420 + }, + { + "epoch": 1.863855126845583, + "grad_norm": 0.0008936987142078578, + "learning_rate": 0.24356557046774852, + "loss": 0.3089, + "num_input_tokens_seen": 23876384, + "step": 11425 + }, + { + "epoch": 1.8646708540664, + "grad_norm": 0.0004099891521036625, + "learning_rate": 0.24351952276790606, + "loss": 0.3498, + "num_input_tokens_seen": 23887792, + "step": 11430 + }, + { + "epoch": 1.8654865812872177, + "grad_norm": 0.0007790329400449991, + "learning_rate": 0.24347346064617797, + "loss": 0.3423, + "num_input_tokens_seen": 23897968, + "step": 11435 + }, + { + "epoch": 1.8663023085080348, + "grad_norm": 0.0003694529877975583, + "learning_rate": 0.24342738410966758, + "loss": 0.3119, + "num_input_tokens_seen": 23907632, + "step": 11440 + }, + { + "epoch": 1.8671180357288524, + "grad_norm": 0.00047863443614915013, + "learning_rate": 0.24338129316548046, + "loss": 0.3216, + "num_input_tokens_seen": 23917440, + "step": 11445 + }, + { + "epoch": 1.8679337629496695, + "grad_norm": 0.0003516806464176625, + "learning_rate": 0.24333518782072444, + "loss": 0.2957, + "num_input_tokens_seen": 23928144, + "step": 11450 + }, + { + "epoch": 1.868749490170487, + "grad_norm": 0.0004897741600871086, + "learning_rate": 0.24328906808250952, + "loss": 0.3008, + "num_input_tokens_seen": 23938480, + "step": 11455 + }, + { + "epoch": 1.8695652173913042, + "grad_norm": 0.00028586408006958663, + "learning_rate": 0.243242933957948, + "loss": 0.3352, + "num_input_tokens_seen": 23948192, + "step": 11460 + }, + { + "epoch": 1.8703809446121218, + "grad_norm": 0.0009155469015240669, + "learning_rate": 0.24319678545415427, + "loss": 0.2961, + "num_input_tokens_seen": 23958784, + "step": 11465 + }, + { + "epoch": 1.871196671832939, + "grad_norm": 0.001088675926439464, + "learning_rate": 0.24315062257824507, + "loss": 0.3302, + "num_input_tokens_seen": 23969008, + "step": 11470 + }, + { + "epoch": 1.8720123990537565, + "grad_norm": 0.0004966749111190438, + "learning_rate": 0.24310444533733921, + "loss": 0.2488, + "num_input_tokens_seen": 23978688, + "step": 11475 + }, + { + "epoch": 1.8728281262745736, + "grad_norm": 0.0008985394379124045, + "learning_rate": 0.2430582537385579, + "loss": 0.3322, + "num_input_tokens_seen": 23987408, + "step": 11480 + }, + { + "epoch": 1.8736438534953912, + "grad_norm": 0.0004696486284956336, + "learning_rate": 0.2430120477890244, + "loss": 0.2678, + "num_input_tokens_seen": 23997328, + "step": 11485 + }, + { + "epoch": 1.8744595807162086, + "grad_norm": 0.0005103228031657636, + "learning_rate": 0.24296582749586426, + "loss": 0.2848, + "num_input_tokens_seen": 24006304, + "step": 11490 + }, + { + "epoch": 1.875275307937026, + "grad_norm": 0.0005440377281047404, + "learning_rate": 0.24291959286620526, + "loss": 0.3738, + "num_input_tokens_seen": 24017856, + "step": 11495 + }, + { + "epoch": 1.8760910351578433, + "grad_norm": 0.0005013240152038634, + "learning_rate": 0.24287334390717738, + "loss": 0.3323, + "num_input_tokens_seen": 24029760, + "step": 11500 + }, + { + "epoch": 1.8769067623786606, + "grad_norm": 0.0011595290852710605, + "learning_rate": 0.24282708062591268, + "loss": 0.3099, + "num_input_tokens_seen": 24040336, + "step": 11505 + }, + { + "epoch": 1.877722489599478, + "grad_norm": 0.0007056013564579189, + "learning_rate": 0.24278080302954563, + "loss": 0.3482, + "num_input_tokens_seen": 24051472, + "step": 11510 + }, + { + "epoch": 1.8785382168202953, + "grad_norm": 0.0012028943747282028, + "learning_rate": 0.24273451112521283, + "loss": 0.261, + "num_input_tokens_seen": 24061920, + "step": 11515 + }, + { + "epoch": 1.8793539440411127, + "grad_norm": 0.0007352998945862055, + "learning_rate": 0.242688204920053, + "loss": 0.3268, + "num_input_tokens_seen": 24071424, + "step": 11520 + }, + { + "epoch": 1.88016967126193, + "grad_norm": 0.000632162147667259, + "learning_rate": 0.24264188442120715, + "loss": 0.2785, + "num_input_tokens_seen": 24081712, + "step": 11525 + }, + { + "epoch": 1.8809853984827474, + "grad_norm": 0.00038524653064087033, + "learning_rate": 0.24259554963581853, + "loss": 0.3391, + "num_input_tokens_seen": 24091248, + "step": 11530 + }, + { + "epoch": 1.8818011257035647, + "grad_norm": 0.0008068302413448691, + "learning_rate": 0.24254920057103257, + "loss": 0.2457, + "num_input_tokens_seen": 24103200, + "step": 11535 + }, + { + "epoch": 1.882616852924382, + "grad_norm": 0.0005644605262205005, + "learning_rate": 0.24250283723399685, + "loss": 0.3183, + "num_input_tokens_seen": 24113200, + "step": 11540 + }, + { + "epoch": 1.8834325801451994, + "grad_norm": 0.0023990555200725794, + "learning_rate": 0.24245645963186108, + "loss": 0.2756, + "num_input_tokens_seen": 24121952, + "step": 11545 + }, + { + "epoch": 1.8842483073660168, + "grad_norm": 0.0016128390561789274, + "learning_rate": 0.2424100677717774, + "loss": 0.3535, + "num_input_tokens_seen": 24131840, + "step": 11550 + }, + { + "epoch": 1.8850640345868341, + "grad_norm": 0.0007452861755155027, + "learning_rate": 0.24236366166090004, + "loss": 0.23, + "num_input_tokens_seen": 24143936, + "step": 11555 + }, + { + "epoch": 1.8858797618076515, + "grad_norm": 0.0012659423518925905, + "learning_rate": 0.24231724130638527, + "loss": 0.4209, + "num_input_tokens_seen": 24153168, + "step": 11560 + }, + { + "epoch": 1.8866954890284688, + "grad_norm": 0.001462847227230668, + "learning_rate": 0.2422708067153917, + "loss": 0.3424, + "num_input_tokens_seen": 24164576, + "step": 11565 + }, + { + "epoch": 1.8875112162492862, + "grad_norm": 0.0005080808186903596, + "learning_rate": 0.24222435789508026, + "loss": 0.3684, + "num_input_tokens_seen": 24176608, + "step": 11570 + }, + { + "epoch": 1.8883269434701035, + "grad_norm": 0.0007898506592027843, + "learning_rate": 0.24217789485261387, + "loss": 0.2779, + "num_input_tokens_seen": 24187696, + "step": 11575 + }, + { + "epoch": 1.8891426706909211, + "grad_norm": 0.00033251073909923434, + "learning_rate": 0.2421314175951577, + "loss": 0.3527, + "num_input_tokens_seen": 24199280, + "step": 11580 + }, + { + "epoch": 1.8899583979117383, + "grad_norm": 0.0031299449037760496, + "learning_rate": 0.2420849261298791, + "loss": 0.3244, + "num_input_tokens_seen": 24209504, + "step": 11585 + }, + { + "epoch": 1.8907741251325558, + "grad_norm": 0.0013507383409887552, + "learning_rate": 0.24203842046394775, + "loss": 0.3086, + "num_input_tokens_seen": 24220576, + "step": 11590 + }, + { + "epoch": 1.891589852353373, + "grad_norm": 0.0007298809941858053, + "learning_rate": 0.24199190060453535, + "loss": 0.2968, + "num_input_tokens_seen": 24230752, + "step": 11595 + }, + { + "epoch": 1.8924055795741905, + "grad_norm": 0.0007765875197947025, + "learning_rate": 0.2419453665588158, + "loss": 0.351, + "num_input_tokens_seen": 24242256, + "step": 11600 + }, + { + "epoch": 1.8924055795741905, + "eval_loss": 0.31863877177238464, + "eval_runtime": 156.0144, + "eval_samples_per_second": 17.466, + "eval_steps_per_second": 8.736, + "num_input_tokens_seen": 24242256, + "step": 11600 + }, + { + "epoch": 1.8932213067950077, + "grad_norm": 0.0003725196002051234, + "learning_rate": 0.24189881833396523, + "loss": 0.3292, + "num_input_tokens_seen": 24252608, + "step": 11605 + }, + { + "epoch": 1.8940370340158252, + "grad_norm": 0.0009923611069098115, + "learning_rate": 0.24185225593716203, + "loss": 0.2586, + "num_input_tokens_seen": 24262416, + "step": 11610 + }, + { + "epoch": 1.8948527612366424, + "grad_norm": 0.0006226606201380491, + "learning_rate": 0.2418056793755867, + "loss": 0.316, + "num_input_tokens_seen": 24273952, + "step": 11615 + }, + { + "epoch": 1.89566848845746, + "grad_norm": 0.0007063756929710507, + "learning_rate": 0.24175908865642187, + "loss": 0.3274, + "num_input_tokens_seen": 24282352, + "step": 11620 + }, + { + "epoch": 1.896484215678277, + "grad_norm": 0.000455596687970683, + "learning_rate": 0.24171248378685248, + "loss": 0.3477, + "num_input_tokens_seen": 24293440, + "step": 11625 + }, + { + "epoch": 1.8972999428990946, + "grad_norm": 0.0010540199000388384, + "learning_rate": 0.24166586477406554, + "loss": 0.3552, + "num_input_tokens_seen": 24304640, + "step": 11630 + }, + { + "epoch": 1.8981156701199118, + "grad_norm": 0.0018309904262423515, + "learning_rate": 0.24161923162525034, + "loss": 0.3345, + "num_input_tokens_seen": 24314496, + "step": 11635 + }, + { + "epoch": 1.8989313973407294, + "grad_norm": 0.0005625904304906726, + "learning_rate": 0.2415725843475982, + "loss": 0.357, + "num_input_tokens_seen": 24324096, + "step": 11640 + }, + { + "epoch": 1.8997471245615465, + "grad_norm": 0.0007604634156450629, + "learning_rate": 0.24152592294830286, + "loss": 0.2782, + "num_input_tokens_seen": 24334672, + "step": 11645 + }, + { + "epoch": 1.900562851782364, + "grad_norm": 0.0003880466101691127, + "learning_rate": 0.24147924743455995, + "loss": 0.3054, + "num_input_tokens_seen": 24344512, + "step": 11650 + }, + { + "epoch": 1.9013785790031812, + "grad_norm": 0.0009974554413929582, + "learning_rate": 0.24143255781356754, + "loss": 0.307, + "num_input_tokens_seen": 24354208, + "step": 11655 + }, + { + "epoch": 1.9021943062239988, + "grad_norm": 0.00047459438792429864, + "learning_rate": 0.24138585409252566, + "loss": 0.3854, + "num_input_tokens_seen": 24364928, + "step": 11660 + }, + { + "epoch": 1.903010033444816, + "grad_norm": 0.0004120371595490724, + "learning_rate": 0.24133913627863662, + "loss": 0.3249, + "num_input_tokens_seen": 24376576, + "step": 11665 + }, + { + "epoch": 1.9038257606656335, + "grad_norm": 0.0007020796183496714, + "learning_rate": 0.241292404379105, + "loss": 0.3463, + "num_input_tokens_seen": 24387504, + "step": 11670 + }, + { + "epoch": 1.9046414878864508, + "grad_norm": 0.0008374791359528899, + "learning_rate": 0.24124565840113735, + "loss": 0.379, + "num_input_tokens_seen": 24398400, + "step": 11675 + }, + { + "epoch": 1.9054572151072682, + "grad_norm": 0.00046592261060141027, + "learning_rate": 0.2411988983519425, + "loss": 0.3067, + "num_input_tokens_seen": 24408896, + "step": 11680 + }, + { + "epoch": 1.9062729423280855, + "grad_norm": 0.0006837643450126052, + "learning_rate": 0.24115212423873145, + "loss": 0.3406, + "num_input_tokens_seen": 24418928, + "step": 11685 + }, + { + "epoch": 1.9070886695489029, + "grad_norm": 0.0006820783601142466, + "learning_rate": 0.24110533606871737, + "loss": 0.2941, + "num_input_tokens_seen": 24429792, + "step": 11690 + }, + { + "epoch": 1.9079043967697202, + "grad_norm": 0.0008958164835348725, + "learning_rate": 0.24105853384911552, + "loss": 0.3445, + "num_input_tokens_seen": 24439808, + "step": 11695 + }, + { + "epoch": 1.9087201239905376, + "grad_norm": 0.0013932357542216778, + "learning_rate": 0.24101171758714346, + "loss": 0.3675, + "num_input_tokens_seen": 24449200, + "step": 11700 + }, + { + "epoch": 1.909535851211355, + "grad_norm": 0.0008598203421570361, + "learning_rate": 0.24096488729002086, + "loss": 0.3044, + "num_input_tokens_seen": 24459552, + "step": 11705 + }, + { + "epoch": 1.9103515784321723, + "grad_norm": 0.0009658256894908845, + "learning_rate": 0.24091804296496946, + "loss": 0.3454, + "num_input_tokens_seen": 24469008, + "step": 11710 + }, + { + "epoch": 1.9111673056529896, + "grad_norm": 0.0009332073968835175, + "learning_rate": 0.2408711846192133, + "loss": 0.3012, + "num_input_tokens_seen": 24480576, + "step": 11715 + }, + { + "epoch": 1.911983032873807, + "grad_norm": 0.0007100806687958539, + "learning_rate": 0.24082431225997855, + "loss": 0.2466, + "num_input_tokens_seen": 24489856, + "step": 11720 + }, + { + "epoch": 1.9127987600946244, + "grad_norm": 0.00099766266066581, + "learning_rate": 0.24077742589449344, + "loss": 0.2957, + "num_input_tokens_seen": 24500080, + "step": 11725 + }, + { + "epoch": 1.9136144873154417, + "grad_norm": 0.001186452922411263, + "learning_rate": 0.24073052552998844, + "loss": 0.3341, + "num_input_tokens_seen": 24510480, + "step": 11730 + }, + { + "epoch": 1.914430214536259, + "grad_norm": 0.0005835727788507938, + "learning_rate": 0.2406836111736963, + "loss": 0.3094, + "num_input_tokens_seen": 24519168, + "step": 11735 + }, + { + "epoch": 1.9152459417570764, + "grad_norm": 0.0006932538817636669, + "learning_rate": 0.2406366828328517, + "loss": 0.3272, + "num_input_tokens_seen": 24528720, + "step": 11740 + }, + { + "epoch": 1.9160616689778938, + "grad_norm": 0.0007186182192526758, + "learning_rate": 0.2405897405146915, + "loss": 0.3207, + "num_input_tokens_seen": 24539408, + "step": 11745 + }, + { + "epoch": 1.9168773961987111, + "grad_norm": 0.0003779531689360738, + "learning_rate": 0.240542784226455, + "loss": 0.3413, + "num_input_tokens_seen": 24549280, + "step": 11750 + }, + { + "epoch": 1.9176931234195285, + "grad_norm": 0.0005244969506748021, + "learning_rate": 0.24049581397538328, + "loss": 0.2908, + "num_input_tokens_seen": 24559536, + "step": 11755 + }, + { + "epoch": 1.9185088506403458, + "grad_norm": 0.0010990076698362827, + "learning_rate": 0.24044882976871984, + "loss": 0.3192, + "num_input_tokens_seen": 24571120, + "step": 11760 + }, + { + "epoch": 1.9193245778611632, + "grad_norm": 0.0018004877492785454, + "learning_rate": 0.2404018316137102, + "loss": 0.3662, + "num_input_tokens_seen": 24582864, + "step": 11765 + }, + { + "epoch": 1.9201403050819805, + "grad_norm": 0.0013074768939986825, + "learning_rate": 0.24035481951760204, + "loss": 0.3127, + "num_input_tokens_seen": 24592016, + "step": 11770 + }, + { + "epoch": 1.920956032302798, + "grad_norm": 0.000627107045147568, + "learning_rate": 0.2403077934876452, + "loss": 0.2996, + "num_input_tokens_seen": 24603728, + "step": 11775 + }, + { + "epoch": 1.9217717595236152, + "grad_norm": 0.0004447213723324239, + "learning_rate": 0.2402607535310918, + "loss": 0.333, + "num_input_tokens_seen": 24612416, + "step": 11780 + }, + { + "epoch": 1.9225874867444328, + "grad_norm": 0.0006266366690397263, + "learning_rate": 0.2402136996551959, + "loss": 0.3074, + "num_input_tokens_seen": 24624080, + "step": 11785 + }, + { + "epoch": 1.92340321396525, + "grad_norm": 0.0007717427215538919, + "learning_rate": 0.24016663186721376, + "loss": 0.3188, + "num_input_tokens_seen": 24633392, + "step": 11790 + }, + { + "epoch": 1.9242189411860675, + "grad_norm": 0.0006746497238054872, + "learning_rate": 0.24011955017440395, + "loss": 0.3412, + "num_input_tokens_seen": 24643712, + "step": 11795 + }, + { + "epoch": 1.9250346684068846, + "grad_norm": 0.0012356452643871307, + "learning_rate": 0.24007245458402696, + "loss": 0.3605, + "num_input_tokens_seen": 24655408, + "step": 11800 + }, + { + "epoch": 1.9250346684068846, + "eval_loss": 0.327572762966156, + "eval_runtime": 155.8092, + "eval_samples_per_second": 17.489, + "eval_steps_per_second": 8.748, + "num_input_tokens_seen": 24655408, + "step": 11800 + }, + { + "epoch": 1.9258503956277022, + "grad_norm": 0.0007325527258217335, + "learning_rate": 0.2400253451033456, + "loss": 0.3434, + "num_input_tokens_seen": 24666144, + "step": 11805 + }, + { + "epoch": 1.9266661228485193, + "grad_norm": 0.00033369308221153915, + "learning_rate": 0.23997822173962463, + "loss": 0.3523, + "num_input_tokens_seen": 24677264, + "step": 11810 + }, + { + "epoch": 1.927481850069337, + "grad_norm": 0.0006018911371938884, + "learning_rate": 0.23993108450013118, + "loss": 0.3202, + "num_input_tokens_seen": 24688864, + "step": 11815 + }, + { + "epoch": 1.928297577290154, + "grad_norm": 0.0008318384643644094, + "learning_rate": 0.2398839333921343, + "loss": 0.294, + "num_input_tokens_seen": 24697600, + "step": 11820 + }, + { + "epoch": 1.9291133045109716, + "grad_norm": 0.0005380494403652847, + "learning_rate": 0.23983676842290536, + "loss": 0.3224, + "num_input_tokens_seen": 24708608, + "step": 11825 + }, + { + "epoch": 1.9299290317317888, + "grad_norm": 0.0009657593327574432, + "learning_rate": 0.2397895895997178, + "loss": 0.2541, + "num_input_tokens_seen": 24718944, + "step": 11830 + }, + { + "epoch": 1.9307447589526063, + "grad_norm": 0.0012934772530570626, + "learning_rate": 0.23974239692984714, + "loss": 0.3209, + "num_input_tokens_seen": 24729152, + "step": 11835 + }, + { + "epoch": 1.9315604861734235, + "grad_norm": 0.0004148321459069848, + "learning_rate": 0.2396951904205711, + "loss": 0.3132, + "num_input_tokens_seen": 24739152, + "step": 11840 + }, + { + "epoch": 1.932376213394241, + "grad_norm": 0.001348873833194375, + "learning_rate": 0.23964797007916952, + "loss": 0.3959, + "num_input_tokens_seen": 24748384, + "step": 11845 + }, + { + "epoch": 1.9331919406150582, + "grad_norm": 0.00036095367977395654, + "learning_rate": 0.23960073591292436, + "loss": 0.2512, + "num_input_tokens_seen": 24760544, + "step": 11850 + }, + { + "epoch": 1.9340076678358757, + "grad_norm": 0.0004492023726925254, + "learning_rate": 0.2395534879291197, + "loss": 0.3553, + "num_input_tokens_seen": 24769376, + "step": 11855 + }, + { + "epoch": 1.934823395056693, + "grad_norm": 0.0009590869885869324, + "learning_rate": 0.23950622613504186, + "loss": 0.3768, + "num_input_tokens_seen": 24778816, + "step": 11860 + }, + { + "epoch": 1.9356391222775104, + "grad_norm": 0.000869792711455375, + "learning_rate": 0.2394589505379791, + "loss": 0.3798, + "num_input_tokens_seen": 24789696, + "step": 11865 + }, + { + "epoch": 1.9364548494983278, + "grad_norm": 0.00043572220602072775, + "learning_rate": 0.23941166114522197, + "loss": 0.3458, + "num_input_tokens_seen": 24799536, + "step": 11870 + }, + { + "epoch": 1.9372705767191452, + "grad_norm": 0.00044214577064849436, + "learning_rate": 0.23936435796406308, + "loss": 0.3099, + "num_input_tokens_seen": 24809232, + "step": 11875 + }, + { + "epoch": 1.9380863039399625, + "grad_norm": 0.001028719125315547, + "learning_rate": 0.23931704100179715, + "loss": 0.2879, + "num_input_tokens_seen": 24819872, + "step": 11880 + }, + { + "epoch": 1.9389020311607799, + "grad_norm": 0.0008773903828114271, + "learning_rate": 0.2392697102657211, + "loss": 0.3141, + "num_input_tokens_seen": 24831072, + "step": 11885 + }, + { + "epoch": 1.9397177583815972, + "grad_norm": 0.00040104956133291125, + "learning_rate": 0.23922236576313388, + "loss": 0.3009, + "num_input_tokens_seen": 24842384, + "step": 11890 + }, + { + "epoch": 1.9405334856024146, + "grad_norm": 0.00044121898827143013, + "learning_rate": 0.2391750075013366, + "loss": 0.3507, + "num_input_tokens_seen": 24852016, + "step": 11895 + }, + { + "epoch": 1.941349212823232, + "grad_norm": 0.0004069545248057693, + "learning_rate": 0.2391276354876326, + "loss": 0.2895, + "num_input_tokens_seen": 24862448, + "step": 11900 + }, + { + "epoch": 1.9421649400440493, + "grad_norm": 0.00029158033430576324, + "learning_rate": 0.23908024972932707, + "loss": 0.3174, + "num_input_tokens_seen": 24871696, + "step": 11905 + }, + { + "epoch": 1.9429806672648666, + "grad_norm": 0.0005405833944678307, + "learning_rate": 0.2390328502337276, + "loss": 0.4059, + "num_input_tokens_seen": 24882912, + "step": 11910 + }, + { + "epoch": 1.943796394485684, + "grad_norm": 0.0004444158112164587, + "learning_rate": 0.23898543700814376, + "loss": 0.282, + "num_input_tokens_seen": 24892256, + "step": 11915 + }, + { + "epoch": 1.9446121217065013, + "grad_norm": 0.0005511933122761548, + "learning_rate": 0.2389380100598873, + "loss": 0.3493, + "num_input_tokens_seen": 24902048, + "step": 11920 + }, + { + "epoch": 1.9454278489273187, + "grad_norm": 0.0009346286533400416, + "learning_rate": 0.23889056939627207, + "loss": 0.3343, + "num_input_tokens_seen": 24912096, + "step": 11925 + }, + { + "epoch": 1.946243576148136, + "grad_norm": 0.0007933284505270422, + "learning_rate": 0.23884311502461386, + "loss": 0.3123, + "num_input_tokens_seen": 24923632, + "step": 11930 + }, + { + "epoch": 1.9470593033689534, + "grad_norm": 0.001036580535583198, + "learning_rate": 0.23879564695223088, + "loss": 0.3798, + "num_input_tokens_seen": 24933408, + "step": 11935 + }, + { + "epoch": 1.9478750305897707, + "grad_norm": 0.0004948544083163142, + "learning_rate": 0.23874816518644332, + "loss": 0.2953, + "num_input_tokens_seen": 24944400, + "step": 11940 + }, + { + "epoch": 1.948690757810588, + "grad_norm": 0.0003407506155781448, + "learning_rate": 0.23870066973457335, + "loss": 0.3144, + "num_input_tokens_seen": 24955440, + "step": 11945 + }, + { + "epoch": 1.9495064850314054, + "grad_norm": 0.000606115092523396, + "learning_rate": 0.23865316060394545, + "loss": 0.3375, + "num_input_tokens_seen": 24966176, + "step": 11950 + }, + { + "epoch": 1.9503222122522228, + "grad_norm": 0.00036320698563940823, + "learning_rate": 0.2386056378018861, + "loss": 0.3241, + "num_input_tokens_seen": 24976736, + "step": 11955 + }, + { + "epoch": 1.9511379394730404, + "grad_norm": 0.00032050072331912816, + "learning_rate": 0.2385581013357239, + "loss": 0.3113, + "num_input_tokens_seen": 24985568, + "step": 11960 + }, + { + "epoch": 1.9519536666938575, + "grad_norm": 0.0009513369295746088, + "learning_rate": 0.23851055121278958, + "loss": 0.2772, + "num_input_tokens_seen": 24996992, + "step": 11965 + }, + { + "epoch": 1.952769393914675, + "grad_norm": 0.0005038348608650267, + "learning_rate": 0.23846298744041594, + "loss": 0.2942, + "num_input_tokens_seen": 25008896, + "step": 11970 + }, + { + "epoch": 1.9535851211354922, + "grad_norm": 0.0011932032648473978, + "learning_rate": 0.23841541002593802, + "loss": 0.3243, + "num_input_tokens_seen": 25020304, + "step": 11975 + }, + { + "epoch": 1.9544008483563098, + "grad_norm": 0.0006781602860428393, + "learning_rate": 0.23836781897669276, + "loss": 0.3388, + "num_input_tokens_seen": 25030384, + "step": 11980 + }, + { + "epoch": 1.955216575577127, + "grad_norm": 0.0005204707849770784, + "learning_rate": 0.23832021430001926, + "loss": 0.3409, + "num_input_tokens_seen": 25040608, + "step": 11985 + }, + { + "epoch": 1.9560323027979445, + "grad_norm": 0.0008048043237067759, + "learning_rate": 0.2382725960032588, + "loss": 0.3018, + "num_input_tokens_seen": 25051680, + "step": 11990 + }, + { + "epoch": 1.9568480300187616, + "grad_norm": 0.0014248902443796396, + "learning_rate": 0.23822496409375482, + "loss": 0.2577, + "num_input_tokens_seen": 25062624, + "step": 11995 + }, + { + "epoch": 1.9576637572395792, + "grad_norm": 0.0005623481702059507, + "learning_rate": 0.2381773185788526, + "loss": 0.2571, + "num_input_tokens_seen": 25074096, + "step": 12000 + }, + { + "epoch": 1.9576637572395792, + "eval_loss": 0.32312074303627014, + "eval_runtime": 155.6195, + "eval_samples_per_second": 17.511, + "eval_steps_per_second": 8.759, + "num_input_tokens_seen": 25074096, + "step": 12000 + }, + { + "epoch": 1.9584794844603963, + "grad_norm": 0.0013754336396232247, + "learning_rate": 0.2381296594658998, + "loss": 0.3185, + "num_input_tokens_seen": 25084000, + "step": 12005 + }, + { + "epoch": 1.959295211681214, + "grad_norm": 0.0008309163968078792, + "learning_rate": 0.238081986762246, + "loss": 0.3718, + "num_input_tokens_seen": 25093232, + "step": 12010 + }, + { + "epoch": 1.960110938902031, + "grad_norm": 0.0011993462685495615, + "learning_rate": 0.23803430047524293, + "loss": 0.2899, + "num_input_tokens_seen": 25103312, + "step": 12015 + }, + { + "epoch": 1.9609266661228486, + "grad_norm": 0.0006691439193673432, + "learning_rate": 0.23798660061224441, + "loss": 0.3664, + "num_input_tokens_seen": 25113136, + "step": 12020 + }, + { + "epoch": 1.9617423933436657, + "grad_norm": 0.0005056218942627311, + "learning_rate": 0.23793888718060632, + "loss": 0.2849, + "num_input_tokens_seen": 25124128, + "step": 12025 + }, + { + "epoch": 1.9625581205644833, + "grad_norm": 0.0005737290484830737, + "learning_rate": 0.23789116018768675, + "loss": 0.3347, + "num_input_tokens_seen": 25134960, + "step": 12030 + }, + { + "epoch": 1.9633738477853004, + "grad_norm": 0.0017425337573513389, + "learning_rate": 0.2378434196408458, + "loss": 0.3885, + "num_input_tokens_seen": 25146896, + "step": 12035 + }, + { + "epoch": 1.964189575006118, + "grad_norm": 0.0009305261773988605, + "learning_rate": 0.23779566554744563, + "loss": 0.3569, + "num_input_tokens_seen": 25157648, + "step": 12040 + }, + { + "epoch": 1.9650053022269351, + "grad_norm": 0.0007111672312021255, + "learning_rate": 0.23774789791485051, + "loss": 0.3547, + "num_input_tokens_seen": 25169040, + "step": 12045 + }, + { + "epoch": 1.9658210294477527, + "grad_norm": 0.0006880752043798566, + "learning_rate": 0.2377001167504268, + "loss": 0.3648, + "num_input_tokens_seen": 25178944, + "step": 12050 + }, + { + "epoch": 1.96663675666857, + "grad_norm": 0.0004984018160030246, + "learning_rate": 0.23765232206154302, + "loss": 0.339, + "num_input_tokens_seen": 25187888, + "step": 12055 + }, + { + "epoch": 1.9674524838893874, + "grad_norm": 0.000775065622292459, + "learning_rate": 0.23760451385556966, + "loss": 0.3358, + "num_input_tokens_seen": 25199440, + "step": 12060 + }, + { + "epoch": 1.9682682111102048, + "grad_norm": 0.0003963966737501323, + "learning_rate": 0.23755669213987932, + "loss": 0.3134, + "num_input_tokens_seen": 25209616, + "step": 12065 + }, + { + "epoch": 1.9690839383310221, + "grad_norm": 0.0007306751213036478, + "learning_rate": 0.23750885692184676, + "loss": 0.2818, + "num_input_tokens_seen": 25220544, + "step": 12070 + }, + { + "epoch": 1.9698996655518395, + "grad_norm": 0.0007001912454143167, + "learning_rate": 0.23746100820884875, + "loss": 0.3273, + "num_input_tokens_seen": 25231424, + "step": 12075 + }, + { + "epoch": 1.9707153927726568, + "grad_norm": 0.0004575422208290547, + "learning_rate": 0.23741314600826421, + "loss": 0.347, + "num_input_tokens_seen": 25241712, + "step": 12080 + }, + { + "epoch": 1.9715311199934742, + "grad_norm": 0.00043905279017053545, + "learning_rate": 0.23736527032747406, + "loss": 0.3095, + "num_input_tokens_seen": 25252192, + "step": 12085 + }, + { + "epoch": 1.9723468472142915, + "grad_norm": 0.00035605093580670655, + "learning_rate": 0.23731738117386128, + "loss": 0.3123, + "num_input_tokens_seen": 25262416, + "step": 12090 + }, + { + "epoch": 1.973162574435109, + "grad_norm": 0.0009841490536928177, + "learning_rate": 0.237269478554811, + "loss": 0.344, + "num_input_tokens_seen": 25272464, + "step": 12095 + }, + { + "epoch": 1.9739783016559262, + "grad_norm": 0.000426364887971431, + "learning_rate": 0.23722156247771053, + "loss": 0.3361, + "num_input_tokens_seen": 25283248, + "step": 12100 + }, + { + "epoch": 1.9747940288767436, + "grad_norm": 0.0008489641477353871, + "learning_rate": 0.23717363294994895, + "loss": 0.3331, + "num_input_tokens_seen": 25292512, + "step": 12105 + }, + { + "epoch": 1.975609756097561, + "grad_norm": 0.001867296057753265, + "learning_rate": 0.2371256899789177, + "loss": 0.34, + "num_input_tokens_seen": 25303888, + "step": 12110 + }, + { + "epoch": 1.9764254833183783, + "grad_norm": 0.0006646796828135848, + "learning_rate": 0.23707773357201017, + "loss": 0.3402, + "num_input_tokens_seen": 25314464, + "step": 12115 + }, + { + "epoch": 1.9772412105391957, + "grad_norm": 0.0006694598705507815, + "learning_rate": 0.2370297637366218, + "loss": 0.321, + "num_input_tokens_seen": 25324080, + "step": 12120 + }, + { + "epoch": 1.978056937760013, + "grad_norm": 0.00063502904959023, + "learning_rate": 0.23698178048015026, + "loss": 0.3109, + "num_input_tokens_seen": 25334816, + "step": 12125 + }, + { + "epoch": 1.9788726649808304, + "grad_norm": 0.0005175643018446863, + "learning_rate": 0.236933783809995, + "loss": 0.3306, + "num_input_tokens_seen": 25345280, + "step": 12130 + }, + { + "epoch": 1.9796883922016477, + "grad_norm": 0.0004520883085206151, + "learning_rate": 0.23688577373355785, + "loss": 0.3311, + "num_input_tokens_seen": 25356624, + "step": 12135 + }, + { + "epoch": 1.980504119422465, + "grad_norm": 0.0006125399377197027, + "learning_rate": 0.23683775025824247, + "loss": 0.3144, + "num_input_tokens_seen": 25367376, + "step": 12140 + }, + { + "epoch": 1.9813198466432826, + "grad_norm": 0.0013344187755137682, + "learning_rate": 0.2367897133914548, + "loss": 0.3198, + "num_input_tokens_seen": 25377584, + "step": 12145 + }, + { + "epoch": 1.9821355738640998, + "grad_norm": 0.0003397375694476068, + "learning_rate": 0.2367416631406026, + "loss": 0.3254, + "num_input_tokens_seen": 25386672, + "step": 12150 + }, + { + "epoch": 1.9829513010849174, + "grad_norm": 0.0007913331501185894, + "learning_rate": 0.23669359951309588, + "loss": 0.3095, + "num_input_tokens_seen": 25396320, + "step": 12155 + }, + { + "epoch": 1.9837670283057345, + "grad_norm": 0.0010570166632533073, + "learning_rate": 0.23664552251634666, + "loss": 0.329, + "num_input_tokens_seen": 25406208, + "step": 12160 + }, + { + "epoch": 1.984582755526552, + "grad_norm": 0.0009891169611364603, + "learning_rate": 0.23659743215776907, + "loss": 0.3306, + "num_input_tokens_seen": 25417344, + "step": 12165 + }, + { + "epoch": 1.9853984827473692, + "grad_norm": 0.000609717215411365, + "learning_rate": 0.23654932844477908, + "loss": 0.3244, + "num_input_tokens_seen": 25427200, + "step": 12170 + }, + { + "epoch": 1.9862142099681868, + "grad_norm": 0.0007397013832814991, + "learning_rate": 0.23650121138479507, + "loss": 0.2973, + "num_input_tokens_seen": 25437872, + "step": 12175 + }, + { + "epoch": 1.987029937189004, + "grad_norm": 0.001087819691747427, + "learning_rate": 0.23645308098523724, + "loss": 0.3577, + "num_input_tokens_seen": 25447520, + "step": 12180 + }, + { + "epoch": 1.9878456644098215, + "grad_norm": 0.0008591049700044096, + "learning_rate": 0.23640493725352785, + "loss": 0.3449, + "num_input_tokens_seen": 25458304, + "step": 12185 + }, + { + "epoch": 1.9886613916306386, + "grad_norm": 0.0005513931391760707, + "learning_rate": 0.2363567801970913, + "loss": 0.309, + "num_input_tokens_seen": 25467952, + "step": 12190 + }, + { + "epoch": 1.9894771188514562, + "grad_norm": 0.0004045482783112675, + "learning_rate": 0.236308609823354, + "loss": 0.3129, + "num_input_tokens_seen": 25478480, + "step": 12195 + }, + { + "epoch": 1.9902928460722733, + "grad_norm": 0.00035746587673202157, + "learning_rate": 0.23626042613974452, + "loss": 0.3465, + "num_input_tokens_seen": 25489056, + "step": 12200 + }, + { + "epoch": 1.9902928460722733, + "eval_loss": 0.3185650110244751, + "eval_runtime": 155.8878, + "eval_samples_per_second": 17.481, + "eval_steps_per_second": 8.743, + "num_input_tokens_seen": 25489056, + "step": 12200 + }, + { + "epoch": 1.9911085732930909, + "grad_norm": 0.0013255660887807608, + "learning_rate": 0.23621222915369325, + "loss": 0.2904, + "num_input_tokens_seen": 25498816, + "step": 12205 + }, + { + "epoch": 1.991924300513908, + "grad_norm": 0.0005808486603200436, + "learning_rate": 0.23616401887263283, + "loss": 0.339, + "num_input_tokens_seen": 25509696, + "step": 12210 + }, + { + "epoch": 1.9927400277347256, + "grad_norm": 0.0005593015812337399, + "learning_rate": 0.23611579530399793, + "loss": 0.3716, + "num_input_tokens_seen": 25519360, + "step": 12215 + }, + { + "epoch": 1.9935557549555427, + "grad_norm": 0.0007167815347202122, + "learning_rate": 0.23606755845522517, + "loss": 0.3632, + "num_input_tokens_seen": 25530288, + "step": 12220 + }, + { + "epoch": 1.9943714821763603, + "grad_norm": 0.001334213768132031, + "learning_rate": 0.23601930833375329, + "loss": 0.3528, + "num_input_tokens_seen": 25539632, + "step": 12225 + }, + { + "epoch": 1.9951872093971774, + "grad_norm": 0.0002740130585152656, + "learning_rate": 0.23597104494702312, + "loss": 0.2776, + "num_input_tokens_seen": 25549744, + "step": 12230 + }, + { + "epoch": 1.996002936617995, + "grad_norm": 0.0018761195242404938, + "learning_rate": 0.23592276830247744, + "loss": 0.3646, + "num_input_tokens_seen": 25558864, + "step": 12235 + }, + { + "epoch": 1.9968186638388123, + "grad_norm": 0.0007563342223875225, + "learning_rate": 0.2358744784075611, + "loss": 0.3196, + "num_input_tokens_seen": 25568784, + "step": 12240 + }, + { + "epoch": 1.9976343910596297, + "grad_norm": 0.0004745338228531182, + "learning_rate": 0.235826175269721, + "loss": 0.2823, + "num_input_tokens_seen": 25578816, + "step": 12245 + }, + { + "epoch": 1.998450118280447, + "grad_norm": 0.0008074515499174595, + "learning_rate": 0.23577785889640612, + "loss": 0.3476, + "num_input_tokens_seen": 25589696, + "step": 12250 + }, + { + "epoch": 1.9992658455012644, + "grad_norm": 0.000496081484016031, + "learning_rate": 0.23572952929506744, + "loss": 0.316, + "num_input_tokens_seen": 25601184, + "step": 12255 + }, + { + "epoch": 2.0, + "grad_norm": 0.001373382518067956, + "learning_rate": 0.23568118647315803, + "loss": 0.4412, + "num_input_tokens_seen": 25610688, + "step": 12260 + }, + { + "epoch": 2.0008157272208176, + "grad_norm": 0.0006567779346369207, + "learning_rate": 0.23563283043813296, + "loss": 0.3484, + "num_input_tokens_seen": 25620368, + "step": 12265 + }, + { + "epoch": 2.0016314544416347, + "grad_norm": 0.0006510523962788284, + "learning_rate": 0.23558446119744922, + "loss": 0.3062, + "num_input_tokens_seen": 25629664, + "step": 12270 + }, + { + "epoch": 2.0024471816624523, + "grad_norm": 0.0006078559090383351, + "learning_rate": 0.23553607875856608, + "loss": 0.3387, + "num_input_tokens_seen": 25640880, + "step": 12275 + }, + { + "epoch": 2.0032629088832694, + "grad_norm": 0.0011048808228224516, + "learning_rate": 0.2354876831289447, + "loss": 0.3356, + "num_input_tokens_seen": 25650480, + "step": 12280 + }, + { + "epoch": 2.004078636104087, + "grad_norm": 0.0008587336051277816, + "learning_rate": 0.23543927431604827, + "loss": 0.3045, + "num_input_tokens_seen": 25659424, + "step": 12285 + }, + { + "epoch": 2.004894363324904, + "grad_norm": 0.0009138171444647014, + "learning_rate": 0.23539085232734203, + "loss": 0.3438, + "num_input_tokens_seen": 25669408, + "step": 12290 + }, + { + "epoch": 2.0057100905457217, + "grad_norm": 0.0005579369608312845, + "learning_rate": 0.2353424171702933, + "loss": 0.2968, + "num_input_tokens_seen": 25678880, + "step": 12295 + }, + { + "epoch": 2.006525817766539, + "grad_norm": 0.0005754713201895356, + "learning_rate": 0.23529396885237133, + "loss": 0.3431, + "num_input_tokens_seen": 25689760, + "step": 12300 + }, + { + "epoch": 2.0073415449873564, + "grad_norm": 0.0009955493733286858, + "learning_rate": 0.2352455073810475, + "loss": 0.3673, + "num_input_tokens_seen": 25700656, + "step": 12305 + }, + { + "epoch": 2.0081572722081735, + "grad_norm": 0.0002910990151576698, + "learning_rate": 0.23519703276379517, + "loss": 0.3098, + "num_input_tokens_seen": 25711920, + "step": 12310 + }, + { + "epoch": 2.008972999428991, + "grad_norm": 0.0006108647212386131, + "learning_rate": 0.2351485450080897, + "loss": 0.3121, + "num_input_tokens_seen": 25721008, + "step": 12315 + }, + { + "epoch": 2.0097887266498082, + "grad_norm": 0.0005000095115974545, + "learning_rate": 0.2351000441214086, + "loss": 0.3324, + "num_input_tokens_seen": 25732000, + "step": 12320 + }, + { + "epoch": 2.010604453870626, + "grad_norm": 0.0003912453830707818, + "learning_rate": 0.23505153011123125, + "loss": 0.2802, + "num_input_tokens_seen": 25742368, + "step": 12325 + }, + { + "epoch": 2.011420181091443, + "grad_norm": 0.0005252975970506668, + "learning_rate": 0.23500300298503912, + "loss": 0.3109, + "num_input_tokens_seen": 25752880, + "step": 12330 + }, + { + "epoch": 2.0122359083122605, + "grad_norm": 0.001405200338922441, + "learning_rate": 0.23495446275031576, + "loss": 0.3533, + "num_input_tokens_seen": 25761968, + "step": 12335 + }, + { + "epoch": 2.0130516355330776, + "grad_norm": 0.0004162219993304461, + "learning_rate": 0.2349059094145466, + "loss": 0.3172, + "num_input_tokens_seen": 25772688, + "step": 12340 + }, + { + "epoch": 2.013867362753895, + "grad_norm": 0.0006479172152467072, + "learning_rate": 0.2348573429852192, + "loss": 0.333, + "num_input_tokens_seen": 25783792, + "step": 12345 + }, + { + "epoch": 2.0146830899747123, + "grad_norm": 0.0009452527738176286, + "learning_rate": 0.23480876346982313, + "loss": 0.294, + "num_input_tokens_seen": 25794128, + "step": 12350 + }, + { + "epoch": 2.01549881719553, + "grad_norm": 0.0007514213211834431, + "learning_rate": 0.23476017087585, + "loss": 0.2856, + "num_input_tokens_seen": 25805056, + "step": 12355 + }, + { + "epoch": 2.016314544416347, + "grad_norm": 0.0008858628571033478, + "learning_rate": 0.23471156521079334, + "loss": 0.2986, + "num_input_tokens_seen": 25814320, + "step": 12360 + }, + { + "epoch": 2.0171302716371646, + "grad_norm": 0.0007719406276009977, + "learning_rate": 0.23466294648214875, + "loss": 0.3443, + "num_input_tokens_seen": 25825520, + "step": 12365 + }, + { + "epoch": 2.0179459988579818, + "grad_norm": 0.0006355312070809305, + "learning_rate": 0.2346143146974139, + "loss": 0.2985, + "num_input_tokens_seen": 25836208, + "step": 12370 + }, + { + "epoch": 2.0187617260787993, + "grad_norm": 0.0009346420411020517, + "learning_rate": 0.23456566986408836, + "loss": 0.3575, + "num_input_tokens_seen": 25844720, + "step": 12375 + }, + { + "epoch": 2.0195774532996165, + "grad_norm": 0.0008880929090082645, + "learning_rate": 0.23451701198967384, + "loss": 0.3286, + "num_input_tokens_seen": 25854864, + "step": 12380 + }, + { + "epoch": 2.020393180520434, + "grad_norm": 0.0005017244257032871, + "learning_rate": 0.23446834108167397, + "loss": 0.3577, + "num_input_tokens_seen": 25865968, + "step": 12385 + }, + { + "epoch": 2.021208907741251, + "grad_norm": 0.0006110953981988132, + "learning_rate": 0.23441965714759438, + "loss": 0.3363, + "num_input_tokens_seen": 25876128, + "step": 12390 + }, + { + "epoch": 2.0220246349620687, + "grad_norm": 0.0009335332433693111, + "learning_rate": 0.23437096019494277, + "loss": 0.3151, + "num_input_tokens_seen": 25886736, + "step": 12395 + }, + { + "epoch": 2.022840362182886, + "grad_norm": 0.0005141337751410902, + "learning_rate": 0.23432225023122885, + "loss": 0.4011, + "num_input_tokens_seen": 25898496, + "step": 12400 + }, + { + "epoch": 2.022840362182886, + "eval_loss": 0.31951773166656494, + "eval_runtime": 155.9079, + "eval_samples_per_second": 17.478, + "eval_steps_per_second": 8.742, + "num_input_tokens_seen": 25898496, + "step": 12400 + }, + { + "epoch": 2.0236560894037035, + "grad_norm": 0.000614762248005718, + "learning_rate": 0.23427352726396428, + "loss": 0.3049, + "num_input_tokens_seen": 25908560, + "step": 12405 + }, + { + "epoch": 2.0244718166245206, + "grad_norm": 0.0006234949687495828, + "learning_rate": 0.2342247913006628, + "loss": 0.3835, + "num_input_tokens_seen": 25919824, + "step": 12410 + }, + { + "epoch": 2.025287543845338, + "grad_norm": 0.0006215699249878526, + "learning_rate": 0.23417604234883999, + "loss": 0.3034, + "num_input_tokens_seen": 25929968, + "step": 12415 + }, + { + "epoch": 2.0261032710661553, + "grad_norm": 0.0007660859846509993, + "learning_rate": 0.23412728041601363, + "loss": 0.3144, + "num_input_tokens_seen": 25941216, + "step": 12420 + }, + { + "epoch": 2.026918998286973, + "grad_norm": 0.000313503056531772, + "learning_rate": 0.23407850550970347, + "loss": 0.3695, + "num_input_tokens_seen": 25950816, + "step": 12425 + }, + { + "epoch": 2.02773472550779, + "grad_norm": 0.0005810604779981077, + "learning_rate": 0.23402971763743116, + "loss": 0.2826, + "num_input_tokens_seen": 25962336, + "step": 12430 + }, + { + "epoch": 2.0285504527286076, + "grad_norm": 0.0006526351207867265, + "learning_rate": 0.23398091680672037, + "loss": 0.3335, + "num_input_tokens_seen": 25971072, + "step": 12435 + }, + { + "epoch": 2.0293661799494247, + "grad_norm": 0.0008115345262922347, + "learning_rate": 0.23393210302509687, + "loss": 0.2912, + "num_input_tokens_seen": 25982448, + "step": 12440 + }, + { + "epoch": 2.0301819071702423, + "grad_norm": 0.0023358259350061417, + "learning_rate": 0.23388327630008832, + "loss": 0.2984, + "num_input_tokens_seen": 25993344, + "step": 12445 + }, + { + "epoch": 2.03099763439106, + "grad_norm": 0.0009343879064545035, + "learning_rate": 0.23383443663922443, + "loss": 0.2714, + "num_input_tokens_seen": 26004288, + "step": 12450 + }, + { + "epoch": 2.031813361611877, + "grad_norm": 0.0005063461139798164, + "learning_rate": 0.23378558405003685, + "loss": 0.4569, + "num_input_tokens_seen": 26014928, + "step": 12455 + }, + { + "epoch": 2.0326290888326946, + "grad_norm": 0.0016168096335604787, + "learning_rate": 0.2337367185400593, + "loss": 0.2746, + "num_input_tokens_seen": 26025088, + "step": 12460 + }, + { + "epoch": 2.0334448160535117, + "grad_norm": 0.0007912159780971706, + "learning_rate": 0.23368784011682747, + "loss": 0.3282, + "num_input_tokens_seen": 26036816, + "step": 12465 + }, + { + "epoch": 2.0342605432743293, + "grad_norm": 0.0006607161485590041, + "learning_rate": 0.23363894878787902, + "loss": 0.3307, + "num_input_tokens_seen": 26047488, + "step": 12470 + }, + { + "epoch": 2.0350762704951464, + "grad_norm": 0.00041180034168064594, + "learning_rate": 0.23359004456075352, + "loss": 0.373, + "num_input_tokens_seen": 26058448, + "step": 12475 + }, + { + "epoch": 2.035891997715964, + "grad_norm": 0.000460305018350482, + "learning_rate": 0.23354112744299277, + "loss": 0.305, + "num_input_tokens_seen": 26069760, + "step": 12480 + }, + { + "epoch": 2.036707724936781, + "grad_norm": 0.0005201891763135791, + "learning_rate": 0.2334921974421403, + "loss": 0.3514, + "num_input_tokens_seen": 26079680, + "step": 12485 + }, + { + "epoch": 2.0375234521575987, + "grad_norm": 0.0004000047338195145, + "learning_rate": 0.23344325456574178, + "loss": 0.3201, + "num_input_tokens_seen": 26090800, + "step": 12490 + }, + { + "epoch": 2.038339179378416, + "grad_norm": 0.0006047594943083823, + "learning_rate": 0.23339429882134477, + "loss": 0.3796, + "num_input_tokens_seen": 26100384, + "step": 12495 + }, + { + "epoch": 2.0391549065992334, + "grad_norm": 0.0004349245282355696, + "learning_rate": 0.23334533021649884, + "loss": 0.329, + "num_input_tokens_seen": 26109712, + "step": 12500 + }, + { + "epoch": 2.0399706338200505, + "grad_norm": 0.0002677442680578679, + "learning_rate": 0.23329634875875566, + "loss": 0.2712, + "num_input_tokens_seen": 26121536, + "step": 12505 + }, + { + "epoch": 2.040786361040868, + "grad_norm": 0.0006410750211216509, + "learning_rate": 0.23324735445566874, + "loss": 0.2931, + "num_input_tokens_seen": 26133024, + "step": 12510 + }, + { + "epoch": 2.041602088261685, + "grad_norm": 0.00064668501727283, + "learning_rate": 0.2331983473147936, + "loss": 0.3569, + "num_input_tokens_seen": 26143376, + "step": 12515 + }, + { + "epoch": 2.042417815482503, + "grad_norm": 0.00042830288293771446, + "learning_rate": 0.23314932734368776, + "loss": 0.3347, + "num_input_tokens_seen": 26154608, + "step": 12520 + }, + { + "epoch": 2.04323354270332, + "grad_norm": 0.0009520506719127297, + "learning_rate": 0.2331002945499107, + "loss": 0.3025, + "num_input_tokens_seen": 26165248, + "step": 12525 + }, + { + "epoch": 2.0440492699241375, + "grad_norm": 0.0004469461855478585, + "learning_rate": 0.23305124894102397, + "loss": 0.3159, + "num_input_tokens_seen": 26175952, + "step": 12530 + }, + { + "epoch": 2.0448649971449546, + "grad_norm": 0.0007055606110952795, + "learning_rate": 0.23300219052459092, + "loss": 0.3634, + "num_input_tokens_seen": 26185984, + "step": 12535 + }, + { + "epoch": 2.045680724365772, + "grad_norm": 0.0007120921509340405, + "learning_rate": 0.23295311930817708, + "loss": 0.3036, + "num_input_tokens_seen": 26196656, + "step": 12540 + }, + { + "epoch": 2.0464964515865893, + "grad_norm": 0.0004857099847868085, + "learning_rate": 0.23290403529934972, + "loss": 0.3544, + "num_input_tokens_seen": 26206848, + "step": 12545 + }, + { + "epoch": 2.047312178807407, + "grad_norm": 0.0004043901280965656, + "learning_rate": 0.23285493850567832, + "loss": 0.3483, + "num_input_tokens_seen": 26216768, + "step": 12550 + }, + { + "epoch": 2.048127906028224, + "grad_norm": 0.0007205711444839835, + "learning_rate": 0.23280582893473414, + "loss": 0.3689, + "num_input_tokens_seen": 26226320, + "step": 12555 + }, + { + "epoch": 2.0489436332490416, + "grad_norm": 0.0007993800682015717, + "learning_rate": 0.2327567065940906, + "loss": 0.3044, + "num_input_tokens_seen": 26236448, + "step": 12560 + }, + { + "epoch": 2.0497593604698587, + "grad_norm": 0.0007322060992009938, + "learning_rate": 0.23270757149132285, + "loss": 0.3092, + "num_input_tokens_seen": 26247392, + "step": 12565 + }, + { + "epoch": 2.0505750876906763, + "grad_norm": 0.0007166403229348361, + "learning_rate": 0.23265842363400827, + "loss": 0.3665, + "num_input_tokens_seen": 26257872, + "step": 12570 + }, + { + "epoch": 2.0513908149114934, + "grad_norm": 0.0010994478361681104, + "learning_rate": 0.23260926302972595, + "loss": 0.3169, + "num_input_tokens_seen": 26267872, + "step": 12575 + }, + { + "epoch": 2.052206542132311, + "grad_norm": 0.0005669129313901067, + "learning_rate": 0.2325600896860572, + "loss": 0.2945, + "num_input_tokens_seen": 26279088, + "step": 12580 + }, + { + "epoch": 2.053022269353128, + "grad_norm": 0.0006533319246955216, + "learning_rate": 0.23251090361058505, + "loss": 0.3406, + "num_input_tokens_seen": 26289664, + "step": 12585 + }, + { + "epoch": 2.0538379965739457, + "grad_norm": 0.0011103353463113308, + "learning_rate": 0.23246170481089476, + "loss": 0.3286, + "num_input_tokens_seen": 26298768, + "step": 12590 + }, + { + "epoch": 2.054653723794763, + "grad_norm": 0.0003881265874952078, + "learning_rate": 0.23241249329457317, + "loss": 0.3062, + "num_input_tokens_seen": 26309216, + "step": 12595 + }, + { + "epoch": 2.0554694510155804, + "grad_norm": 0.0004886484821327031, + "learning_rate": 0.23236326906920957, + "loss": 0.3619, + "num_input_tokens_seen": 26319696, + "step": 12600 + }, + { + "epoch": 2.0554694510155804, + "eval_loss": 0.3205036222934723, + "eval_runtime": 155.8178, + "eval_samples_per_second": 17.488, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 26319696, + "step": 12600 + }, + { + "epoch": 2.0562851782363976, + "grad_norm": 0.00041948442230932415, + "learning_rate": 0.2323140321423948, + "loss": 0.3427, + "num_input_tokens_seen": 26329824, + "step": 12605 + }, + { + "epoch": 2.057100905457215, + "grad_norm": 0.0004952432936988771, + "learning_rate": 0.23226478252172184, + "loss": 0.3324, + "num_input_tokens_seen": 26340432, + "step": 12610 + }, + { + "epoch": 2.0579166326780323, + "grad_norm": 0.00035401133936829865, + "learning_rate": 0.23221552021478561, + "loss": 0.3418, + "num_input_tokens_seen": 26351760, + "step": 12615 + }, + { + "epoch": 2.05873235989885, + "grad_norm": 0.0003429055504966527, + "learning_rate": 0.232166245229183, + "loss": 0.3286, + "num_input_tokens_seen": 26361696, + "step": 12620 + }, + { + "epoch": 2.059548087119667, + "grad_norm": 0.0003586273523978889, + "learning_rate": 0.2321169575725128, + "loss": 0.3418, + "num_input_tokens_seen": 26370592, + "step": 12625 + }, + { + "epoch": 2.0603638143404845, + "grad_norm": 0.0007356209098361433, + "learning_rate": 0.23206765725237577, + "loss": 0.2645, + "num_input_tokens_seen": 26382208, + "step": 12630 + }, + { + "epoch": 2.0611795415613017, + "grad_norm": 0.0015206251991912723, + "learning_rate": 0.2320183442763747, + "loss": 0.3195, + "num_input_tokens_seen": 26392192, + "step": 12635 + }, + { + "epoch": 2.0619952687821193, + "grad_norm": 0.00037177716149017215, + "learning_rate": 0.23196901865211422, + "loss": 0.2666, + "num_input_tokens_seen": 26403872, + "step": 12640 + }, + { + "epoch": 2.062810996002937, + "grad_norm": 0.0005867932923138142, + "learning_rate": 0.231919680387201, + "loss": 0.3021, + "num_input_tokens_seen": 26415312, + "step": 12645 + }, + { + "epoch": 2.063626723223754, + "grad_norm": 0.0005945838056504726, + "learning_rate": 0.23187032948924358, + "loss": 0.3055, + "num_input_tokens_seen": 26426144, + "step": 12650 + }, + { + "epoch": 2.0644424504445715, + "grad_norm": 0.0005410637822933495, + "learning_rate": 0.23182096596585247, + "loss": 0.2657, + "num_input_tokens_seen": 26436112, + "step": 12655 + }, + { + "epoch": 2.0652581776653887, + "grad_norm": 0.0006639778730459511, + "learning_rate": 0.23177158982464025, + "loss": 0.3133, + "num_input_tokens_seen": 26447408, + "step": 12660 + }, + { + "epoch": 2.0660739048862062, + "grad_norm": 0.0006410842179320753, + "learning_rate": 0.23172220107322122, + "loss": 0.3533, + "num_input_tokens_seen": 26459504, + "step": 12665 + }, + { + "epoch": 2.0668896321070234, + "grad_norm": 0.0007228070753626525, + "learning_rate": 0.23167279971921184, + "loss": 0.2744, + "num_input_tokens_seen": 26469904, + "step": 12670 + }, + { + "epoch": 2.067705359327841, + "grad_norm": 0.0010310421930626035, + "learning_rate": 0.23162338577023034, + "loss": 0.3155, + "num_input_tokens_seen": 26479040, + "step": 12675 + }, + { + "epoch": 2.068521086548658, + "grad_norm": 0.0005993798258714378, + "learning_rate": 0.23157395923389704, + "loss": 0.2586, + "num_input_tokens_seen": 26488256, + "step": 12680 + }, + { + "epoch": 2.0693368137694756, + "grad_norm": 0.0001871114654932171, + "learning_rate": 0.2315245201178341, + "loss": 0.3158, + "num_input_tokens_seen": 26500016, + "step": 12685 + }, + { + "epoch": 2.0701525409902928, + "grad_norm": 0.0038453093729913235, + "learning_rate": 0.23147506842966564, + "loss": 0.3836, + "num_input_tokens_seen": 26510464, + "step": 12690 + }, + { + "epoch": 2.0709682682111104, + "grad_norm": 0.0005486314767040312, + "learning_rate": 0.23142560417701774, + "loss": 0.3344, + "num_input_tokens_seen": 26521344, + "step": 12695 + }, + { + "epoch": 2.0717839954319275, + "grad_norm": 0.0006865118630230427, + "learning_rate": 0.23137612736751845, + "loss": 0.268, + "num_input_tokens_seen": 26532416, + "step": 12700 + }, + { + "epoch": 2.072599722652745, + "grad_norm": 0.0020791159477084875, + "learning_rate": 0.23132663800879766, + "loss": 0.3669, + "num_input_tokens_seen": 26543520, + "step": 12705 + }, + { + "epoch": 2.073415449873562, + "grad_norm": 0.00034463670453988016, + "learning_rate": 0.2312771361084873, + "loss": 0.3088, + "num_input_tokens_seen": 26554352, + "step": 12710 + }, + { + "epoch": 2.0742311770943798, + "grad_norm": 0.00032573947100900114, + "learning_rate": 0.23122762167422112, + "loss": 0.3369, + "num_input_tokens_seen": 26565488, + "step": 12715 + }, + { + "epoch": 2.075046904315197, + "grad_norm": 0.0007034028531052172, + "learning_rate": 0.23117809471363493, + "loss": 0.2883, + "num_input_tokens_seen": 26576080, + "step": 12720 + }, + { + "epoch": 2.0758626315360145, + "grad_norm": 0.00020388078701216727, + "learning_rate": 0.23112855523436637, + "loss": 0.3047, + "num_input_tokens_seen": 26585968, + "step": 12725 + }, + { + "epoch": 2.0766783587568316, + "grad_norm": 0.0008867908618412912, + "learning_rate": 0.23107900324405511, + "loss": 0.3428, + "num_input_tokens_seen": 26597760, + "step": 12730 + }, + { + "epoch": 2.077494085977649, + "grad_norm": 0.0006835227832198143, + "learning_rate": 0.2310294387503426, + "loss": 0.2936, + "num_input_tokens_seen": 26607568, + "step": 12735 + }, + { + "epoch": 2.0783098131984663, + "grad_norm": 0.0019036201993003488, + "learning_rate": 0.23097986176087237, + "loss": 0.3587, + "num_input_tokens_seen": 26616912, + "step": 12740 + }, + { + "epoch": 2.079125540419284, + "grad_norm": 0.0003868371422868222, + "learning_rate": 0.23093027228328986, + "loss": 0.3162, + "num_input_tokens_seen": 26628064, + "step": 12745 + }, + { + "epoch": 2.079941267640101, + "grad_norm": 0.0004780509334523231, + "learning_rate": 0.23088067032524226, + "loss": 0.4024, + "num_input_tokens_seen": 26640128, + "step": 12750 + }, + { + "epoch": 2.0807569948609186, + "grad_norm": 0.0002754742163233459, + "learning_rate": 0.23083105589437888, + "loss": 0.2753, + "num_input_tokens_seen": 26650512, + "step": 12755 + }, + { + "epoch": 2.0815727220817357, + "grad_norm": 0.0006988613167777658, + "learning_rate": 0.23078142899835094, + "loss": 0.317, + "num_input_tokens_seen": 26661488, + "step": 12760 + }, + { + "epoch": 2.0823884493025533, + "grad_norm": 0.0005227453075349331, + "learning_rate": 0.23073178964481147, + "loss": 0.3088, + "num_input_tokens_seen": 26671744, + "step": 12765 + }, + { + "epoch": 2.0832041765233704, + "grad_norm": 0.001081263879314065, + "learning_rate": 0.2306821378414155, + "loss": 0.3036, + "num_input_tokens_seen": 26681824, + "step": 12770 + }, + { + "epoch": 2.084019903744188, + "grad_norm": 0.0008438954828307033, + "learning_rate": 0.2306324735958199, + "loss": 0.4119, + "num_input_tokens_seen": 26692400, + "step": 12775 + }, + { + "epoch": 2.084835630965005, + "grad_norm": 0.00046400996507145464, + "learning_rate": 0.23058279691568362, + "loss": 0.3186, + "num_input_tokens_seen": 26703200, + "step": 12780 + }, + { + "epoch": 2.0856513581858227, + "grad_norm": 0.000438651186414063, + "learning_rate": 0.23053310780866745, + "loss": 0.3072, + "num_input_tokens_seen": 26713824, + "step": 12785 + }, + { + "epoch": 2.08646708540664, + "grad_norm": 0.0004488517006393522, + "learning_rate": 0.23048340628243397, + "loss": 0.3111, + "num_input_tokens_seen": 26724832, + "step": 12790 + }, + { + "epoch": 2.0872828126274574, + "grad_norm": 0.0003743824490811676, + "learning_rate": 0.23043369234464783, + "loss": 0.2804, + "num_input_tokens_seen": 26735552, + "step": 12795 + }, + { + "epoch": 2.0880985398482745, + "grad_norm": 0.0007721097208559513, + "learning_rate": 0.2303839660029755, + "loss": 0.3846, + "num_input_tokens_seen": 26744896, + "step": 12800 + }, + { + "epoch": 2.0880985398482745, + "eval_loss": 0.31859099864959717, + "eval_runtime": 155.7211, + "eval_samples_per_second": 17.499, + "eval_steps_per_second": 8.753, + "num_input_tokens_seen": 26744896, + "step": 12800 + }, + { + "epoch": 2.088914267069092, + "grad_norm": 0.0006351171177811921, + "learning_rate": 0.23033422726508548, + "loss": 0.2989, + "num_input_tokens_seen": 26754400, + "step": 12805 + }, + { + "epoch": 2.0897299942899092, + "grad_norm": 0.0008030397002585232, + "learning_rate": 0.23028447613864808, + "loss": 0.4179, + "num_input_tokens_seen": 26765168, + "step": 12810 + }, + { + "epoch": 2.090545721510727, + "grad_norm": 0.00035730120725929737, + "learning_rate": 0.2302347126313355, + "loss": 0.292, + "num_input_tokens_seen": 26775520, + "step": 12815 + }, + { + "epoch": 2.0913614487315444, + "grad_norm": 0.0005502490093931556, + "learning_rate": 0.23018493675082197, + "loss": 0.3215, + "num_input_tokens_seen": 26784816, + "step": 12820 + }, + { + "epoch": 2.0921771759523615, + "grad_norm": 0.0005416418425738811, + "learning_rate": 0.2301351485047835, + "loss": 0.2695, + "num_input_tokens_seen": 26795056, + "step": 12825 + }, + { + "epoch": 2.092992903173179, + "grad_norm": 0.0006709761801175773, + "learning_rate": 0.23008534790089813, + "loss": 0.2658, + "num_input_tokens_seen": 26806992, + "step": 12830 + }, + { + "epoch": 2.0938086303939962, + "grad_norm": 0.0005678535089828074, + "learning_rate": 0.2300355349468457, + "loss": 0.3374, + "num_input_tokens_seen": 26817088, + "step": 12835 + }, + { + "epoch": 2.094624357614814, + "grad_norm": 0.0006063763867132366, + "learning_rate": 0.22998570965030793, + "loss": 0.359, + "num_input_tokens_seen": 26827392, + "step": 12840 + }, + { + "epoch": 2.095440084835631, + "grad_norm": 0.0007431244594044983, + "learning_rate": 0.22993587201896862, + "loss": 0.309, + "num_input_tokens_seen": 26836096, + "step": 12845 + }, + { + "epoch": 2.0962558120564485, + "grad_norm": 0.000945841078646481, + "learning_rate": 0.2298860220605133, + "loss": 0.3801, + "num_input_tokens_seen": 26846768, + "step": 12850 + }, + { + "epoch": 2.0970715392772656, + "grad_norm": 0.0003210263676010072, + "learning_rate": 0.22983615978262942, + "loss": 0.2885, + "num_input_tokens_seen": 26857536, + "step": 12855 + }, + { + "epoch": 2.097887266498083, + "grad_norm": 0.0003848833148367703, + "learning_rate": 0.22978628519300648, + "loss": 0.263, + "num_input_tokens_seen": 26868368, + "step": 12860 + }, + { + "epoch": 2.0987029937189003, + "grad_norm": 0.000829090888146311, + "learning_rate": 0.22973639829933568, + "loss": 0.3629, + "num_input_tokens_seen": 26877584, + "step": 12865 + }, + { + "epoch": 2.099518720939718, + "grad_norm": 0.0004665482265409082, + "learning_rate": 0.22968649910931027, + "loss": 0.3042, + "num_input_tokens_seen": 26889136, + "step": 12870 + }, + { + "epoch": 2.100334448160535, + "grad_norm": 0.0003985076618846506, + "learning_rate": 0.22963658763062528, + "loss": 0.2856, + "num_input_tokens_seen": 26899952, + "step": 12875 + }, + { + "epoch": 2.1011501753813526, + "grad_norm": 0.0008228457882069051, + "learning_rate": 0.22958666387097765, + "loss": 0.2729, + "num_input_tokens_seen": 26910688, + "step": 12880 + }, + { + "epoch": 2.1019659026021698, + "grad_norm": 0.0004041141946800053, + "learning_rate": 0.22953672783806633, + "loss": 0.3955, + "num_input_tokens_seen": 26922000, + "step": 12885 + }, + { + "epoch": 2.1027816298229873, + "grad_norm": 0.000409131491323933, + "learning_rate": 0.22948677953959207, + "loss": 0.2879, + "num_input_tokens_seen": 26932608, + "step": 12890 + }, + { + "epoch": 2.1035973570438045, + "grad_norm": 0.0007209430914372206, + "learning_rate": 0.2294368189832575, + "loss": 0.2794, + "num_input_tokens_seen": 26943520, + "step": 12895 + }, + { + "epoch": 2.104413084264622, + "grad_norm": 0.0005898734671063721, + "learning_rate": 0.2293868461767672, + "loss": 0.2632, + "num_input_tokens_seen": 26954048, + "step": 12900 + }, + { + "epoch": 2.105228811485439, + "grad_norm": 0.000632004695944488, + "learning_rate": 0.22933686112782758, + "loss": 0.3327, + "num_input_tokens_seen": 26965152, + "step": 12905 + }, + { + "epoch": 2.1060445387062567, + "grad_norm": 0.0005846835556440055, + "learning_rate": 0.22928686384414698, + "loss": 0.2769, + "num_input_tokens_seen": 26975120, + "step": 12910 + }, + { + "epoch": 2.106860265927074, + "grad_norm": 0.0007000279147177935, + "learning_rate": 0.22923685433343552, + "loss": 0.3313, + "num_input_tokens_seen": 26986816, + "step": 12915 + }, + { + "epoch": 2.1076759931478914, + "grad_norm": 0.0008653639233671129, + "learning_rate": 0.22918683260340542, + "loss": 0.3476, + "num_input_tokens_seen": 26998400, + "step": 12920 + }, + { + "epoch": 2.1084917203687086, + "grad_norm": 0.0008404765976592898, + "learning_rate": 0.2291367986617706, + "loss": 0.2843, + "num_input_tokens_seen": 27007616, + "step": 12925 + }, + { + "epoch": 2.109307447589526, + "grad_norm": 0.0024212454445660114, + "learning_rate": 0.22908675251624697, + "loss": 0.3689, + "num_input_tokens_seen": 27019168, + "step": 12930 + }, + { + "epoch": 2.1101231748103433, + "grad_norm": 0.0007021475466899574, + "learning_rate": 0.22903669417455216, + "loss": 0.3614, + "num_input_tokens_seen": 27030048, + "step": 12935 + }, + { + "epoch": 2.110938902031161, + "grad_norm": 0.0005748681141994894, + "learning_rate": 0.22898662364440592, + "loss": 0.2553, + "num_input_tokens_seen": 27040080, + "step": 12940 + }, + { + "epoch": 2.111754629251978, + "grad_norm": 0.0003544283681549132, + "learning_rate": 0.2289365409335297, + "loss": 0.3402, + "num_input_tokens_seen": 27050496, + "step": 12945 + }, + { + "epoch": 2.1125703564727956, + "grad_norm": 0.0002904744178522378, + "learning_rate": 0.2288864460496469, + "loss": 0.3091, + "num_input_tokens_seen": 27060784, + "step": 12950 + }, + { + "epoch": 2.1133860836936127, + "grad_norm": 0.00048573012463748455, + "learning_rate": 0.22883633900048272, + "loss": 0.2935, + "num_input_tokens_seen": 27071056, + "step": 12955 + }, + { + "epoch": 2.1142018109144303, + "grad_norm": 0.0003595180460251868, + "learning_rate": 0.2287862197937644, + "loss": 0.3328, + "num_input_tokens_seen": 27080096, + "step": 12960 + }, + { + "epoch": 2.1150175381352474, + "grad_norm": 0.0004215666849631816, + "learning_rate": 0.2287360884372209, + "loss": 0.3072, + "num_input_tokens_seen": 27091264, + "step": 12965 + }, + { + "epoch": 2.115833265356065, + "grad_norm": 0.0007851742557249963, + "learning_rate": 0.22868594493858307, + "loss": 0.352, + "num_input_tokens_seen": 27101040, + "step": 12970 + }, + { + "epoch": 2.116648992576882, + "grad_norm": 0.0017648490611463785, + "learning_rate": 0.2286357893055837, + "loss": 0.4024, + "num_input_tokens_seen": 27113424, + "step": 12975 + }, + { + "epoch": 2.1174647197976997, + "grad_norm": 0.0008731165435165167, + "learning_rate": 0.22858562154595746, + "loss": 0.357, + "num_input_tokens_seen": 27123488, + "step": 12980 + }, + { + "epoch": 2.118280447018517, + "grad_norm": 0.000451364234322682, + "learning_rate": 0.22853544166744078, + "loss": 0.3072, + "num_input_tokens_seen": 27134432, + "step": 12985 + }, + { + "epoch": 2.1190961742393344, + "grad_norm": 0.0003861688310280442, + "learning_rate": 0.22848524967777206, + "loss": 0.3685, + "num_input_tokens_seen": 27144976, + "step": 12990 + }, + { + "epoch": 2.1199119014601515, + "grad_norm": 0.0009379608673043549, + "learning_rate": 0.22843504558469152, + "loss": 0.3324, + "num_input_tokens_seen": 27155152, + "step": 12995 + }, + { + "epoch": 2.120727628680969, + "grad_norm": 0.0007192593184299767, + "learning_rate": 0.2283848293959413, + "loss": 0.3307, + "num_input_tokens_seen": 27166064, + "step": 13000 + }, + { + "epoch": 2.120727628680969, + "eval_loss": 0.3362112045288086, + "eval_runtime": 155.8362, + "eval_samples_per_second": 17.486, + "eval_steps_per_second": 8.746, + "num_input_tokens_seen": 27166064, + "step": 13000 + }, + { + "epoch": 2.121543355901786, + "grad_norm": 0.0018047267803922296, + "learning_rate": 0.22833460111926532, + "loss": 0.2978, + "num_input_tokens_seen": 27177456, + "step": 13005 + }, + { + "epoch": 2.122359083122604, + "grad_norm": 0.0010561547242105007, + "learning_rate": 0.22828436076240946, + "loss": 0.3452, + "num_input_tokens_seen": 27186464, + "step": 13010 + }, + { + "epoch": 2.1231748103434214, + "grad_norm": 0.0004922559019178152, + "learning_rate": 0.22823410833312135, + "loss": 0.3645, + "num_input_tokens_seen": 27196320, + "step": 13015 + }, + { + "epoch": 2.1239905375642385, + "grad_norm": 0.00026085955323651433, + "learning_rate": 0.2281838438391506, + "loss": 0.3267, + "num_input_tokens_seen": 27206992, + "step": 13020 + }, + { + "epoch": 2.124806264785056, + "grad_norm": 0.0006319555686786771, + "learning_rate": 0.22813356728824863, + "loss": 0.3323, + "num_input_tokens_seen": 27218304, + "step": 13025 + }, + { + "epoch": 2.125621992005873, + "grad_norm": 0.0005762625369243324, + "learning_rate": 0.2280832786881687, + "loss": 0.3088, + "num_input_tokens_seen": 27229040, + "step": 13030 + }, + { + "epoch": 2.126437719226691, + "grad_norm": 0.0005520034465007484, + "learning_rate": 0.22803297804666592, + "loss": 0.3509, + "num_input_tokens_seen": 27240064, + "step": 13035 + }, + { + "epoch": 2.127253446447508, + "grad_norm": 0.0004719290300272405, + "learning_rate": 0.22798266537149728, + "loss": 0.3432, + "num_input_tokens_seen": 27249600, + "step": 13040 + }, + { + "epoch": 2.1280691736683255, + "grad_norm": 0.0003501957398839295, + "learning_rate": 0.22793234067042167, + "loss": 0.3631, + "num_input_tokens_seen": 27259264, + "step": 13045 + }, + { + "epoch": 2.1288849008891426, + "grad_norm": 0.0007035158341750503, + "learning_rate": 0.22788200395119979, + "loss": 0.3157, + "num_input_tokens_seen": 27270784, + "step": 13050 + }, + { + "epoch": 2.12970062810996, + "grad_norm": 0.000359459372702986, + "learning_rate": 0.2278316552215942, + "loss": 0.3579, + "num_input_tokens_seen": 27281168, + "step": 13055 + }, + { + "epoch": 2.1305163553307773, + "grad_norm": 0.00033647846430540085, + "learning_rate": 0.22778129448936918, + "loss": 0.3282, + "num_input_tokens_seen": 27290976, + "step": 13060 + }, + { + "epoch": 2.131332082551595, + "grad_norm": 0.0003039133735001087, + "learning_rate": 0.22773092176229118, + "loss": 0.3557, + "num_input_tokens_seen": 27301088, + "step": 13065 + }, + { + "epoch": 2.132147809772412, + "grad_norm": 0.00044378373422659934, + "learning_rate": 0.22768053704812816, + "loss": 0.3255, + "num_input_tokens_seen": 27311856, + "step": 13070 + }, + { + "epoch": 2.1329635369932296, + "grad_norm": 0.00036573768011294305, + "learning_rate": 0.22763014035465018, + "loss": 0.3201, + "num_input_tokens_seen": 27323280, + "step": 13075 + }, + { + "epoch": 2.1337792642140467, + "grad_norm": 0.00025363790336996317, + "learning_rate": 0.22757973168962892, + "loss": 0.3063, + "num_input_tokens_seen": 27332192, + "step": 13080 + }, + { + "epoch": 2.1345949914348643, + "grad_norm": 0.000738753704354167, + "learning_rate": 0.22752931106083818, + "loss": 0.3142, + "num_input_tokens_seen": 27341712, + "step": 13085 + }, + { + "epoch": 2.1354107186556814, + "grad_norm": 0.0004626608861144632, + "learning_rate": 0.22747887847605341, + "loss": 0.2905, + "num_input_tokens_seen": 27352880, + "step": 13090 + }, + { + "epoch": 2.136226445876499, + "grad_norm": 0.0006629357812926173, + "learning_rate": 0.22742843394305184, + "loss": 0.2808, + "num_input_tokens_seen": 27362432, + "step": 13095 + }, + { + "epoch": 2.137042173097316, + "grad_norm": 0.0008255615248344839, + "learning_rate": 0.22737797746961272, + "loss": 0.3169, + "num_input_tokens_seen": 27373728, + "step": 13100 + }, + { + "epoch": 2.1378579003181337, + "grad_norm": 0.0007914722082205117, + "learning_rate": 0.22732750906351712, + "loss": 0.3158, + "num_input_tokens_seen": 27383936, + "step": 13105 + }, + { + "epoch": 2.138673627538951, + "grad_norm": 0.0008127117180265486, + "learning_rate": 0.22727702873254785, + "loss": 0.3199, + "num_input_tokens_seen": 27396240, + "step": 13110 + }, + { + "epoch": 2.1394893547597684, + "grad_norm": 0.00099173525813967, + "learning_rate": 0.22722653648448968, + "loss": 0.2784, + "num_input_tokens_seen": 27406352, + "step": 13115 + }, + { + "epoch": 2.1403050819805856, + "grad_norm": 0.0004840277833864093, + "learning_rate": 0.22717603232712902, + "loss": 0.266, + "num_input_tokens_seen": 27415696, + "step": 13120 + }, + { + "epoch": 2.141120809201403, + "grad_norm": 0.00036814124905504286, + "learning_rate": 0.22712551626825436, + "loss": 0.3237, + "num_input_tokens_seen": 27425504, + "step": 13125 + }, + { + "epoch": 2.1419365364222203, + "grad_norm": 0.0003183314111083746, + "learning_rate": 0.2270749883156559, + "loss": 0.3518, + "num_input_tokens_seen": 27437184, + "step": 13130 + }, + { + "epoch": 2.142752263643038, + "grad_norm": 0.0015132547123357654, + "learning_rate": 0.22702444847712563, + "loss": 0.3244, + "num_input_tokens_seen": 27446608, + "step": 13135 + }, + { + "epoch": 2.143567990863855, + "grad_norm": 0.0005272316047921777, + "learning_rate": 0.22697389676045743, + "loss": 0.2781, + "num_input_tokens_seen": 27456128, + "step": 13140 + }, + { + "epoch": 2.1443837180846725, + "grad_norm": 0.0015206256648525596, + "learning_rate": 0.22692333317344704, + "loss": 0.3327, + "num_input_tokens_seen": 27466640, + "step": 13145 + }, + { + "epoch": 2.1451994453054897, + "grad_norm": 0.0008790856809355319, + "learning_rate": 0.22687275772389198, + "loss": 0.2723, + "num_input_tokens_seen": 27476384, + "step": 13150 + }, + { + "epoch": 2.1460151725263072, + "grad_norm": 0.001119293156079948, + "learning_rate": 0.22682217041959168, + "loss": 0.3484, + "num_input_tokens_seen": 27487184, + "step": 13155 + }, + { + "epoch": 2.1468308997471244, + "grad_norm": 0.00043277614167891443, + "learning_rate": 0.2267715712683473, + "loss": 0.3868, + "num_input_tokens_seen": 27497152, + "step": 13160 + }, + { + "epoch": 2.147646626967942, + "grad_norm": 0.0008389597642235458, + "learning_rate": 0.22672096027796182, + "loss": 0.3434, + "num_input_tokens_seen": 27507504, + "step": 13165 + }, + { + "epoch": 2.148462354188759, + "grad_norm": 0.000822818314190954, + "learning_rate": 0.22667033745624016, + "loss": 0.2698, + "num_input_tokens_seen": 27517008, + "step": 13170 + }, + { + "epoch": 2.1492780814095767, + "grad_norm": 0.0010244951117783785, + "learning_rate": 0.22661970281098895, + "loss": 0.321, + "num_input_tokens_seen": 27527184, + "step": 13175 + }, + { + "epoch": 2.150093808630394, + "grad_norm": 0.0013721812283620238, + "learning_rate": 0.22656905635001667, + "loss": 0.3342, + "num_input_tokens_seen": 27537632, + "step": 13180 + }, + { + "epoch": 2.1509095358512114, + "grad_norm": 0.0010262508876621723, + "learning_rate": 0.2265183980811337, + "loss": 0.3893, + "num_input_tokens_seen": 27550032, + "step": 13185 + }, + { + "epoch": 2.151725263072029, + "grad_norm": 0.0005693990387953818, + "learning_rate": 0.22646772801215218, + "loss": 0.3578, + "num_input_tokens_seen": 27560336, + "step": 13190 + }, + { + "epoch": 2.152540990292846, + "grad_norm": 0.0003469625080469996, + "learning_rate": 0.22641704615088598, + "loss": 0.3173, + "num_input_tokens_seen": 27571200, + "step": 13195 + }, + { + "epoch": 2.153356717513663, + "grad_norm": 0.0007064339006319642, + "learning_rate": 0.22636635250515103, + "loss": 0.3568, + "num_input_tokens_seen": 27581264, + "step": 13200 + }, + { + "epoch": 2.153356717513663, + "eval_loss": 0.32597729563713074, + "eval_runtime": 156.1605, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 27581264, + "step": 13200 + }, + { + "epoch": 2.1541724447344808, + "grad_norm": 0.0034879751037806273, + "learning_rate": 0.2263156470827648, + "loss": 0.38, + "num_input_tokens_seen": 27593168, + "step": 13205 + }, + { + "epoch": 2.1549881719552983, + "grad_norm": 0.001515903975814581, + "learning_rate": 0.22626492989154678, + "loss": 0.3392, + "num_input_tokens_seen": 27602352, + "step": 13210 + }, + { + "epoch": 2.1558038991761155, + "grad_norm": 0.0003441296285018325, + "learning_rate": 0.22621420093931813, + "loss": 0.3202, + "num_input_tokens_seen": 27612032, + "step": 13215 + }, + { + "epoch": 2.156619626396933, + "grad_norm": 0.0014806970721110702, + "learning_rate": 0.22616346023390194, + "loss": 0.371, + "num_input_tokens_seen": 27621792, + "step": 13220 + }, + { + "epoch": 2.15743535361775, + "grad_norm": 0.0003689152654260397, + "learning_rate": 0.22611270778312306, + "loss": 0.3331, + "num_input_tokens_seen": 27633856, + "step": 13225 + }, + { + "epoch": 2.1582510808385678, + "grad_norm": 0.000977076473645866, + "learning_rate": 0.2260619435948081, + "loss": 0.3587, + "num_input_tokens_seen": 27644656, + "step": 13230 + }, + { + "epoch": 2.159066808059385, + "grad_norm": 0.0008063928107731044, + "learning_rate": 0.22601116767678567, + "loss": 0.2769, + "num_input_tokens_seen": 27654736, + "step": 13235 + }, + { + "epoch": 2.1598825352802025, + "grad_norm": 0.00039670729893259704, + "learning_rate": 0.2259603800368859, + "loss": 0.3582, + "num_input_tokens_seen": 27664336, + "step": 13240 + }, + { + "epoch": 2.1606982625010196, + "grad_norm": 0.00036823019036091864, + "learning_rate": 0.22590958068294098, + "loss": 0.326, + "num_input_tokens_seen": 27675728, + "step": 13245 + }, + { + "epoch": 2.161513989721837, + "grad_norm": 0.0006532368715852499, + "learning_rate": 0.22585876962278478, + "loss": 0.2798, + "num_input_tokens_seen": 27687088, + "step": 13250 + }, + { + "epoch": 2.1623297169426543, + "grad_norm": 0.0007041233475320041, + "learning_rate": 0.22580794686425298, + "loss": 0.3305, + "num_input_tokens_seen": 27697088, + "step": 13255 + }, + { + "epoch": 2.163145444163472, + "grad_norm": 0.0002524075098335743, + "learning_rate": 0.22575711241518312, + "loss": 0.3116, + "num_input_tokens_seen": 27707600, + "step": 13260 + }, + { + "epoch": 2.163961171384289, + "grad_norm": 0.0004081691149622202, + "learning_rate": 0.22570626628341453, + "loss": 0.3467, + "num_input_tokens_seen": 27717360, + "step": 13265 + }, + { + "epoch": 2.1647768986051066, + "grad_norm": 0.000995139591395855, + "learning_rate": 0.22565540847678828, + "loss": 0.3051, + "num_input_tokens_seen": 27727952, + "step": 13270 + }, + { + "epoch": 2.1655926258259237, + "grad_norm": 0.0005294294096529484, + "learning_rate": 0.2256045390031473, + "loss": 0.3446, + "num_input_tokens_seen": 27737792, + "step": 13275 + }, + { + "epoch": 2.1664083530467413, + "grad_norm": 0.0007073816377669573, + "learning_rate": 0.22555365787033627, + "loss": 0.3271, + "num_input_tokens_seen": 27747280, + "step": 13280 + }, + { + "epoch": 2.1672240802675584, + "grad_norm": 0.00043037996510975063, + "learning_rate": 0.22550276508620173, + "loss": 0.3132, + "num_input_tokens_seen": 27757504, + "step": 13285 + }, + { + "epoch": 2.168039807488376, + "grad_norm": 0.0004958932986482978, + "learning_rate": 0.22545186065859202, + "loss": 0.3001, + "num_input_tokens_seen": 27767552, + "step": 13290 + }, + { + "epoch": 2.168855534709193, + "grad_norm": 0.0008453234913758934, + "learning_rate": 0.2254009445953572, + "loss": 0.2472, + "num_input_tokens_seen": 27778560, + "step": 13295 + }, + { + "epoch": 2.1696712619300107, + "grad_norm": 0.000563829205930233, + "learning_rate": 0.22535001690434917, + "loss": 0.3013, + "num_input_tokens_seen": 27789344, + "step": 13300 + }, + { + "epoch": 2.170486989150828, + "grad_norm": 0.0004825741343665868, + "learning_rate": 0.22529907759342163, + "loss": 0.2544, + "num_input_tokens_seen": 27799824, + "step": 13305 + }, + { + "epoch": 2.1713027163716454, + "grad_norm": 0.06431534886360168, + "learning_rate": 0.22524812667043007, + "loss": 0.4421, + "num_input_tokens_seen": 27810992, + "step": 13310 + }, + { + "epoch": 2.1721184435924625, + "grad_norm": 0.0005040719988755882, + "learning_rate": 0.22519716414323177, + "loss": 0.3903, + "num_input_tokens_seen": 27821392, + "step": 13315 + }, + { + "epoch": 2.17293417081328, + "grad_norm": 0.00037934916326776147, + "learning_rate": 0.22514619001968567, + "loss": 0.2848, + "num_input_tokens_seen": 27831264, + "step": 13320 + }, + { + "epoch": 2.1737498980340972, + "grad_norm": 0.00048022338887676597, + "learning_rate": 0.2250952043076528, + "loss": 0.2986, + "num_input_tokens_seen": 27839360, + "step": 13325 + }, + { + "epoch": 2.174565625254915, + "grad_norm": 0.0006351610063575208, + "learning_rate": 0.2250442070149957, + "loss": 0.3834, + "num_input_tokens_seen": 27848448, + "step": 13330 + }, + { + "epoch": 2.175381352475732, + "grad_norm": 0.0005179623840376735, + "learning_rate": 0.22499319814957885, + "loss": 0.294, + "num_input_tokens_seen": 27859440, + "step": 13335 + }, + { + "epoch": 2.1761970796965495, + "grad_norm": 0.0010772485984489322, + "learning_rate": 0.2249421777192684, + "loss": 0.2933, + "num_input_tokens_seen": 27869952, + "step": 13340 + }, + { + "epoch": 2.1770128069173666, + "grad_norm": 0.0013897609896957874, + "learning_rate": 0.22489114573193236, + "loss": 0.3938, + "num_input_tokens_seen": 27879376, + "step": 13345 + }, + { + "epoch": 2.1778285341381842, + "grad_norm": 0.0011102643329650164, + "learning_rate": 0.2248401021954405, + "loss": 0.3501, + "num_input_tokens_seen": 27888912, + "step": 13350 + }, + { + "epoch": 2.1786442613590014, + "grad_norm": 0.0010266087483614683, + "learning_rate": 0.22478904711766443, + "loss": 0.2941, + "num_input_tokens_seen": 27898512, + "step": 13355 + }, + { + "epoch": 2.179459988579819, + "grad_norm": 0.0017897728830575943, + "learning_rate": 0.22473798050647734, + "loss": 0.3464, + "num_input_tokens_seen": 27908352, + "step": 13360 + }, + { + "epoch": 2.180275715800636, + "grad_norm": 0.0006857148837298155, + "learning_rate": 0.22468690236975453, + "loss": 0.3209, + "num_input_tokens_seen": 27917456, + "step": 13365 + }, + { + "epoch": 2.1810914430214536, + "grad_norm": 0.0004995284252800047, + "learning_rate": 0.22463581271537272, + "loss": 0.3163, + "num_input_tokens_seen": 27929024, + "step": 13370 + }, + { + "epoch": 2.1819071702422708, + "grad_norm": 0.0016310919309034944, + "learning_rate": 0.22458471155121076, + "loss": 0.3662, + "num_input_tokens_seen": 27940192, + "step": 13375 + }, + { + "epoch": 2.1827228974630883, + "grad_norm": 0.0005848165019415319, + "learning_rate": 0.2245335988851489, + "loss": 0.3175, + "num_input_tokens_seen": 27949648, + "step": 13380 + }, + { + "epoch": 2.183538624683906, + "grad_norm": 0.0009667627746239305, + "learning_rate": 0.2244824747250695, + "loss": 0.321, + "num_input_tokens_seen": 27959968, + "step": 13385 + }, + { + "epoch": 2.184354351904723, + "grad_norm": 0.0005568080232478678, + "learning_rate": 0.22443133907885646, + "loss": 0.3293, + "num_input_tokens_seen": 27968144, + "step": 13390 + }, + { + "epoch": 2.1851700791255406, + "grad_norm": 0.0018446177709847689, + "learning_rate": 0.22438019195439557, + "loss": 0.352, + "num_input_tokens_seen": 27978992, + "step": 13395 + }, + { + "epoch": 2.1859858063463578, + "grad_norm": 0.0012329543242231011, + "learning_rate": 0.22432903335957435, + "loss": 0.3308, + "num_input_tokens_seen": 27988880, + "step": 13400 + }, + { + "epoch": 2.1859858063463578, + "eval_loss": 0.32471781969070435, + "eval_runtime": 156.3288, + "eval_samples_per_second": 17.431, + "eval_steps_per_second": 8.719, + "num_input_tokens_seen": 27988880, + "step": 13400 + }, + { + "epoch": 2.1868015335671753, + "grad_norm": 0.0005952356150373816, + "learning_rate": 0.22427786330228214, + "loss": 0.3375, + "num_input_tokens_seen": 27999520, + "step": 13405 + }, + { + "epoch": 2.1876172607879925, + "grad_norm": 0.000995609792880714, + "learning_rate": 0.22422668179040997, + "loss": 0.3729, + "num_input_tokens_seen": 28010640, + "step": 13410 + }, + { + "epoch": 2.18843298800881, + "grad_norm": 0.0008844530675560236, + "learning_rate": 0.2241754888318507, + "loss": 0.3392, + "num_input_tokens_seen": 28021504, + "step": 13415 + }, + { + "epoch": 2.189248715229627, + "grad_norm": 0.0010066063841804862, + "learning_rate": 0.22412428443449886, + "loss": 0.315, + "num_input_tokens_seen": 28030432, + "step": 13420 + }, + { + "epoch": 2.1900644424504447, + "grad_norm": 0.0016936458414420485, + "learning_rate": 0.22407306860625087, + "loss": 0.3297, + "num_input_tokens_seen": 28040752, + "step": 13425 + }, + { + "epoch": 2.190880169671262, + "grad_norm": 0.0005020322278141975, + "learning_rate": 0.22402184135500483, + "loss": 0.3393, + "num_input_tokens_seen": 28050464, + "step": 13430 + }, + { + "epoch": 2.1916958968920794, + "grad_norm": 0.0012148631503805518, + "learning_rate": 0.22397060268866067, + "loss": 0.3153, + "num_input_tokens_seen": 28060048, + "step": 13435 + }, + { + "epoch": 2.1925116241128966, + "grad_norm": 0.0006234528846107423, + "learning_rate": 0.22391935261511994, + "loss": 0.3169, + "num_input_tokens_seen": 28069696, + "step": 13440 + }, + { + "epoch": 2.193327351333714, + "grad_norm": 0.000508093333337456, + "learning_rate": 0.22386809114228615, + "loss": 0.3293, + "num_input_tokens_seen": 28080464, + "step": 13445 + }, + { + "epoch": 2.1941430785545313, + "grad_norm": 0.0008971308707259595, + "learning_rate": 0.22381681827806446, + "loss": 0.3344, + "num_input_tokens_seen": 28091296, + "step": 13450 + }, + { + "epoch": 2.194958805775349, + "grad_norm": 0.0006281742244027555, + "learning_rate": 0.22376553403036173, + "loss": 0.3496, + "num_input_tokens_seen": 28103232, + "step": 13455 + }, + { + "epoch": 2.195774532996166, + "grad_norm": 0.0006660691578872502, + "learning_rate": 0.22371423840708662, + "loss": 0.3281, + "num_input_tokens_seen": 28114000, + "step": 13460 + }, + { + "epoch": 2.1965902602169836, + "grad_norm": 0.0009596376912668347, + "learning_rate": 0.22366293141614962, + "loss": 0.3555, + "num_input_tokens_seen": 28123856, + "step": 13465 + }, + { + "epoch": 2.1974059874378007, + "grad_norm": 0.0006085739005357027, + "learning_rate": 0.22361161306546287, + "loss": 0.3254, + "num_input_tokens_seen": 28135040, + "step": 13470 + }, + { + "epoch": 2.1982217146586183, + "grad_norm": 0.00041793062700890005, + "learning_rate": 0.22356028336294037, + "loss": 0.2813, + "num_input_tokens_seen": 28143808, + "step": 13475 + }, + { + "epoch": 2.1990374418794354, + "grad_norm": 0.0011676165740936995, + "learning_rate": 0.2235089423164977, + "loss": 0.3013, + "num_input_tokens_seen": 28154528, + "step": 13480 + }, + { + "epoch": 2.199853169100253, + "grad_norm": 0.0005191658274270594, + "learning_rate": 0.22345758993405243, + "loss": 0.3456, + "num_input_tokens_seen": 28164800, + "step": 13485 + }, + { + "epoch": 2.20066889632107, + "grad_norm": 0.0006593929138034582, + "learning_rate": 0.2234062262235236, + "loss": 0.3574, + "num_input_tokens_seen": 28174208, + "step": 13490 + }, + { + "epoch": 2.2014846235418877, + "grad_norm": 0.0005333241424523294, + "learning_rate": 0.22335485119283222, + "loss": 0.3296, + "num_input_tokens_seen": 28183648, + "step": 13495 + }, + { + "epoch": 2.202300350762705, + "grad_norm": 0.0004006357048638165, + "learning_rate": 0.22330346484990093, + "loss": 0.3561, + "num_input_tokens_seen": 28194656, + "step": 13500 + }, + { + "epoch": 2.2031160779835224, + "grad_norm": 0.0005526028689928353, + "learning_rate": 0.22325206720265425, + "loss": 0.3586, + "num_input_tokens_seen": 28204624, + "step": 13505 + }, + { + "epoch": 2.2039318052043395, + "grad_norm": 0.0006171065615490079, + "learning_rate": 0.2232006582590182, + "loss": 0.3389, + "num_input_tokens_seen": 28214304, + "step": 13510 + }, + { + "epoch": 2.204747532425157, + "grad_norm": 0.0013758287532255054, + "learning_rate": 0.22314923802692077, + "loss": 0.3627, + "num_input_tokens_seen": 28225728, + "step": 13515 + }, + { + "epoch": 2.205563259645974, + "grad_norm": 0.0011568976333364844, + "learning_rate": 0.22309780651429156, + "loss": 0.3077, + "num_input_tokens_seen": 28235520, + "step": 13520 + }, + { + "epoch": 2.206378986866792, + "grad_norm": 0.0004951293230988085, + "learning_rate": 0.22304636372906203, + "loss": 0.3343, + "num_input_tokens_seen": 28245232, + "step": 13525 + }, + { + "epoch": 2.207194714087609, + "grad_norm": 0.0005270736292004585, + "learning_rate": 0.22299490967916522, + "loss": 0.3005, + "num_input_tokens_seen": 28255552, + "step": 13530 + }, + { + "epoch": 2.2080104413084265, + "grad_norm": 0.000952571805100888, + "learning_rate": 0.22294344437253602, + "loss": 0.2902, + "num_input_tokens_seen": 28265104, + "step": 13535 + }, + { + "epoch": 2.2088261685292436, + "grad_norm": 0.0011361701181158423, + "learning_rate": 0.22289196781711101, + "loss": 0.3149, + "num_input_tokens_seen": 28275552, + "step": 13540 + }, + { + "epoch": 2.209641895750061, + "grad_norm": 0.000987071543931961, + "learning_rate": 0.2228404800208286, + "loss": 0.358, + "num_input_tokens_seen": 28285296, + "step": 13545 + }, + { + "epoch": 2.2104576229708783, + "grad_norm": 0.0013625657884404063, + "learning_rate": 0.22278898099162875, + "loss": 0.3613, + "num_input_tokens_seen": 28295872, + "step": 13550 + }, + { + "epoch": 2.211273350191696, + "grad_norm": 0.0012053397949784994, + "learning_rate": 0.22273747073745337, + "loss": 0.3539, + "num_input_tokens_seen": 28304816, + "step": 13555 + }, + { + "epoch": 2.2120890774125135, + "grad_norm": 0.0005624975892715156, + "learning_rate": 0.22268594926624588, + "loss": 0.3472, + "num_input_tokens_seen": 28315664, + "step": 13560 + }, + { + "epoch": 2.2129048046333306, + "grad_norm": 0.000519445922691375, + "learning_rate": 0.22263441658595162, + "loss": 0.3438, + "num_input_tokens_seen": 28326160, + "step": 13565 + }, + { + "epoch": 2.2137205318541477, + "grad_norm": 0.0004148939624428749, + "learning_rate": 0.2225828727045175, + "loss": 0.3512, + "num_input_tokens_seen": 28335872, + "step": 13570 + }, + { + "epoch": 2.2145362590749653, + "grad_norm": 0.0007670239428989589, + "learning_rate": 0.22253131762989228, + "loss": 0.3075, + "num_input_tokens_seen": 28346192, + "step": 13575 + }, + { + "epoch": 2.215351986295783, + "grad_norm": 0.000916296208743006, + "learning_rate": 0.2224797513700264, + "loss": 0.3432, + "num_input_tokens_seen": 28355680, + "step": 13580 + }, + { + "epoch": 2.2161677135166, + "grad_norm": 0.000710505002643913, + "learning_rate": 0.22242817393287204, + "loss": 0.3418, + "num_input_tokens_seen": 28366080, + "step": 13585 + }, + { + "epoch": 2.2169834407374176, + "grad_norm": 0.00044835032895207405, + "learning_rate": 0.22237658532638305, + "loss": 0.3202, + "num_input_tokens_seen": 28376176, + "step": 13590 + }, + { + "epoch": 2.2177991679582347, + "grad_norm": 0.0003915154084097594, + "learning_rate": 0.22232498555851513, + "loss": 0.2853, + "num_input_tokens_seen": 28386832, + "step": 13595 + }, + { + "epoch": 2.2186148951790523, + "grad_norm": 0.0007716647232882679, + "learning_rate": 0.22227337463722546, + "loss": 0.3383, + "num_input_tokens_seen": 28397472, + "step": 13600 + }, + { + "epoch": 2.2186148951790523, + "eval_loss": 0.32431650161743164, + "eval_runtime": 156.2083, + "eval_samples_per_second": 17.445, + "eval_steps_per_second": 8.726, + "num_input_tokens_seen": 28397472, + "step": 13600 + }, + { + "epoch": 2.2194306223998694, + "grad_norm": 0.0012347536394372582, + "learning_rate": 0.2222217525704732, + "loss": 0.3201, + "num_input_tokens_seen": 28406608, + "step": 13605 + }, + { + "epoch": 2.220246349620687, + "grad_norm": 0.00044015058665536344, + "learning_rate": 0.22217011936621908, + "loss": 0.3138, + "num_input_tokens_seen": 28416480, + "step": 13610 + }, + { + "epoch": 2.221062076841504, + "grad_norm": 0.0009099284070543945, + "learning_rate": 0.22211847503242566, + "loss": 0.3021, + "num_input_tokens_seen": 28426144, + "step": 13615 + }, + { + "epoch": 2.2218778040623217, + "grad_norm": 0.0005529483314603567, + "learning_rate": 0.22206681957705704, + "loss": 0.3752, + "num_input_tokens_seen": 28438080, + "step": 13620 + }, + { + "epoch": 2.222693531283139, + "grad_norm": 0.0007194819627329707, + "learning_rate": 0.2220151530080792, + "loss": 0.3797, + "num_input_tokens_seen": 28447968, + "step": 13625 + }, + { + "epoch": 2.2235092585039564, + "grad_norm": 0.0008988502086140215, + "learning_rate": 0.2219634753334598, + "loss": 0.2991, + "num_input_tokens_seen": 28458752, + "step": 13630 + }, + { + "epoch": 2.2243249857247736, + "grad_norm": 0.0015706770354881883, + "learning_rate": 0.22191178656116817, + "loss": 0.3502, + "num_input_tokens_seen": 28468912, + "step": 13635 + }, + { + "epoch": 2.225140712945591, + "grad_norm": 0.0009027968626469374, + "learning_rate": 0.2218600866991753, + "loss": 0.3222, + "num_input_tokens_seen": 28478864, + "step": 13640 + }, + { + "epoch": 2.2259564401664083, + "grad_norm": 0.0006058734725229442, + "learning_rate": 0.221808375755454, + "loss": 0.3357, + "num_input_tokens_seen": 28489504, + "step": 13645 + }, + { + "epoch": 2.226772167387226, + "grad_norm": 0.0003019085561390966, + "learning_rate": 0.22175665373797881, + "loss": 0.329, + "num_input_tokens_seen": 28500112, + "step": 13650 + }, + { + "epoch": 2.227587894608043, + "grad_norm": 0.0007105354452505708, + "learning_rate": 0.22170492065472583, + "loss": 0.3635, + "num_input_tokens_seen": 28511216, + "step": 13655 + }, + { + "epoch": 2.2284036218288605, + "grad_norm": 0.0004026175884064287, + "learning_rate": 0.221653176513673, + "loss": 0.2952, + "num_input_tokens_seen": 28521776, + "step": 13660 + }, + { + "epoch": 2.2292193490496777, + "grad_norm": 0.0003323530836496502, + "learning_rate": 0.2216014213227999, + "loss": 0.305, + "num_input_tokens_seen": 28531920, + "step": 13665 + }, + { + "epoch": 2.2300350762704952, + "grad_norm": 0.001610307488590479, + "learning_rate": 0.22154965509008784, + "loss": 0.3815, + "num_input_tokens_seen": 28543856, + "step": 13670 + }, + { + "epoch": 2.2308508034913124, + "grad_norm": 0.0006119011668488383, + "learning_rate": 0.2214978778235198, + "loss": 0.336, + "num_input_tokens_seen": 28554288, + "step": 13675 + }, + { + "epoch": 2.23166653071213, + "grad_norm": 0.0009249509312212467, + "learning_rate": 0.2214460895310805, + "loss": 0.3255, + "num_input_tokens_seen": 28564352, + "step": 13680 + }, + { + "epoch": 2.232482257932947, + "grad_norm": 0.0004388616362120956, + "learning_rate": 0.22139429022075635, + "loss": 0.2913, + "num_input_tokens_seen": 28575712, + "step": 13685 + }, + { + "epoch": 2.2332979851537647, + "grad_norm": 0.0004290194483473897, + "learning_rate": 0.22134247990053546, + "loss": 0.3175, + "num_input_tokens_seen": 28587104, + "step": 13690 + }, + { + "epoch": 2.234113712374582, + "grad_norm": 0.0004554199113044888, + "learning_rate": 0.2212906585784076, + "loss": 0.3657, + "num_input_tokens_seen": 28597840, + "step": 13695 + }, + { + "epoch": 2.2349294395953994, + "grad_norm": 0.0004412807757034898, + "learning_rate": 0.22123882626236432, + "loss": 0.3364, + "num_input_tokens_seen": 28608656, + "step": 13700 + }, + { + "epoch": 2.2357451668162165, + "grad_norm": 0.0005740158376283944, + "learning_rate": 0.2211869829603988, + "loss": 0.327, + "num_input_tokens_seen": 28618576, + "step": 13705 + }, + { + "epoch": 2.236560894037034, + "grad_norm": 0.0006026190239936113, + "learning_rate": 0.22113512868050592, + "loss": 0.3457, + "num_input_tokens_seen": 28629008, + "step": 13710 + }, + { + "epoch": 2.237376621257851, + "grad_norm": 0.001100944122299552, + "learning_rate": 0.2210832634306822, + "loss": 0.3503, + "num_input_tokens_seen": 28639072, + "step": 13715 + }, + { + "epoch": 2.2381923484786688, + "grad_norm": 0.0010547968558967113, + "learning_rate": 0.22103138721892598, + "loss": 0.3146, + "num_input_tokens_seen": 28648304, + "step": 13720 + }, + { + "epoch": 2.239008075699486, + "grad_norm": 0.00036561532760970294, + "learning_rate": 0.22097950005323724, + "loss": 0.2843, + "num_input_tokens_seen": 28659168, + "step": 13725 + }, + { + "epoch": 2.2398238029203035, + "grad_norm": 0.0004958714125677943, + "learning_rate": 0.22092760194161762, + "loss": 0.3271, + "num_input_tokens_seen": 28668560, + "step": 13730 + }, + { + "epoch": 2.2406395301411206, + "grad_norm": 0.00039582705358043313, + "learning_rate": 0.2208756928920704, + "loss": 0.3107, + "num_input_tokens_seen": 28679888, + "step": 13735 + }, + { + "epoch": 2.241455257361938, + "grad_norm": 0.00038048060378059745, + "learning_rate": 0.22082377291260072, + "loss": 0.2696, + "num_input_tokens_seen": 28688992, + "step": 13740 + }, + { + "epoch": 2.2422709845827553, + "grad_norm": 0.000514333660248667, + "learning_rate": 0.2207718420112152, + "loss": 0.3084, + "num_input_tokens_seen": 28700704, + "step": 13745 + }, + { + "epoch": 2.243086711803573, + "grad_norm": 0.0003772930649574846, + "learning_rate": 0.22071990019592228, + "loss": 0.3247, + "num_input_tokens_seen": 28711392, + "step": 13750 + }, + { + "epoch": 2.2439024390243905, + "grad_norm": 0.00029834784800186753, + "learning_rate": 0.22066794747473198, + "loss": 0.3906, + "num_input_tokens_seen": 28721920, + "step": 13755 + }, + { + "epoch": 2.2447181662452076, + "grad_norm": 0.0006337364320643246, + "learning_rate": 0.2206159838556562, + "loss": 0.2525, + "num_input_tokens_seen": 28733392, + "step": 13760 + }, + { + "epoch": 2.245533893466025, + "grad_norm": 0.001480858656577766, + "learning_rate": 0.2205640093467082, + "loss": 0.4321, + "num_input_tokens_seen": 28743088, + "step": 13765 + }, + { + "epoch": 2.2463496206868423, + "grad_norm": 0.0006885475013405085, + "learning_rate": 0.22051202395590322, + "loss": 0.3456, + "num_input_tokens_seen": 28752656, + "step": 13770 + }, + { + "epoch": 2.24716534790766, + "grad_norm": 0.00044739528675563633, + "learning_rate": 0.22046002769125808, + "loss": 0.3149, + "num_input_tokens_seen": 28763440, + "step": 13775 + }, + { + "epoch": 2.247981075128477, + "grad_norm": 0.0006286867428570986, + "learning_rate": 0.2204080205607912, + "loss": 0.3533, + "num_input_tokens_seen": 28774128, + "step": 13780 + }, + { + "epoch": 2.2487968023492946, + "grad_norm": 0.0006241845549084246, + "learning_rate": 0.22035600257252272, + "loss": 0.3037, + "num_input_tokens_seen": 28782768, + "step": 13785 + }, + { + "epoch": 2.2496125295701117, + "grad_norm": 0.0007179924286901951, + "learning_rate": 0.2203039737344745, + "loss": 0.3013, + "num_input_tokens_seen": 28792224, + "step": 13790 + }, + { + "epoch": 2.2504282567909293, + "grad_norm": 0.00036707159597426653, + "learning_rate": 0.22025193405467003, + "loss": 0.3406, + "num_input_tokens_seen": 28801664, + "step": 13795 + }, + { + "epoch": 2.2512439840117464, + "grad_norm": 0.0007381672621704638, + "learning_rate": 0.2201998835411345, + "loss": 0.3414, + "num_input_tokens_seen": 28812800, + "step": 13800 + }, + { + "epoch": 2.2512439840117464, + "eval_loss": 0.32220301032066345, + "eval_runtime": 156.107, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 28812800, + "step": 13800 + }, + { + "epoch": 2.252059711232564, + "grad_norm": 0.000752683961763978, + "learning_rate": 0.22014782220189474, + "loss": 0.3362, + "num_input_tokens_seen": 28822224, + "step": 13805 + }, + { + "epoch": 2.252875438453381, + "grad_norm": 0.0012264405377209187, + "learning_rate": 0.2200957500449793, + "loss": 0.3008, + "num_input_tokens_seen": 28831536, + "step": 13810 + }, + { + "epoch": 2.2536911656741987, + "grad_norm": 0.0009058888535946608, + "learning_rate": 0.22004366707841827, + "loss": 0.3692, + "num_input_tokens_seen": 28841728, + "step": 13815 + }, + { + "epoch": 2.254506892895016, + "grad_norm": 0.0009610836859792471, + "learning_rate": 0.21999157331024358, + "loss": 0.3014, + "num_input_tokens_seen": 28852208, + "step": 13820 + }, + { + "epoch": 2.2553226201158334, + "grad_norm": 0.0006417190306819975, + "learning_rate": 0.21993946874848871, + "loss": 0.3618, + "num_input_tokens_seen": 28862912, + "step": 13825 + }, + { + "epoch": 2.2561383473366505, + "grad_norm": 0.000666026258841157, + "learning_rate": 0.2198873534011888, + "loss": 0.2965, + "num_input_tokens_seen": 28872560, + "step": 13830 + }, + { + "epoch": 2.256954074557468, + "grad_norm": 0.0005622315802611411, + "learning_rate": 0.2198352272763808, + "loss": 0.2992, + "num_input_tokens_seen": 28884400, + "step": 13835 + }, + { + "epoch": 2.2577698017782852, + "grad_norm": 0.00040614791214466095, + "learning_rate": 0.2197830903821031, + "loss": 0.3168, + "num_input_tokens_seen": 28892976, + "step": 13840 + }, + { + "epoch": 2.258585528999103, + "grad_norm": 0.0005646928329952061, + "learning_rate": 0.21973094272639598, + "loss": 0.304, + "num_input_tokens_seen": 28904016, + "step": 13845 + }, + { + "epoch": 2.25940125621992, + "grad_norm": 0.0012025227770209312, + "learning_rate": 0.21967878431730117, + "loss": 0.2904, + "num_input_tokens_seen": 28914288, + "step": 13850 + }, + { + "epoch": 2.2602169834407375, + "grad_norm": 0.0009050164953805506, + "learning_rate": 0.21962661516286217, + "loss": 0.3986, + "num_input_tokens_seen": 28923968, + "step": 13855 + }, + { + "epoch": 2.2610327106615546, + "grad_norm": 0.0008470775792375207, + "learning_rate": 0.21957443527112414, + "loss": 0.3545, + "num_input_tokens_seen": 28936352, + "step": 13860 + }, + { + "epoch": 2.261848437882372, + "grad_norm": 0.0004247610631864518, + "learning_rate": 0.21952224465013384, + "loss": 0.3321, + "num_input_tokens_seen": 28945232, + "step": 13865 + }, + { + "epoch": 2.2626641651031894, + "grad_norm": 0.0008301076013594866, + "learning_rate": 0.21947004330793976, + "loss": 0.3452, + "num_input_tokens_seen": 28954992, + "step": 13870 + }, + { + "epoch": 2.263479892324007, + "grad_norm": 0.0008380187791772187, + "learning_rate": 0.21941783125259198, + "loss": 0.3471, + "num_input_tokens_seen": 28965136, + "step": 13875 + }, + { + "epoch": 2.264295619544824, + "grad_norm": 0.0005294785951264203, + "learning_rate": 0.21936560849214226, + "loss": 0.3359, + "num_input_tokens_seen": 28975520, + "step": 13880 + }, + { + "epoch": 2.2651113467656416, + "grad_norm": 0.001012178836390376, + "learning_rate": 0.21931337503464404, + "loss": 0.3064, + "num_input_tokens_seen": 28985424, + "step": 13885 + }, + { + "epoch": 2.2659270739864588, + "grad_norm": 0.0010440004989504814, + "learning_rate": 0.21926113088815233, + "loss": 0.3433, + "num_input_tokens_seen": 28995040, + "step": 13890 + }, + { + "epoch": 2.2667428012072763, + "grad_norm": 0.0009306969586759806, + "learning_rate": 0.2192088760607238, + "loss": 0.2986, + "num_input_tokens_seen": 29005008, + "step": 13895 + }, + { + "epoch": 2.2675585284280935, + "grad_norm": 0.0004152717301622033, + "learning_rate": 0.2191566105604169, + "loss": 0.3784, + "num_input_tokens_seen": 29014944, + "step": 13900 + }, + { + "epoch": 2.268374255648911, + "grad_norm": 0.0003614372108131647, + "learning_rate": 0.21910433439529153, + "loss": 0.3188, + "num_input_tokens_seen": 29024880, + "step": 13905 + }, + { + "epoch": 2.269189982869728, + "grad_norm": 0.0004885232774540782, + "learning_rate": 0.2190520475734094, + "loss": 0.3168, + "num_input_tokens_seen": 29034784, + "step": 13910 + }, + { + "epoch": 2.2700057100905457, + "grad_norm": 0.00073239766061306, + "learning_rate": 0.2189997501028338, + "loss": 0.3571, + "num_input_tokens_seen": 29043744, + "step": 13915 + }, + { + "epoch": 2.270821437311363, + "grad_norm": 0.0004920100909657776, + "learning_rate": 0.2189474419916296, + "loss": 0.3698, + "num_input_tokens_seen": 29054128, + "step": 13920 + }, + { + "epoch": 2.2716371645321805, + "grad_norm": 0.0009325509308837354, + "learning_rate": 0.21889512324786342, + "loss": 0.3191, + "num_input_tokens_seen": 29063200, + "step": 13925 + }, + { + "epoch": 2.272452891752998, + "grad_norm": 0.0004417099116835743, + "learning_rate": 0.21884279387960345, + "loss": 0.3095, + "num_input_tokens_seen": 29073488, + "step": 13930 + }, + { + "epoch": 2.273268618973815, + "grad_norm": 0.0005979503621347249, + "learning_rate": 0.2187904538949195, + "loss": 0.3462, + "num_input_tokens_seen": 29082896, + "step": 13935 + }, + { + "epoch": 2.2740843461946323, + "grad_norm": 0.00038784320349805057, + "learning_rate": 0.2187381033018831, + "loss": 0.4017, + "num_input_tokens_seen": 29093968, + "step": 13940 + }, + { + "epoch": 2.27490007341545, + "grad_norm": 0.00045280755148269236, + "learning_rate": 0.2186857421085673, + "loss": 0.3391, + "num_input_tokens_seen": 29103984, + "step": 13945 + }, + { + "epoch": 2.2757158006362674, + "grad_norm": 0.0011919826501980424, + "learning_rate": 0.21863337032304697, + "loss": 0.326, + "num_input_tokens_seen": 29115280, + "step": 13950 + }, + { + "epoch": 2.2765315278570846, + "grad_norm": 0.00036269990960136056, + "learning_rate": 0.21858098795339845, + "loss": 0.3017, + "num_input_tokens_seen": 29126464, + "step": 13955 + }, + { + "epoch": 2.2773472550779017, + "grad_norm": 0.0004571315075736493, + "learning_rate": 0.21852859500769975, + "loss": 0.3154, + "num_input_tokens_seen": 29137424, + "step": 13960 + }, + { + "epoch": 2.2781629822987193, + "grad_norm": 0.0008227949147112668, + "learning_rate": 0.21847619149403044, + "loss": 0.3895, + "num_input_tokens_seen": 29148448, + "step": 13965 + }, + { + "epoch": 2.278978709519537, + "grad_norm": 0.0006820275448262691, + "learning_rate": 0.21842377742047195, + "loss": 0.3426, + "num_input_tokens_seen": 29159104, + "step": 13970 + }, + { + "epoch": 2.279794436740354, + "grad_norm": 0.0006459500291384757, + "learning_rate": 0.21837135279510705, + "loss": 0.303, + "num_input_tokens_seen": 29169232, + "step": 13975 + }, + { + "epoch": 2.2806101639611716, + "grad_norm": 0.0006416604737751186, + "learning_rate": 0.21831891762602038, + "loss": 0.3199, + "num_input_tokens_seen": 29179232, + "step": 13980 + }, + { + "epoch": 2.2814258911819887, + "grad_norm": 0.0011990517377853394, + "learning_rate": 0.21826647192129806, + "loss": 0.3227, + "num_input_tokens_seen": 29189408, + "step": 13985 + }, + { + "epoch": 2.2822416184028063, + "grad_norm": 0.00044266070472076535, + "learning_rate": 0.21821401568902787, + "loss": 0.3546, + "num_input_tokens_seen": 29200160, + "step": 13990 + }, + { + "epoch": 2.2830573456236234, + "grad_norm": 0.00045713272993452847, + "learning_rate": 0.21816154893729925, + "loss": 0.3653, + "num_input_tokens_seen": 29211872, + "step": 13995 + }, + { + "epoch": 2.283873072844441, + "grad_norm": 0.0004059972707182169, + "learning_rate": 0.2181090716742032, + "loss": 0.3195, + "num_input_tokens_seen": 29222656, + "step": 14000 + }, + { + "epoch": 2.283873072844441, + "eval_loss": 0.3187328279018402, + "eval_runtime": 155.8901, + "eval_samples_per_second": 17.48, + "eval_steps_per_second": 8.743, + "num_input_tokens_seen": 29222656, + "step": 14000 + }, + { + "epoch": 2.284688800065258, + "grad_norm": 0.000766587967518717, + "learning_rate": 0.21805658390783236, + "loss": 0.3224, + "num_input_tokens_seen": 29233840, + "step": 14005 + }, + { + "epoch": 2.2855045272860757, + "grad_norm": 0.0006906762137077749, + "learning_rate": 0.21800408564628107, + "loss": 0.334, + "num_input_tokens_seen": 29244288, + "step": 14010 + }, + { + "epoch": 2.286320254506893, + "grad_norm": 0.0005164851318113506, + "learning_rate": 0.21795157689764516, + "loss": 0.3497, + "num_input_tokens_seen": 29253968, + "step": 14015 + }, + { + "epoch": 2.2871359817277104, + "grad_norm": 0.0007244653534144163, + "learning_rate": 0.21789905767002216, + "loss": 0.2936, + "num_input_tokens_seen": 29265168, + "step": 14020 + }, + { + "epoch": 2.2879517089485275, + "grad_norm": 0.000551245640963316, + "learning_rate": 0.2178465279715112, + "loss": 0.3296, + "num_input_tokens_seen": 29274496, + "step": 14025 + }, + { + "epoch": 2.288767436169345, + "grad_norm": 0.0003180103958584368, + "learning_rate": 0.21779398781021303, + "loss": 0.3161, + "num_input_tokens_seen": 29284496, + "step": 14030 + }, + { + "epoch": 2.289583163390162, + "grad_norm": 0.0006760053802281618, + "learning_rate": 0.21774143719422998, + "loss": 0.3594, + "num_input_tokens_seen": 29295264, + "step": 14035 + }, + { + "epoch": 2.29039889061098, + "grad_norm": 0.00039831403410062194, + "learning_rate": 0.21768887613166601, + "loss": 0.3151, + "num_input_tokens_seen": 29305280, + "step": 14040 + }, + { + "epoch": 2.291214617831797, + "grad_norm": 0.0005120692076161504, + "learning_rate": 0.2176363046306267, + "loss": 0.3154, + "num_input_tokens_seen": 29314640, + "step": 14045 + }, + { + "epoch": 2.2920303450526145, + "grad_norm": 0.00044513464672490954, + "learning_rate": 0.21758372269921925, + "loss": 0.2795, + "num_input_tokens_seen": 29325504, + "step": 14050 + }, + { + "epoch": 2.2928460722734316, + "grad_norm": 0.0006930839736014605, + "learning_rate": 0.21753113034555244, + "loss": 0.2896, + "num_input_tokens_seen": 29335632, + "step": 14055 + }, + { + "epoch": 2.293661799494249, + "grad_norm": 0.00040326648741029203, + "learning_rate": 0.2174785275777367, + "loss": 0.2909, + "num_input_tokens_seen": 29346896, + "step": 14060 + }, + { + "epoch": 2.2944775267150663, + "grad_norm": 0.000536229694262147, + "learning_rate": 0.21742591440388404, + "loss": 0.3131, + "num_input_tokens_seen": 29357008, + "step": 14065 + }, + { + "epoch": 2.295293253935884, + "grad_norm": 0.0003188415430486202, + "learning_rate": 0.21737329083210802, + "loss": 0.3401, + "num_input_tokens_seen": 29366960, + "step": 14070 + }, + { + "epoch": 2.296108981156701, + "grad_norm": 0.0006602091598324478, + "learning_rate": 0.2173206568705239, + "loss": 0.346, + "num_input_tokens_seen": 29377664, + "step": 14075 + }, + { + "epoch": 2.2969247083775186, + "grad_norm": 0.00094573077512905, + "learning_rate": 0.2172680125272485, + "loss": 0.3744, + "num_input_tokens_seen": 29388528, + "step": 14080 + }, + { + "epoch": 2.2977404355983357, + "grad_norm": 0.0004373564734123647, + "learning_rate": 0.2172153578104002, + "loss": 0.3128, + "num_input_tokens_seen": 29398960, + "step": 14085 + }, + { + "epoch": 2.2985561628191533, + "grad_norm": 0.0004426162922754884, + "learning_rate": 0.21716269272809902, + "loss": 0.3255, + "num_input_tokens_seen": 29409936, + "step": 14090 + }, + { + "epoch": 2.2993718900399704, + "grad_norm": 0.0005198528524488211, + "learning_rate": 0.21711001728846666, + "loss": 0.3608, + "num_input_tokens_seen": 29421136, + "step": 14095 + }, + { + "epoch": 2.300187617260788, + "grad_norm": 0.0007064444944262505, + "learning_rate": 0.21705733149962628, + "loss": 0.3196, + "num_input_tokens_seen": 29431168, + "step": 14100 + }, + { + "epoch": 2.3010033444816056, + "grad_norm": 0.0009363348362967372, + "learning_rate": 0.21700463536970263, + "loss": 0.3161, + "num_input_tokens_seen": 29441936, + "step": 14105 + }, + { + "epoch": 2.3018190717024227, + "grad_norm": 0.0005059175309725106, + "learning_rate": 0.21695192890682222, + "loss": 0.3194, + "num_input_tokens_seen": 29451808, + "step": 14110 + }, + { + "epoch": 2.30263479892324, + "grad_norm": 0.0009135217987932265, + "learning_rate": 0.21689921211911298, + "loss": 0.2423, + "num_input_tokens_seen": 29462464, + "step": 14115 + }, + { + "epoch": 2.3034505261440574, + "grad_norm": 0.0005987958866171539, + "learning_rate": 0.21684648501470452, + "loss": 0.2952, + "num_input_tokens_seen": 29473552, + "step": 14120 + }, + { + "epoch": 2.304266253364875, + "grad_norm": 0.0008000672678463161, + "learning_rate": 0.216793747601728, + "loss": 0.3784, + "num_input_tokens_seen": 29482896, + "step": 14125 + }, + { + "epoch": 2.305081980585692, + "grad_norm": 0.0007527083507739007, + "learning_rate": 0.21674099988831627, + "loss": 0.2903, + "num_input_tokens_seen": 29492912, + "step": 14130 + }, + { + "epoch": 2.3058977078065093, + "grad_norm": 0.0020074560306966305, + "learning_rate": 0.21668824188260363, + "loss": 0.366, + "num_input_tokens_seen": 29503536, + "step": 14135 + }, + { + "epoch": 2.306713435027327, + "grad_norm": 0.0006598015315830708, + "learning_rate": 0.21663547359272606, + "loss": 0.326, + "num_input_tokens_seen": 29513328, + "step": 14140 + }, + { + "epoch": 2.3075291622481444, + "grad_norm": 0.0003107559750787914, + "learning_rate": 0.216582695026821, + "loss": 0.3341, + "num_input_tokens_seen": 29523856, + "step": 14145 + }, + { + "epoch": 2.3083448894689615, + "grad_norm": 0.00036366909625940025, + "learning_rate": 0.21652990619302767, + "loss": 0.2757, + "num_input_tokens_seen": 29535504, + "step": 14150 + }, + { + "epoch": 2.309160616689779, + "grad_norm": 0.0004450998385436833, + "learning_rate": 0.21647710709948673, + "loss": 0.3553, + "num_input_tokens_seen": 29545376, + "step": 14155 + }, + { + "epoch": 2.3099763439105963, + "grad_norm": 0.0005304207443259656, + "learning_rate": 0.2164242977543405, + "loss": 0.3167, + "num_input_tokens_seen": 29554848, + "step": 14160 + }, + { + "epoch": 2.310792071131414, + "grad_norm": 0.00041900851647369564, + "learning_rate": 0.21637147816573277, + "loss": 0.2738, + "num_input_tokens_seen": 29566080, + "step": 14165 + }, + { + "epoch": 2.311607798352231, + "grad_norm": 0.0003885718178935349, + "learning_rate": 0.21631864834180908, + "loss": 0.2651, + "num_input_tokens_seen": 29575968, + "step": 14170 + }, + { + "epoch": 2.3124235255730485, + "grad_norm": 0.0011924317805096507, + "learning_rate": 0.21626580829071637, + "loss": 0.2789, + "num_input_tokens_seen": 29586512, + "step": 14175 + }, + { + "epoch": 2.3132392527938657, + "grad_norm": 0.0008138471748679876, + "learning_rate": 0.21621295802060328, + "loss": 0.3879, + "num_input_tokens_seen": 29598816, + "step": 14180 + }, + { + "epoch": 2.3140549800146832, + "grad_norm": 0.0008354602614417672, + "learning_rate": 0.21616009753961996, + "loss": 0.3306, + "num_input_tokens_seen": 29608768, + "step": 14185 + }, + { + "epoch": 2.3148707072355004, + "grad_norm": 0.0004122811078559607, + "learning_rate": 0.2161072268559182, + "loss": 0.2982, + "num_input_tokens_seen": 29620032, + "step": 14190 + }, + { + "epoch": 2.315686434456318, + "grad_norm": 0.0008757372852414846, + "learning_rate": 0.21605434597765133, + "loss": 0.3286, + "num_input_tokens_seen": 29631728, + "step": 14195 + }, + { + "epoch": 2.316502161677135, + "grad_norm": 0.0004672003269661218, + "learning_rate": 0.21600145491297418, + "loss": 0.3708, + "num_input_tokens_seen": 29642224, + "step": 14200 + }, + { + "epoch": 2.316502161677135, + "eval_loss": 0.31858864426612854, + "eval_runtime": 155.9268, + "eval_samples_per_second": 17.476, + "eval_steps_per_second": 8.741, + "num_input_tokens_seen": 29642224, + "step": 14200 + }, + { + "epoch": 2.3173178888979526, + "grad_norm": 0.0006966868531890213, + "learning_rate": 0.21594855367004326, + "loss": 0.3186, + "num_input_tokens_seen": 29653872, + "step": 14205 + }, + { + "epoch": 2.31813361611877, + "grad_norm": 0.0007643174612894654, + "learning_rate": 0.21589564225701663, + "loss": 0.2716, + "num_input_tokens_seen": 29664560, + "step": 14210 + }, + { + "epoch": 2.3189493433395874, + "grad_norm": 0.0007160252425819635, + "learning_rate": 0.21584272068205385, + "loss": 0.2971, + "num_input_tokens_seen": 29676496, + "step": 14215 + }, + { + "epoch": 2.3197650705604045, + "grad_norm": 0.00038060141378082335, + "learning_rate": 0.2157897889533161, + "loss": 0.2911, + "num_input_tokens_seen": 29687824, + "step": 14220 + }, + { + "epoch": 2.320580797781222, + "grad_norm": 0.0005163239547982812, + "learning_rate": 0.21573684707896612, + "loss": 0.283, + "num_input_tokens_seen": 29698720, + "step": 14225 + }, + { + "epoch": 2.321396525002039, + "grad_norm": 0.0008914459031075239, + "learning_rate": 0.21568389506716826, + "loss": 0.3533, + "num_input_tokens_seen": 29708240, + "step": 14230 + }, + { + "epoch": 2.3222122522228568, + "grad_norm": 0.00137691805139184, + "learning_rate": 0.21563093292608831, + "loss": 0.32, + "num_input_tokens_seen": 29718976, + "step": 14235 + }, + { + "epoch": 2.323027979443674, + "grad_norm": 0.0020514209754765034, + "learning_rate": 0.21557796066389376, + "loss": 0.4261, + "num_input_tokens_seen": 29728784, + "step": 14240 + }, + { + "epoch": 2.3238437066644915, + "grad_norm": 0.000775521679315716, + "learning_rate": 0.21552497828875353, + "loss": 0.3022, + "num_input_tokens_seen": 29738880, + "step": 14245 + }, + { + "epoch": 2.3246594338853086, + "grad_norm": 0.001065208693034947, + "learning_rate": 0.21547198580883828, + "loss": 0.3185, + "num_input_tokens_seen": 29750480, + "step": 14250 + }, + { + "epoch": 2.325475161106126, + "grad_norm": 0.0004345793859101832, + "learning_rate": 0.21541898323232, + "loss": 0.3224, + "num_input_tokens_seen": 29761520, + "step": 14255 + }, + { + "epoch": 2.3262908883269433, + "grad_norm": 0.0007211077027022839, + "learning_rate": 0.2153659705673724, + "loss": 0.3688, + "num_input_tokens_seen": 29771872, + "step": 14260 + }, + { + "epoch": 2.327106615547761, + "grad_norm": 0.0014147392939776182, + "learning_rate": 0.2153129478221707, + "loss": 0.3639, + "num_input_tokens_seen": 29779600, + "step": 14265 + }, + { + "epoch": 2.327922342768578, + "grad_norm": 0.0005335792666301131, + "learning_rate": 0.21525991500489164, + "loss": 0.2872, + "num_input_tokens_seen": 29790640, + "step": 14270 + }, + { + "epoch": 2.3287380699893956, + "grad_norm": 0.0005572998779825866, + "learning_rate": 0.21520687212371362, + "loss": 0.3515, + "num_input_tokens_seen": 29800640, + "step": 14275 + }, + { + "epoch": 2.3295537972102127, + "grad_norm": 0.0003190326096955687, + "learning_rate": 0.21515381918681648, + "loss": 0.3014, + "num_input_tokens_seen": 29811728, + "step": 14280 + }, + { + "epoch": 2.3303695244310303, + "grad_norm": 0.0006011218065395951, + "learning_rate": 0.21510075620238167, + "loss": 0.2866, + "num_input_tokens_seen": 29822288, + "step": 14285 + }, + { + "epoch": 2.3311852516518474, + "grad_norm": 0.0017422784585505724, + "learning_rate": 0.21504768317859208, + "loss": 0.3805, + "num_input_tokens_seen": 29833840, + "step": 14290 + }, + { + "epoch": 2.332000978872665, + "grad_norm": 0.0003820337587967515, + "learning_rate": 0.2149946001236323, + "loss": 0.3075, + "num_input_tokens_seen": 29845760, + "step": 14295 + }, + { + "epoch": 2.3328167060934826, + "grad_norm": 0.0007527418783865869, + "learning_rate": 0.21494150704568848, + "loss": 0.3369, + "num_input_tokens_seen": 29854816, + "step": 14300 + }, + { + "epoch": 2.3336324333142997, + "grad_norm": 0.0006683938554488122, + "learning_rate": 0.21488840395294811, + "loss": 0.3189, + "num_input_tokens_seen": 29864512, + "step": 14305 + }, + { + "epoch": 2.334448160535117, + "grad_norm": 0.00033342785900458694, + "learning_rate": 0.21483529085360042, + "loss": 0.279, + "num_input_tokens_seen": 29874544, + "step": 14310 + }, + { + "epoch": 2.3352638877559344, + "grad_norm": 0.0006154947332106531, + "learning_rate": 0.2147821677558361, + "loss": 0.2819, + "num_input_tokens_seen": 29883760, + "step": 14315 + }, + { + "epoch": 2.336079614976752, + "grad_norm": 0.0006129066459834576, + "learning_rate": 0.2147290346678475, + "loss": 0.2774, + "num_input_tokens_seen": 29894480, + "step": 14320 + }, + { + "epoch": 2.336895342197569, + "grad_norm": 0.0008143150480464101, + "learning_rate": 0.21467589159782827, + "loss": 0.3577, + "num_input_tokens_seen": 29905280, + "step": 14325 + }, + { + "epoch": 2.3377110694183862, + "grad_norm": 0.0007393661071546376, + "learning_rate": 0.21462273855397374, + "loss": 0.3256, + "num_input_tokens_seen": 29914832, + "step": 14330 + }, + { + "epoch": 2.338526796639204, + "grad_norm": 0.000419733754824847, + "learning_rate": 0.21456957554448083, + "loss": 0.2952, + "num_input_tokens_seen": 29925888, + "step": 14335 + }, + { + "epoch": 2.3393425238600214, + "grad_norm": 0.0008994811796583235, + "learning_rate": 0.21451640257754795, + "loss": 0.3123, + "num_input_tokens_seen": 29936368, + "step": 14340 + }, + { + "epoch": 2.3401582510808385, + "grad_norm": 0.0005173497484065592, + "learning_rate": 0.21446321966137508, + "loss": 0.2984, + "num_input_tokens_seen": 29946672, + "step": 14345 + }, + { + "epoch": 2.340973978301656, + "grad_norm": 0.0006168095860630274, + "learning_rate": 0.21441002680416354, + "loss": 0.26, + "num_input_tokens_seen": 29956544, + "step": 14350 + }, + { + "epoch": 2.3417897055224732, + "grad_norm": 0.0024730509612709284, + "learning_rate": 0.21435682401411654, + "loss": 0.3633, + "num_input_tokens_seen": 29966224, + "step": 14355 + }, + { + "epoch": 2.342605432743291, + "grad_norm": 0.0010164702543988824, + "learning_rate": 0.2143036112994385, + "loss": 0.2717, + "num_input_tokens_seen": 29977936, + "step": 14360 + }, + { + "epoch": 2.343421159964108, + "grad_norm": 0.0004578660009428859, + "learning_rate": 0.21425038866833548, + "loss": 0.2518, + "num_input_tokens_seen": 29990224, + "step": 14365 + }, + { + "epoch": 2.3442368871849255, + "grad_norm": 0.000985530437901616, + "learning_rate": 0.21419715612901508, + "loss": 0.3286, + "num_input_tokens_seen": 30000672, + "step": 14370 + }, + { + "epoch": 2.3450526144057426, + "grad_norm": 0.0009031554800458252, + "learning_rate": 0.21414391368968652, + "loss": 0.3527, + "num_input_tokens_seen": 30011664, + "step": 14375 + }, + { + "epoch": 2.34586834162656, + "grad_norm": 0.00044755221460945904, + "learning_rate": 0.21409066135856034, + "loss": 0.281, + "num_input_tokens_seen": 30023536, + "step": 14380 + }, + { + "epoch": 2.3466840688473773, + "grad_norm": 0.0003026303311344236, + "learning_rate": 0.21403739914384878, + "loss": 0.3108, + "num_input_tokens_seen": 30034016, + "step": 14385 + }, + { + "epoch": 2.347499796068195, + "grad_norm": 0.0003990405530203134, + "learning_rate": 0.21398412705376554, + "loss": 0.3134, + "num_input_tokens_seen": 30045520, + "step": 14390 + }, + { + "epoch": 2.348315523289012, + "grad_norm": 0.0010798489674925804, + "learning_rate": 0.2139308450965258, + "loss": 0.3765, + "num_input_tokens_seen": 30055440, + "step": 14395 + }, + { + "epoch": 2.3491312505098296, + "grad_norm": 0.00031945735099725425, + "learning_rate": 0.21387755328034638, + "loss": 0.3423, + "num_input_tokens_seen": 30064704, + "step": 14400 + }, + { + "epoch": 2.3491312505098296, + "eval_loss": 0.32697173953056335, + "eval_runtime": 155.9467, + "eval_samples_per_second": 17.474, + "eval_steps_per_second": 8.74, + "num_input_tokens_seen": 30064704, + "step": 14400 + }, + { + "epoch": 2.3499469777306468, + "grad_norm": 0.000512758968397975, + "learning_rate": 0.2138242516134455, + "loss": 0.3115, + "num_input_tokens_seen": 30074832, + "step": 14405 + }, + { + "epoch": 2.3507627049514643, + "grad_norm": 0.0008115223026834428, + "learning_rate": 0.2137709401040429, + "loss": 0.3513, + "num_input_tokens_seen": 30084592, + "step": 14410 + }, + { + "epoch": 2.3515784321722815, + "grad_norm": 0.001339829876087606, + "learning_rate": 0.21371761876036, + "loss": 0.3149, + "num_input_tokens_seen": 30095216, + "step": 14415 + }, + { + "epoch": 2.352394159393099, + "grad_norm": 0.0003946369106415659, + "learning_rate": 0.21366428759061956, + "loss": 0.3123, + "num_input_tokens_seen": 30106320, + "step": 14420 + }, + { + "epoch": 2.353209886613916, + "grad_norm": 0.0004390110552776605, + "learning_rate": 0.2136109466030459, + "loss": 0.2949, + "num_input_tokens_seen": 30116320, + "step": 14425 + }, + { + "epoch": 2.3540256138347337, + "grad_norm": 0.004234457854181528, + "learning_rate": 0.2135575958058649, + "loss": 0.3924, + "num_input_tokens_seen": 30127712, + "step": 14430 + }, + { + "epoch": 2.354841341055551, + "grad_norm": 0.00036483374424278736, + "learning_rate": 0.2135042352073039, + "loss": 0.2886, + "num_input_tokens_seen": 30137312, + "step": 14435 + }, + { + "epoch": 2.3556570682763684, + "grad_norm": 0.00028845679480582476, + "learning_rate": 0.2134508648155918, + "loss": 0.3918, + "num_input_tokens_seen": 30148848, + "step": 14440 + }, + { + "epoch": 2.3564727954971856, + "grad_norm": 0.0008395476033911109, + "learning_rate": 0.213397484638959, + "loss": 0.3096, + "num_input_tokens_seen": 30158944, + "step": 14445 + }, + { + "epoch": 2.357288522718003, + "grad_norm": 0.0005300433258526027, + "learning_rate": 0.21334409468563728, + "loss": 0.2597, + "num_input_tokens_seen": 30167920, + "step": 14450 + }, + { + "epoch": 2.3581042499388203, + "grad_norm": 0.0004711307992693037, + "learning_rate": 0.2132906949638602, + "loss": 0.3979, + "num_input_tokens_seen": 30179664, + "step": 14455 + }, + { + "epoch": 2.358919977159638, + "grad_norm": 0.0004508869315031916, + "learning_rate": 0.21323728548186255, + "loss": 0.2961, + "num_input_tokens_seen": 30191488, + "step": 14460 + }, + { + "epoch": 2.359735704380455, + "grad_norm": 0.0007526202825829387, + "learning_rate": 0.21318386624788088, + "loss": 0.3052, + "num_input_tokens_seen": 30202240, + "step": 14465 + }, + { + "epoch": 2.3605514316012726, + "grad_norm": 0.000536420731805265, + "learning_rate": 0.21313043727015288, + "loss": 0.2896, + "num_input_tokens_seen": 30212496, + "step": 14470 + }, + { + "epoch": 2.36136715882209, + "grad_norm": 0.0006473761168308556, + "learning_rate": 0.2130769985569182, + "loss": 0.3438, + "num_input_tokens_seen": 30221872, + "step": 14475 + }, + { + "epoch": 2.3621828860429073, + "grad_norm": 0.0006357617676258087, + "learning_rate": 0.21302355011641766, + "loss": 0.2933, + "num_input_tokens_seen": 30232960, + "step": 14480 + }, + { + "epoch": 2.3629986132637244, + "grad_norm": 0.0004215839726384729, + "learning_rate": 0.21297009195689365, + "loss": 0.2692, + "num_input_tokens_seen": 30244256, + "step": 14485 + }, + { + "epoch": 2.363814340484542, + "grad_norm": 0.0004828058881685138, + "learning_rate": 0.21291662408659015, + "loss": 0.2483, + "num_input_tokens_seen": 30253104, + "step": 14490 + }, + { + "epoch": 2.3646300677053596, + "grad_norm": 0.0011717525776475668, + "learning_rate": 0.21286314651375254, + "loss": 0.343, + "num_input_tokens_seen": 30264096, + "step": 14495 + }, + { + "epoch": 2.3654457949261767, + "grad_norm": 0.0006580836488865316, + "learning_rate": 0.2128096592466278, + "loss": 0.3801, + "num_input_tokens_seen": 30273744, + "step": 14500 + }, + { + "epoch": 2.366261522146994, + "grad_norm": 0.0004431513953022659, + "learning_rate": 0.21275616229346428, + "loss": 0.2911, + "num_input_tokens_seen": 30283232, + "step": 14505 + }, + { + "epoch": 2.3670772493678114, + "grad_norm": 0.00040307201561518013, + "learning_rate": 0.21270265566251184, + "loss": 0.3111, + "num_input_tokens_seen": 30293040, + "step": 14510 + }, + { + "epoch": 2.367892976588629, + "grad_norm": 0.0007944535464048386, + "learning_rate": 0.21264913936202193, + "loss": 0.2983, + "num_input_tokens_seen": 30303968, + "step": 14515 + }, + { + "epoch": 2.368708703809446, + "grad_norm": 0.0006719930097460747, + "learning_rate": 0.2125956134002475, + "loss": 0.2903, + "num_input_tokens_seen": 30314576, + "step": 14520 + }, + { + "epoch": 2.3695244310302637, + "grad_norm": 0.0008525322773493826, + "learning_rate": 0.2125420777854428, + "loss": 0.4016, + "num_input_tokens_seen": 30324208, + "step": 14525 + }, + { + "epoch": 2.370340158251081, + "grad_norm": 0.0005327899125404656, + "learning_rate": 0.21248853252586372, + "loss": 0.3109, + "num_input_tokens_seen": 30333680, + "step": 14530 + }, + { + "epoch": 2.3711558854718984, + "grad_norm": 0.0008275792351923883, + "learning_rate": 0.21243497762976774, + "loss": 0.326, + "num_input_tokens_seen": 30342384, + "step": 14535 + }, + { + "epoch": 2.3719716126927155, + "grad_norm": 0.0009037061245180666, + "learning_rate": 0.21238141310541356, + "loss": 0.3274, + "num_input_tokens_seen": 30352784, + "step": 14540 + }, + { + "epoch": 2.372787339913533, + "grad_norm": 0.00040289349271915853, + "learning_rate": 0.21232783896106153, + "loss": 0.3112, + "num_input_tokens_seen": 30362976, + "step": 14545 + }, + { + "epoch": 2.37360306713435, + "grad_norm": 0.0010582611430436373, + "learning_rate": 0.21227425520497345, + "loss": 0.444, + "num_input_tokens_seen": 30373696, + "step": 14550 + }, + { + "epoch": 2.374418794355168, + "grad_norm": 0.0007014498696662486, + "learning_rate": 0.2122206618454127, + "loss": 0.3207, + "num_input_tokens_seen": 30384000, + "step": 14555 + }, + { + "epoch": 2.375234521575985, + "grad_norm": 0.000890367547981441, + "learning_rate": 0.2121670588906439, + "loss": 0.279, + "num_input_tokens_seen": 30394352, + "step": 14560 + }, + { + "epoch": 2.3760502487968025, + "grad_norm": 0.0007217544480226934, + "learning_rate": 0.21211344634893345, + "loss": 0.3293, + "num_input_tokens_seen": 30405232, + "step": 14565 + }, + { + "epoch": 2.3768659760176196, + "grad_norm": 0.0006691768649034202, + "learning_rate": 0.21205982422854897, + "loss": 0.2817, + "num_input_tokens_seen": 30415088, + "step": 14570 + }, + { + "epoch": 2.377681703238437, + "grad_norm": 0.0002647550718393177, + "learning_rate": 0.21200619253775974, + "loss": 0.2966, + "num_input_tokens_seen": 30426272, + "step": 14575 + }, + { + "epoch": 2.3784974304592543, + "grad_norm": 0.00031370072974823415, + "learning_rate": 0.21195255128483637, + "loss": 0.3147, + "num_input_tokens_seen": 30436848, + "step": 14580 + }, + { + "epoch": 2.379313157680072, + "grad_norm": 0.0009749605669640005, + "learning_rate": 0.21189890047805102, + "loss": 0.4061, + "num_input_tokens_seen": 30448144, + "step": 14585 + }, + { + "epoch": 2.380128884900889, + "grad_norm": 0.0008524152799509466, + "learning_rate": 0.21184524012567735, + "loss": 0.3409, + "num_input_tokens_seen": 30459360, + "step": 14590 + }, + { + "epoch": 2.3809446121217066, + "grad_norm": 0.0009002635488286614, + "learning_rate": 0.2117915702359905, + "loss": 0.3716, + "num_input_tokens_seen": 30470560, + "step": 14595 + }, + { + "epoch": 2.3817603393425237, + "grad_norm": 0.0005698651657439768, + "learning_rate": 0.211737890817267, + "loss": 0.3308, + "num_input_tokens_seen": 30481488, + "step": 14600 + }, + { + "epoch": 2.3817603393425237, + "eval_loss": 0.32414165139198303, + "eval_runtime": 156.0266, + "eval_samples_per_second": 17.465, + "eval_steps_per_second": 8.736, + "num_input_tokens_seen": 30481488, + "step": 14600 + }, + { + "epoch": 2.3825760665633413, + "grad_norm": 0.0008987071923911572, + "learning_rate": 0.21168420187778483, + "loss": 0.285, + "num_input_tokens_seen": 30491696, + "step": 14605 + }, + { + "epoch": 2.3833917937841584, + "grad_norm": 0.0004939140635542572, + "learning_rate": 0.21163050342582362, + "loss": 0.3156, + "num_input_tokens_seen": 30501200, + "step": 14610 + }, + { + "epoch": 2.384207521004976, + "grad_norm": 0.0008342473884113133, + "learning_rate": 0.21157679546966426, + "loss": 0.3065, + "num_input_tokens_seen": 30510432, + "step": 14615 + }, + { + "epoch": 2.385023248225793, + "grad_norm": 0.0005994322709739208, + "learning_rate": 0.2115230780175892, + "loss": 0.3302, + "num_input_tokens_seen": 30520592, + "step": 14620 + }, + { + "epoch": 2.3858389754466107, + "grad_norm": 0.0007060667267069221, + "learning_rate": 0.21146935107788237, + "loss": 0.318, + "num_input_tokens_seen": 30531296, + "step": 14625 + }, + { + "epoch": 2.386654702667428, + "grad_norm": 0.0005610733642242849, + "learning_rate": 0.21141561465882916, + "loss": 0.2854, + "num_input_tokens_seen": 30543184, + "step": 14630 + }, + { + "epoch": 2.3874704298882454, + "grad_norm": 0.0004720187862403691, + "learning_rate": 0.21136186876871635, + "loss": 0.2886, + "num_input_tokens_seen": 30553072, + "step": 14635 + }, + { + "epoch": 2.3882861571090626, + "grad_norm": 0.0007974999025464058, + "learning_rate": 0.21130811341583225, + "loss": 0.2826, + "num_input_tokens_seen": 30563600, + "step": 14640 + }, + { + "epoch": 2.38910188432988, + "grad_norm": 0.0007345480844378471, + "learning_rate": 0.21125434860846667, + "loss": 0.2928, + "num_input_tokens_seen": 30574320, + "step": 14645 + }, + { + "epoch": 2.3899176115506973, + "grad_norm": 0.0020395005121827126, + "learning_rate": 0.2112005743549107, + "loss": 0.3828, + "num_input_tokens_seen": 30585232, + "step": 14650 + }, + { + "epoch": 2.390733338771515, + "grad_norm": 0.000513229111675173, + "learning_rate": 0.21114679066345707, + "loss": 0.3359, + "num_input_tokens_seen": 30596096, + "step": 14655 + }, + { + "epoch": 2.391549065992332, + "grad_norm": 0.0007069871644489467, + "learning_rate": 0.21109299754239993, + "loss": 0.3833, + "num_input_tokens_seen": 30607264, + "step": 14660 + }, + { + "epoch": 2.3923647932131495, + "grad_norm": 0.000392142275813967, + "learning_rate": 0.21103919500003482, + "loss": 0.3339, + "num_input_tokens_seen": 30617344, + "step": 14665 + }, + { + "epoch": 2.393180520433967, + "grad_norm": 0.0005552921211346984, + "learning_rate": 0.21098538304465872, + "loss": 0.3501, + "num_input_tokens_seen": 30629600, + "step": 14670 + }, + { + "epoch": 2.3939962476547842, + "grad_norm": 0.0003799316182266921, + "learning_rate": 0.2109315616845702, + "loss": 0.2253, + "num_input_tokens_seen": 30640864, + "step": 14675 + }, + { + "epoch": 2.3948119748756014, + "grad_norm": 0.0003049572405871004, + "learning_rate": 0.21087773092806925, + "loss": 0.3585, + "num_input_tokens_seen": 30650272, + "step": 14680 + }, + { + "epoch": 2.395627702096419, + "grad_norm": 0.0003809739137068391, + "learning_rate": 0.21082389078345704, + "loss": 0.3183, + "num_input_tokens_seen": 30662080, + "step": 14685 + }, + { + "epoch": 2.3964434293172365, + "grad_norm": 0.0007267841137945652, + "learning_rate": 0.2107700412590365, + "loss": 0.3331, + "num_input_tokens_seen": 30672416, + "step": 14690 + }, + { + "epoch": 2.3972591565380537, + "grad_norm": 0.0005271521513350308, + "learning_rate": 0.210716182363112, + "loss": 0.3266, + "num_input_tokens_seen": 30683440, + "step": 14695 + }, + { + "epoch": 2.398074883758871, + "grad_norm": 0.0005679088062606752, + "learning_rate": 0.2106623141039891, + "loss": 0.2795, + "num_input_tokens_seen": 30693136, + "step": 14700 + }, + { + "epoch": 2.3988906109796884, + "grad_norm": 0.00045227151713334024, + "learning_rate": 0.21060843648997507, + "loss": 0.3669, + "num_input_tokens_seen": 30702672, + "step": 14705 + }, + { + "epoch": 2.399706338200506, + "grad_norm": 0.0012834342196583748, + "learning_rate": 0.21055454952937844, + "loss": 0.283, + "num_input_tokens_seen": 30713680, + "step": 14710 + }, + { + "epoch": 2.400522065421323, + "grad_norm": 0.0005400861846283078, + "learning_rate": 0.21050065323050937, + "loss": 0.3067, + "num_input_tokens_seen": 30724016, + "step": 14715 + }, + { + "epoch": 2.4013377926421406, + "grad_norm": 0.0008784040692262352, + "learning_rate": 0.21044674760167928, + "loss": 0.3454, + "num_input_tokens_seen": 30734096, + "step": 14720 + }, + { + "epoch": 2.4021535198629578, + "grad_norm": 0.0003371852508280426, + "learning_rate": 0.210392832651201, + "loss": 0.3331, + "num_input_tokens_seen": 30745136, + "step": 14725 + }, + { + "epoch": 2.4029692470837754, + "grad_norm": 0.00042465439764782786, + "learning_rate": 0.210338908387389, + "loss": 0.3352, + "num_input_tokens_seen": 30756032, + "step": 14730 + }, + { + "epoch": 2.4037849743045925, + "grad_norm": 0.00030482516740448773, + "learning_rate": 0.21028497481855912, + "loss": 0.3073, + "num_input_tokens_seen": 30766208, + "step": 14735 + }, + { + "epoch": 2.40460070152541, + "grad_norm": 0.0004723676247522235, + "learning_rate": 0.21023103195302847, + "loss": 0.2803, + "num_input_tokens_seen": 30776864, + "step": 14740 + }, + { + "epoch": 2.405416428746227, + "grad_norm": 0.0009691065060906112, + "learning_rate": 0.21017707979911582, + "loss": 0.2538, + "num_input_tokens_seen": 30786800, + "step": 14745 + }, + { + "epoch": 2.4062321559670448, + "grad_norm": 0.0005093807703815401, + "learning_rate": 0.21012311836514122, + "loss": 0.3125, + "num_input_tokens_seen": 30797072, + "step": 14750 + }, + { + "epoch": 2.407047883187862, + "grad_norm": 0.0009195592720061541, + "learning_rate": 0.21006914765942622, + "loss": 0.328, + "num_input_tokens_seen": 30808688, + "step": 14755 + }, + { + "epoch": 2.4078636104086795, + "grad_norm": 0.0005274680443108082, + "learning_rate": 0.2100151676902938, + "loss": 0.2999, + "num_input_tokens_seen": 30818016, + "step": 14760 + }, + { + "epoch": 2.4086793376294966, + "grad_norm": 0.0004363912157714367, + "learning_rate": 0.2099611784660683, + "loss": 0.3208, + "num_input_tokens_seen": 30828160, + "step": 14765 + }, + { + "epoch": 2.409495064850314, + "grad_norm": 0.0013481980422511697, + "learning_rate": 0.20990717999507552, + "loss": 0.3885, + "num_input_tokens_seen": 30837824, + "step": 14770 + }, + { + "epoch": 2.4103107920711313, + "grad_norm": 0.0009042919846251607, + "learning_rate": 0.20985317228564276, + "loss": 0.2806, + "num_input_tokens_seen": 30848544, + "step": 14775 + }, + { + "epoch": 2.411126519291949, + "grad_norm": 0.00048241220065392554, + "learning_rate": 0.20979915534609872, + "loss": 0.3076, + "num_input_tokens_seen": 30860304, + "step": 14780 + }, + { + "epoch": 2.411942246512766, + "grad_norm": 0.001312323147431016, + "learning_rate": 0.20974512918477342, + "loss": 0.3638, + "num_input_tokens_seen": 30871360, + "step": 14785 + }, + { + "epoch": 2.4127579737335836, + "grad_norm": 0.0003161607019137591, + "learning_rate": 0.2096910938099984, + "loss": 0.2905, + "num_input_tokens_seen": 30880800, + "step": 14790 + }, + { + "epoch": 2.4135737009544007, + "grad_norm": 0.0006560253095813096, + "learning_rate": 0.2096370492301066, + "loss": 0.334, + "num_input_tokens_seen": 30890192, + "step": 14795 + }, + { + "epoch": 2.4143894281752183, + "grad_norm": 0.0010373437544330955, + "learning_rate": 0.2095829954534323, + "loss": 0.3559, + "num_input_tokens_seen": 30900976, + "step": 14800 + }, + { + "epoch": 2.4143894281752183, + "eval_loss": 0.3218074440956116, + "eval_runtime": 155.9299, + "eval_samples_per_second": 17.476, + "eval_steps_per_second": 8.741, + "num_input_tokens_seen": 30900976, + "step": 14800 + }, + { + "epoch": 2.4152051553960354, + "grad_norm": 0.0009311250760219991, + "learning_rate": 0.2095289324883114, + "loss": 0.3134, + "num_input_tokens_seen": 30911120, + "step": 14805 + }, + { + "epoch": 2.416020882616853, + "grad_norm": 0.00071395852137357, + "learning_rate": 0.20947486034308097, + "loss": 0.3248, + "num_input_tokens_seen": 30921008, + "step": 14810 + }, + { + "epoch": 2.41683660983767, + "grad_norm": 0.0007855726289562881, + "learning_rate": 0.2094207790260797, + "loss": 0.3116, + "num_input_tokens_seen": 30931664, + "step": 14815 + }, + { + "epoch": 2.4176523370584877, + "grad_norm": 0.0004026470414828509, + "learning_rate": 0.20936668854564758, + "loss": 0.3475, + "num_input_tokens_seen": 30941824, + "step": 14820 + }, + { + "epoch": 2.418468064279305, + "grad_norm": 0.0005806844565086067, + "learning_rate": 0.20931258891012602, + "loss": 0.2889, + "num_input_tokens_seen": 30952608, + "step": 14825 + }, + { + "epoch": 2.4192837915001224, + "grad_norm": 0.0010659995023161173, + "learning_rate": 0.20925848012785792, + "loss": 0.3533, + "num_input_tokens_seen": 30960944, + "step": 14830 + }, + { + "epoch": 2.4200995187209395, + "grad_norm": 0.0003534907300490886, + "learning_rate": 0.20920436220718747, + "loss": 0.3455, + "num_input_tokens_seen": 30972512, + "step": 14835 + }, + { + "epoch": 2.420915245941757, + "grad_norm": 0.00038863293593749404, + "learning_rate": 0.20915023515646033, + "loss": 0.3012, + "num_input_tokens_seen": 30983520, + "step": 14840 + }, + { + "epoch": 2.4217309731625742, + "grad_norm": 0.0006031311932019889, + "learning_rate": 0.20909609898402368, + "loss": 0.3407, + "num_input_tokens_seen": 30993792, + "step": 14845 + }, + { + "epoch": 2.422546700383392, + "grad_norm": 0.00039071947685442865, + "learning_rate": 0.2090419536982258, + "loss": 0.3006, + "num_input_tokens_seen": 31003552, + "step": 14850 + }, + { + "epoch": 2.423362427604209, + "grad_norm": 0.0003311352338641882, + "learning_rate": 0.2089877993074168, + "loss": 0.3182, + "num_input_tokens_seen": 31013232, + "step": 14855 + }, + { + "epoch": 2.4241781548250265, + "grad_norm": 0.0008505430887453258, + "learning_rate": 0.20893363581994784, + "loss": 0.3671, + "num_input_tokens_seen": 31022432, + "step": 14860 + }, + { + "epoch": 2.424993882045844, + "grad_norm": 0.0003453787066973746, + "learning_rate": 0.2088794632441716, + "loss": 0.3224, + "num_input_tokens_seen": 31034512, + "step": 14865 + }, + { + "epoch": 2.4258096092666612, + "grad_norm": 0.0003704069531522691, + "learning_rate": 0.20882528158844219, + "loss": 0.2942, + "num_input_tokens_seen": 31044704, + "step": 14870 + }, + { + "epoch": 2.4266253364874784, + "grad_norm": 0.00035422499058768153, + "learning_rate": 0.20877109086111514, + "loss": 0.324, + "num_input_tokens_seen": 31055280, + "step": 14875 + }, + { + "epoch": 2.427441063708296, + "grad_norm": 0.00031895918073132634, + "learning_rate": 0.2087168910705473, + "loss": 0.3182, + "num_input_tokens_seen": 31066656, + "step": 14880 + }, + { + "epoch": 2.4282567909291135, + "grad_norm": 0.0006372383213602006, + "learning_rate": 0.208662682225097, + "loss": 0.2671, + "num_input_tokens_seen": 31079120, + "step": 14885 + }, + { + "epoch": 2.4290725181499306, + "grad_norm": 0.0004229994083289057, + "learning_rate": 0.2086084643331239, + "loss": 0.3513, + "num_input_tokens_seen": 31089872, + "step": 14890 + }, + { + "epoch": 2.429888245370748, + "grad_norm": 0.0006645721150562167, + "learning_rate": 0.20855423740298906, + "loss": 0.3015, + "num_input_tokens_seen": 31099744, + "step": 14895 + }, + { + "epoch": 2.4307039725915653, + "grad_norm": 0.0009326615836471319, + "learning_rate": 0.208500001443055, + "loss": 0.3011, + "num_input_tokens_seen": 31109760, + "step": 14900 + }, + { + "epoch": 2.431519699812383, + "grad_norm": 0.0007778042927384377, + "learning_rate": 0.20844575646168553, + "loss": 0.3259, + "num_input_tokens_seen": 31119664, + "step": 14905 + }, + { + "epoch": 2.4323354270332, + "grad_norm": 0.00039955382817424834, + "learning_rate": 0.20839150246724594, + "loss": 0.3175, + "num_input_tokens_seen": 31132128, + "step": 14910 + }, + { + "epoch": 2.4331511542540176, + "grad_norm": 0.0007631305488757789, + "learning_rate": 0.20833723946810287, + "loss": 0.2798, + "num_input_tokens_seen": 31143344, + "step": 14915 + }, + { + "epoch": 2.4339668814748348, + "grad_norm": 0.0004245859745424241, + "learning_rate": 0.20828296747262437, + "loss": 0.325, + "num_input_tokens_seen": 31153952, + "step": 14920 + }, + { + "epoch": 2.4347826086956523, + "grad_norm": 0.00039076924440450966, + "learning_rate": 0.20822868648917986, + "loss": 0.3421, + "num_input_tokens_seen": 31163680, + "step": 14925 + }, + { + "epoch": 2.4355983359164695, + "grad_norm": 0.0004083890817128122, + "learning_rate": 0.20817439652614017, + "loss": 0.3586, + "num_input_tokens_seen": 31174816, + "step": 14930 + }, + { + "epoch": 2.436414063137287, + "grad_norm": 0.000805168878287077, + "learning_rate": 0.20812009759187744, + "loss": 0.3315, + "num_input_tokens_seen": 31186400, + "step": 14935 + }, + { + "epoch": 2.437229790358104, + "grad_norm": 0.0007684531738050282, + "learning_rate": 0.2080657896947653, + "loss": 0.2877, + "num_input_tokens_seen": 31195712, + "step": 14940 + }, + { + "epoch": 2.4380455175789217, + "grad_norm": 0.0003560335608199239, + "learning_rate": 0.2080114728431787, + "loss": 0.3121, + "num_input_tokens_seen": 31205184, + "step": 14945 + }, + { + "epoch": 2.438861244799739, + "grad_norm": 0.0002653349074535072, + "learning_rate": 0.20795714704549392, + "loss": 0.22, + "num_input_tokens_seen": 31215504, + "step": 14950 + }, + { + "epoch": 2.4396769720205564, + "grad_norm": 0.0012336241779848933, + "learning_rate": 0.20790281231008875, + "loss": 0.3686, + "num_input_tokens_seen": 31226688, + "step": 14955 + }, + { + "epoch": 2.4404926992413736, + "grad_norm": 0.00043933611596003175, + "learning_rate": 0.20784846864534226, + "loss": 0.2528, + "num_input_tokens_seen": 31236048, + "step": 14960 + }, + { + "epoch": 2.441308426462191, + "grad_norm": 0.0003782349522225559, + "learning_rate": 0.20779411605963496, + "loss": 0.3473, + "num_input_tokens_seen": 31246864, + "step": 14965 + }, + { + "epoch": 2.4421241536830083, + "grad_norm": 0.000442318560089916, + "learning_rate": 0.2077397545613487, + "loss": 0.2747, + "num_input_tokens_seen": 31257808, + "step": 14970 + }, + { + "epoch": 2.442939880903826, + "grad_norm": 0.0008934937068261206, + "learning_rate": 0.20768538415886661, + "loss": 0.35, + "num_input_tokens_seen": 31268976, + "step": 14975 + }, + { + "epoch": 2.443755608124643, + "grad_norm": 0.0005046434234827757, + "learning_rate": 0.20763100486057343, + "loss": 0.3352, + "num_input_tokens_seen": 31279488, + "step": 14980 + }, + { + "epoch": 2.4445713353454606, + "grad_norm": 0.0010686106979846954, + "learning_rate": 0.20757661667485502, + "loss": 0.3452, + "num_input_tokens_seen": 31290400, + "step": 14985 + }, + { + "epoch": 2.4453870625662777, + "grad_norm": 0.0003896715643350035, + "learning_rate": 0.2075222196100988, + "loss": 0.2468, + "num_input_tokens_seen": 31301008, + "step": 14990 + }, + { + "epoch": 2.4462027897870953, + "grad_norm": 0.00046024960465729237, + "learning_rate": 0.20746781367469344, + "loss": 0.2608, + "num_input_tokens_seen": 31311360, + "step": 14995 + }, + { + "epoch": 2.4470185170079124, + "grad_norm": 0.0002843124093487859, + "learning_rate": 0.207413398877029, + "loss": 0.308, + "num_input_tokens_seen": 31321184, + "step": 15000 + }, + { + "epoch": 2.4470185170079124, + "eval_loss": 0.321241557598114, + "eval_runtime": 155.7579, + "eval_samples_per_second": 17.495, + "eval_steps_per_second": 8.751, + "num_input_tokens_seen": 31321184, + "step": 15000 + }, + { + "epoch": 2.44783424422873, + "grad_norm": 0.0007383562042377889, + "learning_rate": 0.20735897522549698, + "loss": 0.2959, + "num_input_tokens_seen": 31330304, + "step": 15005 + }, + { + "epoch": 2.448649971449547, + "grad_norm": 0.00035592488711699843, + "learning_rate": 0.2073045427284902, + "loss": 0.2845, + "num_input_tokens_seen": 31339040, + "step": 15010 + }, + { + "epoch": 2.4494656986703647, + "grad_norm": 0.0006959867314435542, + "learning_rate": 0.2072501013944027, + "loss": 0.3885, + "num_input_tokens_seen": 31349296, + "step": 15015 + }, + { + "epoch": 2.450281425891182, + "grad_norm": 0.0005747838295064867, + "learning_rate": 0.20719565123163017, + "loss": 0.3352, + "num_input_tokens_seen": 31358976, + "step": 15020 + }, + { + "epoch": 2.4510971531119994, + "grad_norm": 0.0005864295526407659, + "learning_rate": 0.20714119224856944, + "loss": 0.3514, + "num_input_tokens_seen": 31368864, + "step": 15025 + }, + { + "epoch": 2.4519128803328165, + "grad_norm": 0.0003932503459509462, + "learning_rate": 0.2070867244536188, + "loss": 0.2677, + "num_input_tokens_seen": 31378800, + "step": 15030 + }, + { + "epoch": 2.452728607553634, + "grad_norm": 0.0009164891089312732, + "learning_rate": 0.20703224785517785, + "loss": 0.2981, + "num_input_tokens_seen": 31389072, + "step": 15035 + }, + { + "epoch": 2.4535443347744517, + "grad_norm": 0.00045514979865401983, + "learning_rate": 0.20697776246164754, + "loss": 0.3081, + "num_input_tokens_seen": 31399312, + "step": 15040 + }, + { + "epoch": 2.454360061995269, + "grad_norm": 0.00041159073589369655, + "learning_rate": 0.2069232682814303, + "loss": 0.3208, + "num_input_tokens_seen": 31408864, + "step": 15045 + }, + { + "epoch": 2.455175789216086, + "grad_norm": 0.0007815703866071999, + "learning_rate": 0.20686876532292972, + "loss": 0.3479, + "num_input_tokens_seen": 31420176, + "step": 15050 + }, + { + "epoch": 2.4559915164369035, + "grad_norm": 0.0003058628353755921, + "learning_rate": 0.20681425359455083, + "loss": 0.2818, + "num_input_tokens_seen": 31431296, + "step": 15055 + }, + { + "epoch": 2.456807243657721, + "grad_norm": 0.0008273920975625515, + "learning_rate": 0.20675973310470008, + "loss": 0.3686, + "num_input_tokens_seen": 31441184, + "step": 15060 + }, + { + "epoch": 2.457622970878538, + "grad_norm": 0.0008520522969774902, + "learning_rate": 0.2067052038617852, + "loss": 0.3872, + "num_input_tokens_seen": 31451168, + "step": 15065 + }, + { + "epoch": 2.4584386980993553, + "grad_norm": 0.00038759346352890134, + "learning_rate": 0.2066506658742153, + "loss": 0.3018, + "num_input_tokens_seen": 31462624, + "step": 15070 + }, + { + "epoch": 2.459254425320173, + "grad_norm": 0.00040959729813039303, + "learning_rate": 0.20659611915040077, + "loss": 0.342, + "num_input_tokens_seen": 31473744, + "step": 15075 + }, + { + "epoch": 2.4600701525409905, + "grad_norm": 0.0004809004603885114, + "learning_rate": 0.20654156369875348, + "loss": 0.2864, + "num_input_tokens_seen": 31484640, + "step": 15080 + }, + { + "epoch": 2.4608858797618076, + "grad_norm": 0.0007907967083156109, + "learning_rate": 0.20648699952768648, + "loss": 0.3259, + "num_input_tokens_seen": 31494944, + "step": 15085 + }, + { + "epoch": 2.461701606982625, + "grad_norm": 0.0006529700476676226, + "learning_rate": 0.20643242664561437, + "loss": 0.2797, + "num_input_tokens_seen": 31504032, + "step": 15090 + }, + { + "epoch": 2.4625173342034423, + "grad_norm": 0.0004599677922669798, + "learning_rate": 0.20637784506095277, + "loss": 0.331, + "num_input_tokens_seen": 31514128, + "step": 15095 + }, + { + "epoch": 2.46333306142426, + "grad_norm": 0.0005706630763597786, + "learning_rate": 0.20632325478211908, + "loss": 0.3642, + "num_input_tokens_seen": 31525008, + "step": 15100 + }, + { + "epoch": 2.464148788645077, + "grad_norm": 0.0008962053107097745, + "learning_rate": 0.20626865581753165, + "loss": 0.3266, + "num_input_tokens_seen": 31535568, + "step": 15105 + }, + { + "epoch": 2.4649645158658946, + "grad_norm": 0.0006965374923311174, + "learning_rate": 0.2062140481756104, + "loss": 0.3391, + "num_input_tokens_seen": 31544336, + "step": 15110 + }, + { + "epoch": 2.4657802430867117, + "grad_norm": 0.0006989672547206283, + "learning_rate": 0.20615943186477648, + "loss": 0.3244, + "num_input_tokens_seen": 31553456, + "step": 15115 + }, + { + "epoch": 2.4665959703075293, + "grad_norm": 0.00044165313011035323, + "learning_rate": 0.20610480689345242, + "loss": 0.2747, + "num_input_tokens_seen": 31564816, + "step": 15120 + }, + { + "epoch": 2.4674116975283464, + "grad_norm": 0.0009101385367102921, + "learning_rate": 0.2060501732700621, + "loss": 0.3206, + "num_input_tokens_seen": 31574768, + "step": 15125 + }, + { + "epoch": 2.468227424749164, + "grad_norm": 0.0005127440672367811, + "learning_rate": 0.20599553100303067, + "loss": 0.3168, + "num_input_tokens_seen": 31585808, + "step": 15130 + }, + { + "epoch": 2.469043151969981, + "grad_norm": 0.0004777077410835773, + "learning_rate": 0.20594088010078465, + "loss": 0.3178, + "num_input_tokens_seen": 31595216, + "step": 15135 + }, + { + "epoch": 2.4698588791907987, + "grad_norm": 0.00040159790660254657, + "learning_rate": 0.20588622057175196, + "loss": 0.3371, + "num_input_tokens_seen": 31606384, + "step": 15140 + }, + { + "epoch": 2.470674606411616, + "grad_norm": 0.00026503356639295816, + "learning_rate": 0.20583155242436177, + "loss": 0.3574, + "num_input_tokens_seen": 31616320, + "step": 15145 + }, + { + "epoch": 2.4714903336324334, + "grad_norm": 0.000597055593971163, + "learning_rate": 0.20577687566704453, + "loss": 0.2951, + "num_input_tokens_seen": 31625984, + "step": 15150 + }, + { + "epoch": 2.4723060608532506, + "grad_norm": 0.0005675397696904838, + "learning_rate": 0.20572219030823213, + "loss": 0.3634, + "num_input_tokens_seen": 31634848, + "step": 15155 + }, + { + "epoch": 2.473121788074068, + "grad_norm": 0.0008604584145359695, + "learning_rate": 0.20566749635635775, + "loss": 0.3128, + "num_input_tokens_seen": 31646256, + "step": 15160 + }, + { + "epoch": 2.4739375152948853, + "grad_norm": 0.0008289250545203686, + "learning_rate": 0.20561279381985587, + "loss": 0.3019, + "num_input_tokens_seen": 31656448, + "step": 15165 + }, + { + "epoch": 2.474753242515703, + "grad_norm": 0.00042527433834038675, + "learning_rate": 0.2055580827071623, + "loss": 0.2767, + "num_input_tokens_seen": 31666800, + "step": 15170 + }, + { + "epoch": 2.47556896973652, + "grad_norm": 0.0009093802655115724, + "learning_rate": 0.20550336302671418, + "loss": 0.3386, + "num_input_tokens_seen": 31677056, + "step": 15175 + }, + { + "epoch": 2.4763846969573375, + "grad_norm": 0.0007278951816260815, + "learning_rate": 0.20544863478695, + "loss": 0.3609, + "num_input_tokens_seen": 31688336, + "step": 15180 + }, + { + "epoch": 2.4772004241781547, + "grad_norm": 0.0006014902610331774, + "learning_rate": 0.20539389799630953, + "loss": 0.3368, + "num_input_tokens_seen": 31699296, + "step": 15185 + }, + { + "epoch": 2.4780161513989722, + "grad_norm": 0.0004036260361317545, + "learning_rate": 0.20533915266323388, + "loss": 0.289, + "num_input_tokens_seen": 31710784, + "step": 15190 + }, + { + "epoch": 2.4788318786197894, + "grad_norm": 0.0005632268148474395, + "learning_rate": 0.20528439879616542, + "loss": 0.3122, + "num_input_tokens_seen": 31721616, + "step": 15195 + }, + { + "epoch": 2.479647605840607, + "grad_norm": 0.0005278686294332147, + "learning_rate": 0.20522963640354794, + "loss": 0.3841, + "num_input_tokens_seen": 31730928, + "step": 15200 + }, + { + "epoch": 2.479647605840607, + "eval_loss": 0.3172062039375305, + "eval_runtime": 156.1683, + "eval_samples_per_second": 17.449, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 31730928, + "step": 15200 + }, + { + "epoch": 2.480463333061424, + "grad_norm": 0.0008156888652592897, + "learning_rate": 0.20517486549382644, + "loss": 0.3571, + "num_input_tokens_seen": 31741152, + "step": 15205 + }, + { + "epoch": 2.4812790602822417, + "grad_norm": 0.0003384735609870404, + "learning_rate": 0.20512008607544735, + "loss": 0.3055, + "num_input_tokens_seen": 31750960, + "step": 15210 + }, + { + "epoch": 2.482094787503059, + "grad_norm": 0.0003151812998112291, + "learning_rate": 0.20506529815685826, + "loss": 0.3537, + "num_input_tokens_seen": 31761584, + "step": 15215 + }, + { + "epoch": 2.4829105147238764, + "grad_norm": 0.00032902651582844555, + "learning_rate": 0.2050105017465082, + "loss": 0.3575, + "num_input_tokens_seen": 31772016, + "step": 15220 + }, + { + "epoch": 2.4837262419446935, + "grad_norm": 0.000739592534955591, + "learning_rate": 0.20495569685284754, + "loss": 0.3027, + "num_input_tokens_seen": 31782560, + "step": 15225 + }, + { + "epoch": 2.484541969165511, + "grad_norm": 0.0006022547022439539, + "learning_rate": 0.20490088348432778, + "loss": 0.3186, + "num_input_tokens_seen": 31792752, + "step": 15230 + }, + { + "epoch": 2.4853576963863286, + "grad_norm": 0.0013919300399720669, + "learning_rate": 0.2048460616494018, + "loss": 0.3063, + "num_input_tokens_seen": 31801280, + "step": 15235 + }, + { + "epoch": 2.4861734236071458, + "grad_norm": 0.0003747994778677821, + "learning_rate": 0.2047912313565239, + "loss": 0.3037, + "num_input_tokens_seen": 31812144, + "step": 15240 + }, + { + "epoch": 2.486989150827963, + "grad_norm": 0.00037776559474878013, + "learning_rate": 0.20473639261414958, + "loss": 0.3347, + "num_input_tokens_seen": 31822048, + "step": 15245 + }, + { + "epoch": 2.4878048780487805, + "grad_norm": 0.00030055129900574684, + "learning_rate": 0.2046815454307357, + "loss": 0.2918, + "num_input_tokens_seen": 31832688, + "step": 15250 + }, + { + "epoch": 2.488620605269598, + "grad_norm": 0.0007865098887123168, + "learning_rate": 0.20462668981474028, + "loss": 0.3689, + "num_input_tokens_seen": 31842896, + "step": 15255 + }, + { + "epoch": 2.489436332490415, + "grad_norm": 0.0004554148472379893, + "learning_rate": 0.20457182577462288, + "loss": 0.2604, + "num_input_tokens_seen": 31852768, + "step": 15260 + }, + { + "epoch": 2.4902520597112328, + "grad_norm": 0.0006653736927546561, + "learning_rate": 0.2045169533188441, + "loss": 0.2851, + "num_input_tokens_seen": 31862848, + "step": 15265 + }, + { + "epoch": 2.49106778693205, + "grad_norm": 0.0009075793204829097, + "learning_rate": 0.20446207245586603, + "loss": 0.3417, + "num_input_tokens_seen": 31872800, + "step": 15270 + }, + { + "epoch": 2.4918835141528675, + "grad_norm": 0.0007200983818620443, + "learning_rate": 0.20440718319415196, + "loss": 0.3555, + "num_input_tokens_seen": 31883360, + "step": 15275 + }, + { + "epoch": 2.4926992413736846, + "grad_norm": 0.0005052431370131671, + "learning_rate": 0.20435228554216653, + "loss": 0.3169, + "num_input_tokens_seen": 31893792, + "step": 15280 + }, + { + "epoch": 2.493514968594502, + "grad_norm": 0.0006323256529867649, + "learning_rate": 0.20429737950837565, + "loss": 0.3454, + "num_input_tokens_seen": 31905584, + "step": 15285 + }, + { + "epoch": 2.4943306958153193, + "grad_norm": 0.00035428721457719803, + "learning_rate": 0.20424246510124647, + "loss": 0.3405, + "num_input_tokens_seen": 31916208, + "step": 15290 + }, + { + "epoch": 2.495146423036137, + "grad_norm": 0.0006297047366388142, + "learning_rate": 0.20418754232924755, + "loss": 0.3141, + "num_input_tokens_seen": 31926080, + "step": 15295 + }, + { + "epoch": 2.495962150256954, + "grad_norm": 0.0007781399763189256, + "learning_rate": 0.20413261120084863, + "loss": 0.3648, + "num_input_tokens_seen": 31936576, + "step": 15300 + }, + { + "epoch": 2.4967778774777716, + "grad_norm": 0.0007167700678110123, + "learning_rate": 0.2040776717245208, + "loss": 0.4392, + "num_input_tokens_seen": 31946544, + "step": 15305 + }, + { + "epoch": 2.4975936046985887, + "grad_norm": 0.00038096061325632036, + "learning_rate": 0.2040227239087364, + "loss": 0.2785, + "num_input_tokens_seen": 31956288, + "step": 15310 + }, + { + "epoch": 2.4984093319194063, + "grad_norm": 0.000609835609793663, + "learning_rate": 0.20396776776196904, + "loss": 0.3409, + "num_input_tokens_seen": 31967632, + "step": 15315 + }, + { + "epoch": 2.4992250591402234, + "grad_norm": 0.0007002691272646189, + "learning_rate": 0.20391280329269373, + "loss": 0.3299, + "num_input_tokens_seen": 31977712, + "step": 15320 + }, + { + "epoch": 2.500040786361041, + "grad_norm": 0.0006997073069214821, + "learning_rate": 0.20385783050938663, + "loss": 0.2789, + "num_input_tokens_seen": 31988000, + "step": 15325 + }, + { + "epoch": 2.500856513581858, + "grad_norm": 0.0004246380121912807, + "learning_rate": 0.20380284942052526, + "loss": 0.2703, + "num_input_tokens_seen": 31998304, + "step": 15330 + }, + { + "epoch": 2.5016722408026757, + "grad_norm": 0.0006813960499130189, + "learning_rate": 0.2037478600345884, + "loss": 0.345, + "num_input_tokens_seen": 32009984, + "step": 15335 + }, + { + "epoch": 2.502487968023493, + "grad_norm": 0.0004612632328644395, + "learning_rate": 0.20369286236005604, + "loss": 0.3005, + "num_input_tokens_seen": 32020720, + "step": 15340 + }, + { + "epoch": 2.5033036952443104, + "grad_norm": 0.0006328434683382511, + "learning_rate": 0.20363785640540957, + "loss": 0.3394, + "num_input_tokens_seen": 32031824, + "step": 15345 + }, + { + "epoch": 2.5041194224651275, + "grad_norm": 0.0004935616161674261, + "learning_rate": 0.2035828421791316, + "loss": 0.3326, + "num_input_tokens_seen": 32042816, + "step": 15350 + }, + { + "epoch": 2.504935149685945, + "grad_norm": 0.0004580821259878576, + "learning_rate": 0.20352781968970599, + "loss": 0.3143, + "num_input_tokens_seen": 32052992, + "step": 15355 + }, + { + "epoch": 2.5057508769067622, + "grad_norm": 0.0003406275063753128, + "learning_rate": 0.2034727889456179, + "loss": 0.2816, + "num_input_tokens_seen": 32063216, + "step": 15360 + }, + { + "epoch": 2.50656660412758, + "grad_norm": 0.0012994649587199092, + "learning_rate": 0.2034177499553538, + "loss": 0.3568, + "num_input_tokens_seen": 32075360, + "step": 15365 + }, + { + "epoch": 2.507382331348397, + "grad_norm": 0.0014713505515828729, + "learning_rate": 0.2033627027274014, + "loss": 0.3335, + "num_input_tokens_seen": 32085072, + "step": 15370 + }, + { + "epoch": 2.5081980585692145, + "grad_norm": 0.0005873881746083498, + "learning_rate": 0.20330764727024955, + "loss": 0.3332, + "num_input_tokens_seen": 32095792, + "step": 15375 + }, + { + "epoch": 2.5090137857900316, + "grad_norm": 0.0004256698885001242, + "learning_rate": 0.20325258359238868, + "loss": 0.2649, + "num_input_tokens_seen": 32106880, + "step": 15380 + }, + { + "epoch": 2.5098295130108492, + "grad_norm": 0.00041972266626544297, + "learning_rate": 0.20319751170231018, + "loss": 0.3245, + "num_input_tokens_seen": 32117200, + "step": 15385 + }, + { + "epoch": 2.510645240231667, + "grad_norm": 0.0005575617542490363, + "learning_rate": 0.2031424316085068, + "loss": 0.3248, + "num_input_tokens_seen": 32127072, + "step": 15390 + }, + { + "epoch": 2.511460967452484, + "grad_norm": 0.0004616440273821354, + "learning_rate": 0.20308734331947265, + "loss": 0.2874, + "num_input_tokens_seen": 32136368, + "step": 15395 + }, + { + "epoch": 2.512276694673301, + "grad_norm": 0.0007003787322901189, + "learning_rate": 0.20303224684370305, + "loss": 0.3464, + "num_input_tokens_seen": 32146304, + "step": 15400 + }, + { + "epoch": 2.512276694673301, + "eval_loss": 0.3176092505455017, + "eval_runtime": 155.9832, + "eval_samples_per_second": 17.47, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 32146304, + "step": 15400 + }, + { + "epoch": 2.5130924218941186, + "grad_norm": 0.0003762127598747611, + "learning_rate": 0.20297714218969456, + "loss": 0.2732, + "num_input_tokens_seen": 32154896, + "step": 15405 + }, + { + "epoch": 2.513908149114936, + "grad_norm": 0.0007187857409007847, + "learning_rate": 0.20292202936594497, + "loss": 0.3369, + "num_input_tokens_seen": 32166304, + "step": 15410 + }, + { + "epoch": 2.5147238763357533, + "grad_norm": 0.0002714260190259665, + "learning_rate": 0.2028669083809534, + "loss": 0.2491, + "num_input_tokens_seen": 32176656, + "step": 15415 + }, + { + "epoch": 2.5155396035565705, + "grad_norm": 0.0008097444078885019, + "learning_rate": 0.20281177924322016, + "loss": 0.3384, + "num_input_tokens_seen": 32188800, + "step": 15420 + }, + { + "epoch": 2.516355330777388, + "grad_norm": 0.0025235528592020273, + "learning_rate": 0.2027566419612469, + "loss": 0.3757, + "num_input_tokens_seen": 32200256, + "step": 15425 + }, + { + "epoch": 2.5171710579982056, + "grad_norm": 0.0005851391470059752, + "learning_rate": 0.20270149654353647, + "loss": 0.3235, + "num_input_tokens_seen": 32211104, + "step": 15430 + }, + { + "epoch": 2.5179867852190227, + "grad_norm": 0.0003430467622820288, + "learning_rate": 0.202646342998593, + "loss": 0.334, + "num_input_tokens_seen": 32221104, + "step": 15435 + }, + { + "epoch": 2.51880251243984, + "grad_norm": 0.0007569096051156521, + "learning_rate": 0.20259118133492185, + "loss": 0.2933, + "num_input_tokens_seen": 32231184, + "step": 15440 + }, + { + "epoch": 2.5196182396606575, + "grad_norm": 0.0006272000027820468, + "learning_rate": 0.20253601156102966, + "loss": 0.2829, + "num_input_tokens_seen": 32241536, + "step": 15445 + }, + { + "epoch": 2.520433966881475, + "grad_norm": 0.0007249770569615066, + "learning_rate": 0.20248083368542422, + "loss": 0.3024, + "num_input_tokens_seen": 32252896, + "step": 15450 + }, + { + "epoch": 2.521249694102292, + "grad_norm": 0.0012261526426300406, + "learning_rate": 0.2024256477166147, + "loss": 0.3643, + "num_input_tokens_seen": 32263472, + "step": 15455 + }, + { + "epoch": 2.5220654213231093, + "grad_norm": 0.0009060215088538826, + "learning_rate": 0.2023704536631115, + "loss": 0.3799, + "num_input_tokens_seen": 32273920, + "step": 15460 + }, + { + "epoch": 2.522881148543927, + "grad_norm": 0.00042803381802514195, + "learning_rate": 0.20231525153342625, + "loss": 0.32, + "num_input_tokens_seen": 32282784, + "step": 15465 + }, + { + "epoch": 2.5236968757647444, + "grad_norm": 0.0006604401860386133, + "learning_rate": 0.20226004133607173, + "loss": 0.308, + "num_input_tokens_seen": 32293136, + "step": 15470 + }, + { + "epoch": 2.5245126029855616, + "grad_norm": 0.0009072068496607244, + "learning_rate": 0.20220482307956214, + "loss": 0.3862, + "num_input_tokens_seen": 32303824, + "step": 15475 + }, + { + "epoch": 2.525328330206379, + "grad_norm": 0.000454260065453127, + "learning_rate": 0.20214959677241276, + "loss": 0.2424, + "num_input_tokens_seen": 32314352, + "step": 15480 + }, + { + "epoch": 2.5261440574271963, + "grad_norm": 0.0003819091652985662, + "learning_rate": 0.20209436242314022, + "loss": 0.3511, + "num_input_tokens_seen": 32324912, + "step": 15485 + }, + { + "epoch": 2.526959784648014, + "grad_norm": 0.0007922534714452922, + "learning_rate": 0.2020391200402623, + "loss": 0.3288, + "num_input_tokens_seen": 32335200, + "step": 15490 + }, + { + "epoch": 2.527775511868831, + "grad_norm": 0.0007731466321274638, + "learning_rate": 0.2019838696322981, + "loss": 0.3358, + "num_input_tokens_seen": 32346208, + "step": 15495 + }, + { + "epoch": 2.5285912390896486, + "grad_norm": 0.0006583599024452269, + "learning_rate": 0.20192861120776798, + "loss": 0.3454, + "num_input_tokens_seen": 32356368, + "step": 15500 + }, + { + "epoch": 2.5294069663104657, + "grad_norm": 0.001643983181566, + "learning_rate": 0.20187334477519345, + "loss": 0.3583, + "num_input_tokens_seen": 32367376, + "step": 15505 + }, + { + "epoch": 2.5302226935312833, + "grad_norm": 0.0004022673238068819, + "learning_rate": 0.20181807034309726, + "loss": 0.2877, + "num_input_tokens_seen": 32378048, + "step": 15510 + }, + { + "epoch": 2.5310384207521004, + "grad_norm": 0.0004995728959329426, + "learning_rate": 0.2017627879200034, + "loss": 0.2876, + "num_input_tokens_seen": 32388208, + "step": 15515 + }, + { + "epoch": 2.531854147972918, + "grad_norm": 0.0006757296505384147, + "learning_rate": 0.2017074975144372, + "loss": 0.3129, + "num_input_tokens_seen": 32398720, + "step": 15520 + }, + { + "epoch": 2.532669875193735, + "grad_norm": 0.0007197118829935789, + "learning_rate": 0.20165219913492508, + "loss": 0.3038, + "num_input_tokens_seen": 32409328, + "step": 15525 + }, + { + "epoch": 2.5334856024145527, + "grad_norm": 0.0007384954951703548, + "learning_rate": 0.20159689278999468, + "loss": 0.2773, + "num_input_tokens_seen": 32420800, + "step": 15530 + }, + { + "epoch": 2.53430132963537, + "grad_norm": 0.000875085243023932, + "learning_rate": 0.20154157848817508, + "loss": 0.3556, + "num_input_tokens_seen": 32430784, + "step": 15535 + }, + { + "epoch": 2.5351170568561874, + "grad_norm": 0.00036465993616729975, + "learning_rate": 0.20148625623799632, + "loss": 0.2881, + "num_input_tokens_seen": 32441296, + "step": 15540 + }, + { + "epoch": 2.5359327840770045, + "grad_norm": 0.0007418259629048407, + "learning_rate": 0.20143092604798984, + "loss": 0.2805, + "num_input_tokens_seen": 32451552, + "step": 15545 + }, + { + "epoch": 2.536748511297822, + "grad_norm": 0.00036374927731230855, + "learning_rate": 0.2013755879266883, + "loss": 0.3479, + "num_input_tokens_seen": 32463136, + "step": 15550 + }, + { + "epoch": 2.537564238518639, + "grad_norm": 0.00034119535121135414, + "learning_rate": 0.20132024188262543, + "loss": 0.3513, + "num_input_tokens_seen": 32472416, + "step": 15555 + }, + { + "epoch": 2.538379965739457, + "grad_norm": 0.0003644776879809797, + "learning_rate": 0.2012648879243363, + "loss": 0.3329, + "num_input_tokens_seen": 32481344, + "step": 15560 + }, + { + "epoch": 2.539195692960274, + "grad_norm": 0.0007194355712272227, + "learning_rate": 0.20120952606035725, + "loss": 0.3337, + "num_input_tokens_seen": 32491648, + "step": 15565 + }, + { + "epoch": 2.5400114201810915, + "grad_norm": 0.00045443809358403087, + "learning_rate": 0.20115415629922576, + "loss": 0.315, + "num_input_tokens_seen": 32503072, + "step": 15570 + }, + { + "epoch": 2.5408271474019086, + "grad_norm": 0.0005370359867811203, + "learning_rate": 0.20109877864948048, + "loss": 0.3295, + "num_input_tokens_seen": 32513504, + "step": 15575 + }, + { + "epoch": 2.541642874622726, + "grad_norm": 0.0003316134389024228, + "learning_rate": 0.20104339311966138, + "loss": 0.3009, + "num_input_tokens_seen": 32525040, + "step": 15580 + }, + { + "epoch": 2.5424586018435438, + "grad_norm": 0.0004614816280081868, + "learning_rate": 0.2009879997183097, + "loss": 0.3539, + "num_input_tokens_seen": 32536208, + "step": 15585 + }, + { + "epoch": 2.543274329064361, + "grad_norm": 0.0005758440238423645, + "learning_rate": 0.20093259845396763, + "loss": 0.2676, + "num_input_tokens_seen": 32546080, + "step": 15590 + }, + { + "epoch": 2.544090056285178, + "grad_norm": 0.0019345718901604414, + "learning_rate": 0.20087718933517884, + "loss": 0.361, + "num_input_tokens_seen": 32555104, + "step": 15595 + }, + { + "epoch": 2.5449057835059956, + "grad_norm": 0.0005851486348547041, + "learning_rate": 0.20082177237048807, + "loss": 0.3327, + "num_input_tokens_seen": 32566096, + "step": 15600 + }, + { + "epoch": 2.5449057835059956, + "eval_loss": 0.3167375922203064, + "eval_runtime": 155.9631, + "eval_samples_per_second": 17.472, + "eval_steps_per_second": 8.739, + "num_input_tokens_seen": 32566096, + "step": 15600 + }, + { + "epoch": 2.545721510726813, + "grad_norm": 0.0007156752981245518, + "learning_rate": 0.20076634756844133, + "loss": 0.3428, + "num_input_tokens_seen": 32575488, + "step": 15605 + }, + { + "epoch": 2.5465372379476303, + "grad_norm": 0.000585820060223341, + "learning_rate": 0.20071091493758586, + "loss": 0.3037, + "num_input_tokens_seen": 32586816, + "step": 15610 + }, + { + "epoch": 2.5473529651684474, + "grad_norm": 0.0012071396922692657, + "learning_rate": 0.20065547448647003, + "loss": 0.3874, + "num_input_tokens_seen": 32597904, + "step": 15615 + }, + { + "epoch": 2.548168692389265, + "grad_norm": 0.00038197345566004515, + "learning_rate": 0.20060002622364348, + "loss": 0.3043, + "num_input_tokens_seen": 32607696, + "step": 15620 + }, + { + "epoch": 2.5489844196100826, + "grad_norm": 0.000546945258975029, + "learning_rate": 0.20054457015765695, + "loss": 0.2933, + "num_input_tokens_seen": 32617664, + "step": 15625 + }, + { + "epoch": 2.5498001468308997, + "grad_norm": 0.0006424199091270566, + "learning_rate": 0.20048910629706254, + "loss": 0.3675, + "num_input_tokens_seen": 32626624, + "step": 15630 + }, + { + "epoch": 2.550615874051717, + "grad_norm": 0.0003610679123084992, + "learning_rate": 0.20043363465041347, + "loss": 0.3493, + "num_input_tokens_seen": 32635344, + "step": 15635 + }, + { + "epoch": 2.5514316012725344, + "grad_norm": 0.00041753824916668236, + "learning_rate": 0.2003781552262641, + "loss": 0.2935, + "num_input_tokens_seen": 32646112, + "step": 15640 + }, + { + "epoch": 2.552247328493352, + "grad_norm": 0.00024034189118538052, + "learning_rate": 0.20032266803317014, + "loss": 0.3279, + "num_input_tokens_seen": 32656528, + "step": 15645 + }, + { + "epoch": 2.553063055714169, + "grad_norm": 0.0005767586990259588, + "learning_rate": 0.2002671730796884, + "loss": 0.3607, + "num_input_tokens_seen": 32666688, + "step": 15650 + }, + { + "epoch": 2.5538787829349863, + "grad_norm": 0.00025620762608014047, + "learning_rate": 0.20021167037437684, + "loss": 0.3182, + "num_input_tokens_seen": 32677632, + "step": 15655 + }, + { + "epoch": 2.554694510155804, + "grad_norm": 0.0010504773817956448, + "learning_rate": 0.20015615992579472, + "loss": 0.3398, + "num_input_tokens_seen": 32688016, + "step": 15660 + }, + { + "epoch": 2.5555102373766214, + "grad_norm": 0.0005306322127580643, + "learning_rate": 0.20010064174250244, + "loss": 0.3156, + "num_input_tokens_seen": 32699632, + "step": 15665 + }, + { + "epoch": 2.5563259645974385, + "grad_norm": 0.0002472430933266878, + "learning_rate": 0.2000451158330616, + "loss": 0.3274, + "num_input_tokens_seen": 32709728, + "step": 15670 + }, + { + "epoch": 2.557141691818256, + "grad_norm": 0.00023840487119741738, + "learning_rate": 0.199989582206035, + "loss": 0.328, + "num_input_tokens_seen": 32719408, + "step": 15675 + }, + { + "epoch": 2.5579574190390733, + "grad_norm": 0.00026774388970807195, + "learning_rate": 0.1999340408699866, + "loss": 0.3832, + "num_input_tokens_seen": 32729776, + "step": 15680 + }, + { + "epoch": 2.558773146259891, + "grad_norm": 0.0003843979793600738, + "learning_rate": 0.19987849183348155, + "loss": 0.3163, + "num_input_tokens_seen": 32739136, + "step": 15685 + }, + { + "epoch": 2.559588873480708, + "grad_norm": 0.0002537597611080855, + "learning_rate": 0.19982293510508628, + "loss": 0.2897, + "num_input_tokens_seen": 32750304, + "step": 15690 + }, + { + "epoch": 2.5604046007015255, + "grad_norm": 0.00038675282848998904, + "learning_rate": 0.19976737069336833, + "loss": 0.3495, + "num_input_tokens_seen": 32760592, + "step": 15695 + }, + { + "epoch": 2.5612203279223427, + "grad_norm": 0.00036670107510872185, + "learning_rate": 0.1997117986068964, + "loss": 0.3314, + "num_input_tokens_seen": 32770800, + "step": 15700 + }, + { + "epoch": 2.5620360551431602, + "grad_norm": 0.0009729620651341975, + "learning_rate": 0.19965621885424037, + "loss": 0.3549, + "num_input_tokens_seen": 32781408, + "step": 15705 + }, + { + "epoch": 2.5628517823639774, + "grad_norm": 0.0007895493181422353, + "learning_rate": 0.19960063144397142, + "loss": 0.3108, + "num_input_tokens_seen": 32793712, + "step": 15710 + }, + { + "epoch": 2.563667509584795, + "grad_norm": 0.00043831608491018414, + "learning_rate": 0.19954503638466176, + "loss": 0.3151, + "num_input_tokens_seen": 32804608, + "step": 15715 + }, + { + "epoch": 2.564483236805612, + "grad_norm": 0.0006797827081754804, + "learning_rate": 0.1994894336848848, + "loss": 0.3011, + "num_input_tokens_seen": 32814416, + "step": 15720 + }, + { + "epoch": 2.5652989640264297, + "grad_norm": 0.0007318655261769891, + "learning_rate": 0.1994338233532153, + "loss": 0.3274, + "num_input_tokens_seen": 32824672, + "step": 15725 + }, + { + "epoch": 2.566114691247247, + "grad_norm": 0.0007231347844935954, + "learning_rate": 0.19937820539822904, + "loss": 0.3168, + "num_input_tokens_seen": 32834528, + "step": 15730 + }, + { + "epoch": 2.5669304184680644, + "grad_norm": 0.0005050647887401283, + "learning_rate": 0.199322579828503, + "loss": 0.2483, + "num_input_tokens_seen": 32844496, + "step": 15735 + }, + { + "epoch": 2.5677461456888815, + "grad_norm": 0.0003783351567108184, + "learning_rate": 0.19926694665261527, + "loss": 0.2928, + "num_input_tokens_seen": 32852992, + "step": 15740 + }, + { + "epoch": 2.568561872909699, + "grad_norm": 0.0009064487530849874, + "learning_rate": 0.19921130587914526, + "loss": 0.3533, + "num_input_tokens_seen": 32862832, + "step": 15745 + }, + { + "epoch": 2.569377600130516, + "grad_norm": 0.0004894405137747526, + "learning_rate": 0.19915565751667344, + "loss": 0.3128, + "num_input_tokens_seen": 32873168, + "step": 15750 + }, + { + "epoch": 2.5701933273513338, + "grad_norm": 0.0014366236282512546, + "learning_rate": 0.19910000157378152, + "loss": 0.2797, + "num_input_tokens_seen": 32882624, + "step": 15755 + }, + { + "epoch": 2.5710090545721513, + "grad_norm": 0.00036869445466436446, + "learning_rate": 0.1990443380590523, + "loss": 0.3247, + "num_input_tokens_seen": 32893328, + "step": 15760 + }, + { + "epoch": 2.5718247817929685, + "grad_norm": 0.000746092526242137, + "learning_rate": 0.19898866698106984, + "loss": 0.3234, + "num_input_tokens_seen": 32904288, + "step": 15765 + }, + { + "epoch": 2.5726405090137856, + "grad_norm": 0.0006701387465000153, + "learning_rate": 0.19893298834841933, + "loss": 0.3477, + "num_input_tokens_seen": 32914384, + "step": 15770 + }, + { + "epoch": 2.573456236234603, + "grad_norm": 0.000473203748697415, + "learning_rate": 0.19887730216968705, + "loss": 0.2897, + "num_input_tokens_seen": 32924400, + "step": 15775 + }, + { + "epoch": 2.5742719634554208, + "grad_norm": 0.0003276905626989901, + "learning_rate": 0.19882160845346053, + "loss": 0.3721, + "num_input_tokens_seen": 32935856, + "step": 15780 + }, + { + "epoch": 2.575087690676238, + "grad_norm": 0.000707280938513577, + "learning_rate": 0.1987659072083285, + "loss": 0.3292, + "num_input_tokens_seen": 32948160, + "step": 15785 + }, + { + "epoch": 2.575903417897055, + "grad_norm": 0.00028251539333723485, + "learning_rate": 0.1987101984428807, + "loss": 0.3094, + "num_input_tokens_seen": 32959840, + "step": 15790 + }, + { + "epoch": 2.5767191451178726, + "grad_norm": 0.0005694057908840477, + "learning_rate": 0.19865448216570822, + "loss": 0.2966, + "num_input_tokens_seen": 32969856, + "step": 15795 + }, + { + "epoch": 2.57753487233869, + "grad_norm": 0.00053737114649266, + "learning_rate": 0.19859875838540317, + "loss": 0.2851, + "num_input_tokens_seen": 32981664, + "step": 15800 + }, + { + "epoch": 2.57753487233869, + "eval_loss": 0.31960317492485046, + "eval_runtime": 155.992, + "eval_samples_per_second": 17.469, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 32981664, + "step": 15800 + }, + { + "epoch": 2.5783505995595073, + "grad_norm": 0.0009017531992867589, + "learning_rate": 0.1985430271105588, + "loss": 0.3744, + "num_input_tokens_seen": 32991168, + "step": 15805 + }, + { + "epoch": 2.5791663267803244, + "grad_norm": 0.00033063319278880954, + "learning_rate": 0.19848728834976961, + "loss": 0.3344, + "num_input_tokens_seen": 33001728, + "step": 15810 + }, + { + "epoch": 2.579982054001142, + "grad_norm": 0.00044259970309212804, + "learning_rate": 0.19843154211163128, + "loss": 0.3084, + "num_input_tokens_seen": 33011744, + "step": 15815 + }, + { + "epoch": 2.5807977812219596, + "grad_norm": 0.00037573586450889707, + "learning_rate": 0.1983757884047405, + "loss": 0.2547, + "num_input_tokens_seen": 33022016, + "step": 15820 + }, + { + "epoch": 2.5816135084427767, + "grad_norm": 0.0007027426618151367, + "learning_rate": 0.1983200272376952, + "loss": 0.3085, + "num_input_tokens_seen": 33032480, + "step": 15825 + }, + { + "epoch": 2.582429235663594, + "grad_norm": 0.0007240496925078332, + "learning_rate": 0.1982642586190945, + "loss": 0.3231, + "num_input_tokens_seen": 33042336, + "step": 15830 + }, + { + "epoch": 2.5832449628844114, + "grad_norm": 0.0005786857800558209, + "learning_rate": 0.1982084825575386, + "loss": 0.3174, + "num_input_tokens_seen": 33053264, + "step": 15835 + }, + { + "epoch": 2.584060690105229, + "grad_norm": 0.0003545189101714641, + "learning_rate": 0.19815269906162883, + "loss": 0.3071, + "num_input_tokens_seen": 33064032, + "step": 15840 + }, + { + "epoch": 2.584876417326046, + "grad_norm": 0.0006733209011144936, + "learning_rate": 0.19809690813996775, + "loss": 0.304, + "num_input_tokens_seen": 33074544, + "step": 15845 + }, + { + "epoch": 2.5856921445468637, + "grad_norm": 0.00048395097837783396, + "learning_rate": 0.19804110980115905, + "loss": 0.2857, + "num_input_tokens_seen": 33085008, + "step": 15850 + }, + { + "epoch": 2.586507871767681, + "grad_norm": 0.0004063860105816275, + "learning_rate": 0.19798530405380746, + "loss": 0.2461, + "num_input_tokens_seen": 33095200, + "step": 15855 + }, + { + "epoch": 2.5873235989884984, + "grad_norm": 0.0003768267633859068, + "learning_rate": 0.19792949090651893, + "loss": 0.3669, + "num_input_tokens_seen": 33105120, + "step": 15860 + }, + { + "epoch": 2.5881393262093155, + "grad_norm": 0.0009110927931033075, + "learning_rate": 0.19787367036790066, + "loss": 0.3215, + "num_input_tokens_seen": 33115440, + "step": 15865 + }, + { + "epoch": 2.588955053430133, + "grad_norm": 0.0004353343101684004, + "learning_rate": 0.19781784244656075, + "loss": 0.264, + "num_input_tokens_seen": 33125728, + "step": 15870 + }, + { + "epoch": 2.5897707806509502, + "grad_norm": 0.0003783829160965979, + "learning_rate": 0.19776200715110864, + "loss": 0.2536, + "num_input_tokens_seen": 33136224, + "step": 15875 + }, + { + "epoch": 2.590586507871768, + "grad_norm": 0.0010649573523551226, + "learning_rate": 0.1977061644901548, + "loss": 0.3178, + "num_input_tokens_seen": 33147344, + "step": 15880 + }, + { + "epoch": 2.591402235092585, + "grad_norm": 0.0008438296499662101, + "learning_rate": 0.1976503144723109, + "loss": 0.3119, + "num_input_tokens_seen": 33156096, + "step": 15885 + }, + { + "epoch": 2.5922179623134025, + "grad_norm": 0.00047708326019346714, + "learning_rate": 0.19759445710618967, + "loss": 0.2952, + "num_input_tokens_seen": 33165744, + "step": 15890 + }, + { + "epoch": 2.5930336895342196, + "grad_norm": 0.0006298149819485843, + "learning_rate": 0.19753859240040508, + "loss": 0.331, + "num_input_tokens_seen": 33176192, + "step": 15895 + }, + { + "epoch": 2.593849416755037, + "grad_norm": 0.0006209854036569595, + "learning_rate": 0.1974827203635721, + "loss": 0.3497, + "num_input_tokens_seen": 33186432, + "step": 15900 + }, + { + "epoch": 2.5946651439758543, + "grad_norm": 0.0004700461868196726, + "learning_rate": 0.19742684100430694, + "loss": 0.3437, + "num_input_tokens_seen": 33197696, + "step": 15905 + }, + { + "epoch": 2.595480871196672, + "grad_norm": 0.0003588953404687345, + "learning_rate": 0.19737095433122692, + "loss": 0.3299, + "num_input_tokens_seen": 33206864, + "step": 15910 + }, + { + "epoch": 2.596296598417489, + "grad_norm": 0.0005997338448651135, + "learning_rate": 0.19731506035295046, + "loss": 0.3053, + "num_input_tokens_seen": 33218080, + "step": 15915 + }, + { + "epoch": 2.5971123256383066, + "grad_norm": 0.0008736556628718972, + "learning_rate": 0.19725915907809702, + "loss": 0.3554, + "num_input_tokens_seen": 33228848, + "step": 15920 + }, + { + "epoch": 2.5979280528591238, + "grad_norm": 0.00034578837221488357, + "learning_rate": 0.1972032505152874, + "loss": 0.3384, + "num_input_tokens_seen": 33239872, + "step": 15925 + }, + { + "epoch": 2.5987437800799413, + "grad_norm": 0.0005071292398497462, + "learning_rate": 0.19714733467314338, + "loss": 0.2856, + "num_input_tokens_seen": 33250592, + "step": 15930 + }, + { + "epoch": 2.5995595073007585, + "grad_norm": 0.0003130122786387801, + "learning_rate": 0.19709141156028784, + "loss": 0.3171, + "num_input_tokens_seen": 33261488, + "step": 15935 + }, + { + "epoch": 2.600375234521576, + "grad_norm": 0.00041566966683603823, + "learning_rate": 0.1970354811853448, + "loss": 0.337, + "num_input_tokens_seen": 33273872, + "step": 15940 + }, + { + "epoch": 2.601190961742393, + "grad_norm": 0.0003304651181679219, + "learning_rate": 0.19697954355693953, + "loss": 0.333, + "num_input_tokens_seen": 33284208, + "step": 15945 + }, + { + "epoch": 2.6020066889632107, + "grad_norm": 0.0010207765735685825, + "learning_rate": 0.19692359868369827, + "loss": 0.2817, + "num_input_tokens_seen": 33295584, + "step": 15950 + }, + { + "epoch": 2.6028224161840283, + "grad_norm": 0.0007177323568612337, + "learning_rate": 0.1968676465742484, + "loss": 0.268, + "num_input_tokens_seen": 33305872, + "step": 15955 + }, + { + "epoch": 2.6036381434048455, + "grad_norm": 0.0009596054442226887, + "learning_rate": 0.19681168723721845, + "loss": 0.301, + "num_input_tokens_seen": 33317712, + "step": 15960 + }, + { + "epoch": 2.6044538706256626, + "grad_norm": 0.0003979514876846224, + "learning_rate": 0.19675572068123803, + "loss": 0.4232, + "num_input_tokens_seen": 33328944, + "step": 15965 + }, + { + "epoch": 2.60526959784648, + "grad_norm": 0.0008553373045288026, + "learning_rate": 0.19669974691493794, + "loss": 0.3421, + "num_input_tokens_seen": 33340816, + "step": 15970 + }, + { + "epoch": 2.6060853250672977, + "grad_norm": 0.0008446500869467854, + "learning_rate": 0.19664376594695002, + "loss": 0.3249, + "num_input_tokens_seen": 33351152, + "step": 15975 + }, + { + "epoch": 2.606901052288115, + "grad_norm": 0.0005272452835924923, + "learning_rate": 0.19658777778590722, + "loss": 0.384, + "num_input_tokens_seen": 33361680, + "step": 15980 + }, + { + "epoch": 2.607716779508932, + "grad_norm": 0.001073053223080933, + "learning_rate": 0.19653178244044364, + "loss": 0.2933, + "num_input_tokens_seen": 33372992, + "step": 15985 + }, + { + "epoch": 2.6085325067297496, + "grad_norm": 0.0006877362611703575, + "learning_rate": 0.19647577991919443, + "loss": 0.3346, + "num_input_tokens_seen": 33382656, + "step": 15990 + }, + { + "epoch": 2.609348233950567, + "grad_norm": 0.0004776517453137785, + "learning_rate": 0.1964197702307959, + "loss": 0.3555, + "num_input_tokens_seen": 33393824, + "step": 15995 + }, + { + "epoch": 2.6101639611713843, + "grad_norm": 0.0006880429573357105, + "learning_rate": 0.19636375338388545, + "loss": 0.2903, + "num_input_tokens_seen": 33403328, + "step": 16000 + }, + { + "epoch": 2.6101639611713843, + "eval_loss": 0.31848761439323425, + "eval_runtime": 155.7505, + "eval_samples_per_second": 17.496, + "eval_steps_per_second": 8.751, + "num_input_tokens_seen": 33403328, + "step": 16000 + }, + { + "epoch": 2.6109796883922014, + "grad_norm": 0.0004531819140538573, + "learning_rate": 0.1963077293871016, + "loss": 0.3281, + "num_input_tokens_seen": 33413568, + "step": 16005 + }, + { + "epoch": 2.611795415613019, + "grad_norm": 0.00032918102806434035, + "learning_rate": 0.19625169824908395, + "loss": 0.3211, + "num_input_tokens_seen": 33424032, + "step": 16010 + }, + { + "epoch": 2.6126111428338366, + "grad_norm": 0.0008967044996097684, + "learning_rate": 0.19619565997847319, + "loss": 0.3735, + "num_input_tokens_seen": 33435056, + "step": 16015 + }, + { + "epoch": 2.6134268700546537, + "grad_norm": 0.000516149855684489, + "learning_rate": 0.19613961458391113, + "loss": 0.3218, + "num_input_tokens_seen": 33447312, + "step": 16020 + }, + { + "epoch": 2.614242597275471, + "grad_norm": 0.0008115245727822185, + "learning_rate": 0.19608356207404065, + "loss": 0.2892, + "num_input_tokens_seen": 33458480, + "step": 16025 + }, + { + "epoch": 2.6150583244962884, + "grad_norm": 0.0008920795517042279, + "learning_rate": 0.1960275024575058, + "loss": 0.3062, + "num_input_tokens_seen": 33468512, + "step": 16030 + }, + { + "epoch": 2.615874051717106, + "grad_norm": 0.000795153493527323, + "learning_rate": 0.19597143574295164, + "loss": 0.3839, + "num_input_tokens_seen": 33479328, + "step": 16035 + }, + { + "epoch": 2.616689778937923, + "grad_norm": 0.0007488247938454151, + "learning_rate": 0.1959153619390244, + "loss": 0.3358, + "num_input_tokens_seen": 33490416, + "step": 16040 + }, + { + "epoch": 2.6175055061587407, + "grad_norm": 0.000566784874536097, + "learning_rate": 0.1958592810543713, + "loss": 0.3027, + "num_input_tokens_seen": 33499504, + "step": 16045 + }, + { + "epoch": 2.618321233379558, + "grad_norm": 0.0003691113961394876, + "learning_rate": 0.19580319309764077, + "loss": 0.3076, + "num_input_tokens_seen": 33509056, + "step": 16050 + }, + { + "epoch": 2.6191369606003754, + "grad_norm": 0.0005269371322356164, + "learning_rate": 0.1957470980774823, + "loss": 0.292, + "num_input_tokens_seen": 33519584, + "step": 16055 + }, + { + "epoch": 2.6199526878211925, + "grad_norm": 0.0003648402926046401, + "learning_rate": 0.19569099600254639, + "loss": 0.2921, + "num_input_tokens_seen": 33528752, + "step": 16060 + }, + { + "epoch": 2.62076841504201, + "grad_norm": 0.0006452787201851606, + "learning_rate": 0.1956348868814847, + "loss": 0.2926, + "num_input_tokens_seen": 33538528, + "step": 16065 + }, + { + "epoch": 2.621584142262827, + "grad_norm": 0.0005617998540401459, + "learning_rate": 0.19557877072295, + "loss": 0.2887, + "num_input_tokens_seen": 33549104, + "step": 16070 + }, + { + "epoch": 2.622399869483645, + "grad_norm": 0.0005032222834415734, + "learning_rate": 0.19552264753559603, + "loss": 0.2536, + "num_input_tokens_seen": 33559680, + "step": 16075 + }, + { + "epoch": 2.623215596704462, + "grad_norm": 0.0008669317467138171, + "learning_rate": 0.19546651732807774, + "loss": 0.3974, + "num_input_tokens_seen": 33570480, + "step": 16080 + }, + { + "epoch": 2.6240313239252795, + "grad_norm": 0.0011244193883612752, + "learning_rate": 0.19541038010905112, + "loss": 0.3346, + "num_input_tokens_seen": 33580912, + "step": 16085 + }, + { + "epoch": 2.6248470511460966, + "grad_norm": 0.0002992569061461836, + "learning_rate": 0.19535423588717324, + "loss": 0.3293, + "num_input_tokens_seen": 33592112, + "step": 16090 + }, + { + "epoch": 2.625662778366914, + "grad_norm": 0.00046901722089387476, + "learning_rate": 0.19529808467110224, + "loss": 0.3039, + "num_input_tokens_seen": 33602320, + "step": 16095 + }, + { + "epoch": 2.6264785055877313, + "grad_norm": 0.0018824241124093533, + "learning_rate": 0.19524192646949734, + "loss": 0.4091, + "num_input_tokens_seen": 33613648, + "step": 16100 + }, + { + "epoch": 2.627294232808549, + "grad_norm": 0.0003343445423524827, + "learning_rate": 0.19518576129101878, + "loss": 0.3729, + "num_input_tokens_seen": 33623328, + "step": 16105 + }, + { + "epoch": 2.628109960029366, + "grad_norm": 0.0009196455939672887, + "learning_rate": 0.19512958914432804, + "loss": 0.2874, + "num_input_tokens_seen": 33632848, + "step": 16110 + }, + { + "epoch": 2.6289256872501836, + "grad_norm": 0.00043126314994879067, + "learning_rate": 0.1950734100380875, + "loss": 0.2771, + "num_input_tokens_seen": 33641904, + "step": 16115 + }, + { + "epoch": 2.6297414144710007, + "grad_norm": 0.00033471573260612786, + "learning_rate": 0.19501722398096066, + "loss": 0.3272, + "num_input_tokens_seen": 33653152, + "step": 16120 + }, + { + "epoch": 2.6305571416918183, + "grad_norm": 0.0005158674903213978, + "learning_rate": 0.1949610309816122, + "loss": 0.3607, + "num_input_tokens_seen": 33663984, + "step": 16125 + }, + { + "epoch": 2.631372868912636, + "grad_norm": 0.00026015847106464207, + "learning_rate": 0.1949048310487078, + "loss": 0.3108, + "num_input_tokens_seen": 33675664, + "step": 16130 + }, + { + "epoch": 2.632188596133453, + "grad_norm": 0.000249437871389091, + "learning_rate": 0.19484862419091406, + "loss": 0.3255, + "num_input_tokens_seen": 33686080, + "step": 16135 + }, + { + "epoch": 2.63300432335427, + "grad_norm": 0.000198497116798535, + "learning_rate": 0.19479241041689893, + "loss": 0.3163, + "num_input_tokens_seen": 33695568, + "step": 16140 + }, + { + "epoch": 2.6338200505750877, + "grad_norm": 0.00034798242268152535, + "learning_rate": 0.19473618973533116, + "loss": 0.325, + "num_input_tokens_seen": 33708096, + "step": 16145 + }, + { + "epoch": 2.6346357777959053, + "grad_norm": 0.0006481058662757277, + "learning_rate": 0.19467996215488076, + "loss": 0.3285, + "num_input_tokens_seen": 33718976, + "step": 16150 + }, + { + "epoch": 2.6354515050167224, + "grad_norm": 0.0004409108078107238, + "learning_rate": 0.1946237276842187, + "loss": 0.3648, + "num_input_tokens_seen": 33730048, + "step": 16155 + }, + { + "epoch": 2.6362672322375396, + "grad_norm": 0.0007816875004209578, + "learning_rate": 0.19456748633201712, + "loss": 0.3509, + "num_input_tokens_seen": 33740688, + "step": 16160 + }, + { + "epoch": 2.637082959458357, + "grad_norm": 0.001932273036800325, + "learning_rate": 0.194511238106949, + "loss": 0.321, + "num_input_tokens_seen": 33750912, + "step": 16165 + }, + { + "epoch": 2.6378986866791747, + "grad_norm": 0.0008084149449132383, + "learning_rate": 0.19445498301768863, + "loss": 0.3088, + "num_input_tokens_seen": 33761520, + "step": 16170 + }, + { + "epoch": 2.638714413899992, + "grad_norm": 0.0003260998928453773, + "learning_rate": 0.19439872107291126, + "loss": 0.2923, + "num_input_tokens_seen": 33772096, + "step": 16175 + }, + { + "epoch": 2.639530141120809, + "grad_norm": 0.0005583719466812909, + "learning_rate": 0.1943424522812931, + "loss": 0.2719, + "num_input_tokens_seen": 33783264, + "step": 16180 + }, + { + "epoch": 2.6403458683416265, + "grad_norm": 0.000695676717441529, + "learning_rate": 0.19428617665151157, + "loss": 0.3557, + "num_input_tokens_seen": 33793840, + "step": 16185 + }, + { + "epoch": 2.641161595562444, + "grad_norm": 0.0005025587161071599, + "learning_rate": 0.19422989419224507, + "loss": 0.2847, + "num_input_tokens_seen": 33804768, + "step": 16190 + }, + { + "epoch": 2.6419773227832613, + "grad_norm": 0.0006864862516522408, + "learning_rate": 0.19417360491217303, + "loss": 0.2967, + "num_input_tokens_seen": 33815648, + "step": 16195 + }, + { + "epoch": 2.6427930500040784, + "grad_norm": 0.0008754546288400888, + "learning_rate": 0.19411730881997605, + "loss": 0.3954, + "num_input_tokens_seen": 33827808, + "step": 16200 + }, + { + "epoch": 2.6427930500040784, + "eval_loss": 0.3186524510383606, + "eval_runtime": 156.1635, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 33827808, + "step": 16200 + }, + { + "epoch": 2.643608777224896, + "grad_norm": 0.0014819735661149025, + "learning_rate": 0.1940610059243356, + "loss": 0.2982, + "num_input_tokens_seen": 33838544, + "step": 16205 + }, + { + "epoch": 2.6444245044457135, + "grad_norm": 0.00038299657171592116, + "learning_rate": 0.19400469623393435, + "loss": 0.2416, + "num_input_tokens_seen": 33849280, + "step": 16210 + }, + { + "epoch": 2.6452402316665307, + "grad_norm": 0.0004270474018994719, + "learning_rate": 0.1939483797574559, + "loss": 0.2758, + "num_input_tokens_seen": 33859808, + "step": 16215 + }, + { + "epoch": 2.6460559588873482, + "grad_norm": 0.001112726517021656, + "learning_rate": 0.19389205650358504, + "loss": 0.3964, + "num_input_tokens_seen": 33870112, + "step": 16220 + }, + { + "epoch": 2.6468716861081654, + "grad_norm": 0.0004114792391192168, + "learning_rate": 0.19383572648100747, + "loss": 0.3126, + "num_input_tokens_seen": 33880832, + "step": 16225 + }, + { + "epoch": 2.647687413328983, + "grad_norm": 0.00030409847386181355, + "learning_rate": 0.19377938969841, + "loss": 0.3003, + "num_input_tokens_seen": 33890960, + "step": 16230 + }, + { + "epoch": 2.6485031405498, + "grad_norm": 0.00034073847928084433, + "learning_rate": 0.1937230461644805, + "loss": 0.3573, + "num_input_tokens_seen": 33900816, + "step": 16235 + }, + { + "epoch": 2.6493188677706176, + "grad_norm": 0.00042230356484651566, + "learning_rate": 0.19366669588790777, + "loss": 0.3479, + "num_input_tokens_seen": 33910464, + "step": 16240 + }, + { + "epoch": 2.6501345949914348, + "grad_norm": 0.00030073325615376234, + "learning_rate": 0.19361033887738185, + "loss": 0.3357, + "num_input_tokens_seen": 33921520, + "step": 16245 + }, + { + "epoch": 2.6509503222122524, + "grad_norm": 0.0008086194284260273, + "learning_rate": 0.19355397514159361, + "loss": 0.3236, + "num_input_tokens_seen": 33931552, + "step": 16250 + }, + { + "epoch": 2.6517660494330695, + "grad_norm": 0.00029997609090059996, + "learning_rate": 0.19349760468923508, + "loss": 0.3363, + "num_input_tokens_seen": 33943072, + "step": 16255 + }, + { + "epoch": 2.652581776653887, + "grad_norm": 0.001159815350547433, + "learning_rate": 0.19344122752899925, + "loss": 0.3359, + "num_input_tokens_seen": 33952448, + "step": 16260 + }, + { + "epoch": 2.653397503874704, + "grad_norm": 0.0009911480592563748, + "learning_rate": 0.1933848436695802, + "loss": 0.349, + "num_input_tokens_seen": 33963264, + "step": 16265 + }, + { + "epoch": 2.6542132310955218, + "grad_norm": 0.0012653777375817299, + "learning_rate": 0.1933284531196731, + "loss": 0.3639, + "num_input_tokens_seen": 33974368, + "step": 16270 + }, + { + "epoch": 2.655028958316339, + "grad_norm": 0.0003862916200887412, + "learning_rate": 0.19327205588797403, + "loss": 0.2993, + "num_input_tokens_seen": 33985248, + "step": 16275 + }, + { + "epoch": 2.6558446855371565, + "grad_norm": 0.0006881604786030948, + "learning_rate": 0.19321565198318014, + "loss": 0.3589, + "num_input_tokens_seen": 33995632, + "step": 16280 + }, + { + "epoch": 2.6566604127579736, + "grad_norm": 0.0005148484488017857, + "learning_rate": 0.1931592414139896, + "loss": 0.2831, + "num_input_tokens_seen": 34006160, + "step": 16285 + }, + { + "epoch": 2.657476139978791, + "grad_norm": 0.00038578329258598387, + "learning_rate": 0.19310282418910169, + "loss": 0.3234, + "num_input_tokens_seen": 34016576, + "step": 16290 + }, + { + "epoch": 2.6582918671996083, + "grad_norm": 0.0003473304386716336, + "learning_rate": 0.1930464003172166, + "loss": 0.3177, + "num_input_tokens_seen": 34027776, + "step": 16295 + }, + { + "epoch": 2.659107594420426, + "grad_norm": 0.00044559809612110257, + "learning_rate": 0.19298996980703567, + "loss": 0.3303, + "num_input_tokens_seen": 34038256, + "step": 16300 + }, + { + "epoch": 2.659923321641243, + "grad_norm": 0.000633994466625154, + "learning_rate": 0.19293353266726113, + "loss": 0.3494, + "num_input_tokens_seen": 34046896, + "step": 16305 + }, + { + "epoch": 2.6607390488620606, + "grad_norm": 0.0011159294517710805, + "learning_rate": 0.19287708890659633, + "loss": 0.3979, + "num_input_tokens_seen": 34056896, + "step": 16310 + }, + { + "epoch": 2.6615547760828777, + "grad_norm": 0.0006646146648563445, + "learning_rate": 0.19282063853374556, + "loss": 0.3332, + "num_input_tokens_seen": 34066672, + "step": 16315 + }, + { + "epoch": 2.6623705033036953, + "grad_norm": 0.0006059612496756017, + "learning_rate": 0.19276418155741423, + "loss": 0.2688, + "num_input_tokens_seen": 34078384, + "step": 16320 + }, + { + "epoch": 2.663186230524513, + "grad_norm": 0.00034861546009778976, + "learning_rate": 0.19270771798630867, + "loss": 0.3401, + "num_input_tokens_seen": 34087568, + "step": 16325 + }, + { + "epoch": 2.66400195774533, + "grad_norm": 0.00047682650620117784, + "learning_rate": 0.1926512478291363, + "loss": 0.3293, + "num_input_tokens_seen": 34098896, + "step": 16330 + }, + { + "epoch": 2.664817684966147, + "grad_norm": 0.0002595184778328985, + "learning_rate": 0.19259477109460557, + "loss": 0.3885, + "num_input_tokens_seen": 34108032, + "step": 16335 + }, + { + "epoch": 2.6656334121869647, + "grad_norm": 0.00024547515204176307, + "learning_rate": 0.19253828779142584, + "loss": 0.3026, + "num_input_tokens_seen": 34118448, + "step": 16340 + }, + { + "epoch": 2.6664491394077823, + "grad_norm": 0.0003280538658145815, + "learning_rate": 0.19248179792830755, + "loss": 0.3216, + "num_input_tokens_seen": 34129936, + "step": 16345 + }, + { + "epoch": 2.6672648666285994, + "grad_norm": 0.0008201033342629671, + "learning_rate": 0.19242530151396217, + "loss": 0.2958, + "num_input_tokens_seen": 34141632, + "step": 16350 + }, + { + "epoch": 2.6680805938494165, + "grad_norm": 0.0004074828466400504, + "learning_rate": 0.19236879855710215, + "loss": 0.3317, + "num_input_tokens_seen": 34150912, + "step": 16355 + }, + { + "epoch": 2.668896321070234, + "grad_norm": 0.00041510621667839587, + "learning_rate": 0.19231228906644096, + "loss": 0.332, + "num_input_tokens_seen": 34161312, + "step": 16360 + }, + { + "epoch": 2.6697120482910517, + "grad_norm": 0.0005664521595463157, + "learning_rate": 0.19225577305069302, + "loss": 0.3449, + "num_input_tokens_seen": 34173104, + "step": 16365 + }, + { + "epoch": 2.670527775511869, + "grad_norm": 0.00034729429171420634, + "learning_rate": 0.1921992505185739, + "loss": 0.3428, + "num_input_tokens_seen": 34183376, + "step": 16370 + }, + { + "epoch": 2.671343502732686, + "grad_norm": 0.0003043870092369616, + "learning_rate": 0.19214272147880004, + "loss": 0.326, + "num_input_tokens_seen": 34193168, + "step": 16375 + }, + { + "epoch": 2.6721592299535035, + "grad_norm": 0.0006663418025709689, + "learning_rate": 0.19208618594008892, + "loss": 0.3341, + "num_input_tokens_seen": 34203664, + "step": 16380 + }, + { + "epoch": 2.672974957174321, + "grad_norm": 0.0007036480237729847, + "learning_rate": 0.19202964391115904, + "loss": 0.3537, + "num_input_tokens_seen": 34214544, + "step": 16385 + }, + { + "epoch": 2.6737906843951382, + "grad_norm": 0.0005677424487657845, + "learning_rate": 0.1919730954007299, + "loss": 0.3644, + "num_input_tokens_seen": 34225424, + "step": 16390 + }, + { + "epoch": 2.6746064116159554, + "grad_norm": 0.000640353886410594, + "learning_rate": 0.19191654041752199, + "loss": 0.2884, + "num_input_tokens_seen": 34236496, + "step": 16395 + }, + { + "epoch": 2.675422138836773, + "grad_norm": 0.0003767431480810046, + "learning_rate": 0.19185997897025678, + "loss": 0.2707, + "num_input_tokens_seen": 34245456, + "step": 16400 + }, + { + "epoch": 2.675422138836773, + "eval_loss": 0.3170260190963745, + "eval_runtime": 156.0356, + "eval_samples_per_second": 17.464, + "eval_steps_per_second": 8.735, + "num_input_tokens_seen": 34245456, + "step": 16400 + }, + { + "epoch": 2.6762378660575905, + "grad_norm": 0.0006025170441716909, + "learning_rate": 0.19180341106765672, + "loss": 0.3109, + "num_input_tokens_seen": 34256464, + "step": 16405 + }, + { + "epoch": 2.6770535932784076, + "grad_norm": 0.00029324996285140514, + "learning_rate": 0.19174683671844536, + "loss": 0.3602, + "num_input_tokens_seen": 34267728, + "step": 16410 + }, + { + "epoch": 2.677869320499225, + "grad_norm": 0.0004684047307819128, + "learning_rate": 0.19169025593134717, + "loss": 0.3376, + "num_input_tokens_seen": 34278192, + "step": 16415 + }, + { + "epoch": 2.6786850477200423, + "grad_norm": 0.00033679389161989093, + "learning_rate": 0.19163366871508764, + "loss": 0.3799, + "num_input_tokens_seen": 34288624, + "step": 16420 + }, + { + "epoch": 2.67950077494086, + "grad_norm": 0.000594426179304719, + "learning_rate": 0.19157707507839317, + "loss": 0.3182, + "num_input_tokens_seen": 34299536, + "step": 16425 + }, + { + "epoch": 2.680316502161677, + "grad_norm": 0.00028609801665879786, + "learning_rate": 0.19152047502999123, + "loss": 0.3512, + "num_input_tokens_seen": 34310640, + "step": 16430 + }, + { + "epoch": 2.6811322293824946, + "grad_norm": 0.00024250143906101584, + "learning_rate": 0.19146386857861025, + "loss": 0.2878, + "num_input_tokens_seen": 34321392, + "step": 16435 + }, + { + "epoch": 2.6819479566033118, + "grad_norm": 0.00024697757908143103, + "learning_rate": 0.19140725573297968, + "loss": 0.3809, + "num_input_tokens_seen": 34332544, + "step": 16440 + }, + { + "epoch": 2.6827636838241293, + "grad_norm": 0.0005050341715104878, + "learning_rate": 0.19135063650182987, + "loss": 0.2928, + "num_input_tokens_seen": 34342800, + "step": 16445 + }, + { + "epoch": 2.6835794110449465, + "grad_norm": 0.0005961167044006288, + "learning_rate": 0.19129401089389234, + "loss": 0.3311, + "num_input_tokens_seen": 34353200, + "step": 16450 + }, + { + "epoch": 2.684395138265764, + "grad_norm": 0.00033180610626004636, + "learning_rate": 0.19123737891789938, + "loss": 0.3066, + "num_input_tokens_seen": 34364784, + "step": 16455 + }, + { + "epoch": 2.685210865486581, + "grad_norm": 0.00043016611016355455, + "learning_rate": 0.19118074058258439, + "loss": 0.2989, + "num_input_tokens_seen": 34375824, + "step": 16460 + }, + { + "epoch": 2.6860265927073987, + "grad_norm": 0.0003791759954765439, + "learning_rate": 0.1911240958966816, + "loss": 0.2823, + "num_input_tokens_seen": 34387168, + "step": 16465 + }, + { + "epoch": 2.686842319928216, + "grad_norm": 0.0005360979121178389, + "learning_rate": 0.19106744486892652, + "loss": 0.2817, + "num_input_tokens_seen": 34398128, + "step": 16470 + }, + { + "epoch": 2.6876580471490334, + "grad_norm": 0.0007753453101031482, + "learning_rate": 0.1910107875080553, + "loss": 0.215, + "num_input_tokens_seen": 34409296, + "step": 16475 + }, + { + "epoch": 2.6884737743698506, + "grad_norm": 0.0003998517931904644, + "learning_rate": 0.19095412382280533, + "loss": 0.3213, + "num_input_tokens_seen": 34418224, + "step": 16480 + }, + { + "epoch": 2.689289501590668, + "grad_norm": 0.0002238836168544367, + "learning_rate": 0.19089745382191473, + "loss": 0.3202, + "num_input_tokens_seen": 34428720, + "step": 16485 + }, + { + "epoch": 2.6901052288114853, + "grad_norm": 0.00031653314363211393, + "learning_rate": 0.19084077751412284, + "loss": 0.3308, + "num_input_tokens_seen": 34438752, + "step": 16490 + }, + { + "epoch": 2.690920956032303, + "grad_norm": 0.0006231372826732695, + "learning_rate": 0.19078409490816986, + "loss": 0.3593, + "num_input_tokens_seen": 34449840, + "step": 16495 + }, + { + "epoch": 2.6917366832531204, + "grad_norm": 0.0003698734799399972, + "learning_rate": 0.19072740601279686, + "loss": 0.2844, + "num_input_tokens_seen": 34460352, + "step": 16500 + }, + { + "epoch": 2.6925524104739376, + "grad_norm": 0.0005866384599357843, + "learning_rate": 0.19067071083674605, + "loss": 0.2886, + "num_input_tokens_seen": 34472368, + "step": 16505 + }, + { + "epoch": 2.6933681376947547, + "grad_norm": 0.0006890042568556964, + "learning_rate": 0.19061400938876052, + "loss": 0.2562, + "num_input_tokens_seen": 34484160, + "step": 16510 + }, + { + "epoch": 2.6941838649155723, + "grad_norm": 0.001789048663340509, + "learning_rate": 0.1905573016775844, + "loss": 0.3785, + "num_input_tokens_seen": 34494000, + "step": 16515 + }, + { + "epoch": 2.69499959213639, + "grad_norm": 0.0004699740675278008, + "learning_rate": 0.19050058771196263, + "loss": 0.3579, + "num_input_tokens_seen": 34503968, + "step": 16520 + }, + { + "epoch": 2.695815319357207, + "grad_norm": 0.00099766138009727, + "learning_rate": 0.19044386750064132, + "loss": 0.346, + "num_input_tokens_seen": 34514832, + "step": 16525 + }, + { + "epoch": 2.696631046578024, + "grad_norm": 0.000747030833736062, + "learning_rate": 0.19038714105236737, + "loss": 0.3938, + "num_input_tokens_seen": 34525824, + "step": 16530 + }, + { + "epoch": 2.6974467737988417, + "grad_norm": 0.0006763843703083694, + "learning_rate": 0.19033040837588874, + "loss": 0.3557, + "num_input_tokens_seen": 34536832, + "step": 16535 + }, + { + "epoch": 2.6982625010196593, + "grad_norm": 0.0004039858467876911, + "learning_rate": 0.1902736694799543, + "loss": 0.2537, + "num_input_tokens_seen": 34546720, + "step": 16540 + }, + { + "epoch": 2.6990782282404764, + "grad_norm": 0.0002986262261401862, + "learning_rate": 0.19021692437331392, + "loss": 0.3033, + "num_input_tokens_seen": 34555904, + "step": 16545 + }, + { + "epoch": 2.6998939554612935, + "grad_norm": 0.0006128324894234538, + "learning_rate": 0.1901601730647184, + "loss": 0.3531, + "num_input_tokens_seen": 34566944, + "step": 16550 + }, + { + "epoch": 2.700709682682111, + "grad_norm": 0.0002950016350951046, + "learning_rate": 0.19010341556291954, + "loss": 0.2773, + "num_input_tokens_seen": 34578752, + "step": 16555 + }, + { + "epoch": 2.7015254099029287, + "grad_norm": 0.0006634688470512629, + "learning_rate": 0.19004665187667, + "loss": 0.3216, + "num_input_tokens_seen": 34589856, + "step": 16560 + }, + { + "epoch": 2.702341137123746, + "grad_norm": 0.000346878485288471, + "learning_rate": 0.1899898820147235, + "loss": 0.2978, + "num_input_tokens_seen": 34600896, + "step": 16565 + }, + { + "epoch": 2.703156864344563, + "grad_norm": 0.00021371530601754785, + "learning_rate": 0.18993310598583465, + "loss": 0.3124, + "num_input_tokens_seen": 34611376, + "step": 16570 + }, + { + "epoch": 2.7039725915653805, + "grad_norm": 0.0008687076042406261, + "learning_rate": 0.18987632379875904, + "loss": 0.3773, + "num_input_tokens_seen": 34622896, + "step": 16575 + }, + { + "epoch": 2.704788318786198, + "grad_norm": 0.0005249055102467537, + "learning_rate": 0.18981953546225314, + "loss": 0.2759, + "num_input_tokens_seen": 34634352, + "step": 16580 + }, + { + "epoch": 2.705604046007015, + "grad_norm": 0.0004408776294440031, + "learning_rate": 0.18976274098507445, + "loss": 0.3084, + "num_input_tokens_seen": 34645040, + "step": 16585 + }, + { + "epoch": 2.7064197732278323, + "grad_norm": 0.0003420943976379931, + "learning_rate": 0.18970594037598146, + "loss": 0.3789, + "num_input_tokens_seen": 34654080, + "step": 16590 + }, + { + "epoch": 2.70723550044865, + "grad_norm": 0.0007888608379289508, + "learning_rate": 0.1896491336437335, + "loss": 0.3154, + "num_input_tokens_seen": 34663200, + "step": 16595 + }, + { + "epoch": 2.7080512276694675, + "grad_norm": 0.0009591656853444874, + "learning_rate": 0.18959232079709085, + "loss": 0.3526, + "num_input_tokens_seen": 34673616, + "step": 16600 + }, + { + "epoch": 2.7080512276694675, + "eval_loss": 0.32488399744033813, + "eval_runtime": 156.1065, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 34673616, + "step": 16600 + }, + { + "epoch": 2.7088669548902846, + "grad_norm": 0.0002301844651810825, + "learning_rate": 0.18953550184481477, + "loss": 0.36, + "num_input_tokens_seen": 34684144, + "step": 16605 + }, + { + "epoch": 2.709682682111102, + "grad_norm": 0.0004882966750301421, + "learning_rate": 0.18947867679566752, + "loss": 0.3295, + "num_input_tokens_seen": 34695328, + "step": 16610 + }, + { + "epoch": 2.7104984093319193, + "grad_norm": 0.0004353416443336755, + "learning_rate": 0.18942184565841216, + "loss": 0.3044, + "num_input_tokens_seen": 34706288, + "step": 16615 + }, + { + "epoch": 2.711314136552737, + "grad_norm": 0.0008300227927975357, + "learning_rate": 0.18936500844181278, + "loss": 0.3362, + "num_input_tokens_seen": 34717520, + "step": 16620 + }, + { + "epoch": 2.712129863773554, + "grad_norm": 0.0004456038004718721, + "learning_rate": 0.18930816515463436, + "loss": 0.3903, + "num_input_tokens_seen": 34728480, + "step": 16625 + }, + { + "epoch": 2.7129455909943716, + "grad_norm": 0.0005217691068537533, + "learning_rate": 0.18925131580564297, + "loss": 0.2865, + "num_input_tokens_seen": 34738880, + "step": 16630 + }, + { + "epoch": 2.7137613182151887, + "grad_norm": 0.000511404185090214, + "learning_rate": 0.1891944604036054, + "loss": 0.3234, + "num_input_tokens_seen": 34749200, + "step": 16635 + }, + { + "epoch": 2.7145770454360063, + "grad_norm": 0.00028384008328430355, + "learning_rate": 0.1891375989572895, + "loss": 0.3001, + "num_input_tokens_seen": 34759968, + "step": 16640 + }, + { + "epoch": 2.7153927726568234, + "grad_norm": 0.0005133106606081128, + "learning_rate": 0.18908073147546398, + "loss": 0.3592, + "num_input_tokens_seen": 34769664, + "step": 16645 + }, + { + "epoch": 2.716208499877641, + "grad_norm": 0.0006672985036857426, + "learning_rate": 0.18902385796689858, + "loss": 0.3376, + "num_input_tokens_seen": 34780544, + "step": 16650 + }, + { + "epoch": 2.717024227098458, + "grad_norm": 0.000530361954588443, + "learning_rate": 0.18896697844036384, + "loss": 0.3143, + "num_input_tokens_seen": 34790144, + "step": 16655 + }, + { + "epoch": 2.7178399543192757, + "grad_norm": 0.00034511301782913506, + "learning_rate": 0.18891009290463137, + "loss": 0.3374, + "num_input_tokens_seen": 34800832, + "step": 16660 + }, + { + "epoch": 2.718655681540093, + "grad_norm": 0.00029530367464758456, + "learning_rate": 0.18885320136847353, + "loss": 0.302, + "num_input_tokens_seen": 34810576, + "step": 16665 + }, + { + "epoch": 2.7194714087609104, + "grad_norm": 0.00023189815692603588, + "learning_rate": 0.1887963038406639, + "loss": 0.2983, + "num_input_tokens_seen": 34820976, + "step": 16670 + }, + { + "epoch": 2.7202871359817276, + "grad_norm": 0.00032677367562428117, + "learning_rate": 0.18873940032997658, + "loss": 0.3512, + "num_input_tokens_seen": 34830256, + "step": 16675 + }, + { + "epoch": 2.721102863202545, + "grad_norm": 0.0003563058562576771, + "learning_rate": 0.18868249084518693, + "loss": 0.3368, + "num_input_tokens_seen": 34840848, + "step": 16680 + }, + { + "epoch": 2.7219185904233623, + "grad_norm": 0.0005410537705756724, + "learning_rate": 0.18862557539507102, + "loss": 0.2594, + "num_input_tokens_seen": 34851184, + "step": 16685 + }, + { + "epoch": 2.72273431764418, + "grad_norm": 0.0005221136379987001, + "learning_rate": 0.18856865398840605, + "loss": 0.298, + "num_input_tokens_seen": 34861232, + "step": 16690 + }, + { + "epoch": 2.7235500448649974, + "grad_norm": 0.0008349309209734201, + "learning_rate": 0.18851172663396995, + "loss": 0.2795, + "num_input_tokens_seen": 34869968, + "step": 16695 + }, + { + "epoch": 2.7243657720858145, + "grad_norm": 0.00033490185160189867, + "learning_rate": 0.1884547933405416, + "loss": 0.2946, + "num_input_tokens_seen": 34881120, + "step": 16700 + }, + { + "epoch": 2.7251814993066317, + "grad_norm": 0.0005071200430393219, + "learning_rate": 0.1883978541169009, + "loss": 0.3442, + "num_input_tokens_seen": 34891696, + "step": 16705 + }, + { + "epoch": 2.7259972265274492, + "grad_norm": 0.0011299606412649155, + "learning_rate": 0.18834090897182854, + "loss": 0.3543, + "num_input_tokens_seen": 34901808, + "step": 16710 + }, + { + "epoch": 2.726812953748267, + "grad_norm": 0.0008093455689959228, + "learning_rate": 0.1882839579141062, + "loss": 0.3089, + "num_input_tokens_seen": 34911216, + "step": 16715 + }, + { + "epoch": 2.727628680969084, + "grad_norm": 0.00033012256608344615, + "learning_rate": 0.18822700095251646, + "loss": 0.3389, + "num_input_tokens_seen": 34922304, + "step": 16720 + }, + { + "epoch": 2.728444408189901, + "grad_norm": 0.0004774095432367176, + "learning_rate": 0.18817003809584273, + "loss": 0.2854, + "num_input_tokens_seen": 34932336, + "step": 16725 + }, + { + "epoch": 2.7292601354107187, + "grad_norm": 0.00071330537321046, + "learning_rate": 0.1881130693528695, + "loss": 0.4336, + "num_input_tokens_seen": 34942880, + "step": 16730 + }, + { + "epoch": 2.7300758626315362, + "grad_norm": 0.0005883621051907539, + "learning_rate": 0.18805609473238197, + "loss": 0.343, + "num_input_tokens_seen": 34952592, + "step": 16735 + }, + { + "epoch": 2.7308915898523534, + "grad_norm": 0.0004114277253393084, + "learning_rate": 0.18799911424316643, + "loss": 0.2829, + "num_input_tokens_seen": 34964592, + "step": 16740 + }, + { + "epoch": 2.7317073170731705, + "grad_norm": 0.00040299154352396727, + "learning_rate": 0.18794212789400994, + "loss": 0.2846, + "num_input_tokens_seen": 34975328, + "step": 16745 + }, + { + "epoch": 2.732523044293988, + "grad_norm": 0.00037631537998095155, + "learning_rate": 0.18788513569370052, + "loss": 0.3092, + "num_input_tokens_seen": 34985088, + "step": 16750 + }, + { + "epoch": 2.7333387715148056, + "grad_norm": 0.00045382342068478465, + "learning_rate": 0.1878281376510271, + "loss": 0.361, + "num_input_tokens_seen": 34994960, + "step": 16755 + }, + { + "epoch": 2.7341544987356228, + "grad_norm": 0.0007335554109886289, + "learning_rate": 0.18777113377477941, + "loss": 0.3697, + "num_input_tokens_seen": 35005184, + "step": 16760 + }, + { + "epoch": 2.73497022595644, + "grad_norm": 0.0003903168544638902, + "learning_rate": 0.1877141240737483, + "loss": 0.3489, + "num_input_tokens_seen": 35018112, + "step": 16765 + }, + { + "epoch": 2.7357859531772575, + "grad_norm": 0.0004197876260150224, + "learning_rate": 0.18765710855672527, + "loss": 0.2494, + "num_input_tokens_seen": 35029856, + "step": 16770 + }, + { + "epoch": 2.736601680398075, + "grad_norm": 0.0004007467068731785, + "learning_rate": 0.18760008723250288, + "loss": 0.2812, + "num_input_tokens_seen": 35040416, + "step": 16775 + }, + { + "epoch": 2.737417407618892, + "grad_norm": 0.00038373187999241054, + "learning_rate": 0.18754306010987457, + "loss": 0.3368, + "num_input_tokens_seen": 35051920, + "step": 16780 + }, + { + "epoch": 2.7382331348397098, + "grad_norm": 0.0004743965982925147, + "learning_rate": 0.18748602719763457, + "loss": 0.3463, + "num_input_tokens_seen": 35060656, + "step": 16785 + }, + { + "epoch": 2.739048862060527, + "grad_norm": 0.00039735130849294364, + "learning_rate": 0.18742898850457804, + "loss": 0.3184, + "num_input_tokens_seen": 35072064, + "step": 16790 + }, + { + "epoch": 2.7398645892813445, + "grad_norm": 0.0004222822608426213, + "learning_rate": 0.1873719440395012, + "loss": 0.2788, + "num_input_tokens_seen": 35081360, + "step": 16795 + }, + { + "epoch": 2.7406803165021616, + "grad_norm": 0.000935245247092098, + "learning_rate": 0.1873148938112009, + "loss": 0.3808, + "num_input_tokens_seen": 35089872, + "step": 16800 + }, + { + "epoch": 2.7406803165021616, + "eval_loss": 0.3165464699268341, + "eval_runtime": 156.1027, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 35089872, + "step": 16800 + }, + { + "epoch": 2.741496043722979, + "grad_norm": 0.0005090018385089934, + "learning_rate": 0.18725783782847508, + "loss": 0.3572, + "num_input_tokens_seen": 35100608, + "step": 16805 + }, + { + "epoch": 2.7423117709437963, + "grad_norm": 0.0003739780804608017, + "learning_rate": 0.1872007761001224, + "loss": 0.3227, + "num_input_tokens_seen": 35111216, + "step": 16810 + }, + { + "epoch": 2.743127498164614, + "grad_norm": 0.00031287240562960505, + "learning_rate": 0.1871437086349426, + "loss": 0.3393, + "num_input_tokens_seen": 35120848, + "step": 16815 + }, + { + "epoch": 2.743943225385431, + "grad_norm": 0.0007218412938527763, + "learning_rate": 0.18708663544173615, + "loss": 0.3046, + "num_input_tokens_seen": 35130416, + "step": 16820 + }, + { + "epoch": 2.7447589526062486, + "grad_norm": 0.000642957747913897, + "learning_rate": 0.18702955652930442, + "loss": 0.3431, + "num_input_tokens_seen": 35141152, + "step": 16825 + }, + { + "epoch": 2.7455746798270657, + "grad_norm": 0.0002768718695733696, + "learning_rate": 0.18697247190644972, + "loss": 0.3239, + "num_input_tokens_seen": 35151424, + "step": 16830 + }, + { + "epoch": 2.7463904070478833, + "grad_norm": 0.0004325930494815111, + "learning_rate": 0.18691538158197527, + "loss": 0.2714, + "num_input_tokens_seen": 35162128, + "step": 16835 + }, + { + "epoch": 2.7472061342687004, + "grad_norm": 0.00030190605320967734, + "learning_rate": 0.1868582855646851, + "loss": 0.3186, + "num_input_tokens_seen": 35172688, + "step": 16840 + }, + { + "epoch": 2.748021861489518, + "grad_norm": 0.00042972961091436446, + "learning_rate": 0.18680118386338404, + "loss": 0.2744, + "num_input_tokens_seen": 35182672, + "step": 16845 + }, + { + "epoch": 2.748837588710335, + "grad_norm": 0.0012734206393361092, + "learning_rate": 0.18674407648687794, + "loss": 0.3493, + "num_input_tokens_seen": 35193920, + "step": 16850 + }, + { + "epoch": 2.7496533159311527, + "grad_norm": 0.0004247003816999495, + "learning_rate": 0.1866869634439736, + "loss": 0.3056, + "num_input_tokens_seen": 35204576, + "step": 16855 + }, + { + "epoch": 2.75046904315197, + "grad_norm": 0.00055412802612409, + "learning_rate": 0.18662984474347838, + "loss": 0.3221, + "num_input_tokens_seen": 35215056, + "step": 16860 + }, + { + "epoch": 2.7512847703727874, + "grad_norm": 0.0006085578352212906, + "learning_rate": 0.1865727203942008, + "loss": 0.3798, + "num_input_tokens_seen": 35223456, + "step": 16865 + }, + { + "epoch": 2.7521004975936045, + "grad_norm": 0.0003050450759474188, + "learning_rate": 0.1865155904049501, + "loss": 0.3111, + "num_input_tokens_seen": 35235344, + "step": 16870 + }, + { + "epoch": 2.752916224814422, + "grad_norm": 0.00022891565458849072, + "learning_rate": 0.1864584547845365, + "loss": 0.308, + "num_input_tokens_seen": 35245824, + "step": 16875 + }, + { + "epoch": 2.7537319520352392, + "grad_norm": 0.00033613009145483375, + "learning_rate": 0.186401313541771, + "loss": 0.3291, + "num_input_tokens_seen": 35255808, + "step": 16880 + }, + { + "epoch": 2.754547679256057, + "grad_norm": 0.0003718797815963626, + "learning_rate": 0.18634416668546552, + "loss": 0.3285, + "num_input_tokens_seen": 35265232, + "step": 16885 + }, + { + "epoch": 2.7553634064768744, + "grad_norm": 0.0005830345908179879, + "learning_rate": 0.1862870142244328, + "loss": 0.3188, + "num_input_tokens_seen": 35274480, + "step": 16890 + }, + { + "epoch": 2.7561791336976915, + "grad_norm": 0.00034927757224068046, + "learning_rate": 0.1862298561674865, + "loss": 0.2885, + "num_input_tokens_seen": 35284816, + "step": 16895 + }, + { + "epoch": 2.7569948609185086, + "grad_norm": 0.0005817347555421293, + "learning_rate": 0.18617269252344104, + "loss": 0.246, + "num_input_tokens_seen": 35296592, + "step": 16900 + }, + { + "epoch": 2.7578105881393262, + "grad_norm": 0.0008022933034226298, + "learning_rate": 0.18611552330111186, + "loss": 0.3971, + "num_input_tokens_seen": 35307152, + "step": 16905 + }, + { + "epoch": 2.758626315360144, + "grad_norm": 0.00022234313655644655, + "learning_rate": 0.18605834850931507, + "loss": 0.2719, + "num_input_tokens_seen": 35317264, + "step": 16910 + }, + { + "epoch": 2.759442042580961, + "grad_norm": 0.0004001515917479992, + "learning_rate": 0.18600116815686787, + "loss": 0.3554, + "num_input_tokens_seen": 35328528, + "step": 16915 + }, + { + "epoch": 2.760257769801778, + "grad_norm": 0.00026154995430260897, + "learning_rate": 0.1859439822525881, + "loss": 0.3419, + "num_input_tokens_seen": 35339936, + "step": 16920 + }, + { + "epoch": 2.7610734970225956, + "grad_norm": 0.00034034281270578504, + "learning_rate": 0.18588679080529455, + "loss": 0.3521, + "num_input_tokens_seen": 35350704, + "step": 16925 + }, + { + "epoch": 2.761889224243413, + "grad_norm": 0.00038429832784458995, + "learning_rate": 0.1858295938238069, + "loss": 0.4206, + "num_input_tokens_seen": 35361584, + "step": 16930 + }, + { + "epoch": 2.7627049514642303, + "grad_norm": 0.0004660432750824839, + "learning_rate": 0.18577239131694562, + "loss": 0.3179, + "num_input_tokens_seen": 35371440, + "step": 16935 + }, + { + "epoch": 2.7635206786850475, + "grad_norm": 0.0005983537994325161, + "learning_rate": 0.18571518329353204, + "loss": 0.3432, + "num_input_tokens_seen": 35382048, + "step": 16940 + }, + { + "epoch": 2.764336405905865, + "grad_norm": 0.000502077687997371, + "learning_rate": 0.18565796976238838, + "loss": 0.2863, + "num_input_tokens_seen": 35392880, + "step": 16945 + }, + { + "epoch": 2.7651521331266826, + "grad_norm": 0.0003123941132798791, + "learning_rate": 0.18560075073233764, + "loss": 0.3552, + "num_input_tokens_seen": 35402944, + "step": 16950 + }, + { + "epoch": 2.7659678603474998, + "grad_norm": 0.00028749092598445714, + "learning_rate": 0.18554352621220377, + "loss": 0.3321, + "num_input_tokens_seen": 35413168, + "step": 16955 + }, + { + "epoch": 2.766783587568317, + "grad_norm": 0.00023230246733874083, + "learning_rate": 0.18548629621081153, + "loss": 0.3394, + "num_input_tokens_seen": 35423888, + "step": 16960 + }, + { + "epoch": 2.7675993147891345, + "grad_norm": 0.0002869931631721556, + "learning_rate": 0.18542906073698645, + "loss": 0.2944, + "num_input_tokens_seen": 35434368, + "step": 16965 + }, + { + "epoch": 2.768415042009952, + "grad_norm": 0.000583695131354034, + "learning_rate": 0.18537181979955494, + "loss": 0.2854, + "num_input_tokens_seen": 35444096, + "step": 16970 + }, + { + "epoch": 2.769230769230769, + "grad_norm": 0.0006391534698195755, + "learning_rate": 0.18531457340734434, + "loss": 0.376, + "num_input_tokens_seen": 35455280, + "step": 16975 + }, + { + "epoch": 2.7700464964515867, + "grad_norm": 0.0006468944484367967, + "learning_rate": 0.1852573215691827, + "loss": 0.2925, + "num_input_tokens_seen": 35466160, + "step": 16980 + }, + { + "epoch": 2.770862223672404, + "grad_norm": 0.00045306948595680296, + "learning_rate": 0.18520006429389904, + "loss": 0.278, + "num_input_tokens_seen": 35475824, + "step": 16985 + }, + { + "epoch": 2.7716779508932214, + "grad_norm": 0.0004949498106725514, + "learning_rate": 0.1851428015903231, + "loss": 0.2984, + "num_input_tokens_seen": 35486272, + "step": 16990 + }, + { + "epoch": 2.7724936781140386, + "grad_norm": 0.0005756798782385886, + "learning_rate": 0.1850855334672855, + "loss": 0.309, + "num_input_tokens_seen": 35497424, + "step": 16995 + }, + { + "epoch": 2.773309405334856, + "grad_norm": 0.0006174396839924157, + "learning_rate": 0.1850282599336178, + "loss": 0.343, + "num_input_tokens_seen": 35508944, + "step": 17000 + }, + { + "epoch": 2.773309405334856, + "eval_loss": 0.31778526306152344, + "eval_runtime": 156.036, + "eval_samples_per_second": 17.464, + "eval_steps_per_second": 8.735, + "num_input_tokens_seen": 35508944, + "step": 17000 + }, + { + "epoch": 2.7741251325556733, + "grad_norm": 0.0005779413622803986, + "learning_rate": 0.18497098099815215, + "loss": 0.4025, + "num_input_tokens_seen": 35519872, + "step": 17005 + }, + { + "epoch": 2.774940859776491, + "grad_norm": 0.00033362157410010695, + "learning_rate": 0.18491369666972174, + "loss": 0.3052, + "num_input_tokens_seen": 35530032, + "step": 17010 + }, + { + "epoch": 2.775756586997308, + "grad_norm": 0.0003982487542089075, + "learning_rate": 0.1848564069571606, + "loss": 0.272, + "num_input_tokens_seen": 35538944, + "step": 17015 + }, + { + "epoch": 2.7765723142181256, + "grad_norm": 0.0006615103338845074, + "learning_rate": 0.18479911186930348, + "loss": 0.3451, + "num_input_tokens_seen": 35549136, + "step": 17020 + }, + { + "epoch": 2.7773880414389427, + "grad_norm": 0.0006302974070422351, + "learning_rate": 0.18474181141498597, + "loss": 0.2834, + "num_input_tokens_seen": 35559376, + "step": 17025 + }, + { + "epoch": 2.7782037686597603, + "grad_norm": 0.0005075373919680715, + "learning_rate": 0.18468450560304453, + "loss": 0.3539, + "num_input_tokens_seen": 35570352, + "step": 17030 + }, + { + "epoch": 2.7790194958805774, + "grad_norm": 0.0002032132906606421, + "learning_rate": 0.1846271944423165, + "loss": 0.282, + "num_input_tokens_seen": 35582144, + "step": 17035 + }, + { + "epoch": 2.779835223101395, + "grad_norm": 0.0003404257004149258, + "learning_rate": 0.18456987794163993, + "loss": 0.3311, + "num_input_tokens_seen": 35592304, + "step": 17040 + }, + { + "epoch": 2.780650950322212, + "grad_norm": 0.001022293115966022, + "learning_rate": 0.18451255610985373, + "loss": 0.3457, + "num_input_tokens_seen": 35603264, + "step": 17045 + }, + { + "epoch": 2.7814666775430297, + "grad_norm": 0.0002676509029697627, + "learning_rate": 0.18445522895579766, + "loss": 0.328, + "num_input_tokens_seen": 35613904, + "step": 17050 + }, + { + "epoch": 2.782282404763847, + "grad_norm": 0.0002372012531850487, + "learning_rate": 0.1843978964883123, + "loss": 0.2637, + "num_input_tokens_seen": 35625424, + "step": 17055 + }, + { + "epoch": 2.7830981319846644, + "grad_norm": 0.0002931032213382423, + "learning_rate": 0.18434055871623906, + "loss": 0.374, + "num_input_tokens_seen": 35636720, + "step": 17060 + }, + { + "epoch": 2.783913859205482, + "grad_norm": 0.00026620953576639295, + "learning_rate": 0.18428321564842007, + "loss": 0.324, + "num_input_tokens_seen": 35645360, + "step": 17065 + }, + { + "epoch": 2.784729586426299, + "grad_norm": 0.0006530903046950698, + "learning_rate": 0.18422586729369841, + "loss": 0.3276, + "num_input_tokens_seen": 35655600, + "step": 17070 + }, + { + "epoch": 2.785545313647116, + "grad_norm": 0.0008659678278490901, + "learning_rate": 0.1841685136609179, + "loss": 0.3513, + "num_input_tokens_seen": 35664736, + "step": 17075 + }, + { + "epoch": 2.786361040867934, + "grad_norm": 0.0005457377410493791, + "learning_rate": 0.18411115475892326, + "loss": 0.3021, + "num_input_tokens_seen": 35675040, + "step": 17080 + }, + { + "epoch": 2.7871767680887514, + "grad_norm": 0.001483965665102005, + "learning_rate": 0.18405379059655982, + "loss": 0.3353, + "num_input_tokens_seen": 35684384, + "step": 17085 + }, + { + "epoch": 2.7879924953095685, + "grad_norm": 0.0005437543732114136, + "learning_rate": 0.1839964211826739, + "loss": 0.3124, + "num_input_tokens_seen": 35695216, + "step": 17090 + }, + { + "epoch": 2.7888082225303856, + "grad_norm": 0.0004608062736224383, + "learning_rate": 0.18393904652611265, + "loss": 0.3038, + "num_input_tokens_seen": 35705920, + "step": 17095 + }, + { + "epoch": 2.789623949751203, + "grad_norm": 0.0005466967122629285, + "learning_rate": 0.18388166663572392, + "loss": 0.3127, + "num_input_tokens_seen": 35715104, + "step": 17100 + }, + { + "epoch": 2.7904396769720208, + "grad_norm": 0.0002703836071304977, + "learning_rate": 0.18382428152035643, + "loss": 0.3028, + "num_input_tokens_seen": 35724960, + "step": 17105 + }, + { + "epoch": 2.791255404192838, + "grad_norm": 0.00026399444323033094, + "learning_rate": 0.1837668911888596, + "loss": 0.3267, + "num_input_tokens_seen": 35734432, + "step": 17110 + }, + { + "epoch": 2.792071131413655, + "grad_norm": 0.0003823181614279747, + "learning_rate": 0.18370949565008388, + "loss": 0.4076, + "num_input_tokens_seen": 35745392, + "step": 17115 + }, + { + "epoch": 2.7928868586344726, + "grad_norm": 0.0005957210669294, + "learning_rate": 0.1836520949128803, + "loss": 0.4104, + "num_input_tokens_seen": 35756096, + "step": 17120 + }, + { + "epoch": 2.79370258585529, + "grad_norm": 0.0005126815522089601, + "learning_rate": 0.18359468898610076, + "loss": 0.3109, + "num_input_tokens_seen": 35767056, + "step": 17125 + }, + { + "epoch": 2.7945183130761073, + "grad_norm": 0.00048722224892117083, + "learning_rate": 0.18353727787859797, + "loss": 0.269, + "num_input_tokens_seen": 35775424, + "step": 17130 + }, + { + "epoch": 2.7953340402969244, + "grad_norm": 0.0005755027523264289, + "learning_rate": 0.18347986159922552, + "loss": 0.2635, + "num_input_tokens_seen": 35785376, + "step": 17135 + }, + { + "epoch": 2.796149767517742, + "grad_norm": 0.0005954491789452732, + "learning_rate": 0.1834224401568377, + "loss": 0.3778, + "num_input_tokens_seen": 35796928, + "step": 17140 + }, + { + "epoch": 2.7969654947385596, + "grad_norm": 0.00031485053477808833, + "learning_rate": 0.1833650135602896, + "loss": 0.3566, + "num_input_tokens_seen": 35807216, + "step": 17145 + }, + { + "epoch": 2.7977812219593767, + "grad_norm": 0.00029119086684659123, + "learning_rate": 0.18330758181843707, + "loss": 0.3056, + "num_input_tokens_seen": 35817088, + "step": 17150 + }, + { + "epoch": 2.7985969491801943, + "grad_norm": 0.0004288619093131274, + "learning_rate": 0.18325014494013686, + "loss": 0.3592, + "num_input_tokens_seen": 35828160, + "step": 17155 + }, + { + "epoch": 2.7994126764010114, + "grad_norm": 0.00023306807270273566, + "learning_rate": 0.18319270293424647, + "loss": 0.2641, + "num_input_tokens_seen": 35839072, + "step": 17160 + }, + { + "epoch": 2.800228403621829, + "grad_norm": 0.0005877825315110385, + "learning_rate": 0.18313525580962417, + "loss": 0.3411, + "num_input_tokens_seen": 35849472, + "step": 17165 + }, + { + "epoch": 2.801044130842646, + "grad_norm": 0.0003957065346185118, + "learning_rate": 0.18307780357512896, + "loss": 0.337, + "num_input_tokens_seen": 35858000, + "step": 17170 + }, + { + "epoch": 2.8018598580634637, + "grad_norm": 0.0006370706250891089, + "learning_rate": 0.1830203462396208, + "loss": 0.2838, + "num_input_tokens_seen": 35867488, + "step": 17175 + }, + { + "epoch": 2.802675585284281, + "grad_norm": 0.0012543922057375312, + "learning_rate": 0.18296288381196033, + "loss": 0.3856, + "num_input_tokens_seen": 35878608, + "step": 17180 + }, + { + "epoch": 2.8034913125050984, + "grad_norm": 0.0002879262319765985, + "learning_rate": 0.1829054163010089, + "loss": 0.3111, + "num_input_tokens_seen": 35889840, + "step": 17185 + }, + { + "epoch": 2.8043070397259156, + "grad_norm": 0.0006897313287481666, + "learning_rate": 0.18284794371562874, + "loss": 0.3029, + "num_input_tokens_seen": 35901072, + "step": 17190 + }, + { + "epoch": 2.805122766946733, + "grad_norm": 0.00035443209344521165, + "learning_rate": 0.18279046606468288, + "loss": 0.3076, + "num_input_tokens_seen": 35911616, + "step": 17195 + }, + { + "epoch": 2.8059384941675503, + "grad_norm": 0.0004815393767785281, + "learning_rate": 0.1827329833570351, + "loss": 0.3599, + "num_input_tokens_seen": 35922144, + "step": 17200 + }, + { + "epoch": 2.8059384941675503, + "eval_loss": 0.3173592686653137, + "eval_runtime": 156.0706, + "eval_samples_per_second": 17.46, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 35922144, + "step": 17200 + }, + { + "epoch": 2.806754221388368, + "grad_norm": 0.00029707790235988796, + "learning_rate": 0.18267549560154991, + "loss": 0.3167, + "num_input_tokens_seen": 35933024, + "step": 17205 + }, + { + "epoch": 2.807569948609185, + "grad_norm": 0.0009553282870911062, + "learning_rate": 0.18261800280709267, + "loss": 0.3651, + "num_input_tokens_seen": 35944192, + "step": 17210 + }, + { + "epoch": 2.8083856758300025, + "grad_norm": 0.0003318752278573811, + "learning_rate": 0.18256050498252957, + "loss": 0.3408, + "num_input_tokens_seen": 35954208, + "step": 17215 + }, + { + "epoch": 2.8092014030508197, + "grad_norm": 0.00031180318910628557, + "learning_rate": 0.18250300213672735, + "loss": 0.3137, + "num_input_tokens_seen": 35963440, + "step": 17220 + }, + { + "epoch": 2.8100171302716372, + "grad_norm": 0.000575736747123301, + "learning_rate": 0.18244549427855378, + "loss": 0.3172, + "num_input_tokens_seen": 35974224, + "step": 17225 + }, + { + "epoch": 2.8108328574924544, + "grad_norm": 0.0006883296882733703, + "learning_rate": 0.1823879814168772, + "loss": 0.343, + "num_input_tokens_seen": 35982752, + "step": 17230 + }, + { + "epoch": 2.811648584713272, + "grad_norm": 0.0005071866326034069, + "learning_rate": 0.18233046356056692, + "loss": 0.3062, + "num_input_tokens_seen": 35991984, + "step": 17235 + }, + { + "epoch": 2.812464311934089, + "grad_norm": 0.00036015972727909684, + "learning_rate": 0.18227294071849284, + "loss": 0.3682, + "num_input_tokens_seen": 36002720, + "step": 17240 + }, + { + "epoch": 2.8132800391549067, + "grad_norm": 0.0008670278475619853, + "learning_rate": 0.18221541289952578, + "loss": 0.3117, + "num_input_tokens_seen": 36014592, + "step": 17245 + }, + { + "epoch": 2.814095766375724, + "grad_norm": 0.0004111742600798607, + "learning_rate": 0.18215788011253717, + "loss": 0.3214, + "num_input_tokens_seen": 36026720, + "step": 17250 + }, + { + "epoch": 2.8149114935965414, + "grad_norm": 0.0005245543434284627, + "learning_rate": 0.18210034236639935, + "loss": 0.3508, + "num_input_tokens_seen": 36037200, + "step": 17255 + }, + { + "epoch": 2.815727220817359, + "grad_norm": 0.00040616377373225987, + "learning_rate": 0.1820427996699853, + "loss": 0.3033, + "num_input_tokens_seen": 36048144, + "step": 17260 + }, + { + "epoch": 2.816542948038176, + "grad_norm": 0.00031509323162026703, + "learning_rate": 0.1819852520321689, + "loss": 0.3222, + "num_input_tokens_seen": 36059312, + "step": 17265 + }, + { + "epoch": 2.817358675258993, + "grad_norm": 0.0004446137463673949, + "learning_rate": 0.18192769946182466, + "loss": 0.3059, + "num_input_tokens_seen": 36070640, + "step": 17270 + }, + { + "epoch": 2.8181744024798108, + "grad_norm": 0.0006222766824066639, + "learning_rate": 0.18187014196782794, + "loss": 0.3319, + "num_input_tokens_seen": 36081104, + "step": 17275 + }, + { + "epoch": 2.8189901297006283, + "grad_norm": 0.0008836279739625752, + "learning_rate": 0.18181257955905486, + "loss": 0.2913, + "num_input_tokens_seen": 36092160, + "step": 17280 + }, + { + "epoch": 2.8198058569214455, + "grad_norm": 0.0011137756519019604, + "learning_rate": 0.18175501224438217, + "loss": 0.2876, + "num_input_tokens_seen": 36103200, + "step": 17285 + }, + { + "epoch": 2.8206215841422626, + "grad_norm": 0.0005235695862211287, + "learning_rate": 0.18169744003268756, + "loss": 0.3097, + "num_input_tokens_seen": 36114512, + "step": 17290 + }, + { + "epoch": 2.82143731136308, + "grad_norm": 0.0006263513932935894, + "learning_rate": 0.18163986293284937, + "loss": 0.3234, + "num_input_tokens_seen": 36125808, + "step": 17295 + }, + { + "epoch": 2.8222530385838978, + "grad_norm": 0.0003369559708517045, + "learning_rate": 0.18158228095374673, + "loss": 0.325, + "num_input_tokens_seen": 36136112, + "step": 17300 + }, + { + "epoch": 2.823068765804715, + "grad_norm": 0.0002842480025719851, + "learning_rate": 0.18152469410425945, + "loss": 0.3525, + "num_input_tokens_seen": 36147872, + "step": 17305 + }, + { + "epoch": 2.823884493025532, + "grad_norm": 0.0004271330835763365, + "learning_rate": 0.18146710239326813, + "loss": 0.289, + "num_input_tokens_seen": 36157840, + "step": 17310 + }, + { + "epoch": 2.8247002202463496, + "grad_norm": 0.0010086038382723927, + "learning_rate": 0.18140950582965423, + "loss": 0.2823, + "num_input_tokens_seen": 36167696, + "step": 17315 + }, + { + "epoch": 2.825515947467167, + "grad_norm": 0.0005275433650240302, + "learning_rate": 0.1813519044222998, + "loss": 0.3517, + "num_input_tokens_seen": 36179024, + "step": 17320 + }, + { + "epoch": 2.8263316746879843, + "grad_norm": 0.00032703112810850143, + "learning_rate": 0.18129429818008772, + "loss": 0.3237, + "num_input_tokens_seen": 36189584, + "step": 17325 + }, + { + "epoch": 2.8271474019088014, + "grad_norm": 0.0004193760978523642, + "learning_rate": 0.18123668711190163, + "loss": 0.3204, + "num_input_tokens_seen": 36201184, + "step": 17330 + }, + { + "epoch": 2.827963129129619, + "grad_norm": 0.0003717286163009703, + "learning_rate": 0.18117907122662583, + "loss": 0.3095, + "num_input_tokens_seen": 36212912, + "step": 17335 + }, + { + "epoch": 2.8287788563504366, + "grad_norm": 0.0003493674157653004, + "learning_rate": 0.1811214505331454, + "loss": 0.326, + "num_input_tokens_seen": 36222608, + "step": 17340 + }, + { + "epoch": 2.8295945835712537, + "grad_norm": 0.0005864085396751761, + "learning_rate": 0.1810638250403462, + "loss": 0.3321, + "num_input_tokens_seen": 36232848, + "step": 17345 + }, + { + "epoch": 2.8304103107920713, + "grad_norm": 0.0004814234853256494, + "learning_rate": 0.1810061947571148, + "loss": 0.2986, + "num_input_tokens_seen": 36243776, + "step": 17350 + }, + { + "epoch": 2.8312260380128884, + "grad_norm": 0.0003809748450294137, + "learning_rate": 0.1809485596923385, + "loss": 0.3519, + "num_input_tokens_seen": 36252960, + "step": 17355 + }, + { + "epoch": 2.832041765233706, + "grad_norm": 0.00039593438850715756, + "learning_rate": 0.18089091985490546, + "loss": 0.2735, + "num_input_tokens_seen": 36262832, + "step": 17360 + }, + { + "epoch": 2.832857492454523, + "grad_norm": 0.0005779835046269, + "learning_rate": 0.18083327525370432, + "loss": 0.349, + "num_input_tokens_seen": 36272544, + "step": 17365 + }, + { + "epoch": 2.8336732196753407, + "grad_norm": 0.00044878257904201746, + "learning_rate": 0.18077562589762464, + "loss": 0.2879, + "num_input_tokens_seen": 36283392, + "step": 17370 + }, + { + "epoch": 2.834488946896158, + "grad_norm": 0.0002274888684041798, + "learning_rate": 0.1807179717955567, + "loss": 0.2773, + "num_input_tokens_seen": 36292272, + "step": 17375 + }, + { + "epoch": 2.8353046741169754, + "grad_norm": 0.0008399909129366279, + "learning_rate": 0.1806603129563915, + "loss": 0.3867, + "num_input_tokens_seen": 36302400, + "step": 17380 + }, + { + "epoch": 2.8361204013377925, + "grad_norm": 0.0005058366223238409, + "learning_rate": 0.1806026493890208, + "loss": 0.3143, + "num_input_tokens_seen": 36313072, + "step": 17385 + }, + { + "epoch": 2.83693612855861, + "grad_norm": 0.0005359407514333725, + "learning_rate": 0.18054498110233688, + "loss": 0.3048, + "num_input_tokens_seen": 36324048, + "step": 17390 + }, + { + "epoch": 2.8377518557794272, + "grad_norm": 0.00034357127151452005, + "learning_rate": 0.1804873081052331, + "loss": 0.3672, + "num_input_tokens_seen": 36335360, + "step": 17395 + }, + { + "epoch": 2.838567583000245, + "grad_norm": 0.0007493679295293987, + "learning_rate": 0.18042963040660326, + "loss": 0.3173, + "num_input_tokens_seen": 36345856, + "step": 17400 + }, + { + "epoch": 2.838567583000245, + "eval_loss": 0.3167881369590759, + "eval_runtime": 155.9931, + "eval_samples_per_second": 17.469, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 36345856, + "step": 17400 + }, + { + "epoch": 2.839383310221062, + "grad_norm": 0.00027907168259844184, + "learning_rate": 0.180371948015342, + "loss": 0.3498, + "num_input_tokens_seen": 36357392, + "step": 17405 + }, + { + "epoch": 2.8401990374418795, + "grad_norm": 0.0003678477369248867, + "learning_rate": 0.18031426094034472, + "loss": 0.3133, + "num_input_tokens_seen": 36367664, + "step": 17410 + }, + { + "epoch": 2.8410147646626966, + "grad_norm": 0.000592562835663557, + "learning_rate": 0.18025656919050737, + "loss": 0.3585, + "num_input_tokens_seen": 36377856, + "step": 17415 + }, + { + "epoch": 2.841830491883514, + "grad_norm": 0.0006572415004484355, + "learning_rate": 0.18019887277472688, + "loss": 0.3434, + "num_input_tokens_seen": 36387200, + "step": 17420 + }, + { + "epoch": 2.8426462191043314, + "grad_norm": 0.0004166710132267326, + "learning_rate": 0.18014117170190067, + "loss": 0.3231, + "num_input_tokens_seen": 36397888, + "step": 17425 + }, + { + "epoch": 2.843461946325149, + "grad_norm": 0.00028392591048032045, + "learning_rate": 0.18008346598092703, + "loss": 0.3282, + "num_input_tokens_seen": 36408752, + "step": 17430 + }, + { + "epoch": 2.8442776735459665, + "grad_norm": 0.0002004958369070664, + "learning_rate": 0.18002575562070489, + "loss": 0.3022, + "num_input_tokens_seen": 36420272, + "step": 17435 + }, + { + "epoch": 2.8450934007667836, + "grad_norm": 0.0005526596214622259, + "learning_rate": 0.1799680406301339, + "loss": 0.302, + "num_input_tokens_seen": 36431040, + "step": 17440 + }, + { + "epoch": 2.8459091279876008, + "grad_norm": 0.00047366891521960497, + "learning_rate": 0.17991032101811447, + "loss": 0.365, + "num_input_tokens_seen": 36441520, + "step": 17445 + }, + { + "epoch": 2.8467248552084183, + "grad_norm": 0.00036752503365278244, + "learning_rate": 0.1798525967935476, + "loss": 0.3141, + "num_input_tokens_seen": 36452272, + "step": 17450 + }, + { + "epoch": 2.847540582429236, + "grad_norm": 0.00033164702472276986, + "learning_rate": 0.17979486796533517, + "loss": 0.3371, + "num_input_tokens_seen": 36462832, + "step": 17455 + }, + { + "epoch": 2.848356309650053, + "grad_norm": 0.0009222259395755827, + "learning_rate": 0.1797371345423797, + "loss": 0.3538, + "num_input_tokens_seen": 36473952, + "step": 17460 + }, + { + "epoch": 2.84917203687087, + "grad_norm": 0.0004421777557581663, + "learning_rate": 0.17967939653358436, + "loss": 0.308, + "num_input_tokens_seen": 36484384, + "step": 17465 + }, + { + "epoch": 2.8499877640916877, + "grad_norm": 0.00036750241997651756, + "learning_rate": 0.17962165394785315, + "loss": 0.3401, + "num_input_tokens_seen": 36493968, + "step": 17470 + }, + { + "epoch": 2.8508034913125053, + "grad_norm": 0.00029554544016718864, + "learning_rate": 0.17956390679409057, + "loss": 0.2677, + "num_input_tokens_seen": 36504144, + "step": 17475 + }, + { + "epoch": 2.8516192185333225, + "grad_norm": 0.0004883355577476323, + "learning_rate": 0.1795061550812021, + "loss": 0.3368, + "num_input_tokens_seen": 36514512, + "step": 17480 + }, + { + "epoch": 2.8524349457541396, + "grad_norm": 0.0004022153152618557, + "learning_rate": 0.1794483988180937, + "loss": 0.3313, + "num_input_tokens_seen": 36525648, + "step": 17485 + }, + { + "epoch": 2.853250672974957, + "grad_norm": 0.0002752228465396911, + "learning_rate": 0.17939063801367214, + "loss": 0.3547, + "num_input_tokens_seen": 36536144, + "step": 17490 + }, + { + "epoch": 2.8540664001957747, + "grad_norm": 0.0004917054320685565, + "learning_rate": 0.17933287267684483, + "loss": 0.292, + "num_input_tokens_seen": 36546000, + "step": 17495 + }, + { + "epoch": 2.854882127416592, + "grad_norm": 0.00046438659774139524, + "learning_rate": 0.17927510281651995, + "loss": 0.2669, + "num_input_tokens_seen": 36556688, + "step": 17500 + }, + { + "epoch": 2.855697854637409, + "grad_norm": 0.0006143227219581604, + "learning_rate": 0.17921732844160634, + "loss": 0.2924, + "num_input_tokens_seen": 36567232, + "step": 17505 + }, + { + "epoch": 2.8565135818582266, + "grad_norm": 0.0010229962645098567, + "learning_rate": 0.17915954956101351, + "loss": 0.3598, + "num_input_tokens_seen": 36576832, + "step": 17510 + }, + { + "epoch": 2.857329309079044, + "grad_norm": 0.0015061988960951567, + "learning_rate": 0.17910176618365165, + "loss": 0.3897, + "num_input_tokens_seen": 36586576, + "step": 17515 + }, + { + "epoch": 2.8581450362998613, + "grad_norm": 0.00031489436514675617, + "learning_rate": 0.17904397831843177, + "loss": 0.3294, + "num_input_tokens_seen": 36598400, + "step": 17520 + }, + { + "epoch": 2.858960763520679, + "grad_norm": 0.0004966185661032796, + "learning_rate": 0.17898618597426547, + "loss": 0.312, + "num_input_tokens_seen": 36609696, + "step": 17525 + }, + { + "epoch": 2.859776490741496, + "grad_norm": 0.0005934697692282498, + "learning_rate": 0.17892838916006495, + "loss": 0.3252, + "num_input_tokens_seen": 36620944, + "step": 17530 + }, + { + "epoch": 2.8605922179623136, + "grad_norm": 0.0009120721952058375, + "learning_rate": 0.17887058788474333, + "loss": 0.3183, + "num_input_tokens_seen": 36630976, + "step": 17535 + }, + { + "epoch": 2.8614079451831307, + "grad_norm": 0.000451945757959038, + "learning_rate": 0.17881278215721427, + "loss": 0.3123, + "num_input_tokens_seen": 36642704, + "step": 17540 + }, + { + "epoch": 2.8622236724039483, + "grad_norm": 0.00035026631667278707, + "learning_rate": 0.1787549719863921, + "loss": 0.2883, + "num_input_tokens_seen": 36652464, + "step": 17545 + }, + { + "epoch": 2.8630393996247654, + "grad_norm": 0.0004345875058788806, + "learning_rate": 0.17869715738119188, + "loss": 0.3612, + "num_input_tokens_seen": 36663120, + "step": 17550 + }, + { + "epoch": 2.863855126845583, + "grad_norm": 0.0009781557600945234, + "learning_rate": 0.17863933835052936, + "loss": 0.298, + "num_input_tokens_seen": 36673872, + "step": 17555 + }, + { + "epoch": 2.8646708540664, + "grad_norm": 0.0008067041635513306, + "learning_rate": 0.17858151490332097, + "loss": 0.3672, + "num_input_tokens_seen": 36683760, + "step": 17560 + }, + { + "epoch": 2.8654865812872177, + "grad_norm": 0.00044802986667491496, + "learning_rate": 0.17852368704848381, + "loss": 0.3085, + "num_input_tokens_seen": 36694112, + "step": 17565 + }, + { + "epoch": 2.866302308508035, + "grad_norm": 0.0008153209346346557, + "learning_rate": 0.17846585479493565, + "loss": 0.3135, + "num_input_tokens_seen": 36705472, + "step": 17570 + }, + { + "epoch": 2.8671180357288524, + "grad_norm": 0.0006567270611412823, + "learning_rate": 0.178408018151595, + "loss": 0.3637, + "num_input_tokens_seen": 36716288, + "step": 17575 + }, + { + "epoch": 2.8679337629496695, + "grad_norm": 0.0002819421933963895, + "learning_rate": 0.17835017712738085, + "loss": 0.3193, + "num_input_tokens_seen": 36727216, + "step": 17580 + }, + { + "epoch": 2.868749490170487, + "grad_norm": 0.0007351613021455705, + "learning_rate": 0.17829233173121323, + "loss": 0.3546, + "num_input_tokens_seen": 36737648, + "step": 17585 + }, + { + "epoch": 2.869565217391304, + "grad_norm": 0.0003487511130515486, + "learning_rate": 0.17823448197201244, + "loss": 0.3067, + "num_input_tokens_seen": 36748784, + "step": 17590 + }, + { + "epoch": 2.870380944612122, + "grad_norm": 0.0005476651713252068, + "learning_rate": 0.1781766278586997, + "loss": 0.3184, + "num_input_tokens_seen": 36758976, + "step": 17595 + }, + { + "epoch": 2.871196671832939, + "grad_norm": 0.00041904294630512595, + "learning_rate": 0.1781187694001969, + "loss": 0.3271, + "num_input_tokens_seen": 36770688, + "step": 17600 + }, + { + "epoch": 2.871196671832939, + "eval_loss": 0.3186143636703491, + "eval_runtime": 155.8561, + "eval_samples_per_second": 17.484, + "eval_steps_per_second": 8.745, + "num_input_tokens_seen": 36770688, + "step": 17600 + }, + { + "epoch": 2.8720123990537565, + "grad_norm": 0.0005233008414506912, + "learning_rate": 0.1780609066054265, + "loss": 0.2843, + "num_input_tokens_seen": 36782448, + "step": 17605 + }, + { + "epoch": 2.8728281262745736, + "grad_norm": 0.0005571604706346989, + "learning_rate": 0.17800303948331164, + "loss": 0.3426, + "num_input_tokens_seen": 36793232, + "step": 17610 + }, + { + "epoch": 2.873643853495391, + "grad_norm": 0.0005514048971235752, + "learning_rate": 0.1779451680427762, + "loss": 0.3127, + "num_input_tokens_seen": 36803664, + "step": 17615 + }, + { + "epoch": 2.8744595807162083, + "grad_norm": 0.0005888240993954241, + "learning_rate": 0.17788729229274464, + "loss": 0.3597, + "num_input_tokens_seen": 36815504, + "step": 17620 + }, + { + "epoch": 2.875275307937026, + "grad_norm": 0.0006132165435701609, + "learning_rate": 0.17782941224214222, + "loss": 0.3286, + "num_input_tokens_seen": 36826224, + "step": 17625 + }, + { + "epoch": 2.8760910351578435, + "grad_norm": 0.00032384376390837133, + "learning_rate": 0.17777152789989464, + "loss": 0.3058, + "num_input_tokens_seen": 36835872, + "step": 17630 + }, + { + "epoch": 2.8769067623786606, + "grad_norm": 0.0003404467715881765, + "learning_rate": 0.17771363927492845, + "loss": 0.2799, + "num_input_tokens_seen": 36846880, + "step": 17635 + }, + { + "epoch": 2.8777224895994777, + "grad_norm": 0.0006744189304299653, + "learning_rate": 0.17765574637617085, + "loss": 0.3167, + "num_input_tokens_seen": 36856528, + "step": 17640 + }, + { + "epoch": 2.8785382168202953, + "grad_norm": 0.0003776420489884913, + "learning_rate": 0.17759784921254962, + "loss": 0.3588, + "num_input_tokens_seen": 36866592, + "step": 17645 + }, + { + "epoch": 2.879353944041113, + "grad_norm": 0.00036026854650117457, + "learning_rate": 0.1775399477929932, + "loss": 0.3265, + "num_input_tokens_seen": 36875232, + "step": 17650 + }, + { + "epoch": 2.88016967126193, + "grad_norm": 0.0008356474572792649, + "learning_rate": 0.17748204212643076, + "loss": 0.3672, + "num_input_tokens_seen": 36885696, + "step": 17655 + }, + { + "epoch": 2.880985398482747, + "grad_norm": 0.0006096410215832293, + "learning_rate": 0.17742413222179204, + "loss": 0.3132, + "num_input_tokens_seen": 36895456, + "step": 17660 + }, + { + "epoch": 2.8818011257035647, + "grad_norm": 0.0005814191536046565, + "learning_rate": 0.17736621808800754, + "loss": 0.3711, + "num_input_tokens_seen": 36905728, + "step": 17665 + }, + { + "epoch": 2.8826168529243823, + "grad_norm": 0.0003615849418565631, + "learning_rate": 0.17730829973400827, + "loss": 0.328, + "num_input_tokens_seen": 36916784, + "step": 17670 + }, + { + "epoch": 2.8834325801451994, + "grad_norm": 0.0004771658277604729, + "learning_rate": 0.17725037716872602, + "loss": 0.3327, + "num_input_tokens_seen": 36927488, + "step": 17675 + }, + { + "epoch": 2.8842483073660166, + "grad_norm": 0.0003583767975214869, + "learning_rate": 0.17719245040109313, + "loss": 0.3108, + "num_input_tokens_seen": 36938704, + "step": 17680 + }, + { + "epoch": 2.885064034586834, + "grad_norm": 0.00042159829172305763, + "learning_rate": 0.17713451944004271, + "loss": 0.3107, + "num_input_tokens_seen": 36947904, + "step": 17685 + }, + { + "epoch": 2.8858797618076517, + "grad_norm": 0.0003205976390745491, + "learning_rate": 0.17707658429450843, + "loss": 0.2856, + "num_input_tokens_seen": 36958672, + "step": 17690 + }, + { + "epoch": 2.886695489028469, + "grad_norm": 0.0005337584880180657, + "learning_rate": 0.1770186449734245, + "loss": 0.2973, + "num_input_tokens_seen": 36970336, + "step": 17695 + }, + { + "epoch": 2.887511216249286, + "grad_norm": 0.0003804836014751345, + "learning_rate": 0.17696070148572599, + "loss": 0.3306, + "num_input_tokens_seen": 36979680, + "step": 17700 + }, + { + "epoch": 2.8883269434701035, + "grad_norm": 0.0003480128652881831, + "learning_rate": 0.17690275384034856, + "loss": 0.3815, + "num_input_tokens_seen": 36990912, + "step": 17705 + }, + { + "epoch": 2.889142670690921, + "grad_norm": 0.0002606328926049173, + "learning_rate": 0.17684480204622835, + "loss": 0.2721, + "num_input_tokens_seen": 37001408, + "step": 17710 + }, + { + "epoch": 2.8899583979117383, + "grad_norm": 0.0006130418041720986, + "learning_rate": 0.1767868461123023, + "loss": 0.3634, + "num_input_tokens_seen": 37011600, + "step": 17715 + }, + { + "epoch": 2.890774125132556, + "grad_norm": 0.0003394945524632931, + "learning_rate": 0.176728886047508, + "loss": 0.2856, + "num_input_tokens_seen": 37022160, + "step": 17720 + }, + { + "epoch": 2.891589852353373, + "grad_norm": 0.0004577224317472428, + "learning_rate": 0.17667092186078362, + "loss": 0.3165, + "num_input_tokens_seen": 37033504, + "step": 17725 + }, + { + "epoch": 2.8924055795741905, + "grad_norm": 0.00024793422198854387, + "learning_rate": 0.17661295356106785, + "loss": 0.3332, + "num_input_tokens_seen": 37045696, + "step": 17730 + }, + { + "epoch": 2.8932213067950077, + "grad_norm": 0.00026724557392299175, + "learning_rate": 0.1765549811573002, + "loss": 0.3426, + "num_input_tokens_seen": 37055840, + "step": 17735 + }, + { + "epoch": 2.8940370340158252, + "grad_norm": 0.0009228740818798542, + "learning_rate": 0.17649700465842078, + "loss": 0.2896, + "num_input_tokens_seen": 37066144, + "step": 17740 + }, + { + "epoch": 2.8948527612366424, + "grad_norm": 0.0004105556581635028, + "learning_rate": 0.17643902407337023, + "loss": 0.2849, + "num_input_tokens_seen": 37076144, + "step": 17745 + }, + { + "epoch": 2.89566848845746, + "grad_norm": 0.0003481419407762587, + "learning_rate": 0.17638103941108993, + "loss": 0.3145, + "num_input_tokens_seen": 37087440, + "step": 17750 + }, + { + "epoch": 2.896484215678277, + "grad_norm": 0.000526929390616715, + "learning_rate": 0.1763230506805218, + "loss": 0.3134, + "num_input_tokens_seen": 37097520, + "step": 17755 + }, + { + "epoch": 2.8972999428990946, + "grad_norm": 0.0002850063901860267, + "learning_rate": 0.1762650578906085, + "loss": 0.2566, + "num_input_tokens_seen": 37109024, + "step": 17760 + }, + { + "epoch": 2.898115670119912, + "grad_norm": 0.0006548782112076879, + "learning_rate": 0.1762070610502932, + "loss": 0.2728, + "num_input_tokens_seen": 37118336, + "step": 17765 + }, + { + "epoch": 2.8989313973407294, + "grad_norm": 0.0003520918544381857, + "learning_rate": 0.17614906016851975, + "loss": 0.2733, + "num_input_tokens_seen": 37128624, + "step": 17770 + }, + { + "epoch": 2.8997471245615465, + "grad_norm": 0.0007843513740226626, + "learning_rate": 0.17609105525423258, + "loss": 0.3422, + "num_input_tokens_seen": 37139824, + "step": 17775 + }, + { + "epoch": 2.900562851782364, + "grad_norm": 0.000647136417683214, + "learning_rate": 0.1760330463163768, + "loss": 0.3164, + "num_input_tokens_seen": 37150288, + "step": 17780 + }, + { + "epoch": 2.901378579003181, + "grad_norm": 0.0005892838235013187, + "learning_rate": 0.17597503336389816, + "loss": 0.3859, + "num_input_tokens_seen": 37162576, + "step": 17785 + }, + { + "epoch": 2.9021943062239988, + "grad_norm": 0.0005481570842675865, + "learning_rate": 0.17591701640574298, + "loss": 0.3883, + "num_input_tokens_seen": 37173104, + "step": 17790 + }, + { + "epoch": 2.903010033444816, + "grad_norm": 0.0004316032864153385, + "learning_rate": 0.17585899545085815, + "loss": 0.3125, + "num_input_tokens_seen": 37184544, + "step": 17795 + }, + { + "epoch": 2.9038257606656335, + "grad_norm": 0.0007048550178296864, + "learning_rate": 0.17580097050819124, + "loss": 0.3495, + "num_input_tokens_seen": 37194864, + "step": 17800 + }, + { + "epoch": 2.9038257606656335, + "eval_loss": 0.3217899799346924, + "eval_runtime": 156.1585, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 37194864, + "step": 17800 + }, + { + "epoch": 2.904641487886451, + "grad_norm": 0.00034963092184625566, + "learning_rate": 0.17574294158669046, + "loss": 0.3422, + "num_input_tokens_seen": 37204528, + "step": 17805 + }, + { + "epoch": 2.905457215107268, + "grad_norm": 0.0004390310205053538, + "learning_rate": 0.17568490869530456, + "loss": 0.3139, + "num_input_tokens_seen": 37214704, + "step": 17810 + }, + { + "epoch": 2.9062729423280853, + "grad_norm": 0.0019696750678122044, + "learning_rate": 0.17562687184298295, + "loss": 0.4008, + "num_input_tokens_seen": 37225456, + "step": 17815 + }, + { + "epoch": 2.907088669548903, + "grad_norm": 0.0004067425325047225, + "learning_rate": 0.1755688310386757, + "loss": 0.3264, + "num_input_tokens_seen": 37235872, + "step": 17820 + }, + { + "epoch": 2.9079043967697205, + "grad_norm": 0.00021354161435738206, + "learning_rate": 0.17551078629133335, + "loss": 0.3582, + "num_input_tokens_seen": 37246032, + "step": 17825 + }, + { + "epoch": 2.9087201239905376, + "grad_norm": 0.0005658926093019545, + "learning_rate": 0.17545273760990718, + "loss": 0.323, + "num_input_tokens_seen": 37256912, + "step": 17830 + }, + { + "epoch": 2.9095358512113547, + "grad_norm": 0.0003338792303111404, + "learning_rate": 0.17539468500334904, + "loss": 0.3031, + "num_input_tokens_seen": 37268896, + "step": 17835 + }, + { + "epoch": 2.9103515784321723, + "grad_norm": 0.00030081463046371937, + "learning_rate": 0.17533662848061132, + "loss": 0.3386, + "num_input_tokens_seen": 37278672, + "step": 17840 + }, + { + "epoch": 2.91116730565299, + "grad_norm": 0.0005128480843268335, + "learning_rate": 0.1752785680506471, + "loss": 0.3483, + "num_input_tokens_seen": 37288208, + "step": 17845 + }, + { + "epoch": 2.911983032873807, + "grad_norm": 0.00028985083918087184, + "learning_rate": 0.17522050372241, + "loss": 0.263, + "num_input_tokens_seen": 37298112, + "step": 17850 + }, + { + "epoch": 2.912798760094624, + "grad_norm": 0.0006078563746996224, + "learning_rate": 0.17516243550485425, + "loss": 0.3461, + "num_input_tokens_seen": 37308032, + "step": 17855 + }, + { + "epoch": 2.9136144873154417, + "grad_norm": 0.0005329691339284182, + "learning_rate": 0.17510436340693478, + "loss": 0.3135, + "num_input_tokens_seen": 37320720, + "step": 17860 + }, + { + "epoch": 2.9144302145362593, + "grad_norm": 0.00041225150926038623, + "learning_rate": 0.175046287437607, + "loss": 0.3148, + "num_input_tokens_seen": 37329840, + "step": 17865 + }, + { + "epoch": 2.9152459417570764, + "grad_norm": 0.0005686948425136507, + "learning_rate": 0.17498820760582695, + "loss": 0.3482, + "num_input_tokens_seen": 37339488, + "step": 17870 + }, + { + "epoch": 2.9160616689778935, + "grad_norm": 0.0006820844137109816, + "learning_rate": 0.1749301239205512, + "loss": 0.313, + "num_input_tokens_seen": 37348608, + "step": 17875 + }, + { + "epoch": 2.916877396198711, + "grad_norm": 0.00027223219512961805, + "learning_rate": 0.1748720363907371, + "loss": 0.3764, + "num_input_tokens_seen": 37358672, + "step": 17880 + }, + { + "epoch": 2.9176931234195287, + "grad_norm": 0.0005073713837191463, + "learning_rate": 0.17481394502534242, + "loss": 0.2993, + "num_input_tokens_seen": 37368752, + "step": 17885 + }, + { + "epoch": 2.918508850640346, + "grad_norm": 0.0010817628353834152, + "learning_rate": 0.17475584983332562, + "loss": 0.3789, + "num_input_tokens_seen": 37379168, + "step": 17890 + }, + { + "epoch": 2.919324577861163, + "grad_norm": 0.0002871961914934218, + "learning_rate": 0.17469775082364558, + "loss": 0.3046, + "num_input_tokens_seen": 37388944, + "step": 17895 + }, + { + "epoch": 2.9201403050819805, + "grad_norm": 0.0003718585358001292, + "learning_rate": 0.17463964800526205, + "loss": 0.3499, + "num_input_tokens_seen": 37399216, + "step": 17900 + }, + { + "epoch": 2.920956032302798, + "grad_norm": 0.00041176131344400346, + "learning_rate": 0.17458154138713522, + "loss": 0.3546, + "num_input_tokens_seen": 37410144, + "step": 17905 + }, + { + "epoch": 2.9217717595236152, + "grad_norm": 0.00030043660080991685, + "learning_rate": 0.17452343097822576, + "loss": 0.3256, + "num_input_tokens_seen": 37421264, + "step": 17910 + }, + { + "epoch": 2.922587486744433, + "grad_norm": 0.0004883660585619509, + "learning_rate": 0.17446531678749497, + "loss": 0.3603, + "num_input_tokens_seen": 37431408, + "step": 17915 + }, + { + "epoch": 2.92340321396525, + "grad_norm": 0.0004995402414351702, + "learning_rate": 0.17440719882390496, + "loss": 0.3251, + "num_input_tokens_seen": 37442976, + "step": 17920 + }, + { + "epoch": 2.9242189411860675, + "grad_norm": 0.00021883746376261115, + "learning_rate": 0.17434907709641814, + "loss": 0.3279, + "num_input_tokens_seen": 37452608, + "step": 17925 + }, + { + "epoch": 2.9250346684068846, + "grad_norm": 0.0002430077875033021, + "learning_rate": 0.17429095161399769, + "loss": 0.3408, + "num_input_tokens_seen": 37462960, + "step": 17930 + }, + { + "epoch": 2.925850395627702, + "grad_norm": 0.0005045098951086402, + "learning_rate": 0.1742328223856072, + "loss": 0.3038, + "num_input_tokens_seen": 37473776, + "step": 17935 + }, + { + "epoch": 2.9266661228485193, + "grad_norm": 0.00040135037852451205, + "learning_rate": 0.174174689420211, + "loss": 0.3458, + "num_input_tokens_seen": 37484896, + "step": 17940 + }, + { + "epoch": 2.927481850069337, + "grad_norm": 0.00022356506087817252, + "learning_rate": 0.1741165527267739, + "loss": 0.334, + "num_input_tokens_seen": 37496272, + "step": 17945 + }, + { + "epoch": 2.928297577290154, + "grad_norm": 0.0006706552812829614, + "learning_rate": 0.17405841231426125, + "loss": 0.3289, + "num_input_tokens_seen": 37505872, + "step": 17950 + }, + { + "epoch": 2.9291133045109716, + "grad_norm": 0.00046017864951863885, + "learning_rate": 0.1740002681916391, + "loss": 0.3546, + "num_input_tokens_seen": 37516864, + "step": 17955 + }, + { + "epoch": 2.9299290317317888, + "grad_norm": 0.0002839392109308392, + "learning_rate": 0.17394212036787401, + "loss": 0.3128, + "num_input_tokens_seen": 37528080, + "step": 17960 + }, + { + "epoch": 2.9307447589526063, + "grad_norm": 0.000761874602176249, + "learning_rate": 0.1738839688519331, + "loss": 0.3399, + "num_input_tokens_seen": 37539600, + "step": 17965 + }, + { + "epoch": 2.9315604861734235, + "grad_norm": 0.0006957595469430089, + "learning_rate": 0.17382581365278402, + "loss": 0.3261, + "num_input_tokens_seen": 37550176, + "step": 17970 + }, + { + "epoch": 2.932376213394241, + "grad_norm": 0.00019814513507299125, + "learning_rate": 0.17376765477939507, + "loss": 0.3116, + "num_input_tokens_seen": 37560304, + "step": 17975 + }, + { + "epoch": 2.933191940615058, + "grad_norm": 0.0002917087113019079, + "learning_rate": 0.1737094922407351, + "loss": 0.3413, + "num_input_tokens_seen": 37571488, + "step": 17980 + }, + { + "epoch": 2.9340076678358757, + "grad_norm": 0.00026613069348968565, + "learning_rate": 0.1736513260457734, + "loss": 0.356, + "num_input_tokens_seen": 37581136, + "step": 17985 + }, + { + "epoch": 2.934823395056693, + "grad_norm": 0.00032306788489222527, + "learning_rate": 0.17359315620348006, + "loss": 0.308, + "num_input_tokens_seen": 37591968, + "step": 17990 + }, + { + "epoch": 2.9356391222775104, + "grad_norm": 0.0006360172992572188, + "learning_rate": 0.17353498272282547, + "loss": 0.2967, + "num_input_tokens_seen": 37603072, + "step": 17995 + }, + { + "epoch": 2.936454849498328, + "grad_norm": 0.0003674353938549757, + "learning_rate": 0.17347680561278087, + "loss": 0.2759, + "num_input_tokens_seen": 37615344, + "step": 18000 + }, + { + "epoch": 2.936454849498328, + "eval_loss": 0.3171684145927429, + "eval_runtime": 155.8023, + "eval_samples_per_second": 17.49, + "eval_steps_per_second": 8.748, + "num_input_tokens_seen": 37615344, + "step": 18000 + }, + { + "epoch": 2.937270576719145, + "grad_norm": 0.00027232334832660854, + "learning_rate": 0.1734186248823178, + "loss": 0.2917, + "num_input_tokens_seen": 37626608, + "step": 18005 + }, + { + "epoch": 2.9380863039399623, + "grad_norm": 0.0003112062404397875, + "learning_rate": 0.17336044054040844, + "loss": 0.3214, + "num_input_tokens_seen": 37637136, + "step": 18010 + }, + { + "epoch": 2.93890203116078, + "grad_norm": 0.00031301044509746134, + "learning_rate": 0.1733022525960256, + "loss": 0.2354, + "num_input_tokens_seen": 37647664, + "step": 18015 + }, + { + "epoch": 2.9397177583815974, + "grad_norm": 0.00046649997239001095, + "learning_rate": 0.1732440610581426, + "loss": 0.3207, + "num_input_tokens_seen": 37658352, + "step": 18020 + }, + { + "epoch": 2.9405334856024146, + "grad_norm": 0.0009460591245442629, + "learning_rate": 0.17318586593573326, + "loss": 0.2596, + "num_input_tokens_seen": 37669504, + "step": 18025 + }, + { + "epoch": 2.9413492128232317, + "grad_norm": 0.00041483648237772286, + "learning_rate": 0.17312766723777204, + "loss": 0.2691, + "num_input_tokens_seen": 37679520, + "step": 18030 + }, + { + "epoch": 2.9421649400440493, + "grad_norm": 0.0003619686176534742, + "learning_rate": 0.1730694649732339, + "loss": 0.314, + "num_input_tokens_seen": 37688176, + "step": 18035 + }, + { + "epoch": 2.942980667264867, + "grad_norm": 0.0008659118320792913, + "learning_rate": 0.17301125915109428, + "loss": 0.2833, + "num_input_tokens_seen": 37698912, + "step": 18040 + }, + { + "epoch": 2.943796394485684, + "grad_norm": 0.0008122857543639839, + "learning_rate": 0.17295304978032938, + "loss": 0.2826, + "num_input_tokens_seen": 37710480, + "step": 18045 + }, + { + "epoch": 2.944612121706501, + "grad_norm": 0.0008761372300796211, + "learning_rate": 0.17289483686991577, + "loss": 0.3268, + "num_input_tokens_seen": 37721216, + "step": 18050 + }, + { + "epoch": 2.9454278489273187, + "grad_norm": 0.00039798146462999284, + "learning_rate": 0.1728366204288306, + "loss": 0.2973, + "num_input_tokens_seen": 37732480, + "step": 18055 + }, + { + "epoch": 2.9462435761481363, + "grad_norm": 0.0005517982644960284, + "learning_rate": 0.17277840046605153, + "loss": 0.3516, + "num_input_tokens_seen": 37743744, + "step": 18060 + }, + { + "epoch": 2.9470593033689534, + "grad_norm": 0.0008783839293755591, + "learning_rate": 0.17272017699055686, + "loss": 0.3433, + "num_input_tokens_seen": 37754848, + "step": 18065 + }, + { + "epoch": 2.9478750305897705, + "grad_norm": 0.00041712672100402415, + "learning_rate": 0.17266195001132542, + "loss": 0.2878, + "num_input_tokens_seen": 37764960, + "step": 18070 + }, + { + "epoch": 2.948690757810588, + "grad_norm": 0.0006325640715658665, + "learning_rate": 0.17260371953733647, + "loss": 0.3649, + "num_input_tokens_seen": 37775856, + "step": 18075 + }, + { + "epoch": 2.9495064850314057, + "grad_norm": 0.0002706728409975767, + "learning_rate": 0.1725454855775699, + "loss": 0.2896, + "num_input_tokens_seen": 37786416, + "step": 18080 + }, + { + "epoch": 2.950322212252223, + "grad_norm": 0.0008034890051931143, + "learning_rate": 0.17248724814100616, + "loss": 0.2812, + "num_input_tokens_seen": 37795104, + "step": 18085 + }, + { + "epoch": 2.9511379394730404, + "grad_norm": 0.00040829196223057806, + "learning_rate": 0.17242900723662619, + "loss": 0.2955, + "num_input_tokens_seen": 37805536, + "step": 18090 + }, + { + "epoch": 2.9519536666938575, + "grad_norm": 0.00029405916575342417, + "learning_rate": 0.1723707628734114, + "loss": 0.368, + "num_input_tokens_seen": 37813488, + "step": 18095 + }, + { + "epoch": 2.952769393914675, + "grad_norm": 0.00023383986263070256, + "learning_rate": 0.1723125150603438, + "loss": 0.2823, + "num_input_tokens_seen": 37823648, + "step": 18100 + }, + { + "epoch": 2.953585121135492, + "grad_norm": 0.0007295131799764931, + "learning_rate": 0.1722542638064061, + "loss": 0.281, + "num_input_tokens_seen": 37833744, + "step": 18105 + }, + { + "epoch": 2.95440084835631, + "grad_norm": 0.000268375501036644, + "learning_rate": 0.17219600912058117, + "loss": 0.2735, + "num_input_tokens_seen": 37844752, + "step": 18110 + }, + { + "epoch": 2.955216575577127, + "grad_norm": 0.0006426885374821723, + "learning_rate": 0.17213775101185272, + "loss": 0.3598, + "num_input_tokens_seen": 37854704, + "step": 18115 + }, + { + "epoch": 2.9560323027979445, + "grad_norm": 0.0016529904678463936, + "learning_rate": 0.17207948948920485, + "loss": 0.348, + "num_input_tokens_seen": 37864976, + "step": 18120 + }, + { + "epoch": 2.9568480300187616, + "grad_norm": 0.0013324158499017358, + "learning_rate": 0.17202122456162228, + "loss": 0.4003, + "num_input_tokens_seen": 37876352, + "step": 18125 + }, + { + "epoch": 2.957663757239579, + "grad_norm": 0.00021396158263087273, + "learning_rate": 0.17196295623809013, + "loss": 0.2734, + "num_input_tokens_seen": 37886000, + "step": 18130 + }, + { + "epoch": 2.9584794844603963, + "grad_norm": 0.0004963969113305211, + "learning_rate": 0.1719046845275941, + "loss": 0.3065, + "num_input_tokens_seen": 37896960, + "step": 18135 + }, + { + "epoch": 2.959295211681214, + "grad_norm": 0.00041696373955346644, + "learning_rate": 0.17184640943912044, + "loss": 0.2717, + "num_input_tokens_seen": 37906224, + "step": 18140 + }, + { + "epoch": 2.960110938902031, + "grad_norm": 0.0004657376848626882, + "learning_rate": 0.1717881309816559, + "loss": 0.2758, + "num_input_tokens_seen": 37917632, + "step": 18145 + }, + { + "epoch": 2.9609266661228486, + "grad_norm": 0.00044472020817920566, + "learning_rate": 0.1717298491641878, + "loss": 0.2919, + "num_input_tokens_seen": 37929440, + "step": 18150 + }, + { + "epoch": 2.9617423933436657, + "grad_norm": 0.0003991824050899595, + "learning_rate": 0.17167156399570385, + "loss": 0.3144, + "num_input_tokens_seen": 37939968, + "step": 18155 + }, + { + "epoch": 2.9625581205644833, + "grad_norm": 0.000584136345423758, + "learning_rate": 0.17161327548519242, + "loss": 0.3644, + "num_input_tokens_seen": 37949728, + "step": 18160 + }, + { + "epoch": 2.9633738477853004, + "grad_norm": 0.0007070123101584613, + "learning_rate": 0.1715549836416423, + "loss": 0.3366, + "num_input_tokens_seen": 37960832, + "step": 18165 + }, + { + "epoch": 2.964189575006118, + "grad_norm": 0.0020120847038924694, + "learning_rate": 0.17149668847404279, + "loss": 0.3293, + "num_input_tokens_seen": 37972704, + "step": 18170 + }, + { + "epoch": 2.965005302226935, + "grad_norm": 0.0006057844730094075, + "learning_rate": 0.1714383899913838, + "loss": 0.3084, + "num_input_tokens_seen": 37983120, + "step": 18175 + }, + { + "epoch": 2.9658210294477527, + "grad_norm": 0.0005001164972782135, + "learning_rate": 0.17138008820265563, + "loss": 0.3723, + "num_input_tokens_seen": 37992672, + "step": 18180 + }, + { + "epoch": 2.96663675666857, + "grad_norm": 0.0008127804612740874, + "learning_rate": 0.17132178311684917, + "loss": 0.3083, + "num_input_tokens_seen": 38002944, + "step": 18185 + }, + { + "epoch": 2.9674524838893874, + "grad_norm": 0.0007944502867758274, + "learning_rate": 0.1712634747429559, + "loss": 0.2997, + "num_input_tokens_seen": 38012096, + "step": 18190 + }, + { + "epoch": 2.968268211110205, + "grad_norm": 0.0003743039269465953, + "learning_rate": 0.17120516308996753, + "loss": 0.3034, + "num_input_tokens_seen": 38020576, + "step": 18195 + }, + { + "epoch": 2.969083938331022, + "grad_norm": 0.0006473443354479969, + "learning_rate": 0.17114684816687653, + "loss": 0.3311, + "num_input_tokens_seen": 38030400, + "step": 18200 + }, + { + "epoch": 2.969083938331022, + "eval_loss": 0.3162723481655121, + "eval_runtime": 156.1101, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 38030400, + "step": 18200 + }, + { + "epoch": 2.9698996655518393, + "grad_norm": 0.0009566736989654601, + "learning_rate": 0.17108852998267585, + "loss": 0.3005, + "num_input_tokens_seen": 38039856, + "step": 18205 + }, + { + "epoch": 2.970715392772657, + "grad_norm": 0.0007238294347189367, + "learning_rate": 0.17103020854635878, + "loss": 0.322, + "num_input_tokens_seen": 38049600, + "step": 18210 + }, + { + "epoch": 2.9715311199934744, + "grad_norm": 0.0008198287687264383, + "learning_rate": 0.1709718838669193, + "loss": 0.246, + "num_input_tokens_seen": 38060512, + "step": 18215 + }, + { + "epoch": 2.9723468472142915, + "grad_norm": 0.000653385475743562, + "learning_rate": 0.17091355595335173, + "loss": 0.4017, + "num_input_tokens_seen": 38069968, + "step": 18220 + }, + { + "epoch": 2.9731625744351087, + "grad_norm": 0.00031299504917114973, + "learning_rate": 0.17085522481465107, + "loss": 0.2859, + "num_input_tokens_seen": 38080560, + "step": 18225 + }, + { + "epoch": 2.9739783016559262, + "grad_norm": 0.0005991485668346286, + "learning_rate": 0.17079689045981264, + "loss": 0.372, + "num_input_tokens_seen": 38089328, + "step": 18230 + }, + { + "epoch": 2.974794028876744, + "grad_norm": 0.0005568366614170372, + "learning_rate": 0.17073855289783238, + "loss": 0.3618, + "num_input_tokens_seen": 38098864, + "step": 18235 + }, + { + "epoch": 2.975609756097561, + "grad_norm": 0.00025347311748191714, + "learning_rate": 0.1706802121377066, + "loss": 0.342, + "num_input_tokens_seen": 38108464, + "step": 18240 + }, + { + "epoch": 2.976425483318378, + "grad_norm": 0.0004895497695542872, + "learning_rate": 0.17062186818843225, + "loss": 0.2708, + "num_input_tokens_seen": 38118272, + "step": 18245 + }, + { + "epoch": 2.9772412105391957, + "grad_norm": 0.0005860217497684062, + "learning_rate": 0.17056352105900668, + "loss": 0.3952, + "num_input_tokens_seen": 38128688, + "step": 18250 + }, + { + "epoch": 2.9780569377600132, + "grad_norm": 0.0006658206111751497, + "learning_rate": 0.17050517075842772, + "loss": 0.3353, + "num_input_tokens_seen": 38137520, + "step": 18255 + }, + { + "epoch": 2.9788726649808304, + "grad_norm": 0.0005004447302781045, + "learning_rate": 0.17044681729569375, + "loss": 0.342, + "num_input_tokens_seen": 38148480, + "step": 18260 + }, + { + "epoch": 2.9796883922016475, + "grad_norm": 0.00020122774003539234, + "learning_rate": 0.17038846067980365, + "loss": 0.2974, + "num_input_tokens_seen": 38158160, + "step": 18265 + }, + { + "epoch": 2.980504119422465, + "grad_norm": 0.0004244815499987453, + "learning_rate": 0.17033010091975664, + "loss": 0.3771, + "num_input_tokens_seen": 38168544, + "step": 18270 + }, + { + "epoch": 2.9813198466432826, + "grad_norm": 0.000903750245925039, + "learning_rate": 0.17027173802455262, + "loss": 0.3546, + "num_input_tokens_seen": 38178848, + "step": 18275 + }, + { + "epoch": 2.9821355738640998, + "grad_norm": 0.0004987167194485664, + "learning_rate": 0.1702133720031918, + "loss": 0.3576, + "num_input_tokens_seen": 38188288, + "step": 18280 + }, + { + "epoch": 2.9829513010849174, + "grad_norm": 0.00031372642843052745, + "learning_rate": 0.17015500286467503, + "loss": 0.3594, + "num_input_tokens_seen": 38197904, + "step": 18285 + }, + { + "epoch": 2.9837670283057345, + "grad_norm": 0.0006714295595884323, + "learning_rate": 0.17009663061800354, + "loss": 0.336, + "num_input_tokens_seen": 38208640, + "step": 18290 + }, + { + "epoch": 2.984582755526552, + "grad_norm": 0.00042176389251835644, + "learning_rate": 0.17003825527217903, + "loss": 0.3033, + "num_input_tokens_seen": 38221136, + "step": 18295 + }, + { + "epoch": 2.985398482747369, + "grad_norm": 0.00027240501367487013, + "learning_rate": 0.16997987683620377, + "loss": 0.3121, + "num_input_tokens_seen": 38230800, + "step": 18300 + }, + { + "epoch": 2.9862142099681868, + "grad_norm": 0.0002220288588432595, + "learning_rate": 0.16992149531908043, + "loss": 0.3429, + "num_input_tokens_seen": 38240128, + "step": 18305 + }, + { + "epoch": 2.987029937189004, + "grad_norm": 0.00047619006363675, + "learning_rate": 0.16986311072981214, + "loss": 0.3104, + "num_input_tokens_seen": 38251888, + "step": 18310 + }, + { + "epoch": 2.9878456644098215, + "grad_norm": 0.0006784395081922412, + "learning_rate": 0.16980472307740255, + "loss": 0.3391, + "num_input_tokens_seen": 38262384, + "step": 18315 + }, + { + "epoch": 2.9886613916306386, + "grad_norm": 0.0005219795275479555, + "learning_rate": 0.1697463323708558, + "loss": 0.311, + "num_input_tokens_seen": 38270528, + "step": 18320 + }, + { + "epoch": 2.989477118851456, + "grad_norm": 0.00028159006615169346, + "learning_rate": 0.16968793861917641, + "loss": 0.3693, + "num_input_tokens_seen": 38282320, + "step": 18325 + }, + { + "epoch": 2.9902928460722733, + "grad_norm": 0.000503373856190592, + "learning_rate": 0.16962954183136952, + "loss": 0.3372, + "num_input_tokens_seen": 38293984, + "step": 18330 + }, + { + "epoch": 2.991108573293091, + "grad_norm": 0.0004526688135229051, + "learning_rate": 0.16957114201644058, + "loss": 0.3132, + "num_input_tokens_seen": 38304048, + "step": 18335 + }, + { + "epoch": 2.991924300513908, + "grad_norm": 0.0005070853512734175, + "learning_rate": 0.16951273918339563, + "loss": 0.3128, + "num_input_tokens_seen": 38314816, + "step": 18340 + }, + { + "epoch": 2.9927400277347256, + "grad_norm": 0.0007993340841494501, + "learning_rate": 0.16945433334124105, + "loss": 0.3137, + "num_input_tokens_seen": 38325424, + "step": 18345 + }, + { + "epoch": 2.9935557549555427, + "grad_norm": 0.00026462896494194865, + "learning_rate": 0.1693959244989838, + "loss": 0.3226, + "num_input_tokens_seen": 38335776, + "step": 18350 + }, + { + "epoch": 2.9943714821763603, + "grad_norm": 0.000406573322834447, + "learning_rate": 0.16933751266563127, + "loss": 0.3383, + "num_input_tokens_seen": 38346608, + "step": 18355 + }, + { + "epoch": 2.9951872093971774, + "grad_norm": 0.00039546508924104273, + "learning_rate": 0.16927909785019118, + "loss": 0.3293, + "num_input_tokens_seen": 38357200, + "step": 18360 + }, + { + "epoch": 2.996002936617995, + "grad_norm": 0.00043273690971545875, + "learning_rate": 0.169220680061672, + "loss": 0.338, + "num_input_tokens_seen": 38367104, + "step": 18365 + }, + { + "epoch": 2.9968186638388126, + "grad_norm": 0.00045721899368800223, + "learning_rate": 0.16916225930908244, + "loss": 0.3044, + "num_input_tokens_seen": 38377984, + "step": 18370 + }, + { + "epoch": 2.9976343910596297, + "grad_norm": 0.0007624165154993534, + "learning_rate": 0.16910383560143163, + "loss": 0.3392, + "num_input_tokens_seen": 38387936, + "step": 18375 + }, + { + "epoch": 2.998450118280447, + "grad_norm": 0.0004064014065079391, + "learning_rate": 0.16904540894772935, + "loss": 0.3577, + "num_input_tokens_seen": 38397024, + "step": 18380 + }, + { + "epoch": 2.9992658455012644, + "grad_norm": 0.0003887888742610812, + "learning_rate": 0.16898697935698562, + "loss": 0.3271, + "num_input_tokens_seen": 38405936, + "step": 18385 + }, + { + "epoch": 3.0, + "grad_norm": 0.000313139142235741, + "learning_rate": 0.1689285468382111, + "loss": 0.3388, + "num_input_tokens_seen": 38413760, + "step": 18390 + }, + { + "epoch": 3.0008157272208176, + "grad_norm": 0.0006595521699637175, + "learning_rate": 0.16887011140041677, + "loss": 0.3273, + "num_input_tokens_seen": 38423248, + "step": 18395 + }, + { + "epoch": 3.0016314544416347, + "grad_norm": 0.00031788318301551044, + "learning_rate": 0.1688116730526141, + "loss": 0.2723, + "num_input_tokens_seen": 38435312, + "step": 18400 + }, + { + "epoch": 3.0016314544416347, + "eval_loss": 0.3174293041229248, + "eval_runtime": 155.9367, + "eval_samples_per_second": 17.475, + "eval_steps_per_second": 8.741, + "num_input_tokens_seen": 38435312, + "step": 18400 + }, + { + "epoch": 3.0024471816624523, + "grad_norm": 0.0002888654125854373, + "learning_rate": 0.1687532318038151, + "loss": 0.3209, + "num_input_tokens_seen": 38446352, + "step": 18405 + }, + { + "epoch": 3.0032629088832694, + "grad_norm": 0.00034232690813951194, + "learning_rate": 0.16869478766303206, + "loss": 0.2709, + "num_input_tokens_seen": 38455520, + "step": 18410 + }, + { + "epoch": 3.004078636104087, + "grad_norm": 0.00035412711440585554, + "learning_rate": 0.16863634063927788, + "loss": 0.3143, + "num_input_tokens_seen": 38467056, + "step": 18415 + }, + { + "epoch": 3.004894363324904, + "grad_norm": 0.0005151231307536364, + "learning_rate": 0.16857789074156568, + "loss": 0.3002, + "num_input_tokens_seen": 38476688, + "step": 18420 + }, + { + "epoch": 3.0057100905457217, + "grad_norm": 0.0008949890616349876, + "learning_rate": 0.16851943797890928, + "loss": 0.3321, + "num_input_tokens_seen": 38487936, + "step": 18425 + }, + { + "epoch": 3.006525817766539, + "grad_norm": 0.0008114202064462006, + "learning_rate": 0.16846098236032284, + "loss": 0.3211, + "num_input_tokens_seen": 38497984, + "step": 18430 + }, + { + "epoch": 3.0073415449873564, + "grad_norm": 0.0007253034273162484, + "learning_rate": 0.16840252389482097, + "loss": 0.2727, + "num_input_tokens_seen": 38508944, + "step": 18435 + }, + { + "epoch": 3.0081572722081735, + "grad_norm": 0.00038874385063536465, + "learning_rate": 0.16834406259141857, + "loss": 0.3509, + "num_input_tokens_seen": 38519808, + "step": 18440 + }, + { + "epoch": 3.008972999428991, + "grad_norm": 0.0005544577143155038, + "learning_rate": 0.16828559845913124, + "loss": 0.3213, + "num_input_tokens_seen": 38530928, + "step": 18445 + }, + { + "epoch": 3.0097887266498082, + "grad_norm": 0.0003757265512831509, + "learning_rate": 0.16822713150697488, + "loss": 0.2962, + "num_input_tokens_seen": 38542160, + "step": 18450 + }, + { + "epoch": 3.010604453870626, + "grad_norm": 0.0003714165068231523, + "learning_rate": 0.16816866174396575, + "loss": 0.4017, + "num_input_tokens_seen": 38554224, + "step": 18455 + }, + { + "epoch": 3.011420181091443, + "grad_norm": 0.00019574942416511476, + "learning_rate": 0.16811018917912057, + "loss": 0.3038, + "num_input_tokens_seen": 38565952, + "step": 18460 + }, + { + "epoch": 3.0122359083122605, + "grad_norm": 0.0002877725346479565, + "learning_rate": 0.16805171382145673, + "loss": 0.3686, + "num_input_tokens_seen": 38576800, + "step": 18465 + }, + { + "epoch": 3.0130516355330776, + "grad_norm": 0.00027531865634955466, + "learning_rate": 0.16799323567999175, + "loss": 0.2733, + "num_input_tokens_seen": 38587584, + "step": 18470 + }, + { + "epoch": 3.013867362753895, + "grad_norm": 0.00020670446974691004, + "learning_rate": 0.16793475476374367, + "loss": 0.3484, + "num_input_tokens_seen": 38597520, + "step": 18475 + }, + { + "epoch": 3.0146830899747123, + "grad_norm": 0.000401382683776319, + "learning_rate": 0.1678762710817311, + "loss": 0.3064, + "num_input_tokens_seen": 38608288, + "step": 18480 + }, + { + "epoch": 3.01549881719553, + "grad_norm": 0.00033096791594289243, + "learning_rate": 0.1678177846429728, + "loss": 0.3167, + "num_input_tokens_seen": 38617328, + "step": 18485 + }, + { + "epoch": 3.016314544416347, + "grad_norm": 0.00038292192039079964, + "learning_rate": 0.16775929545648827, + "loss": 0.2968, + "num_input_tokens_seen": 38627536, + "step": 18490 + }, + { + "epoch": 3.0171302716371646, + "grad_norm": 0.00031424229382537305, + "learning_rate": 0.16770080353129715, + "loss": 0.2982, + "num_input_tokens_seen": 38638480, + "step": 18495 + }, + { + "epoch": 3.0179459988579818, + "grad_norm": 0.0004577132349368185, + "learning_rate": 0.16764230887641968, + "loss": 0.2859, + "num_input_tokens_seen": 38649872, + "step": 18500 + }, + { + "epoch": 3.0187617260787993, + "grad_norm": 0.00039415349601767957, + "learning_rate": 0.1675838115008765, + "loss": 0.3561, + "num_input_tokens_seen": 38660688, + "step": 18505 + }, + { + "epoch": 3.0195774532996165, + "grad_norm": 0.0002561057626735419, + "learning_rate": 0.1675253114136886, + "loss": 0.2889, + "num_input_tokens_seen": 38672800, + "step": 18510 + }, + { + "epoch": 3.020393180520434, + "grad_norm": 0.0003745905414689332, + "learning_rate": 0.16746680862387747, + "loss": 0.396, + "num_input_tokens_seen": 38684128, + "step": 18515 + }, + { + "epoch": 3.021208907741251, + "grad_norm": 0.00033842475386336446, + "learning_rate": 0.16740830314046493, + "loss": 0.2854, + "num_input_tokens_seen": 38694736, + "step": 18520 + }, + { + "epoch": 3.0220246349620687, + "grad_norm": 0.0007092177984304726, + "learning_rate": 0.1673497949724733, + "loss": 0.3001, + "num_input_tokens_seen": 38706064, + "step": 18525 + }, + { + "epoch": 3.022840362182886, + "grad_norm": 0.00041231809882447124, + "learning_rate": 0.16729128412892522, + "loss": 0.3072, + "num_input_tokens_seen": 38717632, + "step": 18530 + }, + { + "epoch": 3.0236560894037035, + "grad_norm": 0.0004480659554246813, + "learning_rate": 0.16723277061884384, + "loss": 0.2951, + "num_input_tokens_seen": 38728528, + "step": 18535 + }, + { + "epoch": 3.0244718166245206, + "grad_norm": 0.0003416385152377188, + "learning_rate": 0.16717425445125267, + "loss": 0.3513, + "num_input_tokens_seen": 38739536, + "step": 18540 + }, + { + "epoch": 3.025287543845338, + "grad_norm": 0.0009759812382981181, + "learning_rate": 0.16711573563517565, + "loss": 0.3164, + "num_input_tokens_seen": 38750880, + "step": 18545 + }, + { + "epoch": 3.0261032710661553, + "grad_norm": 0.0006627426482737064, + "learning_rate": 0.1670572141796371, + "loss": 0.3643, + "num_input_tokens_seen": 38760576, + "step": 18550 + }, + { + "epoch": 3.026918998286973, + "grad_norm": 0.0002891905605792999, + "learning_rate": 0.16699869009366175, + "loss": 0.3335, + "num_input_tokens_seen": 38771152, + "step": 18555 + }, + { + "epoch": 3.02773472550779, + "grad_norm": 0.0005891499458812177, + "learning_rate": 0.1669401633862748, + "loss": 0.3333, + "num_input_tokens_seen": 38780672, + "step": 18560 + }, + { + "epoch": 3.0285504527286076, + "grad_norm": 0.0004115173651371151, + "learning_rate": 0.16688163406650178, + "loss": 0.3345, + "num_input_tokens_seen": 38789696, + "step": 18565 + }, + { + "epoch": 3.0293661799494247, + "grad_norm": 0.00032121664844453335, + "learning_rate": 0.1668231021433686, + "loss": 0.2311, + "num_input_tokens_seen": 38801872, + "step": 18570 + }, + { + "epoch": 3.0301819071702423, + "grad_norm": 0.000561719061806798, + "learning_rate": 0.1667645676259017, + "loss": 0.2578, + "num_input_tokens_seen": 38813200, + "step": 18575 + }, + { + "epoch": 3.03099763439106, + "grad_norm": 0.0006849258788861334, + "learning_rate": 0.1667060305231277, + "loss": 0.2755, + "num_input_tokens_seen": 38824160, + "step": 18580 + }, + { + "epoch": 3.031813361611877, + "grad_norm": 0.0005560816498473287, + "learning_rate": 0.16664749084407396, + "loss": 0.3757, + "num_input_tokens_seen": 38835856, + "step": 18585 + }, + { + "epoch": 3.0326290888326946, + "grad_norm": 0.0003768438473343849, + "learning_rate": 0.16658894859776788, + "loss": 0.2014, + "num_input_tokens_seen": 38846672, + "step": 18590 + }, + { + "epoch": 3.0334448160535117, + "grad_norm": 0.0004787658399436623, + "learning_rate": 0.16653040379323752, + "loss": 0.3657, + "num_input_tokens_seen": 38856576, + "step": 18595 + }, + { + "epoch": 3.0342605432743293, + "grad_norm": 0.00043427475611679256, + "learning_rate": 0.16647185643951107, + "loss": 0.2748, + "num_input_tokens_seen": 38869040, + "step": 18600 + }, + { + "epoch": 3.0342605432743293, + "eval_loss": 0.3171112537384033, + "eval_runtime": 156.0946, + "eval_samples_per_second": 17.457, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 38869040, + "step": 18600 + }, + { + "epoch": 3.0350762704951464, + "grad_norm": 0.0007742477464489639, + "learning_rate": 0.1664133065456174, + "loss": 0.3416, + "num_input_tokens_seen": 38879216, + "step": 18605 + }, + { + "epoch": 3.035891997715964, + "grad_norm": 0.00044989967136643827, + "learning_rate": 0.1663547541205856, + "loss": 0.3045, + "num_input_tokens_seen": 38890208, + "step": 18610 + }, + { + "epoch": 3.036707724936781, + "grad_norm": 0.0006649309070780873, + "learning_rate": 0.16629619917344518, + "loss": 0.3509, + "num_input_tokens_seen": 38900800, + "step": 18615 + }, + { + "epoch": 3.0375234521575987, + "grad_norm": 0.0003501271130517125, + "learning_rate": 0.16623764171322605, + "loss": 0.3441, + "num_input_tokens_seen": 38912096, + "step": 18620 + }, + { + "epoch": 3.038339179378416, + "grad_norm": 0.0004447241371963173, + "learning_rate": 0.1661790817489585, + "loss": 0.3393, + "num_input_tokens_seen": 38922192, + "step": 18625 + }, + { + "epoch": 3.0391549065992334, + "grad_norm": 0.0003575164591893554, + "learning_rate": 0.16612051928967328, + "loss": 0.3697, + "num_input_tokens_seen": 38932528, + "step": 18630 + }, + { + "epoch": 3.0399706338200505, + "grad_norm": 0.00024675627355463803, + "learning_rate": 0.16606195434440138, + "loss": 0.2697, + "num_input_tokens_seen": 38943424, + "step": 18635 + }, + { + "epoch": 3.040786361040868, + "grad_norm": 0.00048616155982017517, + "learning_rate": 0.16600338692217426, + "loss": 0.3332, + "num_input_tokens_seen": 38953312, + "step": 18640 + }, + { + "epoch": 3.041602088261685, + "grad_norm": 0.0007577137439511716, + "learning_rate": 0.16594481703202374, + "loss": 0.3062, + "num_input_tokens_seen": 38963648, + "step": 18645 + }, + { + "epoch": 3.042417815482503, + "grad_norm": 0.00041997016523964703, + "learning_rate": 0.1658862446829821, + "loss": 0.2696, + "num_input_tokens_seen": 38975440, + "step": 18650 + }, + { + "epoch": 3.04323354270332, + "grad_norm": 0.0005715566221624613, + "learning_rate": 0.16582766988408187, + "loss": 0.2957, + "num_input_tokens_seen": 38985712, + "step": 18655 + }, + { + "epoch": 3.0440492699241375, + "grad_norm": 0.0009312601177953184, + "learning_rate": 0.16576909264435608, + "loss": 0.3349, + "num_input_tokens_seen": 38996432, + "step": 18660 + }, + { + "epoch": 3.0448649971449546, + "grad_norm": 0.00019619098748080432, + "learning_rate": 0.16571051297283798, + "loss": 0.3175, + "num_input_tokens_seen": 39006176, + "step": 18665 + }, + { + "epoch": 3.045680724365772, + "grad_norm": 0.00038556120125576854, + "learning_rate": 0.16565193087856137, + "loss": 0.2787, + "num_input_tokens_seen": 39016032, + "step": 18670 + }, + { + "epoch": 3.0464964515865893, + "grad_norm": 0.0004255737876519561, + "learning_rate": 0.16559334637056033, + "loss": 0.2875, + "num_input_tokens_seen": 39028112, + "step": 18675 + }, + { + "epoch": 3.047312178807407, + "grad_norm": 0.00028882434708066285, + "learning_rate": 0.16553475945786933, + "loss": 0.2958, + "num_input_tokens_seen": 39039104, + "step": 18680 + }, + { + "epoch": 3.048127906028224, + "grad_norm": 0.0003480685409158468, + "learning_rate": 0.16547617014952318, + "loss": 0.3323, + "num_input_tokens_seen": 39050944, + "step": 18685 + }, + { + "epoch": 3.0489436332490416, + "grad_norm": 0.0004127322172280401, + "learning_rate": 0.1654175784545571, + "loss": 0.31, + "num_input_tokens_seen": 39062608, + "step": 18690 + }, + { + "epoch": 3.0497593604698587, + "grad_norm": 0.00040460051968693733, + "learning_rate": 0.1653589843820067, + "loss": 0.2311, + "num_input_tokens_seen": 39073088, + "step": 18695 + }, + { + "epoch": 3.0505750876906763, + "grad_norm": 0.0006879012216813862, + "learning_rate": 0.1653003879409079, + "loss": 0.342, + "num_input_tokens_seen": 39082720, + "step": 18700 + }, + { + "epoch": 3.0513908149114934, + "grad_norm": 0.0007521272054873407, + "learning_rate": 0.165241789140297, + "loss": 0.3327, + "num_input_tokens_seen": 39094272, + "step": 18705 + }, + { + "epoch": 3.052206542132311, + "grad_norm": 0.0006431689253076911, + "learning_rate": 0.16518318798921064, + "loss": 0.507, + "num_input_tokens_seen": 39104816, + "step": 18710 + }, + { + "epoch": 3.053022269353128, + "grad_norm": 0.0004837385204154998, + "learning_rate": 0.16512458449668593, + "loss": 0.3475, + "num_input_tokens_seen": 39114880, + "step": 18715 + }, + { + "epoch": 3.0538379965739457, + "grad_norm": 0.0005178381106816232, + "learning_rate": 0.1650659786717602, + "loss": 0.3148, + "num_input_tokens_seen": 39126880, + "step": 18720 + }, + { + "epoch": 3.054653723794763, + "grad_norm": 0.000386690313462168, + "learning_rate": 0.1650073705234712, + "loss": 0.3027, + "num_input_tokens_seen": 39136016, + "step": 18725 + }, + { + "epoch": 3.0554694510155804, + "grad_norm": 0.0008809607243165374, + "learning_rate": 0.16494876006085712, + "loss": 0.2953, + "num_input_tokens_seen": 39146688, + "step": 18730 + }, + { + "epoch": 3.0562851782363976, + "grad_norm": 0.0006010729121044278, + "learning_rate": 0.16489014729295634, + "loss": 0.2761, + "num_input_tokens_seen": 39156928, + "step": 18735 + }, + { + "epoch": 3.057100905457215, + "grad_norm": 0.0005260596517473459, + "learning_rate": 0.16483153222880775, + "loss": 0.3176, + "num_input_tokens_seen": 39168096, + "step": 18740 + }, + { + "epoch": 3.0579166326780323, + "grad_norm": 0.0004400604811962694, + "learning_rate": 0.16477291487745052, + "loss": 0.2844, + "num_input_tokens_seen": 39178272, + "step": 18745 + }, + { + "epoch": 3.05873235989885, + "grad_norm": 0.0003942506154999137, + "learning_rate": 0.16471429524792416, + "loss": 0.3139, + "num_input_tokens_seen": 39188880, + "step": 18750 + }, + { + "epoch": 3.059548087119667, + "grad_norm": 0.0004929438582621515, + "learning_rate": 0.16465567334926856, + "loss": 0.2891, + "num_input_tokens_seen": 39199712, + "step": 18755 + }, + { + "epoch": 3.0603638143404845, + "grad_norm": 0.0006461934535764158, + "learning_rate": 0.16459704919052395, + "loss": 0.2867, + "num_input_tokens_seen": 39210928, + "step": 18760 + }, + { + "epoch": 3.0611795415613017, + "grad_norm": 0.00045745985698886216, + "learning_rate": 0.16453842278073086, + "loss": 0.2841, + "num_input_tokens_seen": 39221392, + "step": 18765 + }, + { + "epoch": 3.0619952687821193, + "grad_norm": 0.00042979110730811954, + "learning_rate": 0.16447979412893038, + "loss": 0.3506, + "num_input_tokens_seen": 39230624, + "step": 18770 + }, + { + "epoch": 3.062810996002937, + "grad_norm": 0.0003883663157466799, + "learning_rate": 0.16442116324416367, + "loss": 0.2818, + "num_input_tokens_seen": 39241776, + "step": 18775 + }, + { + "epoch": 3.063626723223754, + "grad_norm": 0.0008465160499326885, + "learning_rate": 0.1643625301354723, + "loss": 0.3512, + "num_input_tokens_seen": 39251312, + "step": 18780 + }, + { + "epoch": 3.0644424504445715, + "grad_norm": 0.000387604784918949, + "learning_rate": 0.16430389481189828, + "loss": 0.3008, + "num_input_tokens_seen": 39262512, + "step": 18785 + }, + { + "epoch": 3.0652581776653887, + "grad_norm": 0.0006240197108127177, + "learning_rate": 0.164245257282484, + "loss": 0.3037, + "num_input_tokens_seen": 39272320, + "step": 18790 + }, + { + "epoch": 3.0660739048862062, + "grad_norm": 0.0004592328623402864, + "learning_rate": 0.16418661755627195, + "loss": 0.3819, + "num_input_tokens_seen": 39283888, + "step": 18795 + }, + { + "epoch": 3.0668896321070234, + "grad_norm": 0.0004524483811110258, + "learning_rate": 0.16412797564230527, + "loss": 0.3487, + "num_input_tokens_seen": 39294832, + "step": 18800 + }, + { + "epoch": 3.0668896321070234, + "eval_loss": 0.3222704529762268, + "eval_runtime": 155.9757, + "eval_samples_per_second": 17.471, + "eval_steps_per_second": 8.739, + "num_input_tokens_seen": 39294832, + "step": 18800 + }, + { + "epoch": 3.067705359327841, + "grad_norm": 0.00036097000702284276, + "learning_rate": 0.16406933154962713, + "loss": 0.3257, + "num_input_tokens_seen": 39305520, + "step": 18805 + }, + { + "epoch": 3.068521086548658, + "grad_norm": 0.00044154853094369173, + "learning_rate": 0.16401068528728133, + "loss": 0.3083, + "num_input_tokens_seen": 39314640, + "step": 18810 + }, + { + "epoch": 3.0693368137694756, + "grad_norm": 0.000463344797026366, + "learning_rate": 0.16395203686431173, + "loss": 0.3741, + "num_input_tokens_seen": 39324720, + "step": 18815 + }, + { + "epoch": 3.0701525409902928, + "grad_norm": 0.0004275323008187115, + "learning_rate": 0.16389338628976277, + "loss": 0.3264, + "num_input_tokens_seen": 39335504, + "step": 18820 + }, + { + "epoch": 3.0709682682111104, + "grad_norm": 0.0004669659538194537, + "learning_rate": 0.163834733572679, + "loss": 0.3637, + "num_input_tokens_seen": 39346160, + "step": 18825 + }, + { + "epoch": 3.0717839954319275, + "grad_norm": 0.000284833658952266, + "learning_rate": 0.16377607872210545, + "loss": 0.3436, + "num_input_tokens_seen": 39354576, + "step": 18830 + }, + { + "epoch": 3.072599722652745, + "grad_norm": 0.0005350438877940178, + "learning_rate": 0.16371742174708748, + "loss": 0.3076, + "num_input_tokens_seen": 39363920, + "step": 18835 + }, + { + "epoch": 3.073415449873562, + "grad_norm": 0.0005998249398544431, + "learning_rate": 0.16365876265667065, + "loss": 0.3073, + "num_input_tokens_seen": 39374368, + "step": 18840 + }, + { + "epoch": 3.0742311770943798, + "grad_norm": 0.0002473719941917807, + "learning_rate": 0.163600101459901, + "loss": 0.3312, + "num_input_tokens_seen": 39386048, + "step": 18845 + }, + { + "epoch": 3.075046904315197, + "grad_norm": 0.00028641746030189097, + "learning_rate": 0.16354143816582484, + "loss": 0.3316, + "num_input_tokens_seen": 39397056, + "step": 18850 + }, + { + "epoch": 3.0758626315360145, + "grad_norm": 0.0002896037476602942, + "learning_rate": 0.1634827727834887, + "loss": 0.327, + "num_input_tokens_seen": 39406592, + "step": 18855 + }, + { + "epoch": 3.0766783587568316, + "grad_norm": 0.00048761526704765856, + "learning_rate": 0.16342410532193954, + "loss": 0.3188, + "num_input_tokens_seen": 39416016, + "step": 18860 + }, + { + "epoch": 3.077494085977649, + "grad_norm": 0.0006156627205200493, + "learning_rate": 0.16336543579022464, + "loss": 0.2922, + "num_input_tokens_seen": 39426144, + "step": 18865 + }, + { + "epoch": 3.0783098131984663, + "grad_norm": 0.0004484282690100372, + "learning_rate": 0.16330676419739157, + "loss": 0.2953, + "num_input_tokens_seen": 39436880, + "step": 18870 + }, + { + "epoch": 3.079125540419284, + "grad_norm": 0.0005745477392338216, + "learning_rate": 0.1632480905524883, + "loss": 0.3098, + "num_input_tokens_seen": 39447856, + "step": 18875 + }, + { + "epoch": 3.079941267640101, + "grad_norm": 0.0004825856303796172, + "learning_rate": 0.16318941486456293, + "loss": 0.2847, + "num_input_tokens_seen": 39456720, + "step": 18880 + }, + { + "epoch": 3.0807569948609186, + "grad_norm": 0.00037000628071837127, + "learning_rate": 0.16313073714266405, + "loss": 0.249, + "num_input_tokens_seen": 39466992, + "step": 18885 + }, + { + "epoch": 3.0815727220817357, + "grad_norm": 0.0005863634869456291, + "learning_rate": 0.16307205739584052, + "loss": 0.2562, + "num_input_tokens_seen": 39477488, + "step": 18890 + }, + { + "epoch": 3.0823884493025533, + "grad_norm": 0.0008006913121789694, + "learning_rate": 0.16301337563314144, + "loss": 0.3697, + "num_input_tokens_seen": 39486432, + "step": 18895 + }, + { + "epoch": 3.0832041765233704, + "grad_norm": 0.0006044563488103449, + "learning_rate": 0.1629546918636163, + "loss": 0.3392, + "num_input_tokens_seen": 39498208, + "step": 18900 + }, + { + "epoch": 3.084019903744188, + "grad_norm": 0.0005393032333813608, + "learning_rate": 0.16289600609631485, + "loss": 0.3392, + "num_input_tokens_seen": 39508832, + "step": 18905 + }, + { + "epoch": 3.084835630965005, + "grad_norm": 0.001243896665982902, + "learning_rate": 0.16283731834028722, + "loss": 0.2767, + "num_input_tokens_seen": 39519248, + "step": 18910 + }, + { + "epoch": 3.0856513581858227, + "grad_norm": 0.0005052417400293052, + "learning_rate": 0.16277862860458378, + "loss": 0.3116, + "num_input_tokens_seen": 39529936, + "step": 18915 + }, + { + "epoch": 3.08646708540664, + "grad_norm": 0.0007110195001587272, + "learning_rate": 0.16271993689825526, + "loss": 0.3401, + "num_input_tokens_seen": 39540752, + "step": 18920 + }, + { + "epoch": 3.0872828126274574, + "grad_norm": 0.0010813164990395308, + "learning_rate": 0.1626612432303526, + "loss": 0.4231, + "num_input_tokens_seen": 39550272, + "step": 18925 + }, + { + "epoch": 3.0880985398482745, + "grad_norm": 0.0006408400367945433, + "learning_rate": 0.1626025476099271, + "loss": 0.2762, + "num_input_tokens_seen": 39559920, + "step": 18930 + }, + { + "epoch": 3.088914267069092, + "grad_norm": 0.00056412850972265, + "learning_rate": 0.1625438500460304, + "loss": 0.2986, + "num_input_tokens_seen": 39569632, + "step": 18935 + }, + { + "epoch": 3.0897299942899092, + "grad_norm": 0.0004933489253744483, + "learning_rate": 0.16248515054771442, + "loss": 0.3147, + "num_input_tokens_seen": 39580704, + "step": 18940 + }, + { + "epoch": 3.090545721510727, + "grad_norm": 0.0002873046032618731, + "learning_rate": 0.16242644912403123, + "loss": 0.3563, + "num_input_tokens_seen": 39590160, + "step": 18945 + }, + { + "epoch": 3.0913614487315444, + "grad_norm": 0.0004196064255665988, + "learning_rate": 0.1623677457840335, + "loss": 0.3318, + "num_input_tokens_seen": 39600768, + "step": 18950 + }, + { + "epoch": 3.0921771759523615, + "grad_norm": 0.0004636904923245311, + "learning_rate": 0.16230904053677397, + "loss": 0.2673, + "num_input_tokens_seen": 39612256, + "step": 18955 + }, + { + "epoch": 3.092992903173179, + "grad_norm": 0.0010057749459519982, + "learning_rate": 0.16225033339130568, + "loss": 0.4379, + "num_input_tokens_seen": 39621536, + "step": 18960 + }, + { + "epoch": 3.0938086303939962, + "grad_norm": 0.0002784507523756474, + "learning_rate": 0.16219162435668197, + "loss": 0.3196, + "num_input_tokens_seen": 39631904, + "step": 18965 + }, + { + "epoch": 3.094624357614814, + "grad_norm": 0.00039307671249844134, + "learning_rate": 0.16213291344195666, + "loss": 0.2442, + "num_input_tokens_seen": 39641872, + "step": 18970 + }, + { + "epoch": 3.095440084835631, + "grad_norm": 0.00031868048245087266, + "learning_rate": 0.16207420065618358, + "loss": 0.3699, + "num_input_tokens_seen": 39652656, + "step": 18975 + }, + { + "epoch": 3.0962558120564485, + "grad_norm": 0.0006578473839908838, + "learning_rate": 0.16201548600841706, + "loss": 0.3296, + "num_input_tokens_seen": 39664400, + "step": 18980 + }, + { + "epoch": 3.0970715392772656, + "grad_norm": 0.00031959119951352477, + "learning_rate": 0.16195676950771154, + "loss": 0.3552, + "num_input_tokens_seen": 39674944, + "step": 18985 + }, + { + "epoch": 3.097887266498083, + "grad_norm": 0.00037912424886599183, + "learning_rate": 0.16189805116312198, + "loss": 0.2836, + "num_input_tokens_seen": 39685440, + "step": 18990 + }, + { + "epoch": 3.0987029937189003, + "grad_norm": 0.0002592800301499665, + "learning_rate": 0.16183933098370337, + "loss": 0.4206, + "num_input_tokens_seen": 39696608, + "step": 18995 + }, + { + "epoch": 3.099518720939718, + "grad_norm": 0.00020796847820747644, + "learning_rate": 0.16178060897851115, + "loss": 0.3459, + "num_input_tokens_seen": 39706928, + "step": 19000 + }, + { + "epoch": 3.099518720939718, + "eval_loss": 0.31838974356651306, + "eval_runtime": 156.064, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 39706928, + "step": 19000 + }, + { + "epoch": 3.100334448160535, + "grad_norm": 0.0002677248266991228, + "learning_rate": 0.16172188515660096, + "loss": 0.2744, + "num_input_tokens_seen": 39717904, + "step": 19005 + }, + { + "epoch": 3.1011501753813526, + "grad_norm": 0.000265840208157897, + "learning_rate": 0.16166315952702878, + "loss": 0.365, + "num_input_tokens_seen": 39728368, + "step": 19010 + }, + { + "epoch": 3.1019659026021698, + "grad_norm": 0.00020220840815454721, + "learning_rate": 0.16160443209885084, + "loss": 0.3706, + "num_input_tokens_seen": 39738544, + "step": 19015 + }, + { + "epoch": 3.1027816298229873, + "grad_norm": 0.00024346438294742256, + "learning_rate": 0.16154570288112363, + "loss": 0.2899, + "num_input_tokens_seen": 39748848, + "step": 19020 + }, + { + "epoch": 3.1035973570438045, + "grad_norm": 0.0005922979325987399, + "learning_rate": 0.16148697188290395, + "loss": 0.2832, + "num_input_tokens_seen": 39759024, + "step": 19025 + }, + { + "epoch": 3.104413084264622, + "grad_norm": 0.0004210302431602031, + "learning_rate": 0.16142823911324888, + "loss": 0.2632, + "num_input_tokens_seen": 39770576, + "step": 19030 + }, + { + "epoch": 3.105228811485439, + "grad_norm": 0.00023029837757349014, + "learning_rate": 0.16136950458121568, + "loss": 0.2955, + "num_input_tokens_seen": 39780272, + "step": 19035 + }, + { + "epoch": 3.1060445387062567, + "grad_norm": 0.00046236737398430705, + "learning_rate": 0.16131076829586205, + "loss": 0.32, + "num_input_tokens_seen": 39793072, + "step": 19040 + }, + { + "epoch": 3.106860265927074, + "grad_norm": 0.00038004416273906827, + "learning_rate": 0.1612520302662457, + "loss": 0.3294, + "num_input_tokens_seen": 39803264, + "step": 19045 + }, + { + "epoch": 3.1076759931478914, + "grad_norm": 0.00023636213154532015, + "learning_rate": 0.16119329050142497, + "loss": 0.3046, + "num_input_tokens_seen": 39813824, + "step": 19050 + }, + { + "epoch": 3.1084917203687086, + "grad_norm": 0.0007706411997787654, + "learning_rate": 0.16113454901045818, + "loss": 0.3205, + "num_input_tokens_seen": 39822560, + "step": 19055 + }, + { + "epoch": 3.109307447589526, + "grad_norm": 0.000340128579409793, + "learning_rate": 0.16107580580240397, + "loss": 0.3735, + "num_input_tokens_seen": 39832432, + "step": 19060 + }, + { + "epoch": 3.1101231748103433, + "grad_norm": 0.0004636592057067901, + "learning_rate": 0.16101706088632134, + "loss": 0.3672, + "num_input_tokens_seen": 39843424, + "step": 19065 + }, + { + "epoch": 3.110938902031161, + "grad_norm": 0.0002470928302500397, + "learning_rate": 0.16095831427126947, + "loss": 0.3203, + "num_input_tokens_seen": 39852128, + "step": 19070 + }, + { + "epoch": 3.111754629251978, + "grad_norm": 0.0006092395051382482, + "learning_rate": 0.16089956596630783, + "loss": 0.2702, + "num_input_tokens_seen": 39863744, + "step": 19075 + }, + { + "epoch": 3.1125703564727956, + "grad_norm": 0.00023938289086800069, + "learning_rate": 0.16084081598049618, + "loss": 0.3569, + "num_input_tokens_seen": 39873584, + "step": 19080 + }, + { + "epoch": 3.1133860836936127, + "grad_norm": 0.00025590157019905746, + "learning_rate": 0.1607820643228944, + "loss": 0.3112, + "num_input_tokens_seen": 39882704, + "step": 19085 + }, + { + "epoch": 3.1142018109144303, + "grad_norm": 0.000462969095679, + "learning_rate": 0.16072331100256285, + "loss": 0.2366, + "num_input_tokens_seen": 39893568, + "step": 19090 + }, + { + "epoch": 3.1150175381352474, + "grad_norm": 0.0006350572220981121, + "learning_rate": 0.16066455602856197, + "loss": 0.3443, + "num_input_tokens_seen": 39905648, + "step": 19095 + }, + { + "epoch": 3.115833265356065, + "grad_norm": 0.0003344239084981382, + "learning_rate": 0.16060579940995257, + "loss": 0.2717, + "num_input_tokens_seen": 39914608, + "step": 19100 + }, + { + "epoch": 3.116648992576882, + "grad_norm": 0.0003633512824308127, + "learning_rate": 0.16054704115579557, + "loss": 0.3137, + "num_input_tokens_seen": 39925312, + "step": 19105 + }, + { + "epoch": 3.1174647197976997, + "grad_norm": 0.00031355046667158604, + "learning_rate": 0.1604882812751523, + "loss": 0.3004, + "num_input_tokens_seen": 39934672, + "step": 19110 + }, + { + "epoch": 3.118280447018517, + "grad_norm": 0.00047063021338544786, + "learning_rate": 0.16042951977708425, + "loss": 0.3299, + "num_input_tokens_seen": 39945040, + "step": 19115 + }, + { + "epoch": 3.1190961742393344, + "grad_norm": 0.00027512534870766103, + "learning_rate": 0.16037075667065318, + "loss": 0.277, + "num_input_tokens_seen": 39955840, + "step": 19120 + }, + { + "epoch": 3.1199119014601515, + "grad_norm": 0.0003326097794342786, + "learning_rate": 0.1603119919649211, + "loss": 0.3149, + "num_input_tokens_seen": 39966688, + "step": 19125 + }, + { + "epoch": 3.120727628680969, + "grad_norm": 0.0002251083787996322, + "learning_rate": 0.16025322566895028, + "loss": 0.2832, + "num_input_tokens_seen": 39977392, + "step": 19130 + }, + { + "epoch": 3.121543355901786, + "grad_norm": 0.0003538521414157003, + "learning_rate": 0.16019445779180322, + "loss": 0.2702, + "num_input_tokens_seen": 39987968, + "step": 19135 + }, + { + "epoch": 3.122359083122604, + "grad_norm": 0.0006169155240058899, + "learning_rate": 0.16013568834254271, + "loss": 0.2991, + "num_input_tokens_seen": 39998272, + "step": 19140 + }, + { + "epoch": 3.1231748103434214, + "grad_norm": 0.0003531251277308911, + "learning_rate": 0.1600769173302316, + "loss": 0.3468, + "num_input_tokens_seen": 40009344, + "step": 19145 + }, + { + "epoch": 3.1239905375642385, + "grad_norm": 0.0005792954470962286, + "learning_rate": 0.16001814476393322, + "loss": 0.4337, + "num_input_tokens_seen": 40019200, + "step": 19150 + }, + { + "epoch": 3.124806264785056, + "grad_norm": 0.00042472800123505294, + "learning_rate": 0.15995937065271104, + "loss": 0.3565, + "num_input_tokens_seen": 40028880, + "step": 19155 + }, + { + "epoch": 3.125621992005873, + "grad_norm": 0.0003121462359558791, + "learning_rate": 0.15990059500562873, + "loss": 0.3044, + "num_input_tokens_seen": 40039824, + "step": 19160 + }, + { + "epoch": 3.126437719226691, + "grad_norm": 0.0002741119242273271, + "learning_rate": 0.15984181783175025, + "loss": 0.304, + "num_input_tokens_seen": 40048944, + "step": 19165 + }, + { + "epoch": 3.127253446447508, + "grad_norm": 0.00029821612406522036, + "learning_rate": 0.1597830391401398, + "loss": 0.2977, + "num_input_tokens_seen": 40059312, + "step": 19170 + }, + { + "epoch": 3.1280691736683255, + "grad_norm": 0.0005865884013473988, + "learning_rate": 0.15972425893986178, + "loss": 0.3421, + "num_input_tokens_seen": 40070928, + "step": 19175 + }, + { + "epoch": 3.1288849008891426, + "grad_norm": 0.0003548384120222181, + "learning_rate": 0.15966547723998084, + "loss": 0.2838, + "num_input_tokens_seen": 40080368, + "step": 19180 + }, + { + "epoch": 3.12970062810996, + "grad_norm": 0.000329698552377522, + "learning_rate": 0.15960669404956176, + "loss": 0.3757, + "num_input_tokens_seen": 40091376, + "step": 19185 + }, + { + "epoch": 3.1305163553307773, + "grad_norm": 0.00030219554901123047, + "learning_rate": 0.1595479093776698, + "loss": 0.3173, + "num_input_tokens_seen": 40101408, + "step": 19190 + }, + { + "epoch": 3.131332082551595, + "grad_norm": 0.00045206115464679897, + "learning_rate": 0.15948912323337022, + "loss": 0.3088, + "num_input_tokens_seen": 40112880, + "step": 19195 + }, + { + "epoch": 3.132147809772412, + "grad_norm": 0.00024859816767275333, + "learning_rate": 0.1594303356257286, + "loss": 0.3048, + "num_input_tokens_seen": 40121472, + "step": 19200 + }, + { + "epoch": 3.132147809772412, + "eval_loss": 0.31717145442962646, + "eval_runtime": 156.1619, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 40121472, + "step": 19200 + }, + { + "epoch": 3.1329635369932296, + "grad_norm": 0.0003464036271907389, + "learning_rate": 0.15937154656381072, + "loss": 0.3524, + "num_input_tokens_seen": 40131728, + "step": 19205 + }, + { + "epoch": 3.1337792642140467, + "grad_norm": 0.0012415219098329544, + "learning_rate": 0.15931275605668258, + "loss": 0.3599, + "num_input_tokens_seen": 40141408, + "step": 19210 + }, + { + "epoch": 3.1345949914348643, + "grad_norm": 0.0003178457554895431, + "learning_rate": 0.1592539641134104, + "loss": 0.321, + "num_input_tokens_seen": 40150880, + "step": 19215 + }, + { + "epoch": 3.1354107186556814, + "grad_norm": 0.0005508855683729053, + "learning_rate": 0.1591951707430607, + "loss": 0.2812, + "num_input_tokens_seen": 40160608, + "step": 19220 + }, + { + "epoch": 3.136226445876499, + "grad_norm": 0.0005190112278796732, + "learning_rate": 0.15913637595470007, + "loss": 0.3132, + "num_input_tokens_seen": 40170480, + "step": 19225 + }, + { + "epoch": 3.137042173097316, + "grad_norm": 0.0002959250414278358, + "learning_rate": 0.15907757975739548, + "loss": 0.3532, + "num_input_tokens_seen": 40181536, + "step": 19230 + }, + { + "epoch": 3.1378579003181337, + "grad_norm": 0.00033186873770318925, + "learning_rate": 0.159018782160214, + "loss": 0.2995, + "num_input_tokens_seen": 40191952, + "step": 19235 + }, + { + "epoch": 3.138673627538951, + "grad_norm": 0.00031678512459620833, + "learning_rate": 0.158959983172223, + "loss": 0.3164, + "num_input_tokens_seen": 40201808, + "step": 19240 + }, + { + "epoch": 3.1394893547597684, + "grad_norm": 0.0005786349647678435, + "learning_rate": 0.15890118280249, + "loss": 0.3385, + "num_input_tokens_seen": 40214720, + "step": 19245 + }, + { + "epoch": 3.1403050819805856, + "grad_norm": 0.00033864815486595035, + "learning_rate": 0.15884238106008275, + "loss": 0.3133, + "num_input_tokens_seen": 40225920, + "step": 19250 + }, + { + "epoch": 3.141120809201403, + "grad_norm": 0.000642747909296304, + "learning_rate": 0.15878357795406922, + "loss": 0.3146, + "num_input_tokens_seen": 40236416, + "step": 19255 + }, + { + "epoch": 3.1419365364222203, + "grad_norm": 0.0002998071431647986, + "learning_rate": 0.15872477349351757, + "loss": 0.3217, + "num_input_tokens_seen": 40245776, + "step": 19260 + }, + { + "epoch": 3.142752263643038, + "grad_norm": 0.0005323613877408206, + "learning_rate": 0.15866596768749622, + "loss": 0.3074, + "num_input_tokens_seen": 40256160, + "step": 19265 + }, + { + "epoch": 3.143567990863855, + "grad_norm": 0.0003773793869186193, + "learning_rate": 0.15860716054507373, + "loss": 0.3455, + "num_input_tokens_seen": 40267136, + "step": 19270 + }, + { + "epoch": 3.1443837180846725, + "grad_norm": 0.0007163326372392476, + "learning_rate": 0.1585483520753189, + "loss": 0.3102, + "num_input_tokens_seen": 40277360, + "step": 19275 + }, + { + "epoch": 3.1451994453054897, + "grad_norm": 0.0003728134906850755, + "learning_rate": 0.1584895422873008, + "loss": 0.259, + "num_input_tokens_seen": 40287152, + "step": 19280 + }, + { + "epoch": 3.1460151725263072, + "grad_norm": 0.0005064102006144822, + "learning_rate": 0.1584307311900886, + "loss": 0.3522, + "num_input_tokens_seen": 40297744, + "step": 19285 + }, + { + "epoch": 3.1468308997471244, + "grad_norm": 0.0005526600289158523, + "learning_rate": 0.1583719187927517, + "loss": 0.2761, + "num_input_tokens_seen": 40307808, + "step": 19290 + }, + { + "epoch": 3.147646626967942, + "grad_norm": 0.0005925050936639309, + "learning_rate": 0.15831310510435967, + "loss": 0.3107, + "num_input_tokens_seen": 40319152, + "step": 19295 + }, + { + "epoch": 3.148462354188759, + "grad_norm": 0.0003707563446369022, + "learning_rate": 0.15825429013398243, + "loss": 0.2494, + "num_input_tokens_seen": 40330512, + "step": 19300 + }, + { + "epoch": 3.1492780814095767, + "grad_norm": 0.00041566594154573977, + "learning_rate": 0.15819547389068986, + "loss": 0.3168, + "num_input_tokens_seen": 40341744, + "step": 19305 + }, + { + "epoch": 3.150093808630394, + "grad_norm": 0.00019081399659626186, + "learning_rate": 0.1581366563835522, + "loss": 0.2901, + "num_input_tokens_seen": 40353744, + "step": 19310 + }, + { + "epoch": 3.1509095358512114, + "grad_norm": 0.000516811734996736, + "learning_rate": 0.15807783762163993, + "loss": 0.2537, + "num_input_tokens_seen": 40364288, + "step": 19315 + }, + { + "epoch": 3.151725263072029, + "grad_norm": 0.00021302922687027603, + "learning_rate": 0.15801901761402365, + "loss": 0.3052, + "num_input_tokens_seen": 40375152, + "step": 19320 + }, + { + "epoch": 3.152540990292846, + "grad_norm": 0.0008109085028991103, + "learning_rate": 0.157960196369774, + "loss": 0.3695, + "num_input_tokens_seen": 40384496, + "step": 19325 + }, + { + "epoch": 3.153356717513663, + "grad_norm": 0.0004753133107442409, + "learning_rate": 0.157901373897962, + "loss": 0.3382, + "num_input_tokens_seen": 40394080, + "step": 19330 + }, + { + "epoch": 3.1541724447344808, + "grad_norm": 0.00044839578913524747, + "learning_rate": 0.15784255020765892, + "loss": 0.3177, + "num_input_tokens_seen": 40404240, + "step": 19335 + }, + { + "epoch": 3.1549881719552983, + "grad_norm": 0.0004952616873197258, + "learning_rate": 0.157783725307936, + "loss": 0.3718, + "num_input_tokens_seen": 40415696, + "step": 19340 + }, + { + "epoch": 3.1558038991761155, + "grad_norm": 0.0018029491184279323, + "learning_rate": 0.15772489920786484, + "loss": 0.3856, + "num_input_tokens_seen": 40425376, + "step": 19345 + }, + { + "epoch": 3.156619626396933, + "grad_norm": 0.0004595150821842253, + "learning_rate": 0.15766607191651713, + "loss": 0.2639, + "num_input_tokens_seen": 40435040, + "step": 19350 + }, + { + "epoch": 3.15743535361775, + "grad_norm": 0.0003075501008424908, + "learning_rate": 0.1576072434429648, + "loss": 0.3203, + "num_input_tokens_seen": 40445680, + "step": 19355 + }, + { + "epoch": 3.1582510808385678, + "grad_norm": 0.0005009357118979096, + "learning_rate": 0.15754841379627998, + "loss": 0.2821, + "num_input_tokens_seen": 40457040, + "step": 19360 + }, + { + "epoch": 3.159066808059385, + "grad_norm": 0.0005162059096619487, + "learning_rate": 0.15748958298553484, + "loss": 0.2751, + "num_input_tokens_seen": 40467776, + "step": 19365 + }, + { + "epoch": 3.1598825352802025, + "grad_norm": 0.0003243596584070474, + "learning_rate": 0.1574307510198019, + "loss": 0.3781, + "num_input_tokens_seen": 40478464, + "step": 19370 + }, + { + "epoch": 3.1606982625010196, + "grad_norm": 0.00024243044026661664, + "learning_rate": 0.15737191790815375, + "loss": 0.3299, + "num_input_tokens_seen": 40487760, + "step": 19375 + }, + { + "epoch": 3.161513989721837, + "grad_norm": 0.0002786560507956892, + "learning_rate": 0.15731308365966323, + "loss": 0.408, + "num_input_tokens_seen": 40497584, + "step": 19380 + }, + { + "epoch": 3.1623297169426543, + "grad_norm": 0.0005120671703480184, + "learning_rate": 0.15725424828340331, + "loss": 0.3426, + "num_input_tokens_seen": 40507808, + "step": 19385 + }, + { + "epoch": 3.163145444163472, + "grad_norm": 0.00030126527417451143, + "learning_rate": 0.15719541178844715, + "loss": 0.3416, + "num_input_tokens_seen": 40518256, + "step": 19390 + }, + { + "epoch": 3.163961171384289, + "grad_norm": 0.000773948966525495, + "learning_rate": 0.15713657418386806, + "loss": 0.302, + "num_input_tokens_seen": 40528144, + "step": 19395 + }, + { + "epoch": 3.1647768986051066, + "grad_norm": 0.00033396558137610555, + "learning_rate": 0.15707773547873957, + "loss": 0.373, + "num_input_tokens_seen": 40537792, + "step": 19400 + }, + { + "epoch": 3.1647768986051066, + "eval_loss": 0.3200557827949524, + "eval_runtime": 156.099, + "eval_samples_per_second": 17.457, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 40537792, + "step": 19400 + }, + { + "epoch": 3.1655926258259237, + "grad_norm": 0.000681693316437304, + "learning_rate": 0.1570188956821353, + "loss": 0.3421, + "num_input_tokens_seen": 40548320, + "step": 19405 + }, + { + "epoch": 3.1664083530467413, + "grad_norm": 0.0006075255805626512, + "learning_rate": 0.1569600548031291, + "loss": 0.2799, + "num_input_tokens_seen": 40559536, + "step": 19410 + }, + { + "epoch": 3.1672240802675584, + "grad_norm": 0.00028003103216178715, + "learning_rate": 0.156901212850795, + "loss": 0.2909, + "num_input_tokens_seen": 40571136, + "step": 19415 + }, + { + "epoch": 3.168039807488376, + "grad_norm": 0.0004443876096047461, + "learning_rate": 0.15684236983420716, + "loss": 0.3597, + "num_input_tokens_seen": 40581776, + "step": 19420 + }, + { + "epoch": 3.168855534709193, + "grad_norm": 0.0003070196253247559, + "learning_rate": 0.1567835257624399, + "loss": 0.2723, + "num_input_tokens_seen": 40592656, + "step": 19425 + }, + { + "epoch": 3.1696712619300107, + "grad_norm": 0.00028951570857316256, + "learning_rate": 0.1567246806445677, + "loss": 0.3057, + "num_input_tokens_seen": 40601984, + "step": 19430 + }, + { + "epoch": 3.170486989150828, + "grad_norm": 0.0002939459518529475, + "learning_rate": 0.15666583448966526, + "loss": 0.2956, + "num_input_tokens_seen": 40612368, + "step": 19435 + }, + { + "epoch": 3.1713027163716454, + "grad_norm": 0.0002775135508272797, + "learning_rate": 0.1566069873068074, + "loss": 0.3513, + "num_input_tokens_seen": 40623376, + "step": 19440 + }, + { + "epoch": 3.1721184435924625, + "grad_norm": 0.00018151284893974662, + "learning_rate": 0.156548139105069, + "loss": 0.2883, + "num_input_tokens_seen": 40633616, + "step": 19445 + }, + { + "epoch": 3.17293417081328, + "grad_norm": 0.0003760932304430753, + "learning_rate": 0.15648928989352529, + "loss": 0.2727, + "num_input_tokens_seen": 40644192, + "step": 19450 + }, + { + "epoch": 3.1737498980340972, + "grad_norm": 0.00025030417600646615, + "learning_rate": 0.15643043968125156, + "loss": 0.3817, + "num_input_tokens_seen": 40655584, + "step": 19455 + }, + { + "epoch": 3.174565625254915, + "grad_norm": 0.00034455914283171296, + "learning_rate": 0.15637158847732316, + "loss": 0.3947, + "num_input_tokens_seen": 40664544, + "step": 19460 + }, + { + "epoch": 3.175381352475732, + "grad_norm": 0.0006244547548703849, + "learning_rate": 0.15631273629081582, + "loss": 0.3067, + "num_input_tokens_seen": 40675968, + "step": 19465 + }, + { + "epoch": 3.1761970796965495, + "grad_norm": 0.00030589121161028743, + "learning_rate": 0.15625388313080518, + "loss": 0.2754, + "num_input_tokens_seen": 40686736, + "step": 19470 + }, + { + "epoch": 3.1770128069173666, + "grad_norm": 0.00027418480021879077, + "learning_rate": 0.15619502900636714, + "loss": 0.3208, + "num_input_tokens_seen": 40696656, + "step": 19475 + }, + { + "epoch": 3.1778285341381842, + "grad_norm": 0.0008736494346521795, + "learning_rate": 0.15613617392657783, + "loss": 0.3666, + "num_input_tokens_seen": 40707856, + "step": 19480 + }, + { + "epoch": 3.1786442613590014, + "grad_norm": 0.0002656077267602086, + "learning_rate": 0.15607731790051335, + "loss": 0.3394, + "num_input_tokens_seen": 40718752, + "step": 19485 + }, + { + "epoch": 3.179459988579819, + "grad_norm": 0.0003565521619748324, + "learning_rate": 0.15601846093725008, + "loss": 0.3244, + "num_input_tokens_seen": 40730240, + "step": 19490 + }, + { + "epoch": 3.180275715800636, + "grad_norm": 0.0005045137368142605, + "learning_rate": 0.1559596030458645, + "loss": 0.3033, + "num_input_tokens_seen": 40740416, + "step": 19495 + }, + { + "epoch": 3.1810914430214536, + "grad_norm": 0.0004301271983422339, + "learning_rate": 0.1559007442354333, + "loss": 0.2998, + "num_input_tokens_seen": 40749968, + "step": 19500 + }, + { + "epoch": 3.1819071702422708, + "grad_norm": 0.0002720743359532207, + "learning_rate": 0.15584188451503314, + "loss": 0.4113, + "num_input_tokens_seen": 40759088, + "step": 19505 + }, + { + "epoch": 3.1827228974630883, + "grad_norm": 0.0003032355452887714, + "learning_rate": 0.15578302389374094, + "loss": 0.2981, + "num_input_tokens_seen": 40769504, + "step": 19510 + }, + { + "epoch": 3.183538624683906, + "grad_norm": 0.00021735978953074664, + "learning_rate": 0.1557241623806338, + "loss": 0.2986, + "num_input_tokens_seen": 40780048, + "step": 19515 + }, + { + "epoch": 3.184354351904723, + "grad_norm": 0.000396856659790501, + "learning_rate": 0.15566529998478887, + "loss": 0.2701, + "num_input_tokens_seen": 40790256, + "step": 19520 + }, + { + "epoch": 3.1851700791255406, + "grad_norm": 0.00035589528852142394, + "learning_rate": 0.15560643671528354, + "loss": 0.2765, + "num_input_tokens_seen": 40802192, + "step": 19525 + }, + { + "epoch": 3.1859858063463578, + "grad_norm": 0.000589165254496038, + "learning_rate": 0.15554757258119514, + "loss": 0.344, + "num_input_tokens_seen": 40812880, + "step": 19530 + }, + { + "epoch": 3.1868015335671753, + "grad_norm": 0.00024226360255852342, + "learning_rate": 0.1554887075916014, + "loss": 0.3493, + "num_input_tokens_seen": 40824320, + "step": 19535 + }, + { + "epoch": 3.1876172607879925, + "grad_norm": 0.0004958976642228663, + "learning_rate": 0.15542984175558, + "loss": 0.2712, + "num_input_tokens_seen": 40835888, + "step": 19540 + }, + { + "epoch": 3.18843298800881, + "grad_norm": 0.0003066488716285676, + "learning_rate": 0.1553709750822087, + "loss": 0.2989, + "num_input_tokens_seen": 40847680, + "step": 19545 + }, + { + "epoch": 3.189248715229627, + "grad_norm": 0.0004815817519556731, + "learning_rate": 0.15531210758056554, + "loss": 0.3116, + "num_input_tokens_seen": 40857424, + "step": 19550 + }, + { + "epoch": 3.1900644424504447, + "grad_norm": 0.0004923653323203325, + "learning_rate": 0.15525323925972867, + "loss": 0.2777, + "num_input_tokens_seen": 40868864, + "step": 19555 + }, + { + "epoch": 3.190880169671262, + "grad_norm": 0.0003849011554848403, + "learning_rate": 0.15519437012877627, + "loss": 0.3071, + "num_input_tokens_seen": 40878672, + "step": 19560 + }, + { + "epoch": 3.1916958968920794, + "grad_norm": 0.00019462130148895085, + "learning_rate": 0.15513550019678676, + "loss": 0.3736, + "num_input_tokens_seen": 40888256, + "step": 19565 + }, + { + "epoch": 3.1925116241128966, + "grad_norm": 0.0013062905054539442, + "learning_rate": 0.15507662947283854, + "loss": 0.3568, + "num_input_tokens_seen": 40899104, + "step": 19570 + }, + { + "epoch": 3.193327351333714, + "grad_norm": 0.0005122317234054208, + "learning_rate": 0.15501775796601028, + "loss": 0.2984, + "num_input_tokens_seen": 40909856, + "step": 19575 + }, + { + "epoch": 3.1941430785545313, + "grad_norm": 0.000385764375096187, + "learning_rate": 0.15495888568538066, + "loss": 0.2576, + "num_input_tokens_seen": 40920000, + "step": 19580 + }, + { + "epoch": 3.194958805775349, + "grad_norm": 0.00038755935383960605, + "learning_rate": 0.1549000126400286, + "loss": 0.367, + "num_input_tokens_seen": 40929168, + "step": 19585 + }, + { + "epoch": 3.195774532996166, + "grad_norm": 0.00021642666251864284, + "learning_rate": 0.15484113883903294, + "loss": 0.2935, + "num_input_tokens_seen": 40939328, + "step": 19590 + }, + { + "epoch": 3.1965902602169836, + "grad_norm": 0.0004425928054843098, + "learning_rate": 0.15478226429147288, + "loss": 0.2453, + "num_input_tokens_seen": 40949664, + "step": 19595 + }, + { + "epoch": 3.1974059874378007, + "grad_norm": 0.0003397221153136343, + "learning_rate": 0.15472338900642757, + "loss": 0.3033, + "num_input_tokens_seen": 40958928, + "step": 19600 + }, + { + "epoch": 3.1974059874378007, + "eval_loss": 0.3187048137187958, + "eval_runtime": 155.6365, + "eval_samples_per_second": 17.509, + "eval_steps_per_second": 8.758, + "num_input_tokens_seen": 40958928, + "step": 19600 + }, + { + "epoch": 3.1982217146586183, + "grad_norm": 0.0004506147815845907, + "learning_rate": 0.15466451299297632, + "loss": 0.2575, + "num_input_tokens_seen": 40969952, + "step": 19605 + }, + { + "epoch": 3.1990374418794354, + "grad_norm": 0.00043260862003080547, + "learning_rate": 0.15460563626019852, + "loss": 0.2441, + "num_input_tokens_seen": 40978208, + "step": 19610 + }, + { + "epoch": 3.199853169100253, + "grad_norm": 0.0002641906321514398, + "learning_rate": 0.15454675881717375, + "loss": 0.3558, + "num_input_tokens_seen": 40990016, + "step": 19615 + }, + { + "epoch": 3.20066889632107, + "grad_norm": 0.00043551556882448494, + "learning_rate": 0.1544878806729816, + "loss": 0.3408, + "num_input_tokens_seen": 41000672, + "step": 19620 + }, + { + "epoch": 3.2014846235418877, + "grad_norm": 0.0002372012531850487, + "learning_rate": 0.1544290018367019, + "loss": 0.3051, + "num_input_tokens_seen": 41011200, + "step": 19625 + }, + { + "epoch": 3.202300350762705, + "grad_norm": 0.000490324106067419, + "learning_rate": 0.15437012231741445, + "loss": 0.4242, + "num_input_tokens_seen": 41022320, + "step": 19630 + }, + { + "epoch": 3.2031160779835224, + "grad_norm": 0.00090167042799294, + "learning_rate": 0.1543112421241992, + "loss": 0.3324, + "num_input_tokens_seen": 41031760, + "step": 19635 + }, + { + "epoch": 3.2039318052043395, + "grad_norm": 0.0003486888308543712, + "learning_rate": 0.15425236126613626, + "loss": 0.3039, + "num_input_tokens_seen": 41042448, + "step": 19640 + }, + { + "epoch": 3.204747532425157, + "grad_norm": 0.000512682250700891, + "learning_rate": 0.15419347975230577, + "loss": 0.3124, + "num_input_tokens_seen": 41051344, + "step": 19645 + }, + { + "epoch": 3.205563259645974, + "grad_norm": 0.00038516923086717725, + "learning_rate": 0.154134597591788, + "loss": 0.3762, + "num_input_tokens_seen": 41062000, + "step": 19650 + }, + { + "epoch": 3.206378986866792, + "grad_norm": 0.0002642756444402039, + "learning_rate": 0.1540757147936633, + "loss": 0.3188, + "num_input_tokens_seen": 41070912, + "step": 19655 + }, + { + "epoch": 3.207194714087609, + "grad_norm": 0.00030894356314092875, + "learning_rate": 0.1540168313670122, + "loss": 0.2994, + "num_input_tokens_seen": 41082080, + "step": 19660 + }, + { + "epoch": 3.2080104413084265, + "grad_norm": 0.00036217676824890077, + "learning_rate": 0.1539579473209152, + "loss": 0.2882, + "num_input_tokens_seen": 41092096, + "step": 19665 + }, + { + "epoch": 3.2088261685292436, + "grad_norm": 0.0005831841262988746, + "learning_rate": 0.15389906266445294, + "loss": 0.3318, + "num_input_tokens_seen": 41102816, + "step": 19670 + }, + { + "epoch": 3.209641895750061, + "grad_norm": 0.00036336068296805024, + "learning_rate": 0.15384017740670627, + "loss": 0.3331, + "num_input_tokens_seen": 41112032, + "step": 19675 + }, + { + "epoch": 3.2104576229708783, + "grad_norm": 0.00044789156527258456, + "learning_rate": 0.15378129155675602, + "loss": 0.4059, + "num_input_tokens_seen": 41123952, + "step": 19680 + }, + { + "epoch": 3.211273350191696, + "grad_norm": 0.0005686490330845118, + "learning_rate": 0.15372240512368307, + "loss": 0.3223, + "num_input_tokens_seen": 41133648, + "step": 19685 + }, + { + "epoch": 3.2120890774125135, + "grad_norm": 0.00045372688327915967, + "learning_rate": 0.1536635181165684, + "loss": 0.3473, + "num_input_tokens_seen": 41146128, + "step": 19690 + }, + { + "epoch": 3.2129048046333306, + "grad_norm": 0.0006515560089610517, + "learning_rate": 0.15360463054449328, + "loss": 0.3023, + "num_input_tokens_seen": 41156144, + "step": 19695 + }, + { + "epoch": 3.2137205318541477, + "grad_norm": 0.0003153875295538455, + "learning_rate": 0.1535457424165388, + "loss": 0.3087, + "num_input_tokens_seen": 41165552, + "step": 19700 + }, + { + "epoch": 3.2145362590749653, + "grad_norm": 0.0007426017546094954, + "learning_rate": 0.15348685374178628, + "loss": 0.3159, + "num_input_tokens_seen": 41175968, + "step": 19705 + }, + { + "epoch": 3.215351986295783, + "grad_norm": 0.0002639228478074074, + "learning_rate": 0.1534279645293171, + "loss": 0.3345, + "num_input_tokens_seen": 41187264, + "step": 19710 + }, + { + "epoch": 3.2161677135166, + "grad_norm": 0.00034231229801662266, + "learning_rate": 0.1533690747882127, + "loss": 0.3251, + "num_input_tokens_seen": 41197616, + "step": 19715 + }, + { + "epoch": 3.2169834407374176, + "grad_norm": 0.00046319718239828944, + "learning_rate": 0.15331018452755465, + "loss": 0.2893, + "num_input_tokens_seen": 41208896, + "step": 19720 + }, + { + "epoch": 3.2177991679582347, + "grad_norm": 0.0006041408632881939, + "learning_rate": 0.15325129375642457, + "loss": 0.2842, + "num_input_tokens_seen": 41219552, + "step": 19725 + }, + { + "epoch": 3.2186148951790523, + "grad_norm": 0.0002711009292397648, + "learning_rate": 0.15319240248390406, + "loss": 0.2892, + "num_input_tokens_seen": 41230448, + "step": 19730 + }, + { + "epoch": 3.2194306223998694, + "grad_norm": 0.00040814789826981723, + "learning_rate": 0.153133510719075, + "loss": 0.2917, + "num_input_tokens_seen": 41241104, + "step": 19735 + }, + { + "epoch": 3.220246349620687, + "grad_norm": 0.0004323154571466148, + "learning_rate": 0.15307461847101922, + "loss": 0.2644, + "num_input_tokens_seen": 41253232, + "step": 19740 + }, + { + "epoch": 3.221062076841504, + "grad_norm": 0.0006582319037988782, + "learning_rate": 0.15301572574881864, + "loss": 0.3287, + "num_input_tokens_seen": 41264640, + "step": 19745 + }, + { + "epoch": 3.2218778040623217, + "grad_norm": 0.0015339356614276767, + "learning_rate": 0.15295683256155523, + "loss": 0.3659, + "num_input_tokens_seen": 41275200, + "step": 19750 + }, + { + "epoch": 3.222693531283139, + "grad_norm": 0.000413595320424065, + "learning_rate": 0.15289793891831113, + "loss": 0.3097, + "num_input_tokens_seen": 41285568, + "step": 19755 + }, + { + "epoch": 3.2235092585039564, + "grad_norm": 0.0004800644819624722, + "learning_rate": 0.15283904482816837, + "loss": 0.3153, + "num_input_tokens_seen": 41295600, + "step": 19760 + }, + { + "epoch": 3.2243249857247736, + "grad_norm": 0.0006771219777874649, + "learning_rate": 0.15278015030020928, + "loss": 0.2951, + "num_input_tokens_seen": 41304880, + "step": 19765 + }, + { + "epoch": 3.225140712945591, + "grad_norm": 0.0003588136751204729, + "learning_rate": 0.152721255343516, + "loss": 0.3385, + "num_input_tokens_seen": 41313616, + "step": 19770 + }, + { + "epoch": 3.2259564401664083, + "grad_norm": 0.00027644800138659775, + "learning_rate": 0.15266235996717098, + "loss": 0.2545, + "num_input_tokens_seen": 41324176, + "step": 19775 + }, + { + "epoch": 3.226772167387226, + "grad_norm": 0.0006106523214839399, + "learning_rate": 0.15260346418025664, + "loss": 0.2988, + "num_input_tokens_seen": 41335824, + "step": 19780 + }, + { + "epoch": 3.227587894608043, + "grad_norm": 0.00044736755080521107, + "learning_rate": 0.15254456799185537, + "loss": 0.3275, + "num_input_tokens_seen": 41346272, + "step": 19785 + }, + { + "epoch": 3.2284036218288605, + "grad_norm": 0.00036798324435949326, + "learning_rate": 0.15248567141104974, + "loss": 0.3577, + "num_input_tokens_seen": 41356880, + "step": 19790 + }, + { + "epoch": 3.2292193490496777, + "grad_norm": 0.00023069468443281949, + "learning_rate": 0.15242677444692232, + "loss": 0.3423, + "num_input_tokens_seen": 41366608, + "step": 19795 + }, + { + "epoch": 3.2300350762704952, + "grad_norm": 0.00029958118102513254, + "learning_rate": 0.15236787710855584, + "loss": 0.3176, + "num_input_tokens_seen": 41378128, + "step": 19800 + }, + { + "epoch": 3.2300350762704952, + "eval_loss": 0.3197885751724243, + "eval_runtime": 156.1743, + "eval_samples_per_second": 17.448, + "eval_steps_per_second": 8.727, + "num_input_tokens_seen": 41378128, + "step": 19800 + }, + { + "epoch": 3.2308508034913124, + "grad_norm": 0.0002867418515961617, + "learning_rate": 0.1523089794050329, + "loss": 0.2769, + "num_input_tokens_seen": 41387392, + "step": 19805 + }, + { + "epoch": 3.23166653071213, + "grad_norm": 0.0007009010878391564, + "learning_rate": 0.15225008134543633, + "loss": 0.3641, + "num_input_tokens_seen": 41397776, + "step": 19810 + }, + { + "epoch": 3.232482257932947, + "grad_norm": 0.0004935573088005185, + "learning_rate": 0.15219118293884895, + "loss": 0.3179, + "num_input_tokens_seen": 41408080, + "step": 19815 + }, + { + "epoch": 3.2332979851537647, + "grad_norm": 0.0004191804910078645, + "learning_rate": 0.15213228419435362, + "loss": 0.2739, + "num_input_tokens_seen": 41418032, + "step": 19820 + }, + { + "epoch": 3.234113712374582, + "grad_norm": 0.0007964320830069482, + "learning_rate": 0.15207338512103327, + "loss": 0.3336, + "num_input_tokens_seen": 41428048, + "step": 19825 + }, + { + "epoch": 3.2349294395953994, + "grad_norm": 0.0003529639798216522, + "learning_rate": 0.1520144857279709, + "loss": 0.3081, + "num_input_tokens_seen": 41438704, + "step": 19830 + }, + { + "epoch": 3.2357451668162165, + "grad_norm": 0.0003863846941385418, + "learning_rate": 0.1519555860242495, + "loss": 0.2844, + "num_input_tokens_seen": 41449008, + "step": 19835 + }, + { + "epoch": 3.236560894037034, + "grad_norm": 0.0005969122285023332, + "learning_rate": 0.15189668601895218, + "loss": 0.3211, + "num_input_tokens_seen": 41460176, + "step": 19840 + }, + { + "epoch": 3.237376621257851, + "grad_norm": 0.0003833108930848539, + "learning_rate": 0.151837785721162, + "loss": 0.3271, + "num_input_tokens_seen": 41471168, + "step": 19845 + }, + { + "epoch": 3.2381923484786688, + "grad_norm": 0.0003312979533802718, + "learning_rate": 0.15177888513996218, + "loss": 0.361, + "num_input_tokens_seen": 41482192, + "step": 19850 + }, + { + "epoch": 3.239008075699486, + "grad_norm": 0.0002741198695730418, + "learning_rate": 0.15171998428443592, + "loss": 0.336, + "num_input_tokens_seen": 41492384, + "step": 19855 + }, + { + "epoch": 3.2398238029203035, + "grad_norm": 0.0009784703142940998, + "learning_rate": 0.1516610831636665, + "loss": 0.3888, + "num_input_tokens_seen": 41500896, + "step": 19860 + }, + { + "epoch": 3.2406395301411206, + "grad_norm": 0.0012576543958857656, + "learning_rate": 0.15160218178673715, + "loss": 0.3745, + "num_input_tokens_seen": 41511920, + "step": 19865 + }, + { + "epoch": 3.241455257361938, + "grad_norm": 0.0008276299340650439, + "learning_rate": 0.15154328016273122, + "loss": 0.3088, + "num_input_tokens_seen": 41522720, + "step": 19870 + }, + { + "epoch": 3.2422709845827553, + "grad_norm": 0.0007415956351906061, + "learning_rate": 0.1514843783007321, + "loss": 0.3648, + "num_input_tokens_seen": 41533840, + "step": 19875 + }, + { + "epoch": 3.243086711803573, + "grad_norm": 0.00035091518657281995, + "learning_rate": 0.15142547620982322, + "loss": 0.3482, + "num_input_tokens_seen": 41544368, + "step": 19880 + }, + { + "epoch": 3.2439024390243905, + "grad_norm": 0.001708440133370459, + "learning_rate": 0.15136657389908797, + "loss": 0.3679, + "num_input_tokens_seen": 41555952, + "step": 19885 + }, + { + "epoch": 3.2447181662452076, + "grad_norm": 0.0004015575395897031, + "learning_rate": 0.15130767137760986, + "loss": 0.2948, + "num_input_tokens_seen": 41566096, + "step": 19890 + }, + { + "epoch": 3.245533893466025, + "grad_norm": 0.00033263477962464094, + "learning_rate": 0.15124876865447243, + "loss": 0.3454, + "num_input_tokens_seen": 41576928, + "step": 19895 + }, + { + "epoch": 3.2463496206868423, + "grad_norm": 0.0002502793213352561, + "learning_rate": 0.15118986573875912, + "loss": 0.3047, + "num_input_tokens_seen": 41588112, + "step": 19900 + }, + { + "epoch": 3.24716534790766, + "grad_norm": 0.00034483589115552604, + "learning_rate": 0.15113096263955358, + "loss": 0.2761, + "num_input_tokens_seen": 41597744, + "step": 19905 + }, + { + "epoch": 3.247981075128477, + "grad_norm": 0.0002546483010519296, + "learning_rate": 0.1510720593659394, + "loss": 0.3179, + "num_input_tokens_seen": 41609232, + "step": 19910 + }, + { + "epoch": 3.2487968023492946, + "grad_norm": 0.0003870599903166294, + "learning_rate": 0.15101315592700015, + "loss": 0.2766, + "num_input_tokens_seen": 41619408, + "step": 19915 + }, + { + "epoch": 3.2496125295701117, + "grad_norm": 0.0003617483307607472, + "learning_rate": 0.15095425233181956, + "loss": 0.2701, + "num_input_tokens_seen": 41630112, + "step": 19920 + }, + { + "epoch": 3.2504282567909293, + "grad_norm": 0.0006344463326968253, + "learning_rate": 0.15089534858948128, + "loss": 0.3264, + "num_input_tokens_seen": 41639920, + "step": 19925 + }, + { + "epoch": 3.2512439840117464, + "grad_norm": 0.0003488803922664374, + "learning_rate": 0.15083644470906898, + "loss": 0.3121, + "num_input_tokens_seen": 41650608, + "step": 19930 + }, + { + "epoch": 3.252059711232564, + "grad_norm": 0.0003312467597424984, + "learning_rate": 0.1507775406996664, + "loss": 0.3231, + "num_input_tokens_seen": 41661808, + "step": 19935 + }, + { + "epoch": 3.252875438453381, + "grad_norm": 0.0008319343323819339, + "learning_rate": 0.15071863657035725, + "loss": 0.4348, + "num_input_tokens_seen": 41672320, + "step": 19940 + }, + { + "epoch": 3.2536911656741987, + "grad_norm": 0.0004867111274506897, + "learning_rate": 0.15065973233022534, + "loss": 0.2732, + "num_input_tokens_seen": 41683392, + "step": 19945 + }, + { + "epoch": 3.254506892895016, + "grad_norm": 0.0005729782278649509, + "learning_rate": 0.15060082798835442, + "loss": 0.3254, + "num_input_tokens_seen": 41692528, + "step": 19950 + }, + { + "epoch": 3.2553226201158334, + "grad_norm": 0.00028561949147842824, + "learning_rate": 0.15054192355382823, + "loss": 0.3265, + "num_input_tokens_seen": 41701984, + "step": 19955 + }, + { + "epoch": 3.2561383473366505, + "grad_norm": 0.0003589988627936691, + "learning_rate": 0.15048301903573066, + "loss": 0.3078, + "num_input_tokens_seen": 41711760, + "step": 19960 + }, + { + "epoch": 3.256954074557468, + "grad_norm": 0.00027391535695642233, + "learning_rate": 0.15042411444314546, + "loss": 0.2965, + "num_input_tokens_seen": 41723376, + "step": 19965 + }, + { + "epoch": 3.2577698017782852, + "grad_norm": 0.0002724505902733654, + "learning_rate": 0.1503652097851565, + "loss": 0.3175, + "num_input_tokens_seen": 41733392, + "step": 19970 + }, + { + "epoch": 3.258585528999103, + "grad_norm": 0.0005578369600698352, + "learning_rate": 0.15030630507084758, + "loss": 0.2827, + "num_input_tokens_seen": 41744704, + "step": 19975 + }, + { + "epoch": 3.25940125621992, + "grad_norm": 0.0006049057701602578, + "learning_rate": 0.1502474003093026, + "loss": 0.3046, + "num_input_tokens_seen": 41754176, + "step": 19980 + }, + { + "epoch": 3.2602169834407375, + "grad_norm": 0.0003493686090223491, + "learning_rate": 0.15018849550960536, + "loss": 0.3224, + "num_input_tokens_seen": 41763776, + "step": 19985 + }, + { + "epoch": 3.2610327106615546, + "grad_norm": 0.00038926658453419805, + "learning_rate": 0.15012959068083975, + "loss": 0.347, + "num_input_tokens_seen": 41775120, + "step": 19990 + }, + { + "epoch": 3.261848437882372, + "grad_norm": 0.0005182258901186287, + "learning_rate": 0.1500706858320896, + "loss": 0.307, + "num_input_tokens_seen": 41783792, + "step": 19995 + }, + { + "epoch": 3.2626641651031894, + "grad_norm": 0.0005430023884400725, + "learning_rate": 0.15001178097243886, + "loss": 0.3015, + "num_input_tokens_seen": 41794480, + "step": 20000 + }, + { + "epoch": 3.2626641651031894, + "eval_loss": 0.31724241375923157, + "eval_runtime": 156.0451, + "eval_samples_per_second": 17.463, + "eval_steps_per_second": 8.735, + "num_input_tokens_seen": 41794480, + "step": 20000 + }, + { + "epoch": 3.263479892324007, + "grad_norm": 0.0004937872872687876, + "learning_rate": 0.1499528761109713, + "loss": 0.3495, + "num_input_tokens_seen": 41804528, + "step": 20005 + }, + { + "epoch": 3.264295619544824, + "grad_norm": 0.001001682598143816, + "learning_rate": 0.14989397125677087, + "loss": 0.3783, + "num_input_tokens_seen": 41814624, + "step": 20010 + }, + { + "epoch": 3.2651113467656416, + "grad_norm": 0.0012522273464128375, + "learning_rate": 0.14983506641892141, + "loss": 0.3749, + "num_input_tokens_seen": 41825856, + "step": 20015 + }, + { + "epoch": 3.2659270739864588, + "grad_norm": 0.0009309692541137338, + "learning_rate": 0.14977616160650672, + "loss": 0.3042, + "num_input_tokens_seen": 41834864, + "step": 20020 + }, + { + "epoch": 3.2667428012072763, + "grad_norm": 0.00042165580089204013, + "learning_rate": 0.14971725682861076, + "loss": 0.2567, + "num_input_tokens_seen": 41845344, + "step": 20025 + }, + { + "epoch": 3.2675585284280935, + "grad_norm": 0.0008053000783547759, + "learning_rate": 0.14965835209431738, + "loss": 0.3349, + "num_input_tokens_seen": 41855808, + "step": 20030 + }, + { + "epoch": 3.268374255648911, + "grad_norm": 0.00031878240406513214, + "learning_rate": 0.14959944741271036, + "loss": 0.3441, + "num_input_tokens_seen": 41866176, + "step": 20035 + }, + { + "epoch": 3.269189982869728, + "grad_norm": 0.001257614465430379, + "learning_rate": 0.14954054279287363, + "loss": 0.3961, + "num_input_tokens_seen": 41878032, + "step": 20040 + }, + { + "epoch": 3.2700057100905457, + "grad_norm": 0.0007021718192845583, + "learning_rate": 0.14948163824389094, + "loss": 0.3115, + "num_input_tokens_seen": 41889280, + "step": 20045 + }, + { + "epoch": 3.270821437311363, + "grad_norm": 0.00047925784019753337, + "learning_rate": 0.14942273377484613, + "loss": 0.3258, + "num_input_tokens_seen": 41898304, + "step": 20050 + }, + { + "epoch": 3.2716371645321805, + "grad_norm": 0.0003291078028269112, + "learning_rate": 0.1493638293948231, + "loss": 0.3119, + "num_input_tokens_seen": 41907136, + "step": 20055 + }, + { + "epoch": 3.272452891752998, + "grad_norm": 0.0005627247737720609, + "learning_rate": 0.14930492511290547, + "loss": 0.3181, + "num_input_tokens_seen": 41917536, + "step": 20060 + }, + { + "epoch": 3.273268618973815, + "grad_norm": 0.0003520725585985929, + "learning_rate": 0.14924602093817715, + "loss": 0.2475, + "num_input_tokens_seen": 41927856, + "step": 20065 + }, + { + "epoch": 3.2740843461946323, + "grad_norm": 0.00038841404602862895, + "learning_rate": 0.14918711687972194, + "loss": 0.2915, + "num_input_tokens_seen": 41937584, + "step": 20070 + }, + { + "epoch": 3.27490007341545, + "grad_norm": 0.0005315814050845802, + "learning_rate": 0.14912821294662346, + "loss": 0.342, + "num_input_tokens_seen": 41946448, + "step": 20075 + }, + { + "epoch": 3.2757158006362674, + "grad_norm": 0.0007179573294706643, + "learning_rate": 0.14906930914796554, + "loss": 0.3424, + "num_input_tokens_seen": 41957904, + "step": 20080 + }, + { + "epoch": 3.2765315278570846, + "grad_norm": 0.00026938639348372817, + "learning_rate": 0.14901040549283182, + "loss": 0.3271, + "num_input_tokens_seen": 41968816, + "step": 20085 + }, + { + "epoch": 3.2773472550779017, + "grad_norm": 0.0005058101960457861, + "learning_rate": 0.148951501990306, + "loss": 0.3317, + "num_input_tokens_seen": 41979344, + "step": 20090 + }, + { + "epoch": 3.2781629822987193, + "grad_norm": 0.0003876287373714149, + "learning_rate": 0.14889259864947177, + "loss": 0.3145, + "num_input_tokens_seen": 41990896, + "step": 20095 + }, + { + "epoch": 3.278978709519537, + "grad_norm": 0.0006357879028655589, + "learning_rate": 0.14883369547941272, + "loss": 0.3159, + "num_input_tokens_seen": 42000880, + "step": 20100 + }, + { + "epoch": 3.279794436740354, + "grad_norm": 0.000780373637098819, + "learning_rate": 0.14877479248921247, + "loss": 0.3457, + "num_input_tokens_seen": 42012592, + "step": 20105 + }, + { + "epoch": 3.2806101639611716, + "grad_norm": 0.0006318976520560682, + "learning_rate": 0.14871588968795468, + "loss": 0.3173, + "num_input_tokens_seen": 42022608, + "step": 20110 + }, + { + "epoch": 3.2814258911819887, + "grad_norm": 0.0005149224889464676, + "learning_rate": 0.1486569870847228, + "loss": 0.2932, + "num_input_tokens_seen": 42034208, + "step": 20115 + }, + { + "epoch": 3.2822416184028063, + "grad_norm": 0.00034389435313642025, + "learning_rate": 0.1485980846886004, + "loss": 0.2861, + "num_input_tokens_seen": 42043744, + "step": 20120 + }, + { + "epoch": 3.2830573456236234, + "grad_norm": 0.00035175905213691294, + "learning_rate": 0.14853918250867096, + "loss": 0.3166, + "num_input_tokens_seen": 42054272, + "step": 20125 + }, + { + "epoch": 3.283873072844441, + "grad_norm": 0.0004547625139821321, + "learning_rate": 0.1484802805540179, + "loss": 0.261, + "num_input_tokens_seen": 42065472, + "step": 20130 + }, + { + "epoch": 3.284688800065258, + "grad_norm": 0.00039514576201327145, + "learning_rate": 0.14842137883372472, + "loss": 0.3578, + "num_input_tokens_seen": 42075008, + "step": 20135 + }, + { + "epoch": 3.2855045272860757, + "grad_norm": 0.00045054193469695747, + "learning_rate": 0.14836247735687474, + "loss": 0.2326, + "num_input_tokens_seen": 42084736, + "step": 20140 + }, + { + "epoch": 3.286320254506893, + "grad_norm": 0.0013092664303258061, + "learning_rate": 0.14830357613255132, + "loss": 0.3449, + "num_input_tokens_seen": 42095104, + "step": 20145 + }, + { + "epoch": 3.2871359817277104, + "grad_norm": 0.00045014277566224337, + "learning_rate": 0.1482446751698378, + "loss": 0.2728, + "num_input_tokens_seen": 42104928, + "step": 20150 + }, + { + "epoch": 3.2879517089485275, + "grad_norm": 0.0006974060670472682, + "learning_rate": 0.14818577447781744, + "loss": 0.4354, + "num_input_tokens_seen": 42114704, + "step": 20155 + }, + { + "epoch": 3.288767436169345, + "grad_norm": 0.0006045638001523912, + "learning_rate": 0.14812687406557346, + "loss": 0.3601, + "num_input_tokens_seen": 42124720, + "step": 20160 + }, + { + "epoch": 3.289583163390162, + "grad_norm": 0.0011508562602102757, + "learning_rate": 0.14806797394218899, + "loss": 0.4243, + "num_input_tokens_seen": 42134256, + "step": 20165 + }, + { + "epoch": 3.29039889061098, + "grad_norm": 0.0007592430338263512, + "learning_rate": 0.1480090741167472, + "loss": 0.3286, + "num_input_tokens_seen": 42144736, + "step": 20170 + }, + { + "epoch": 3.291214617831797, + "grad_norm": 0.00027297859196551144, + "learning_rate": 0.1479501745983313, + "loss": 0.2927, + "num_input_tokens_seen": 42155008, + "step": 20175 + }, + { + "epoch": 3.2920303450526145, + "grad_norm": 0.0006504260818473995, + "learning_rate": 0.14789127539602415, + "loss": 0.3611, + "num_input_tokens_seen": 42164896, + "step": 20180 + }, + { + "epoch": 3.2928460722734316, + "grad_norm": 0.00025473866844549775, + "learning_rate": 0.14783237651890885, + "loss": 0.3515, + "num_input_tokens_seen": 42175216, + "step": 20185 + }, + { + "epoch": 3.293661799494249, + "grad_norm": 0.0002970833156723529, + "learning_rate": 0.14777347797606838, + "loss": 0.3361, + "num_input_tokens_seen": 42186368, + "step": 20190 + }, + { + "epoch": 3.2944775267150663, + "grad_norm": 0.0004677297256421298, + "learning_rate": 0.14771457977658553, + "loss": 0.3224, + "num_input_tokens_seen": 42198400, + "step": 20195 + }, + { + "epoch": 3.295293253935884, + "grad_norm": 0.0005236223223619163, + "learning_rate": 0.14765568192954326, + "loss": 0.3611, + "num_input_tokens_seen": 42208144, + "step": 20200 + }, + { + "epoch": 3.295293253935884, + "eval_loss": 0.32120656967163086, + "eval_runtime": 156.2073, + "eval_samples_per_second": 17.445, + "eval_steps_per_second": 8.726, + "num_input_tokens_seen": 42208144, + "step": 20200 + }, + { + "epoch": 3.296108981156701, + "grad_norm": 0.0005901111289858818, + "learning_rate": 0.14759678444402421, + "loss": 0.3103, + "num_input_tokens_seen": 42219360, + "step": 20205 + }, + { + "epoch": 3.2969247083775186, + "grad_norm": 0.0003055954584851861, + "learning_rate": 0.14753788732911122, + "loss": 0.3241, + "num_input_tokens_seen": 42229456, + "step": 20210 + }, + { + "epoch": 3.2977404355983357, + "grad_norm": 0.00038286615745164454, + "learning_rate": 0.147478990593887, + "loss": 0.3062, + "num_input_tokens_seen": 42240688, + "step": 20215 + }, + { + "epoch": 3.2985561628191533, + "grad_norm": 0.0003877217823173851, + "learning_rate": 0.14742009424743405, + "loss": 0.2685, + "num_input_tokens_seen": 42249680, + "step": 20220 + }, + { + "epoch": 3.2993718900399704, + "grad_norm": 0.0005540474667213857, + "learning_rate": 0.14736119829883504, + "loss": 0.3315, + "num_input_tokens_seen": 42259936, + "step": 20225 + }, + { + "epoch": 3.300187617260788, + "grad_norm": 0.0003956314467359334, + "learning_rate": 0.14730230275717243, + "loss": 0.2454, + "num_input_tokens_seen": 42271776, + "step": 20230 + }, + { + "epoch": 3.3010033444816056, + "grad_norm": 0.000709167739842087, + "learning_rate": 0.14724340763152854, + "loss": 0.3927, + "num_input_tokens_seen": 42283536, + "step": 20235 + }, + { + "epoch": 3.3018190717024227, + "grad_norm": 0.0002544232993386686, + "learning_rate": 0.14718451293098594, + "loss": 0.3082, + "num_input_tokens_seen": 42294464, + "step": 20240 + }, + { + "epoch": 3.30263479892324, + "grad_norm": 0.0005507220630533993, + "learning_rate": 0.14712561866462676, + "loss": 0.3076, + "num_input_tokens_seen": 42304464, + "step": 20245 + }, + { + "epoch": 3.3034505261440574, + "grad_norm": 0.00022829484078101814, + "learning_rate": 0.1470667248415333, + "loss": 0.2848, + "num_input_tokens_seen": 42316048, + "step": 20250 + }, + { + "epoch": 3.304266253364875, + "grad_norm": 0.00038390018744394183, + "learning_rate": 0.1470078314707878, + "loss": 0.2644, + "num_input_tokens_seen": 42327616, + "step": 20255 + }, + { + "epoch": 3.305081980585692, + "grad_norm": 0.0003586740349419415, + "learning_rate": 0.14694893856147223, + "loss": 0.2898, + "num_input_tokens_seen": 42336704, + "step": 20260 + }, + { + "epoch": 3.3058977078065093, + "grad_norm": 0.0007635607616975904, + "learning_rate": 0.14689004612266868, + "loss": 0.3607, + "num_input_tokens_seen": 42346672, + "step": 20265 + }, + { + "epoch": 3.306713435027327, + "grad_norm": 0.0018260296201333404, + "learning_rate": 0.14683115416345913, + "loss": 0.3724, + "num_input_tokens_seen": 42356848, + "step": 20270 + }, + { + "epoch": 3.3075291622481444, + "grad_norm": 0.0006409902707673609, + "learning_rate": 0.1467722626929254, + "loss": 0.4209, + "num_input_tokens_seen": 42367392, + "step": 20275 + }, + { + "epoch": 3.3083448894689615, + "grad_norm": 0.000637591234408319, + "learning_rate": 0.14671337172014937, + "loss": 0.3022, + "num_input_tokens_seen": 42377856, + "step": 20280 + }, + { + "epoch": 3.309160616689779, + "grad_norm": 0.00033252721186727285, + "learning_rate": 0.14665448125421265, + "loss": 0.2837, + "num_input_tokens_seen": 42387808, + "step": 20285 + }, + { + "epoch": 3.3099763439105963, + "grad_norm": 0.0005193236866034567, + "learning_rate": 0.146595591304197, + "loss": 0.3619, + "num_input_tokens_seen": 42399008, + "step": 20290 + }, + { + "epoch": 3.310792071131414, + "grad_norm": 0.0003134818689431995, + "learning_rate": 0.14653670187918397, + "loss": 0.3025, + "num_input_tokens_seen": 42409408, + "step": 20295 + }, + { + "epoch": 3.311607798352231, + "grad_norm": 0.0006766076548956335, + "learning_rate": 0.14647781298825502, + "loss": 0.2573, + "num_input_tokens_seen": 42420736, + "step": 20300 + }, + { + "epoch": 3.3124235255730485, + "grad_norm": 0.00061030249344185, + "learning_rate": 0.14641892464049153, + "loss": 0.2714, + "num_input_tokens_seen": 42430352, + "step": 20305 + }, + { + "epoch": 3.3132392527938657, + "grad_norm": 0.0003514347190503031, + "learning_rate": 0.14636003684497495, + "loss": 0.3001, + "num_input_tokens_seen": 42440464, + "step": 20310 + }, + { + "epoch": 3.3140549800146832, + "grad_norm": 0.000753148808144033, + "learning_rate": 0.14630114961078636, + "loss": 0.3284, + "num_input_tokens_seen": 42450896, + "step": 20315 + }, + { + "epoch": 3.3148707072355004, + "grad_norm": 0.0005497104721143842, + "learning_rate": 0.14624226294700704, + "loss": 0.3578, + "num_input_tokens_seen": 42461856, + "step": 20320 + }, + { + "epoch": 3.315686434456318, + "grad_norm": 0.00039224422653205693, + "learning_rate": 0.14618337686271793, + "loss": 0.3065, + "num_input_tokens_seen": 42472768, + "step": 20325 + }, + { + "epoch": 3.316502161677135, + "grad_norm": 0.0006945045897737145, + "learning_rate": 0.1461244913670001, + "loss": 0.3206, + "num_input_tokens_seen": 42482928, + "step": 20330 + }, + { + "epoch": 3.3173178888979526, + "grad_norm": 0.00040341608109883964, + "learning_rate": 0.1460656064689344, + "loss": 0.3361, + "num_input_tokens_seen": 42493696, + "step": 20335 + }, + { + "epoch": 3.31813361611877, + "grad_norm": 0.0005012764013372362, + "learning_rate": 0.14600672217760163, + "loss": 0.3486, + "num_input_tokens_seen": 42503920, + "step": 20340 + }, + { + "epoch": 3.3189493433395874, + "grad_norm": 0.0005320647032931447, + "learning_rate": 0.14594783850208248, + "loss": 0.3031, + "num_input_tokens_seen": 42514560, + "step": 20345 + }, + { + "epoch": 3.3197650705604045, + "grad_norm": 0.0005647196085192263, + "learning_rate": 0.14588895545145758, + "loss": 0.3581, + "num_input_tokens_seen": 42524272, + "step": 20350 + }, + { + "epoch": 3.320580797781222, + "grad_norm": 0.00034849403891712427, + "learning_rate": 0.14583007303480738, + "loss": 0.3449, + "num_input_tokens_seen": 42534704, + "step": 20355 + }, + { + "epoch": 3.321396525002039, + "grad_norm": 0.0002826167328748852, + "learning_rate": 0.14577119126121235, + "loss": 0.3249, + "num_input_tokens_seen": 42545616, + "step": 20360 + }, + { + "epoch": 3.3222122522228568, + "grad_norm": 0.0008839423535391688, + "learning_rate": 0.14571231013975272, + "loss": 0.2825, + "num_input_tokens_seen": 42555136, + "step": 20365 + }, + { + "epoch": 3.323027979443674, + "grad_norm": 0.00033541093580424786, + "learning_rate": 0.1456534296795088, + "loss": 0.2851, + "num_input_tokens_seen": 42565456, + "step": 20370 + }, + { + "epoch": 3.3238437066644915, + "grad_norm": 0.0003382720169611275, + "learning_rate": 0.14559454988956066, + "loss": 0.3053, + "num_input_tokens_seen": 42577072, + "step": 20375 + }, + { + "epoch": 3.3246594338853086, + "grad_norm": 0.00041261891601607203, + "learning_rate": 0.1455356707789882, + "loss": 0.2489, + "num_input_tokens_seen": 42587424, + "step": 20380 + }, + { + "epoch": 3.325475161106126, + "grad_norm": 0.0009514846024103463, + "learning_rate": 0.14547679235687147, + "loss": 0.3681, + "num_input_tokens_seen": 42597888, + "step": 20385 + }, + { + "epoch": 3.3262908883269433, + "grad_norm": 0.000466479774331674, + "learning_rate": 0.14541791463229023, + "loss": 0.3185, + "num_input_tokens_seen": 42606192, + "step": 20390 + }, + { + "epoch": 3.327106615547761, + "grad_norm": 0.0009570640977472067, + "learning_rate": 0.14535903761432406, + "loss": 0.326, + "num_input_tokens_seen": 42616096, + "step": 20395 + }, + { + "epoch": 3.327922342768578, + "grad_norm": 0.0004992742324247956, + "learning_rate": 0.1453001613120527, + "loss": 0.3039, + "num_input_tokens_seen": 42625520, + "step": 20400 + }, + { + "epoch": 3.327922342768578, + "eval_loss": 0.3169015944004059, + "eval_runtime": 156.1611, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 42625520, + "step": 20400 + }, + { + "epoch": 3.3287380699893956, + "grad_norm": 0.0004331798991188407, + "learning_rate": 0.14524128573455547, + "loss": 0.3841, + "num_input_tokens_seen": 42636448, + "step": 20405 + }, + { + "epoch": 3.3295537972102127, + "grad_norm": 0.00045809379662387073, + "learning_rate": 0.14518241089091177, + "loss": 0.369, + "num_input_tokens_seen": 42647328, + "step": 20410 + }, + { + "epoch": 3.3303695244310303, + "grad_norm": 0.0010867544915527105, + "learning_rate": 0.1451235367902009, + "loss": 0.3424, + "num_input_tokens_seen": 42657552, + "step": 20415 + }, + { + "epoch": 3.3311852516518474, + "grad_norm": 0.0004648061003535986, + "learning_rate": 0.1450646634415019, + "loss": 0.3378, + "num_input_tokens_seen": 42668800, + "step": 20420 + }, + { + "epoch": 3.332000978872665, + "grad_norm": 0.0005386329721659422, + "learning_rate": 0.1450057908538938, + "loss": 0.2655, + "num_input_tokens_seen": 42679600, + "step": 20425 + }, + { + "epoch": 3.3328167060934826, + "grad_norm": 0.00033697710023261607, + "learning_rate": 0.14494691903645557, + "loss": 0.2899, + "num_input_tokens_seen": 42690832, + "step": 20430 + }, + { + "epoch": 3.3336324333142997, + "grad_norm": 0.00025698530953377485, + "learning_rate": 0.14488804799826588, + "loss": 0.3158, + "num_input_tokens_seen": 42700912, + "step": 20435 + }, + { + "epoch": 3.334448160535117, + "grad_norm": 0.0001881359057733789, + "learning_rate": 0.14482917774840348, + "loss": 0.375, + "num_input_tokens_seen": 42712016, + "step": 20440 + }, + { + "epoch": 3.3352638877559344, + "grad_norm": 0.0004837728338316083, + "learning_rate": 0.14477030829594684, + "loss": 0.2759, + "num_input_tokens_seen": 42722336, + "step": 20445 + }, + { + "epoch": 3.336079614976752, + "grad_norm": 0.00024087616475299, + "learning_rate": 0.14471143964997432, + "loss": 0.3095, + "num_input_tokens_seen": 42731808, + "step": 20450 + }, + { + "epoch": 3.336895342197569, + "grad_norm": 0.0002908949099946767, + "learning_rate": 0.14465257181956434, + "loss": 0.2996, + "num_input_tokens_seen": 42742512, + "step": 20455 + }, + { + "epoch": 3.3377110694183862, + "grad_norm": 0.0004498311609495431, + "learning_rate": 0.1445937048137949, + "loss": 0.3116, + "num_input_tokens_seen": 42753680, + "step": 20460 + }, + { + "epoch": 3.338526796639204, + "grad_norm": 0.000497231085319072, + "learning_rate": 0.14453483864174416, + "loss": 0.3594, + "num_input_tokens_seen": 42765376, + "step": 20465 + }, + { + "epoch": 3.3393425238600214, + "grad_norm": 0.0005670637474395335, + "learning_rate": 0.14447597331249, + "loss": 0.3395, + "num_input_tokens_seen": 42775072, + "step": 20470 + }, + { + "epoch": 3.3401582510808385, + "grad_norm": 0.0004945266991853714, + "learning_rate": 0.1444171088351102, + "loss": 0.3022, + "num_input_tokens_seen": 42785872, + "step": 20475 + }, + { + "epoch": 3.340973978301656, + "grad_norm": 0.0003761481202673167, + "learning_rate": 0.14435824521868235, + "loss": 0.3059, + "num_input_tokens_seen": 42796800, + "step": 20480 + }, + { + "epoch": 3.3417897055224732, + "grad_norm": 0.00034094491275027394, + "learning_rate": 0.14429938247228397, + "loss": 0.3052, + "num_input_tokens_seen": 42807312, + "step": 20485 + }, + { + "epoch": 3.342605432743291, + "grad_norm": 0.0005419516237452626, + "learning_rate": 0.14424052060499243, + "loss": 0.3285, + "num_input_tokens_seen": 42818704, + "step": 20490 + }, + { + "epoch": 3.343421159964108, + "grad_norm": 0.00036622441257350147, + "learning_rate": 0.14418165962588506, + "loss": 0.361, + "num_input_tokens_seen": 42829088, + "step": 20495 + }, + { + "epoch": 3.3442368871849255, + "grad_norm": 0.00035822016070596874, + "learning_rate": 0.1441227995440388, + "loss": 0.3646, + "num_input_tokens_seen": 42840000, + "step": 20500 + }, + { + "epoch": 3.3450526144057426, + "grad_norm": 0.000578985083848238, + "learning_rate": 0.14406394036853082, + "loss": 0.3687, + "num_input_tokens_seen": 42850240, + "step": 20505 + }, + { + "epoch": 3.34586834162656, + "grad_norm": 0.0005968459299765527, + "learning_rate": 0.14400508210843774, + "loss": 0.2818, + "num_input_tokens_seen": 42862704, + "step": 20510 + }, + { + "epoch": 3.3466840688473773, + "grad_norm": 0.0004989058361388743, + "learning_rate": 0.1439462247728364, + "loss": 0.3408, + "num_input_tokens_seen": 42873392, + "step": 20515 + }, + { + "epoch": 3.347499796068195, + "grad_norm": 0.00037064915522933006, + "learning_rate": 0.14388736837080326, + "loss": 0.3371, + "num_input_tokens_seen": 42883440, + "step": 20520 + }, + { + "epoch": 3.348315523289012, + "grad_norm": 0.0003086285141762346, + "learning_rate": 0.14382851291141469, + "loss": 0.2821, + "num_input_tokens_seen": 42893728, + "step": 20525 + }, + { + "epoch": 3.3491312505098296, + "grad_norm": 0.00030844518914818764, + "learning_rate": 0.14376965840374697, + "loss": 0.3378, + "num_input_tokens_seen": 42903968, + "step": 20530 + }, + { + "epoch": 3.3499469777306468, + "grad_norm": 0.00024835634394548833, + "learning_rate": 0.14371080485687632, + "loss": 0.3182, + "num_input_tokens_seen": 42914464, + "step": 20535 + }, + { + "epoch": 3.3507627049514643, + "grad_norm": 0.000514082727022469, + "learning_rate": 0.1436519522798785, + "loss": 0.3258, + "num_input_tokens_seen": 42925872, + "step": 20540 + }, + { + "epoch": 3.3515784321722815, + "grad_norm": 0.0005201274761930108, + "learning_rate": 0.14359310068182948, + "loss": 0.336, + "num_input_tokens_seen": 42936208, + "step": 20545 + }, + { + "epoch": 3.352394159393099, + "grad_norm": 0.00029679882572963834, + "learning_rate": 0.14353425007180484, + "loss": 0.2961, + "num_input_tokens_seen": 42947456, + "step": 20550 + }, + { + "epoch": 3.353209886613916, + "grad_norm": 0.0003777243255171925, + "learning_rate": 0.14347540045888005, + "loss": 0.3326, + "num_input_tokens_seen": 42959088, + "step": 20555 + }, + { + "epoch": 3.3540256138347337, + "grad_norm": 0.000293458258965984, + "learning_rate": 0.14341655185213056, + "loss": 0.3499, + "num_input_tokens_seen": 42969424, + "step": 20560 + }, + { + "epoch": 3.354841341055551, + "grad_norm": 0.00043215134064666927, + "learning_rate": 0.14335770426063144, + "loss": 0.3767, + "num_input_tokens_seen": 42980192, + "step": 20565 + }, + { + "epoch": 3.3556570682763684, + "grad_norm": 0.0002719943877309561, + "learning_rate": 0.1432988576934578, + "loss": 0.244, + "num_input_tokens_seen": 42991168, + "step": 20570 + }, + { + "epoch": 3.3564727954971856, + "grad_norm": 0.0004427152161952108, + "learning_rate": 0.14324001215968457, + "loss": 0.2899, + "num_input_tokens_seen": 43000928, + "step": 20575 + }, + { + "epoch": 3.357288522718003, + "grad_norm": 0.00039532044320367277, + "learning_rate": 0.14318116766838637, + "loss": 0.277, + "num_input_tokens_seen": 43011936, + "step": 20580 + }, + { + "epoch": 3.3581042499388203, + "grad_norm": 0.00024143762129824609, + "learning_rate": 0.14312232422863788, + "loss": 0.3148, + "num_input_tokens_seen": 43023840, + "step": 20585 + }, + { + "epoch": 3.358919977159638, + "grad_norm": 0.0005241032340563834, + "learning_rate": 0.14306348184951334, + "loss": 0.2937, + "num_input_tokens_seen": 43034288, + "step": 20590 + }, + { + "epoch": 3.359735704380455, + "grad_norm": 0.00031577193294651806, + "learning_rate": 0.1430046405400871, + "loss": 0.3373, + "num_input_tokens_seen": 43044768, + "step": 20595 + }, + { + "epoch": 3.3605514316012726, + "grad_norm": 0.000280142790870741, + "learning_rate": 0.14294580030943324, + "loss": 0.3327, + "num_input_tokens_seen": 43054848, + "step": 20600 + }, + { + "epoch": 3.3605514316012726, + "eval_loss": 0.31586754322052, + "eval_runtime": 156.0642, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 43054848, + "step": 20600 + }, + { + "epoch": 3.36136715882209, + "grad_norm": 0.0003569799882825464, + "learning_rate": 0.14288696116662553, + "loss": 0.4144, + "num_input_tokens_seen": 43064880, + "step": 20605 + }, + { + "epoch": 3.3621828860429073, + "grad_norm": 0.00032326378277502954, + "learning_rate": 0.1428281231207378, + "loss": 0.3415, + "num_input_tokens_seen": 43076000, + "step": 20610 + }, + { + "epoch": 3.3629986132637244, + "grad_norm": 0.0003975844010710716, + "learning_rate": 0.1427692861808437, + "loss": 0.2424, + "num_input_tokens_seen": 43085296, + "step": 20615 + }, + { + "epoch": 3.363814340484542, + "grad_norm": 0.0003025557962246239, + "learning_rate": 0.1427104503560165, + "loss": 0.2791, + "num_input_tokens_seen": 43096432, + "step": 20620 + }, + { + "epoch": 3.3646300677053596, + "grad_norm": 0.00036292258300818503, + "learning_rate": 0.14265161565532947, + "loss": 0.2976, + "num_input_tokens_seen": 43108512, + "step": 20625 + }, + { + "epoch": 3.3654457949261767, + "grad_norm": 0.0004162581462878734, + "learning_rate": 0.14259278208785564, + "loss": 0.3572, + "num_input_tokens_seen": 43118416, + "step": 20630 + }, + { + "epoch": 3.366261522146994, + "grad_norm": 0.0006890145596116781, + "learning_rate": 0.14253394966266789, + "loss": 0.3173, + "num_input_tokens_seen": 43129488, + "step": 20635 + }, + { + "epoch": 3.3670772493678114, + "grad_norm": 0.00037303639692254364, + "learning_rate": 0.14247511838883894, + "loss": 0.2772, + "num_input_tokens_seen": 43140688, + "step": 20640 + }, + { + "epoch": 3.367892976588629, + "grad_norm": 0.0003970491816289723, + "learning_rate": 0.14241628827544126, + "loss": 0.2557, + "num_input_tokens_seen": 43149344, + "step": 20645 + }, + { + "epoch": 3.368708703809446, + "grad_norm": 0.00017719944298733026, + "learning_rate": 0.14235745933154723, + "loss": 0.2927, + "num_input_tokens_seen": 43160080, + "step": 20650 + }, + { + "epoch": 3.3695244310302637, + "grad_norm": 0.0007188184536062181, + "learning_rate": 0.14229863156622907, + "loss": 0.2923, + "num_input_tokens_seen": 43171104, + "step": 20655 + }, + { + "epoch": 3.370340158251081, + "grad_norm": 0.0005342025542631745, + "learning_rate": 0.14223980498855868, + "loss": 0.2991, + "num_input_tokens_seen": 43181728, + "step": 20660 + }, + { + "epoch": 3.3711558854718984, + "grad_norm": 0.00036900892155244946, + "learning_rate": 0.14218097960760792, + "loss": 0.2893, + "num_input_tokens_seen": 43191984, + "step": 20665 + }, + { + "epoch": 3.3719716126927155, + "grad_norm": 0.0004598720988724381, + "learning_rate": 0.1421221554324483, + "loss": 0.2699, + "num_input_tokens_seen": 43202240, + "step": 20670 + }, + { + "epoch": 3.372787339913533, + "grad_norm": 0.0007950398721732199, + "learning_rate": 0.1420633324721513, + "loss": 0.3565, + "num_input_tokens_seen": 43213408, + "step": 20675 + }, + { + "epoch": 3.37360306713435, + "grad_norm": 0.0003521368489600718, + "learning_rate": 0.14200451073578824, + "loss": 0.3153, + "num_input_tokens_seen": 43223568, + "step": 20680 + }, + { + "epoch": 3.374418794355168, + "grad_norm": 0.00035211798967793584, + "learning_rate": 0.14194569023243003, + "loss": 0.2518, + "num_input_tokens_seen": 43235024, + "step": 20685 + }, + { + "epoch": 3.375234521575985, + "grad_norm": 0.0004924519453197718, + "learning_rate": 0.14188687097114766, + "loss": 0.2998, + "num_input_tokens_seen": 43246640, + "step": 20690 + }, + { + "epoch": 3.3760502487968025, + "grad_norm": 0.0005836089258082211, + "learning_rate": 0.14182805296101172, + "loss": 0.3135, + "num_input_tokens_seen": 43258448, + "step": 20695 + }, + { + "epoch": 3.3768659760176196, + "grad_norm": 0.00039012666093185544, + "learning_rate": 0.14176923621109272, + "loss": 0.3548, + "num_input_tokens_seen": 43268800, + "step": 20700 + }, + { + "epoch": 3.377681703238437, + "grad_norm": 0.0003998494939878583, + "learning_rate": 0.14171042073046097, + "loss": 0.3422, + "num_input_tokens_seen": 43279584, + "step": 20705 + }, + { + "epoch": 3.3784974304592543, + "grad_norm": 0.0003962362534366548, + "learning_rate": 0.14165160652818642, + "loss": 0.3384, + "num_input_tokens_seen": 43290352, + "step": 20710 + }, + { + "epoch": 3.379313157680072, + "grad_norm": 0.00040174758760258555, + "learning_rate": 0.14159279361333907, + "loss": 0.2678, + "num_input_tokens_seen": 43300704, + "step": 20715 + }, + { + "epoch": 3.380128884900889, + "grad_norm": 0.0005926314042881131, + "learning_rate": 0.14153398199498868, + "loss": 0.3278, + "num_input_tokens_seen": 43309936, + "step": 20720 + }, + { + "epoch": 3.3809446121217066, + "grad_norm": 0.00041792760021053255, + "learning_rate": 0.14147517168220458, + "loss": 0.3351, + "num_input_tokens_seen": 43320720, + "step": 20725 + }, + { + "epoch": 3.3817603393425237, + "grad_norm": 0.0004921128274872899, + "learning_rate": 0.14141636268405616, + "loss": 0.2762, + "num_input_tokens_seen": 43332080, + "step": 20730 + }, + { + "epoch": 3.3825760665633413, + "grad_norm": 0.0006992571870796382, + "learning_rate": 0.14135755500961253, + "loss": 0.2327, + "num_input_tokens_seen": 43342944, + "step": 20735 + }, + { + "epoch": 3.3833917937841584, + "grad_norm": 0.0007958585629239678, + "learning_rate": 0.14129874866794245, + "loss": 0.3794, + "num_input_tokens_seen": 43352864, + "step": 20740 + }, + { + "epoch": 3.384207521004976, + "grad_norm": 0.0005712095880880952, + "learning_rate": 0.14123994366811476, + "loss": 0.2653, + "num_input_tokens_seen": 43362480, + "step": 20745 + }, + { + "epoch": 3.385023248225793, + "grad_norm": 0.0009051788365468383, + "learning_rate": 0.14118114001919774, + "loss": 0.3029, + "num_input_tokens_seen": 43371840, + "step": 20750 + }, + { + "epoch": 3.3858389754466107, + "grad_norm": 0.000887187896296382, + "learning_rate": 0.14112233773025978, + "loss": 0.255, + "num_input_tokens_seen": 43383200, + "step": 20755 + }, + { + "epoch": 3.386654702667428, + "grad_norm": 0.00035409373231232166, + "learning_rate": 0.14106353681036896, + "loss": 0.3388, + "num_input_tokens_seen": 43393792, + "step": 20760 + }, + { + "epoch": 3.3874704298882454, + "grad_norm": 0.0004107528366148472, + "learning_rate": 0.14100473726859303, + "loss": 0.3046, + "num_input_tokens_seen": 43404704, + "step": 20765 + }, + { + "epoch": 3.3882861571090626, + "grad_norm": 0.0003149022231809795, + "learning_rate": 0.14094593911399964, + "loss": 0.308, + "num_input_tokens_seen": 43413344, + "step": 20770 + }, + { + "epoch": 3.38910188432988, + "grad_norm": 0.0006372390780597925, + "learning_rate": 0.14088714235565625, + "loss": 0.3387, + "num_input_tokens_seen": 43423200, + "step": 20775 + }, + { + "epoch": 3.3899176115506973, + "grad_norm": 0.00048471667105332017, + "learning_rate": 0.14082834700263, + "loss": 0.2374, + "num_input_tokens_seen": 43432320, + "step": 20780 + }, + { + "epoch": 3.390733338771515, + "grad_norm": 0.0002392954775132239, + "learning_rate": 0.14076955306398795, + "loss": 0.3836, + "num_input_tokens_seen": 43441616, + "step": 20785 + }, + { + "epoch": 3.391549065992332, + "grad_norm": 0.0005668253288604319, + "learning_rate": 0.14071076054879675, + "loss": 0.3618, + "num_input_tokens_seen": 43452752, + "step": 20790 + }, + { + "epoch": 3.3923647932131495, + "grad_norm": 0.0004810095706488937, + "learning_rate": 0.14065196946612302, + "loss": 0.3466, + "num_input_tokens_seen": 43463200, + "step": 20795 + }, + { + "epoch": 3.393180520433967, + "grad_norm": 0.00038178617251105607, + "learning_rate": 0.1405931798250331, + "loss": 0.338, + "num_input_tokens_seen": 43472928, + "step": 20800 + }, + { + "epoch": 3.393180520433967, + "eval_loss": 0.3215060532093048, + "eval_runtime": 156.0018, + "eval_samples_per_second": 17.468, + "eval_steps_per_second": 8.737, + "num_input_tokens_seen": 43472928, + "step": 20800 + }, + { + "epoch": 3.3939962476547842, + "grad_norm": 0.00047529011499136686, + "learning_rate": 0.14053439163459308, + "loss": 0.3038, + "num_input_tokens_seen": 43482976, + "step": 20805 + }, + { + "epoch": 3.3948119748756014, + "grad_norm": 0.0007318668649531901, + "learning_rate": 0.14047560490386876, + "loss": 0.3279, + "num_input_tokens_seen": 43493936, + "step": 20810 + }, + { + "epoch": 3.395627702096419, + "grad_norm": 0.000457880028989166, + "learning_rate": 0.14041681964192593, + "loss": 0.3142, + "num_input_tokens_seen": 43503936, + "step": 20815 + }, + { + "epoch": 3.3964434293172365, + "grad_norm": 0.0008624106412753463, + "learning_rate": 0.14035803585782988, + "loss": 0.4265, + "num_input_tokens_seen": 43514416, + "step": 20820 + }, + { + "epoch": 3.3972591565380537, + "grad_norm": 0.0006646325346082449, + "learning_rate": 0.14029925356064593, + "loss": 0.369, + "num_input_tokens_seen": 43525536, + "step": 20825 + }, + { + "epoch": 3.398074883758871, + "grad_norm": 0.000498749956022948, + "learning_rate": 0.1402404727594389, + "loss": 0.3151, + "num_input_tokens_seen": 43537152, + "step": 20830 + }, + { + "epoch": 3.3988906109796884, + "grad_norm": 0.00035234741517342627, + "learning_rate": 0.1401816934632737, + "loss": 0.3165, + "num_input_tokens_seen": 43546944, + "step": 20835 + }, + { + "epoch": 3.399706338200506, + "grad_norm": 0.0003223724488634616, + "learning_rate": 0.1401229156812147, + "loss": 0.2977, + "num_input_tokens_seen": 43556096, + "step": 20840 + }, + { + "epoch": 3.400522065421323, + "grad_norm": 0.00039090801146812737, + "learning_rate": 0.14006413942232626, + "loss": 0.2886, + "num_input_tokens_seen": 43566256, + "step": 20845 + }, + { + "epoch": 3.4013377926421406, + "grad_norm": 0.0006410626228898764, + "learning_rate": 0.14000536469567235, + "loss": 0.3569, + "num_input_tokens_seen": 43576896, + "step": 20850 + }, + { + "epoch": 3.4021535198629578, + "grad_norm": 0.0004047675465699285, + "learning_rate": 0.13994659151031685, + "loss": 0.3208, + "num_input_tokens_seen": 43587392, + "step": 20855 + }, + { + "epoch": 3.4029692470837754, + "grad_norm": 0.0007119845831766725, + "learning_rate": 0.13988781987532323, + "loss": 0.2554, + "num_input_tokens_seen": 43596608, + "step": 20860 + }, + { + "epoch": 3.4037849743045925, + "grad_norm": 0.0006801943527534604, + "learning_rate": 0.1398290497997549, + "loss": 0.3168, + "num_input_tokens_seen": 43607840, + "step": 20865 + }, + { + "epoch": 3.40460070152541, + "grad_norm": 0.00044203957077115774, + "learning_rate": 0.13977028129267488, + "loss": 0.3519, + "num_input_tokens_seen": 43616752, + "step": 20870 + }, + { + "epoch": 3.405416428746227, + "grad_norm": 0.0004757892165798694, + "learning_rate": 0.13971151436314605, + "loss": 0.3203, + "num_input_tokens_seen": 43628080, + "step": 20875 + }, + { + "epoch": 3.4062321559670448, + "grad_norm": 0.0007728030905127525, + "learning_rate": 0.13965274902023103, + "loss": 0.3347, + "num_input_tokens_seen": 43638208, + "step": 20880 + }, + { + "epoch": 3.407047883187862, + "grad_norm": 0.0009200425120070577, + "learning_rate": 0.13959398527299208, + "loss": 0.3267, + "num_input_tokens_seen": 43648688, + "step": 20885 + }, + { + "epoch": 3.4078636104086795, + "grad_norm": 0.00047386492951773107, + "learning_rate": 0.13953522313049138, + "loss": 0.3189, + "num_input_tokens_seen": 43659616, + "step": 20890 + }, + { + "epoch": 3.4086793376294966, + "grad_norm": 0.0004152601177338511, + "learning_rate": 0.13947646260179083, + "loss": 0.263, + "num_input_tokens_seen": 43669744, + "step": 20895 + }, + { + "epoch": 3.409495064850314, + "grad_norm": 0.0007792055839672685, + "learning_rate": 0.13941770369595194, + "loss": 0.3542, + "num_input_tokens_seen": 43682080, + "step": 20900 + }, + { + "epoch": 3.4103107920711313, + "grad_norm": 0.0005212662508711219, + "learning_rate": 0.1393589464220362, + "loss": 0.3082, + "num_input_tokens_seen": 43692640, + "step": 20905 + }, + { + "epoch": 3.411126519291949, + "grad_norm": 0.0007872026180848479, + "learning_rate": 0.13930019078910455, + "loss": 0.338, + "num_input_tokens_seen": 43702496, + "step": 20910 + }, + { + "epoch": 3.411942246512766, + "grad_norm": 0.0004752592649310827, + "learning_rate": 0.139241436806218, + "loss": 0.2966, + "num_input_tokens_seen": 43713552, + "step": 20915 + }, + { + "epoch": 3.4127579737335836, + "grad_norm": 0.0008231193060055375, + "learning_rate": 0.13918268448243712, + "loss": 0.3074, + "num_input_tokens_seen": 43725264, + "step": 20920 + }, + { + "epoch": 3.4135737009544007, + "grad_norm": 0.0007769163348712027, + "learning_rate": 0.13912393382682217, + "loss": 0.3425, + "num_input_tokens_seen": 43734864, + "step": 20925 + }, + { + "epoch": 3.4143894281752183, + "grad_norm": 0.0003926569188479334, + "learning_rate": 0.1390651848484333, + "loss": 0.3072, + "num_input_tokens_seen": 43746240, + "step": 20930 + }, + { + "epoch": 3.4152051553960354, + "grad_norm": 0.00039343704702332616, + "learning_rate": 0.1390064375563304, + "loss": 0.3572, + "num_input_tokens_seen": 43757072, + "step": 20935 + }, + { + "epoch": 3.416020882616853, + "grad_norm": 0.00043841535807587206, + "learning_rate": 0.13894769195957293, + "loss": 0.354, + "num_input_tokens_seen": 43766576, + "step": 20940 + }, + { + "epoch": 3.41683660983767, + "grad_norm": 0.001006046892143786, + "learning_rate": 0.13888894806722032, + "loss": 0.3458, + "num_input_tokens_seen": 43777056, + "step": 20945 + }, + { + "epoch": 3.4176523370584877, + "grad_norm": 0.00045305705862119794, + "learning_rate": 0.1388302058883315, + "loss": 0.2857, + "num_input_tokens_seen": 43786704, + "step": 20950 + }, + { + "epoch": 3.418468064279305, + "grad_norm": 0.00046146701788529754, + "learning_rate": 0.13877146543196528, + "loss": 0.3637, + "num_input_tokens_seen": 43797760, + "step": 20955 + }, + { + "epoch": 3.4192837915001224, + "grad_norm": 0.0003483077452983707, + "learning_rate": 0.13871272670718027, + "loss": 0.3056, + "num_input_tokens_seen": 43807680, + "step": 20960 + }, + { + "epoch": 3.4200995187209395, + "grad_norm": 0.0009394650114700198, + "learning_rate": 0.13865398972303455, + "loss": 0.2855, + "num_input_tokens_seen": 43818064, + "step": 20965 + }, + { + "epoch": 3.420915245941757, + "grad_norm": 0.0006311503821052611, + "learning_rate": 0.13859525448858623, + "loss": 0.3355, + "num_input_tokens_seen": 43829376, + "step": 20970 + }, + { + "epoch": 3.4217309731625742, + "grad_norm": 0.0007850069086998701, + "learning_rate": 0.13853652101289304, + "loss": 0.3445, + "num_input_tokens_seen": 43839792, + "step": 20975 + }, + { + "epoch": 3.422546700383392, + "grad_norm": 0.0034751815255731344, + "learning_rate": 0.13847778930501234, + "loss": 0.2839, + "num_input_tokens_seen": 43850624, + "step": 20980 + }, + { + "epoch": 3.423362427604209, + "grad_norm": 0.0006284848786890507, + "learning_rate": 0.1384190593740013, + "loss": 0.283, + "num_input_tokens_seen": 43862416, + "step": 20985 + }, + { + "epoch": 3.4241781548250265, + "grad_norm": 0.0003857780247926712, + "learning_rate": 0.13836033122891686, + "loss": 0.2712, + "num_input_tokens_seen": 43872432, + "step": 20990 + }, + { + "epoch": 3.424993882045844, + "grad_norm": 0.0005456936778500676, + "learning_rate": 0.1383016048788156, + "loss": 0.3427, + "num_input_tokens_seen": 43882976, + "step": 20995 + }, + { + "epoch": 3.4258096092666612, + "grad_norm": 0.0007885062950663269, + "learning_rate": 0.13824288033275392, + "loss": 0.3392, + "num_input_tokens_seen": 43892704, + "step": 21000 + }, + { + "epoch": 3.4258096092666612, + "eval_loss": 0.3171202540397644, + "eval_runtime": 156.1047, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 43892704, + "step": 21000 + }, + { + "epoch": 3.4266253364874784, + "grad_norm": 0.00044395297300070524, + "learning_rate": 0.1381841575997878, + "loss": 0.2929, + "num_input_tokens_seen": 43903520, + "step": 21005 + }, + { + "epoch": 3.427441063708296, + "grad_norm": 0.0008458448573946953, + "learning_rate": 0.13812543668897306, + "loss": 0.3076, + "num_input_tokens_seen": 43912960, + "step": 21010 + }, + { + "epoch": 3.4282567909291135, + "grad_norm": 0.0005778043414466083, + "learning_rate": 0.13806671760936526, + "loss": 0.3698, + "num_input_tokens_seen": 43924288, + "step": 21015 + }, + { + "epoch": 3.4290725181499306, + "grad_norm": 0.0007362011820077896, + "learning_rate": 0.13800800037001956, + "loss": 0.3282, + "num_input_tokens_seen": 43934240, + "step": 21020 + }, + { + "epoch": 3.429888245370748, + "grad_norm": 0.00042501732241362333, + "learning_rate": 0.13794928497999087, + "loss": 0.3167, + "num_input_tokens_seen": 43943504, + "step": 21025 + }, + { + "epoch": 3.4307039725915653, + "grad_norm": 0.0006727923173457384, + "learning_rate": 0.1378905714483339, + "loss": 0.2817, + "num_input_tokens_seen": 43953264, + "step": 21030 + }, + { + "epoch": 3.431519699812383, + "grad_norm": 0.0009252275340259075, + "learning_rate": 0.13783185978410295, + "loss": 0.3935, + "num_input_tokens_seen": 43964608, + "step": 21035 + }, + { + "epoch": 3.4323354270332, + "grad_norm": 0.0007266944157890975, + "learning_rate": 0.13777314999635218, + "loss": 0.3066, + "num_input_tokens_seen": 43973664, + "step": 21040 + }, + { + "epoch": 3.4331511542540176, + "grad_norm": 0.0008683079504407942, + "learning_rate": 0.1377144420941353, + "loss": 0.2889, + "num_input_tokens_seen": 43984960, + "step": 21045 + }, + { + "epoch": 3.4339668814748348, + "grad_norm": 0.0003583536308724433, + "learning_rate": 0.13765573608650586, + "loss": 0.2999, + "num_input_tokens_seen": 43995424, + "step": 21050 + }, + { + "epoch": 3.4347826086956523, + "grad_norm": 0.0007915825117379427, + "learning_rate": 0.13759703198251702, + "loss": 0.3285, + "num_input_tokens_seen": 44006160, + "step": 21055 + }, + { + "epoch": 3.4355983359164695, + "grad_norm": 0.0009121345356106758, + "learning_rate": 0.13753832979122174, + "loss": 0.2769, + "num_input_tokens_seen": 44016960, + "step": 21060 + }, + { + "epoch": 3.436414063137287, + "grad_norm": 0.0007292252266779542, + "learning_rate": 0.13747962952167264, + "loss": 0.3563, + "num_input_tokens_seen": 44027328, + "step": 21065 + }, + { + "epoch": 3.437229790358104, + "grad_norm": 0.00036544475005939603, + "learning_rate": 0.13742093118292192, + "loss": 0.2713, + "num_input_tokens_seen": 44037824, + "step": 21070 + }, + { + "epoch": 3.4380455175789217, + "grad_norm": 0.0007334748515859246, + "learning_rate": 0.13736223478402174, + "loss": 0.3285, + "num_input_tokens_seen": 44046448, + "step": 21075 + }, + { + "epoch": 3.438861244799739, + "grad_norm": 0.000875803001690656, + "learning_rate": 0.1373035403340238, + "loss": 0.304, + "num_input_tokens_seen": 44056896, + "step": 21080 + }, + { + "epoch": 3.4396769720205564, + "grad_norm": 0.0004908247501589358, + "learning_rate": 0.13724484784197943, + "loss": 0.3516, + "num_input_tokens_seen": 44066624, + "step": 21085 + }, + { + "epoch": 3.4404926992413736, + "grad_norm": 0.0003920189046766609, + "learning_rate": 0.13718615731693987, + "loss": 0.3025, + "num_input_tokens_seen": 44076384, + "step": 21090 + }, + { + "epoch": 3.441308426462191, + "grad_norm": 0.0004275843675713986, + "learning_rate": 0.13712746876795587, + "loss": 0.3511, + "num_input_tokens_seen": 44086704, + "step": 21095 + }, + { + "epoch": 3.4421241536830083, + "grad_norm": 0.0003513682459015399, + "learning_rate": 0.13706878220407792, + "loss": 0.3508, + "num_input_tokens_seen": 44097616, + "step": 21100 + }, + { + "epoch": 3.442939880903826, + "grad_norm": 0.0006657222984358668, + "learning_rate": 0.13701009763435631, + "loss": 0.3164, + "num_input_tokens_seen": 44108928, + "step": 21105 + }, + { + "epoch": 3.443755608124643, + "grad_norm": 0.0006538729649037123, + "learning_rate": 0.13695141506784084, + "loss": 0.309, + "num_input_tokens_seen": 44119888, + "step": 21110 + }, + { + "epoch": 3.4445713353454606, + "grad_norm": 0.0005224343040026724, + "learning_rate": 0.13689273451358114, + "loss": 0.3774, + "num_input_tokens_seen": 44130656, + "step": 21115 + }, + { + "epoch": 3.4453870625662777, + "grad_norm": 0.0004241164424456656, + "learning_rate": 0.13683405598062653, + "loss": 0.3587, + "num_input_tokens_seen": 44143360, + "step": 21120 + }, + { + "epoch": 3.4462027897870953, + "grad_norm": 0.0006497004069387913, + "learning_rate": 0.1367753794780259, + "loss": 0.3481, + "num_input_tokens_seen": 44154752, + "step": 21125 + }, + { + "epoch": 3.4470185170079124, + "grad_norm": 0.0004739404539577663, + "learning_rate": 0.13671670501482802, + "loss": 0.3315, + "num_input_tokens_seen": 44163728, + "step": 21130 + }, + { + "epoch": 3.44783424422873, + "grad_norm": 0.0002988146443385631, + "learning_rate": 0.1366580326000811, + "loss": 0.3165, + "num_input_tokens_seen": 44174928, + "step": 21135 + }, + { + "epoch": 3.448649971449547, + "grad_norm": 0.0002709797117859125, + "learning_rate": 0.1365993622428332, + "loss": 0.3219, + "num_input_tokens_seen": 44185776, + "step": 21140 + }, + { + "epoch": 3.4494656986703647, + "grad_norm": 0.0005856452626176178, + "learning_rate": 0.13654069395213211, + "loss": 0.329, + "num_input_tokens_seen": 44195920, + "step": 21145 + }, + { + "epoch": 3.450281425891182, + "grad_norm": 0.0006126192747615278, + "learning_rate": 0.13648202773702509, + "loss": 0.3303, + "num_input_tokens_seen": 44206000, + "step": 21150 + }, + { + "epoch": 3.4510971531119994, + "grad_norm": 0.000779840222094208, + "learning_rate": 0.13642336360655927, + "loss": 0.3211, + "num_input_tokens_seen": 44216032, + "step": 21155 + }, + { + "epoch": 3.4519128803328165, + "grad_norm": 0.0004913080483675003, + "learning_rate": 0.13636470156978145, + "loss": 0.3024, + "num_input_tokens_seen": 44226368, + "step": 21160 + }, + { + "epoch": 3.452728607553634, + "grad_norm": 0.0004450819978956133, + "learning_rate": 0.13630604163573798, + "loss": 0.2978, + "num_input_tokens_seen": 44237008, + "step": 21165 + }, + { + "epoch": 3.4535443347744517, + "grad_norm": 0.0003171354765072465, + "learning_rate": 0.13624738381347495, + "loss": 0.3349, + "num_input_tokens_seen": 44248144, + "step": 21170 + }, + { + "epoch": 3.454360061995269, + "grad_norm": 0.00025749794440343976, + "learning_rate": 0.1361887281120382, + "loss": 0.3225, + "num_input_tokens_seen": 44259040, + "step": 21175 + }, + { + "epoch": 3.455175789216086, + "grad_norm": 0.0011567221954464912, + "learning_rate": 0.13613007454047307, + "loss": 0.358, + "num_input_tokens_seen": 44269184, + "step": 21180 + }, + { + "epoch": 3.4559915164369035, + "grad_norm": 0.0002855679776985198, + "learning_rate": 0.13607142310782486, + "loss": 0.3282, + "num_input_tokens_seen": 44279808, + "step": 21185 + }, + { + "epoch": 3.456807243657721, + "grad_norm": 0.0005772073636762798, + "learning_rate": 0.13601277382313814, + "loss": 0.2833, + "num_input_tokens_seen": 44289072, + "step": 21190 + }, + { + "epoch": 3.457622970878538, + "grad_norm": 0.0009106198558583856, + "learning_rate": 0.1359541266954575, + "loss": 0.3773, + "num_input_tokens_seen": 44298752, + "step": 21195 + }, + { + "epoch": 3.4584386980993553, + "grad_norm": 0.00033340652589686215, + "learning_rate": 0.13589548173382707, + "loss": 0.3119, + "num_input_tokens_seen": 44309408, + "step": 21200 + }, + { + "epoch": 3.4584386980993553, + "eval_loss": 0.3145348131656647, + "eval_runtime": 156.0703, + "eval_samples_per_second": 17.46, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 44309408, + "step": 21200 + }, + { + "epoch": 3.459254425320173, + "grad_norm": 0.0003568716929294169, + "learning_rate": 0.1358368389472906, + "loss": 0.3423, + "num_input_tokens_seen": 44320160, + "step": 21205 + }, + { + "epoch": 3.4600701525409905, + "grad_norm": 0.00049501663306728, + "learning_rate": 0.13577819834489155, + "loss": 0.3642, + "num_input_tokens_seen": 44330528, + "step": 21210 + }, + { + "epoch": 3.4608858797618076, + "grad_norm": 0.0004372868570499122, + "learning_rate": 0.135719559935673, + "loss": 0.3421, + "num_input_tokens_seen": 44340304, + "step": 21215 + }, + { + "epoch": 3.461701606982625, + "grad_norm": 0.0004500061913859099, + "learning_rate": 0.13566092372867775, + "loss": 0.2909, + "num_input_tokens_seen": 44351648, + "step": 21220 + }, + { + "epoch": 3.4625173342034423, + "grad_norm": 0.0003222195664420724, + "learning_rate": 0.13560228973294833, + "loss": 0.3642, + "num_input_tokens_seen": 44361584, + "step": 21225 + }, + { + "epoch": 3.46333306142426, + "grad_norm": 0.0005374932079575956, + "learning_rate": 0.13554365795752668, + "loss": 0.3217, + "num_input_tokens_seen": 44371200, + "step": 21230 + }, + { + "epoch": 3.464148788645077, + "grad_norm": 0.0003682489041239023, + "learning_rate": 0.1354850284114547, + "loss": 0.3271, + "num_input_tokens_seen": 44380544, + "step": 21235 + }, + { + "epoch": 3.4649645158658946, + "grad_norm": 0.001996224047616124, + "learning_rate": 0.13542640110377374, + "loss": 0.4022, + "num_input_tokens_seen": 44389472, + "step": 21240 + }, + { + "epoch": 3.4657802430867117, + "grad_norm": 0.0003251934249419719, + "learning_rate": 0.13536777604352487, + "loss": 0.3479, + "num_input_tokens_seen": 44400704, + "step": 21245 + }, + { + "epoch": 3.4665959703075293, + "grad_norm": 0.0003102507907897234, + "learning_rate": 0.13530915323974887, + "loss": 0.2599, + "num_input_tokens_seen": 44411120, + "step": 21250 + }, + { + "epoch": 3.4674116975283464, + "grad_norm": 0.0005330419517122209, + "learning_rate": 0.13525053270148596, + "loss": 0.328, + "num_input_tokens_seen": 44421040, + "step": 21255 + }, + { + "epoch": 3.468227424749164, + "grad_norm": 0.0004117292119190097, + "learning_rate": 0.13519191443777628, + "loss": 0.2953, + "num_input_tokens_seen": 44431920, + "step": 21260 + }, + { + "epoch": 3.469043151969981, + "grad_norm": 0.0005865278071723878, + "learning_rate": 0.13513329845765953, + "loss": 0.3788, + "num_input_tokens_seen": 44441328, + "step": 21265 + }, + { + "epoch": 3.4698588791907987, + "grad_norm": 0.0005439002998173237, + "learning_rate": 0.13507468477017495, + "loss": 0.3061, + "num_input_tokens_seen": 44452080, + "step": 21270 + }, + { + "epoch": 3.470674606411616, + "grad_norm": 0.0005642804899252951, + "learning_rate": 0.13501607338436153, + "loss": 0.315, + "num_input_tokens_seen": 44463232, + "step": 21275 + }, + { + "epoch": 3.4714903336324334, + "grad_norm": 0.00033696298487484455, + "learning_rate": 0.13495746430925798, + "loss": 0.3518, + "num_input_tokens_seen": 44473232, + "step": 21280 + }, + { + "epoch": 3.4723060608532506, + "grad_norm": 0.00033797265496104956, + "learning_rate": 0.13489885755390238, + "loss": 0.3106, + "num_input_tokens_seen": 44485232, + "step": 21285 + }, + { + "epoch": 3.473121788074068, + "grad_norm": 0.00044909058487974107, + "learning_rate": 0.13484025312733275, + "loss": 0.2341, + "num_input_tokens_seen": 44493840, + "step": 21290 + }, + { + "epoch": 3.4739375152948853, + "grad_norm": 0.0007225880981422961, + "learning_rate": 0.13478165103858658, + "loss": 0.3246, + "num_input_tokens_seen": 44504560, + "step": 21295 + }, + { + "epoch": 3.474753242515703, + "grad_norm": 0.000794969848357141, + "learning_rate": 0.13472305129670106, + "loss": 0.352, + "num_input_tokens_seen": 44515136, + "step": 21300 + }, + { + "epoch": 3.47556896973652, + "grad_norm": 0.00047490079305134714, + "learning_rate": 0.13466445391071305, + "loss": 0.3614, + "num_input_tokens_seen": 44526256, + "step": 21305 + }, + { + "epoch": 3.4763846969573375, + "grad_norm": 0.0005033068591728806, + "learning_rate": 0.13460585888965895, + "loss": 0.2697, + "num_input_tokens_seen": 44535824, + "step": 21310 + }, + { + "epoch": 3.4772004241781547, + "grad_norm": 0.0006683338433504105, + "learning_rate": 0.13454726624257482, + "loss": 0.2835, + "num_input_tokens_seen": 44547472, + "step": 21315 + }, + { + "epoch": 3.4780161513989722, + "grad_norm": 0.0003555740404408425, + "learning_rate": 0.1344886759784965, + "loss": 0.2971, + "num_input_tokens_seen": 44557328, + "step": 21320 + }, + { + "epoch": 3.4788318786197894, + "grad_norm": 0.0007085768738761544, + "learning_rate": 0.13443008810645923, + "loss": 0.3326, + "num_input_tokens_seen": 44567168, + "step": 21325 + }, + { + "epoch": 3.479647605840607, + "grad_norm": 0.0003781465347856283, + "learning_rate": 0.13437150263549807, + "loss": 0.3192, + "num_input_tokens_seen": 44577984, + "step": 21330 + }, + { + "epoch": 3.480463333061424, + "grad_norm": 0.0002111326903104782, + "learning_rate": 0.13431291957464755, + "loss": 0.3244, + "num_input_tokens_seen": 44589184, + "step": 21335 + }, + { + "epoch": 3.4812790602822417, + "grad_norm": 0.0005791664589196444, + "learning_rate": 0.13425433893294197, + "loss": 0.2464, + "num_input_tokens_seen": 44598624, + "step": 21340 + }, + { + "epoch": 3.482094787503059, + "grad_norm": 0.0005075957742519677, + "learning_rate": 0.13419576071941525, + "loss": 0.3858, + "num_input_tokens_seen": 44609584, + "step": 21345 + }, + { + "epoch": 3.4829105147238764, + "grad_norm": 0.0005458530504256487, + "learning_rate": 0.1341371849431008, + "loss": 0.3799, + "num_input_tokens_seen": 44621760, + "step": 21350 + }, + { + "epoch": 3.4837262419446935, + "grad_norm": 0.0006439578137360513, + "learning_rate": 0.13407861161303178, + "loss": 0.3201, + "num_input_tokens_seen": 44632832, + "step": 21355 + }, + { + "epoch": 3.484541969165511, + "grad_norm": 0.0008561733411625028, + "learning_rate": 0.13402004073824098, + "loss": 0.3826, + "num_input_tokens_seen": 44642464, + "step": 21360 + }, + { + "epoch": 3.4853576963863286, + "grad_norm": 0.0007407403900288045, + "learning_rate": 0.13396147232776062, + "loss": 0.3633, + "num_input_tokens_seen": 44652192, + "step": 21365 + }, + { + "epoch": 3.4861734236071458, + "grad_norm": 0.0006255265907384455, + "learning_rate": 0.13390290639062288, + "loss": 0.3292, + "num_input_tokens_seen": 44663072, + "step": 21370 + }, + { + "epoch": 3.486989150827963, + "grad_norm": 0.0005521899438463151, + "learning_rate": 0.13384434293585917, + "loss": 0.3742, + "num_input_tokens_seen": 44674384, + "step": 21375 + }, + { + "epoch": 3.4878048780487805, + "grad_norm": 0.0008192152017727494, + "learning_rate": 0.13378578197250088, + "loss": 0.3292, + "num_input_tokens_seen": 44684848, + "step": 21380 + }, + { + "epoch": 3.488620605269598, + "grad_norm": 0.00038296496495604515, + "learning_rate": 0.13372722350957872, + "loss": 0.3356, + "num_input_tokens_seen": 44695120, + "step": 21385 + }, + { + "epoch": 3.489436332490415, + "grad_norm": 0.0006935525452718139, + "learning_rate": 0.13366866755612322, + "loss": 0.3229, + "num_input_tokens_seen": 44705456, + "step": 21390 + }, + { + "epoch": 3.4902520597112328, + "grad_norm": 0.0002639455778989941, + "learning_rate": 0.13361011412116436, + "loss": 0.3146, + "num_input_tokens_seen": 44715360, + "step": 21395 + }, + { + "epoch": 3.49106778693205, + "grad_norm": 0.0005798477795906365, + "learning_rate": 0.13355156321373196, + "loss": 0.3679, + "num_input_tokens_seen": 44724144, + "step": 21400 + }, + { + "epoch": 3.49106778693205, + "eval_loss": 0.3164065182209015, + "eval_runtime": 156.1621, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 44724144, + "step": 21400 + }, + { + "epoch": 3.4918835141528675, + "grad_norm": 0.0005041967961005867, + "learning_rate": 0.13349301484285514, + "loss": 0.3024, + "num_input_tokens_seen": 44734976, + "step": 21405 + }, + { + "epoch": 3.4926992413736846, + "grad_norm": 0.00032465491676703095, + "learning_rate": 0.13343446901756295, + "loss": 0.2996, + "num_input_tokens_seen": 44744560, + "step": 21410 + }, + { + "epoch": 3.493514968594502, + "grad_norm": 0.0005937282112427056, + "learning_rate": 0.13337592574688376, + "loss": 0.2815, + "num_input_tokens_seen": 44753984, + "step": 21415 + }, + { + "epoch": 3.4943306958153193, + "grad_norm": 0.00040630149305798113, + "learning_rate": 0.13331738503984572, + "loss": 0.3404, + "num_input_tokens_seen": 44763552, + "step": 21420 + }, + { + "epoch": 3.495146423036137, + "grad_norm": 0.0012689067516475916, + "learning_rate": 0.1332588469054766, + "loss": 0.3803, + "num_input_tokens_seen": 44774832, + "step": 21425 + }, + { + "epoch": 3.495962150256954, + "grad_norm": 0.0006571935955435038, + "learning_rate": 0.1332003113528036, + "loss": 0.3013, + "num_input_tokens_seen": 44784448, + "step": 21430 + }, + { + "epoch": 3.4967778774777716, + "grad_norm": 0.0002253599086543545, + "learning_rate": 0.13314177839085373, + "loss": 0.3015, + "num_input_tokens_seen": 44795360, + "step": 21435 + }, + { + "epoch": 3.4975936046985887, + "grad_norm": 0.00036882684798911214, + "learning_rate": 0.13308324802865354, + "loss": 0.3393, + "num_input_tokens_seen": 44806224, + "step": 21440 + }, + { + "epoch": 3.4984093319194063, + "grad_norm": 0.0006661381339654326, + "learning_rate": 0.13302472027522905, + "loss": 0.31, + "num_input_tokens_seen": 44816992, + "step": 21445 + }, + { + "epoch": 3.4992250591402234, + "grad_norm": 0.0006770908366888762, + "learning_rate": 0.13296619513960606, + "loss": 0.3123, + "num_input_tokens_seen": 44827984, + "step": 21450 + }, + { + "epoch": 3.500040786361041, + "grad_norm": 0.00018831936176866293, + "learning_rate": 0.1329076726308098, + "loss": 0.3389, + "num_input_tokens_seen": 44838480, + "step": 21455 + }, + { + "epoch": 3.500856513581858, + "grad_norm": 0.00026876054471358657, + "learning_rate": 0.13284915275786519, + "loss": 0.3758, + "num_input_tokens_seen": 44847680, + "step": 21460 + }, + { + "epoch": 3.5016722408026757, + "grad_norm": 0.0007761467131786048, + "learning_rate": 0.1327906355297968, + "loss": 0.3316, + "num_input_tokens_seen": 44857872, + "step": 21465 + }, + { + "epoch": 3.502487968023493, + "grad_norm": 0.0002995729446411133, + "learning_rate": 0.13273212095562867, + "loss": 0.3386, + "num_input_tokens_seen": 44868096, + "step": 21470 + }, + { + "epoch": 3.5033036952443104, + "grad_norm": 0.00040398709825240076, + "learning_rate": 0.13267360904438444, + "loss": 0.2655, + "num_input_tokens_seen": 44878592, + "step": 21475 + }, + { + "epoch": 3.5041194224651275, + "grad_norm": 0.00044996757060289383, + "learning_rate": 0.1326150998050875, + "loss": 0.336, + "num_input_tokens_seen": 44888304, + "step": 21480 + }, + { + "epoch": 3.504935149685945, + "grad_norm": 0.00044821525807492435, + "learning_rate": 0.1325565932467606, + "loss": 0.3376, + "num_input_tokens_seen": 44897872, + "step": 21485 + }, + { + "epoch": 3.5057508769067622, + "grad_norm": 0.0003986148221883923, + "learning_rate": 0.13249808937842628, + "loss": 0.3286, + "num_input_tokens_seen": 44908912, + "step": 21490 + }, + { + "epoch": 3.50656660412758, + "grad_norm": 0.0008491904591210186, + "learning_rate": 0.1324395882091065, + "loss": 0.356, + "num_input_tokens_seen": 44919760, + "step": 21495 + }, + { + "epoch": 3.507382331348397, + "grad_norm": 0.0005912845954298973, + "learning_rate": 0.13238108974782284, + "loss": 0.3336, + "num_input_tokens_seen": 44930912, + "step": 21500 + }, + { + "epoch": 3.5081980585692145, + "grad_norm": 0.00042215822031721473, + "learning_rate": 0.13232259400359664, + "loss": 0.2584, + "num_input_tokens_seen": 44942144, + "step": 21505 + }, + { + "epoch": 3.5090137857900316, + "grad_norm": 0.0006810381892137229, + "learning_rate": 0.13226410098544852, + "loss": 0.3179, + "num_input_tokens_seen": 44953280, + "step": 21510 + }, + { + "epoch": 3.5098295130108492, + "grad_norm": 0.00042393378680571914, + "learning_rate": 0.13220561070239892, + "loss": 0.3913, + "num_input_tokens_seen": 44963136, + "step": 21515 + }, + { + "epoch": 3.510645240231667, + "grad_norm": 0.000748101097997278, + "learning_rate": 0.13214712316346783, + "loss": 0.3153, + "num_input_tokens_seen": 44973664, + "step": 21520 + }, + { + "epoch": 3.511460967452484, + "grad_norm": 0.0007714561652392149, + "learning_rate": 0.13208863837767465, + "loss": 0.3182, + "num_input_tokens_seen": 44983936, + "step": 21525 + }, + { + "epoch": 3.512276694673301, + "grad_norm": 0.00021466305770445615, + "learning_rate": 0.13203015635403856, + "loss": 0.3161, + "num_input_tokens_seen": 44994272, + "step": 21530 + }, + { + "epoch": 3.5130924218941186, + "grad_norm": 0.0005806322442367673, + "learning_rate": 0.13197167710157817, + "loss": 0.2777, + "num_input_tokens_seen": 45004848, + "step": 21535 + }, + { + "epoch": 3.513908149114936, + "grad_norm": 0.0010914687300100923, + "learning_rate": 0.13191320062931167, + "loss": 0.3771, + "num_input_tokens_seen": 45015872, + "step": 21540 + }, + { + "epoch": 3.5147238763357533, + "grad_norm": 0.00042828635196201503, + "learning_rate": 0.13185472694625702, + "loss": 0.3097, + "num_input_tokens_seen": 45027344, + "step": 21545 + }, + { + "epoch": 3.5155396035565705, + "grad_norm": 0.0010854130377992988, + "learning_rate": 0.13179625606143142, + "loss": 0.377, + "num_input_tokens_seen": 45037744, + "step": 21550 + }, + { + "epoch": 3.516355330777388, + "grad_norm": 0.00030180145404301584, + "learning_rate": 0.13173778798385188, + "loss": 0.3304, + "num_input_tokens_seen": 45048704, + "step": 21555 + }, + { + "epoch": 3.5171710579982056, + "grad_norm": 0.00042976587428711355, + "learning_rate": 0.13167932272253505, + "loss": 0.3178, + "num_input_tokens_seen": 45058864, + "step": 21560 + }, + { + "epoch": 3.5179867852190227, + "grad_norm": 0.0007669001934118569, + "learning_rate": 0.1316208602864968, + "loss": 0.3282, + "num_input_tokens_seen": 45069760, + "step": 21565 + }, + { + "epoch": 3.51880251243984, + "grad_norm": 0.00027483838493935764, + "learning_rate": 0.13156240068475292, + "loss": 0.3405, + "num_input_tokens_seen": 45078912, + "step": 21570 + }, + { + "epoch": 3.5196182396606575, + "grad_norm": 0.0010215984657406807, + "learning_rate": 0.1315039439263185, + "loss": 0.3395, + "num_input_tokens_seen": 45090800, + "step": 21575 + }, + { + "epoch": 3.520433966881475, + "grad_norm": 0.00035597500391304493, + "learning_rate": 0.13144549002020833, + "loss": 0.3371, + "num_input_tokens_seen": 45101264, + "step": 21580 + }, + { + "epoch": 3.521249694102292, + "grad_norm": 0.0004295765538699925, + "learning_rate": 0.13138703897543688, + "loss": 0.2878, + "num_input_tokens_seen": 45110288, + "step": 21585 + }, + { + "epoch": 3.5220654213231093, + "grad_norm": 0.0006396842654794455, + "learning_rate": 0.1313285908010178, + "loss": 0.2839, + "num_input_tokens_seen": 45121680, + "step": 21590 + }, + { + "epoch": 3.522881148543927, + "grad_norm": 0.0005071081686764956, + "learning_rate": 0.13127014550596475, + "loss": 0.3259, + "num_input_tokens_seen": 45132944, + "step": 21595 + }, + { + "epoch": 3.5236968757647444, + "grad_norm": 0.00045271997805684805, + "learning_rate": 0.1312117030992906, + "loss": 0.3277, + "num_input_tokens_seen": 45143632, + "step": 21600 + }, + { + "epoch": 3.5236968757647444, + "eval_loss": 0.3268171548843384, + "eval_runtime": 156.1711, + "eval_samples_per_second": 17.449, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 45143632, + "step": 21600 + }, + { + "epoch": 3.5245126029855616, + "grad_norm": 0.0003038057475350797, + "learning_rate": 0.13115326359000795, + "loss": 0.3404, + "num_input_tokens_seen": 45153424, + "step": 21605 + }, + { + "epoch": 3.525328330206379, + "grad_norm": 0.0013838739832863212, + "learning_rate": 0.13109482698712896, + "loss": 0.3329, + "num_input_tokens_seen": 45162688, + "step": 21610 + }, + { + "epoch": 3.5261440574271963, + "grad_norm": 0.0012370510958135128, + "learning_rate": 0.1310363932996651, + "loss": 0.3381, + "num_input_tokens_seen": 45173392, + "step": 21615 + }, + { + "epoch": 3.526959784648014, + "grad_norm": 0.0007720473222434521, + "learning_rate": 0.13097796253662775, + "loss": 0.285, + "num_input_tokens_seen": 45182640, + "step": 21620 + }, + { + "epoch": 3.527775511868831, + "grad_norm": 0.00028782698791474104, + "learning_rate": 0.1309195347070277, + "loss": 0.2989, + "num_input_tokens_seen": 45192848, + "step": 21625 + }, + { + "epoch": 3.5285912390896486, + "grad_norm": 0.00047488705604337156, + "learning_rate": 0.13086110981987506, + "loss": 0.3294, + "num_input_tokens_seen": 45203920, + "step": 21630 + }, + { + "epoch": 3.5294069663104657, + "grad_norm": 0.0009320095996372402, + "learning_rate": 0.13080268788417987, + "loss": 0.3202, + "num_input_tokens_seen": 45213280, + "step": 21635 + }, + { + "epoch": 3.5302226935312833, + "grad_norm": 0.000571827229578048, + "learning_rate": 0.1307442689089515, + "loss": 0.3233, + "num_input_tokens_seen": 45223712, + "step": 21640 + }, + { + "epoch": 3.5310384207521004, + "grad_norm": 0.0004982694517821074, + "learning_rate": 0.13068585290319873, + "loss": 0.2415, + "num_input_tokens_seen": 45234368, + "step": 21645 + }, + { + "epoch": 3.531854147972918, + "grad_norm": 0.0025416475255042315, + "learning_rate": 0.13062743987593026, + "loss": 0.3746, + "num_input_tokens_seen": 45244656, + "step": 21650 + }, + { + "epoch": 3.532669875193735, + "grad_norm": 0.0007005073130130768, + "learning_rate": 0.13056902983615395, + "loss": 0.3405, + "num_input_tokens_seen": 45255104, + "step": 21655 + }, + { + "epoch": 3.5334856024145527, + "grad_norm": 0.0013080146163702011, + "learning_rate": 0.13051062279287742, + "loss": 0.2916, + "num_input_tokens_seen": 45266464, + "step": 21660 + }, + { + "epoch": 3.53430132963537, + "grad_norm": 0.0008266912191174924, + "learning_rate": 0.13045221875510782, + "loss": 0.3493, + "num_input_tokens_seen": 45276544, + "step": 21665 + }, + { + "epoch": 3.5351170568561874, + "grad_norm": 0.0003221490769647062, + "learning_rate": 0.13039381773185174, + "loss": 0.3183, + "num_input_tokens_seen": 45287312, + "step": 21670 + }, + { + "epoch": 3.5359327840770045, + "grad_norm": 0.0005063117132522166, + "learning_rate": 0.1303354197321153, + "loss": 0.3346, + "num_input_tokens_seen": 45298960, + "step": 21675 + }, + { + "epoch": 3.536748511297822, + "grad_norm": 0.0007172428304329515, + "learning_rate": 0.13027702476490433, + "loss": 0.2761, + "num_input_tokens_seen": 45310352, + "step": 21680 + }, + { + "epoch": 3.537564238518639, + "grad_norm": 0.0002949567569885403, + "learning_rate": 0.1302186328392239, + "loss": 0.3044, + "num_input_tokens_seen": 45321376, + "step": 21685 + }, + { + "epoch": 3.538379965739457, + "grad_norm": 0.0003850517387036234, + "learning_rate": 0.130160243964079, + "loss": 0.2803, + "num_input_tokens_seen": 45332192, + "step": 21690 + }, + { + "epoch": 3.539195692960274, + "grad_norm": 0.0008541707065887749, + "learning_rate": 0.13010185814847372, + "loss": 0.3314, + "num_input_tokens_seen": 45342800, + "step": 21695 + }, + { + "epoch": 3.5400114201810915, + "grad_norm": 0.001318878261372447, + "learning_rate": 0.13004347540141192, + "loss": 0.4596, + "num_input_tokens_seen": 45354384, + "step": 21700 + }, + { + "epoch": 3.5408271474019086, + "grad_norm": 0.0002619923325255513, + "learning_rate": 0.12998509573189712, + "loss": 0.3695, + "num_input_tokens_seen": 45364752, + "step": 21705 + }, + { + "epoch": 3.541642874622726, + "grad_norm": 0.00023006365518085659, + "learning_rate": 0.12992671914893203, + "loss": 0.341, + "num_input_tokens_seen": 45374640, + "step": 21710 + }, + { + "epoch": 3.5424586018435438, + "grad_norm": 0.00023471086751669645, + "learning_rate": 0.12986834566151909, + "loss": 0.3136, + "num_input_tokens_seen": 45385920, + "step": 21715 + }, + { + "epoch": 3.543274329064361, + "grad_norm": 0.0005298955366015434, + "learning_rate": 0.12980997527866028, + "loss": 0.2738, + "num_input_tokens_seen": 45396848, + "step": 21720 + }, + { + "epoch": 3.544090056285178, + "grad_norm": 0.00041324732592329383, + "learning_rate": 0.12975160800935692, + "loss": 0.3121, + "num_input_tokens_seen": 45405856, + "step": 21725 + }, + { + "epoch": 3.5449057835059956, + "grad_norm": 0.00035318126901984215, + "learning_rate": 0.12969324386261016, + "loss": 0.2891, + "num_input_tokens_seen": 45417600, + "step": 21730 + }, + { + "epoch": 3.545721510726813, + "grad_norm": 0.00035708665382117033, + "learning_rate": 0.12963488284742034, + "loss": 0.3636, + "num_input_tokens_seen": 45427856, + "step": 21735 + }, + { + "epoch": 3.5465372379476303, + "grad_norm": 0.0006272736354731023, + "learning_rate": 0.12957652497278752, + "loss": 0.338, + "num_input_tokens_seen": 45437600, + "step": 21740 + }, + { + "epoch": 3.5473529651684474, + "grad_norm": 0.00041252278606407344, + "learning_rate": 0.12951817024771117, + "loss": 0.2741, + "num_input_tokens_seen": 45449312, + "step": 21745 + }, + { + "epoch": 3.548168692389265, + "grad_norm": 0.0005829560686834157, + "learning_rate": 0.12945981868119041, + "loss": 0.3368, + "num_input_tokens_seen": 45459520, + "step": 21750 + }, + { + "epoch": 3.5489844196100826, + "grad_norm": 0.00031470158137381077, + "learning_rate": 0.12940147028222376, + "loss": 0.2877, + "num_input_tokens_seen": 45470320, + "step": 21755 + }, + { + "epoch": 3.5498001468308997, + "grad_norm": 0.0006097813020460308, + "learning_rate": 0.12934312505980916, + "loss": 0.2817, + "num_input_tokens_seen": 45480576, + "step": 21760 + }, + { + "epoch": 3.550615874051717, + "grad_norm": 0.0017175842076539993, + "learning_rate": 0.1292847830229443, + "loss": 0.4002, + "num_input_tokens_seen": 45489344, + "step": 21765 + }, + { + "epoch": 3.5514316012725344, + "grad_norm": 0.0002628655347507447, + "learning_rate": 0.12922644418062626, + "loss": 0.2997, + "num_input_tokens_seen": 45500592, + "step": 21770 + }, + { + "epoch": 3.552247328493352, + "grad_norm": 0.0004094647883903235, + "learning_rate": 0.1291681085418515, + "loss": 0.2826, + "num_input_tokens_seen": 45511568, + "step": 21775 + }, + { + "epoch": 3.553063055714169, + "grad_norm": 0.0002092776558129117, + "learning_rate": 0.12910977611561628, + "loss": 0.3447, + "num_input_tokens_seen": 45521840, + "step": 21780 + }, + { + "epoch": 3.5538787829349863, + "grad_norm": 0.0003813588118646294, + "learning_rate": 0.1290514469109161, + "loss": 0.2985, + "num_input_tokens_seen": 45533632, + "step": 21785 + }, + { + "epoch": 3.554694510155804, + "grad_norm": 0.0006310080643743277, + "learning_rate": 0.128993120936746, + "loss": 0.3205, + "num_input_tokens_seen": 45544752, + "step": 21790 + }, + { + "epoch": 3.5555102373766214, + "grad_norm": 0.0002810269361361861, + "learning_rate": 0.12893479820210071, + "loss": 0.315, + "num_input_tokens_seen": 45556336, + "step": 21795 + }, + { + "epoch": 3.5563259645974385, + "grad_norm": 0.00031305215088650584, + "learning_rate": 0.1288764787159742, + "loss": 0.2847, + "num_input_tokens_seen": 45567152, + "step": 21800 + }, + { + "epoch": 3.5563259645974385, + "eval_loss": 0.3166733980178833, + "eval_runtime": 155.9684, + "eval_samples_per_second": 17.471, + "eval_steps_per_second": 8.739, + "num_input_tokens_seen": 45567152, + "step": 21800 + }, + { + "epoch": 3.557141691818256, + "grad_norm": 0.00032979718525893986, + "learning_rate": 0.1288181624873601, + "loss": 0.2867, + "num_input_tokens_seen": 45578368, + "step": 21805 + }, + { + "epoch": 3.5579574190390733, + "grad_norm": 0.0006463605095632374, + "learning_rate": 0.12875984952525163, + "loss": 0.3838, + "num_input_tokens_seen": 45588992, + "step": 21810 + }, + { + "epoch": 3.558773146259891, + "grad_norm": 0.0008221117313951254, + "learning_rate": 0.12870153983864122, + "loss": 0.3217, + "num_input_tokens_seen": 45599840, + "step": 21815 + }, + { + "epoch": 3.559588873480708, + "grad_norm": 0.0005640015588141978, + "learning_rate": 0.12864323343652104, + "loss": 0.302, + "num_input_tokens_seen": 45609632, + "step": 21820 + }, + { + "epoch": 3.5604046007015255, + "grad_norm": 0.0007584189297631383, + "learning_rate": 0.12858493032788268, + "loss": 0.2796, + "num_input_tokens_seen": 45620832, + "step": 21825 + }, + { + "epoch": 3.5612203279223427, + "grad_norm": 0.00028439308516681194, + "learning_rate": 0.12852663052171714, + "loss": 0.3497, + "num_input_tokens_seen": 45631616, + "step": 21830 + }, + { + "epoch": 3.5620360551431602, + "grad_norm": 0.000949226669035852, + "learning_rate": 0.12846833402701507, + "loss": 0.3205, + "num_input_tokens_seen": 45642048, + "step": 21835 + }, + { + "epoch": 3.5628517823639774, + "grad_norm": 0.00029514613561332226, + "learning_rate": 0.12841004085276642, + "loss": 0.3799, + "num_input_tokens_seen": 45653040, + "step": 21840 + }, + { + "epoch": 3.563667509584795, + "grad_norm": 0.00043300248216837645, + "learning_rate": 0.12835175100796076, + "loss": 0.31, + "num_input_tokens_seen": 45663344, + "step": 21845 + }, + { + "epoch": 3.564483236805612, + "grad_norm": 0.000487352313939482, + "learning_rate": 0.12829346450158724, + "loss": 0.2939, + "num_input_tokens_seen": 45672896, + "step": 21850 + }, + { + "epoch": 3.5652989640264297, + "grad_norm": 0.0006417327094823122, + "learning_rate": 0.12823518134263423, + "loss": 0.3039, + "num_input_tokens_seen": 45684048, + "step": 21855 + }, + { + "epoch": 3.566114691247247, + "grad_norm": 0.00033979484578594565, + "learning_rate": 0.12817690154008973, + "loss": 0.3117, + "num_input_tokens_seen": 45695088, + "step": 21860 + }, + { + "epoch": 3.5669304184680644, + "grad_norm": 0.00020859345386270434, + "learning_rate": 0.12811862510294134, + "loss": 0.2908, + "num_input_tokens_seen": 45705232, + "step": 21865 + }, + { + "epoch": 3.5677461456888815, + "grad_norm": 0.000392543530324474, + "learning_rate": 0.12806035204017585, + "loss": 0.3439, + "num_input_tokens_seen": 45715856, + "step": 21870 + }, + { + "epoch": 3.568561872909699, + "grad_norm": 0.00038417341420426965, + "learning_rate": 0.12800208236077987, + "loss": 0.3092, + "num_input_tokens_seen": 45725552, + "step": 21875 + }, + { + "epoch": 3.569377600130516, + "grad_norm": 0.00042664044303819537, + "learning_rate": 0.12794381607373917, + "loss": 0.3161, + "num_input_tokens_seen": 45737472, + "step": 21880 + }, + { + "epoch": 3.5701933273513338, + "grad_norm": 0.0010943347588181496, + "learning_rate": 0.12788555318803924, + "loss": 0.3739, + "num_input_tokens_seen": 45748224, + "step": 21885 + }, + { + "epoch": 3.5710090545721513, + "grad_norm": 0.0005741174682043493, + "learning_rate": 0.1278272937126649, + "loss": 0.305, + "num_input_tokens_seen": 45757696, + "step": 21890 + }, + { + "epoch": 3.5718247817929685, + "grad_norm": 0.0004123554390389472, + "learning_rate": 0.1277690376566005, + "loss": 0.3383, + "num_input_tokens_seen": 45769200, + "step": 21895 + }, + { + "epoch": 3.5726405090137856, + "grad_norm": 0.0003137555031571537, + "learning_rate": 0.12771078502882985, + "loss": 0.3005, + "num_input_tokens_seen": 45779520, + "step": 21900 + }, + { + "epoch": 3.573456236234603, + "grad_norm": 0.0006370576447807252, + "learning_rate": 0.12765253583833633, + "loss": 0.3273, + "num_input_tokens_seen": 45790048, + "step": 21905 + }, + { + "epoch": 3.5742719634554208, + "grad_norm": 0.0005005900165997446, + "learning_rate": 0.12759429009410256, + "loss": 0.3124, + "num_input_tokens_seen": 45800144, + "step": 21910 + }, + { + "epoch": 3.575087690676238, + "grad_norm": 0.00033893444924615324, + "learning_rate": 0.12753604780511085, + "loss": 0.3012, + "num_input_tokens_seen": 45809344, + "step": 21915 + }, + { + "epoch": 3.575903417897055, + "grad_norm": 0.0002895543584600091, + "learning_rate": 0.12747780898034283, + "loss": 0.2622, + "num_input_tokens_seen": 45820096, + "step": 21920 + }, + { + "epoch": 3.5767191451178726, + "grad_norm": 0.0009882779559120536, + "learning_rate": 0.12741957362877973, + "loss": 0.3407, + "num_input_tokens_seen": 45830448, + "step": 21925 + }, + { + "epoch": 3.57753487233869, + "grad_norm": 0.00029949116287752986, + "learning_rate": 0.12736134175940214, + "loss": 0.3345, + "num_input_tokens_seen": 45840032, + "step": 21930 + }, + { + "epoch": 3.5783505995595073, + "grad_norm": 0.0002730423875618726, + "learning_rate": 0.12730311338119016, + "loss": 0.3356, + "num_input_tokens_seen": 45850704, + "step": 21935 + }, + { + "epoch": 3.5791663267803244, + "grad_norm": 0.00038654039963148534, + "learning_rate": 0.12724488850312327, + "loss": 0.2968, + "num_input_tokens_seen": 45861296, + "step": 21940 + }, + { + "epoch": 3.579982054001142, + "grad_norm": 0.0005076323868706822, + "learning_rate": 0.1271866671341806, + "loss": 0.2766, + "num_input_tokens_seen": 45869984, + "step": 21945 + }, + { + "epoch": 3.5807977812219596, + "grad_norm": 0.0003296982613392174, + "learning_rate": 0.12712844928334047, + "loss": 0.3102, + "num_input_tokens_seen": 45879968, + "step": 21950 + }, + { + "epoch": 3.5816135084427767, + "grad_norm": 0.00030893119401298463, + "learning_rate": 0.12707023495958095, + "loss": 0.3093, + "num_input_tokens_seen": 45890608, + "step": 21955 + }, + { + "epoch": 3.582429235663594, + "grad_norm": 0.00027349081938154995, + "learning_rate": 0.12701202417187932, + "loss": 0.3107, + "num_input_tokens_seen": 45901504, + "step": 21960 + }, + { + "epoch": 3.5832449628844114, + "grad_norm": 0.00038330804090946913, + "learning_rate": 0.12695381692921243, + "loss": 0.3372, + "num_input_tokens_seen": 45911232, + "step": 21965 + }, + { + "epoch": 3.584060690105229, + "grad_norm": 0.0004729986540041864, + "learning_rate": 0.12689561324055665, + "loss": 0.352, + "num_input_tokens_seen": 45922368, + "step": 21970 + }, + { + "epoch": 3.584876417326046, + "grad_norm": 0.0006264089024625719, + "learning_rate": 0.12683741311488758, + "loss": 0.3208, + "num_input_tokens_seen": 45932448, + "step": 21975 + }, + { + "epoch": 3.5856921445468637, + "grad_norm": 0.0006246205302886665, + "learning_rate": 0.1267792165611805, + "loss": 0.2227, + "num_input_tokens_seen": 45942560, + "step": 21980 + }, + { + "epoch": 3.586507871767681, + "grad_norm": 0.0004290815268177539, + "learning_rate": 0.1267210235884101, + "loss": 0.3457, + "num_input_tokens_seen": 45951824, + "step": 21985 + }, + { + "epoch": 3.5873235989884984, + "grad_norm": 0.0008867810247465968, + "learning_rate": 0.12666283420555033, + "loss": 0.3347, + "num_input_tokens_seen": 45962640, + "step": 21990 + }, + { + "epoch": 3.5881393262093155, + "grad_norm": 0.0009704211843200028, + "learning_rate": 0.12660464842157487, + "loss": 0.3522, + "num_input_tokens_seen": 45972624, + "step": 21995 + }, + { + "epoch": 3.588955053430133, + "grad_norm": 0.0008914792560972273, + "learning_rate": 0.1265464662454566, + "loss": 0.3947, + "num_input_tokens_seen": 45983168, + "step": 22000 + }, + { + "epoch": 3.588955053430133, + "eval_loss": 0.3164173662662506, + "eval_runtime": 156.0881, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 45983168, + "step": 22000 + }, + { + "epoch": 3.5897707806509502, + "grad_norm": 0.00028484826907515526, + "learning_rate": 0.12648828768616793, + "loss": 0.3446, + "num_input_tokens_seen": 45994816, + "step": 22005 + }, + { + "epoch": 3.590586507871768, + "grad_norm": 0.0005037600640207529, + "learning_rate": 0.12643011275268085, + "loss": 0.2929, + "num_input_tokens_seen": 46004384, + "step": 22010 + }, + { + "epoch": 3.591402235092585, + "grad_norm": 0.00033522190642543137, + "learning_rate": 0.1263719414539665, + "loss": 0.3572, + "num_input_tokens_seen": 46015552, + "step": 22015 + }, + { + "epoch": 3.5922179623134025, + "grad_norm": 0.0006595855811610818, + "learning_rate": 0.1263137737989957, + "loss": 0.3207, + "num_input_tokens_seen": 46025024, + "step": 22020 + }, + { + "epoch": 3.5930336895342196, + "grad_norm": 0.000564955233130604, + "learning_rate": 0.1262556097967387, + "loss": 0.374, + "num_input_tokens_seen": 46035584, + "step": 22025 + }, + { + "epoch": 3.593849416755037, + "grad_norm": 0.000742718402761966, + "learning_rate": 0.126197449456165, + "loss": 0.2897, + "num_input_tokens_seen": 46046112, + "step": 22030 + }, + { + "epoch": 3.5946651439758543, + "grad_norm": 0.0004086294211447239, + "learning_rate": 0.12613929278624378, + "loss": 0.3167, + "num_input_tokens_seen": 46057376, + "step": 22035 + }, + { + "epoch": 3.595480871196672, + "grad_norm": 0.00046951932017691433, + "learning_rate": 0.12608113979594343, + "loss": 0.3178, + "num_input_tokens_seen": 46068944, + "step": 22040 + }, + { + "epoch": 3.596296598417489, + "grad_norm": 0.0007878296310082078, + "learning_rate": 0.1260229904942319, + "loss": 0.3899, + "num_input_tokens_seen": 46080352, + "step": 22045 + }, + { + "epoch": 3.5971123256383066, + "grad_norm": 0.0003881620359607041, + "learning_rate": 0.12596484489007662, + "loss": 0.3484, + "num_input_tokens_seen": 46089872, + "step": 22050 + }, + { + "epoch": 3.5979280528591238, + "grad_norm": 0.0010594954947009683, + "learning_rate": 0.1259067029924442, + "loss": 0.3408, + "num_input_tokens_seen": 46098160, + "step": 22055 + }, + { + "epoch": 3.5987437800799413, + "grad_norm": 0.0010443313512951136, + "learning_rate": 0.12584856481030096, + "loss": 0.3802, + "num_input_tokens_seen": 46108704, + "step": 22060 + }, + { + "epoch": 3.5995595073007585, + "grad_norm": 0.0003208141715731472, + "learning_rate": 0.12579043035261261, + "loss": 0.2824, + "num_input_tokens_seen": 46119664, + "step": 22065 + }, + { + "epoch": 3.600375234521576, + "grad_norm": 0.00028068566462025046, + "learning_rate": 0.1257322996283441, + "loss": 0.2773, + "num_input_tokens_seen": 46128512, + "step": 22070 + }, + { + "epoch": 3.601190961742393, + "grad_norm": 0.0004784950288012624, + "learning_rate": 0.12567417264645994, + "loss": 0.3683, + "num_input_tokens_seen": 46137680, + "step": 22075 + }, + { + "epoch": 3.6020066889632107, + "grad_norm": 0.0007454033475369215, + "learning_rate": 0.12561604941592408, + "loss": 0.2934, + "num_input_tokens_seen": 46148512, + "step": 22080 + }, + { + "epoch": 3.6028224161840283, + "grad_norm": 0.0009585680090822279, + "learning_rate": 0.12555792994569978, + "loss": 0.2644, + "num_input_tokens_seen": 46160000, + "step": 22085 + }, + { + "epoch": 3.6036381434048455, + "grad_norm": 0.000532169418875128, + "learning_rate": 0.1254998142447499, + "loss": 0.3084, + "num_input_tokens_seen": 46170320, + "step": 22090 + }, + { + "epoch": 3.6044538706256626, + "grad_norm": 0.0005534370429813862, + "learning_rate": 0.1254417023220365, + "loss": 0.3102, + "num_input_tokens_seen": 46181552, + "step": 22095 + }, + { + "epoch": 3.60526959784648, + "grad_norm": 0.0004121322126593441, + "learning_rate": 0.12538359418652126, + "loss": 0.3563, + "num_input_tokens_seen": 46191984, + "step": 22100 + }, + { + "epoch": 3.6060853250672977, + "grad_norm": 0.00066607870394364, + "learning_rate": 0.12532548984716513, + "loss": 0.326, + "num_input_tokens_seen": 46201616, + "step": 22105 + }, + { + "epoch": 3.606901052288115, + "grad_norm": 0.0006586306262761354, + "learning_rate": 0.12526738931292855, + "loss": 0.3719, + "num_input_tokens_seen": 46211504, + "step": 22110 + }, + { + "epoch": 3.607716779508932, + "grad_norm": 0.0006702328682877123, + "learning_rate": 0.1252092925927714, + "loss": 0.2944, + "num_input_tokens_seen": 46222768, + "step": 22115 + }, + { + "epoch": 3.6085325067297496, + "grad_norm": 0.00021205875964369625, + "learning_rate": 0.12515119969565278, + "loss": 0.3058, + "num_input_tokens_seen": 46233168, + "step": 22120 + }, + { + "epoch": 3.609348233950567, + "grad_norm": 0.000649003719445318, + "learning_rate": 0.12509311063053144, + "loss": 0.2989, + "num_input_tokens_seen": 46243104, + "step": 22125 + }, + { + "epoch": 3.6101639611713843, + "grad_norm": 0.00036499687121249735, + "learning_rate": 0.1250350254063655, + "loss": 0.3044, + "num_input_tokens_seen": 46254512, + "step": 22130 + }, + { + "epoch": 3.6109796883922014, + "grad_norm": 0.000842325680423528, + "learning_rate": 0.1249769440321123, + "loss": 0.3278, + "num_input_tokens_seen": 46264352, + "step": 22135 + }, + { + "epoch": 3.611795415613019, + "grad_norm": 0.0010757397394627333, + "learning_rate": 0.12491886651672884, + "loss": 0.3526, + "num_input_tokens_seen": 46274736, + "step": 22140 + }, + { + "epoch": 3.6126111428338366, + "grad_norm": 0.0007459368789568543, + "learning_rate": 0.12486079286917139, + "loss": 0.3746, + "num_input_tokens_seen": 46285680, + "step": 22145 + }, + { + "epoch": 3.6134268700546537, + "grad_norm": 0.0006586355739273131, + "learning_rate": 0.12480272309839553, + "loss": 0.3122, + "num_input_tokens_seen": 46296784, + "step": 22150 + }, + { + "epoch": 3.614242597275471, + "grad_norm": 0.00029965213616378605, + "learning_rate": 0.12474465721335648, + "loss": 0.3117, + "num_input_tokens_seen": 46307888, + "step": 22155 + }, + { + "epoch": 3.6150583244962884, + "grad_norm": 0.0003224940737709403, + "learning_rate": 0.12468659522300861, + "loss": 0.3417, + "num_input_tokens_seen": 46319136, + "step": 22160 + }, + { + "epoch": 3.615874051717106, + "grad_norm": 0.0003270031011197716, + "learning_rate": 0.12462853713630584, + "loss": 0.3504, + "num_input_tokens_seen": 46329120, + "step": 22165 + }, + { + "epoch": 3.616689778937923, + "grad_norm": 0.0002832117897924036, + "learning_rate": 0.12457048296220156, + "loss": 0.3235, + "num_input_tokens_seen": 46339168, + "step": 22170 + }, + { + "epoch": 3.6175055061587407, + "grad_norm": 0.0004096023621968925, + "learning_rate": 0.12451243270964832, + "loss": 0.3725, + "num_input_tokens_seen": 46350352, + "step": 22175 + }, + { + "epoch": 3.618321233379558, + "grad_norm": 0.0004354656848590821, + "learning_rate": 0.12445438638759827, + "loss": 0.3499, + "num_input_tokens_seen": 46361728, + "step": 22180 + }, + { + "epoch": 3.6191369606003754, + "grad_norm": 0.00024123743060044944, + "learning_rate": 0.1243963440050029, + "loss": 0.3073, + "num_input_tokens_seen": 46372224, + "step": 22185 + }, + { + "epoch": 3.6199526878211925, + "grad_norm": 0.0003374142979737371, + "learning_rate": 0.12433830557081298, + "loss": 0.3382, + "num_input_tokens_seen": 46382016, + "step": 22190 + }, + { + "epoch": 3.62076841504201, + "grad_norm": 0.0005770173738710582, + "learning_rate": 0.12428027109397889, + "loss": 0.3134, + "num_input_tokens_seen": 46391904, + "step": 22195 + }, + { + "epoch": 3.621584142262827, + "grad_norm": 0.0004850990080740303, + "learning_rate": 0.12422224058345015, + "loss": 0.3045, + "num_input_tokens_seen": 46401184, + "step": 22200 + }, + { + "epoch": 3.621584142262827, + "eval_loss": 0.31893888115882874, + "eval_runtime": 156.0912, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 46401184, + "step": 22200 + }, + { + "epoch": 3.622399869483645, + "grad_norm": 0.0003882818855345249, + "learning_rate": 0.12416421404817583, + "loss": 0.3329, + "num_input_tokens_seen": 46411840, + "step": 22205 + }, + { + "epoch": 3.623215596704462, + "grad_norm": 0.0003572822315618396, + "learning_rate": 0.12410619149710447, + "loss": 0.3706, + "num_input_tokens_seen": 46423152, + "step": 22210 + }, + { + "epoch": 3.6240313239252795, + "grad_norm": 0.0003153097932226956, + "learning_rate": 0.12404817293918374, + "loss": 0.2926, + "num_input_tokens_seen": 46432656, + "step": 22215 + }, + { + "epoch": 3.6248470511460966, + "grad_norm": 0.00038452877197414637, + "learning_rate": 0.12399015838336086, + "loss": 0.2916, + "num_input_tokens_seen": 46443392, + "step": 22220 + }, + { + "epoch": 3.625662778366914, + "grad_norm": 0.0005591498920693994, + "learning_rate": 0.12393214783858246, + "loss": 0.2827, + "num_input_tokens_seen": 46453296, + "step": 22225 + }, + { + "epoch": 3.6264785055877313, + "grad_norm": 0.0005141918081790209, + "learning_rate": 0.1238741413137944, + "loss": 0.296, + "num_input_tokens_seen": 46464000, + "step": 22230 + }, + { + "epoch": 3.627294232808549, + "grad_norm": 0.0003797431418206543, + "learning_rate": 0.12381613881794212, + "loss": 0.387, + "num_input_tokens_seen": 46475120, + "step": 22235 + }, + { + "epoch": 3.628109960029366, + "grad_norm": 0.00027089647483080626, + "learning_rate": 0.12375814035997022, + "loss": 0.3165, + "num_input_tokens_seen": 46486000, + "step": 22240 + }, + { + "epoch": 3.6289256872501836, + "grad_norm": 0.00029402683139778674, + "learning_rate": 0.12370014594882285, + "loss": 0.3335, + "num_input_tokens_seen": 46496288, + "step": 22245 + }, + { + "epoch": 3.6297414144710007, + "grad_norm": 0.0005969941266812384, + "learning_rate": 0.12364215559344356, + "loss": 0.327, + "num_input_tokens_seen": 46504000, + "step": 22250 + }, + { + "epoch": 3.6305571416918183, + "grad_norm": 0.0006705625564791262, + "learning_rate": 0.12358416930277506, + "loss": 0.3757, + "num_input_tokens_seen": 46513056, + "step": 22255 + }, + { + "epoch": 3.631372868912636, + "grad_norm": 0.00027562680770643055, + "learning_rate": 0.1235261870857596, + "loss": 0.3015, + "num_input_tokens_seen": 46523488, + "step": 22260 + }, + { + "epoch": 3.632188596133453, + "grad_norm": 0.0002522999420762062, + "learning_rate": 0.12346820895133884, + "loss": 0.3309, + "num_input_tokens_seen": 46533808, + "step": 22265 + }, + { + "epoch": 3.63300432335427, + "grad_norm": 0.0004432762216310948, + "learning_rate": 0.12341023490845361, + "loss": 0.3442, + "num_input_tokens_seen": 46544672, + "step": 22270 + }, + { + "epoch": 3.6338200505750877, + "grad_norm": 0.0006239071954041719, + "learning_rate": 0.12335226496604437, + "loss": 0.2868, + "num_input_tokens_seen": 46554160, + "step": 22275 + }, + { + "epoch": 3.6346357777959053, + "grad_norm": 0.0006272261962294579, + "learning_rate": 0.12329429913305069, + "loss": 0.3679, + "num_input_tokens_seen": 46564672, + "step": 22280 + }, + { + "epoch": 3.6354515050167224, + "grad_norm": 0.0005595466354861856, + "learning_rate": 0.12323633741841171, + "loss": 0.2973, + "num_input_tokens_seen": 46574400, + "step": 22285 + }, + { + "epoch": 3.6362672322375396, + "grad_norm": 0.00024690982536412776, + "learning_rate": 0.12317837983106583, + "loss": 0.3411, + "num_input_tokens_seen": 46583776, + "step": 22290 + }, + { + "epoch": 3.637082959458357, + "grad_norm": 0.000486353732412681, + "learning_rate": 0.12312042637995087, + "loss": 0.3427, + "num_input_tokens_seen": 46593552, + "step": 22295 + }, + { + "epoch": 3.6378986866791747, + "grad_norm": 0.0005953838117420673, + "learning_rate": 0.12306247707400389, + "loss": 0.3672, + "num_input_tokens_seen": 46603408, + "step": 22300 + }, + { + "epoch": 3.638714413899992, + "grad_norm": 0.0003890554653480649, + "learning_rate": 0.12300453192216154, + "loss": 0.3088, + "num_input_tokens_seen": 46614000, + "step": 22305 + }, + { + "epoch": 3.639530141120809, + "grad_norm": 0.00023177887487690896, + "learning_rate": 0.12294659093335956, + "loss": 0.3622, + "num_input_tokens_seen": 46624192, + "step": 22310 + }, + { + "epoch": 3.6403458683416265, + "grad_norm": 0.0006611443241126835, + "learning_rate": 0.12288865411653327, + "loss": 0.3695, + "num_input_tokens_seen": 46635600, + "step": 22315 + }, + { + "epoch": 3.641161595562444, + "grad_norm": 0.00039202201878651977, + "learning_rate": 0.12283072148061717, + "loss": 0.3088, + "num_input_tokens_seen": 46646656, + "step": 22320 + }, + { + "epoch": 3.6419773227832613, + "grad_norm": 0.00033711688593029976, + "learning_rate": 0.12277279303454529, + "loss": 0.3213, + "num_input_tokens_seen": 46657600, + "step": 22325 + }, + { + "epoch": 3.6427930500040784, + "grad_norm": 0.0007511196308769286, + "learning_rate": 0.12271486878725091, + "loss": 0.312, + "num_input_tokens_seen": 46669040, + "step": 22330 + }, + { + "epoch": 3.643608777224896, + "grad_norm": 0.000831638288218528, + "learning_rate": 0.12265694874766658, + "loss": 0.3269, + "num_input_tokens_seen": 46679792, + "step": 22335 + }, + { + "epoch": 3.6444245044457135, + "grad_norm": 0.0002955761447083205, + "learning_rate": 0.12259903292472435, + "loss": 0.382, + "num_input_tokens_seen": 46689280, + "step": 22340 + }, + { + "epoch": 3.6452402316665307, + "grad_norm": 0.000590614799875766, + "learning_rate": 0.12254112132735567, + "loss": 0.3698, + "num_input_tokens_seen": 46697968, + "step": 22345 + }, + { + "epoch": 3.6460559588873482, + "grad_norm": 0.0002927101159002632, + "learning_rate": 0.12248321396449108, + "loss": 0.3257, + "num_input_tokens_seen": 46708416, + "step": 22350 + }, + { + "epoch": 3.6468716861081654, + "grad_norm": 0.00040521653136238456, + "learning_rate": 0.12242531084506075, + "loss": 0.3168, + "num_input_tokens_seen": 46718432, + "step": 22355 + }, + { + "epoch": 3.647687413328983, + "grad_norm": 0.0003300705866422504, + "learning_rate": 0.122367411977994, + "loss": 0.3096, + "num_input_tokens_seen": 46729008, + "step": 22360 + }, + { + "epoch": 3.6485031405498, + "grad_norm": 0.000900304177775979, + "learning_rate": 0.12230951737221954, + "loss": 0.3518, + "num_input_tokens_seen": 46740064, + "step": 22365 + }, + { + "epoch": 3.6493188677706176, + "grad_norm": 0.0006376633537001908, + "learning_rate": 0.12225162703666555, + "loss": 0.3326, + "num_input_tokens_seen": 46749568, + "step": 22370 + }, + { + "epoch": 3.6501345949914348, + "grad_norm": 0.0003573862777557224, + "learning_rate": 0.1221937409802593, + "loss": 0.342, + "num_input_tokens_seen": 46760304, + "step": 22375 + }, + { + "epoch": 3.6509503222122524, + "grad_norm": 0.00021360567188821733, + "learning_rate": 0.12213585921192768, + "loss": 0.3059, + "num_input_tokens_seen": 46771072, + "step": 22380 + }, + { + "epoch": 3.6517660494330695, + "grad_norm": 0.00041278163553215563, + "learning_rate": 0.1220779817405967, + "loss": 0.3518, + "num_input_tokens_seen": 46781856, + "step": 22385 + }, + { + "epoch": 3.652581776653887, + "grad_norm": 0.0018922945018857718, + "learning_rate": 0.12202010857519181, + "loss": 0.3425, + "num_input_tokens_seen": 46792112, + "step": 22390 + }, + { + "epoch": 3.653397503874704, + "grad_norm": 0.0006462865858338773, + "learning_rate": 0.12196223972463785, + "loss": 0.2615, + "num_input_tokens_seen": 46801664, + "step": 22395 + }, + { + "epoch": 3.6542132310955218, + "grad_norm": 0.0003618149785324931, + "learning_rate": 0.12190437519785885, + "loss": 0.32, + "num_input_tokens_seen": 46813008, + "step": 22400 + }, + { + "epoch": 3.6542132310955218, + "eval_loss": 0.32016995549201965, + "eval_runtime": 156.0846, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 46813008, + "step": 22400 + }, + { + "epoch": 3.655028958316339, + "grad_norm": 0.0007280407007783651, + "learning_rate": 0.12184651500377823, + "loss": 0.3442, + "num_input_tokens_seen": 46824208, + "step": 22405 + }, + { + "epoch": 3.6558446855371565, + "grad_norm": 0.0005488039460033178, + "learning_rate": 0.12178865915131885, + "loss": 0.2778, + "num_input_tokens_seen": 46833312, + "step": 22410 + }, + { + "epoch": 3.6566604127579736, + "grad_norm": 0.00031363306334242225, + "learning_rate": 0.1217308076494027, + "loss": 0.341, + "num_input_tokens_seen": 46843360, + "step": 22415 + }, + { + "epoch": 3.657476139978791, + "grad_norm": 0.0008659733575768769, + "learning_rate": 0.12167296050695134, + "loss": 0.3355, + "num_input_tokens_seen": 46854240, + "step": 22420 + }, + { + "epoch": 3.6582918671996083, + "grad_norm": 0.0004197793605271727, + "learning_rate": 0.12161511773288536, + "loss": 0.3054, + "num_input_tokens_seen": 46864784, + "step": 22425 + }, + { + "epoch": 3.659107594420426, + "grad_norm": 0.0008280001347884536, + "learning_rate": 0.121557279336125, + "loss": 0.3619, + "num_input_tokens_seen": 46876352, + "step": 22430 + }, + { + "epoch": 3.659923321641243, + "grad_norm": 0.00034267615410499275, + "learning_rate": 0.12149944532558957, + "loss": 0.2952, + "num_input_tokens_seen": 46886896, + "step": 22435 + }, + { + "epoch": 3.6607390488620606, + "grad_norm": 0.0006626941030845046, + "learning_rate": 0.12144161571019785, + "loss": 0.2997, + "num_input_tokens_seen": 46898576, + "step": 22440 + }, + { + "epoch": 3.6615547760828777, + "grad_norm": 0.0002700702752918005, + "learning_rate": 0.12138379049886781, + "loss": 0.2964, + "num_input_tokens_seen": 46908752, + "step": 22445 + }, + { + "epoch": 3.6623705033036953, + "grad_norm": 0.0006833079387433827, + "learning_rate": 0.12132596970051697, + "loss": 0.2818, + "num_input_tokens_seen": 46918848, + "step": 22450 + }, + { + "epoch": 3.663186230524513, + "grad_norm": 0.0007215318037196994, + "learning_rate": 0.12126815332406189, + "loss": 0.2737, + "num_input_tokens_seen": 46929664, + "step": 22455 + }, + { + "epoch": 3.66400195774533, + "grad_norm": 0.000601316278334707, + "learning_rate": 0.12121034137841868, + "loss": 0.312, + "num_input_tokens_seen": 46939664, + "step": 22460 + }, + { + "epoch": 3.664817684966147, + "grad_norm": 0.0010093547170981765, + "learning_rate": 0.12115253387250258, + "loss": 0.3356, + "num_input_tokens_seen": 46950912, + "step": 22465 + }, + { + "epoch": 3.6656334121869647, + "grad_norm": 0.0006390360067598522, + "learning_rate": 0.12109473081522831, + "loss": 0.2683, + "num_input_tokens_seen": 46960272, + "step": 22470 + }, + { + "epoch": 3.6664491394077823, + "grad_norm": 0.00034099718322977424, + "learning_rate": 0.12103693221550982, + "loss": 0.3873, + "num_input_tokens_seen": 46970144, + "step": 22475 + }, + { + "epoch": 3.6672648666285994, + "grad_norm": 0.0013111162697896361, + "learning_rate": 0.12097913808226027, + "loss": 0.307, + "num_input_tokens_seen": 46981040, + "step": 22480 + }, + { + "epoch": 3.6680805938494165, + "grad_norm": 0.0006426605978049338, + "learning_rate": 0.12092134842439234, + "loss": 0.3363, + "num_input_tokens_seen": 46989744, + "step": 22485 + }, + { + "epoch": 3.668896321070234, + "grad_norm": 0.0003331816114950925, + "learning_rate": 0.12086356325081798, + "loss": 0.2665, + "num_input_tokens_seen": 47001376, + "step": 22490 + }, + { + "epoch": 3.6697120482910517, + "grad_norm": 0.00039508056943304837, + "learning_rate": 0.12080578257044824, + "loss": 0.3036, + "num_input_tokens_seen": 47012816, + "step": 22495 + }, + { + "epoch": 3.670527775511869, + "grad_norm": 0.0002714544825721532, + "learning_rate": 0.12074800639219378, + "loss": 0.3741, + "num_input_tokens_seen": 47023488, + "step": 22500 + }, + { + "epoch": 3.671343502732686, + "grad_norm": 0.0003124819486401975, + "learning_rate": 0.12069023472496428, + "loss": 0.3331, + "num_input_tokens_seen": 47033200, + "step": 22505 + }, + { + "epoch": 3.6721592299535035, + "grad_norm": 0.0005304437363520265, + "learning_rate": 0.12063246757766893, + "loss": 0.3399, + "num_input_tokens_seen": 47043936, + "step": 22510 + }, + { + "epoch": 3.672974957174321, + "grad_norm": 0.00040548323886469007, + "learning_rate": 0.12057470495921618, + "loss": 0.3033, + "num_input_tokens_seen": 47053712, + "step": 22515 + }, + { + "epoch": 3.6737906843951382, + "grad_norm": 0.00039360052323900163, + "learning_rate": 0.12051694687851364, + "loss": 0.3368, + "num_input_tokens_seen": 47063168, + "step": 22520 + }, + { + "epoch": 3.6746064116159554, + "grad_norm": 0.0003363913274370134, + "learning_rate": 0.12045919334446839, + "loss": 0.2881, + "num_input_tokens_seen": 47073968, + "step": 22525 + }, + { + "epoch": 3.675422138836773, + "grad_norm": 0.000851708697155118, + "learning_rate": 0.12040144436598683, + "loss": 0.3083, + "num_input_tokens_seen": 47084944, + "step": 22530 + }, + { + "epoch": 3.6762378660575905, + "grad_norm": 0.0007169767050072551, + "learning_rate": 0.12034369995197444, + "loss": 0.3199, + "num_input_tokens_seen": 47095888, + "step": 22535 + }, + { + "epoch": 3.6770535932784076, + "grad_norm": 0.0005994650418870151, + "learning_rate": 0.12028596011133627, + "loss": 0.2768, + "num_input_tokens_seen": 47106384, + "step": 22540 + }, + { + "epoch": 3.677869320499225, + "grad_norm": 0.0002904511638917029, + "learning_rate": 0.12022822485297643, + "loss": 0.3054, + "num_input_tokens_seen": 47117376, + "step": 22545 + }, + { + "epoch": 3.6786850477200423, + "grad_norm": 0.00036681865458376706, + "learning_rate": 0.12017049418579843, + "loss": 0.3409, + "num_input_tokens_seen": 47127712, + "step": 22550 + }, + { + "epoch": 3.67950077494086, + "grad_norm": 0.000678130891174078, + "learning_rate": 0.12011276811870514, + "loss": 0.3101, + "num_input_tokens_seen": 47136112, + "step": 22555 + }, + { + "epoch": 3.680316502161677, + "grad_norm": 0.00030551644158549607, + "learning_rate": 0.12005504666059852, + "loss": 0.2739, + "num_input_tokens_seen": 47147904, + "step": 22560 + }, + { + "epoch": 3.6811322293824946, + "grad_norm": 0.0004742716555483639, + "learning_rate": 0.11999732982038003, + "loss": 0.2687, + "num_input_tokens_seen": 47158112, + "step": 22565 + }, + { + "epoch": 3.6819479566033118, + "grad_norm": 0.0006257694913074374, + "learning_rate": 0.11993961760695038, + "loss": 0.2998, + "num_input_tokens_seen": 47168512, + "step": 22570 + }, + { + "epoch": 3.6827636838241293, + "grad_norm": 0.0005915833753533661, + "learning_rate": 0.11988191002920942, + "loss": 0.3112, + "num_input_tokens_seen": 47178176, + "step": 22575 + }, + { + "epoch": 3.6835794110449465, + "grad_norm": 0.0006608418188989162, + "learning_rate": 0.11982420709605641, + "loss": 0.3249, + "num_input_tokens_seen": 47190304, + "step": 22580 + }, + { + "epoch": 3.684395138265764, + "grad_norm": 0.0004464154480956495, + "learning_rate": 0.11976650881638991, + "loss": 0.3165, + "num_input_tokens_seen": 47201680, + "step": 22585 + }, + { + "epoch": 3.685210865486581, + "grad_norm": 0.000527604075614363, + "learning_rate": 0.11970881519910764, + "loss": 0.3209, + "num_input_tokens_seen": 47213200, + "step": 22590 + }, + { + "epoch": 3.6860265927073987, + "grad_norm": 0.00036331749288365245, + "learning_rate": 0.1196511262531068, + "loss": 0.3058, + "num_input_tokens_seen": 47224736, + "step": 22595 + }, + { + "epoch": 3.686842319928216, + "grad_norm": 0.0008520230185240507, + "learning_rate": 0.11959344198728361, + "loss": 0.3964, + "num_input_tokens_seen": 47233968, + "step": 22600 + }, + { + "epoch": 3.686842319928216, + "eval_loss": 0.3140667676925659, + "eval_runtime": 156.2095, + "eval_samples_per_second": 17.445, + "eval_steps_per_second": 8.725, + "num_input_tokens_seen": 47233968, + "step": 22600 + }, + { + "epoch": 3.6876580471490334, + "grad_norm": 0.0004810822429135442, + "learning_rate": 0.11953576241053378, + "loss": 0.3141, + "num_input_tokens_seen": 47242496, + "step": 22605 + }, + { + "epoch": 3.6884737743698506, + "grad_norm": 0.0005439265514723957, + "learning_rate": 0.11947808753175228, + "loss": 0.3737, + "num_input_tokens_seen": 47254336, + "step": 22610 + }, + { + "epoch": 3.689289501590668, + "grad_norm": 0.0005817650817334652, + "learning_rate": 0.1194204173598332, + "loss": 0.3663, + "num_input_tokens_seen": 47264384, + "step": 22615 + }, + { + "epoch": 3.6901052288114853, + "grad_norm": 0.0004885298549197614, + "learning_rate": 0.11936275190367007, + "loss": 0.3637, + "num_input_tokens_seen": 47276400, + "step": 22620 + }, + { + "epoch": 3.690920956032303, + "grad_norm": 0.0003492825198918581, + "learning_rate": 0.11930509117215563, + "loss": 0.3133, + "num_input_tokens_seen": 47287040, + "step": 22625 + }, + { + "epoch": 3.6917366832531204, + "grad_norm": 0.0002794082392938435, + "learning_rate": 0.11924743517418179, + "loss": 0.3262, + "num_input_tokens_seen": 47298864, + "step": 22630 + }, + { + "epoch": 3.6925524104739376, + "grad_norm": 0.0005049144383519888, + "learning_rate": 0.11918978391864, + "loss": 0.3685, + "num_input_tokens_seen": 47307920, + "step": 22635 + }, + { + "epoch": 3.6933681376947547, + "grad_norm": 0.00020936237706337124, + "learning_rate": 0.11913213741442065, + "loss": 0.332, + "num_input_tokens_seen": 47319520, + "step": 22640 + }, + { + "epoch": 3.6941838649155723, + "grad_norm": 0.0006876729894429445, + "learning_rate": 0.11907449567041364, + "loss": 0.3216, + "num_input_tokens_seen": 47330448, + "step": 22645 + }, + { + "epoch": 3.69499959213639, + "grad_norm": 0.00036391496541909873, + "learning_rate": 0.11901685869550803, + "loss": 0.3783, + "num_input_tokens_seen": 47339728, + "step": 22650 + }, + { + "epoch": 3.695815319357207, + "grad_norm": 0.0004207253805361688, + "learning_rate": 0.1189592264985922, + "loss": 0.3095, + "num_input_tokens_seen": 47350000, + "step": 22655 + }, + { + "epoch": 3.696631046578024, + "grad_norm": 0.0008644493645988405, + "learning_rate": 0.11890159908855373, + "loss": 0.4064, + "num_input_tokens_seen": 47360432, + "step": 22660 + }, + { + "epoch": 3.6974467737988417, + "grad_norm": 0.00037439176230691373, + "learning_rate": 0.11884397647427941, + "loss": 0.306, + "num_input_tokens_seen": 47371264, + "step": 22665 + }, + { + "epoch": 3.6982625010196593, + "grad_norm": 0.0002674860297702253, + "learning_rate": 0.11878635866465546, + "loss": 0.3175, + "num_input_tokens_seen": 47382144, + "step": 22670 + }, + { + "epoch": 3.6990782282404764, + "grad_norm": 0.0005654924898408353, + "learning_rate": 0.11872874566856734, + "loss": 0.3254, + "num_input_tokens_seen": 47391712, + "step": 22675 + }, + { + "epoch": 3.6998939554612935, + "grad_norm": 0.0008198421564884484, + "learning_rate": 0.11867113749489955, + "loss": 0.3182, + "num_input_tokens_seen": 47402080, + "step": 22680 + }, + { + "epoch": 3.700709682682111, + "grad_norm": 0.0006011492223478854, + "learning_rate": 0.11861353415253607, + "loss": 0.3294, + "num_input_tokens_seen": 47412720, + "step": 22685 + }, + { + "epoch": 3.7015254099029287, + "grad_norm": 0.00030989281367510557, + "learning_rate": 0.11855593565036011, + "loss": 0.3398, + "num_input_tokens_seen": 47422368, + "step": 22690 + }, + { + "epoch": 3.702341137123746, + "grad_norm": 0.0004505499091465026, + "learning_rate": 0.11849834199725394, + "loss": 0.2871, + "num_input_tokens_seen": 47433600, + "step": 22695 + }, + { + "epoch": 3.703156864344563, + "grad_norm": 0.0002990545763168484, + "learning_rate": 0.1184407532020994, + "loss": 0.3283, + "num_input_tokens_seen": 47444624, + "step": 22700 + }, + { + "epoch": 3.7039725915653805, + "grad_norm": 0.0003217622870579362, + "learning_rate": 0.11838316927377723, + "loss": 0.4056, + "num_input_tokens_seen": 47456112, + "step": 22705 + }, + { + "epoch": 3.704788318786198, + "grad_norm": 0.0008387398556806147, + "learning_rate": 0.11832559022116766, + "loss": 0.3336, + "num_input_tokens_seen": 47466080, + "step": 22710 + }, + { + "epoch": 3.705604046007015, + "grad_norm": 0.0003159634943585843, + "learning_rate": 0.11826801605315022, + "loss": 0.2844, + "num_input_tokens_seen": 47476688, + "step": 22715 + }, + { + "epoch": 3.7064197732278323, + "grad_norm": 0.00035760586615651846, + "learning_rate": 0.1182104467786034, + "loss": 0.3092, + "num_input_tokens_seen": 47487744, + "step": 22720 + }, + { + "epoch": 3.70723550044865, + "grad_norm": 0.00035225952160544693, + "learning_rate": 0.1181528824064052, + "loss": 0.3509, + "num_input_tokens_seen": 47497392, + "step": 22725 + }, + { + "epoch": 3.7080512276694675, + "grad_norm": 0.00043457496212795377, + "learning_rate": 0.11809532294543279, + "loss": 0.374, + "num_input_tokens_seen": 47508064, + "step": 22730 + }, + { + "epoch": 3.7088669548902846, + "grad_norm": 0.00028485135408118367, + "learning_rate": 0.11803776840456245, + "loss": 0.2908, + "num_input_tokens_seen": 47518080, + "step": 22735 + }, + { + "epoch": 3.709682682111102, + "grad_norm": 0.000595108256675303, + "learning_rate": 0.11798021879266997, + "loss": 0.3166, + "num_input_tokens_seen": 47529040, + "step": 22740 + }, + { + "epoch": 3.7104984093319193, + "grad_norm": 0.0005873765330761671, + "learning_rate": 0.11792267411863006, + "loss": 0.355, + "num_input_tokens_seen": 47537760, + "step": 22745 + }, + { + "epoch": 3.711314136552737, + "grad_norm": 0.000345634703990072, + "learning_rate": 0.1178651343913169, + "loss": 0.3537, + "num_input_tokens_seen": 47548544, + "step": 22750 + }, + { + "epoch": 3.712129863773554, + "grad_norm": 0.00021975113486405462, + "learning_rate": 0.11780759961960392, + "loss": 0.3062, + "num_input_tokens_seen": 47558368, + "step": 22755 + }, + { + "epoch": 3.7129455909943716, + "grad_norm": 0.00036478121182881296, + "learning_rate": 0.1177500698123636, + "loss": 0.3388, + "num_input_tokens_seen": 47567744, + "step": 22760 + }, + { + "epoch": 3.7137613182151887, + "grad_norm": 0.00036975793773308396, + "learning_rate": 0.11769254497846778, + "loss": 0.3041, + "num_input_tokens_seen": 47577488, + "step": 22765 + }, + { + "epoch": 3.7145770454360063, + "grad_norm": 0.0006303248228505254, + "learning_rate": 0.11763502512678758, + "loss": 0.3425, + "num_input_tokens_seen": 47587760, + "step": 22770 + }, + { + "epoch": 3.7153927726568234, + "grad_norm": 0.0005971788777969778, + "learning_rate": 0.11757751026619315, + "loss": 0.3266, + "num_input_tokens_seen": 47598960, + "step": 22775 + }, + { + "epoch": 3.716208499877641, + "grad_norm": 0.0005005800048820674, + "learning_rate": 0.11752000040555416, + "loss": 0.2681, + "num_input_tokens_seen": 47608400, + "step": 22780 + }, + { + "epoch": 3.717024227098458, + "grad_norm": 0.0006809242768213153, + "learning_rate": 0.11746249555373921, + "loss": 0.3285, + "num_input_tokens_seen": 47620192, + "step": 22785 + }, + { + "epoch": 3.7178399543192757, + "grad_norm": 0.0004270919889677316, + "learning_rate": 0.11740499571961638, + "loss": 0.3124, + "num_input_tokens_seen": 47629360, + "step": 22790 + }, + { + "epoch": 3.718655681540093, + "grad_norm": 0.0003458283899817616, + "learning_rate": 0.11734750091205279, + "loss": 0.3208, + "num_input_tokens_seen": 47639456, + "step": 22795 + }, + { + "epoch": 3.7194714087609104, + "grad_norm": 0.0003795636002905667, + "learning_rate": 0.11729001113991493, + "loss": 0.2685, + "num_input_tokens_seen": 47650016, + "step": 22800 + }, + { + "epoch": 3.7194714087609104, + "eval_loss": 0.31385818123817444, + "eval_runtime": 156.0834, + "eval_samples_per_second": 17.459, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 47650016, + "step": 22800 + }, + { + "epoch": 3.7202871359817276, + "grad_norm": 0.0007936984184198081, + "learning_rate": 0.11723252641206837, + "loss": 0.3005, + "num_input_tokens_seen": 47661056, + "step": 22805 + }, + { + "epoch": 3.721102863202545, + "grad_norm": 0.00038526527350768447, + "learning_rate": 0.11717504673737808, + "loss": 0.3139, + "num_input_tokens_seen": 47671200, + "step": 22810 + }, + { + "epoch": 3.7219185904233623, + "grad_norm": 0.00024153913545887917, + "learning_rate": 0.11711757212470802, + "loss": 0.3115, + "num_input_tokens_seen": 47681408, + "step": 22815 + }, + { + "epoch": 3.72273431764418, + "grad_norm": 0.00028190878219902515, + "learning_rate": 0.11706010258292165, + "loss": 0.3088, + "num_input_tokens_seen": 47691872, + "step": 22820 + }, + { + "epoch": 3.7235500448649974, + "grad_norm": 0.0005979188135825098, + "learning_rate": 0.11700263812088131, + "loss": 0.3041, + "num_input_tokens_seen": 47703248, + "step": 22825 + }, + { + "epoch": 3.7243657720858145, + "grad_norm": 0.0005642702453769743, + "learning_rate": 0.11694517874744892, + "loss": 0.2813, + "num_input_tokens_seen": 47715152, + "step": 22830 + }, + { + "epoch": 3.7251814993066317, + "grad_norm": 0.001250224420800805, + "learning_rate": 0.11688772447148532, + "loss": 0.3902, + "num_input_tokens_seen": 47725680, + "step": 22835 + }, + { + "epoch": 3.7259972265274492, + "grad_norm": 0.0010286528849974275, + "learning_rate": 0.11683027530185074, + "loss": 0.3388, + "num_input_tokens_seen": 47737184, + "step": 22840 + }, + { + "epoch": 3.726812953748267, + "grad_norm": 0.0004167563165538013, + "learning_rate": 0.11677283124740451, + "loss": 0.2869, + "num_input_tokens_seen": 47747424, + "step": 22845 + }, + { + "epoch": 3.727628680969084, + "grad_norm": 0.00020362272334750742, + "learning_rate": 0.11671539231700531, + "loss": 0.3381, + "num_input_tokens_seen": 47757664, + "step": 22850 + }, + { + "epoch": 3.728444408189901, + "grad_norm": 0.0011927024461328983, + "learning_rate": 0.11665795851951084, + "loss": 0.3043, + "num_input_tokens_seen": 47767664, + "step": 22855 + }, + { + "epoch": 3.7292601354107187, + "grad_norm": 0.0007581932004541159, + "learning_rate": 0.11660052986377825, + "loss": 0.345, + "num_input_tokens_seen": 47778208, + "step": 22860 + }, + { + "epoch": 3.7300758626315362, + "grad_norm": 0.0007581675890833139, + "learning_rate": 0.1165431063586636, + "loss": 0.3688, + "num_input_tokens_seen": 47789456, + "step": 22865 + }, + { + "epoch": 3.7308915898523534, + "grad_norm": 0.0009688013815321028, + "learning_rate": 0.11648568801302245, + "loss": 0.348, + "num_input_tokens_seen": 47800480, + "step": 22870 + }, + { + "epoch": 3.7317073170731705, + "grad_norm": 0.0003530373505782336, + "learning_rate": 0.11642827483570937, + "loss": 0.2633, + "num_input_tokens_seen": 47810048, + "step": 22875 + }, + { + "epoch": 3.732523044293988, + "grad_norm": 0.0010148345027118921, + "learning_rate": 0.11637086683557815, + "loss": 0.3117, + "num_input_tokens_seen": 47820320, + "step": 22880 + }, + { + "epoch": 3.7333387715148056, + "grad_norm": 0.00039433062192983925, + "learning_rate": 0.11631346402148188, + "loss": 0.2757, + "num_input_tokens_seen": 47830800, + "step": 22885 + }, + { + "epoch": 3.7341544987356228, + "grad_norm": 0.0007453811122104526, + "learning_rate": 0.11625606640227285, + "loss": 0.3508, + "num_input_tokens_seen": 47840464, + "step": 22890 + }, + { + "epoch": 3.73497022595644, + "grad_norm": 0.0007175299106165767, + "learning_rate": 0.11619867398680238, + "loss": 0.3875, + "num_input_tokens_seen": 47849600, + "step": 22895 + }, + { + "epoch": 3.7357859531772575, + "grad_norm": 0.0003594547160901129, + "learning_rate": 0.11614128678392119, + "loss": 0.3753, + "num_input_tokens_seen": 47859616, + "step": 22900 + }, + { + "epoch": 3.736601680398075, + "grad_norm": 0.00029666401678696275, + "learning_rate": 0.11608390480247906, + "loss": 0.3206, + "num_input_tokens_seen": 47869712, + "step": 22905 + }, + { + "epoch": 3.737417407618892, + "grad_norm": 0.0004737533163279295, + "learning_rate": 0.11602652805132499, + "loss": 0.2811, + "num_input_tokens_seen": 47879472, + "step": 22910 + }, + { + "epoch": 3.7382331348397098, + "grad_norm": 0.0010581094538792968, + "learning_rate": 0.11596915653930731, + "loss": 0.373, + "num_input_tokens_seen": 47889008, + "step": 22915 + }, + { + "epoch": 3.739048862060527, + "grad_norm": 0.00030815141508355737, + "learning_rate": 0.11591179027527328, + "loss": 0.2974, + "num_input_tokens_seen": 47899936, + "step": 22920 + }, + { + "epoch": 3.7398645892813445, + "grad_norm": 0.00036021103733219206, + "learning_rate": 0.11585442926806956, + "loss": 0.344, + "num_input_tokens_seen": 47910496, + "step": 22925 + }, + { + "epoch": 3.7406803165021616, + "grad_norm": 0.0003662409435492009, + "learning_rate": 0.11579707352654202, + "loss": 0.3854, + "num_input_tokens_seen": 47918832, + "step": 22930 + }, + { + "epoch": 3.741496043722979, + "grad_norm": 0.0003198127669747919, + "learning_rate": 0.11573972305953548, + "loss": 0.3303, + "num_input_tokens_seen": 47929264, + "step": 22935 + }, + { + "epoch": 3.7423117709437963, + "grad_norm": 0.0006805963930673897, + "learning_rate": 0.11568237787589426, + "loss": 0.3112, + "num_input_tokens_seen": 47940032, + "step": 22940 + }, + { + "epoch": 3.743127498164614, + "grad_norm": 0.0002786907134577632, + "learning_rate": 0.11562503798446161, + "loss": 0.2908, + "num_input_tokens_seen": 47949600, + "step": 22945 + }, + { + "epoch": 3.743943225385431, + "grad_norm": 0.0004604039713740349, + "learning_rate": 0.11556770339408005, + "loss": 0.2868, + "num_input_tokens_seen": 47959792, + "step": 22950 + }, + { + "epoch": 3.7447589526062486, + "grad_norm": 0.00037229713052511215, + "learning_rate": 0.1155103741135914, + "loss": 0.3349, + "num_input_tokens_seen": 47969680, + "step": 22955 + }, + { + "epoch": 3.7455746798270657, + "grad_norm": 0.0003607496328186244, + "learning_rate": 0.1154530501518364, + "loss": 0.31, + "num_input_tokens_seen": 47980032, + "step": 22960 + }, + { + "epoch": 3.7463904070478833, + "grad_norm": 0.0007102751405909657, + "learning_rate": 0.11539573151765523, + "loss": 0.3674, + "num_input_tokens_seen": 47991488, + "step": 22965 + }, + { + "epoch": 3.7472061342687004, + "grad_norm": 0.0005855255876667798, + "learning_rate": 0.11533841821988719, + "loss": 0.3548, + "num_input_tokens_seen": 48001632, + "step": 22970 + }, + { + "epoch": 3.748021861489518, + "grad_norm": 0.0006686741253361106, + "learning_rate": 0.11528111026737059, + "loss": 0.3222, + "num_input_tokens_seen": 48012864, + "step": 22975 + }, + { + "epoch": 3.748837588710335, + "grad_norm": 0.000610657618381083, + "learning_rate": 0.11522380766894312, + "loss": 0.3358, + "num_input_tokens_seen": 48023408, + "step": 22980 + }, + { + "epoch": 3.7496533159311527, + "grad_norm": 0.0004703848098870367, + "learning_rate": 0.11516651043344152, + "loss": 0.3195, + "num_input_tokens_seen": 48033072, + "step": 22985 + }, + { + "epoch": 3.75046904315197, + "grad_norm": 0.00029009065474383533, + "learning_rate": 0.11510921856970172, + "loss": 0.311, + "num_input_tokens_seen": 48043264, + "step": 22990 + }, + { + "epoch": 3.7512847703727874, + "grad_norm": 0.00045496560051105917, + "learning_rate": 0.11505193208655895, + "loss": 0.3146, + "num_input_tokens_seen": 48053616, + "step": 22995 + }, + { + "epoch": 3.7521004975936045, + "grad_norm": 0.0004781213356181979, + "learning_rate": 0.11499465099284738, + "loss": 0.313, + "num_input_tokens_seen": 48064160, + "step": 23000 + }, + { + "epoch": 3.7521004975936045, + "eval_loss": 0.316121906042099, + "eval_runtime": 156.1242, + "eval_samples_per_second": 17.454, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 48064160, + "step": 23000 + }, + { + "epoch": 3.752916224814422, + "grad_norm": 0.00071271340129897, + "learning_rate": 0.1149373752974006, + "loss": 0.2956, + "num_input_tokens_seen": 48074288, + "step": 23005 + }, + { + "epoch": 3.7537319520352392, + "grad_norm": 0.000412879919167608, + "learning_rate": 0.11488010500905109, + "loss": 0.3128, + "num_input_tokens_seen": 48085456, + "step": 23010 + }, + { + "epoch": 3.754547679256057, + "grad_norm": 0.00031706574372947216, + "learning_rate": 0.11482284013663077, + "loss": 0.3393, + "num_input_tokens_seen": 48096896, + "step": 23015 + }, + { + "epoch": 3.7553634064768744, + "grad_norm": 0.00042995542753487825, + "learning_rate": 0.11476558068897061, + "loss": 0.2901, + "num_input_tokens_seen": 48108224, + "step": 23020 + }, + { + "epoch": 3.7561791336976915, + "grad_norm": 0.0011370141291990876, + "learning_rate": 0.11470832667490061, + "loss": 0.4216, + "num_input_tokens_seen": 48118192, + "step": 23025 + }, + { + "epoch": 3.7569948609185086, + "grad_norm": 0.000269682495854795, + "learning_rate": 0.11465107810325013, + "loss": 0.354, + "num_input_tokens_seen": 48128848, + "step": 23030 + }, + { + "epoch": 3.7578105881393262, + "grad_norm": 0.00044404136133380234, + "learning_rate": 0.11459383498284771, + "loss": 0.2966, + "num_input_tokens_seen": 48139056, + "step": 23035 + }, + { + "epoch": 3.758626315360144, + "grad_norm": 0.0004888621624559164, + "learning_rate": 0.11453659732252082, + "loss": 0.3538, + "num_input_tokens_seen": 48148928, + "step": 23040 + }, + { + "epoch": 3.759442042580961, + "grad_norm": 0.0012659867061302066, + "learning_rate": 0.11447936513109633, + "loss": 0.3395, + "num_input_tokens_seen": 48159760, + "step": 23045 + }, + { + "epoch": 3.760257769801778, + "grad_norm": 0.0009149573743343353, + "learning_rate": 0.11442213841740011, + "loss": 0.2817, + "num_input_tokens_seen": 48169120, + "step": 23050 + }, + { + "epoch": 3.7610734970225956, + "grad_norm": 0.00034920882899314165, + "learning_rate": 0.1143649171902572, + "loss": 0.3621, + "num_input_tokens_seen": 48180832, + "step": 23055 + }, + { + "epoch": 3.761889224243413, + "grad_norm": 0.000441107404185459, + "learning_rate": 0.11430770145849194, + "loss": 0.2614, + "num_input_tokens_seen": 48189840, + "step": 23060 + }, + { + "epoch": 3.7627049514642303, + "grad_norm": 0.0006045686895959079, + "learning_rate": 0.11425049123092756, + "loss": 0.3235, + "num_input_tokens_seen": 48200272, + "step": 23065 + }, + { + "epoch": 3.7635206786850475, + "grad_norm": 0.0002613466349430382, + "learning_rate": 0.11419328651638674, + "loss": 0.3674, + "num_input_tokens_seen": 48211040, + "step": 23070 + }, + { + "epoch": 3.764336405905865, + "grad_norm": 0.00037060712929815054, + "learning_rate": 0.11413608732369115, + "loss": 0.2749, + "num_input_tokens_seen": 48222480, + "step": 23075 + }, + { + "epoch": 3.7651521331266826, + "grad_norm": 0.000872961652930826, + "learning_rate": 0.11407889366166153, + "loss": 0.343, + "num_input_tokens_seen": 48233664, + "step": 23080 + }, + { + "epoch": 3.7659678603474998, + "grad_norm": 0.0002604596083983779, + "learning_rate": 0.11402170553911797, + "loss": 0.2882, + "num_input_tokens_seen": 48243680, + "step": 23085 + }, + { + "epoch": 3.766783587568317, + "grad_norm": 0.0003939753514714539, + "learning_rate": 0.11396452296487955, + "loss": 0.3852, + "num_input_tokens_seen": 48254128, + "step": 23090 + }, + { + "epoch": 3.7675993147891345, + "grad_norm": 0.0004285108589101583, + "learning_rate": 0.11390734594776449, + "loss": 0.3481, + "num_input_tokens_seen": 48263856, + "step": 23095 + }, + { + "epoch": 3.768415042009952, + "grad_norm": 0.0005789677379652858, + "learning_rate": 0.11385017449659031, + "loss": 0.3136, + "num_input_tokens_seen": 48276144, + "step": 23100 + }, + { + "epoch": 3.769230769230769, + "grad_norm": 0.00042495023808442056, + "learning_rate": 0.11379300862017344, + "loss": 0.3593, + "num_input_tokens_seen": 48286768, + "step": 23105 + }, + { + "epoch": 3.7700464964515867, + "grad_norm": 0.00034333067014813423, + "learning_rate": 0.11373584832732966, + "loss": 0.2987, + "num_input_tokens_seen": 48296896, + "step": 23110 + }, + { + "epoch": 3.770862223672404, + "grad_norm": 0.0005105629679746926, + "learning_rate": 0.11367869362687386, + "loss": 0.2868, + "num_input_tokens_seen": 48306576, + "step": 23115 + }, + { + "epoch": 3.7716779508932214, + "grad_norm": 0.0003145936643704772, + "learning_rate": 0.11362154452761988, + "loss": 0.3062, + "num_input_tokens_seen": 48317200, + "step": 23120 + }, + { + "epoch": 3.7724936781140386, + "grad_norm": 0.00030986356432549655, + "learning_rate": 0.11356440103838095, + "loss": 0.3196, + "num_input_tokens_seen": 48328080, + "step": 23125 + }, + { + "epoch": 3.773309405334856, + "grad_norm": 0.0003788000321947038, + "learning_rate": 0.11350726316796922, + "loss": 0.2676, + "num_input_tokens_seen": 48338848, + "step": 23130 + }, + { + "epoch": 3.7741251325556733, + "grad_norm": 0.00035888334969058633, + "learning_rate": 0.11345013092519607, + "loss": 0.2659, + "num_input_tokens_seen": 48347888, + "step": 23135 + }, + { + "epoch": 3.774940859776491, + "grad_norm": 0.0005492912605404854, + "learning_rate": 0.11339300431887213, + "loss": 0.3787, + "num_input_tokens_seen": 48357728, + "step": 23140 + }, + { + "epoch": 3.775756586997308, + "grad_norm": 0.0006858946871943772, + "learning_rate": 0.11333588335780687, + "loss": 0.3849, + "num_input_tokens_seen": 48368928, + "step": 23145 + }, + { + "epoch": 3.7765723142181256, + "grad_norm": 0.0006667522247880697, + "learning_rate": 0.11327876805080916, + "loss": 0.2914, + "num_input_tokens_seen": 48378848, + "step": 23150 + }, + { + "epoch": 3.7773880414389427, + "grad_norm": 0.0004014163860119879, + "learning_rate": 0.11322165840668696, + "loss": 0.3003, + "num_input_tokens_seen": 48389664, + "step": 23155 + }, + { + "epoch": 3.7782037686597603, + "grad_norm": 0.0006043196190148592, + "learning_rate": 0.11316455443424717, + "loss": 0.322, + "num_input_tokens_seen": 48400928, + "step": 23160 + }, + { + "epoch": 3.7790194958805774, + "grad_norm": 0.00040100273326970637, + "learning_rate": 0.11310745614229603, + "loss": 0.2852, + "num_input_tokens_seen": 48412032, + "step": 23165 + }, + { + "epoch": 3.779835223101395, + "grad_norm": 0.0003480259911157191, + "learning_rate": 0.1130503635396387, + "loss": 0.291, + "num_input_tokens_seen": 48422704, + "step": 23170 + }, + { + "epoch": 3.780650950322212, + "grad_norm": 0.0008477790397591889, + "learning_rate": 0.11299327663507966, + "loss": 0.3065, + "num_input_tokens_seen": 48433104, + "step": 23175 + }, + { + "epoch": 3.7814666775430297, + "grad_norm": 0.0006902751047164202, + "learning_rate": 0.11293619543742246, + "loss": 0.364, + "num_input_tokens_seen": 48443632, + "step": 23180 + }, + { + "epoch": 3.782282404763847, + "grad_norm": 0.000836253515444696, + "learning_rate": 0.11287911995546965, + "loss": 0.3654, + "num_input_tokens_seen": 48453504, + "step": 23185 + }, + { + "epoch": 3.7830981319846644, + "grad_norm": 0.00025766415637917817, + "learning_rate": 0.11282205019802308, + "loss": 0.3524, + "num_input_tokens_seen": 48463680, + "step": 23190 + }, + { + "epoch": 3.783913859205482, + "grad_norm": 0.0003022443561349064, + "learning_rate": 0.11276498617388354, + "loss": 0.3228, + "num_input_tokens_seen": 48473712, + "step": 23195 + }, + { + "epoch": 3.784729586426299, + "grad_norm": 0.0003401450230740011, + "learning_rate": 0.11270792789185109, + "loss": 0.3128, + "num_input_tokens_seen": 48484384, + "step": 23200 + }, + { + "epoch": 3.784729586426299, + "eval_loss": 0.31450438499450684, + "eval_runtime": 155.9684, + "eval_samples_per_second": 17.471, + "eval_steps_per_second": 8.739, + "num_input_tokens_seen": 48484384, + "step": 23200 + }, + { + "epoch": 3.785545313647116, + "grad_norm": 0.000587977992836386, + "learning_rate": 0.11265087536072482, + "loss": 0.3196, + "num_input_tokens_seen": 48495936, + "step": 23205 + }, + { + "epoch": 3.786361040867934, + "grad_norm": 0.00038425513776019216, + "learning_rate": 0.11259382858930288, + "loss": 0.3162, + "num_input_tokens_seen": 48506656, + "step": 23210 + }, + { + "epoch": 3.7871767680887514, + "grad_norm": 0.0004102051898371428, + "learning_rate": 0.11253678758638262, + "loss": 0.3529, + "num_input_tokens_seen": 48516624, + "step": 23215 + }, + { + "epoch": 3.7879924953095685, + "grad_norm": 0.00108595029450953, + "learning_rate": 0.11247975236076059, + "loss": 0.3555, + "num_input_tokens_seen": 48526752, + "step": 23220 + }, + { + "epoch": 3.7888082225303856, + "grad_norm": 0.0005795384058728814, + "learning_rate": 0.11242272292123218, + "loss": 0.3221, + "num_input_tokens_seen": 48537664, + "step": 23225 + }, + { + "epoch": 3.789623949751203, + "grad_norm": 0.0005390078295022249, + "learning_rate": 0.11236569927659217, + "loss": 0.3536, + "num_input_tokens_seen": 48548656, + "step": 23230 + }, + { + "epoch": 3.7904396769720208, + "grad_norm": 0.0007965471013449132, + "learning_rate": 0.11230868143563429, + "loss": 0.3048, + "num_input_tokens_seen": 48559888, + "step": 23235 + }, + { + "epoch": 3.791255404192838, + "grad_norm": 0.0005937626701779664, + "learning_rate": 0.11225166940715131, + "loss": 0.3321, + "num_input_tokens_seen": 48570768, + "step": 23240 + }, + { + "epoch": 3.792071131413655, + "grad_norm": 0.0010757347336038947, + "learning_rate": 0.11219466319993537, + "loss": 0.3419, + "num_input_tokens_seen": 48581280, + "step": 23245 + }, + { + "epoch": 3.7928868586344726, + "grad_norm": 0.0005621684831567109, + "learning_rate": 0.11213766282277739, + "loss": 0.3315, + "num_input_tokens_seen": 48592032, + "step": 23250 + }, + { + "epoch": 3.79370258585529, + "grad_norm": 0.0006928109796717763, + "learning_rate": 0.11208066828446761, + "loss": 0.3118, + "num_input_tokens_seen": 48601072, + "step": 23255 + }, + { + "epoch": 3.7945183130761073, + "grad_norm": 0.0006080032326281071, + "learning_rate": 0.11202367959379537, + "loss": 0.3328, + "num_input_tokens_seen": 48611392, + "step": 23260 + }, + { + "epoch": 3.7953340402969244, + "grad_norm": 0.0003554025897756219, + "learning_rate": 0.11196669675954894, + "loss": 0.3249, + "num_input_tokens_seen": 48621984, + "step": 23265 + }, + { + "epoch": 3.796149767517742, + "grad_norm": 0.000527488358784467, + "learning_rate": 0.1119097197905158, + "loss": 0.2505, + "num_input_tokens_seen": 48632864, + "step": 23270 + }, + { + "epoch": 3.7969654947385596, + "grad_norm": 0.00022553290182258934, + "learning_rate": 0.11185274869548259, + "loss": 0.3511, + "num_input_tokens_seen": 48643520, + "step": 23275 + }, + { + "epoch": 3.7977812219593767, + "grad_norm": 0.0005026492872275412, + "learning_rate": 0.11179578348323486, + "loss": 0.277, + "num_input_tokens_seen": 48653104, + "step": 23280 + }, + { + "epoch": 3.7985969491801943, + "grad_norm": 0.00033920654095709324, + "learning_rate": 0.1117388241625575, + "loss": 0.3258, + "num_input_tokens_seen": 48663536, + "step": 23285 + }, + { + "epoch": 3.7994126764010114, + "grad_norm": 0.000259226537309587, + "learning_rate": 0.11168187074223421, + "loss": 0.3217, + "num_input_tokens_seen": 48674272, + "step": 23290 + }, + { + "epoch": 3.800228403621829, + "grad_norm": 0.0006012494559399784, + "learning_rate": 0.11162492323104796, + "loss": 0.3895, + "num_input_tokens_seen": 48684848, + "step": 23295 + }, + { + "epoch": 3.801044130842646, + "grad_norm": 0.000870527874212712, + "learning_rate": 0.11156798163778091, + "loss": 0.2927, + "num_input_tokens_seen": 48694048, + "step": 23300 + }, + { + "epoch": 3.8018598580634637, + "grad_norm": 0.00040228900616057217, + "learning_rate": 0.11151104597121399, + "loss": 0.3137, + "num_input_tokens_seen": 48704640, + "step": 23305 + }, + { + "epoch": 3.802675585284281, + "grad_norm": 0.0005085703451186419, + "learning_rate": 0.11145411624012742, + "loss": 0.3065, + "num_input_tokens_seen": 48716880, + "step": 23310 + }, + { + "epoch": 3.8034913125050984, + "grad_norm": 0.0003494714619591832, + "learning_rate": 0.11139719245330063, + "loss": 0.4347, + "num_input_tokens_seen": 48726336, + "step": 23315 + }, + { + "epoch": 3.8043070397259156, + "grad_norm": 0.0002877835067920387, + "learning_rate": 0.11134027461951179, + "loss": 0.3206, + "num_input_tokens_seen": 48736192, + "step": 23320 + }, + { + "epoch": 3.805122766946733, + "grad_norm": 0.0007902098586782813, + "learning_rate": 0.11128336274753849, + "loss": 0.3483, + "num_input_tokens_seen": 48746864, + "step": 23325 + }, + { + "epoch": 3.8059384941675503, + "grad_norm": 0.0007062124786898494, + "learning_rate": 0.11122645684615715, + "loss": 0.319, + "num_input_tokens_seen": 48756432, + "step": 23330 + }, + { + "epoch": 3.806754221388368, + "grad_norm": 0.00023675437842030078, + "learning_rate": 0.11116955692414345, + "loss": 0.3083, + "num_input_tokens_seen": 48766672, + "step": 23335 + }, + { + "epoch": 3.807569948609185, + "grad_norm": 0.0005287716048769653, + "learning_rate": 0.11111266299027203, + "loss": 0.3157, + "num_input_tokens_seen": 48776800, + "step": 23340 + }, + { + "epoch": 3.8083856758300025, + "grad_norm": 0.0005388166755437851, + "learning_rate": 0.11105577505331668, + "loss": 0.3413, + "num_input_tokens_seen": 48786752, + "step": 23345 + }, + { + "epoch": 3.8092014030508197, + "grad_norm": 0.00034578941995278, + "learning_rate": 0.11099889312205018, + "loss": 0.3457, + "num_input_tokens_seen": 48795408, + "step": 23350 + }, + { + "epoch": 3.8100171302716372, + "grad_norm": 0.0005431053577922285, + "learning_rate": 0.11094201720524455, + "loss": 0.4033, + "num_input_tokens_seen": 48806016, + "step": 23355 + }, + { + "epoch": 3.8108328574924544, + "grad_norm": 0.00021880320855416358, + "learning_rate": 0.11088514731167064, + "loss": 0.3245, + "num_input_tokens_seen": 48817088, + "step": 23360 + }, + { + "epoch": 3.811648584713272, + "grad_norm": 0.0002710201370064169, + "learning_rate": 0.11082828345009862, + "loss": 0.3236, + "num_input_tokens_seen": 48828416, + "step": 23365 + }, + { + "epoch": 3.812464311934089, + "grad_norm": 0.0005445056012831628, + "learning_rate": 0.11077142562929748, + "loss": 0.3575, + "num_input_tokens_seen": 48838320, + "step": 23370 + }, + { + "epoch": 3.8132800391549067, + "grad_norm": 0.00041566279833205044, + "learning_rate": 0.11071457385803554, + "loss": 0.2976, + "num_input_tokens_seen": 48848288, + "step": 23375 + }, + { + "epoch": 3.814095766375724, + "grad_norm": 0.000684981350786984, + "learning_rate": 0.11065772814508001, + "loss": 0.2805, + "num_input_tokens_seen": 48858672, + "step": 23380 + }, + { + "epoch": 3.8149114935965414, + "grad_norm": 0.00023338482424151152, + "learning_rate": 0.11060088849919715, + "loss": 0.315, + "num_input_tokens_seen": 48868720, + "step": 23385 + }, + { + "epoch": 3.815727220817359, + "grad_norm": 0.00025299409753642976, + "learning_rate": 0.11054405492915244, + "loss": 0.3402, + "num_input_tokens_seen": 48878528, + "step": 23390 + }, + { + "epoch": 3.816542948038176, + "grad_norm": 0.0003865258186124265, + "learning_rate": 0.11048722744371031, + "loss": 0.3182, + "num_input_tokens_seen": 48888032, + "step": 23395 + }, + { + "epoch": 3.817358675258993, + "grad_norm": 0.0006580277695320547, + "learning_rate": 0.1104304060516342, + "loss": 0.3507, + "num_input_tokens_seen": 48897744, + "step": 23400 + }, + { + "epoch": 3.817358675258993, + "eval_loss": 0.3209737539291382, + "eval_runtime": 155.8549, + "eval_samples_per_second": 17.484, + "eval_steps_per_second": 8.745, + "num_input_tokens_seen": 48897744, + "step": 23400 + }, + { + "epoch": 3.8181744024798108, + "grad_norm": 0.0002915352815762162, + "learning_rate": 0.11037359076168682, + "loss": 0.2691, + "num_input_tokens_seen": 48907904, + "step": 23405 + }, + { + "epoch": 3.8189901297006283, + "grad_norm": 0.0003229460271541029, + "learning_rate": 0.11031678158262966, + "loss": 0.3123, + "num_input_tokens_seen": 48917888, + "step": 23410 + }, + { + "epoch": 3.8198058569214455, + "grad_norm": 0.0006855239626020193, + "learning_rate": 0.11025997852322349, + "loss": 0.3374, + "num_input_tokens_seen": 48926976, + "step": 23415 + }, + { + "epoch": 3.8206215841422626, + "grad_norm": 0.0003805782471317798, + "learning_rate": 0.11020318159222807, + "loss": 0.3156, + "num_input_tokens_seen": 48935440, + "step": 23420 + }, + { + "epoch": 3.82143731136308, + "grad_norm": 0.0003422145382501185, + "learning_rate": 0.1101463907984021, + "loss": 0.3325, + "num_input_tokens_seen": 48945712, + "step": 23425 + }, + { + "epoch": 3.8222530385838978, + "grad_norm": 0.0003806010936386883, + "learning_rate": 0.11008960615050352, + "loss": 0.2793, + "num_input_tokens_seen": 48956768, + "step": 23430 + }, + { + "epoch": 3.823068765804715, + "grad_norm": 0.00034248476731590927, + "learning_rate": 0.11003282765728925, + "loss": 0.3466, + "num_input_tokens_seen": 48968144, + "step": 23435 + }, + { + "epoch": 3.823884493025532, + "grad_norm": 0.0013128285063430667, + "learning_rate": 0.10997605532751518, + "loss": 0.3168, + "num_input_tokens_seen": 48980144, + "step": 23440 + }, + { + "epoch": 3.8247002202463496, + "grad_norm": 0.0003037426504306495, + "learning_rate": 0.1099192891699364, + "loss": 0.341, + "num_input_tokens_seen": 48991328, + "step": 23445 + }, + { + "epoch": 3.825515947467167, + "grad_norm": 0.0005769141716882586, + "learning_rate": 0.10986252919330687, + "loss": 0.4022, + "num_input_tokens_seen": 49001984, + "step": 23450 + }, + { + "epoch": 3.8263316746879843, + "grad_norm": 0.0007138108485378325, + "learning_rate": 0.10980577540637973, + "loss": 0.3625, + "num_input_tokens_seen": 49011888, + "step": 23455 + }, + { + "epoch": 3.8271474019088014, + "grad_norm": 0.00047376647125929594, + "learning_rate": 0.10974902781790719, + "loss": 0.2878, + "num_input_tokens_seen": 49021584, + "step": 23460 + }, + { + "epoch": 3.827963129129619, + "grad_norm": 0.0005657591973431408, + "learning_rate": 0.10969228643664032, + "loss": 0.2884, + "num_input_tokens_seen": 49033184, + "step": 23465 + }, + { + "epoch": 3.8287788563504366, + "grad_norm": 0.00028129853308200836, + "learning_rate": 0.10963555127132942, + "loss": 0.2912, + "num_input_tokens_seen": 49044064, + "step": 23470 + }, + { + "epoch": 3.8295945835712537, + "grad_norm": 0.0002328073896933347, + "learning_rate": 0.10957882233072382, + "loss": 0.3267, + "num_input_tokens_seen": 49054320, + "step": 23475 + }, + { + "epoch": 3.8304103107920713, + "grad_norm": 0.0004983869148418307, + "learning_rate": 0.10952209962357176, + "loss": 0.412, + "num_input_tokens_seen": 49063520, + "step": 23480 + }, + { + "epoch": 3.8312260380128884, + "grad_norm": 0.00022823097242508084, + "learning_rate": 0.10946538315862062, + "loss": 0.2931, + "num_input_tokens_seen": 49074080, + "step": 23485 + }, + { + "epoch": 3.832041765233706, + "grad_norm": 0.00045950498315505683, + "learning_rate": 0.10940867294461679, + "loss": 0.2959, + "num_input_tokens_seen": 49084432, + "step": 23490 + }, + { + "epoch": 3.832857492454523, + "grad_norm": 0.0004550247685983777, + "learning_rate": 0.10935196899030565, + "loss": 0.2776, + "num_input_tokens_seen": 49094560, + "step": 23495 + }, + { + "epoch": 3.8336732196753407, + "grad_norm": 0.0005668136291205883, + "learning_rate": 0.10929527130443177, + "loss": 0.3624, + "num_input_tokens_seen": 49105216, + "step": 23500 + }, + { + "epoch": 3.834488946896158, + "grad_norm": 0.0006497748545370996, + "learning_rate": 0.1092385798957385, + "loss": 0.3614, + "num_input_tokens_seen": 49115440, + "step": 23505 + }, + { + "epoch": 3.8353046741169754, + "grad_norm": 0.00028505668160505593, + "learning_rate": 0.10918189477296848, + "loss": 0.3253, + "num_input_tokens_seen": 49125200, + "step": 23510 + }, + { + "epoch": 3.8361204013377925, + "grad_norm": 0.0002744256053119898, + "learning_rate": 0.1091252159448633, + "loss": 0.3394, + "num_input_tokens_seen": 49135328, + "step": 23515 + }, + { + "epoch": 3.83693612855861, + "grad_norm": 0.0002573148813098669, + "learning_rate": 0.10906854342016345, + "loss": 0.2676, + "num_input_tokens_seen": 49146016, + "step": 23520 + }, + { + "epoch": 3.8377518557794272, + "grad_norm": 0.00041262098238803446, + "learning_rate": 0.10901187720760858, + "loss": 0.31, + "num_input_tokens_seen": 49154416, + "step": 23525 + }, + { + "epoch": 3.838567583000245, + "grad_norm": 0.0005562860169447958, + "learning_rate": 0.10895521731593734, + "loss": 0.3107, + "num_input_tokens_seen": 49163728, + "step": 23530 + }, + { + "epoch": 3.839383310221062, + "grad_norm": 0.000637976685538888, + "learning_rate": 0.10889856375388733, + "loss": 0.3151, + "num_input_tokens_seen": 49174960, + "step": 23535 + }, + { + "epoch": 3.8401990374418795, + "grad_norm": 0.00038426287937909365, + "learning_rate": 0.1088419165301954, + "loss": 0.2439, + "num_input_tokens_seen": 49185648, + "step": 23540 + }, + { + "epoch": 3.8410147646626966, + "grad_norm": 0.00033235264709219337, + "learning_rate": 0.1087852756535971, + "loss": 0.328, + "num_input_tokens_seen": 49196896, + "step": 23545 + }, + { + "epoch": 3.841830491883514, + "grad_norm": 0.00046511099208146334, + "learning_rate": 0.10872864113282725, + "loss": 0.2864, + "num_input_tokens_seen": 49205728, + "step": 23550 + }, + { + "epoch": 3.8426462191043314, + "grad_norm": 0.00033490968053229153, + "learning_rate": 0.10867201297661958, + "loss": 0.352, + "num_input_tokens_seen": 49214816, + "step": 23555 + }, + { + "epoch": 3.843461946325149, + "grad_norm": 0.0003181143547408283, + "learning_rate": 0.10861539119370689, + "loss": 0.3564, + "num_input_tokens_seen": 49224288, + "step": 23560 + }, + { + "epoch": 3.8442776735459665, + "grad_norm": 0.0008387233247049153, + "learning_rate": 0.10855877579282096, + "loss": 0.2951, + "num_input_tokens_seen": 49234000, + "step": 23565 + }, + { + "epoch": 3.8450934007667836, + "grad_norm": 0.00028594379546120763, + "learning_rate": 0.10850216678269252, + "loss": 0.3509, + "num_input_tokens_seen": 49244144, + "step": 23570 + }, + { + "epoch": 3.8459091279876008, + "grad_norm": 0.0003709422016981989, + "learning_rate": 0.10844556417205146, + "loss": 0.3189, + "num_input_tokens_seen": 49252848, + "step": 23575 + }, + { + "epoch": 3.8467248552084183, + "grad_norm": 0.0006065424531698227, + "learning_rate": 0.10838896796962669, + "loss": 0.2828, + "num_input_tokens_seen": 49263920, + "step": 23580 + }, + { + "epoch": 3.847540582429236, + "grad_norm": 0.0007432725979015231, + "learning_rate": 0.1083323781841459, + "loss": 0.3593, + "num_input_tokens_seen": 49274608, + "step": 23585 + }, + { + "epoch": 3.848356309650053, + "grad_norm": 0.0002883337438106537, + "learning_rate": 0.10827579482433607, + "loss": 0.3132, + "num_input_tokens_seen": 49284992, + "step": 23590 + }, + { + "epoch": 3.84917203687087, + "grad_norm": 0.00025323222507722676, + "learning_rate": 0.10821921789892304, + "loss": 0.2868, + "num_input_tokens_seen": 49297072, + "step": 23595 + }, + { + "epoch": 3.8499877640916877, + "grad_norm": 0.00028249435126781464, + "learning_rate": 0.10816264741663158, + "loss": 0.3455, + "num_input_tokens_seen": 49308304, + "step": 23600 + }, + { + "epoch": 3.8499877640916877, + "eval_loss": 0.3147221505641937, + "eval_runtime": 155.8792, + "eval_samples_per_second": 17.481, + "eval_steps_per_second": 8.744, + "num_input_tokens_seen": 49308304, + "step": 23600 + }, + { + "epoch": 3.8508034913125053, + "grad_norm": 0.00041210019844584167, + "learning_rate": 0.10810608338618573, + "loss": 0.3498, + "num_input_tokens_seen": 49318272, + "step": 23605 + }, + { + "epoch": 3.8516192185333225, + "grad_norm": 0.00031938686151988804, + "learning_rate": 0.10804952581630821, + "loss": 0.3156, + "num_input_tokens_seen": 49329904, + "step": 23610 + }, + { + "epoch": 3.8524349457541396, + "grad_norm": 0.00036206477670930326, + "learning_rate": 0.10799297471572102, + "loss": 0.3649, + "num_input_tokens_seen": 49339984, + "step": 23615 + }, + { + "epoch": 3.853250672974957, + "grad_norm": 0.0007501378422603011, + "learning_rate": 0.10793643009314507, + "loss": 0.3108, + "num_input_tokens_seen": 49351680, + "step": 23620 + }, + { + "epoch": 3.8540664001957747, + "grad_norm": 0.00023385403619613498, + "learning_rate": 0.10787989195730015, + "loss": 0.305, + "num_input_tokens_seen": 49360576, + "step": 23625 + }, + { + "epoch": 3.854882127416592, + "grad_norm": 0.0006384584121406078, + "learning_rate": 0.10782336031690525, + "loss": 0.388, + "num_input_tokens_seen": 49371840, + "step": 23630 + }, + { + "epoch": 3.855697854637409, + "grad_norm": 0.000285941205220297, + "learning_rate": 0.10776683518067821, + "loss": 0.2991, + "num_input_tokens_seen": 49381968, + "step": 23635 + }, + { + "epoch": 3.8565135818582266, + "grad_norm": 0.0004965749103575945, + "learning_rate": 0.10771031655733587, + "loss": 0.2829, + "num_input_tokens_seen": 49392896, + "step": 23640 + }, + { + "epoch": 3.857329309079044, + "grad_norm": 0.0003526492801029235, + "learning_rate": 0.10765380445559422, + "loss": 0.3792, + "num_input_tokens_seen": 49402352, + "step": 23645 + }, + { + "epoch": 3.8581450362998613, + "grad_norm": 0.0003384179435670376, + "learning_rate": 0.10759729888416801, + "loss": 0.348, + "num_input_tokens_seen": 49413504, + "step": 23650 + }, + { + "epoch": 3.858960763520679, + "grad_norm": 0.00032170090707950294, + "learning_rate": 0.10754079985177119, + "loss": 0.3086, + "num_input_tokens_seen": 49423952, + "step": 23655 + }, + { + "epoch": 3.859776490741496, + "grad_norm": 0.00022944882221054286, + "learning_rate": 0.10748430736711667, + "loss": 0.2747, + "num_input_tokens_seen": 49433808, + "step": 23660 + }, + { + "epoch": 3.8605922179623136, + "grad_norm": 0.000344053260050714, + "learning_rate": 0.10742782143891623, + "loss": 0.3129, + "num_input_tokens_seen": 49445152, + "step": 23665 + }, + { + "epoch": 3.8614079451831307, + "grad_norm": 0.0003525945940054953, + "learning_rate": 0.10737134207588069, + "loss": 0.3357, + "num_input_tokens_seen": 49456064, + "step": 23670 + }, + { + "epoch": 3.8622236724039483, + "grad_norm": 0.00027732516173273325, + "learning_rate": 0.10731486928671992, + "loss": 0.3603, + "num_input_tokens_seen": 49467360, + "step": 23675 + }, + { + "epoch": 3.8630393996247654, + "grad_norm": 0.0005007845466025174, + "learning_rate": 0.10725840308014269, + "loss": 0.295, + "num_input_tokens_seen": 49478224, + "step": 23680 + }, + { + "epoch": 3.863855126845583, + "grad_norm": 0.0005974859814159572, + "learning_rate": 0.10720194346485688, + "loss": 0.3374, + "num_input_tokens_seen": 49488960, + "step": 23685 + }, + { + "epoch": 3.8646708540664, + "grad_norm": 0.0005659757880493999, + "learning_rate": 0.10714549044956918, + "loss": 0.3473, + "num_input_tokens_seen": 49499552, + "step": 23690 + }, + { + "epoch": 3.8654865812872177, + "grad_norm": 0.0010282864095643163, + "learning_rate": 0.10708904404298542, + "loss": 0.3715, + "num_input_tokens_seen": 49509696, + "step": 23695 + }, + { + "epoch": 3.866302308508035, + "grad_norm": 0.00042827296420000494, + "learning_rate": 0.1070326042538103, + "loss": 0.3024, + "num_input_tokens_seen": 49520336, + "step": 23700 + }, + { + "epoch": 3.8671180357288524, + "grad_norm": 0.0002890771720558405, + "learning_rate": 0.10697617109074758, + "loss": 0.2901, + "num_input_tokens_seen": 49530320, + "step": 23705 + }, + { + "epoch": 3.8679337629496695, + "grad_norm": 0.00027470028726384044, + "learning_rate": 0.10691974456249999, + "loss": 0.3382, + "num_input_tokens_seen": 49540944, + "step": 23710 + }, + { + "epoch": 3.868749490170487, + "grad_norm": 0.0002879149396903813, + "learning_rate": 0.10686332467776909, + "loss": 0.3474, + "num_input_tokens_seen": 49551168, + "step": 23715 + }, + { + "epoch": 3.869565217391304, + "grad_norm": 0.0005710828118026257, + "learning_rate": 0.10680691144525563, + "loss": 0.3549, + "num_input_tokens_seen": 49561136, + "step": 23720 + }, + { + "epoch": 3.870380944612122, + "grad_norm": 0.0002570179640315473, + "learning_rate": 0.10675050487365928, + "loss": 0.3551, + "num_input_tokens_seen": 49570864, + "step": 23725 + }, + { + "epoch": 3.871196671832939, + "grad_norm": 0.0006856517284177244, + "learning_rate": 0.10669410497167851, + "loss": 0.3791, + "num_input_tokens_seen": 49581824, + "step": 23730 + }, + { + "epoch": 3.8720123990537565, + "grad_norm": 0.0003549254615791142, + "learning_rate": 0.10663771174801102, + "loss": 0.2646, + "num_input_tokens_seen": 49591712, + "step": 23735 + }, + { + "epoch": 3.8728281262745736, + "grad_norm": 0.00039653846761211753, + "learning_rate": 0.10658132521135329, + "loss": 0.3166, + "num_input_tokens_seen": 49602192, + "step": 23740 + }, + { + "epoch": 3.873643853495391, + "grad_norm": 0.00037786876782774925, + "learning_rate": 0.10652494537040084, + "loss": 0.3455, + "num_input_tokens_seen": 49612704, + "step": 23745 + }, + { + "epoch": 3.8744595807162083, + "grad_norm": 0.0006549863610416651, + "learning_rate": 0.1064685722338482, + "loss": 0.3321, + "num_input_tokens_seen": 49623184, + "step": 23750 + }, + { + "epoch": 3.875275307937026, + "grad_norm": 0.0005652153049595654, + "learning_rate": 0.10641220581038871, + "loss": 0.3005, + "num_input_tokens_seen": 49634720, + "step": 23755 + }, + { + "epoch": 3.8760910351578435, + "grad_norm": 0.0005203622276894748, + "learning_rate": 0.10635584610871483, + "loss": 0.2737, + "num_input_tokens_seen": 49644640, + "step": 23760 + }, + { + "epoch": 3.8769067623786606, + "grad_norm": 0.00039493723306804895, + "learning_rate": 0.10629949313751803, + "loss": 0.3525, + "num_input_tokens_seen": 49654160, + "step": 23765 + }, + { + "epoch": 3.8777224895994777, + "grad_norm": 0.00028531160205602646, + "learning_rate": 0.10624314690548849, + "loss": 0.2764, + "num_input_tokens_seen": 49664816, + "step": 23770 + }, + { + "epoch": 3.8785382168202953, + "grad_norm": 0.00047665846068412066, + "learning_rate": 0.1061868074213156, + "loss": 0.3133, + "num_input_tokens_seen": 49675968, + "step": 23775 + }, + { + "epoch": 3.879353944041113, + "grad_norm": 0.0005551131325773895, + "learning_rate": 0.10613047469368765, + "loss": 0.3358, + "num_input_tokens_seen": 49686992, + "step": 23780 + }, + { + "epoch": 3.88016967126193, + "grad_norm": 0.0003485802444629371, + "learning_rate": 0.10607414873129171, + "loss": 0.2621, + "num_input_tokens_seen": 49697392, + "step": 23785 + }, + { + "epoch": 3.880985398482747, + "grad_norm": 0.00033161105238832533, + "learning_rate": 0.10601782954281413, + "loss": 0.3118, + "num_input_tokens_seen": 49707184, + "step": 23790 + }, + { + "epoch": 3.8818011257035647, + "grad_norm": 0.00029621535213664174, + "learning_rate": 0.1059615171369399, + "loss": 0.3238, + "num_input_tokens_seen": 49717424, + "step": 23795 + }, + { + "epoch": 3.8826168529243823, + "grad_norm": 0.00045826766290701926, + "learning_rate": 0.10590521152235312, + "loss": 0.3713, + "num_input_tokens_seen": 49728368, + "step": 23800 + }, + { + "epoch": 3.8826168529243823, + "eval_loss": 0.3208862543106079, + "eval_runtime": 156.1527, + "eval_samples_per_second": 17.451, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 49728368, + "step": 23800 + }, + { + "epoch": 3.8834325801451994, + "grad_norm": 0.0008295593434013426, + "learning_rate": 0.1058489127077369, + "loss": 0.389, + "num_input_tokens_seen": 49738752, + "step": 23805 + }, + { + "epoch": 3.8842483073660166, + "grad_norm": 0.000357204582542181, + "learning_rate": 0.1057926207017732, + "loss": 0.2719, + "num_input_tokens_seen": 49746480, + "step": 23810 + }, + { + "epoch": 3.885064034586834, + "grad_norm": 0.0002977530821226537, + "learning_rate": 0.10573633551314285, + "loss": 0.4212, + "num_input_tokens_seen": 49758128, + "step": 23815 + }, + { + "epoch": 3.8858797618076517, + "grad_norm": 0.0007333290413953364, + "learning_rate": 0.1056800571505259, + "loss": 0.283, + "num_input_tokens_seen": 49769072, + "step": 23820 + }, + { + "epoch": 3.886695489028469, + "grad_norm": 0.00041989979217760265, + "learning_rate": 0.10562378562260105, + "loss": 0.3012, + "num_input_tokens_seen": 49779104, + "step": 23825 + }, + { + "epoch": 3.887511216249286, + "grad_norm": 0.0006341583793982863, + "learning_rate": 0.10556752093804615, + "loss": 0.379, + "num_input_tokens_seen": 49790000, + "step": 23830 + }, + { + "epoch": 3.8883269434701035, + "grad_norm": 0.0005099824629724026, + "learning_rate": 0.10551126310553786, + "loss": 0.365, + "num_input_tokens_seen": 49800256, + "step": 23835 + }, + { + "epoch": 3.889142670690921, + "grad_norm": 0.00031178220524452627, + "learning_rate": 0.10545501213375187, + "loss": 0.282, + "num_input_tokens_seen": 49810848, + "step": 23840 + }, + { + "epoch": 3.8899583979117383, + "grad_norm": 0.00043834224925376475, + "learning_rate": 0.10539876803136287, + "loss": 0.3176, + "num_input_tokens_seen": 49821600, + "step": 23845 + }, + { + "epoch": 3.890774125132556, + "grad_norm": 0.000382529862690717, + "learning_rate": 0.10534253080704428, + "loss": 0.3029, + "num_input_tokens_seen": 49829904, + "step": 23850 + }, + { + "epoch": 3.891589852353373, + "grad_norm": 0.0009994582505896688, + "learning_rate": 0.10528630046946862, + "loss": 0.3318, + "num_input_tokens_seen": 49839024, + "step": 23855 + }, + { + "epoch": 3.8924055795741905, + "grad_norm": 0.0003702800313476473, + "learning_rate": 0.1052300770273074, + "loss": 0.3332, + "num_input_tokens_seen": 49850304, + "step": 23860 + }, + { + "epoch": 3.8932213067950077, + "grad_norm": 0.000337316159857437, + "learning_rate": 0.10517386048923086, + "loss": 0.3225, + "num_input_tokens_seen": 49860608, + "step": 23865 + }, + { + "epoch": 3.8940370340158252, + "grad_norm": 0.0007560758967883885, + "learning_rate": 0.10511765086390841, + "loss": 0.3304, + "num_input_tokens_seen": 49870848, + "step": 23870 + }, + { + "epoch": 3.8948527612366424, + "grad_norm": 0.00033517650444991887, + "learning_rate": 0.10506144816000816, + "loss": 0.2967, + "num_input_tokens_seen": 49882112, + "step": 23875 + }, + { + "epoch": 3.89566848845746, + "grad_norm": 0.0008174122194759548, + "learning_rate": 0.10500525238619736, + "loss": 0.3411, + "num_input_tokens_seen": 49890944, + "step": 23880 + }, + { + "epoch": 3.896484215678277, + "grad_norm": 0.00047607399756088853, + "learning_rate": 0.10494906355114209, + "loss": 0.3413, + "num_input_tokens_seen": 49901520, + "step": 23885 + }, + { + "epoch": 3.8972999428990946, + "grad_norm": 0.00040891769458539784, + "learning_rate": 0.10489288166350737, + "loss": 0.2805, + "num_input_tokens_seen": 49911712, + "step": 23890 + }, + { + "epoch": 3.898115670119912, + "grad_norm": 0.0003773495845962316, + "learning_rate": 0.10483670673195711, + "loss": 0.3019, + "num_input_tokens_seen": 49923344, + "step": 23895 + }, + { + "epoch": 3.8989313973407294, + "grad_norm": 0.0003005442558787763, + "learning_rate": 0.10478053876515431, + "loss": 0.3322, + "num_input_tokens_seen": 49933520, + "step": 23900 + }, + { + "epoch": 3.8997471245615465, + "grad_norm": 0.00044470708235166967, + "learning_rate": 0.10472437777176061, + "loss": 0.3465, + "num_input_tokens_seen": 49943152, + "step": 23905 + }, + { + "epoch": 3.900562851782364, + "grad_norm": 0.0005890839383937418, + "learning_rate": 0.1046682237604369, + "loss": 0.3429, + "num_input_tokens_seen": 49951840, + "step": 23910 + }, + { + "epoch": 3.901378579003181, + "grad_norm": 0.0009404154261574149, + "learning_rate": 0.1046120767398427, + "loss": 0.3978, + "num_input_tokens_seen": 49962128, + "step": 23915 + }, + { + "epoch": 3.9021943062239988, + "grad_norm": 0.0007303063757717609, + "learning_rate": 0.10455593671863667, + "loss": 0.3191, + "num_input_tokens_seen": 49972528, + "step": 23920 + }, + { + "epoch": 3.903010033444816, + "grad_norm": 0.0005524644511751831, + "learning_rate": 0.1044998037054763, + "loss": 0.3181, + "num_input_tokens_seen": 49982096, + "step": 23925 + }, + { + "epoch": 3.9038257606656335, + "grad_norm": 0.000587640271987766, + "learning_rate": 0.10444367770901794, + "loss": 0.3201, + "num_input_tokens_seen": 49992864, + "step": 23930 + }, + { + "epoch": 3.904641487886451, + "grad_norm": 0.0004186738806311041, + "learning_rate": 0.10438755873791698, + "loss": 0.3011, + "num_input_tokens_seen": 50002864, + "step": 23935 + }, + { + "epoch": 3.905457215107268, + "grad_norm": 0.0008721032063476741, + "learning_rate": 0.10433144680082775, + "loss": 0.3434, + "num_input_tokens_seen": 50014064, + "step": 23940 + }, + { + "epoch": 3.9062729423280853, + "grad_norm": 0.0003846075269393623, + "learning_rate": 0.10427534190640322, + "loss": 0.3543, + "num_input_tokens_seen": 50025088, + "step": 23945 + }, + { + "epoch": 3.907088669548903, + "grad_norm": 0.00034458411391824484, + "learning_rate": 0.10421924406329568, + "loss": 0.3066, + "num_input_tokens_seen": 50034912, + "step": 23950 + }, + { + "epoch": 3.9079043967697205, + "grad_norm": 0.0003841824072878808, + "learning_rate": 0.10416315328015598, + "loss": 0.2872, + "num_input_tokens_seen": 50044912, + "step": 23955 + }, + { + "epoch": 3.9087201239905376, + "grad_norm": 0.00027794690686278045, + "learning_rate": 0.10410706956563402, + "loss": 0.3444, + "num_input_tokens_seen": 50055728, + "step": 23960 + }, + { + "epoch": 3.9095358512113547, + "grad_norm": 0.0004975217743776739, + "learning_rate": 0.10405099292837874, + "loss": 0.3251, + "num_input_tokens_seen": 50065168, + "step": 23965 + }, + { + "epoch": 3.9103515784321723, + "grad_norm": 0.00034601782681420445, + "learning_rate": 0.10399492337703771, + "loss": 0.3743, + "num_input_tokens_seen": 50076352, + "step": 23970 + }, + { + "epoch": 3.91116730565299, + "grad_norm": 0.00036112501402385533, + "learning_rate": 0.10393886092025764, + "loss": 0.2974, + "num_input_tokens_seen": 50088432, + "step": 23975 + }, + { + "epoch": 3.911983032873807, + "grad_norm": 0.000500144436955452, + "learning_rate": 0.10388280556668412, + "loss": 0.3502, + "num_input_tokens_seen": 50100112, + "step": 23980 + }, + { + "epoch": 3.912798760094624, + "grad_norm": 0.0004970676964148879, + "learning_rate": 0.10382675732496145, + "loss": 0.3234, + "num_input_tokens_seen": 50110320, + "step": 23985 + }, + { + "epoch": 3.9136144873154417, + "grad_norm": 0.0010092465672641993, + "learning_rate": 0.10377071620373311, + "loss": 0.2899, + "num_input_tokens_seen": 50120464, + "step": 23990 + }, + { + "epoch": 3.9144302145362593, + "grad_norm": 0.0003194937889929861, + "learning_rate": 0.10371468221164128, + "loss": 0.3407, + "num_input_tokens_seen": 50129680, + "step": 23995 + }, + { + "epoch": 3.9152459417570764, + "grad_norm": 0.001287082675844431, + "learning_rate": 0.10365865535732706, + "loss": 0.3627, + "num_input_tokens_seen": 50140272, + "step": 24000 + }, + { + "epoch": 3.9152459417570764, + "eval_loss": 0.31985530257225037, + "eval_runtime": 156.1219, + "eval_samples_per_second": 17.454, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 50140272, + "step": 24000 + }, + { + "epoch": 3.9160616689778935, + "grad_norm": 0.00025738487602211535, + "learning_rate": 0.10360263564943062, + "loss": 0.2815, + "num_input_tokens_seen": 50149872, + "step": 24005 + }, + { + "epoch": 3.916877396198711, + "grad_norm": 0.0007808018708601594, + "learning_rate": 0.10354662309659075, + "loss": 0.3194, + "num_input_tokens_seen": 50161136, + "step": 24010 + }, + { + "epoch": 3.9176931234195287, + "grad_norm": 0.000429430918302387, + "learning_rate": 0.10349061770744537, + "loss": 0.3253, + "num_input_tokens_seen": 50169104, + "step": 24015 + }, + { + "epoch": 3.918508850640346, + "grad_norm": 0.00028689231839962304, + "learning_rate": 0.10343461949063128, + "loss": 0.3348, + "num_input_tokens_seen": 50179104, + "step": 24020 + }, + { + "epoch": 3.919324577861163, + "grad_norm": 0.00048514973605051637, + "learning_rate": 0.103378628454784, + "loss": 0.3207, + "num_input_tokens_seen": 50189920, + "step": 24025 + }, + { + "epoch": 3.9201403050819805, + "grad_norm": 0.0005361350486055017, + "learning_rate": 0.10332264460853811, + "loss": 0.3162, + "num_input_tokens_seen": 50200608, + "step": 24030 + }, + { + "epoch": 3.920956032302798, + "grad_norm": 0.0004661700804717839, + "learning_rate": 0.10326666796052701, + "loss": 0.3576, + "num_input_tokens_seen": 50210560, + "step": 24035 + }, + { + "epoch": 3.9217717595236152, + "grad_norm": 0.0006381311104632914, + "learning_rate": 0.10321069851938296, + "loss": 0.2983, + "num_input_tokens_seen": 50221568, + "step": 24040 + }, + { + "epoch": 3.922587486744433, + "grad_norm": 0.0005487120943143964, + "learning_rate": 0.10315473629373724, + "loss": 0.3586, + "num_input_tokens_seen": 50230848, + "step": 24045 + }, + { + "epoch": 3.92340321396525, + "grad_norm": 0.00033456613891758025, + "learning_rate": 0.10309878129221982, + "loss": 0.3814, + "num_input_tokens_seen": 50241840, + "step": 24050 + }, + { + "epoch": 3.9242189411860675, + "grad_norm": 0.00040442138561047614, + "learning_rate": 0.10304283352345973, + "loss": 0.3363, + "num_input_tokens_seen": 50252432, + "step": 24055 + }, + { + "epoch": 3.9250346684068846, + "grad_norm": 0.00031711708288639784, + "learning_rate": 0.10298689299608486, + "loss": 0.2942, + "num_input_tokens_seen": 50262704, + "step": 24060 + }, + { + "epoch": 3.925850395627702, + "grad_norm": 0.00024319012300111353, + "learning_rate": 0.10293095971872188, + "loss": 0.2629, + "num_input_tokens_seen": 50273184, + "step": 24065 + }, + { + "epoch": 3.9266661228485193, + "grad_norm": 0.0003221405786462128, + "learning_rate": 0.10287503369999645, + "loss": 0.3141, + "num_input_tokens_seen": 50283296, + "step": 24070 + }, + { + "epoch": 3.927481850069337, + "grad_norm": 0.000324793451000005, + "learning_rate": 0.10281911494853295, + "loss": 0.3299, + "num_input_tokens_seen": 50293536, + "step": 24075 + }, + { + "epoch": 3.928297577290154, + "grad_norm": 0.000359523284714669, + "learning_rate": 0.10276320347295485, + "loss": 0.2735, + "num_input_tokens_seen": 50305104, + "step": 24080 + }, + { + "epoch": 3.9291133045109716, + "grad_norm": 0.0004903314984403551, + "learning_rate": 0.10270729928188446, + "loss": 0.2897, + "num_input_tokens_seen": 50315648, + "step": 24085 + }, + { + "epoch": 3.9299290317317888, + "grad_norm": 0.0007489905692636967, + "learning_rate": 0.10265140238394276, + "loss": 0.3034, + "num_input_tokens_seen": 50326352, + "step": 24090 + }, + { + "epoch": 3.9307447589526063, + "grad_norm": 0.0003435113758314401, + "learning_rate": 0.10259551278774988, + "loss": 0.346, + "num_input_tokens_seen": 50337520, + "step": 24095 + }, + { + "epoch": 3.9315604861734235, + "grad_norm": 0.0009083332843147218, + "learning_rate": 0.10253963050192462, + "loss": 0.3852, + "num_input_tokens_seen": 50348880, + "step": 24100 + }, + { + "epoch": 3.932376213394241, + "grad_norm": 0.0006664209649898112, + "learning_rate": 0.10248375553508478, + "loss": 0.2885, + "num_input_tokens_seen": 50359840, + "step": 24105 + }, + { + "epoch": 3.933191940615058, + "grad_norm": 0.0006749642197974026, + "learning_rate": 0.102427887895847, + "loss": 0.2733, + "num_input_tokens_seen": 50371248, + "step": 24110 + }, + { + "epoch": 3.9340076678358757, + "grad_norm": 0.000526508258190006, + "learning_rate": 0.10237202759282668, + "loss": 0.3984, + "num_input_tokens_seen": 50382672, + "step": 24115 + }, + { + "epoch": 3.934823395056693, + "grad_norm": 0.00026573354261927307, + "learning_rate": 0.10231617463463821, + "loss": 0.2964, + "num_input_tokens_seen": 50393584, + "step": 24120 + }, + { + "epoch": 3.9356391222775104, + "grad_norm": 0.0005595492548309267, + "learning_rate": 0.10226032902989492, + "loss": 0.3134, + "num_input_tokens_seen": 50404016, + "step": 24125 + }, + { + "epoch": 3.936454849498328, + "grad_norm": 0.00023596534447278827, + "learning_rate": 0.10220449078720877, + "loss": 0.3088, + "num_input_tokens_seen": 50415488, + "step": 24130 + }, + { + "epoch": 3.937270576719145, + "grad_norm": 0.0007047923281788826, + "learning_rate": 0.1021486599151908, + "loss": 0.3284, + "num_input_tokens_seen": 50426464, + "step": 24135 + }, + { + "epoch": 3.9380863039399623, + "grad_norm": 0.0004291152872610837, + "learning_rate": 0.10209283642245084, + "loss": 0.329, + "num_input_tokens_seen": 50435312, + "step": 24140 + }, + { + "epoch": 3.93890203116078, + "grad_norm": 0.0003412847581785172, + "learning_rate": 0.10203702031759748, + "loss": 0.3342, + "num_input_tokens_seen": 50445792, + "step": 24145 + }, + { + "epoch": 3.9397177583815974, + "grad_norm": 0.00023965239233803004, + "learning_rate": 0.1019812116092384, + "loss": 0.2733, + "num_input_tokens_seen": 50455808, + "step": 24150 + }, + { + "epoch": 3.9405334856024146, + "grad_norm": 0.0003537907905410975, + "learning_rate": 0.10192541030597986, + "loss": 0.3027, + "num_input_tokens_seen": 50464960, + "step": 24155 + }, + { + "epoch": 3.9413492128232317, + "grad_norm": 0.000742763455491513, + "learning_rate": 0.1018696164164272, + "loss": 0.3479, + "num_input_tokens_seen": 50475760, + "step": 24160 + }, + { + "epoch": 3.9421649400440493, + "grad_norm": 0.0003090511600021273, + "learning_rate": 0.10181382994918459, + "loss": 0.3443, + "num_input_tokens_seen": 50487600, + "step": 24165 + }, + { + "epoch": 3.942980667264867, + "grad_norm": 0.0007476829341612756, + "learning_rate": 0.10175805091285492, + "loss": 0.3305, + "num_input_tokens_seen": 50497200, + "step": 24170 + }, + { + "epoch": 3.943796394485684, + "grad_norm": 0.0003563184873200953, + "learning_rate": 0.10170227931603999, + "loss": 0.3017, + "num_input_tokens_seen": 50508416, + "step": 24175 + }, + { + "epoch": 3.944612121706501, + "grad_norm": 0.00029393541626632214, + "learning_rate": 0.10164651516734062, + "loss": 0.3219, + "num_input_tokens_seen": 50518992, + "step": 24180 + }, + { + "epoch": 3.9454278489273187, + "grad_norm": 0.00032148335594683886, + "learning_rate": 0.1015907584753562, + "loss": 0.3145, + "num_input_tokens_seen": 50529664, + "step": 24185 + }, + { + "epoch": 3.9462435761481363, + "grad_norm": 0.00026686786441132426, + "learning_rate": 0.10153500924868523, + "loss": 0.3575, + "num_input_tokens_seen": 50540192, + "step": 24190 + }, + { + "epoch": 3.9470593033689534, + "grad_norm": 0.0006172740831971169, + "learning_rate": 0.10147926749592483, + "loss": 0.3095, + "num_input_tokens_seen": 50549568, + "step": 24195 + }, + { + "epoch": 3.9478750305897705, + "grad_norm": 0.00037493440322577953, + "learning_rate": 0.10142353322567112, + "loss": 0.2905, + "num_input_tokens_seen": 50557680, + "step": 24200 + }, + { + "epoch": 3.9478750305897705, + "eval_loss": 0.31602007150650024, + "eval_runtime": 155.8533, + "eval_samples_per_second": 17.484, + "eval_steps_per_second": 8.745, + "num_input_tokens_seen": 50557680, + "step": 24200 + }, + { + "epoch": 3.948690757810588, + "grad_norm": 0.00037727810558862984, + "learning_rate": 0.1013678064465191, + "loss": 0.3272, + "num_input_tokens_seen": 50569904, + "step": 24205 + }, + { + "epoch": 3.9495064850314057, + "grad_norm": 0.0004006448434665799, + "learning_rate": 0.10131208716706244, + "loss": 0.2842, + "num_input_tokens_seen": 50580304, + "step": 24210 + }, + { + "epoch": 3.950322212252223, + "grad_norm": 0.0002854076155927032, + "learning_rate": 0.10125637539589379, + "loss": 0.3145, + "num_input_tokens_seen": 50591760, + "step": 24215 + }, + { + "epoch": 3.9511379394730404, + "grad_norm": 0.0005399743095040321, + "learning_rate": 0.10120067114160464, + "loss": 0.3226, + "num_input_tokens_seen": 50602928, + "step": 24220 + }, + { + "epoch": 3.9519536666938575, + "grad_norm": 0.0006617774488404393, + "learning_rate": 0.10114497441278517, + "loss": 0.3316, + "num_input_tokens_seen": 50613200, + "step": 24225 + }, + { + "epoch": 3.952769393914675, + "grad_norm": 0.0003312548215035349, + "learning_rate": 0.10108928521802468, + "loss": 0.315, + "num_input_tokens_seen": 50623984, + "step": 24230 + }, + { + "epoch": 3.953585121135492, + "grad_norm": 0.0005665997741743922, + "learning_rate": 0.101033603565911, + "loss": 0.2236, + "num_input_tokens_seen": 50634160, + "step": 24235 + }, + { + "epoch": 3.95440084835631, + "grad_norm": 0.0007106759585440159, + "learning_rate": 0.10097792946503102, + "loss": 0.327, + "num_input_tokens_seen": 50644512, + "step": 24240 + }, + { + "epoch": 3.955216575577127, + "grad_norm": 0.0003804648877121508, + "learning_rate": 0.10092226292397039, + "loss": 0.3367, + "num_input_tokens_seen": 50655456, + "step": 24245 + }, + { + "epoch": 3.9560323027979445, + "grad_norm": 0.0006231848383322358, + "learning_rate": 0.10086660395131354, + "loss": 0.3227, + "num_input_tokens_seen": 50665584, + "step": 24250 + }, + { + "epoch": 3.9568480300187616, + "grad_norm": 0.0004303572641219944, + "learning_rate": 0.10081095255564385, + "loss": 0.2151, + "num_input_tokens_seen": 50675328, + "step": 24255 + }, + { + "epoch": 3.957663757239579, + "grad_norm": 0.0003100406320299953, + "learning_rate": 0.10075530874554335, + "loss": 0.2742, + "num_input_tokens_seen": 50685232, + "step": 24260 + }, + { + "epoch": 3.9584794844603963, + "grad_norm": 0.00027331896126270294, + "learning_rate": 0.10069967252959311, + "loss": 0.301, + "num_input_tokens_seen": 50696144, + "step": 24265 + }, + { + "epoch": 3.959295211681214, + "grad_norm": 0.0004893155419267714, + "learning_rate": 0.10064404391637297, + "loss": 0.3182, + "num_input_tokens_seen": 50706224, + "step": 24270 + }, + { + "epoch": 3.960110938902031, + "grad_norm": 0.0006598885520361364, + "learning_rate": 0.10058842291446145, + "loss": 0.292, + "num_input_tokens_seen": 50716144, + "step": 24275 + }, + { + "epoch": 3.9609266661228486, + "grad_norm": 0.001144138048402965, + "learning_rate": 0.10053280953243608, + "loss": 0.3718, + "num_input_tokens_seen": 50726448, + "step": 24280 + }, + { + "epoch": 3.9617423933436657, + "grad_norm": 0.00044936398626305163, + "learning_rate": 0.10047720377887315, + "loss": 0.3236, + "num_input_tokens_seen": 50736144, + "step": 24285 + }, + { + "epoch": 3.9625581205644833, + "grad_norm": 0.000530094956047833, + "learning_rate": 0.10042160566234767, + "loss": 0.37, + "num_input_tokens_seen": 50746080, + "step": 24290 + }, + { + "epoch": 3.9633738477853004, + "grad_norm": 0.0004623386776074767, + "learning_rate": 0.10036601519143372, + "loss": 0.3294, + "num_input_tokens_seen": 50756960, + "step": 24295 + }, + { + "epoch": 3.964189575006118, + "grad_norm": 0.0004822012269869447, + "learning_rate": 0.1003104323747039, + "loss": 0.2698, + "num_input_tokens_seen": 50767632, + "step": 24300 + }, + { + "epoch": 3.965005302226935, + "grad_norm": 0.00048721322673372924, + "learning_rate": 0.10025485722072984, + "loss": 0.3248, + "num_input_tokens_seen": 50777072, + "step": 24305 + }, + { + "epoch": 3.9658210294477527, + "grad_norm": 0.0004885740345343947, + "learning_rate": 0.10019928973808201, + "loss": 0.3115, + "num_input_tokens_seen": 50788672, + "step": 24310 + }, + { + "epoch": 3.96663675666857, + "grad_norm": 0.0006934190168976784, + "learning_rate": 0.10014372993532945, + "loss": 0.3502, + "num_input_tokens_seen": 50799888, + "step": 24315 + }, + { + "epoch": 3.9674524838893874, + "grad_norm": 0.0005906325532123446, + "learning_rate": 0.1000881778210403, + "loss": 0.3199, + "num_input_tokens_seen": 50809600, + "step": 24320 + }, + { + "epoch": 3.968268211110205, + "grad_norm": 0.0006147132953628898, + "learning_rate": 0.10003263340378142, + "loss": 0.288, + "num_input_tokens_seen": 50817744, + "step": 24325 + }, + { + "epoch": 3.969083938331022, + "grad_norm": 0.0006004309398122132, + "learning_rate": 0.09997709669211834, + "loss": 0.3257, + "num_input_tokens_seen": 50830240, + "step": 24330 + }, + { + "epoch": 3.9698996655518393, + "grad_norm": 0.0003265351406298578, + "learning_rate": 0.0999215676946156, + "loss": 0.3434, + "num_input_tokens_seen": 50840800, + "step": 24335 + }, + { + "epoch": 3.970715392772657, + "grad_norm": 0.0005064026336185634, + "learning_rate": 0.0998660464198364, + "loss": 0.3307, + "num_input_tokens_seen": 50851440, + "step": 24340 + }, + { + "epoch": 3.9715311199934744, + "grad_norm": 0.0006408164626918733, + "learning_rate": 0.09981053287634288, + "loss": 0.3298, + "num_input_tokens_seen": 50861856, + "step": 24345 + }, + { + "epoch": 3.9723468472142915, + "grad_norm": 0.0003111921832896769, + "learning_rate": 0.09975502707269596, + "loss": 0.334, + "num_input_tokens_seen": 50872528, + "step": 24350 + }, + { + "epoch": 3.9731625744351087, + "grad_norm": 0.0007341749151237309, + "learning_rate": 0.09969952901745524, + "loss": 0.2957, + "num_input_tokens_seen": 50882928, + "step": 24355 + }, + { + "epoch": 3.9739783016559262, + "grad_norm": 0.0005310339038260281, + "learning_rate": 0.09964403871917925, + "loss": 0.3787, + "num_input_tokens_seen": 50894336, + "step": 24360 + }, + { + "epoch": 3.974794028876744, + "grad_norm": 0.0003157273749820888, + "learning_rate": 0.09958855618642536, + "loss": 0.2479, + "num_input_tokens_seen": 50904112, + "step": 24365 + }, + { + "epoch": 3.975609756097561, + "grad_norm": 0.0009535637800581753, + "learning_rate": 0.09953308142774955, + "loss": 0.3803, + "num_input_tokens_seen": 50914752, + "step": 24370 + }, + { + "epoch": 3.976425483318378, + "grad_norm": 0.0011203686008229852, + "learning_rate": 0.09947761445170686, + "loss": 0.3829, + "num_input_tokens_seen": 50926304, + "step": 24375 + }, + { + "epoch": 3.9772412105391957, + "grad_norm": 0.0004012413846794516, + "learning_rate": 0.09942215526685086, + "loss": 0.3609, + "num_input_tokens_seen": 50936400, + "step": 24380 + }, + { + "epoch": 3.9780569377600132, + "grad_norm": 0.0005851704045198858, + "learning_rate": 0.09936670388173414, + "loss": 0.2519, + "num_input_tokens_seen": 50946176, + "step": 24385 + }, + { + "epoch": 3.9788726649808304, + "grad_norm": 0.0002112486690748483, + "learning_rate": 0.09931126030490799, + "loss": 0.3157, + "num_input_tokens_seen": 50957424, + "step": 24390 + }, + { + "epoch": 3.9796883922016475, + "grad_norm": 0.00043076969450339675, + "learning_rate": 0.0992558245449225, + "loss": 0.3, + "num_input_tokens_seen": 50968656, + "step": 24395 + }, + { + "epoch": 3.980504119422465, + "grad_norm": 0.00028884338098578155, + "learning_rate": 0.09920039661032651, + "loss": 0.3251, + "num_input_tokens_seen": 50978512, + "step": 24400 + }, + { + "epoch": 3.980504119422465, + "eval_loss": 0.31618034839630127, + "eval_runtime": 156.0884, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 50978512, + "step": 24400 + }, + { + "epoch": 3.9813198466432826, + "grad_norm": 0.0005849683657288551, + "learning_rate": 0.09914497650966782, + "loss": 0.328, + "num_input_tokens_seen": 50990544, + "step": 24405 + }, + { + "epoch": 3.9821355738640998, + "grad_norm": 0.000640656566247344, + "learning_rate": 0.09908956425149276, + "loss": 0.3286, + "num_input_tokens_seen": 50999664, + "step": 24410 + }, + { + "epoch": 3.9829513010849174, + "grad_norm": 0.0003165229572914541, + "learning_rate": 0.09903415984434677, + "loss": 0.294, + "num_input_tokens_seen": 51009920, + "step": 24415 + }, + { + "epoch": 3.9837670283057345, + "grad_norm": 0.0006254612235352397, + "learning_rate": 0.09897876329677373, + "loss": 0.323, + "num_input_tokens_seen": 51020544, + "step": 24420 + }, + { + "epoch": 3.984582755526552, + "grad_norm": 0.001107700983993709, + "learning_rate": 0.09892337461731658, + "loss": 0.3156, + "num_input_tokens_seen": 51030672, + "step": 24425 + }, + { + "epoch": 3.985398482747369, + "grad_norm": 0.007886660285294056, + "learning_rate": 0.09886799381451693, + "loss": 0.3364, + "num_input_tokens_seen": 51040544, + "step": 24430 + }, + { + "epoch": 3.9862142099681868, + "grad_norm": 0.0005736545426771045, + "learning_rate": 0.09881262089691521, + "loss": 0.3384, + "num_input_tokens_seen": 51050672, + "step": 24435 + }, + { + "epoch": 3.987029937189004, + "grad_norm": 0.00028702805866487324, + "learning_rate": 0.09875725587305059, + "loss": 0.3667, + "num_input_tokens_seen": 51061808, + "step": 24440 + }, + { + "epoch": 3.9878456644098215, + "grad_norm": 0.0004812183615285903, + "learning_rate": 0.09870189875146111, + "loss": 0.3179, + "num_input_tokens_seen": 51071872, + "step": 24445 + }, + { + "epoch": 3.9886613916306386, + "grad_norm": 0.00029118001111783087, + "learning_rate": 0.09864654954068346, + "loss": 0.2932, + "num_input_tokens_seen": 51081600, + "step": 24450 + }, + { + "epoch": 3.989477118851456, + "grad_norm": 0.00046348091564141214, + "learning_rate": 0.09859120824925326, + "loss": 0.3674, + "num_input_tokens_seen": 51091584, + "step": 24455 + }, + { + "epoch": 3.9902928460722733, + "grad_norm": 0.0003777245001401752, + "learning_rate": 0.09853587488570474, + "loss": 0.2837, + "num_input_tokens_seen": 51101568, + "step": 24460 + }, + { + "epoch": 3.991108573293091, + "grad_norm": 0.0004976758500561118, + "learning_rate": 0.09848054945857107, + "loss": 0.2894, + "num_input_tokens_seen": 51112080, + "step": 24465 + }, + { + "epoch": 3.991924300513908, + "grad_norm": 0.0005789493443444371, + "learning_rate": 0.09842523197638416, + "loss": 0.306, + "num_input_tokens_seen": 51122960, + "step": 24470 + }, + { + "epoch": 3.9927400277347256, + "grad_norm": 0.0005520779523067176, + "learning_rate": 0.09836992244767452, + "loss": 0.3399, + "num_input_tokens_seen": 51133232, + "step": 24475 + }, + { + "epoch": 3.9935557549555427, + "grad_norm": 0.00036815842031501234, + "learning_rate": 0.09831462088097168, + "loss": 0.2768, + "num_input_tokens_seen": 51143872, + "step": 24480 + }, + { + "epoch": 3.9943714821763603, + "grad_norm": 0.0008592416997998953, + "learning_rate": 0.09825932728480385, + "loss": 0.3342, + "num_input_tokens_seen": 51153024, + "step": 24485 + }, + { + "epoch": 3.9951872093971774, + "grad_norm": 0.0004710450302809477, + "learning_rate": 0.09820404166769794, + "loss": 0.2489, + "num_input_tokens_seen": 51163504, + "step": 24490 + }, + { + "epoch": 3.996002936617995, + "grad_norm": 0.0006625722744502127, + "learning_rate": 0.09814876403817978, + "loss": 0.294, + "num_input_tokens_seen": 51174416, + "step": 24495 + }, + { + "epoch": 3.9968186638388126, + "grad_norm": 0.00037280211108736694, + "learning_rate": 0.09809349440477376, + "loss": 0.332, + "num_input_tokens_seen": 51185584, + "step": 24500 + }, + { + "epoch": 3.9976343910596297, + "grad_norm": 0.000659565906971693, + "learning_rate": 0.09803823277600317, + "loss": 0.2869, + "num_input_tokens_seen": 51195856, + "step": 24505 + }, + { + "epoch": 3.998450118280447, + "grad_norm": 0.0005632743122987449, + "learning_rate": 0.09798297916039014, + "loss": 0.3855, + "num_input_tokens_seen": 51205760, + "step": 24510 + }, + { + "epoch": 3.9992658455012644, + "grad_norm": 0.00023171700013335794, + "learning_rate": 0.09792773356645534, + "loss": 0.3149, + "num_input_tokens_seen": 51216512, + "step": 24515 + }, + { + "epoch": 4.0, + "grad_norm": 0.0004957966157235205, + "learning_rate": 0.09787249600271843, + "loss": 0.3343, + "num_input_tokens_seen": 51225984, + "step": 24520 + }, + { + "epoch": 4.000815727220817, + "grad_norm": 0.00027313039754517376, + "learning_rate": 0.09781726647769776, + "loss": 0.3583, + "num_input_tokens_seen": 51237728, + "step": 24525 + }, + { + "epoch": 4.001631454441635, + "grad_norm": 0.0004982620012015104, + "learning_rate": 0.0977620449999103, + "loss": 0.2912, + "num_input_tokens_seen": 51249056, + "step": 24530 + }, + { + "epoch": 4.002447181662452, + "grad_norm": 0.0002721749187912792, + "learning_rate": 0.09770683157787204, + "loss": 0.3368, + "num_input_tokens_seen": 51257376, + "step": 24535 + }, + { + "epoch": 4.003262908883269, + "grad_norm": 0.0006349445320665836, + "learning_rate": 0.09765162622009745, + "loss": 0.3211, + "num_input_tokens_seen": 51265952, + "step": 24540 + }, + { + "epoch": 4.0040786361040865, + "grad_norm": 0.00037745197187177837, + "learning_rate": 0.09759642893509995, + "loss": 0.3674, + "num_input_tokens_seen": 51277120, + "step": 24545 + }, + { + "epoch": 4.004894363324905, + "grad_norm": 0.0002768064441625029, + "learning_rate": 0.09754123973139169, + "loss": 0.2681, + "num_input_tokens_seen": 51287936, + "step": 24550 + }, + { + "epoch": 4.005710090545722, + "grad_norm": 0.0005306398961693048, + "learning_rate": 0.09748605861748345, + "loss": 0.3416, + "num_input_tokens_seen": 51298240, + "step": 24555 + }, + { + "epoch": 4.006525817766539, + "grad_norm": 0.0011574930977076292, + "learning_rate": 0.0974308856018849, + "loss": 0.3324, + "num_input_tokens_seen": 51307616, + "step": 24560 + }, + { + "epoch": 4.007341544987356, + "grad_norm": 0.000263286754488945, + "learning_rate": 0.09737572069310449, + "loss": 0.319, + "num_input_tokens_seen": 51318096, + "step": 24565 + }, + { + "epoch": 4.008157272208174, + "grad_norm": 0.0003758057428058237, + "learning_rate": 0.09732056389964922, + "loss": 0.3488, + "num_input_tokens_seen": 51329808, + "step": 24570 + }, + { + "epoch": 4.008972999428991, + "grad_norm": 0.0005490386974997818, + "learning_rate": 0.097265415230025, + "loss": 0.3606, + "num_input_tokens_seen": 51341056, + "step": 24575 + }, + { + "epoch": 4.009788726649808, + "grad_norm": 0.0003591825079638511, + "learning_rate": 0.09721027469273648, + "loss": 0.3501, + "num_input_tokens_seen": 51351568, + "step": 24580 + }, + { + "epoch": 4.010604453870625, + "grad_norm": 0.0002666655636858195, + "learning_rate": 0.09715514229628695, + "loss": 0.3267, + "num_input_tokens_seen": 51362640, + "step": 24585 + }, + { + "epoch": 4.011420181091443, + "grad_norm": 0.0005095743108540773, + "learning_rate": 0.09710001804917864, + "loss": 0.3402, + "num_input_tokens_seen": 51372368, + "step": 24590 + }, + { + "epoch": 4.0122359083122605, + "grad_norm": 0.0004520126967690885, + "learning_rate": 0.09704490195991226, + "loss": 0.3357, + "num_input_tokens_seen": 51382528, + "step": 24595 + }, + { + "epoch": 4.013051635533078, + "grad_norm": 0.000365445768693462, + "learning_rate": 0.09698979403698753, + "loss": 0.3528, + "num_input_tokens_seen": 51394160, + "step": 24600 + }, + { + "epoch": 4.013051635533078, + "eval_loss": 0.3200281262397766, + "eval_runtime": 156.0623, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 51394160, + "step": 24600 + }, + { + "epoch": 4.013867362753895, + "grad_norm": 0.0009157260064966977, + "learning_rate": 0.0969346942889027, + "loss": 0.3387, + "num_input_tokens_seen": 51405728, + "step": 24605 + }, + { + "epoch": 4.014683089974713, + "grad_norm": 0.00031711155315861106, + "learning_rate": 0.09687960272415487, + "loss": 0.352, + "num_input_tokens_seen": 51415616, + "step": 24610 + }, + { + "epoch": 4.01549881719553, + "grad_norm": 0.0006321932887658477, + "learning_rate": 0.0968245193512399, + "loss": 0.2852, + "num_input_tokens_seen": 51426240, + "step": 24615 + }, + { + "epoch": 4.016314544416347, + "grad_norm": 0.0003621437062975019, + "learning_rate": 0.09676944417865221, + "loss": 0.2567, + "num_input_tokens_seen": 51436048, + "step": 24620 + }, + { + "epoch": 4.017130271637164, + "grad_norm": 0.0004551849269773811, + "learning_rate": 0.09671437721488517, + "loss": 0.3287, + "num_input_tokens_seen": 51446320, + "step": 24625 + }, + { + "epoch": 4.017945998857982, + "grad_norm": 0.0002765927347354591, + "learning_rate": 0.09665931846843086, + "loss": 0.269, + "num_input_tokens_seen": 51457104, + "step": 24630 + }, + { + "epoch": 4.018761726078799, + "grad_norm": 0.00026887113926932216, + "learning_rate": 0.0966042679477799, + "loss": 0.3478, + "num_input_tokens_seen": 51469088, + "step": 24635 + }, + { + "epoch": 4.0195774532996165, + "grad_norm": 0.00047480573994107544, + "learning_rate": 0.09654922566142186, + "loss": 0.3565, + "num_input_tokens_seen": 51480320, + "step": 24640 + }, + { + "epoch": 4.020393180520434, + "grad_norm": 0.0003422429144848138, + "learning_rate": 0.09649419161784498, + "loss": 0.3153, + "num_input_tokens_seen": 51490976, + "step": 24645 + }, + { + "epoch": 4.021208907741252, + "grad_norm": 0.0005521015264093876, + "learning_rate": 0.09643916582553606, + "loss": 0.2951, + "num_input_tokens_seen": 51500784, + "step": 24650 + }, + { + "epoch": 4.022024634962069, + "grad_norm": 0.00026544462889432907, + "learning_rate": 0.09638414829298093, + "loss": 0.3912, + "num_input_tokens_seen": 51510432, + "step": 24655 + }, + { + "epoch": 4.022840362182886, + "grad_norm": 0.000508505676407367, + "learning_rate": 0.09632913902866386, + "loss": 0.2745, + "num_input_tokens_seen": 51520896, + "step": 24660 + }, + { + "epoch": 4.023656089403703, + "grad_norm": 0.00022300021373666823, + "learning_rate": 0.096274138041068, + "loss": 0.2994, + "num_input_tokens_seen": 51531872, + "step": 24665 + }, + { + "epoch": 4.024471816624521, + "grad_norm": 0.0005674546700902283, + "learning_rate": 0.09621914533867527, + "loss": 0.3221, + "num_input_tokens_seen": 51542816, + "step": 24670 + }, + { + "epoch": 4.025287543845338, + "grad_norm": 0.00037375185638666153, + "learning_rate": 0.09616416092996616, + "loss": 0.3094, + "num_input_tokens_seen": 51553568, + "step": 24675 + }, + { + "epoch": 4.026103271066155, + "grad_norm": 0.0002995021641254425, + "learning_rate": 0.09610918482342, + "loss": 0.2955, + "num_input_tokens_seen": 51563920, + "step": 24680 + }, + { + "epoch": 4.026918998286972, + "grad_norm": 0.00029081711545586586, + "learning_rate": 0.09605421702751478, + "loss": 0.3185, + "num_input_tokens_seen": 51575584, + "step": 24685 + }, + { + "epoch": 4.02773472550779, + "grad_norm": 0.0004682003927882761, + "learning_rate": 0.09599925755072718, + "loss": 0.3495, + "num_input_tokens_seen": 51587680, + "step": 24690 + }, + { + "epoch": 4.028550452728608, + "grad_norm": 0.0004088472342118621, + "learning_rate": 0.09594430640153273, + "loss": 0.2501, + "num_input_tokens_seen": 51598160, + "step": 24695 + }, + { + "epoch": 4.029366179949425, + "grad_norm": 0.00044462186633609235, + "learning_rate": 0.09588936358840547, + "loss": 0.3136, + "num_input_tokens_seen": 51609072, + "step": 24700 + }, + { + "epoch": 4.030181907170242, + "grad_norm": 0.000694530492182821, + "learning_rate": 0.09583442911981836, + "loss": 0.3594, + "num_input_tokens_seen": 51619664, + "step": 24705 + }, + { + "epoch": 4.03099763439106, + "grad_norm": 0.0002507923054508865, + "learning_rate": 0.09577950300424302, + "loss": 0.3316, + "num_input_tokens_seen": 51630016, + "step": 24710 + }, + { + "epoch": 4.031813361611877, + "grad_norm": 0.00023254948609974235, + "learning_rate": 0.09572458525014967, + "loss": 0.2172, + "num_input_tokens_seen": 51639664, + "step": 24715 + }, + { + "epoch": 4.032629088832694, + "grad_norm": 0.0004445068188942969, + "learning_rate": 0.0956696758660073, + "loss": 0.3278, + "num_input_tokens_seen": 51650096, + "step": 24720 + }, + { + "epoch": 4.033444816053512, + "grad_norm": 0.0007793641998432577, + "learning_rate": 0.09561477486028373, + "loss": 0.3591, + "num_input_tokens_seen": 51659952, + "step": 24725 + }, + { + "epoch": 4.034260543274329, + "grad_norm": 0.0005889206076972187, + "learning_rate": 0.09555988224144528, + "loss": 0.3432, + "num_input_tokens_seen": 51669488, + "step": 24730 + }, + { + "epoch": 4.035076270495146, + "grad_norm": 0.0003162597422488034, + "learning_rate": 0.09550499801795717, + "loss": 0.3323, + "num_input_tokens_seen": 51681232, + "step": 24735 + }, + { + "epoch": 4.0358919977159635, + "grad_norm": 0.0003030872903764248, + "learning_rate": 0.09545012219828314, + "loss": 0.3266, + "num_input_tokens_seen": 51692112, + "step": 24740 + }, + { + "epoch": 4.0367077249367815, + "grad_norm": 0.0004384132334962487, + "learning_rate": 0.09539525479088577, + "loss": 0.3213, + "num_input_tokens_seen": 51702096, + "step": 24745 + }, + { + "epoch": 4.037523452157599, + "grad_norm": 0.0003468972281552851, + "learning_rate": 0.0953403958042264, + "loss": 0.2522, + "num_input_tokens_seen": 51713936, + "step": 24750 + }, + { + "epoch": 4.038339179378416, + "grad_norm": 0.0004239145200699568, + "learning_rate": 0.09528554524676484, + "loss": 0.3009, + "num_input_tokens_seen": 51722512, + "step": 24755 + }, + { + "epoch": 4.039154906599233, + "grad_norm": 0.00020332005806267262, + "learning_rate": 0.09523070312695978, + "loss": 0.2997, + "num_input_tokens_seen": 51734720, + "step": 24760 + }, + { + "epoch": 4.039970633820051, + "grad_norm": 0.0006081934552639723, + "learning_rate": 0.09517586945326863, + "loss": 0.3172, + "num_input_tokens_seen": 51745600, + "step": 24765 + }, + { + "epoch": 4.040786361040868, + "grad_norm": 0.0004366060020402074, + "learning_rate": 0.0951210442341473, + "loss": 0.3077, + "num_input_tokens_seen": 51757568, + "step": 24770 + }, + { + "epoch": 4.041602088261685, + "grad_norm": 0.0004068066191393882, + "learning_rate": 0.09506622747805066, + "loss": 0.3134, + "num_input_tokens_seen": 51767328, + "step": 24775 + }, + { + "epoch": 4.042417815482502, + "grad_norm": 0.00025602694950066507, + "learning_rate": 0.09501141919343203, + "loss": 0.2751, + "num_input_tokens_seen": 51777536, + "step": 24780 + }, + { + "epoch": 4.04323354270332, + "grad_norm": 0.00039215051219798625, + "learning_rate": 0.09495661938874361, + "loss": 0.2817, + "num_input_tokens_seen": 51788096, + "step": 24785 + }, + { + "epoch": 4.0440492699241375, + "grad_norm": 0.00038506960845552385, + "learning_rate": 0.0949018280724362, + "loss": 0.298, + "num_input_tokens_seen": 51799056, + "step": 24790 + }, + { + "epoch": 4.044864997144955, + "grad_norm": 0.00025436049327254295, + "learning_rate": 0.09484704525295934, + "loss": 0.2967, + "num_input_tokens_seen": 51810320, + "step": 24795 + }, + { + "epoch": 4.045680724365772, + "grad_norm": 0.000591679650824517, + "learning_rate": 0.09479227093876112, + "loss": 0.2819, + "num_input_tokens_seen": 51821712, + "step": 24800 + }, + { + "epoch": 4.045680724365772, + "eval_loss": 0.3154638409614563, + "eval_runtime": 156.1167, + "eval_samples_per_second": 17.455, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 51821712, + "step": 24800 + }, + { + "epoch": 4.04649645158659, + "grad_norm": 0.00036271088174544275, + "learning_rate": 0.0947375051382886, + "loss": 0.3316, + "num_input_tokens_seen": 51831136, + "step": 24805 + }, + { + "epoch": 4.047312178807407, + "grad_norm": 0.00041047437116503716, + "learning_rate": 0.09468274785998718, + "loss": 0.2754, + "num_input_tokens_seen": 51841488, + "step": 24810 + }, + { + "epoch": 4.048127906028224, + "grad_norm": 0.0004826137737836689, + "learning_rate": 0.09462799911230127, + "loss": 0.2773, + "num_input_tokens_seen": 51852528, + "step": 24815 + }, + { + "epoch": 4.048943633249041, + "grad_norm": 0.0006950558163225651, + "learning_rate": 0.0945732589036737, + "loss": 0.3404, + "num_input_tokens_seen": 51864240, + "step": 24820 + }, + { + "epoch": 4.049759360469859, + "grad_norm": 0.0005899274256080389, + "learning_rate": 0.09451852724254614, + "loss": 0.2395, + "num_input_tokens_seen": 51873168, + "step": 24825 + }, + { + "epoch": 4.050575087690676, + "grad_norm": 0.0005288200336508453, + "learning_rate": 0.09446380413735894, + "loss": 0.3633, + "num_input_tokens_seen": 51884496, + "step": 24830 + }, + { + "epoch": 4.051390814911493, + "grad_norm": 0.0011175203835591674, + "learning_rate": 0.09440908959655099, + "loss": 0.3452, + "num_input_tokens_seen": 51896304, + "step": 24835 + }, + { + "epoch": 4.052206542132311, + "grad_norm": 0.0005875289789400995, + "learning_rate": 0.09435438362856004, + "loss": 0.3744, + "num_input_tokens_seen": 51906608, + "step": 24840 + }, + { + "epoch": 4.053022269353129, + "grad_norm": 0.00047089095460250974, + "learning_rate": 0.0942996862418225, + "loss": 0.3662, + "num_input_tokens_seen": 51916880, + "step": 24845 + }, + { + "epoch": 4.053837996573946, + "grad_norm": 0.0008407302666455507, + "learning_rate": 0.09424499744477322, + "loss": 0.2844, + "num_input_tokens_seen": 51926672, + "step": 24850 + }, + { + "epoch": 4.054653723794763, + "grad_norm": 0.0005057814996689558, + "learning_rate": 0.09419031724584608, + "loss": 0.3118, + "num_input_tokens_seen": 51937280, + "step": 24855 + }, + { + "epoch": 4.05546945101558, + "grad_norm": 0.0002566901966929436, + "learning_rate": 0.09413564565347331, + "loss": 0.3462, + "num_input_tokens_seen": 51948752, + "step": 24860 + }, + { + "epoch": 4.056285178236398, + "grad_norm": 0.0006897934945300221, + "learning_rate": 0.094080982676086, + "loss": 0.3041, + "num_input_tokens_seen": 51959760, + "step": 24865 + }, + { + "epoch": 4.057100905457215, + "grad_norm": 0.0014107782626524568, + "learning_rate": 0.09402632832211395, + "loss": 0.3831, + "num_input_tokens_seen": 51970000, + "step": 24870 + }, + { + "epoch": 4.057916632678032, + "grad_norm": 0.00037099680048413575, + "learning_rate": 0.09397168259998541, + "loss": 0.2893, + "num_input_tokens_seen": 51981456, + "step": 24875 + }, + { + "epoch": 4.058732359898849, + "grad_norm": 0.0005480166873894632, + "learning_rate": 0.09391704551812759, + "loss": 0.2906, + "num_input_tokens_seen": 51991584, + "step": 24880 + }, + { + "epoch": 4.059548087119667, + "grad_norm": 0.0005172427045181394, + "learning_rate": 0.09386241708496605, + "loss": 0.3122, + "num_input_tokens_seen": 52001152, + "step": 24885 + }, + { + "epoch": 4.0603638143404845, + "grad_norm": 0.0004179145907983184, + "learning_rate": 0.09380779730892527, + "loss": 0.4082, + "num_input_tokens_seen": 52011600, + "step": 24890 + }, + { + "epoch": 4.061179541561302, + "grad_norm": 0.0005145521718077362, + "learning_rate": 0.09375318619842836, + "loss": 0.3955, + "num_input_tokens_seen": 52022496, + "step": 24895 + }, + { + "epoch": 4.06199526878212, + "grad_norm": 0.0005606449558399618, + "learning_rate": 0.09369858376189696, + "loss": 0.3193, + "num_input_tokens_seen": 52033472, + "step": 24900 + }, + { + "epoch": 4.062810996002937, + "grad_norm": 0.00037468408118002117, + "learning_rate": 0.09364399000775143, + "loss": 0.3, + "num_input_tokens_seen": 52043280, + "step": 24905 + }, + { + "epoch": 4.063626723223754, + "grad_norm": 0.0004927097470499575, + "learning_rate": 0.09358940494441093, + "loss": 0.3307, + "num_input_tokens_seen": 52053680, + "step": 24910 + }, + { + "epoch": 4.064442450444571, + "grad_norm": 0.0002984025049954653, + "learning_rate": 0.09353482858029301, + "loss": 0.3456, + "num_input_tokens_seen": 52063584, + "step": 24915 + }, + { + "epoch": 4.065258177665389, + "grad_norm": 0.00027922209119424224, + "learning_rate": 0.09348026092381419, + "loss": 0.3561, + "num_input_tokens_seen": 52073488, + "step": 24920 + }, + { + "epoch": 4.066073904886206, + "grad_norm": 0.0003141142369713634, + "learning_rate": 0.09342570198338931, + "loss": 0.2285, + "num_input_tokens_seen": 52082688, + "step": 24925 + }, + { + "epoch": 4.066889632107023, + "grad_norm": 0.00035526364808902144, + "learning_rate": 0.0933711517674322, + "loss": 0.3142, + "num_input_tokens_seen": 52094176, + "step": 24930 + }, + { + "epoch": 4.0677053593278405, + "grad_norm": 0.0003340706171002239, + "learning_rate": 0.09331661028435513, + "loss": 0.2994, + "num_input_tokens_seen": 52105200, + "step": 24935 + }, + { + "epoch": 4.0685210865486585, + "grad_norm": 0.0006023877649568021, + "learning_rate": 0.09326207754256909, + "loss": 0.3342, + "num_input_tokens_seen": 52115024, + "step": 24940 + }, + { + "epoch": 4.069336813769476, + "grad_norm": 0.001083452720195055, + "learning_rate": 0.09320755355048366, + "loss": 0.3089, + "num_input_tokens_seen": 52125360, + "step": 24945 + }, + { + "epoch": 4.070152540990293, + "grad_norm": 0.0008135351235978305, + "learning_rate": 0.09315303831650722, + "loss": 0.3066, + "num_input_tokens_seen": 52135376, + "step": 24950 + }, + { + "epoch": 4.07096826821111, + "grad_norm": 0.00027824233984574676, + "learning_rate": 0.09309853184904661, + "loss": 0.3627, + "num_input_tokens_seen": 52147040, + "step": 24955 + }, + { + "epoch": 4.071783995431928, + "grad_norm": 0.00040946368244476616, + "learning_rate": 0.09304403415650753, + "loss": 0.3228, + "num_input_tokens_seen": 52157680, + "step": 24960 + }, + { + "epoch": 4.072599722652745, + "grad_norm": 0.0003307468432467431, + "learning_rate": 0.09298954524729405, + "loss": 0.3077, + "num_input_tokens_seen": 52167888, + "step": 24965 + }, + { + "epoch": 4.073415449873562, + "grad_norm": 0.0008833098108880222, + "learning_rate": 0.09293506512980916, + "loss": 0.3205, + "num_input_tokens_seen": 52178752, + "step": 24970 + }, + { + "epoch": 4.074231177094379, + "grad_norm": 0.0006407975452020764, + "learning_rate": 0.0928805938124544, + "loss": 0.3236, + "num_input_tokens_seen": 52189200, + "step": 24975 + }, + { + "epoch": 4.075046904315197, + "grad_norm": 0.0003345493460074067, + "learning_rate": 0.09282613130362982, + "loss": 0.289, + "num_input_tokens_seen": 52199984, + "step": 24980 + }, + { + "epoch": 4.0758626315360145, + "grad_norm": 0.000475308857858181, + "learning_rate": 0.09277167761173427, + "loss": 0.2827, + "num_input_tokens_seen": 52211584, + "step": 24985 + }, + { + "epoch": 4.076678358756832, + "grad_norm": 0.0005264716455712914, + "learning_rate": 0.0927172327451653, + "loss": 0.2907, + "num_input_tokens_seen": 52222816, + "step": 24990 + }, + { + "epoch": 4.077494085977649, + "grad_norm": 0.0005072479252703488, + "learning_rate": 0.09266279671231882, + "loss": 0.3111, + "num_input_tokens_seen": 52234016, + "step": 24995 + }, + { + "epoch": 4.078309813198467, + "grad_norm": 0.0003974897845182568, + "learning_rate": 0.09260836952158967, + "loss": 0.2867, + "num_input_tokens_seen": 52244608, + "step": 25000 + }, + { + "epoch": 4.078309813198467, + "eval_loss": 0.31988462805747986, + "eval_runtime": 156.0593, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 52244608, + "step": 25000 + }, + { + "epoch": 4.079125540419284, + "grad_norm": 0.0008271404076367617, + "learning_rate": 0.09255395118137114, + "loss": 0.3274, + "num_input_tokens_seen": 52254960, + "step": 25005 + }, + { + "epoch": 4.079941267640101, + "grad_norm": 0.0003697453357744962, + "learning_rate": 0.09249954170005527, + "loss": 0.2719, + "num_input_tokens_seen": 52265136, + "step": 25010 + }, + { + "epoch": 4.080756994860918, + "grad_norm": 0.00033174757845699787, + "learning_rate": 0.0924451410860327, + "loss": 0.3249, + "num_input_tokens_seen": 52276048, + "step": 25015 + }, + { + "epoch": 4.081572722081736, + "grad_norm": 0.0005706774536520243, + "learning_rate": 0.09239074934769258, + "loss": 0.2535, + "num_input_tokens_seen": 52286256, + "step": 25020 + }, + { + "epoch": 4.082388449302553, + "grad_norm": 0.0005687539232894778, + "learning_rate": 0.09233636649342288, + "loss": 0.3352, + "num_input_tokens_seen": 52295872, + "step": 25025 + }, + { + "epoch": 4.08320417652337, + "grad_norm": 0.00031027651857584715, + "learning_rate": 0.09228199253161017, + "loss": 0.3807, + "num_input_tokens_seen": 52305616, + "step": 25030 + }, + { + "epoch": 4.0840199037441876, + "grad_norm": 0.0006051124655641615, + "learning_rate": 0.09222762747063949, + "loss": 0.3574, + "num_input_tokens_seen": 52317376, + "step": 25035 + }, + { + "epoch": 4.084835630965006, + "grad_norm": 0.0007140432135201991, + "learning_rate": 0.09217327131889473, + "loss": 0.3248, + "num_input_tokens_seen": 52327792, + "step": 25040 + }, + { + "epoch": 4.085651358185823, + "grad_norm": 0.0004075261822436005, + "learning_rate": 0.09211892408475818, + "loss": 0.2704, + "num_input_tokens_seen": 52338112, + "step": 25045 + }, + { + "epoch": 4.08646708540664, + "grad_norm": 0.0002884493151213974, + "learning_rate": 0.09206458577661089, + "loss": 0.3443, + "num_input_tokens_seen": 52347840, + "step": 25050 + }, + { + "epoch": 4.087282812627457, + "grad_norm": 0.0003436188562773168, + "learning_rate": 0.09201025640283263, + "loss": 0.3053, + "num_input_tokens_seen": 52358000, + "step": 25055 + }, + { + "epoch": 4.088098539848275, + "grad_norm": 0.0003004195459652692, + "learning_rate": 0.09195593597180148, + "loss": 0.3113, + "num_input_tokens_seen": 52368512, + "step": 25060 + }, + { + "epoch": 4.088914267069092, + "grad_norm": 0.0005068281898275018, + "learning_rate": 0.09190162449189444, + "loss": 0.3031, + "num_input_tokens_seen": 52379568, + "step": 25065 + }, + { + "epoch": 4.089729994289909, + "grad_norm": 0.0005602976307272911, + "learning_rate": 0.09184732197148705, + "loss": 0.3249, + "num_input_tokens_seen": 52388832, + "step": 25070 + }, + { + "epoch": 4.090545721510727, + "grad_norm": 0.000429609848652035, + "learning_rate": 0.09179302841895343, + "loss": 0.3326, + "num_input_tokens_seen": 52399056, + "step": 25075 + }, + { + "epoch": 4.091361448731544, + "grad_norm": 0.00045774149475619197, + "learning_rate": 0.09173874384266625, + "loss": 0.304, + "num_input_tokens_seen": 52411456, + "step": 25080 + }, + { + "epoch": 4.0921771759523615, + "grad_norm": 0.0003096503496635705, + "learning_rate": 0.09168446825099695, + "loss": 0.2866, + "num_input_tokens_seen": 52422176, + "step": 25085 + }, + { + "epoch": 4.092992903173179, + "grad_norm": 0.0005397468339651823, + "learning_rate": 0.09163020165231545, + "loss": 0.2831, + "num_input_tokens_seen": 52432880, + "step": 25090 + }, + { + "epoch": 4.093808630393997, + "grad_norm": 0.0004695923125836998, + "learning_rate": 0.09157594405499044, + "loss": 0.3436, + "num_input_tokens_seen": 52444256, + "step": 25095 + }, + { + "epoch": 4.094624357614814, + "grad_norm": 0.0003369297191966325, + "learning_rate": 0.09152169546738899, + "loss": 0.29, + "num_input_tokens_seen": 52453856, + "step": 25100 + }, + { + "epoch": 4.095440084835631, + "grad_norm": 0.0007473513833247125, + "learning_rate": 0.09146745589787698, + "loss": 0.3571, + "num_input_tokens_seen": 52463568, + "step": 25105 + }, + { + "epoch": 4.096255812056448, + "grad_norm": 0.000545073882676661, + "learning_rate": 0.09141322535481891, + "loss": 0.2769, + "num_input_tokens_seen": 52473232, + "step": 25110 + }, + { + "epoch": 4.097071539277266, + "grad_norm": 0.00024771536118350923, + "learning_rate": 0.0913590038465777, + "loss": 0.3712, + "num_input_tokens_seen": 52482672, + "step": 25115 + }, + { + "epoch": 4.097887266498083, + "grad_norm": 0.0008510515908710659, + "learning_rate": 0.09130479138151505, + "loss": 0.2646, + "num_input_tokens_seen": 52494656, + "step": 25120 + }, + { + "epoch": 4.0987029937189, + "grad_norm": 0.0003205212706234306, + "learning_rate": 0.09125058796799114, + "loss": 0.3306, + "num_input_tokens_seen": 52504064, + "step": 25125 + }, + { + "epoch": 4.0995187209397175, + "grad_norm": 0.00031103185028769076, + "learning_rate": 0.09119639361436485, + "loss": 0.3308, + "num_input_tokens_seen": 52515760, + "step": 25130 + }, + { + "epoch": 4.1003344481605355, + "grad_norm": 0.0005210265517234802, + "learning_rate": 0.09114220832899368, + "loss": 0.3595, + "num_input_tokens_seen": 52526400, + "step": 25135 + }, + { + "epoch": 4.101150175381353, + "grad_norm": 0.00023747340310364962, + "learning_rate": 0.0910880321202336, + "loss": 0.26, + "num_input_tokens_seen": 52537456, + "step": 25140 + }, + { + "epoch": 4.10196590260217, + "grad_norm": 0.000731165986508131, + "learning_rate": 0.09103386499643933, + "loss": 0.3109, + "num_input_tokens_seen": 52547792, + "step": 25145 + }, + { + "epoch": 4.102781629822987, + "grad_norm": 0.0005796723999083042, + "learning_rate": 0.09097970696596407, + "loss": 0.2945, + "num_input_tokens_seen": 52558288, + "step": 25150 + }, + { + "epoch": 4.103597357043805, + "grad_norm": 0.0005363106029108167, + "learning_rate": 0.09092555803715971, + "loss": 0.2746, + "num_input_tokens_seen": 52568000, + "step": 25155 + }, + { + "epoch": 4.104413084264622, + "grad_norm": 0.000537465384695679, + "learning_rate": 0.0908714182183767, + "loss": 0.2742, + "num_input_tokens_seen": 52577920, + "step": 25160 + }, + { + "epoch": 4.105228811485439, + "grad_norm": 0.0003845940809696913, + "learning_rate": 0.090817287517964, + "loss": 0.3091, + "num_input_tokens_seen": 52587536, + "step": 25165 + }, + { + "epoch": 4.106044538706256, + "grad_norm": 0.000402840378228575, + "learning_rate": 0.09076316594426931, + "loss": 0.2326, + "num_input_tokens_seen": 52597008, + "step": 25170 + }, + { + "epoch": 4.106860265927074, + "grad_norm": 0.0007012481219135225, + "learning_rate": 0.09070905350563888, + "loss": 0.3802, + "num_input_tokens_seen": 52606624, + "step": 25175 + }, + { + "epoch": 4.1076759931478914, + "grad_norm": 0.00036421194090507925, + "learning_rate": 0.09065495021041745, + "loss": 0.2795, + "num_input_tokens_seen": 52617376, + "step": 25180 + }, + { + "epoch": 4.108491720368709, + "grad_norm": 0.0003354646032676101, + "learning_rate": 0.09060085606694851, + "loss": 0.2723, + "num_input_tokens_seen": 52628400, + "step": 25185 + }, + { + "epoch": 4.109307447589526, + "grad_norm": 0.000474188884254545, + "learning_rate": 0.09054677108357405, + "loss": 0.2999, + "num_input_tokens_seen": 52638192, + "step": 25190 + }, + { + "epoch": 4.110123174810344, + "grad_norm": 0.0004235246160533279, + "learning_rate": 0.09049269526863457, + "loss": 0.3619, + "num_input_tokens_seen": 52649008, + "step": 25195 + }, + { + "epoch": 4.110938902031161, + "grad_norm": 0.0004393476410768926, + "learning_rate": 0.09043862863046935, + "loss": 0.3124, + "num_input_tokens_seen": 52659888, + "step": 25200 + }, + { + "epoch": 4.110938902031161, + "eval_loss": 0.3160848617553711, + "eval_runtime": 156.1432, + "eval_samples_per_second": 17.452, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 52659888, + "step": 25200 + }, + { + "epoch": 4.111754629251978, + "grad_norm": 0.0004670302732847631, + "learning_rate": 0.09038457117741602, + "loss": 0.3303, + "num_input_tokens_seen": 52668528, + "step": 25205 + }, + { + "epoch": 4.112570356472795, + "grad_norm": 0.0005957602406851947, + "learning_rate": 0.09033052291781099, + "loss": 0.3992, + "num_input_tokens_seen": 52679824, + "step": 25210 + }, + { + "epoch": 4.113386083693613, + "grad_norm": 0.0003415996325202286, + "learning_rate": 0.09027648385998926, + "loss": 0.3745, + "num_input_tokens_seen": 52688496, + "step": 25215 + }, + { + "epoch": 4.11420181091443, + "grad_norm": 0.0003108804230578244, + "learning_rate": 0.09022245401228417, + "loss": 0.3213, + "num_input_tokens_seen": 52700544, + "step": 25220 + }, + { + "epoch": 4.115017538135247, + "grad_norm": 0.0007459704065695405, + "learning_rate": 0.09016843338302792, + "loss": 0.2669, + "num_input_tokens_seen": 52710304, + "step": 25225 + }, + { + "epoch": 4.1158332653560645, + "grad_norm": 0.0003280231903772801, + "learning_rate": 0.09011442198055115, + "loss": 0.3695, + "num_input_tokens_seen": 52720528, + "step": 25230 + }, + { + "epoch": 4.1166489925768825, + "grad_norm": 0.00038550232420675457, + "learning_rate": 0.09006041981318305, + "loss": 0.2844, + "num_input_tokens_seen": 52731280, + "step": 25235 + }, + { + "epoch": 4.1174647197977, + "grad_norm": 0.00034778245026245713, + "learning_rate": 0.09000642688925149, + "loss": 0.2856, + "num_input_tokens_seen": 52742448, + "step": 25240 + }, + { + "epoch": 4.118280447018517, + "grad_norm": 0.0001928535639308393, + "learning_rate": 0.0899524432170828, + "loss": 0.2872, + "num_input_tokens_seen": 52753328, + "step": 25245 + }, + { + "epoch": 4.119096174239334, + "grad_norm": 0.0007297443808056414, + "learning_rate": 0.08989846880500196, + "loss": 0.3203, + "num_input_tokens_seen": 52764416, + "step": 25250 + }, + { + "epoch": 4.119911901460152, + "grad_norm": 0.0002873956400435418, + "learning_rate": 0.08984450366133256, + "loss": 0.286, + "num_input_tokens_seen": 52775616, + "step": 25255 + }, + { + "epoch": 4.120727628680969, + "grad_norm": 0.0003148837713524699, + "learning_rate": 0.08979054779439664, + "loss": 0.2476, + "num_input_tokens_seen": 52787056, + "step": 25260 + }, + { + "epoch": 4.121543355901786, + "grad_norm": 0.00029562116833403707, + "learning_rate": 0.08973660121251485, + "loss": 0.3398, + "num_input_tokens_seen": 52796016, + "step": 25265 + }, + { + "epoch": 4.122359083122603, + "grad_norm": 0.0007061199867166579, + "learning_rate": 0.08968266392400655, + "loss": 0.3206, + "num_input_tokens_seen": 52807248, + "step": 25270 + }, + { + "epoch": 4.123174810343421, + "grad_norm": 0.0005096991662867367, + "learning_rate": 0.0896287359371894, + "loss": 0.2924, + "num_input_tokens_seen": 52816848, + "step": 25275 + }, + { + "epoch": 4.1239905375642385, + "grad_norm": 0.0010127167915925384, + "learning_rate": 0.08957481726037989, + "loss": 0.3387, + "num_input_tokens_seen": 52826816, + "step": 25280 + }, + { + "epoch": 4.124806264785056, + "grad_norm": 0.0003525955253280699, + "learning_rate": 0.08952090790189286, + "loss": 0.3332, + "num_input_tokens_seen": 52836528, + "step": 25285 + }, + { + "epoch": 4.125621992005874, + "grad_norm": 0.00036904486478306353, + "learning_rate": 0.08946700787004187, + "loss": 0.3165, + "num_input_tokens_seen": 52846576, + "step": 25290 + }, + { + "epoch": 4.126437719226691, + "grad_norm": 0.00035211635986343026, + "learning_rate": 0.08941311717313899, + "loss": 0.3469, + "num_input_tokens_seen": 52857712, + "step": 25295 + }, + { + "epoch": 4.127253446447508, + "grad_norm": 0.0006735860952176154, + "learning_rate": 0.08935923581949483, + "loss": 0.2921, + "num_input_tokens_seen": 52868576, + "step": 25300 + }, + { + "epoch": 4.128069173668325, + "grad_norm": 0.000416646187659353, + "learning_rate": 0.0893053638174185, + "loss": 0.3174, + "num_input_tokens_seen": 52879552, + "step": 25305 + }, + { + "epoch": 4.128884900889143, + "grad_norm": 0.00034885446075350046, + "learning_rate": 0.0892515011752179, + "loss": 0.3166, + "num_input_tokens_seen": 52888144, + "step": 25310 + }, + { + "epoch": 4.12970062810996, + "grad_norm": 0.00043418671702966094, + "learning_rate": 0.08919764790119918, + "loss": 0.3334, + "num_input_tokens_seen": 52898208, + "step": 25315 + }, + { + "epoch": 4.130516355330777, + "grad_norm": 0.0005151101504452527, + "learning_rate": 0.08914380400366727, + "loss": 0.3575, + "num_input_tokens_seen": 52909136, + "step": 25320 + }, + { + "epoch": 4.1313320825515945, + "grad_norm": 0.0004883884103037417, + "learning_rate": 0.08908996949092551, + "loss": 0.2925, + "num_input_tokens_seen": 52918720, + "step": 25325 + }, + { + "epoch": 4.1321478097724125, + "grad_norm": 0.0008558817789889872, + "learning_rate": 0.08903614437127592, + "loss": 0.3042, + "num_input_tokens_seen": 52928464, + "step": 25330 + }, + { + "epoch": 4.13296353699323, + "grad_norm": 0.00046532045234926045, + "learning_rate": 0.088982328653019, + "loss": 0.2887, + "num_input_tokens_seen": 52939008, + "step": 25335 + }, + { + "epoch": 4.133779264214047, + "grad_norm": 0.0008133762748911977, + "learning_rate": 0.0889285223444538, + "loss": 0.4038, + "num_input_tokens_seen": 52949664, + "step": 25340 + }, + { + "epoch": 4.134594991434864, + "grad_norm": 0.0005499979597516358, + "learning_rate": 0.08887472545387787, + "loss": 0.3517, + "num_input_tokens_seen": 52959760, + "step": 25345 + }, + { + "epoch": 4.135410718655682, + "grad_norm": 0.00036075845127925277, + "learning_rate": 0.08882093798958751, + "loss": 0.2691, + "num_input_tokens_seen": 52969664, + "step": 25350 + }, + { + "epoch": 4.136226445876499, + "grad_norm": 0.0004334926779847592, + "learning_rate": 0.08876715995987726, + "loss": 0.343, + "num_input_tokens_seen": 52979200, + "step": 25355 + }, + { + "epoch": 4.137042173097316, + "grad_norm": 0.0009057038696482778, + "learning_rate": 0.08871339137304052, + "loss": 0.3531, + "num_input_tokens_seen": 52988976, + "step": 25360 + }, + { + "epoch": 4.137857900318133, + "grad_norm": 0.001354277366772294, + "learning_rate": 0.0886596322373689, + "loss": 0.3637, + "num_input_tokens_seen": 52999936, + "step": 25365 + }, + { + "epoch": 4.138673627538951, + "grad_norm": 0.0005419871304184198, + "learning_rate": 0.08860588256115293, + "loss": 0.3535, + "num_input_tokens_seen": 53010512, + "step": 25370 + }, + { + "epoch": 4.139489354759768, + "grad_norm": 0.0005913680652156472, + "learning_rate": 0.0885521423526814, + "loss": 0.3411, + "num_input_tokens_seen": 53020880, + "step": 25375 + }, + { + "epoch": 4.1403050819805856, + "grad_norm": 0.0005985705065540969, + "learning_rate": 0.08849841162024165, + "loss": 0.3181, + "num_input_tokens_seen": 53031280, + "step": 25380 + }, + { + "epoch": 4.141120809201403, + "grad_norm": 0.00030358010553754866, + "learning_rate": 0.08844469037211973, + "loss": 0.3055, + "num_input_tokens_seen": 53042176, + "step": 25385 + }, + { + "epoch": 4.141936536422221, + "grad_norm": 0.0007025215891189873, + "learning_rate": 0.08839097861660014, + "loss": 0.2956, + "num_input_tokens_seen": 53051424, + "step": 25390 + }, + { + "epoch": 4.142752263643038, + "grad_norm": 0.00039768763235770166, + "learning_rate": 0.08833727636196585, + "loss": 0.3455, + "num_input_tokens_seen": 53062576, + "step": 25395 + }, + { + "epoch": 4.143567990863855, + "grad_norm": 0.0003141411580145359, + "learning_rate": 0.08828358361649848, + "loss": 0.2975, + "num_input_tokens_seen": 53073648, + "step": 25400 + }, + { + "epoch": 4.143567990863855, + "eval_loss": 0.31847384572029114, + "eval_runtime": 156.1427, + "eval_samples_per_second": 17.452, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 53073648, + "step": 25400 + }, + { + "epoch": 4.144383718084672, + "grad_norm": 0.00054167426424101, + "learning_rate": 0.08822990038847807, + "loss": 0.306, + "num_input_tokens_seen": 53083344, + "step": 25405 + }, + { + "epoch": 4.14519944530549, + "grad_norm": 0.00035435595782473683, + "learning_rate": 0.08817622668618325, + "loss": 0.3231, + "num_input_tokens_seen": 53093280, + "step": 25410 + }, + { + "epoch": 4.146015172526307, + "grad_norm": 0.0004925322136841714, + "learning_rate": 0.08812256251789125, + "loss": 0.3102, + "num_input_tokens_seen": 53103952, + "step": 25415 + }, + { + "epoch": 4.146830899747124, + "grad_norm": 0.0006699520745314658, + "learning_rate": 0.08806890789187766, + "loss": 0.3084, + "num_input_tokens_seen": 53114256, + "step": 25420 + }, + { + "epoch": 4.1476466269679415, + "grad_norm": 0.00034314210643060505, + "learning_rate": 0.08801526281641672, + "loss": 0.347, + "num_input_tokens_seen": 53125136, + "step": 25425 + }, + { + "epoch": 4.1484623541887595, + "grad_norm": 0.0005175905534997582, + "learning_rate": 0.0879616272997813, + "loss": 0.3421, + "num_input_tokens_seen": 53136720, + "step": 25430 + }, + { + "epoch": 4.149278081409577, + "grad_norm": 0.00029491528403013945, + "learning_rate": 0.08790800135024247, + "loss": 0.3449, + "num_input_tokens_seen": 53147456, + "step": 25435 + }, + { + "epoch": 4.150093808630394, + "grad_norm": 0.0007504560053348541, + "learning_rate": 0.08785438497607023, + "loss": 0.3412, + "num_input_tokens_seen": 53156704, + "step": 25440 + }, + { + "epoch": 4.150909535851211, + "grad_norm": 0.00035721604945138097, + "learning_rate": 0.08780077818553277, + "loss": 0.2981, + "num_input_tokens_seen": 53166528, + "step": 25445 + }, + { + "epoch": 4.151725263072029, + "grad_norm": 0.00040215562330558896, + "learning_rate": 0.0877471809868969, + "loss": 0.2615, + "num_input_tokens_seen": 53176992, + "step": 25450 + }, + { + "epoch": 4.152540990292846, + "grad_norm": 0.0008556385873816907, + "learning_rate": 0.08769359338842811, + "loss": 0.2823, + "num_input_tokens_seen": 53188688, + "step": 25455 + }, + { + "epoch": 4.153356717513663, + "grad_norm": 0.00043219365761615336, + "learning_rate": 0.08764001539839016, + "loss": 0.2771, + "num_input_tokens_seen": 53198672, + "step": 25460 + }, + { + "epoch": 4.154172444734481, + "grad_norm": 0.0002825654810294509, + "learning_rate": 0.08758644702504548, + "loss": 0.3672, + "num_input_tokens_seen": 53208640, + "step": 25465 + }, + { + "epoch": 4.154988171955298, + "grad_norm": 0.0005159733118489385, + "learning_rate": 0.0875328882766551, + "loss": 0.3135, + "num_input_tokens_seen": 53219648, + "step": 25470 + }, + { + "epoch": 4.1558038991761155, + "grad_norm": 0.00039516540709882975, + "learning_rate": 0.08747933916147828, + "loss": 0.2601, + "num_input_tokens_seen": 53231152, + "step": 25475 + }, + { + "epoch": 4.156619626396933, + "grad_norm": 0.000337409321218729, + "learning_rate": 0.0874257996877731, + "loss": 0.2889, + "num_input_tokens_seen": 53242784, + "step": 25480 + }, + { + "epoch": 4.157435353617751, + "grad_norm": 0.0006774658686481416, + "learning_rate": 0.08737226986379593, + "loss": 0.3196, + "num_input_tokens_seen": 53252704, + "step": 25485 + }, + { + "epoch": 4.158251080838568, + "grad_norm": 0.0005861903191544116, + "learning_rate": 0.08731874969780173, + "loss": 0.2415, + "num_input_tokens_seen": 53263424, + "step": 25490 + }, + { + "epoch": 4.159066808059385, + "grad_norm": 0.0004111573798581958, + "learning_rate": 0.08726523919804412, + "loss": 0.2849, + "num_input_tokens_seen": 53272240, + "step": 25495 + }, + { + "epoch": 4.159882535280202, + "grad_norm": 0.0005080556147731841, + "learning_rate": 0.08721173837277492, + "loss": 0.315, + "num_input_tokens_seen": 53283120, + "step": 25500 + }, + { + "epoch": 4.16069826250102, + "grad_norm": 0.0006558899767696857, + "learning_rate": 0.08715824723024479, + "loss": 0.3861, + "num_input_tokens_seen": 53294464, + "step": 25505 + }, + { + "epoch": 4.161513989721837, + "grad_norm": 0.0005897200317122042, + "learning_rate": 0.08710476577870258, + "loss": 0.3421, + "num_input_tokens_seen": 53304464, + "step": 25510 + }, + { + "epoch": 4.162329716942654, + "grad_norm": 0.0004640291153918952, + "learning_rate": 0.08705129402639587, + "loss": 0.3655, + "num_input_tokens_seen": 53314688, + "step": 25515 + }, + { + "epoch": 4.163145444163471, + "grad_norm": 0.0004122414975427091, + "learning_rate": 0.08699783198157078, + "loss": 0.3501, + "num_input_tokens_seen": 53325152, + "step": 25520 + }, + { + "epoch": 4.1639611713842895, + "grad_norm": 0.00032037036726251245, + "learning_rate": 0.08694437965247163, + "loss": 0.3015, + "num_input_tokens_seen": 53336240, + "step": 25525 + }, + { + "epoch": 4.164776898605107, + "grad_norm": 0.0002822873357217759, + "learning_rate": 0.08689093704734165, + "loss": 0.3463, + "num_input_tokens_seen": 53346800, + "step": 25530 + }, + { + "epoch": 4.165592625825924, + "grad_norm": 0.0004602801927831024, + "learning_rate": 0.08683750417442222, + "loss": 0.296, + "num_input_tokens_seen": 53357168, + "step": 25535 + }, + { + "epoch": 4.166408353046741, + "grad_norm": 0.0003450413350947201, + "learning_rate": 0.08678408104195334, + "loss": 0.3075, + "num_input_tokens_seen": 53367120, + "step": 25540 + }, + { + "epoch": 4.167224080267559, + "grad_norm": 0.0006566400406882167, + "learning_rate": 0.08673066765817365, + "loss": 0.3902, + "num_input_tokens_seen": 53376864, + "step": 25545 + }, + { + "epoch": 4.168039807488376, + "grad_norm": 0.0005914088105782866, + "learning_rate": 0.08667726403132005, + "loss": 0.2973, + "num_input_tokens_seen": 53386192, + "step": 25550 + }, + { + "epoch": 4.168855534709193, + "grad_norm": 0.0004122977261431515, + "learning_rate": 0.0866238701696281, + "loss": 0.308, + "num_input_tokens_seen": 53397664, + "step": 25555 + }, + { + "epoch": 4.16967126193001, + "grad_norm": 0.0005841234233230352, + "learning_rate": 0.08657048608133185, + "loss": 0.3092, + "num_input_tokens_seen": 53408752, + "step": 25560 + }, + { + "epoch": 4.170486989150828, + "grad_norm": 0.0005028573796153069, + "learning_rate": 0.08651711177466369, + "loss": 0.3487, + "num_input_tokens_seen": 53420560, + "step": 25565 + }, + { + "epoch": 4.171302716371645, + "grad_norm": 0.00031725395820103586, + "learning_rate": 0.08646374725785466, + "loss": 0.2468, + "num_input_tokens_seen": 53429136, + "step": 25570 + }, + { + "epoch": 4.1721184435924625, + "grad_norm": 0.000665303785353899, + "learning_rate": 0.08641039253913434, + "loss": 0.3179, + "num_input_tokens_seen": 53440016, + "step": 25575 + }, + { + "epoch": 4.17293417081328, + "grad_norm": 0.0010408122325316072, + "learning_rate": 0.08635704762673052, + "loss": 0.3598, + "num_input_tokens_seen": 53450816, + "step": 25580 + }, + { + "epoch": 4.173749898034098, + "grad_norm": 0.00023504605633206666, + "learning_rate": 0.08630371252886981, + "loss": 0.3432, + "num_input_tokens_seen": 53461024, + "step": 25585 + }, + { + "epoch": 4.174565625254915, + "grad_norm": 0.0005827974528074265, + "learning_rate": 0.08625038725377704, + "loss": 0.307, + "num_input_tokens_seen": 53470912, + "step": 25590 + }, + { + "epoch": 4.175381352475732, + "grad_norm": 0.00047179593821056187, + "learning_rate": 0.08619707180967566, + "loss": 0.3225, + "num_input_tokens_seen": 53482736, + "step": 25595 + }, + { + "epoch": 4.176197079696549, + "grad_norm": 0.0004283924645278603, + "learning_rate": 0.08614376620478768, + "loss": 0.4165, + "num_input_tokens_seen": 53493696, + "step": 25600 + }, + { + "epoch": 4.176197079696549, + "eval_loss": 0.3147547245025635, + "eval_runtime": 156.2566, + "eval_samples_per_second": 17.439, + "eval_steps_per_second": 8.723, + "num_input_tokens_seen": 53493696, + "step": 25600 + }, + { + "epoch": 4.177012806917367, + "grad_norm": 0.0007238955004140735, + "learning_rate": 0.08609047044733344, + "loss": 0.3088, + "num_input_tokens_seen": 53503328, + "step": 25605 + }, + { + "epoch": 4.177828534138184, + "grad_norm": 0.0005391501472331583, + "learning_rate": 0.08603718454553168, + "loss": 0.2992, + "num_input_tokens_seen": 53514032, + "step": 25610 + }, + { + "epoch": 4.178644261359001, + "grad_norm": 0.0005209394148550928, + "learning_rate": 0.08598390850759997, + "loss": 0.33, + "num_input_tokens_seen": 53524368, + "step": 25615 + }, + { + "epoch": 4.1794599885798185, + "grad_norm": 0.0007695420645177364, + "learning_rate": 0.08593064234175397, + "loss": 0.2576, + "num_input_tokens_seen": 53534976, + "step": 25620 + }, + { + "epoch": 4.1802757158006365, + "grad_norm": 0.00030298190540634096, + "learning_rate": 0.08587738605620815, + "loss": 0.2912, + "num_input_tokens_seen": 53545472, + "step": 25625 + }, + { + "epoch": 4.181091443021454, + "grad_norm": 0.00036514594103209674, + "learning_rate": 0.08582413965917512, + "loss": 0.3711, + "num_input_tokens_seen": 53557344, + "step": 25630 + }, + { + "epoch": 4.181907170242271, + "grad_norm": 0.00031967429094947875, + "learning_rate": 0.08577090315886628, + "loss": 0.3284, + "num_input_tokens_seen": 53567248, + "step": 25635 + }, + { + "epoch": 4.182722897463089, + "grad_norm": 0.00041268550558015704, + "learning_rate": 0.08571767656349136, + "loss": 0.253, + "num_input_tokens_seen": 53576720, + "step": 25640 + }, + { + "epoch": 4.183538624683906, + "grad_norm": 0.0008701176848262548, + "learning_rate": 0.08566445988125847, + "loss": 0.3791, + "num_input_tokens_seen": 53586848, + "step": 25645 + }, + { + "epoch": 4.184354351904723, + "grad_norm": 0.0003843301674351096, + "learning_rate": 0.08561125312037436, + "loss": 0.3524, + "num_input_tokens_seen": 53596352, + "step": 25650 + }, + { + "epoch": 4.18517007912554, + "grad_norm": 0.0003169100673403591, + "learning_rate": 0.08555805628904424, + "loss": 0.2821, + "num_input_tokens_seen": 53606768, + "step": 25655 + }, + { + "epoch": 4.185985806346358, + "grad_norm": 0.00034888446680270135, + "learning_rate": 0.08550486939547161, + "loss": 0.3131, + "num_input_tokens_seen": 53616768, + "step": 25660 + }, + { + "epoch": 4.186801533567175, + "grad_norm": 0.0005012729088775814, + "learning_rate": 0.08545169244785869, + "loss": 0.306, + "num_input_tokens_seen": 53626944, + "step": 25665 + }, + { + "epoch": 4.1876172607879925, + "grad_norm": 0.00036845257272943854, + "learning_rate": 0.08539852545440589, + "loss": 0.3274, + "num_input_tokens_seen": 53636016, + "step": 25670 + }, + { + "epoch": 4.18843298800881, + "grad_norm": 0.0003930449893232435, + "learning_rate": 0.08534536842331235, + "loss": 0.2938, + "num_input_tokens_seen": 53648112, + "step": 25675 + }, + { + "epoch": 4.189248715229628, + "grad_norm": 0.0002413669863017276, + "learning_rate": 0.08529222136277545, + "loss": 0.3685, + "num_input_tokens_seen": 53658112, + "step": 25680 + }, + { + "epoch": 4.190064442450445, + "grad_norm": 0.0006719594239257276, + "learning_rate": 0.08523908428099125, + "loss": 0.3755, + "num_input_tokens_seen": 53669696, + "step": 25685 + }, + { + "epoch": 4.190880169671262, + "grad_norm": 0.00035909470170736313, + "learning_rate": 0.08518595718615402, + "loss": 0.3333, + "num_input_tokens_seen": 53679856, + "step": 25690 + }, + { + "epoch": 4.191695896892079, + "grad_norm": 0.00032300548627972603, + "learning_rate": 0.08513284008645675, + "loss": 0.2592, + "num_input_tokens_seen": 53690176, + "step": 25695 + }, + { + "epoch": 4.192511624112897, + "grad_norm": 0.0005676541477441788, + "learning_rate": 0.08507973299009065, + "loss": 0.2916, + "num_input_tokens_seen": 53699744, + "step": 25700 + }, + { + "epoch": 4.193327351333714, + "grad_norm": 0.0005183383473195136, + "learning_rate": 0.08502663590524563, + "loss": 0.2485, + "num_input_tokens_seen": 53710144, + "step": 25705 + }, + { + "epoch": 4.194143078554531, + "grad_norm": 0.0004840015317313373, + "learning_rate": 0.08497354884010981, + "loss": 0.3071, + "num_input_tokens_seen": 53721808, + "step": 25710 + }, + { + "epoch": 4.194958805775348, + "grad_norm": 0.0002953489311039448, + "learning_rate": 0.0849204718028699, + "loss": 0.3402, + "num_input_tokens_seen": 53732848, + "step": 25715 + }, + { + "epoch": 4.195774532996166, + "grad_norm": 0.00026919011725112796, + "learning_rate": 0.08486740480171118, + "loss": 0.3201, + "num_input_tokens_seen": 53742944, + "step": 25720 + }, + { + "epoch": 4.196590260216984, + "grad_norm": 0.000396048097172752, + "learning_rate": 0.08481434784481706, + "loss": 0.3161, + "num_input_tokens_seen": 53753152, + "step": 25725 + }, + { + "epoch": 4.197405987437801, + "grad_norm": 0.00067865289747715, + "learning_rate": 0.08476130094036968, + "loss": 0.2413, + "num_input_tokens_seen": 53764384, + "step": 25730 + }, + { + "epoch": 4.198221714658618, + "grad_norm": 0.00043929534149356186, + "learning_rate": 0.08470826409654961, + "loss": 0.3248, + "num_input_tokens_seen": 53775360, + "step": 25735 + }, + { + "epoch": 4.199037441879436, + "grad_norm": 0.00042595379636622965, + "learning_rate": 0.08465523732153564, + "loss": 0.2595, + "num_input_tokens_seen": 53784400, + "step": 25740 + }, + { + "epoch": 4.199853169100253, + "grad_norm": 0.00040629348950460553, + "learning_rate": 0.08460222062350532, + "loss": 0.2865, + "num_input_tokens_seen": 53794048, + "step": 25745 + }, + { + "epoch": 4.20066889632107, + "grad_norm": 0.0004259705019649118, + "learning_rate": 0.08454921401063442, + "loss": 0.3174, + "num_input_tokens_seen": 53804560, + "step": 25750 + }, + { + "epoch": 4.201484623541887, + "grad_norm": 0.0007851631962694228, + "learning_rate": 0.08449621749109716, + "loss": 0.3756, + "num_input_tokens_seen": 53815536, + "step": 25755 + }, + { + "epoch": 4.202300350762705, + "grad_norm": 0.0002703365171328187, + "learning_rate": 0.08444323107306641, + "loss": 0.29, + "num_input_tokens_seen": 53826400, + "step": 25760 + }, + { + "epoch": 4.203116077983522, + "grad_norm": 0.000604555825702846, + "learning_rate": 0.0843902547647132, + "loss": 0.3109, + "num_input_tokens_seen": 53835808, + "step": 25765 + }, + { + "epoch": 4.2039318052043395, + "grad_norm": 0.0009204238303937018, + "learning_rate": 0.0843372885742072, + "loss": 0.4416, + "num_input_tokens_seen": 53844880, + "step": 25770 + }, + { + "epoch": 4.204747532425157, + "grad_norm": 0.0005215967539697886, + "learning_rate": 0.08428433250971652, + "loss": 0.2721, + "num_input_tokens_seen": 53856352, + "step": 25775 + }, + { + "epoch": 4.205563259645975, + "grad_norm": 0.00038315041456371546, + "learning_rate": 0.08423138657940757, + "loss": 0.2789, + "num_input_tokens_seen": 53867232, + "step": 25780 + }, + { + "epoch": 4.206378986866792, + "grad_norm": 0.0010117094498127699, + "learning_rate": 0.08417845079144536, + "loss": 0.3783, + "num_input_tokens_seen": 53877248, + "step": 25785 + }, + { + "epoch": 4.207194714087609, + "grad_norm": 0.0003584550286177546, + "learning_rate": 0.08412552515399314, + "loss": 0.2824, + "num_input_tokens_seen": 53887328, + "step": 25790 + }, + { + "epoch": 4.208010441308426, + "grad_norm": 0.0007056255708448589, + "learning_rate": 0.08407260967521278, + "loss": 0.3254, + "num_input_tokens_seen": 53897504, + "step": 25795 + }, + { + "epoch": 4.208826168529244, + "grad_norm": 0.0004885845119133592, + "learning_rate": 0.08401970436326454, + "loss": 0.3362, + "num_input_tokens_seen": 53907648, + "step": 25800 + }, + { + "epoch": 4.208826168529244, + "eval_loss": 0.31577369570732117, + "eval_runtime": 156.0886, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 53907648, + "step": 25800 + }, + { + "epoch": 4.209641895750061, + "grad_norm": 0.000640285958070308, + "learning_rate": 0.08396680922630702, + "loss": 0.3484, + "num_input_tokens_seen": 53917760, + "step": 25805 + }, + { + "epoch": 4.210457622970878, + "grad_norm": 0.0003450109506957233, + "learning_rate": 0.08391392427249732, + "loss": 0.3327, + "num_input_tokens_seen": 53928512, + "step": 25810 + }, + { + "epoch": 4.211273350191696, + "grad_norm": 0.0003647210542112589, + "learning_rate": 0.08386104950999107, + "loss": 0.3247, + "num_input_tokens_seen": 53939424, + "step": 25815 + }, + { + "epoch": 4.2120890774125135, + "grad_norm": 0.0004128518339712173, + "learning_rate": 0.0838081849469421, + "loss": 0.3012, + "num_input_tokens_seen": 53949408, + "step": 25820 + }, + { + "epoch": 4.212904804633331, + "grad_norm": 0.000559865846298635, + "learning_rate": 0.08375533059150281, + "loss": 0.2849, + "num_input_tokens_seen": 53960928, + "step": 25825 + }, + { + "epoch": 4.213720531854148, + "grad_norm": 0.00034903723280876875, + "learning_rate": 0.08370248645182406, + "loss": 0.2979, + "num_input_tokens_seen": 53970144, + "step": 25830 + }, + { + "epoch": 4.214536259074965, + "grad_norm": 0.0006485744379460812, + "learning_rate": 0.083649652536055, + "loss": 0.3554, + "num_input_tokens_seen": 53979808, + "step": 25835 + }, + { + "epoch": 4.215351986295783, + "grad_norm": 0.0004660431877709925, + "learning_rate": 0.08359682885234339, + "loss": 0.3138, + "num_input_tokens_seen": 53990992, + "step": 25840 + }, + { + "epoch": 4.2161677135166, + "grad_norm": 0.0006468167994171381, + "learning_rate": 0.08354401540883516, + "loss": 0.2818, + "num_input_tokens_seen": 54001200, + "step": 25845 + }, + { + "epoch": 4.216983440737417, + "grad_norm": 0.0003911625244654715, + "learning_rate": 0.0834912122136749, + "loss": 0.2896, + "num_input_tokens_seen": 54011120, + "step": 25850 + }, + { + "epoch": 4.217799167958235, + "grad_norm": 0.0006547868251800537, + "learning_rate": 0.0834384192750056, + "loss": 0.325, + "num_input_tokens_seen": 54021888, + "step": 25855 + }, + { + "epoch": 4.218614895179052, + "grad_norm": 0.0003388152690604329, + "learning_rate": 0.08338563660096844, + "loss": 0.3617, + "num_input_tokens_seen": 54032800, + "step": 25860 + }, + { + "epoch": 4.219430622399869, + "grad_norm": 0.0005179644213058054, + "learning_rate": 0.08333286419970329, + "loss": 0.2775, + "num_input_tokens_seen": 54044784, + "step": 25865 + }, + { + "epoch": 4.220246349620687, + "grad_norm": 0.0005170778604224324, + "learning_rate": 0.08328010207934824, + "loss": 0.2946, + "num_input_tokens_seen": 54055664, + "step": 25870 + }, + { + "epoch": 4.221062076841505, + "grad_norm": 0.0004552812606561929, + "learning_rate": 0.08322735024803989, + "loss": 0.2805, + "num_input_tokens_seen": 54066416, + "step": 25875 + }, + { + "epoch": 4.221877804062322, + "grad_norm": 0.0010310668731108308, + "learning_rate": 0.08317460871391331, + "loss": 0.3142, + "num_input_tokens_seen": 54077792, + "step": 25880 + }, + { + "epoch": 4.222693531283139, + "grad_norm": 0.0009827424073591828, + "learning_rate": 0.08312187748510179, + "loss": 0.3398, + "num_input_tokens_seen": 54088192, + "step": 25885 + }, + { + "epoch": 4.223509258503956, + "grad_norm": 0.0004271532816346735, + "learning_rate": 0.08306915656973726, + "loss": 0.3373, + "num_input_tokens_seen": 54098160, + "step": 25890 + }, + { + "epoch": 4.224324985724774, + "grad_norm": 0.0008982185972854495, + "learning_rate": 0.08301644597594988, + "loss": 0.3301, + "num_input_tokens_seen": 54108704, + "step": 25895 + }, + { + "epoch": 4.225140712945591, + "grad_norm": 0.001379090128466487, + "learning_rate": 0.08296374571186826, + "loss": 0.3632, + "num_input_tokens_seen": 54120736, + "step": 25900 + }, + { + "epoch": 4.225956440166408, + "grad_norm": 0.0007692537037655711, + "learning_rate": 0.08291105578561955, + "loss": 0.2727, + "num_input_tokens_seen": 54131696, + "step": 25905 + }, + { + "epoch": 4.226772167387225, + "grad_norm": 0.00030037437682040036, + "learning_rate": 0.08285837620532904, + "loss": 0.3155, + "num_input_tokens_seen": 54141552, + "step": 25910 + }, + { + "epoch": 4.227587894608043, + "grad_norm": 0.000677332456689328, + "learning_rate": 0.0828057069791207, + "loss": 0.3996, + "num_input_tokens_seen": 54152112, + "step": 25915 + }, + { + "epoch": 4.2284036218288605, + "grad_norm": 0.0010643729474395514, + "learning_rate": 0.0827530481151168, + "loss": 0.2828, + "num_input_tokens_seen": 54162400, + "step": 25920 + }, + { + "epoch": 4.229219349049678, + "grad_norm": 0.000350664253346622, + "learning_rate": 0.08270039962143792, + "loss": 0.3389, + "num_input_tokens_seen": 54173472, + "step": 25925 + }, + { + "epoch": 4.230035076270495, + "grad_norm": 0.0005240938626229763, + "learning_rate": 0.08264776150620314, + "loss": 0.3377, + "num_input_tokens_seen": 54185312, + "step": 25930 + }, + { + "epoch": 4.230850803491313, + "grad_norm": 0.00048085671733133495, + "learning_rate": 0.08259513377753, + "loss": 0.2853, + "num_input_tokens_seen": 54196304, + "step": 25935 + }, + { + "epoch": 4.23166653071213, + "grad_norm": 0.00037587014958262444, + "learning_rate": 0.08254251644353423, + "loss": 0.2921, + "num_input_tokens_seen": 54206400, + "step": 25940 + }, + { + "epoch": 4.232482257932947, + "grad_norm": 0.0011010380694642663, + "learning_rate": 0.08248990951233022, + "loss": 0.3447, + "num_input_tokens_seen": 54215760, + "step": 25945 + }, + { + "epoch": 4.233297985153764, + "grad_norm": 0.0005066032172180712, + "learning_rate": 0.08243731299203048, + "loss": 0.2729, + "num_input_tokens_seen": 54226768, + "step": 25950 + }, + { + "epoch": 4.234113712374582, + "grad_norm": 0.00039809412555769086, + "learning_rate": 0.08238472689074612, + "loss": 0.2981, + "num_input_tokens_seen": 54237120, + "step": 25955 + }, + { + "epoch": 4.234929439595399, + "grad_norm": 0.00045700412010774016, + "learning_rate": 0.08233215121658666, + "loss": 0.3355, + "num_input_tokens_seen": 54247664, + "step": 25960 + }, + { + "epoch": 4.2357451668162165, + "grad_norm": 0.0004697238327935338, + "learning_rate": 0.08227958597765982, + "loss": 0.2903, + "num_input_tokens_seen": 54257248, + "step": 25965 + }, + { + "epoch": 4.236560894037034, + "grad_norm": 0.0004368131922092289, + "learning_rate": 0.08222703118207181, + "loss": 0.3484, + "num_input_tokens_seen": 54265808, + "step": 25970 + }, + { + "epoch": 4.237376621257852, + "grad_norm": 0.0005123066948726773, + "learning_rate": 0.08217448683792734, + "loss": 0.2901, + "num_input_tokens_seen": 54277552, + "step": 25975 + }, + { + "epoch": 4.238192348478669, + "grad_norm": 0.0006019356660544872, + "learning_rate": 0.08212195295332926, + "loss": 0.3093, + "num_input_tokens_seen": 54288288, + "step": 25980 + }, + { + "epoch": 4.239008075699486, + "grad_norm": 0.0005717556923627853, + "learning_rate": 0.08206942953637915, + "loss": 0.3319, + "num_input_tokens_seen": 54297600, + "step": 25985 + }, + { + "epoch": 4.239823802920303, + "grad_norm": 0.0008001414826139808, + "learning_rate": 0.08201691659517658, + "loss": 0.2931, + "num_input_tokens_seen": 54307440, + "step": 25990 + }, + { + "epoch": 4.240639530141121, + "grad_norm": 0.0005656671128235757, + "learning_rate": 0.08196441413781981, + "loss": 0.2992, + "num_input_tokens_seen": 54317024, + "step": 25995 + }, + { + "epoch": 4.241455257361938, + "grad_norm": 0.000853347301017493, + "learning_rate": 0.08191192217240544, + "loss": 0.258, + "num_input_tokens_seen": 54327568, + "step": 26000 + }, + { + "epoch": 4.241455257361938, + "eval_loss": 0.3151492476463318, + "eval_runtime": 156.172, + "eval_samples_per_second": 17.449, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 54327568, + "step": 26000 + }, + { + "epoch": 4.242270984582755, + "grad_norm": 0.0005234795389696956, + "learning_rate": 0.08185944070702823, + "loss": 0.3018, + "num_input_tokens_seen": 54337136, + "step": 26005 + }, + { + "epoch": 4.243086711803572, + "grad_norm": 0.00036557327257469296, + "learning_rate": 0.08180696974978159, + "loss": 0.2922, + "num_input_tokens_seen": 54348752, + "step": 26010 + }, + { + "epoch": 4.2439024390243905, + "grad_norm": 0.0010469018016010523, + "learning_rate": 0.08175450930875724, + "loss": 0.3172, + "num_input_tokens_seen": 54359840, + "step": 26015 + }, + { + "epoch": 4.244718166245208, + "grad_norm": 0.0006486761849373579, + "learning_rate": 0.08170205939204513, + "loss": 0.2593, + "num_input_tokens_seen": 54370656, + "step": 26020 + }, + { + "epoch": 4.245533893466025, + "grad_norm": 0.000621859566308558, + "learning_rate": 0.08164962000773379, + "loss": 0.364, + "num_input_tokens_seen": 54380768, + "step": 26025 + }, + { + "epoch": 4.246349620686843, + "grad_norm": 0.0004491420404519886, + "learning_rate": 0.08159719116390995, + "loss": 0.364, + "num_input_tokens_seen": 54390688, + "step": 26030 + }, + { + "epoch": 4.24716534790766, + "grad_norm": 0.0006951657123863697, + "learning_rate": 0.08154477286865887, + "loss": 0.3353, + "num_input_tokens_seen": 54400368, + "step": 26035 + }, + { + "epoch": 4.247981075128477, + "grad_norm": 0.0006350920884869993, + "learning_rate": 0.08149236513006404, + "loss": 0.2913, + "num_input_tokens_seen": 54411168, + "step": 26040 + }, + { + "epoch": 4.248796802349294, + "grad_norm": 0.000798349326942116, + "learning_rate": 0.08143996795620746, + "loss": 0.2985, + "num_input_tokens_seen": 54421344, + "step": 26045 + }, + { + "epoch": 4.249612529570112, + "grad_norm": 0.0004021915956400335, + "learning_rate": 0.08138758135516938, + "loss": 0.3065, + "num_input_tokens_seen": 54433344, + "step": 26050 + }, + { + "epoch": 4.250428256790929, + "grad_norm": 0.000352782808477059, + "learning_rate": 0.08133520533502851, + "loss": 0.2925, + "num_input_tokens_seen": 54444256, + "step": 26055 + }, + { + "epoch": 4.251243984011746, + "grad_norm": 0.0006619796040467918, + "learning_rate": 0.08128283990386184, + "loss": 0.3188, + "num_input_tokens_seen": 54454384, + "step": 26060 + }, + { + "epoch": 4.2520597112325635, + "grad_norm": 0.0009318329975940287, + "learning_rate": 0.08123048506974488, + "loss": 0.3097, + "num_input_tokens_seen": 54466032, + "step": 26065 + }, + { + "epoch": 4.252875438453382, + "grad_norm": 0.0024002769496291876, + "learning_rate": 0.08117814084075124, + "loss": 0.3382, + "num_input_tokens_seen": 54476704, + "step": 26070 + }, + { + "epoch": 4.253691165674199, + "grad_norm": 0.0005038862582296133, + "learning_rate": 0.08112580722495318, + "loss": 0.3494, + "num_input_tokens_seen": 54488320, + "step": 26075 + }, + { + "epoch": 4.254506892895016, + "grad_norm": 0.00038922467501834035, + "learning_rate": 0.08107348423042122, + "loss": 0.2666, + "num_input_tokens_seen": 54497456, + "step": 26080 + }, + { + "epoch": 4.255322620115833, + "grad_norm": 0.00037241788231767714, + "learning_rate": 0.08102117186522413, + "loss": 0.31, + "num_input_tokens_seen": 54508944, + "step": 26085 + }, + { + "epoch": 4.256138347336651, + "grad_norm": 0.0003280637611169368, + "learning_rate": 0.08096887013742916, + "loss": 0.3181, + "num_input_tokens_seen": 54518960, + "step": 26090 + }, + { + "epoch": 4.256954074557468, + "grad_norm": 0.00039539826684631407, + "learning_rate": 0.08091657905510198, + "loss": 0.3273, + "num_input_tokens_seen": 54527920, + "step": 26095 + }, + { + "epoch": 4.257769801778285, + "grad_norm": 0.00040072709089145064, + "learning_rate": 0.08086429862630642, + "loss": 0.2829, + "num_input_tokens_seen": 54539904, + "step": 26100 + }, + { + "epoch": 4.258585528999102, + "grad_norm": 0.000967279658652842, + "learning_rate": 0.08081202885910488, + "loss": 0.302, + "num_input_tokens_seen": 54550416, + "step": 26105 + }, + { + "epoch": 4.25940125621992, + "grad_norm": 0.0003858056734316051, + "learning_rate": 0.08075976976155795, + "loss": 0.2701, + "num_input_tokens_seen": 54560640, + "step": 26110 + }, + { + "epoch": 4.2602169834407375, + "grad_norm": 0.0006105828797444701, + "learning_rate": 0.08070752134172461, + "loss": 0.2633, + "num_input_tokens_seen": 54570752, + "step": 26115 + }, + { + "epoch": 4.261032710661555, + "grad_norm": 0.0006483561010099947, + "learning_rate": 0.08065528360766229, + "loss": 0.3054, + "num_input_tokens_seen": 54580224, + "step": 26120 + }, + { + "epoch": 4.261848437882372, + "grad_norm": 0.001047066762112081, + "learning_rate": 0.08060305656742664, + "loss": 0.3555, + "num_input_tokens_seen": 54589488, + "step": 26125 + }, + { + "epoch": 4.26266416510319, + "grad_norm": 0.000762227748055011, + "learning_rate": 0.08055084022907182, + "loss": 0.3811, + "num_input_tokens_seen": 54599072, + "step": 26130 + }, + { + "epoch": 4.263479892324007, + "grad_norm": 0.0005796291516162455, + "learning_rate": 0.08049863460065014, + "loss": 0.3435, + "num_input_tokens_seen": 54609312, + "step": 26135 + }, + { + "epoch": 4.264295619544824, + "grad_norm": 0.0006439991993829608, + "learning_rate": 0.0804464396902124, + "loss": 0.3093, + "num_input_tokens_seen": 54619472, + "step": 26140 + }, + { + "epoch": 4.265111346765641, + "grad_norm": 0.001123769674450159, + "learning_rate": 0.08039425550580777, + "loss": 0.3789, + "num_input_tokens_seen": 54627840, + "step": 26145 + }, + { + "epoch": 4.265927073986459, + "grad_norm": 0.0005702669732272625, + "learning_rate": 0.08034208205548363, + "loss": 0.294, + "num_input_tokens_seen": 54638848, + "step": 26150 + }, + { + "epoch": 4.266742801207276, + "grad_norm": 0.0007892863941378891, + "learning_rate": 0.08028991934728581, + "loss": 0.3347, + "num_input_tokens_seen": 54649680, + "step": 26155 + }, + { + "epoch": 4.2675585284280935, + "grad_norm": 0.00036872667260468006, + "learning_rate": 0.0802377673892585, + "loss": 0.4044, + "num_input_tokens_seen": 54658608, + "step": 26160 + }, + { + "epoch": 4.268374255648911, + "grad_norm": 0.0005566354375332594, + "learning_rate": 0.0801856261894441, + "loss": 0.3114, + "num_input_tokens_seen": 54670272, + "step": 26165 + }, + { + "epoch": 4.269189982869729, + "grad_norm": 0.00045803270768374205, + "learning_rate": 0.08013349575588354, + "loss": 0.2989, + "num_input_tokens_seen": 54682048, + "step": 26170 + }, + { + "epoch": 4.270005710090546, + "grad_norm": 0.00028528564143925905, + "learning_rate": 0.08008137609661586, + "loss": 0.3227, + "num_input_tokens_seen": 54691856, + "step": 26175 + }, + { + "epoch": 4.270821437311363, + "grad_norm": 0.0006649026763625443, + "learning_rate": 0.08002926721967872, + "loss": 0.3066, + "num_input_tokens_seen": 54701552, + "step": 26180 + }, + { + "epoch": 4.27163716453218, + "grad_norm": 0.0006486261845566332, + "learning_rate": 0.07997716913310782, + "loss": 0.313, + "num_input_tokens_seen": 54712528, + "step": 26185 + }, + { + "epoch": 4.272452891752998, + "grad_norm": 0.0004175821377430111, + "learning_rate": 0.07992508184493745, + "loss": 0.307, + "num_input_tokens_seen": 54723360, + "step": 26190 + }, + { + "epoch": 4.273268618973815, + "grad_norm": 0.000675372255500406, + "learning_rate": 0.07987300536320001, + "loss": 0.3575, + "num_input_tokens_seen": 54733856, + "step": 26195 + }, + { + "epoch": 4.274084346194632, + "grad_norm": 0.00023539191170129925, + "learning_rate": 0.07982093969592649, + "loss": 0.3329, + "num_input_tokens_seen": 54743840, + "step": 26200 + }, + { + "epoch": 4.274084346194632, + "eval_loss": 0.3232954442501068, + "eval_runtime": 155.8265, + "eval_samples_per_second": 17.487, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 54743840, + "step": 26200 + }, + { + "epoch": 4.27490007341545, + "grad_norm": 0.0006516753346659243, + "learning_rate": 0.07976888485114592, + "loss": 0.3535, + "num_input_tokens_seen": 54753920, + "step": 26205 + }, + { + "epoch": 4.275715800636267, + "grad_norm": 0.0004350942908786237, + "learning_rate": 0.07971684083688595, + "loss": 0.3203, + "num_input_tokens_seen": 54763408, + "step": 26210 + }, + { + "epoch": 4.276531527857085, + "grad_norm": 0.00045268022222444415, + "learning_rate": 0.0796648076611723, + "loss": 0.2638, + "num_input_tokens_seen": 54774096, + "step": 26215 + }, + { + "epoch": 4.277347255077902, + "grad_norm": 0.0006090457318350673, + "learning_rate": 0.07961278533202922, + "loss": 0.3292, + "num_input_tokens_seen": 54784560, + "step": 26220 + }, + { + "epoch": 4.27816298229872, + "grad_norm": 0.0006249453872442245, + "learning_rate": 0.07956077385747919, + "loss": 0.3356, + "num_input_tokens_seen": 54795056, + "step": 26225 + }, + { + "epoch": 4.278978709519537, + "grad_norm": 0.0006293043261393905, + "learning_rate": 0.079508773245543, + "loss": 0.3147, + "num_input_tokens_seen": 54804816, + "step": 26230 + }, + { + "epoch": 4.279794436740354, + "grad_norm": 0.0006704065017402172, + "learning_rate": 0.07945678350423982, + "loss": 0.3073, + "num_input_tokens_seen": 54815184, + "step": 26235 + }, + { + "epoch": 4.280610163961171, + "grad_norm": 0.0004711865622084588, + "learning_rate": 0.07940480464158717, + "loss": 0.2978, + "num_input_tokens_seen": 54826192, + "step": 26240 + }, + { + "epoch": 4.281425891181989, + "grad_norm": 0.0004535960906650871, + "learning_rate": 0.07935283666560076, + "loss": 0.3174, + "num_input_tokens_seen": 54837392, + "step": 26245 + }, + { + "epoch": 4.282241618402806, + "grad_norm": 0.000718028168193996, + "learning_rate": 0.07930087958429478, + "loss": 0.3135, + "num_input_tokens_seen": 54847600, + "step": 26250 + }, + { + "epoch": 4.283057345623623, + "grad_norm": 0.0004083929234184325, + "learning_rate": 0.07924893340568159, + "loss": 0.3151, + "num_input_tokens_seen": 54858128, + "step": 26255 + }, + { + "epoch": 4.2838730728444405, + "grad_norm": 0.00032344527426175773, + "learning_rate": 0.07919699813777205, + "loss": 0.3271, + "num_input_tokens_seen": 54868896, + "step": 26260 + }, + { + "epoch": 4.2846888000652585, + "grad_norm": 0.0005210945964790881, + "learning_rate": 0.07914507378857515, + "loss": 0.387, + "num_input_tokens_seen": 54879280, + "step": 26265 + }, + { + "epoch": 4.285504527286076, + "grad_norm": 0.000279987754765898, + "learning_rate": 0.07909316036609822, + "loss": 0.3183, + "num_input_tokens_seen": 54890368, + "step": 26270 + }, + { + "epoch": 4.286320254506893, + "grad_norm": 0.0005813715397380292, + "learning_rate": 0.07904125787834704, + "loss": 0.3771, + "num_input_tokens_seen": 54900016, + "step": 26275 + }, + { + "epoch": 4.28713598172771, + "grad_norm": 0.0004490804858505726, + "learning_rate": 0.07898936633332569, + "loss": 0.2752, + "num_input_tokens_seen": 54910112, + "step": 26280 + }, + { + "epoch": 4.287951708948528, + "grad_norm": 0.0006125876680016518, + "learning_rate": 0.07893748573903635, + "loss": 0.4596, + "num_input_tokens_seen": 54921200, + "step": 26285 + }, + { + "epoch": 4.288767436169345, + "grad_norm": 0.0006024452159181237, + "learning_rate": 0.0788856161034798, + "loss": 0.3648, + "num_input_tokens_seen": 54932144, + "step": 26290 + }, + { + "epoch": 4.289583163390162, + "grad_norm": 0.001028039725497365, + "learning_rate": 0.07883375743465487, + "loss": 0.3282, + "num_input_tokens_seen": 54941744, + "step": 26295 + }, + { + "epoch": 4.290398890610979, + "grad_norm": 0.0006652727606706321, + "learning_rate": 0.07878190974055888, + "loss": 0.3302, + "num_input_tokens_seen": 54950800, + "step": 26300 + }, + { + "epoch": 4.291214617831797, + "grad_norm": 0.0004021104541607201, + "learning_rate": 0.07873007302918746, + "loss": 0.3181, + "num_input_tokens_seen": 54961936, + "step": 26305 + }, + { + "epoch": 4.2920303450526145, + "grad_norm": 0.00044794954010285437, + "learning_rate": 0.07867824730853433, + "loss": 0.3626, + "num_input_tokens_seen": 54972512, + "step": 26310 + }, + { + "epoch": 4.292846072273432, + "grad_norm": 0.0007059010094963014, + "learning_rate": 0.07862643258659176, + "loss": 0.3419, + "num_input_tokens_seen": 54983040, + "step": 26315 + }, + { + "epoch": 4.293661799494249, + "grad_norm": 0.000360235630068928, + "learning_rate": 0.07857462887135026, + "loss": 0.3159, + "num_input_tokens_seen": 54993328, + "step": 26320 + }, + { + "epoch": 4.294477526715067, + "grad_norm": 0.000855427875649184, + "learning_rate": 0.0785228361707986, + "loss": 0.3698, + "num_input_tokens_seen": 55003680, + "step": 26325 + }, + { + "epoch": 4.295293253935884, + "grad_norm": 0.0005398907233029604, + "learning_rate": 0.07847105449292378, + "loss": 0.3459, + "num_input_tokens_seen": 55014032, + "step": 26330 + }, + { + "epoch": 4.296108981156701, + "grad_norm": 0.000576831866055727, + "learning_rate": 0.0784192838457113, + "loss": 0.289, + "num_input_tokens_seen": 55023472, + "step": 26335 + }, + { + "epoch": 4.296924708377518, + "grad_norm": 0.0006284540868364275, + "learning_rate": 0.07836752423714473, + "loss": 0.2898, + "num_input_tokens_seen": 55033664, + "step": 26340 + }, + { + "epoch": 4.297740435598336, + "grad_norm": 0.00029658203129656613, + "learning_rate": 0.07831577567520616, + "loss": 0.3591, + "num_input_tokens_seen": 55044480, + "step": 26345 + }, + { + "epoch": 4.298556162819153, + "grad_norm": 0.0004182902630418539, + "learning_rate": 0.07826403816787579, + "loss": 0.3499, + "num_input_tokens_seen": 55055472, + "step": 26350 + }, + { + "epoch": 4.2993718900399704, + "grad_norm": 0.0003520615864545107, + "learning_rate": 0.0782123117231322, + "loss": 0.3488, + "num_input_tokens_seen": 55064512, + "step": 26355 + }, + { + "epoch": 4.300187617260788, + "grad_norm": 0.0003985015500802547, + "learning_rate": 0.07816059634895237, + "loss": 0.3203, + "num_input_tokens_seen": 55075408, + "step": 26360 + }, + { + "epoch": 4.301003344481606, + "grad_norm": 0.0003649583668448031, + "learning_rate": 0.0781088920533113, + "loss": 0.3479, + "num_input_tokens_seen": 55085760, + "step": 26365 + }, + { + "epoch": 4.301819071702423, + "grad_norm": 0.001060200622305274, + "learning_rate": 0.07805719884418257, + "loss": 0.316, + "num_input_tokens_seen": 55095936, + "step": 26370 + }, + { + "epoch": 4.30263479892324, + "grad_norm": 0.0007021415513008833, + "learning_rate": 0.07800551672953779, + "loss": 0.3505, + "num_input_tokens_seen": 55106064, + "step": 26375 + }, + { + "epoch": 4.303450526144058, + "grad_norm": 0.0006962674087844789, + "learning_rate": 0.07795384571734709, + "loss": 0.3347, + "num_input_tokens_seen": 55118080, + "step": 26380 + }, + { + "epoch": 4.304266253364875, + "grad_norm": 0.0004144226259086281, + "learning_rate": 0.07790218581557883, + "loss": 0.3152, + "num_input_tokens_seen": 55128464, + "step": 26385 + }, + { + "epoch": 4.305081980585692, + "grad_norm": 0.0008601060253567994, + "learning_rate": 0.07785053703219949, + "loss": 0.3662, + "num_input_tokens_seen": 55137632, + "step": 26390 + }, + { + "epoch": 4.305897707806509, + "grad_norm": 0.0005052294000051916, + "learning_rate": 0.07779889937517409, + "loss": 0.3069, + "num_input_tokens_seen": 55148192, + "step": 26395 + }, + { + "epoch": 4.306713435027326, + "grad_norm": 0.0005424702540040016, + "learning_rate": 0.0777472728524657, + "loss": 0.3172, + "num_input_tokens_seen": 55158912, + "step": 26400 + }, + { + "epoch": 4.306713435027326, + "eval_loss": 0.31920480728149414, + "eval_runtime": 156.2931, + "eval_samples_per_second": 17.435, + "eval_steps_per_second": 8.721, + "num_input_tokens_seen": 55158912, + "step": 26400 + }, + { + "epoch": 4.307529162248144, + "grad_norm": 0.00038102376856841147, + "learning_rate": 0.07769565747203584, + "loss": 0.3335, + "num_input_tokens_seen": 55169312, + "step": 26405 + }, + { + "epoch": 4.3083448894689615, + "grad_norm": 0.0005863065598532557, + "learning_rate": 0.07764405324184427, + "loss": 0.3476, + "num_input_tokens_seen": 55180208, + "step": 26410 + }, + { + "epoch": 4.309160616689779, + "grad_norm": 0.0005972943617962301, + "learning_rate": 0.07759246016984889, + "loss": 0.256, + "num_input_tokens_seen": 55190016, + "step": 26415 + }, + { + "epoch": 4.309976343910597, + "grad_norm": 0.0005646183853968978, + "learning_rate": 0.07754087826400609, + "loss": 0.3327, + "num_input_tokens_seen": 55199984, + "step": 26420 + }, + { + "epoch": 4.310792071131414, + "grad_norm": 0.0006402673316188157, + "learning_rate": 0.0774893075322705, + "loss": 0.3403, + "num_input_tokens_seen": 55209968, + "step": 26425 + }, + { + "epoch": 4.311607798352231, + "grad_norm": 0.00044056991464458406, + "learning_rate": 0.07743774798259484, + "loss": 0.3116, + "num_input_tokens_seen": 55220064, + "step": 26430 + }, + { + "epoch": 4.312423525573048, + "grad_norm": 0.000376744574168697, + "learning_rate": 0.07738619962293032, + "loss": 0.3264, + "num_input_tokens_seen": 55231184, + "step": 26435 + }, + { + "epoch": 4.313239252793866, + "grad_norm": 0.0003292388282716274, + "learning_rate": 0.0773346624612264, + "loss": 0.3269, + "num_input_tokens_seen": 55241264, + "step": 26440 + }, + { + "epoch": 4.314054980014683, + "grad_norm": 0.0003096950822509825, + "learning_rate": 0.07728313650543066, + "loss": 0.3235, + "num_input_tokens_seen": 55252336, + "step": 26445 + }, + { + "epoch": 4.3148707072355, + "grad_norm": 0.0006906756316311657, + "learning_rate": 0.07723162176348913, + "loss": 0.3483, + "num_input_tokens_seen": 55262736, + "step": 26450 + }, + { + "epoch": 4.3156864344563175, + "grad_norm": 0.00048694832366891205, + "learning_rate": 0.07718011824334593, + "loss": 0.3256, + "num_input_tokens_seen": 55272512, + "step": 26455 + }, + { + "epoch": 4.3165021616771355, + "grad_norm": 0.0005158008425496519, + "learning_rate": 0.07712862595294363, + "loss": 0.3646, + "num_input_tokens_seen": 55283200, + "step": 26460 + }, + { + "epoch": 4.317317888897953, + "grad_norm": 0.0004375748394522816, + "learning_rate": 0.07707714490022301, + "loss": 0.2616, + "num_input_tokens_seen": 55293536, + "step": 26465 + }, + { + "epoch": 4.31813361611877, + "grad_norm": 0.0006056930287741125, + "learning_rate": 0.07702567509312298, + "loss": 0.3488, + "num_input_tokens_seen": 55303952, + "step": 26470 + }, + { + "epoch": 4.318949343339587, + "grad_norm": 0.0004539176297839731, + "learning_rate": 0.07697421653958098, + "loss": 0.3288, + "num_input_tokens_seen": 55315136, + "step": 26475 + }, + { + "epoch": 4.319765070560405, + "grad_norm": 0.0002988300402648747, + "learning_rate": 0.07692276924753247, + "loss": 0.3388, + "num_input_tokens_seen": 55325264, + "step": 26480 + }, + { + "epoch": 4.320580797781222, + "grad_norm": 0.0005114374216645956, + "learning_rate": 0.07687133322491124, + "loss": 0.333, + "num_input_tokens_seen": 55335040, + "step": 26485 + }, + { + "epoch": 4.321396525002039, + "grad_norm": 0.00045154881081543863, + "learning_rate": 0.07681990847964948, + "loss": 0.3086, + "num_input_tokens_seen": 55345648, + "step": 26490 + }, + { + "epoch": 4.322212252222856, + "grad_norm": 0.0004610335163306445, + "learning_rate": 0.0767684950196774, + "loss": 0.2683, + "num_input_tokens_seen": 55356592, + "step": 26495 + }, + { + "epoch": 4.323027979443674, + "grad_norm": 0.00037870105006732047, + "learning_rate": 0.0767170928529237, + "loss": 0.3057, + "num_input_tokens_seen": 55368272, + "step": 26500 + }, + { + "epoch": 4.3238437066644915, + "grad_norm": 0.0008536350796930492, + "learning_rate": 0.07666570198731526, + "loss": 0.3349, + "num_input_tokens_seen": 55379008, + "step": 26505 + }, + { + "epoch": 4.324659433885309, + "grad_norm": 0.0005280011100694537, + "learning_rate": 0.07661432243077708, + "loss": 0.3218, + "num_input_tokens_seen": 55389216, + "step": 26510 + }, + { + "epoch": 4.325475161106126, + "grad_norm": 0.00033133768010884523, + "learning_rate": 0.0765629541912326, + "loss": 0.315, + "num_input_tokens_seen": 55401024, + "step": 26515 + }, + { + "epoch": 4.326290888326944, + "grad_norm": 0.0003452693053986877, + "learning_rate": 0.07651159727660352, + "loss": 0.2576, + "num_input_tokens_seen": 55410432, + "step": 26520 + }, + { + "epoch": 4.327106615547761, + "grad_norm": 0.0006842846632935107, + "learning_rate": 0.07646025169480959, + "loss": 0.3363, + "num_input_tokens_seen": 55420480, + "step": 26525 + }, + { + "epoch": 4.327922342768578, + "grad_norm": 0.00046815641690045595, + "learning_rate": 0.07640891745376908, + "loss": 0.3017, + "num_input_tokens_seen": 55429792, + "step": 26530 + }, + { + "epoch": 4.328738069989395, + "grad_norm": 0.0002890178875532001, + "learning_rate": 0.07635759456139822, + "loss": 0.3426, + "num_input_tokens_seen": 55439600, + "step": 26535 + }, + { + "epoch": 4.329553797210213, + "grad_norm": 0.0006813342333771288, + "learning_rate": 0.0763062830256118, + "loss": 0.2715, + "num_input_tokens_seen": 55450320, + "step": 26540 + }, + { + "epoch": 4.33036952443103, + "grad_norm": 0.00045147043420001864, + "learning_rate": 0.07625498285432258, + "loss": 0.31, + "num_input_tokens_seen": 55462160, + "step": 26545 + }, + { + "epoch": 4.331185251651847, + "grad_norm": 0.0006826806929893792, + "learning_rate": 0.07620369405544176, + "loss": 0.321, + "num_input_tokens_seen": 55472288, + "step": 26550 + }, + { + "epoch": 4.332000978872665, + "grad_norm": 0.0004876784805674106, + "learning_rate": 0.07615241663687868, + "loss": 0.3357, + "num_input_tokens_seen": 55482848, + "step": 26555 + }, + { + "epoch": 4.332816706093483, + "grad_norm": 0.0004954770556651056, + "learning_rate": 0.07610115060654106, + "loss": 0.3291, + "num_input_tokens_seen": 55492480, + "step": 26560 + }, + { + "epoch": 4.3336324333143, + "grad_norm": 0.00035619118716567755, + "learning_rate": 0.07604989597233458, + "loss": 0.3347, + "num_input_tokens_seen": 55503328, + "step": 26565 + }, + { + "epoch": 4.334448160535117, + "grad_norm": 0.0004227271710988134, + "learning_rate": 0.07599865274216352, + "loss": 0.3187, + "num_input_tokens_seen": 55513392, + "step": 26570 + }, + { + "epoch": 4.335263887755934, + "grad_norm": 0.0002806941920425743, + "learning_rate": 0.07594742092393013, + "loss": 0.295, + "num_input_tokens_seen": 55523808, + "step": 26575 + }, + { + "epoch": 4.336079614976752, + "grad_norm": 0.00042092552757821977, + "learning_rate": 0.07589620052553503, + "loss": 0.3294, + "num_input_tokens_seen": 55534256, + "step": 26580 + }, + { + "epoch": 4.336895342197569, + "grad_norm": 0.000675124756526202, + "learning_rate": 0.0758449915548771, + "loss": 0.2844, + "num_input_tokens_seen": 55545360, + "step": 26585 + }, + { + "epoch": 4.337711069418386, + "grad_norm": 0.00042865786235779524, + "learning_rate": 0.07579379401985332, + "loss": 0.3441, + "num_input_tokens_seen": 55554896, + "step": 26590 + }, + { + "epoch": 4.338526796639204, + "grad_norm": 0.0005330693675205112, + "learning_rate": 0.07574260792835905, + "loss": 0.3926, + "num_input_tokens_seen": 55565904, + "step": 26595 + }, + { + "epoch": 4.339342523860021, + "grad_norm": 0.0009888058993965387, + "learning_rate": 0.07569143328828784, + "loss": 0.3416, + "num_input_tokens_seen": 55575232, + "step": 26600 + }, + { + "epoch": 4.339342523860021, + "eval_loss": 0.31923818588256836, + "eval_runtime": 156.1912, + "eval_samples_per_second": 17.447, + "eval_steps_per_second": 8.726, + "num_input_tokens_seen": 55575232, + "step": 26600 + }, + { + "epoch": 4.3401582510808385, + "grad_norm": 0.00039836057112552226, + "learning_rate": 0.0756402701075314, + "loss": 0.3797, + "num_input_tokens_seen": 55586864, + "step": 26605 + }, + { + "epoch": 4.340973978301656, + "grad_norm": 0.0004619379178620875, + "learning_rate": 0.07558911839397982, + "loss": 0.3306, + "num_input_tokens_seen": 55597040, + "step": 26610 + }, + { + "epoch": 4.341789705522474, + "grad_norm": 0.00028013443807139993, + "learning_rate": 0.07553797815552123, + "loss": 0.3274, + "num_input_tokens_seen": 55607040, + "step": 26615 + }, + { + "epoch": 4.342605432743291, + "grad_norm": 0.000625727407168597, + "learning_rate": 0.07548684940004222, + "loss": 0.2972, + "num_input_tokens_seen": 55617872, + "step": 26620 + }, + { + "epoch": 4.343421159964108, + "grad_norm": 0.0006775731453672051, + "learning_rate": 0.07543573213542744, + "loss": 0.2738, + "num_input_tokens_seen": 55629680, + "step": 26625 + }, + { + "epoch": 4.344236887184925, + "grad_norm": 0.0007525925757363439, + "learning_rate": 0.0753846263695597, + "loss": 0.3247, + "num_input_tokens_seen": 55641296, + "step": 26630 + }, + { + "epoch": 4.345052614405743, + "grad_norm": 0.0005382676608860493, + "learning_rate": 0.07533353211032029, + "loss": 0.3259, + "num_input_tokens_seen": 55652576, + "step": 26635 + }, + { + "epoch": 4.34586834162656, + "grad_norm": 0.0007069528219290078, + "learning_rate": 0.07528244936558857, + "loss": 0.3676, + "num_input_tokens_seen": 55662992, + "step": 26640 + }, + { + "epoch": 4.346684068847377, + "grad_norm": 0.0007515751058235765, + "learning_rate": 0.07523137814324206, + "loss": 0.3155, + "num_input_tokens_seen": 55673120, + "step": 26645 + }, + { + "epoch": 4.3474997960681945, + "grad_norm": 0.0011327568208798766, + "learning_rate": 0.07518031845115672, + "loss": 0.2692, + "num_input_tokens_seen": 55683552, + "step": 26650 + }, + { + "epoch": 4.3483155232890125, + "grad_norm": 0.0003292927867732942, + "learning_rate": 0.07512927029720647, + "loss": 0.3571, + "num_input_tokens_seen": 55695008, + "step": 26655 + }, + { + "epoch": 4.34913125050983, + "grad_norm": 0.00047828946844674647, + "learning_rate": 0.0750782336892636, + "loss": 0.2914, + "num_input_tokens_seen": 55705360, + "step": 26660 + }, + { + "epoch": 4.349946977730647, + "grad_norm": 0.00036303215892985463, + "learning_rate": 0.0750272086351987, + "loss": 0.3272, + "num_input_tokens_seen": 55715680, + "step": 26665 + }, + { + "epoch": 4.350762704951464, + "grad_norm": 0.0006069666123948991, + "learning_rate": 0.07497619514288031, + "loss": 0.303, + "num_input_tokens_seen": 55726368, + "step": 26670 + }, + { + "epoch": 4.351578432172282, + "grad_norm": 0.0004939312348142266, + "learning_rate": 0.07492519322017545, + "loss": 0.3733, + "num_input_tokens_seen": 55736416, + "step": 26675 + }, + { + "epoch": 4.352394159393099, + "grad_norm": 0.0006996985175646842, + "learning_rate": 0.0748742028749493, + "loss": 0.338, + "num_input_tokens_seen": 55747552, + "step": 26680 + }, + { + "epoch": 4.353209886613916, + "grad_norm": 0.001002976205199957, + "learning_rate": 0.0748232241150651, + "loss": 0.3315, + "num_input_tokens_seen": 55758752, + "step": 26685 + }, + { + "epoch": 4.354025613834733, + "grad_norm": 0.0003762083943001926, + "learning_rate": 0.07477225694838453, + "loss": 0.347, + "num_input_tokens_seen": 55769520, + "step": 26690 + }, + { + "epoch": 4.354841341055551, + "grad_norm": 0.000458630733191967, + "learning_rate": 0.07472130138276731, + "loss": 0.3803, + "num_input_tokens_seen": 55779296, + "step": 26695 + }, + { + "epoch": 4.3556570682763684, + "grad_norm": 0.0009381452691741288, + "learning_rate": 0.07467035742607138, + "loss": 0.3311, + "num_input_tokens_seen": 55790112, + "step": 26700 + }, + { + "epoch": 4.356472795497186, + "grad_norm": 0.0005140206776559353, + "learning_rate": 0.07461942508615303, + "loss": 0.2509, + "num_input_tokens_seen": 55800848, + "step": 26705 + }, + { + "epoch": 4.357288522718003, + "grad_norm": 0.0007862396887503564, + "learning_rate": 0.07456850437086657, + "loss": 0.2884, + "num_input_tokens_seen": 55810736, + "step": 26710 + }, + { + "epoch": 4.358104249938821, + "grad_norm": 0.0005944161093793809, + "learning_rate": 0.07451759528806468, + "loss": 0.3519, + "num_input_tokens_seen": 55821168, + "step": 26715 + }, + { + "epoch": 4.358919977159638, + "grad_norm": 0.0007219486869871616, + "learning_rate": 0.0744666978455982, + "loss": 0.2984, + "num_input_tokens_seen": 55831744, + "step": 26720 + }, + { + "epoch": 4.359735704380455, + "grad_norm": 0.000676034833304584, + "learning_rate": 0.07441581205131609, + "loss": 0.3214, + "num_input_tokens_seen": 55840800, + "step": 26725 + }, + { + "epoch": 4.360551431601272, + "grad_norm": 0.0005878261872567236, + "learning_rate": 0.07436493791306566, + "loss": 0.3263, + "num_input_tokens_seen": 55851472, + "step": 26730 + }, + { + "epoch": 4.36136715882209, + "grad_norm": 0.0003624252276495099, + "learning_rate": 0.07431407543869223, + "loss": 0.2984, + "num_input_tokens_seen": 55861216, + "step": 26735 + }, + { + "epoch": 4.362182886042907, + "grad_norm": 0.0007726629846729338, + "learning_rate": 0.0742632246360395, + "loss": 0.3264, + "num_input_tokens_seen": 55872448, + "step": 26740 + }, + { + "epoch": 4.362998613263724, + "grad_norm": 0.0006119739264249802, + "learning_rate": 0.07421238551294934, + "loss": 0.3563, + "num_input_tokens_seen": 55883360, + "step": 26745 + }, + { + "epoch": 4.3638143404845415, + "grad_norm": 0.0002689459070097655, + "learning_rate": 0.07416155807726171, + "loss": 0.3161, + "num_input_tokens_seen": 55892480, + "step": 26750 + }, + { + "epoch": 4.3646300677053596, + "grad_norm": 0.0006831246428191662, + "learning_rate": 0.07411074233681492, + "loss": 0.2853, + "num_input_tokens_seen": 55903360, + "step": 26755 + }, + { + "epoch": 4.365445794926177, + "grad_norm": 0.0004258951230440289, + "learning_rate": 0.07405993829944528, + "loss": 0.3835, + "num_input_tokens_seen": 55912720, + "step": 26760 + }, + { + "epoch": 4.366261522146994, + "grad_norm": 0.00045989122008904815, + "learning_rate": 0.07400914597298755, + "loss": 0.3472, + "num_input_tokens_seen": 55922912, + "step": 26765 + }, + { + "epoch": 4.367077249367812, + "grad_norm": 0.0005358275957405567, + "learning_rate": 0.07395836536527445, + "loss": 0.3053, + "num_input_tokens_seen": 55933952, + "step": 26770 + }, + { + "epoch": 4.367892976588629, + "grad_norm": 0.00043204647954553366, + "learning_rate": 0.07390759648413696, + "loss": 0.3231, + "num_input_tokens_seen": 55943728, + "step": 26775 + }, + { + "epoch": 4.368708703809446, + "grad_norm": 0.00041678702109493315, + "learning_rate": 0.07385683933740435, + "loss": 0.2942, + "num_input_tokens_seen": 55955536, + "step": 26780 + }, + { + "epoch": 4.369524431030263, + "grad_norm": 0.0006367117748595774, + "learning_rate": 0.07380609393290402, + "loss": 0.3447, + "num_input_tokens_seen": 55966016, + "step": 26785 + }, + { + "epoch": 4.370340158251081, + "grad_norm": 0.0005737565224990249, + "learning_rate": 0.07375536027846147, + "loss": 0.2728, + "num_input_tokens_seen": 55975376, + "step": 26790 + }, + { + "epoch": 4.371155885471898, + "grad_norm": 0.0010552378371357918, + "learning_rate": 0.07370463838190057, + "loss": 0.2983, + "num_input_tokens_seen": 55984848, + "step": 26795 + }, + { + "epoch": 4.3719716126927155, + "grad_norm": 0.0008431118912994862, + "learning_rate": 0.07365392825104317, + "loss": 0.4005, + "num_input_tokens_seen": 55994160, + "step": 26800 + }, + { + "epoch": 4.3719716126927155, + "eval_loss": 0.3154870569705963, + "eval_runtime": 156.2497, + "eval_samples_per_second": 17.44, + "eval_steps_per_second": 8.723, + "num_input_tokens_seen": 55994160, + "step": 26800 + }, + { + "epoch": 4.372787339913533, + "grad_norm": 0.0003280571545474231, + "learning_rate": 0.07360322989370945, + "loss": 0.2971, + "num_input_tokens_seen": 56004288, + "step": 26805 + }, + { + "epoch": 4.373603067134351, + "grad_norm": 0.00033022035495378077, + "learning_rate": 0.07355254331771781, + "loss": 0.3256, + "num_input_tokens_seen": 56015504, + "step": 26810 + }, + { + "epoch": 4.374418794355168, + "grad_norm": 0.0003301891265437007, + "learning_rate": 0.07350186853088461, + "loss": 0.2937, + "num_input_tokens_seen": 56025184, + "step": 26815 + }, + { + "epoch": 4.375234521575985, + "grad_norm": 0.0006288377335295081, + "learning_rate": 0.07345120554102462, + "loss": 0.286, + "num_input_tokens_seen": 56036560, + "step": 26820 + }, + { + "epoch": 4.376050248796802, + "grad_norm": 0.0004100099904462695, + "learning_rate": 0.07340055435595079, + "loss": 0.3428, + "num_input_tokens_seen": 56046208, + "step": 26825 + }, + { + "epoch": 4.37686597601762, + "grad_norm": 0.0004921951913274825, + "learning_rate": 0.07334991498347401, + "loss": 0.3212, + "num_input_tokens_seen": 56056560, + "step": 26830 + }, + { + "epoch": 4.377681703238437, + "grad_norm": 0.0004551723541226238, + "learning_rate": 0.07329928743140365, + "loss": 0.3194, + "num_input_tokens_seen": 56067680, + "step": 26835 + }, + { + "epoch": 4.378497430459254, + "grad_norm": 0.0010634601349011064, + "learning_rate": 0.07324867170754705, + "loss": 0.3491, + "num_input_tokens_seen": 56077904, + "step": 26840 + }, + { + "epoch": 4.3793131576800715, + "grad_norm": 0.0007919012568891048, + "learning_rate": 0.07319806781970974, + "loss": 0.2726, + "num_input_tokens_seen": 56089232, + "step": 26845 + }, + { + "epoch": 4.3801288849008895, + "grad_norm": 0.000733269436750561, + "learning_rate": 0.07314747577569555, + "loss": 0.4108, + "num_input_tokens_seen": 56100208, + "step": 26850 + }, + { + "epoch": 4.380944612121707, + "grad_norm": 0.00027575562126003206, + "learning_rate": 0.07309689558330636, + "loss": 0.2739, + "num_input_tokens_seen": 56109968, + "step": 26855 + }, + { + "epoch": 4.381760339342524, + "grad_norm": 0.0003694015322253108, + "learning_rate": 0.0730463272503423, + "loss": 0.3009, + "num_input_tokens_seen": 56120176, + "step": 26860 + }, + { + "epoch": 4.382576066563341, + "grad_norm": 0.00033308041747659445, + "learning_rate": 0.07299577078460168, + "loss": 0.3221, + "num_input_tokens_seen": 56131232, + "step": 26865 + }, + { + "epoch": 4.383391793784159, + "grad_norm": 0.0005680904723703861, + "learning_rate": 0.07294522619388083, + "loss": 0.3793, + "num_input_tokens_seen": 56141776, + "step": 26870 + }, + { + "epoch": 4.384207521004976, + "grad_norm": 0.0009030813234858215, + "learning_rate": 0.07289469348597452, + "loss": 0.264, + "num_input_tokens_seen": 56151232, + "step": 26875 + }, + { + "epoch": 4.385023248225793, + "grad_norm": 0.0005359470378607512, + "learning_rate": 0.07284417266867535, + "loss": 0.3208, + "num_input_tokens_seen": 56161152, + "step": 26880 + }, + { + "epoch": 4.38583897544661, + "grad_norm": 0.0007189130992628634, + "learning_rate": 0.07279366374977439, + "loss": 0.3701, + "num_input_tokens_seen": 56172384, + "step": 26885 + }, + { + "epoch": 4.386654702667428, + "grad_norm": 0.0009487156639806926, + "learning_rate": 0.07274316673706074, + "loss": 0.3236, + "num_input_tokens_seen": 56182960, + "step": 26890 + }, + { + "epoch": 4.387470429888245, + "grad_norm": 0.00036773073952645063, + "learning_rate": 0.07269268163832161, + "loss": 0.3198, + "num_input_tokens_seen": 56194512, + "step": 26895 + }, + { + "epoch": 4.388286157109063, + "grad_norm": 0.0005177340353839099, + "learning_rate": 0.07264220846134248, + "loss": 0.3234, + "num_input_tokens_seen": 56204816, + "step": 26900 + }, + { + "epoch": 4.38910188432988, + "grad_norm": 0.0006236720946617424, + "learning_rate": 0.07259174721390699, + "loss": 0.3116, + "num_input_tokens_seen": 56216240, + "step": 26905 + }, + { + "epoch": 4.389917611550698, + "grad_norm": 0.0005226426292210817, + "learning_rate": 0.07254129790379686, + "loss": 0.2787, + "num_input_tokens_seen": 56226304, + "step": 26910 + }, + { + "epoch": 4.390733338771515, + "grad_norm": 0.00038778691668994725, + "learning_rate": 0.072490860538792, + "loss": 0.3304, + "num_input_tokens_seen": 56237264, + "step": 26915 + }, + { + "epoch": 4.391549065992332, + "grad_norm": 0.001294393907301128, + "learning_rate": 0.07244043512667042, + "loss": 0.3356, + "num_input_tokens_seen": 56246848, + "step": 26920 + }, + { + "epoch": 4.392364793213149, + "grad_norm": 0.00040199514478445053, + "learning_rate": 0.07239002167520843, + "loss": 0.316, + "num_input_tokens_seen": 56256800, + "step": 26925 + }, + { + "epoch": 4.393180520433967, + "grad_norm": 0.000766476325225085, + "learning_rate": 0.07233962019218045, + "loss": 0.3315, + "num_input_tokens_seen": 56266320, + "step": 26930 + }, + { + "epoch": 4.393996247654784, + "grad_norm": 0.0005060647963546216, + "learning_rate": 0.07228923068535892, + "loss": 0.2646, + "num_input_tokens_seen": 56276896, + "step": 26935 + }, + { + "epoch": 4.394811974875601, + "grad_norm": 0.0010777259012684226, + "learning_rate": 0.0722388531625146, + "loss": 0.3436, + "num_input_tokens_seen": 56287424, + "step": 26940 + }, + { + "epoch": 4.395627702096419, + "grad_norm": 0.0006616482860408723, + "learning_rate": 0.07218848763141639, + "loss": 0.3544, + "num_input_tokens_seen": 56298624, + "step": 26945 + }, + { + "epoch": 4.3964434293172365, + "grad_norm": 0.00033871422056108713, + "learning_rate": 0.07213813409983118, + "loss": 0.3508, + "num_input_tokens_seen": 56307536, + "step": 26950 + }, + { + "epoch": 4.397259156538054, + "grad_norm": 0.0006609113770537078, + "learning_rate": 0.0720877925755242, + "loss": 0.3664, + "num_input_tokens_seen": 56317504, + "step": 26955 + }, + { + "epoch": 4.398074883758871, + "grad_norm": 0.0006575039587914944, + "learning_rate": 0.07203746306625866, + "loss": 0.2827, + "num_input_tokens_seen": 56328816, + "step": 26960 + }, + { + "epoch": 4.398890610979688, + "grad_norm": 0.0011975514935329556, + "learning_rate": 0.07198714557979606, + "loss": 0.3527, + "num_input_tokens_seen": 56339744, + "step": 26965 + }, + { + "epoch": 4.399706338200506, + "grad_norm": 0.000996650429442525, + "learning_rate": 0.07193684012389602, + "loss": 0.3641, + "num_input_tokens_seen": 56348320, + "step": 26970 + }, + { + "epoch": 4.400522065421323, + "grad_norm": 0.00033843080746009946, + "learning_rate": 0.07188654670631621, + "loss": 0.2777, + "num_input_tokens_seen": 56358512, + "step": 26975 + }, + { + "epoch": 4.40133779264214, + "grad_norm": 0.00028953145374543965, + "learning_rate": 0.07183626533481258, + "loss": 0.2797, + "num_input_tokens_seen": 56369104, + "step": 26980 + }, + { + "epoch": 4.402153519862958, + "grad_norm": 0.0003358162066433579, + "learning_rate": 0.07178599601713909, + "loss": 0.2913, + "num_input_tokens_seen": 56379824, + "step": 26985 + }, + { + "epoch": 4.402969247083775, + "grad_norm": 0.0005156345432624221, + "learning_rate": 0.07173573876104786, + "loss": 0.3188, + "num_input_tokens_seen": 56389936, + "step": 26990 + }, + { + "epoch": 4.4037849743045925, + "grad_norm": 0.0005266615771688521, + "learning_rate": 0.0716854935742893, + "loss": 0.3302, + "num_input_tokens_seen": 56400336, + "step": 26995 + }, + { + "epoch": 4.40460070152541, + "grad_norm": 0.00044112777686677873, + "learning_rate": 0.07163526046461174, + "loss": 0.3505, + "num_input_tokens_seen": 56410736, + "step": 27000 + }, + { + "epoch": 4.40460070152541, + "eval_loss": 0.3154727518558502, + "eval_runtime": 156.1627, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 56410736, + "step": 27000 + }, + { + "epoch": 4.405416428746228, + "grad_norm": 0.0004575115453917533, + "learning_rate": 0.07158503943976181, + "loss": 0.3264, + "num_input_tokens_seen": 56421248, + "step": 27005 + }, + { + "epoch": 4.406232155967045, + "grad_norm": 0.0006090269307605922, + "learning_rate": 0.07153483050748427, + "loss": 0.3089, + "num_input_tokens_seen": 56432560, + "step": 27010 + }, + { + "epoch": 4.407047883187862, + "grad_norm": 0.00036630156682804227, + "learning_rate": 0.07148463367552188, + "loss": 0.3162, + "num_input_tokens_seen": 56444432, + "step": 27015 + }, + { + "epoch": 4.407863610408679, + "grad_norm": 0.0006240297225303948, + "learning_rate": 0.07143444895161565, + "loss": 0.3019, + "num_input_tokens_seen": 56455584, + "step": 27020 + }, + { + "epoch": 4.408679337629497, + "grad_norm": 0.0004837456508539617, + "learning_rate": 0.07138427634350476, + "loss": 0.3044, + "num_input_tokens_seen": 56465632, + "step": 27025 + }, + { + "epoch": 4.409495064850314, + "grad_norm": 0.0004074128228239715, + "learning_rate": 0.07133411585892636, + "loss": 0.3012, + "num_input_tokens_seen": 56475120, + "step": 27030 + }, + { + "epoch": 4.410310792071131, + "grad_norm": 0.00034893726115114987, + "learning_rate": 0.07128396750561593, + "loss": 0.3186, + "num_input_tokens_seen": 56486160, + "step": 27035 + }, + { + "epoch": 4.411126519291948, + "grad_norm": 0.000507241755258292, + "learning_rate": 0.07123383129130685, + "loss": 0.2634, + "num_input_tokens_seen": 56495968, + "step": 27040 + }, + { + "epoch": 4.4119422465127665, + "grad_norm": 0.0003136048326268792, + "learning_rate": 0.07118370722373084, + "loss": 0.2935, + "num_input_tokens_seen": 56506464, + "step": 27045 + }, + { + "epoch": 4.412757973733584, + "grad_norm": 0.0004268854681868106, + "learning_rate": 0.07113359531061769, + "loss": 0.286, + "num_input_tokens_seen": 56515680, + "step": 27050 + }, + { + "epoch": 4.413573700954401, + "grad_norm": 0.0007975343614816666, + "learning_rate": 0.07108349555969525, + "loss": 0.3244, + "num_input_tokens_seen": 56526528, + "step": 27055 + }, + { + "epoch": 4.414389428175218, + "grad_norm": 0.00040600125794298947, + "learning_rate": 0.07103340797868944, + "loss": 0.3094, + "num_input_tokens_seen": 56536720, + "step": 27060 + }, + { + "epoch": 4.415205155396036, + "grad_norm": 0.0003776569210458547, + "learning_rate": 0.07098333257532453, + "loss": 0.2539, + "num_input_tokens_seen": 56546432, + "step": 27065 + }, + { + "epoch": 4.416020882616853, + "grad_norm": 0.001082566799595952, + "learning_rate": 0.07093326935732269, + "loss": 0.3235, + "num_input_tokens_seen": 56557248, + "step": 27070 + }, + { + "epoch": 4.41683660983767, + "grad_norm": 0.0006616957834921777, + "learning_rate": 0.0708832183324044, + "loss": 0.3075, + "num_input_tokens_seen": 56568784, + "step": 27075 + }, + { + "epoch": 4.417652337058487, + "grad_norm": 0.00036411636392585933, + "learning_rate": 0.07083317950828799, + "loss": 0.2985, + "num_input_tokens_seen": 56580256, + "step": 27080 + }, + { + "epoch": 4.418468064279305, + "grad_norm": 0.0003103472990915179, + "learning_rate": 0.0707831528926902, + "loss": 0.3151, + "num_input_tokens_seen": 56590144, + "step": 27085 + }, + { + "epoch": 4.419283791500122, + "grad_norm": 0.0008564237505197525, + "learning_rate": 0.07073313849332578, + "loss": 0.3005, + "num_input_tokens_seen": 56601408, + "step": 27090 + }, + { + "epoch": 4.4200995187209395, + "grad_norm": 0.0006915461272001266, + "learning_rate": 0.07068313631790749, + "loss": 0.3026, + "num_input_tokens_seen": 56611200, + "step": 27095 + }, + { + "epoch": 4.420915245941757, + "grad_norm": 0.00033487408654764295, + "learning_rate": 0.07063314637414632, + "loss": 0.2908, + "num_input_tokens_seen": 56622416, + "step": 27100 + }, + { + "epoch": 4.421730973162575, + "grad_norm": 0.0006882957532070577, + "learning_rate": 0.07058316866975144, + "loss": 0.3028, + "num_input_tokens_seen": 56633504, + "step": 27105 + }, + { + "epoch": 4.422546700383392, + "grad_norm": 0.0004252373764757067, + "learning_rate": 0.0705332032124299, + "loss": 0.2755, + "num_input_tokens_seen": 56645328, + "step": 27110 + }, + { + "epoch": 4.423362427604209, + "grad_norm": 0.0006513758562505245, + "learning_rate": 0.0704832500098871, + "loss": 0.3568, + "num_input_tokens_seen": 56655616, + "step": 27115 + }, + { + "epoch": 4.424178154825027, + "grad_norm": 0.0005090294871479273, + "learning_rate": 0.07043330906982641, + "loss": 0.2653, + "num_input_tokens_seen": 56666256, + "step": 27120 + }, + { + "epoch": 4.424993882045844, + "grad_norm": 0.0002431945176795125, + "learning_rate": 0.07038338039994936, + "loss": 0.2679, + "num_input_tokens_seen": 56676848, + "step": 27125 + }, + { + "epoch": 4.425809609266661, + "grad_norm": 0.00046673775068484247, + "learning_rate": 0.07033346400795562, + "loss": 0.3675, + "num_input_tokens_seen": 56687472, + "step": 27130 + }, + { + "epoch": 4.426625336487478, + "grad_norm": 0.0004875586600974202, + "learning_rate": 0.07028355990154282, + "loss": 0.3257, + "num_input_tokens_seen": 56697760, + "step": 27135 + }, + { + "epoch": 4.4274410637082955, + "grad_norm": 0.0006831102655269206, + "learning_rate": 0.07023366808840685, + "loss": 0.3403, + "num_input_tokens_seen": 56708736, + "step": 27140 + }, + { + "epoch": 4.4282567909291135, + "grad_norm": 0.0009103293996304274, + "learning_rate": 0.07018378857624172, + "loss": 0.3563, + "num_input_tokens_seen": 56719504, + "step": 27145 + }, + { + "epoch": 4.429072518149931, + "grad_norm": 0.0008847378776408732, + "learning_rate": 0.0701339213727394, + "loss": 0.2569, + "num_input_tokens_seen": 56728608, + "step": 27150 + }, + { + "epoch": 4.429888245370748, + "grad_norm": 0.0004659552942030132, + "learning_rate": 0.07008406648559008, + "loss": 0.3387, + "num_input_tokens_seen": 56738656, + "step": 27155 + }, + { + "epoch": 4.430703972591566, + "grad_norm": 0.0004258238186594099, + "learning_rate": 0.07003422392248196, + "loss": 0.2818, + "num_input_tokens_seen": 56749344, + "step": 27160 + }, + { + "epoch": 4.431519699812383, + "grad_norm": 0.0006289277225732803, + "learning_rate": 0.06998439369110142, + "loss": 0.2529, + "num_input_tokens_seen": 56759008, + "step": 27165 + }, + { + "epoch": 4.4323354270332, + "grad_norm": 0.0004332080716267228, + "learning_rate": 0.06993457579913295, + "loss": 0.2532, + "num_input_tokens_seen": 56769616, + "step": 27170 + }, + { + "epoch": 4.433151154254017, + "grad_norm": 0.0005278109456412494, + "learning_rate": 0.06988477025425903, + "loss": 0.4152, + "num_input_tokens_seen": 56781232, + "step": 27175 + }, + { + "epoch": 4.433966881474835, + "grad_norm": 0.0005894998903386295, + "learning_rate": 0.06983497706416032, + "loss": 0.3268, + "num_input_tokens_seen": 56793568, + "step": 27180 + }, + { + "epoch": 4.434782608695652, + "grad_norm": 0.0004480787902139127, + "learning_rate": 0.0697851962365156, + "loss": 0.3089, + "num_input_tokens_seen": 56805280, + "step": 27185 + }, + { + "epoch": 4.4355983359164695, + "grad_norm": 0.0008993821102194488, + "learning_rate": 0.06973542777900163, + "loss": 0.3145, + "num_input_tokens_seen": 56816720, + "step": 27190 + }, + { + "epoch": 4.436414063137287, + "grad_norm": 0.0003979617904406041, + "learning_rate": 0.06968567169929342, + "loss": 0.2661, + "num_input_tokens_seen": 56828112, + "step": 27195 + }, + { + "epoch": 4.437229790358105, + "grad_norm": 0.0006597454776056111, + "learning_rate": 0.06963592800506392, + "loss": 0.3187, + "num_input_tokens_seen": 56838864, + "step": 27200 + }, + { + "epoch": 4.437229790358105, + "eval_loss": 0.31747475266456604, + "eval_runtime": 156.0273, + "eval_samples_per_second": 17.465, + "eval_steps_per_second": 8.736, + "num_input_tokens_seen": 56838864, + "step": 27200 + }, + { + "epoch": 4.438045517578922, + "grad_norm": 0.0009596596355549991, + "learning_rate": 0.06958619670398417, + "loss": 0.3581, + "num_input_tokens_seen": 56848512, + "step": 27205 + }, + { + "epoch": 4.438861244799739, + "grad_norm": 0.0004452296707313508, + "learning_rate": 0.0695364778037235, + "loss": 0.347, + "num_input_tokens_seen": 56857600, + "step": 27210 + }, + { + "epoch": 4.439676972020556, + "grad_norm": 0.0005379258655011654, + "learning_rate": 0.06948677131194907, + "loss": 0.2933, + "num_input_tokens_seen": 56867824, + "step": 27215 + }, + { + "epoch": 4.440492699241374, + "grad_norm": 0.0004101393569726497, + "learning_rate": 0.06943707723632629, + "loss": 0.3136, + "num_input_tokens_seen": 56877632, + "step": 27220 + }, + { + "epoch": 4.441308426462191, + "grad_norm": 0.0006874561659060419, + "learning_rate": 0.06938739558451867, + "loss": 0.3584, + "num_input_tokens_seen": 56889136, + "step": 27225 + }, + { + "epoch": 4.442124153683008, + "grad_norm": 0.00034378969576209784, + "learning_rate": 0.06933772636418763, + "loss": 0.3188, + "num_input_tokens_seen": 56900160, + "step": 27230 + }, + { + "epoch": 4.442939880903825, + "grad_norm": 0.0007102651288732886, + "learning_rate": 0.06928806958299293, + "loss": 0.217, + "num_input_tokens_seen": 56910752, + "step": 27235 + }, + { + "epoch": 4.443755608124643, + "grad_norm": 0.0005568863707594573, + "learning_rate": 0.06923842524859211, + "loss": 0.3545, + "num_input_tokens_seen": 56920096, + "step": 27240 + }, + { + "epoch": 4.444571335345461, + "grad_norm": 0.0006302451365627348, + "learning_rate": 0.06918879336864105, + "loss": 0.3018, + "num_input_tokens_seen": 56929808, + "step": 27245 + }, + { + "epoch": 4.445387062566278, + "grad_norm": 0.001214398886077106, + "learning_rate": 0.06913917395079362, + "loss": 0.3334, + "num_input_tokens_seen": 56940016, + "step": 27250 + }, + { + "epoch": 4.446202789787095, + "grad_norm": 0.001261401572264731, + "learning_rate": 0.0690895670027017, + "loss": 0.3122, + "num_input_tokens_seen": 56950176, + "step": 27255 + }, + { + "epoch": 4.447018517007913, + "grad_norm": 0.0006583509966731071, + "learning_rate": 0.06903997253201531, + "loss": 0.333, + "num_input_tokens_seen": 56960656, + "step": 27260 + }, + { + "epoch": 4.44783424422873, + "grad_norm": 0.0009152496932074428, + "learning_rate": 0.06899039054638263, + "loss": 0.352, + "num_input_tokens_seen": 56970800, + "step": 27265 + }, + { + "epoch": 4.448649971449547, + "grad_norm": 0.00048479490214958787, + "learning_rate": 0.06894082105344976, + "loss": 0.2888, + "num_input_tokens_seen": 56981984, + "step": 27270 + }, + { + "epoch": 4.449465698670364, + "grad_norm": 0.0005060622934252024, + "learning_rate": 0.06889126406086087, + "loss": 0.3297, + "num_input_tokens_seen": 56991312, + "step": 27275 + }, + { + "epoch": 4.450281425891182, + "grad_norm": 0.00035980981192551553, + "learning_rate": 0.0688417195762584, + "loss": 0.3192, + "num_input_tokens_seen": 57002032, + "step": 27280 + }, + { + "epoch": 4.451097153111999, + "grad_norm": 0.0008928682073019445, + "learning_rate": 0.06879218760728262, + "loss": 0.3239, + "num_input_tokens_seen": 57013296, + "step": 27285 + }, + { + "epoch": 4.4519128803328165, + "grad_norm": 0.0006566328229382634, + "learning_rate": 0.06874266816157207, + "loss": 0.3183, + "num_input_tokens_seen": 57023104, + "step": 27290 + }, + { + "epoch": 4.4527286075536345, + "grad_norm": 0.000489449012093246, + "learning_rate": 0.06869316124676321, + "loss": 0.3324, + "num_input_tokens_seen": 57033696, + "step": 27295 + }, + { + "epoch": 4.453544334774452, + "grad_norm": 0.00041221873834729195, + "learning_rate": 0.06864366687049062, + "loss": 0.2742, + "num_input_tokens_seen": 57044320, + "step": 27300 + }, + { + "epoch": 4.454360061995269, + "grad_norm": 0.0005007268628105521, + "learning_rate": 0.06859418504038704, + "loss": 0.2644, + "num_input_tokens_seen": 57053712, + "step": 27305 + }, + { + "epoch": 4.455175789216086, + "grad_norm": 0.0003493914846330881, + "learning_rate": 0.06854471576408311, + "loss": 0.329, + "num_input_tokens_seen": 57063376, + "step": 27310 + }, + { + "epoch": 4.455991516436903, + "grad_norm": 0.0004425600345712155, + "learning_rate": 0.06849525904920767, + "loss": 0.3001, + "num_input_tokens_seen": 57074512, + "step": 27315 + }, + { + "epoch": 4.456807243657721, + "grad_norm": 0.00047024470404721797, + "learning_rate": 0.06844581490338748, + "loss": 0.3066, + "num_input_tokens_seen": 57083952, + "step": 27320 + }, + { + "epoch": 4.457622970878538, + "grad_norm": 0.0005862897960469127, + "learning_rate": 0.06839638333424752, + "loss": 0.2735, + "num_input_tokens_seen": 57094256, + "step": 27325 + }, + { + "epoch": 4.458438698099355, + "grad_norm": 0.0006819262052886188, + "learning_rate": 0.06834696434941082, + "loss": 0.3048, + "num_input_tokens_seen": 57106000, + "step": 27330 + }, + { + "epoch": 4.459254425320173, + "grad_norm": 0.00027963140746578574, + "learning_rate": 0.06829755795649824, + "loss": 0.2728, + "num_input_tokens_seen": 57115504, + "step": 27335 + }, + { + "epoch": 4.4600701525409905, + "grad_norm": 0.00046454189578071237, + "learning_rate": 0.06824816416312904, + "loss": 0.2388, + "num_input_tokens_seen": 57125808, + "step": 27340 + }, + { + "epoch": 4.460885879761808, + "grad_norm": 0.0005979996640235186, + "learning_rate": 0.06819878297692027, + "loss": 0.3489, + "num_input_tokens_seen": 57134688, + "step": 27345 + }, + { + "epoch": 4.461701606982625, + "grad_norm": 0.00047942413948476315, + "learning_rate": 0.0681494144054871, + "loss": 0.3947, + "num_input_tokens_seen": 57144032, + "step": 27350 + }, + { + "epoch": 4.462517334203443, + "grad_norm": 0.0004611215554177761, + "learning_rate": 0.06810005845644286, + "loss": 0.3455, + "num_input_tokens_seen": 57153712, + "step": 27355 + }, + { + "epoch": 4.46333306142426, + "grad_norm": 0.0005662112962454557, + "learning_rate": 0.06805071513739878, + "loss": 0.3871, + "num_input_tokens_seen": 57164560, + "step": 27360 + }, + { + "epoch": 4.464148788645077, + "grad_norm": 0.0007699862471781671, + "learning_rate": 0.06800138445596428, + "loss": 0.369, + "num_input_tokens_seen": 57174048, + "step": 27365 + }, + { + "epoch": 4.464964515865894, + "grad_norm": 0.0003632446750998497, + "learning_rate": 0.06795206641974678, + "loss": 0.3143, + "num_input_tokens_seen": 57184304, + "step": 27370 + }, + { + "epoch": 4.465780243086712, + "grad_norm": 0.00044579102541320026, + "learning_rate": 0.06790276103635169, + "loss": 0.2906, + "num_input_tokens_seen": 57194240, + "step": 27375 + }, + { + "epoch": 4.466595970307529, + "grad_norm": 0.0008378512575291097, + "learning_rate": 0.0678534683133826, + "loss": 0.2949, + "num_input_tokens_seen": 57202928, + "step": 27380 + }, + { + "epoch": 4.467411697528346, + "grad_norm": 0.0005174413090571761, + "learning_rate": 0.06780418825844095, + "loss": 0.3434, + "num_input_tokens_seen": 57212272, + "step": 27385 + }, + { + "epoch": 4.468227424749164, + "grad_norm": 0.0004462622164282948, + "learning_rate": 0.0677549208791264, + "loss": 0.3583, + "num_input_tokens_seen": 57223424, + "step": 27390 + }, + { + "epoch": 4.469043151969982, + "grad_norm": 0.000377959426259622, + "learning_rate": 0.06770566618303668, + "loss": 0.3004, + "num_input_tokens_seen": 57234608, + "step": 27395 + }, + { + "epoch": 4.469858879190799, + "grad_norm": 0.0002619597944431007, + "learning_rate": 0.06765642417776736, + "loss": 0.339, + "num_input_tokens_seen": 57245776, + "step": 27400 + }, + { + "epoch": 4.469858879190799, + "eval_loss": 0.3152846097946167, + "eval_runtime": 155.7405, + "eval_samples_per_second": 17.497, + "eval_steps_per_second": 8.752, + "num_input_tokens_seen": 57245776, + "step": 27400 + }, + { + "epoch": 4.470674606411616, + "grad_norm": 0.0006551764090545475, + "learning_rate": 0.0676071948709122, + "loss": 0.4106, + "num_input_tokens_seen": 57256528, + "step": 27405 + }, + { + "epoch": 4.471490333632433, + "grad_norm": 0.00040469589293934405, + "learning_rate": 0.06755797827006307, + "loss": 0.2839, + "num_input_tokens_seen": 57266272, + "step": 27410 + }, + { + "epoch": 4.472306060853251, + "grad_norm": 0.00037820980651304126, + "learning_rate": 0.06750877438280974, + "loss": 0.2921, + "num_input_tokens_seen": 57277008, + "step": 27415 + }, + { + "epoch": 4.473121788074068, + "grad_norm": 0.0008461694233119488, + "learning_rate": 0.06745958321673998, + "loss": 0.2975, + "num_input_tokens_seen": 57287632, + "step": 27420 + }, + { + "epoch": 4.473937515294885, + "grad_norm": 0.0004459690535441041, + "learning_rate": 0.0674104047794398, + "loss": 0.3035, + "num_input_tokens_seen": 57297104, + "step": 27425 + }, + { + "epoch": 4.474753242515702, + "grad_norm": 0.0006625402020290494, + "learning_rate": 0.06736123907849303, + "loss": 0.326, + "num_input_tokens_seen": 57305968, + "step": 27430 + }, + { + "epoch": 4.47556896973652, + "grad_norm": 0.0003586360835470259, + "learning_rate": 0.06731208612148178, + "loss": 0.3319, + "num_input_tokens_seen": 57316400, + "step": 27435 + }, + { + "epoch": 4.4763846969573375, + "grad_norm": 0.0005483888671733439, + "learning_rate": 0.0672629459159859, + "loss": 0.344, + "num_input_tokens_seen": 57327376, + "step": 27440 + }, + { + "epoch": 4.477200424178155, + "grad_norm": 0.00030641278135590255, + "learning_rate": 0.0672138184695835, + "loss": 0.3249, + "num_input_tokens_seen": 57336704, + "step": 27445 + }, + { + "epoch": 4.478016151398972, + "grad_norm": 0.0003315695794299245, + "learning_rate": 0.0671647037898507, + "loss": 0.3291, + "num_input_tokens_seen": 57347056, + "step": 27450 + }, + { + "epoch": 4.47883187861979, + "grad_norm": 0.0005551692447625101, + "learning_rate": 0.0671156018843615, + "loss": 0.3307, + "num_input_tokens_seen": 57357792, + "step": 27455 + }, + { + "epoch": 4.479647605840607, + "grad_norm": 0.0003131765406578779, + "learning_rate": 0.06706651276068812, + "loss": 0.2631, + "num_input_tokens_seen": 57368576, + "step": 27460 + }, + { + "epoch": 4.480463333061424, + "grad_norm": 0.000511825957801193, + "learning_rate": 0.06701743642640064, + "loss": 0.2611, + "num_input_tokens_seen": 57379616, + "step": 27465 + }, + { + "epoch": 4.481279060282241, + "grad_norm": 0.00043479932355694473, + "learning_rate": 0.06696837288906729, + "loss": 0.2839, + "num_input_tokens_seen": 57389840, + "step": 27470 + }, + { + "epoch": 4.482094787503059, + "grad_norm": 0.0005345606477931142, + "learning_rate": 0.06691932215625432, + "loss": 0.2688, + "num_input_tokens_seen": 57399840, + "step": 27475 + }, + { + "epoch": 4.482910514723876, + "grad_norm": 0.0004037805483676493, + "learning_rate": 0.06687028423552589, + "loss": 0.3047, + "num_input_tokens_seen": 57409568, + "step": 27480 + }, + { + "epoch": 4.4837262419446935, + "grad_norm": 0.0007616623770445585, + "learning_rate": 0.06682125913444435, + "loss": 0.374, + "num_input_tokens_seen": 57418560, + "step": 27485 + }, + { + "epoch": 4.484541969165511, + "grad_norm": 0.00038866919931024313, + "learning_rate": 0.0667722468605699, + "loss": 0.3435, + "num_input_tokens_seen": 57427904, + "step": 27490 + }, + { + "epoch": 4.485357696386329, + "grad_norm": 0.0002957389806397259, + "learning_rate": 0.06672324742146094, + "loss": 0.3376, + "num_input_tokens_seen": 57437408, + "step": 27495 + }, + { + "epoch": 4.486173423607146, + "grad_norm": 0.0006715520285069942, + "learning_rate": 0.06667426082467373, + "loss": 0.3029, + "num_input_tokens_seen": 57447664, + "step": 27500 + }, + { + "epoch": 4.486989150827963, + "grad_norm": 0.0003701407113112509, + "learning_rate": 0.0666252870777626, + "loss": 0.3143, + "num_input_tokens_seen": 57457424, + "step": 27505 + }, + { + "epoch": 4.487804878048781, + "grad_norm": 0.0003816764219664037, + "learning_rate": 0.06657632618827995, + "loss": 0.3096, + "num_input_tokens_seen": 57466784, + "step": 27510 + }, + { + "epoch": 4.488620605269598, + "grad_norm": 0.00046725396532565355, + "learning_rate": 0.06652737816377623, + "loss": 0.3211, + "num_input_tokens_seen": 57475296, + "step": 27515 + }, + { + "epoch": 4.489436332490415, + "grad_norm": 0.0003549342800397426, + "learning_rate": 0.06647844301179971, + "loss": 0.2372, + "num_input_tokens_seen": 57486400, + "step": 27520 + }, + { + "epoch": 4.490252059711232, + "grad_norm": 0.00043245372944511473, + "learning_rate": 0.06642952073989689, + "loss": 0.2717, + "num_input_tokens_seen": 57498304, + "step": 27525 + }, + { + "epoch": 4.49106778693205, + "grad_norm": 0.00046380978892557323, + "learning_rate": 0.06638061135561223, + "loss": 0.2681, + "num_input_tokens_seen": 57507664, + "step": 27530 + }, + { + "epoch": 4.4918835141528675, + "grad_norm": 0.0005041753756813705, + "learning_rate": 0.06633171486648808, + "loss": 0.2944, + "num_input_tokens_seen": 57518720, + "step": 27535 + }, + { + "epoch": 4.492699241373685, + "grad_norm": 0.0005098702386021614, + "learning_rate": 0.06628283128006499, + "loss": 0.3451, + "num_input_tokens_seen": 57528496, + "step": 27540 + }, + { + "epoch": 4.493514968594502, + "grad_norm": 0.0004021340573672205, + "learning_rate": 0.0662339606038813, + "loss": 0.3427, + "num_input_tokens_seen": 57538416, + "step": 27545 + }, + { + "epoch": 4.49433069581532, + "grad_norm": 0.0003734708880074322, + "learning_rate": 0.06618510284547358, + "loss": 0.3282, + "num_input_tokens_seen": 57548560, + "step": 27550 + }, + { + "epoch": 4.495146423036137, + "grad_norm": 0.0009267837158404291, + "learning_rate": 0.06613625801237633, + "loss": 0.3505, + "num_input_tokens_seen": 57558400, + "step": 27555 + }, + { + "epoch": 4.495962150256954, + "grad_norm": 0.00035761581966653466, + "learning_rate": 0.066087426112122, + "loss": 0.3019, + "num_input_tokens_seen": 57568496, + "step": 27560 + }, + { + "epoch": 4.496777877477771, + "grad_norm": 0.000962444581091404, + "learning_rate": 0.06603860715224101, + "loss": 0.3023, + "num_input_tokens_seen": 57578704, + "step": 27565 + }, + { + "epoch": 4.497593604698589, + "grad_norm": 0.0005547392065636814, + "learning_rate": 0.06598980114026198, + "loss": 0.3152, + "num_input_tokens_seen": 57590128, + "step": 27570 + }, + { + "epoch": 4.498409331919406, + "grad_norm": 0.0005162959569133818, + "learning_rate": 0.06594100808371128, + "loss": 0.3046, + "num_input_tokens_seen": 57601472, + "step": 27575 + }, + { + "epoch": 4.499225059140223, + "grad_norm": 0.00028360640862956643, + "learning_rate": 0.06589222799011357, + "loss": 0.3397, + "num_input_tokens_seen": 57611648, + "step": 27580 + }, + { + "epoch": 4.5000407863610405, + "grad_norm": 0.0006496589630842209, + "learning_rate": 0.0658434608669912, + "loss": 0.3873, + "num_input_tokens_seen": 57622560, + "step": 27585 + }, + { + "epoch": 4.500856513581859, + "grad_norm": 0.001005358761176467, + "learning_rate": 0.06579470672186473, + "loss": 0.3382, + "num_input_tokens_seen": 57632656, + "step": 27590 + }, + { + "epoch": 4.501672240802676, + "grad_norm": 0.00045750103890895844, + "learning_rate": 0.06574596556225275, + "loss": 0.2953, + "num_input_tokens_seen": 57642704, + "step": 27595 + }, + { + "epoch": 4.502487968023493, + "grad_norm": 0.0007818405865691602, + "learning_rate": 0.06569723739567161, + "loss": 0.3405, + "num_input_tokens_seen": 57651824, + "step": 27600 + }, + { + "epoch": 4.502487968023493, + "eval_loss": 0.3173266053199768, + "eval_runtime": 156.0651, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 57651824, + "step": 27600 + }, + { + "epoch": 4.50330369524431, + "grad_norm": 0.0008765995153225958, + "learning_rate": 0.06564852222963588, + "loss": 0.3041, + "num_input_tokens_seen": 57662976, + "step": 27605 + }, + { + "epoch": 4.504119422465128, + "grad_norm": 0.0007171145989559591, + "learning_rate": 0.06559982007165813, + "loss": 0.2688, + "num_input_tokens_seen": 57673952, + "step": 27610 + }, + { + "epoch": 4.504935149685945, + "grad_norm": 0.0005004492704756558, + "learning_rate": 0.06555113092924868, + "loss": 0.2896, + "num_input_tokens_seen": 57684704, + "step": 27615 + }, + { + "epoch": 4.505750876906762, + "grad_norm": 0.00042944453889504075, + "learning_rate": 0.06550245480991615, + "loss": 0.2815, + "num_input_tokens_seen": 57695008, + "step": 27620 + }, + { + "epoch": 4.506566604127579, + "grad_norm": 0.0005179873551242054, + "learning_rate": 0.0654537917211669, + "loss": 0.3495, + "num_input_tokens_seen": 57704624, + "step": 27625 + }, + { + "epoch": 4.507382331348397, + "grad_norm": 0.0008994975942187011, + "learning_rate": 0.0654051416705055, + "loss": 0.2744, + "num_input_tokens_seen": 57714224, + "step": 27630 + }, + { + "epoch": 4.5081980585692145, + "grad_norm": 0.00032986258156597614, + "learning_rate": 0.06535650466543427, + "loss": 0.276, + "num_input_tokens_seen": 57723680, + "step": 27635 + }, + { + "epoch": 4.509013785790032, + "grad_norm": 0.0004554312617983669, + "learning_rate": 0.0653078807134538, + "loss": 0.3102, + "num_input_tokens_seen": 57732576, + "step": 27640 + }, + { + "epoch": 4.50982951301085, + "grad_norm": 0.00046842408482916653, + "learning_rate": 0.06525926982206236, + "loss": 0.3113, + "num_input_tokens_seen": 57741872, + "step": 27645 + }, + { + "epoch": 4.510645240231667, + "grad_norm": 0.0005285517545416951, + "learning_rate": 0.06521067199875648, + "loss": 0.2592, + "num_input_tokens_seen": 57752272, + "step": 27650 + }, + { + "epoch": 4.511460967452484, + "grad_norm": 0.0007782935863360763, + "learning_rate": 0.06516208725103047, + "loss": 0.3275, + "num_input_tokens_seen": 57761488, + "step": 27655 + }, + { + "epoch": 4.512276694673301, + "grad_norm": 0.0006652522133663297, + "learning_rate": 0.06511351558637678, + "loss": 0.3264, + "num_input_tokens_seen": 57770704, + "step": 27660 + }, + { + "epoch": 4.513092421894118, + "grad_norm": 0.0007094059255905449, + "learning_rate": 0.06506495701228569, + "loss": 0.2256, + "num_input_tokens_seen": 57781632, + "step": 27665 + }, + { + "epoch": 4.513908149114936, + "grad_norm": 0.000550207681953907, + "learning_rate": 0.06501641153624559, + "loss": 0.3002, + "num_input_tokens_seen": 57791264, + "step": 27670 + }, + { + "epoch": 4.514723876335753, + "grad_norm": 0.0003159786865580827, + "learning_rate": 0.06496787916574286, + "loss": 0.3482, + "num_input_tokens_seen": 57803632, + "step": 27675 + }, + { + "epoch": 4.5155396035565705, + "grad_norm": 0.00038969595334492624, + "learning_rate": 0.06491935990826168, + "loss": 0.2856, + "num_input_tokens_seen": 57813504, + "step": 27680 + }, + { + "epoch": 4.5163553307773885, + "grad_norm": 0.00046871043741703033, + "learning_rate": 0.0648708537712844, + "loss": 0.2493, + "num_input_tokens_seen": 57824816, + "step": 27685 + }, + { + "epoch": 4.517171057998206, + "grad_norm": 0.0002671059046406299, + "learning_rate": 0.06482236076229132, + "loss": 0.3308, + "num_input_tokens_seen": 57833760, + "step": 27690 + }, + { + "epoch": 4.517986785219023, + "grad_norm": 0.0009313413756899536, + "learning_rate": 0.06477388088876056, + "loss": 0.4085, + "num_input_tokens_seen": 57843328, + "step": 27695 + }, + { + "epoch": 4.51880251243984, + "grad_norm": 0.0007770788506604731, + "learning_rate": 0.06472541415816846, + "loss": 0.2905, + "num_input_tokens_seen": 57853504, + "step": 27700 + }, + { + "epoch": 4.519618239660657, + "grad_norm": 0.00043294255738146603, + "learning_rate": 0.06467696057798909, + "loss": 0.3521, + "num_input_tokens_seen": 57863632, + "step": 27705 + }, + { + "epoch": 4.520433966881475, + "grad_norm": 0.00022584735415875912, + "learning_rate": 0.0646285201556946, + "loss": 0.3435, + "num_input_tokens_seen": 57872704, + "step": 27710 + }, + { + "epoch": 4.521249694102292, + "grad_norm": 0.0004112667520530522, + "learning_rate": 0.06458009289875521, + "loss": 0.2951, + "num_input_tokens_seen": 57882272, + "step": 27715 + }, + { + "epoch": 4.522065421323109, + "grad_norm": 0.0005232463590800762, + "learning_rate": 0.0645316788146389, + "loss": 0.2916, + "num_input_tokens_seen": 57893712, + "step": 27720 + }, + { + "epoch": 4.522881148543927, + "grad_norm": 0.00034924858482554555, + "learning_rate": 0.06448327791081175, + "loss": 0.2885, + "num_input_tokens_seen": 57903520, + "step": 27725 + }, + { + "epoch": 4.523696875764744, + "grad_norm": 0.0006840704008936882, + "learning_rate": 0.0644348901947379, + "loss": 0.322, + "num_input_tokens_seen": 57913808, + "step": 27730 + }, + { + "epoch": 4.524512602985562, + "grad_norm": 0.000588861876167357, + "learning_rate": 0.06438651567387917, + "loss": 0.2992, + "num_input_tokens_seen": 57924064, + "step": 27735 + }, + { + "epoch": 4.525328330206379, + "grad_norm": 0.0005445186397992074, + "learning_rate": 0.0643381543556957, + "loss": 0.4269, + "num_input_tokens_seen": 57935104, + "step": 27740 + }, + { + "epoch": 4.526144057427197, + "grad_norm": 0.0010205291910097003, + "learning_rate": 0.06428980624764526, + "loss": 0.3454, + "num_input_tokens_seen": 57946800, + "step": 27745 + }, + { + "epoch": 4.526959784648014, + "grad_norm": 0.001497355755418539, + "learning_rate": 0.06424147135718378, + "loss": 0.3643, + "num_input_tokens_seen": 57956608, + "step": 27750 + }, + { + "epoch": 4.527775511868831, + "grad_norm": 0.0005369537393562496, + "learning_rate": 0.06419314969176519, + "loss": 0.3565, + "num_input_tokens_seen": 57966464, + "step": 27755 + }, + { + "epoch": 4.528591239089648, + "grad_norm": 0.00039336972986347973, + "learning_rate": 0.06414484125884118, + "loss": 0.3397, + "num_input_tokens_seen": 57977344, + "step": 27760 + }, + { + "epoch": 4.529406966310466, + "grad_norm": 0.0004099195066373795, + "learning_rate": 0.06409654606586157, + "loss": 0.2763, + "num_input_tokens_seen": 57986752, + "step": 27765 + }, + { + "epoch": 4.530222693531283, + "grad_norm": 0.0005039112875238061, + "learning_rate": 0.06404826412027415, + "loss": 0.2667, + "num_input_tokens_seen": 57997104, + "step": 27770 + }, + { + "epoch": 4.5310384207521, + "grad_norm": 0.00035301659954711795, + "learning_rate": 0.06399999542952453, + "loss": 0.3034, + "num_input_tokens_seen": 58008464, + "step": 27775 + }, + { + "epoch": 4.5318541479729175, + "grad_norm": 0.0006500568124465644, + "learning_rate": 0.0639517400010563, + "loss": 0.3157, + "num_input_tokens_seen": 58019488, + "step": 27780 + }, + { + "epoch": 4.5326698751937355, + "grad_norm": 0.0010242497082799673, + "learning_rate": 0.06390349784231118, + "loss": 0.2856, + "num_input_tokens_seen": 58029504, + "step": 27785 + }, + { + "epoch": 4.533485602414553, + "grad_norm": 0.0006112410919740796, + "learning_rate": 0.06385526896072859, + "loss": 0.3493, + "num_input_tokens_seen": 58039744, + "step": 27790 + }, + { + "epoch": 4.53430132963537, + "grad_norm": 0.0007417788146995008, + "learning_rate": 0.06380705336374613, + "loss": 0.2771, + "num_input_tokens_seen": 58050432, + "step": 27795 + }, + { + "epoch": 4.535117056856187, + "grad_norm": 0.00043339733383618295, + "learning_rate": 0.06375885105879918, + "loss": 0.2684, + "num_input_tokens_seen": 58060672, + "step": 27800 + }, + { + "epoch": 4.535117056856187, + "eval_loss": 0.31762340664863586, + "eval_runtime": 156.0707, + "eval_samples_per_second": 17.46, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 58060672, + "step": 27800 + }, + { + "epoch": 4.535932784077005, + "grad_norm": 0.0004501728108152747, + "learning_rate": 0.06371066205332115, + "loss": 0.3057, + "num_input_tokens_seen": 58070560, + "step": 27805 + }, + { + "epoch": 4.536748511297822, + "grad_norm": 0.0003027570783160627, + "learning_rate": 0.06366248635474347, + "loss": 0.3402, + "num_input_tokens_seen": 58079824, + "step": 27810 + }, + { + "epoch": 4.537564238518639, + "grad_norm": 0.0009920786833390594, + "learning_rate": 0.06361432397049532, + "loss": 0.2929, + "num_input_tokens_seen": 58092144, + "step": 27815 + }, + { + "epoch": 4.538379965739456, + "grad_norm": 0.000898390484508127, + "learning_rate": 0.06356617490800408, + "loss": 0.2704, + "num_input_tokens_seen": 58103568, + "step": 27820 + }, + { + "epoch": 4.539195692960274, + "grad_norm": 0.000855075370054692, + "learning_rate": 0.06351803917469478, + "loss": 0.3795, + "num_input_tokens_seen": 58113744, + "step": 27825 + }, + { + "epoch": 4.5400114201810915, + "grad_norm": 0.0005661520408466458, + "learning_rate": 0.06346991677799067, + "loss": 0.3895, + "num_input_tokens_seen": 58123520, + "step": 27830 + }, + { + "epoch": 4.540827147401909, + "grad_norm": 0.000694344169460237, + "learning_rate": 0.06342180772531283, + "loss": 0.2667, + "num_input_tokens_seen": 58134720, + "step": 27835 + }, + { + "epoch": 4.541642874622726, + "grad_norm": 0.0007283947779797018, + "learning_rate": 0.06337371202408021, + "loss": 0.3294, + "num_input_tokens_seen": 58143776, + "step": 27840 + }, + { + "epoch": 4.542458601843544, + "grad_norm": 0.00043832173105329275, + "learning_rate": 0.06332562968170984, + "loss": 0.332, + "num_input_tokens_seen": 58153520, + "step": 27845 + }, + { + "epoch": 4.543274329064361, + "grad_norm": 0.00039895542431622744, + "learning_rate": 0.06327756070561656, + "loss": 0.3512, + "num_input_tokens_seen": 58163696, + "step": 27850 + }, + { + "epoch": 4.544090056285178, + "grad_norm": 0.0006387625471688807, + "learning_rate": 0.06322950510321329, + "loss": 0.3218, + "num_input_tokens_seen": 58174080, + "step": 27855 + }, + { + "epoch": 4.544905783505996, + "grad_norm": 0.000583952700253576, + "learning_rate": 0.06318146288191076, + "loss": 0.3211, + "num_input_tokens_seen": 58185104, + "step": 27860 + }, + { + "epoch": 4.545721510726813, + "grad_norm": 0.0004390173126012087, + "learning_rate": 0.06313343404911763, + "loss": 0.2619, + "num_input_tokens_seen": 58195200, + "step": 27865 + }, + { + "epoch": 4.54653723794763, + "grad_norm": 0.00031467765802517533, + "learning_rate": 0.0630854186122406, + "loss": 0.2844, + "num_input_tokens_seen": 58205632, + "step": 27870 + }, + { + "epoch": 4.5473529651684474, + "grad_norm": 0.00040815165266394615, + "learning_rate": 0.06303741657868431, + "loss": 0.3076, + "num_input_tokens_seen": 58218112, + "step": 27875 + }, + { + "epoch": 4.548168692389265, + "grad_norm": 0.0010871640406548977, + "learning_rate": 0.06298942795585115, + "loss": 0.3809, + "num_input_tokens_seen": 58226848, + "step": 27880 + }, + { + "epoch": 4.548984419610083, + "grad_norm": 0.0003925535420421511, + "learning_rate": 0.06294145275114167, + "loss": 0.2904, + "num_input_tokens_seen": 58238080, + "step": 27885 + }, + { + "epoch": 4.5498001468309, + "grad_norm": 0.0003381349379196763, + "learning_rate": 0.06289349097195428, + "loss": 0.3701, + "num_input_tokens_seen": 58248400, + "step": 27890 + }, + { + "epoch": 4.550615874051717, + "grad_norm": 0.00032930553425103426, + "learning_rate": 0.06284554262568516, + "loss": 0.3786, + "num_input_tokens_seen": 58258560, + "step": 27895 + }, + { + "epoch": 4.551431601272535, + "grad_norm": 0.0007206554873846471, + "learning_rate": 0.06279760771972868, + "loss": 0.2912, + "num_input_tokens_seen": 58268800, + "step": 27900 + }, + { + "epoch": 4.552247328493352, + "grad_norm": 0.0004510356520768255, + "learning_rate": 0.06274968626147688, + "loss": 0.2838, + "num_input_tokens_seen": 58278592, + "step": 27905 + }, + { + "epoch": 4.553063055714169, + "grad_norm": 0.0010824654018506408, + "learning_rate": 0.06270177825831993, + "loss": 0.3268, + "num_input_tokens_seen": 58289936, + "step": 27910 + }, + { + "epoch": 4.553878782934986, + "grad_norm": 0.0005333981243893504, + "learning_rate": 0.06265388371764587, + "loss": 0.2814, + "num_input_tokens_seen": 58300576, + "step": 27915 + }, + { + "epoch": 4.554694510155803, + "grad_norm": 0.001106349634937942, + "learning_rate": 0.0626060026468406, + "loss": 0.39, + "num_input_tokens_seen": 58310160, + "step": 27920 + }, + { + "epoch": 4.555510237376621, + "grad_norm": 0.00038614190998487175, + "learning_rate": 0.06255813505328794, + "loss": 0.3327, + "num_input_tokens_seen": 58319808, + "step": 27925 + }, + { + "epoch": 4.5563259645974385, + "grad_norm": 0.0005870744935236871, + "learning_rate": 0.06251028094436978, + "loss": 0.3176, + "num_input_tokens_seen": 58331360, + "step": 27930 + }, + { + "epoch": 4.557141691818256, + "grad_norm": 0.0002804561227094382, + "learning_rate": 0.06246244032746568, + "loss": 0.2799, + "num_input_tokens_seen": 58341152, + "step": 27935 + }, + { + "epoch": 4.557957419039074, + "grad_norm": 0.0004372968105599284, + "learning_rate": 0.06241461320995342, + "loss": 0.3116, + "num_input_tokens_seen": 58351472, + "step": 27940 + }, + { + "epoch": 4.558773146259891, + "grad_norm": 0.0008362994412891567, + "learning_rate": 0.062366799599208426, + "loss": 0.3476, + "num_input_tokens_seen": 58359664, + "step": 27945 + }, + { + "epoch": 4.559588873480708, + "grad_norm": 0.001292986678890884, + "learning_rate": 0.06231899950260418, + "loss": 0.385, + "num_input_tokens_seen": 58370112, + "step": 27950 + }, + { + "epoch": 4.560404600701525, + "grad_norm": 0.0003627216792665422, + "learning_rate": 0.06227121292751214, + "loss": 0.2912, + "num_input_tokens_seen": 58380560, + "step": 27955 + }, + { + "epoch": 4.561220327922343, + "grad_norm": 0.0005793889285996556, + "learning_rate": 0.062223439881301496, + "loss": 0.297, + "num_input_tokens_seen": 58391968, + "step": 27960 + }, + { + "epoch": 4.56203605514316, + "grad_norm": 0.0007426101365126669, + "learning_rate": 0.06217568037133948, + "loss": 0.3392, + "num_input_tokens_seen": 58401984, + "step": 27965 + }, + { + "epoch": 4.562851782363977, + "grad_norm": 0.0013344207545742393, + "learning_rate": 0.06212793440499126, + "loss": 0.2834, + "num_input_tokens_seen": 58412048, + "step": 27970 + }, + { + "epoch": 4.5636675095847945, + "grad_norm": 0.0004227184981573373, + "learning_rate": 0.062080201989619783, + "loss": 0.3326, + "num_input_tokens_seen": 58422864, + "step": 27975 + }, + { + "epoch": 4.5644832368056125, + "grad_norm": 0.00062164495466277, + "learning_rate": 0.062032483132586094, + "loss": 0.3144, + "num_input_tokens_seen": 58433808, + "step": 27980 + }, + { + "epoch": 4.56529896402643, + "grad_norm": 0.0006962307379581034, + "learning_rate": 0.0619847778412489, + "loss": 0.3146, + "num_input_tokens_seen": 58443376, + "step": 27985 + }, + { + "epoch": 4.566114691247247, + "grad_norm": 0.00033114850521087646, + "learning_rate": 0.06193708612296509, + "loss": 0.2933, + "num_input_tokens_seen": 58453936, + "step": 27990 + }, + { + "epoch": 4.566930418468064, + "grad_norm": 0.0004998814547434449, + "learning_rate": 0.06188940798508923, + "loss": 0.3483, + "num_input_tokens_seen": 58464416, + "step": 27995 + }, + { + "epoch": 4.567746145688882, + "grad_norm": 0.0007230261107906699, + "learning_rate": 0.06184174343497397, + "loss": 0.3248, + "num_input_tokens_seen": 58475488, + "step": 28000 + }, + { + "epoch": 4.567746145688882, + "eval_loss": 0.3164209723472595, + "eval_runtime": 155.7744, + "eval_samples_per_second": 17.493, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 58475488, + "step": 28000 + }, + { + "epoch": 4.568561872909699, + "grad_norm": 0.0003601922944653779, + "learning_rate": 0.061794092479969726, + "loss": 0.2851, + "num_input_tokens_seen": 58484720, + "step": 28005 + }, + { + "epoch": 4.569377600130516, + "grad_norm": 0.0004062915686517954, + "learning_rate": 0.06174645512742485, + "loss": 0.2492, + "num_input_tokens_seen": 58495184, + "step": 28010 + }, + { + "epoch": 4.570193327351333, + "grad_norm": 0.00026371210697107017, + "learning_rate": 0.06169883138468565, + "loss": 0.2987, + "num_input_tokens_seen": 58506336, + "step": 28015 + }, + { + "epoch": 4.571009054572151, + "grad_norm": 0.0004514685133472085, + "learning_rate": 0.06165122125909637, + "loss": 0.2639, + "num_input_tokens_seen": 58516768, + "step": 28020 + }, + { + "epoch": 4.5718247817929685, + "grad_norm": 0.0004864537622779608, + "learning_rate": 0.061603624757998965, + "loss": 0.3, + "num_input_tokens_seen": 58528272, + "step": 28025 + }, + { + "epoch": 4.572640509013786, + "grad_norm": 0.0006987181259319186, + "learning_rate": 0.0615560418887335, + "loss": 0.3179, + "num_input_tokens_seen": 58539232, + "step": 28030 + }, + { + "epoch": 4.573456236234604, + "grad_norm": 0.0007012615678831935, + "learning_rate": 0.06150847265863787, + "loss": 0.2688, + "num_input_tokens_seen": 58550400, + "step": 28035 + }, + { + "epoch": 4.574271963455421, + "grad_norm": 0.0010184170678257942, + "learning_rate": 0.061460917075047757, + "loss": 0.3867, + "num_input_tokens_seen": 58560288, + "step": 28040 + }, + { + "epoch": 4.575087690676238, + "grad_norm": 0.0005127341137267649, + "learning_rate": 0.06141337514529694, + "loss": 0.3003, + "num_input_tokens_seen": 58570048, + "step": 28045 + }, + { + "epoch": 4.575903417897055, + "grad_norm": 0.0004894917365163565, + "learning_rate": 0.06136584687671687, + "loss": 0.2747, + "num_input_tokens_seen": 58580240, + "step": 28050 + }, + { + "epoch": 4.576719145117872, + "grad_norm": 0.00042890041368082166, + "learning_rate": 0.061318332276637064, + "loss": 0.3111, + "num_input_tokens_seen": 58589200, + "step": 28055 + }, + { + "epoch": 4.57753487233869, + "grad_norm": 0.0009014893439598382, + "learning_rate": 0.06127083135238491, + "loss": 0.3026, + "num_input_tokens_seen": 58600768, + "step": 28060 + }, + { + "epoch": 4.578350599559507, + "grad_norm": 0.0004283388261683285, + "learning_rate": 0.06122334411128555, + "loss": 0.2932, + "num_input_tokens_seen": 58611376, + "step": 28065 + }, + { + "epoch": 4.579166326780324, + "grad_norm": 0.00031273619970306754, + "learning_rate": 0.06117587056066223, + "loss": 0.3215, + "num_input_tokens_seen": 58620496, + "step": 28070 + }, + { + "epoch": 4.5799820540011424, + "grad_norm": 0.0007057571201585233, + "learning_rate": 0.06112841070783589, + "loss": 0.3023, + "num_input_tokens_seen": 58631264, + "step": 28075 + }, + { + "epoch": 4.58079778122196, + "grad_norm": 0.0007054224261082709, + "learning_rate": 0.061080964560125406, + "loss": 0.3616, + "num_input_tokens_seen": 58640272, + "step": 28080 + }, + { + "epoch": 4.581613508442777, + "grad_norm": 0.000702817807905376, + "learning_rate": 0.06103353212484766, + "loss": 0.305, + "num_input_tokens_seen": 58649280, + "step": 28085 + }, + { + "epoch": 4.582429235663594, + "grad_norm": 0.00042660589679144323, + "learning_rate": 0.06098611340931722, + "loss": 0.3132, + "num_input_tokens_seen": 58658960, + "step": 28090 + }, + { + "epoch": 4.583244962884411, + "grad_norm": 0.000777060107793659, + "learning_rate": 0.06093870842084672, + "loss": 0.3075, + "num_input_tokens_seen": 58671856, + "step": 28095 + }, + { + "epoch": 4.584060690105229, + "grad_norm": 0.00055519217858091, + "learning_rate": 0.06089131716674666, + "loss": 0.3263, + "num_input_tokens_seen": 58681456, + "step": 28100 + }, + { + "epoch": 4.584876417326046, + "grad_norm": 0.0008849296136759222, + "learning_rate": 0.060843939654325226, + "loss": 0.3438, + "num_input_tokens_seen": 58692080, + "step": 28105 + }, + { + "epoch": 4.585692144546863, + "grad_norm": 0.0003217872872482985, + "learning_rate": 0.06079657589088873, + "loss": 0.2512, + "num_input_tokens_seen": 58703024, + "step": 28110 + }, + { + "epoch": 4.586507871767681, + "grad_norm": 0.0008413592586293817, + "learning_rate": 0.06074922588374126, + "loss": 0.326, + "num_input_tokens_seen": 58714960, + "step": 28115 + }, + { + "epoch": 4.587323598988498, + "grad_norm": 0.0011546361492946744, + "learning_rate": 0.06070188964018472, + "loss": 0.3246, + "num_input_tokens_seen": 58725920, + "step": 28120 + }, + { + "epoch": 4.5881393262093155, + "grad_norm": 0.0003845160244964063, + "learning_rate": 0.06065456716751902, + "loss": 0.3114, + "num_input_tokens_seen": 58737584, + "step": 28125 + }, + { + "epoch": 4.588955053430133, + "grad_norm": 0.0009933680994436145, + "learning_rate": 0.06060725847304182, + "loss": 0.2857, + "num_input_tokens_seen": 58748016, + "step": 28130 + }, + { + "epoch": 4.589770780650951, + "grad_norm": 0.0009420475107617676, + "learning_rate": 0.06055996356404877, + "loss": 0.2977, + "num_input_tokens_seen": 58758800, + "step": 28135 + }, + { + "epoch": 4.590586507871768, + "grad_norm": 0.00032632541842758656, + "learning_rate": 0.06051268244783327, + "loss": 0.3121, + "num_input_tokens_seen": 58769904, + "step": 28140 + }, + { + "epoch": 4.591402235092585, + "grad_norm": 0.00038610430783592165, + "learning_rate": 0.06046541513168676, + "loss": 0.256, + "num_input_tokens_seen": 58779696, + "step": 28145 + }, + { + "epoch": 4.592217962313402, + "grad_norm": 0.00046589059638790786, + "learning_rate": 0.060418161622898356, + "loss": 0.2644, + "num_input_tokens_seen": 58790816, + "step": 28150 + }, + { + "epoch": 4.59303368953422, + "grad_norm": 0.0007782959146425128, + "learning_rate": 0.06037092192875521, + "loss": 0.3332, + "num_input_tokens_seen": 58802208, + "step": 28155 + }, + { + "epoch": 4.593849416755037, + "grad_norm": 0.00034636558848433197, + "learning_rate": 0.060323696056542225, + "loss": 0.2952, + "num_input_tokens_seen": 58813168, + "step": 28160 + }, + { + "epoch": 4.594665143975854, + "grad_norm": 0.000590705662034452, + "learning_rate": 0.06027648401354229, + "loss": 0.2914, + "num_input_tokens_seen": 58824880, + "step": 28165 + }, + { + "epoch": 4.5954808711966715, + "grad_norm": 0.00036300579085946083, + "learning_rate": 0.06022928580703601, + "loss": 0.3087, + "num_input_tokens_seen": 58836016, + "step": 28170 + }, + { + "epoch": 4.5962965984174895, + "grad_norm": 0.0009595865267328918, + "learning_rate": 0.060182101444301986, + "loss": 0.2773, + "num_input_tokens_seen": 58846768, + "step": 28175 + }, + { + "epoch": 4.597112325638307, + "grad_norm": 0.000802867638412863, + "learning_rate": 0.06013493093261669, + "loss": 0.3391, + "num_input_tokens_seen": 58856816, + "step": 28180 + }, + { + "epoch": 4.597928052859124, + "grad_norm": 0.0005444218404591084, + "learning_rate": 0.06008777427925432, + "loss": 0.3402, + "num_input_tokens_seen": 58867648, + "step": 28185 + }, + { + "epoch": 4.598743780079941, + "grad_norm": 0.000708250212483108, + "learning_rate": 0.06004063149148705, + "loss": 0.2918, + "num_input_tokens_seen": 58877792, + "step": 28190 + }, + { + "epoch": 4.599559507300759, + "grad_norm": 0.0007077451446093619, + "learning_rate": 0.05999350257658497, + "loss": 0.3103, + "num_input_tokens_seen": 58888176, + "step": 28195 + }, + { + "epoch": 4.600375234521576, + "grad_norm": 0.0007845821091905236, + "learning_rate": 0.05994638754181582, + "loss": 0.3213, + "num_input_tokens_seen": 58898896, + "step": 28200 + }, + { + "epoch": 4.600375234521576, + "eval_loss": 0.31769484281539917, + "eval_runtime": 155.9897, + "eval_samples_per_second": 17.469, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 58898896, + "step": 28200 + }, + { + "epoch": 4.601190961742393, + "grad_norm": 0.0006249836878851056, + "learning_rate": 0.059899286394445445, + "loss": 0.2577, + "num_input_tokens_seen": 58909616, + "step": 28205 + }, + { + "epoch": 4.602006688963211, + "grad_norm": 0.0005907629965804517, + "learning_rate": 0.059852199141737346, + "loss": 0.2985, + "num_input_tokens_seen": 58919776, + "step": 28210 + }, + { + "epoch": 4.602822416184028, + "grad_norm": 0.001056305249221623, + "learning_rate": 0.05980512579095304, + "loss": 0.3492, + "num_input_tokens_seen": 58930704, + "step": 28215 + }, + { + "epoch": 4.6036381434048455, + "grad_norm": 0.0006096430006437004, + "learning_rate": 0.05975806634935181, + "loss": 0.3045, + "num_input_tokens_seen": 58940608, + "step": 28220 + }, + { + "epoch": 4.604453870625663, + "grad_norm": 0.0007649794570170343, + "learning_rate": 0.05971102082419076, + "loss": 0.2613, + "num_input_tokens_seen": 58952352, + "step": 28225 + }, + { + "epoch": 4.60526959784648, + "grad_norm": 0.0007744599715806544, + "learning_rate": 0.05966398922272492, + "loss": 0.299, + "num_input_tokens_seen": 58962528, + "step": 28230 + }, + { + "epoch": 4.606085325067298, + "grad_norm": 0.0006833907100372016, + "learning_rate": 0.059616971552207236, + "loss": 0.3096, + "num_input_tokens_seen": 58971472, + "step": 28235 + }, + { + "epoch": 4.606901052288115, + "grad_norm": 0.0007548312423750758, + "learning_rate": 0.059569967819888305, + "loss": 0.2947, + "num_input_tokens_seen": 58982672, + "step": 28240 + }, + { + "epoch": 4.607716779508932, + "grad_norm": 0.00045123635209165514, + "learning_rate": 0.05952297803301681, + "loss": 0.2573, + "num_input_tokens_seen": 58993904, + "step": 28245 + }, + { + "epoch": 4.60853250672975, + "grad_norm": 0.00042251957347616553, + "learning_rate": 0.059476002198839056, + "loss": 0.2773, + "num_input_tokens_seen": 59004352, + "step": 28250 + }, + { + "epoch": 4.609348233950567, + "grad_norm": 0.0004480051575228572, + "learning_rate": 0.05942904032459935, + "loss": 0.2836, + "num_input_tokens_seen": 59015344, + "step": 28255 + }, + { + "epoch": 4.610163961171384, + "grad_norm": 0.0013310645008459687, + "learning_rate": 0.05938209241753987, + "loss": 0.3649, + "num_input_tokens_seen": 59024640, + "step": 28260 + }, + { + "epoch": 4.610979688392201, + "grad_norm": 0.0005866946303285658, + "learning_rate": 0.05933515848490046, + "loss": 0.2475, + "num_input_tokens_seen": 59034672, + "step": 28265 + }, + { + "epoch": 4.6117954156130185, + "grad_norm": 0.0008309065597131848, + "learning_rate": 0.059288238533918985, + "loss": 0.3014, + "num_input_tokens_seen": 59044112, + "step": 28270 + }, + { + "epoch": 4.6126111428338366, + "grad_norm": 0.0009638108313083649, + "learning_rate": 0.05924133257183113, + "loss": 0.3266, + "num_input_tokens_seen": 59053808, + "step": 28275 + }, + { + "epoch": 4.613426870054654, + "grad_norm": 0.0006156001472845674, + "learning_rate": 0.059194440605870285, + "loss": 0.3134, + "num_input_tokens_seen": 59064640, + "step": 28280 + }, + { + "epoch": 4.614242597275471, + "grad_norm": 0.0006810183986090124, + "learning_rate": 0.059147562643267884, + "loss": 0.3111, + "num_input_tokens_seen": 59076512, + "step": 28285 + }, + { + "epoch": 4.615058324496289, + "grad_norm": 0.0007411534897983074, + "learning_rate": 0.059100698691253055, + "loss": 0.4161, + "num_input_tokens_seen": 59086432, + "step": 28290 + }, + { + "epoch": 4.615874051717106, + "grad_norm": 0.001995920669287443, + "learning_rate": 0.05905384875705273, + "loss": 0.3507, + "num_input_tokens_seen": 59096240, + "step": 28295 + }, + { + "epoch": 4.616689778937923, + "grad_norm": 0.0007896916358731687, + "learning_rate": 0.05900701284789189, + "loss": 0.28, + "num_input_tokens_seen": 59106080, + "step": 28300 + }, + { + "epoch": 4.61750550615874, + "grad_norm": 0.00033187298686243594, + "learning_rate": 0.058960190970993115, + "loss": 0.2938, + "num_input_tokens_seen": 59116816, + "step": 28305 + }, + { + "epoch": 4.618321233379558, + "grad_norm": 0.0005196444108150899, + "learning_rate": 0.058913383133576955, + "loss": 0.3568, + "num_input_tokens_seen": 59128208, + "step": 28310 + }, + { + "epoch": 4.619136960600375, + "grad_norm": 0.0008766053360886872, + "learning_rate": 0.05886658934286185, + "loss": 0.3778, + "num_input_tokens_seen": 59138208, + "step": 28315 + }, + { + "epoch": 4.6199526878211925, + "grad_norm": 0.0005842511309310794, + "learning_rate": 0.058819809606063846, + "loss": 0.3113, + "num_input_tokens_seen": 59149728, + "step": 28320 + }, + { + "epoch": 4.62076841504201, + "grad_norm": 0.0007282238220795989, + "learning_rate": 0.05877304393039711, + "loss": 0.301, + "num_input_tokens_seen": 59160272, + "step": 28325 + }, + { + "epoch": 4.621584142262828, + "grad_norm": 0.0010130490409210324, + "learning_rate": 0.05872629232307338, + "loss": 0.3639, + "num_input_tokens_seen": 59172320, + "step": 28330 + }, + { + "epoch": 4.622399869483645, + "grad_norm": 0.0007750572985969484, + "learning_rate": 0.05867955479130239, + "loss": 0.2954, + "num_input_tokens_seen": 59182816, + "step": 28335 + }, + { + "epoch": 4.623215596704462, + "grad_norm": 0.0012368615716695786, + "learning_rate": 0.058632831342291705, + "loss": 0.3507, + "num_input_tokens_seen": 59193328, + "step": 28340 + }, + { + "epoch": 4.624031323925279, + "grad_norm": 0.0006528382655233145, + "learning_rate": 0.05858612198324655, + "loss": 0.3416, + "num_input_tokens_seen": 59203392, + "step": 28345 + }, + { + "epoch": 4.624847051146097, + "grad_norm": 0.0004955873009748757, + "learning_rate": 0.05853942672137025, + "loss": 0.3561, + "num_input_tokens_seen": 59212800, + "step": 28350 + }, + { + "epoch": 4.625662778366914, + "grad_norm": 0.0007314556278288364, + "learning_rate": 0.05849274556386363, + "loss": 0.2969, + "num_input_tokens_seen": 59222464, + "step": 28355 + }, + { + "epoch": 4.626478505587731, + "grad_norm": 0.00046144117368385196, + "learning_rate": 0.05844607851792567, + "loss": 0.31, + "num_input_tokens_seen": 59233248, + "step": 28360 + }, + { + "epoch": 4.6272942328085485, + "grad_norm": 0.000332203897414729, + "learning_rate": 0.058399425590752924, + "loss": 0.3205, + "num_input_tokens_seen": 59243392, + "step": 28365 + }, + { + "epoch": 4.6281099600293665, + "grad_norm": 0.0006785792065784335, + "learning_rate": 0.05835278678953985, + "loss": 0.269, + "num_input_tokens_seen": 59254432, + "step": 28370 + }, + { + "epoch": 4.628925687250184, + "grad_norm": 0.0003702484827954322, + "learning_rate": 0.05830616212147874, + "loss": 0.3188, + "num_input_tokens_seen": 59265408, + "step": 28375 + }, + { + "epoch": 4.629741414471001, + "grad_norm": 0.00039052992360666394, + "learning_rate": 0.058259551593759784, + "loss": 0.2825, + "num_input_tokens_seen": 59276224, + "step": 28380 + }, + { + "epoch": 4.630557141691818, + "grad_norm": 0.0005101470742374659, + "learning_rate": 0.058212955213570804, + "loss": 0.3906, + "num_input_tokens_seen": 59286560, + "step": 28385 + }, + { + "epoch": 4.631372868912636, + "grad_norm": 0.00043065729551017284, + "learning_rate": 0.0581663729880976, + "loss": 0.2623, + "num_input_tokens_seen": 59297920, + "step": 28390 + }, + { + "epoch": 4.632188596133453, + "grad_norm": 0.0008272643317468464, + "learning_rate": 0.05811980492452379, + "loss": 0.3375, + "num_input_tokens_seen": 59308704, + "step": 28395 + }, + { + "epoch": 4.63300432335427, + "grad_norm": 0.0007697375258430839, + "learning_rate": 0.058073251030030644, + "loss": 0.3105, + "num_input_tokens_seen": 59318976, + "step": 28400 + }, + { + "epoch": 4.63300432335427, + "eval_loss": 0.31559979915618896, + "eval_runtime": 155.8202, + "eval_samples_per_second": 17.488, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 59318976, + "step": 28400 + }, + { + "epoch": 4.633820050575087, + "grad_norm": 0.000651804031804204, + "learning_rate": 0.05802671131179747, + "loss": 0.3235, + "num_input_tokens_seen": 59328672, + "step": 28405 + }, + { + "epoch": 4.634635777795905, + "grad_norm": 0.0003584953665267676, + "learning_rate": 0.057980185777001154, + "loss": 0.3264, + "num_input_tokens_seen": 59338848, + "step": 28410 + }, + { + "epoch": 4.635451505016722, + "grad_norm": 0.0005929706967435777, + "learning_rate": 0.057933674432816606, + "loss": 0.3183, + "num_input_tokens_seen": 59348848, + "step": 28415 + }, + { + "epoch": 4.63626723223754, + "grad_norm": 0.0005470045143738389, + "learning_rate": 0.05788717728641648, + "loss": 0.3228, + "num_input_tokens_seen": 59359568, + "step": 28420 + }, + { + "epoch": 4.637082959458358, + "grad_norm": 0.0012005290482193232, + "learning_rate": 0.057840694344971126, + "loss": 0.3615, + "num_input_tokens_seen": 59371504, + "step": 28425 + }, + { + "epoch": 4.637898686679175, + "grad_norm": 0.0005090049817226827, + "learning_rate": 0.0577942256156489, + "loss": 0.2161, + "num_input_tokens_seen": 59381904, + "step": 28430 + }, + { + "epoch": 4.638714413899992, + "grad_norm": 0.0005892263143323362, + "learning_rate": 0.057747771105615804, + "loss": 0.3166, + "num_input_tokens_seen": 59392896, + "step": 28435 + }, + { + "epoch": 4.639530141120809, + "grad_norm": 0.0004685699241235852, + "learning_rate": 0.05770133082203568, + "loss": 0.2633, + "num_input_tokens_seen": 59403712, + "step": 28440 + }, + { + "epoch": 4.640345868341626, + "grad_norm": 0.0006221800576895475, + "learning_rate": 0.0576549047720703, + "loss": 0.3796, + "num_input_tokens_seen": 59413664, + "step": 28445 + }, + { + "epoch": 4.641161595562444, + "grad_norm": 0.0004159490345045924, + "learning_rate": 0.05760849296287902, + "loss": 0.2732, + "num_input_tokens_seen": 59423408, + "step": 28450 + }, + { + "epoch": 4.641977322783261, + "grad_norm": 0.0005201485473662615, + "learning_rate": 0.05756209540161919, + "loss": 0.3925, + "num_input_tokens_seen": 59433808, + "step": 28455 + }, + { + "epoch": 4.642793050004078, + "grad_norm": 0.0005282842903397977, + "learning_rate": 0.05751571209544595, + "loss": 0.304, + "num_input_tokens_seen": 59442560, + "step": 28460 + }, + { + "epoch": 4.643608777224896, + "grad_norm": 0.0007309484644792974, + "learning_rate": 0.057469343051512085, + "loss": 0.3113, + "num_input_tokens_seen": 59453696, + "step": 28465 + }, + { + "epoch": 4.6444245044457135, + "grad_norm": 0.0005107767647132277, + "learning_rate": 0.057422988276968324, + "loss": 0.2988, + "num_input_tokens_seen": 59464592, + "step": 28470 + }, + { + "epoch": 4.645240231666531, + "grad_norm": 0.0006333562778308988, + "learning_rate": 0.05737664777896323, + "loss": 0.276, + "num_input_tokens_seen": 59475568, + "step": 28475 + }, + { + "epoch": 4.646055958887348, + "grad_norm": 0.0003649734426289797, + "learning_rate": 0.057330321564642975, + "loss": 0.336, + "num_input_tokens_seen": 59486560, + "step": 28480 + }, + { + "epoch": 4.646871686108166, + "grad_norm": 0.0006661544321104884, + "learning_rate": 0.05728400964115174, + "loss": 0.3009, + "num_input_tokens_seen": 59495312, + "step": 28485 + }, + { + "epoch": 4.647687413328983, + "grad_norm": 0.0005427655996754766, + "learning_rate": 0.057237712015631305, + "loss": 0.312, + "num_input_tokens_seen": 59505408, + "step": 28490 + }, + { + "epoch": 4.6485031405498, + "grad_norm": 0.00041955127380788326, + "learning_rate": 0.057191428695221425, + "loss": 0.2886, + "num_input_tokens_seen": 59516096, + "step": 28495 + }, + { + "epoch": 4.649318867770617, + "grad_norm": 0.0007575751515105367, + "learning_rate": 0.05714515968705958, + "loss": 0.3177, + "num_input_tokens_seen": 59527264, + "step": 28500 + }, + { + "epoch": 4.650134594991435, + "grad_norm": 0.000346664572134614, + "learning_rate": 0.05709890499828099, + "loss": 0.356, + "num_input_tokens_seen": 59537808, + "step": 28505 + }, + { + "epoch": 4.650950322212252, + "grad_norm": 0.0007263495353981853, + "learning_rate": 0.05705266463601868, + "loss": 0.2626, + "num_input_tokens_seen": 59549408, + "step": 28510 + }, + { + "epoch": 4.6517660494330695, + "grad_norm": 0.0005969009944237769, + "learning_rate": 0.057006438607403565, + "loss": 0.2673, + "num_input_tokens_seen": 59560496, + "step": 28515 + }, + { + "epoch": 4.652581776653887, + "grad_norm": 0.0008994486997835338, + "learning_rate": 0.056960226919564205, + "loss": 0.2872, + "num_input_tokens_seen": 59570624, + "step": 28520 + }, + { + "epoch": 4.653397503874705, + "grad_norm": 0.001177743193693459, + "learning_rate": 0.05691402957962713, + "loss": 0.4302, + "num_input_tokens_seen": 59580192, + "step": 28525 + }, + { + "epoch": 4.654213231095522, + "grad_norm": 0.00029499552329070866, + "learning_rate": 0.05686784659471642, + "loss": 0.3316, + "num_input_tokens_seen": 59590752, + "step": 28530 + }, + { + "epoch": 4.655028958316339, + "grad_norm": 0.0005447511794045568, + "learning_rate": 0.056821677971954136, + "loss": 0.3559, + "num_input_tokens_seen": 59601440, + "step": 28535 + }, + { + "epoch": 4.655844685537156, + "grad_norm": 0.0006788413738831878, + "learning_rate": 0.05677552371846012, + "loss": 0.3679, + "num_input_tokens_seen": 59611824, + "step": 28540 + }, + { + "epoch": 4.656660412757974, + "grad_norm": 0.00048949068877846, + "learning_rate": 0.05672938384135182, + "loss": 0.299, + "num_input_tokens_seen": 59622880, + "step": 28545 + }, + { + "epoch": 4.657476139978791, + "grad_norm": 0.0006274607731029391, + "learning_rate": 0.05668325834774465, + "loss": 0.2627, + "num_input_tokens_seen": 59633904, + "step": 28550 + }, + { + "epoch": 4.658291867199608, + "grad_norm": 0.0004451198619790375, + "learning_rate": 0.05663714724475177, + "loss": 0.4176, + "num_input_tokens_seen": 59644992, + "step": 28555 + }, + { + "epoch": 4.659107594420425, + "grad_norm": 0.00034649172448553145, + "learning_rate": 0.05659105053948403, + "loss": 0.3073, + "num_input_tokens_seen": 59655392, + "step": 28560 + }, + { + "epoch": 4.6599233216412435, + "grad_norm": 0.000819963461253792, + "learning_rate": 0.056544968239050176, + "loss": 0.3445, + "num_input_tokens_seen": 59665728, + "step": 28565 + }, + { + "epoch": 4.660739048862061, + "grad_norm": 0.0003618358459789306, + "learning_rate": 0.056498900350556616, + "loss": 0.3371, + "num_input_tokens_seen": 59676656, + "step": 28570 + }, + { + "epoch": 4.661554776082878, + "grad_norm": 0.0007079920032992959, + "learning_rate": 0.05645284688110766, + "loss": 0.3347, + "num_input_tokens_seen": 59686800, + "step": 28575 + }, + { + "epoch": 4.662370503303695, + "grad_norm": 0.00037583502125926316, + "learning_rate": 0.05640680783780532, + "loss": 0.3282, + "num_input_tokens_seen": 59696768, + "step": 28580 + }, + { + "epoch": 4.663186230524513, + "grad_norm": 0.0003921617171727121, + "learning_rate": 0.056360783227749324, + "loss": 0.3248, + "num_input_tokens_seen": 59707760, + "step": 28585 + }, + { + "epoch": 4.66400195774533, + "grad_norm": 0.0003850639332085848, + "learning_rate": 0.05631477305803728, + "loss": 0.312, + "num_input_tokens_seen": 59717952, + "step": 28590 + }, + { + "epoch": 4.664817684966147, + "grad_norm": 0.0007122641545720398, + "learning_rate": 0.05626877733576462, + "loss": 0.3357, + "num_input_tokens_seen": 59728832, + "step": 28595 + }, + { + "epoch": 4.665633412186965, + "grad_norm": 0.0007564148982055485, + "learning_rate": 0.05622279606802435, + "loss": 0.3098, + "num_input_tokens_seen": 59739680, + "step": 28600 + }, + { + "epoch": 4.665633412186965, + "eval_loss": 0.31601062417030334, + "eval_runtime": 156.0892, + "eval_samples_per_second": 17.458, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 59739680, + "step": 28600 + }, + { + "epoch": 4.666449139407782, + "grad_norm": 0.0007065091049298644, + "learning_rate": 0.05617682926190744, + "loss": 0.3055, + "num_input_tokens_seen": 59751520, + "step": 28605 + }, + { + "epoch": 4.667264866628599, + "grad_norm": 0.0005177210550755262, + "learning_rate": 0.05613087692450248, + "loss": 0.3438, + "num_input_tokens_seen": 59762512, + "step": 28610 + }, + { + "epoch": 4.6680805938494165, + "grad_norm": 0.0007463916554115713, + "learning_rate": 0.05608493906289592, + "loss": 0.3426, + "num_input_tokens_seen": 59773120, + "step": 28615 + }, + { + "epoch": 4.668896321070234, + "grad_norm": 0.0006722761318087578, + "learning_rate": 0.05603901568417201, + "loss": 0.2929, + "num_input_tokens_seen": 59783056, + "step": 28620 + }, + { + "epoch": 4.669712048291052, + "grad_norm": 0.00040721200639382005, + "learning_rate": 0.055993106795412625, + "loss": 0.2914, + "num_input_tokens_seen": 59793744, + "step": 28625 + }, + { + "epoch": 4.670527775511869, + "grad_norm": 0.0005578635027632117, + "learning_rate": 0.05594721240369759, + "loss": 0.3271, + "num_input_tokens_seen": 59804752, + "step": 28630 + }, + { + "epoch": 4.671343502732686, + "grad_norm": 0.00047490501310676336, + "learning_rate": 0.055901332516104296, + "loss": 0.2941, + "num_input_tokens_seen": 59815392, + "step": 28635 + }, + { + "epoch": 4.672159229953504, + "grad_norm": 0.0004544336989056319, + "learning_rate": 0.05585546713970804, + "loss": 0.3275, + "num_input_tokens_seen": 59826784, + "step": 28640 + }, + { + "epoch": 4.672974957174321, + "grad_norm": 0.0005231259274296463, + "learning_rate": 0.05580961628158189, + "loss": 0.3318, + "num_input_tokens_seen": 59836560, + "step": 28645 + }, + { + "epoch": 4.673790684395138, + "grad_norm": 0.0005036260699853301, + "learning_rate": 0.05576377994879659, + "loss": 0.3461, + "num_input_tokens_seen": 59846848, + "step": 28650 + }, + { + "epoch": 4.674606411615955, + "grad_norm": 0.00035158824175596237, + "learning_rate": 0.05571795814842063, + "loss": 0.3097, + "num_input_tokens_seen": 59857152, + "step": 28655 + }, + { + "epoch": 4.6754221388367725, + "grad_norm": 0.0011605450417846441, + "learning_rate": 0.05567215088752037, + "loss": 0.2674, + "num_input_tokens_seen": 59867328, + "step": 28660 + }, + { + "epoch": 4.6762378660575905, + "grad_norm": 0.0007682900759391487, + "learning_rate": 0.05562635817315981, + "loss": 0.3105, + "num_input_tokens_seen": 59877776, + "step": 28665 + }, + { + "epoch": 4.677053593278408, + "grad_norm": 0.0004004265647381544, + "learning_rate": 0.05558058001240083, + "loss": 0.3203, + "num_input_tokens_seen": 59887696, + "step": 28670 + }, + { + "epoch": 4.677869320499225, + "grad_norm": 0.0005049781175330281, + "learning_rate": 0.055534816412302915, + "loss": 0.2996, + "num_input_tokens_seen": 59897952, + "step": 28675 + }, + { + "epoch": 4.678685047720043, + "grad_norm": 0.000518265413120389, + "learning_rate": 0.055489067379923436, + "loss": 0.3193, + "num_input_tokens_seen": 59908192, + "step": 28680 + }, + { + "epoch": 4.67950077494086, + "grad_norm": 0.0006867299089208245, + "learning_rate": 0.055443332922317505, + "loss": 0.3684, + "num_input_tokens_seen": 59918096, + "step": 28685 + }, + { + "epoch": 4.680316502161677, + "grad_norm": 0.0003396507236175239, + "learning_rate": 0.055397613046537876, + "loss": 0.3574, + "num_input_tokens_seen": 59928672, + "step": 28690 + }, + { + "epoch": 4.681132229382494, + "grad_norm": 0.0005555694806389511, + "learning_rate": 0.055351907759635145, + "loss": 0.2594, + "num_input_tokens_seen": 59938976, + "step": 28695 + }, + { + "epoch": 4.681947956603312, + "grad_norm": 0.0004884928930550814, + "learning_rate": 0.05530621706865772, + "loss": 0.3422, + "num_input_tokens_seen": 59948224, + "step": 28700 + }, + { + "epoch": 4.682763683824129, + "grad_norm": 0.0007369964732788503, + "learning_rate": 0.055260540980651564, + "loss": 0.3325, + "num_input_tokens_seen": 59956928, + "step": 28705 + }, + { + "epoch": 4.6835794110449465, + "grad_norm": 0.00024284869141411036, + "learning_rate": 0.05521487950266062, + "loss": 0.374, + "num_input_tokens_seen": 59967552, + "step": 28710 + }, + { + "epoch": 4.684395138265764, + "grad_norm": 0.00044791382970288396, + "learning_rate": 0.055169232641726344, + "loss": 0.2429, + "num_input_tokens_seen": 59979072, + "step": 28715 + }, + { + "epoch": 4.685210865486582, + "grad_norm": 0.0004687142209149897, + "learning_rate": 0.055123600404888166, + "loss": 0.2991, + "num_input_tokens_seen": 59989504, + "step": 28720 + }, + { + "epoch": 4.686026592707399, + "grad_norm": 0.0004825633659493178, + "learning_rate": 0.05507798279918309, + "loss": 0.2624, + "num_input_tokens_seen": 59999856, + "step": 28725 + }, + { + "epoch": 4.686842319928216, + "grad_norm": 0.00048356890329159796, + "learning_rate": 0.0550323798316459, + "loss": 0.3302, + "num_input_tokens_seen": 60011200, + "step": 28730 + }, + { + "epoch": 4.687658047149033, + "grad_norm": 0.000517644512001425, + "learning_rate": 0.05498679150930916, + "loss": 0.2811, + "num_input_tokens_seen": 60021824, + "step": 28735 + }, + { + "epoch": 4.688473774369851, + "grad_norm": 0.0007031125132925808, + "learning_rate": 0.05494121783920323, + "loss": 0.3434, + "num_input_tokens_seen": 60031136, + "step": 28740 + }, + { + "epoch": 4.689289501590668, + "grad_norm": 0.0006728919688612223, + "learning_rate": 0.05489565882835605, + "loss": 0.307, + "num_input_tokens_seen": 60041344, + "step": 28745 + }, + { + "epoch": 4.690105228811485, + "grad_norm": 0.00047960568917915225, + "learning_rate": 0.05485011448379348, + "loss": 0.2771, + "num_input_tokens_seen": 60052272, + "step": 28750 + }, + { + "epoch": 4.690920956032302, + "grad_norm": 0.00044191969209350646, + "learning_rate": 0.05480458481253893, + "loss": 0.3175, + "num_input_tokens_seen": 60063376, + "step": 28755 + }, + { + "epoch": 4.69173668325312, + "grad_norm": 0.0005593036185018718, + "learning_rate": 0.054759069821613715, + "loss": 0.4007, + "num_input_tokens_seen": 60073776, + "step": 28760 + }, + { + "epoch": 4.692552410473938, + "grad_norm": 0.0008108899928629398, + "learning_rate": 0.05471356951803683, + "loss": 0.3208, + "num_input_tokens_seen": 60084032, + "step": 28765 + }, + { + "epoch": 4.693368137694755, + "grad_norm": 0.0005970069323666394, + "learning_rate": 0.054668083908824945, + "loss": 0.2108, + "num_input_tokens_seen": 60094944, + "step": 28770 + }, + { + "epoch": 4.694183864915573, + "grad_norm": 0.0007614505011588335, + "learning_rate": 0.054622613000992526, + "loss": 0.3722, + "num_input_tokens_seen": 60106592, + "step": 28775 + }, + { + "epoch": 4.69499959213639, + "grad_norm": 0.0008580884896218777, + "learning_rate": 0.05457715680155182, + "loss": 0.2496, + "num_input_tokens_seen": 60116976, + "step": 28780 + }, + { + "epoch": 4.695815319357207, + "grad_norm": 0.00048438014346174896, + "learning_rate": 0.05453171531751265, + "loss": 0.3239, + "num_input_tokens_seen": 60128064, + "step": 28785 + }, + { + "epoch": 4.696631046578024, + "grad_norm": 0.0004677229735534638, + "learning_rate": 0.05448628855588276, + "loss": 0.3242, + "num_input_tokens_seen": 60139104, + "step": 28790 + }, + { + "epoch": 4.697446773798841, + "grad_norm": 0.0005420728120952845, + "learning_rate": 0.05444087652366746, + "loss": 0.2911, + "num_input_tokens_seen": 60148912, + "step": 28795 + }, + { + "epoch": 4.698262501019659, + "grad_norm": 0.0003461106971371919, + "learning_rate": 0.05439547922786984, + "loss": 0.4193, + "num_input_tokens_seen": 60159984, + "step": 28800 + }, + { + "epoch": 4.698262501019659, + "eval_loss": 0.314362496137619, + "eval_runtime": 156.1364, + "eval_samples_per_second": 17.453, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 60159984, + "step": 28800 + }, + { + "epoch": 4.699078228240476, + "grad_norm": 0.00028505720547400415, + "learning_rate": 0.0543500966754908, + "loss": 0.3003, + "num_input_tokens_seen": 60169904, + "step": 28805 + }, + { + "epoch": 4.6998939554612935, + "grad_norm": 0.0005345356767065823, + "learning_rate": 0.05430472887352882, + "loss": 0.3462, + "num_input_tokens_seen": 60181232, + "step": 28810 + }, + { + "epoch": 4.7007096826821115, + "grad_norm": 0.00039761431980878115, + "learning_rate": 0.05425937582898023, + "loss": 0.3478, + "num_input_tokens_seen": 60192144, + "step": 28815 + }, + { + "epoch": 4.701525409902929, + "grad_norm": 0.0005932115600444376, + "learning_rate": 0.054214037548839085, + "loss": 0.3121, + "num_input_tokens_seen": 60202992, + "step": 28820 + }, + { + "epoch": 4.702341137123746, + "grad_norm": 0.000722425349522382, + "learning_rate": 0.05416871404009703, + "loss": 0.3738, + "num_input_tokens_seen": 60212768, + "step": 28825 + }, + { + "epoch": 4.703156864344563, + "grad_norm": 0.0005892776534892619, + "learning_rate": 0.054123405309743605, + "loss": 0.2808, + "num_input_tokens_seen": 60224016, + "step": 28830 + }, + { + "epoch": 4.70397259156538, + "grad_norm": 0.00036080574500374496, + "learning_rate": 0.0540781113647659, + "loss": 0.2956, + "num_input_tokens_seen": 60234128, + "step": 28835 + }, + { + "epoch": 4.704788318786198, + "grad_norm": 0.0006279073422774673, + "learning_rate": 0.054032832212148836, + "loss": 0.2826, + "num_input_tokens_seen": 60244272, + "step": 28840 + }, + { + "epoch": 4.705604046007015, + "grad_norm": 0.0005234169075265527, + "learning_rate": 0.0539875678588751, + "loss": 0.2974, + "num_input_tokens_seen": 60255376, + "step": 28845 + }, + { + "epoch": 4.706419773227832, + "grad_norm": 0.0005061023402959108, + "learning_rate": 0.05394231831192492, + "loss": 0.3188, + "num_input_tokens_seen": 60265616, + "step": 28850 + }, + { + "epoch": 4.70723550044865, + "grad_norm": 0.0003658255736809224, + "learning_rate": 0.05389708357827639, + "loss": 0.3247, + "num_input_tokens_seen": 60276656, + "step": 28855 + }, + { + "epoch": 4.7080512276694675, + "grad_norm": 0.0005649380036629736, + "learning_rate": 0.05385186366490533, + "loss": 0.3552, + "num_input_tokens_seen": 60286624, + "step": 28860 + }, + { + "epoch": 4.708866954890285, + "grad_norm": 0.000609665468800813, + "learning_rate": 0.053806658578785166, + "loss": 0.3661, + "num_input_tokens_seen": 60297120, + "step": 28865 + }, + { + "epoch": 4.709682682111102, + "grad_norm": 0.0004928806447423995, + "learning_rate": 0.05376146832688705, + "loss": 0.2646, + "num_input_tokens_seen": 60309456, + "step": 28870 + }, + { + "epoch": 4.71049840933192, + "grad_norm": 0.00035321939503774047, + "learning_rate": 0.053716292916179964, + "loss": 0.2537, + "num_input_tokens_seen": 60319680, + "step": 28875 + }, + { + "epoch": 4.711314136552737, + "grad_norm": 0.000459537171991542, + "learning_rate": 0.05367113235363045, + "loss": 0.3257, + "num_input_tokens_seen": 60331120, + "step": 28880 + }, + { + "epoch": 4.712129863773554, + "grad_norm": 0.0005116679822094738, + "learning_rate": 0.05362598664620289, + "loss": 0.3655, + "num_input_tokens_seen": 60340560, + "step": 28885 + }, + { + "epoch": 4.712945590994371, + "grad_norm": 0.00026018390781246126, + "learning_rate": 0.053580855800859285, + "loss": 0.2958, + "num_input_tokens_seen": 60349776, + "step": 28890 + }, + { + "epoch": 4.713761318215189, + "grad_norm": 0.0007358406437560916, + "learning_rate": 0.05353573982455938, + "loss": 0.332, + "num_input_tokens_seen": 60359824, + "step": 28895 + }, + { + "epoch": 4.714577045436006, + "grad_norm": 0.0004770316882058978, + "learning_rate": 0.053490638724260686, + "loss": 0.3131, + "num_input_tokens_seen": 60369952, + "step": 28900 + }, + { + "epoch": 4.715392772656823, + "grad_norm": 0.0007244322914630175, + "learning_rate": 0.05344555250691827, + "loss": 0.3178, + "num_input_tokens_seen": 60380368, + "step": 28905 + }, + { + "epoch": 4.716208499877641, + "grad_norm": 0.0006766328006051481, + "learning_rate": 0.053400481179485086, + "loss": 0.3742, + "num_input_tokens_seen": 60390400, + "step": 28910 + }, + { + "epoch": 4.717024227098459, + "grad_norm": 0.0005186509806662798, + "learning_rate": 0.05335542474891159, + "loss": 0.3331, + "num_input_tokens_seen": 60402256, + "step": 28915 + }, + { + "epoch": 4.717839954319276, + "grad_norm": 0.0007042267825454473, + "learning_rate": 0.053310383222146124, + "loss": 0.3043, + "num_input_tokens_seen": 60412128, + "step": 28920 + }, + { + "epoch": 4.718655681540093, + "grad_norm": 0.0004486609832383692, + "learning_rate": 0.053265356606134684, + "loss": 0.3384, + "num_input_tokens_seen": 60421648, + "step": 28925 + }, + { + "epoch": 4.71947140876091, + "grad_norm": 0.00041363731725141406, + "learning_rate": 0.053220344907820856, + "loss": 0.3409, + "num_input_tokens_seen": 60432384, + "step": 28930 + }, + { + "epoch": 4.720287135981728, + "grad_norm": 0.0003929815429728478, + "learning_rate": 0.05317534813414608, + "loss": 0.3724, + "num_input_tokens_seen": 60441088, + "step": 28935 + }, + { + "epoch": 4.721102863202545, + "grad_norm": 0.00044470230932347476, + "learning_rate": 0.05313036629204942, + "loss": 0.2917, + "num_input_tokens_seen": 60451392, + "step": 28940 + }, + { + "epoch": 4.721918590423362, + "grad_norm": 0.0006500431918539107, + "learning_rate": 0.05308539938846756, + "loss": 0.3844, + "num_input_tokens_seen": 60461024, + "step": 28945 + }, + { + "epoch": 4.72273431764418, + "grad_norm": 0.0006282167159952223, + "learning_rate": 0.05304044743033507, + "loss": 0.3549, + "num_input_tokens_seen": 60471680, + "step": 28950 + }, + { + "epoch": 4.723550044864997, + "grad_norm": 0.0007960713119246066, + "learning_rate": 0.05299551042458401, + "loss": 0.2917, + "num_input_tokens_seen": 60483488, + "step": 28955 + }, + { + "epoch": 4.7243657720858145, + "grad_norm": 0.0003939070156775415, + "learning_rate": 0.052950588378144266, + "loss": 0.3346, + "num_input_tokens_seen": 60494080, + "step": 28960 + }, + { + "epoch": 4.725181499306632, + "grad_norm": 0.00040068215457722545, + "learning_rate": 0.052905681297943465, + "loss": 0.3414, + "num_input_tokens_seen": 60504592, + "step": 28965 + }, + { + "epoch": 4.725997226527449, + "grad_norm": 0.00047620231634937227, + "learning_rate": 0.0528607891909067, + "loss": 0.3233, + "num_input_tokens_seen": 60515984, + "step": 28970 + }, + { + "epoch": 4.726812953748267, + "grad_norm": 0.00041688751662150025, + "learning_rate": 0.05281591206395697, + "loss": 0.3261, + "num_input_tokens_seen": 60526352, + "step": 28975 + }, + { + "epoch": 4.727628680969084, + "grad_norm": 0.0008051852928474545, + "learning_rate": 0.05277104992401496, + "loss": 0.3092, + "num_input_tokens_seen": 60535920, + "step": 28980 + }, + { + "epoch": 4.728444408189901, + "grad_norm": 0.001117694191634655, + "learning_rate": 0.05272620277799884, + "loss": 0.3495, + "num_input_tokens_seen": 60546976, + "step": 28985 + }, + { + "epoch": 4.729260135410719, + "grad_norm": 0.0008874270715750754, + "learning_rate": 0.05268137063282473, + "loss": 0.2878, + "num_input_tokens_seen": 60557520, + "step": 28990 + }, + { + "epoch": 4.730075862631536, + "grad_norm": 0.0005602039746008813, + "learning_rate": 0.0526365534954062, + "loss": 0.3299, + "num_input_tokens_seen": 60568208, + "step": 28995 + }, + { + "epoch": 4.730891589852353, + "grad_norm": 0.0010774100665003061, + "learning_rate": 0.052591751372654656, + "loss": 0.3198, + "num_input_tokens_seen": 60579344, + "step": 29000 + }, + { + "epoch": 4.730891589852353, + "eval_loss": 0.31704726815223694, + "eval_runtime": 156.0993, + "eval_samples_per_second": 17.457, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 60579344, + "step": 29000 + }, + { + "epoch": 4.7317073170731705, + "grad_norm": 0.0003370216873008758, + "learning_rate": 0.05254696427147921, + "loss": 0.3297, + "num_input_tokens_seen": 60590496, + "step": 29005 + }, + { + "epoch": 4.732523044293988, + "grad_norm": 0.00047079692012630403, + "learning_rate": 0.052502192198786546, + "loss": 0.2879, + "num_input_tokens_seen": 60600928, + "step": 29010 + }, + { + "epoch": 4.733338771514806, + "grad_norm": 0.0005156921106390655, + "learning_rate": 0.05245743516148103, + "loss": 0.3207, + "num_input_tokens_seen": 60609712, + "step": 29015 + }, + { + "epoch": 4.734154498735623, + "grad_norm": 0.00033655340666882694, + "learning_rate": 0.05241269316646486, + "loss": 0.3075, + "num_input_tokens_seen": 60620160, + "step": 29020 + }, + { + "epoch": 4.73497022595644, + "grad_norm": 0.00033528151107020676, + "learning_rate": 0.052367966220637725, + "loss": 0.3098, + "num_input_tokens_seen": 60630288, + "step": 29025 + }, + { + "epoch": 4.735785953177258, + "grad_norm": 0.0006347715971060097, + "learning_rate": 0.05232325433089716, + "loss": 0.3461, + "num_input_tokens_seen": 60640128, + "step": 29030 + }, + { + "epoch": 4.736601680398075, + "grad_norm": 0.00046063167974352837, + "learning_rate": 0.052278557504138214, + "loss": 0.4291, + "num_input_tokens_seen": 60649488, + "step": 29035 + }, + { + "epoch": 4.737417407618892, + "grad_norm": 0.0006626800750382245, + "learning_rate": 0.05223387574725372, + "loss": 0.3362, + "num_input_tokens_seen": 60660448, + "step": 29040 + }, + { + "epoch": 4.738233134839709, + "grad_norm": 0.000423479825258255, + "learning_rate": 0.05218920906713428, + "loss": 0.3372, + "num_input_tokens_seen": 60671168, + "step": 29045 + }, + { + "epoch": 4.739048862060527, + "grad_norm": 0.0004416152078192681, + "learning_rate": 0.05214455747066789, + "loss": 0.3123, + "num_input_tokens_seen": 60679776, + "step": 29050 + }, + { + "epoch": 4.7398645892813445, + "grad_norm": 0.0007441012421622872, + "learning_rate": 0.05209992096474048, + "loss": 0.328, + "num_input_tokens_seen": 60689792, + "step": 29055 + }, + { + "epoch": 4.740680316502162, + "grad_norm": 0.0004939091741107404, + "learning_rate": 0.05205529955623559, + "loss": 0.3258, + "num_input_tokens_seen": 60701568, + "step": 29060 + }, + { + "epoch": 4.741496043722979, + "grad_norm": 0.0005962155410088599, + "learning_rate": 0.052010693252034314, + "loss": 0.3362, + "num_input_tokens_seen": 60711200, + "step": 29065 + }, + { + "epoch": 4.742311770943797, + "grad_norm": 0.00037043652264401317, + "learning_rate": 0.0519661020590156, + "loss": 0.3325, + "num_input_tokens_seen": 60721456, + "step": 29070 + }, + { + "epoch": 4.743127498164614, + "grad_norm": 0.0005553844966925681, + "learning_rate": 0.05192152598405586, + "loss": 0.3063, + "num_input_tokens_seen": 60730432, + "step": 29075 + }, + { + "epoch": 4.743943225385431, + "grad_norm": 0.00029810736305080354, + "learning_rate": 0.05187696503402941, + "loss": 0.3117, + "num_input_tokens_seen": 60740768, + "step": 29080 + }, + { + "epoch": 4.744758952606248, + "grad_norm": 0.0002753867011051625, + "learning_rate": 0.05183241921580798, + "loss": 0.328, + "num_input_tokens_seen": 60751680, + "step": 29085 + }, + { + "epoch": 4.745574679827066, + "grad_norm": 0.0005826213746331632, + "learning_rate": 0.051787888536261206, + "loss": 0.3467, + "num_input_tokens_seen": 60762928, + "step": 29090 + }, + { + "epoch": 4.746390407047883, + "grad_norm": 0.0004742025048471987, + "learning_rate": 0.051743373002256184, + "loss": 0.3202, + "num_input_tokens_seen": 60772320, + "step": 29095 + }, + { + "epoch": 4.7472061342687, + "grad_norm": 0.0017282812623307109, + "learning_rate": 0.05169887262065787, + "loss": 0.3206, + "num_input_tokens_seen": 60782304, + "step": 29100 + }, + { + "epoch": 4.7480218614895175, + "grad_norm": 0.0005210715462453663, + "learning_rate": 0.051654387398328665, + "loss": 0.31, + "num_input_tokens_seen": 60793824, + "step": 29105 + }, + { + "epoch": 4.748837588710336, + "grad_norm": 0.0004647592722903937, + "learning_rate": 0.05160991734212888, + "loss": 0.2896, + "num_input_tokens_seen": 60803936, + "step": 29110 + }, + { + "epoch": 4.749653315931153, + "grad_norm": 0.00039865297731012106, + "learning_rate": 0.051565462458916224, + "loss": 0.2934, + "num_input_tokens_seen": 60814064, + "step": 29115 + }, + { + "epoch": 4.75046904315197, + "grad_norm": 0.00044762156903743744, + "learning_rate": 0.05152102275554627, + "loss": 0.2904, + "num_input_tokens_seen": 60825728, + "step": 29120 + }, + { + "epoch": 4.751284770372787, + "grad_norm": 0.00040416119736619294, + "learning_rate": 0.05147659823887222, + "loss": 0.3291, + "num_input_tokens_seen": 60836592, + "step": 29125 + }, + { + "epoch": 4.752100497593605, + "grad_norm": 0.000413714034948498, + "learning_rate": 0.05143218891574479, + "loss": 0.3009, + "num_input_tokens_seen": 60846224, + "step": 29130 + }, + { + "epoch": 4.752916224814422, + "grad_norm": 0.0004053845477756113, + "learning_rate": 0.0513877947930125, + "loss": 0.3555, + "num_input_tokens_seen": 60856576, + "step": 29135 + }, + { + "epoch": 4.753731952035239, + "grad_norm": 0.0010924128582701087, + "learning_rate": 0.051343415877521566, + "loss": 0.3448, + "num_input_tokens_seen": 60866896, + "step": 29140 + }, + { + "epoch": 4.754547679256056, + "grad_norm": 0.00036627339432016015, + "learning_rate": 0.051299052176115634, + "loss": 0.305, + "num_input_tokens_seen": 60878880, + "step": 29145 + }, + { + "epoch": 4.755363406476874, + "grad_norm": 0.000486369855934754, + "learning_rate": 0.051254703695636256, + "loss": 0.3546, + "num_input_tokens_seen": 60890768, + "step": 29150 + }, + { + "epoch": 4.7561791336976915, + "grad_norm": 0.00032211310463026166, + "learning_rate": 0.05121037044292249, + "loss": 0.306, + "num_input_tokens_seen": 60900944, + "step": 29155 + }, + { + "epoch": 4.756994860918509, + "grad_norm": 0.0005989089841023088, + "learning_rate": 0.05116605242481101, + "loss": 0.324, + "num_input_tokens_seen": 60911200, + "step": 29160 + }, + { + "epoch": 4.757810588139327, + "grad_norm": 0.0006572072743438184, + "learning_rate": 0.05112174964813634, + "loss": 0.291, + "num_input_tokens_seen": 60919824, + "step": 29165 + }, + { + "epoch": 4.758626315360144, + "grad_norm": 0.000299772247672081, + "learning_rate": 0.05107746211973038, + "loss": 0.2963, + "num_input_tokens_seen": 60930640, + "step": 29170 + }, + { + "epoch": 4.759442042580961, + "grad_norm": 0.0005814430769532919, + "learning_rate": 0.05103318984642291, + "loss": 0.3375, + "num_input_tokens_seen": 60941984, + "step": 29175 + }, + { + "epoch": 4.760257769801778, + "grad_norm": 0.00036873132921755314, + "learning_rate": 0.05098893283504131, + "loss": 0.3371, + "num_input_tokens_seen": 60952032, + "step": 29180 + }, + { + "epoch": 4.761073497022595, + "grad_norm": 0.0005413482431322336, + "learning_rate": 0.050944691092410475, + "loss": 0.3042, + "num_input_tokens_seen": 60962800, + "step": 29185 + }, + { + "epoch": 4.761889224243413, + "grad_norm": 0.00042002691770903766, + "learning_rate": 0.05090046462535313, + "loss": 0.3055, + "num_input_tokens_seen": 60972320, + "step": 29190 + }, + { + "epoch": 4.76270495146423, + "grad_norm": 0.0006365273729898036, + "learning_rate": 0.050856253440689454, + "loss": 0.3132, + "num_input_tokens_seen": 60983072, + "step": 29195 + }, + { + "epoch": 4.7635206786850475, + "grad_norm": 0.0005321009666658938, + "learning_rate": 0.050812057545237405, + "loss": 0.2599, + "num_input_tokens_seen": 60992736, + "step": 29200 + }, + { + "epoch": 4.7635206786850475, + "eval_loss": 0.31774213910102844, + "eval_runtime": 156.0835, + "eval_samples_per_second": 17.459, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 60992736, + "step": 29200 + }, + { + "epoch": 4.7643364059058655, + "grad_norm": 0.001050389837473631, + "learning_rate": 0.0507678769458126, + "loss": 0.2948, + "num_input_tokens_seen": 61003536, + "step": 29205 + }, + { + "epoch": 4.765152133126683, + "grad_norm": 0.0005643151816911995, + "learning_rate": 0.050723711649228155, + "loss": 0.3595, + "num_input_tokens_seen": 61013056, + "step": 29210 + }, + { + "epoch": 4.7659678603475, + "grad_norm": 0.0009357279050163925, + "learning_rate": 0.05067956166229496, + "loss": 0.331, + "num_input_tokens_seen": 61022816, + "step": 29215 + }, + { + "epoch": 4.766783587568317, + "grad_norm": 0.0007363299955613911, + "learning_rate": 0.05063542699182155, + "loss": 0.3275, + "num_input_tokens_seen": 61033536, + "step": 29220 + }, + { + "epoch": 4.767599314789135, + "grad_norm": 0.0005172114470042288, + "learning_rate": 0.050591307644613996, + "loss": 0.33, + "num_input_tokens_seen": 61043920, + "step": 29225 + }, + { + "epoch": 4.768415042009952, + "grad_norm": 0.00038138090167194605, + "learning_rate": 0.05054720362747599, + "loss": 0.3352, + "num_input_tokens_seen": 61053872, + "step": 29230 + }, + { + "epoch": 4.769230769230769, + "grad_norm": 0.000493427156470716, + "learning_rate": 0.050503114947209035, + "loss": 0.3141, + "num_input_tokens_seen": 61065536, + "step": 29235 + }, + { + "epoch": 4.770046496451586, + "grad_norm": 0.0003196925681550056, + "learning_rate": 0.05045904161061207, + "loss": 0.335, + "num_input_tokens_seen": 61075280, + "step": 29240 + }, + { + "epoch": 4.770862223672404, + "grad_norm": 0.0004910837160423398, + "learning_rate": 0.05041498362448185, + "loss": 0.3059, + "num_input_tokens_seen": 61085776, + "step": 29245 + }, + { + "epoch": 4.771677950893221, + "grad_norm": 0.0005033467314206064, + "learning_rate": 0.05037094099561256, + "loss": 0.326, + "num_input_tokens_seen": 61095328, + "step": 29250 + }, + { + "epoch": 4.772493678114039, + "grad_norm": 0.00030962080927565694, + "learning_rate": 0.05032691373079624, + "loss": 0.2854, + "num_input_tokens_seen": 61105744, + "step": 29255 + }, + { + "epoch": 4.773309405334856, + "grad_norm": 0.0005830866284668446, + "learning_rate": 0.05028290183682234, + "loss": 0.3329, + "num_input_tokens_seen": 61115200, + "step": 29260 + }, + { + "epoch": 4.774125132555674, + "grad_norm": 0.000470676546683535, + "learning_rate": 0.050238905320478096, + "loss": 0.3266, + "num_input_tokens_seen": 61125024, + "step": 29265 + }, + { + "epoch": 4.774940859776491, + "grad_norm": 0.0005459269741550088, + "learning_rate": 0.05019492418854838, + "loss": 0.3647, + "num_input_tokens_seen": 61136960, + "step": 29270 + }, + { + "epoch": 4.775756586997308, + "grad_norm": 0.0005134140374138951, + "learning_rate": 0.05015095844781554, + "loss": 0.324, + "num_input_tokens_seen": 61147888, + "step": 29275 + }, + { + "epoch": 4.776572314218125, + "grad_norm": 0.0003367543686181307, + "learning_rate": 0.05010700810505968, + "loss": 0.299, + "num_input_tokens_seen": 61159184, + "step": 29280 + }, + { + "epoch": 4.777388041438943, + "grad_norm": 0.0004112502501811832, + "learning_rate": 0.05006307316705856, + "loss": 0.3474, + "num_input_tokens_seen": 61169680, + "step": 29285 + }, + { + "epoch": 4.77820376865976, + "grad_norm": 0.0003611066786106676, + "learning_rate": 0.0500191536405874, + "loss": 0.2947, + "num_input_tokens_seen": 61179376, + "step": 29290 + }, + { + "epoch": 4.779019495880577, + "grad_norm": 0.0003978620807174593, + "learning_rate": 0.04997524953241922, + "loss": 0.2496, + "num_input_tokens_seen": 61189840, + "step": 29295 + }, + { + "epoch": 4.7798352231013945, + "grad_norm": 0.0005882870173081756, + "learning_rate": 0.049931360849324556, + "loss": 0.2802, + "num_input_tokens_seen": 61199856, + "step": 29300 + }, + { + "epoch": 4.7806509503222125, + "grad_norm": 0.00026678384165279567, + "learning_rate": 0.04988748759807155, + "loss": 0.2787, + "num_input_tokens_seen": 61209936, + "step": 29305 + }, + { + "epoch": 4.78146667754303, + "grad_norm": 0.0005186992348171771, + "learning_rate": 0.0498436297854261, + "loss": 0.2962, + "num_input_tokens_seen": 61220880, + "step": 29310 + }, + { + "epoch": 4.782282404763847, + "grad_norm": 0.0006851070793345571, + "learning_rate": 0.04979978741815152, + "loss": 0.3552, + "num_input_tokens_seen": 61232064, + "step": 29315 + }, + { + "epoch": 4.783098131984664, + "grad_norm": 0.0003492686664685607, + "learning_rate": 0.04975596050300891, + "loss": 0.3406, + "num_input_tokens_seen": 61241216, + "step": 29320 + }, + { + "epoch": 4.783913859205482, + "grad_norm": 0.0008608084172010422, + "learning_rate": 0.049712149046757005, + "loss": 0.3534, + "num_input_tokens_seen": 61250272, + "step": 29325 + }, + { + "epoch": 4.784729586426299, + "grad_norm": 0.001214818563312292, + "learning_rate": 0.04966835305615194, + "loss": 0.3422, + "num_input_tokens_seen": 61261872, + "step": 29330 + }, + { + "epoch": 4.785545313647116, + "grad_norm": 0.000372574373614043, + "learning_rate": 0.049624572537947755, + "loss": 0.3283, + "num_input_tokens_seen": 61272016, + "step": 29335 + }, + { + "epoch": 4.786361040867934, + "grad_norm": 0.0003585568629205227, + "learning_rate": 0.04958080749889582, + "loss": 0.3139, + "num_input_tokens_seen": 61283808, + "step": 29340 + }, + { + "epoch": 4.787176768088751, + "grad_norm": 0.0005340364295989275, + "learning_rate": 0.049537057945745304, + "loss": 0.3298, + "num_input_tokens_seen": 61295744, + "step": 29345 + }, + { + "epoch": 4.7879924953095685, + "grad_norm": 0.0006202636286616325, + "learning_rate": 0.049493323885243, + "loss": 0.3187, + "num_input_tokens_seen": 61307520, + "step": 29350 + }, + { + "epoch": 4.788808222530386, + "grad_norm": 0.00033395408536307514, + "learning_rate": 0.04944960532413318, + "loss": 0.3006, + "num_input_tokens_seen": 61318064, + "step": 29355 + }, + { + "epoch": 4.789623949751203, + "grad_norm": 0.000534099293872714, + "learning_rate": 0.049405902269157774, + "loss": 0.2915, + "num_input_tokens_seen": 61329200, + "step": 29360 + }, + { + "epoch": 4.790439676972021, + "grad_norm": 0.0005332050495781004, + "learning_rate": 0.04936221472705646, + "loss": 0.3026, + "num_input_tokens_seen": 61338976, + "step": 29365 + }, + { + "epoch": 4.791255404192838, + "grad_norm": 0.00028300995472818613, + "learning_rate": 0.04931854270456632, + "loss": 0.3295, + "num_input_tokens_seen": 61350368, + "step": 29370 + }, + { + "epoch": 4.792071131413655, + "grad_norm": 0.0004913609009236097, + "learning_rate": 0.049274886208422075, + "loss": 0.2753, + "num_input_tokens_seen": 61360192, + "step": 29375 + }, + { + "epoch": 4.792886858634473, + "grad_norm": 0.00036977598210796714, + "learning_rate": 0.049231245245356235, + "loss": 0.3462, + "num_input_tokens_seen": 61370912, + "step": 29380 + }, + { + "epoch": 4.79370258585529, + "grad_norm": 0.0005808906862512231, + "learning_rate": 0.049187619822098655, + "loss": 0.351, + "num_input_tokens_seen": 61381712, + "step": 29385 + }, + { + "epoch": 4.794518313076107, + "grad_norm": 0.00032057889620773494, + "learning_rate": 0.04914400994537705, + "loss": 0.2855, + "num_input_tokens_seen": 61392752, + "step": 29390 + }, + { + "epoch": 4.7953340402969244, + "grad_norm": 0.0002563381567597389, + "learning_rate": 0.049100415621916485, + "loss": 0.396, + "num_input_tokens_seen": 61405296, + "step": 29395 + }, + { + "epoch": 4.796149767517742, + "grad_norm": 0.0005042675184085965, + "learning_rate": 0.04905683685843981, + "loss": 0.2971, + "num_input_tokens_seen": 61414080, + "step": 29400 + }, + { + "epoch": 4.796149767517742, + "eval_loss": 0.3152347207069397, + "eval_runtime": 155.8222, + "eval_samples_per_second": 17.488, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 61414080, + "step": 29400 + }, + { + "epoch": 4.79696549473856, + "grad_norm": 0.00044317045831121504, + "learning_rate": 0.049013273661667495, + "loss": 0.2782, + "num_input_tokens_seen": 61423616, + "step": 29405 + }, + { + "epoch": 4.797781221959377, + "grad_norm": 0.0007735502440482378, + "learning_rate": 0.048969726038317396, + "loss": 0.2882, + "num_input_tokens_seen": 61434048, + "step": 29410 + }, + { + "epoch": 4.798596949180194, + "grad_norm": 0.0003241390222683549, + "learning_rate": 0.048926193995105206, + "loss": 0.3334, + "num_input_tokens_seen": 61442848, + "step": 29415 + }, + { + "epoch": 4.799412676401012, + "grad_norm": 0.0005415347986854613, + "learning_rate": 0.048882677538744035, + "loss": 0.3085, + "num_input_tokens_seen": 61453664, + "step": 29420 + }, + { + "epoch": 4.800228403621829, + "grad_norm": 0.0004593120829667896, + "learning_rate": 0.048839176675944715, + "loss": 0.2833, + "num_input_tokens_seen": 61464880, + "step": 29425 + }, + { + "epoch": 4.801044130842646, + "grad_norm": 0.0003674389736261219, + "learning_rate": 0.04879569141341566, + "loss": 0.3726, + "num_input_tokens_seen": 61474464, + "step": 29430 + }, + { + "epoch": 4.801859858063463, + "grad_norm": 0.0003486679634079337, + "learning_rate": 0.04875222175786274, + "loss": 0.2841, + "num_input_tokens_seen": 61483520, + "step": 29435 + }, + { + "epoch": 4.802675585284281, + "grad_norm": 0.0007173597696237266, + "learning_rate": 0.04870876771598966, + "loss": 0.3325, + "num_input_tokens_seen": 61494000, + "step": 29440 + }, + { + "epoch": 4.803491312505098, + "grad_norm": 0.0005473200581036508, + "learning_rate": 0.04866532929449744, + "loss": 0.3241, + "num_input_tokens_seen": 61503728, + "step": 29445 + }, + { + "epoch": 4.8043070397259156, + "grad_norm": 0.00041920423973351717, + "learning_rate": 0.048621906500084945, + "loss": 0.1982, + "num_input_tokens_seen": 61514912, + "step": 29450 + }, + { + "epoch": 4.805122766946733, + "grad_norm": 0.00048059862456284463, + "learning_rate": 0.04857849933944845, + "loss": 0.2952, + "num_input_tokens_seen": 61526080, + "step": 29455 + }, + { + "epoch": 4.805938494167551, + "grad_norm": 0.0008329306147061288, + "learning_rate": 0.048535107819281866, + "loss": 0.3562, + "num_input_tokens_seen": 61535920, + "step": 29460 + }, + { + "epoch": 4.806754221388368, + "grad_norm": 0.0003904237528331578, + "learning_rate": 0.04849173194627675, + "loss": 0.2523, + "num_input_tokens_seen": 61547280, + "step": 29465 + }, + { + "epoch": 4.807569948609185, + "grad_norm": 0.0007740150904282928, + "learning_rate": 0.04844837172712223, + "loss": 0.2606, + "num_input_tokens_seen": 61559104, + "step": 29470 + }, + { + "epoch": 4.808385675830002, + "grad_norm": 0.0006681016529910266, + "learning_rate": 0.04840502716850494, + "loss": 0.2654, + "num_input_tokens_seen": 61569760, + "step": 29475 + }, + { + "epoch": 4.80920140305082, + "grad_norm": 0.00045528364717029035, + "learning_rate": 0.04836169827710916, + "loss": 0.2975, + "num_input_tokens_seen": 61580144, + "step": 29480 + }, + { + "epoch": 4.810017130271637, + "grad_norm": 0.0006631092983298004, + "learning_rate": 0.04831838505961684, + "loss": 0.3289, + "num_input_tokens_seen": 61591488, + "step": 29485 + }, + { + "epoch": 4.810832857492454, + "grad_norm": 0.0008063683053478599, + "learning_rate": 0.048275087522707295, + "loss": 0.3788, + "num_input_tokens_seen": 61601280, + "step": 29490 + }, + { + "epoch": 4.8116485847132715, + "grad_norm": 0.0007933492306619883, + "learning_rate": 0.04823180567305766, + "loss": 0.2373, + "num_input_tokens_seen": 61612080, + "step": 29495 + }, + { + "epoch": 4.8124643119340895, + "grad_norm": 0.0007014950970187783, + "learning_rate": 0.04818853951734244, + "loss": 0.2839, + "num_input_tokens_seen": 61622896, + "step": 29500 + }, + { + "epoch": 4.813280039154907, + "grad_norm": 0.0005369660211727023, + "learning_rate": 0.04814528906223387, + "loss": 0.349, + "num_input_tokens_seen": 61633136, + "step": 29505 + }, + { + "epoch": 4.814095766375724, + "grad_norm": 0.000672119960654527, + "learning_rate": 0.04810205431440177, + "loss": 0.3568, + "num_input_tokens_seen": 61643440, + "step": 29510 + }, + { + "epoch": 4.814911493596542, + "grad_norm": 0.0005419381777755916, + "learning_rate": 0.04805883528051341, + "loss": 0.3442, + "num_input_tokens_seen": 61654592, + "step": 29515 + }, + { + "epoch": 4.815727220817359, + "grad_norm": 0.0003842492587864399, + "learning_rate": 0.048015631967233685, + "loss": 0.2934, + "num_input_tokens_seen": 61664112, + "step": 29520 + }, + { + "epoch": 4.816542948038176, + "grad_norm": 0.0008205775520764291, + "learning_rate": 0.04797244438122517, + "loss": 0.3536, + "num_input_tokens_seen": 61674432, + "step": 29525 + }, + { + "epoch": 4.817358675258993, + "grad_norm": 0.00045217914157547057, + "learning_rate": 0.04792927252914784, + "loss": 0.3757, + "num_input_tokens_seen": 61684576, + "step": 29530 + }, + { + "epoch": 4.81817440247981, + "grad_norm": 0.0006892029196023941, + "learning_rate": 0.04788611641765944, + "loss": 0.3385, + "num_input_tokens_seen": 61696240, + "step": 29535 + }, + { + "epoch": 4.818990129700628, + "grad_norm": 0.0003207214758731425, + "learning_rate": 0.04784297605341508, + "loss": 0.3484, + "num_input_tokens_seen": 61706752, + "step": 29540 + }, + { + "epoch": 4.8198058569214455, + "grad_norm": 0.00046605398529209197, + "learning_rate": 0.04779985144306761, + "loss": 0.2501, + "num_input_tokens_seen": 61716576, + "step": 29545 + }, + { + "epoch": 4.820621584142263, + "grad_norm": 0.00042856347863562405, + "learning_rate": 0.047756742593267405, + "loss": 0.2985, + "num_input_tokens_seen": 61726368, + "step": 29550 + }, + { + "epoch": 4.821437311363081, + "grad_norm": 0.0005497458623722196, + "learning_rate": 0.047713649510662315, + "loss": 0.3242, + "num_input_tokens_seen": 61736704, + "step": 29555 + }, + { + "epoch": 4.822253038583898, + "grad_norm": 0.0002745494421105832, + "learning_rate": 0.04767057220189789, + "loss": 0.2422, + "num_input_tokens_seen": 61747296, + "step": 29560 + }, + { + "epoch": 4.823068765804715, + "grad_norm": 0.0005566403269767761, + "learning_rate": 0.04762751067361722, + "loss": 0.3512, + "num_input_tokens_seen": 61757616, + "step": 29565 + }, + { + "epoch": 4.823884493025532, + "grad_norm": 0.0005178602295927703, + "learning_rate": 0.04758446493246086, + "loss": 0.3461, + "num_input_tokens_seen": 61767120, + "step": 29570 + }, + { + "epoch": 4.824700220246349, + "grad_norm": 0.0005771779688075185, + "learning_rate": 0.047541434985067084, + "loss": 0.3089, + "num_input_tokens_seen": 61777584, + "step": 29575 + }, + { + "epoch": 4.825515947467167, + "grad_norm": 0.0008034518687054515, + "learning_rate": 0.047498420838071556, + "loss": 0.3576, + "num_input_tokens_seen": 61787792, + "step": 29580 + }, + { + "epoch": 4.826331674687984, + "grad_norm": 0.0010350322118028998, + "learning_rate": 0.04745542249810772, + "loss": 0.3249, + "num_input_tokens_seen": 61797760, + "step": 29585 + }, + { + "epoch": 4.827147401908801, + "grad_norm": 0.0006625870009884238, + "learning_rate": 0.047412439971806324, + "loss": 0.3365, + "num_input_tokens_seen": 61808720, + "step": 29590 + }, + { + "epoch": 4.8279631291296194, + "grad_norm": 0.0004565221897792071, + "learning_rate": 0.04736947326579592, + "loss": 0.2638, + "num_input_tokens_seen": 61818624, + "step": 29595 + }, + { + "epoch": 4.828778856350437, + "grad_norm": 0.00038053205935284495, + "learning_rate": 0.04732652238670245, + "loss": 0.2985, + "num_input_tokens_seen": 61829776, + "step": 29600 + }, + { + "epoch": 4.828778856350437, + "eval_loss": 0.3153003752231598, + "eval_runtime": 156.1474, + "eval_samples_per_second": 17.451, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 61829776, + "step": 29600 + }, + { + "epoch": 4.829594583571254, + "grad_norm": 0.0006630767020396888, + "learning_rate": 0.04728358734114952, + "loss": 0.3493, + "num_input_tokens_seen": 61839296, + "step": 29605 + }, + { + "epoch": 4.830410310792071, + "grad_norm": 0.0012942333705723286, + "learning_rate": 0.04724066813575821, + "loss": 0.3465, + "num_input_tokens_seen": 61850144, + "step": 29610 + }, + { + "epoch": 4.831226038012889, + "grad_norm": 0.0004076463810633868, + "learning_rate": 0.04719776477714729, + "loss": 0.3075, + "num_input_tokens_seen": 61859952, + "step": 29615 + }, + { + "epoch": 4.832041765233706, + "grad_norm": 0.00041381060145795345, + "learning_rate": 0.047154877271932856, + "loss": 0.3273, + "num_input_tokens_seen": 61870560, + "step": 29620 + }, + { + "epoch": 4.832857492454523, + "grad_norm": 0.0015325323911383748, + "learning_rate": 0.0471120056267288, + "loss": 0.375, + "num_input_tokens_seen": 61882064, + "step": 29625 + }, + { + "epoch": 4.83367321967534, + "grad_norm": 0.0005009424057789147, + "learning_rate": 0.047069149848146495, + "loss": 0.3604, + "num_input_tokens_seen": 61892912, + "step": 29630 + }, + { + "epoch": 4.834488946896158, + "grad_norm": 0.0006497757276520133, + "learning_rate": 0.04702630994279473, + "loss": 0.3512, + "num_input_tokens_seen": 61902944, + "step": 29635 + }, + { + "epoch": 4.835304674116975, + "grad_norm": 0.0003941053291782737, + "learning_rate": 0.046983485917280035, + "loss": 0.3188, + "num_input_tokens_seen": 61912416, + "step": 29640 + }, + { + "epoch": 4.8361204013377925, + "grad_norm": 0.0007463697111234069, + "learning_rate": 0.04694067777820644, + "loss": 0.2878, + "num_input_tokens_seen": 61923968, + "step": 29645 + }, + { + "epoch": 4.83693612855861, + "grad_norm": 0.0004226358432788402, + "learning_rate": 0.046897885532175415, + "loss": 0.3238, + "num_input_tokens_seen": 61935376, + "step": 29650 + }, + { + "epoch": 4.837751855779428, + "grad_norm": 0.0008548024925403297, + "learning_rate": 0.04685510918578613, + "loss": 0.4092, + "num_input_tokens_seen": 61946224, + "step": 29655 + }, + { + "epoch": 4.838567583000245, + "grad_norm": 0.00043872641981579363, + "learning_rate": 0.04681234874563519, + "loss": 0.3415, + "num_input_tokens_seen": 61955136, + "step": 29660 + }, + { + "epoch": 4.839383310221062, + "grad_norm": 0.00026455961051397026, + "learning_rate": 0.046769604218316836, + "loss": 0.2881, + "num_input_tokens_seen": 61966608, + "step": 29665 + }, + { + "epoch": 4.840199037441879, + "grad_norm": 0.0002567099581938237, + "learning_rate": 0.04672687561042279, + "loss": 0.2931, + "num_input_tokens_seen": 61977280, + "step": 29670 + }, + { + "epoch": 4.841014764662697, + "grad_norm": 0.0003779797989409417, + "learning_rate": 0.046684162928542286, + "loss": 0.3137, + "num_input_tokens_seen": 61988224, + "step": 29675 + }, + { + "epoch": 4.841830491883514, + "grad_norm": 0.0003957259759772569, + "learning_rate": 0.04664146617926222, + "loss": 0.3213, + "num_input_tokens_seen": 62000544, + "step": 29680 + }, + { + "epoch": 4.842646219104331, + "grad_norm": 0.0003248801804147661, + "learning_rate": 0.046598785369167, + "loss": 0.3207, + "num_input_tokens_seen": 62011632, + "step": 29685 + }, + { + "epoch": 4.8434619463251485, + "grad_norm": 0.0005825653206557035, + "learning_rate": 0.046556120504838434, + "loss": 0.3223, + "num_input_tokens_seen": 62022848, + "step": 29690 + }, + { + "epoch": 4.8442776735459665, + "grad_norm": 0.0005241502658464015, + "learning_rate": 0.04651347159285609, + "loss": 0.406, + "num_input_tokens_seen": 62034448, + "step": 29695 + }, + { + "epoch": 4.845093400766784, + "grad_norm": 0.0006193496519699693, + "learning_rate": 0.04647083863979688, + "loss": 0.3631, + "num_input_tokens_seen": 62045136, + "step": 29700 + }, + { + "epoch": 4.845909127987601, + "grad_norm": 0.0003034551627933979, + "learning_rate": 0.04642822165223538, + "loss": 0.3865, + "num_input_tokens_seen": 62055600, + "step": 29705 + }, + { + "epoch": 4.846724855208418, + "grad_norm": 0.0005139242857694626, + "learning_rate": 0.046385620636743716, + "loss": 0.3467, + "num_input_tokens_seen": 62065824, + "step": 29710 + }, + { + "epoch": 4.847540582429236, + "grad_norm": 0.0004324107721913606, + "learning_rate": 0.04634303559989141, + "loss": 0.3664, + "num_input_tokens_seen": 62076368, + "step": 29715 + }, + { + "epoch": 4.848356309650053, + "grad_norm": 0.00032559834653511643, + "learning_rate": 0.046300466548245635, + "loss": 0.3254, + "num_input_tokens_seen": 62086064, + "step": 29720 + }, + { + "epoch": 4.84917203687087, + "grad_norm": 0.0004048266273457557, + "learning_rate": 0.04625791348837114, + "loss": 0.2583, + "num_input_tokens_seen": 62095968, + "step": 29725 + }, + { + "epoch": 4.849987764091688, + "grad_norm": 0.000605641573201865, + "learning_rate": 0.046215376426830095, + "loss": 0.2976, + "num_input_tokens_seen": 62107440, + "step": 29730 + }, + { + "epoch": 4.850803491312505, + "grad_norm": 0.0004392681294120848, + "learning_rate": 0.04617285537018219, + "loss": 0.3403, + "num_input_tokens_seen": 62116608, + "step": 29735 + }, + { + "epoch": 4.8516192185333225, + "grad_norm": 0.0005991292418912053, + "learning_rate": 0.046130350324984803, + "loss": 0.2968, + "num_input_tokens_seen": 62126384, + "step": 29740 + }, + { + "epoch": 4.85243494575414, + "grad_norm": 0.0008436781354248524, + "learning_rate": 0.046087861297792666, + "loss": 0.3222, + "num_input_tokens_seen": 62137760, + "step": 29745 + }, + { + "epoch": 4.853250672974957, + "grad_norm": 0.0010933715384453535, + "learning_rate": 0.0460453882951582, + "loss": 0.3236, + "num_input_tokens_seen": 62148448, + "step": 29750 + }, + { + "epoch": 4.854066400195775, + "grad_norm": 0.0007567487773485482, + "learning_rate": 0.04600293132363119, + "loss": 0.3358, + "num_input_tokens_seen": 62158992, + "step": 29755 + }, + { + "epoch": 4.854882127416592, + "grad_norm": 0.0007325925398617983, + "learning_rate": 0.045960490389759086, + "loss": 0.2889, + "num_input_tokens_seen": 62170272, + "step": 29760 + }, + { + "epoch": 4.855697854637409, + "grad_norm": 0.00034419764415360987, + "learning_rate": 0.04591806550008685, + "loss": 0.3113, + "num_input_tokens_seen": 62180288, + "step": 29765 + }, + { + "epoch": 4.856513581858227, + "grad_norm": 0.0004322176391724497, + "learning_rate": 0.045875656661156825, + "loss": 0.4001, + "num_input_tokens_seen": 62190176, + "step": 29770 + }, + { + "epoch": 4.857329309079044, + "grad_norm": 0.0003299899981357157, + "learning_rate": 0.04583326387950911, + "loss": 0.303, + "num_input_tokens_seen": 62199616, + "step": 29775 + }, + { + "epoch": 4.858145036299861, + "grad_norm": 0.000729050487279892, + "learning_rate": 0.0457908871616811, + "loss": 0.3427, + "num_input_tokens_seen": 62209456, + "step": 29780 + }, + { + "epoch": 4.858960763520678, + "grad_norm": 0.0004306466435082257, + "learning_rate": 0.04574852651420786, + "loss": 0.2961, + "num_input_tokens_seen": 62219408, + "step": 29785 + }, + { + "epoch": 4.859776490741496, + "grad_norm": 0.0006425231113098562, + "learning_rate": 0.045706181943621985, + "loss": 0.3198, + "num_input_tokens_seen": 62229968, + "step": 29790 + }, + { + "epoch": 4.8605922179623136, + "grad_norm": 0.0007424099021591246, + "learning_rate": 0.04566385345645344, + "loss": 0.2912, + "num_input_tokens_seen": 62239616, + "step": 29795 + }, + { + "epoch": 4.861407945183131, + "grad_norm": 0.0005117899854667485, + "learning_rate": 0.04562154105922993, + "loss": 0.3175, + "num_input_tokens_seen": 62250704, + "step": 29800 + }, + { + "epoch": 4.861407945183131, + "eval_loss": 0.3161299228668213, + "eval_runtime": 156.0559, + "eval_samples_per_second": 17.462, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 62250704, + "step": 29800 + }, + { + "epoch": 4.862223672403948, + "grad_norm": 0.0021409608889371157, + "learning_rate": 0.04557924475847642, + "loss": 0.3524, + "num_input_tokens_seen": 62262048, + "step": 29805 + }, + { + "epoch": 4.863039399624766, + "grad_norm": 0.0003838100819848478, + "learning_rate": 0.04553696456071567, + "loss": 0.3369, + "num_input_tokens_seen": 62271216, + "step": 29810 + }, + { + "epoch": 4.863855126845583, + "grad_norm": 0.00033262919168919325, + "learning_rate": 0.045494700472467724, + "loss": 0.301, + "num_input_tokens_seen": 62282272, + "step": 29815 + }, + { + "epoch": 4.8646708540664, + "grad_norm": 0.0005434969207271934, + "learning_rate": 0.04545245250025024, + "loss": 0.3226, + "num_input_tokens_seen": 62293152, + "step": 29820 + }, + { + "epoch": 4.865486581287217, + "grad_norm": 0.0006499222945421934, + "learning_rate": 0.045410220650578384, + "loss": 0.3223, + "num_input_tokens_seen": 62303536, + "step": 29825 + }, + { + "epoch": 4.866302308508035, + "grad_norm": 0.0014621001901105046, + "learning_rate": 0.04536800492996492, + "loss": 0.3262, + "num_input_tokens_seen": 62315040, + "step": 29830 + }, + { + "epoch": 4.867118035728852, + "grad_norm": 0.0005580916767939925, + "learning_rate": 0.04532580534491994, + "loss": 0.3141, + "num_input_tokens_seen": 62325584, + "step": 29835 + }, + { + "epoch": 4.8679337629496695, + "grad_norm": 0.00030852362397126853, + "learning_rate": 0.045283621901951183, + "loss": 0.3523, + "num_input_tokens_seen": 62335856, + "step": 29840 + }, + { + "epoch": 4.868749490170487, + "grad_norm": 0.0010074294405058026, + "learning_rate": 0.04524145460756393, + "loss": 0.3765, + "num_input_tokens_seen": 62346928, + "step": 29845 + }, + { + "epoch": 4.869565217391305, + "grad_norm": 0.0008233481785282493, + "learning_rate": 0.045199303468260794, + "loss": 0.3698, + "num_input_tokens_seen": 62356368, + "step": 29850 + }, + { + "epoch": 4.870380944612122, + "grad_norm": 0.00044610039913095534, + "learning_rate": 0.04515716849054214, + "loss": 0.3165, + "num_input_tokens_seen": 62366816, + "step": 29855 + }, + { + "epoch": 4.871196671832939, + "grad_norm": 0.0010251495987176895, + "learning_rate": 0.04511504968090558, + "loss": 0.3595, + "num_input_tokens_seen": 62377952, + "step": 29860 + }, + { + "epoch": 4.872012399053756, + "grad_norm": 0.0008402874227613211, + "learning_rate": 0.04507294704584644, + "loss": 0.2757, + "num_input_tokens_seen": 62388736, + "step": 29865 + }, + { + "epoch": 4.872828126274574, + "grad_norm": 0.00042283249786123633, + "learning_rate": 0.04503086059185749, + "loss": 0.2757, + "num_input_tokens_seen": 62397968, + "step": 29870 + }, + { + "epoch": 4.873643853495391, + "grad_norm": 0.00038493183092214167, + "learning_rate": 0.04498879032542893, + "loss": 0.3162, + "num_input_tokens_seen": 62408064, + "step": 29875 + }, + { + "epoch": 4.874459580716208, + "grad_norm": 0.00034605813561938703, + "learning_rate": 0.0449467362530486, + "loss": 0.3502, + "num_input_tokens_seen": 62418992, + "step": 29880 + }, + { + "epoch": 4.8752753079370255, + "grad_norm": 0.0008234247216023505, + "learning_rate": 0.04490469838120171, + "loss": 0.292, + "num_input_tokens_seen": 62429392, + "step": 29885 + }, + { + "epoch": 4.8760910351578435, + "grad_norm": 0.00037170175346545875, + "learning_rate": 0.04486267671637101, + "loss": 0.3149, + "num_input_tokens_seen": 62438288, + "step": 29890 + }, + { + "epoch": 4.876906762378661, + "grad_norm": 0.0009402279392816126, + "learning_rate": 0.04482067126503683, + "loss": 0.3486, + "num_input_tokens_seen": 62447328, + "step": 29895 + }, + { + "epoch": 4.877722489599478, + "grad_norm": 0.0008330834098160267, + "learning_rate": 0.04477868203367687, + "loss": 0.3855, + "num_input_tokens_seen": 62457248, + "step": 29900 + }, + { + "epoch": 4.878538216820296, + "grad_norm": 0.0003836054529529065, + "learning_rate": 0.044736709028766426, + "loss": 0.3155, + "num_input_tokens_seen": 62467056, + "step": 29905 + }, + { + "epoch": 4.879353944041113, + "grad_norm": 0.0005039718234911561, + "learning_rate": 0.04469475225677832, + "loss": 0.3241, + "num_input_tokens_seen": 62477072, + "step": 29910 + }, + { + "epoch": 4.88016967126193, + "grad_norm": 0.0003037917776964605, + "learning_rate": 0.04465281172418273, + "loss": 0.3621, + "num_input_tokens_seen": 62487728, + "step": 29915 + }, + { + "epoch": 4.880985398482747, + "grad_norm": 0.00036002948763780296, + "learning_rate": 0.044610887437447476, + "loss": 0.3307, + "num_input_tokens_seen": 62498880, + "step": 29920 + }, + { + "epoch": 4.881801125703564, + "grad_norm": 0.00039331582956947386, + "learning_rate": 0.044568979403037744, + "loss": 0.3068, + "num_input_tokens_seen": 62508048, + "step": 29925 + }, + { + "epoch": 4.882616852924382, + "grad_norm": 0.00038322556065395474, + "learning_rate": 0.04452708762741631, + "loss": 0.3487, + "num_input_tokens_seen": 62518464, + "step": 29930 + }, + { + "epoch": 4.883432580145199, + "grad_norm": 0.0007643406861461699, + "learning_rate": 0.044485212117043475, + "loss": 0.3818, + "num_input_tokens_seen": 62528144, + "step": 29935 + }, + { + "epoch": 4.884248307366017, + "grad_norm": 0.0005777118494734168, + "learning_rate": 0.04444335287837687, + "loss": 0.3072, + "num_input_tokens_seen": 62538560, + "step": 29940 + }, + { + "epoch": 4.885064034586835, + "grad_norm": 0.0007700955029577017, + "learning_rate": 0.04440150991787179, + "loss": 0.301, + "num_input_tokens_seen": 62549648, + "step": 29945 + }, + { + "epoch": 4.885879761807652, + "grad_norm": 0.0005659381276927888, + "learning_rate": 0.04435968324198088, + "loss": 0.2738, + "num_input_tokens_seen": 62560480, + "step": 29950 + }, + { + "epoch": 4.886695489028469, + "grad_norm": 0.000354816293111071, + "learning_rate": 0.04431787285715442, + "loss": 0.3212, + "num_input_tokens_seen": 62569552, + "step": 29955 + }, + { + "epoch": 4.887511216249286, + "grad_norm": 0.0003790987830143422, + "learning_rate": 0.04427607876984004, + "loss": 0.3296, + "num_input_tokens_seen": 62580816, + "step": 29960 + }, + { + "epoch": 4.888326943470103, + "grad_norm": 0.0008863159455358982, + "learning_rate": 0.044234300986482886, + "loss": 0.3316, + "num_input_tokens_seen": 62590096, + "step": 29965 + }, + { + "epoch": 4.889142670690921, + "grad_norm": 0.00048591019003652036, + "learning_rate": 0.04419253951352566, + "loss": 0.2598, + "num_input_tokens_seen": 62600256, + "step": 29970 + }, + { + "epoch": 4.889958397911738, + "grad_norm": 0.0004736761620733887, + "learning_rate": 0.044150794357408533, + "loss": 0.2757, + "num_input_tokens_seen": 62609792, + "step": 29975 + }, + { + "epoch": 4.890774125132555, + "grad_norm": 0.0011436005588620901, + "learning_rate": 0.044109065524569065, + "loss": 0.3789, + "num_input_tokens_seen": 62620736, + "step": 29980 + }, + { + "epoch": 4.891589852353373, + "grad_norm": 0.0005485496367327869, + "learning_rate": 0.0440673530214424, + "loss": 0.3335, + "num_input_tokens_seen": 62631904, + "step": 29985 + }, + { + "epoch": 4.8924055795741905, + "grad_norm": 0.0003383290022611618, + "learning_rate": 0.04402565685446117, + "loss": 0.3308, + "num_input_tokens_seen": 62642752, + "step": 29990 + }, + { + "epoch": 4.893221306795008, + "grad_norm": 0.00055567646631971, + "learning_rate": 0.04398397703005536, + "loss": 0.3249, + "num_input_tokens_seen": 62653248, + "step": 29995 + }, + { + "epoch": 4.894037034015825, + "grad_norm": 0.0005562871228903532, + "learning_rate": 0.043942313554652626, + "loss": 0.3234, + "num_input_tokens_seen": 62662656, + "step": 30000 + }, + { + "epoch": 4.894037034015825, + "eval_loss": 0.3150690793991089, + "eval_runtime": 156.1071, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 62662656, + "step": 30000 + }, + { + "epoch": 4.894852761236643, + "grad_norm": 0.00032751820981502533, + "learning_rate": 0.0439006664346779, + "loss": 0.3346, + "num_input_tokens_seen": 62671264, + "step": 30005 + }, + { + "epoch": 4.89566848845746, + "grad_norm": 0.0005586258484981954, + "learning_rate": 0.043859035676553755, + "loss": 0.2695, + "num_input_tokens_seen": 62682720, + "step": 30010 + }, + { + "epoch": 4.896484215678277, + "grad_norm": 0.0004202354175504297, + "learning_rate": 0.043817421286700194, + "loss": 0.2998, + "num_input_tokens_seen": 62694144, + "step": 30015 + }, + { + "epoch": 4.897299942899094, + "grad_norm": 0.00035642090369947255, + "learning_rate": 0.043775823271534585, + "loss": 0.3436, + "num_input_tokens_seen": 62706368, + "step": 30020 + }, + { + "epoch": 4.898115670119912, + "grad_norm": 0.0004905051318928599, + "learning_rate": 0.04373424163747197, + "loss": 0.3062, + "num_input_tokens_seen": 62717952, + "step": 30025 + }, + { + "epoch": 4.898931397340729, + "grad_norm": 0.00045757595216855407, + "learning_rate": 0.04369267639092473, + "loss": 0.3332, + "num_input_tokens_seen": 62729552, + "step": 30030 + }, + { + "epoch": 4.8997471245615465, + "grad_norm": 0.0002624374465085566, + "learning_rate": 0.04365112753830268, + "loss": 0.2951, + "num_input_tokens_seen": 62740064, + "step": 30035 + }, + { + "epoch": 4.900562851782364, + "grad_norm": 0.0006411675130948424, + "learning_rate": 0.04360959508601327, + "loss": 0.2872, + "num_input_tokens_seen": 62751072, + "step": 30040 + }, + { + "epoch": 4.901378579003182, + "grad_norm": 0.0006366918096318841, + "learning_rate": 0.04356807904046123, + "loss": 0.3169, + "num_input_tokens_seen": 62762304, + "step": 30045 + }, + { + "epoch": 4.902194306223999, + "grad_norm": 0.000709483923856169, + "learning_rate": 0.04352657940804892, + "loss": 0.345, + "num_input_tokens_seen": 62772432, + "step": 30050 + }, + { + "epoch": 4.903010033444816, + "grad_norm": 0.0005289185792207718, + "learning_rate": 0.04348509619517613, + "loss": 0.2928, + "num_input_tokens_seen": 62782064, + "step": 30055 + }, + { + "epoch": 4.903825760665633, + "grad_norm": 0.0005533218500204384, + "learning_rate": 0.04344362940824002, + "loss": 0.216, + "num_input_tokens_seen": 62791744, + "step": 30060 + }, + { + "epoch": 4.904641487886451, + "grad_norm": 0.0008202927419915795, + "learning_rate": 0.04340217905363533, + "loss": 0.3137, + "num_input_tokens_seen": 62802240, + "step": 30065 + }, + { + "epoch": 4.905457215107268, + "grad_norm": 0.0003431173099670559, + "learning_rate": 0.04336074513775425, + "loss": 0.2599, + "num_input_tokens_seen": 62813120, + "step": 30070 + }, + { + "epoch": 4.906272942328085, + "grad_norm": 0.0005954777589067817, + "learning_rate": 0.04331932766698636, + "loss": 0.3049, + "num_input_tokens_seen": 62823792, + "step": 30075 + }, + { + "epoch": 4.907088669548903, + "grad_norm": 0.000742180272936821, + "learning_rate": 0.0432779266477188, + "loss": 0.3911, + "num_input_tokens_seen": 62833440, + "step": 30080 + }, + { + "epoch": 4.9079043967697205, + "grad_norm": 0.00037867005448788404, + "learning_rate": 0.04323654208633607, + "loss": 0.2851, + "num_input_tokens_seen": 62844544, + "step": 30085 + }, + { + "epoch": 4.908720123990538, + "grad_norm": 0.0004118001670576632, + "learning_rate": 0.04319517398922024, + "loss": 0.3393, + "num_input_tokens_seen": 62854832, + "step": 30090 + }, + { + "epoch": 4.909535851211355, + "grad_norm": 0.0004484154924284667, + "learning_rate": 0.04315382236275079, + "loss": 0.3156, + "num_input_tokens_seen": 62865248, + "step": 30095 + }, + { + "epoch": 4.910351578432172, + "grad_norm": 0.0005375085165724158, + "learning_rate": 0.043112487213304664, + "loss": 0.2913, + "num_input_tokens_seen": 62876384, + "step": 30100 + }, + { + "epoch": 4.91116730565299, + "grad_norm": 0.0006033725803717971, + "learning_rate": 0.04307116854725618, + "loss": 0.3046, + "num_input_tokens_seen": 62888000, + "step": 30105 + }, + { + "epoch": 4.911983032873807, + "grad_norm": 0.00068679079413414, + "learning_rate": 0.043029866370977325, + "loss": 0.3357, + "num_input_tokens_seen": 62899568, + "step": 30110 + }, + { + "epoch": 4.912798760094624, + "grad_norm": 0.0005362546653486788, + "learning_rate": 0.04298858069083728, + "loss": 0.3138, + "num_input_tokens_seen": 62910944, + "step": 30115 + }, + { + "epoch": 4.913614487315442, + "grad_norm": 0.0005679057212546468, + "learning_rate": 0.04294731151320295, + "loss": 0.295, + "num_input_tokens_seen": 62921744, + "step": 30120 + }, + { + "epoch": 4.914430214536259, + "grad_norm": 0.0008212907705456018, + "learning_rate": 0.04290605884443841, + "loss": 0.3319, + "num_input_tokens_seen": 62931952, + "step": 30125 + }, + { + "epoch": 4.915245941757076, + "grad_norm": 0.00033821785473264754, + "learning_rate": 0.04286482269090545, + "loss": 0.3304, + "num_input_tokens_seen": 62943808, + "step": 30130 + }, + { + "epoch": 4.9160616689778935, + "grad_norm": 0.0003885548503603786, + "learning_rate": 0.04282360305896323, + "loss": 0.3462, + "num_input_tokens_seen": 62955200, + "step": 30135 + }, + { + "epoch": 4.916877396198711, + "grad_norm": 0.0006925287889316678, + "learning_rate": 0.04278239995496822, + "loss": 0.3483, + "num_input_tokens_seen": 62964256, + "step": 30140 + }, + { + "epoch": 4.917693123419529, + "grad_norm": 0.0003936440625693649, + "learning_rate": 0.042741213385274514, + "loss": 0.3407, + "num_input_tokens_seen": 62975280, + "step": 30145 + }, + { + "epoch": 4.918508850640346, + "grad_norm": 0.00046649196883663535, + "learning_rate": 0.04270004335623366, + "loss": 0.3615, + "num_input_tokens_seen": 62986288, + "step": 30150 + }, + { + "epoch": 4.919324577861163, + "grad_norm": 0.0006409296183846891, + "learning_rate": 0.04265888987419448, + "loss": 0.3693, + "num_input_tokens_seen": 62997504, + "step": 30155 + }, + { + "epoch": 4.920140305081981, + "grad_norm": 0.0004993949551135302, + "learning_rate": 0.04261775294550346, + "loss": 0.3355, + "num_input_tokens_seen": 63007344, + "step": 30160 + }, + { + "epoch": 4.920956032302798, + "grad_norm": 0.0005359802162274718, + "learning_rate": 0.042576632576504354, + "loss": 0.3362, + "num_input_tokens_seen": 63017920, + "step": 30165 + }, + { + "epoch": 4.921771759523615, + "grad_norm": 0.00027822391712106764, + "learning_rate": 0.0425355287735385, + "loss": 0.3048, + "num_input_tokens_seen": 63027088, + "step": 30170 + }, + { + "epoch": 4.922587486744432, + "grad_norm": 0.00037657920620404184, + "learning_rate": 0.0424944415429446, + "loss": 0.33, + "num_input_tokens_seen": 63037840, + "step": 30175 + }, + { + "epoch": 4.92340321396525, + "grad_norm": 0.0004388471716083586, + "learning_rate": 0.04245337089105877, + "loss": 0.395, + "num_input_tokens_seen": 63047552, + "step": 30180 + }, + { + "epoch": 4.9242189411860675, + "grad_norm": 0.0007243382278829813, + "learning_rate": 0.04241231682421467, + "loss": 0.3319, + "num_input_tokens_seen": 63057456, + "step": 30185 + }, + { + "epoch": 4.925034668406885, + "grad_norm": 0.0009359829127788544, + "learning_rate": 0.04237127934874337, + "loss": 0.2734, + "num_input_tokens_seen": 63068464, + "step": 30190 + }, + { + "epoch": 4.925850395627702, + "grad_norm": 0.0005120497080497444, + "learning_rate": 0.042330258470973305, + "loss": 0.3394, + "num_input_tokens_seen": 63078112, + "step": 30195 + }, + { + "epoch": 4.92666612284852, + "grad_norm": 0.0003317716473247856, + "learning_rate": 0.042289254197230515, + "loss": 0.3708, + "num_input_tokens_seen": 63088352, + "step": 30200 + }, + { + "epoch": 4.92666612284852, + "eval_loss": 0.3169708251953125, + "eval_runtime": 156.0833, + "eval_samples_per_second": 17.459, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 63088352, + "step": 30200 + }, + { + "epoch": 4.927481850069337, + "grad_norm": 0.00041330381645821035, + "learning_rate": 0.04224826653383823, + "loss": 0.3314, + "num_input_tokens_seen": 63098240, + "step": 30205 + }, + { + "epoch": 4.928297577290154, + "grad_norm": 0.00036510435165837407, + "learning_rate": 0.04220729548711735, + "loss": 0.2735, + "num_input_tokens_seen": 63108928, + "step": 30210 + }, + { + "epoch": 4.929113304510971, + "grad_norm": 0.0005914294742979109, + "learning_rate": 0.04216634106338616, + "loss": 0.3008, + "num_input_tokens_seen": 63119520, + "step": 30215 + }, + { + "epoch": 4.929929031731789, + "grad_norm": 0.0004046468820888549, + "learning_rate": 0.04212540326896025, + "loss": 0.2701, + "num_input_tokens_seen": 63130384, + "step": 30220 + }, + { + "epoch": 4.930744758952606, + "grad_norm": 0.0009120145696215332, + "learning_rate": 0.0420844821101528, + "loss": 0.3713, + "num_input_tokens_seen": 63141024, + "step": 30225 + }, + { + "epoch": 4.9315604861734235, + "grad_norm": 0.0003146443050354719, + "learning_rate": 0.04204357759327441, + "loss": 0.272, + "num_input_tokens_seen": 63151424, + "step": 30230 + }, + { + "epoch": 4.932376213394241, + "grad_norm": 0.0005417903303168714, + "learning_rate": 0.042002689724632954, + "loss": 0.3054, + "num_input_tokens_seen": 63161952, + "step": 30235 + }, + { + "epoch": 4.933191940615059, + "grad_norm": 0.0006046365015208721, + "learning_rate": 0.04196181851053398, + "loss": 0.2322, + "num_input_tokens_seen": 63172496, + "step": 30240 + }, + { + "epoch": 4.934007667835876, + "grad_norm": 0.0005541631835512817, + "learning_rate": 0.041920963957280295, + "loss": 0.3819, + "num_input_tokens_seen": 63181344, + "step": 30245 + }, + { + "epoch": 4.934823395056693, + "grad_norm": 0.0005028998130001128, + "learning_rate": 0.04188012607117212, + "loss": 0.3414, + "num_input_tokens_seen": 63190512, + "step": 30250 + }, + { + "epoch": 4.935639122277511, + "grad_norm": 0.0005864985287189484, + "learning_rate": 0.04183930485850725, + "loss": 0.3461, + "num_input_tokens_seen": 63201712, + "step": 30255 + }, + { + "epoch": 4.936454849498328, + "grad_norm": 0.0003537957672961056, + "learning_rate": 0.04179850032558078, + "loss": 0.3252, + "num_input_tokens_seen": 63212976, + "step": 30260 + }, + { + "epoch": 4.937270576719145, + "grad_norm": 0.000517234206199646, + "learning_rate": 0.041757712478685295, + "loss": 0.3055, + "num_input_tokens_seen": 63224096, + "step": 30265 + }, + { + "epoch": 4.938086303939962, + "grad_norm": 0.0003289131273049861, + "learning_rate": 0.04171694132411085, + "loss": 0.2925, + "num_input_tokens_seen": 63234976, + "step": 30270 + }, + { + "epoch": 4.938902031160779, + "grad_norm": 0.0005363922100514174, + "learning_rate": 0.04167618686814479, + "loss": 0.3849, + "num_input_tokens_seen": 63244480, + "step": 30275 + }, + { + "epoch": 4.939717758381597, + "grad_norm": 0.0004371714894659817, + "learning_rate": 0.041635449117072024, + "loss": 0.3279, + "num_input_tokens_seen": 63255152, + "step": 30280 + }, + { + "epoch": 4.940533485602415, + "grad_norm": 0.0005839768564328551, + "learning_rate": 0.04159472807717477, + "loss": 0.3041, + "num_input_tokens_seen": 63266544, + "step": 30285 + }, + { + "epoch": 4.941349212823232, + "grad_norm": 0.0007184273563325405, + "learning_rate": 0.041554023754732744, + "loss": 0.3671, + "num_input_tokens_seen": 63277792, + "step": 30290 + }, + { + "epoch": 4.94216494004405, + "grad_norm": 0.00030332658207044005, + "learning_rate": 0.04151333615602311, + "loss": 0.3234, + "num_input_tokens_seen": 63288912, + "step": 30295 + }, + { + "epoch": 4.942980667264867, + "grad_norm": 0.00038698656135238707, + "learning_rate": 0.04147266528732034, + "loss": 0.3256, + "num_input_tokens_seen": 63298928, + "step": 30300 + }, + { + "epoch": 4.943796394485684, + "grad_norm": 0.0004148791776970029, + "learning_rate": 0.0414320111548964, + "loss": 0.2999, + "num_input_tokens_seen": 63309328, + "step": 30305 + }, + { + "epoch": 4.944612121706501, + "grad_norm": 0.00025896841543726623, + "learning_rate": 0.04139137376502076, + "loss": 0.3364, + "num_input_tokens_seen": 63320864, + "step": 30310 + }, + { + "epoch": 4.945427848927318, + "grad_norm": 0.00036162015749141574, + "learning_rate": 0.04135075312396014, + "loss": 0.3114, + "num_input_tokens_seen": 63330432, + "step": 30315 + }, + { + "epoch": 4.946243576148136, + "grad_norm": 0.0006863628514111042, + "learning_rate": 0.04131014923797875, + "loss": 0.3457, + "num_input_tokens_seen": 63340848, + "step": 30320 + }, + { + "epoch": 4.947059303368953, + "grad_norm": 0.0003519937163218856, + "learning_rate": 0.04126956211333819, + "loss": 0.2975, + "num_input_tokens_seen": 63350048, + "step": 30325 + }, + { + "epoch": 4.9478750305897705, + "grad_norm": 0.0006699992809444666, + "learning_rate": 0.041228991756297545, + "loss": 0.3338, + "num_input_tokens_seen": 63360608, + "step": 30330 + }, + { + "epoch": 4.9486907578105885, + "grad_norm": 0.0010202622506767511, + "learning_rate": 0.04118843817311332, + "loss": 0.3463, + "num_input_tokens_seen": 63371632, + "step": 30335 + }, + { + "epoch": 4.949506485031406, + "grad_norm": 0.0003526063810568303, + "learning_rate": 0.0411479013700393, + "loss": 0.3038, + "num_input_tokens_seen": 63380816, + "step": 30340 + }, + { + "epoch": 4.950322212252223, + "grad_norm": 0.00044544864795170724, + "learning_rate": 0.0411073813533268, + "loss": 0.2324, + "num_input_tokens_seen": 63390720, + "step": 30345 + }, + { + "epoch": 4.95113793947304, + "grad_norm": 0.000462528521893546, + "learning_rate": 0.04106687812922456, + "loss": 0.2552, + "num_input_tokens_seen": 63400592, + "step": 30350 + }, + { + "epoch": 4.951953666693858, + "grad_norm": 0.00048008275916799903, + "learning_rate": 0.041026391703978635, + "loss": 0.3348, + "num_input_tokens_seen": 63409728, + "step": 30355 + }, + { + "epoch": 4.952769393914675, + "grad_norm": 0.0004011821001768112, + "learning_rate": 0.04098592208383259, + "loss": 0.3221, + "num_input_tokens_seen": 63420544, + "step": 30360 + }, + { + "epoch": 4.953585121135492, + "grad_norm": 0.0007469919510185719, + "learning_rate": 0.040945469275027256, + "loss": 0.33, + "num_input_tokens_seen": 63429984, + "step": 30365 + }, + { + "epoch": 4.954400848356309, + "grad_norm": 0.000844883150421083, + "learning_rate": 0.04090503328380104, + "loss": 0.3137, + "num_input_tokens_seen": 63440512, + "step": 30370 + }, + { + "epoch": 4.955216575577127, + "grad_norm": 0.0006769768078811467, + "learning_rate": 0.04086461411638971, + "loss": 0.3151, + "num_input_tokens_seen": 63451424, + "step": 30375 + }, + { + "epoch": 4.9560323027979445, + "grad_norm": 0.000578313076402992, + "learning_rate": 0.04082421177902631, + "loss": 0.2941, + "num_input_tokens_seen": 63462256, + "step": 30380 + }, + { + "epoch": 4.956848030018762, + "grad_norm": 0.0003717863583005965, + "learning_rate": 0.04078382627794149, + "loss": 0.3031, + "num_input_tokens_seen": 63473568, + "step": 30385 + }, + { + "epoch": 4.957663757239579, + "grad_norm": 0.0005667862133122981, + "learning_rate": 0.04074345761936316, + "loss": 0.3385, + "num_input_tokens_seen": 63485152, + "step": 30390 + }, + { + "epoch": 4.958479484460397, + "grad_norm": 0.0006762261618860066, + "learning_rate": 0.04070310580951663, + "loss": 0.3195, + "num_input_tokens_seen": 63495200, + "step": 30395 + }, + { + "epoch": 4.959295211681214, + "grad_norm": 0.00048415231867693365, + "learning_rate": 0.040662770854624726, + "loss": 0.2391, + "num_input_tokens_seen": 63504960, + "step": 30400 + }, + { + "epoch": 4.959295211681214, + "eval_loss": 0.3179698586463928, + "eval_runtime": 156.1788, + "eval_samples_per_second": 17.448, + "eval_steps_per_second": 8.727, + "num_input_tokens_seen": 63504960, + "step": 30400 + }, + { + "epoch": 4.960110938902031, + "grad_norm": 0.0004007055249530822, + "learning_rate": 0.040622452760907535, + "loss": 0.3007, + "num_input_tokens_seen": 63516224, + "step": 30405 + }, + { + "epoch": 4.960926666122848, + "grad_norm": 0.0006214901804924011, + "learning_rate": 0.04058215153458265, + "loss": 0.2938, + "num_input_tokens_seen": 63527232, + "step": 30410 + }, + { + "epoch": 4.961742393343666, + "grad_norm": 0.000584778084885329, + "learning_rate": 0.04054186718186507, + "loss": 0.3673, + "num_input_tokens_seen": 63537936, + "step": 30415 + }, + { + "epoch": 4.962558120564483, + "grad_norm": 0.0005826468695886433, + "learning_rate": 0.04050159970896708, + "loss": 0.3065, + "num_input_tokens_seen": 63547776, + "step": 30420 + }, + { + "epoch": 4.9633738477853, + "grad_norm": 0.00075041368836537, + "learning_rate": 0.04046134912209843, + "loss": 0.3366, + "num_input_tokens_seen": 63558144, + "step": 30425 + }, + { + "epoch": 4.964189575006118, + "grad_norm": 0.0005585853359661996, + "learning_rate": 0.040421115427466354, + "loss": 0.3509, + "num_input_tokens_seen": 63568336, + "step": 30430 + }, + { + "epoch": 4.965005302226936, + "grad_norm": 0.0006349313189275563, + "learning_rate": 0.04038089863127529, + "loss": 0.3176, + "num_input_tokens_seen": 63578944, + "step": 30435 + }, + { + "epoch": 4.965821029447753, + "grad_norm": 0.0002548351767472923, + "learning_rate": 0.04034069873972727, + "loss": 0.3346, + "num_input_tokens_seen": 63589280, + "step": 30440 + }, + { + "epoch": 4.96663675666857, + "grad_norm": 0.0003565814986359328, + "learning_rate": 0.040300515759021514, + "loss": 0.2691, + "num_input_tokens_seen": 63598912, + "step": 30445 + }, + { + "epoch": 4.967452483889387, + "grad_norm": 0.0004950627335347235, + "learning_rate": 0.04026034969535478, + "loss": 0.3031, + "num_input_tokens_seen": 63611200, + "step": 30450 + }, + { + "epoch": 4.968268211110205, + "grad_norm": 0.0005156346596777439, + "learning_rate": 0.040220200554921266, + "loss": 0.2843, + "num_input_tokens_seen": 63621056, + "step": 30455 + }, + { + "epoch": 4.969083938331022, + "grad_norm": 0.00042455687071196735, + "learning_rate": 0.0401800683439124, + "loss": 0.3239, + "num_input_tokens_seen": 63631312, + "step": 30460 + }, + { + "epoch": 4.969899665551839, + "grad_norm": 0.000436878704931587, + "learning_rate": 0.04013995306851704, + "loss": 0.3707, + "num_input_tokens_seen": 63642624, + "step": 30465 + }, + { + "epoch": 4.970715392772657, + "grad_norm": 0.0006308098090812564, + "learning_rate": 0.040099854734921545, + "loss": 0.3531, + "num_input_tokens_seen": 63652912, + "step": 30470 + }, + { + "epoch": 4.971531119993474, + "grad_norm": 0.00042364615364931524, + "learning_rate": 0.0400597733493095, + "loss": 0.346, + "num_input_tokens_seen": 63663584, + "step": 30475 + }, + { + "epoch": 4.9723468472142915, + "grad_norm": 0.00047209285548888147, + "learning_rate": 0.04001970891786203, + "loss": 0.2817, + "num_input_tokens_seen": 63675504, + "step": 30480 + }, + { + "epoch": 4.973162574435109, + "grad_norm": 0.0003903531178366393, + "learning_rate": 0.03997966144675752, + "loss": 0.2585, + "num_input_tokens_seen": 63685776, + "step": 30485 + }, + { + "epoch": 4.973978301655926, + "grad_norm": 0.00035854606539942324, + "learning_rate": 0.039939630942171796, + "loss": 0.3355, + "num_input_tokens_seen": 63695376, + "step": 30490 + }, + { + "epoch": 4.974794028876744, + "grad_norm": 0.0006948731606826186, + "learning_rate": 0.03989961741027815, + "loss": 0.2979, + "num_input_tokens_seen": 63705200, + "step": 30495 + }, + { + "epoch": 4.975609756097561, + "grad_norm": 0.00041226233588531613, + "learning_rate": 0.03985962085724704, + "loss": 0.3152, + "num_input_tokens_seen": 63715568, + "step": 30500 + }, + { + "epoch": 4.976425483318378, + "grad_norm": 0.0008579884306527674, + "learning_rate": 0.03981964128924656, + "loss": 0.3219, + "num_input_tokens_seen": 63724816, + "step": 30505 + }, + { + "epoch": 4.977241210539196, + "grad_norm": 0.0006923860055394471, + "learning_rate": 0.03977967871244197, + "loss": 0.3017, + "num_input_tokens_seen": 63735072, + "step": 30510 + }, + { + "epoch": 4.978056937760013, + "grad_norm": 0.0006414463277906179, + "learning_rate": 0.03973973313299602, + "loss": 0.3265, + "num_input_tokens_seen": 63744944, + "step": 30515 + }, + { + "epoch": 4.97887266498083, + "grad_norm": 0.0006827125325798988, + "learning_rate": 0.0396998045570689, + "loss": 0.3563, + "num_input_tokens_seen": 63755520, + "step": 30520 + }, + { + "epoch": 4.9796883922016475, + "grad_norm": 0.0006058980943635106, + "learning_rate": 0.03965989299081798, + "loss": 0.3437, + "num_input_tokens_seen": 63765152, + "step": 30525 + }, + { + "epoch": 4.9805041194224655, + "grad_norm": 0.000595856283325702, + "learning_rate": 0.039619998440398235, + "loss": 0.2907, + "num_input_tokens_seen": 63775360, + "step": 30530 + }, + { + "epoch": 4.981319846643283, + "grad_norm": 0.00046803156146779656, + "learning_rate": 0.03958012091196184, + "loss": 0.3312, + "num_input_tokens_seen": 63786800, + "step": 30535 + }, + { + "epoch": 4.9821355738641, + "grad_norm": 0.0005712243728339672, + "learning_rate": 0.039540260411658396, + "loss": 0.3458, + "num_input_tokens_seen": 63797600, + "step": 30540 + }, + { + "epoch": 4.982951301084917, + "grad_norm": 0.0005898498930037022, + "learning_rate": 0.03950041694563496, + "loss": 0.3528, + "num_input_tokens_seen": 63808816, + "step": 30545 + }, + { + "epoch": 4.983767028305735, + "grad_norm": 0.00035936973290517926, + "learning_rate": 0.0394605905200358, + "loss": 0.3054, + "num_input_tokens_seen": 63819040, + "step": 30550 + }, + { + "epoch": 4.984582755526552, + "grad_norm": 0.0008004331612028182, + "learning_rate": 0.03942078114100272, + "loss": 0.2774, + "num_input_tokens_seen": 63829824, + "step": 30555 + }, + { + "epoch": 4.985398482747369, + "grad_norm": 0.00044139460078440607, + "learning_rate": 0.03938098881467485, + "loss": 0.3125, + "num_input_tokens_seen": 63839456, + "step": 30560 + }, + { + "epoch": 4.986214209968186, + "grad_norm": 0.0006863137823529541, + "learning_rate": 0.039341213547188586, + "loss": 0.3479, + "num_input_tokens_seen": 63849904, + "step": 30565 + }, + { + "epoch": 4.987029937189004, + "grad_norm": 0.0003892704553436488, + "learning_rate": 0.03930145534467782, + "loss": 0.3183, + "num_input_tokens_seen": 63861792, + "step": 30570 + }, + { + "epoch": 4.9878456644098215, + "grad_norm": 0.0003335669171065092, + "learning_rate": 0.0392617142132738, + "loss": 0.2884, + "num_input_tokens_seen": 63872528, + "step": 30575 + }, + { + "epoch": 4.988661391630639, + "grad_norm": 0.0008212125976569951, + "learning_rate": 0.03922199015910504, + "loss": 0.3276, + "num_input_tokens_seen": 63883104, + "step": 30580 + }, + { + "epoch": 4.989477118851456, + "grad_norm": 0.0005600991426035762, + "learning_rate": 0.039182283188297556, + "loss": 0.2902, + "num_input_tokens_seen": 63893264, + "step": 30585 + }, + { + "epoch": 4.990292846072274, + "grad_norm": 0.00031830053194426, + "learning_rate": 0.039142593306974595, + "loss": 0.2688, + "num_input_tokens_seen": 63904400, + "step": 30590 + }, + { + "epoch": 4.991108573293091, + "grad_norm": 0.00036382893449626863, + "learning_rate": 0.039102920521256856, + "loss": 0.3354, + "num_input_tokens_seen": 63915328, + "step": 30595 + }, + { + "epoch": 4.991924300513908, + "grad_norm": 0.0009444318129681051, + "learning_rate": 0.03906326483726243, + "loss": 0.3714, + "num_input_tokens_seen": 63926432, + "step": 30600 + }, + { + "epoch": 4.991924300513908, + "eval_loss": 0.3148377537727356, + "eval_runtime": 155.7635, + "eval_samples_per_second": 17.494, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 63926432, + "step": 30600 + }, + { + "epoch": 4.992740027734725, + "grad_norm": 0.000471162551548332, + "learning_rate": 0.039023626261106704, + "loss": 0.3463, + "num_input_tokens_seen": 63936720, + "step": 30605 + }, + { + "epoch": 4.993555754955543, + "grad_norm": 0.0004649118927773088, + "learning_rate": 0.03898400479890237, + "loss": 0.3462, + "num_input_tokens_seen": 63946576, + "step": 30610 + }, + { + "epoch": 4.99437148217636, + "grad_norm": 0.0010957315098494291, + "learning_rate": 0.038944400456759655, + "loss": 0.3284, + "num_input_tokens_seen": 63957312, + "step": 30615 + }, + { + "epoch": 4.995187209397177, + "grad_norm": 0.0004063040541950613, + "learning_rate": 0.038904813240785964, + "loss": 0.3127, + "num_input_tokens_seen": 63967360, + "step": 30620 + }, + { + "epoch": 4.9960029366179945, + "grad_norm": 0.0014765687519684434, + "learning_rate": 0.03886524315708621, + "loss": 0.3786, + "num_input_tokens_seen": 63976832, + "step": 30625 + }, + { + "epoch": 4.996818663838813, + "grad_norm": 0.0005052686901763082, + "learning_rate": 0.03882569021176255, + "loss": 0.3869, + "num_input_tokens_seen": 63988352, + "step": 30630 + }, + { + "epoch": 4.99763439105963, + "grad_norm": 0.0005545157473534346, + "learning_rate": 0.038786154410914535, + "loss": 0.3227, + "num_input_tokens_seen": 63998512, + "step": 30635 + }, + { + "epoch": 4.998450118280447, + "grad_norm": 0.0003288758161943406, + "learning_rate": 0.03874663576063917, + "loss": 0.2876, + "num_input_tokens_seen": 64007648, + "step": 30640 + }, + { + "epoch": 4.999265845501265, + "grad_norm": 0.0007316241390071809, + "learning_rate": 0.038707134267030624, + "loss": 0.3263, + "num_input_tokens_seen": 64019360, + "step": 30645 + }, + { + "epoch": 5.0, + "grad_norm": 0.0005328620318323374, + "learning_rate": 0.038667649936180555, + "loss": 0.2987, + "num_input_tokens_seen": 64029024, + "step": 30650 + }, + { + "epoch": 5.000815727220817, + "grad_norm": 0.0007875771261751652, + "learning_rate": 0.038628182774178, + "loss": 0.3363, + "num_input_tokens_seen": 64040032, + "step": 30655 + }, + { + "epoch": 5.001631454441635, + "grad_norm": 0.0013038733741268516, + "learning_rate": 0.038588732787109226, + "loss": 0.3927, + "num_input_tokens_seen": 64050192, + "step": 30660 + }, + { + "epoch": 5.002447181662452, + "grad_norm": 0.000550361699424684, + "learning_rate": 0.03854929998105795, + "loss": 0.292, + "num_input_tokens_seen": 64061088, + "step": 30665 + }, + { + "epoch": 5.003262908883269, + "grad_norm": 0.0004184274293947965, + "learning_rate": 0.03850988436210518, + "loss": 0.2837, + "num_input_tokens_seen": 64072320, + "step": 30670 + }, + { + "epoch": 5.0040786361040865, + "grad_norm": 0.0004637759120669216, + "learning_rate": 0.03847048593632933, + "loss": 0.3183, + "num_input_tokens_seen": 64083360, + "step": 30675 + }, + { + "epoch": 5.004894363324905, + "grad_norm": 0.00041515164775773883, + "learning_rate": 0.038431104709806096, + "loss": 0.3257, + "num_input_tokens_seen": 64093664, + "step": 30680 + }, + { + "epoch": 5.005710090545722, + "grad_norm": 0.00035500075318850577, + "learning_rate": 0.0383917406886086, + "loss": 0.2961, + "num_input_tokens_seen": 64104320, + "step": 30685 + }, + { + "epoch": 5.006525817766539, + "grad_norm": 0.00035562674747779965, + "learning_rate": 0.03835239387880722, + "loss": 0.3233, + "num_input_tokens_seen": 64114320, + "step": 30690 + }, + { + "epoch": 5.007341544987356, + "grad_norm": 0.0010657968232408166, + "learning_rate": 0.03831306428646979, + "loss": 0.3051, + "num_input_tokens_seen": 64126544, + "step": 30695 + }, + { + "epoch": 5.008157272208174, + "grad_norm": 0.0004323760513216257, + "learning_rate": 0.03827375191766135, + "loss": 0.3253, + "num_input_tokens_seen": 64136672, + "step": 30700 + }, + { + "epoch": 5.008972999428991, + "grad_norm": 0.00040081943734548986, + "learning_rate": 0.03823445677844446, + "loss": 0.3195, + "num_input_tokens_seen": 64147600, + "step": 30705 + }, + { + "epoch": 5.009788726649808, + "grad_norm": 0.0005742998328059912, + "learning_rate": 0.03819517887487881, + "loss": 0.3303, + "num_input_tokens_seen": 64157408, + "step": 30710 + }, + { + "epoch": 5.010604453870625, + "grad_norm": 0.0007049173582345247, + "learning_rate": 0.03815591821302161, + "loss": 0.3082, + "num_input_tokens_seen": 64166864, + "step": 30715 + }, + { + "epoch": 5.011420181091443, + "grad_norm": 0.0004499243514146656, + "learning_rate": 0.03811667479892739, + "loss": 0.2856, + "num_input_tokens_seen": 64176080, + "step": 30720 + }, + { + "epoch": 5.0122359083122605, + "grad_norm": 0.00047426746459677815, + "learning_rate": 0.03807744863864788, + "loss": 0.3356, + "num_input_tokens_seen": 64187056, + "step": 30725 + }, + { + "epoch": 5.013051635533078, + "grad_norm": 0.0004675767559092492, + "learning_rate": 0.03803823973823229, + "loss": 0.3067, + "num_input_tokens_seen": 64197456, + "step": 30730 + }, + { + "epoch": 5.013867362753895, + "grad_norm": 0.0003900045994669199, + "learning_rate": 0.03799904810372719, + "loss": 0.2914, + "num_input_tokens_seen": 64208176, + "step": 30735 + }, + { + "epoch": 5.014683089974713, + "grad_norm": 0.0005091322818771005, + "learning_rate": 0.03795987374117632, + "loss": 0.3092, + "num_input_tokens_seen": 64218672, + "step": 30740 + }, + { + "epoch": 5.01549881719553, + "grad_norm": 0.0002452031767461449, + "learning_rate": 0.03792071665662093, + "loss": 0.331, + "num_input_tokens_seen": 64228672, + "step": 30745 + }, + { + "epoch": 5.016314544416347, + "grad_norm": 0.000638594850897789, + "learning_rate": 0.03788157685609952, + "loss": 0.2622, + "num_input_tokens_seen": 64239488, + "step": 30750 + }, + { + "epoch": 5.017130271637164, + "grad_norm": 0.0005021225661039352, + "learning_rate": 0.037842454345647876, + "loss": 0.2962, + "num_input_tokens_seen": 64250384, + "step": 30755 + }, + { + "epoch": 5.017945998857982, + "grad_norm": 0.0004004590737167746, + "learning_rate": 0.03780334913129929, + "loss": 0.2943, + "num_input_tokens_seen": 64261456, + "step": 30760 + }, + { + "epoch": 5.018761726078799, + "grad_norm": 0.0005961099523119628, + "learning_rate": 0.037764261219084175, + "loss": 0.2967, + "num_input_tokens_seen": 64273424, + "step": 30765 + }, + { + "epoch": 5.0195774532996165, + "grad_norm": 0.0006669195718131959, + "learning_rate": 0.037725190615030414, + "loss": 0.2816, + "num_input_tokens_seen": 64283088, + "step": 30770 + }, + { + "epoch": 5.020393180520434, + "grad_norm": 0.00043717565131373703, + "learning_rate": 0.037686137325163224, + "loss": 0.289, + "num_input_tokens_seen": 64294176, + "step": 30775 + }, + { + "epoch": 5.021208907741252, + "grad_norm": 0.0006017321138642728, + "learning_rate": 0.037647101355505065, + "loss": 0.309, + "num_input_tokens_seen": 64303936, + "step": 30780 + }, + { + "epoch": 5.022024634962069, + "grad_norm": 0.000256339437328279, + "learning_rate": 0.03760808271207581, + "loss": 0.3767, + "num_input_tokens_seen": 64314736, + "step": 30785 + }, + { + "epoch": 5.022840362182886, + "grad_norm": 0.00031584384851157665, + "learning_rate": 0.03756908140089258, + "loss": 0.3894, + "num_input_tokens_seen": 64324976, + "step": 30790 + }, + { + "epoch": 5.023656089403703, + "grad_norm": 0.000547256728168577, + "learning_rate": 0.03753009742796989, + "loss": 0.3151, + "num_input_tokens_seen": 64337424, + "step": 30795 + }, + { + "epoch": 5.024471816624521, + "grad_norm": 0.0004399017780087888, + "learning_rate": 0.037491130799319615, + "loss": 0.32, + "num_input_tokens_seen": 64346032, + "step": 30800 + }, + { + "epoch": 5.024471816624521, + "eval_loss": 0.31425270438194275, + "eval_runtime": 155.8799, + "eval_samples_per_second": 17.481, + "eval_steps_per_second": 8.744, + "num_input_tokens_seen": 64346032, + "step": 30800 + }, + { + "epoch": 5.025287543845338, + "grad_norm": 0.000520929868798703, + "learning_rate": 0.03745218152095079, + "loss": 0.3012, + "num_input_tokens_seen": 64357072, + "step": 30805 + }, + { + "epoch": 5.026103271066155, + "grad_norm": 0.0005091555067338049, + "learning_rate": 0.037413249598869935, + "loss": 0.3136, + "num_input_tokens_seen": 64366352, + "step": 30810 + }, + { + "epoch": 5.026918998286972, + "grad_norm": 0.00037358520785346627, + "learning_rate": 0.037374335039080886, + "loss": 0.3397, + "num_input_tokens_seen": 64377504, + "step": 30815 + }, + { + "epoch": 5.02773472550779, + "grad_norm": 0.00041204874287359416, + "learning_rate": 0.037335437847584724, + "loss": 0.3048, + "num_input_tokens_seen": 64386512, + "step": 30820 + }, + { + "epoch": 5.028550452728608, + "grad_norm": 0.0005729135009460151, + "learning_rate": 0.03729655803037983, + "loss": 0.2725, + "num_input_tokens_seen": 64396400, + "step": 30825 + }, + { + "epoch": 5.029366179949425, + "grad_norm": 0.00046895683044567704, + "learning_rate": 0.03725769559346207, + "loss": 0.3269, + "num_input_tokens_seen": 64406864, + "step": 30830 + }, + { + "epoch": 5.030181907170242, + "grad_norm": 0.0007247003959491849, + "learning_rate": 0.03721885054282439, + "loss": 0.347, + "num_input_tokens_seen": 64415088, + "step": 30835 + }, + { + "epoch": 5.03099763439106, + "grad_norm": 0.0002780561335384846, + "learning_rate": 0.03718002288445731, + "loss": 0.329, + "num_input_tokens_seen": 64425744, + "step": 30840 + }, + { + "epoch": 5.031813361611877, + "grad_norm": 0.0004429547698237002, + "learning_rate": 0.03714121262434844, + "loss": 0.2664, + "num_input_tokens_seen": 64437008, + "step": 30845 + }, + { + "epoch": 5.032629088832694, + "grad_norm": 0.0006620916537940502, + "learning_rate": 0.037102419768482844, + "loss": 0.369, + "num_input_tokens_seen": 64446976, + "step": 30850 + }, + { + "epoch": 5.033444816053512, + "grad_norm": 0.0005593750975094736, + "learning_rate": 0.03706364432284293, + "loss": 0.3988, + "num_input_tokens_seen": 64457040, + "step": 30855 + }, + { + "epoch": 5.034260543274329, + "grad_norm": 0.0005859402008354664, + "learning_rate": 0.03702488629340828, + "loss": 0.3509, + "num_input_tokens_seen": 64467440, + "step": 30860 + }, + { + "epoch": 5.035076270495146, + "grad_norm": 0.0007041585049591959, + "learning_rate": 0.036986145686155915, + "loss": 0.3216, + "num_input_tokens_seen": 64478704, + "step": 30865 + }, + { + "epoch": 5.0358919977159635, + "grad_norm": 0.0005601344746537507, + "learning_rate": 0.036947422507060075, + "loss": 0.2797, + "num_input_tokens_seen": 64488992, + "step": 30870 + }, + { + "epoch": 5.0367077249367815, + "grad_norm": 0.0007772606331855059, + "learning_rate": 0.0369087167620924, + "loss": 0.2809, + "num_input_tokens_seen": 64500496, + "step": 30875 + }, + { + "epoch": 5.037523452157599, + "grad_norm": 0.0003932616673409939, + "learning_rate": 0.03687002845722183, + "loss": 0.358, + "num_input_tokens_seen": 64510416, + "step": 30880 + }, + { + "epoch": 5.038339179378416, + "grad_norm": 0.000310016650473699, + "learning_rate": 0.03683135759841451, + "loss": 0.2531, + "num_input_tokens_seen": 64521760, + "step": 30885 + }, + { + "epoch": 5.039154906599233, + "grad_norm": 0.0006241786759346724, + "learning_rate": 0.03679270419163406, + "loss": 0.3622, + "num_input_tokens_seen": 64531664, + "step": 30890 + }, + { + "epoch": 5.039970633820051, + "grad_norm": 0.0003739777603186667, + "learning_rate": 0.03675406824284127, + "loss": 0.3375, + "num_input_tokens_seen": 64541952, + "step": 30895 + }, + { + "epoch": 5.040786361040868, + "grad_norm": 0.000563561450690031, + "learning_rate": 0.03671544975799425, + "loss": 0.3091, + "num_input_tokens_seen": 64551488, + "step": 30900 + }, + { + "epoch": 5.041602088261685, + "grad_norm": 0.000466126628452912, + "learning_rate": 0.03667684874304854, + "loss": 0.2802, + "num_input_tokens_seen": 64561616, + "step": 30905 + }, + { + "epoch": 5.042417815482502, + "grad_norm": 0.00036121284938417375, + "learning_rate": 0.03663826520395683, + "loss": 0.3418, + "num_input_tokens_seen": 64571952, + "step": 30910 + }, + { + "epoch": 5.04323354270332, + "grad_norm": 0.0006005833274684846, + "learning_rate": 0.03659969914666922, + "loss": 0.2916, + "num_input_tokens_seen": 64582112, + "step": 30915 + }, + { + "epoch": 5.0440492699241375, + "grad_norm": 0.00041323804180137813, + "learning_rate": 0.036561150577133106, + "loss": 0.3463, + "num_input_tokens_seen": 64593200, + "step": 30920 + }, + { + "epoch": 5.044864997144955, + "grad_norm": 0.0005667421501129866, + "learning_rate": 0.036522619501293103, + "loss": 0.3283, + "num_input_tokens_seen": 64603936, + "step": 30925 + }, + { + "epoch": 5.045680724365772, + "grad_norm": 0.00033171859104186296, + "learning_rate": 0.03648410592509122, + "loss": 0.3065, + "num_input_tokens_seen": 64614096, + "step": 30930 + }, + { + "epoch": 5.04649645158659, + "grad_norm": 0.0006548987003043294, + "learning_rate": 0.03644560985446676, + "loss": 0.3156, + "num_input_tokens_seen": 64624928, + "step": 30935 + }, + { + "epoch": 5.047312178807407, + "grad_norm": 0.00101521797478199, + "learning_rate": 0.036407131295356256, + "loss": 0.3667, + "num_input_tokens_seen": 64635184, + "step": 30940 + }, + { + "epoch": 5.048127906028224, + "grad_norm": 0.000508954341057688, + "learning_rate": 0.03636867025369362, + "loss": 0.3389, + "num_input_tokens_seen": 64647344, + "step": 30945 + }, + { + "epoch": 5.048943633249041, + "grad_norm": 0.0004358003498055041, + "learning_rate": 0.03633022673540999, + "loss": 0.2666, + "num_input_tokens_seen": 64657680, + "step": 30950 + }, + { + "epoch": 5.049759360469859, + "grad_norm": 0.00030494286329485476, + "learning_rate": 0.03629180074643385, + "loss": 0.3272, + "num_input_tokens_seen": 64667856, + "step": 30955 + }, + { + "epoch": 5.050575087690676, + "grad_norm": 0.00033818010706454515, + "learning_rate": 0.03625339229269102, + "loss": 0.2848, + "num_input_tokens_seen": 64677952, + "step": 30960 + }, + { + "epoch": 5.051390814911493, + "grad_norm": 0.00038435327587649226, + "learning_rate": 0.036215001380104535, + "loss": 0.271, + "num_input_tokens_seen": 64690304, + "step": 30965 + }, + { + "epoch": 5.052206542132311, + "grad_norm": 0.0003110641264356673, + "learning_rate": 0.03617662801459471, + "loss": 0.3273, + "num_input_tokens_seen": 64700528, + "step": 30970 + }, + { + "epoch": 5.053022269353129, + "grad_norm": 0.0007848163368180394, + "learning_rate": 0.036138272202079276, + "loss": 0.281, + "num_input_tokens_seen": 64712144, + "step": 30975 + }, + { + "epoch": 5.053837996573946, + "grad_norm": 0.0006519880262203515, + "learning_rate": 0.036099933948473106, + "loss": 0.3715, + "num_input_tokens_seen": 64722720, + "step": 30980 + }, + { + "epoch": 5.054653723794763, + "grad_norm": 0.0005235272110439837, + "learning_rate": 0.03606161325968851, + "loss": 0.2786, + "num_input_tokens_seen": 64732896, + "step": 30985 + }, + { + "epoch": 5.05546945101558, + "grad_norm": 0.00034461598261259496, + "learning_rate": 0.03602331014163496, + "loss": 0.2791, + "num_input_tokens_seen": 64743184, + "step": 30990 + }, + { + "epoch": 5.056285178236398, + "grad_norm": 0.0007048335392028093, + "learning_rate": 0.035985024600219295, + "loss": 0.3738, + "num_input_tokens_seen": 64754608, + "step": 30995 + }, + { + "epoch": 5.057100905457215, + "grad_norm": 0.000488322926685214, + "learning_rate": 0.03594675664134569, + "loss": 0.3243, + "num_input_tokens_seen": 64764608, + "step": 31000 + }, + { + "epoch": 5.057100905457215, + "eval_loss": 0.3148408830165863, + "eval_runtime": 156.1366, + "eval_samples_per_second": 17.453, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 64764608, + "step": 31000 + }, + { + "epoch": 5.057916632678032, + "grad_norm": 0.0004280077409930527, + "learning_rate": 0.03590850627091545, + "loss": 0.3187, + "num_input_tokens_seen": 64774208, + "step": 31005 + }, + { + "epoch": 5.058732359898849, + "grad_norm": 0.0004988795262761414, + "learning_rate": 0.03587027349482731, + "loss": 0.3121, + "num_input_tokens_seen": 64785216, + "step": 31010 + }, + { + "epoch": 5.059548087119667, + "grad_norm": 0.00037917474401183426, + "learning_rate": 0.035832058318977275, + "loss": 0.3359, + "num_input_tokens_seen": 64796016, + "step": 31015 + }, + { + "epoch": 5.0603638143404845, + "grad_norm": 0.0005833249306306243, + "learning_rate": 0.03579386074925853, + "loss": 0.25, + "num_input_tokens_seen": 64806640, + "step": 31020 + }, + { + "epoch": 5.061179541561302, + "grad_norm": 0.0005600120057351887, + "learning_rate": 0.035755680791561696, + "loss": 0.2772, + "num_input_tokens_seen": 64815616, + "step": 31025 + }, + { + "epoch": 5.06199526878212, + "grad_norm": 0.0005808355053886771, + "learning_rate": 0.03571751845177454, + "loss": 0.3687, + "num_input_tokens_seen": 64825360, + "step": 31030 + }, + { + "epoch": 5.062810996002937, + "grad_norm": 0.0002984190359711647, + "learning_rate": 0.03567937373578225, + "loss": 0.3061, + "num_input_tokens_seen": 64836768, + "step": 31035 + }, + { + "epoch": 5.063626723223754, + "grad_norm": 0.0003696864587254822, + "learning_rate": 0.03564124664946711, + "loss": 0.3266, + "num_input_tokens_seen": 64845504, + "step": 31040 + }, + { + "epoch": 5.064442450444571, + "grad_norm": 0.0006675364566035569, + "learning_rate": 0.035603137198708924, + "loss": 0.3008, + "num_input_tokens_seen": 64856016, + "step": 31045 + }, + { + "epoch": 5.065258177665389, + "grad_norm": 0.0005519710830412805, + "learning_rate": 0.035565045389384514, + "loss": 0.266, + "num_input_tokens_seen": 64866528, + "step": 31050 + }, + { + "epoch": 5.066073904886206, + "grad_norm": 0.0004326421476434916, + "learning_rate": 0.03552697122736823, + "loss": 0.2187, + "num_input_tokens_seen": 64877344, + "step": 31055 + }, + { + "epoch": 5.066889632107023, + "grad_norm": 0.000543954607564956, + "learning_rate": 0.03548891471853153, + "loss": 0.2973, + "num_input_tokens_seen": 64887504, + "step": 31060 + }, + { + "epoch": 5.0677053593278405, + "grad_norm": 0.0004172314947936684, + "learning_rate": 0.03545087586874322, + "loss": 0.3073, + "num_input_tokens_seen": 64898288, + "step": 31065 + }, + { + "epoch": 5.0685210865486585, + "grad_norm": 0.00032277443096973, + "learning_rate": 0.03541285468386935, + "loss": 0.3099, + "num_input_tokens_seen": 64908656, + "step": 31070 + }, + { + "epoch": 5.069336813769476, + "grad_norm": 0.00035779093741439283, + "learning_rate": 0.03537485116977327, + "loss": 0.2907, + "num_input_tokens_seen": 64919536, + "step": 31075 + }, + { + "epoch": 5.070152540990293, + "grad_norm": 0.0006063788896426558, + "learning_rate": 0.03533686533231565, + "loss": 0.2347, + "num_input_tokens_seen": 64930032, + "step": 31080 + }, + { + "epoch": 5.07096826821111, + "grad_norm": 0.0004892715369351208, + "learning_rate": 0.0352988971773543, + "loss": 0.3113, + "num_input_tokens_seen": 64940064, + "step": 31085 + }, + { + "epoch": 5.071783995431928, + "grad_norm": 0.0004729666979983449, + "learning_rate": 0.03526094671074443, + "loss": 0.3409, + "num_input_tokens_seen": 64950464, + "step": 31090 + }, + { + "epoch": 5.072599722652745, + "grad_norm": 0.0005590454093180597, + "learning_rate": 0.03522301393833852, + "loss": 0.3319, + "num_input_tokens_seen": 64961280, + "step": 31095 + }, + { + "epoch": 5.073415449873562, + "grad_norm": 0.0009594811708666384, + "learning_rate": 0.035185098865986204, + "loss": 0.2784, + "num_input_tokens_seen": 64972640, + "step": 31100 + }, + { + "epoch": 5.074231177094379, + "grad_norm": 0.0004579368105623871, + "learning_rate": 0.03514720149953453, + "loss": 0.348, + "num_input_tokens_seen": 64983888, + "step": 31105 + }, + { + "epoch": 5.075046904315197, + "grad_norm": 0.00034792933729477227, + "learning_rate": 0.03510932184482773, + "loss": 0.3171, + "num_input_tokens_seen": 64994560, + "step": 31110 + }, + { + "epoch": 5.0758626315360145, + "grad_norm": 0.0004949025460518897, + "learning_rate": 0.03507145990770724, + "loss": 0.3232, + "num_input_tokens_seen": 65004768, + "step": 31115 + }, + { + "epoch": 5.076678358756832, + "grad_norm": 0.0006865697796456516, + "learning_rate": 0.035033615694011984, + "loss": 0.3088, + "num_input_tokens_seen": 65014816, + "step": 31120 + }, + { + "epoch": 5.077494085977649, + "grad_norm": 0.0004806236829608679, + "learning_rate": 0.03499578920957788, + "loss": 0.3441, + "num_input_tokens_seen": 65024832, + "step": 31125 + }, + { + "epoch": 5.078309813198467, + "grad_norm": 0.000985453138127923, + "learning_rate": 0.034957980460238375, + "loss": 0.3977, + "num_input_tokens_seen": 65034272, + "step": 31130 + }, + { + "epoch": 5.079125540419284, + "grad_norm": 0.0004907424445264041, + "learning_rate": 0.03492018945182393, + "loss": 0.3094, + "num_input_tokens_seen": 65044480, + "step": 31135 + }, + { + "epoch": 5.079941267640101, + "grad_norm": 0.0011780713684856892, + "learning_rate": 0.03488241619016247, + "loss": 0.3131, + "num_input_tokens_seen": 65054144, + "step": 31140 + }, + { + "epoch": 5.080756994860918, + "grad_norm": 0.0005467661540023983, + "learning_rate": 0.03484466068107913, + "loss": 0.3727, + "num_input_tokens_seen": 65065792, + "step": 31145 + }, + { + "epoch": 5.081572722081736, + "grad_norm": 0.0004193173663225025, + "learning_rate": 0.034806922930396195, + "loss": 0.3399, + "num_input_tokens_seen": 65075840, + "step": 31150 + }, + { + "epoch": 5.082388449302553, + "grad_norm": 0.0003204015374649316, + "learning_rate": 0.03476920294393337, + "loss": 0.3207, + "num_input_tokens_seen": 65086080, + "step": 31155 + }, + { + "epoch": 5.08320417652337, + "grad_norm": 0.0009305334533564746, + "learning_rate": 0.03473150072750755, + "loss": 0.3578, + "num_input_tokens_seen": 65096384, + "step": 31160 + }, + { + "epoch": 5.0840199037441876, + "grad_norm": 0.0004838818567804992, + "learning_rate": 0.03469381628693284, + "loss": 0.3545, + "num_input_tokens_seen": 65105280, + "step": 31165 + }, + { + "epoch": 5.084835630965006, + "grad_norm": 0.001175466226413846, + "learning_rate": 0.03465614962802072, + "loss": 0.3832, + "num_input_tokens_seen": 65115184, + "step": 31170 + }, + { + "epoch": 5.085651358185823, + "grad_norm": 0.0006345369038172066, + "learning_rate": 0.0346185007565798, + "loss": 0.2991, + "num_input_tokens_seen": 65127168, + "step": 31175 + }, + { + "epoch": 5.08646708540664, + "grad_norm": 0.00044317447463981807, + "learning_rate": 0.03458086967841609, + "loss": 0.3215, + "num_input_tokens_seen": 65138080, + "step": 31180 + }, + { + "epoch": 5.087282812627457, + "grad_norm": 0.0003891483065672219, + "learning_rate": 0.03454325639933266, + "loss": 0.3037, + "num_input_tokens_seen": 65147600, + "step": 31185 + }, + { + "epoch": 5.088098539848275, + "grad_norm": 0.0004967173445038497, + "learning_rate": 0.03450566092513007, + "loss": 0.2929, + "num_input_tokens_seen": 65158544, + "step": 31190 + }, + { + "epoch": 5.088914267069092, + "grad_norm": 0.000522166199516505, + "learning_rate": 0.034468083261605914, + "loss": 0.3495, + "num_input_tokens_seen": 65169168, + "step": 31195 + }, + { + "epoch": 5.089729994289909, + "grad_norm": 0.0003015281108673662, + "learning_rate": 0.03443052341455522, + "loss": 0.3121, + "num_input_tokens_seen": 65180560, + "step": 31200 + }, + { + "epoch": 5.089729994289909, + "eval_loss": 0.3152436912059784, + "eval_runtime": 156.1757, + "eval_samples_per_second": 17.448, + "eval_steps_per_second": 8.727, + "num_input_tokens_seen": 65180560, + "step": 31200 + }, + { + "epoch": 5.090545721510727, + "grad_norm": 0.0004716264083981514, + "learning_rate": 0.0343929813897701, + "loss": 0.3351, + "num_input_tokens_seen": 65191952, + "step": 31205 + }, + { + "epoch": 5.091361448731544, + "grad_norm": 0.0008114222437143326, + "learning_rate": 0.034355457193040125, + "loss": 0.3063, + "num_input_tokens_seen": 65202112, + "step": 31210 + }, + { + "epoch": 5.0921771759523615, + "grad_norm": 0.0003571318811737001, + "learning_rate": 0.03431795083015186, + "loss": 0.3118, + "num_input_tokens_seen": 65213296, + "step": 31215 + }, + { + "epoch": 5.092992903173179, + "grad_norm": 0.0007701806607656181, + "learning_rate": 0.03428046230688936, + "loss": 0.3182, + "num_input_tokens_seen": 65224384, + "step": 31220 + }, + { + "epoch": 5.093808630393997, + "grad_norm": 0.00046375952661037445, + "learning_rate": 0.034242991629033805, + "loss": 0.2918, + "num_input_tokens_seen": 65235296, + "step": 31225 + }, + { + "epoch": 5.094624357614814, + "grad_norm": 0.0006700679077766836, + "learning_rate": 0.03420553880236362, + "loss": 0.3116, + "num_input_tokens_seen": 65246192, + "step": 31230 + }, + { + "epoch": 5.095440084835631, + "grad_norm": 0.00065540277864784, + "learning_rate": 0.03416810383265449, + "loss": 0.3322, + "num_input_tokens_seen": 65256608, + "step": 31235 + }, + { + "epoch": 5.096255812056448, + "grad_norm": 0.0005809549475088716, + "learning_rate": 0.03413068672567944, + "loss": 0.3352, + "num_input_tokens_seen": 65266624, + "step": 31240 + }, + { + "epoch": 5.097071539277266, + "grad_norm": 0.0004725689650513232, + "learning_rate": 0.034093287487208565, + "loss": 0.3133, + "num_input_tokens_seen": 65277120, + "step": 31245 + }, + { + "epoch": 5.097887266498083, + "grad_norm": 0.000446677440777421, + "learning_rate": 0.03405590612300937, + "loss": 0.3147, + "num_input_tokens_seen": 65287280, + "step": 31250 + }, + { + "epoch": 5.0987029937189, + "grad_norm": 0.00046634775935672224, + "learning_rate": 0.03401854263884646, + "loss": 0.2869, + "num_input_tokens_seen": 65298608, + "step": 31255 + }, + { + "epoch": 5.0995187209397175, + "grad_norm": 0.0004196800000499934, + "learning_rate": 0.033981197040481824, + "loss": 0.3334, + "num_input_tokens_seen": 65309600, + "step": 31260 + }, + { + "epoch": 5.1003344481605355, + "grad_norm": 0.00041289773071184754, + "learning_rate": 0.03394386933367459, + "loss": 0.2964, + "num_input_tokens_seen": 65319648, + "step": 31265 + }, + { + "epoch": 5.101150175381353, + "grad_norm": 0.00032104618730954826, + "learning_rate": 0.033906559524181104, + "loss": 0.2959, + "num_input_tokens_seen": 65329232, + "step": 31270 + }, + { + "epoch": 5.10196590260217, + "grad_norm": 0.0005340961506590247, + "learning_rate": 0.033869267617755085, + "loss": 0.2958, + "num_input_tokens_seen": 65339040, + "step": 31275 + }, + { + "epoch": 5.102781629822987, + "grad_norm": 0.0004323841421864927, + "learning_rate": 0.0338319936201474, + "loss": 0.3191, + "num_input_tokens_seen": 65349104, + "step": 31280 + }, + { + "epoch": 5.103597357043805, + "grad_norm": 0.0008463225676678121, + "learning_rate": 0.033794737537106136, + "loss": 0.3816, + "num_input_tokens_seen": 65359648, + "step": 31285 + }, + { + "epoch": 5.104413084264622, + "grad_norm": 0.000616702251136303, + "learning_rate": 0.03375749937437671, + "loss": 0.3268, + "num_input_tokens_seen": 65368528, + "step": 31290 + }, + { + "epoch": 5.105228811485439, + "grad_norm": 0.00037578126648440957, + "learning_rate": 0.033720279137701634, + "loss": 0.2783, + "num_input_tokens_seen": 65378976, + "step": 31295 + }, + { + "epoch": 5.106044538706256, + "grad_norm": 0.00033302675001323223, + "learning_rate": 0.03368307683282078, + "loss": 0.2903, + "num_input_tokens_seen": 65389056, + "step": 31300 + }, + { + "epoch": 5.106860265927074, + "grad_norm": 0.0005300997290760279, + "learning_rate": 0.033645892465471235, + "loss": 0.3297, + "num_input_tokens_seen": 65399936, + "step": 31305 + }, + { + "epoch": 5.1076759931478914, + "grad_norm": 0.0005241132457740605, + "learning_rate": 0.03360872604138724, + "loss": 0.2699, + "num_input_tokens_seen": 65410416, + "step": 31310 + }, + { + "epoch": 5.108491720368709, + "grad_norm": 0.0006032430683262646, + "learning_rate": 0.03357157756630034, + "loss": 0.2825, + "num_input_tokens_seen": 65420656, + "step": 31315 + }, + { + "epoch": 5.109307447589526, + "grad_norm": 0.0006346912123262882, + "learning_rate": 0.033534447045939365, + "loss": 0.3092, + "num_input_tokens_seen": 65432000, + "step": 31320 + }, + { + "epoch": 5.110123174810344, + "grad_norm": 0.00039267135434783995, + "learning_rate": 0.03349733448603026, + "loss": 0.2763, + "num_input_tokens_seen": 65443632, + "step": 31325 + }, + { + "epoch": 5.110938902031161, + "grad_norm": 0.0007847754750400782, + "learning_rate": 0.03346023989229619, + "loss": 0.3358, + "num_input_tokens_seen": 65453312, + "step": 31330 + }, + { + "epoch": 5.111754629251978, + "grad_norm": 0.0007788154180161655, + "learning_rate": 0.03342316327045769, + "loss": 0.3208, + "num_input_tokens_seen": 65464128, + "step": 31335 + }, + { + "epoch": 5.112570356472795, + "grad_norm": 0.0010450661648064852, + "learning_rate": 0.033386104626232385, + "loss": 0.3004, + "num_input_tokens_seen": 65474128, + "step": 31340 + }, + { + "epoch": 5.113386083693613, + "grad_norm": 0.0008403785177506506, + "learning_rate": 0.03334906396533525, + "loss": 0.247, + "num_input_tokens_seen": 65484512, + "step": 31345 + }, + { + "epoch": 5.11420181091443, + "grad_norm": 0.0005386772681958973, + "learning_rate": 0.033312041293478326, + "loss": 0.3649, + "num_input_tokens_seen": 65495776, + "step": 31350 + }, + { + "epoch": 5.115017538135247, + "grad_norm": 0.00047031414578668773, + "learning_rate": 0.03327503661637103, + "loss": 0.3301, + "num_input_tokens_seen": 65506848, + "step": 31355 + }, + { + "epoch": 5.1158332653560645, + "grad_norm": 0.0004988271975889802, + "learning_rate": 0.03323804993971998, + "loss": 0.3686, + "num_input_tokens_seen": 65518288, + "step": 31360 + }, + { + "epoch": 5.1166489925768825, + "grad_norm": 0.0005217836005613208, + "learning_rate": 0.033201081269228924, + "loss": 0.3291, + "num_input_tokens_seen": 65529024, + "step": 31365 + }, + { + "epoch": 5.1174647197977, + "grad_norm": 0.00041359930764883757, + "learning_rate": 0.03316413061059895, + "loss": 0.2486, + "num_input_tokens_seen": 65538688, + "step": 31370 + }, + { + "epoch": 5.118280447018517, + "grad_norm": 0.0003885856713168323, + "learning_rate": 0.03312719796952827, + "loss": 0.2733, + "num_input_tokens_seen": 65547984, + "step": 31375 + }, + { + "epoch": 5.119096174239334, + "grad_norm": 0.0004335390403866768, + "learning_rate": 0.03309028335171236, + "loss": 0.2608, + "num_input_tokens_seen": 65558400, + "step": 31380 + }, + { + "epoch": 5.119911901460152, + "grad_norm": 0.0006170457345433533, + "learning_rate": 0.03305338676284398, + "loss": 0.2509, + "num_input_tokens_seen": 65568816, + "step": 31385 + }, + { + "epoch": 5.120727628680969, + "grad_norm": 0.0003919457085430622, + "learning_rate": 0.03301650820861296, + "loss": 0.275, + "num_input_tokens_seen": 65579536, + "step": 31390 + }, + { + "epoch": 5.121543355901786, + "grad_norm": 0.0005701230838894844, + "learning_rate": 0.03297964769470652, + "loss": 0.2936, + "num_input_tokens_seen": 65589984, + "step": 31395 + }, + { + "epoch": 5.122359083122603, + "grad_norm": 0.0007385958451777697, + "learning_rate": 0.032942805226808945, + "loss": 0.3838, + "num_input_tokens_seen": 65600032, + "step": 31400 + }, + { + "epoch": 5.122359083122603, + "eval_loss": 0.3142872750759125, + "eval_runtime": 156.1658, + "eval_samples_per_second": 17.449, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 65600032, + "step": 31400 + }, + { + "epoch": 5.123174810343421, + "grad_norm": 0.0009834232041612267, + "learning_rate": 0.03290598081060187, + "loss": 0.3667, + "num_input_tokens_seen": 65609696, + "step": 31405 + }, + { + "epoch": 5.1239905375642385, + "grad_norm": 0.00039845751598477364, + "learning_rate": 0.03286917445176407, + "loss": 0.3362, + "num_input_tokens_seen": 65620784, + "step": 31410 + }, + { + "epoch": 5.124806264785056, + "grad_norm": 0.0006502459291368723, + "learning_rate": 0.032832386155971456, + "loss": 0.3173, + "num_input_tokens_seen": 65631760, + "step": 31415 + }, + { + "epoch": 5.125621992005874, + "grad_norm": 0.00042925364687107503, + "learning_rate": 0.032795615928897334, + "loss": 0.3499, + "num_input_tokens_seen": 65642784, + "step": 31420 + }, + { + "epoch": 5.126437719226691, + "grad_norm": 0.00038189507904462516, + "learning_rate": 0.03275886377621215, + "loss": 0.3209, + "num_input_tokens_seen": 65653392, + "step": 31425 + }, + { + "epoch": 5.127253446447508, + "grad_norm": 0.0009967803489416838, + "learning_rate": 0.03272212970358348, + "loss": 0.2872, + "num_input_tokens_seen": 65662032, + "step": 31430 + }, + { + "epoch": 5.128069173668325, + "grad_norm": 0.0006051569944247603, + "learning_rate": 0.032685413716676215, + "loss": 0.2561, + "num_input_tokens_seen": 65672000, + "step": 31435 + }, + { + "epoch": 5.128884900889143, + "grad_norm": 0.0006751267937943339, + "learning_rate": 0.032648715821152474, + "loss": 0.2995, + "num_input_tokens_seen": 65681360, + "step": 31440 + }, + { + "epoch": 5.12970062810996, + "grad_norm": 0.0002422290126560256, + "learning_rate": 0.03261203602267143, + "loss": 0.2687, + "num_input_tokens_seen": 65691824, + "step": 31445 + }, + { + "epoch": 5.130516355330777, + "grad_norm": 0.0004638251557480544, + "learning_rate": 0.03257537432688966, + "loss": 0.31, + "num_input_tokens_seen": 65701424, + "step": 31450 + }, + { + "epoch": 5.1313320825515945, + "grad_norm": 0.000538808002602309, + "learning_rate": 0.03253873073946077, + "loss": 0.3239, + "num_input_tokens_seen": 65710192, + "step": 31455 + }, + { + "epoch": 5.1321478097724125, + "grad_norm": 0.00048721441999077797, + "learning_rate": 0.03250210526603572, + "loss": 0.343, + "num_input_tokens_seen": 65718128, + "step": 31460 + }, + { + "epoch": 5.13296353699323, + "grad_norm": 0.00028008664958178997, + "learning_rate": 0.03246549791226266, + "loss": 0.3377, + "num_input_tokens_seen": 65729200, + "step": 31465 + }, + { + "epoch": 5.133779264214047, + "grad_norm": 0.0007110710721462965, + "learning_rate": 0.03242890868378679, + "loss": 0.3402, + "num_input_tokens_seen": 65739136, + "step": 31470 + }, + { + "epoch": 5.134594991434864, + "grad_norm": 0.0009087944054044783, + "learning_rate": 0.03239233758625074, + "loss": 0.3172, + "num_input_tokens_seen": 65750576, + "step": 31475 + }, + { + "epoch": 5.135410718655682, + "grad_norm": 0.0006424824241548777, + "learning_rate": 0.032355784625294204, + "loss": 0.2527, + "num_input_tokens_seen": 65762304, + "step": 31480 + }, + { + "epoch": 5.136226445876499, + "grad_norm": 0.0009774903301149607, + "learning_rate": 0.03231924980655402, + "loss": 0.3106, + "num_input_tokens_seen": 65772672, + "step": 31485 + }, + { + "epoch": 5.137042173097316, + "grad_norm": 0.0005263859638944268, + "learning_rate": 0.032282733135664446, + "loss": 0.2885, + "num_input_tokens_seen": 65781568, + "step": 31490 + }, + { + "epoch": 5.137857900318133, + "grad_norm": 0.0019667113665491343, + "learning_rate": 0.03224623461825669, + "loss": 0.4109, + "num_input_tokens_seen": 65791568, + "step": 31495 + }, + { + "epoch": 5.138673627538951, + "grad_norm": 0.000354056857759133, + "learning_rate": 0.03220975425995937, + "loss": 0.302, + "num_input_tokens_seen": 65801920, + "step": 31500 + }, + { + "epoch": 5.139489354759768, + "grad_norm": 0.0005941757117398083, + "learning_rate": 0.032173292066398206, + "loss": 0.2541, + "num_input_tokens_seen": 65811280, + "step": 31505 + }, + { + "epoch": 5.1403050819805856, + "grad_norm": 0.00046009779907763004, + "learning_rate": 0.03213684804319606, + "loss": 0.3226, + "num_input_tokens_seen": 65821712, + "step": 31510 + }, + { + "epoch": 5.141120809201403, + "grad_norm": 0.0004886167589575052, + "learning_rate": 0.03210042219597312, + "loss": 0.3626, + "num_input_tokens_seen": 65832400, + "step": 31515 + }, + { + "epoch": 5.141936536422221, + "grad_norm": 0.0006054890109226108, + "learning_rate": 0.03206401453034675, + "loss": 0.3234, + "num_input_tokens_seen": 65843712, + "step": 31520 + }, + { + "epoch": 5.142752263643038, + "grad_norm": 0.0006856732652522624, + "learning_rate": 0.03202762505193136, + "loss": 0.3056, + "num_input_tokens_seen": 65853920, + "step": 31525 + }, + { + "epoch": 5.143567990863855, + "grad_norm": 0.0006548957317136228, + "learning_rate": 0.031991253766338754, + "loss": 0.2865, + "num_input_tokens_seen": 65863536, + "step": 31530 + }, + { + "epoch": 5.144383718084672, + "grad_norm": 0.00041536072967574, + "learning_rate": 0.03195490067917778, + "loss": 0.2793, + "num_input_tokens_seen": 65874544, + "step": 31535 + }, + { + "epoch": 5.14519944530549, + "grad_norm": 0.0007716960972175002, + "learning_rate": 0.03191856579605461, + "loss": 0.2953, + "num_input_tokens_seen": 65884400, + "step": 31540 + }, + { + "epoch": 5.146015172526307, + "grad_norm": 0.0004534078179858625, + "learning_rate": 0.031882249122572454, + "loss": 0.2988, + "num_input_tokens_seen": 65895632, + "step": 31545 + }, + { + "epoch": 5.146830899747124, + "grad_norm": 0.0008355012978427112, + "learning_rate": 0.03184595066433188, + "loss": 0.3083, + "num_input_tokens_seen": 65906416, + "step": 31550 + }, + { + "epoch": 5.1476466269679415, + "grad_norm": 0.0002945057931356132, + "learning_rate": 0.03180967042693049, + "loss": 0.3289, + "num_input_tokens_seen": 65917248, + "step": 31555 + }, + { + "epoch": 5.1484623541887595, + "grad_norm": 0.0004574393096845597, + "learning_rate": 0.03177340841596323, + "loss": 0.2374, + "num_input_tokens_seen": 65926992, + "step": 31560 + }, + { + "epoch": 5.149278081409577, + "grad_norm": 0.0003797830722760409, + "learning_rate": 0.03173716463702209, + "loss": 0.2893, + "num_input_tokens_seen": 65937008, + "step": 31565 + }, + { + "epoch": 5.150093808630394, + "grad_norm": 0.0005642495234496891, + "learning_rate": 0.03170093909569638, + "loss": 0.3303, + "num_input_tokens_seen": 65946864, + "step": 31570 + }, + { + "epoch": 5.150909535851211, + "grad_norm": 0.0005176176782697439, + "learning_rate": 0.03166473179757246, + "loss": 0.393, + "num_input_tokens_seen": 65958320, + "step": 31575 + }, + { + "epoch": 5.151725263072029, + "grad_norm": 0.0004848108219448477, + "learning_rate": 0.031628542748234005, + "loss": 0.3147, + "num_input_tokens_seen": 65966784, + "step": 31580 + }, + { + "epoch": 5.152540990292846, + "grad_norm": 0.000546481751371175, + "learning_rate": 0.03159237195326184, + "loss": 0.2705, + "num_input_tokens_seen": 65977376, + "step": 31585 + }, + { + "epoch": 5.153356717513663, + "grad_norm": 0.0005029173335060477, + "learning_rate": 0.031556219418233875, + "loss": 0.2624, + "num_input_tokens_seen": 65987472, + "step": 31590 + }, + { + "epoch": 5.154172444734481, + "grad_norm": 0.0005185109912417829, + "learning_rate": 0.03152008514872533, + "loss": 0.2986, + "num_input_tokens_seen": 65997968, + "step": 31595 + }, + { + "epoch": 5.154988171955298, + "grad_norm": 0.0004441578930709511, + "learning_rate": 0.03148396915030862, + "loss": 0.2727, + "num_input_tokens_seen": 66007440, + "step": 31600 + }, + { + "epoch": 5.154988171955298, + "eval_loss": 0.31366169452667236, + "eval_runtime": 155.8191, + "eval_samples_per_second": 17.488, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 66007440, + "step": 31600 + }, + { + "epoch": 5.1558038991761155, + "grad_norm": 0.0008033380145207047, + "learning_rate": 0.03144787142855318, + "loss": 0.2932, + "num_input_tokens_seen": 66018320, + "step": 31605 + }, + { + "epoch": 5.156619626396933, + "grad_norm": 0.0004687289474532008, + "learning_rate": 0.031411791989025835, + "loss": 0.2416, + "num_input_tokens_seen": 66027536, + "step": 31610 + }, + { + "epoch": 5.157435353617751, + "grad_norm": 0.0009154343279078603, + "learning_rate": 0.031375730837290394, + "loss": 0.3164, + "num_input_tokens_seen": 66036640, + "step": 31615 + }, + { + "epoch": 5.158251080838568, + "grad_norm": 0.00047512727905996144, + "learning_rate": 0.031339687978908015, + "loss": 0.2539, + "num_input_tokens_seen": 66046416, + "step": 31620 + }, + { + "epoch": 5.159066808059385, + "grad_norm": 0.000499594840221107, + "learning_rate": 0.03130366341943694, + "loss": 0.3785, + "num_input_tokens_seen": 66056624, + "step": 31625 + }, + { + "epoch": 5.159882535280202, + "grad_norm": 0.000521934824064374, + "learning_rate": 0.031267657164432555, + "loss": 0.3251, + "num_input_tokens_seen": 66066896, + "step": 31630 + }, + { + "epoch": 5.16069826250102, + "grad_norm": 0.0007666356978006661, + "learning_rate": 0.03123166921944752, + "loss": 0.4489, + "num_input_tokens_seen": 66077888, + "step": 31635 + }, + { + "epoch": 5.161513989721837, + "grad_norm": 0.0008163235033862293, + "learning_rate": 0.031195699590031666, + "loss": 0.2886, + "num_input_tokens_seen": 66088368, + "step": 31640 + }, + { + "epoch": 5.162329716942654, + "grad_norm": 0.0004674381925724447, + "learning_rate": 0.031159748281731885, + "loss": 0.3221, + "num_input_tokens_seen": 66099520, + "step": 31645 + }, + { + "epoch": 5.163145444163471, + "grad_norm": 0.0003649006539490074, + "learning_rate": 0.031123815300092394, + "loss": 0.2403, + "num_input_tokens_seen": 66110880, + "step": 31650 + }, + { + "epoch": 5.1639611713842895, + "grad_norm": 0.0009954925626516342, + "learning_rate": 0.031087900650654424, + "loss": 0.3723, + "num_input_tokens_seen": 66121072, + "step": 31655 + }, + { + "epoch": 5.164776898605107, + "grad_norm": 0.000436890113633126, + "learning_rate": 0.031052004338956534, + "loss": 0.2549, + "num_input_tokens_seen": 66130272, + "step": 31660 + }, + { + "epoch": 5.165592625825924, + "grad_norm": 0.0004597794613800943, + "learning_rate": 0.031016126370534407, + "loss": 0.3317, + "num_input_tokens_seen": 66140704, + "step": 31665 + }, + { + "epoch": 5.166408353046741, + "grad_norm": 0.00040982000064104795, + "learning_rate": 0.030980266750920804, + "loss": 0.321, + "num_input_tokens_seen": 66151952, + "step": 31670 + }, + { + "epoch": 5.167224080267559, + "grad_norm": 0.00033026939490810037, + "learning_rate": 0.030944425485645747, + "loss": 0.307, + "num_input_tokens_seen": 66161168, + "step": 31675 + }, + { + "epoch": 5.168039807488376, + "grad_norm": 0.0007900429191067815, + "learning_rate": 0.03090860258023647, + "loss": 0.4053, + "num_input_tokens_seen": 66170128, + "step": 31680 + }, + { + "epoch": 5.168855534709193, + "grad_norm": 0.0009317818912677467, + "learning_rate": 0.030872798040217236, + "loss": 0.3471, + "num_input_tokens_seen": 66181056, + "step": 31685 + }, + { + "epoch": 5.16967126193001, + "grad_norm": 0.0005661106551997364, + "learning_rate": 0.03083701187110964, + "loss": 0.2858, + "num_input_tokens_seen": 66192928, + "step": 31690 + }, + { + "epoch": 5.170486989150828, + "grad_norm": 0.0006307458970695734, + "learning_rate": 0.030801244078432294, + "loss": 0.3174, + "num_input_tokens_seen": 66201744, + "step": 31695 + }, + { + "epoch": 5.171302716371645, + "grad_norm": 0.0005020490498282015, + "learning_rate": 0.030765494667701024, + "loss": 0.2721, + "num_input_tokens_seen": 66211488, + "step": 31700 + }, + { + "epoch": 5.1721184435924625, + "grad_norm": 0.0007833954878151417, + "learning_rate": 0.030729763644428913, + "loss": 0.3778, + "num_input_tokens_seen": 66222560, + "step": 31705 + }, + { + "epoch": 5.17293417081328, + "grad_norm": 0.00037420442095026374, + "learning_rate": 0.030694051014126048, + "loss": 0.3198, + "num_input_tokens_seen": 66233088, + "step": 31710 + }, + { + "epoch": 5.173749898034098, + "grad_norm": 0.0009544520871713758, + "learning_rate": 0.030658356782299792, + "loss": 0.3665, + "num_input_tokens_seen": 66243952, + "step": 31715 + }, + { + "epoch": 5.174565625254915, + "grad_norm": 0.0011854746844619513, + "learning_rate": 0.030622680954454726, + "loss": 0.4378, + "num_input_tokens_seen": 66253872, + "step": 31720 + }, + { + "epoch": 5.175381352475732, + "grad_norm": 0.00037472377880476415, + "learning_rate": 0.030587023536092398, + "loss": 0.3848, + "num_input_tokens_seen": 66263152, + "step": 31725 + }, + { + "epoch": 5.176197079696549, + "grad_norm": 0.0005437866784632206, + "learning_rate": 0.03055138453271171, + "loss": 0.248, + "num_input_tokens_seen": 66273792, + "step": 31730 + }, + { + "epoch": 5.177012806917367, + "grad_norm": 0.0009171907440759242, + "learning_rate": 0.03051576394980858, + "loss": 0.361, + "num_input_tokens_seen": 66284208, + "step": 31735 + }, + { + "epoch": 5.177828534138184, + "grad_norm": 0.000468193058623001, + "learning_rate": 0.030480161792876187, + "loss": 0.3037, + "num_input_tokens_seen": 66294992, + "step": 31740 + }, + { + "epoch": 5.178644261359001, + "grad_norm": 0.0007505734683945775, + "learning_rate": 0.030444578067404846, + "loss": 0.389, + "num_input_tokens_seen": 66304032, + "step": 31745 + }, + { + "epoch": 5.1794599885798185, + "grad_norm": 0.0006172592402435839, + "learning_rate": 0.030409012778881975, + "loss": 0.336, + "num_input_tokens_seen": 66313328, + "step": 31750 + }, + { + "epoch": 5.1802757158006365, + "grad_norm": 0.000571569602470845, + "learning_rate": 0.030373465932792235, + "loss": 0.2816, + "num_input_tokens_seen": 66321696, + "step": 31755 + }, + { + "epoch": 5.181091443021454, + "grad_norm": 0.0003711612371262163, + "learning_rate": 0.030337937534617342, + "loss": 0.3251, + "num_input_tokens_seen": 66331744, + "step": 31760 + }, + { + "epoch": 5.181907170242271, + "grad_norm": 0.00042032761848531663, + "learning_rate": 0.030302427589836277, + "loss": 0.3393, + "num_input_tokens_seen": 66341456, + "step": 31765 + }, + { + "epoch": 5.182722897463089, + "grad_norm": 0.0004411601403262466, + "learning_rate": 0.030266936103925095, + "loss": 0.288, + "num_input_tokens_seen": 66350864, + "step": 31770 + }, + { + "epoch": 5.183538624683906, + "grad_norm": 0.00027004742878489196, + "learning_rate": 0.030231463082356982, + "loss": 0.2991, + "num_input_tokens_seen": 66361136, + "step": 31775 + }, + { + "epoch": 5.184354351904723, + "grad_norm": 0.0007879862678237259, + "learning_rate": 0.030196008530602367, + "loss": 0.2879, + "num_input_tokens_seen": 66371056, + "step": 31780 + }, + { + "epoch": 5.18517007912554, + "grad_norm": 0.0004063095839228481, + "learning_rate": 0.030160572454128842, + "loss": 0.287, + "num_input_tokens_seen": 66381744, + "step": 31785 + }, + { + "epoch": 5.185985806346358, + "grad_norm": 0.0006167874089442194, + "learning_rate": 0.03012515485840098, + "loss": 0.2863, + "num_input_tokens_seen": 66393056, + "step": 31790 + }, + { + "epoch": 5.186801533567175, + "grad_norm": 0.0004836043808609247, + "learning_rate": 0.030089755748880734, + "loss": 0.2593, + "num_input_tokens_seen": 66405392, + "step": 31795 + }, + { + "epoch": 5.1876172607879925, + "grad_norm": 0.0004360912134870887, + "learning_rate": 0.030054375131027003, + "loss": 0.3052, + "num_input_tokens_seen": 66416480, + "step": 31800 + }, + { + "epoch": 5.1876172607879925, + "eval_loss": 0.31376245617866516, + "eval_runtime": 155.7773, + "eval_samples_per_second": 17.493, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 66416480, + "step": 31800 + }, + { + "epoch": 5.18843298800881, + "grad_norm": 0.0007375528221018612, + "learning_rate": 0.030019013010295942, + "loss": 0.3204, + "num_input_tokens_seen": 66427040, + "step": 31805 + }, + { + "epoch": 5.189248715229628, + "grad_norm": 0.00044108089059591293, + "learning_rate": 0.029983669392140897, + "loss": 0.2561, + "num_input_tokens_seen": 66436864, + "step": 31810 + }, + { + "epoch": 5.190064442450445, + "grad_norm": 0.0003589593106880784, + "learning_rate": 0.029948344282012217, + "loss": 0.3504, + "num_input_tokens_seen": 66446400, + "step": 31815 + }, + { + "epoch": 5.190880169671262, + "grad_norm": 0.00047809878014959395, + "learning_rate": 0.029913037685357507, + "loss": 0.3224, + "num_input_tokens_seen": 66456784, + "step": 31820 + }, + { + "epoch": 5.191695896892079, + "grad_norm": 0.0005380227812565863, + "learning_rate": 0.029877749607621528, + "loss": 0.2946, + "num_input_tokens_seen": 66466704, + "step": 31825 + }, + { + "epoch": 5.192511624112897, + "grad_norm": 0.0006297735380940139, + "learning_rate": 0.029842480054246077, + "loss": 0.2565, + "num_input_tokens_seen": 66477280, + "step": 31830 + }, + { + "epoch": 5.193327351333714, + "grad_norm": 0.0006591947749257088, + "learning_rate": 0.02980722903067022, + "loss": 0.305, + "num_input_tokens_seen": 66487744, + "step": 31835 + }, + { + "epoch": 5.194143078554531, + "grad_norm": 0.0008920709369704127, + "learning_rate": 0.029771996542330113, + "loss": 0.3192, + "num_input_tokens_seen": 66498080, + "step": 31840 + }, + { + "epoch": 5.194958805775348, + "grad_norm": 0.00044085262925364077, + "learning_rate": 0.029736782594658954, + "loss": 0.2778, + "num_input_tokens_seen": 66509040, + "step": 31845 + }, + { + "epoch": 5.195774532996166, + "grad_norm": 0.000537365791387856, + "learning_rate": 0.029701587193087284, + "loss": 0.3393, + "num_input_tokens_seen": 66517952, + "step": 31850 + }, + { + "epoch": 5.196590260216984, + "grad_norm": 0.0004290697106625885, + "learning_rate": 0.0296664103430426, + "loss": 0.281, + "num_input_tokens_seen": 66529056, + "step": 31855 + }, + { + "epoch": 5.197405987437801, + "grad_norm": 0.0006101600592955947, + "learning_rate": 0.029631252049949652, + "loss": 0.3554, + "num_input_tokens_seen": 66538912, + "step": 31860 + }, + { + "epoch": 5.198221714658618, + "grad_norm": 0.0007331144297495484, + "learning_rate": 0.02959611231923031, + "loss": 0.3557, + "num_input_tokens_seen": 66548272, + "step": 31865 + }, + { + "epoch": 5.199037441879436, + "grad_norm": 0.0004577718209475279, + "learning_rate": 0.029560991156303507, + "loss": 0.3726, + "num_input_tokens_seen": 66559104, + "step": 31870 + }, + { + "epoch": 5.199853169100253, + "grad_norm": 0.0004315838741604239, + "learning_rate": 0.02952588856658544, + "loss": 0.3361, + "num_input_tokens_seen": 66568928, + "step": 31875 + }, + { + "epoch": 5.20066889632107, + "grad_norm": 0.0006561105255968869, + "learning_rate": 0.029490804555489296, + "loss": 0.3343, + "num_input_tokens_seen": 66579952, + "step": 31880 + }, + { + "epoch": 5.201484623541887, + "grad_norm": 0.00033292616717517376, + "learning_rate": 0.029455739128425484, + "loss": 0.3647, + "num_input_tokens_seen": 66589232, + "step": 31885 + }, + { + "epoch": 5.202300350762705, + "grad_norm": 0.000701139448210597, + "learning_rate": 0.029420692290801607, + "loss": 0.3139, + "num_input_tokens_seen": 66599728, + "step": 31890 + }, + { + "epoch": 5.203116077983522, + "grad_norm": 0.0005583761376328766, + "learning_rate": 0.02938566404802223, + "loss": 0.2909, + "num_input_tokens_seen": 66609392, + "step": 31895 + }, + { + "epoch": 5.2039318052043395, + "grad_norm": 0.0007335121626965702, + "learning_rate": 0.029350654405489195, + "loss": 0.4259, + "num_input_tokens_seen": 66618816, + "step": 31900 + }, + { + "epoch": 5.204747532425157, + "grad_norm": 0.0006309885066002607, + "learning_rate": 0.02931566336860145, + "loss": 0.3065, + "num_input_tokens_seen": 66629344, + "step": 31905 + }, + { + "epoch": 5.205563259645975, + "grad_norm": 0.0003847787738777697, + "learning_rate": 0.02928069094275505, + "loss": 0.3949, + "num_input_tokens_seen": 66641392, + "step": 31910 + }, + { + "epoch": 5.206378986866792, + "grad_norm": 0.0005778991035185754, + "learning_rate": 0.02924573713334314, + "loss": 0.3837, + "num_input_tokens_seen": 66652272, + "step": 31915 + }, + { + "epoch": 5.207194714087609, + "grad_norm": 0.0005062980926595628, + "learning_rate": 0.02921080194575603, + "loss": 0.3838, + "num_input_tokens_seen": 66662016, + "step": 31920 + }, + { + "epoch": 5.208010441308426, + "grad_norm": 0.0005735517479479313, + "learning_rate": 0.029175885385381177, + "loss": 0.3483, + "num_input_tokens_seen": 66672032, + "step": 31925 + }, + { + "epoch": 5.208826168529244, + "grad_norm": 0.0004659242113120854, + "learning_rate": 0.029140987457603223, + "loss": 0.3308, + "num_input_tokens_seen": 66684560, + "step": 31930 + }, + { + "epoch": 5.209641895750061, + "grad_norm": 0.000596717232838273, + "learning_rate": 0.029106108167803763, + "loss": 0.294, + "num_input_tokens_seen": 66694736, + "step": 31935 + }, + { + "epoch": 5.210457622970878, + "grad_norm": 0.0003193223965354264, + "learning_rate": 0.029071247521361674, + "loss": 0.393, + "num_input_tokens_seen": 66705392, + "step": 31940 + }, + { + "epoch": 5.211273350191696, + "grad_norm": 0.0004049596609547734, + "learning_rate": 0.029036405523652945, + "loss": 0.3441, + "num_input_tokens_seen": 66716144, + "step": 31945 + }, + { + "epoch": 5.2120890774125135, + "grad_norm": 0.0008771470165811479, + "learning_rate": 0.029001582180050577, + "loss": 0.3534, + "num_input_tokens_seen": 66726768, + "step": 31950 + }, + { + "epoch": 5.212904804633331, + "grad_norm": 0.0004737790732178837, + "learning_rate": 0.02896677749592482, + "loss": 0.2725, + "num_input_tokens_seen": 66737952, + "step": 31955 + }, + { + "epoch": 5.213720531854148, + "grad_norm": 0.00045871478505432606, + "learning_rate": 0.028931991476642938, + "loss": 0.3035, + "num_input_tokens_seen": 66746976, + "step": 31960 + }, + { + "epoch": 5.214536259074965, + "grad_norm": 0.00036834535421803594, + "learning_rate": 0.028897224127569412, + "loss": 0.3059, + "num_input_tokens_seen": 66757648, + "step": 31965 + }, + { + "epoch": 5.215351986295783, + "grad_norm": 0.0004691238282248378, + "learning_rate": 0.028862475454065832, + "loss": 0.3095, + "num_input_tokens_seen": 66768464, + "step": 31970 + }, + { + "epoch": 5.2161677135166, + "grad_norm": 0.0004378090088721365, + "learning_rate": 0.028827745461490806, + "loss": 0.3343, + "num_input_tokens_seen": 66778624, + "step": 31975 + }, + { + "epoch": 5.216983440737417, + "grad_norm": 0.0007211029878817499, + "learning_rate": 0.028793034155200212, + "loss": 0.2757, + "num_input_tokens_seen": 66789776, + "step": 31980 + }, + { + "epoch": 5.217799167958235, + "grad_norm": 0.0004106978594791144, + "learning_rate": 0.028758341540546944, + "loss": 0.3331, + "num_input_tokens_seen": 66798656, + "step": 31985 + }, + { + "epoch": 5.218614895179052, + "grad_norm": 0.0002678557066246867, + "learning_rate": 0.02872366762288098, + "loss": 0.3235, + "num_input_tokens_seen": 66808448, + "step": 31990 + }, + { + "epoch": 5.219430622399869, + "grad_norm": 0.0005803332314826548, + "learning_rate": 0.028689012407549567, + "loss": 0.3417, + "num_input_tokens_seen": 66818384, + "step": 31995 + }, + { + "epoch": 5.220246349620687, + "grad_norm": 0.0003857281117234379, + "learning_rate": 0.028654375899896892, + "loss": 0.3068, + "num_input_tokens_seen": 66829712, + "step": 32000 + }, + { + "epoch": 5.220246349620687, + "eval_loss": 0.3147255778312683, + "eval_runtime": 155.7848, + "eval_samples_per_second": 17.492, + "eval_steps_per_second": 8.749, + "num_input_tokens_seen": 66829712, + "step": 32000 + }, + { + "epoch": 5.221062076841505, + "grad_norm": 0.00044464273378252983, + "learning_rate": 0.02861975810526437, + "loss": 0.2453, + "num_input_tokens_seen": 66839920, + "step": 32005 + }, + { + "epoch": 5.221877804062322, + "grad_norm": 0.0004052765143569559, + "learning_rate": 0.02858515902899056, + "loss": 0.2775, + "num_input_tokens_seen": 66851664, + "step": 32010 + }, + { + "epoch": 5.222693531283139, + "grad_norm": 0.0005939858383499086, + "learning_rate": 0.028550578676410976, + "loss": 0.3328, + "num_input_tokens_seen": 66861888, + "step": 32015 + }, + { + "epoch": 5.223509258503956, + "grad_norm": 0.0004535141633823514, + "learning_rate": 0.02851601705285837, + "loss": 0.2924, + "num_input_tokens_seen": 66873440, + "step": 32020 + }, + { + "epoch": 5.224324985724774, + "grad_norm": 0.00035496504278853536, + "learning_rate": 0.028481474163662666, + "loss": 0.3802, + "num_input_tokens_seen": 66884336, + "step": 32025 + }, + { + "epoch": 5.225140712945591, + "grad_norm": 0.00036046584136784077, + "learning_rate": 0.028446950014150683, + "loss": 0.3167, + "num_input_tokens_seen": 66894608, + "step": 32030 + }, + { + "epoch": 5.225956440166408, + "grad_norm": 0.0006901885499246418, + "learning_rate": 0.028412444609646596, + "loss": 0.2914, + "num_input_tokens_seen": 66904624, + "step": 32035 + }, + { + "epoch": 5.226772167387225, + "grad_norm": 0.00043836692930199206, + "learning_rate": 0.028377957955471465, + "loss": 0.354, + "num_input_tokens_seen": 66916560, + "step": 32040 + }, + { + "epoch": 5.227587894608043, + "grad_norm": 0.0004171842592768371, + "learning_rate": 0.0283434900569436, + "loss": 0.3356, + "num_input_tokens_seen": 66926144, + "step": 32045 + }, + { + "epoch": 5.2284036218288605, + "grad_norm": 0.0003621653304435313, + "learning_rate": 0.028309040919378456, + "loss": 0.2616, + "num_input_tokens_seen": 66936800, + "step": 32050 + }, + { + "epoch": 5.229219349049678, + "grad_norm": 0.00055979989701882, + "learning_rate": 0.02827461054808848, + "loss": 0.3419, + "num_input_tokens_seen": 66947168, + "step": 32055 + }, + { + "epoch": 5.230035076270495, + "grad_norm": 0.00042580082663334906, + "learning_rate": 0.028240198948383186, + "loss": 0.2859, + "num_input_tokens_seen": 66958176, + "step": 32060 + }, + { + "epoch": 5.230850803491313, + "grad_norm": 0.0006555629661306739, + "learning_rate": 0.028205806125569402, + "loss": 0.3126, + "num_input_tokens_seen": 66969808, + "step": 32065 + }, + { + "epoch": 5.23166653071213, + "grad_norm": 0.0004258155240677297, + "learning_rate": 0.028171432084950834, + "loss": 0.3037, + "num_input_tokens_seen": 66981632, + "step": 32070 + }, + { + "epoch": 5.232482257932947, + "grad_norm": 0.0005562669830396771, + "learning_rate": 0.028137076831828478, + "loss": 0.3216, + "num_input_tokens_seen": 66992208, + "step": 32075 + }, + { + "epoch": 5.233297985153764, + "grad_norm": 0.0003749783500097692, + "learning_rate": 0.028102740371500238, + "loss": 0.3381, + "num_input_tokens_seen": 67002880, + "step": 32080 + }, + { + "epoch": 5.234113712374582, + "grad_norm": 0.00037974893348291516, + "learning_rate": 0.0280684227092613, + "loss": 0.3579, + "num_input_tokens_seen": 67014272, + "step": 32085 + }, + { + "epoch": 5.234929439595399, + "grad_norm": 0.0009637001785449684, + "learning_rate": 0.02803412385040392, + "loss": 0.3208, + "num_input_tokens_seen": 67024096, + "step": 32090 + }, + { + "epoch": 5.2357451668162165, + "grad_norm": 0.0006422097212634981, + "learning_rate": 0.027999843800217306, + "loss": 0.3477, + "num_input_tokens_seen": 67033792, + "step": 32095 + }, + { + "epoch": 5.236560894037034, + "grad_norm": 0.0006315424107015133, + "learning_rate": 0.027965582563987932, + "loss": 0.3652, + "num_input_tokens_seen": 67044784, + "step": 32100 + }, + { + "epoch": 5.237376621257852, + "grad_norm": 0.0006825654418207705, + "learning_rate": 0.027931340146999346, + "loss": 0.3023, + "num_input_tokens_seen": 67056784, + "step": 32105 + }, + { + "epoch": 5.238192348478669, + "grad_norm": 0.00043789460323750973, + "learning_rate": 0.02789711655453208, + "loss": 0.3026, + "num_input_tokens_seen": 67067072, + "step": 32110 + }, + { + "epoch": 5.239008075699486, + "grad_norm": 0.0003104073693975806, + "learning_rate": 0.02786291179186392, + "loss": 0.3103, + "num_input_tokens_seen": 67076736, + "step": 32115 + }, + { + "epoch": 5.239823802920303, + "grad_norm": 0.0005406615091487765, + "learning_rate": 0.02782872586426961, + "loss": 0.2941, + "num_input_tokens_seen": 67087712, + "step": 32120 + }, + { + "epoch": 5.240639530141121, + "grad_norm": 0.0004122167592868209, + "learning_rate": 0.027794558777021083, + "loss": 0.3274, + "num_input_tokens_seen": 67097968, + "step": 32125 + }, + { + "epoch": 5.241455257361938, + "grad_norm": 0.0007172158802859485, + "learning_rate": 0.02776041053538734, + "loss": 0.3151, + "num_input_tokens_seen": 67108800, + "step": 32130 + }, + { + "epoch": 5.242270984582755, + "grad_norm": 0.0008071644115261734, + "learning_rate": 0.027726281144634407, + "loss": 0.2995, + "num_input_tokens_seen": 67119088, + "step": 32135 + }, + { + "epoch": 5.243086711803572, + "grad_norm": 0.00047389112296514213, + "learning_rate": 0.02769217061002552, + "loss": 0.3425, + "num_input_tokens_seen": 67130176, + "step": 32140 + }, + { + "epoch": 5.2439024390243905, + "grad_norm": 0.0006457612616941333, + "learning_rate": 0.027658078936820967, + "loss": 0.3025, + "num_input_tokens_seen": 67140512, + "step": 32145 + }, + { + "epoch": 5.244718166245208, + "grad_norm": 0.0003817129763774574, + "learning_rate": 0.02762400613027805, + "loss": 0.3368, + "num_input_tokens_seen": 67150560, + "step": 32150 + }, + { + "epoch": 5.245533893466025, + "grad_norm": 0.0003844605525955558, + "learning_rate": 0.027589952195651295, + "loss": 0.321, + "num_input_tokens_seen": 67160880, + "step": 32155 + }, + { + "epoch": 5.246349620686843, + "grad_norm": 0.0006442402373068035, + "learning_rate": 0.027555917138192186, + "loss": 0.256, + "num_input_tokens_seen": 67171072, + "step": 32160 + }, + { + "epoch": 5.24716534790766, + "grad_norm": 0.0004357521247584373, + "learning_rate": 0.027521900963149375, + "loss": 0.2976, + "num_input_tokens_seen": 67182512, + "step": 32165 + }, + { + "epoch": 5.247981075128477, + "grad_norm": 0.000362969673005864, + "learning_rate": 0.027487903675768633, + "loss": 0.3445, + "num_input_tokens_seen": 67192528, + "step": 32170 + }, + { + "epoch": 5.248796802349294, + "grad_norm": 0.0004629853938240558, + "learning_rate": 0.027453925281292677, + "loss": 0.325, + "num_input_tokens_seen": 67201712, + "step": 32175 + }, + { + "epoch": 5.249612529570112, + "grad_norm": 0.0009802752174437046, + "learning_rate": 0.027419965784961475, + "loss": 0.3112, + "num_input_tokens_seen": 67211488, + "step": 32180 + }, + { + "epoch": 5.250428256790929, + "grad_norm": 0.000595934921875596, + "learning_rate": 0.027386025192012015, + "loss": 0.3298, + "num_input_tokens_seen": 67223168, + "step": 32185 + }, + { + "epoch": 5.251243984011746, + "grad_norm": 0.0006576524465344846, + "learning_rate": 0.027352103507678277, + "loss": 0.3842, + "num_input_tokens_seen": 67233920, + "step": 32190 + }, + { + "epoch": 5.2520597112325635, + "grad_norm": 0.00040694745257496834, + "learning_rate": 0.027318200737191527, + "loss": 0.3192, + "num_input_tokens_seen": 67244016, + "step": 32195 + }, + { + "epoch": 5.252875438453382, + "grad_norm": 0.0003454200050327927, + "learning_rate": 0.027284316885779935, + "loss": 0.294, + "num_input_tokens_seen": 67253936, + "step": 32200 + }, + { + "epoch": 5.252875438453382, + "eval_loss": 0.3162870407104492, + "eval_runtime": 155.8784, + "eval_samples_per_second": 17.482, + "eval_steps_per_second": 8.744, + "num_input_tokens_seen": 67253936, + "step": 32200 + }, + { + "epoch": 5.253691165674199, + "grad_norm": 0.0003570525732357055, + "learning_rate": 0.027250451958668785, + "loss": 0.3416, + "num_input_tokens_seen": 67263344, + "step": 32205 + }, + { + "epoch": 5.254506892895016, + "grad_norm": 0.00036706411628983915, + "learning_rate": 0.027216605961080536, + "loss": 0.3358, + "num_input_tokens_seen": 67272144, + "step": 32210 + }, + { + "epoch": 5.255322620115833, + "grad_norm": 0.00043645038385875523, + "learning_rate": 0.02718277889823461, + "loss": 0.322, + "num_input_tokens_seen": 67282128, + "step": 32215 + }, + { + "epoch": 5.256138347336651, + "grad_norm": 0.0005165535840205848, + "learning_rate": 0.027148970775347604, + "loss": 0.3014, + "num_input_tokens_seen": 67293008, + "step": 32220 + }, + { + "epoch": 5.256954074557468, + "grad_norm": 0.00036051744245924056, + "learning_rate": 0.027115181597633174, + "loss": 0.3098, + "num_input_tokens_seen": 67303792, + "step": 32225 + }, + { + "epoch": 5.257769801778285, + "grad_norm": 0.0006750362226739526, + "learning_rate": 0.027081411370301976, + "loss": 0.3074, + "num_input_tokens_seen": 67315088, + "step": 32230 + }, + { + "epoch": 5.258585528999102, + "grad_norm": 0.0006757008959539235, + "learning_rate": 0.027047660098561875, + "loss": 0.2914, + "num_input_tokens_seen": 67325984, + "step": 32235 + }, + { + "epoch": 5.25940125621992, + "grad_norm": 0.001047865953296423, + "learning_rate": 0.02701392778761766, + "loss": 0.3579, + "num_input_tokens_seen": 67336736, + "step": 32240 + }, + { + "epoch": 5.2602169834407375, + "grad_norm": 0.0006482604658231139, + "learning_rate": 0.02698021444267133, + "loss": 0.3166, + "num_input_tokens_seen": 67348336, + "step": 32245 + }, + { + "epoch": 5.261032710661555, + "grad_norm": 0.0004054179007653147, + "learning_rate": 0.026946520068921915, + "loss": 0.2696, + "num_input_tokens_seen": 67358048, + "step": 32250 + }, + { + "epoch": 5.261848437882372, + "grad_norm": 0.00046109757386147976, + "learning_rate": 0.02691284467156547, + "loss": 0.2899, + "num_input_tokens_seen": 67367520, + "step": 32255 + }, + { + "epoch": 5.26266416510319, + "grad_norm": 0.0005775429308414459, + "learning_rate": 0.026879188255795182, + "loss": 0.3464, + "num_input_tokens_seen": 67379680, + "step": 32260 + }, + { + "epoch": 5.263479892324007, + "grad_norm": 0.0002920945407822728, + "learning_rate": 0.026845550826801328, + "loss": 0.3131, + "num_input_tokens_seen": 67389632, + "step": 32265 + }, + { + "epoch": 5.264295619544824, + "grad_norm": 0.00041142298141494393, + "learning_rate": 0.02681193238977121, + "loss": 0.2923, + "num_input_tokens_seen": 67400288, + "step": 32270 + }, + { + "epoch": 5.265111346765641, + "grad_norm": 0.0004878930631093681, + "learning_rate": 0.026778332949889145, + "loss": 0.3392, + "num_input_tokens_seen": 67410336, + "step": 32275 + }, + { + "epoch": 5.265927073986459, + "grad_norm": 0.0005380587535910308, + "learning_rate": 0.026744752512336673, + "loss": 0.3443, + "num_input_tokens_seen": 67420944, + "step": 32280 + }, + { + "epoch": 5.266742801207276, + "grad_norm": 0.0007341621094383299, + "learning_rate": 0.02671119108229225, + "loss": 0.2552, + "num_input_tokens_seen": 67431184, + "step": 32285 + }, + { + "epoch": 5.2675585284280935, + "grad_norm": 0.0004207367601338774, + "learning_rate": 0.026677648664931556, + "loss": 0.3027, + "num_input_tokens_seen": 67442512, + "step": 32290 + }, + { + "epoch": 5.268374255648911, + "grad_norm": 0.0003748438321053982, + "learning_rate": 0.026644125265427154, + "loss": 0.3705, + "num_input_tokens_seen": 67454128, + "step": 32295 + }, + { + "epoch": 5.269189982869729, + "grad_norm": 0.0005186726921238005, + "learning_rate": 0.026610620888948822, + "loss": 0.3332, + "num_input_tokens_seen": 67462240, + "step": 32300 + }, + { + "epoch": 5.270005710090546, + "grad_norm": 0.0007008736720308661, + "learning_rate": 0.026577135540663408, + "loss": 0.2914, + "num_input_tokens_seen": 67473200, + "step": 32305 + }, + { + "epoch": 5.270821437311363, + "grad_norm": 0.00038446616963483393, + "learning_rate": 0.026543669225734673, + "loss": 0.2974, + "num_input_tokens_seen": 67484400, + "step": 32310 + }, + { + "epoch": 5.27163716453218, + "grad_norm": 0.00033952342346310616, + "learning_rate": 0.02651022194932363, + "loss": 0.3366, + "num_input_tokens_seen": 67494656, + "step": 32315 + }, + { + "epoch": 5.272452891752998, + "grad_norm": 0.0005345280515030026, + "learning_rate": 0.026476793716588194, + "loss": 0.3102, + "num_input_tokens_seen": 67504944, + "step": 32320 + }, + { + "epoch": 5.273268618973815, + "grad_norm": 0.0006953041884116828, + "learning_rate": 0.026443384532683467, + "loss": 0.2947, + "num_input_tokens_seen": 67516272, + "step": 32325 + }, + { + "epoch": 5.274084346194632, + "grad_norm": 0.0008469204767607152, + "learning_rate": 0.026409994402761584, + "loss": 0.3189, + "num_input_tokens_seen": 67526672, + "step": 32330 + }, + { + "epoch": 5.27490007341545, + "grad_norm": 0.0004655686498153955, + "learning_rate": 0.026376623331971653, + "loss": 0.2908, + "num_input_tokens_seen": 67537008, + "step": 32335 + }, + { + "epoch": 5.275715800636267, + "grad_norm": 0.0005568927736021578, + "learning_rate": 0.026343271325459997, + "loss": 0.4266, + "num_input_tokens_seen": 67548528, + "step": 32340 + }, + { + "epoch": 5.276531527857085, + "grad_norm": 0.00038840511115267873, + "learning_rate": 0.02630993838836987, + "loss": 0.409, + "num_input_tokens_seen": 67558592, + "step": 32345 + }, + { + "epoch": 5.277347255077902, + "grad_norm": 0.0005034767091274261, + "learning_rate": 0.026276624525841584, + "loss": 0.3039, + "num_input_tokens_seen": 67567600, + "step": 32350 + }, + { + "epoch": 5.27816298229872, + "grad_norm": 0.0003189335693605244, + "learning_rate": 0.026243329743012637, + "loss": 0.2956, + "num_input_tokens_seen": 67579440, + "step": 32355 + }, + { + "epoch": 5.278978709519537, + "grad_norm": 0.0004148586594965309, + "learning_rate": 0.026210054045017438, + "loss": 0.292, + "num_input_tokens_seen": 67591344, + "step": 32360 + }, + { + "epoch": 5.279794436740354, + "grad_norm": 0.0005790753639303148, + "learning_rate": 0.02617679743698755, + "loss": 0.2791, + "num_input_tokens_seen": 67602144, + "step": 32365 + }, + { + "epoch": 5.280610163961171, + "grad_norm": 0.0004099970974493772, + "learning_rate": 0.02614355992405158, + "loss": 0.314, + "num_input_tokens_seen": 67612032, + "step": 32370 + }, + { + "epoch": 5.281425891181989, + "grad_norm": 0.0005106358439661562, + "learning_rate": 0.026110341511335115, + "loss": 0.3426, + "num_input_tokens_seen": 67621120, + "step": 32375 + }, + { + "epoch": 5.282241618402806, + "grad_norm": 0.0004010779084637761, + "learning_rate": 0.02607714220396093, + "loss": 0.3219, + "num_input_tokens_seen": 67631680, + "step": 32380 + }, + { + "epoch": 5.283057345623623, + "grad_norm": 0.0004339400620665401, + "learning_rate": 0.02604396200704869, + "loss": 0.2907, + "num_input_tokens_seen": 67641568, + "step": 32385 + }, + { + "epoch": 5.2838730728444405, + "grad_norm": 0.0004326174675952643, + "learning_rate": 0.02601080092571523, + "loss": 0.3245, + "num_input_tokens_seen": 67652336, + "step": 32390 + }, + { + "epoch": 5.2846888000652585, + "grad_norm": 0.0004256491956766695, + "learning_rate": 0.025977658965074455, + "loss": 0.2955, + "num_input_tokens_seen": 67662848, + "step": 32395 + }, + { + "epoch": 5.285504527286076, + "grad_norm": 0.00046442385064437985, + "learning_rate": 0.02594453613023719, + "loss": 0.3119, + "num_input_tokens_seen": 67674048, + "step": 32400 + }, + { + "epoch": 5.285504527286076, + "eval_loss": 0.31453070044517517, + "eval_runtime": 156.1295, + "eval_samples_per_second": 17.453, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 67674048, + "step": 32400 + }, + { + "epoch": 5.286320254506893, + "grad_norm": 0.0006521029863506556, + "learning_rate": 0.025911432426311443, + "loss": 0.3269, + "num_input_tokens_seen": 67685024, + "step": 32405 + }, + { + "epoch": 5.28713598172771, + "grad_norm": 0.00034990633139386773, + "learning_rate": 0.025878347858402234, + "loss": 0.2806, + "num_input_tokens_seen": 67695840, + "step": 32410 + }, + { + "epoch": 5.287951708948528, + "grad_norm": 0.0008494284702464938, + "learning_rate": 0.025845282431611598, + "loss": 0.2977, + "num_input_tokens_seen": 67705408, + "step": 32415 + }, + { + "epoch": 5.288767436169345, + "grad_norm": 0.0004823036433663219, + "learning_rate": 0.025812236151038608, + "loss": 0.3567, + "num_input_tokens_seen": 67715168, + "step": 32420 + }, + { + "epoch": 5.289583163390162, + "grad_norm": 0.0004665862943511456, + "learning_rate": 0.025779209021779468, + "loss": 0.329, + "num_input_tokens_seen": 67725184, + "step": 32425 + }, + { + "epoch": 5.290398890610979, + "grad_norm": 0.0005366674158722162, + "learning_rate": 0.025746201048927324, + "loss": 0.3283, + "num_input_tokens_seen": 67735760, + "step": 32430 + }, + { + "epoch": 5.291214617831797, + "grad_norm": 0.0005132885999046266, + "learning_rate": 0.025713212237572485, + "loss": 0.3011, + "num_input_tokens_seen": 67746560, + "step": 32435 + }, + { + "epoch": 5.2920303450526145, + "grad_norm": 0.0003913214022759348, + "learning_rate": 0.025680242592802164, + "loss": 0.312, + "num_input_tokens_seen": 67757488, + "step": 32440 + }, + { + "epoch": 5.292846072273432, + "grad_norm": 0.00037423119647428393, + "learning_rate": 0.02564729211970073, + "loss": 0.3298, + "num_input_tokens_seen": 67768752, + "step": 32445 + }, + { + "epoch": 5.293661799494249, + "grad_norm": 0.00047431830898858607, + "learning_rate": 0.025614360823349617, + "loss": 0.2944, + "num_input_tokens_seen": 67778272, + "step": 32450 + }, + { + "epoch": 5.294477526715067, + "grad_norm": 0.0007155067287385464, + "learning_rate": 0.025581448708827146, + "loss": 0.2611, + "num_input_tokens_seen": 67790016, + "step": 32455 + }, + { + "epoch": 5.295293253935884, + "grad_norm": 0.0007125789998099208, + "learning_rate": 0.025548555781208876, + "loss": 0.3452, + "num_input_tokens_seen": 67799584, + "step": 32460 + }, + { + "epoch": 5.296108981156701, + "grad_norm": 0.0008429413428530097, + "learning_rate": 0.02551568204556721, + "loss": 0.3182, + "num_input_tokens_seen": 67809680, + "step": 32465 + }, + { + "epoch": 5.296924708377518, + "grad_norm": 0.0005256925360299647, + "learning_rate": 0.02548282750697173, + "loss": 0.2913, + "num_input_tokens_seen": 67819696, + "step": 32470 + }, + { + "epoch": 5.297740435598336, + "grad_norm": 0.0003255569317843765, + "learning_rate": 0.02544999217048909, + "loss": 0.322, + "num_input_tokens_seen": 67829616, + "step": 32475 + }, + { + "epoch": 5.298556162819153, + "grad_norm": 0.00037170457653701305, + "learning_rate": 0.025417176041182793, + "loss": 0.3076, + "num_input_tokens_seen": 67838944, + "step": 32480 + }, + { + "epoch": 5.2993718900399704, + "grad_norm": 0.0006436161929741502, + "learning_rate": 0.025384379124113596, + "loss": 0.3344, + "num_input_tokens_seen": 67848096, + "step": 32485 + }, + { + "epoch": 5.300187617260788, + "grad_norm": 0.00040432572131976485, + "learning_rate": 0.025351601424339124, + "loss": 0.2854, + "num_input_tokens_seen": 67859296, + "step": 32490 + }, + { + "epoch": 5.301003344481606, + "grad_norm": 0.0007360586896538734, + "learning_rate": 0.025318842946914184, + "loss": 0.2883, + "num_input_tokens_seen": 67870448, + "step": 32495 + }, + { + "epoch": 5.301819071702423, + "grad_norm": 0.002506938064470887, + "learning_rate": 0.025286103696890494, + "loss": 0.3311, + "num_input_tokens_seen": 67882608, + "step": 32500 + }, + { + "epoch": 5.30263479892324, + "grad_norm": 0.00035522429971024394, + "learning_rate": 0.025253383679316836, + "loss": 0.2928, + "num_input_tokens_seen": 67893648, + "step": 32505 + }, + { + "epoch": 5.303450526144058, + "grad_norm": 0.0005998575361445546, + "learning_rate": 0.025220682899239077, + "loss": 0.3863, + "num_input_tokens_seen": 67903344, + "step": 32510 + }, + { + "epoch": 5.304266253364875, + "grad_norm": 0.0003091816324740648, + "learning_rate": 0.02518800136170013, + "loss": 0.3039, + "num_input_tokens_seen": 67915104, + "step": 32515 + }, + { + "epoch": 5.305081980585692, + "grad_norm": 0.0007363061886280775, + "learning_rate": 0.02515533907173981, + "loss": 0.2869, + "num_input_tokens_seen": 67924016, + "step": 32520 + }, + { + "epoch": 5.305897707806509, + "grad_norm": 0.0004755309782922268, + "learning_rate": 0.025122696034395115, + "loss": 0.3587, + "num_input_tokens_seen": 67935280, + "step": 32525 + }, + { + "epoch": 5.306713435027326, + "grad_norm": 0.0006008914206176996, + "learning_rate": 0.025090072254700023, + "loss": 0.3566, + "num_input_tokens_seen": 67945904, + "step": 32530 + }, + { + "epoch": 5.307529162248144, + "grad_norm": 0.0003729382005985826, + "learning_rate": 0.025057467737685468, + "loss": 0.3232, + "num_input_tokens_seen": 67956576, + "step": 32535 + }, + { + "epoch": 5.3083448894689615, + "grad_norm": 0.00030785088893026114, + "learning_rate": 0.025024882488379557, + "loss": 0.2759, + "num_input_tokens_seen": 67966880, + "step": 32540 + }, + { + "epoch": 5.309160616689779, + "grad_norm": 0.0007814913988113403, + "learning_rate": 0.02499231651180727, + "loss": 0.2905, + "num_input_tokens_seen": 67977120, + "step": 32545 + }, + { + "epoch": 5.309976343910597, + "grad_norm": 0.0008360439678654075, + "learning_rate": 0.024959769812990713, + "loss": 0.3511, + "num_input_tokens_seen": 67987424, + "step": 32550 + }, + { + "epoch": 5.310792071131414, + "grad_norm": 0.0002954817609861493, + "learning_rate": 0.024927242396949045, + "loss": 0.3604, + "num_input_tokens_seen": 67997840, + "step": 32555 + }, + { + "epoch": 5.311607798352231, + "grad_norm": 0.0008417306817136705, + "learning_rate": 0.02489473426869836, + "loss": 0.3185, + "num_input_tokens_seen": 68008352, + "step": 32560 + }, + { + "epoch": 5.312423525573048, + "grad_norm": 0.0010501035721972585, + "learning_rate": 0.024862245433251776, + "loss": 0.3516, + "num_input_tokens_seen": 68018960, + "step": 32565 + }, + { + "epoch": 5.313239252793866, + "grad_norm": 0.000790701771620661, + "learning_rate": 0.024829775895619577, + "loss": 0.3013, + "num_input_tokens_seen": 68029760, + "step": 32570 + }, + { + "epoch": 5.314054980014683, + "grad_norm": 0.0005017669755034149, + "learning_rate": 0.024797325660808882, + "loss": 0.2661, + "num_input_tokens_seen": 68041344, + "step": 32575 + }, + { + "epoch": 5.3148707072355, + "grad_norm": 0.0005845595151185989, + "learning_rate": 0.02476489473382401, + "loss": 0.328, + "num_input_tokens_seen": 68052192, + "step": 32580 + }, + { + "epoch": 5.3156864344563175, + "grad_norm": 0.0010892378631979227, + "learning_rate": 0.024732483119666127, + "loss": 0.34, + "num_input_tokens_seen": 68063104, + "step": 32585 + }, + { + "epoch": 5.3165021616771355, + "grad_norm": 0.0005194342229515314, + "learning_rate": 0.024700090823333548, + "loss": 0.3317, + "num_input_tokens_seen": 68074704, + "step": 32590 + }, + { + "epoch": 5.317317888897953, + "grad_norm": 0.00035080802626907825, + "learning_rate": 0.02466771784982163, + "loss": 0.2635, + "num_input_tokens_seen": 68085408, + "step": 32595 + }, + { + "epoch": 5.31813361611877, + "grad_norm": 0.0005537371034733951, + "learning_rate": 0.024635364204122594, + "loss": 0.3581, + "num_input_tokens_seen": 68096656, + "step": 32600 + }, + { + "epoch": 5.31813361611877, + "eval_loss": 0.31552234292030334, + "eval_runtime": 156.1043, + "eval_samples_per_second": 17.456, + "eval_steps_per_second": 8.731, + "num_input_tokens_seen": 68096656, + "step": 32600 + }, + { + "epoch": 5.318949343339587, + "grad_norm": 0.00047554151387885213, + "learning_rate": 0.024603029891225852, + "loss": 0.3247, + "num_input_tokens_seen": 68107840, + "step": 32605 + }, + { + "epoch": 5.319765070560405, + "grad_norm": 0.0010257211979478598, + "learning_rate": 0.024570714916117748, + "loss": 0.2813, + "num_input_tokens_seen": 68119552, + "step": 32610 + }, + { + "epoch": 5.320580797781222, + "grad_norm": 0.000777583394665271, + "learning_rate": 0.024538419283781625, + "loss": 0.3687, + "num_input_tokens_seen": 68131072, + "step": 32615 + }, + { + "epoch": 5.321396525002039, + "grad_norm": 0.00044655491365119815, + "learning_rate": 0.024506142999197938, + "loss": 0.2831, + "num_input_tokens_seen": 68141904, + "step": 32620 + }, + { + "epoch": 5.322212252222856, + "grad_norm": 0.0003790041373576969, + "learning_rate": 0.024473886067344002, + "loss": 0.3115, + "num_input_tokens_seen": 68153152, + "step": 32625 + }, + { + "epoch": 5.323027979443674, + "grad_norm": 0.0006529239472001791, + "learning_rate": 0.02444164849319434, + "loss": 0.27, + "num_input_tokens_seen": 68162144, + "step": 32630 + }, + { + "epoch": 5.3238437066644915, + "grad_norm": 0.0006427898770198226, + "learning_rate": 0.024409430281720306, + "loss": 0.326, + "num_input_tokens_seen": 68173824, + "step": 32635 + }, + { + "epoch": 5.324659433885309, + "grad_norm": 0.0007967166020534933, + "learning_rate": 0.024377231437890428, + "loss": 0.2974, + "num_input_tokens_seen": 68185104, + "step": 32640 + }, + { + "epoch": 5.325475161106126, + "grad_norm": 0.0007620585383847356, + "learning_rate": 0.024345051966670115, + "loss": 0.3032, + "num_input_tokens_seen": 68195792, + "step": 32645 + }, + { + "epoch": 5.326290888326944, + "grad_norm": 0.0003330581239424646, + "learning_rate": 0.024312891873021884, + "loss": 0.3208, + "num_input_tokens_seen": 68206432, + "step": 32650 + }, + { + "epoch": 5.327106615547761, + "grad_norm": 0.00037632600287906826, + "learning_rate": 0.024280751161905183, + "loss": 0.3279, + "num_input_tokens_seen": 68216560, + "step": 32655 + }, + { + "epoch": 5.327922342768578, + "grad_norm": 0.0006988947279751301, + "learning_rate": 0.02424862983827658, + "loss": 0.2804, + "num_input_tokens_seen": 68226208, + "step": 32660 + }, + { + "epoch": 5.328738069989395, + "grad_norm": 0.0005600601434707642, + "learning_rate": 0.024216527907089495, + "loss": 0.3227, + "num_input_tokens_seen": 68237168, + "step": 32665 + }, + { + "epoch": 5.329553797210213, + "grad_norm": 0.00044355381396599114, + "learning_rate": 0.024184445373294505, + "loss": 0.3034, + "num_input_tokens_seen": 68246800, + "step": 32670 + }, + { + "epoch": 5.33036952443103, + "grad_norm": 0.0006094063865020871, + "learning_rate": 0.02415238224183918, + "loss": 0.311, + "num_input_tokens_seen": 68256720, + "step": 32675 + }, + { + "epoch": 5.331185251651847, + "grad_norm": 0.0004268065677024424, + "learning_rate": 0.024120338517667973, + "loss": 0.297, + "num_input_tokens_seen": 68268000, + "step": 32680 + }, + { + "epoch": 5.332000978872665, + "grad_norm": 0.0003517907753121108, + "learning_rate": 0.02408831420572247, + "loss": 0.3028, + "num_input_tokens_seen": 68278608, + "step": 32685 + }, + { + "epoch": 5.332816706093483, + "grad_norm": 0.001875571208074689, + "learning_rate": 0.024056309310941264, + "loss": 0.3389, + "num_input_tokens_seen": 68288064, + "step": 32690 + }, + { + "epoch": 5.3336324333143, + "grad_norm": 0.0006199294584803283, + "learning_rate": 0.02402432383825982, + "loss": 0.2932, + "num_input_tokens_seen": 68300560, + "step": 32695 + }, + { + "epoch": 5.334448160535117, + "grad_norm": 0.0007699491106905043, + "learning_rate": 0.023992357792610792, + "loss": 0.3157, + "num_input_tokens_seen": 68312400, + "step": 32700 + }, + { + "epoch": 5.335263887755934, + "grad_norm": 0.00039233770803548396, + "learning_rate": 0.0239604111789237, + "loss": 0.2757, + "num_input_tokens_seen": 68323984, + "step": 32705 + }, + { + "epoch": 5.336079614976752, + "grad_norm": 0.0005588364438153803, + "learning_rate": 0.023928484002125095, + "loss": 0.3371, + "num_input_tokens_seen": 68334384, + "step": 32710 + }, + { + "epoch": 5.336895342197569, + "grad_norm": 0.0004732238594442606, + "learning_rate": 0.023896576267138595, + "loss": 0.3141, + "num_input_tokens_seen": 68345152, + "step": 32715 + }, + { + "epoch": 5.337711069418386, + "grad_norm": 0.0008343725930899382, + "learning_rate": 0.02386468797888471, + "loss": 0.3526, + "num_input_tokens_seen": 68355280, + "step": 32720 + }, + { + "epoch": 5.338526796639204, + "grad_norm": 0.0006766126025468111, + "learning_rate": 0.023832819142281057, + "loss": 0.3097, + "num_input_tokens_seen": 68364992, + "step": 32725 + }, + { + "epoch": 5.339342523860021, + "grad_norm": 0.00046892149839550257, + "learning_rate": 0.02380096976224225, + "loss": 0.3294, + "num_input_tokens_seen": 68375824, + "step": 32730 + }, + { + "epoch": 5.3401582510808385, + "grad_norm": 0.00035650000791065395, + "learning_rate": 0.023769139843679777, + "loss": 0.2593, + "num_input_tokens_seen": 68384864, + "step": 32735 + }, + { + "epoch": 5.340973978301656, + "grad_norm": 0.0008853041799739003, + "learning_rate": 0.023737329391502287, + "loss": 0.3783, + "num_input_tokens_seen": 68395744, + "step": 32740 + }, + { + "epoch": 5.341789705522474, + "grad_norm": 0.0005938054528087378, + "learning_rate": 0.023705538410615293, + "loss": 0.2551, + "num_input_tokens_seen": 68405536, + "step": 32745 + }, + { + "epoch": 5.342605432743291, + "grad_norm": 0.0008316584280692041, + "learning_rate": 0.023673766905921396, + "loss": 0.3469, + "num_input_tokens_seen": 68415440, + "step": 32750 + }, + { + "epoch": 5.343421159964108, + "grad_norm": 0.0003751579497475177, + "learning_rate": 0.0236420148823202, + "loss": 0.3123, + "num_input_tokens_seen": 68426080, + "step": 32755 + }, + { + "epoch": 5.344236887184925, + "grad_norm": 0.0007653225329704583, + "learning_rate": 0.02361028234470816, + "loss": 0.3322, + "num_input_tokens_seen": 68437056, + "step": 32760 + }, + { + "epoch": 5.345052614405743, + "grad_norm": 0.0006158651667647064, + "learning_rate": 0.023578569297978913, + "loss": 0.3088, + "num_input_tokens_seen": 68447056, + "step": 32765 + }, + { + "epoch": 5.34586834162656, + "grad_norm": 0.00043597284820862114, + "learning_rate": 0.023546875747023025, + "loss": 0.2745, + "num_input_tokens_seen": 68457936, + "step": 32770 + }, + { + "epoch": 5.346684068847377, + "grad_norm": 0.0010256202658638358, + "learning_rate": 0.02351520169672801, + "loss": 0.3728, + "num_input_tokens_seen": 68468528, + "step": 32775 + }, + { + "epoch": 5.3474997960681945, + "grad_norm": 0.00035363037022762, + "learning_rate": 0.023483547151978357, + "loss": 0.2715, + "num_input_tokens_seen": 68480448, + "step": 32780 + }, + { + "epoch": 5.3483155232890125, + "grad_norm": 0.0004602066474035382, + "learning_rate": 0.023451912117655675, + "loss": 0.2957, + "num_input_tokens_seen": 68492144, + "step": 32785 + }, + { + "epoch": 5.34913125050983, + "grad_norm": 0.00038286010385490954, + "learning_rate": 0.023420296598638417, + "loss": 0.2774, + "num_input_tokens_seen": 68501632, + "step": 32790 + }, + { + "epoch": 5.349946977730647, + "grad_norm": 0.0005481119733303785, + "learning_rate": 0.023388700599802165, + "loss": 0.2746, + "num_input_tokens_seen": 68512016, + "step": 32795 + }, + { + "epoch": 5.350762704951464, + "grad_norm": 0.0007237319950945675, + "learning_rate": 0.023357124126019334, + "loss": 0.3607, + "num_input_tokens_seen": 68521600, + "step": 32800 + }, + { + "epoch": 5.350762704951464, + "eval_loss": 0.31374478340148926, + "eval_runtime": 155.6972, + "eval_samples_per_second": 17.502, + "eval_steps_per_second": 8.754, + "num_input_tokens_seen": 68521600, + "step": 32800 + }, + { + "epoch": 5.351578432172282, + "grad_norm": 0.0005472971824929118, + "learning_rate": 0.02332556718215945, + "loss": 0.3249, + "num_input_tokens_seen": 68532432, + "step": 32805 + }, + { + "epoch": 5.352394159393099, + "grad_norm": 0.0004706378385890275, + "learning_rate": 0.023294029773089035, + "loss": 0.3683, + "num_input_tokens_seen": 68542336, + "step": 32810 + }, + { + "epoch": 5.353209886613916, + "grad_norm": 0.0005092447972856462, + "learning_rate": 0.023262511903671484, + "loss": 0.3366, + "num_input_tokens_seen": 68553520, + "step": 32815 + }, + { + "epoch": 5.354025613834733, + "grad_norm": 0.0006696992204524577, + "learning_rate": 0.023231013578767324, + "loss": 0.2932, + "num_input_tokens_seen": 68564048, + "step": 32820 + }, + { + "epoch": 5.354841341055551, + "grad_norm": 0.00041500551742501557, + "learning_rate": 0.0231995348032339, + "loss": 0.3493, + "num_input_tokens_seen": 68574256, + "step": 32825 + }, + { + "epoch": 5.3556570682763684, + "grad_norm": 0.0012570745311677456, + "learning_rate": 0.023168075581925685, + "loss": 0.3542, + "num_input_tokens_seen": 68585120, + "step": 32830 + }, + { + "epoch": 5.356472795497186, + "grad_norm": 0.00046153791481629014, + "learning_rate": 0.023136635919694126, + "loss": 0.2744, + "num_input_tokens_seen": 68594976, + "step": 32835 + }, + { + "epoch": 5.357288522718003, + "grad_norm": 0.0006924188928678632, + "learning_rate": 0.02310521582138753, + "loss": 0.3851, + "num_input_tokens_seen": 68605920, + "step": 32840 + }, + { + "epoch": 5.358104249938821, + "grad_norm": 0.0004283167945686728, + "learning_rate": 0.023073815291851357, + "loss": 0.33, + "num_input_tokens_seen": 68615888, + "step": 32845 + }, + { + "epoch": 5.358919977159638, + "grad_norm": 0.0007701045833528042, + "learning_rate": 0.02304243433592788, + "loss": 0.3351, + "num_input_tokens_seen": 68626112, + "step": 32850 + }, + { + "epoch": 5.359735704380455, + "grad_norm": 0.0005286188097670674, + "learning_rate": 0.023011072958456513, + "loss": 0.3502, + "num_input_tokens_seen": 68636736, + "step": 32855 + }, + { + "epoch": 5.360551431601272, + "grad_norm": 0.00044128333684057, + "learning_rate": 0.022979731164273536, + "loss": 0.3072, + "num_input_tokens_seen": 68647936, + "step": 32860 + }, + { + "epoch": 5.36136715882209, + "grad_norm": 0.0005425781710073352, + "learning_rate": 0.022948408958212218, + "loss": 0.3618, + "num_input_tokens_seen": 68658336, + "step": 32865 + }, + { + "epoch": 5.362182886042907, + "grad_norm": 0.0005162459565326571, + "learning_rate": 0.022917106345102876, + "loss": 0.2865, + "num_input_tokens_seen": 68669232, + "step": 32870 + }, + { + "epoch": 5.362998613263724, + "grad_norm": 0.0004011390556115657, + "learning_rate": 0.022885823329772785, + "loss": 0.3373, + "num_input_tokens_seen": 68681264, + "step": 32875 + }, + { + "epoch": 5.3638143404845415, + "grad_norm": 0.0005178343271836638, + "learning_rate": 0.02285455991704612, + "loss": 0.3005, + "num_input_tokens_seen": 68690752, + "step": 32880 + }, + { + "epoch": 5.3646300677053596, + "grad_norm": 0.000647915352601558, + "learning_rate": 0.022823316111744117, + "loss": 0.3149, + "num_input_tokens_seen": 68701584, + "step": 32885 + }, + { + "epoch": 5.365445794926177, + "grad_norm": 0.0004697332333307713, + "learning_rate": 0.022792091918685014, + "loss": 0.3216, + "num_input_tokens_seen": 68711952, + "step": 32890 + }, + { + "epoch": 5.366261522146994, + "grad_norm": 0.00047058984637260437, + "learning_rate": 0.022760887342683906, + "loss": 0.306, + "num_input_tokens_seen": 68721808, + "step": 32895 + }, + { + "epoch": 5.367077249367812, + "grad_norm": 0.0006202487275004387, + "learning_rate": 0.022729702388552975, + "loss": 0.2683, + "num_input_tokens_seen": 68732656, + "step": 32900 + }, + { + "epoch": 5.367892976588629, + "grad_norm": 0.0004808719386346638, + "learning_rate": 0.022698537061101292, + "loss": 0.3001, + "num_input_tokens_seen": 68744704, + "step": 32905 + }, + { + "epoch": 5.368708703809446, + "grad_norm": 0.0005038009257987142, + "learning_rate": 0.022667391365134962, + "loss": 0.2808, + "num_input_tokens_seen": 68755264, + "step": 32910 + }, + { + "epoch": 5.369524431030263, + "grad_norm": 0.0006500156596302986, + "learning_rate": 0.022636265305457065, + "loss": 0.3516, + "num_input_tokens_seen": 68765280, + "step": 32915 + }, + { + "epoch": 5.370340158251081, + "grad_norm": 0.0005245216307230294, + "learning_rate": 0.02260515888686764, + "loss": 0.3606, + "num_input_tokens_seen": 68775232, + "step": 32920 + }, + { + "epoch": 5.371155885471898, + "grad_norm": 0.0008450545719824731, + "learning_rate": 0.022574072114163596, + "loss": 0.3126, + "num_input_tokens_seen": 68785760, + "step": 32925 + }, + { + "epoch": 5.3719716126927155, + "grad_norm": 0.0008642810862511396, + "learning_rate": 0.022543004992139005, + "loss": 0.3155, + "num_input_tokens_seen": 68795424, + "step": 32930 + }, + { + "epoch": 5.372787339913533, + "grad_norm": 0.0006237569614313543, + "learning_rate": 0.022511957525584745, + "loss": 0.3106, + "num_input_tokens_seen": 68805504, + "step": 32935 + }, + { + "epoch": 5.373603067134351, + "grad_norm": 0.0004782045725733042, + "learning_rate": 0.022480929719288778, + "loss": 0.3034, + "num_input_tokens_seen": 68816416, + "step": 32940 + }, + { + "epoch": 5.374418794355168, + "grad_norm": 0.0003936954599339515, + "learning_rate": 0.02244992157803592, + "loss": 0.2847, + "num_input_tokens_seen": 68825600, + "step": 32945 + }, + { + "epoch": 5.375234521575985, + "grad_norm": 0.0008605166804045439, + "learning_rate": 0.022418933106608047, + "loss": 0.2776, + "num_input_tokens_seen": 68836592, + "step": 32950 + }, + { + "epoch": 5.376050248796802, + "grad_norm": 0.000995707348920405, + "learning_rate": 0.022387964309784018, + "loss": 0.3359, + "num_input_tokens_seen": 68846784, + "step": 32955 + }, + { + "epoch": 5.37686597601762, + "grad_norm": 0.0004613129422068596, + "learning_rate": 0.022357015192339517, + "loss": 0.3984, + "num_input_tokens_seen": 68859424, + "step": 32960 + }, + { + "epoch": 5.377681703238437, + "grad_norm": 0.0005081254639662802, + "learning_rate": 0.02232608575904734, + "loss": 0.3281, + "num_input_tokens_seen": 68869312, + "step": 32965 + }, + { + "epoch": 5.378497430459254, + "grad_norm": 0.0007430273108184338, + "learning_rate": 0.022295176014677225, + "loss": 0.3413, + "num_input_tokens_seen": 68880224, + "step": 32970 + }, + { + "epoch": 5.3793131576800715, + "grad_norm": 0.0004840785404667258, + "learning_rate": 0.02226428596399577, + "loss": 0.3506, + "num_input_tokens_seen": 68892112, + "step": 32975 + }, + { + "epoch": 5.3801288849008895, + "grad_norm": 0.0007137857028283179, + "learning_rate": 0.02223341561176669, + "loss": 0.3032, + "num_input_tokens_seen": 68904640, + "step": 32980 + }, + { + "epoch": 5.380944612121707, + "grad_norm": 0.000709670246578753, + "learning_rate": 0.0222025649627505, + "loss": 0.2962, + "num_input_tokens_seen": 68915632, + "step": 32985 + }, + { + "epoch": 5.381760339342524, + "grad_norm": 0.0005600260337814689, + "learning_rate": 0.022171734021704814, + "loss": 0.2616, + "num_input_tokens_seen": 68926288, + "step": 32990 + }, + { + "epoch": 5.382576066563341, + "grad_norm": 0.006863137241452932, + "learning_rate": 0.022140922793384116, + "loss": 0.344, + "num_input_tokens_seen": 68937328, + "step": 32995 + }, + { + "epoch": 5.383391793784159, + "grad_norm": 0.0005161588778719306, + "learning_rate": 0.022110131282539934, + "loss": 0.2925, + "num_input_tokens_seen": 68948064, + "step": 33000 + }, + { + "epoch": 5.383391793784159, + "eval_loss": 0.3153829872608185, + "eval_runtime": 156.0961, + "eval_samples_per_second": 17.457, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 68948064, + "step": 33000 + }, + { + "epoch": 5.384207521004976, + "grad_norm": 0.0005816013435833156, + "learning_rate": 0.022079359493920675, + "loss": 0.3203, + "num_input_tokens_seen": 68958288, + "step": 33005 + }, + { + "epoch": 5.385023248225793, + "grad_norm": 0.0007093854364939034, + "learning_rate": 0.02204860743227169, + "loss": 0.319, + "num_input_tokens_seen": 68968576, + "step": 33010 + }, + { + "epoch": 5.38583897544661, + "grad_norm": 0.0005660208407789469, + "learning_rate": 0.022017875102335365, + "loss": 0.3232, + "num_input_tokens_seen": 68978176, + "step": 33015 + }, + { + "epoch": 5.386654702667428, + "grad_norm": 0.0005107233882881701, + "learning_rate": 0.02198716250885108, + "loss": 0.2892, + "num_input_tokens_seen": 68988560, + "step": 33020 + }, + { + "epoch": 5.387470429888245, + "grad_norm": 0.0005749469855800271, + "learning_rate": 0.021956469656555, + "loss": 0.2998, + "num_input_tokens_seen": 68998704, + "step": 33025 + }, + { + "epoch": 5.388286157109063, + "grad_norm": 0.00047812264529056847, + "learning_rate": 0.0219257965501804, + "loss": 0.3012, + "num_input_tokens_seen": 69009760, + "step": 33030 + }, + { + "epoch": 5.38910188432988, + "grad_norm": 0.00045351358130574226, + "learning_rate": 0.021895143194457494, + "loss": 0.368, + "num_input_tokens_seen": 69020944, + "step": 33035 + }, + { + "epoch": 5.389917611550698, + "grad_norm": 0.00045235283323563635, + "learning_rate": 0.021864509594113322, + "loss": 0.3092, + "num_input_tokens_seen": 69031296, + "step": 33040 + }, + { + "epoch": 5.390733338771515, + "grad_norm": 0.0007505902322009206, + "learning_rate": 0.02183389575387207, + "loss": 0.3199, + "num_input_tokens_seen": 69042720, + "step": 33045 + }, + { + "epoch": 5.391549065992332, + "grad_norm": 0.0005359457572922111, + "learning_rate": 0.021803301678454682, + "loss": 0.3061, + "num_input_tokens_seen": 69052672, + "step": 33050 + }, + { + "epoch": 5.392364793213149, + "grad_norm": 0.0004310861404519528, + "learning_rate": 0.021772727372579213, + "loss": 0.3601, + "num_input_tokens_seen": 69063568, + "step": 33055 + }, + { + "epoch": 5.393180520433967, + "grad_norm": 0.0005672852857969701, + "learning_rate": 0.02174217284096061, + "loss": 0.2638, + "num_input_tokens_seen": 69073744, + "step": 33060 + }, + { + "epoch": 5.393996247654784, + "grad_norm": 0.000581924628932029, + "learning_rate": 0.0217116380883107, + "loss": 0.3222, + "num_input_tokens_seen": 69082464, + "step": 33065 + }, + { + "epoch": 5.394811974875601, + "grad_norm": 0.000861417327541858, + "learning_rate": 0.021681123119338425, + "loss": 0.3027, + "num_input_tokens_seen": 69093296, + "step": 33070 + }, + { + "epoch": 5.395627702096419, + "grad_norm": 0.0005722224595956504, + "learning_rate": 0.02165062793874951, + "loss": 0.2842, + "num_input_tokens_seen": 69104224, + "step": 33075 + }, + { + "epoch": 5.3964434293172365, + "grad_norm": 0.0006943875923752785, + "learning_rate": 0.021620152551246666, + "loss": 0.3127, + "num_input_tokens_seen": 69113376, + "step": 33080 + }, + { + "epoch": 5.397259156538054, + "grad_norm": 0.0006763713899999857, + "learning_rate": 0.02158969696152967, + "loss": 0.2896, + "num_input_tokens_seen": 69124336, + "step": 33085 + }, + { + "epoch": 5.398074883758871, + "grad_norm": 0.0003438839630689472, + "learning_rate": 0.021559261174295057, + "loss": 0.2658, + "num_input_tokens_seen": 69135296, + "step": 33090 + }, + { + "epoch": 5.398890610979688, + "grad_norm": 0.000910153379663825, + "learning_rate": 0.02152884519423646, + "loss": 0.2937, + "num_input_tokens_seen": 69146496, + "step": 33095 + }, + { + "epoch": 5.399706338200506, + "grad_norm": 0.0007924582459963858, + "learning_rate": 0.021498449026044447, + "loss": 0.3471, + "num_input_tokens_seen": 69155952, + "step": 33100 + }, + { + "epoch": 5.400522065421323, + "grad_norm": 0.0014556945534422994, + "learning_rate": 0.021468072674406414, + "loss": 0.4127, + "num_input_tokens_seen": 69165904, + "step": 33105 + }, + { + "epoch": 5.40133779264214, + "grad_norm": 0.0005345948156900704, + "learning_rate": 0.021437716144006795, + "loss": 0.3657, + "num_input_tokens_seen": 69176560, + "step": 33110 + }, + { + "epoch": 5.402153519862958, + "grad_norm": 0.0004974445910193026, + "learning_rate": 0.021407379439527002, + "loss": 0.2992, + "num_input_tokens_seen": 69186592, + "step": 33115 + }, + { + "epoch": 5.402969247083775, + "grad_norm": 0.0005334231536835432, + "learning_rate": 0.021377062565645255, + "loss": 0.3039, + "num_input_tokens_seen": 69195792, + "step": 33120 + }, + { + "epoch": 5.4037849743045925, + "grad_norm": 0.0005341392243281007, + "learning_rate": 0.02134676552703688, + "loss": 0.3006, + "num_input_tokens_seen": 69205552, + "step": 33125 + }, + { + "epoch": 5.40460070152541, + "grad_norm": 0.00045310636050999165, + "learning_rate": 0.02131648832837398, + "loss": 0.2891, + "num_input_tokens_seen": 69216192, + "step": 33130 + }, + { + "epoch": 5.405416428746228, + "grad_norm": 0.0003765394212678075, + "learning_rate": 0.02128623097432574, + "loss": 0.3283, + "num_input_tokens_seen": 69226304, + "step": 33135 + }, + { + "epoch": 5.406232155967045, + "grad_norm": 0.0004346759233158082, + "learning_rate": 0.021255993469558192, + "loss": 0.289, + "num_input_tokens_seen": 69236736, + "step": 33140 + }, + { + "epoch": 5.407047883187862, + "grad_norm": 0.00043981653288938105, + "learning_rate": 0.021225775818734364, + "loss": 0.2994, + "num_input_tokens_seen": 69247776, + "step": 33145 + }, + { + "epoch": 5.407863610408679, + "grad_norm": 0.0006464622565545142, + "learning_rate": 0.021195578026514166, + "loss": 0.3242, + "num_input_tokens_seen": 69255792, + "step": 33150 + }, + { + "epoch": 5.408679337629497, + "grad_norm": 0.0005539258709177375, + "learning_rate": 0.02116540009755452, + "loss": 0.2684, + "num_input_tokens_seen": 69265888, + "step": 33155 + }, + { + "epoch": 5.409495064850314, + "grad_norm": 0.0004554849583655596, + "learning_rate": 0.021135242036509173, + "loss": 0.3529, + "num_input_tokens_seen": 69275840, + "step": 33160 + }, + { + "epoch": 5.410310792071131, + "grad_norm": 0.001068034442141652, + "learning_rate": 0.021105103848028967, + "loss": 0.3519, + "num_input_tokens_seen": 69285936, + "step": 33165 + }, + { + "epoch": 5.411126519291948, + "grad_norm": 0.0005641066818498075, + "learning_rate": 0.021074985536761504, + "loss": 0.3289, + "num_input_tokens_seen": 69296416, + "step": 33170 + }, + { + "epoch": 5.4119422465127665, + "grad_norm": 0.0005025541177019477, + "learning_rate": 0.021044887107351435, + "loss": 0.3101, + "num_input_tokens_seen": 69305872, + "step": 33175 + }, + { + "epoch": 5.412757973733584, + "grad_norm": 0.00033758257632143795, + "learning_rate": 0.021014808564440362, + "loss": 0.2771, + "num_input_tokens_seen": 69316816, + "step": 33180 + }, + { + "epoch": 5.413573700954401, + "grad_norm": 0.00042940949788317084, + "learning_rate": 0.02098474991266671, + "loss": 0.2502, + "num_input_tokens_seen": 69326928, + "step": 33185 + }, + { + "epoch": 5.414389428175218, + "grad_norm": 0.0005671408725902438, + "learning_rate": 0.02095471115666592, + "loss": 0.3355, + "num_input_tokens_seen": 69338448, + "step": 33190 + }, + { + "epoch": 5.415205155396036, + "grad_norm": 0.000760889844968915, + "learning_rate": 0.020924692301070406, + "loss": 0.3293, + "num_input_tokens_seen": 69347312, + "step": 33195 + }, + { + "epoch": 5.416020882616853, + "grad_norm": 0.0006250061560422182, + "learning_rate": 0.020894693350509346, + "loss": 0.2777, + "num_input_tokens_seen": 69357008, + "step": 33200 + }, + { + "epoch": 5.416020882616853, + "eval_loss": 0.3144669234752655, + "eval_runtime": 155.7346, + "eval_samples_per_second": 17.498, + "eval_steps_per_second": 8.752, + "num_input_tokens_seen": 69357008, + "step": 33200 + }, + { + "epoch": 5.41683660983767, + "grad_norm": 0.0008062694687396288, + "learning_rate": 0.020864714309609057, + "loss": 0.3672, + "num_input_tokens_seen": 69367024, + "step": 33205 + }, + { + "epoch": 5.417652337058487, + "grad_norm": 0.0006789707113057375, + "learning_rate": 0.020834755182992604, + "loss": 0.3301, + "num_input_tokens_seen": 69376848, + "step": 33210 + }, + { + "epoch": 5.418468064279305, + "grad_norm": 0.000649005058221519, + "learning_rate": 0.02080481597528011, + "loss": 0.2542, + "num_input_tokens_seen": 69386992, + "step": 33215 + }, + { + "epoch": 5.419283791500122, + "grad_norm": 0.0007456055609509349, + "learning_rate": 0.020774896691088583, + "loss": 0.3593, + "num_input_tokens_seen": 69395536, + "step": 33220 + }, + { + "epoch": 5.4200995187209395, + "grad_norm": 0.0006427406915463507, + "learning_rate": 0.020744997335031882, + "loss": 0.3626, + "num_input_tokens_seen": 69406144, + "step": 33225 + }, + { + "epoch": 5.420915245941757, + "grad_norm": 0.0011627055937424302, + "learning_rate": 0.02071511791172092, + "loss": 0.2857, + "num_input_tokens_seen": 69417696, + "step": 33230 + }, + { + "epoch": 5.421730973162575, + "grad_norm": 0.0006728776497766376, + "learning_rate": 0.02068525842576351, + "loss": 0.3079, + "num_input_tokens_seen": 69427536, + "step": 33235 + }, + { + "epoch": 5.422546700383392, + "grad_norm": 0.0007935905596241355, + "learning_rate": 0.020655418881764264, + "loss": 0.3056, + "num_input_tokens_seen": 69437968, + "step": 33240 + }, + { + "epoch": 5.423362427604209, + "grad_norm": 0.0005553739611059427, + "learning_rate": 0.020625599284324923, + "loss": 0.3534, + "num_input_tokens_seen": 69447936, + "step": 33245 + }, + { + "epoch": 5.424178154825027, + "grad_norm": 0.0005036382935941219, + "learning_rate": 0.02059579963804396, + "loss": 0.3309, + "num_input_tokens_seen": 69458528, + "step": 33250 + }, + { + "epoch": 5.424993882045844, + "grad_norm": 0.0005157593986950815, + "learning_rate": 0.02056601994751688, + "loss": 0.3464, + "num_input_tokens_seen": 69470064, + "step": 33255 + }, + { + "epoch": 5.425809609266661, + "grad_norm": 0.0007480886997655034, + "learning_rate": 0.02053626021733614, + "loss": 0.2802, + "num_input_tokens_seen": 69480944, + "step": 33260 + }, + { + "epoch": 5.426625336487478, + "grad_norm": 0.0003803317085839808, + "learning_rate": 0.02050652045209097, + "loss": 0.3267, + "num_input_tokens_seen": 69491008, + "step": 33265 + }, + { + "epoch": 5.4274410637082955, + "grad_norm": 0.0003815782256424427, + "learning_rate": 0.020476800656367672, + "loss": 0.3096, + "num_input_tokens_seen": 69501648, + "step": 33270 + }, + { + "epoch": 5.4282567909291135, + "grad_norm": 0.00043827146873809397, + "learning_rate": 0.020447100834749425, + "loss": 0.2931, + "num_input_tokens_seen": 69512304, + "step": 33275 + }, + { + "epoch": 5.429072518149931, + "grad_norm": 0.0007266983157023787, + "learning_rate": 0.02041742099181627, + "loss": 0.3032, + "num_input_tokens_seen": 69522528, + "step": 33280 + }, + { + "epoch": 5.429888245370748, + "grad_norm": 0.0004883180372416973, + "learning_rate": 0.02038776113214526, + "loss": 0.3026, + "num_input_tokens_seen": 69532704, + "step": 33285 + }, + { + "epoch": 5.430703972591566, + "grad_norm": 0.00040917532169260085, + "learning_rate": 0.0203581212603103, + "loss": 0.2516, + "num_input_tokens_seen": 69542064, + "step": 33290 + }, + { + "epoch": 5.431519699812383, + "grad_norm": 0.0006701766396872699, + "learning_rate": 0.02032850138088219, + "loss": 0.2555, + "num_input_tokens_seen": 69552368, + "step": 33295 + }, + { + "epoch": 5.4323354270332, + "grad_norm": 0.0006435977411456406, + "learning_rate": 0.020298901498428754, + "loss": 0.2892, + "num_input_tokens_seen": 69562224, + "step": 33300 + }, + { + "epoch": 5.433151154254017, + "grad_norm": 0.00047868656110949814, + "learning_rate": 0.020269321617514595, + "loss": 0.3205, + "num_input_tokens_seen": 69572112, + "step": 33305 + }, + { + "epoch": 5.433966881474835, + "grad_norm": 0.0006549336249008775, + "learning_rate": 0.020239761742701343, + "loss": 0.3445, + "num_input_tokens_seen": 69584000, + "step": 33310 + }, + { + "epoch": 5.434782608695652, + "grad_norm": 0.0004597057995852083, + "learning_rate": 0.02021022187854754, + "loss": 0.2889, + "num_input_tokens_seen": 69594048, + "step": 33315 + }, + { + "epoch": 5.4355983359164695, + "grad_norm": 0.0006162559147924185, + "learning_rate": 0.020180702029608522, + "loss": 0.3436, + "num_input_tokens_seen": 69604560, + "step": 33320 + }, + { + "epoch": 5.436414063137287, + "grad_norm": 0.00045969142229296267, + "learning_rate": 0.020151202200436695, + "loss": 0.2802, + "num_input_tokens_seen": 69614336, + "step": 33325 + }, + { + "epoch": 5.437229790358105, + "grad_norm": 0.0004913303419016302, + "learning_rate": 0.020121722395581226, + "loss": 0.2772, + "num_input_tokens_seen": 69625216, + "step": 33330 + }, + { + "epoch": 5.438045517578922, + "grad_norm": 0.0009047970524989069, + "learning_rate": 0.020092262619588342, + "loss": 0.2774, + "num_input_tokens_seen": 69635632, + "step": 33335 + }, + { + "epoch": 5.438861244799739, + "grad_norm": 0.00047880742931738496, + "learning_rate": 0.02006282287700109, + "loss": 0.2929, + "num_input_tokens_seen": 69645936, + "step": 33340 + }, + { + "epoch": 5.439676972020556, + "grad_norm": 0.0003834362723864615, + "learning_rate": 0.020033403172359427, + "loss": 0.3885, + "num_input_tokens_seen": 69656928, + "step": 33345 + }, + { + "epoch": 5.440492699241374, + "grad_norm": 0.0004321725864429027, + "learning_rate": 0.020004003510200284, + "loss": 0.3458, + "num_input_tokens_seen": 69668128, + "step": 33350 + }, + { + "epoch": 5.441308426462191, + "grad_norm": 0.0004767276404891163, + "learning_rate": 0.019974623895057407, + "loss": 0.3164, + "num_input_tokens_seen": 69678864, + "step": 33355 + }, + { + "epoch": 5.442124153683008, + "grad_norm": 0.0007869285182096064, + "learning_rate": 0.019945264331461553, + "loss": 0.323, + "num_input_tokens_seen": 69689216, + "step": 33360 + }, + { + "epoch": 5.442939880903825, + "grad_norm": 0.0007457199390046299, + "learning_rate": 0.019915924823940317, + "loss": 0.3132, + "num_input_tokens_seen": 69699936, + "step": 33365 + }, + { + "epoch": 5.443755608124643, + "grad_norm": 0.0006524266209453344, + "learning_rate": 0.01988660537701816, + "loss": 0.3193, + "num_input_tokens_seen": 69710640, + "step": 33370 + }, + { + "epoch": 5.444571335345461, + "grad_norm": 0.0007563949911855161, + "learning_rate": 0.01985730599521659, + "loss": 0.3585, + "num_input_tokens_seen": 69721824, + "step": 33375 + }, + { + "epoch": 5.445387062566278, + "grad_norm": 0.0007152993930503726, + "learning_rate": 0.019828026683053918, + "loss": 0.3194, + "num_input_tokens_seen": 69730608, + "step": 33380 + }, + { + "epoch": 5.446202789787095, + "grad_norm": 0.00031635945197194815, + "learning_rate": 0.01979876744504535, + "loss": 0.3395, + "num_input_tokens_seen": 69742128, + "step": 33385 + }, + { + "epoch": 5.447018517007913, + "grad_norm": 0.0005179584259167314, + "learning_rate": 0.019769528285703046, + "loss": 0.3836, + "num_input_tokens_seen": 69751952, + "step": 33390 + }, + { + "epoch": 5.44783424422873, + "grad_norm": 0.0004995448398403823, + "learning_rate": 0.019740309209536098, + "loss": 0.2609, + "num_input_tokens_seen": 69762464, + "step": 33395 + }, + { + "epoch": 5.448649971449547, + "grad_norm": 0.0005342772346921265, + "learning_rate": 0.019711110221050387, + "loss": 0.2747, + "num_input_tokens_seen": 69771824, + "step": 33400 + }, + { + "epoch": 5.448649971449547, + "eval_loss": 0.3145916759967804, + "eval_runtime": 155.9899, + "eval_samples_per_second": 17.469, + "eval_steps_per_second": 8.738, + "num_input_tokens_seen": 69771824, + "step": 33400 + }, + { + "epoch": 5.449465698670364, + "grad_norm": 0.0005774666205979884, + "learning_rate": 0.019681931324748825, + "loss": 0.371, + "num_input_tokens_seen": 69781600, + "step": 33405 + }, + { + "epoch": 5.450281425891182, + "grad_norm": 0.000534180726390332, + "learning_rate": 0.019652772525131094, + "loss": 0.2917, + "num_input_tokens_seen": 69791936, + "step": 33410 + }, + { + "epoch": 5.451097153111999, + "grad_norm": 0.0007314815884456038, + "learning_rate": 0.019623633826693885, + "loss": 0.3098, + "num_input_tokens_seen": 69801552, + "step": 33415 + }, + { + "epoch": 5.4519128803328165, + "grad_norm": 0.0004109461442567408, + "learning_rate": 0.019594515233930788, + "loss": 0.3196, + "num_input_tokens_seen": 69812304, + "step": 33420 + }, + { + "epoch": 5.4527286075536345, + "grad_norm": 0.0003878641873598099, + "learning_rate": 0.019565416751332186, + "loss": 0.342, + "num_input_tokens_seen": 69822688, + "step": 33425 + }, + { + "epoch": 5.453544334774452, + "grad_norm": 0.0005588119965977967, + "learning_rate": 0.019536338383385497, + "loss": 0.4093, + "num_input_tokens_seen": 69833120, + "step": 33430 + }, + { + "epoch": 5.454360061995269, + "grad_norm": 0.0009519964805804193, + "learning_rate": 0.019507280134574933, + "loss": 0.2858, + "num_input_tokens_seen": 69844432, + "step": 33435 + }, + { + "epoch": 5.455175789216086, + "grad_norm": 0.0007088609854690731, + "learning_rate": 0.019478242009381624, + "loss": 0.3393, + "num_input_tokens_seen": 69854976, + "step": 33440 + }, + { + "epoch": 5.455991516436903, + "grad_norm": 0.000926793145481497, + "learning_rate": 0.01944922401228367, + "loss": 0.322, + "num_input_tokens_seen": 69864944, + "step": 33445 + }, + { + "epoch": 5.456807243657721, + "grad_norm": 0.00042998912977054715, + "learning_rate": 0.01942022614775593, + "loss": 0.3126, + "num_input_tokens_seen": 69874608, + "step": 33450 + }, + { + "epoch": 5.457622970878538, + "grad_norm": 0.0005299998447299004, + "learning_rate": 0.01939124842027029, + "loss": 0.334, + "num_input_tokens_seen": 69884512, + "step": 33455 + }, + { + "epoch": 5.458438698099355, + "grad_norm": 0.0006390133639797568, + "learning_rate": 0.01936229083429551, + "loss": 0.2781, + "num_input_tokens_seen": 69894128, + "step": 33460 + }, + { + "epoch": 5.459254425320173, + "grad_norm": 0.0007583443657495081, + "learning_rate": 0.019333353394297148, + "loss": 0.2911, + "num_input_tokens_seen": 69903952, + "step": 33465 + }, + { + "epoch": 5.4600701525409905, + "grad_norm": 0.0007408655364997685, + "learning_rate": 0.019304436104737754, + "loss": 0.2892, + "num_input_tokens_seen": 69915072, + "step": 33470 + }, + { + "epoch": 5.460885879761808, + "grad_norm": 0.0003469556977506727, + "learning_rate": 0.019275538970076778, + "loss": 0.2534, + "num_input_tokens_seen": 69926032, + "step": 33475 + }, + { + "epoch": 5.461701606982625, + "grad_norm": 0.00043499961611814797, + "learning_rate": 0.019246661994770434, + "loss": 0.3141, + "num_input_tokens_seen": 69936800, + "step": 33480 + }, + { + "epoch": 5.462517334203443, + "grad_norm": 0.0011932312045246363, + "learning_rate": 0.019217805183271985, + "loss": 0.3073, + "num_input_tokens_seen": 69948208, + "step": 33485 + }, + { + "epoch": 5.46333306142426, + "grad_norm": 0.00045081102871336043, + "learning_rate": 0.019188968540031465, + "loss": 0.2652, + "num_input_tokens_seen": 69959552, + "step": 33490 + }, + { + "epoch": 5.464148788645077, + "grad_norm": 0.0006896810373291373, + "learning_rate": 0.019160152069495867, + "loss": 0.3488, + "num_input_tokens_seen": 69968208, + "step": 33495 + }, + { + "epoch": 5.464964515865894, + "grad_norm": 0.0005065254517830908, + "learning_rate": 0.019131355776109103, + "loss": 0.2647, + "num_input_tokens_seen": 69978928, + "step": 33500 + }, + { + "epoch": 5.465780243086712, + "grad_norm": 0.00046836771070957184, + "learning_rate": 0.019102579664311857, + "loss": 0.3878, + "num_input_tokens_seen": 69988816, + "step": 33505 + }, + { + "epoch": 5.466595970307529, + "grad_norm": 0.0005854279734194279, + "learning_rate": 0.019073823738541763, + "loss": 0.3101, + "num_input_tokens_seen": 69998800, + "step": 33510 + }, + { + "epoch": 5.467411697528346, + "grad_norm": 0.000485365220811218, + "learning_rate": 0.0190450880032334, + "loss": 0.3774, + "num_input_tokens_seen": 70008688, + "step": 33515 + }, + { + "epoch": 5.468227424749164, + "grad_norm": 0.0010944157838821411, + "learning_rate": 0.019016372462818114, + "loss": 0.3387, + "num_input_tokens_seen": 70019232, + "step": 33520 + }, + { + "epoch": 5.469043151969982, + "grad_norm": 0.0004198939132038504, + "learning_rate": 0.018987677121724278, + "loss": 0.3697, + "num_input_tokens_seen": 70030400, + "step": 33525 + }, + { + "epoch": 5.469858879190799, + "grad_norm": 0.0006542615010403097, + "learning_rate": 0.018959001984377, + "loss": 0.285, + "num_input_tokens_seen": 70041184, + "step": 33530 + }, + { + "epoch": 5.470674606411616, + "grad_norm": 0.0003455092082731426, + "learning_rate": 0.018930347055198377, + "loss": 0.3766, + "num_input_tokens_seen": 70052048, + "step": 33535 + }, + { + "epoch": 5.471490333632433, + "grad_norm": 0.0009140128968283534, + "learning_rate": 0.01890171233860739, + "loss": 0.2723, + "num_input_tokens_seen": 70062160, + "step": 33540 + }, + { + "epoch": 5.472306060853251, + "grad_norm": 0.0007531836745329201, + "learning_rate": 0.018873097839019807, + "loss": 0.3041, + "num_input_tokens_seen": 70071664, + "step": 33545 + }, + { + "epoch": 5.473121788074068, + "grad_norm": 0.0004007796524092555, + "learning_rate": 0.0188445035608484, + "loss": 0.2908, + "num_input_tokens_seen": 70081184, + "step": 33550 + }, + { + "epoch": 5.473937515294885, + "grad_norm": 0.00048610876547172666, + "learning_rate": 0.018815929508502777, + "loss": 0.2807, + "num_input_tokens_seen": 70092528, + "step": 33555 + }, + { + "epoch": 5.474753242515702, + "grad_norm": 0.0009474455728195608, + "learning_rate": 0.01878737568638934, + "loss": 0.288, + "num_input_tokens_seen": 70104464, + "step": 33560 + }, + { + "epoch": 5.47556896973652, + "grad_norm": 0.000713318819180131, + "learning_rate": 0.01875884209891152, + "loss": 0.3434, + "num_input_tokens_seen": 70115648, + "step": 33565 + }, + { + "epoch": 5.4763846969573375, + "grad_norm": 0.000915067212190479, + "learning_rate": 0.018730328750469514, + "loss": 0.3511, + "num_input_tokens_seen": 70126800, + "step": 33570 + }, + { + "epoch": 5.477200424178155, + "grad_norm": 0.0003633495362009853, + "learning_rate": 0.018701835645460473, + "loss": 0.3289, + "num_input_tokens_seen": 70137504, + "step": 33575 + }, + { + "epoch": 5.478016151398972, + "grad_norm": 0.0012639269698411226, + "learning_rate": 0.01867336278827838, + "loss": 0.4202, + "num_input_tokens_seen": 70148528, + "step": 33580 + }, + { + "epoch": 5.47883187861979, + "grad_norm": 0.0004074903845321387, + "learning_rate": 0.018644910183314056, + "loss": 0.3109, + "num_input_tokens_seen": 70159392, + "step": 33585 + }, + { + "epoch": 5.479647605840607, + "grad_norm": 0.0008454523631371558, + "learning_rate": 0.01861647783495531, + "loss": 0.3658, + "num_input_tokens_seen": 70169040, + "step": 33590 + }, + { + "epoch": 5.480463333061424, + "grad_norm": 0.0004478857445064932, + "learning_rate": 0.01858806574758676, + "loss": 0.3322, + "num_input_tokens_seen": 70178832, + "step": 33595 + }, + { + "epoch": 5.481279060282241, + "grad_norm": 0.0005698925233446062, + "learning_rate": 0.01855967392558988, + "loss": 0.3595, + "num_input_tokens_seen": 70189824, + "step": 33600 + }, + { + "epoch": 5.481279060282241, + "eval_loss": 0.3148590326309204, + "eval_runtime": 155.7609, + "eval_samples_per_second": 17.495, + "eval_steps_per_second": 8.751, + "num_input_tokens_seen": 70189824, + "step": 33600 + }, + { + "epoch": 5.482094787503059, + "grad_norm": 0.0004956421325914562, + "learning_rate": 0.018531302373343096, + "loss": 0.2857, + "num_input_tokens_seen": 70200560, + "step": 33605 + }, + { + "epoch": 5.482910514723876, + "grad_norm": 0.0008160844445228577, + "learning_rate": 0.018502951095221588, + "loss": 0.3439, + "num_input_tokens_seen": 70211088, + "step": 33610 + }, + { + "epoch": 5.4837262419446935, + "grad_norm": 0.0006024445174261928, + "learning_rate": 0.01847462009559751, + "loss": 0.2899, + "num_input_tokens_seen": 70222144, + "step": 33615 + }, + { + "epoch": 5.484541969165511, + "grad_norm": 0.000924219551961869, + "learning_rate": 0.01844630937883992, + "loss": 0.4149, + "num_input_tokens_seen": 70233568, + "step": 33620 + }, + { + "epoch": 5.485357696386329, + "grad_norm": 0.0005470383330248296, + "learning_rate": 0.018418018949314573, + "loss": 0.3076, + "num_input_tokens_seen": 70244336, + "step": 33625 + }, + { + "epoch": 5.486173423607146, + "grad_norm": 0.0005179574945941567, + "learning_rate": 0.018389748811384315, + "loss": 0.336, + "num_input_tokens_seen": 70255184, + "step": 33630 + }, + { + "epoch": 5.486989150827963, + "grad_norm": 0.0006700403173454106, + "learning_rate": 0.018361498969408658, + "loss": 0.3202, + "num_input_tokens_seen": 70265920, + "step": 33635 + }, + { + "epoch": 5.487804878048781, + "grad_norm": 0.00046198643394745886, + "learning_rate": 0.01833326942774415, + "loss": 0.3113, + "num_input_tokens_seen": 70276672, + "step": 33640 + }, + { + "epoch": 5.488620605269598, + "grad_norm": 0.0004927188856527209, + "learning_rate": 0.018305060190744155, + "loss": 0.3083, + "num_input_tokens_seen": 70287824, + "step": 33645 + }, + { + "epoch": 5.489436332490415, + "grad_norm": 0.0005418746150098741, + "learning_rate": 0.018276871262758846, + "loss": 0.2759, + "num_input_tokens_seen": 70298064, + "step": 33650 + }, + { + "epoch": 5.490252059711232, + "grad_norm": 0.0006736043724231422, + "learning_rate": 0.0182487026481353, + "loss": 0.3117, + "num_input_tokens_seen": 70308208, + "step": 33655 + }, + { + "epoch": 5.49106778693205, + "grad_norm": 0.00047117393114604056, + "learning_rate": 0.018220554351217538, + "loss": 0.3447, + "num_input_tokens_seen": 70317648, + "step": 33660 + }, + { + "epoch": 5.4918835141528675, + "grad_norm": 0.0006623180233873427, + "learning_rate": 0.01819242637634629, + "loss": 0.3214, + "num_input_tokens_seen": 70328688, + "step": 33665 + }, + { + "epoch": 5.492699241373685, + "grad_norm": 0.0013623784761875868, + "learning_rate": 0.01816431872785933, + "loss": 0.3208, + "num_input_tokens_seen": 70339792, + "step": 33670 + }, + { + "epoch": 5.493514968594502, + "grad_norm": 0.00043276476208120584, + "learning_rate": 0.018136231410091148, + "loss": 0.2984, + "num_input_tokens_seen": 70350032, + "step": 33675 + }, + { + "epoch": 5.49433069581532, + "grad_norm": 0.0011197348358109593, + "learning_rate": 0.018108164427373175, + "loss": 0.353, + "num_input_tokens_seen": 70360688, + "step": 33680 + }, + { + "epoch": 5.495146423036137, + "grad_norm": 0.00039907972677610815, + "learning_rate": 0.01808011778403375, + "loss": 0.3423, + "num_input_tokens_seen": 70370272, + "step": 33685 + }, + { + "epoch": 5.495962150256954, + "grad_norm": 0.0006033574463799596, + "learning_rate": 0.01805209148439793, + "loss": 0.2698, + "num_input_tokens_seen": 70380080, + "step": 33690 + }, + { + "epoch": 5.496777877477771, + "grad_norm": 0.0007406800286844373, + "learning_rate": 0.018024085532787757, + "loss": 0.2789, + "num_input_tokens_seen": 70390080, + "step": 33695 + }, + { + "epoch": 5.497593604698589, + "grad_norm": 0.0007010294357314706, + "learning_rate": 0.017996099933522164, + "loss": 0.3375, + "num_input_tokens_seen": 70400720, + "step": 33700 + }, + { + "epoch": 5.498409331919406, + "grad_norm": 0.0006793543579988182, + "learning_rate": 0.017968134690916775, + "loss": 0.3531, + "num_input_tokens_seen": 70410736, + "step": 33705 + }, + { + "epoch": 5.499225059140223, + "grad_norm": 0.0004989821463823318, + "learning_rate": 0.017940189809284263, + "loss": 0.3052, + "num_input_tokens_seen": 70420976, + "step": 33710 + }, + { + "epoch": 5.5000407863610405, + "grad_norm": 0.0004055177269037813, + "learning_rate": 0.017912265292934024, + "loss": 0.2688, + "num_input_tokens_seen": 70431280, + "step": 33715 + }, + { + "epoch": 5.500856513581859, + "grad_norm": 0.0005866278079338372, + "learning_rate": 0.017884361146172423, + "loss": 0.2847, + "num_input_tokens_seen": 70441936, + "step": 33720 + }, + { + "epoch": 5.501672240802676, + "grad_norm": 0.0013265348970890045, + "learning_rate": 0.01785647737330261, + "loss": 0.3375, + "num_input_tokens_seen": 70450240, + "step": 33725 + }, + { + "epoch": 5.502487968023493, + "grad_norm": 0.0005161869921721518, + "learning_rate": 0.017828613978624563, + "loss": 0.323, + "num_input_tokens_seen": 70458880, + "step": 33730 + }, + { + "epoch": 5.50330369524431, + "grad_norm": 0.0005572992959059775, + "learning_rate": 0.01780077096643523, + "loss": 0.278, + "num_input_tokens_seen": 70468288, + "step": 33735 + }, + { + "epoch": 5.504119422465128, + "grad_norm": 0.0004420837794896215, + "learning_rate": 0.017772948341028345, + "loss": 0.3674, + "num_input_tokens_seen": 70478384, + "step": 33740 + }, + { + "epoch": 5.504935149685945, + "grad_norm": 0.000720930052921176, + "learning_rate": 0.01774514610669447, + "loss": 0.3434, + "num_input_tokens_seen": 70489344, + "step": 33745 + }, + { + "epoch": 5.505750876906762, + "grad_norm": 0.0007316754199564457, + "learning_rate": 0.017717364267721112, + "loss": 0.2806, + "num_input_tokens_seen": 70500352, + "step": 33750 + }, + { + "epoch": 5.506566604127579, + "grad_norm": 0.00044801432522945106, + "learning_rate": 0.017689602828392513, + "loss": 0.3538, + "num_input_tokens_seen": 70510560, + "step": 33755 + }, + { + "epoch": 5.507382331348397, + "grad_norm": 0.000732945220079273, + "learning_rate": 0.017661861792989897, + "loss": 0.3165, + "num_input_tokens_seen": 70519920, + "step": 33760 + }, + { + "epoch": 5.5081980585692145, + "grad_norm": 0.0007110829465091228, + "learning_rate": 0.017634141165791272, + "loss": 0.3058, + "num_input_tokens_seen": 70530768, + "step": 33765 + }, + { + "epoch": 5.509013785790032, + "grad_norm": 0.0006311091710813344, + "learning_rate": 0.017606440951071455, + "loss": 0.2961, + "num_input_tokens_seen": 70541920, + "step": 33770 + }, + { + "epoch": 5.50982951301085, + "grad_norm": 0.0009822123683989048, + "learning_rate": 0.017578761153102213, + "loss": 0.3128, + "num_input_tokens_seen": 70551472, + "step": 33775 + }, + { + "epoch": 5.510645240231667, + "grad_norm": 0.000610718154348433, + "learning_rate": 0.017551101776152146, + "loss": 0.3512, + "num_input_tokens_seen": 70562112, + "step": 33780 + }, + { + "epoch": 5.511460967452484, + "grad_norm": 0.00048409614828415215, + "learning_rate": 0.017523462824486608, + "loss": 0.3325, + "num_input_tokens_seen": 70571520, + "step": 33785 + }, + { + "epoch": 5.512276694673301, + "grad_norm": 0.00042639614548534155, + "learning_rate": 0.01749584430236794, + "loss": 0.2925, + "num_input_tokens_seen": 70581776, + "step": 33790 + }, + { + "epoch": 5.513092421894118, + "grad_norm": 0.0008432192844338715, + "learning_rate": 0.01746824621405524, + "loss": 0.3004, + "num_input_tokens_seen": 70592048, + "step": 33795 + }, + { + "epoch": 5.513908149114936, + "grad_norm": 0.0003406519826967269, + "learning_rate": 0.017440668563804412, + "loss": 0.2622, + "num_input_tokens_seen": 70602704, + "step": 33800 + }, + { + "epoch": 5.513908149114936, + "eval_loss": 0.31445664167404175, + "eval_runtime": 155.8361, + "eval_samples_per_second": 17.486, + "eval_steps_per_second": 8.746, + "num_input_tokens_seen": 70602704, + "step": 33800 + }, + { + "epoch": 5.514723876335753, + "grad_norm": 0.0005837138742208481, + "learning_rate": 0.017413111355868392, + "loss": 0.3129, + "num_input_tokens_seen": 70614112, + "step": 33805 + }, + { + "epoch": 5.5155396035565705, + "grad_norm": 0.00041055481415241957, + "learning_rate": 0.017385574594496748, + "loss": 0.3539, + "num_input_tokens_seen": 70624400, + "step": 33810 + }, + { + "epoch": 5.5163553307773885, + "grad_norm": 0.0007029402768239379, + "learning_rate": 0.01735805828393605, + "loss": 0.3546, + "num_input_tokens_seen": 70636048, + "step": 33815 + }, + { + "epoch": 5.517171057998206, + "grad_norm": 0.0003558016032911837, + "learning_rate": 0.017330562428429667, + "loss": 0.2138, + "num_input_tokens_seen": 70647168, + "step": 33820 + }, + { + "epoch": 5.517986785219023, + "grad_norm": 0.00035311191459186375, + "learning_rate": 0.01730308703221776, + "loss": 0.3679, + "num_input_tokens_seen": 70658976, + "step": 33825 + }, + { + "epoch": 5.51880251243984, + "grad_norm": 0.0005759859923273325, + "learning_rate": 0.01727563209953744, + "loss": 0.336, + "num_input_tokens_seen": 70670608, + "step": 33830 + }, + { + "epoch": 5.519618239660657, + "grad_norm": 0.0005483304848894477, + "learning_rate": 0.017248197634622535, + "loss": 0.3264, + "num_input_tokens_seen": 70681008, + "step": 33835 + }, + { + "epoch": 5.520433966881475, + "grad_norm": 0.00048610285739414394, + "learning_rate": 0.01722078364170383, + "loss": 0.3167, + "num_input_tokens_seen": 70692576, + "step": 33840 + }, + { + "epoch": 5.521249694102292, + "grad_norm": 0.0006118621677160263, + "learning_rate": 0.017193390125008905, + "loss": 0.3235, + "num_input_tokens_seen": 70703616, + "step": 33845 + }, + { + "epoch": 5.522065421323109, + "grad_norm": 0.0006683837855234742, + "learning_rate": 0.017166017088762153, + "loss": 0.2586, + "num_input_tokens_seen": 70714848, + "step": 33850 + }, + { + "epoch": 5.522881148543927, + "grad_norm": 0.0010525111574679613, + "learning_rate": 0.017138664537184878, + "loss": 0.3216, + "num_input_tokens_seen": 70726000, + "step": 33855 + }, + { + "epoch": 5.523696875764744, + "grad_norm": 0.0008790604188106954, + "learning_rate": 0.017111332474495172, + "loss": 0.3489, + "num_input_tokens_seen": 70736960, + "step": 33860 + }, + { + "epoch": 5.524512602985562, + "grad_norm": 0.000578471168410033, + "learning_rate": 0.017084020904907998, + "loss": 0.3818, + "num_input_tokens_seen": 70746688, + "step": 33865 + }, + { + "epoch": 5.525328330206379, + "grad_norm": 0.0008136756368912756, + "learning_rate": 0.017056729832635103, + "loss": 0.3616, + "num_input_tokens_seen": 70758304, + "step": 33870 + }, + { + "epoch": 5.526144057427197, + "grad_norm": 0.0005618312279693782, + "learning_rate": 0.017029459261885153, + "loss": 0.3141, + "num_input_tokens_seen": 70768304, + "step": 33875 + }, + { + "epoch": 5.526959784648014, + "grad_norm": 0.000612565956544131, + "learning_rate": 0.01700220919686359, + "loss": 0.2894, + "num_input_tokens_seen": 70778944, + "step": 33880 + }, + { + "epoch": 5.527775511868831, + "grad_norm": 0.0005215267301537097, + "learning_rate": 0.016974979641772723, + "loss": 0.2817, + "num_input_tokens_seen": 70790960, + "step": 33885 + }, + { + "epoch": 5.528591239089648, + "grad_norm": 0.0007174051715992391, + "learning_rate": 0.01694777060081169, + "loss": 0.2876, + "num_input_tokens_seen": 70801744, + "step": 33890 + }, + { + "epoch": 5.529406966310466, + "grad_norm": 0.0006421355064958334, + "learning_rate": 0.016920582078176444, + "loss": 0.3594, + "num_input_tokens_seen": 70811344, + "step": 33895 + }, + { + "epoch": 5.530222693531283, + "grad_norm": 0.0007630245527252555, + "learning_rate": 0.016893414078059863, + "loss": 0.3496, + "num_input_tokens_seen": 70820608, + "step": 33900 + }, + { + "epoch": 5.5310384207521, + "grad_norm": 0.001171381794847548, + "learning_rate": 0.016866266604651535, + "loss": 0.358, + "num_input_tokens_seen": 70832208, + "step": 33905 + }, + { + "epoch": 5.5318541479729175, + "grad_norm": 0.00039431359618902206, + "learning_rate": 0.016839139662137976, + "loss": 0.2814, + "num_input_tokens_seen": 70842064, + "step": 33910 + }, + { + "epoch": 5.5326698751937355, + "grad_norm": 0.0005011894972994924, + "learning_rate": 0.01681203325470245, + "loss": 0.2755, + "num_input_tokens_seen": 70851280, + "step": 33915 + }, + { + "epoch": 5.533485602414553, + "grad_norm": 0.000753512722440064, + "learning_rate": 0.016784947386525157, + "loss": 0.322, + "num_input_tokens_seen": 70860832, + "step": 33920 + }, + { + "epoch": 5.53430132963537, + "grad_norm": 0.0007049458217807114, + "learning_rate": 0.01675788206178308, + "loss": 0.2763, + "num_input_tokens_seen": 70871840, + "step": 33925 + }, + { + "epoch": 5.535117056856187, + "grad_norm": 0.0007919640047475696, + "learning_rate": 0.016730837284649986, + "loss": 0.3014, + "num_input_tokens_seen": 70881504, + "step": 33930 + }, + { + "epoch": 5.535932784077005, + "grad_norm": 0.00037148947012610734, + "learning_rate": 0.016703813059296583, + "loss": 0.2884, + "num_input_tokens_seen": 70892464, + "step": 33935 + }, + { + "epoch": 5.536748511297822, + "grad_norm": 0.00043235349585302174, + "learning_rate": 0.016676809389890294, + "loss": 0.3166, + "num_input_tokens_seen": 70903888, + "step": 33940 + }, + { + "epoch": 5.537564238518639, + "grad_norm": 0.0005255614523775876, + "learning_rate": 0.016649826280595435, + "loss": 0.3659, + "num_input_tokens_seen": 70914432, + "step": 33945 + }, + { + "epoch": 5.538379965739456, + "grad_norm": 0.0005344055825844407, + "learning_rate": 0.016622863735573163, + "loss": 0.3459, + "num_input_tokens_seen": 70926496, + "step": 33950 + }, + { + "epoch": 5.539195692960274, + "grad_norm": 0.0005529720219783485, + "learning_rate": 0.016595921758981395, + "loss": 0.2864, + "num_input_tokens_seen": 70935280, + "step": 33955 + }, + { + "epoch": 5.5400114201810915, + "grad_norm": 0.0008366044494323432, + "learning_rate": 0.01656900035497495, + "loss": 0.3148, + "num_input_tokens_seen": 70943968, + "step": 33960 + }, + { + "epoch": 5.540827147401909, + "grad_norm": 0.00045827985741198063, + "learning_rate": 0.016542099527705485, + "loss": 0.2803, + "num_input_tokens_seen": 70955440, + "step": 33965 + }, + { + "epoch": 5.541642874622726, + "grad_norm": 0.000516161322593689, + "learning_rate": 0.01651521928132138, + "loss": 0.2695, + "num_input_tokens_seen": 70965936, + "step": 33970 + }, + { + "epoch": 5.542458601843544, + "grad_norm": 0.0008285950752906501, + "learning_rate": 0.01648835961996794, + "loss": 0.3268, + "num_input_tokens_seen": 70976880, + "step": 33975 + }, + { + "epoch": 5.543274329064361, + "grad_norm": 0.0004153219342697412, + "learning_rate": 0.016461520547787285, + "loss": 0.2925, + "num_input_tokens_seen": 70988896, + "step": 33980 + }, + { + "epoch": 5.544090056285178, + "grad_norm": 0.0008425276027992368, + "learning_rate": 0.016434702068918266, + "loss": 0.3181, + "num_input_tokens_seen": 70999376, + "step": 33985 + }, + { + "epoch": 5.544905783505996, + "grad_norm": 0.0012939333682879806, + "learning_rate": 0.01640790418749673, + "loss": 0.3211, + "num_input_tokens_seen": 71011360, + "step": 33990 + }, + { + "epoch": 5.545721510726813, + "grad_norm": 0.0006058299331925809, + "learning_rate": 0.016381126907655134, + "loss": 0.2934, + "num_input_tokens_seen": 71022528, + "step": 33995 + }, + { + "epoch": 5.54653723794763, + "grad_norm": 0.0006593712023459375, + "learning_rate": 0.016354370233522948, + "loss": 0.3138, + "num_input_tokens_seen": 71032768, + "step": 34000 + }, + { + "epoch": 5.54653723794763, + "eval_loss": 0.31373628973960876, + "eval_runtime": 155.8091, + "eval_samples_per_second": 17.489, + "eval_steps_per_second": 8.748, + "num_input_tokens_seen": 71032768, + "step": 34000 + }, + { + "epoch": 5.5473529651684474, + "grad_norm": 0.0005112604121677577, + "learning_rate": 0.016327634169226394, + "loss": 0.3454, + "num_input_tokens_seen": 71043328, + "step": 34005 + }, + { + "epoch": 5.548168692389265, + "grad_norm": 0.0013843702618032694, + "learning_rate": 0.016300918718888485, + "loss": 0.3498, + "num_input_tokens_seen": 71054096, + "step": 34010 + }, + { + "epoch": 5.548984419610083, + "grad_norm": 0.0006464109756052494, + "learning_rate": 0.016274223886629052, + "loss": 0.3374, + "num_input_tokens_seen": 71064384, + "step": 34015 + }, + { + "epoch": 5.5498001468309, + "grad_norm": 0.0005588604253716767, + "learning_rate": 0.01624754967656482, + "loss": 0.2471, + "num_input_tokens_seen": 71074272, + "step": 34020 + }, + { + "epoch": 5.550615874051717, + "grad_norm": 0.00038853927981108427, + "learning_rate": 0.016220896092809235, + "loss": 0.2784, + "num_input_tokens_seen": 71083584, + "step": 34025 + }, + { + "epoch": 5.551431601272535, + "grad_norm": 0.0004356972349341959, + "learning_rate": 0.01619426313947267, + "loss": 0.3509, + "num_input_tokens_seen": 71092560, + "step": 34030 + }, + { + "epoch": 5.552247328493352, + "grad_norm": 0.0006385737797245383, + "learning_rate": 0.016167650820662228, + "loss": 0.2869, + "num_input_tokens_seen": 71103216, + "step": 34035 + }, + { + "epoch": 5.553063055714169, + "grad_norm": 0.0007261291611939669, + "learning_rate": 0.016141059140481855, + "loss": 0.36, + "num_input_tokens_seen": 71113568, + "step": 34040 + }, + { + "epoch": 5.553878782934986, + "grad_norm": 0.0005294921575114131, + "learning_rate": 0.016114488103032374, + "loss": 0.2632, + "num_input_tokens_seen": 71123872, + "step": 34045 + }, + { + "epoch": 5.554694510155803, + "grad_norm": 0.0005421992391347885, + "learning_rate": 0.016087937712411293, + "loss": 0.2293, + "num_input_tokens_seen": 71133792, + "step": 34050 + }, + { + "epoch": 5.555510237376621, + "grad_norm": 0.0009569602552801371, + "learning_rate": 0.01606140797271308, + "loss": 0.3195, + "num_input_tokens_seen": 71143472, + "step": 34055 + }, + { + "epoch": 5.5563259645974385, + "grad_norm": 0.0007519181235693395, + "learning_rate": 0.01603489888802897, + "loss": 0.3583, + "num_input_tokens_seen": 71154528, + "step": 34060 + }, + { + "epoch": 5.557141691818256, + "grad_norm": 0.00047419662587344646, + "learning_rate": 0.016008410462446918, + "loss": 0.2881, + "num_input_tokens_seen": 71166400, + "step": 34065 + }, + { + "epoch": 5.557957419039074, + "grad_norm": 0.00047724336036480963, + "learning_rate": 0.01598194270005185, + "loss": 0.3362, + "num_input_tokens_seen": 71178000, + "step": 34070 + }, + { + "epoch": 5.558773146259891, + "grad_norm": 0.0006809478509239852, + "learning_rate": 0.015955495604925356, + "loss": 0.2899, + "num_input_tokens_seen": 71187856, + "step": 34075 + }, + { + "epoch": 5.559588873480708, + "grad_norm": 0.0007833741838112473, + "learning_rate": 0.01592906918114598, + "loss": 0.3937, + "num_input_tokens_seen": 71198256, + "step": 34080 + }, + { + "epoch": 5.560404600701525, + "grad_norm": 0.0008499902323819697, + "learning_rate": 0.015902663432788965, + "loss": 0.2838, + "num_input_tokens_seen": 71208928, + "step": 34085 + }, + { + "epoch": 5.561220327922343, + "grad_norm": 0.0014819451607763767, + "learning_rate": 0.01587627836392643, + "loss": 0.4158, + "num_input_tokens_seen": 71219168, + "step": 34090 + }, + { + "epoch": 5.56203605514316, + "grad_norm": 0.0008818808710202575, + "learning_rate": 0.01584991397862726, + "loss": 0.2548, + "num_input_tokens_seen": 71229696, + "step": 34095 + }, + { + "epoch": 5.562851782363977, + "grad_norm": 0.0004775257548317313, + "learning_rate": 0.015823570280957214, + "loss": 0.3066, + "num_input_tokens_seen": 71240736, + "step": 34100 + }, + { + "epoch": 5.5636675095847945, + "grad_norm": 0.00043328627361916006, + "learning_rate": 0.015797247274978766, + "loss": 0.2883, + "num_input_tokens_seen": 71249376, + "step": 34105 + }, + { + "epoch": 5.5644832368056125, + "grad_norm": 0.0005070827319286764, + "learning_rate": 0.015770944964751326, + "loss": 0.3055, + "num_input_tokens_seen": 71258848, + "step": 34110 + }, + { + "epoch": 5.56529896402643, + "grad_norm": 0.0006735521019436419, + "learning_rate": 0.015744663354330956, + "loss": 0.2945, + "num_input_tokens_seen": 71267808, + "step": 34115 + }, + { + "epoch": 5.566114691247247, + "grad_norm": 0.0007266960456036031, + "learning_rate": 0.015718402447770664, + "loss": 0.2854, + "num_input_tokens_seen": 71279072, + "step": 34120 + }, + { + "epoch": 5.566930418468064, + "grad_norm": 0.0005613576504401863, + "learning_rate": 0.015692162249120224, + "loss": 0.3393, + "num_input_tokens_seen": 71289504, + "step": 34125 + }, + { + "epoch": 5.567746145688882, + "grad_norm": 0.0006027127965353429, + "learning_rate": 0.01566594276242615, + "loss": 0.3118, + "num_input_tokens_seen": 71299392, + "step": 34130 + }, + { + "epoch": 5.568561872909699, + "grad_norm": 0.0004897345788776875, + "learning_rate": 0.015639743991731857, + "loss": 0.3395, + "num_input_tokens_seen": 71309664, + "step": 34135 + }, + { + "epoch": 5.569377600130516, + "grad_norm": 0.0003109626704826951, + "learning_rate": 0.01561356594107755, + "loss": 0.2363, + "num_input_tokens_seen": 71320400, + "step": 34140 + }, + { + "epoch": 5.570193327351333, + "grad_norm": 0.0003853575326502323, + "learning_rate": 0.015587408614500147, + "loss": 0.2471, + "num_input_tokens_seen": 71331072, + "step": 34145 + }, + { + "epoch": 5.571009054572151, + "grad_norm": 0.0002692725684028119, + "learning_rate": 0.015561272016033505, + "loss": 0.3055, + "num_input_tokens_seen": 71341824, + "step": 34150 + }, + { + "epoch": 5.5718247817929685, + "grad_norm": 0.0008325760718435049, + "learning_rate": 0.015535156149708167, + "loss": 0.3582, + "num_input_tokens_seen": 71352608, + "step": 34155 + }, + { + "epoch": 5.572640509013786, + "grad_norm": 0.0005437797517515719, + "learning_rate": 0.015509061019551528, + "loss": 0.2917, + "num_input_tokens_seen": 71362656, + "step": 34160 + }, + { + "epoch": 5.573456236234604, + "grad_norm": 0.000911963579710573, + "learning_rate": 0.015482986629587818, + "loss": 0.2897, + "num_input_tokens_seen": 71372496, + "step": 34165 + }, + { + "epoch": 5.574271963455421, + "grad_norm": 0.0006871519144624472, + "learning_rate": 0.01545693298383799, + "loss": 0.2652, + "num_input_tokens_seen": 71382528, + "step": 34170 + }, + { + "epoch": 5.575087690676238, + "grad_norm": 0.0007813041447661817, + "learning_rate": 0.015430900086319858, + "loss": 0.2975, + "num_input_tokens_seen": 71393696, + "step": 34175 + }, + { + "epoch": 5.575903417897055, + "grad_norm": 0.0014982395805418491, + "learning_rate": 0.015404887941048084, + "loss": 0.3467, + "num_input_tokens_seen": 71403216, + "step": 34180 + }, + { + "epoch": 5.576719145117872, + "grad_norm": 0.0006423849263228476, + "learning_rate": 0.01537889655203397, + "loss": 0.3115, + "num_input_tokens_seen": 71414112, + "step": 34185 + }, + { + "epoch": 5.57753487233869, + "grad_norm": 0.0005337302573025227, + "learning_rate": 0.015352925923285798, + "loss": 0.2742, + "num_input_tokens_seen": 71425344, + "step": 34190 + }, + { + "epoch": 5.578350599559507, + "grad_norm": 0.000681750534567982, + "learning_rate": 0.015326976058808511, + "loss": 0.27, + "num_input_tokens_seen": 71435504, + "step": 34195 + }, + { + "epoch": 5.579166326780324, + "grad_norm": 0.0006889811484143138, + "learning_rate": 0.015301046962603908, + "loss": 0.2645, + "num_input_tokens_seen": 71445488, + "step": 34200 + }, + { + "epoch": 5.579166326780324, + "eval_loss": 0.3147622346878052, + "eval_runtime": 156.0983, + "eval_samples_per_second": 17.457, + "eval_steps_per_second": 8.732, + "num_input_tokens_seen": 71445488, + "step": 34200 + }, + { + "epoch": 5.5799820540011424, + "grad_norm": 0.0008826743578538299, + "learning_rate": 0.015275138638670626, + "loss": 0.3046, + "num_input_tokens_seen": 71455584, + "step": 34205 + }, + { + "epoch": 5.58079778122196, + "grad_norm": 0.000639694684650749, + "learning_rate": 0.015249251091004001, + "loss": 0.4104, + "num_input_tokens_seen": 71466432, + "step": 34210 + }, + { + "epoch": 5.581613508442777, + "grad_norm": 0.0006734102498739958, + "learning_rate": 0.01522338432359624, + "loss": 0.3093, + "num_input_tokens_seen": 71476048, + "step": 34215 + }, + { + "epoch": 5.582429235663594, + "grad_norm": 0.00042182623292319477, + "learning_rate": 0.01519753834043635, + "loss": 0.2847, + "num_input_tokens_seen": 71486128, + "step": 34220 + }, + { + "epoch": 5.583244962884411, + "grad_norm": 0.0006450879154726863, + "learning_rate": 0.015171713145510095, + "loss": 0.3026, + "num_input_tokens_seen": 71497328, + "step": 34225 + }, + { + "epoch": 5.584060690105229, + "grad_norm": 0.0011773273581638932, + "learning_rate": 0.01514590874279999, + "loss": 0.3048, + "num_input_tokens_seen": 71507792, + "step": 34230 + }, + { + "epoch": 5.584876417326046, + "grad_norm": 0.0003762225969694555, + "learning_rate": 0.015120125136285467, + "loss": 0.3055, + "num_input_tokens_seen": 71517600, + "step": 34235 + }, + { + "epoch": 5.585692144546863, + "grad_norm": 0.0011323558865115047, + "learning_rate": 0.015094362329942629, + "loss": 0.3405, + "num_input_tokens_seen": 71529728, + "step": 34240 + }, + { + "epoch": 5.586507871767681, + "grad_norm": 0.0005180801963433623, + "learning_rate": 0.01506862032774448, + "loss": 0.29, + "num_input_tokens_seen": 71540784, + "step": 34245 + }, + { + "epoch": 5.587323598988498, + "grad_norm": 0.0004725630860775709, + "learning_rate": 0.015042899133660697, + "loss": 0.3245, + "num_input_tokens_seen": 71550576, + "step": 34250 + }, + { + "epoch": 5.5881393262093155, + "grad_norm": 0.00047499279025942087, + "learning_rate": 0.01501719875165789, + "loss": 0.3014, + "num_input_tokens_seen": 71560656, + "step": 34255 + }, + { + "epoch": 5.588955053430133, + "grad_norm": 0.00114241533447057, + "learning_rate": 0.014991519185699286, + "loss": 0.2681, + "num_input_tokens_seen": 71571792, + "step": 34260 + }, + { + "epoch": 5.589770780650951, + "grad_norm": 0.000837477098684758, + "learning_rate": 0.014965860439745054, + "loss": 0.3302, + "num_input_tokens_seen": 71579712, + "step": 34265 + }, + { + "epoch": 5.590586507871768, + "grad_norm": 0.0012907389318570495, + "learning_rate": 0.01494022251775211, + "loss": 0.2983, + "num_input_tokens_seen": 71590704, + "step": 34270 + }, + { + "epoch": 5.591402235092585, + "grad_norm": 0.0007022881181910634, + "learning_rate": 0.014914605423674109, + "loss": 0.2511, + "num_input_tokens_seen": 71600912, + "step": 34275 + }, + { + "epoch": 5.592217962313402, + "grad_norm": 0.0006420380668714643, + "learning_rate": 0.014889009161461525, + "loss": 0.2997, + "num_input_tokens_seen": 71610848, + "step": 34280 + }, + { + "epoch": 5.59303368953422, + "grad_norm": 0.0005132773076184094, + "learning_rate": 0.014863433735061665, + "loss": 0.2735, + "num_input_tokens_seen": 71620688, + "step": 34285 + }, + { + "epoch": 5.593849416755037, + "grad_norm": 0.0005669415113516152, + "learning_rate": 0.014837879148418541, + "loss": 0.3062, + "num_input_tokens_seen": 71631792, + "step": 34290 + }, + { + "epoch": 5.594665143975854, + "grad_norm": 0.001701171393506229, + "learning_rate": 0.01481234540547302, + "loss": 0.2896, + "num_input_tokens_seen": 71641664, + "step": 34295 + }, + { + "epoch": 5.5954808711966715, + "grad_norm": 0.0009402764262631536, + "learning_rate": 0.014786832510162717, + "loss": 0.3052, + "num_input_tokens_seen": 71652096, + "step": 34300 + }, + { + "epoch": 5.5962965984174895, + "grad_norm": 0.0008545407326892018, + "learning_rate": 0.014761340466422017, + "loss": 0.2912, + "num_input_tokens_seen": 71662576, + "step": 34305 + }, + { + "epoch": 5.597112325638307, + "grad_norm": 0.0006146224332042038, + "learning_rate": 0.014735869278182144, + "loss": 0.2392, + "num_input_tokens_seen": 71672560, + "step": 34310 + }, + { + "epoch": 5.597928052859124, + "grad_norm": 0.0003701621026266366, + "learning_rate": 0.014710418949371057, + "loss": 0.3636, + "num_input_tokens_seen": 71682288, + "step": 34315 + }, + { + "epoch": 5.598743780079941, + "grad_norm": 0.0006804680451750755, + "learning_rate": 0.014684989483913495, + "loss": 0.2918, + "num_input_tokens_seen": 71693792, + "step": 34320 + }, + { + "epoch": 5.599559507300759, + "grad_norm": 0.000680643948726356, + "learning_rate": 0.014659580885731077, + "loss": 0.3558, + "num_input_tokens_seen": 71702880, + "step": 34325 + }, + { + "epoch": 5.600375234521576, + "grad_norm": 0.0007998219225555658, + "learning_rate": 0.014634193158742047, + "loss": 0.3582, + "num_input_tokens_seen": 71714128, + "step": 34330 + }, + { + "epoch": 5.601190961742393, + "grad_norm": 0.0003575154405552894, + "learning_rate": 0.014608826306861576, + "loss": 0.3278, + "num_input_tokens_seen": 71724640, + "step": 34335 + }, + { + "epoch": 5.602006688963211, + "grad_norm": 0.0005299598560668528, + "learning_rate": 0.014583480334001486, + "loss": 0.3253, + "num_input_tokens_seen": 71733648, + "step": 34340 + }, + { + "epoch": 5.602822416184028, + "grad_norm": 0.0010040561901405454, + "learning_rate": 0.014558155244070496, + "loss": 0.3241, + "num_input_tokens_seen": 71744624, + "step": 34345 + }, + { + "epoch": 5.6036381434048455, + "grad_norm": 0.0009512904798611999, + "learning_rate": 0.014532851040974036, + "loss": 0.3012, + "num_input_tokens_seen": 71754624, + "step": 34350 + }, + { + "epoch": 5.604453870625663, + "grad_norm": 0.0011211420642212033, + "learning_rate": 0.014507567728614335, + "loss": 0.3502, + "num_input_tokens_seen": 71764720, + "step": 34355 + }, + { + "epoch": 5.60526959784648, + "grad_norm": 0.00038447827682830393, + "learning_rate": 0.01448230531089037, + "loss": 0.2326, + "num_input_tokens_seen": 71774736, + "step": 34360 + }, + { + "epoch": 5.606085325067298, + "grad_norm": 0.0007091696024872363, + "learning_rate": 0.014457063791697993, + "loss": 0.3387, + "num_input_tokens_seen": 71785712, + "step": 34365 + }, + { + "epoch": 5.606901052288115, + "grad_norm": 0.00035841428325511515, + "learning_rate": 0.01443184317492971, + "loss": 0.3309, + "num_input_tokens_seen": 71796608, + "step": 34370 + }, + { + "epoch": 5.607716779508932, + "grad_norm": 0.0008069264586083591, + "learning_rate": 0.014406643464474822, + "loss": 0.3531, + "num_input_tokens_seen": 71806048, + "step": 34375 + }, + { + "epoch": 5.60853250672975, + "grad_norm": 0.0004788900550920516, + "learning_rate": 0.014381464664219539, + "loss": 0.3051, + "num_input_tokens_seen": 71816384, + "step": 34380 + }, + { + "epoch": 5.609348233950567, + "grad_norm": 0.000791023310739547, + "learning_rate": 0.014356306778046656, + "loss": 0.2879, + "num_input_tokens_seen": 71827088, + "step": 34385 + }, + { + "epoch": 5.610163961171384, + "grad_norm": 0.0006306893192231655, + "learning_rate": 0.014331169809835885, + "loss": 0.3366, + "num_input_tokens_seen": 71838288, + "step": 34390 + }, + { + "epoch": 5.610979688392201, + "grad_norm": 0.001454720040783286, + "learning_rate": 0.014306053763463644, + "loss": 0.3555, + "num_input_tokens_seen": 71848496, + "step": 34395 + }, + { + "epoch": 5.6117954156130185, + "grad_norm": 0.0006934268167242408, + "learning_rate": 0.014280958642803147, + "loss": 0.283, + "num_input_tokens_seen": 71858096, + "step": 34400 + }, + { + "epoch": 5.6117954156130185, + "eval_loss": 0.3143464922904968, + "eval_runtime": 155.7549, + "eval_samples_per_second": 17.495, + "eval_steps_per_second": 8.751, + "num_input_tokens_seen": 71858096, + "step": 34400 + }, + { + "epoch": 5.6126111428338366, + "grad_norm": 0.0006499727605842054, + "learning_rate": 0.014255884451724404, + "loss": 0.3034, + "num_input_tokens_seen": 71868224, + "step": 34405 + }, + { + "epoch": 5.613426870054654, + "grad_norm": 0.0005363639793358743, + "learning_rate": 0.014230831194094101, + "loss": 0.2975, + "num_input_tokens_seen": 71877264, + "step": 34410 + }, + { + "epoch": 5.614242597275471, + "grad_norm": 0.000708153413143009, + "learning_rate": 0.014205798873775865, + "loss": 0.2835, + "num_input_tokens_seen": 71888592, + "step": 34415 + }, + { + "epoch": 5.615058324496289, + "grad_norm": 0.00042649521492421627, + "learning_rate": 0.014180787494629893, + "loss": 0.3152, + "num_input_tokens_seen": 71898640, + "step": 34420 + }, + { + "epoch": 5.615874051717106, + "grad_norm": 0.00043598024058155715, + "learning_rate": 0.014155797060513314, + "loss": 0.3172, + "num_input_tokens_seen": 71909040, + "step": 34425 + }, + { + "epoch": 5.616689778937923, + "grad_norm": 0.00112501485273242, + "learning_rate": 0.014130827575279963, + "loss": 0.35, + "num_input_tokens_seen": 71919440, + "step": 34430 + }, + { + "epoch": 5.61750550615874, + "grad_norm": 0.00048210506793111563, + "learning_rate": 0.014105879042780427, + "loss": 0.294, + "num_input_tokens_seen": 71930384, + "step": 34435 + }, + { + "epoch": 5.618321233379558, + "grad_norm": 0.0006475599948316813, + "learning_rate": 0.014080951466862113, + "loss": 0.2921, + "num_input_tokens_seen": 71940016, + "step": 34440 + }, + { + "epoch": 5.619136960600375, + "grad_norm": 0.00040711258770897985, + "learning_rate": 0.014056044851369126, + "loss": 0.2942, + "num_input_tokens_seen": 71951776, + "step": 34445 + }, + { + "epoch": 5.6199526878211925, + "grad_norm": 0.0004299394495319575, + "learning_rate": 0.014031159200142428, + "loss": 0.2694, + "num_input_tokens_seen": 71962688, + "step": 34450 + }, + { + "epoch": 5.62076841504201, + "grad_norm": 0.0008743057260289788, + "learning_rate": 0.014006294517019667, + "loss": 0.2944, + "num_input_tokens_seen": 71973040, + "step": 34455 + }, + { + "epoch": 5.621584142262828, + "grad_norm": 0.0005206374917179346, + "learning_rate": 0.013981450805835276, + "loss": 0.2494, + "num_input_tokens_seen": 71982976, + "step": 34460 + }, + { + "epoch": 5.622399869483645, + "grad_norm": 0.0003405621100682765, + "learning_rate": 0.01395662807042049, + "loss": 0.3107, + "num_input_tokens_seen": 71994320, + "step": 34465 + }, + { + "epoch": 5.623215596704462, + "grad_norm": 0.0006261837552301586, + "learning_rate": 0.013931826314603296, + "loss": 0.3771, + "num_input_tokens_seen": 72004960, + "step": 34470 + }, + { + "epoch": 5.624031323925279, + "grad_norm": 0.0010303851449862123, + "learning_rate": 0.013907045542208401, + "loss": 0.376, + "num_input_tokens_seen": 72015216, + "step": 34475 + }, + { + "epoch": 5.624847051146097, + "grad_norm": 0.0005550393252633512, + "learning_rate": 0.013882285757057333, + "loss": 0.2699, + "num_input_tokens_seen": 72027328, + "step": 34480 + }, + { + "epoch": 5.625662778366914, + "grad_norm": 0.0005753745208494365, + "learning_rate": 0.013857546962968403, + "loss": 0.3359, + "num_input_tokens_seen": 72037600, + "step": 34485 + }, + { + "epoch": 5.626478505587731, + "grad_norm": 0.0018323056865483522, + "learning_rate": 0.013832829163756577, + "loss": 0.3598, + "num_input_tokens_seen": 72045760, + "step": 34490 + }, + { + "epoch": 5.6272942328085485, + "grad_norm": 0.00046372783253900707, + "learning_rate": 0.013808132363233689, + "loss": 0.3517, + "num_input_tokens_seen": 72056608, + "step": 34495 + }, + { + "epoch": 5.6281099600293665, + "grad_norm": 0.0005047958111390471, + "learning_rate": 0.013783456565208256, + "loss": 0.269, + "num_input_tokens_seen": 72066976, + "step": 34500 + }, + { + "epoch": 5.628925687250184, + "grad_norm": 0.0008163308957591653, + "learning_rate": 0.01375880177348564, + "loss": 0.3488, + "num_input_tokens_seen": 72077504, + "step": 34505 + }, + { + "epoch": 5.629741414471001, + "grad_norm": 0.0005979674169793725, + "learning_rate": 0.013734167991867928, + "loss": 0.2954, + "num_input_tokens_seen": 72086000, + "step": 34510 + }, + { + "epoch": 5.630557141691818, + "grad_norm": 0.0005898991948924959, + "learning_rate": 0.013709555224153935, + "loss": 0.3003, + "num_input_tokens_seen": 72095824, + "step": 34515 + }, + { + "epoch": 5.631372868912636, + "grad_norm": 0.0007781829917803407, + "learning_rate": 0.013684963474139222, + "loss": 0.3049, + "num_input_tokens_seen": 72106224, + "step": 34520 + }, + { + "epoch": 5.632188596133453, + "grad_norm": 0.0008064210996963084, + "learning_rate": 0.013660392745616224, + "loss": 0.2851, + "num_input_tokens_seen": 72117584, + "step": 34525 + }, + { + "epoch": 5.63300432335427, + "grad_norm": 0.0006214092136360705, + "learning_rate": 0.013635843042373974, + "loss": 0.3424, + "num_input_tokens_seen": 72126896, + "step": 34530 + }, + { + "epoch": 5.633820050575087, + "grad_norm": 0.000378641503630206, + "learning_rate": 0.01361131436819843, + "loss": 0.3315, + "num_input_tokens_seen": 72138912, + "step": 34535 + }, + { + "epoch": 5.634635777795905, + "grad_norm": 0.000659390352666378, + "learning_rate": 0.013586806726872147, + "loss": 0.2442, + "num_input_tokens_seen": 72150160, + "step": 34540 + }, + { + "epoch": 5.635451505016722, + "grad_norm": 0.00038931393646635115, + "learning_rate": 0.013562320122174537, + "loss": 0.2815, + "num_input_tokens_seen": 72161008, + "step": 34545 + }, + { + "epoch": 5.63626723223754, + "grad_norm": 0.0005731654819101095, + "learning_rate": 0.013537854557881762, + "loss": 0.327, + "num_input_tokens_seen": 72172768, + "step": 34550 + }, + { + "epoch": 5.637082959458358, + "grad_norm": 0.00042194969137199223, + "learning_rate": 0.013513410037766687, + "loss": 0.3596, + "num_input_tokens_seen": 72183552, + "step": 34555 + }, + { + "epoch": 5.637898686679175, + "grad_norm": 0.0005959411500953138, + "learning_rate": 0.013488986565598998, + "loss": 0.2849, + "num_input_tokens_seen": 72193664, + "step": 34560 + }, + { + "epoch": 5.638714413899992, + "grad_norm": 0.00036562790046446025, + "learning_rate": 0.013464584145145097, + "loss": 0.2739, + "num_input_tokens_seen": 72204592, + "step": 34565 + }, + { + "epoch": 5.639530141120809, + "grad_norm": 0.000903002277482301, + "learning_rate": 0.013440202780168109, + "loss": 0.3062, + "num_input_tokens_seen": 72215760, + "step": 34570 + }, + { + "epoch": 5.640345868341626, + "grad_norm": 0.0010639277752488852, + "learning_rate": 0.01341584247442799, + "loss": 0.3051, + "num_input_tokens_seen": 72226368, + "step": 34575 + }, + { + "epoch": 5.641161595562444, + "grad_norm": 0.000771414372138679, + "learning_rate": 0.013391503231681355, + "loss": 0.2287, + "num_input_tokens_seen": 72237360, + "step": 34580 + }, + { + "epoch": 5.641977322783261, + "grad_norm": 0.00038566641160286963, + "learning_rate": 0.013367185055681685, + "loss": 0.3497, + "num_input_tokens_seen": 72246928, + "step": 34585 + }, + { + "epoch": 5.642793050004078, + "grad_norm": 0.00048499691183678806, + "learning_rate": 0.013342887950179095, + "loss": 0.3186, + "num_input_tokens_seen": 72256400, + "step": 34590 + }, + { + "epoch": 5.643608777224896, + "grad_norm": 0.00048270795377902687, + "learning_rate": 0.013318611918920554, + "loss": 0.2165, + "num_input_tokens_seen": 72265744, + "step": 34595 + }, + { + "epoch": 5.6444245044457135, + "grad_norm": 0.0004962876555509865, + "learning_rate": 0.01329435696564965, + "loss": 0.3269, + "num_input_tokens_seen": 72276272, + "step": 34600 + }, + { + "epoch": 5.6444245044457135, + "eval_loss": 0.3150556981563568, + "eval_runtime": 155.8524, + "eval_samples_per_second": 17.484, + "eval_steps_per_second": 8.745, + "num_input_tokens_seen": 72276272, + "step": 34600 + }, + { + "epoch": 5.645240231666531, + "grad_norm": 0.0007212595082819462, + "learning_rate": 0.013270123094106894, + "loss": 0.3269, + "num_input_tokens_seen": 72287232, + "step": 34605 + }, + { + "epoch": 5.646055958887348, + "grad_norm": 0.00102612457703799, + "learning_rate": 0.013245910308029395, + "loss": 0.3147, + "num_input_tokens_seen": 72298224, + "step": 34610 + }, + { + "epoch": 5.646871686108166, + "grad_norm": 0.001351755578070879, + "learning_rate": 0.0132217186111511, + "loss": 0.4544, + "num_input_tokens_seen": 72308384, + "step": 34615 + }, + { + "epoch": 5.647687413328983, + "grad_norm": 0.0010777597781270742, + "learning_rate": 0.013197548007202626, + "loss": 0.3455, + "num_input_tokens_seen": 72318208, + "step": 34620 + }, + { + "epoch": 5.6485031405498, + "grad_norm": 0.0005048444727435708, + "learning_rate": 0.01317339849991142, + "loss": 0.328, + "num_input_tokens_seen": 72329152, + "step": 34625 + }, + { + "epoch": 5.649318867770617, + "grad_norm": 0.001082432339899242, + "learning_rate": 0.013149270093001675, + "loss": 0.2866, + "num_input_tokens_seen": 72340016, + "step": 34630 + }, + { + "epoch": 5.650134594991435, + "grad_norm": 0.0006255211774259806, + "learning_rate": 0.013125162790194227, + "loss": 0.299, + "num_input_tokens_seen": 72351392, + "step": 34635 + }, + { + "epoch": 5.650950322212252, + "grad_norm": 0.0004777397552970797, + "learning_rate": 0.01310107659520674, + "loss": 0.2857, + "num_input_tokens_seen": 72361888, + "step": 34640 + }, + { + "epoch": 5.6517660494330695, + "grad_norm": 0.0005293307476677, + "learning_rate": 0.013077011511753655, + "loss": 0.3097, + "num_input_tokens_seen": 72372320, + "step": 34645 + }, + { + "epoch": 5.652581776653887, + "grad_norm": 0.0007857251330278814, + "learning_rate": 0.013052967543546056, + "loss": 0.2705, + "num_input_tokens_seen": 72382960, + "step": 34650 + }, + { + "epoch": 5.653397503874705, + "grad_norm": 0.0009260557708330452, + "learning_rate": 0.01302894469429186, + "loss": 0.3755, + "num_input_tokens_seen": 72392912, + "step": 34655 + }, + { + "epoch": 5.654213231095522, + "grad_norm": 0.0008033743361011147, + "learning_rate": 0.013004942967695653, + "loss": 0.3462, + "num_input_tokens_seen": 72402496, + "step": 34660 + }, + { + "epoch": 5.655028958316339, + "grad_norm": 0.0006975589203648269, + "learning_rate": 0.012980962367458859, + "loss": 0.3527, + "num_input_tokens_seen": 72413264, + "step": 34665 + }, + { + "epoch": 5.655844685537156, + "grad_norm": 0.0005523458239622414, + "learning_rate": 0.012957002897279567, + "loss": 0.318, + "num_input_tokens_seen": 72422864, + "step": 34670 + }, + { + "epoch": 5.656660412757974, + "grad_norm": 0.0004765316261909902, + "learning_rate": 0.012933064560852576, + "loss": 0.3341, + "num_input_tokens_seen": 72432944, + "step": 34675 + }, + { + "epoch": 5.657476139978791, + "grad_norm": 0.0004632220952771604, + "learning_rate": 0.012909147361869527, + "loss": 0.3182, + "num_input_tokens_seen": 72443168, + "step": 34680 + }, + { + "epoch": 5.658291867199608, + "grad_norm": 0.0007375223794952035, + "learning_rate": 0.012885251304018774, + "loss": 0.291, + "num_input_tokens_seen": 72452128, + "step": 34685 + }, + { + "epoch": 5.659107594420425, + "grad_norm": 0.0007189548923633993, + "learning_rate": 0.012861376390985335, + "loss": 0.3331, + "num_input_tokens_seen": 72463520, + "step": 34690 + }, + { + "epoch": 5.6599233216412435, + "grad_norm": 0.00038061695522628725, + "learning_rate": 0.012837522626451063, + "loss": 0.3297, + "num_input_tokens_seen": 72474160, + "step": 34695 + }, + { + "epoch": 5.660739048862061, + "grad_norm": 0.0005148261552676558, + "learning_rate": 0.01281369001409447, + "loss": 0.3299, + "num_input_tokens_seen": 72483936, + "step": 34700 + }, + { + "epoch": 5.661554776082878, + "grad_norm": 0.001398957334458828, + "learning_rate": 0.012789878557590877, + "loss": 0.3366, + "num_input_tokens_seen": 72496208, + "step": 34705 + }, + { + "epoch": 5.662370503303695, + "grad_norm": 0.0006506264908239245, + "learning_rate": 0.012766088260612334, + "loss": 0.3281, + "num_input_tokens_seen": 72506976, + "step": 34710 + }, + { + "epoch": 5.663186230524513, + "grad_norm": 0.0008148052729666233, + "learning_rate": 0.012742319126827523, + "loss": 0.3176, + "num_input_tokens_seen": 72517504, + "step": 34715 + }, + { + "epoch": 5.66400195774533, + "grad_norm": 0.001291960827074945, + "learning_rate": 0.012718571159902008, + "loss": 0.3841, + "num_input_tokens_seen": 72527792, + "step": 34720 + }, + { + "epoch": 5.664817684966147, + "grad_norm": 0.0007176150684244931, + "learning_rate": 0.01269484436349803, + "loss": 0.2847, + "num_input_tokens_seen": 72538128, + "step": 34725 + }, + { + "epoch": 5.665633412186965, + "grad_norm": 0.0007276099058799446, + "learning_rate": 0.012671138741274528, + "loss": 0.3294, + "num_input_tokens_seen": 72548000, + "step": 34730 + }, + { + "epoch": 5.666449139407782, + "grad_norm": 0.000603605411015451, + "learning_rate": 0.012647454296887194, + "loss": 0.2552, + "num_input_tokens_seen": 72558096, + "step": 34735 + }, + { + "epoch": 5.667264866628599, + "grad_norm": 0.0007073305314406753, + "learning_rate": 0.012623791033988507, + "loss": 0.375, + "num_input_tokens_seen": 72568112, + "step": 34740 + }, + { + "epoch": 5.6680805938494165, + "grad_norm": 0.0004807489749509841, + "learning_rate": 0.012600148956227597, + "loss": 0.2666, + "num_input_tokens_seen": 72578112, + "step": 34745 + }, + { + "epoch": 5.668896321070234, + "grad_norm": 0.0003993730351794511, + "learning_rate": 0.012576528067250414, + "loss": 0.281, + "num_input_tokens_seen": 72588816, + "step": 34750 + }, + { + "epoch": 5.669712048291052, + "grad_norm": 0.0011174515821039677, + "learning_rate": 0.012552928370699561, + "loss": 0.2993, + "num_input_tokens_seen": 72598864, + "step": 34755 + }, + { + "epoch": 5.670527775511869, + "grad_norm": 0.0009788552997633815, + "learning_rate": 0.012529349870214411, + "loss": 0.3494, + "num_input_tokens_seen": 72608976, + "step": 34760 + }, + { + "epoch": 5.671343502732686, + "grad_norm": 0.0005955533124506474, + "learning_rate": 0.012505792569431106, + "loss": 0.3163, + "num_input_tokens_seen": 72620032, + "step": 34765 + }, + { + "epoch": 5.672159229953504, + "grad_norm": 0.0006619886844418943, + "learning_rate": 0.012482256471982422, + "loss": 0.2752, + "num_input_tokens_seen": 72630608, + "step": 34770 + }, + { + "epoch": 5.672974957174321, + "grad_norm": 0.0006165800150483847, + "learning_rate": 0.012458741581497956, + "loss": 0.3193, + "num_input_tokens_seen": 72642416, + "step": 34775 + }, + { + "epoch": 5.673790684395138, + "grad_norm": 0.0015325681306421757, + "learning_rate": 0.012435247901603974, + "loss": 0.2952, + "num_input_tokens_seen": 72653840, + "step": 34780 + }, + { + "epoch": 5.674606411615955, + "grad_norm": 0.0004267682961653918, + "learning_rate": 0.012411775435923528, + "loss": 0.3363, + "num_input_tokens_seen": 72663408, + "step": 34785 + }, + { + "epoch": 5.6754221388367725, + "grad_norm": 0.0006211249274201691, + "learning_rate": 0.012388324188076354, + "loss": 0.2817, + "num_input_tokens_seen": 72673824, + "step": 34790 + }, + { + "epoch": 5.6762378660575905, + "grad_norm": 0.0007444993243552744, + "learning_rate": 0.012364894161678913, + "loss": 0.2999, + "num_input_tokens_seen": 72684000, + "step": 34795 + }, + { + "epoch": 5.677053593278408, + "grad_norm": 0.00204839906655252, + "learning_rate": 0.012341485360344445, + "loss": 0.3394, + "num_input_tokens_seen": 72694032, + "step": 34800 + }, + { + "epoch": 5.677053593278408, + "eval_loss": 0.31490325927734375, + "eval_runtime": 155.7767, + "eval_samples_per_second": 17.493, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 72694032, + "step": 34800 + }, + { + "epoch": 5.677869320499225, + "grad_norm": 0.000701370183378458, + "learning_rate": 0.01231809778768283, + "loss": 0.3212, + "num_input_tokens_seen": 72704208, + "step": 34805 + }, + { + "epoch": 5.678685047720043, + "grad_norm": 0.00038536672946065664, + "learning_rate": 0.012294731447300799, + "loss": 0.3363, + "num_input_tokens_seen": 72717472, + "step": 34810 + }, + { + "epoch": 5.67950077494086, + "grad_norm": 0.000855572463478893, + "learning_rate": 0.012271386342801671, + "loss": 0.3337, + "num_input_tokens_seen": 72728048, + "step": 34815 + }, + { + "epoch": 5.680316502161677, + "grad_norm": 0.0008921264670789242, + "learning_rate": 0.012248062477785565, + "loss": 0.2661, + "num_input_tokens_seen": 72737696, + "step": 34820 + }, + { + "epoch": 5.681132229382494, + "grad_norm": 0.0006008072523400187, + "learning_rate": 0.012224759855849305, + "loss": 0.2513, + "num_input_tokens_seen": 72748304, + "step": 34825 + }, + { + "epoch": 5.681947956603312, + "grad_norm": 0.0007353870896622539, + "learning_rate": 0.012201478480586513, + "loss": 0.405, + "num_input_tokens_seen": 72757024, + "step": 34830 + }, + { + "epoch": 5.682763683824129, + "grad_norm": 0.0009005352621898055, + "learning_rate": 0.012178218355587389, + "loss": 0.3456, + "num_input_tokens_seen": 72768320, + "step": 34835 + }, + { + "epoch": 5.6835794110449465, + "grad_norm": 0.0005363347008824348, + "learning_rate": 0.01215497948443896, + "loss": 0.2586, + "num_input_tokens_seen": 72778656, + "step": 34840 + }, + { + "epoch": 5.684395138265764, + "grad_norm": 0.0007214929792098701, + "learning_rate": 0.012131761870724993, + "loss": 0.3064, + "num_input_tokens_seen": 72788544, + "step": 34845 + }, + { + "epoch": 5.685210865486582, + "grad_norm": 0.0005988184711895883, + "learning_rate": 0.012108565518025893, + "loss": 0.3339, + "num_input_tokens_seen": 72798352, + "step": 34850 + }, + { + "epoch": 5.686026592707399, + "grad_norm": 0.0007499027415178716, + "learning_rate": 0.012085390429918862, + "loss": 0.3179, + "num_input_tokens_seen": 72809360, + "step": 34855 + }, + { + "epoch": 5.686842319928216, + "grad_norm": 0.00045106615289114416, + "learning_rate": 0.012062236609977744, + "loss": 0.347, + "num_input_tokens_seen": 72820496, + "step": 34860 + }, + { + "epoch": 5.687658047149033, + "grad_norm": 0.000651970854960382, + "learning_rate": 0.01203910406177318, + "loss": 0.3118, + "num_input_tokens_seen": 72830000, + "step": 34865 + }, + { + "epoch": 5.688473774369851, + "grad_norm": 0.0004732600355055183, + "learning_rate": 0.01201599278887252, + "loss": 0.3648, + "num_input_tokens_seen": 72841584, + "step": 34870 + }, + { + "epoch": 5.689289501590668, + "grad_norm": 0.0004402030317578465, + "learning_rate": 0.011992902794839744, + "loss": 0.325, + "num_input_tokens_seen": 72853072, + "step": 34875 + }, + { + "epoch": 5.690105228811485, + "grad_norm": 0.0005827402928844094, + "learning_rate": 0.011969834083235703, + "loss": 0.3657, + "num_input_tokens_seen": 72863376, + "step": 34880 + }, + { + "epoch": 5.690920956032302, + "grad_norm": 0.0005634665139950812, + "learning_rate": 0.011946786657617836, + "loss": 0.3005, + "num_input_tokens_seen": 72875280, + "step": 34885 + }, + { + "epoch": 5.69173668325312, + "grad_norm": 0.000587351736612618, + "learning_rate": 0.011923760521540332, + "loss": 0.3023, + "num_input_tokens_seen": 72884912, + "step": 34890 + }, + { + "epoch": 5.692552410473938, + "grad_norm": 0.0005539737176150084, + "learning_rate": 0.011900755678554153, + "loss": 0.3414, + "num_input_tokens_seen": 72895264, + "step": 34895 + }, + { + "epoch": 5.693368137694755, + "grad_norm": 0.0005999650456942618, + "learning_rate": 0.011877772132206893, + "loss": 0.2285, + "num_input_tokens_seen": 72906160, + "step": 34900 + }, + { + "epoch": 5.694183864915573, + "grad_norm": 0.000760076567530632, + "learning_rate": 0.011854809886042915, + "loss": 0.2482, + "num_input_tokens_seen": 72918880, + "step": 34905 + }, + { + "epoch": 5.69499959213639, + "grad_norm": 0.0006094647687859833, + "learning_rate": 0.011831868943603325, + "loss": 0.335, + "num_input_tokens_seen": 72930736, + "step": 34910 + }, + { + "epoch": 5.695815319357207, + "grad_norm": 0.0009065186604857445, + "learning_rate": 0.011808949308425836, + "loss": 0.311, + "num_input_tokens_seen": 72941216, + "step": 34915 + }, + { + "epoch": 5.696631046578024, + "grad_norm": 0.0007282978622242808, + "learning_rate": 0.01178605098404501, + "loss": 0.2749, + "num_input_tokens_seen": 72952560, + "step": 34920 + }, + { + "epoch": 5.697446773798841, + "grad_norm": 0.0007997517241165042, + "learning_rate": 0.011763173973992002, + "loss": 0.2689, + "num_input_tokens_seen": 72962800, + "step": 34925 + }, + { + "epoch": 5.698262501019659, + "grad_norm": 0.001181241706945002, + "learning_rate": 0.011740318281794776, + "loss": 0.3798, + "num_input_tokens_seen": 72973664, + "step": 34930 + }, + { + "epoch": 5.699078228240476, + "grad_norm": 0.0004439943586476147, + "learning_rate": 0.01171748391097796, + "loss": 0.2583, + "num_input_tokens_seen": 72985504, + "step": 34935 + }, + { + "epoch": 5.6998939554612935, + "grad_norm": 0.00035715094418264925, + "learning_rate": 0.011694670865062873, + "loss": 0.3295, + "num_input_tokens_seen": 72995904, + "step": 34940 + }, + { + "epoch": 5.7007096826821115, + "grad_norm": 0.0015565187204629183, + "learning_rate": 0.011671879147567616, + "loss": 0.2821, + "num_input_tokens_seen": 73006592, + "step": 34945 + }, + { + "epoch": 5.701525409902929, + "grad_norm": 0.00044012797297909856, + "learning_rate": 0.011649108762006893, + "loss": 0.278, + "num_input_tokens_seen": 73018432, + "step": 34950 + }, + { + "epoch": 5.702341137123746, + "grad_norm": 0.000799807021394372, + "learning_rate": 0.011626359711892265, + "loss": 0.265, + "num_input_tokens_seen": 73028768, + "step": 34955 + }, + { + "epoch": 5.703156864344563, + "grad_norm": 0.000695557042490691, + "learning_rate": 0.01160363200073189, + "loss": 0.3065, + "num_input_tokens_seen": 73040016, + "step": 34960 + }, + { + "epoch": 5.70397259156538, + "grad_norm": 0.0004741507873404771, + "learning_rate": 0.011580925632030614, + "loss": 0.2767, + "num_input_tokens_seen": 73049792, + "step": 34965 + }, + { + "epoch": 5.704788318786198, + "grad_norm": 0.0008181220619007945, + "learning_rate": 0.011558240609290104, + "loss": 0.3222, + "num_input_tokens_seen": 73058704, + "step": 34970 + }, + { + "epoch": 5.705604046007015, + "grad_norm": 0.0009123493800871074, + "learning_rate": 0.011535576936008679, + "loss": 0.2951, + "num_input_tokens_seen": 73067696, + "step": 34975 + }, + { + "epoch": 5.706419773227832, + "grad_norm": 0.0007187851588241756, + "learning_rate": 0.011512934615681309, + "loss": 0.3079, + "num_input_tokens_seen": 73079440, + "step": 34980 + }, + { + "epoch": 5.70723550044865, + "grad_norm": 0.0004947743145748973, + "learning_rate": 0.011490313651799765, + "loss": 0.2894, + "num_input_tokens_seen": 73089152, + "step": 34985 + }, + { + "epoch": 5.7080512276694675, + "grad_norm": 0.0005351406871341169, + "learning_rate": 0.011467714047852512, + "loss": 0.3005, + "num_input_tokens_seen": 73100144, + "step": 34990 + }, + { + "epoch": 5.708866954890285, + "grad_norm": 0.0009195485035888851, + "learning_rate": 0.011445135807324624, + "loss": 0.3776, + "num_input_tokens_seen": 73110544, + "step": 34995 + }, + { + "epoch": 5.709682682111102, + "grad_norm": 0.0007893331930972636, + "learning_rate": 0.011422578933698002, + "loss": 0.3687, + "num_input_tokens_seen": 73119856, + "step": 35000 + }, + { + "epoch": 5.709682682111102, + "eval_loss": 0.31392788887023926, + "eval_runtime": 156.0624, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 73119856, + "step": 35000 + }, + { + "epoch": 5.71049840933192, + "grad_norm": 0.00044323207112029195, + "learning_rate": 0.011400043430451161, + "loss": 0.3128, + "num_input_tokens_seen": 73129952, + "step": 35005 + }, + { + "epoch": 5.711314136552737, + "grad_norm": 0.00043080426985397935, + "learning_rate": 0.011377529301059392, + "loss": 0.3356, + "num_input_tokens_seen": 73138528, + "step": 35010 + }, + { + "epoch": 5.712129863773554, + "grad_norm": 0.0005891550681553781, + "learning_rate": 0.011355036548994646, + "loss": 0.3305, + "num_input_tokens_seen": 73148288, + "step": 35015 + }, + { + "epoch": 5.712945590994371, + "grad_norm": 0.0006996144657023251, + "learning_rate": 0.011332565177725584, + "loss": 0.3223, + "num_input_tokens_seen": 73158304, + "step": 35020 + }, + { + "epoch": 5.713761318215189, + "grad_norm": 0.0006921231979504228, + "learning_rate": 0.011310115190717585, + "loss": 0.3502, + "num_input_tokens_seen": 73167824, + "step": 35025 + }, + { + "epoch": 5.714577045436006, + "grad_norm": 0.00036425181315280497, + "learning_rate": 0.01128768659143271, + "loss": 0.2792, + "num_input_tokens_seen": 73179744, + "step": 35030 + }, + { + "epoch": 5.715392772656823, + "grad_norm": 0.0010017883032560349, + "learning_rate": 0.011265279383329713, + "loss": 0.3191, + "num_input_tokens_seen": 73189840, + "step": 35035 + }, + { + "epoch": 5.716208499877641, + "grad_norm": 0.0006017452105879784, + "learning_rate": 0.01124289356986411, + "loss": 0.3694, + "num_input_tokens_seen": 73199648, + "step": 35040 + }, + { + "epoch": 5.717024227098459, + "grad_norm": 0.0005013193585909903, + "learning_rate": 0.011220529154488023, + "loss": 0.3326, + "num_input_tokens_seen": 73208720, + "step": 35045 + }, + { + "epoch": 5.717839954319276, + "grad_norm": 0.0006249439902603626, + "learning_rate": 0.011198186140650346, + "loss": 0.37, + "num_input_tokens_seen": 73220592, + "step": 35050 + }, + { + "epoch": 5.718655681540093, + "grad_norm": 0.0005211267271079123, + "learning_rate": 0.011175864531796685, + "loss": 0.3632, + "num_input_tokens_seen": 73230896, + "step": 35055 + }, + { + "epoch": 5.71947140876091, + "grad_norm": 0.0006034680991433561, + "learning_rate": 0.011153564331369258, + "loss": 0.3835, + "num_input_tokens_seen": 73242704, + "step": 35060 + }, + { + "epoch": 5.720287135981728, + "grad_norm": 0.0006245932308956981, + "learning_rate": 0.011131285542807078, + "loss": 0.2864, + "num_input_tokens_seen": 73252992, + "step": 35065 + }, + { + "epoch": 5.721102863202545, + "grad_norm": 0.0006316693616099656, + "learning_rate": 0.011109028169545815, + "loss": 0.2806, + "num_input_tokens_seen": 73263776, + "step": 35070 + }, + { + "epoch": 5.721918590423362, + "grad_norm": 0.0006025254842825234, + "learning_rate": 0.011086792215017804, + "loss": 0.3016, + "num_input_tokens_seen": 73275184, + "step": 35075 + }, + { + "epoch": 5.72273431764418, + "grad_norm": 0.0004460051713977009, + "learning_rate": 0.011064577682652137, + "loss": 0.2701, + "num_input_tokens_seen": 73285600, + "step": 35080 + }, + { + "epoch": 5.723550044864997, + "grad_norm": 0.0012533703120425344, + "learning_rate": 0.011042384575874559, + "loss": 0.3476, + "num_input_tokens_seen": 73294976, + "step": 35085 + }, + { + "epoch": 5.7243657720858145, + "grad_norm": 0.00040815307875163853, + "learning_rate": 0.011020212898107512, + "loss": 0.297, + "num_input_tokens_seen": 73306224, + "step": 35090 + }, + { + "epoch": 5.725181499306632, + "grad_norm": 0.0004438407195266336, + "learning_rate": 0.010998062652770197, + "loss": 0.3125, + "num_input_tokens_seen": 73316448, + "step": 35095 + }, + { + "epoch": 5.725997226527449, + "grad_norm": 0.0003903878678102046, + "learning_rate": 0.010975933843278428, + "loss": 0.3153, + "num_input_tokens_seen": 73327552, + "step": 35100 + }, + { + "epoch": 5.726812953748267, + "grad_norm": 0.0008084762957878411, + "learning_rate": 0.010953826473044714, + "loss": 0.3213, + "num_input_tokens_seen": 73338384, + "step": 35105 + }, + { + "epoch": 5.727628680969084, + "grad_norm": 0.0005067907040938735, + "learning_rate": 0.010931740545478357, + "loss": 0.3231, + "num_input_tokens_seen": 73349216, + "step": 35110 + }, + { + "epoch": 5.728444408189901, + "grad_norm": 0.0007244272856041789, + "learning_rate": 0.010909676063985218, + "loss": 0.2931, + "num_input_tokens_seen": 73359792, + "step": 35115 + }, + { + "epoch": 5.729260135410719, + "grad_norm": 0.0006199191557243466, + "learning_rate": 0.010887633031967974, + "loss": 0.3853, + "num_input_tokens_seen": 73370544, + "step": 35120 + }, + { + "epoch": 5.730075862631536, + "grad_norm": 0.0010711773065850139, + "learning_rate": 0.01086561145282589, + "loss": 0.3233, + "num_input_tokens_seen": 73379712, + "step": 35125 + }, + { + "epoch": 5.730891589852353, + "grad_norm": 0.000852659170050174, + "learning_rate": 0.010843611329954983, + "loss": 0.3304, + "num_input_tokens_seen": 73390752, + "step": 35130 + }, + { + "epoch": 5.7317073170731705, + "grad_norm": 0.0008603577734902501, + "learning_rate": 0.010821632666747988, + "loss": 0.2704, + "num_input_tokens_seen": 73400832, + "step": 35135 + }, + { + "epoch": 5.732523044293988, + "grad_norm": 0.0005857625510543585, + "learning_rate": 0.010799675466594244, + "loss": 0.3816, + "num_input_tokens_seen": 73412800, + "step": 35140 + }, + { + "epoch": 5.733338771514806, + "grad_norm": 0.0010333063546568155, + "learning_rate": 0.010777739732879826, + "loss": 0.3641, + "num_input_tokens_seen": 73423920, + "step": 35145 + }, + { + "epoch": 5.734154498735623, + "grad_norm": 0.000755393470171839, + "learning_rate": 0.010755825468987562, + "loss": 0.3168, + "num_input_tokens_seen": 73435200, + "step": 35150 + }, + { + "epoch": 5.73497022595644, + "grad_norm": 0.00045553813106380403, + "learning_rate": 0.010733932678296814, + "loss": 0.2758, + "num_input_tokens_seen": 73447296, + "step": 35155 + }, + { + "epoch": 5.735785953177258, + "grad_norm": 0.0008838336798362434, + "learning_rate": 0.010712061364183817, + "loss": 0.2851, + "num_input_tokens_seen": 73458016, + "step": 35160 + }, + { + "epoch": 5.736601680398075, + "grad_norm": 0.0016480702906847, + "learning_rate": 0.010690211530021337, + "loss": 0.3035, + "num_input_tokens_seen": 73468784, + "step": 35165 + }, + { + "epoch": 5.737417407618892, + "grad_norm": 0.00038990008761174977, + "learning_rate": 0.01066838317917893, + "loss": 0.3323, + "num_input_tokens_seen": 73479536, + "step": 35170 + }, + { + "epoch": 5.738233134839709, + "grad_norm": 0.00047980251838453114, + "learning_rate": 0.010646576315022787, + "loss": 0.3127, + "num_input_tokens_seen": 73489024, + "step": 35175 + }, + { + "epoch": 5.739048862060527, + "grad_norm": 0.0004746770719066262, + "learning_rate": 0.010624790940915785, + "loss": 0.3362, + "num_input_tokens_seen": 73499216, + "step": 35180 + }, + { + "epoch": 5.7398645892813445, + "grad_norm": 0.0007148345466703176, + "learning_rate": 0.0106030270602175, + "loss": 0.3712, + "num_input_tokens_seen": 73509232, + "step": 35185 + }, + { + "epoch": 5.740680316502162, + "grad_norm": 0.0004915815661661327, + "learning_rate": 0.010581284676284252, + "loss": 0.3223, + "num_input_tokens_seen": 73518752, + "step": 35190 + }, + { + "epoch": 5.741496043722979, + "grad_norm": 0.0006988761015236378, + "learning_rate": 0.010559563792468923, + "loss": 0.3177, + "num_input_tokens_seen": 73527936, + "step": 35195 + }, + { + "epoch": 5.742311770943797, + "grad_norm": 0.0003559200558811426, + "learning_rate": 0.010537864412121217, + "loss": 0.33, + "num_input_tokens_seen": 73537984, + "step": 35200 + }, + { + "epoch": 5.742311770943797, + "eval_loss": 0.31461116671562195, + "eval_runtime": 155.7693, + "eval_samples_per_second": 17.494, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 73537984, + "step": 35200 + }, + { + "epoch": 5.743127498164614, + "grad_norm": 0.00048327946569770575, + "learning_rate": 0.010516186538587357, + "loss": 0.2746, + "num_input_tokens_seen": 73548880, + "step": 35205 + }, + { + "epoch": 5.743943225385431, + "grad_norm": 0.0005227135261520743, + "learning_rate": 0.01049453017521042, + "loss": 0.3574, + "num_input_tokens_seen": 73560624, + "step": 35210 + }, + { + "epoch": 5.744758952606248, + "grad_norm": 0.00048245396465063095, + "learning_rate": 0.010472895325330083, + "loss": 0.3424, + "num_input_tokens_seen": 73571536, + "step": 35215 + }, + { + "epoch": 5.745574679827066, + "grad_norm": 0.0005335484747774899, + "learning_rate": 0.010451281992282662, + "loss": 0.2997, + "num_input_tokens_seen": 73582240, + "step": 35220 + }, + { + "epoch": 5.746390407047883, + "grad_norm": 0.00045093096559867263, + "learning_rate": 0.01042969017940124, + "loss": 0.2824, + "num_input_tokens_seen": 73593120, + "step": 35225 + }, + { + "epoch": 5.7472061342687, + "grad_norm": 0.0007320239092223346, + "learning_rate": 0.01040811989001557, + "loss": 0.3001, + "num_input_tokens_seen": 73603680, + "step": 35230 + }, + { + "epoch": 5.7480218614895175, + "grad_norm": 0.00036370402085594833, + "learning_rate": 0.010386571127451992, + "loss": 0.2787, + "num_input_tokens_seen": 73615024, + "step": 35235 + }, + { + "epoch": 5.748837588710336, + "grad_norm": 0.0020547197200357914, + "learning_rate": 0.010365043895033682, + "loss": 0.4403, + "num_input_tokens_seen": 73625232, + "step": 35240 + }, + { + "epoch": 5.749653315931153, + "grad_norm": 0.0005299302865751088, + "learning_rate": 0.010343538196080365, + "loss": 0.3269, + "num_input_tokens_seen": 73636096, + "step": 35245 + }, + { + "epoch": 5.75046904315197, + "grad_norm": 0.0005352839943952858, + "learning_rate": 0.010322054033908457, + "loss": 0.3401, + "num_input_tokens_seen": 73645248, + "step": 35250 + }, + { + "epoch": 5.751284770372787, + "grad_norm": 0.0006306275608949363, + "learning_rate": 0.010300591411831156, + "loss": 0.2967, + "num_input_tokens_seen": 73656912, + "step": 35255 + }, + { + "epoch": 5.752100497593605, + "grad_norm": 0.0015247655101120472, + "learning_rate": 0.010279150333158198, + "loss": 0.3702, + "num_input_tokens_seen": 73667040, + "step": 35260 + }, + { + "epoch": 5.752916224814422, + "grad_norm": 0.000683815567754209, + "learning_rate": 0.010257730801196107, + "loss": 0.3141, + "num_input_tokens_seen": 73677072, + "step": 35265 + }, + { + "epoch": 5.753731952035239, + "grad_norm": 0.0007589667220599949, + "learning_rate": 0.010236332819248056, + "loss": 0.2989, + "num_input_tokens_seen": 73686224, + "step": 35270 + }, + { + "epoch": 5.754547679256056, + "grad_norm": 0.000684393453411758, + "learning_rate": 0.010214956390613854, + "loss": 0.3142, + "num_input_tokens_seen": 73697168, + "step": 35275 + }, + { + "epoch": 5.755363406476874, + "grad_norm": 0.0007642772397957742, + "learning_rate": 0.010193601518590034, + "loss": 0.3066, + "num_input_tokens_seen": 73707232, + "step": 35280 + }, + { + "epoch": 5.7561791336976915, + "grad_norm": 0.0005337047041393816, + "learning_rate": 0.010172268206469758, + "loss": 0.2779, + "num_input_tokens_seen": 73718160, + "step": 35285 + }, + { + "epoch": 5.756994860918509, + "grad_norm": 0.0006109747919254005, + "learning_rate": 0.010150956457542897, + "loss": 0.2797, + "num_input_tokens_seen": 73729648, + "step": 35290 + }, + { + "epoch": 5.757810588139327, + "grad_norm": 0.0006805998273193836, + "learning_rate": 0.010129666275096054, + "loss": 0.2463, + "num_input_tokens_seen": 73739568, + "step": 35295 + }, + { + "epoch": 5.758626315360144, + "grad_norm": 0.0007953579188324511, + "learning_rate": 0.010108397662412338, + "loss": 0.3792, + "num_input_tokens_seen": 73749776, + "step": 35300 + }, + { + "epoch": 5.759442042580961, + "grad_norm": 0.0007511694566346705, + "learning_rate": 0.010087150622771707, + "loss": 0.3094, + "num_input_tokens_seen": 73759600, + "step": 35305 + }, + { + "epoch": 5.760257769801778, + "grad_norm": 0.0004771939420606941, + "learning_rate": 0.010065925159450739, + "loss": 0.291, + "num_input_tokens_seen": 73768608, + "step": 35310 + }, + { + "epoch": 5.761073497022595, + "grad_norm": 0.000560948858037591, + "learning_rate": 0.010044721275722618, + "loss": 0.4421, + "num_input_tokens_seen": 73779168, + "step": 35315 + }, + { + "epoch": 5.761889224243413, + "grad_norm": 0.0005263272323645651, + "learning_rate": 0.01002353897485726, + "loss": 0.2737, + "num_input_tokens_seen": 73789184, + "step": 35320 + }, + { + "epoch": 5.76270495146423, + "grad_norm": 0.000756799359805882, + "learning_rate": 0.010002378260121236, + "loss": 0.3375, + "num_input_tokens_seen": 73800160, + "step": 35325 + }, + { + "epoch": 5.7635206786850475, + "grad_norm": 0.0009304005652666092, + "learning_rate": 0.009981239134777786, + "loss": 0.3103, + "num_input_tokens_seen": 73809072, + "step": 35330 + }, + { + "epoch": 5.7643364059058655, + "grad_norm": 0.0005213677068240941, + "learning_rate": 0.009960121602086884, + "loss": 0.3141, + "num_input_tokens_seen": 73819504, + "step": 35335 + }, + { + "epoch": 5.765152133126683, + "grad_norm": 0.0007873069844208658, + "learning_rate": 0.009939025665305062, + "loss": 0.3814, + "num_input_tokens_seen": 73829472, + "step": 35340 + }, + { + "epoch": 5.7659678603475, + "grad_norm": 0.0008556140237487853, + "learning_rate": 0.009917951327685597, + "loss": 0.3561, + "num_input_tokens_seen": 73840576, + "step": 35345 + }, + { + "epoch": 5.766783587568317, + "grad_norm": 0.000687521300278604, + "learning_rate": 0.009896898592478425, + "loss": 0.4021, + "num_input_tokens_seen": 73850752, + "step": 35350 + }, + { + "epoch": 5.767599314789135, + "grad_norm": 0.0006791619234718382, + "learning_rate": 0.009875867462930132, + "loss": 0.3441, + "num_input_tokens_seen": 73859776, + "step": 35355 + }, + { + "epoch": 5.768415042009952, + "grad_norm": 0.0005532680079340935, + "learning_rate": 0.009854857942284006, + "loss": 0.3572, + "num_input_tokens_seen": 73869104, + "step": 35360 + }, + { + "epoch": 5.769230769230769, + "grad_norm": 0.00046032798127271235, + "learning_rate": 0.009833870033779923, + "loss": 0.3465, + "num_input_tokens_seen": 73880352, + "step": 35365 + }, + { + "epoch": 5.770046496451586, + "grad_norm": 0.0009044627659022808, + "learning_rate": 0.009812903740654527, + "loss": 0.3248, + "num_input_tokens_seen": 73891024, + "step": 35370 + }, + { + "epoch": 5.770862223672404, + "grad_norm": 0.0005751281860284507, + "learning_rate": 0.009791959066141097, + "loss": 0.3324, + "num_input_tokens_seen": 73902336, + "step": 35375 + }, + { + "epoch": 5.771677950893221, + "grad_norm": 0.0006692697061225772, + "learning_rate": 0.009771036013469537, + "loss": 0.3246, + "num_input_tokens_seen": 73912832, + "step": 35380 + }, + { + "epoch": 5.772493678114039, + "grad_norm": 0.0005851921741850674, + "learning_rate": 0.00975013458586646, + "loss": 0.3081, + "num_input_tokens_seen": 73924496, + "step": 35385 + }, + { + "epoch": 5.773309405334856, + "grad_norm": 0.00036544035538099706, + "learning_rate": 0.009729254786555107, + "loss": 0.3191, + "num_input_tokens_seen": 73933136, + "step": 35390 + }, + { + "epoch": 5.774125132555674, + "grad_norm": 0.00044743347098119557, + "learning_rate": 0.009708396618755421, + "loss": 0.3085, + "num_input_tokens_seen": 73944048, + "step": 35395 + }, + { + "epoch": 5.774940859776491, + "grad_norm": 0.0005911492044106126, + "learning_rate": 0.009687560085683994, + "loss": 0.314, + "num_input_tokens_seen": 73955216, + "step": 35400 + }, + { + "epoch": 5.774940859776491, + "eval_loss": 0.315165251493454, + "eval_runtime": 156.0559, + "eval_samples_per_second": 17.462, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 73955216, + "step": 35400 + }, + { + "epoch": 5.775756586997308, + "grad_norm": 0.0004694922245107591, + "learning_rate": 0.009666745190554054, + "loss": 0.3009, + "num_input_tokens_seen": 73965456, + "step": 35405 + }, + { + "epoch": 5.776572314218125, + "grad_norm": 0.000453125307103619, + "learning_rate": 0.009645951936575553, + "loss": 0.2688, + "num_input_tokens_seen": 73976192, + "step": 35410 + }, + { + "epoch": 5.777388041438943, + "grad_norm": 0.0005166502087377012, + "learning_rate": 0.00962518032695509, + "loss": 0.3424, + "num_input_tokens_seen": 73985856, + "step": 35415 + }, + { + "epoch": 5.77820376865976, + "grad_norm": 0.0007497056503780186, + "learning_rate": 0.009604430364895855, + "loss": 0.3347, + "num_input_tokens_seen": 73996528, + "step": 35420 + }, + { + "epoch": 5.779019495880577, + "grad_norm": 0.0006654361495748162, + "learning_rate": 0.00958370205359777, + "loss": 0.2688, + "num_input_tokens_seen": 74005680, + "step": 35425 + }, + { + "epoch": 5.7798352231013945, + "grad_norm": 0.00065552385058254, + "learning_rate": 0.009562995396257445, + "loss": 0.2599, + "num_input_tokens_seen": 74016384, + "step": 35430 + }, + { + "epoch": 5.7806509503222125, + "grad_norm": 0.00039933828520588577, + "learning_rate": 0.009542310396068026, + "loss": 0.3485, + "num_input_tokens_seen": 74026704, + "step": 35435 + }, + { + "epoch": 5.78146667754303, + "grad_norm": 0.0014191524824127555, + "learning_rate": 0.009521647056219495, + "loss": 0.4326, + "num_input_tokens_seen": 74036544, + "step": 35440 + }, + { + "epoch": 5.782282404763847, + "grad_norm": 0.0007109034922905266, + "learning_rate": 0.00950100537989832, + "loss": 0.342, + "num_input_tokens_seen": 74047072, + "step": 35445 + }, + { + "epoch": 5.783098131984664, + "grad_norm": 0.0005678036250174046, + "learning_rate": 0.00948038537028772, + "loss": 0.2763, + "num_input_tokens_seen": 74058192, + "step": 35450 + }, + { + "epoch": 5.783913859205482, + "grad_norm": 0.0005505996523424983, + "learning_rate": 0.009459787030567617, + "loss": 0.3305, + "num_input_tokens_seen": 74068992, + "step": 35455 + }, + { + "epoch": 5.784729586426299, + "grad_norm": 0.0005255833966657519, + "learning_rate": 0.00943921036391449, + "loss": 0.3105, + "num_input_tokens_seen": 74080048, + "step": 35460 + }, + { + "epoch": 5.785545313647116, + "grad_norm": 0.0004331892414484173, + "learning_rate": 0.009418655373501483, + "loss": 0.3271, + "num_input_tokens_seen": 74090960, + "step": 35465 + }, + { + "epoch": 5.786361040867934, + "grad_norm": 0.0007969856960698962, + "learning_rate": 0.00939812206249851, + "loss": 0.3227, + "num_input_tokens_seen": 74101232, + "step": 35470 + }, + { + "epoch": 5.787176768088751, + "grad_norm": 0.0005381247028708458, + "learning_rate": 0.009377610434072004, + "loss": 0.2947, + "num_input_tokens_seen": 74111344, + "step": 35475 + }, + { + "epoch": 5.7879924953095685, + "grad_norm": 0.0005397476488724351, + "learning_rate": 0.009357120491385167, + "loss": 0.3324, + "num_input_tokens_seen": 74121104, + "step": 35480 + }, + { + "epoch": 5.788808222530386, + "grad_norm": 0.0005089764017611742, + "learning_rate": 0.009336652237597743, + "loss": 0.362, + "num_input_tokens_seen": 74131136, + "step": 35485 + }, + { + "epoch": 5.789623949751203, + "grad_norm": 0.0005364279495552182, + "learning_rate": 0.009316205675866251, + "loss": 0.2972, + "num_input_tokens_seen": 74141280, + "step": 35490 + }, + { + "epoch": 5.790439676972021, + "grad_norm": 0.00046738781384192407, + "learning_rate": 0.00929578080934379, + "loss": 0.3335, + "num_input_tokens_seen": 74153408, + "step": 35495 + }, + { + "epoch": 5.791255404192838, + "grad_norm": 0.0004679044068325311, + "learning_rate": 0.00927537764118012, + "loss": 0.2494, + "num_input_tokens_seen": 74165248, + "step": 35500 + }, + { + "epoch": 5.792071131413655, + "grad_norm": 0.00048771806177683175, + "learning_rate": 0.009254996174521678, + "loss": 0.3228, + "num_input_tokens_seen": 74175312, + "step": 35505 + }, + { + "epoch": 5.792886858634473, + "grad_norm": 0.0011964656878262758, + "learning_rate": 0.009234636412511531, + "loss": 0.3504, + "num_input_tokens_seen": 74186016, + "step": 35510 + }, + { + "epoch": 5.79370258585529, + "grad_norm": 0.0006700964295305312, + "learning_rate": 0.009214298358289418, + "loss": 0.3624, + "num_input_tokens_seen": 74196992, + "step": 35515 + }, + { + "epoch": 5.794518313076107, + "grad_norm": 0.0004685766762122512, + "learning_rate": 0.00919398201499173, + "loss": 0.3088, + "num_input_tokens_seen": 74207792, + "step": 35520 + }, + { + "epoch": 5.7953340402969244, + "grad_norm": 0.0012603967916220427, + "learning_rate": 0.009173687385751495, + "loss": 0.3365, + "num_input_tokens_seen": 74217008, + "step": 35525 + }, + { + "epoch": 5.796149767517742, + "grad_norm": 0.0004406542284414172, + "learning_rate": 0.009153414473698407, + "loss": 0.2765, + "num_input_tokens_seen": 74227120, + "step": 35530 + }, + { + "epoch": 5.79696549473856, + "grad_norm": 0.0006617713370360434, + "learning_rate": 0.009133163281958784, + "loss": 0.3526, + "num_input_tokens_seen": 74238304, + "step": 35535 + }, + { + "epoch": 5.797781221959377, + "grad_norm": 0.0006860520225018263, + "learning_rate": 0.009112933813655627, + "loss": 0.3206, + "num_input_tokens_seen": 74249136, + "step": 35540 + }, + { + "epoch": 5.798596949180194, + "grad_norm": 0.0005877961521036923, + "learning_rate": 0.009092726071908573, + "loss": 0.2205, + "num_input_tokens_seen": 74258720, + "step": 35545 + }, + { + "epoch": 5.799412676401012, + "grad_norm": 0.0008590991492383182, + "learning_rate": 0.0090725400598339, + "loss": 0.3437, + "num_input_tokens_seen": 74269584, + "step": 35550 + }, + { + "epoch": 5.800228403621829, + "grad_norm": 0.0005133010563440621, + "learning_rate": 0.009052375780544563, + "loss": 0.3446, + "num_input_tokens_seen": 74279088, + "step": 35555 + }, + { + "epoch": 5.801044130842646, + "grad_norm": 0.0005700091132894158, + "learning_rate": 0.009032233237150144, + "loss": 0.2583, + "num_input_tokens_seen": 74289312, + "step": 35560 + }, + { + "epoch": 5.801859858063463, + "grad_norm": 0.0004973296890966594, + "learning_rate": 0.009012112432756875, + "loss": 0.3289, + "num_input_tokens_seen": 74298640, + "step": 35565 + }, + { + "epoch": 5.802675585284281, + "grad_norm": 0.00099308998323977, + "learning_rate": 0.008992013370467605, + "loss": 0.3547, + "num_input_tokens_seen": 74308160, + "step": 35570 + }, + { + "epoch": 5.803491312505098, + "grad_norm": 0.001433575409464538, + "learning_rate": 0.008971936053381924, + "loss": 0.3554, + "num_input_tokens_seen": 74319088, + "step": 35575 + }, + { + "epoch": 5.8043070397259156, + "grad_norm": 0.000512292783241719, + "learning_rate": 0.008951880484595953, + "loss": 0.312, + "num_input_tokens_seen": 74329712, + "step": 35580 + }, + { + "epoch": 5.805122766946733, + "grad_norm": 0.0006865602917969227, + "learning_rate": 0.008931846667202552, + "loss": 0.236, + "num_input_tokens_seen": 74340640, + "step": 35585 + }, + { + "epoch": 5.805938494167551, + "grad_norm": 0.0007350603118538857, + "learning_rate": 0.008911834604291152, + "loss": 0.3387, + "num_input_tokens_seen": 74350544, + "step": 35590 + }, + { + "epoch": 5.806754221388368, + "grad_norm": 0.0005766748217865825, + "learning_rate": 0.008891844298947882, + "loss": 0.3261, + "num_input_tokens_seen": 74360256, + "step": 35595 + }, + { + "epoch": 5.807569948609185, + "grad_norm": 0.0005179001018404961, + "learning_rate": 0.008871875754255508, + "loss": 0.2967, + "num_input_tokens_seen": 74371040, + "step": 35600 + }, + { + "epoch": 5.807569948609185, + "eval_loss": 0.3164384067058563, + "eval_runtime": 156.1255, + "eval_samples_per_second": 17.454, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 74371040, + "step": 35600 + }, + { + "epoch": 5.808385675830002, + "grad_norm": 0.0011609967332333326, + "learning_rate": 0.008851928973293422, + "loss": 0.3295, + "num_input_tokens_seen": 74381664, + "step": 35605 + }, + { + "epoch": 5.80920140305082, + "grad_norm": 0.00038200317067094147, + "learning_rate": 0.00883200395913764, + "loss": 0.3179, + "num_input_tokens_seen": 74392912, + "step": 35610 + }, + { + "epoch": 5.810017130271637, + "grad_norm": 0.0004984520492143929, + "learning_rate": 0.00881210071486091, + "loss": 0.347, + "num_input_tokens_seen": 74404320, + "step": 35615 + }, + { + "epoch": 5.810832857492454, + "grad_norm": 0.00035072301398031414, + "learning_rate": 0.008792219243532505, + "loss": 0.297, + "num_input_tokens_seen": 74413904, + "step": 35620 + }, + { + "epoch": 5.8116485847132715, + "grad_norm": 0.0008546868921257555, + "learning_rate": 0.008772359548218428, + "loss": 0.3005, + "num_input_tokens_seen": 74425440, + "step": 35625 + }, + { + "epoch": 5.8124643119340895, + "grad_norm": 0.00040886743227019906, + "learning_rate": 0.008752521631981274, + "loss": 0.3163, + "num_input_tokens_seen": 74435184, + "step": 35630 + }, + { + "epoch": 5.813280039154907, + "grad_norm": 0.0012810835614800453, + "learning_rate": 0.008732705497880315, + "loss": 0.3582, + "num_input_tokens_seen": 74444448, + "step": 35635 + }, + { + "epoch": 5.814095766375724, + "grad_norm": 0.0004846003430429846, + "learning_rate": 0.008712911148971459, + "loss": 0.3567, + "num_input_tokens_seen": 74454960, + "step": 35640 + }, + { + "epoch": 5.814911493596542, + "grad_norm": 0.0005301802302710712, + "learning_rate": 0.008693138588307208, + "loss": 0.2949, + "num_input_tokens_seen": 74464624, + "step": 35645 + }, + { + "epoch": 5.815727220817359, + "grad_norm": 0.0008985087042674422, + "learning_rate": 0.008673387818936762, + "loss": 0.3273, + "num_input_tokens_seen": 74474800, + "step": 35650 + }, + { + "epoch": 5.816542948038176, + "grad_norm": 0.0018371597398072481, + "learning_rate": 0.008653658843905948, + "loss": 0.4198, + "num_input_tokens_seen": 74484224, + "step": 35655 + }, + { + "epoch": 5.817358675258993, + "grad_norm": 0.000784798467066139, + "learning_rate": 0.0086339516662572, + "loss": 0.3052, + "num_input_tokens_seen": 74493856, + "step": 35660 + }, + { + "epoch": 5.81817440247981, + "grad_norm": 0.0009203413501381874, + "learning_rate": 0.008614266289029638, + "loss": 0.2822, + "num_input_tokens_seen": 74506032, + "step": 35665 + }, + { + "epoch": 5.818990129700628, + "grad_norm": 0.0005061683477833867, + "learning_rate": 0.008594602715258965, + "loss": 0.3151, + "num_input_tokens_seen": 74515440, + "step": 35670 + }, + { + "epoch": 5.8198058569214455, + "grad_norm": 0.0004772766260430217, + "learning_rate": 0.008574960947977573, + "loss": 0.2695, + "num_input_tokens_seen": 74526096, + "step": 35675 + }, + { + "epoch": 5.820621584142263, + "grad_norm": 0.0010528747225180268, + "learning_rate": 0.008555340990214438, + "loss": 0.3492, + "num_input_tokens_seen": 74536512, + "step": 35680 + }, + { + "epoch": 5.821437311363081, + "grad_norm": 0.000720408686902374, + "learning_rate": 0.008535742844995258, + "loss": 0.2865, + "num_input_tokens_seen": 74548288, + "step": 35685 + }, + { + "epoch": 5.822253038583898, + "grad_norm": 0.0007323569152504206, + "learning_rate": 0.008516166515342266, + "loss": 0.3093, + "num_input_tokens_seen": 74558192, + "step": 35690 + }, + { + "epoch": 5.823068765804715, + "grad_norm": 0.0007031335262581706, + "learning_rate": 0.008496612004274411, + "loss": 0.257, + "num_input_tokens_seen": 74569312, + "step": 35695 + }, + { + "epoch": 5.823884493025532, + "grad_norm": 0.0004096550401300192, + "learning_rate": 0.008477079314807201, + "loss": 0.3021, + "num_input_tokens_seen": 74576928, + "step": 35700 + }, + { + "epoch": 5.824700220246349, + "grad_norm": 0.00038519990630447865, + "learning_rate": 0.008457568449952874, + "loss": 0.3587, + "num_input_tokens_seen": 74587664, + "step": 35705 + }, + { + "epoch": 5.825515947467167, + "grad_norm": 0.00039101039874367416, + "learning_rate": 0.008438079412720189, + "loss": 0.3506, + "num_input_tokens_seen": 74600128, + "step": 35710 + }, + { + "epoch": 5.826331674687984, + "grad_norm": 0.0008017680374905467, + "learning_rate": 0.00841861220611466, + "loss": 0.3443, + "num_input_tokens_seen": 74611136, + "step": 35715 + }, + { + "epoch": 5.827147401908801, + "grad_norm": 0.0011980487033724785, + "learning_rate": 0.008399166833138355, + "loss": 0.3099, + "num_input_tokens_seen": 74623008, + "step": 35720 + }, + { + "epoch": 5.8279631291296194, + "grad_norm": 0.0004982821410521865, + "learning_rate": 0.008379743296789987, + "loss": 0.3239, + "num_input_tokens_seen": 74634848, + "step": 35725 + }, + { + "epoch": 5.828778856350437, + "grad_norm": 0.0007967339479364455, + "learning_rate": 0.008360341600064896, + "loss": 0.3033, + "num_input_tokens_seen": 74644288, + "step": 35730 + }, + { + "epoch": 5.829594583571254, + "grad_norm": 0.0018949523800984025, + "learning_rate": 0.008340961745955121, + "loss": 0.3877, + "num_input_tokens_seen": 74656336, + "step": 35735 + }, + { + "epoch": 5.830410310792071, + "grad_norm": 0.0007735922699794173, + "learning_rate": 0.008321603737449224, + "loss": 0.3362, + "num_input_tokens_seen": 74667408, + "step": 35740 + }, + { + "epoch": 5.831226038012889, + "grad_norm": 0.0004524930554907769, + "learning_rate": 0.008302267577532479, + "loss": 0.269, + "num_input_tokens_seen": 74678768, + "step": 35745 + }, + { + "epoch": 5.832041765233706, + "grad_norm": 0.0010047060204669833, + "learning_rate": 0.008282953269186771, + "loss": 0.3337, + "num_input_tokens_seen": 74688944, + "step": 35750 + }, + { + "epoch": 5.832857492454523, + "grad_norm": 0.00045443823910318315, + "learning_rate": 0.008263660815390567, + "loss": 0.3768, + "num_input_tokens_seen": 74698720, + "step": 35755 + }, + { + "epoch": 5.83367321967534, + "grad_norm": 0.00044065233669243753, + "learning_rate": 0.008244390219119069, + "loss": 0.3474, + "num_input_tokens_seen": 74709936, + "step": 35760 + }, + { + "epoch": 5.834488946896158, + "grad_norm": 0.0009204763919115067, + "learning_rate": 0.008225141483343967, + "loss": 0.3429, + "num_input_tokens_seen": 74719792, + "step": 35765 + }, + { + "epoch": 5.835304674116975, + "grad_norm": 0.0005483684362843633, + "learning_rate": 0.00820591461103372, + "loss": 0.3479, + "num_input_tokens_seen": 74731040, + "step": 35770 + }, + { + "epoch": 5.8361204013377925, + "grad_norm": 0.0005283438367769122, + "learning_rate": 0.008186709605153358, + "loss": 0.2949, + "num_input_tokens_seen": 74742256, + "step": 35775 + }, + { + "epoch": 5.83693612855861, + "grad_norm": 0.0004721153527498245, + "learning_rate": 0.008167526468664492, + "loss": 0.2453, + "num_input_tokens_seen": 74754048, + "step": 35780 + }, + { + "epoch": 5.837751855779428, + "grad_norm": 0.0005103845032863319, + "learning_rate": 0.008148365204525443, + "loss": 0.3255, + "num_input_tokens_seen": 74763952, + "step": 35785 + }, + { + "epoch": 5.838567583000245, + "grad_norm": 0.0005714594735763967, + "learning_rate": 0.00812922581569106, + "loss": 0.3311, + "num_input_tokens_seen": 74772224, + "step": 35790 + }, + { + "epoch": 5.839383310221062, + "grad_norm": 0.0004182934935670346, + "learning_rate": 0.008110108305112934, + "loss": 0.3771, + "num_input_tokens_seen": 74784368, + "step": 35795 + }, + { + "epoch": 5.840199037441879, + "grad_norm": 0.0009556979057379067, + "learning_rate": 0.008091012675739223, + "loss": 0.2874, + "num_input_tokens_seen": 74795680, + "step": 35800 + }, + { + "epoch": 5.840199037441879, + "eval_loss": 0.31687024235725403, + "eval_runtime": 155.7408, + "eval_samples_per_second": 17.497, + "eval_steps_per_second": 8.752, + "num_input_tokens_seen": 74795680, + "step": 35800 + }, + { + "epoch": 5.841014764662697, + "grad_norm": 0.0007996095228008926, + "learning_rate": 0.008071938930514671, + "loss": 0.3838, + "num_input_tokens_seen": 74806624, + "step": 35805 + }, + { + "epoch": 5.841830491883514, + "grad_norm": 0.0007975493208505213, + "learning_rate": 0.008052887072380726, + "loss": 0.3012, + "num_input_tokens_seen": 74817232, + "step": 35810 + }, + { + "epoch": 5.842646219104331, + "grad_norm": 0.0005635333945974708, + "learning_rate": 0.008033857104275437, + "loss": 0.2634, + "num_input_tokens_seen": 74827904, + "step": 35815 + }, + { + "epoch": 5.8434619463251485, + "grad_norm": 0.001547626219689846, + "learning_rate": 0.008014849029133424, + "loss": 0.3863, + "num_input_tokens_seen": 74838080, + "step": 35820 + }, + { + "epoch": 5.8442776735459665, + "grad_norm": 0.0007189220632426441, + "learning_rate": 0.007995862849885975, + "loss": 0.2909, + "num_input_tokens_seen": 74847616, + "step": 35825 + }, + { + "epoch": 5.845093400766784, + "grad_norm": 0.00032633423688821495, + "learning_rate": 0.007976898569461032, + "loss": 0.2582, + "num_input_tokens_seen": 74857664, + "step": 35830 + }, + { + "epoch": 5.845909127987601, + "grad_norm": 0.0005800472572445869, + "learning_rate": 0.007957956190783088, + "loss": 0.3254, + "num_input_tokens_seen": 74867824, + "step": 35835 + }, + { + "epoch": 5.846724855208418, + "grad_norm": 0.0007627442828379571, + "learning_rate": 0.007939035716773324, + "loss": 0.2881, + "num_input_tokens_seen": 74876944, + "step": 35840 + }, + { + "epoch": 5.847540582429236, + "grad_norm": 0.0006175412563607097, + "learning_rate": 0.007920137150349487, + "loss": 0.3314, + "num_input_tokens_seen": 74887088, + "step": 35845 + }, + { + "epoch": 5.848356309650053, + "grad_norm": 0.0005459830863401294, + "learning_rate": 0.007901260494425981, + "loss": 0.2859, + "num_input_tokens_seen": 74896656, + "step": 35850 + }, + { + "epoch": 5.84917203687087, + "grad_norm": 0.0005719672772102058, + "learning_rate": 0.007882405751913861, + "loss": 0.3309, + "num_input_tokens_seen": 74906880, + "step": 35855 + }, + { + "epoch": 5.849987764091688, + "grad_norm": 0.0005574115202762187, + "learning_rate": 0.007863572925720702, + "loss": 0.3127, + "num_input_tokens_seen": 74917232, + "step": 35860 + }, + { + "epoch": 5.850803491312505, + "grad_norm": 0.0012510429369285703, + "learning_rate": 0.007844762018750827, + "loss": 0.2787, + "num_input_tokens_seen": 74926736, + "step": 35865 + }, + { + "epoch": 5.8516192185333225, + "grad_norm": 0.0005779169732704759, + "learning_rate": 0.007825973033905054, + "loss": 0.365, + "num_input_tokens_seen": 74937584, + "step": 35870 + }, + { + "epoch": 5.85243494575414, + "grad_norm": 0.0006133401184342802, + "learning_rate": 0.007807205974080927, + "loss": 0.2802, + "num_input_tokens_seen": 74947008, + "step": 35875 + }, + { + "epoch": 5.853250672974957, + "grad_norm": 0.000363765109796077, + "learning_rate": 0.007788460842172551, + "loss": 0.2943, + "num_input_tokens_seen": 74958480, + "step": 35880 + }, + { + "epoch": 5.854066400195775, + "grad_norm": 0.0007193564088083804, + "learning_rate": 0.0077697376410706285, + "loss": 0.3182, + "num_input_tokens_seen": 74967792, + "step": 35885 + }, + { + "epoch": 5.854882127416592, + "grad_norm": 0.0007013407885096967, + "learning_rate": 0.007751036373662567, + "loss": 0.2742, + "num_input_tokens_seen": 74977520, + "step": 35890 + }, + { + "epoch": 5.855697854637409, + "grad_norm": 0.0007016074378043413, + "learning_rate": 0.00773235704283231, + "loss": 0.3304, + "num_input_tokens_seen": 74988288, + "step": 35895 + }, + { + "epoch": 5.856513581858227, + "grad_norm": 0.0005344474338926375, + "learning_rate": 0.007713699651460437, + "loss": 0.2945, + "num_input_tokens_seen": 74999408, + "step": 35900 + }, + { + "epoch": 5.857329309079044, + "grad_norm": 0.0007244416628964245, + "learning_rate": 0.007695064202424162, + "loss": 0.3121, + "num_input_tokens_seen": 75008240, + "step": 35905 + }, + { + "epoch": 5.858145036299861, + "grad_norm": 0.0014298547757789493, + "learning_rate": 0.007676450698597286, + "loss": 0.3266, + "num_input_tokens_seen": 75018480, + "step": 35910 + }, + { + "epoch": 5.858960763520678, + "grad_norm": 0.0007681566057726741, + "learning_rate": 0.007657859142850265, + "loss": 0.2591, + "num_input_tokens_seen": 75029088, + "step": 35915 + }, + { + "epoch": 5.859776490741496, + "grad_norm": 0.0006457787822000682, + "learning_rate": 0.0076392895380501535, + "loss": 0.3255, + "num_input_tokens_seen": 75040048, + "step": 35920 + }, + { + "epoch": 5.8605922179623136, + "grad_norm": 0.00047364295460283756, + "learning_rate": 0.007620741887060611, + "loss": 0.2482, + "num_input_tokens_seen": 75048928, + "step": 35925 + }, + { + "epoch": 5.861407945183131, + "grad_norm": 0.0003644612734206021, + "learning_rate": 0.007602216192741901, + "loss": 0.315, + "num_input_tokens_seen": 75060000, + "step": 35930 + }, + { + "epoch": 5.862223672403948, + "grad_norm": 0.0005843281396664679, + "learning_rate": 0.007583712457950969, + "loss": 0.3018, + "num_input_tokens_seen": 75071168, + "step": 35935 + }, + { + "epoch": 5.863039399624766, + "grad_norm": 0.0004321521264500916, + "learning_rate": 0.007565230685541269, + "loss": 0.3863, + "num_input_tokens_seen": 75082272, + "step": 35940 + }, + { + "epoch": 5.863855126845583, + "grad_norm": 0.0011029160814359784, + "learning_rate": 0.007546770878362968, + "loss": 0.3144, + "num_input_tokens_seen": 75094512, + "step": 35945 + }, + { + "epoch": 5.8646708540664, + "grad_norm": 0.0004284303868189454, + "learning_rate": 0.0075283330392627405, + "loss": 0.3568, + "num_input_tokens_seen": 75105952, + "step": 35950 + }, + { + "epoch": 5.865486581287217, + "grad_norm": 0.0005690768593922257, + "learning_rate": 0.007509917171083979, + "loss": 0.3548, + "num_input_tokens_seen": 75115920, + "step": 35955 + }, + { + "epoch": 5.866302308508035, + "grad_norm": 0.0007531496812589467, + "learning_rate": 0.007491523276666662, + "loss": 0.3387, + "num_input_tokens_seen": 75126304, + "step": 35960 + }, + { + "epoch": 5.867118035728852, + "grad_norm": 0.000532948353793472, + "learning_rate": 0.007473151358847318, + "loss": 0.3479, + "num_input_tokens_seen": 75136544, + "step": 35965 + }, + { + "epoch": 5.8679337629496695, + "grad_norm": 0.0008182238088920712, + "learning_rate": 0.007454801420459117, + "loss": 0.3102, + "num_input_tokens_seen": 75147392, + "step": 35970 + }, + { + "epoch": 5.868749490170487, + "grad_norm": 0.0006535073043778539, + "learning_rate": 0.0074364734643319105, + "loss": 0.2947, + "num_input_tokens_seen": 75157632, + "step": 35975 + }, + { + "epoch": 5.869565217391305, + "grad_norm": 0.0004202075651846826, + "learning_rate": 0.007418167493292022, + "loss": 0.3122, + "num_input_tokens_seen": 75168016, + "step": 35980 + }, + { + "epoch": 5.870380944612122, + "grad_norm": 0.0005148935015313327, + "learning_rate": 0.0073998835101625245, + "loss": 0.2639, + "num_input_tokens_seen": 75178096, + "step": 35985 + }, + { + "epoch": 5.871196671832939, + "grad_norm": 0.0007136096246540546, + "learning_rate": 0.007381621517762998, + "loss": 0.2614, + "num_input_tokens_seen": 75188704, + "step": 35990 + }, + { + "epoch": 5.872012399053756, + "grad_norm": 0.0005651957471854985, + "learning_rate": 0.007363381518909689, + "loss": 0.2888, + "num_input_tokens_seen": 75200320, + "step": 35995 + }, + { + "epoch": 5.872828126274574, + "grad_norm": 0.0008657586295157671, + "learning_rate": 0.007345163516415448, + "loss": 0.2811, + "num_input_tokens_seen": 75209824, + "step": 36000 + }, + { + "epoch": 5.872828126274574, + "eval_loss": 0.31587091088294983, + "eval_runtime": 155.7764, + "eval_samples_per_second": 17.493, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 75209824, + "step": 36000 + }, + { + "epoch": 5.873643853495391, + "grad_norm": 0.0006036505801603198, + "learning_rate": 0.007326967513089693, + "loss": 0.2972, + "num_input_tokens_seen": 75220640, + "step": 36005 + }, + { + "epoch": 5.874459580716208, + "grad_norm": 0.0004228803445585072, + "learning_rate": 0.0073087935117384815, + "loss": 0.3182, + "num_input_tokens_seen": 75231136, + "step": 36010 + }, + { + "epoch": 5.8752753079370255, + "grad_norm": 0.0005340758361853659, + "learning_rate": 0.007290641515164503, + "loss": 0.391, + "num_input_tokens_seen": 75243488, + "step": 36015 + }, + { + "epoch": 5.8760910351578435, + "grad_norm": 0.0008026442374102771, + "learning_rate": 0.007272511526166986, + "loss": 0.274, + "num_input_tokens_seen": 75255216, + "step": 36020 + }, + { + "epoch": 5.876906762378661, + "grad_norm": 0.0004201968258712441, + "learning_rate": 0.0072544035475418265, + "loss": 0.3103, + "num_input_tokens_seen": 75265920, + "step": 36025 + }, + { + "epoch": 5.877722489599478, + "grad_norm": 0.0007512951269745827, + "learning_rate": 0.007236317582081475, + "loss": 0.3145, + "num_input_tokens_seen": 75275520, + "step": 36030 + }, + { + "epoch": 5.878538216820296, + "grad_norm": 0.0006669285940006375, + "learning_rate": 0.007218253632575066, + "loss": 0.3109, + "num_input_tokens_seen": 75284864, + "step": 36035 + }, + { + "epoch": 5.879353944041113, + "grad_norm": 0.0008752928697504103, + "learning_rate": 0.007200211701808223, + "loss": 0.2906, + "num_input_tokens_seen": 75296912, + "step": 36040 + }, + { + "epoch": 5.88016967126193, + "grad_norm": 0.0009017452830448747, + "learning_rate": 0.007182191792563286, + "loss": 0.3307, + "num_input_tokens_seen": 75306672, + "step": 36045 + }, + { + "epoch": 5.880985398482747, + "grad_norm": 0.0006965919747017324, + "learning_rate": 0.0071641939076191145, + "loss": 0.3189, + "num_input_tokens_seen": 75316064, + "step": 36050 + }, + { + "epoch": 5.881801125703564, + "grad_norm": 0.00035325699718669057, + "learning_rate": 0.007146218049751257, + "loss": 0.2968, + "num_input_tokens_seen": 75325568, + "step": 36055 + }, + { + "epoch": 5.882616852924382, + "grad_norm": 0.001068013720214367, + "learning_rate": 0.0071282642217317775, + "loss": 0.3312, + "num_input_tokens_seen": 75336288, + "step": 36060 + }, + { + "epoch": 5.883432580145199, + "grad_norm": 0.0004028325201943517, + "learning_rate": 0.007110332426329396, + "loss": 0.3122, + "num_input_tokens_seen": 75347280, + "step": 36065 + }, + { + "epoch": 5.884248307366017, + "grad_norm": 0.0007817238802090287, + "learning_rate": 0.007092422666309417, + "loss": 0.306, + "num_input_tokens_seen": 75358304, + "step": 36070 + }, + { + "epoch": 5.885064034586835, + "grad_norm": 0.0011090667685493827, + "learning_rate": 0.0070745349444337295, + "loss": 0.3093, + "num_input_tokens_seen": 75369184, + "step": 36075 + }, + { + "epoch": 5.885879761807652, + "grad_norm": 0.0005844812840223312, + "learning_rate": 0.007056669263460913, + "loss": 0.2887, + "num_input_tokens_seen": 75379824, + "step": 36080 + }, + { + "epoch": 5.886695489028469, + "grad_norm": 0.0007428310927934945, + "learning_rate": 0.007038825626145995, + "loss": 0.3373, + "num_input_tokens_seen": 75390688, + "step": 36085 + }, + { + "epoch": 5.887511216249286, + "grad_norm": 0.0003866975021082908, + "learning_rate": 0.007021004035240724, + "loss": 0.3568, + "num_input_tokens_seen": 75400640, + "step": 36090 + }, + { + "epoch": 5.888326943470103, + "grad_norm": 0.0007939612842164934, + "learning_rate": 0.007003204493493453, + "loss": 0.2989, + "num_input_tokens_seen": 75411088, + "step": 36095 + }, + { + "epoch": 5.889142670690921, + "grad_norm": 0.0006436624680645764, + "learning_rate": 0.006985427003649036, + "loss": 0.2692, + "num_input_tokens_seen": 75422512, + "step": 36100 + }, + { + "epoch": 5.889958397911738, + "grad_norm": 0.00035853954614140093, + "learning_rate": 0.006967671568449013, + "loss": 0.271, + "num_input_tokens_seen": 75433360, + "step": 36105 + }, + { + "epoch": 5.890774125132555, + "grad_norm": 0.0008834113832563162, + "learning_rate": 0.006949938190631511, + "loss": 0.3193, + "num_input_tokens_seen": 75444400, + "step": 36110 + }, + { + "epoch": 5.891589852353373, + "grad_norm": 0.00043961452320218086, + "learning_rate": 0.0069322268729311905, + "loss": 0.2878, + "num_input_tokens_seen": 75454544, + "step": 36115 + }, + { + "epoch": 5.8924055795741905, + "grad_norm": 0.0004135681956540793, + "learning_rate": 0.006914537618079403, + "loss": 0.3582, + "num_input_tokens_seen": 75464624, + "step": 36120 + }, + { + "epoch": 5.893221306795008, + "grad_norm": 0.0010565209668129683, + "learning_rate": 0.006896870428804031, + "loss": 0.3126, + "num_input_tokens_seen": 75475104, + "step": 36125 + }, + { + "epoch": 5.894037034015825, + "grad_norm": 0.00045224151108413935, + "learning_rate": 0.006879225307829595, + "loss": 0.2747, + "num_input_tokens_seen": 75485472, + "step": 36130 + }, + { + "epoch": 5.894852761236643, + "grad_norm": 0.000600895844399929, + "learning_rate": 0.00686160225787717, + "loss": 0.3217, + "num_input_tokens_seen": 75496656, + "step": 36135 + }, + { + "epoch": 5.89566848845746, + "grad_norm": 0.0007862052298150957, + "learning_rate": 0.006844001281664463, + "loss": 0.3697, + "num_input_tokens_seen": 75507232, + "step": 36140 + }, + { + "epoch": 5.896484215678277, + "grad_norm": 0.0007931160507723689, + "learning_rate": 0.006826422381905789, + "loss": 0.2938, + "num_input_tokens_seen": 75516800, + "step": 36145 + }, + { + "epoch": 5.897299942899094, + "grad_norm": 0.0011518708197399974, + "learning_rate": 0.006808865561311994, + "loss": 0.349, + "num_input_tokens_seen": 75528800, + "step": 36150 + }, + { + "epoch": 5.898115670119912, + "grad_norm": 0.0005361664807423949, + "learning_rate": 0.00679133082259058, + "loss": 0.3407, + "num_input_tokens_seen": 75540256, + "step": 36155 + }, + { + "epoch": 5.898931397340729, + "grad_norm": 0.0009690735605545342, + "learning_rate": 0.00677381816844565, + "loss": 0.2525, + "num_input_tokens_seen": 75552112, + "step": 36160 + }, + { + "epoch": 5.8997471245615465, + "grad_norm": 0.00040401736623607576, + "learning_rate": 0.0067563276015778434, + "loss": 0.3155, + "num_input_tokens_seen": 75561888, + "step": 36165 + }, + { + "epoch": 5.900562851782364, + "grad_norm": 0.000716913549695164, + "learning_rate": 0.006738859124684437, + "loss": 0.2834, + "num_input_tokens_seen": 75571888, + "step": 36170 + }, + { + "epoch": 5.901378579003182, + "grad_norm": 0.00042200309690088034, + "learning_rate": 0.006721412740459259, + "loss": 0.2892, + "num_input_tokens_seen": 75581872, + "step": 36175 + }, + { + "epoch": 5.902194306223999, + "grad_norm": 0.000676547409966588, + "learning_rate": 0.006703988451592824, + "loss": 0.3021, + "num_input_tokens_seen": 75591600, + "step": 36180 + }, + { + "epoch": 5.903010033444816, + "grad_norm": 0.0008035122300498188, + "learning_rate": 0.006686586260772114, + "loss": 0.3324, + "num_input_tokens_seen": 75602912, + "step": 36185 + }, + { + "epoch": 5.903825760665633, + "grad_norm": 0.0007208614842966199, + "learning_rate": 0.006669206170680819, + "loss": 0.2553, + "num_input_tokens_seen": 75613792, + "step": 36190 + }, + { + "epoch": 5.904641487886451, + "grad_norm": 0.001005398342385888, + "learning_rate": 0.0066518481839991095, + "loss": 0.2661, + "num_input_tokens_seen": 75624720, + "step": 36195 + }, + { + "epoch": 5.905457215107268, + "grad_norm": 0.0005352403968572617, + "learning_rate": 0.006634512303403861, + "loss": 0.2885, + "num_input_tokens_seen": 75634096, + "step": 36200 + }, + { + "epoch": 5.905457215107268, + "eval_loss": 0.31549128890037537, + "eval_runtime": 155.6322, + "eval_samples_per_second": 17.509, + "eval_steps_per_second": 8.758, + "num_input_tokens_seen": 75634096, + "step": 36200 + }, + { + "epoch": 5.906272942328085, + "grad_norm": 0.0007199848187156022, + "learning_rate": 0.0066171985315684355, + "loss": 0.2804, + "num_input_tokens_seen": 75644848, + "step": 36205 + }, + { + "epoch": 5.907088669548903, + "grad_norm": 0.0008089984767138958, + "learning_rate": 0.0065999068711628806, + "loss": 0.3187, + "num_input_tokens_seen": 75654928, + "step": 36210 + }, + { + "epoch": 5.9079043967697205, + "grad_norm": 0.00046765219303779304, + "learning_rate": 0.0065826373248537295, + "loss": 0.3158, + "num_input_tokens_seen": 75663952, + "step": 36215 + }, + { + "epoch": 5.908720123990538, + "grad_norm": 0.0007413563434965909, + "learning_rate": 0.006565389895304218, + "loss": 0.2885, + "num_input_tokens_seen": 75674032, + "step": 36220 + }, + { + "epoch": 5.909535851211355, + "grad_norm": 0.0008502723067067564, + "learning_rate": 0.006548164585174104, + "loss": 0.3666, + "num_input_tokens_seen": 75684448, + "step": 36225 + }, + { + "epoch": 5.910351578432172, + "grad_norm": 0.0005442890687845647, + "learning_rate": 0.006530961397119728, + "loss": 0.2885, + "num_input_tokens_seen": 75694512, + "step": 36230 + }, + { + "epoch": 5.91116730565299, + "grad_norm": 0.0006709861918352544, + "learning_rate": 0.00651378033379405, + "loss": 0.381, + "num_input_tokens_seen": 75705120, + "step": 36235 + }, + { + "epoch": 5.911983032873807, + "grad_norm": 0.0010504391975700855, + "learning_rate": 0.006496621397846619, + "loss": 0.3157, + "num_input_tokens_seen": 75714496, + "step": 36240 + }, + { + "epoch": 5.912798760094624, + "grad_norm": 0.0003687688149511814, + "learning_rate": 0.006479484591923518, + "loss": 0.2929, + "num_input_tokens_seen": 75726192, + "step": 36245 + }, + { + "epoch": 5.913614487315442, + "grad_norm": 0.0004911335418000817, + "learning_rate": 0.006462369918667515, + "loss": 0.3414, + "num_input_tokens_seen": 75736304, + "step": 36250 + }, + { + "epoch": 5.914430214536259, + "grad_norm": 0.0004924662644043565, + "learning_rate": 0.006445277380717851, + "loss": 0.333, + "num_input_tokens_seen": 75745488, + "step": 36255 + }, + { + "epoch": 5.915245941757076, + "grad_norm": 0.0006582245114259422, + "learning_rate": 0.006428206980710466, + "loss": 0.3239, + "num_input_tokens_seen": 75756176, + "step": 36260 + }, + { + "epoch": 5.9160616689778935, + "grad_norm": 0.0006099045276641846, + "learning_rate": 0.006411158721277788, + "loss": 0.2877, + "num_input_tokens_seen": 75767296, + "step": 36265 + }, + { + "epoch": 5.916877396198711, + "grad_norm": 0.0006285603740252554, + "learning_rate": 0.00639413260504888, + "loss": 0.3255, + "num_input_tokens_seen": 75778384, + "step": 36270 + }, + { + "epoch": 5.917693123419529, + "grad_norm": 0.0005098359542898834, + "learning_rate": 0.006377128634649376, + "loss": 0.2579, + "num_input_tokens_seen": 75787296, + "step": 36275 + }, + { + "epoch": 5.918508850640346, + "grad_norm": 0.0006794242653995752, + "learning_rate": 0.006360146812701528, + "loss": 0.3424, + "num_input_tokens_seen": 75797280, + "step": 36280 + }, + { + "epoch": 5.919324577861163, + "grad_norm": 0.000703529454767704, + "learning_rate": 0.006343187141824125, + "loss": 0.3325, + "num_input_tokens_seen": 75807408, + "step": 36285 + }, + { + "epoch": 5.920140305081981, + "grad_norm": 0.0004437017778400332, + "learning_rate": 0.00632624962463259, + "loss": 0.2741, + "num_input_tokens_seen": 75816800, + "step": 36290 + }, + { + "epoch": 5.920956032302798, + "grad_norm": 0.0005678822053596377, + "learning_rate": 0.006309334263738853, + "loss": 0.3215, + "num_input_tokens_seen": 75827808, + "step": 36295 + }, + { + "epoch": 5.921771759523615, + "grad_norm": 0.00035464513348415494, + "learning_rate": 0.006292441061751508, + "loss": 0.2493, + "num_input_tokens_seen": 75839584, + "step": 36300 + }, + { + "epoch": 5.922587486744432, + "grad_norm": 0.0004545227566268295, + "learning_rate": 0.0062755700212757054, + "loss": 0.3214, + "num_input_tokens_seen": 75849248, + "step": 36305 + }, + { + "epoch": 5.92340321396525, + "grad_norm": 0.0004852983693126589, + "learning_rate": 0.006258721144913148, + "loss": 0.3115, + "num_input_tokens_seen": 75860112, + "step": 36310 + }, + { + "epoch": 5.9242189411860675, + "grad_norm": 0.0010400256142020226, + "learning_rate": 0.0062418944352621575, + "loss": 0.3636, + "num_input_tokens_seen": 75871200, + "step": 36315 + }, + { + "epoch": 5.925034668406885, + "grad_norm": 0.0005354845197871327, + "learning_rate": 0.0062250898949176405, + "loss": 0.3181, + "num_input_tokens_seen": 75882256, + "step": 36320 + }, + { + "epoch": 5.925850395627702, + "grad_norm": 0.0007719629211351275, + "learning_rate": 0.006208307526471041, + "loss": 0.3226, + "num_input_tokens_seen": 75893184, + "step": 36325 + }, + { + "epoch": 5.92666612284852, + "grad_norm": 0.001782671664841473, + "learning_rate": 0.006191547332510405, + "loss": 0.3705, + "num_input_tokens_seen": 75902352, + "step": 36330 + }, + { + "epoch": 5.927481850069337, + "grad_norm": 0.0004726468469016254, + "learning_rate": 0.006174809315620416, + "loss": 0.3847, + "num_input_tokens_seen": 75913216, + "step": 36335 + }, + { + "epoch": 5.928297577290154, + "grad_norm": 0.0004043170774821192, + "learning_rate": 0.00615809347838221, + "loss": 0.2997, + "num_input_tokens_seen": 75924304, + "step": 36340 + }, + { + "epoch": 5.929113304510971, + "grad_norm": 0.000898653466720134, + "learning_rate": 0.006141399823373655, + "loss": 0.3575, + "num_input_tokens_seen": 75934400, + "step": 36345 + }, + { + "epoch": 5.929929031731789, + "grad_norm": 0.00043022469617426395, + "learning_rate": 0.0061247283531690455, + "loss": 0.2635, + "num_input_tokens_seen": 75943616, + "step": 36350 + }, + { + "epoch": 5.930744758952606, + "grad_norm": 0.0009009650093503296, + "learning_rate": 0.0061080790703393895, + "loss": 0.3246, + "num_input_tokens_seen": 75954736, + "step": 36355 + }, + { + "epoch": 5.9315604861734235, + "grad_norm": 0.0006832076469436288, + "learning_rate": 0.006091451977452217, + "loss": 0.3106, + "num_input_tokens_seen": 75965424, + "step": 36360 + }, + { + "epoch": 5.932376213394241, + "grad_norm": 0.0007424769573844969, + "learning_rate": 0.00607484707707161, + "loss": 0.3371, + "num_input_tokens_seen": 75975680, + "step": 36365 + }, + { + "epoch": 5.933191940615059, + "grad_norm": 0.0015245408285409212, + "learning_rate": 0.006058264371758254, + "loss": 0.2893, + "num_input_tokens_seen": 75985360, + "step": 36370 + }, + { + "epoch": 5.934007667835876, + "grad_norm": 0.0008049853495322168, + "learning_rate": 0.00604170386406942, + "loss": 0.2854, + "num_input_tokens_seen": 75994256, + "step": 36375 + }, + { + "epoch": 5.934823395056693, + "grad_norm": 0.0005875753704458475, + "learning_rate": 0.006025165556558931, + "loss": 0.3178, + "num_input_tokens_seen": 76004416, + "step": 36380 + }, + { + "epoch": 5.935639122277511, + "grad_norm": 0.0005095448577776551, + "learning_rate": 0.006008649451777248, + "loss": 0.2762, + "num_input_tokens_seen": 76013728, + "step": 36385 + }, + { + "epoch": 5.936454849498328, + "grad_norm": 0.0003198268823325634, + "learning_rate": 0.005992155552271283, + "loss": 0.3878, + "num_input_tokens_seen": 76025168, + "step": 36390 + }, + { + "epoch": 5.937270576719145, + "grad_norm": 0.000432274944614619, + "learning_rate": 0.005975683860584685, + "loss": 0.272, + "num_input_tokens_seen": 76035776, + "step": 36395 + }, + { + "epoch": 5.938086303939962, + "grad_norm": 0.0005637682625092566, + "learning_rate": 0.0059592343792575385, + "loss": 0.3083, + "num_input_tokens_seen": 76046144, + "step": 36400 + }, + { + "epoch": 5.938086303939962, + "eval_loss": 0.31605619192123413, + "eval_runtime": 155.6859, + "eval_samples_per_second": 17.503, + "eval_steps_per_second": 8.755, + "num_input_tokens_seen": 76046144, + "step": 36400 + }, + { + "epoch": 5.938902031160779, + "grad_norm": 0.0006765479338355362, + "learning_rate": 0.0059428071108265975, + "loss": 0.2828, + "num_input_tokens_seen": 76057040, + "step": 36405 + }, + { + "epoch": 5.939717758381597, + "grad_norm": 0.0005003520636819303, + "learning_rate": 0.005926402057825136, + "loss": 0.2821, + "num_input_tokens_seen": 76066784, + "step": 36410 + }, + { + "epoch": 5.940533485602415, + "grad_norm": 0.0006555879372172058, + "learning_rate": 0.005910019222782997, + "loss": 0.2576, + "num_input_tokens_seen": 76076880, + "step": 36415 + }, + { + "epoch": 5.941349212823232, + "grad_norm": 0.0008314935839734972, + "learning_rate": 0.005893658608226643, + "loss": 0.3196, + "num_input_tokens_seen": 76086960, + "step": 36420 + }, + { + "epoch": 5.94216494004405, + "grad_norm": 0.0006550178513862193, + "learning_rate": 0.0058773202166791045, + "loss": 0.3467, + "num_input_tokens_seen": 76098240, + "step": 36425 + }, + { + "epoch": 5.942980667264867, + "grad_norm": 0.0012326715514063835, + "learning_rate": 0.005861004050659918, + "loss": 0.3645, + "num_input_tokens_seen": 76105760, + "step": 36430 + }, + { + "epoch": 5.943796394485684, + "grad_norm": 0.0006157014286145568, + "learning_rate": 0.005844710112685286, + "loss": 0.282, + "num_input_tokens_seen": 76114832, + "step": 36435 + }, + { + "epoch": 5.944612121706501, + "grad_norm": 0.00044220505515113473, + "learning_rate": 0.005828438405267933, + "loss": 0.2644, + "num_input_tokens_seen": 76124800, + "step": 36440 + }, + { + "epoch": 5.945427848927318, + "grad_norm": 0.0007760493317618966, + "learning_rate": 0.00581218893091715, + "loss": 0.3367, + "num_input_tokens_seen": 76136336, + "step": 36445 + }, + { + "epoch": 5.946243576148136, + "grad_norm": 0.0003988494863733649, + "learning_rate": 0.005795961692138801, + "loss": 0.2473, + "num_input_tokens_seen": 76147744, + "step": 36450 + }, + { + "epoch": 5.947059303368953, + "grad_norm": 0.0004322379536461085, + "learning_rate": 0.00577975669143535, + "loss": 0.3054, + "num_input_tokens_seen": 76157040, + "step": 36455 + }, + { + "epoch": 5.9478750305897705, + "grad_norm": 0.0004554407496470958, + "learning_rate": 0.005763573931305782, + "loss": 0.2677, + "num_input_tokens_seen": 76167424, + "step": 36460 + }, + { + "epoch": 5.9486907578105885, + "grad_norm": 0.00046809480409137905, + "learning_rate": 0.005747413414245733, + "loss": 0.306, + "num_input_tokens_seen": 76177216, + "step": 36465 + }, + { + "epoch": 5.949506485031406, + "grad_norm": 0.000514293962623924, + "learning_rate": 0.005731275142747294, + "loss": 0.2619, + "num_input_tokens_seen": 76186768, + "step": 36470 + }, + { + "epoch": 5.950322212252223, + "grad_norm": 0.0005655235145241022, + "learning_rate": 0.005715159119299256, + "loss": 0.2605, + "num_input_tokens_seen": 76197264, + "step": 36475 + }, + { + "epoch": 5.95113793947304, + "grad_norm": 0.0005387559067457914, + "learning_rate": 0.005699065346386867, + "loss": 0.3164, + "num_input_tokens_seen": 76207232, + "step": 36480 + }, + { + "epoch": 5.951953666693858, + "grad_norm": 0.0006387574831023812, + "learning_rate": 0.0056829938264919885, + "loss": 0.3162, + "num_input_tokens_seen": 76217920, + "step": 36485 + }, + { + "epoch": 5.952769393914675, + "grad_norm": 0.00042529404163360596, + "learning_rate": 0.005666944562093074, + "loss": 0.3183, + "num_input_tokens_seen": 76228608, + "step": 36490 + }, + { + "epoch": 5.953585121135492, + "grad_norm": 0.0008324864320456982, + "learning_rate": 0.005650917555665108, + "loss": 0.3148, + "num_input_tokens_seen": 76238656, + "step": 36495 + }, + { + "epoch": 5.954400848356309, + "grad_norm": 0.000556051847524941, + "learning_rate": 0.005634912809679632, + "loss": 0.299, + "num_input_tokens_seen": 76247696, + "step": 36500 + }, + { + "epoch": 5.955216575577127, + "grad_norm": 0.0003925525234080851, + "learning_rate": 0.005618930326604854, + "loss": 0.3161, + "num_input_tokens_seen": 76258256, + "step": 36505 + }, + { + "epoch": 5.9560323027979445, + "grad_norm": 0.0005812462186440825, + "learning_rate": 0.005602970108905386, + "loss": 0.2682, + "num_input_tokens_seen": 76268544, + "step": 36510 + }, + { + "epoch": 5.956848030018762, + "grad_norm": 0.0008239729795604944, + "learning_rate": 0.005587032159042543, + "loss": 0.3012, + "num_input_tokens_seen": 76277440, + "step": 36515 + }, + { + "epoch": 5.957663757239579, + "grad_norm": 0.000507341290358454, + "learning_rate": 0.005571116479474158, + "loss": 0.3194, + "num_input_tokens_seen": 76288512, + "step": 36520 + }, + { + "epoch": 5.958479484460397, + "grad_norm": 0.0006320161046460271, + "learning_rate": 0.005555223072654619, + "loss": 0.2909, + "num_input_tokens_seen": 76297424, + "step": 36525 + }, + { + "epoch": 5.959295211681214, + "grad_norm": 0.0004144490812905133, + "learning_rate": 0.005539351941034881, + "loss": 0.2805, + "num_input_tokens_seen": 76308160, + "step": 36530 + }, + { + "epoch": 5.960110938902031, + "grad_norm": 0.0004319490399211645, + "learning_rate": 0.0055235030870624865, + "loss": 0.2763, + "num_input_tokens_seen": 76320032, + "step": 36535 + }, + { + "epoch": 5.960926666122848, + "grad_norm": 0.000527695519849658, + "learning_rate": 0.005507676513181514, + "loss": 0.3595, + "num_input_tokens_seen": 76330368, + "step": 36540 + }, + { + "epoch": 5.961742393343666, + "grad_norm": 0.0005940515547990799, + "learning_rate": 0.005491872221832628, + "loss": 0.2382, + "num_input_tokens_seen": 76340896, + "step": 36545 + }, + { + "epoch": 5.962558120564483, + "grad_norm": 0.0008136474061757326, + "learning_rate": 0.005476090215453061, + "loss": 0.3252, + "num_input_tokens_seen": 76350032, + "step": 36550 + }, + { + "epoch": 5.9633738477853, + "grad_norm": 0.0005868250736966729, + "learning_rate": 0.0054603304964765675, + "loss": 0.2613, + "num_input_tokens_seen": 76361312, + "step": 36555 + }, + { + "epoch": 5.964189575006118, + "grad_norm": 0.0013613286428153515, + "learning_rate": 0.005444593067333519, + "loss": 0.3723, + "num_input_tokens_seen": 76370128, + "step": 36560 + }, + { + "epoch": 5.965005302226936, + "grad_norm": 0.000831242126878351, + "learning_rate": 0.00542887793045081, + "loss": 0.321, + "num_input_tokens_seen": 76379952, + "step": 36565 + }, + { + "epoch": 5.965821029447753, + "grad_norm": 0.000649085093755275, + "learning_rate": 0.005413185088251932, + "loss": 0.2641, + "num_input_tokens_seen": 76390992, + "step": 36570 + }, + { + "epoch": 5.96663675666857, + "grad_norm": 0.0004458021139726043, + "learning_rate": 0.005397514543156884, + "loss": 0.3038, + "num_input_tokens_seen": 76402128, + "step": 36575 + }, + { + "epoch": 5.967452483889387, + "grad_norm": 0.0007371202809736133, + "learning_rate": 0.0053818662975822825, + "loss": 0.3157, + "num_input_tokens_seen": 76412672, + "step": 36580 + }, + { + "epoch": 5.968268211110205, + "grad_norm": 0.0010300613939762115, + "learning_rate": 0.005366240353941315, + "loss": 0.3312, + "num_input_tokens_seen": 76422704, + "step": 36585 + }, + { + "epoch": 5.969083938331022, + "grad_norm": 0.0005303160869516432, + "learning_rate": 0.005350636714643636, + "loss": 0.3315, + "num_input_tokens_seen": 76432464, + "step": 36590 + }, + { + "epoch": 5.969899665551839, + "grad_norm": 0.0006614881567656994, + "learning_rate": 0.005335055382095555, + "loss": 0.2781, + "num_input_tokens_seen": 76443008, + "step": 36595 + }, + { + "epoch": 5.970715392772657, + "grad_norm": 0.00044573357445187867, + "learning_rate": 0.005319496358699915, + "loss": 0.2806, + "num_input_tokens_seen": 76453936, + "step": 36600 + }, + { + "epoch": 5.970715392772657, + "eval_loss": 0.31565406918525696, + "eval_runtime": 155.7692, + "eval_samples_per_second": 17.494, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 76453936, + "step": 36600 + }, + { + "epoch": 5.971531119993474, + "grad_norm": 0.0007011384004727006, + "learning_rate": 0.005303959646856099, + "loss": 0.2668, + "num_input_tokens_seen": 76464688, + "step": 36605 + }, + { + "epoch": 5.9723468472142915, + "grad_norm": 0.0008564831223338842, + "learning_rate": 0.005288445248960089, + "loss": 0.3511, + "num_input_tokens_seen": 76474640, + "step": 36610 + }, + { + "epoch": 5.973162574435109, + "grad_norm": 0.0007051386055536568, + "learning_rate": 0.005272953167404354, + "loss": 0.2539, + "num_input_tokens_seen": 76484944, + "step": 36615 + }, + { + "epoch": 5.973978301655926, + "grad_norm": 0.0006410479545593262, + "learning_rate": 0.005257483404578017, + "loss": 0.3528, + "num_input_tokens_seen": 76495760, + "step": 36620 + }, + { + "epoch": 5.974794028876744, + "grad_norm": 0.0009166172239929438, + "learning_rate": 0.0052420359628666865, + "loss": 0.3514, + "num_input_tokens_seen": 76505008, + "step": 36625 + }, + { + "epoch": 5.975609756097561, + "grad_norm": 0.0008018937078304589, + "learning_rate": 0.00522661084465254, + "loss": 0.3377, + "num_input_tokens_seen": 76516160, + "step": 36630 + }, + { + "epoch": 5.976425483318378, + "grad_norm": 0.001341049442999065, + "learning_rate": 0.005211208052314326, + "loss": 0.2926, + "num_input_tokens_seen": 76527360, + "step": 36635 + }, + { + "epoch": 5.977241210539196, + "grad_norm": 0.0007283168961293995, + "learning_rate": 0.005195827588227391, + "loss": 0.2474, + "num_input_tokens_seen": 76537792, + "step": 36640 + }, + { + "epoch": 5.978056937760013, + "grad_norm": 0.0011945064179599285, + "learning_rate": 0.0051804694547635255, + "loss": 0.3982, + "num_input_tokens_seen": 76548192, + "step": 36645 + }, + { + "epoch": 5.97887266498083, + "grad_norm": 0.0006776383961550891, + "learning_rate": 0.005165133654291232, + "loss": 0.3274, + "num_input_tokens_seen": 76558736, + "step": 36650 + }, + { + "epoch": 5.9796883922016475, + "grad_norm": 0.0009050153312273324, + "learning_rate": 0.005149820189175402, + "loss": 0.3272, + "num_input_tokens_seen": 76569808, + "step": 36655 + }, + { + "epoch": 5.9805041194224655, + "grad_norm": 0.0010063295485451818, + "learning_rate": 0.005134529061777598, + "loss": 0.3272, + "num_input_tokens_seen": 76580848, + "step": 36660 + }, + { + "epoch": 5.981319846643283, + "grad_norm": 0.0010047449031844735, + "learning_rate": 0.005119260274455933, + "loss": 0.3297, + "num_input_tokens_seen": 76591184, + "step": 36665 + }, + { + "epoch": 5.9821355738641, + "grad_norm": 0.0012771589681506157, + "learning_rate": 0.005104013829565007, + "loss": 0.316, + "num_input_tokens_seen": 76600016, + "step": 36670 + }, + { + "epoch": 5.982951301084917, + "grad_norm": 0.0007266694447025657, + "learning_rate": 0.005088789729456006, + "loss": 0.3675, + "num_input_tokens_seen": 76611648, + "step": 36675 + }, + { + "epoch": 5.983767028305735, + "grad_norm": 0.0003714004415087402, + "learning_rate": 0.005073587976476735, + "loss": 0.2888, + "num_input_tokens_seen": 76620976, + "step": 36680 + }, + { + "epoch": 5.984582755526552, + "grad_norm": 0.001503273961134255, + "learning_rate": 0.005058408572971418, + "loss": 0.312, + "num_input_tokens_seen": 76631728, + "step": 36685 + }, + { + "epoch": 5.985398482747369, + "grad_norm": 0.0005905559519305825, + "learning_rate": 0.005043251521280983, + "loss": 0.3376, + "num_input_tokens_seen": 76642432, + "step": 36690 + }, + { + "epoch": 5.986214209968186, + "grad_norm": 0.0014386632246896625, + "learning_rate": 0.005028116823742795, + "loss": 0.4297, + "num_input_tokens_seen": 76653600, + "step": 36695 + }, + { + "epoch": 5.987029937189004, + "grad_norm": 0.0006699323421344161, + "learning_rate": 0.005013004482690819, + "loss": 0.2978, + "num_input_tokens_seen": 76663904, + "step": 36700 + }, + { + "epoch": 5.9878456644098215, + "grad_norm": 0.000835960207041353, + "learning_rate": 0.0049979145004555746, + "loss": 0.3024, + "num_input_tokens_seen": 76674880, + "step": 36705 + }, + { + "epoch": 5.988661391630639, + "grad_norm": 0.000633834395557642, + "learning_rate": 0.004982846879364116, + "loss": 0.3214, + "num_input_tokens_seen": 76685600, + "step": 36710 + }, + { + "epoch": 5.989477118851456, + "grad_norm": 0.0005787134869024158, + "learning_rate": 0.0049678016217400535, + "loss": 0.2913, + "num_input_tokens_seen": 76695392, + "step": 36715 + }, + { + "epoch": 5.990292846072274, + "grad_norm": 0.0006937838043086231, + "learning_rate": 0.004952778729903595, + "loss": 0.3588, + "num_input_tokens_seen": 76704784, + "step": 36720 + }, + { + "epoch": 5.991108573293091, + "grad_norm": 0.0006052165408618748, + "learning_rate": 0.004937778206171422, + "loss": 0.295, + "num_input_tokens_seen": 76715152, + "step": 36725 + }, + { + "epoch": 5.991924300513908, + "grad_norm": 0.0010055617894977331, + "learning_rate": 0.004922800052856835, + "loss": 0.3605, + "num_input_tokens_seen": 76726512, + "step": 36730 + }, + { + "epoch": 5.992740027734725, + "grad_norm": 0.0009621756034903228, + "learning_rate": 0.004907844272269602, + "loss": 0.3338, + "num_input_tokens_seen": 76736720, + "step": 36735 + }, + { + "epoch": 5.993555754955543, + "grad_norm": 0.001094267936423421, + "learning_rate": 0.004892910866716144, + "loss": 0.294, + "num_input_tokens_seen": 76748640, + "step": 36740 + }, + { + "epoch": 5.99437148217636, + "grad_norm": 0.0005049263127148151, + "learning_rate": 0.004877999838499369, + "loss": 0.2925, + "num_input_tokens_seen": 76758736, + "step": 36745 + }, + { + "epoch": 5.995187209397177, + "grad_norm": 0.00041519440128467977, + "learning_rate": 0.0048631111899187065, + "loss": 0.352, + "num_input_tokens_seen": 76767664, + "step": 36750 + }, + { + "epoch": 5.9960029366179945, + "grad_norm": 0.0005008794250898063, + "learning_rate": 0.0048482449232702335, + "loss": 0.34, + "num_input_tokens_seen": 76778704, + "step": 36755 + }, + { + "epoch": 5.996818663838813, + "grad_norm": 0.0006471873493865132, + "learning_rate": 0.004833401040846469, + "loss": 0.2669, + "num_input_tokens_seen": 76789360, + "step": 36760 + }, + { + "epoch": 5.99763439105963, + "grad_norm": 0.0008065662696026266, + "learning_rate": 0.004818579544936546, + "loss": 0.2994, + "num_input_tokens_seen": 76799728, + "step": 36765 + }, + { + "epoch": 5.998450118280447, + "grad_norm": 0.0004576406499836594, + "learning_rate": 0.004803780437826121, + "loss": 0.2633, + "num_input_tokens_seen": 76809392, + "step": 36770 + }, + { + "epoch": 5.999265845501265, + "grad_norm": 0.001148347626440227, + "learning_rate": 0.004789003721797402, + "loss": 0.3495, + "num_input_tokens_seen": 76818672, + "step": 36775 + }, + { + "epoch": 6.0, + "grad_norm": 0.0005134973907843232, + "learning_rate": 0.004774249399129132, + "loss": 0.2945, + "num_input_tokens_seen": 76827984, + "step": 36780 + }, + { + "epoch": 6.000815727220817, + "grad_norm": 0.0004986735875718296, + "learning_rate": 0.004759517472096642, + "loss": 0.2344, + "num_input_tokens_seen": 76838816, + "step": 36785 + }, + { + "epoch": 6.001631454441635, + "grad_norm": 0.0009692031308077276, + "learning_rate": 0.004744807942971746, + "loss": 0.3527, + "num_input_tokens_seen": 76850432, + "step": 36790 + }, + { + "epoch": 6.002447181662452, + "grad_norm": 0.0005822222447022796, + "learning_rate": 0.004730120814022881, + "loss": 0.2868, + "num_input_tokens_seen": 76862288, + "step": 36795 + }, + { + "epoch": 6.003262908883269, + "grad_norm": 0.0007175618084147573, + "learning_rate": 0.004715456087514935, + "loss": 0.2728, + "num_input_tokens_seen": 76873152, + "step": 36800 + }, + { + "epoch": 6.003262908883269, + "eval_loss": 0.3154376149177551, + "eval_runtime": 155.6838, + "eval_samples_per_second": 17.503, + "eval_steps_per_second": 8.755, + "num_input_tokens_seen": 76873152, + "step": 36800 + }, + { + "epoch": 6.0040786361040865, + "grad_norm": 0.000771168852224946, + "learning_rate": 0.004700813765709432, + "loss": 0.2751, + "num_input_tokens_seen": 76883280, + "step": 36805 + }, + { + "epoch": 6.004894363324905, + "grad_norm": 0.0007804621127434075, + "learning_rate": 0.004686193850864401, + "loss": 0.3313, + "num_input_tokens_seen": 76893440, + "step": 36810 + }, + { + "epoch": 6.005710090545722, + "grad_norm": 0.0005143180605955422, + "learning_rate": 0.004671596345234385, + "loss": 0.3669, + "num_input_tokens_seen": 76903248, + "step": 36815 + }, + { + "epoch": 6.006525817766539, + "grad_norm": 0.0013364697806537151, + "learning_rate": 0.00465702125107052, + "loss": 0.2797, + "num_input_tokens_seen": 76912240, + "step": 36820 + }, + { + "epoch": 6.007341544987356, + "grad_norm": 0.0008321262430399656, + "learning_rate": 0.004642468570620506, + "loss": 0.3636, + "num_input_tokens_seen": 76923536, + "step": 36825 + }, + { + "epoch": 6.008157272208174, + "grad_norm": 0.0003847806656267494, + "learning_rate": 0.004627938306128482, + "loss": 0.3034, + "num_input_tokens_seen": 76933200, + "step": 36830 + }, + { + "epoch": 6.008972999428991, + "grad_norm": 0.0007489518611691892, + "learning_rate": 0.004613430459835255, + "loss": 0.3343, + "num_input_tokens_seen": 76941984, + "step": 36835 + }, + { + "epoch": 6.009788726649808, + "grad_norm": 0.0005279916804283857, + "learning_rate": 0.004598945033978085, + "loss": 0.3789, + "num_input_tokens_seen": 76952912, + "step": 36840 + }, + { + "epoch": 6.010604453870625, + "grad_norm": 0.0006823279545642436, + "learning_rate": 0.004584482030790804, + "loss": 0.4093, + "num_input_tokens_seen": 76962512, + "step": 36845 + }, + { + "epoch": 6.011420181091443, + "grad_norm": 0.0005192599492147565, + "learning_rate": 0.004570041452503826, + "loss": 0.3295, + "num_input_tokens_seen": 76973392, + "step": 36850 + }, + { + "epoch": 6.0122359083122605, + "grad_norm": 0.0006554268766194582, + "learning_rate": 0.004555623301344003, + "loss": 0.341, + "num_input_tokens_seen": 76984064, + "step": 36855 + }, + { + "epoch": 6.013051635533078, + "grad_norm": 0.0008002616232261062, + "learning_rate": 0.004541227579534857, + "loss": 0.3017, + "num_input_tokens_seen": 76995280, + "step": 36860 + }, + { + "epoch": 6.013867362753895, + "grad_norm": 0.0003830831265076995, + "learning_rate": 0.004526854289296378, + "loss": 0.295, + "num_input_tokens_seen": 77006688, + "step": 36865 + }, + { + "epoch": 6.014683089974713, + "grad_norm": 0.000715323374606669, + "learning_rate": 0.004512503432845078, + "loss": 0.3298, + "num_input_tokens_seen": 77018864, + "step": 36870 + }, + { + "epoch": 6.01549881719553, + "grad_norm": 0.0008125831955112517, + "learning_rate": 0.004498175012394068, + "loss": 0.3314, + "num_input_tokens_seen": 77030208, + "step": 36875 + }, + { + "epoch": 6.016314544416347, + "grad_norm": 0.0006619712221436203, + "learning_rate": 0.004483869030152965, + "loss": 0.2987, + "num_input_tokens_seen": 77041664, + "step": 36880 + }, + { + "epoch": 6.017130271637164, + "grad_norm": 0.0006027640192769468, + "learning_rate": 0.004469585488327904, + "loss": 0.2629, + "num_input_tokens_seen": 77053616, + "step": 36885 + }, + { + "epoch": 6.017945998857982, + "grad_norm": 0.0006710284505970776, + "learning_rate": 0.0044553243891216395, + "loss": 0.3699, + "num_input_tokens_seen": 77062064, + "step": 36890 + }, + { + "epoch": 6.018761726078799, + "grad_norm": 0.000507176446262747, + "learning_rate": 0.004441085734733363, + "loss": 0.3026, + "num_input_tokens_seen": 77072272, + "step": 36895 + }, + { + "epoch": 6.0195774532996165, + "grad_norm": 0.00047193767386488616, + "learning_rate": 0.004426869527358884, + "loss": 0.2508, + "num_input_tokens_seen": 77082320, + "step": 36900 + }, + { + "epoch": 6.020393180520434, + "grad_norm": 0.000691449735313654, + "learning_rate": 0.0044126757691905156, + "loss": 0.2611, + "num_input_tokens_seen": 77091840, + "step": 36905 + }, + { + "epoch": 6.021208907741252, + "grad_norm": 0.0005865743733011186, + "learning_rate": 0.004398504462417107, + "loss": 0.3022, + "num_input_tokens_seen": 77101872, + "step": 36910 + }, + { + "epoch": 6.022024634962069, + "grad_norm": 0.0007216845988295972, + "learning_rate": 0.0043843556092240605, + "loss": 0.2644, + "num_input_tokens_seen": 77110336, + "step": 36915 + }, + { + "epoch": 6.022840362182886, + "grad_norm": 0.0009081993484869599, + "learning_rate": 0.004370229211793281, + "loss": 0.3307, + "num_input_tokens_seen": 77120096, + "step": 36920 + }, + { + "epoch": 6.023656089403703, + "grad_norm": 0.0004077151825185865, + "learning_rate": 0.0043561252723032405, + "loss": 0.3556, + "num_input_tokens_seen": 77131136, + "step": 36925 + }, + { + "epoch": 6.024471816624521, + "grad_norm": 0.0006989827961660922, + "learning_rate": 0.004342043792929001, + "loss": 0.261, + "num_input_tokens_seen": 77141296, + "step": 36930 + }, + { + "epoch": 6.025287543845338, + "grad_norm": 0.0006579973269253969, + "learning_rate": 0.004327984775842025, + "loss": 0.3167, + "num_input_tokens_seen": 77153008, + "step": 36935 + }, + { + "epoch": 6.026103271066155, + "grad_norm": 0.000504465657286346, + "learning_rate": 0.004313948223210428, + "loss": 0.3304, + "num_input_tokens_seen": 77162848, + "step": 36940 + }, + { + "epoch": 6.026918998286972, + "grad_norm": 0.0009242350934073329, + "learning_rate": 0.004299934137198846, + "loss": 0.3316, + "num_input_tokens_seen": 77173920, + "step": 36945 + }, + { + "epoch": 6.02773472550779, + "grad_norm": 0.0008425858104601502, + "learning_rate": 0.004285942519968383, + "loss": 0.3378, + "num_input_tokens_seen": 77183936, + "step": 36950 + }, + { + "epoch": 6.028550452728608, + "grad_norm": 0.00032840276253409684, + "learning_rate": 0.004271973373676746, + "loss": 0.3316, + "num_input_tokens_seen": 77194256, + "step": 36955 + }, + { + "epoch": 6.029366179949425, + "grad_norm": 0.0006101444596424699, + "learning_rate": 0.004258026700478146, + "loss": 0.3533, + "num_input_tokens_seen": 77205296, + "step": 36960 + }, + { + "epoch": 6.030181907170242, + "grad_norm": 0.0012066058116033673, + "learning_rate": 0.004244102502523328, + "loss": 0.3139, + "num_input_tokens_seen": 77216608, + "step": 36965 + }, + { + "epoch": 6.03099763439106, + "grad_norm": 0.0006816160166636109, + "learning_rate": 0.004230200781959592, + "loss": 0.289, + "num_input_tokens_seen": 77225840, + "step": 36970 + }, + { + "epoch": 6.031813361611877, + "grad_norm": 0.0003618893679231405, + "learning_rate": 0.004216321540930756, + "loss": 0.2302, + "num_input_tokens_seen": 77237120, + "step": 36975 + }, + { + "epoch": 6.032629088832694, + "grad_norm": 0.0006793481879867613, + "learning_rate": 0.004202464781577175, + "loss": 0.3599, + "num_input_tokens_seen": 77246736, + "step": 36980 + }, + { + "epoch": 6.033444816053512, + "grad_norm": 0.0007537544588558376, + "learning_rate": 0.00418863050603574, + "loss": 0.355, + "num_input_tokens_seen": 77258144, + "step": 36985 + }, + { + "epoch": 6.034260543274329, + "grad_norm": 0.0006311398465186357, + "learning_rate": 0.004174818716439843, + "loss": 0.2841, + "num_input_tokens_seen": 77268016, + "step": 36990 + }, + { + "epoch": 6.035076270495146, + "grad_norm": 0.0004402613267302513, + "learning_rate": 0.004161029414919464, + "loss": 0.2911, + "num_input_tokens_seen": 77279504, + "step": 36995 + }, + { + "epoch": 6.0358919977159635, + "grad_norm": 0.0016428560484200716, + "learning_rate": 0.004147262603601071, + "loss": 0.357, + "num_input_tokens_seen": 77290000, + "step": 37000 + }, + { + "epoch": 6.0358919977159635, + "eval_loss": 0.3161553144454956, + "eval_runtime": 155.8354, + "eval_samples_per_second": 17.486, + "eval_steps_per_second": 8.746, + "num_input_tokens_seen": 77290000, + "step": 37000 + }, + { + "epoch": 6.0367077249367815, + "grad_norm": 0.0008852860191836953, + "learning_rate": 0.004133518284607679, + "loss": 0.3927, + "num_input_tokens_seen": 77298624, + "step": 37005 + }, + { + "epoch": 6.037523452157599, + "grad_norm": 0.000863156805280596, + "learning_rate": 0.004119796460058861, + "loss": 0.3014, + "num_input_tokens_seen": 77308416, + "step": 37010 + }, + { + "epoch": 6.038339179378416, + "grad_norm": 0.0015824656002223492, + "learning_rate": 0.00410609713207064, + "loss": 0.3295, + "num_input_tokens_seen": 77319728, + "step": 37015 + }, + { + "epoch": 6.039154906599233, + "grad_norm": 0.00046139530604705215, + "learning_rate": 0.004092420302755678, + "loss": 0.2826, + "num_input_tokens_seen": 77330864, + "step": 37020 + }, + { + "epoch": 6.039970633820051, + "grad_norm": 0.0013155442429706454, + "learning_rate": 0.004078765974223103, + "loss": 0.3163, + "num_input_tokens_seen": 77341616, + "step": 37025 + }, + { + "epoch": 6.040786361040868, + "grad_norm": 0.001527190557681024, + "learning_rate": 0.004065134148578564, + "loss": 0.3621, + "num_input_tokens_seen": 77354208, + "step": 37030 + }, + { + "epoch": 6.041602088261685, + "grad_norm": 0.00042234163265675306, + "learning_rate": 0.004051524827924279, + "loss": 0.2683, + "num_input_tokens_seen": 77365424, + "step": 37035 + }, + { + "epoch": 6.042417815482502, + "grad_norm": 0.0005729419644922018, + "learning_rate": 0.004037938014358955, + "loss": 0.3089, + "num_input_tokens_seen": 77374656, + "step": 37040 + }, + { + "epoch": 6.04323354270332, + "grad_norm": 0.000673779402859509, + "learning_rate": 0.004024373709977863, + "loss": 0.2397, + "num_input_tokens_seen": 77385168, + "step": 37045 + }, + { + "epoch": 6.0440492699241375, + "grad_norm": 0.0007428499520756304, + "learning_rate": 0.004010831916872814, + "loss": 0.3666, + "num_input_tokens_seen": 77396656, + "step": 37050 + }, + { + "epoch": 6.044864997144955, + "grad_norm": 0.0007267192704603076, + "learning_rate": 0.003997312637132089, + "loss": 0.2777, + "num_input_tokens_seen": 77406192, + "step": 37055 + }, + { + "epoch": 6.045680724365772, + "grad_norm": 0.000525350624229759, + "learning_rate": 0.003983815872840535, + "loss": 0.2815, + "num_input_tokens_seen": 77418224, + "step": 37060 + }, + { + "epoch": 6.04649645158659, + "grad_norm": 0.0007196424412541091, + "learning_rate": 0.003970341626079521, + "loss": 0.3516, + "num_input_tokens_seen": 77429408, + "step": 37065 + }, + { + "epoch": 6.047312178807407, + "grad_norm": 0.0007257215329445899, + "learning_rate": 0.003956889898926952, + "loss": 0.331, + "num_input_tokens_seen": 77439424, + "step": 37070 + }, + { + "epoch": 6.048127906028224, + "grad_norm": 0.0011060984106734395, + "learning_rate": 0.0039434606934572675, + "loss": 0.4118, + "num_input_tokens_seen": 77450400, + "step": 37075 + }, + { + "epoch": 6.048943633249041, + "grad_norm": 0.0009273673640564084, + "learning_rate": 0.003930054011741396, + "loss": 0.3027, + "num_input_tokens_seen": 77461232, + "step": 37080 + }, + { + "epoch": 6.049759360469859, + "grad_norm": 0.0007414421997964382, + "learning_rate": 0.0039166698558468155, + "loss": 0.2986, + "num_input_tokens_seen": 77472656, + "step": 37085 + }, + { + "epoch": 6.050575087690676, + "grad_norm": 0.0004251625796314329, + "learning_rate": 0.0039033082278375594, + "loss": 0.3028, + "num_input_tokens_seen": 77482944, + "step": 37090 + }, + { + "epoch": 6.051390814911493, + "grad_norm": 0.0008568436605855823, + "learning_rate": 0.003889969129774112, + "loss": 0.2764, + "num_input_tokens_seen": 77494544, + "step": 37095 + }, + { + "epoch": 6.052206542132311, + "grad_norm": 0.0004472802975215018, + "learning_rate": 0.0038766525637135784, + "loss": 0.3229, + "num_input_tokens_seen": 77503920, + "step": 37100 + }, + { + "epoch": 6.053022269353129, + "grad_norm": 0.0008783168741501868, + "learning_rate": 0.0038633585317095318, + "loss": 0.3327, + "num_input_tokens_seen": 77513344, + "step": 37105 + }, + { + "epoch": 6.053837996573946, + "grad_norm": 0.000745507248211652, + "learning_rate": 0.00385008703581205, + "loss": 0.3342, + "num_input_tokens_seen": 77523312, + "step": 37110 + }, + { + "epoch": 6.054653723794763, + "grad_norm": 0.0011857160134240985, + "learning_rate": 0.0038368380780677944, + "loss": 0.3088, + "num_input_tokens_seen": 77533088, + "step": 37115 + }, + { + "epoch": 6.05546945101558, + "grad_norm": 0.00045264893560670316, + "learning_rate": 0.003823611660519882, + "loss": 0.3091, + "num_input_tokens_seen": 77542928, + "step": 37120 + }, + { + "epoch": 6.056285178236398, + "grad_norm": 0.0010289200581610203, + "learning_rate": 0.0038104077852080475, + "loss": 0.3231, + "num_input_tokens_seen": 77553328, + "step": 37125 + }, + { + "epoch": 6.057100905457215, + "grad_norm": 0.0006056306883692741, + "learning_rate": 0.003797226454168462, + "loss": 0.3228, + "num_input_tokens_seen": 77563808, + "step": 37130 + }, + { + "epoch": 6.057916632678032, + "grad_norm": 0.00042818303336389363, + "learning_rate": 0.003784067669433849, + "loss": 0.2433, + "num_input_tokens_seen": 77574160, + "step": 37135 + }, + { + "epoch": 6.058732359898849, + "grad_norm": 0.000673524453304708, + "learning_rate": 0.0037709314330334528, + "loss": 0.3191, + "num_input_tokens_seen": 77585888, + "step": 37140 + }, + { + "epoch": 6.059548087119667, + "grad_norm": 0.0006812526844441891, + "learning_rate": 0.003757817746993086, + "loss": 0.2388, + "num_input_tokens_seen": 77597664, + "step": 37145 + }, + { + "epoch": 6.0603638143404845, + "grad_norm": 0.0007938789785839617, + "learning_rate": 0.0037447266133349977, + "loss": 0.2979, + "num_input_tokens_seen": 77608592, + "step": 37150 + }, + { + "epoch": 6.061179541561302, + "grad_norm": 0.0005184844485484064, + "learning_rate": 0.003731658034078039, + "loss": 0.3086, + "num_input_tokens_seen": 77618736, + "step": 37155 + }, + { + "epoch": 6.06199526878212, + "grad_norm": 0.00046926617505960166, + "learning_rate": 0.0037186120112375153, + "loss": 0.2437, + "num_input_tokens_seen": 77629808, + "step": 37160 + }, + { + "epoch": 6.062810996002937, + "grad_norm": 0.0005344604724086821, + "learning_rate": 0.003705588546825317, + "loss": 0.3059, + "num_input_tokens_seen": 77640416, + "step": 37165 + }, + { + "epoch": 6.063626723223754, + "grad_norm": 0.001001137774437666, + "learning_rate": 0.0036925876428498205, + "loss": 0.309, + "num_input_tokens_seen": 77651008, + "step": 37170 + }, + { + "epoch": 6.064442450444571, + "grad_norm": 0.0005249250098131597, + "learning_rate": 0.0036796093013159057, + "loss": 0.3421, + "num_input_tokens_seen": 77660224, + "step": 37175 + }, + { + "epoch": 6.065258177665389, + "grad_norm": 0.0005435140919871628, + "learning_rate": 0.0036666535242250217, + "loss": 0.2953, + "num_input_tokens_seen": 77669808, + "step": 37180 + }, + { + "epoch": 6.066073904886206, + "grad_norm": 0.0013597480719909072, + "learning_rate": 0.003653720313575104, + "loss": 0.2823, + "num_input_tokens_seen": 77679744, + "step": 37185 + }, + { + "epoch": 6.066889632107023, + "grad_norm": 0.0005922373384237289, + "learning_rate": 0.003640809671360623, + "loss": 0.2835, + "num_input_tokens_seen": 77688512, + "step": 37190 + }, + { + "epoch": 6.0677053593278405, + "grad_norm": 0.0003677695640362799, + "learning_rate": 0.003627921599572553, + "loss": 0.3058, + "num_input_tokens_seen": 77698336, + "step": 37195 + }, + { + "epoch": 6.0685210865486585, + "grad_norm": 0.0005819974467158318, + "learning_rate": 0.003615056100198405, + "loss": 0.32, + "num_input_tokens_seen": 77708416, + "step": 37200 + }, + { + "epoch": 6.0685210865486585, + "eval_loss": 0.31542208790779114, + "eval_runtime": 156.0613, + "eval_samples_per_second": 17.461, + "eval_steps_per_second": 8.734, + "num_input_tokens_seen": 77708416, + "step": 37200 + }, + { + "epoch": 6.069336813769476, + "grad_norm": 0.0005028264713473618, + "learning_rate": 0.003602213175222174, + "loss": 0.3107, + "num_input_tokens_seen": 77718320, + "step": 37205 + }, + { + "epoch": 6.070152540990293, + "grad_norm": 0.000682084821164608, + "learning_rate": 0.0035893928266244432, + "loss": 0.2578, + "num_input_tokens_seen": 77728624, + "step": 37210 + }, + { + "epoch": 6.07096826821111, + "grad_norm": 0.00047022380749695003, + "learning_rate": 0.003576595056382248, + "loss": 0.348, + "num_input_tokens_seen": 77738368, + "step": 37215 + }, + { + "epoch": 6.071783995431928, + "grad_norm": 0.0005536889657378197, + "learning_rate": 0.0035638198664691423, + "loss": 0.3045, + "num_input_tokens_seen": 77749792, + "step": 37220 + }, + { + "epoch": 6.072599722652745, + "grad_norm": 0.0009908846113830805, + "learning_rate": 0.003551067258855267, + "loss": 0.2838, + "num_input_tokens_seen": 77759264, + "step": 37225 + }, + { + "epoch": 6.073415449873562, + "grad_norm": 0.0010588932782411575, + "learning_rate": 0.0035383372355071996, + "loss": 0.2922, + "num_input_tokens_seen": 77769888, + "step": 37230 + }, + { + "epoch": 6.074231177094379, + "grad_norm": 0.0004099419747944921, + "learning_rate": 0.0035256297983881023, + "loss": 0.2991, + "num_input_tokens_seen": 77779120, + "step": 37235 + }, + { + "epoch": 6.075046904315197, + "grad_norm": 0.0006013864767737687, + "learning_rate": 0.0035129449494575747, + "loss": 0.3287, + "num_input_tokens_seen": 77788288, + "step": 37240 + }, + { + "epoch": 6.0758626315360145, + "grad_norm": 0.0004866982635576278, + "learning_rate": 0.0035002826906718187, + "loss": 0.3009, + "num_input_tokens_seen": 77798288, + "step": 37245 + }, + { + "epoch": 6.076678358756832, + "grad_norm": 0.0007729962817393243, + "learning_rate": 0.003487643023983522, + "loss": 0.3384, + "num_input_tokens_seen": 77809072, + "step": 37250 + }, + { + "epoch": 6.077494085977649, + "grad_norm": 0.0011034111957997084, + "learning_rate": 0.003475025951341842, + "loss": 0.3673, + "num_input_tokens_seen": 77819184, + "step": 37255 + }, + { + "epoch": 6.078309813198467, + "grad_norm": 0.0006160447956062853, + "learning_rate": 0.00346243147469249, + "loss": 0.2488, + "num_input_tokens_seen": 77830048, + "step": 37260 + }, + { + "epoch": 6.079125540419284, + "grad_norm": 0.0007962443633005023, + "learning_rate": 0.0034498595959777446, + "loss": 0.3026, + "num_input_tokens_seen": 77840368, + "step": 37265 + }, + { + "epoch": 6.079941267640101, + "grad_norm": 0.0006562072085216641, + "learning_rate": 0.003437310317136305, + "loss": 0.3757, + "num_input_tokens_seen": 77851552, + "step": 37270 + }, + { + "epoch": 6.080756994860918, + "grad_norm": 0.0006932303658686578, + "learning_rate": 0.0034247836401034236, + "loss": 0.3113, + "num_input_tokens_seen": 77862192, + "step": 37275 + }, + { + "epoch": 6.081572722081736, + "grad_norm": 0.0007567377761006355, + "learning_rate": 0.003412279566810905, + "loss": 0.3473, + "num_input_tokens_seen": 77872464, + "step": 37280 + }, + { + "epoch": 6.082388449302553, + "grad_norm": 0.0008004591218195856, + "learning_rate": 0.00339979809918699, + "loss": 0.3077, + "num_input_tokens_seen": 77883024, + "step": 37285 + }, + { + "epoch": 6.08320417652337, + "grad_norm": 0.000591331860050559, + "learning_rate": 0.0033873392391565228, + "loss": 0.309, + "num_input_tokens_seen": 77893824, + "step": 37290 + }, + { + "epoch": 6.0840199037441876, + "grad_norm": 0.0007315069669857621, + "learning_rate": 0.003374902988640782, + "loss": 0.2726, + "num_input_tokens_seen": 77903056, + "step": 37295 + }, + { + "epoch": 6.084835630965006, + "grad_norm": 0.0005198602448217571, + "learning_rate": 0.0033624893495576014, + "loss": 0.2485, + "num_input_tokens_seen": 77914144, + "step": 37300 + }, + { + "epoch": 6.085651358185823, + "grad_norm": 0.0006354361539706588, + "learning_rate": 0.0033500983238213323, + "loss": 0.254, + "num_input_tokens_seen": 77924352, + "step": 37305 + }, + { + "epoch": 6.08646708540664, + "grad_norm": 0.000401119701564312, + "learning_rate": 0.0033377299133428126, + "loss": 0.2748, + "num_input_tokens_seen": 77935680, + "step": 37310 + }, + { + "epoch": 6.087282812627457, + "grad_norm": 0.0007558850920759141, + "learning_rate": 0.003325384120029434, + "loss": 0.2784, + "num_input_tokens_seen": 77946192, + "step": 37315 + }, + { + "epoch": 6.088098539848275, + "grad_norm": 0.0008726358064450324, + "learning_rate": 0.0033130609457850233, + "loss": 0.3525, + "num_input_tokens_seen": 77957408, + "step": 37320 + }, + { + "epoch": 6.088914267069092, + "grad_norm": 0.0005779850180260837, + "learning_rate": 0.0033007603925100104, + "loss": 0.2781, + "num_input_tokens_seen": 77966896, + "step": 37325 + }, + { + "epoch": 6.089729994289909, + "grad_norm": 0.0007031439454294741, + "learning_rate": 0.003288482462101294, + "loss": 0.3487, + "num_input_tokens_seen": 77978240, + "step": 37330 + }, + { + "epoch": 6.090545721510727, + "grad_norm": 0.00035680682049132884, + "learning_rate": 0.0032762271564522605, + "loss": 0.2731, + "num_input_tokens_seen": 77988144, + "step": 37335 + }, + { + "epoch": 6.091361448731544, + "grad_norm": 0.0004970598383806646, + "learning_rate": 0.003263994477452864, + "loss": 0.3108, + "num_input_tokens_seen": 77997536, + "step": 37340 + }, + { + "epoch": 6.0921771759523615, + "grad_norm": 0.0008182222954928875, + "learning_rate": 0.0032517844269895125, + "loss": 0.2974, + "num_input_tokens_seen": 78007664, + "step": 37345 + }, + { + "epoch": 6.092992903173179, + "grad_norm": 0.0005435607163235545, + "learning_rate": 0.0032395970069451496, + "loss": 0.304, + "num_input_tokens_seen": 78018256, + "step": 37350 + }, + { + "epoch": 6.093808630393997, + "grad_norm": 0.0008044097921811044, + "learning_rate": 0.0032274322191992388, + "loss": 0.318, + "num_input_tokens_seen": 78028672, + "step": 37355 + }, + { + "epoch": 6.094624357614814, + "grad_norm": 0.0007977175409905612, + "learning_rate": 0.0032152900656277294, + "loss": 0.3073, + "num_input_tokens_seen": 78039440, + "step": 37360 + }, + { + "epoch": 6.095440084835631, + "grad_norm": 0.0006483711767941713, + "learning_rate": 0.0032031705481030902, + "loss": 0.3398, + "num_input_tokens_seen": 78050848, + "step": 37365 + }, + { + "epoch": 6.096255812056448, + "grad_norm": 0.000618780788499862, + "learning_rate": 0.0031910736684943428, + "loss": 0.2778, + "num_input_tokens_seen": 78060528, + "step": 37370 + }, + { + "epoch": 6.097071539277266, + "grad_norm": 0.0008666878566145897, + "learning_rate": 0.0031789994286669453, + "loss": 0.342, + "num_input_tokens_seen": 78071344, + "step": 37375 + }, + { + "epoch": 6.097887266498083, + "grad_norm": 0.0005567519110627472, + "learning_rate": 0.003166947830482908, + "loss": 0.3007, + "num_input_tokens_seen": 78082832, + "step": 37380 + }, + { + "epoch": 6.0987029937189, + "grad_norm": 0.0006166702369228005, + "learning_rate": 0.003154918875800727, + "loss": 0.2798, + "num_input_tokens_seen": 78092896, + "step": 37385 + }, + { + "epoch": 6.0995187209397175, + "grad_norm": 0.0006385628948919475, + "learning_rate": 0.00314291256647542, + "loss": 0.3294, + "num_input_tokens_seen": 78103872, + "step": 37390 + }, + { + "epoch": 6.1003344481605355, + "grad_norm": 0.000437468639574945, + "learning_rate": 0.0031309289043585375, + "loss": 0.3854, + "num_input_tokens_seen": 78113824, + "step": 37395 + }, + { + "epoch": 6.101150175381353, + "grad_norm": 0.0018058340065181255, + "learning_rate": 0.003118967891298069, + "loss": 0.3433, + "num_input_tokens_seen": 78124432, + "step": 37400 + }, + { + "epoch": 6.101150175381353, + "eval_loss": 0.31546273827552795, + "eval_runtime": 155.7703, + "eval_samples_per_second": 17.494, + "eval_steps_per_second": 8.75, + "num_input_tokens_seen": 78124432, + "step": 37400 + }, + { + "epoch": 6.10196590260217, + "grad_norm": 0.0008789340499788523, + "learning_rate": 0.003107029529138572, + "loss": 0.2805, + "num_input_tokens_seen": 78133600, + "step": 37405 + }, + { + "epoch": 6.102781629822987, + "grad_norm": 0.0009538076119497418, + "learning_rate": 0.0030951138197211235, + "loss": 0.3313, + "num_input_tokens_seen": 78144688, + "step": 37410 + }, + { + "epoch": 6.103597357043805, + "grad_norm": 0.0007016001618467271, + "learning_rate": 0.0030832207648832377, + "loss": 0.2822, + "num_input_tokens_seen": 78155968, + "step": 37415 + }, + { + "epoch": 6.104413084264622, + "grad_norm": 0.0006940207094885409, + "learning_rate": 0.0030713503664589635, + "loss": 0.3255, + "num_input_tokens_seen": 78165616, + "step": 37420 + }, + { + "epoch": 6.105228811485439, + "grad_norm": 0.0006749214953742921, + "learning_rate": 0.0030595026262788872, + "loss": 0.2499, + "num_input_tokens_seen": 78175168, + "step": 37425 + }, + { + "epoch": 6.106044538706256, + "grad_norm": 0.0006859297282062471, + "learning_rate": 0.00304767754617008, + "loss": 0.3369, + "num_input_tokens_seen": 78184784, + "step": 37430 + }, + { + "epoch": 6.106860265927074, + "grad_norm": 0.0008841755334287882, + "learning_rate": 0.003035875127956117, + "loss": 0.3233, + "num_input_tokens_seen": 78195328, + "step": 37435 + }, + { + "epoch": 6.1076759931478914, + "grad_norm": 0.0006492807879112661, + "learning_rate": 0.0030240953734570752, + "loss": 0.2998, + "num_input_tokens_seen": 78206128, + "step": 37440 + }, + { + "epoch": 6.108491720368709, + "grad_norm": 0.0006822214927524328, + "learning_rate": 0.003012338284489535, + "loss": 0.3375, + "num_input_tokens_seen": 78215264, + "step": 37445 + }, + { + "epoch": 6.109307447589526, + "grad_norm": 0.0007495224708691239, + "learning_rate": 0.0030006038628665964, + "loss": 0.2652, + "num_input_tokens_seen": 78224640, + "step": 37450 + }, + { + "epoch": 6.110123174810344, + "grad_norm": 0.0009947916259989142, + "learning_rate": 0.002988892110397845, + "loss": 0.3684, + "num_input_tokens_seen": 78235056, + "step": 37455 + }, + { + "epoch": 6.110938902031161, + "grad_norm": 0.0006443532183766365, + "learning_rate": 0.0029772030288894025, + "loss": 0.3475, + "num_input_tokens_seen": 78245984, + "step": 37460 + }, + { + "epoch": 6.111754629251978, + "grad_norm": 0.000735436740797013, + "learning_rate": 0.0029655366201438438, + "loss": 0.2918, + "num_input_tokens_seen": 78256080, + "step": 37465 + }, + { + "epoch": 6.112570356472795, + "grad_norm": 0.0005480244290083647, + "learning_rate": 0.0029538928859602965, + "loss": 0.2414, + "num_input_tokens_seen": 78265104, + "step": 37470 + }, + { + "epoch": 6.113386083693613, + "grad_norm": 0.0009262060630135238, + "learning_rate": 0.002942271828134374, + "loss": 0.372, + "num_input_tokens_seen": 78274992, + "step": 37475 + }, + { + "epoch": 6.11420181091443, + "grad_norm": 0.0005990703939460218, + "learning_rate": 0.00293067344845816, + "loss": 0.2685, + "num_input_tokens_seen": 78286320, + "step": 37480 + }, + { + "epoch": 6.115017538135247, + "grad_norm": 0.00045911726192571223, + "learning_rate": 0.0029190977487202896, + "loss": 0.3449, + "num_input_tokens_seen": 78298160, + "step": 37485 + }, + { + "epoch": 6.1158332653560645, + "grad_norm": 0.00047111036838032305, + "learning_rate": 0.0029075447307058853, + "loss": 0.2564, + "num_input_tokens_seen": 78308624, + "step": 37490 + }, + { + "epoch": 6.1166489925768825, + "grad_norm": 0.0007661999552510679, + "learning_rate": 0.0028960143961965722, + "loss": 0.2847, + "num_input_tokens_seen": 78318560, + "step": 37495 + }, + { + "epoch": 6.1174647197977, + "grad_norm": 0.0004829551326110959, + "learning_rate": 0.002884506746970461, + "loss": 0.3229, + "num_input_tokens_seen": 78329296, + "step": 37500 + }, + { + "epoch": 6.118280447018517, + "grad_norm": 0.000712225039023906, + "learning_rate": 0.0028730217848021654, + "loss": 0.3227, + "num_input_tokens_seen": 78339648, + "step": 37505 + }, + { + "epoch": 6.119096174239334, + "grad_norm": 0.00057779107009992, + "learning_rate": 0.0028615595114628188, + "loss": 0.3241, + "num_input_tokens_seen": 78350320, + "step": 37510 + }, + { + "epoch": 6.119911901460152, + "grad_norm": 0.0004930337308906019, + "learning_rate": 0.002850119928720074, + "loss": 0.2711, + "num_input_tokens_seen": 78361168, + "step": 37515 + }, + { + "epoch": 6.120727628680969, + "grad_norm": 0.0004998705699108541, + "learning_rate": 0.0028387030383380195, + "loss": 0.3489, + "num_input_tokens_seen": 78371328, + "step": 37520 + }, + { + "epoch": 6.121543355901786, + "grad_norm": 0.0006736535578966141, + "learning_rate": 0.0028273088420772974, + "loss": 0.3451, + "num_input_tokens_seen": 78382928, + "step": 37525 + }, + { + "epoch": 6.122359083122603, + "grad_norm": 0.0005783177330158651, + "learning_rate": 0.002815937341695068, + "loss": 0.3342, + "num_input_tokens_seen": 78393872, + "step": 37530 + }, + { + "epoch": 6.123174810343421, + "grad_norm": 0.0008642030879855156, + "learning_rate": 0.0028045885389448963, + "loss": 0.2868, + "num_input_tokens_seen": 78405536, + "step": 37535 + }, + { + "epoch": 6.1239905375642385, + "grad_norm": 0.0012648946139961481, + "learning_rate": 0.002793262435576965, + "loss": 0.3545, + "num_input_tokens_seen": 78415200, + "step": 37540 + }, + { + "epoch": 6.124806264785056, + "grad_norm": 0.0004770503146573901, + "learning_rate": 0.0027819590333378772, + "loss": 0.2884, + "num_input_tokens_seen": 78425984, + "step": 37545 + }, + { + "epoch": 6.125621992005874, + "grad_norm": 0.0005109467310830951, + "learning_rate": 0.002770678333970755, + "loss": 0.3388, + "num_input_tokens_seen": 78436480, + "step": 37550 + }, + { + "epoch": 6.126437719226691, + "grad_norm": 0.0009199987398460507, + "learning_rate": 0.0027594203392152573, + "loss": 0.302, + "num_input_tokens_seen": 78447504, + "step": 37555 + }, + { + "epoch": 6.127253446447508, + "grad_norm": 0.0005069015314802527, + "learning_rate": 0.002748185050807478, + "loss": 0.299, + "num_input_tokens_seen": 78457632, + "step": 37560 + }, + { + "epoch": 6.128069173668325, + "grad_norm": 0.0009584052604623139, + "learning_rate": 0.002736972470480031, + "loss": 0.2417, + "num_input_tokens_seen": 78469408, + "step": 37565 + }, + { + "epoch": 6.128884900889143, + "grad_norm": 0.0016016705194488168, + "learning_rate": 0.002725782599962068, + "loss": 0.3398, + "num_input_tokens_seen": 78481472, + "step": 37570 + }, + { + "epoch": 6.12970062810996, + "grad_norm": 0.00039681611815467477, + "learning_rate": 0.0027146154409791734, + "loss": 0.2896, + "num_input_tokens_seen": 78490992, + "step": 37575 + }, + { + "epoch": 6.130516355330777, + "grad_norm": 0.00037320758565329015, + "learning_rate": 0.002703470995253504, + "loss": 0.3603, + "num_input_tokens_seen": 78500864, + "step": 37580 + }, + { + "epoch": 6.1313320825515945, + "grad_norm": 0.0011951100314036012, + "learning_rate": 0.0026923492645036184, + "loss": 0.3394, + "num_input_tokens_seen": 78511696, + "step": 37585 + }, + { + "epoch": 6.1321478097724125, + "grad_norm": 0.0008893910562619567, + "learning_rate": 0.0026812502504446776, + "loss": 0.3034, + "num_input_tokens_seen": 78522112, + "step": 37590 + }, + { + "epoch": 6.13296353699323, + "grad_norm": 0.0010979811195284128, + "learning_rate": 0.0026701739547882798, + "loss": 0.3619, + "num_input_tokens_seen": 78532192, + "step": 37595 + }, + { + "epoch": 6.133779264214047, + "grad_norm": 0.0006584410439245403, + "learning_rate": 0.0026591203792425077, + "loss": 0.252, + "num_input_tokens_seen": 78542400, + "step": 37600 + }, + { + "epoch": 6.133779264214047, + "eval_loss": 0.31545624136924744, + "eval_runtime": 155.8054, + "eval_samples_per_second": 17.49, + "eval_steps_per_second": 8.748, + "num_input_tokens_seen": 78542400, + "step": 37600 + }, + { + "epoch": 6.134594991434864, + "grad_norm": 0.0011483065318316221, + "learning_rate": 0.0026480895255119818, + "loss": 0.37, + "num_input_tokens_seen": 78552080, + "step": 37605 + }, + { + "epoch": 6.135410718655682, + "grad_norm": 0.0006931668613106012, + "learning_rate": 0.002637081395297791, + "loss": 0.3212, + "num_input_tokens_seen": 78561856, + "step": 37610 + }, + { + "epoch": 6.136226445876499, + "grad_norm": 0.0005954872467555106, + "learning_rate": 0.0026260959902975113, + "loss": 0.3416, + "num_input_tokens_seen": 78573328, + "step": 37615 + }, + { + "epoch": 6.137042173097316, + "grad_norm": 0.0006126216030679643, + "learning_rate": 0.00261513331220527, + "loss": 0.2763, + "num_input_tokens_seen": 78582816, + "step": 37620 + }, + { + "epoch": 6.137857900318133, + "grad_norm": 0.0005592160159721971, + "learning_rate": 0.0026041933627116154, + "loss": 0.2623, + "num_input_tokens_seen": 78593616, + "step": 37625 + }, + { + "epoch": 6.138673627538951, + "grad_norm": 0.0005705709336325526, + "learning_rate": 0.0025932761435036476, + "loss": 0.2836, + "num_input_tokens_seen": 78605536, + "step": 37630 + }, + { + "epoch": 6.139489354759768, + "grad_norm": 0.0007039845222607255, + "learning_rate": 0.002582381656264904, + "loss": 0.3033, + "num_input_tokens_seen": 78617152, + "step": 37635 + }, + { + "epoch": 6.1403050819805856, + "grad_norm": 0.00044408245594240725, + "learning_rate": 0.0025715099026754895, + "loss": 0.3429, + "num_input_tokens_seen": 78629088, + "step": 37640 + }, + { + "epoch": 6.141120809201403, + "grad_norm": 0.0006004236056469381, + "learning_rate": 0.002560660884411947, + "loss": 0.3156, + "num_input_tokens_seen": 78639680, + "step": 37645 + }, + { + "epoch": 6.141936536422221, + "grad_norm": 0.0007278035627678037, + "learning_rate": 0.0025498346031473385, + "loss": 0.3541, + "num_input_tokens_seen": 78649680, + "step": 37650 + }, + { + "epoch": 6.142752263643038, + "grad_norm": 0.0008646927890367806, + "learning_rate": 0.0025390310605511945, + "loss": 0.3423, + "num_input_tokens_seen": 78661200, + "step": 37655 + }, + { + "epoch": 6.143567990863855, + "grad_norm": 0.0006466704653576016, + "learning_rate": 0.0025282502582895995, + "loss": 0.2432, + "num_input_tokens_seen": 78671248, + "step": 37660 + }, + { + "epoch": 6.144383718084672, + "grad_norm": 0.0012740344973281026, + "learning_rate": 0.002517492198025023, + "loss": 0.3532, + "num_input_tokens_seen": 78681440, + "step": 37665 + }, + { + "epoch": 6.14519944530549, + "grad_norm": 0.0007384260534308851, + "learning_rate": 0.0025067568814165554, + "loss": 0.3659, + "num_input_tokens_seen": 78693152, + "step": 37670 + }, + { + "epoch": 6.146015172526307, + "grad_norm": 0.0006125232903286815, + "learning_rate": 0.0024960443101196884, + "loss": 0.2876, + "num_input_tokens_seen": 78703936, + "step": 37675 + }, + { + "epoch": 6.146830899747124, + "grad_norm": 0.0006061093299649656, + "learning_rate": 0.002485354485786434, + "loss": 0.3341, + "num_input_tokens_seen": 78714448, + "step": 37680 + }, + { + "epoch": 6.1476466269679415, + "grad_norm": 0.00048387428978458047, + "learning_rate": 0.002474687410065307, + "loss": 0.3069, + "num_input_tokens_seen": 78725568, + "step": 37685 + }, + { + "epoch": 6.1484623541887595, + "grad_norm": 0.000761691655497998, + "learning_rate": 0.002464043084601308, + "loss": 0.3111, + "num_input_tokens_seen": 78736144, + "step": 37690 + }, + { + "epoch": 6.149278081409577, + "grad_norm": 0.0007681539282202721, + "learning_rate": 0.0024534215110358915, + "loss": 0.3574, + "num_input_tokens_seen": 78745888, + "step": 37695 + }, + { + "epoch": 6.150093808630394, + "grad_norm": 0.0008400016813538969, + "learning_rate": 0.002442822691007096, + "loss": 0.2777, + "num_input_tokens_seen": 78757104, + "step": 37700 + }, + { + "epoch": 6.150909535851211, + "grad_norm": 0.0005452588084153831, + "learning_rate": 0.002432246626149348, + "loss": 0.3051, + "num_input_tokens_seen": 78767344, + "step": 37705 + }, + { + "epoch": 6.151725263072029, + "grad_norm": 0.0005588463973253965, + "learning_rate": 0.002421693318093626, + "loss": 0.2559, + "num_input_tokens_seen": 78777984, + "step": 37710 + }, + { + "epoch": 6.152540990292846, + "grad_norm": 0.0009750047465786338, + "learning_rate": 0.0024111627684673784, + "loss": 0.3394, + "num_input_tokens_seen": 78788976, + "step": 37715 + }, + { + "epoch": 6.153356717513663, + "grad_norm": 0.0007261957507580519, + "learning_rate": 0.0024006549788945395, + "loss": 0.3547, + "num_input_tokens_seen": 78800800, + "step": 37720 + }, + { + "epoch": 6.154172444734481, + "grad_norm": 0.000551793840713799, + "learning_rate": 0.0023901699509955463, + "loss": 0.2925, + "num_input_tokens_seen": 78811568, + "step": 37725 + }, + { + "epoch": 6.154988171955298, + "grad_norm": 0.0004633610660675913, + "learning_rate": 0.0023797076863873554, + "loss": 0.2946, + "num_input_tokens_seen": 78821792, + "step": 37730 + }, + { + "epoch": 6.1558038991761155, + "grad_norm": 0.0006732348119840026, + "learning_rate": 0.0023692681866833262, + "loss": 0.2786, + "num_input_tokens_seen": 78832368, + "step": 37735 + }, + { + "epoch": 6.156619626396933, + "grad_norm": 0.0006948859081603587, + "learning_rate": 0.0023588514534934046, + "loss": 0.2778, + "num_input_tokens_seen": 78842160, + "step": 37740 + }, + { + "epoch": 6.157435353617751, + "grad_norm": 0.0011241453466936946, + "learning_rate": 0.002348457488423955, + "loss": 0.2691, + "num_input_tokens_seen": 78852560, + "step": 37745 + }, + { + "epoch": 6.158251080838568, + "grad_norm": 0.0006150839617475867, + "learning_rate": 0.0023380862930778624, + "loss": 0.3108, + "num_input_tokens_seen": 78863808, + "step": 37750 + }, + { + "epoch": 6.159066808059385, + "grad_norm": 0.000665110070258379, + "learning_rate": 0.0023277378690545135, + "loss": 0.334, + "num_input_tokens_seen": 78875072, + "step": 37755 + }, + { + "epoch": 6.159882535280202, + "grad_norm": 0.0006403200095519423, + "learning_rate": 0.0023174122179497325, + "loss": 0.3478, + "num_input_tokens_seen": 78884928, + "step": 37760 + }, + { + "epoch": 6.16069826250102, + "grad_norm": 0.0012112330878153443, + "learning_rate": 0.0023071093413558784, + "loss": 0.291, + "num_input_tokens_seen": 78895376, + "step": 37765 + }, + { + "epoch": 6.161513989721837, + "grad_norm": 0.0006345485453493893, + "learning_rate": 0.002296829240861814, + "loss": 0.3055, + "num_input_tokens_seen": 78906896, + "step": 37770 + }, + { + "epoch": 6.162329716942654, + "grad_norm": 0.0008591646328568459, + "learning_rate": 0.002286571918052821, + "loss": 0.3375, + "num_input_tokens_seen": 78916768, + "step": 37775 + }, + { + "epoch": 6.163145444163471, + "grad_norm": 0.000424445781391114, + "learning_rate": 0.0022763373745107174, + "loss": 0.261, + "num_input_tokens_seen": 78927424, + "step": 37780 + }, + { + "epoch": 6.1639611713842895, + "grad_norm": 0.0005652191466651857, + "learning_rate": 0.0022661256118138074, + "loss": 0.2941, + "num_input_tokens_seen": 78938368, + "step": 37785 + }, + { + "epoch": 6.164776898605107, + "grad_norm": 0.000514440587721765, + "learning_rate": 0.0022559366315368645, + "loss": 0.3126, + "num_input_tokens_seen": 78947824, + "step": 37790 + }, + { + "epoch": 6.165592625825924, + "grad_norm": 0.0006658504134975374, + "learning_rate": 0.002245770435251182, + "loss": 0.2684, + "num_input_tokens_seen": 78958352, + "step": 37795 + }, + { + "epoch": 6.166408353046741, + "grad_norm": 0.0008488808525726199, + "learning_rate": 0.002235627024524456, + "loss": 0.3092, + "num_input_tokens_seen": 78968368, + "step": 37800 + }, + { + "epoch": 6.166408353046741, + "eval_loss": 0.31496521830558777, + "eval_runtime": 155.8423, + "eval_samples_per_second": 17.486, + "eval_steps_per_second": 8.746, + "num_input_tokens_seen": 78968368, + "step": 37800 + }, + { + "epoch": 6.167224080267559, + "grad_norm": 0.0005602281889878213, + "learning_rate": 0.0022255064009209847, + "loss": 0.3082, + "num_input_tokens_seen": 78979024, + "step": 37805 + }, + { + "epoch": 6.168039807488376, + "grad_norm": 0.0006702117971144617, + "learning_rate": 0.0022154085660014864, + "loss": 0.325, + "num_input_tokens_seen": 78988768, + "step": 37810 + }, + { + "epoch": 6.168855534709193, + "grad_norm": 0.0006694725016131997, + "learning_rate": 0.0022053335213231494, + "loss": 0.2831, + "num_input_tokens_seen": 78998128, + "step": 37815 + }, + { + "epoch": 6.16967126193001, + "grad_norm": 0.00045731678255833685, + "learning_rate": 0.002195281268439697, + "loss": 0.3161, + "num_input_tokens_seen": 79009536, + "step": 37820 + }, + { + "epoch": 6.170486989150828, + "grad_norm": 0.0005128413322381675, + "learning_rate": 0.002185251808901306, + "loss": 0.3445, + "num_input_tokens_seen": 79020048, + "step": 37825 + }, + { + "epoch": 6.171302716371645, + "grad_norm": 0.0005686835502274334, + "learning_rate": 0.0021752451442546227, + "loss": 0.4125, + "num_input_tokens_seen": 79030624, + "step": 37830 + }, + { + "epoch": 6.1721184435924625, + "grad_norm": 0.0007458358304575086, + "learning_rate": 0.0021652612760428456, + "loss": 0.291, + "num_input_tokens_seen": 79040512, + "step": 37835 + }, + { + "epoch": 6.17293417081328, + "grad_norm": 0.0011234736302867532, + "learning_rate": 0.0021553002058055603, + "loss": 0.2858, + "num_input_tokens_seen": 79050864, + "step": 37840 + }, + { + "epoch": 6.173749898034098, + "grad_norm": 0.0004909763811156154, + "learning_rate": 0.0021453619350789376, + "loss": 0.3583, + "num_input_tokens_seen": 79061008, + "step": 37845 + }, + { + "epoch": 6.174565625254915, + "grad_norm": 0.000804874871391803, + "learning_rate": 0.0021354464653955516, + "loss": 0.2984, + "num_input_tokens_seen": 79071696, + "step": 37850 + }, + { + "epoch": 6.175381352475732, + "grad_norm": 0.0004528005374595523, + "learning_rate": 0.002125553798284513, + "loss": 0.3399, + "num_input_tokens_seen": 79081568, + "step": 37855 + }, + { + "epoch": 6.176197079696549, + "grad_norm": 0.0004367143556009978, + "learning_rate": 0.002115683935271384, + "loss": 0.2071, + "num_input_tokens_seen": 79090976, + "step": 37860 + }, + { + "epoch": 6.177012806917367, + "grad_norm": 0.0005160470609553158, + "learning_rate": 0.0021058368778782144, + "loss": 0.2928, + "num_input_tokens_seen": 79101808, + "step": 37865 + }, + { + "epoch": 6.177828534138184, + "grad_norm": 0.0008289122488349676, + "learning_rate": 0.002096012627623539, + "loss": 0.2813, + "num_input_tokens_seen": 79112272, + "step": 37870 + }, + { + "epoch": 6.178644261359001, + "grad_norm": 0.0005137823754921556, + "learning_rate": 0.00208621118602243, + "loss": 0.3147, + "num_input_tokens_seen": 79123088, + "step": 37875 + }, + { + "epoch": 6.1794599885798185, + "grad_norm": 0.0010453517315909266, + "learning_rate": 0.002076432554586327, + "loss": 0.3048, + "num_input_tokens_seen": 79133488, + "step": 37880 + }, + { + "epoch": 6.1802757158006365, + "grad_norm": 0.0005871920729987323, + "learning_rate": 0.002066676734823258, + "loss": 0.2997, + "num_input_tokens_seen": 79144224, + "step": 37885 + }, + { + "epoch": 6.181091443021454, + "grad_norm": 0.0007071997970342636, + "learning_rate": 0.0020569437282376866, + "loss": 0.331, + "num_input_tokens_seen": 79153888, + "step": 37890 + }, + { + "epoch": 6.181907170242271, + "grad_norm": 0.0008339528576470912, + "learning_rate": 0.002047233536330545, + "loss": 0.2906, + "num_input_tokens_seen": 79164576, + "step": 37895 + }, + { + "epoch": 6.182722897463089, + "grad_norm": 0.0006483588949777186, + "learning_rate": 0.0020375461605993015, + "loss": 0.2501, + "num_input_tokens_seen": 79175104, + "step": 37900 + }, + { + "epoch": 6.183538624683906, + "grad_norm": 0.0005768515402451158, + "learning_rate": 0.002027881602537845, + "loss": 0.256, + "num_input_tokens_seen": 79185472, + "step": 37905 + }, + { + "epoch": 6.184354351904723, + "grad_norm": 0.0010021708440035582, + "learning_rate": 0.002018239863636567, + "loss": 0.3136, + "num_input_tokens_seen": 79196848, + "step": 37910 + }, + { + "epoch": 6.18517007912554, + "grad_norm": 0.0005349140265025198, + "learning_rate": 0.002008620945382378, + "loss": 0.2505, + "num_input_tokens_seen": 79206096, + "step": 37915 + }, + { + "epoch": 6.185985806346358, + "grad_norm": 0.0007703242590650916, + "learning_rate": 0.001999024849258607, + "loss": 0.3389, + "num_input_tokens_seen": 79215904, + "step": 37920 + }, + { + "epoch": 6.186801533567175, + "grad_norm": 0.0009690594160929322, + "learning_rate": 0.001989451576745105, + "loss": 0.3042, + "num_input_tokens_seen": 79226176, + "step": 37925 + }, + { + "epoch": 6.1876172607879925, + "grad_norm": 0.0006603389047086239, + "learning_rate": 0.00197990112931819, + "loss": 0.3113, + "num_input_tokens_seen": 79236272, + "step": 37930 + }, + { + "epoch": 6.18843298800881, + "grad_norm": 0.0006136813317425549, + "learning_rate": 0.0019703735084506345, + "loss": 0.2487, + "num_input_tokens_seen": 79246736, + "step": 37935 + }, + { + "epoch": 6.189248715229628, + "grad_norm": 0.0007109259604476392, + "learning_rate": 0.001960868715611763, + "loss": 0.3408, + "num_input_tokens_seen": 79256384, + "step": 37940 + }, + { + "epoch": 6.190064442450445, + "grad_norm": 0.0009074500412680209, + "learning_rate": 0.0019513867522673034, + "loss": 0.3339, + "num_input_tokens_seen": 79268032, + "step": 37945 + }, + { + "epoch": 6.190880169671262, + "grad_norm": 0.0006436952971853316, + "learning_rate": 0.001941927619879502, + "loss": 0.2569, + "num_input_tokens_seen": 79279328, + "step": 37950 + }, + { + "epoch": 6.191695896892079, + "grad_norm": 0.000580022984649986, + "learning_rate": 0.0019324913199070758, + "loss": 0.2858, + "num_input_tokens_seen": 79289616, + "step": 37955 + }, + { + "epoch": 6.192511624112897, + "grad_norm": 0.0006666498957201838, + "learning_rate": 0.0019230778538052106, + "loss": 0.3312, + "num_input_tokens_seen": 79298368, + "step": 37960 + }, + { + "epoch": 6.193327351333714, + "grad_norm": 0.000508948927745223, + "learning_rate": 0.0019136872230255952, + "loss": 0.308, + "num_input_tokens_seen": 79306544, + "step": 37965 + }, + { + "epoch": 6.194143078554531, + "grad_norm": 0.0006110290996730328, + "learning_rate": 0.0019043194290164045, + "loss": 0.3148, + "num_input_tokens_seen": 79316144, + "step": 37970 + }, + { + "epoch": 6.194958805775348, + "grad_norm": 0.0007732802187092602, + "learning_rate": 0.0018949744732222162, + "loss": 0.292, + "num_input_tokens_seen": 79325568, + "step": 37975 + }, + { + "epoch": 6.195774532996166, + "grad_norm": 0.0012619496555998921, + "learning_rate": 0.0018856523570841776, + "loss": 0.338, + "num_input_tokens_seen": 79337152, + "step": 37980 + }, + { + "epoch": 6.196590260216984, + "grad_norm": 0.0005066294688731432, + "learning_rate": 0.0018763530820398555, + "loss": 0.4199, + "num_input_tokens_seen": 79346832, + "step": 37985 + }, + { + "epoch": 6.197405987437801, + "grad_norm": 0.0005698896711692214, + "learning_rate": 0.0018670766495233525, + "loss": 0.2755, + "num_input_tokens_seen": 79357984, + "step": 37990 + }, + { + "epoch": 6.198221714658618, + "grad_norm": 0.0009243098902516067, + "learning_rate": 0.001857823060965158, + "loss": 0.3464, + "num_input_tokens_seen": 79368224, + "step": 37995 + }, + { + "epoch": 6.199037441879436, + "grad_norm": 0.0008099837577901781, + "learning_rate": 0.0018485923177923467, + "loss": 0.2739, + "num_input_tokens_seen": 79378528, + "step": 38000 + }, + { + "epoch": 6.199037441879436, + "eval_loss": 0.31492581963539124, + "eval_runtime": 156.1636, + "eval_samples_per_second": 17.45, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 79378528, + "step": 38000 + }, + { + "epoch": 6.199853169100253, + "grad_norm": 0.0013346612686291337, + "learning_rate": 0.001839384421428364, + "loss": 0.3349, + "num_input_tokens_seen": 79389168, + "step": 38005 + }, + { + "epoch": 6.20066889632107, + "grad_norm": 0.0007451685378327966, + "learning_rate": 0.0018301993732932065, + "loss": 0.297, + "num_input_tokens_seen": 79399120, + "step": 38010 + }, + { + "epoch": 6.201484623541887, + "grad_norm": 0.0007509324350394309, + "learning_rate": 0.0018210371748033248, + "loss": 0.3983, + "num_input_tokens_seen": 79411024, + "step": 38015 + }, + { + "epoch": 6.202300350762705, + "grad_norm": 0.0005155462422408164, + "learning_rate": 0.0018118978273716556, + "loss": 0.3166, + "num_input_tokens_seen": 79422688, + "step": 38020 + }, + { + "epoch": 6.203116077983522, + "grad_norm": 0.00046685096458531916, + "learning_rate": 0.001802781332407588, + "loss": 0.3411, + "num_input_tokens_seen": 79432832, + "step": 38025 + }, + { + "epoch": 6.2039318052043395, + "grad_norm": 0.0004795882268808782, + "learning_rate": 0.0017936876913169806, + "loss": 0.2671, + "num_input_tokens_seen": 79444064, + "step": 38030 + }, + { + "epoch": 6.204747532425157, + "grad_norm": 0.00044060873915441334, + "learning_rate": 0.0017846169055022287, + "loss": 0.2622, + "num_input_tokens_seen": 79453808, + "step": 38035 + }, + { + "epoch": 6.205563259645975, + "grad_norm": 0.00036746388650499284, + "learning_rate": 0.0017755689763621295, + "loss": 0.292, + "num_input_tokens_seen": 79463552, + "step": 38040 + }, + { + "epoch": 6.206378986866792, + "grad_norm": 0.0009520177845843136, + "learning_rate": 0.0017665439052920173, + "loss": 0.3383, + "num_input_tokens_seen": 79474240, + "step": 38045 + }, + { + "epoch": 6.207194714087609, + "grad_norm": 0.00113862962462008, + "learning_rate": 0.0017575416936836286, + "loss": 0.423, + "num_input_tokens_seen": 79484096, + "step": 38050 + }, + { + "epoch": 6.208010441308426, + "grad_norm": 0.00045988443889655173, + "learning_rate": 0.0017485623429252528, + "loss": 0.3632, + "num_input_tokens_seen": 79494608, + "step": 38055 + }, + { + "epoch": 6.208826168529244, + "grad_norm": 0.000931491726078093, + "learning_rate": 0.0017396058544016156, + "loss": 0.3042, + "num_input_tokens_seen": 79505360, + "step": 38060 + }, + { + "epoch": 6.209641895750061, + "grad_norm": 0.0007166760042309761, + "learning_rate": 0.0017306722294938958, + "loss": 0.3996, + "num_input_tokens_seen": 79515488, + "step": 38065 + }, + { + "epoch": 6.210457622970878, + "grad_norm": 0.0009049908840097487, + "learning_rate": 0.0017217614695798078, + "loss": 0.3501, + "num_input_tokens_seen": 79525840, + "step": 38070 + }, + { + "epoch": 6.211273350191696, + "grad_norm": 0.0006680923979729414, + "learning_rate": 0.001712873576033469, + "loss": 0.3324, + "num_input_tokens_seen": 79536112, + "step": 38075 + }, + { + "epoch": 6.2120890774125135, + "grad_norm": 0.0007798359147273004, + "learning_rate": 0.0017040085502255163, + "loss": 0.3564, + "num_input_tokens_seen": 79546720, + "step": 38080 + }, + { + "epoch": 6.212904804633331, + "grad_norm": 0.0007897571194916964, + "learning_rate": 0.0016951663935230565, + "loss": 0.3024, + "num_input_tokens_seen": 79557936, + "step": 38085 + }, + { + "epoch": 6.213720531854148, + "grad_norm": 0.0007054396555759013, + "learning_rate": 0.0016863471072896485, + "loss": 0.3243, + "num_input_tokens_seen": 79569584, + "step": 38090 + }, + { + "epoch": 6.214536259074965, + "grad_norm": 0.00043754573562182486, + "learning_rate": 0.0016775506928853377, + "loss": 0.3005, + "num_input_tokens_seen": 79579840, + "step": 38095 + }, + { + "epoch": 6.215351986295783, + "grad_norm": 0.0011712198611348867, + "learning_rate": 0.001668777151666656, + "loss": 0.3186, + "num_input_tokens_seen": 79590464, + "step": 38100 + }, + { + "epoch": 6.2161677135166, + "grad_norm": 0.0004733704845421016, + "learning_rate": 0.0016600264849865709, + "loss": 0.2774, + "num_input_tokens_seen": 79600848, + "step": 38105 + }, + { + "epoch": 6.216983440737417, + "grad_norm": 0.00044701495789922774, + "learning_rate": 0.0016512986941945695, + "loss": 0.3011, + "num_input_tokens_seen": 79611408, + "step": 38110 + }, + { + "epoch": 6.217799167958235, + "grad_norm": 0.0013623323757201433, + "learning_rate": 0.0016425937806365753, + "loss": 0.3639, + "num_input_tokens_seen": 79621056, + "step": 38115 + }, + { + "epoch": 6.218614895179052, + "grad_norm": 0.0004107379645574838, + "learning_rate": 0.0016339117456549979, + "loss": 0.3385, + "num_input_tokens_seen": 79631456, + "step": 38120 + }, + { + "epoch": 6.219430622399869, + "grad_norm": 0.0008590451907366514, + "learning_rate": 0.0016252525905886995, + "loss": 0.3575, + "num_input_tokens_seen": 79642448, + "step": 38125 + }, + { + "epoch": 6.220246349620687, + "grad_norm": 0.0004889736883342266, + "learning_rate": 0.0016166163167730617, + "loss": 0.2683, + "num_input_tokens_seen": 79652016, + "step": 38130 + }, + { + "epoch": 6.221062076841505, + "grad_norm": 0.0007602449622936547, + "learning_rate": 0.0016080029255398864, + "loss": 0.2852, + "num_input_tokens_seen": 79663264, + "step": 38135 + }, + { + "epoch": 6.221877804062322, + "grad_norm": 0.0005467255832627416, + "learning_rate": 0.0015994124182174606, + "loss": 0.2865, + "num_input_tokens_seen": 79673168, + "step": 38140 + }, + { + "epoch": 6.222693531283139, + "grad_norm": 0.000924167106859386, + "learning_rate": 0.001590844796130575, + "loss": 0.3063, + "num_input_tokens_seen": 79683728, + "step": 38145 + }, + { + "epoch": 6.223509258503956, + "grad_norm": 0.0006239272188395262, + "learning_rate": 0.001582300060600439, + "loss": 0.3197, + "num_input_tokens_seen": 79694672, + "step": 38150 + }, + { + "epoch": 6.224324985724774, + "grad_norm": 0.0009954475099220872, + "learning_rate": 0.0015737782129447652, + "loss": 0.3338, + "num_input_tokens_seen": 79705216, + "step": 38155 + }, + { + "epoch": 6.225140712945591, + "grad_norm": 0.0005872679757885635, + "learning_rate": 0.0015652792544777361, + "loss": 0.3004, + "num_input_tokens_seen": 79715984, + "step": 38160 + }, + { + "epoch": 6.225956440166408, + "grad_norm": 0.0007305988692678511, + "learning_rate": 0.0015568031865099863, + "loss": 0.309, + "num_input_tokens_seen": 79725904, + "step": 38165 + }, + { + "epoch": 6.226772167387225, + "grad_norm": 0.0008477731025777757, + "learning_rate": 0.0015483500103486369, + "loss": 0.32, + "num_input_tokens_seen": 79736944, + "step": 38170 + }, + { + "epoch": 6.227587894608043, + "grad_norm": 0.0004489447164814919, + "learning_rate": 0.0015399197272972787, + "loss": 0.2709, + "num_input_tokens_seen": 79748304, + "step": 38175 + }, + { + "epoch": 6.2284036218288605, + "grad_norm": 0.0005923401331529021, + "learning_rate": 0.0015315123386559714, + "loss": 0.3066, + "num_input_tokens_seen": 79759728, + "step": 38180 + }, + { + "epoch": 6.229219349049678, + "grad_norm": 0.0005939591210335493, + "learning_rate": 0.0015231278457212283, + "loss": 0.3025, + "num_input_tokens_seen": 79769776, + "step": 38185 + }, + { + "epoch": 6.230035076270495, + "grad_norm": 0.0007422372582368553, + "learning_rate": 0.001514766249786048, + "loss": 0.4076, + "num_input_tokens_seen": 79780208, + "step": 38190 + }, + { + "epoch": 6.230850803491313, + "grad_norm": 0.0005468311719596386, + "learning_rate": 0.0015064275521398994, + "loss": 0.3539, + "num_input_tokens_seen": 79791344, + "step": 38195 + }, + { + "epoch": 6.23166653071213, + "grad_norm": 0.0004992285976186395, + "learning_rate": 0.0014981117540686872, + "loss": 0.3555, + "num_input_tokens_seen": 79802112, + "step": 38200 + }, + { + "epoch": 6.23166653071213, + "eval_loss": 0.3149404525756836, + "eval_runtime": 156.1354, + "eval_samples_per_second": 17.453, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 79802112, + "step": 38200 + }, + { + "epoch": 6.232482257932947, + "grad_norm": 0.000658057106193155, + "learning_rate": 0.0014898188568548687, + "loss": 0.2917, + "num_input_tokens_seen": 79812512, + "step": 38205 + }, + { + "epoch": 6.233297985153764, + "grad_norm": 0.0007559819496236742, + "learning_rate": 0.0014815488617772542, + "loss": 0.318, + "num_input_tokens_seen": 79823568, + "step": 38210 + }, + { + "epoch": 6.234113712374582, + "grad_norm": 0.0006593858706764877, + "learning_rate": 0.0014733017701112072, + "loss": 0.3012, + "num_input_tokens_seen": 79834336, + "step": 38215 + }, + { + "epoch": 6.234929439595399, + "grad_norm": 0.0008905697613954544, + "learning_rate": 0.0014650775831285435, + "loss": 0.2856, + "num_input_tokens_seen": 79844960, + "step": 38220 + }, + { + "epoch": 6.2357451668162165, + "grad_norm": 0.0008744321530684829, + "learning_rate": 0.001456876302097515, + "loss": 0.3149, + "num_input_tokens_seen": 79855664, + "step": 38225 + }, + { + "epoch": 6.236560894037034, + "grad_norm": 0.0010950168361887336, + "learning_rate": 0.0014486979282828604, + "loss": 0.3913, + "num_input_tokens_seen": 79866848, + "step": 38230 + }, + { + "epoch": 6.237376621257852, + "grad_norm": 0.0008838390349410474, + "learning_rate": 0.001440542462945804, + "loss": 0.3332, + "num_input_tokens_seen": 79877536, + "step": 38235 + }, + { + "epoch": 6.238192348478669, + "grad_norm": 0.0011399427894502878, + "learning_rate": 0.0014324099073440232, + "loss": 0.3158, + "num_input_tokens_seen": 79887696, + "step": 38240 + }, + { + "epoch": 6.239008075699486, + "grad_norm": 0.0006294588674791157, + "learning_rate": 0.0014243002627316482, + "loss": 0.2791, + "num_input_tokens_seen": 79897888, + "step": 38245 + }, + { + "epoch": 6.239823802920303, + "grad_norm": 0.0005365620017983019, + "learning_rate": 0.0014162135303592781, + "loss": 0.2904, + "num_input_tokens_seen": 79908800, + "step": 38250 + }, + { + "epoch": 6.240639530141121, + "grad_norm": 0.000672456226311624, + "learning_rate": 0.001408149711474016, + "loss": 0.3273, + "num_input_tokens_seen": 79919232, + "step": 38255 + }, + { + "epoch": 6.241455257361938, + "grad_norm": 0.0009037895360961556, + "learning_rate": 0.0014001088073193834, + "loss": 0.3855, + "num_input_tokens_seen": 79930480, + "step": 38260 + }, + { + "epoch": 6.242270984582755, + "grad_norm": 0.0005187397473491728, + "learning_rate": 0.0013920908191354052, + "loss": 0.32, + "num_input_tokens_seen": 79940784, + "step": 38265 + }, + { + "epoch": 6.243086711803572, + "grad_norm": 0.0005998917622491717, + "learning_rate": 0.001384095748158526, + "loss": 0.3145, + "num_input_tokens_seen": 79950208, + "step": 38270 + }, + { + "epoch": 6.2439024390243905, + "grad_norm": 0.00042927125468850136, + "learning_rate": 0.0013761235956217255, + "loss": 0.2405, + "num_input_tokens_seen": 79960944, + "step": 38275 + }, + { + "epoch": 6.244718166245208, + "grad_norm": 0.0007045553647913039, + "learning_rate": 0.0013681743627543873, + "loss": 0.3117, + "num_input_tokens_seen": 79970528, + "step": 38280 + }, + { + "epoch": 6.245533893466025, + "grad_norm": 0.0013051205314695835, + "learning_rate": 0.001360248050782381, + "loss": 0.4206, + "num_input_tokens_seen": 79979456, + "step": 38285 + }, + { + "epoch": 6.246349620686843, + "grad_norm": 0.0007331764209084213, + "learning_rate": 0.001352344660928062, + "loss": 0.2607, + "num_input_tokens_seen": 79990864, + "step": 38290 + }, + { + "epoch": 6.24716534790766, + "grad_norm": 0.0006755173671990633, + "learning_rate": 0.0013444641944102052, + "loss": 0.2307, + "num_input_tokens_seen": 80001856, + "step": 38295 + }, + { + "epoch": 6.247981075128477, + "grad_norm": 0.0003786296001635492, + "learning_rate": 0.0013366066524441056, + "loss": 0.3339, + "num_input_tokens_seen": 80013392, + "step": 38300 + }, + { + "epoch": 6.248796802349294, + "grad_norm": 0.0008950642077252269, + "learning_rate": 0.0013287720362414768, + "loss": 0.3334, + "num_input_tokens_seen": 80024736, + "step": 38305 + }, + { + "epoch": 6.249612529570112, + "grad_norm": 0.0008242495241574943, + "learning_rate": 0.0013209603470105025, + "loss": 0.3308, + "num_input_tokens_seen": 80034800, + "step": 38310 + }, + { + "epoch": 6.250428256790929, + "grad_norm": 0.0009113829000853002, + "learning_rate": 0.0013131715859558857, + "loss": 0.3004, + "num_input_tokens_seen": 80044960, + "step": 38315 + }, + { + "epoch": 6.251243984011746, + "grad_norm": 0.0006671666633337736, + "learning_rate": 0.001305405754278699, + "loss": 0.2825, + "num_input_tokens_seen": 80056800, + "step": 38320 + }, + { + "epoch": 6.2520597112325635, + "grad_norm": 0.0004645452427212149, + "learning_rate": 0.0012976628531765843, + "loss": 0.3335, + "num_input_tokens_seen": 80067328, + "step": 38325 + }, + { + "epoch": 6.252875438453382, + "grad_norm": 0.0005919227842241526, + "learning_rate": 0.0012899428838435533, + "loss": 0.3065, + "num_input_tokens_seen": 80077776, + "step": 38330 + }, + { + "epoch": 6.253691165674199, + "grad_norm": 0.0007678791880607605, + "learning_rate": 0.001282245847470137, + "loss": 0.3418, + "num_input_tokens_seen": 80086320, + "step": 38335 + }, + { + "epoch": 6.254506892895016, + "grad_norm": 0.0010621537221595645, + "learning_rate": 0.001274571745243319, + "loss": 0.3007, + "num_input_tokens_seen": 80097920, + "step": 38340 + }, + { + "epoch": 6.255322620115833, + "grad_norm": 0.0005188320646993816, + "learning_rate": 0.0012669205783465364, + "loss": 0.3674, + "num_input_tokens_seen": 80110304, + "step": 38345 + }, + { + "epoch": 6.256138347336651, + "grad_norm": 0.0005231964751146734, + "learning_rate": 0.001259292347959695, + "loss": 0.2815, + "num_input_tokens_seen": 80120256, + "step": 38350 + }, + { + "epoch": 6.256954074557468, + "grad_norm": 0.00037996817263774574, + "learning_rate": 0.0012516870552591707, + "loss": 0.3106, + "num_input_tokens_seen": 80131072, + "step": 38355 + }, + { + "epoch": 6.257769801778285, + "grad_norm": 0.0006179750780574977, + "learning_rate": 0.001244104701417792, + "loss": 0.2868, + "num_input_tokens_seen": 80142256, + "step": 38360 + }, + { + "epoch": 6.258585528999102, + "grad_norm": 0.0005382677190937102, + "learning_rate": 0.0012365452876048565, + "loss": 0.323, + "num_input_tokens_seen": 80152928, + "step": 38365 + }, + { + "epoch": 6.25940125621992, + "grad_norm": 0.0005021754186600447, + "learning_rate": 0.001229008814986099, + "loss": 0.2888, + "num_input_tokens_seen": 80163472, + "step": 38370 + }, + { + "epoch": 6.2602169834407375, + "grad_norm": 0.0008080568513832986, + "learning_rate": 0.0012214952847237725, + "loss": 0.3648, + "num_input_tokens_seen": 80174640, + "step": 38375 + }, + { + "epoch": 6.261032710661555, + "grad_norm": 0.0008156350813806057, + "learning_rate": 0.0012140046979765339, + "loss": 0.3333, + "num_input_tokens_seen": 80184272, + "step": 38380 + }, + { + "epoch": 6.261848437882372, + "grad_norm": 0.0005855854833498597, + "learning_rate": 0.0012065370558995258, + "loss": 0.3304, + "num_input_tokens_seen": 80196304, + "step": 38385 + }, + { + "epoch": 6.26266416510319, + "grad_norm": 0.0005717053427360952, + "learning_rate": 0.0011990923596443602, + "loss": 0.299, + "num_input_tokens_seen": 80206544, + "step": 38390 + }, + { + "epoch": 6.263479892324007, + "grad_norm": 0.000800871814135462, + "learning_rate": 0.001191670610359119, + "loss": 0.3278, + "num_input_tokens_seen": 80217744, + "step": 38395 + }, + { + "epoch": 6.264295619544824, + "grad_norm": 0.0009885503677651286, + "learning_rate": 0.0011842718091882865, + "loss": 0.2865, + "num_input_tokens_seen": 80229344, + "step": 38400 + }, + { + "epoch": 6.264295619544824, + "eval_loss": 0.31514087319374084, + "eval_runtime": 155.9027, + "eval_samples_per_second": 17.479, + "eval_steps_per_second": 8.743, + "num_input_tokens_seen": 80229344, + "step": 38400 + }, + { + "epoch": 6.265111346765641, + "grad_norm": 0.0009502753964625299, + "learning_rate": 0.0011768959572729, + "loss": 0.3171, + "num_input_tokens_seen": 80238736, + "step": 38405 + }, + { + "epoch": 6.265927073986459, + "grad_norm": 0.0005187371280044317, + "learning_rate": 0.001169543055750366, + "loss": 0.2718, + "num_input_tokens_seen": 80248368, + "step": 38410 + }, + { + "epoch": 6.266742801207276, + "grad_norm": 0.0005628484068438411, + "learning_rate": 0.0011622131057546115, + "loss": 0.3424, + "num_input_tokens_seen": 80259360, + "step": 38415 + }, + { + "epoch": 6.2675585284280935, + "grad_norm": 0.001130969263613224, + "learning_rate": 0.0011549061084160316, + "loss": 0.378, + "num_input_tokens_seen": 80270256, + "step": 38420 + }, + { + "epoch": 6.268374255648911, + "grad_norm": 0.0003641486109700054, + "learning_rate": 0.0011476220648614088, + "loss": 0.282, + "num_input_tokens_seen": 80280272, + "step": 38425 + }, + { + "epoch": 6.269189982869729, + "grad_norm": 0.000525073439348489, + "learning_rate": 0.0011403609762140777, + "loss": 0.3424, + "num_input_tokens_seen": 80290000, + "step": 38430 + }, + { + "epoch": 6.270005710090546, + "grad_norm": 0.0008562714792788029, + "learning_rate": 0.0011331228435937756, + "loss": 0.3221, + "num_input_tokens_seen": 80301968, + "step": 38435 + }, + { + "epoch": 6.270821437311363, + "grad_norm": 0.0005080850096419454, + "learning_rate": 0.0011259076681166935, + "loss": 0.3337, + "num_input_tokens_seen": 80311936, + "step": 38440 + }, + { + "epoch": 6.27163716453218, + "grad_norm": 0.0006301525863818824, + "learning_rate": 0.0011187154508955244, + "loss": 0.3018, + "num_input_tokens_seen": 80322560, + "step": 38445 + }, + { + "epoch": 6.272452891752998, + "grad_norm": 0.0012037730775773525, + "learning_rate": 0.001111546193039381, + "loss": 0.3176, + "num_input_tokens_seen": 80333280, + "step": 38450 + }, + { + "epoch": 6.273268618973815, + "grad_norm": 0.0003987706149928272, + "learning_rate": 0.0011043998956538792, + "loss": 0.2684, + "num_input_tokens_seen": 80342384, + "step": 38455 + }, + { + "epoch": 6.274084346194632, + "grad_norm": 0.000654694449622184, + "learning_rate": 0.0010972765598410538, + "loss": 0.3075, + "num_input_tokens_seen": 80352576, + "step": 38460 + }, + { + "epoch": 6.27490007341545, + "grad_norm": 0.0004976459313184023, + "learning_rate": 0.0010901761866993931, + "loss": 0.2622, + "num_input_tokens_seen": 80361920, + "step": 38465 + }, + { + "epoch": 6.275715800636267, + "grad_norm": 0.0006987896631471813, + "learning_rate": 0.0010830987773238876, + "loss": 0.3589, + "num_input_tokens_seen": 80372112, + "step": 38470 + }, + { + "epoch": 6.276531527857085, + "grad_norm": 0.000591026502661407, + "learning_rate": 0.0010760443328059644, + "loss": 0.3378, + "num_input_tokens_seen": 80382368, + "step": 38475 + }, + { + "epoch": 6.277347255077902, + "grad_norm": 0.00034667577710933983, + "learning_rate": 0.001069012854233503, + "loss": 0.2731, + "num_input_tokens_seen": 80391328, + "step": 38480 + }, + { + "epoch": 6.27816298229872, + "grad_norm": 0.0006590319098904729, + "learning_rate": 0.0010620043426908365, + "loss": 0.2951, + "num_input_tokens_seen": 80402800, + "step": 38485 + }, + { + "epoch": 6.278978709519537, + "grad_norm": 0.0007984111434780061, + "learning_rate": 0.0010550187992587833, + "loss": 0.3352, + "num_input_tokens_seen": 80413136, + "step": 38490 + }, + { + "epoch": 6.279794436740354, + "grad_norm": 0.0007474272861145437, + "learning_rate": 0.0010480562250145653, + "loss": 0.2968, + "num_input_tokens_seen": 80423584, + "step": 38495 + }, + { + "epoch": 6.280610163961171, + "grad_norm": 0.00040962855564430356, + "learning_rate": 0.0010411166210319567, + "loss": 0.3023, + "num_input_tokens_seen": 80435376, + "step": 38500 + }, + { + "epoch": 6.281425891181989, + "grad_norm": 0.00047710788203403354, + "learning_rate": 0.0010341999883810848, + "loss": 0.2931, + "num_input_tokens_seen": 80445904, + "step": 38505 + }, + { + "epoch": 6.282241618402806, + "grad_norm": 0.0009343388373963535, + "learning_rate": 0.0010273063281285965, + "loss": 0.3181, + "num_input_tokens_seen": 80455344, + "step": 38510 + }, + { + "epoch": 6.283057345623623, + "grad_norm": 0.0007229306502267718, + "learning_rate": 0.0010204356413375747, + "loss": 0.3106, + "num_input_tokens_seen": 80466160, + "step": 38515 + }, + { + "epoch": 6.2838730728444405, + "grad_norm": 0.00042712505091913044, + "learning_rate": 0.001013587929067572, + "loss": 0.2914, + "num_input_tokens_seen": 80476128, + "step": 38520 + }, + { + "epoch": 6.2846888000652585, + "grad_norm": 0.00033280954812653363, + "learning_rate": 0.00100676319237461, + "loss": 0.2746, + "num_input_tokens_seen": 80485872, + "step": 38525 + }, + { + "epoch": 6.285504527286076, + "grad_norm": 0.0010057920590043068, + "learning_rate": 0.0009999614323110972, + "loss": 0.285, + "num_input_tokens_seen": 80496016, + "step": 38530 + }, + { + "epoch": 6.286320254506893, + "grad_norm": 0.0008515917579643428, + "learning_rate": 0.000993182649926011, + "loss": 0.3114, + "num_input_tokens_seen": 80505536, + "step": 38535 + }, + { + "epoch": 6.28713598172771, + "grad_norm": 0.0004842259222641587, + "learning_rate": 0.000986426846264682, + "loss": 0.2662, + "num_input_tokens_seen": 80515776, + "step": 38540 + }, + { + "epoch": 6.287951708948528, + "grad_norm": 0.00036734298919327557, + "learning_rate": 0.00097969402236896, + "loss": 0.3591, + "num_input_tokens_seen": 80526752, + "step": 38545 + }, + { + "epoch": 6.288767436169345, + "grad_norm": 0.0005710048135370016, + "learning_rate": 0.0009729841792771143, + "loss": 0.3495, + "num_input_tokens_seen": 80536496, + "step": 38550 + }, + { + "epoch": 6.289583163390162, + "grad_norm": 0.0005679423338733613, + "learning_rate": 0.0009662973180239176, + "loss": 0.3291, + "num_input_tokens_seen": 80547424, + "step": 38555 + }, + { + "epoch": 6.290398890610979, + "grad_norm": 0.0004829487297683954, + "learning_rate": 0.0009596334396405448, + "loss": 0.3274, + "num_input_tokens_seen": 80557984, + "step": 38560 + }, + { + "epoch": 6.291214617831797, + "grad_norm": 0.00128430244512856, + "learning_rate": 0.0009529925451546406, + "loss": 0.2572, + "num_input_tokens_seen": 80569104, + "step": 38565 + }, + { + "epoch": 6.2920303450526145, + "grad_norm": 0.0007068493869155645, + "learning_rate": 0.0009463746355903357, + "loss": 0.2969, + "num_input_tokens_seen": 80579328, + "step": 38570 + }, + { + "epoch": 6.292846072273432, + "grad_norm": 0.0009408715413883328, + "learning_rate": 0.0009397797119681971, + "loss": 0.3459, + "num_input_tokens_seen": 80589808, + "step": 38575 + }, + { + "epoch": 6.293661799494249, + "grad_norm": 0.0008675414137542248, + "learning_rate": 0.0009332077753052281, + "loss": 0.2681, + "num_input_tokens_seen": 80599456, + "step": 38580 + }, + { + "epoch": 6.294477526715067, + "grad_norm": 0.0006956875440664589, + "learning_rate": 0.0009266588266149011, + "loss": 0.2832, + "num_input_tokens_seen": 80609312, + "step": 38585 + }, + { + "epoch": 6.295293253935884, + "grad_norm": 0.0005069420440122485, + "learning_rate": 0.0009201328669071584, + "loss": 0.3733, + "num_input_tokens_seen": 80619904, + "step": 38590 + }, + { + "epoch": 6.296108981156701, + "grad_norm": 0.000467701320303604, + "learning_rate": 0.0009136298971883949, + "loss": 0.3459, + "num_input_tokens_seen": 80632800, + "step": 38595 + }, + { + "epoch": 6.296924708377518, + "grad_norm": 0.0015474656829610467, + "learning_rate": 0.0009071499184614251, + "loss": 0.3473, + "num_input_tokens_seen": 80643632, + "step": 38600 + }, + { + "epoch": 6.296924708377518, + "eval_loss": 0.31544286012649536, + "eval_runtime": 155.8204, + "eval_samples_per_second": 17.488, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 80643632, + "step": 38600 + }, + { + "epoch": 6.297740435598336, + "grad_norm": 0.0007100090733729303, + "learning_rate": 0.0009006929317255663, + "loss": 0.278, + "num_input_tokens_seen": 80653280, + "step": 38605 + }, + { + "epoch": 6.298556162819153, + "grad_norm": 0.000592104101087898, + "learning_rate": 0.0008942589379765387, + "loss": 0.3067, + "num_input_tokens_seen": 80664608, + "step": 38610 + }, + { + "epoch": 6.2993718900399704, + "grad_norm": 0.0006602386711165309, + "learning_rate": 0.0008878479382065817, + "loss": 0.3033, + "num_input_tokens_seen": 80675408, + "step": 38615 + }, + { + "epoch": 6.300187617260788, + "grad_norm": 0.000669284607283771, + "learning_rate": 0.0008814599334043215, + "loss": 0.3101, + "num_input_tokens_seen": 80686720, + "step": 38620 + }, + { + "epoch": 6.301003344481606, + "grad_norm": 0.000492060265969485, + "learning_rate": 0.0008750949245548866, + "loss": 0.2964, + "num_input_tokens_seen": 80696592, + "step": 38625 + }, + { + "epoch": 6.301819071702423, + "grad_norm": 0.0007082542288117111, + "learning_rate": 0.0008687529126398252, + "loss": 0.2525, + "num_input_tokens_seen": 80707376, + "step": 38630 + }, + { + "epoch": 6.30263479892324, + "grad_norm": 0.0012334297643974423, + "learning_rate": 0.0008624338986371715, + "loss": 0.2788, + "num_input_tokens_seen": 80717056, + "step": 38635 + }, + { + "epoch": 6.303450526144058, + "grad_norm": 0.0005633212276734412, + "learning_rate": 0.0008561378835213962, + "loss": 0.2757, + "num_input_tokens_seen": 80727728, + "step": 38640 + }, + { + "epoch": 6.304266253364875, + "grad_norm": 0.0007808125228621066, + "learning_rate": 0.0008498648682634058, + "loss": 0.3414, + "num_input_tokens_seen": 80736944, + "step": 38645 + }, + { + "epoch": 6.305081980585692, + "grad_norm": 0.0012203236110508442, + "learning_rate": 0.0008436148538306099, + "loss": 0.2776, + "num_input_tokens_seen": 80747520, + "step": 38650 + }, + { + "epoch": 6.305897707806509, + "grad_norm": 0.00168227672111243, + "learning_rate": 0.0008373878411868041, + "loss": 0.4042, + "num_input_tokens_seen": 80757600, + "step": 38655 + }, + { + "epoch": 6.306713435027326, + "grad_norm": 0.00032976994407363236, + "learning_rate": 0.000831183831292287, + "loss": 0.2695, + "num_input_tokens_seen": 80767072, + "step": 38660 + }, + { + "epoch": 6.307529162248144, + "grad_norm": 0.0007627608138136566, + "learning_rate": 0.0008250028251037933, + "loss": 0.3739, + "num_input_tokens_seen": 80777776, + "step": 38665 + }, + { + "epoch": 6.3083448894689615, + "grad_norm": 0.0006807459285482764, + "learning_rate": 0.0008188448235745271, + "loss": 0.2991, + "num_input_tokens_seen": 80787712, + "step": 38670 + }, + { + "epoch": 6.309160616689779, + "grad_norm": 0.0004983011749573052, + "learning_rate": 0.0008127098276541122, + "loss": 0.2551, + "num_input_tokens_seen": 80798736, + "step": 38675 + }, + { + "epoch": 6.309976343910597, + "grad_norm": 0.00039130024379119277, + "learning_rate": 0.0008065978382886418, + "loss": 0.3275, + "num_input_tokens_seen": 80809040, + "step": 38680 + }, + { + "epoch": 6.310792071131414, + "grad_norm": 0.0010510891443118453, + "learning_rate": 0.0008005088564206785, + "loss": 0.3852, + "num_input_tokens_seen": 80818320, + "step": 38685 + }, + { + "epoch": 6.311607798352231, + "grad_norm": 0.0005103163421154022, + "learning_rate": 0.0007944428829891881, + "loss": 0.2754, + "num_input_tokens_seen": 80828656, + "step": 38690 + }, + { + "epoch": 6.312423525573048, + "grad_norm": 0.0007164290873333812, + "learning_rate": 0.0007883999189296386, + "loss": 0.3211, + "num_input_tokens_seen": 80839120, + "step": 38695 + }, + { + "epoch": 6.313239252793866, + "grad_norm": 0.0006502043106593192, + "learning_rate": 0.0007823799651739515, + "loss": 0.3226, + "num_input_tokens_seen": 80850912, + "step": 38700 + }, + { + "epoch": 6.314054980014683, + "grad_norm": 0.0007116328342817724, + "learning_rate": 0.0007763830226504509, + "loss": 0.2973, + "num_input_tokens_seen": 80860240, + "step": 38705 + }, + { + "epoch": 6.3148707072355, + "grad_norm": 0.0007214891957119107, + "learning_rate": 0.0007704090922839468, + "loss": 0.3161, + "num_input_tokens_seen": 80868112, + "step": 38710 + }, + { + "epoch": 6.3156864344563175, + "grad_norm": 0.0006229957798495889, + "learning_rate": 0.0007644581749957025, + "loss": 0.3027, + "num_input_tokens_seen": 80878592, + "step": 38715 + }, + { + "epoch": 6.3165021616771355, + "grad_norm": 0.0006408337503671646, + "learning_rate": 0.000758530271703417, + "loss": 0.2517, + "num_input_tokens_seen": 80887792, + "step": 38720 + }, + { + "epoch": 6.317317888897953, + "grad_norm": 0.0004782795440405607, + "learning_rate": 0.0007526253833212426, + "loss": 0.3528, + "num_input_tokens_seen": 80898208, + "step": 38725 + }, + { + "epoch": 6.31813361611877, + "grad_norm": 0.0004764498444274068, + "learning_rate": 0.0007467435107598008, + "loss": 0.3071, + "num_input_tokens_seen": 80908928, + "step": 38730 + }, + { + "epoch": 6.318949343339587, + "grad_norm": 0.0007834061980247498, + "learning_rate": 0.0007408846549261328, + "loss": 0.3, + "num_input_tokens_seen": 80919424, + "step": 38735 + }, + { + "epoch": 6.319765070560405, + "grad_norm": 0.0003942474431823939, + "learning_rate": 0.0007350488167237656, + "loss": 0.2875, + "num_input_tokens_seen": 80930128, + "step": 38740 + }, + { + "epoch": 6.320580797781222, + "grad_norm": 0.0006534642889164388, + "learning_rate": 0.0007292359970526629, + "loss": 0.2715, + "num_input_tokens_seen": 80940208, + "step": 38745 + }, + { + "epoch": 6.321396525002039, + "grad_norm": 0.00039196800207719207, + "learning_rate": 0.0007234461968092076, + "loss": 0.262, + "num_input_tokens_seen": 80949808, + "step": 38750 + }, + { + "epoch": 6.322212252222856, + "grad_norm": 0.0009444701136089861, + "learning_rate": 0.0007176794168862854, + "loss": 0.4181, + "num_input_tokens_seen": 80960672, + "step": 38755 + }, + { + "epoch": 6.323027979443674, + "grad_norm": 0.0005190189112909138, + "learning_rate": 0.000711935658173185, + "loss": 0.2933, + "num_input_tokens_seen": 80971040, + "step": 38760 + }, + { + "epoch": 6.3238437066644915, + "grad_norm": 0.0008812563610263169, + "learning_rate": 0.0007062149215556812, + "loss": 0.2874, + "num_input_tokens_seen": 80980080, + "step": 38765 + }, + { + "epoch": 6.324659433885309, + "grad_norm": 0.001110413228161633, + "learning_rate": 0.0007005172079159849, + "loss": 0.3513, + "num_input_tokens_seen": 80990080, + "step": 38770 + }, + { + "epoch": 6.325475161106126, + "grad_norm": 0.000612541800364852, + "learning_rate": 0.0006948425181327267, + "loss": 0.271, + "num_input_tokens_seen": 81000416, + "step": 38775 + }, + { + "epoch": 6.326290888326944, + "grad_norm": 0.001445271191187203, + "learning_rate": 0.000689190853081073, + "loss": 0.3141, + "num_input_tokens_seen": 81011376, + "step": 38780 + }, + { + "epoch": 6.327106615547761, + "grad_norm": 0.0005050156614743173, + "learning_rate": 0.000683562213632527, + "loss": 0.2923, + "num_input_tokens_seen": 81021760, + "step": 38785 + }, + { + "epoch": 6.327922342768578, + "grad_norm": 0.0007200168329291046, + "learning_rate": 0.0006779566006551108, + "loss": 0.2843, + "num_input_tokens_seen": 81032352, + "step": 38790 + }, + { + "epoch": 6.328738069989395, + "grad_norm": 0.0006159144104458392, + "learning_rate": 0.0006723740150132995, + "loss": 0.3632, + "num_input_tokens_seen": 81042832, + "step": 38795 + }, + { + "epoch": 6.329553797210213, + "grad_norm": 0.0005485979490913451, + "learning_rate": 0.0006668144575679713, + "loss": 0.2548, + "num_input_tokens_seen": 81051936, + "step": 38800 + }, + { + "epoch": 6.329553797210213, + "eval_loss": 0.31503960490226746, + "eval_runtime": 155.8189, + "eval_samples_per_second": 17.488, + "eval_steps_per_second": 8.747, + "num_input_tokens_seen": 81051936, + "step": 38800 + }, + { + "epoch": 6.33036952443103, + "grad_norm": 0.0007295549148693681, + "learning_rate": 0.0006612779291765069, + "loss": 0.3252, + "num_input_tokens_seen": 81062848, + "step": 38805 + }, + { + "epoch": 6.331185251651847, + "grad_norm": 0.0008849591249600053, + "learning_rate": 0.0006557644306926736, + "loss": 0.3688, + "num_input_tokens_seen": 81072736, + "step": 38810 + }, + { + "epoch": 6.332000978872665, + "grad_norm": 0.0007160521927289665, + "learning_rate": 0.0006502739629667575, + "loss": 0.3351, + "num_input_tokens_seen": 81083248, + "step": 38815 + }, + { + "epoch": 6.332816706093483, + "grad_norm": 0.0011200662702322006, + "learning_rate": 0.0006448065268454317, + "loss": 0.3899, + "num_input_tokens_seen": 81093264, + "step": 38820 + }, + { + "epoch": 6.3336324333143, + "grad_norm": 0.0007347761420533061, + "learning_rate": 0.0006393621231718549, + "loss": 0.3328, + "num_input_tokens_seen": 81105168, + "step": 38825 + }, + { + "epoch": 6.334448160535117, + "grad_norm": 0.0012763963313773274, + "learning_rate": 0.0006339407527856389, + "loss": 0.3387, + "num_input_tokens_seen": 81116016, + "step": 38830 + }, + { + "epoch": 6.335263887755934, + "grad_norm": 0.0004856048908550292, + "learning_rate": 0.0006285424165227982, + "loss": 0.2833, + "num_input_tokens_seen": 81128192, + "step": 38835 + }, + { + "epoch": 6.336079614976752, + "grad_norm": 0.0007053075823932886, + "learning_rate": 0.0006231671152158169, + "loss": 0.3319, + "num_input_tokens_seen": 81138640, + "step": 38840 + }, + { + "epoch": 6.336895342197569, + "grad_norm": 0.000907109584659338, + "learning_rate": 0.0006178148496936819, + "loss": 0.2777, + "num_input_tokens_seen": 81150960, + "step": 38845 + }, + { + "epoch": 6.337711069418386, + "grad_norm": 0.0012054290855303407, + "learning_rate": 0.000612485620781733, + "loss": 0.2951, + "num_input_tokens_seen": 81162768, + "step": 38850 + }, + { + "epoch": 6.338526796639204, + "grad_norm": 0.000728236569557339, + "learning_rate": 0.0006071794293018296, + "loss": 0.3111, + "num_input_tokens_seen": 81173856, + "step": 38855 + }, + { + "epoch": 6.339342523860021, + "grad_norm": 0.0005419243243522942, + "learning_rate": 0.0006018962760722501, + "loss": 0.2748, + "num_input_tokens_seen": 81185152, + "step": 38860 + }, + { + "epoch": 6.3401582510808385, + "grad_norm": 0.000557577412109822, + "learning_rate": 0.0005966361619077098, + "loss": 0.2586, + "num_input_tokens_seen": 81196720, + "step": 38865 + }, + { + "epoch": 6.340973978301656, + "grad_norm": 0.0007134695188142359, + "learning_rate": 0.000591399087619393, + "loss": 0.3501, + "num_input_tokens_seen": 81207584, + "step": 38870 + }, + { + "epoch": 6.341789705522474, + "grad_norm": 0.000525235605891794, + "learning_rate": 0.0005861850540149371, + "loss": 0.3382, + "num_input_tokens_seen": 81217744, + "step": 38875 + }, + { + "epoch": 6.342605432743291, + "grad_norm": 0.0005059729446657002, + "learning_rate": 0.0005809940618983822, + "loss": 0.3487, + "num_input_tokens_seen": 81226160, + "step": 38880 + }, + { + "epoch": 6.343421159964108, + "grad_norm": 0.0006261159433051944, + "learning_rate": 0.0005758261120702712, + "loss": 0.2774, + "num_input_tokens_seen": 81237264, + "step": 38885 + }, + { + "epoch": 6.344236887184925, + "grad_norm": 0.00047389703104272485, + "learning_rate": 0.0005706812053275501, + "loss": 0.339, + "num_input_tokens_seen": 81247472, + "step": 38890 + }, + { + "epoch": 6.345052614405743, + "grad_norm": 0.0005021836259402335, + "learning_rate": 0.0005655593424636173, + "loss": 0.28, + "num_input_tokens_seen": 81259264, + "step": 38895 + }, + { + "epoch": 6.34586834162656, + "grad_norm": 0.0004018813488073647, + "learning_rate": 0.0005604605242683746, + "loss": 0.3266, + "num_input_tokens_seen": 81269632, + "step": 38900 + }, + { + "epoch": 6.346684068847377, + "grad_norm": 0.0005225669010542333, + "learning_rate": 0.0005553847515280596, + "loss": 0.2775, + "num_input_tokens_seen": 81280368, + "step": 38905 + }, + { + "epoch": 6.3474997960681945, + "grad_norm": 0.0004088550340384245, + "learning_rate": 0.0005503320250254795, + "loss": 0.2787, + "num_input_tokens_seen": 81291488, + "step": 38910 + }, + { + "epoch": 6.3483155232890125, + "grad_norm": 0.0004670937778428197, + "learning_rate": 0.0005453023455397943, + "loss": 0.32, + "num_input_tokens_seen": 81300912, + "step": 38915 + }, + { + "epoch": 6.34913125050983, + "grad_norm": 0.0007702261791564524, + "learning_rate": 0.0005402957138466502, + "loss": 0.2786, + "num_input_tokens_seen": 81310944, + "step": 38920 + }, + { + "epoch": 6.349946977730647, + "grad_norm": 0.0005088961333967745, + "learning_rate": 0.0005353121307181463, + "loss": 0.2937, + "num_input_tokens_seen": 81321136, + "step": 38925 + }, + { + "epoch": 6.350762704951464, + "grad_norm": 0.0005238084122538567, + "learning_rate": 0.0005303515969227845, + "loss": 0.2825, + "num_input_tokens_seen": 81332144, + "step": 38930 + }, + { + "epoch": 6.351578432172282, + "grad_norm": 0.0005492718191817403, + "learning_rate": 0.0005254141132255862, + "loss": 0.3027, + "num_input_tokens_seen": 81343024, + "step": 38935 + }, + { + "epoch": 6.352394159393099, + "grad_norm": 0.0008215358830057085, + "learning_rate": 0.0005204996803879258, + "loss": 0.2871, + "num_input_tokens_seen": 81354464, + "step": 38940 + }, + { + "epoch": 6.353209886613916, + "grad_norm": 0.0008283748175017536, + "learning_rate": 0.0005156082991676969, + "loss": 0.2963, + "num_input_tokens_seen": 81365216, + "step": 38945 + }, + { + "epoch": 6.354025613834733, + "grad_norm": 0.0005027849110774696, + "learning_rate": 0.0005107399703192127, + "loss": 0.3675, + "num_input_tokens_seen": 81374624, + "step": 38950 + }, + { + "epoch": 6.354841341055551, + "grad_norm": 0.0005232151597738266, + "learning_rate": 0.0005058946945932063, + "loss": 0.3131, + "num_input_tokens_seen": 81384048, + "step": 38955 + }, + { + "epoch": 6.3556570682763684, + "grad_norm": 0.0004603637498803437, + "learning_rate": 0.0005010724727369131, + "loss": 0.3202, + "num_input_tokens_seen": 81393536, + "step": 38960 + }, + { + "epoch": 6.356472795497186, + "grad_norm": 0.0010347275529056787, + "learning_rate": 0.000496273305493955, + "loss": 0.3296, + "num_input_tokens_seen": 81404128, + "step": 38965 + }, + { + "epoch": 6.357288522718003, + "grad_norm": 0.000775116728618741, + "learning_rate": 0.0004914971936044399, + "loss": 0.3453, + "num_input_tokens_seen": 81415680, + "step": 38970 + }, + { + "epoch": 6.358104249938821, + "grad_norm": 0.0004931698786094785, + "learning_rate": 0.00048674413780491196, + "loss": 0.2764, + "num_input_tokens_seen": 81425568, + "step": 38975 + }, + { + "epoch": 6.358919977159638, + "grad_norm": 0.001099402317777276, + "learning_rate": 0.0004820141388283183, + "loss": 0.4094, + "num_input_tokens_seen": 81434080, + "step": 38980 + }, + { + "epoch": 6.359735704380455, + "grad_norm": 0.001270241686142981, + "learning_rate": 0.00047730719740410874, + "loss": 0.3475, + "num_input_tokens_seen": 81445184, + "step": 38985 + }, + { + "epoch": 6.360551431601272, + "grad_norm": 0.0010489001870155334, + "learning_rate": 0.00047262331425816927, + "loss": 0.3154, + "num_input_tokens_seen": 81456352, + "step": 38990 + }, + { + "epoch": 6.36136715882209, + "grad_norm": 0.0006498139700852334, + "learning_rate": 0.00046796249011277213, + "loss": 0.3111, + "num_input_tokens_seen": 81466752, + "step": 38995 + }, + { + "epoch": 6.362182886042907, + "grad_norm": 0.00031966075766831636, + "learning_rate": 0.00046332472568669236, + "loss": 0.2403, + "num_input_tokens_seen": 81475504, + "step": 39000 + }, + { + "epoch": 6.362182886042907, + "eval_loss": 0.3153197467327118, + "eval_runtime": 156.1446, + "eval_samples_per_second": 17.452, + "eval_steps_per_second": 8.729, + "num_input_tokens_seen": 81475504, + "step": 39000 + }, + { + "epoch": 6.362998613263724, + "grad_norm": 0.0010212219785898924, + "learning_rate": 0.0004587100216951578, + "loss": 0.3357, + "num_input_tokens_seen": 81486336, + "step": 39005 + }, + { + "epoch": 6.3638143404845415, + "grad_norm": 0.0005154759855940938, + "learning_rate": 0.00045411837884978265, + "loss": 0.2933, + "num_input_tokens_seen": 81496464, + "step": 39010 + }, + { + "epoch": 6.3646300677053596, + "grad_norm": 0.000950606307014823, + "learning_rate": 0.00044954979785865045, + "loss": 0.286, + "num_input_tokens_seen": 81507248, + "step": 39015 + }, + { + "epoch": 6.365445794926177, + "grad_norm": 0.0005062778946012259, + "learning_rate": 0.00044500427942631426, + "loss": 0.2509, + "num_input_tokens_seen": 81517984, + "step": 39020 + }, + { + "epoch": 6.366261522146994, + "grad_norm": 0.0005651578539982438, + "learning_rate": 0.0004404818242537467, + "loss": 0.3332, + "num_input_tokens_seen": 81527440, + "step": 39025 + }, + { + "epoch": 6.367077249367812, + "grad_norm": 0.0014172764495015144, + "learning_rate": 0.00043598243303837324, + "loss": 0.3187, + "num_input_tokens_seen": 81537648, + "step": 39030 + }, + { + "epoch": 6.367892976588629, + "grad_norm": 0.0005809378926642239, + "learning_rate": 0.00043150610647403885, + "loss": 0.2586, + "num_input_tokens_seen": 81548528, + "step": 39035 + }, + { + "epoch": 6.368708703809446, + "grad_norm": 0.0005645537748932838, + "learning_rate": 0.00042705284525104134, + "loss": 0.3239, + "num_input_tokens_seen": 81557632, + "step": 39040 + }, + { + "epoch": 6.369524431030263, + "grad_norm": 0.0008336346363648772, + "learning_rate": 0.0004226226500561647, + "loss": 0.3648, + "num_input_tokens_seen": 81568448, + "step": 39045 + }, + { + "epoch": 6.370340158251081, + "grad_norm": 0.0007804553606547415, + "learning_rate": 0.0004182155215725791, + "loss": 0.3265, + "num_input_tokens_seen": 81577360, + "step": 39050 + }, + { + "epoch": 6.371155885471898, + "grad_norm": 0.0010432477574795485, + "learning_rate": 0.00041383146047992424, + "loss": 0.3748, + "num_input_tokens_seen": 81588080, + "step": 39055 + }, + { + "epoch": 6.3719716126927155, + "grad_norm": 0.0005262398044578731, + "learning_rate": 0.00040947046745427597, + "loss": 0.3255, + "num_input_tokens_seen": 81598816, + "step": 39060 + }, + { + "epoch": 6.372787339913533, + "grad_norm": 0.0008613765821792185, + "learning_rate": 0.00040513254316814625, + "loss": 0.3407, + "num_input_tokens_seen": 81609840, + "step": 39065 + }, + { + "epoch": 6.373603067134351, + "grad_norm": 0.000819257867988199, + "learning_rate": 0.0004008176882905168, + "loss": 0.3446, + "num_input_tokens_seen": 81620064, + "step": 39070 + }, + { + "epoch": 6.374418794355168, + "grad_norm": 0.0007060188800096512, + "learning_rate": 0.00039652590348677184, + "loss": 0.3299, + "num_input_tokens_seen": 81630000, + "step": 39075 + }, + { + "epoch": 6.375234521575985, + "grad_norm": 0.0003700531960930675, + "learning_rate": 0.00039225718941878206, + "loss": 0.2675, + "num_input_tokens_seen": 81641088, + "step": 39080 + }, + { + "epoch": 6.376050248796802, + "grad_norm": 0.0007365020574070513, + "learning_rate": 0.00038801154674480417, + "loss": 0.2741, + "num_input_tokens_seen": 81651056, + "step": 39085 + }, + { + "epoch": 6.37686597601762, + "grad_norm": 0.000451906758826226, + "learning_rate": 0.00038378897611959784, + "loss": 0.2508, + "num_input_tokens_seen": 81661632, + "step": 39090 + }, + { + "epoch": 6.377681703238437, + "grad_norm": 0.0004062453517690301, + "learning_rate": 0.00037958947819430875, + "loss": 0.3161, + "num_input_tokens_seen": 81670640, + "step": 39095 + }, + { + "epoch": 6.378497430459254, + "grad_norm": 0.0009403462754562497, + "learning_rate": 0.0003754130536165856, + "loss": 0.2587, + "num_input_tokens_seen": 81679648, + "step": 39100 + }, + { + "epoch": 6.3793131576800715, + "grad_norm": 0.0005047236336395144, + "learning_rate": 0.0003712597030304632, + "loss": 0.3217, + "num_input_tokens_seen": 81690768, + "step": 39105 + }, + { + "epoch": 6.3801288849008895, + "grad_norm": 0.000657399941701442, + "learning_rate": 0.00036712942707646247, + "loss": 0.3071, + "num_input_tokens_seen": 81699984, + "step": 39110 + }, + { + "epoch": 6.380944612121707, + "grad_norm": 0.0008338289335370064, + "learning_rate": 0.00036302222639149063, + "loss": 0.2835, + "num_input_tokens_seen": 81709840, + "step": 39115 + }, + { + "epoch": 6.381760339342524, + "grad_norm": 0.0006552012637257576, + "learning_rate": 0.000358938101608941, + "loss": 0.3112, + "num_input_tokens_seen": 81720080, + "step": 39120 + }, + { + "epoch": 6.382576066563341, + "grad_norm": 0.0005326196551322937, + "learning_rate": 0.0003548770533586598, + "loss": 0.2736, + "num_input_tokens_seen": 81730688, + "step": 39125 + }, + { + "epoch": 6.383391793784159, + "grad_norm": 0.0005538160912692547, + "learning_rate": 0.0003508390822668961, + "loss": 0.2937, + "num_input_tokens_seen": 81741984, + "step": 39130 + }, + { + "epoch": 6.384207521004976, + "grad_norm": 0.0006378117250278592, + "learning_rate": 0.00034682418895633503, + "loss": 0.2935, + "num_input_tokens_seen": 81753808, + "step": 39135 + }, + { + "epoch": 6.385023248225793, + "grad_norm": 0.001274051028303802, + "learning_rate": 0.0003428323740461647, + "loss": 0.3394, + "num_input_tokens_seen": 81764848, + "step": 39140 + }, + { + "epoch": 6.38583897544661, + "grad_norm": 0.0007004071376286447, + "learning_rate": 0.00033886363815194276, + "loss": 0.2944, + "num_input_tokens_seen": 81776176, + "step": 39145 + }, + { + "epoch": 6.386654702667428, + "grad_norm": 0.0013255535159260035, + "learning_rate": 0.0003349179818857129, + "loss": 0.336, + "num_input_tokens_seen": 81786448, + "step": 39150 + }, + { + "epoch": 6.387470429888245, + "grad_norm": 0.0005247555091045797, + "learning_rate": 0.0003309954058559383, + "loss": 0.3239, + "num_input_tokens_seen": 81797184, + "step": 39155 + }, + { + "epoch": 6.388286157109063, + "grad_norm": 0.0009513176628388464, + "learning_rate": 0.0003270959106675186, + "loss": 0.3542, + "num_input_tokens_seen": 81807152, + "step": 39160 + }, + { + "epoch": 6.38910188432988, + "grad_norm": 0.00048562997835688293, + "learning_rate": 0.0003232194969218227, + "loss": 0.2857, + "num_input_tokens_seen": 81818192, + "step": 39165 + }, + { + "epoch": 6.389917611550698, + "grad_norm": 0.0007941331132315099, + "learning_rate": 0.00031936616521663905, + "loss": 0.2907, + "num_input_tokens_seen": 81828800, + "step": 39170 + }, + { + "epoch": 6.390733338771515, + "grad_norm": 0.0006692854221910238, + "learning_rate": 0.00031553591614619236, + "loss": 0.2808, + "num_input_tokens_seen": 81839456, + "step": 39175 + }, + { + "epoch": 6.391549065992332, + "grad_norm": 0.0009848462650552392, + "learning_rate": 0.00031172875030117676, + "loss": 0.3107, + "num_input_tokens_seen": 81849616, + "step": 39180 + }, + { + "epoch": 6.392364793213149, + "grad_norm": 0.0005402906099334359, + "learning_rate": 0.0003079446682686726, + "loss": 0.3408, + "num_input_tokens_seen": 81860432, + "step": 39185 + }, + { + "epoch": 6.393180520433967, + "grad_norm": 0.0010473772417753935, + "learning_rate": 0.0003041836706322465, + "loss": 0.3804, + "num_input_tokens_seen": 81869312, + "step": 39190 + }, + { + "epoch": 6.393996247654784, + "grad_norm": 0.0007492630975320935, + "learning_rate": 0.0003004457579719011, + "loss": 0.3233, + "num_input_tokens_seen": 81880000, + "step": 39195 + }, + { + "epoch": 6.394811974875601, + "grad_norm": 0.0008718514582142234, + "learning_rate": 0.00029673093086405867, + "loss": 0.3694, + "num_input_tokens_seen": 81889856, + "step": 39200 + }, + { + "epoch": 6.394811974875601, + "eval_loss": 0.3150635361671448, + "eval_runtime": 156.1337, + "eval_samples_per_second": 17.453, + "eval_steps_per_second": 8.73, + "num_input_tokens_seen": 81889856, + "step": 39200 + }, + { + "epoch": 6.395627702096419, + "grad_norm": 0.0007621835684403777, + "learning_rate": 0.00029303918988159426, + "loss": 0.3084, + "num_input_tokens_seen": 81900016, + "step": 39205 + }, + { + "epoch": 6.3964434293172365, + "grad_norm": 0.00044591526966542006, + "learning_rate": 0.0002893705355938192, + "loss": 0.2945, + "num_input_tokens_seen": 81910944, + "step": 39210 + }, + { + "epoch": 6.397259156538054, + "grad_norm": 0.000602608488406986, + "learning_rate": 0.0002857249685664975, + "loss": 0.319, + "num_input_tokens_seen": 81921936, + "step": 39215 + }, + { + "epoch": 6.398074883758871, + "grad_norm": 0.0011920274700969458, + "learning_rate": 0.0002821024893618129, + "loss": 0.3432, + "num_input_tokens_seen": 81932640, + "step": 39220 + }, + { + "epoch": 6.398890610979688, + "grad_norm": 0.000626692664809525, + "learning_rate": 0.0002785030985383852, + "loss": 0.3202, + "num_input_tokens_seen": 81941872, + "step": 39225 + }, + { + "epoch": 6.399706338200506, + "grad_norm": 0.000619218684732914, + "learning_rate": 0.00027492679665130356, + "loss": 0.3388, + "num_input_tokens_seen": 81951376, + "step": 39230 + }, + { + "epoch": 6.400522065421323, + "grad_norm": 0.0005634370027109981, + "learning_rate": 0.000271373584252077, + "loss": 0.2902, + "num_input_tokens_seen": 81960416, + "step": 39235 + }, + { + "epoch": 6.40133779264214, + "grad_norm": 0.0004694470262620598, + "learning_rate": 0.00026784346188865046, + "loss": 0.2884, + "num_input_tokens_seen": 81970560, + "step": 39240 + }, + { + "epoch": 6.402153519862958, + "grad_norm": 0.0004422247293405235, + "learning_rate": 0.0002643364301054218, + "loss": 0.3026, + "num_input_tokens_seen": 81980416, + "step": 39245 + }, + { + "epoch": 6.402969247083775, + "grad_norm": 0.0007623861311003566, + "learning_rate": 0.0002608524894431918, + "loss": 0.3183, + "num_input_tokens_seen": 81990944, + "step": 39250 + }, + { + "epoch": 6.4037849743045925, + "grad_norm": 0.00041047995910048485, + "learning_rate": 0.000257391640439264, + "loss": 0.2209, + "num_input_tokens_seen": 82002032, + "step": 39255 + }, + { + "epoch": 6.40460070152541, + "grad_norm": 0.0005300354678183794, + "learning_rate": 0.00025395388362732806, + "loss": 0.3158, + "num_input_tokens_seen": 82012352, + "step": 39260 + }, + { + "epoch": 6.405416428746228, + "grad_norm": 0.0004696928081102669, + "learning_rate": 0.00025053921953751, + "loss": 0.374, + "num_input_tokens_seen": 82022416, + "step": 39265 + }, + { + "epoch": 6.406232155967045, + "grad_norm": 0.0005670296959578991, + "learning_rate": 0.00024714764869643855, + "loss": 0.3486, + "num_input_tokens_seen": 82032992, + "step": 39270 + }, + { + "epoch": 6.407047883187862, + "grad_norm": 0.0006365524604916573, + "learning_rate": 0.0002437791716270954, + "loss": 0.3172, + "num_input_tokens_seen": 82042000, + "step": 39275 + }, + { + "epoch": 6.407863610408679, + "grad_norm": 0.00046513983397744596, + "learning_rate": 0.00024043378884896493, + "loss": 0.2956, + "num_input_tokens_seen": 82052288, + "step": 39280 + }, + { + "epoch": 6.408679337629497, + "grad_norm": 0.0007449216209352016, + "learning_rate": 0.00023711150087793453, + "loss": 0.3275, + "num_input_tokens_seen": 82060864, + "step": 39285 + }, + { + "epoch": 6.409495064850314, + "grad_norm": 0.0009507283102720976, + "learning_rate": 0.000233812308226361, + "loss": 0.365, + "num_input_tokens_seen": 82071424, + "step": 39290 + }, + { + "epoch": 6.410310792071131, + "grad_norm": 0.0007003292557783425, + "learning_rate": 0.00023053621140300406, + "loss": 0.3482, + "num_input_tokens_seen": 82081456, + "step": 39295 + }, + { + "epoch": 6.411126519291948, + "grad_norm": 0.0005953767104074359, + "learning_rate": 0.00022728321091307623, + "loss": 0.2956, + "num_input_tokens_seen": 82091264, + "step": 39300 + }, + { + "epoch": 6.4119422465127665, + "grad_norm": 0.000529298500623554, + "learning_rate": 0.0002240533072582429, + "loss": 0.3162, + "num_input_tokens_seen": 82100880, + "step": 39305 + }, + { + "epoch": 6.412757973733584, + "grad_norm": 0.0005841782549396157, + "learning_rate": 0.00022084650093658897, + "loss": 0.329, + "num_input_tokens_seen": 82112976, + "step": 39310 + }, + { + "epoch": 6.413573700954401, + "grad_norm": 0.000515828316565603, + "learning_rate": 0.0002176627924426522, + "loss": 0.2876, + "num_input_tokens_seen": 82123040, + "step": 39315 + }, + { + "epoch": 6.414389428175218, + "grad_norm": 0.0007078232010826468, + "learning_rate": 0.0002145021822673898, + "loss": 0.2985, + "num_input_tokens_seen": 82134928, + "step": 39320 + }, + { + "epoch": 6.415205155396036, + "grad_norm": 0.0007500796928070486, + "learning_rate": 0.00021136467089822862, + "loss": 0.42, + "num_input_tokens_seen": 82145008, + "step": 39325 + }, + { + "epoch": 6.416020882616853, + "grad_norm": 0.0006684844847768545, + "learning_rate": 0.00020825025881898162, + "loss": 0.3601, + "num_input_tokens_seen": 82155824, + "step": 39330 + }, + { + "epoch": 6.41683660983767, + "grad_norm": 0.0006941516767255962, + "learning_rate": 0.0002051589465099479, + "loss": 0.3739, + "num_input_tokens_seen": 82166352, + "step": 39335 + }, + { + "epoch": 6.417652337058487, + "grad_norm": 0.0005852780886925757, + "learning_rate": 0.0002020907344478462, + "loss": 0.3078, + "num_input_tokens_seen": 82177872, + "step": 39340 + }, + { + "epoch": 6.418468064279305, + "grad_norm": 0.0008414423791691661, + "learning_rate": 0.0001990456231058313, + "loss": 0.2684, + "num_input_tokens_seen": 82188000, + "step": 39345 + }, + { + "epoch": 6.419283791500122, + "grad_norm": 0.0005316517781466246, + "learning_rate": 0.00019602361295349423, + "loss": 0.2868, + "num_input_tokens_seen": 82199440, + "step": 39350 + }, + { + "epoch": 6.4200995187209395, + "grad_norm": 0.00048571248771622777, + "learning_rate": 0.0001930247044568789, + "loss": 0.2837, + "num_input_tokens_seen": 82208624, + "step": 39355 + }, + { + "epoch": 6.420915245941757, + "grad_norm": 0.0015189734986051917, + "learning_rate": 0.00019004889807843205, + "loss": 0.3186, + "num_input_tokens_seen": 82218160, + "step": 39360 + }, + { + "epoch": 6.421730973162575, + "grad_norm": 0.000493094848934561, + "learning_rate": 0.00018709619427708656, + "loss": 0.2458, + "num_input_tokens_seen": 82230352, + "step": 39365 + }, + { + "epoch": 6.422546700383392, + "grad_norm": 0.000600673898588866, + "learning_rate": 0.00018416659350817822, + "loss": 0.374, + "num_input_tokens_seen": 82241536, + "step": 39370 + }, + { + "epoch": 6.423362427604209, + "grad_norm": 0.0005785468965768814, + "learning_rate": 0.00018126009622346229, + "loss": 0.3092, + "num_input_tokens_seen": 82252832, + "step": 39375 + }, + { + "epoch": 6.424178154825027, + "grad_norm": 0.0006951984250918031, + "learning_rate": 0.00017837670287119687, + "loss": 0.2791, + "num_input_tokens_seen": 82263392, + "step": 39380 + }, + { + "epoch": 6.424993882045844, + "grad_norm": 0.00043441695743240416, + "learning_rate": 0.00017551641389602633, + "loss": 0.33, + "num_input_tokens_seen": 82274112, + "step": 39385 + }, + { + "epoch": 6.425809609266661, + "grad_norm": 0.0006530254613608122, + "learning_rate": 0.00017267922973903115, + "loss": 0.359, + "num_input_tokens_seen": 82284160, + "step": 39390 + }, + { + "epoch": 6.426625336487478, + "grad_norm": 0.000638237630482763, + "learning_rate": 0.00016986515083774467, + "loss": 0.2829, + "num_input_tokens_seen": 82294448, + "step": 39395 + }, + { + "epoch": 6.4274410637082955, + "grad_norm": 0.0008342136279679835, + "learning_rate": 0.00016707417762611975, + "loss": 0.3268, + "num_input_tokens_seen": 82305408, + "step": 39400 + }, + { + "epoch": 6.4274410637082955, + "eval_loss": 0.31500184535980225, + "eval_runtime": 156.0721, + "eval_samples_per_second": 17.46, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 82305408, + "step": 39400 + }, + { + "epoch": 6.4282567909291135, + "grad_norm": 0.0008033013436943293, + "learning_rate": 0.00016430631053459543, + "loss": 0.3002, + "num_input_tokens_seen": 82315808, + "step": 39405 + }, + { + "epoch": 6.429072518149931, + "grad_norm": 0.000524776813108474, + "learning_rate": 0.0001615615499899803, + "loss": 0.301, + "num_input_tokens_seen": 82326832, + "step": 39410 + }, + { + "epoch": 6.429888245370748, + "grad_norm": 0.0006121171172708273, + "learning_rate": 0.00015883989641556905, + "loss": 0.2932, + "num_input_tokens_seen": 82337600, + "step": 39415 + }, + { + "epoch": 6.430703972591566, + "grad_norm": 0.0004655921657104045, + "learning_rate": 0.00015614135023105934, + "loss": 0.2679, + "num_input_tokens_seen": 82347360, + "step": 39420 + }, + { + "epoch": 6.431519699812383, + "grad_norm": 0.0007239928818307817, + "learning_rate": 0.00015346591185261827, + "loss": 0.3654, + "num_input_tokens_seen": 82358544, + "step": 39425 + }, + { + "epoch": 6.4323354270332, + "grad_norm": 0.0007355465204454958, + "learning_rate": 0.00015081358169281576, + "loss": 0.3339, + "num_input_tokens_seen": 82367168, + "step": 39430 + }, + { + "epoch": 6.433151154254017, + "grad_norm": 0.0005835682968609035, + "learning_rate": 0.00014818436016069135, + "loss": 0.2914, + "num_input_tokens_seen": 82375824, + "step": 39435 + }, + { + "epoch": 6.433966881474835, + "grad_norm": 0.0007427348173223436, + "learning_rate": 0.00014557824766168735, + "loss": 0.3338, + "num_input_tokens_seen": 82384320, + "step": 39440 + }, + { + "epoch": 6.434782608695652, + "grad_norm": 0.0007639218238182366, + "learning_rate": 0.00014299524459769896, + "loss": 0.3845, + "num_input_tokens_seen": 82394512, + "step": 39445 + }, + { + "epoch": 6.4355983359164695, + "grad_norm": 0.00047853286378085613, + "learning_rate": 0.0001404353513670742, + "loss": 0.3091, + "num_input_tokens_seen": 82403376, + "step": 39450 + }, + { + "epoch": 6.436414063137287, + "grad_norm": 0.0006476407288573682, + "learning_rate": 0.0001378985683645806, + "loss": 0.2882, + "num_input_tokens_seen": 82414784, + "step": 39455 + }, + { + "epoch": 6.437229790358105, + "grad_norm": 0.0004624886205419898, + "learning_rate": 0.0001353848959813886, + "loss": 0.3126, + "num_input_tokens_seen": 82427136, + "step": 39460 + }, + { + "epoch": 6.438045517578922, + "grad_norm": 0.0007643849239684641, + "learning_rate": 0.00013289433460517142, + "loss": 0.3299, + "num_input_tokens_seen": 82436800, + "step": 39465 + }, + { + "epoch": 6.438861244799739, + "grad_norm": 0.0007792774122208357, + "learning_rate": 0.00013042688462000518, + "loss": 0.3064, + "num_input_tokens_seen": 82446528, + "step": 39470 + }, + { + "epoch": 6.439676972020556, + "grad_norm": 0.0008739111362956464, + "learning_rate": 0.0001279825464063855, + "loss": 0.3507, + "num_input_tokens_seen": 82455792, + "step": 39475 + }, + { + "epoch": 6.440492699241374, + "grad_norm": 0.0007490101852454245, + "learning_rate": 0.00012556132034126087, + "loss": 0.3083, + "num_input_tokens_seen": 82466240, + "step": 39480 + }, + { + "epoch": 6.441308426462191, + "grad_norm": 0.0011407288257032633, + "learning_rate": 0.0001231632067980326, + "loss": 0.3208, + "num_input_tokens_seen": 82475008, + "step": 39485 + }, + { + "epoch": 6.442124153683008, + "grad_norm": 0.001252554589882493, + "learning_rate": 0.00012078820614650486, + "loss": 0.3728, + "num_input_tokens_seen": 82484880, + "step": 39490 + }, + { + "epoch": 6.442939880903825, + "grad_norm": 0.0010092540178447962, + "learning_rate": 0.00011843631875291804, + "loss": 0.3014, + "num_input_tokens_seen": 82494848, + "step": 39495 + }, + { + "epoch": 6.443755608124643, + "grad_norm": 0.0007020958582870662, + "learning_rate": 0.00011610754497999863, + "loss": 0.2729, + "num_input_tokens_seen": 82506272, + "step": 39500 + }, + { + "epoch": 6.444571335345461, + "grad_norm": 0.0005525056621991098, + "learning_rate": 0.0001138018851868594, + "loss": 0.3461, + "num_input_tokens_seen": 82518704, + "step": 39505 + }, + { + "epoch": 6.445387062566278, + "grad_norm": 0.0005263853818178177, + "learning_rate": 0.0001115193397290326, + "loss": 0.2719, + "num_input_tokens_seen": 82527280, + "step": 39510 + }, + { + "epoch": 6.446202789787095, + "grad_norm": 0.000619308790192008, + "learning_rate": 0.00010925990895856996, + "loss": 0.2972, + "num_input_tokens_seen": 82538816, + "step": 39515 + }, + { + "epoch": 6.447018517007913, + "grad_norm": 0.0009642438380979002, + "learning_rate": 0.00010702359322385946, + "loss": 0.3568, + "num_input_tokens_seen": 82548464, + "step": 39520 + }, + { + "epoch": 6.44783424422873, + "grad_norm": 0.001217894023284316, + "learning_rate": 0.00010481039286977523, + "loss": 0.3273, + "num_input_tokens_seen": 82559328, + "step": 39525 + }, + { + "epoch": 6.448649971449547, + "grad_norm": 0.0006176697206683457, + "learning_rate": 0.00010262030823764423, + "loss": 0.3725, + "num_input_tokens_seen": 82569888, + "step": 39530 + }, + { + "epoch": 6.449465698670364, + "grad_norm": 0.0008866743301041424, + "learning_rate": 0.00010045333966517966, + "loss": 0.3578, + "num_input_tokens_seen": 82580416, + "step": 39535 + }, + { + "epoch": 6.450281425891182, + "grad_norm": 0.0008718570461496711, + "learning_rate": 9.83094874865642e-05, + "loss": 0.2789, + "num_input_tokens_seen": 82588496, + "step": 39540 + }, + { + "epoch": 6.451097153111999, + "grad_norm": 0.0007709929486736655, + "learning_rate": 9.618875203241672e-05, + "loss": 0.3248, + "num_input_tokens_seen": 82598880, + "step": 39545 + }, + { + "epoch": 6.4519128803328165, + "grad_norm": 0.0004078499914612621, + "learning_rate": 9.409113362977561e-05, + "loss": 0.3509, + "num_input_tokens_seen": 82609360, + "step": 39550 + }, + { + "epoch": 6.4527286075536345, + "grad_norm": 0.0007074158638715744, + "learning_rate": 9.20166326020988e-05, + "loss": 0.4024, + "num_input_tokens_seen": 82618112, + "step": 39555 + }, + { + "epoch": 6.453544334774452, + "grad_norm": 0.0009939165320247412, + "learning_rate": 8.996524926933035e-05, + "loss": 0.394, + "num_input_tokens_seen": 82628160, + "step": 39560 + }, + { + "epoch": 6.454360061995269, + "grad_norm": 0.0007131172460503876, + "learning_rate": 8.793698394781723e-05, + "loss": 0.3767, + "num_input_tokens_seen": 82639168, + "step": 39565 + }, + { + "epoch": 6.455175789216086, + "grad_norm": 0.0004758960858453065, + "learning_rate": 8.593183695030926e-05, + "loss": 0.3183, + "num_input_tokens_seen": 82649872, + "step": 39570 + }, + { + "epoch": 6.455991516436903, + "grad_norm": 0.0005761744687333703, + "learning_rate": 8.39498085860757e-05, + "loss": 0.3011, + "num_input_tokens_seen": 82661040, + "step": 39575 + }, + { + "epoch": 6.456807243657721, + "grad_norm": 0.0008664620690979064, + "learning_rate": 8.199089916072211e-05, + "loss": 0.3097, + "num_input_tokens_seen": 82671712, + "step": 39580 + }, + { + "epoch": 6.457622970878538, + "grad_norm": 0.0005587355699390173, + "learning_rate": 8.005510897637346e-05, + "loss": 0.3487, + "num_input_tokens_seen": 82681328, + "step": 39585 + }, + { + "epoch": 6.458438698099355, + "grad_norm": 0.0006920701125636697, + "learning_rate": 7.8142438331541e-05, + "loss": 0.3439, + "num_input_tokens_seen": 82691968, + "step": 39590 + }, + { + "epoch": 6.459254425320173, + "grad_norm": 0.0006221246439963579, + "learning_rate": 7.625288752117209e-05, + "loss": 0.3291, + "num_input_tokens_seen": 82702032, + "step": 39595 + }, + { + "epoch": 6.4600701525409905, + "grad_norm": 0.0005563319427892566, + "learning_rate": 7.4386456836667e-05, + "loss": 0.2828, + "num_input_tokens_seen": 82712800, + "step": 39600 + }, + { + "epoch": 6.4600701525409905, + "eval_loss": 0.3149668872356415, + "eval_runtime": 156.1695, + "eval_samples_per_second": 17.449, + "eval_steps_per_second": 8.728, + "num_input_tokens_seen": 82712800, + "step": 39600 + }, + { + "epoch": 6.460885879761808, + "grad_norm": 0.000391895737266168, + "learning_rate": 7.254314656586214e-05, + "loss": 0.2152, + "num_input_tokens_seen": 82724720, + "step": 39605 + }, + { + "epoch": 6.461701606982625, + "grad_norm": 0.0008152606314979494, + "learning_rate": 7.07229569929968e-05, + "loss": 0.3186, + "num_input_tokens_seen": 82734720, + "step": 39610 + }, + { + "epoch": 6.462517334203443, + "grad_norm": 0.0011583176674321294, + "learning_rate": 6.892588839879643e-05, + "loss": 0.288, + "num_input_tokens_seen": 82746320, + "step": 39615 + }, + { + "epoch": 6.46333306142426, + "grad_norm": 0.000794461346231401, + "learning_rate": 6.71519410603727e-05, + "loss": 0.2884, + "num_input_tokens_seen": 82756992, + "step": 39620 + }, + { + "epoch": 6.464148788645077, + "grad_norm": 0.001155988429673016, + "learning_rate": 6.540111525129011e-05, + "loss": 0.2818, + "num_input_tokens_seen": 82768560, + "step": 39625 + }, + { + "epoch": 6.464964515865894, + "grad_norm": 0.000775418127886951, + "learning_rate": 6.367341124154934e-05, + "loss": 0.3299, + "num_input_tokens_seen": 82777632, + "step": 39630 + }, + { + "epoch": 6.465780243086712, + "grad_norm": 0.0005500339320860803, + "learning_rate": 6.19688292975873e-05, + "loss": 0.319, + "num_input_tokens_seen": 82787616, + "step": 39635 + }, + { + "epoch": 6.466595970307529, + "grad_norm": 0.0004428365791682154, + "learning_rate": 6.0287369682260336e-05, + "loss": 0.3002, + "num_input_tokens_seen": 82796992, + "step": 39640 + }, + { + "epoch": 6.467411697528346, + "grad_norm": 0.0004968596622347832, + "learning_rate": 5.8629032654894384e-05, + "loss": 0.3435, + "num_input_tokens_seen": 82807456, + "step": 39645 + }, + { + "epoch": 6.468227424749164, + "grad_norm": 0.0007029340486042202, + "learning_rate": 5.699381847120155e-05, + "loss": 0.3478, + "num_input_tokens_seen": 82818016, + "step": 39650 + }, + { + "epoch": 6.469043151969982, + "grad_norm": 0.0003491346724331379, + "learning_rate": 5.5381727383380094e-05, + "loss": 0.2973, + "num_input_tokens_seen": 82827872, + "step": 39655 + }, + { + "epoch": 6.469858879190799, + "grad_norm": 0.001071788021363318, + "learning_rate": 5.379275964001451e-05, + "loss": 0.3107, + "num_input_tokens_seen": 82837984, + "step": 39660 + }, + { + "epoch": 6.470674606411616, + "grad_norm": 0.0005642874166369438, + "learning_rate": 5.222691548614211e-05, + "loss": 0.2645, + "num_input_tokens_seen": 82847584, + "step": 39665 + }, + { + "epoch": 6.471490333632433, + "grad_norm": 0.0004817089647985995, + "learning_rate": 5.068419516323641e-05, + "loss": 0.3007, + "num_input_tokens_seen": 82857392, + "step": 39670 + }, + { + "epoch": 6.472306060853251, + "grad_norm": 0.0006676298216916621, + "learning_rate": 4.91645989092071e-05, + "loss": 0.3051, + "num_input_tokens_seen": 82868144, + "step": 39675 + }, + { + "epoch": 6.473121788074068, + "grad_norm": 0.0005413759499788284, + "learning_rate": 4.7668126958400056e-05, + "loss": 0.29, + "num_input_tokens_seen": 82879424, + "step": 39680 + }, + { + "epoch": 6.473937515294885, + "grad_norm": 0.0007954402826726437, + "learning_rate": 4.619477954159734e-05, + "loss": 0.3549, + "num_input_tokens_seen": 82888720, + "step": 39685 + }, + { + "epoch": 6.474753242515702, + "grad_norm": 0.0007724366732873023, + "learning_rate": 4.4744556885983884e-05, + "loss": 0.3152, + "num_input_tokens_seen": 82899440, + "step": 39690 + }, + { + "epoch": 6.47556896973652, + "grad_norm": 0.0006250036531127989, + "learning_rate": 4.331745921523078e-05, + "loss": 0.3224, + "num_input_tokens_seen": 82908896, + "step": 39695 + }, + { + "epoch": 6.4763846969573375, + "grad_norm": 0.0010331996018067002, + "learning_rate": 4.191348674937867e-05, + "loss": 0.3018, + "num_input_tokens_seen": 82919536, + "step": 39700 + }, + { + "epoch": 6.477200424178155, + "grad_norm": 0.00037559535121545196, + "learning_rate": 4.0532639704971006e-05, + "loss": 0.2785, + "num_input_tokens_seen": 82930080, + "step": 39705 + }, + { + "epoch": 6.478016151398972, + "grad_norm": 0.0007754644029773772, + "learning_rate": 3.917491829493747e-05, + "loss": 0.2249, + "num_input_tokens_seen": 82940800, + "step": 39710 + }, + { + "epoch": 6.47883187861979, + "grad_norm": 0.0004157435323577374, + "learning_rate": 3.78403227286439e-05, + "loss": 0.3264, + "num_input_tokens_seen": 82949920, + "step": 39715 + }, + { + "epoch": 6.479647605840607, + "grad_norm": 0.00046569318510591984, + "learning_rate": 3.652885321192567e-05, + "loss": 0.2633, + "num_input_tokens_seen": 82961248, + "step": 39720 + }, + { + "epoch": 6.480463333061424, + "grad_norm": 0.0004686380852945149, + "learning_rate": 3.524050994702099e-05, + "loss": 0.2681, + "num_input_tokens_seen": 82972816, + "step": 39725 + }, + { + "epoch": 6.481279060282241, + "grad_norm": 0.0005811029695905745, + "learning_rate": 3.3975293132604276e-05, + "loss": 0.2506, + "num_input_tokens_seen": 82983440, + "step": 39730 + }, + { + "epoch": 6.482094787503059, + "grad_norm": 0.00033019910915754735, + "learning_rate": 3.2733202963786125e-05, + "loss": 0.2567, + "num_input_tokens_seen": 82994512, + "step": 39735 + }, + { + "epoch": 6.482910514723876, + "grad_norm": 0.0004762313619721681, + "learning_rate": 3.15142396321133e-05, + "loss": 0.3134, + "num_input_tokens_seen": 83005104, + "step": 39740 + }, + { + "epoch": 6.4837262419446935, + "grad_norm": 0.0006498000002466142, + "learning_rate": 3.0318403325552132e-05, + "loss": 0.2738, + "num_input_tokens_seen": 83017120, + "step": 39745 + }, + { + "epoch": 6.484541969165511, + "grad_norm": 0.0008157661650329828, + "learning_rate": 2.914569422855506e-05, + "loss": 0.2849, + "num_input_tokens_seen": 83025888, + "step": 39750 + }, + { + "epoch": 6.485357696386329, + "grad_norm": 0.0005480166291818023, + "learning_rate": 2.7996112521927462e-05, + "loss": 0.2651, + "num_input_tokens_seen": 83036416, + "step": 39755 + }, + { + "epoch": 6.486173423607146, + "grad_norm": 0.0005370262661017478, + "learning_rate": 2.68696583829775e-05, + "loss": 0.31, + "num_input_tokens_seen": 83046304, + "step": 39760 + }, + { + "epoch": 6.486989150827963, + "grad_norm": 0.0006016622064635158, + "learning_rate": 2.576633198539957e-05, + "loss": 0.307, + "num_input_tokens_seen": 83055904, + "step": 39765 + }, + { + "epoch": 6.487804878048781, + "grad_norm": 0.0007461003260686994, + "learning_rate": 2.46861334993409e-05, + "loss": 0.4111, + "num_input_tokens_seen": 83065600, + "step": 39770 + }, + { + "epoch": 6.488620605269598, + "grad_norm": 0.0011827106354758143, + "learning_rate": 2.3629063091384903e-05, + "loss": 0.3858, + "num_input_tokens_seen": 83076960, + "step": 39775 + }, + { + "epoch": 6.489436332490415, + "grad_norm": 0.0006682027014903724, + "learning_rate": 2.2595120924567834e-05, + "loss": 0.3664, + "num_input_tokens_seen": 83088112, + "step": 39780 + }, + { + "epoch": 6.490252059711232, + "grad_norm": 0.0006258853245526552, + "learning_rate": 2.158430715829551e-05, + "loss": 0.2996, + "num_input_tokens_seen": 83097584, + "step": 39785 + }, + { + "epoch": 6.49106778693205, + "grad_norm": 0.0004446088569238782, + "learning_rate": 2.059662194849321e-05, + "loss": 0.2838, + "num_input_tokens_seen": 83106912, + "step": 39790 + }, + { + "epoch": 6.4918835141528675, + "grad_norm": 0.0007365207420662045, + "learning_rate": 1.9632065447422463e-05, + "loss": 0.3145, + "num_input_tokens_seen": 83117280, + "step": 39795 + }, + { + "epoch": 6.492699241373685, + "grad_norm": 0.0024166549555957317, + "learning_rate": 1.8690637803880916e-05, + "loss": 0.3382, + "num_input_tokens_seen": 83128704, + "step": 39800 + }, + { + "epoch": 6.492699241373685, + "eval_loss": 0.31521594524383545, + "eval_runtime": 156.078, + "eval_samples_per_second": 17.459, + "eval_steps_per_second": 8.733, + "num_input_tokens_seen": 83128704, + "step": 39800 + }, + { + "epoch": 6.493514968594502, + "grad_norm": 0.000413313799072057, + "learning_rate": 1.7772339163019123e-05, + "loss": 0.3129, + "num_input_tokens_seen": 83139264, + "step": 39805 + }, + { + "epoch": 6.49433069581532, + "grad_norm": 0.00046519265742972493, + "learning_rate": 1.6877169666457138e-05, + "loss": 0.3981, + "num_input_tokens_seen": 83149616, + "step": 39810 + }, + { + "epoch": 6.495146423036137, + "grad_norm": 0.0005832380848005414, + "learning_rate": 1.6005129452234532e-05, + "loss": 0.3205, + "num_input_tokens_seen": 83160880, + "step": 39815 + }, + { + "epoch": 6.495962150256954, + "grad_norm": 0.001189239090308547, + "learning_rate": 1.5156218654843733e-05, + "loss": 0.3856, + "num_input_tokens_seen": 83171664, + "step": 39820 + }, + { + "epoch": 6.496777877477771, + "grad_norm": 0.0005426821298897266, + "learning_rate": 1.4330437405196683e-05, + "loss": 0.3042, + "num_input_tokens_seen": 83182000, + "step": 39825 + }, + { + "epoch": 6.497593604698589, + "grad_norm": 0.0006493647815659642, + "learning_rate": 1.352778583062486e-05, + "loss": 0.3844, + "num_input_tokens_seen": 83190688, + "step": 39830 + }, + { + "epoch": 6.498409331919406, + "grad_norm": 0.0007370312814600766, + "learning_rate": 1.2748264054929237e-05, + "loss": 0.2926, + "num_input_tokens_seen": 83199120, + "step": 39835 + }, + { + "epoch": 6.499225059140223, + "grad_norm": 0.00048091847565956414, + "learning_rate": 1.1991872198297004e-05, + "loss": 0.397, + "num_input_tokens_seen": 83210832, + "step": 39840 + }, + { + "epoch": 6.5000407863610405, + "grad_norm": 0.00037730459007434547, + "learning_rate": 1.1258610377384847e-05, + "loss": 0.296, + "num_input_tokens_seen": 83222880, + "step": 39845 + }, + { + "epoch": 6.500856513581859, + "grad_norm": 0.0005381732480600476, + "learning_rate": 1.0548478705268982e-05, + "loss": 0.3033, + "num_input_tokens_seen": 83233184, + "step": 39850 + }, + { + "epoch": 6.501672240802676, + "grad_norm": 0.0014376319013535976, + "learning_rate": 9.86147729147846e-06, + "loss": 0.3344, + "num_input_tokens_seen": 83243648, + "step": 39855 + }, + { + "epoch": 6.502487968023493, + "grad_norm": 0.0013876563170924783, + "learning_rate": 9.197606241928557e-06, + "loss": 0.2919, + "num_input_tokens_seen": 83254464, + "step": 39860 + }, + { + "epoch": 6.50330369524431, + "grad_norm": 0.0005650285165756941, + "learning_rate": 8.556865659004042e-06, + "loss": 0.295, + "num_input_tokens_seen": 83264800, + "step": 39865 + }, + { + "epoch": 6.504119422465128, + "grad_norm": 0.0007203166023828089, + "learning_rate": 7.939255641525867e-06, + "loss": 0.3775, + "num_input_tokens_seen": 83275616, + "step": 39870 + }, + { + "epoch": 6.504935149685945, + "grad_norm": 0.0003828671178780496, + "learning_rate": 7.344776284751164e-06, + "loss": 0.3501, + "num_input_tokens_seen": 83285712, + "step": 39875 + }, + { + "epoch": 6.505750876906762, + "grad_norm": 0.000645027554128319, + "learning_rate": 6.773427680323296e-06, + "loss": 0.3387, + "num_input_tokens_seen": 83297248, + "step": 39880 + }, + { + "epoch": 6.506566604127579, + "grad_norm": 0.0008872900507412851, + "learning_rate": 6.225209916355112e-06, + "loss": 0.3645, + "num_input_tokens_seen": 83308752, + "step": 39885 + }, + { + "epoch": 6.507382331348397, + "grad_norm": 0.0010180266108363867, + "learning_rate": 5.7001230774123e-06, + "loss": 0.3132, + "num_input_tokens_seen": 83317312, + "step": 39890 + }, + { + "epoch": 6.5081980585692145, + "grad_norm": 0.0007164361304603517, + "learning_rate": 5.198167244446772e-06, + "loss": 0.3491, + "num_input_tokens_seen": 83327600, + "step": 39895 + }, + { + "epoch": 6.509013785790032, + "grad_norm": 0.0006242027156986296, + "learning_rate": 4.71934249487993e-06, + "loss": 0.2696, + "num_input_tokens_seen": 83338448, + "step": 39900 + }, + { + "epoch": 6.50982951301085, + "grad_norm": 0.0006228282582014799, + "learning_rate": 4.2636489025527075e-06, + "loss": 0.2819, + "num_input_tokens_seen": 83348896, + "step": 39905 + }, + { + "epoch": 6.510645240231667, + "grad_norm": 0.0007171735633164644, + "learning_rate": 3.831086537742223e-06, + "loss": 0.2953, + "num_input_tokens_seen": 83359104, + "step": 39910 + }, + { + "epoch": 6.511460967452484, + "grad_norm": 0.0004338730068411678, + "learning_rate": 3.4216554671451236e-06, + "loss": 0.3382, + "num_input_tokens_seen": 83368384, + "step": 39915 + }, + { + "epoch": 6.512276694673301, + "grad_norm": 0.000699989905115217, + "learning_rate": 3.035355753894242e-06, + "loss": 0.2941, + "num_input_tokens_seen": 83378352, + "step": 39920 + }, + { + "epoch": 6.513092421894118, + "grad_norm": 0.0004414504219312221, + "learning_rate": 2.6721874575752477e-06, + "loss": 0.2854, + "num_input_tokens_seen": 83388128, + "step": 39925 + }, + { + "epoch": 6.513908149114936, + "grad_norm": 0.0007216911762952805, + "learning_rate": 2.3321506341933418e-06, + "loss": 0.3535, + "num_input_tokens_seen": 83398288, + "step": 39930 + }, + { + "epoch": 6.514723876335753, + "grad_norm": 0.0007060697535052896, + "learning_rate": 2.0152453361732546e-06, + "loss": 0.3134, + "num_input_tokens_seen": 83408256, + "step": 39935 + }, + { + "epoch": 6.5155396035565705, + "grad_norm": 0.0005846042768098414, + "learning_rate": 1.7214716123925554e-06, + "loss": 0.3362, + "num_input_tokens_seen": 83418560, + "step": 39940 + }, + { + "epoch": 6.5163553307773885, + "grad_norm": 0.00042387054418213665, + "learning_rate": 1.4508295081649968e-06, + "loss": 0.3013, + "num_input_tokens_seen": 83429440, + "step": 39945 + }, + { + "epoch": 6.517171057998206, + "grad_norm": 0.0004649171023629606, + "learning_rate": 1.2033190652238623e-06, + "loss": 0.2943, + "num_input_tokens_seen": 83439312, + "step": 39950 + }, + { + "epoch": 6.517986785219023, + "grad_norm": 0.0004693925438914448, + "learning_rate": 9.78940321721966e-07, + "loss": 0.2957, + "num_input_tokens_seen": 83449776, + "step": 39955 + }, + { + "epoch": 6.51880251243984, + "grad_norm": 0.0007705246098339558, + "learning_rate": 7.776933122816132e-07, + "loss": 0.3523, + "num_input_tokens_seen": 83460704, + "step": 39960 + }, + { + "epoch": 6.519618239660657, + "grad_norm": 0.0005077666719444096, + "learning_rate": 5.99578067927986e-07, + "loss": 0.3477, + "num_input_tokens_seen": 83470784, + "step": 39965 + }, + { + "epoch": 6.520433966881475, + "grad_norm": 0.0005825280677527189, + "learning_rate": 4.445946161224512e-07, + "loss": 0.2752, + "num_input_tokens_seen": 83481760, + "step": 39970 + }, + { + "epoch": 6.521249694102292, + "grad_norm": 0.00048119391431100667, + "learning_rate": 3.127429807792126e-07, + "loss": 0.2927, + "num_input_tokens_seen": 83490128, + "step": 39975 + }, + { + "epoch": 6.522065421323109, + "grad_norm": 0.0012440041173249483, + "learning_rate": 2.040231822320049e-07, + "loss": 0.2386, + "num_input_tokens_seen": 83499984, + "step": 39980 + }, + { + "epoch": 6.522881148543927, + "grad_norm": 0.000647709530312568, + "learning_rate": 1.1843523723409354e-07, + "loss": 0.2995, + "num_input_tokens_seen": 83511088, + "step": 39985 + }, + { + "epoch": 6.523696875764744, + "grad_norm": 0.0005797874764539301, + "learning_rate": 5.597915897492811e-08, + "loss": 0.3238, + "num_input_tokens_seen": 83522192, + "step": 39990 + }, + { + "epoch": 6.524512602985562, + "grad_norm": 0.0007255621603690088, + "learning_rate": 1.6654957113448885e-08, + "loss": 0.3434, + "num_input_tokens_seen": 83532624, + "step": 39995 + }, + { + "epoch": 6.525328330206379, + "grad_norm": 0.0007402482442557812, + "learning_rate": 4.626377114735902e-10, + "loss": 0.3072, + "num_input_tokens_seen": 83543088, + "step": 40000 + }, + { + "epoch": 6.525328330206379, + "eval_loss": 0.315355122089386, + "eval_runtime": 155.8402, + "eval_samples_per_second": 17.486, + "eval_steps_per_second": 8.746, + "num_input_tokens_seen": 83543088, + "step": 40000 + }, + { + "epoch": 6.525328330206379, + "num_input_tokens_seen": 83543088, + "step": 40000, + "total_flos": 3.565855835816067e+18, + "train_loss": 0.324408847618103, + "train_runtime": 51929.5354, + "train_samples_per_second": 3.081, + "train_steps_per_second": 0.77 + } + ], + "logging_steps": 5, + "max_steps": 40000, + "num_input_tokens_seen": 83543088, + "num_train_epochs": 7, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 3.565855835816067e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}