diff --git "a/checkpoint-2289587/trainer_state.json" "b/checkpoint-2289587/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-2289587/trainer_state.json" @@ -0,0 +1,27091 @@ +{ + "best_metric": 3.9781768321990967, + "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/2/checkpoints/checkpoint-2289587", + "epoch": 1.0250006060157382, + "eval_steps": 10, + "global_step": 2289587, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "learning_rate": 4.999998362119627e-05, + "loss": 10.8204, + "step": 1 + }, + { + "epoch": 0.0, + "learning_rate": 4.999161405248948e-05, + "loss": 7.5681, + "step": 512 + }, + { + "epoch": 0.0, + "learning_rate": 4.998322810497896e-05, + "loss": 7.0604, + "step": 1024 + }, + { + "epoch": 0.0, + "learning_rate": 4.997484215746844e-05, + "loss": 6.9893, + "step": 1536 + }, + { + "epoch": 0.0, + "learning_rate": 4.996645620995792e-05, + "loss": 6.9585, + "step": 2048 + }, + { + "epoch": 0.0, + "learning_rate": 4.99580702624474e-05, + "loss": 6.8857, + "step": 2560 + }, + { + "epoch": 0.0, + "learning_rate": 4.994968431493688e-05, + "loss": 6.7423, + "step": 3072 + }, + { + "epoch": 0.0, + "learning_rate": 4.994129836742636e-05, + "loss": 6.6348, + "step": 3584 + }, + { + "epoch": 0.0, + "learning_rate": 4.993291241991584e-05, + "loss": 6.5475, + "step": 4096 + }, + { + "epoch": 0.0, + "learning_rate": 4.992452647240532e-05, + "loss": 6.4551, + "step": 4608 + }, + { + "epoch": 0.0, + "learning_rate": 4.99161405248948e-05, + "loss": 6.406, + "step": 5120 + }, + { + "epoch": 0.0, + "learning_rate": 4.990775457738428e-05, + "loss": 6.3398, + "step": 5632 + }, + { + "epoch": 0.0, + "learning_rate": 4.989938500867749e-05, + "loss": 6.2764, + "step": 6144 + }, + { + "epoch": 0.0, + "learning_rate": 4.989099906116697e-05, + "loss": 6.2064, + "step": 6656 + }, + { + "epoch": 0.0, + "learning_rate": 4.988261311365645e-05, + "loss": 6.1503, + "step": 7168 + }, + { + "epoch": 0.0, + "learning_rate": 4.987422716614593e-05, + "loss": 6.103, + "step": 7680 + }, + { + "epoch": 0.0, + "learning_rate": 4.986584121863541e-05, + "loss": 6.0533, + "step": 8192 + }, + { + "epoch": 0.0, + "learning_rate": 4.985745527112489e-05, + "loss": 5.9976, + "step": 8704 + }, + { + "epoch": 0.0, + "learning_rate": 4.984906932361437e-05, + "loss": 5.9613, + "step": 9216 + }, + { + "epoch": 0.0, + "learning_rate": 4.984068337610385e-05, + "loss": 5.9256, + "step": 9728 + }, + { + "epoch": 0.0, + "learning_rate": 4.983229742859333e-05, + "loss": 5.8795, + "step": 10240 + }, + { + "epoch": 0.0, + "learning_rate": 4.982391148108281e-05, + "loss": 5.8432, + "step": 10752 + }, + { + "epoch": 0.0, + "learning_rate": 4.9815541912376026e-05, + "loss": 5.8223, + "step": 11264 + }, + { + "epoch": 0.0, + "learning_rate": 4.9807155964865506e-05, + "loss": 5.7856, + "step": 11776 + }, + { + "epoch": 0.0, + "learning_rate": 4.9798770017354986e-05, + "loss": 5.7637, + "step": 12288 + }, + { + "epoch": 0.0, + "learning_rate": 4.9790384069844466e-05, + "loss": 5.7367, + "step": 12800 + }, + { + "epoch": 0.0, + "learning_rate": 4.9781998122333946e-05, + "loss": 5.7145, + "step": 13312 + }, + { + "epoch": 0.0, + "learning_rate": 4.9773628553627155e-05, + "loss": 5.6845, + "step": 13824 + }, + { + "epoch": 0.0, + "learning_rate": 4.9765242606116635e-05, + "loss": 5.6507, + "step": 14336 + }, + { + "epoch": 0.0, + "learning_rate": 4.9756856658606115e-05, + "loss": 5.645, + "step": 14848 + }, + { + "epoch": 0.01, + "learning_rate": 4.9748470711095595e-05, + "loss": 5.6112, + "step": 15360 + }, + { + "epoch": 0.01, + "learning_rate": 4.974008476358507e-05, + "loss": 5.5953, + "step": 15872 + }, + { + "epoch": 0.01, + "learning_rate": 4.973169881607455e-05, + "loss": 5.5703, + "step": 16384 + }, + { + "epoch": 0.01, + "learning_rate": 4.972331286856403e-05, + "loss": 5.5656, + "step": 16896 + }, + { + "epoch": 0.01, + "learning_rate": 4.9714943299857244e-05, + "loss": 5.5349, + "step": 17408 + }, + { + "epoch": 0.01, + "learning_rate": 4.970657373115046e-05, + "loss": 5.5254, + "step": 17920 + }, + { + "epoch": 0.01, + "learning_rate": 4.969818778363994e-05, + "loss": 5.5123, + "step": 18432 + }, + { + "epoch": 0.01, + "learning_rate": 4.968980183612942e-05, + "loss": 5.4926, + "step": 18944 + }, + { + "epoch": 0.01, + "learning_rate": 4.96814158886189e-05, + "loss": 5.4825, + "step": 19456 + }, + { + "epoch": 0.01, + "learning_rate": 4.967302994110837e-05, + "loss": 5.4603, + "step": 19968 + }, + { + "epoch": 0.01, + "learning_rate": 4.966464399359785e-05, + "loss": 5.4357, + "step": 20480 + }, + { + "epoch": 0.01, + "learning_rate": 4.965625804608733e-05, + "loss": 5.4296, + "step": 20992 + }, + { + "epoch": 0.01, + "learning_rate": 4.964788847738054e-05, + "loss": 5.4028, + "step": 21504 + }, + { + "epoch": 0.01, + "learning_rate": 4.963950252987002e-05, + "loss": 5.4151, + "step": 22016 + }, + { + "epoch": 0.01, + "learning_rate": 4.96311165823595e-05, + "loss": 5.3867, + "step": 22528 + }, + { + "epoch": 0.01, + "learning_rate": 4.962274701365272e-05, + "loss": 5.3832, + "step": 23040 + }, + { + "epoch": 0.01, + "learning_rate": 4.96143610661422e-05, + "loss": 5.3678, + "step": 23552 + }, + { + "epoch": 0.01, + "learning_rate": 4.960597511863168e-05, + "loss": 5.3657, + "step": 24064 + }, + { + "epoch": 0.01, + "learning_rate": 4.959758917112116e-05, + "loss": 5.3572, + "step": 24576 + }, + { + "epoch": 0.01, + "learning_rate": 4.958920322361064e-05, + "loss": 5.3291, + "step": 25088 + }, + { + "epoch": 0.01, + "learning_rate": 4.958081727610012e-05, + "loss": 5.3171, + "step": 25600 + }, + { + "epoch": 0.01, + "learning_rate": 4.957244770739333e-05, + "loss": 5.3093, + "step": 26112 + }, + { + "epoch": 0.01, + "learning_rate": 4.956406175988281e-05, + "loss": 5.316, + "step": 26624 + }, + { + "epoch": 0.01, + "learning_rate": 4.955567581237229e-05, + "loss": 5.2922, + "step": 27136 + }, + { + "epoch": 0.01, + "learning_rate": 4.954728986486177e-05, + "loss": 5.3001, + "step": 27648 + }, + { + "epoch": 0.01, + "learning_rate": 4.953890391735125e-05, + "loss": 5.2701, + "step": 28160 + }, + { + "epoch": 0.01, + "learning_rate": 4.9530517969840727e-05, + "loss": 5.2667, + "step": 28672 + }, + { + "epoch": 0.01, + "learning_rate": 4.9522132022330207e-05, + "loss": 5.2549, + "step": 29184 + }, + { + "epoch": 0.01, + "learning_rate": 4.951376245362342e-05, + "loss": 5.2482, + "step": 29696 + }, + { + "epoch": 0.01, + "learning_rate": 4.95053765061129e-05, + "loss": 5.2455, + "step": 30208 + }, + { + "epoch": 0.01, + "learning_rate": 4.949699055860238e-05, + "loss": 5.2247, + "step": 30720 + }, + { + "epoch": 0.01, + "learning_rate": 4.948860461109186e-05, + "loss": 5.2189, + "step": 31232 + }, + { + "epoch": 0.01, + "learning_rate": 4.94802514211888e-05, + "loss": 5.2231, + "step": 31744 + }, + { + "epoch": 0.01, + "learning_rate": 4.947186547367828e-05, + "loss": 5.202, + "step": 32256 + }, + { + "epoch": 0.01, + "learning_rate": 4.946347952616776e-05, + "loss": 5.1997, + "step": 32768 + }, + { + "epoch": 0.01, + "learning_rate": 4.945509357865724e-05, + "loss": 5.1885, + "step": 33280 + }, + { + "epoch": 0.01, + "learning_rate": 4.944670763114672e-05, + "loss": 5.1889, + "step": 33792 + }, + { + "epoch": 0.01, + "learning_rate": 4.94383216836362e-05, + "loss": 5.1736, + "step": 34304 + }, + { + "epoch": 0.01, + "learning_rate": 4.942993573612568e-05, + "loss": 5.1751, + "step": 34816 + }, + { + "epoch": 0.01, + "learning_rate": 4.942156616741889e-05, + "loss": 5.1639, + "step": 35328 + }, + { + "epoch": 0.01, + "learning_rate": 4.941318021990837e-05, + "loss": 5.1534, + "step": 35840 + }, + { + "epoch": 0.01, + "learning_rate": 4.9404794272397856e-05, + "loss": 5.1556, + "step": 36352 + }, + { + "epoch": 0.01, + "learning_rate": 4.9396408324887336e-05, + "loss": 5.1407, + "step": 36864 + }, + { + "epoch": 0.01, + "learning_rate": 4.9388022377376816e-05, + "loss": 5.1494, + "step": 37376 + }, + { + "epoch": 0.01, + "learning_rate": 4.9379636429866296e-05, + "loss": 5.1466, + "step": 37888 + }, + { + "epoch": 0.01, + "learning_rate": 4.9371250482355776e-05, + "loss": 5.1375, + "step": 38400 + }, + { + "epoch": 0.01, + "learning_rate": 4.9362880913648985e-05, + "loss": 5.1181, + "step": 38912 + }, + { + "epoch": 0.01, + "learning_rate": 4.9354511344942194e-05, + "loss": 5.1031, + "step": 39424 + }, + { + "epoch": 0.01, + "learning_rate": 4.9346125397431674e-05, + "loss": 5.0962, + "step": 39936 + }, + { + "epoch": 0.01, + "learning_rate": 4.9337739449921154e-05, + "loss": 5.1009, + "step": 40448 + }, + { + "epoch": 0.01, + "learning_rate": 4.9329353502410634e-05, + "loss": 5.105, + "step": 40960 + }, + { + "epoch": 0.01, + "learning_rate": 4.9320967554900114e-05, + "loss": 5.0984, + "step": 41472 + }, + { + "epoch": 0.01, + "learning_rate": 4.9312581607389594e-05, + "loss": 5.0732, + "step": 41984 + }, + { + "epoch": 0.01, + "learning_rate": 4.9304195659879074e-05, + "loss": 5.0678, + "step": 42496 + }, + { + "epoch": 0.01, + "learning_rate": 4.9295809712368554e-05, + "loss": 5.0758, + "step": 43008 + }, + { + "epoch": 0.01, + "learning_rate": 4.928744014366177e-05, + "loss": 5.0612, + "step": 43520 + }, + { + "epoch": 0.01, + "learning_rate": 4.927907057495498e-05, + "loss": 5.0598, + "step": 44032 + }, + { + "epoch": 0.01, + "learning_rate": 4.927068462744446e-05, + "loss": 5.0538, + "step": 44544 + }, + { + "epoch": 0.01, + "learning_rate": 4.926229867993394e-05, + "loss": 5.0557, + "step": 45056 + }, + { + "epoch": 0.01, + "learning_rate": 4.925391273242342e-05, + "loss": 5.0432, + "step": 45568 + }, + { + "epoch": 0.02, + "learning_rate": 4.92455267849129e-05, + "loss": 5.0361, + "step": 46080 + }, + { + "epoch": 0.02, + "learning_rate": 4.923714083740238e-05, + "loss": 5.0219, + "step": 46592 + }, + { + "epoch": 0.02, + "learning_rate": 4.922877126869559e-05, + "loss": 5.0354, + "step": 47104 + }, + { + "epoch": 0.02, + "learning_rate": 4.922038532118507e-05, + "loss": 5.0204, + "step": 47616 + }, + { + "epoch": 0.02, + "learning_rate": 4.921199937367455e-05, + "loss": 5.0045, + "step": 48128 + }, + { + "epoch": 0.02, + "learning_rate": 4.920361342616403e-05, + "loss": 5.004, + "step": 48640 + }, + { + "epoch": 0.02, + "learning_rate": 4.919524385745724e-05, + "loss": 5.0162, + "step": 49152 + }, + { + "epoch": 0.02, + "learning_rate": 4.918685790994672e-05, + "loss": 5.0063, + "step": 49664 + }, + { + "epoch": 0.02, + "learning_rate": 4.91784719624362e-05, + "loss": 4.9989, + "step": 50176 + }, + { + "epoch": 0.02, + "learning_rate": 4.9170086014925676e-05, + "loss": 4.9854, + "step": 50688 + }, + { + "epoch": 0.02, + "learning_rate": 4.9161700067415156e-05, + "loss": 4.9901, + "step": 51200 + }, + { + "epoch": 0.02, + "learning_rate": 4.915333049870837e-05, + "loss": 4.9692, + "step": 51712 + }, + { + "epoch": 0.02, + "learning_rate": 4.914494455119785e-05, + "loss": 4.9842, + "step": 52224 + }, + { + "epoch": 0.02, + "learning_rate": 4.9136558603687325e-05, + "loss": 4.9856, + "step": 52736 + }, + { + "epoch": 0.02, + "learning_rate": 4.9128172656176805e-05, + "loss": 4.9622, + "step": 53248 + }, + { + "epoch": 0.02, + "learning_rate": 4.9119786708666285e-05, + "loss": 4.9544, + "step": 53760 + }, + { + "epoch": 0.02, + "learning_rate": 4.9111400761155765e-05, + "loss": 4.9322, + "step": 54272 + }, + { + "epoch": 0.02, + "learning_rate": 4.9103014813645245e-05, + "loss": 4.943, + "step": 54784 + }, + { + "epoch": 0.02, + "learning_rate": 4.909462886613473e-05, + "loss": 4.9443, + "step": 55296 + }, + { + "epoch": 0.02, + "learning_rate": 4.908625929742794e-05, + "loss": 4.9379, + "step": 55808 + }, + { + "epoch": 0.02, + "learning_rate": 4.907787334991742e-05, + "loss": 4.9413, + "step": 56320 + }, + { + "epoch": 0.02, + "learning_rate": 4.90694874024069e-05, + "loss": 4.9229, + "step": 56832 + }, + { + "epoch": 0.02, + "learning_rate": 4.906110145489638e-05, + "loss": 4.9241, + "step": 57344 + }, + { + "epoch": 0.02, + "learning_rate": 4.905271550738586e-05, + "loss": 4.9263, + "step": 57856 + }, + { + "epoch": 0.02, + "learning_rate": 4.904434593867907e-05, + "loss": 4.9277, + "step": 58368 + }, + { + "epoch": 0.02, + "learning_rate": 4.903595999116855e-05, + "loss": 4.9146, + "step": 58880 + }, + { + "epoch": 0.02, + "learning_rate": 4.902757404365803e-05, + "loss": 4.9041, + "step": 59392 + }, + { + "epoch": 0.02, + "learning_rate": 4.901918809614751e-05, + "loss": 4.9048, + "step": 59904 + }, + { + "epoch": 0.02, + "learning_rate": 4.901080214863699e-05, + "loss": 4.9077, + "step": 60416 + }, + { + "epoch": 0.02, + "learning_rate": 4.900241620112647e-05, + "loss": 4.9035, + "step": 60928 + }, + { + "epoch": 0.02, + "learning_rate": 4.8994046632419686e-05, + "loss": 4.8985, + "step": 61440 + }, + { + "epoch": 0.02, + "learning_rate": 4.8985660684909166e-05, + "loss": 4.8885, + "step": 61952 + }, + { + "epoch": 0.02, + "learning_rate": 4.8977274737398646e-05, + "loss": 4.8801, + "step": 62464 + }, + { + "epoch": 0.02, + "learning_rate": 4.8968888789888125e-05, + "loss": 4.8719, + "step": 62976 + }, + { + "epoch": 0.02, + "learning_rate": 4.8960502842377605e-05, + "loss": 4.8796, + "step": 63488 + }, + { + "epoch": 0.02, + "learning_rate": 4.8952116894867085e-05, + "loss": 4.8783, + "step": 64000 + }, + { + "epoch": 0.02, + "learning_rate": 4.8943730947356565e-05, + "loss": 4.88, + "step": 64512 + }, + { + "epoch": 0.02, + "learning_rate": 4.893534499984604e-05, + "loss": 4.8544, + "step": 65024 + }, + { + "epoch": 0.02, + "learning_rate": 4.8926991809942984e-05, + "loss": 4.8673, + "step": 65536 + }, + { + "epoch": 0.02, + "learning_rate": 4.8918605862432463e-05, + "loss": 4.8672, + "step": 66048 + }, + { + "epoch": 0.02, + "learning_rate": 4.8910219914921943e-05, + "loss": 4.8537, + "step": 66560 + }, + { + "epoch": 0.02, + "learning_rate": 4.8901833967411423e-05, + "loss": 4.8684, + "step": 67072 + }, + { + "epoch": 0.02, + "learning_rate": 4.88934480199009e-05, + "loss": 4.8469, + "step": 67584 + }, + { + "epoch": 0.02, + "learning_rate": 4.888506207239038e-05, + "loss": 4.8611, + "step": 68096 + }, + { + "epoch": 0.02, + "learning_rate": 4.887667612487986e-05, + "loss": 4.8507, + "step": 68608 + }, + { + "epoch": 0.02, + "learning_rate": 4.886830655617308e-05, + "loss": 4.8302, + "step": 69120 + }, + { + "epoch": 0.02, + "learning_rate": 4.885992060866256e-05, + "loss": 4.8292, + "step": 69632 + }, + { + "epoch": 0.02, + "learning_rate": 4.885153466115204e-05, + "loss": 4.8293, + "step": 70144 + }, + { + "epoch": 0.02, + "learning_rate": 4.884316509244525e-05, + "loss": 4.8315, + "step": 70656 + }, + { + "epoch": 0.02, + "learning_rate": 4.883477914493473e-05, + "loss": 4.8373, + "step": 71168 + }, + { + "epoch": 0.02, + "learning_rate": 4.882639319742421e-05, + "loss": 4.8324, + "step": 71680 + }, + { + "epoch": 0.02, + "learning_rate": 4.881800724991369e-05, + "loss": 4.8258, + "step": 72192 + }, + { + "epoch": 0.02, + "learning_rate": 4.880962130240316e-05, + "loss": 4.8156, + "step": 72704 + }, + { + "epoch": 0.02, + "learning_rate": 4.880123535489264e-05, + "loss": 4.8045, + "step": 73216 + }, + { + "epoch": 0.02, + "learning_rate": 4.879284940738212e-05, + "loss": 4.8034, + "step": 73728 + }, + { + "epoch": 0.02, + "learning_rate": 4.87844634598716e-05, + "loss": 4.811, + "step": 74240 + }, + { + "epoch": 0.02, + "learning_rate": 4.877607751236109e-05, + "loss": 4.8169, + "step": 74752 + }, + { + "epoch": 0.02, + "learning_rate": 4.876769156485057e-05, + "loss": 4.8081, + "step": 75264 + }, + { + "epoch": 0.02, + "learning_rate": 4.875930561734005e-05, + "loss": 4.7918, + "step": 75776 + }, + { + "epoch": 0.02, + "learning_rate": 4.875091966982953e-05, + "loss": 4.7917, + "step": 76288 + }, + { + "epoch": 0.03, + "eval_loss": 4.756728172302246, + "eval_runtime": 527.8044, + "eval_samples_per_second": 722.978, + "eval_steps_per_second": 22.594, + "step": 76319 + }, + { + "epoch": 1.0, + "learning_rate": 4.874255010112274e-05, + "loss": 4.7996, + "step": 76800 + }, + { + "epoch": 1.0, + "learning_rate": 4.873416415361222e-05, + "loss": 4.7917, + "step": 77312 + }, + { + "epoch": 1.0, + "learning_rate": 4.8725794584905426e-05, + "loss": 4.7837, + "step": 77824 + }, + { + "epoch": 1.0, + "learning_rate": 4.8717408637394906e-05, + "loss": 4.7826, + "step": 78336 + }, + { + "epoch": 1.0, + "learning_rate": 4.8709022689884386e-05, + "loss": 4.7726, + "step": 78848 + }, + { + "epoch": 1.0, + "learning_rate": 4.8700636742373866e-05, + "loss": 4.7569, + "step": 79360 + }, + { + "epoch": 1.0, + "learning_rate": 4.8692250794863346e-05, + "loss": 4.7672, + "step": 79872 + }, + { + "epoch": 1.0, + "learning_rate": 4.8683864847352826e-05, + "loss": 4.7749, + "step": 80384 + }, + { + "epoch": 1.0, + "learning_rate": 4.8675478899842306e-05, + "loss": 4.7578, + "step": 80896 + }, + { + "epoch": 1.0, + "learning_rate": 4.866710933113552e-05, + "loss": 4.7657, + "step": 81408 + }, + { + "epoch": 1.0, + "learning_rate": 4.8658723383625e-05, + "loss": 4.7682, + "step": 81920 + }, + { + "epoch": 1.0, + "learning_rate": 4.865033743611448e-05, + "loss": 4.7687, + "step": 82432 + }, + { + "epoch": 1.0, + "learning_rate": 4.864195148860396e-05, + "loss": 4.7568, + "step": 82944 + }, + { + "epoch": 1.0, + "learning_rate": 4.863356554109344e-05, + "loss": 4.7542, + "step": 83456 + }, + { + "epoch": 1.0, + "learning_rate": 4.862519597238665e-05, + "loss": 4.7512, + "step": 83968 + }, + { + "epoch": 1.0, + "learning_rate": 4.861681002487613e-05, + "loss": 4.7415, + "step": 84480 + }, + { + "epoch": 1.0, + "learning_rate": 4.860842407736561e-05, + "loss": 4.7437, + "step": 84992 + }, + { + "epoch": 1.0, + "learning_rate": 4.860003812985509e-05, + "loss": 4.7457, + "step": 85504 + }, + { + "epoch": 1.0, + "learning_rate": 4.859165218234457e-05, + "loss": 4.7443, + "step": 86016 + }, + { + "epoch": 1.0, + "learning_rate": 4.858326623483405e-05, + "loss": 4.7428, + "step": 86528 + }, + { + "epoch": 1.0, + "learning_rate": 4.857488028732352e-05, + "loss": 4.7293, + "step": 87040 + }, + { + "epoch": 1.0, + "learning_rate": 4.856649433981301e-05, + "loss": 4.7408, + "step": 87552 + }, + { + "epoch": 1.0, + "learning_rate": 4.8558141149909955e-05, + "loss": 4.7278, + "step": 88064 + }, + { + "epoch": 1.0, + "learning_rate": 4.8549755202399435e-05, + "loss": 4.7368, + "step": 88576 + }, + { + "epoch": 1.0, + "learning_rate": 4.8541369254888915e-05, + "loss": 4.726, + "step": 89088 + }, + { + "epoch": 1.0, + "learning_rate": 4.8532983307378395e-05, + "loss": 4.7245, + "step": 89600 + }, + { + "epoch": 1.0, + "learning_rate": 4.8524613738671604e-05, + "loss": 4.7266, + "step": 90112 + }, + { + "epoch": 1.0, + "learning_rate": 4.8516227791161084e-05, + "loss": 4.7079, + "step": 90624 + }, + { + "epoch": 1.0, + "learning_rate": 4.8507841843650564e-05, + "loss": 4.7216, + "step": 91136 + }, + { + "epoch": 1.01, + "learning_rate": 4.849947227494377e-05, + "loss": 4.7026, + "step": 91648 + }, + { + "epoch": 1.01, + "learning_rate": 4.849108632743325e-05, + "loss": 4.7089, + "step": 92160 + }, + { + "epoch": 1.01, + "learning_rate": 4.848270037992273e-05, + "loss": 4.7054, + "step": 92672 + }, + { + "epoch": 1.01, + "learning_rate": 4.847431443241221e-05, + "loss": 4.7136, + "step": 93184 + }, + { + "epoch": 1.01, + "learning_rate": 4.846592848490169e-05, + "loss": 4.7009, + "step": 93696 + }, + { + "epoch": 1.01, + "learning_rate": 4.845754253739117e-05, + "loss": 4.6935, + "step": 94208 + }, + { + "epoch": 1.01, + "learning_rate": 4.844915658988065e-05, + "loss": 4.7101, + "step": 94720 + }, + { + "epoch": 1.01, + "learning_rate": 4.844077064237013e-05, + "loss": 4.6965, + "step": 95232 + }, + { + "epoch": 1.01, + "learning_rate": 4.843238469485961e-05, + "loss": 4.7036, + "step": 95744 + }, + { + "epoch": 1.01, + "learning_rate": 4.842399874734909e-05, + "loss": 4.692, + "step": 96256 + }, + { + "epoch": 1.01, + "learning_rate": 4.841561279983857e-05, + "loss": 4.6775, + "step": 96768 + }, + { + "epoch": 1.01, + "learning_rate": 4.840722685232805e-05, + "loss": 4.6814, + "step": 97280 + }, + { + "epoch": 1.01, + "learning_rate": 4.839885728362126e-05, + "loss": 4.6729, + "step": 97792 + }, + { + "epoch": 1.01, + "learning_rate": 4.839047133611074e-05, + "loss": 4.6871, + "step": 98304 + }, + { + "epoch": 1.01, + "learning_rate": 4.838208538860022e-05, + "loss": 4.6764, + "step": 98816 + }, + { + "epoch": 1.01, + "learning_rate": 4.837371581989343e-05, + "loss": 4.6829, + "step": 99328 + }, + { + "epoch": 1.01, + "learning_rate": 4.836532987238292e-05, + "loss": 4.6749, + "step": 99840 + }, + { + "epoch": 1.01, + "learning_rate": 4.83569439248724e-05, + "loss": 4.6811, + "step": 100352 + }, + { + "epoch": 1.01, + "learning_rate": 4.834855797736188e-05, + "loss": 4.6767, + "step": 100864 + }, + { + "epoch": 1.01, + "learning_rate": 4.834017202985136e-05, + "loss": 4.6531, + "step": 101376 + }, + { + "epoch": 1.01, + "learning_rate": 4.8331802461144566e-05, + "loss": 4.6575, + "step": 101888 + }, + { + "epoch": 1.01, + "learning_rate": 4.8323416513634046e-05, + "loss": 4.6556, + "step": 102400 + }, + { + "epoch": 1.01, + "learning_rate": 4.8315030566123526e-05, + "loss": 4.6685, + "step": 102912 + }, + { + "epoch": 1.01, + "learning_rate": 4.8306644618613006e-05, + "loss": 4.6476, + "step": 103424 + }, + { + "epoch": 1.01, + "learning_rate": 4.8298275049906215e-05, + "loss": 4.67, + "step": 103936 + }, + { + "epoch": 1.01, + "learning_rate": 4.8289889102395695e-05, + "loss": 4.6461, + "step": 104448 + }, + { + "epoch": 1.01, + "learning_rate": 4.8281503154885175e-05, + "loss": 4.6456, + "step": 104960 + }, + { + "epoch": 1.01, + "learning_rate": 4.8273117207374655e-05, + "loss": 4.6474, + "step": 105472 + }, + { + "epoch": 1.01, + "learning_rate": 4.8264731259864135e-05, + "loss": 4.643, + "step": 105984 + }, + { + "epoch": 1.01, + "learning_rate": 4.8256345312353615e-05, + "loss": 4.6445, + "step": 106496 + }, + { + "epoch": 1.01, + "learning_rate": 4.8247959364843095e-05, + "loss": 4.6368, + "step": 107008 + }, + { + "epoch": 1.01, + "learning_rate": 4.823957341733258e-05, + "loss": 4.6331, + "step": 107520 + }, + { + "epoch": 1.01, + "learning_rate": 4.823120384862579e-05, + "loss": 4.6438, + "step": 108032 + }, + { + "epoch": 1.01, + "learning_rate": 4.822281790111527e-05, + "loss": 4.6294, + "step": 108544 + }, + { + "epoch": 1.01, + "learning_rate": 4.821443195360475e-05, + "loss": 4.63, + "step": 109056 + }, + { + "epoch": 1.01, + "learning_rate": 4.820607876370169e-05, + "loss": 4.6273, + "step": 109568 + }, + { + "epoch": 1.01, + "learning_rate": 4.819769281619117e-05, + "loss": 4.6335, + "step": 110080 + }, + { + "epoch": 1.01, + "learning_rate": 4.818930686868065e-05, + "loss": 4.6161, + "step": 110592 + }, + { + "epoch": 1.01, + "learning_rate": 4.818092092117013e-05, + "loss": 4.6358, + "step": 111104 + }, + { + "epoch": 1.01, + "learning_rate": 4.817253497365961e-05, + "loss": 4.6224, + "step": 111616 + }, + { + "epoch": 1.01, + "learning_rate": 4.816414902614909e-05, + "loss": 4.6219, + "step": 112128 + }, + { + "epoch": 1.01, + "learning_rate": 4.815576307863857e-05, + "loss": 4.6259, + "step": 112640 + }, + { + "epoch": 1.01, + "learning_rate": 4.814737713112805e-05, + "loss": 4.6146, + "step": 113152 + }, + { + "epoch": 1.01, + "learning_rate": 4.8138991183617535e-05, + "loss": 4.6332, + "step": 113664 + }, + { + "epoch": 1.01, + "learning_rate": 4.8130621614910745e-05, + "loss": 4.6339, + "step": 114176 + }, + { + "epoch": 1.01, + "learning_rate": 4.8122235667400224e-05, + "loss": 4.6308, + "step": 114688 + }, + { + "epoch": 1.01, + "learning_rate": 4.8113849719889704e-05, + "loss": 4.6108, + "step": 115200 + }, + { + "epoch": 1.01, + "learning_rate": 4.8105463772379184e-05, + "loss": 4.6034, + "step": 115712 + }, + { + "epoch": 1.01, + "learning_rate": 4.809707782486866e-05, + "loss": 4.6007, + "step": 116224 + }, + { + "epoch": 1.01, + "learning_rate": 4.808869187735814e-05, + "loss": 4.6107, + "step": 116736 + }, + { + "epoch": 1.01, + "learning_rate": 4.808033868745508e-05, + "loss": 4.6221, + "step": 117248 + }, + { + "epoch": 1.01, + "learning_rate": 4.807195273994456e-05, + "loss": 4.6134, + "step": 117760 + }, + { + "epoch": 1.01, + "learning_rate": 4.806356679243404e-05, + "loss": 4.6005, + "step": 118272 + }, + { + "epoch": 1.01, + "learning_rate": 4.805518084492352e-05, + "loss": 4.5877, + "step": 118784 + }, + { + "epoch": 1.01, + "learning_rate": 4.8046794897413e-05, + "loss": 4.6079, + "step": 119296 + }, + { + "epoch": 1.01, + "learning_rate": 4.803840894990248e-05, + "loss": 4.5907, + "step": 119808 + }, + { + "epoch": 1.01, + "learning_rate": 4.803002300239196e-05, + "loss": 4.5951, + "step": 120320 + }, + { + "epoch": 1.01, + "learning_rate": 4.802163705488144e-05, + "loss": 4.5976, + "step": 120832 + }, + { + "epoch": 1.01, + "learning_rate": 4.801325110737092e-05, + "loss": 4.6065, + "step": 121344 + }, + { + "epoch": 1.01, + "learning_rate": 4.80048651598604e-05, + "loss": 4.5929, + "step": 121856 + }, + { + "epoch": 1.02, + "learning_rate": 4.799649559115361e-05, + "loss": 4.5897, + "step": 122368 + }, + { + "epoch": 1.02, + "learning_rate": 4.798810964364309e-05, + "loss": 4.5805, + "step": 122880 + }, + { + "epoch": 1.02, + "learning_rate": 4.797972369613257e-05, + "loss": 4.5923, + "step": 123392 + }, + { + "epoch": 1.02, + "learning_rate": 4.797133774862205e-05, + "loss": 4.5917, + "step": 123904 + }, + { + "epoch": 1.02, + "learning_rate": 4.796295180111153e-05, + "loss": 4.568, + "step": 124416 + }, + { + "epoch": 1.02, + "learning_rate": 4.795456585360101e-05, + "loss": 4.5767, + "step": 124928 + }, + { + "epoch": 1.02, + "learning_rate": 4.794617990609049e-05, + "loss": 4.591, + "step": 125440 + }, + { + "epoch": 1.02, + "learning_rate": 4.793779395857997e-05, + "loss": 4.5918, + "step": 125952 + }, + { + "epoch": 1.02, + "learning_rate": 4.792942438987319e-05, + "loss": 4.5789, + "step": 126464 + }, + { + "epoch": 1.02, + "learning_rate": 4.7921054821166396e-05, + "loss": 4.5736, + "step": 126976 + }, + { + "epoch": 1.02, + "learning_rate": 4.7912668873655876e-05, + "loss": 4.5832, + "step": 127488 + }, + { + "epoch": 1.02, + "learning_rate": 4.7904282926145356e-05, + "loss": 4.5626, + "step": 128000 + }, + { + "epoch": 1.02, + "learning_rate": 4.7895896978634836e-05, + "loss": 4.5787, + "step": 128512 + }, + { + "epoch": 1.02, + "learning_rate": 4.7887511031124316e-05, + "loss": 4.5832, + "step": 129024 + }, + { + "epoch": 1.02, + "learning_rate": 4.7879125083613796e-05, + "loss": 4.57, + "step": 129536 + }, + { + "epoch": 1.02, + "learning_rate": 4.7870755514907005e-05, + "loss": 4.5682, + "step": 130048 + }, + { + "epoch": 1.02, + "learning_rate": 4.7862369567396485e-05, + "loss": 4.5469, + "step": 130560 + }, + { + "epoch": 1.02, + "learning_rate": 4.7853983619885965e-05, + "loss": 4.5514, + "step": 131072 + }, + { + "epoch": 1.02, + "learning_rate": 4.7845597672375445e-05, + "loss": 4.5592, + "step": 131584 + }, + { + "epoch": 1.02, + "learning_rate": 4.7837211724864925e-05, + "loss": 4.5602, + "step": 132096 + }, + { + "epoch": 1.02, + "learning_rate": 4.782882577735441e-05, + "loss": 4.5639, + "step": 132608 + }, + { + "epoch": 1.02, + "learning_rate": 4.782043982984389e-05, + "loss": 4.5505, + "step": 133120 + }, + { + "epoch": 1.02, + "learning_rate": 4.781205388233337e-05, + "loss": 4.5495, + "step": 133632 + }, + { + "epoch": 1.02, + "learning_rate": 4.780370069243031e-05, + "loss": 4.5566, + "step": 134144 + }, + { + "epoch": 1.02, + "learning_rate": 4.779531474491979e-05, + "loss": 4.5633, + "step": 134656 + }, + { + "epoch": 1.02, + "learning_rate": 4.778692879740927e-05, + "loss": 4.5511, + "step": 135168 + }, + { + "epoch": 1.02, + "learning_rate": 4.777854284989875e-05, + "loss": 4.5509, + "step": 135680 + }, + { + "epoch": 1.02, + "learning_rate": 4.777015690238823e-05, + "loss": 4.5482, + "step": 136192 + }, + { + "epoch": 1.02, + "learning_rate": 4.776177095487771e-05, + "loss": 4.5518, + "step": 136704 + }, + { + "epoch": 1.02, + "learning_rate": 4.775338500736719e-05, + "loss": 4.5514, + "step": 137216 + }, + { + "epoch": 1.02, + "learning_rate": 4.77450154386604e-05, + "loss": 4.5454, + "step": 137728 + }, + { + "epoch": 1.02, + "learning_rate": 4.773662949114988e-05, + "loss": 4.5423, + "step": 138240 + }, + { + "epoch": 1.02, + "learning_rate": 4.772824354363936e-05, + "loss": 4.5387, + "step": 138752 + }, + { + "epoch": 1.02, + "learning_rate": 4.7719857596128845e-05, + "loss": 4.5325, + "step": 139264 + }, + { + "epoch": 1.02, + "learning_rate": 4.7711488027422054e-05, + "loss": 4.5394, + "step": 139776 + }, + { + "epoch": 1.02, + "learning_rate": 4.7703102079911534e-05, + "loss": 4.5425, + "step": 140288 + }, + { + "epoch": 1.02, + "learning_rate": 4.7694716132401014e-05, + "loss": 4.5439, + "step": 140800 + }, + { + "epoch": 1.02, + "learning_rate": 4.7686330184890494e-05, + "loss": 4.5257, + "step": 141312 + }, + { + "epoch": 1.02, + "learning_rate": 4.76779606161837e-05, + "loss": 4.5343, + "step": 141824 + }, + { + "epoch": 1.02, + "learning_rate": 4.766957466867318e-05, + "loss": 4.5462, + "step": 142336 + }, + { + "epoch": 1.02, + "learning_rate": 4.766120509996639e-05, + "loss": 4.53, + "step": 142848 + }, + { + "epoch": 1.02, + "learning_rate": 4.765281915245587e-05, + "loss": 4.543, + "step": 143360 + }, + { + "epoch": 1.02, + "learning_rate": 4.764443320494535e-05, + "loss": 4.5281, + "step": 143872 + }, + { + "epoch": 1.02, + "learning_rate": 4.763604725743483e-05, + "loss": 4.5447, + "step": 144384 + }, + { + "epoch": 1.02, + "learning_rate": 4.762766130992431e-05, + "loss": 4.5322, + "step": 144896 + }, + { + "epoch": 1.02, + "learning_rate": 4.761927536241379e-05, + "loss": 4.5179, + "step": 145408 + }, + { + "epoch": 1.02, + "learning_rate": 4.761088941490327e-05, + "loss": 4.5205, + "step": 145920 + }, + { + "epoch": 1.02, + "learning_rate": 4.760250346739275e-05, + "loss": 4.5174, + "step": 146432 + }, + { + "epoch": 1.02, + "learning_rate": 4.759413389868596e-05, + "loss": 4.5278, + "step": 146944 + }, + { + "epoch": 1.02, + "learning_rate": 4.758574795117544e-05, + "loss": 4.5284, + "step": 147456 + }, + { + "epoch": 1.02, + "learning_rate": 4.757736200366492e-05, + "loss": 4.5304, + "step": 147968 + }, + { + "epoch": 1.02, + "learning_rate": 4.75689760561544e-05, + "loss": 4.5245, + "step": 148480 + }, + { + "epoch": 1.02, + "learning_rate": 4.756060648744761e-05, + "loss": 4.5162, + "step": 148992 + }, + { + "epoch": 1.02, + "learning_rate": 4.755222053993709e-05, + "loss": 4.511, + "step": 149504 + }, + { + "epoch": 1.02, + "learning_rate": 4.754383459242657e-05, + "loss": 4.5078, + "step": 150016 + }, + { + "epoch": 1.02, + "learning_rate": 4.753544864491605e-05, + "loss": 4.5183, + "step": 150528 + }, + { + "epoch": 1.02, + "learning_rate": 4.7527079076209266e-05, + "loss": 4.5301, + "step": 151040 + }, + { + "epoch": 1.02, + "learning_rate": 4.7518693128698745e-05, + "loss": 4.5147, + "step": 151552 + }, + { + "epoch": 1.02, + "learning_rate": 4.7510307181188225e-05, + "loss": 4.5066, + "step": 152064 + }, + { + "epoch": 1.02, + "learning_rate": 4.7501921233677705e-05, + "loss": 4.5064, + "step": 152576 + }, + { + "epoch": 1.03, + "eval_loss": 4.475635528564453, + "eval_runtime": 527.4249, + "eval_samples_per_second": 723.498, + "eval_steps_per_second": 22.61, + "step": 152638 + }, + { + "epoch": 0.0, + "learning_rate": 4.7493535286167185e-05, + "loss": 4.5181, + "step": 153088 + }, + { + "epoch": 0.0, + "learning_rate": 4.7485149338656665e-05, + "loss": 4.5126, + "step": 153600 + }, + { + "epoch": 0.0, + "learning_rate": 4.7476763391146145e-05, + "loss": 4.5005, + "step": 154112 + }, + { + "epoch": 0.0, + "learning_rate": 4.7468377443635625e-05, + "loss": 4.5026, + "step": 154624 + }, + { + "epoch": 0.0, + "learning_rate": 4.7459991496125105e-05, + "loss": 4.4975, + "step": 155136 + }, + { + "epoch": 0.0, + "learning_rate": 4.7451605548614585e-05, + "loss": 4.4844, + "step": 155648 + }, + { + "epoch": 0.0, + "learning_rate": 4.7443219601104065e-05, + "loss": 4.4942, + "step": 156160 + }, + { + "epoch": 0.0, + "learning_rate": 4.7434833653593545e-05, + "loss": 4.5052, + "step": 156672 + }, + { + "epoch": 0.0, + "learning_rate": 4.7426447706083025e-05, + "loss": 4.4844, + "step": 157184 + }, + { + "epoch": 0.0, + "learning_rate": 4.7418061758572505e-05, + "loss": 4.4965, + "step": 157696 + }, + { + "epoch": 0.0, + "learning_rate": 4.7409675811061985e-05, + "loss": 4.5059, + "step": 158208 + }, + { + "epoch": 0.0, + "learning_rate": 4.7401289863551465e-05, + "loss": 4.5019, + "step": 158720 + }, + { + "epoch": 0.0, + "learning_rate": 4.7392903916040945e-05, + "loss": 4.4908, + "step": 159232 + }, + { + "epoch": 0.0, + "learning_rate": 4.7384517968530425e-05, + "loss": 4.4909, + "step": 159744 + }, + { + "epoch": 0.0, + "learning_rate": 4.7376148399823634e-05, + "loss": 4.49, + "step": 160256 + }, + { + "epoch": 0.0, + "learning_rate": 4.7367762452313114e-05, + "loss": 4.4848, + "step": 160768 + }, + { + "epoch": 0.0, + "learning_rate": 4.7359376504802594e-05, + "loss": 4.4835, + "step": 161280 + }, + { + "epoch": 0.0, + "learning_rate": 4.7350990557292074e-05, + "loss": 4.4901, + "step": 161792 + }, + { + "epoch": 0.0, + "learning_rate": 4.7342604609781554e-05, + "loss": 4.4911, + "step": 162304 + }, + { + "epoch": 0.0, + "learning_rate": 4.7334218662271034e-05, + "loss": 4.489, + "step": 162816 + }, + { + "epoch": 0.0, + "learning_rate": 4.7325832714760514e-05, + "loss": 4.4792, + "step": 163328 + }, + { + "epoch": 0.0, + "learning_rate": 4.7317446767249994e-05, + "loss": 4.4901, + "step": 163840 + }, + { + "epoch": 0.0, + "learning_rate": 4.7309060819739474e-05, + "loss": 4.4802, + "step": 164352 + }, + { + "epoch": 0.0, + "learning_rate": 4.730069125103269e-05, + "loss": 4.4865, + "step": 164864 + }, + { + "epoch": 0.0, + "learning_rate": 4.729230530352217e-05, + "loss": 4.4808, + "step": 165376 + }, + { + "epoch": 0.0, + "learning_rate": 4.728391935601165e-05, + "loss": 4.4781, + "step": 165888 + }, + { + "epoch": 0.0, + "learning_rate": 4.727553340850113e-05, + "loss": 4.4836, + "step": 166400 + }, + { + "epoch": 0.0, + "learning_rate": 4.726716383979434e-05, + "loss": 4.4667, + "step": 166912 + }, + { + "epoch": 0.0, + "learning_rate": 4.725877789228382e-05, + "loss": 4.482, + "step": 167424 + }, + { + "epoch": 0.01, + "learning_rate": 4.72503919447733e-05, + "loss": 4.4673, + "step": 167936 + }, + { + "epoch": 0.01, + "learning_rate": 4.724202237606651e-05, + "loss": 4.4696, + "step": 168448 + }, + { + "epoch": 0.01, + "learning_rate": 4.723363642855599e-05, + "loss": 4.47, + "step": 168960 + }, + { + "epoch": 0.01, + "learning_rate": 4.722525048104547e-05, + "loss": 4.475, + "step": 169472 + }, + { + "epoch": 0.01, + "learning_rate": 4.721686453353495e-05, + "loss": 4.4687, + "step": 169984 + }, + { + "epoch": 0.01, + "learning_rate": 4.720847858602443e-05, + "loss": 4.4592, + "step": 170496 + }, + { + "epoch": 0.01, + "learning_rate": 4.720009263851391e-05, + "loss": 4.4793, + "step": 171008 + }, + { + "epoch": 0.01, + "learning_rate": 4.719170669100339e-05, + "loss": 4.466, + "step": 171520 + }, + { + "epoch": 0.01, + "learning_rate": 4.718332074349287e-05, + "loss": 4.4791, + "step": 172032 + }, + { + "epoch": 0.01, + "learning_rate": 4.717493479598235e-05, + "loss": 4.4662, + "step": 172544 + }, + { + "epoch": 0.01, + "learning_rate": 4.716656522727556e-05, + "loss": 4.4512, + "step": 173056 + }, + { + "epoch": 0.01, + "learning_rate": 4.715817927976504e-05, + "loss": 4.4521, + "step": 173568 + }, + { + "epoch": 0.01, + "learning_rate": 4.7149793332254516e-05, + "loss": 4.4527, + "step": 174080 + }, + { + "epoch": 0.01, + "learning_rate": 4.7141407384743996e-05, + "loss": 4.4696, + "step": 174592 + }, + { + "epoch": 0.01, + "learning_rate": 4.7133021437233476e-05, + "loss": 4.448, + "step": 175104 + }, + { + "epoch": 0.01, + "learning_rate": 4.7124635489722956e-05, + "loss": 4.4625, + "step": 175616 + }, + { + "epoch": 0.01, + "learning_rate": 4.7116265921016165e-05, + "loss": 4.4579, + "step": 176128 + }, + { + "epoch": 0.01, + "learning_rate": 4.7107879973505645e-05, + "loss": 4.4562, + "step": 176640 + }, + { + "epoch": 0.01, + "learning_rate": 4.7099494025995125e-05, + "loss": 4.4631, + "step": 177152 + }, + { + "epoch": 0.01, + "learning_rate": 4.7091108078484605e-05, + "loss": 4.4376, + "step": 177664 + }, + { + "epoch": 0.01, + "learning_rate": 4.708272213097409e-05, + "loss": 4.4436, + "step": 178176 + }, + { + "epoch": 0.01, + "learning_rate": 4.707433618346357e-05, + "loss": 4.4363, + "step": 178688 + }, + { + "epoch": 0.01, + "learning_rate": 4.706595023595305e-05, + "loss": 4.4587, + "step": 179200 + }, + { + "epoch": 0.01, + "learning_rate": 4.705756428844253e-05, + "loss": 4.4349, + "step": 179712 + }, + { + "epoch": 0.01, + "learning_rate": 4.704921109853947e-05, + "loss": 4.4607, + "step": 180224 + }, + { + "epoch": 0.01, + "learning_rate": 4.704082515102895e-05, + "loss": 4.4351, + "step": 180736 + }, + { + "epoch": 0.01, + "learning_rate": 4.703243920351843e-05, + "loss": 4.4369, + "step": 181248 + }, + { + "epoch": 0.01, + "learning_rate": 4.702405325600791e-05, + "loss": 4.4373, + "step": 181760 + }, + { + "epoch": 0.01, + "learning_rate": 4.701566730849739e-05, + "loss": 4.4338, + "step": 182272 + }, + { + "epoch": 0.01, + "learning_rate": 4.700728136098687e-05, + "loss": 4.4405, + "step": 182784 + }, + { + "epoch": 0.01, + "learning_rate": 4.699889541347635e-05, + "loss": 4.4347, + "step": 183296 + }, + { + "epoch": 0.01, + "learning_rate": 4.699052584476956e-05, + "loss": 4.4247, + "step": 183808 + }, + { + "epoch": 0.01, + "learning_rate": 4.6982139897259045e-05, + "loss": 4.4414, + "step": 184320 + }, + { + "epoch": 0.01, + "learning_rate": 4.6973753949748525e-05, + "loss": 4.4276, + "step": 184832 + }, + { + "epoch": 0.01, + "learning_rate": 4.6965368002238005e-05, + "loss": 4.4317, + "step": 185344 + }, + { + "epoch": 0.01, + "learning_rate": 4.6956982054727485e-05, + "loss": 4.4276, + "step": 185856 + }, + { + "epoch": 0.01, + "learning_rate": 4.6948612486020694e-05, + "loss": 4.4352, + "step": 186368 + }, + { + "epoch": 0.01, + "learning_rate": 4.6940226538510174e-05, + "loss": 4.4146, + "step": 186880 + }, + { + "epoch": 0.01, + "learning_rate": 4.6931840590999654e-05, + "loss": 4.4395, + "step": 187392 + }, + { + "epoch": 0.01, + "learning_rate": 4.6923454643489134e-05, + "loss": 4.4249, + "step": 187904 + }, + { + "epoch": 0.01, + "learning_rate": 4.6915068695978614e-05, + "loss": 4.4289, + "step": 188416 + }, + { + "epoch": 0.01, + "learning_rate": 4.6906682748468094e-05, + "loss": 4.4304, + "step": 188928 + }, + { + "epoch": 0.01, + "learning_rate": 4.6898296800957574e-05, + "loss": 4.4219, + "step": 189440 + }, + { + "epoch": 0.01, + "learning_rate": 4.688991085344705e-05, + "loss": 4.439, + "step": 189952 + }, + { + "epoch": 0.01, + "learning_rate": 4.688154128474026e-05, + "loss": 4.4408, + "step": 190464 + }, + { + "epoch": 0.01, + "learning_rate": 4.687315533722974e-05, + "loss": 4.4419, + "step": 190976 + }, + { + "epoch": 0.01, + "learning_rate": 4.686476938971923e-05, + "loss": 4.426, + "step": 191488 + }, + { + "epoch": 0.01, + "learning_rate": 4.68563834422087e-05, + "loss": 4.4157, + "step": 192000 + }, + { + "epoch": 0.01, + "learning_rate": 4.684801387350192e-05, + "loss": 4.4074, + "step": 192512 + }, + { + "epoch": 0.01, + "learning_rate": 4.68396279259914e-05, + "loss": 4.426, + "step": 193024 + }, + { + "epoch": 0.01, + "learning_rate": 4.683125835728461e-05, + "loss": 4.4331, + "step": 193536 + }, + { + "epoch": 0.01, + "learning_rate": 4.682287240977409e-05, + "loss": 4.4286, + "step": 194048 + }, + { + "epoch": 0.01, + "learning_rate": 4.681448646226357e-05, + "loss": 4.4152, + "step": 194560 + }, + { + "epoch": 0.01, + "learning_rate": 4.680610051475305e-05, + "loss": 4.4053, + "step": 195072 + }, + { + "epoch": 0.01, + "learning_rate": 4.679771456724252e-05, + "loss": 4.4199, + "step": 195584 + }, + { + "epoch": 0.01, + "learning_rate": 4.6789328619732e-05, + "loss": 4.4066, + "step": 196096 + }, + { + "epoch": 0.01, + "learning_rate": 4.678095905102522e-05, + "loss": 4.4112, + "step": 196608 + }, + { + "epoch": 0.01, + "learning_rate": 4.6772573103514697e-05, + "loss": 4.4196, + "step": 197120 + }, + { + "epoch": 0.01, + "learning_rate": 4.6764187156004177e-05, + "loss": 4.4196, + "step": 197632 + }, + { + "epoch": 0.01, + "learning_rate": 4.6755801208493657e-05, + "loss": 4.4123, + "step": 198144 + }, + { + "epoch": 0.02, + "learning_rate": 4.674743163978687e-05, + "loss": 4.416, + "step": 198656 + }, + { + "epoch": 0.02, + "learning_rate": 4.6739045692276346e-05, + "loss": 4.4005, + "step": 199168 + }, + { + "epoch": 0.02, + "learning_rate": 4.6730659744765826e-05, + "loss": 4.4153, + "step": 199680 + }, + { + "epoch": 0.02, + "learning_rate": 4.6722273797255305e-05, + "loss": 4.4108, + "step": 200192 + }, + { + "epoch": 0.02, + "learning_rate": 4.6713887849744785e-05, + "loss": 4.3975, + "step": 200704 + }, + { + "epoch": 0.02, + "learning_rate": 4.6705501902234265e-05, + "loss": 4.3998, + "step": 201216 + }, + { + "epoch": 0.02, + "learning_rate": 4.6697115954723745e-05, + "loss": 4.416, + "step": 201728 + }, + { + "epoch": 0.02, + "learning_rate": 4.6688730007213225e-05, + "loss": 4.4191, + "step": 202240 + }, + { + "epoch": 0.02, + "learning_rate": 4.6680360438506434e-05, + "loss": 4.4027, + "step": 202752 + }, + { + "epoch": 0.02, + "learning_rate": 4.667199086979965e-05, + "loss": 4.4049, + "step": 203264 + }, + { + "epoch": 0.02, + "learning_rate": 4.666360492228913e-05, + "loss": 4.411, + "step": 203776 + }, + { + "epoch": 0.02, + "learning_rate": 4.665521897477861e-05, + "loss": 4.3933, + "step": 204288 + }, + { + "epoch": 0.02, + "learning_rate": 4.664683302726809e-05, + "loss": 4.4082, + "step": 204800 + }, + { + "epoch": 0.02, + "learning_rate": 4.663844707975757e-05, + "loss": 4.4144, + "step": 205312 + }, + { + "epoch": 0.02, + "learning_rate": 4.663006113224705e-05, + "loss": 4.3996, + "step": 205824 + }, + { + "epoch": 0.02, + "learning_rate": 4.662169156354026e-05, + "loss": 4.4042, + "step": 206336 + }, + { + "epoch": 0.02, + "learning_rate": 4.661330561602974e-05, + "loss": 4.3807, + "step": 206848 + }, + { + "epoch": 0.02, + "learning_rate": 4.660491966851922e-05, + "loss": 4.3808, + "step": 207360 + }, + { + "epoch": 0.02, + "learning_rate": 4.65965337210087e-05, + "loss": 4.3933, + "step": 207872 + }, + { + "epoch": 0.02, + "learning_rate": 4.658814777349818e-05, + "loss": 4.3957, + "step": 208384 + }, + { + "epoch": 0.02, + "learning_rate": 4.657977820479139e-05, + "loss": 4.402, + "step": 208896 + }, + { + "epoch": 0.02, + "learning_rate": 4.657139225728087e-05, + "loss": 4.3859, + "step": 209408 + }, + { + "epoch": 0.02, + "learning_rate": 4.6563006309770355e-05, + "loss": 4.3847, + "step": 209920 + }, + { + "epoch": 0.02, + "learning_rate": 4.6554620362259835e-05, + "loss": 4.39, + "step": 210432 + }, + { + "epoch": 0.02, + "learning_rate": 4.6546234414749315e-05, + "loss": 4.4018, + "step": 210944 + }, + { + "epoch": 0.02, + "learning_rate": 4.6537848467238795e-05, + "loss": 4.3924, + "step": 211456 + }, + { + "epoch": 0.02, + "learning_rate": 4.6529478898532004e-05, + "loss": 4.3956, + "step": 211968 + }, + { + "epoch": 0.02, + "learning_rate": 4.6521092951021484e-05, + "loss": 4.3859, + "step": 212480 + }, + { + "epoch": 0.02, + "learning_rate": 4.6512707003510964e-05, + "loss": 4.3877, + "step": 212992 + }, + { + "epoch": 0.02, + "learning_rate": 4.6504321056000444e-05, + "loss": 4.3978, + "step": 213504 + }, + { + "epoch": 0.02, + "learning_rate": 4.6495935108489924e-05, + "loss": 4.3853, + "step": 214016 + }, + { + "epoch": 0.02, + "learning_rate": 4.648756553978313e-05, + "loss": 4.3843, + "step": 214528 + }, + { + "epoch": 0.02, + "learning_rate": 4.647917959227261e-05, + "loss": 4.3835, + "step": 215040 + }, + { + "epoch": 0.02, + "learning_rate": 4.647079364476209e-05, + "loss": 4.3748, + "step": 215552 + }, + { + "epoch": 0.02, + "learning_rate": 4.646240769725157e-05, + "loss": 4.3827, + "step": 216064 + }, + { + "epoch": 0.02, + "learning_rate": 4.645402174974105e-05, + "loss": 4.3824, + "step": 216576 + }, + { + "epoch": 0.02, + "learning_rate": 4.644565218103427e-05, + "loss": 4.3906, + "step": 217088 + }, + { + "epoch": 0.02, + "learning_rate": 4.643726623352375e-05, + "loss": 4.371, + "step": 217600 + }, + { + "epoch": 0.02, + "learning_rate": 4.642888028601323e-05, + "loss": 4.3779, + "step": 218112 + }, + { + "epoch": 0.02, + "learning_rate": 4.642051071730644e-05, + "loss": 4.3958, + "step": 218624 + }, + { + "epoch": 0.02, + "learning_rate": 4.641212476979592e-05, + "loss": 4.3758, + "step": 219136 + }, + { + "epoch": 0.02, + "learning_rate": 4.64037388222854e-05, + "loss": 4.3907, + "step": 219648 + }, + { + "epoch": 0.02, + "learning_rate": 4.639535287477488e-05, + "loss": 4.3774, + "step": 220160 + }, + { + "epoch": 0.02, + "learning_rate": 4.638696692726436e-05, + "loss": 4.3943, + "step": 220672 + }, + { + "epoch": 0.02, + "learning_rate": 4.6378597358557566e-05, + "loss": 4.3798, + "step": 221184 + }, + { + "epoch": 0.02, + "learning_rate": 4.6370211411047046e-05, + "loss": 4.3713, + "step": 221696 + }, + { + "epoch": 0.02, + "learning_rate": 4.6361825463536526e-05, + "loss": 4.3721, + "step": 222208 + }, + { + "epoch": 0.02, + "learning_rate": 4.6353439516026006e-05, + "loss": 4.3668, + "step": 222720 + }, + { + "epoch": 0.02, + "learning_rate": 4.6345053568515486e-05, + "loss": 4.3782, + "step": 223232 + }, + { + "epoch": 0.02, + "learning_rate": 4.6336667621004966e-05, + "loss": 4.3863, + "step": 223744 + }, + { + "epoch": 0.02, + "learning_rate": 4.6328281673494446e-05, + "loss": 4.3806, + "step": 224256 + }, + { + "epoch": 0.02, + "learning_rate": 4.6319895725983926e-05, + "loss": 4.3766, + "step": 224768 + }, + { + "epoch": 0.02, + "learning_rate": 4.631154253608087e-05, + "loss": 4.3699, + "step": 225280 + }, + { + "epoch": 0.02, + "learning_rate": 4.630315658857035e-05, + "loss": 4.3681, + "step": 225792 + }, + { + "epoch": 0.02, + "learning_rate": 4.629477064105983e-05, + "loss": 4.3624, + "step": 226304 + }, + { + "epoch": 0.02, + "learning_rate": 4.6286384693549304e-05, + "loss": 4.3741, + "step": 226816 + }, + { + "epoch": 0.02, + "learning_rate": 4.6277998746038784e-05, + "loss": 4.3833, + "step": 227328 + }, + { + "epoch": 0.02, + "learning_rate": 4.6269612798528264e-05, + "loss": 4.3719, + "step": 227840 + }, + { + "epoch": 0.02, + "learning_rate": 4.6261226851017744e-05, + "loss": 4.3658, + "step": 228352 + }, + { + "epoch": 0.02, + "learning_rate": 4.625284090350723e-05, + "loss": 4.3631, + "step": 228864 + }, + { + "epoch": 0.03, + "eval_loss": 4.340713977813721, + "eval_runtime": 523.7183, + "eval_samples_per_second": 728.619, + "eval_steps_per_second": 22.77, + "step": 228957 + }, + { + "epoch": 1.0, + "learning_rate": 4.624447133480044e-05, + "loss": 4.3742, + "step": 229376 + }, + { + "epoch": 1.0, + "learning_rate": 4.623608538728992e-05, + "loss": 4.3753, + "step": 229888 + }, + { + "epoch": 1.0, + "learning_rate": 4.62276994397794e-05, + "loss": 4.3591, + "step": 230400 + }, + { + "epoch": 1.0, + "learning_rate": 4.621932987107261e-05, + "loss": 4.3596, + "step": 230912 + }, + { + "epoch": 1.0, + "learning_rate": 4.621094392356209e-05, + "loss": 4.3603, + "step": 231424 + }, + { + "epoch": 1.0, + "learning_rate": 4.620255797605157e-05, + "loss": 4.3431, + "step": 231936 + }, + { + "epoch": 1.0, + "learning_rate": 4.619417202854105e-05, + "loss": 4.3551, + "step": 232448 + }, + { + "epoch": 1.0, + "learning_rate": 4.618578608103053e-05, + "loss": 4.3639, + "step": 232960 + }, + { + "epoch": 1.0, + "learning_rate": 4.617740013352001e-05, + "loss": 4.3539, + "step": 233472 + }, + { + "epoch": 1.0, + "learning_rate": 4.6169046943616954e-05, + "loss": 4.3605, + "step": 233984 + }, + { + "epoch": 1.0, + "learning_rate": 4.6160660996106434e-05, + "loss": 4.3676, + "step": 234496 + }, + { + "epoch": 1.0, + "learning_rate": 4.6152275048595913e-05, + "loss": 4.3628, + "step": 235008 + }, + { + "epoch": 1.0, + "learning_rate": 4.6143889101085393e-05, + "loss": 4.3554, + "step": 235520 + }, + { + "epoch": 1.0, + "learning_rate": 4.613550315357487e-05, + "loss": 4.3596, + "step": 236032 + }, + { + "epoch": 1.0, + "learning_rate": 4.612711720606435e-05, + "loss": 4.3533, + "step": 236544 + }, + { + "epoch": 1.0, + "learning_rate": 4.611873125855383e-05, + "loss": 4.3505, + "step": 237056 + }, + { + "epoch": 1.0, + "learning_rate": 4.611034531104331e-05, + "loss": 4.3468, + "step": 237568 + }, + { + "epoch": 1.0, + "learning_rate": 4.610195936353279e-05, + "loss": 4.3551, + "step": 238080 + }, + { + "epoch": 1.0, + "learning_rate": 4.609357341602227e-05, + "loss": 4.3561, + "step": 238592 + }, + { + "epoch": 1.0, + "learning_rate": 4.608518746851175e-05, + "loss": 4.3599, + "step": 239104 + }, + { + "epoch": 1.0, + "learning_rate": 4.607680152100123e-05, + "loss": 4.3478, + "step": 239616 + }, + { + "epoch": 1.0, + "learning_rate": 4.606843195229444e-05, + "loss": 4.3599, + "step": 240128 + }, + { + "epoch": 1.0, + "learning_rate": 4.606004600478392e-05, + "loss": 4.3492, + "step": 240640 + }, + { + "epoch": 1.0, + "learning_rate": 4.60516600572734e-05, + "loss": 4.3508, + "step": 241152 + }, + { + "epoch": 1.0, + "learning_rate": 4.604329048856662e-05, + "loss": 4.3516, + "step": 241664 + }, + { + "epoch": 1.0, + "learning_rate": 4.60349045410561e-05, + "loss": 4.3469, + "step": 242176 + }, + { + "epoch": 1.0, + "learning_rate": 4.602651859354558e-05, + "loss": 4.3526, + "step": 242688 + }, + { + "epoch": 1.0, + "learning_rate": 4.601813264603506e-05, + "loss": 4.3395, + "step": 243200 + }, + { + "epoch": 1.0, + "learning_rate": 4.600974669852454e-05, + "loss": 4.3536, + "step": 243712 + }, + { + "epoch": 1.01, + "learning_rate": 4.600136075101402e-05, + "loss": 4.3393, + "step": 244224 + }, + { + "epoch": 1.01, + "learning_rate": 4.599297480350349e-05, + "loss": 4.3403, + "step": 244736 + }, + { + "epoch": 1.01, + "learning_rate": 4.598460523479671e-05, + "loss": 4.3454, + "step": 245248 + }, + { + "epoch": 1.01, + "learning_rate": 4.597621928728619e-05, + "loss": 4.3508, + "step": 245760 + }, + { + "epoch": 1.01, + "learning_rate": 4.596783333977567e-05, + "loss": 4.3389, + "step": 246272 + }, + { + "epoch": 1.01, + "learning_rate": 4.595944739226514e-05, + "loss": 4.338, + "step": 246784 + }, + { + "epoch": 1.01, + "learning_rate": 4.595106144475462e-05, + "loss": 4.3473, + "step": 247296 + }, + { + "epoch": 1.01, + "learning_rate": 4.59426754972441e-05, + "loss": 4.3405, + "step": 247808 + }, + { + "epoch": 1.01, + "learning_rate": 4.5934305928537316e-05, + "loss": 4.3583, + "step": 248320 + }, + { + "epoch": 1.01, + "learning_rate": 4.5925919981026796e-05, + "loss": 4.3417, + "step": 248832 + }, + { + "epoch": 1.01, + "learning_rate": 4.5917534033516276e-05, + "loss": 4.3272, + "step": 249344 + }, + { + "epoch": 1.01, + "learning_rate": 4.5909148086005756e-05, + "loss": 4.3291, + "step": 249856 + }, + { + "epoch": 1.01, + "learning_rate": 4.5900778517298965e-05, + "loss": 4.3258, + "step": 250368 + }, + { + "epoch": 1.01, + "learning_rate": 4.5892392569788445e-05, + "loss": 4.3472, + "step": 250880 + }, + { + "epoch": 1.01, + "learning_rate": 4.5884006622277925e-05, + "loss": 4.3292, + "step": 251392 + }, + { + "epoch": 1.01, + "learning_rate": 4.5875620674767404e-05, + "loss": 4.3399, + "step": 251904 + }, + { + "epoch": 1.01, + "learning_rate": 4.5867234727256884e-05, + "loss": 4.3351, + "step": 252416 + }, + { + "epoch": 1.01, + "learning_rate": 4.5858848779746364e-05, + "loss": 4.3356, + "step": 252928 + }, + { + "epoch": 1.01, + "learning_rate": 4.5850462832235844e-05, + "loss": 4.3423, + "step": 253440 + }, + { + "epoch": 1.01, + "learning_rate": 4.5842076884725324e-05, + "loss": 4.3197, + "step": 253952 + }, + { + "epoch": 1.01, + "learning_rate": 4.583370731601854e-05, + "loss": 4.3215, + "step": 254464 + }, + { + "epoch": 1.01, + "learning_rate": 4.582532136850802e-05, + "loss": 4.3232, + "step": 254976 + }, + { + "epoch": 1.01, + "learning_rate": 4.58169354209975e-05, + "loss": 4.337, + "step": 255488 + }, + { + "epoch": 1.01, + "learning_rate": 4.580856585229071e-05, + "loss": 4.3169, + "step": 256000 + }, + { + "epoch": 1.01, + "learning_rate": 4.580017990478019e-05, + "loss": 4.3377, + "step": 256512 + }, + { + "epoch": 1.01, + "learning_rate": 4.579179395726967e-05, + "loss": 4.3177, + "step": 257024 + }, + { + "epoch": 1.01, + "learning_rate": 4.578340800975915e-05, + "loss": 4.3145, + "step": 257536 + }, + { + "epoch": 1.01, + "learning_rate": 4.577502206224863e-05, + "loss": 4.3224, + "step": 258048 + }, + { + "epoch": 1.01, + "learning_rate": 4.576665249354184e-05, + "loss": 4.3115, + "step": 258560 + }, + { + "epoch": 1.01, + "learning_rate": 4.575826654603132e-05, + "loss": 4.3289, + "step": 259072 + }, + { + "epoch": 1.01, + "learning_rate": 4.57498805985208e-05, + "loss": 4.3211, + "step": 259584 + }, + { + "epoch": 1.01, + "learning_rate": 4.574149465101028e-05, + "loss": 4.3067, + "step": 260096 + }, + { + "epoch": 1.01, + "learning_rate": 4.573310870349976e-05, + "loss": 4.3241, + "step": 260608 + }, + { + "epoch": 1.01, + "learning_rate": 4.572472275598924e-05, + "loss": 4.3137, + "step": 261120 + }, + { + "epoch": 1.01, + "learning_rate": 4.5716353187282454e-05, + "loss": 4.3189, + "step": 261632 + }, + { + "epoch": 1.01, + "learning_rate": 4.5707967239771934e-05, + "loss": 4.3109, + "step": 262144 + }, + { + "epoch": 1.01, + "learning_rate": 4.5699581292261414e-05, + "loss": 4.3239, + "step": 262656 + }, + { + "epoch": 1.01, + "learning_rate": 4.5691195344750894e-05, + "loss": 4.3028, + "step": 263168 + }, + { + "epoch": 1.01, + "learning_rate": 4.5682809397240374e-05, + "loss": 4.324, + "step": 263680 + }, + { + "epoch": 1.01, + "learning_rate": 4.567443982853358e-05, + "loss": 4.3108, + "step": 264192 + }, + { + "epoch": 1.01, + "learning_rate": 4.566605388102306e-05, + "loss": 4.312, + "step": 264704 + }, + { + "epoch": 1.01, + "learning_rate": 4.565768431231627e-05, + "loss": 4.3182, + "step": 265216 + }, + { + "epoch": 1.01, + "learning_rate": 4.564929836480575e-05, + "loss": 4.3149, + "step": 265728 + }, + { + "epoch": 1.01, + "learning_rate": 4.564091241729523e-05, + "loss": 4.3217, + "step": 266240 + }, + { + "epoch": 1.01, + "learning_rate": 4.563252646978471e-05, + "loss": 4.3314, + "step": 266752 + }, + { + "epoch": 1.01, + "learning_rate": 4.562414052227419e-05, + "loss": 4.328, + "step": 267264 + }, + { + "epoch": 1.01, + "learning_rate": 4.561575457476368e-05, + "loss": 4.3127, + "step": 267776 + }, + { + "epoch": 1.01, + "learning_rate": 4.560736862725315e-05, + "loss": 4.3074, + "step": 268288 + }, + { + "epoch": 1.01, + "learning_rate": 4.559899905854637e-05, + "loss": 4.2982, + "step": 268800 + }, + { + "epoch": 1.01, + "learning_rate": 4.559061311103585e-05, + "loss": 4.3164, + "step": 269312 + }, + { + "epoch": 1.01, + "learning_rate": 4.558222716352533e-05, + "loss": 4.3236, + "step": 269824 + }, + { + "epoch": 1.01, + "learning_rate": 4.55738412160148e-05, + "loss": 4.3192, + "step": 270336 + }, + { + "epoch": 1.01, + "learning_rate": 4.556545526850428e-05, + "loss": 4.3055, + "step": 270848 + }, + { + "epoch": 1.01, + "learning_rate": 4.555706932099376e-05, + "loss": 4.2964, + "step": 271360 + }, + { + "epoch": 1.01, + "learning_rate": 4.554868337348324e-05, + "loss": 4.316, + "step": 271872 + }, + { + "epoch": 1.01, + "learning_rate": 4.554029742597272e-05, + "loss": 4.2925, + "step": 272384 + }, + { + "epoch": 1.01, + "learning_rate": 4.55319114784622e-05, + "loss": 4.3055, + "step": 272896 + }, + { + "epoch": 1.01, + "learning_rate": 4.552352553095168e-05, + "loss": 4.3092, + "step": 273408 + }, + { + "epoch": 1.01, + "learning_rate": 4.5515155962244896e-05, + "loss": 4.3126, + "step": 273920 + }, + { + "epoch": 1.01, + "learning_rate": 4.5506770014734376e-05, + "loss": 4.3075, + "step": 274432 + }, + { + "epoch": 1.02, + "learning_rate": 4.5498384067223856e-05, + "loss": 4.3124, + "step": 274944 + }, + { + "epoch": 1.02, + "learning_rate": 4.5489998119713336e-05, + "loss": 4.2941, + "step": 275456 + }, + { + "epoch": 1.02, + "learning_rate": 4.5481612172202816e-05, + "loss": 4.306, + "step": 275968 + }, + { + "epoch": 1.02, + "learning_rate": 4.5473226224692296e-05, + "loss": 4.303, + "step": 276480 + }, + { + "epoch": 1.02, + "learning_rate": 4.5464856655985505e-05, + "loss": 4.2953, + "step": 276992 + }, + { + "epoch": 1.02, + "learning_rate": 4.5456470708474985e-05, + "loss": 4.293, + "step": 277504 + }, + { + "epoch": 1.02, + "learning_rate": 4.5448084760964465e-05, + "loss": 4.3142, + "step": 278016 + }, + { + "epoch": 1.02, + "learning_rate": 4.5439698813453945e-05, + "loss": 4.3074, + "step": 278528 + }, + { + "epoch": 1.02, + "learning_rate": 4.5431312865943425e-05, + "loss": 4.305, + "step": 279040 + }, + { + "epoch": 1.02, + "learning_rate": 4.5422926918432905e-05, + "loss": 4.2991, + "step": 279552 + }, + { + "epoch": 1.02, + "learning_rate": 4.5414557349726114e-05, + "loss": 4.3083, + "step": 280064 + }, + { + "epoch": 1.02, + "learning_rate": 4.5406171402215594e-05, + "loss": 4.292, + "step": 280576 + }, + { + "epoch": 1.02, + "learning_rate": 4.539778545470508e-05, + "loss": 4.3053, + "step": 281088 + }, + { + "epoch": 1.02, + "learning_rate": 4.538939950719456e-05, + "loss": 4.3114, + "step": 281600 + }, + { + "epoch": 1.02, + "learning_rate": 4.538102993848777e-05, + "loss": 4.2922, + "step": 282112 + }, + { + "epoch": 1.02, + "learning_rate": 4.537264399097725e-05, + "loss": 4.3087, + "step": 282624 + }, + { + "epoch": 1.02, + "learning_rate": 4.536425804346673e-05, + "loss": 4.2751, + "step": 283136 + }, + { + "epoch": 1.02, + "learning_rate": 4.535587209595621e-05, + "loss": 4.2808, + "step": 283648 + }, + { + "epoch": 1.02, + "learning_rate": 4.534748614844569e-05, + "loss": 4.293, + "step": 284160 + }, + { + "epoch": 1.02, + "learning_rate": 4.533910020093516e-05, + "loss": 4.2938, + "step": 284672 + }, + { + "epoch": 1.02, + "learning_rate": 4.533071425342464e-05, + "loss": 4.3025, + "step": 285184 + }, + { + "epoch": 1.02, + "learning_rate": 4.532232830591412e-05, + "loss": 4.2848, + "step": 285696 + }, + { + "epoch": 1.02, + "learning_rate": 4.531395873720733e-05, + "loss": 4.2846, + "step": 286208 + }, + { + "epoch": 1.02, + "learning_rate": 4.530557278969682e-05, + "loss": 4.2898, + "step": 286720 + }, + { + "epoch": 1.02, + "learning_rate": 4.52971868421863e-05, + "loss": 4.3017, + "step": 287232 + }, + { + "epoch": 1.02, + "learning_rate": 4.528880089467578e-05, + "loss": 4.2964, + "step": 287744 + }, + { + "epoch": 1.02, + "learning_rate": 4.528041494716526e-05, + "loss": 4.2966, + "step": 288256 + }, + { + "epoch": 1.02, + "learning_rate": 4.527202899965474e-05, + "loss": 4.2852, + "step": 288768 + }, + { + "epoch": 1.02, + "learning_rate": 4.526364305214422e-05, + "loss": 4.2874, + "step": 289280 + }, + { + "epoch": 1.02, + "learning_rate": 4.52552571046337e-05, + "loss": 4.3014, + "step": 289792 + }, + { + "epoch": 1.02, + "learning_rate": 4.524688753592691e-05, + "loss": 4.2855, + "step": 290304 + }, + { + "epoch": 1.02, + "learning_rate": 4.523850158841639e-05, + "loss": 4.2827, + "step": 290816 + }, + { + "epoch": 1.02, + "learning_rate": 4.523011564090587e-05, + "loss": 4.2894, + "step": 291328 + }, + { + "epoch": 1.02, + "learning_rate": 4.5221746072199076e-05, + "loss": 4.2742, + "step": 291840 + }, + { + "epoch": 1.02, + "learning_rate": 4.5213360124688556e-05, + "loss": 4.2885, + "step": 292352 + }, + { + "epoch": 1.02, + "learning_rate": 4.5204974177178036e-05, + "loss": 4.2869, + "step": 292864 + }, + { + "epoch": 1.02, + "learning_rate": 4.5196588229667516e-05, + "loss": 4.2958, + "step": 293376 + }, + { + "epoch": 1.02, + "learning_rate": 4.518821866096073e-05, + "loss": 4.2752, + "step": 293888 + }, + { + "epoch": 1.02, + "learning_rate": 4.517983271345021e-05, + "loss": 4.2784, + "step": 294400 + }, + { + "epoch": 1.02, + "learning_rate": 4.517144676593969e-05, + "loss": 4.3017, + "step": 294912 + }, + { + "epoch": 1.02, + "learning_rate": 4.51630771972329e-05, + "loss": 4.2809, + "step": 295424 + }, + { + "epoch": 1.02, + "learning_rate": 4.515469124972238e-05, + "loss": 4.2911, + "step": 295936 + }, + { + "epoch": 1.02, + "learning_rate": 4.514630530221186e-05, + "loss": 4.2875, + "step": 296448 + }, + { + "epoch": 1.02, + "learning_rate": 4.513791935470134e-05, + "loss": 4.2976, + "step": 296960 + }, + { + "epoch": 1.02, + "learning_rate": 4.512953340719082e-05, + "loss": 4.2873, + "step": 297472 + }, + { + "epoch": 1.02, + "learning_rate": 4.512116383848403e-05, + "loss": 4.2804, + "step": 297984 + }, + { + "epoch": 1.02, + "learning_rate": 4.511277789097351e-05, + "loss": 4.2735, + "step": 298496 + }, + { + "epoch": 1.02, + "learning_rate": 4.510439194346299e-05, + "loss": 4.2728, + "step": 299008 + }, + { + "epoch": 1.02, + "learning_rate": 4.509600599595247e-05, + "loss": 4.2822, + "step": 299520 + }, + { + "epoch": 1.02, + "learning_rate": 4.5087620048441956e-05, + "loss": 4.2928, + "step": 300032 + }, + { + "epoch": 1.02, + "learning_rate": 4.5079250479735165e-05, + "loss": 4.286, + "step": 300544 + }, + { + "epoch": 1.02, + "learning_rate": 4.5070864532224645e-05, + "loss": 4.2881, + "step": 301056 + }, + { + "epoch": 1.02, + "learning_rate": 4.5062478584714125e-05, + "loss": 4.2753, + "step": 301568 + }, + { + "epoch": 1.02, + "learning_rate": 4.5054092637203605e-05, + "loss": 4.2772, + "step": 302080 + }, + { + "epoch": 1.02, + "learning_rate": 4.5045706689693085e-05, + "loss": 4.2684, + "step": 302592 + }, + { + "epoch": 1.02, + "learning_rate": 4.5037320742182565e-05, + "loss": 4.2839, + "step": 303104 + }, + { + "epoch": 1.02, + "learning_rate": 4.5028934794672045e-05, + "loss": 4.2856, + "step": 303616 + }, + { + "epoch": 1.02, + "learning_rate": 4.5020565225965254e-05, + "loss": 4.2871, + "step": 304128 + }, + { + "epoch": 1.02, + "learning_rate": 4.5012179278454734e-05, + "loss": 4.2713, + "step": 304640 + }, + { + "epoch": 1.02, + "learning_rate": 4.5003793330944214e-05, + "loss": 4.2711, + "step": 305152 + }, + { + "epoch": 1.03, + "eval_loss": 4.257774829864502, + "eval_runtime": 522.7433, + "eval_samples_per_second": 729.978, + "eval_steps_per_second": 22.812, + "step": 305276 + }, + { + "epoch": 0.0, + "learning_rate": 4.4995407383433694e-05, + "loss": 4.2871, + "step": 305664 + }, + { + "epoch": 0.0, + "learning_rate": 4.4987021435923174e-05, + "loss": 4.2815, + "step": 306176 + }, + { + "epoch": 0.0, + "learning_rate": 4.4978635488412654e-05, + "loss": 4.2698, + "step": 306688 + }, + { + "epoch": 0.0, + "learning_rate": 4.4970249540902134e-05, + "loss": 4.2694, + "step": 307200 + }, + { + "epoch": 0.0, + "learning_rate": 4.4961863593391614e-05, + "loss": 4.2706, + "step": 307712 + }, + { + "epoch": 0.0, + "learning_rate": 4.4953477645881094e-05, + "loss": 4.2574, + "step": 308224 + }, + { + "epoch": 0.0, + "learning_rate": 4.4945091698370574e-05, + "loss": 4.2649, + "step": 308736 + }, + { + "epoch": 0.0, + "learning_rate": 4.4936705750860054e-05, + "loss": 4.2687, + "step": 309248 + }, + { + "epoch": 0.0, + "learning_rate": 4.4928319803349534e-05, + "loss": 4.2686, + "step": 309760 + }, + { + "epoch": 0.0, + "learning_rate": 4.4919933855839014e-05, + "loss": 4.2742, + "step": 310272 + }, + { + "epoch": 0.0, + "learning_rate": 4.4911547908328494e-05, + "loss": 4.2765, + "step": 310784 + }, + { + "epoch": 0.0, + "learning_rate": 4.4903161960817974e-05, + "loss": 4.2713, + "step": 311296 + }, + { + "epoch": 0.0, + "learning_rate": 4.4894776013307454e-05, + "loss": 4.2698, + "step": 311808 + }, + { + "epoch": 0.0, + "learning_rate": 4.4886390065796934e-05, + "loss": 4.2732, + "step": 312320 + }, + { + "epoch": 0.0, + "learning_rate": 4.487800411828641e-05, + "loss": 4.2674, + "step": 312832 + }, + { + "epoch": 0.0, + "learning_rate": 4.486961817077589e-05, + "loss": 4.2615, + "step": 313344 + }, + { + "epoch": 0.0, + "learning_rate": 4.48612486020691e-05, + "loss": 4.2575, + "step": 313856 + }, + { + "epoch": 0.0, + "learning_rate": 4.485286265455859e-05, + "loss": 4.2716, + "step": 314368 + }, + { + "epoch": 0.0, + "learning_rate": 4.484447670704806e-05, + "loss": 4.2664, + "step": 314880 + }, + { + "epoch": 0.0, + "learning_rate": 4.483609075953754e-05, + "loss": 4.2773, + "step": 315392 + }, + { + "epoch": 0.0, + "learning_rate": 4.482772119083076e-05, + "loss": 4.2605, + "step": 315904 + }, + { + "epoch": 0.0, + "learning_rate": 4.481933524332024e-05, + "loss": 4.2716, + "step": 316416 + }, + { + "epoch": 0.0, + "learning_rate": 4.481094929580971e-05, + "loss": 4.2679, + "step": 316928 + }, + { + "epoch": 0.0, + "learning_rate": 4.480256334829919e-05, + "loss": 4.2676, + "step": 317440 + }, + { + "epoch": 0.0, + "learning_rate": 4.479417740078867e-05, + "loss": 4.265, + "step": 317952 + }, + { + "epoch": 0.0, + "learning_rate": 4.478579145327815e-05, + "loss": 4.262, + "step": 318464 + }, + { + "epoch": 0.0, + "learning_rate": 4.477740550576763e-05, + "loss": 4.2672, + "step": 318976 + }, + { + "epoch": 0.0, + "learning_rate": 4.476901955825711e-05, + "loss": 4.2534, + "step": 319488 + }, + { + "epoch": 0.0, + "learning_rate": 4.476063361074659e-05, + "loss": 4.2692, + "step": 320000 + }, + { + "epoch": 0.0, + "learning_rate": 4.475226404203981e-05, + "loss": 4.2575, + "step": 320512 + }, + { + "epoch": 0.01, + "learning_rate": 4.474387809452929e-05, + "loss": 4.2546, + "step": 321024 + }, + { + "epoch": 0.01, + "learning_rate": 4.473549214701877e-05, + "loss": 4.2621, + "step": 321536 + }, + { + "epoch": 0.01, + "learning_rate": 4.472710619950825e-05, + "loss": 4.2664, + "step": 322048 + }, + { + "epoch": 0.01, + "learning_rate": 4.471872025199773e-05, + "loss": 4.259, + "step": 322560 + }, + { + "epoch": 0.01, + "learning_rate": 4.471033430448721e-05, + "loss": 4.2543, + "step": 323072 + }, + { + "epoch": 0.01, + "learning_rate": 4.4701964735780416e-05, + "loss": 4.2618, + "step": 323584 + }, + { + "epoch": 0.01, + "learning_rate": 4.4693578788269896e-05, + "loss": 4.2629, + "step": 324096 + }, + { + "epoch": 0.01, + "learning_rate": 4.4685192840759376e-05, + "loss": 4.2747, + "step": 324608 + }, + { + "epoch": 0.01, + "learning_rate": 4.4676806893248856e-05, + "loss": 4.2588, + "step": 325120 + }, + { + "epoch": 0.01, + "learning_rate": 4.4668420945738336e-05, + "loss": 4.2462, + "step": 325632 + }, + { + "epoch": 0.01, + "learning_rate": 4.4660034998227816e-05, + "loss": 4.2463, + "step": 326144 + }, + { + "epoch": 0.01, + "learning_rate": 4.4651649050717296e-05, + "loss": 4.248, + "step": 326656 + }, + { + "epoch": 0.01, + "learning_rate": 4.464327948201051e-05, + "loss": 4.2619, + "step": 327168 + }, + { + "epoch": 0.01, + "learning_rate": 4.463489353449999e-05, + "loss": 4.249, + "step": 327680 + }, + { + "epoch": 0.01, + "learning_rate": 4.462650758698947e-05, + "loss": 4.2603, + "step": 328192 + }, + { + "epoch": 0.01, + "learning_rate": 4.461812163947895e-05, + "loss": 4.2549, + "step": 328704 + }, + { + "epoch": 0.01, + "learning_rate": 4.4609735691968425e-05, + "loss": 4.2522, + "step": 329216 + }, + { + "epoch": 0.01, + "learning_rate": 4.460136612326164e-05, + "loss": 4.262, + "step": 329728 + }, + { + "epoch": 0.01, + "learning_rate": 4.459298017575112e-05, + "loss": 4.2414, + "step": 330240 + }, + { + "epoch": 0.01, + "learning_rate": 4.4584594228240594e-05, + "loss": 4.237, + "step": 330752 + }, + { + "epoch": 0.01, + "learning_rate": 4.4576208280730074e-05, + "loss": 4.2465, + "step": 331264 + }, + { + "epoch": 0.01, + "learning_rate": 4.4567822333219554e-05, + "loss": 4.2548, + "step": 331776 + }, + { + "epoch": 0.01, + "learning_rate": 4.455945276451277e-05, + "loss": 4.2363, + "step": 332288 + }, + { + "epoch": 0.01, + "learning_rate": 4.455106681700225e-05, + "loss": 4.2605, + "step": 332800 + }, + { + "epoch": 0.01, + "learning_rate": 4.4542697248295465e-05, + "loss": 4.2377, + "step": 333312 + }, + { + "epoch": 0.01, + "learning_rate": 4.4534311300784945e-05, + "loss": 4.234, + "step": 333824 + }, + { + "epoch": 0.01, + "learning_rate": 4.4525925353274425e-05, + "loss": 4.2435, + "step": 334336 + }, + { + "epoch": 0.01, + "learning_rate": 4.45175394057639e-05, + "loss": 4.2345, + "step": 334848 + }, + { + "epoch": 0.01, + "learning_rate": 4.450915345825338e-05, + "loss": 4.2522, + "step": 335360 + }, + { + "epoch": 0.01, + "learning_rate": 4.450076751074286e-05, + "loss": 4.2448, + "step": 335872 + }, + { + "epoch": 0.01, + "learning_rate": 4.449238156323234e-05, + "loss": 4.2274, + "step": 336384 + }, + { + "epoch": 0.01, + "learning_rate": 4.448399561572182e-05, + "loss": 4.2439, + "step": 336896 + }, + { + "epoch": 0.01, + "learning_rate": 4.44756096682113e-05, + "loss": 4.2376, + "step": 337408 + }, + { + "epoch": 0.01, + "learning_rate": 4.446722372070078e-05, + "loss": 4.2476, + "step": 337920 + }, + { + "epoch": 0.01, + "learning_rate": 4.445883777319026e-05, + "loss": 4.2323, + "step": 338432 + }, + { + "epoch": 0.01, + "learning_rate": 4.445045182567974e-05, + "loss": 4.2453, + "step": 338944 + }, + { + "epoch": 0.01, + "learning_rate": 4.444208225697295e-05, + "loss": 4.2265, + "step": 339456 + }, + { + "epoch": 0.01, + "learning_rate": 4.443369630946243e-05, + "loss": 4.2457, + "step": 339968 + }, + { + "epoch": 0.01, + "learning_rate": 4.4425310361951914e-05, + "loss": 4.2364, + "step": 340480 + }, + { + "epoch": 0.01, + "learning_rate": 4.4416924414441394e-05, + "loss": 4.2357, + "step": 340992 + }, + { + "epoch": 0.01, + "learning_rate": 4.4408538466930874e-05, + "loss": 4.2431, + "step": 341504 + }, + { + "epoch": 0.01, + "learning_rate": 4.440016889822408e-05, + "loss": 4.2362, + "step": 342016 + }, + { + "epoch": 0.01, + "learning_rate": 4.439178295071356e-05, + "loss": 4.2465, + "step": 342528 + }, + { + "epoch": 0.01, + "learning_rate": 4.438339700320304e-05, + "loss": 4.2537, + "step": 343040 + }, + { + "epoch": 0.01, + "learning_rate": 4.437502743449625e-05, + "loss": 4.2536, + "step": 343552 + }, + { + "epoch": 0.01, + "learning_rate": 4.436664148698573e-05, + "loss": 4.2369, + "step": 344064 + }, + { + "epoch": 0.01, + "learning_rate": 4.435825553947521e-05, + "loss": 4.233, + "step": 344576 + }, + { + "epoch": 0.01, + "learning_rate": 4.434986959196469e-05, + "loss": 4.2247, + "step": 345088 + }, + { + "epoch": 0.01, + "learning_rate": 4.434148364445417e-05, + "loss": 4.2392, + "step": 345600 + }, + { + "epoch": 0.01, + "learning_rate": 4.433309769694365e-05, + "loss": 4.2528, + "step": 346112 + }, + { + "epoch": 0.01, + "learning_rate": 4.432472812823687e-05, + "loss": 4.2406, + "step": 346624 + }, + { + "epoch": 0.01, + "learning_rate": 4.431634218072635e-05, + "loss": 4.2342, + "step": 347136 + }, + { + "epoch": 0.01, + "learning_rate": 4.430795623321583e-05, + "loss": 4.2201, + "step": 347648 + }, + { + "epoch": 0.01, + "learning_rate": 4.429957028570531e-05, + "loss": 4.2461, + "step": 348160 + }, + { + "epoch": 0.01, + "learning_rate": 4.429118433819479e-05, + "loss": 4.2203, + "step": 348672 + }, + { + "epoch": 0.01, + "learning_rate": 4.428279839068426e-05, + "loss": 4.2298, + "step": 349184 + }, + { + "epoch": 0.01, + "learning_rate": 4.427441244317374e-05, + "loss": 4.2334, + "step": 349696 + }, + { + "epoch": 0.01, + "learning_rate": 4.426602649566322e-05, + "loss": 4.2359, + "step": 350208 + }, + { + "epoch": 0.01, + "learning_rate": 4.425765692695643e-05, + "loss": 4.2377, + "step": 350720 + }, + { + "epoch": 0.02, + "learning_rate": 4.4249287358249645e-05, + "loss": 4.2401, + "step": 351232 + }, + { + "epoch": 0.02, + "learning_rate": 4.4240901410739125e-05, + "loss": 4.2226, + "step": 351744 + }, + { + "epoch": 0.02, + "learning_rate": 4.4232531842032334e-05, + "loss": 4.2325, + "step": 352256 + }, + { + "epoch": 0.02, + "learning_rate": 4.422414589452182e-05, + "loss": 4.2313, + "step": 352768 + }, + { + "epoch": 0.02, + "learning_rate": 4.42157599470113e-05, + "loss": 4.224, + "step": 353280 + }, + { + "epoch": 0.02, + "learning_rate": 4.420737399950078e-05, + "loss": 4.2205, + "step": 353792 + }, + { + "epoch": 0.02, + "learning_rate": 4.419898805199026e-05, + "loss": 4.2427, + "step": 354304 + }, + { + "epoch": 0.02, + "learning_rate": 4.4190602104479734e-05, + "loss": 4.2346, + "step": 354816 + }, + { + "epoch": 0.02, + "learning_rate": 4.4182216156969214e-05, + "loss": 4.234, + "step": 355328 + }, + { + "epoch": 0.02, + "learning_rate": 4.4173830209458694e-05, + "loss": 4.2269, + "step": 355840 + }, + { + "epoch": 0.02, + "learning_rate": 4.4165444261948174e-05, + "loss": 4.2335, + "step": 356352 + }, + { + "epoch": 0.02, + "learning_rate": 4.4157058314437654e-05, + "loss": 4.2212, + "step": 356864 + }, + { + "epoch": 0.02, + "learning_rate": 4.4148672366927134e-05, + "loss": 4.2354, + "step": 357376 + }, + { + "epoch": 0.02, + "learning_rate": 4.4140286419416614e-05, + "loss": 4.2373, + "step": 357888 + }, + { + "epoch": 0.02, + "learning_rate": 4.413191685070982e-05, + "loss": 4.225, + "step": 358400 + }, + { + "epoch": 0.02, + "learning_rate": 4.41235309031993e-05, + "loss": 4.2401, + "step": 358912 + }, + { + "epoch": 0.02, + "learning_rate": 4.411516133449252e-05, + "loss": 4.206, + "step": 359424 + }, + { + "epoch": 0.02, + "learning_rate": 4.4106775386982e-05, + "loss": 4.212, + "step": 359936 + }, + { + "epoch": 0.02, + "learning_rate": 4.409838943947148e-05, + "loss": 4.2195, + "step": 360448 + }, + { + "epoch": 0.02, + "learning_rate": 4.409000349196096e-05, + "loss": 4.2222, + "step": 360960 + }, + { + "epoch": 0.02, + "learning_rate": 4.408161754445044e-05, + "loss": 4.2329, + "step": 361472 + }, + { + "epoch": 0.02, + "learning_rate": 4.407323159693992e-05, + "loss": 4.22, + "step": 361984 + }, + { + "epoch": 0.02, + "learning_rate": 4.40648456494294e-05, + "loss": 4.2113, + "step": 362496 + }, + { + "epoch": 0.02, + "learning_rate": 4.405647608072261e-05, + "loss": 4.2192, + "step": 363008 + }, + { + "epoch": 0.02, + "learning_rate": 4.404809013321209e-05, + "loss": 4.2311, + "step": 363520 + }, + { + "epoch": 0.02, + "learning_rate": 4.403970418570157e-05, + "loss": 4.2269, + "step": 364032 + }, + { + "epoch": 0.02, + "learning_rate": 4.403131823819105e-05, + "loss": 4.2283, + "step": 364544 + }, + { + "epoch": 0.02, + "learning_rate": 4.402293229068053e-05, + "loss": 4.2172, + "step": 365056 + }, + { + "epoch": 0.02, + "learning_rate": 4.401456272197374e-05, + "loss": 4.2187, + "step": 365568 + }, + { + "epoch": 0.02, + "learning_rate": 4.400617677446322e-05, + "loss": 4.2323, + "step": 366080 + }, + { + "epoch": 0.02, + "learning_rate": 4.39977908269527e-05, + "loss": 4.2222, + "step": 366592 + }, + { + "epoch": 0.02, + "learning_rate": 4.398940487944218e-05, + "loss": 4.2155, + "step": 367104 + }, + { + "epoch": 0.02, + "learning_rate": 4.398103531073539e-05, + "loss": 4.2189, + "step": 367616 + }, + { + "epoch": 0.02, + "learning_rate": 4.397264936322487e-05, + "loss": 4.2072, + "step": 368128 + }, + { + "epoch": 0.02, + "learning_rate": 4.396426341571435e-05, + "loss": 4.2193, + "step": 368640 + }, + { + "epoch": 0.02, + "learning_rate": 4.395587746820383e-05, + "loss": 4.2214, + "step": 369152 + }, + { + "epoch": 0.02, + "learning_rate": 4.394749152069331e-05, + "loss": 4.2251, + "step": 369664 + }, + { + "epoch": 0.02, + "learning_rate": 4.393912195198652e-05, + "loss": 4.2075, + "step": 370176 + }, + { + "epoch": 0.02, + "learning_rate": 4.3930736004476e-05, + "loss": 4.2131, + "step": 370688 + }, + { + "epoch": 0.02, + "learning_rate": 4.392235005696548e-05, + "loss": 4.2296, + "step": 371200 + }, + { + "epoch": 0.02, + "learning_rate": 4.391396410945496e-05, + "loss": 4.2179, + "step": 371712 + }, + { + "epoch": 0.02, + "learning_rate": 4.390557816194444e-05, + "loss": 4.2238, + "step": 372224 + }, + { + "epoch": 0.02, + "learning_rate": 4.389719221443392e-05, + "loss": 4.2242, + "step": 372736 + }, + { + "epoch": 0.02, + "learning_rate": 4.38888062669234e-05, + "loss": 4.2256, + "step": 373248 + }, + { + "epoch": 0.02, + "learning_rate": 4.388043669821662e-05, + "loss": 4.2236, + "step": 373760 + }, + { + "epoch": 0.02, + "learning_rate": 4.387205075070609e-05, + "loss": 4.2104, + "step": 374272 + }, + { + "epoch": 0.02, + "learning_rate": 4.386366480319557e-05, + "loss": 4.2072, + "step": 374784 + }, + { + "epoch": 0.02, + "learning_rate": 4.385527885568505e-05, + "loss": 4.2116, + "step": 375296 + }, + { + "epoch": 0.02, + "learning_rate": 4.3846909286978266e-05, + "loss": 4.2169, + "step": 375808 + }, + { + "epoch": 0.02, + "learning_rate": 4.383852333946774e-05, + "loss": 4.2224, + "step": 376320 + }, + { + "epoch": 0.02, + "learning_rate": 4.383013739195722e-05, + "loss": 4.223, + "step": 376832 + }, + { + "epoch": 0.02, + "learning_rate": 4.38217514444467e-05, + "loss": 4.2209, + "step": 377344 + }, + { + "epoch": 0.02, + "learning_rate": 4.381336549693618e-05, + "loss": 4.2113, + "step": 377856 + }, + { + "epoch": 0.02, + "learning_rate": 4.380497954942566e-05, + "loss": 4.2108, + "step": 378368 + }, + { + "epoch": 0.02, + "learning_rate": 4.3796593601915145e-05, + "loss": 4.2002, + "step": 378880 + }, + { + "epoch": 0.02, + "learning_rate": 4.3788224033208355e-05, + "loss": 4.2184, + "step": 379392 + }, + { + "epoch": 0.02, + "learning_rate": 4.3779838085697835e-05, + "loss": 4.2206, + "step": 379904 + }, + { + "epoch": 0.02, + "learning_rate": 4.3771452138187314e-05, + "loss": 4.2214, + "step": 380416 + }, + { + "epoch": 0.02, + "learning_rate": 4.3763066190676794e-05, + "loss": 4.2044, + "step": 380928 + }, + { + "epoch": 0.02, + "learning_rate": 4.3754680243166274e-05, + "loss": 4.2077, + "step": 381440 + }, + { + "epoch": 0.03, + "eval_loss": 4.20143985748291, + "eval_runtime": 556.6591, + "eval_samples_per_second": 685.502, + "eval_steps_per_second": 21.422, + "step": 381595 + }, + { + "epoch": 1.0, + "learning_rate": 4.3746294295655754e-05, + "loss": 4.2238, + "step": 381952 + }, + { + "epoch": 1.0, + "learning_rate": 4.3737908348145234e-05, + "loss": 4.2128, + "step": 382464 + }, + { + "epoch": 1.0, + "learning_rate": 4.3729522400634714e-05, + "loss": 4.2077, + "step": 382976 + }, + { + "epoch": 1.0, + "learning_rate": 4.3721136453124194e-05, + "loss": 4.2025, + "step": 383488 + }, + { + "epoch": 1.0, + "learning_rate": 4.37127668844174e-05, + "loss": 4.2049, + "step": 384000 + }, + { + "epoch": 1.0, + "learning_rate": 4.370439731571061e-05, + "loss": 4.1953, + "step": 384512 + }, + { + "epoch": 1.0, + "learning_rate": 4.36960113682001e-05, + "loss": 4.2037, + "step": 385024 + }, + { + "epoch": 1.0, + "learning_rate": 4.368762542068958e-05, + "loss": 4.2003, + "step": 385536 + }, + { + "epoch": 1.0, + "learning_rate": 4.367923947317906e-05, + "loss": 4.2066, + "step": 386048 + }, + { + "epoch": 1.0, + "learning_rate": 4.3670886283276e-05, + "loss": 4.2152, + "step": 386560 + }, + { + "epoch": 1.0, + "learning_rate": 4.366250033576548e-05, + "loss": 4.2079, + "step": 387072 + }, + { + "epoch": 1.0, + "learning_rate": 4.365411438825496e-05, + "loss": 4.2068, + "step": 387584 + }, + { + "epoch": 1.0, + "learning_rate": 4.364572844074444e-05, + "loss": 4.2091, + "step": 388096 + }, + { + "epoch": 1.0, + "learning_rate": 4.363734249323392e-05, + "loss": 4.2091, + "step": 388608 + }, + { + "epoch": 1.0, + "learning_rate": 4.36289565457234e-05, + "loss": 4.2019, + "step": 389120 + }, + { + "epoch": 1.0, + "learning_rate": 4.3620586977016606e-05, + "loss": 4.1978, + "step": 389632 + }, + { + "epoch": 1.0, + "learning_rate": 4.3612201029506086e-05, + "loss": 4.1959, + "step": 390144 + }, + { + "epoch": 1.0, + "learning_rate": 4.3603815081995566e-05, + "loss": 4.2058, + "step": 390656 + }, + { + "epoch": 1.0, + "learning_rate": 4.359542913448505e-05, + "loss": 4.2056, + "step": 391168 + }, + { + "epoch": 1.0, + "learning_rate": 4.358704318697453e-05, + "loss": 4.2153, + "step": 391680 + }, + { + "epoch": 1.0, + "learning_rate": 4.357865723946401e-05, + "loss": 4.1998, + "step": 392192 + }, + { + "epoch": 1.0, + "learning_rate": 4.357028767075722e-05, + "loss": 4.2093, + "step": 392704 + }, + { + "epoch": 1.0, + "learning_rate": 4.35619017232467e-05, + "loss": 4.2075, + "step": 393216 + }, + { + "epoch": 1.0, + "learning_rate": 4.355351577573618e-05, + "loss": 4.2025, + "step": 393728 + }, + { + "epoch": 1.0, + "learning_rate": 4.354512982822566e-05, + "loss": 4.2037, + "step": 394240 + }, + { + "epoch": 1.0, + "learning_rate": 4.353674388071514e-05, + "loss": 4.1995, + "step": 394752 + }, + { + "epoch": 1.0, + "learning_rate": 4.352835793320462e-05, + "loss": 4.2063, + "step": 395264 + }, + { + "epoch": 1.0, + "learning_rate": 4.35199719856941e-05, + "loss": 4.1896, + "step": 395776 + }, + { + "epoch": 1.0, + "learning_rate": 4.3511586038183575e-05, + "loss": 4.2106, + "step": 396288 + }, + { + "epoch": 1.0, + "learning_rate": 4.3503200090673055e-05, + "loss": 4.195, + "step": 396800 + }, + { + "epoch": 1.01, + "learning_rate": 4.3494814143162535e-05, + "loss": 4.1892, + "step": 397312 + }, + { + "epoch": 1.01, + "learning_rate": 4.348642819565202e-05, + "loss": 4.2042, + "step": 397824 + }, + { + "epoch": 1.01, + "learning_rate": 4.34780422481415e-05, + "loss": 4.2019, + "step": 398336 + }, + { + "epoch": 1.01, + "learning_rate": 4.346967267943471e-05, + "loss": 4.2047, + "step": 398848 + }, + { + "epoch": 1.01, + "learning_rate": 4.346128673192419e-05, + "loss": 4.1945, + "step": 399360 + }, + { + "epoch": 1.01, + "learning_rate": 4.345290078441367e-05, + "loss": 4.197, + "step": 399872 + }, + { + "epoch": 1.01, + "learning_rate": 4.344453121570688e-05, + "loss": 4.2001, + "step": 400384 + }, + { + "epoch": 1.01, + "learning_rate": 4.343614526819636e-05, + "loss": 4.215, + "step": 400896 + }, + { + "epoch": 1.01, + "learning_rate": 4.342775932068584e-05, + "loss": 4.1978, + "step": 401408 + }, + { + "epoch": 1.01, + "learning_rate": 4.341937337317532e-05, + "loss": 4.1864, + "step": 401920 + }, + { + "epoch": 1.01, + "learning_rate": 4.34109874256648e-05, + "loss": 4.1881, + "step": 402432 + }, + { + "epoch": 1.01, + "learning_rate": 4.340260147815428e-05, + "loss": 4.187, + "step": 402944 + }, + { + "epoch": 1.01, + "learning_rate": 4.339421553064376e-05, + "loss": 4.2002, + "step": 403456 + }, + { + "epoch": 1.01, + "learning_rate": 4.338582958313324e-05, + "loss": 4.192, + "step": 403968 + }, + { + "epoch": 1.01, + "learning_rate": 4.3377460014426455e-05, + "loss": 4.198, + "step": 404480 + }, + { + "epoch": 1.01, + "learning_rate": 4.3369074066915935e-05, + "loss": 4.1995, + "step": 404992 + }, + { + "epoch": 1.01, + "learning_rate": 4.3360688119405415e-05, + "loss": 4.1916, + "step": 405504 + }, + { + "epoch": 1.01, + "learning_rate": 4.3352302171894895e-05, + "loss": 4.2036, + "step": 406016 + }, + { + "epoch": 1.01, + "learning_rate": 4.3343932603188104e-05, + "loss": 4.1819, + "step": 406528 + }, + { + "epoch": 1.01, + "learning_rate": 4.3335546655677584e-05, + "loss": 4.1767, + "step": 407040 + }, + { + "epoch": 1.01, + "learning_rate": 4.3327160708167064e-05, + "loss": 4.1915, + "step": 407552 + }, + { + "epoch": 1.01, + "learning_rate": 4.3318774760656544e-05, + "loss": 4.1893, + "step": 408064 + }, + { + "epoch": 1.01, + "learning_rate": 4.331040519194975e-05, + "loss": 4.1812, + "step": 408576 + }, + { + "epoch": 1.01, + "learning_rate": 4.330201924443923e-05, + "loss": 4.2001, + "step": 409088 + }, + { + "epoch": 1.01, + "learning_rate": 4.329364967573244e-05, + "loss": 4.1789, + "step": 409600 + }, + { + "epoch": 1.01, + "learning_rate": 4.328526372822193e-05, + "loss": 4.1792, + "step": 410112 + }, + { + "epoch": 1.01, + "learning_rate": 4.327687778071141e-05, + "loss": 4.184, + "step": 410624 + }, + { + "epoch": 1.01, + "learning_rate": 4.326849183320089e-05, + "loss": 4.1787, + "step": 411136 + }, + { + "epoch": 1.01, + "learning_rate": 4.326010588569037e-05, + "loss": 4.1927, + "step": 411648 + }, + { + "epoch": 1.01, + "learning_rate": 4.325173631698358e-05, + "loss": 4.1837, + "step": 412160 + }, + { + "epoch": 1.01, + "learning_rate": 4.324335036947306e-05, + "loss": 4.1783, + "step": 412672 + }, + { + "epoch": 1.01, + "learning_rate": 4.323496442196254e-05, + "loss": 4.1814, + "step": 413184 + }, + { + "epoch": 1.01, + "learning_rate": 4.322657847445202e-05, + "loss": 4.1723, + "step": 413696 + }, + { + "epoch": 1.01, + "learning_rate": 4.32181925269415e-05, + "loss": 4.195, + "step": 414208 + }, + { + "epoch": 1.01, + "learning_rate": 4.320980657943098e-05, + "loss": 4.1761, + "step": 414720 + }, + { + "epoch": 1.01, + "learning_rate": 4.320142063192046e-05, + "loss": 4.1939, + "step": 415232 + }, + { + "epoch": 1.01, + "learning_rate": 4.3193051063213666e-05, + "loss": 4.1671, + "step": 415744 + }, + { + "epoch": 1.01, + "learning_rate": 4.3184665115703146e-05, + "loss": 4.1862, + "step": 416256 + }, + { + "epoch": 1.01, + "learning_rate": 4.3176279168192626e-05, + "loss": 4.181, + "step": 416768 + }, + { + "epoch": 1.01, + "learning_rate": 4.3167893220682106e-05, + "loss": 4.1774, + "step": 417280 + }, + { + "epoch": 1.01, + "learning_rate": 4.3159507273171586e-05, + "loss": 4.186, + "step": 417792 + }, + { + "epoch": 1.01, + "learning_rate": 4.3151121325661066e-05, + "loss": 4.1818, + "step": 418304 + }, + { + "epoch": 1.01, + "learning_rate": 4.3142735378150546e-05, + "loss": 4.1859, + "step": 418816 + }, + { + "epoch": 1.01, + "learning_rate": 4.3134349430640026e-05, + "loss": 4.1961, + "step": 419328 + }, + { + "epoch": 1.01, + "learning_rate": 4.3125979861933235e-05, + "loss": 4.1953, + "step": 419840 + }, + { + "epoch": 1.01, + "learning_rate": 4.3117593914422715e-05, + "loss": 4.1876, + "step": 420352 + }, + { + "epoch": 1.01, + "learning_rate": 4.3109207966912195e-05, + "loss": 4.1777, + "step": 420864 + }, + { + "epoch": 1.01, + "learning_rate": 4.3100822019401675e-05, + "loss": 4.1704, + "step": 421376 + }, + { + "epoch": 1.01, + "learning_rate": 4.3092452450694884e-05, + "loss": 4.1819, + "step": 421888 + }, + { + "epoch": 1.01, + "learning_rate": 4.3084066503184364e-05, + "loss": 4.1935, + "step": 422400 + }, + { + "epoch": 1.01, + "learning_rate": 4.3075680555673844e-05, + "loss": 4.1866, + "step": 422912 + }, + { + "epoch": 1.01, + "learning_rate": 4.306729460816333e-05, + "loss": 4.179, + "step": 423424 + }, + { + "epoch": 1.01, + "learning_rate": 4.305890866065281e-05, + "loss": 4.1645, + "step": 423936 + }, + { + "epoch": 1.01, + "learning_rate": 4.305052271314229e-05, + "loss": 4.1917, + "step": 424448 + }, + { + "epoch": 1.01, + "learning_rate": 4.304213676563177e-05, + "loss": 4.1614, + "step": 424960 + }, + { + "epoch": 1.01, + "learning_rate": 4.303375081812125e-05, + "loss": 4.1776, + "step": 425472 + }, + { + "epoch": 1.01, + "learning_rate": 4.302538124941446e-05, + "loss": 4.178, + "step": 425984 + }, + { + "epoch": 1.01, + "learning_rate": 4.301699530190394e-05, + "loss": 4.1811, + "step": 426496 + }, + { + "epoch": 1.01, + "learning_rate": 4.300860935439342e-05, + "loss": 4.1812, + "step": 427008 + }, + { + "epoch": 1.02, + "learning_rate": 4.30002234068829e-05, + "loss": 4.1893, + "step": 427520 + }, + { + "epoch": 1.02, + "learning_rate": 4.299183745937238e-05, + "loss": 4.1647, + "step": 428032 + }, + { + "epoch": 1.02, + "learning_rate": 4.298345151186186e-05, + "loss": 4.175, + "step": 428544 + }, + { + "epoch": 1.02, + "learning_rate": 4.297506556435134e-05, + "loss": 4.1797, + "step": 429056 + }, + { + "epoch": 1.02, + "learning_rate": 4.296669599564455e-05, + "loss": 4.1702, + "step": 429568 + }, + { + "epoch": 1.02, + "learning_rate": 4.295831004813403e-05, + "loss": 4.1658, + "step": 430080 + }, + { + "epoch": 1.02, + "learning_rate": 4.2949924100623515e-05, + "loss": 4.1879, + "step": 430592 + }, + { + "epoch": 1.02, + "learning_rate": 4.2941538153112995e-05, + "loss": 4.1793, + "step": 431104 + }, + { + "epoch": 1.02, + "learning_rate": 4.2933168584406204e-05, + "loss": 4.1872, + "step": 431616 + }, + { + "epoch": 1.02, + "learning_rate": 4.2924782636895684e-05, + "loss": 4.1748, + "step": 432128 + }, + { + "epoch": 1.02, + "learning_rate": 4.2916396689385164e-05, + "loss": 4.1778, + "step": 432640 + }, + { + "epoch": 1.02, + "learning_rate": 4.2908010741874644e-05, + "loss": 4.167, + "step": 433152 + }, + { + "epoch": 1.02, + "learning_rate": 4.2899624794364124e-05, + "loss": 4.1815, + "step": 433664 + }, + { + "epoch": 1.02, + "learning_rate": 4.28912388468536e-05, + "loss": 4.1862, + "step": 434176 + }, + { + "epoch": 1.02, + "learning_rate": 4.288286927814681e-05, + "loss": 4.1696, + "step": 434688 + }, + { + "epoch": 1.02, + "learning_rate": 4.287448333063629e-05, + "loss": 4.1858, + "step": 435200 + }, + { + "epoch": 1.02, + "learning_rate": 4.2866097383125766e-05, + "loss": 4.1515, + "step": 435712 + }, + { + "epoch": 1.02, + "learning_rate": 4.285771143561525e-05, + "loss": 4.1628, + "step": 436224 + }, + { + "epoch": 1.02, + "learning_rate": 4.284932548810473e-05, + "loss": 4.1622, + "step": 436736 + }, + { + "epoch": 1.02, + "learning_rate": 4.284093954059421e-05, + "loss": 4.171, + "step": 437248 + }, + { + "epoch": 1.02, + "learning_rate": 4.283255359308369e-05, + "loss": 4.1798, + "step": 437760 + }, + { + "epoch": 1.02, + "learning_rate": 4.282416764557317e-05, + "loss": 4.1737, + "step": 438272 + }, + { + "epoch": 1.02, + "learning_rate": 4.281579807686638e-05, + "loss": 4.1513, + "step": 438784 + }, + { + "epoch": 1.02, + "learning_rate": 4.280741212935586e-05, + "loss": 4.1716, + "step": 439296 + }, + { + "epoch": 1.02, + "learning_rate": 4.279902618184534e-05, + "loss": 4.1781, + "step": 439808 + }, + { + "epoch": 1.02, + "learning_rate": 4.279064023433482e-05, + "loss": 4.1773, + "step": 440320 + }, + { + "epoch": 1.02, + "learning_rate": 4.27822542868243e-05, + "loss": 4.173, + "step": 440832 + }, + { + "epoch": 1.02, + "learning_rate": 4.277386833931378e-05, + "loss": 4.1659, + "step": 441344 + }, + { + "epoch": 1.02, + "learning_rate": 4.276548239180326e-05, + "loss": 4.1645, + "step": 441856 + }, + { + "epoch": 1.02, + "learning_rate": 4.275711282309647e-05, + "loss": 4.182, + "step": 442368 + }, + { + "epoch": 1.02, + "learning_rate": 4.274874325438969e-05, + "loss": 4.1675, + "step": 442880 + }, + { + "epoch": 1.02, + "learning_rate": 4.274035730687917e-05, + "loss": 4.1657, + "step": 443392 + }, + { + "epoch": 1.02, + "learning_rate": 4.273197135936865e-05, + "loss": 4.1648, + "step": 443904 + }, + { + "epoch": 1.02, + "learning_rate": 4.2723585411858127e-05, + "loss": 4.1596, + "step": 444416 + }, + { + "epoch": 1.02, + "learning_rate": 4.2715199464347607e-05, + "loss": 4.1683, + "step": 444928 + }, + { + "epoch": 1.02, + "learning_rate": 4.2706813516837087e-05, + "loss": 4.1703, + "step": 445440 + }, + { + "epoch": 1.02, + "learning_rate": 4.2698427569326566e-05, + "loss": 4.174, + "step": 445952 + }, + { + "epoch": 1.02, + "learning_rate": 4.2690041621816046e-05, + "loss": 4.1607, + "step": 446464 + }, + { + "epoch": 1.02, + "learning_rate": 4.2681655674305526e-05, + "loss": 4.1619, + "step": 446976 + }, + { + "epoch": 1.02, + "learning_rate": 4.2673269726795006e-05, + "loss": 4.1781, + "step": 447488 + }, + { + "epoch": 1.02, + "learning_rate": 4.2664883779284486e-05, + "loss": 4.165, + "step": 448000 + }, + { + "epoch": 1.02, + "learning_rate": 4.265649783177396e-05, + "loss": 4.1715, + "step": 448512 + }, + { + "epoch": 1.02, + "learning_rate": 4.2648128263067175e-05, + "loss": 4.174, + "step": 449024 + }, + { + "epoch": 1.02, + "learning_rate": 4.2639742315556655e-05, + "loss": 4.1787, + "step": 449536 + }, + { + "epoch": 1.02, + "learning_rate": 4.2631356368046135e-05, + "loss": 4.1726, + "step": 450048 + }, + { + "epoch": 1.02, + "learning_rate": 4.262298679933935e-05, + "loss": 4.1631, + "step": 450560 + }, + { + "epoch": 1.02, + "learning_rate": 4.261460085182883e-05, + "loss": 4.1526, + "step": 451072 + }, + { + "epoch": 1.02, + "learning_rate": 4.260621490431831e-05, + "loss": 4.1612, + "step": 451584 + }, + { + "epoch": 1.02, + "learning_rate": 4.2597828956807784e-05, + "loss": 4.167, + "step": 452096 + }, + { + "epoch": 1.02, + "learning_rate": 4.2589459388101e-05, + "loss": 4.176, + "step": 452608 + }, + { + "epoch": 1.02, + "learning_rate": 4.258107344059048e-05, + "loss": 4.1705, + "step": 453120 + }, + { + "epoch": 1.02, + "learning_rate": 4.257268749307996e-05, + "loss": 4.166, + "step": 453632 + }, + { + "epoch": 1.02, + "learning_rate": 4.256430154556943e-05, + "loss": 4.1656, + "step": 454144 + }, + { + "epoch": 1.02, + "learning_rate": 4.255593197686265e-05, + "loss": 4.1644, + "step": 454656 + }, + { + "epoch": 1.02, + "learning_rate": 4.254754602935213e-05, + "loss": 4.1532, + "step": 455168 + }, + { + "epoch": 1.02, + "learning_rate": 4.253916008184161e-05, + "loss": 4.1621, + "step": 455680 + }, + { + "epoch": 1.02, + "learning_rate": 4.253077413433109e-05, + "loss": 4.1744, + "step": 456192 + }, + { + "epoch": 1.02, + "learning_rate": 4.252238818682057e-05, + "loss": 4.1736, + "step": 456704 + }, + { + "epoch": 1.02, + "learning_rate": 4.251400223931005e-05, + "loss": 4.1548, + "step": 457216 + }, + { + "epoch": 1.02, + "learning_rate": 4.250561629179953e-05, + "loss": 4.157, + "step": 457728 + }, + { + "epoch": 1.03, + "eval_loss": 4.159981727600098, + "eval_runtime": 552.1831, + "eval_samples_per_second": 691.059, + "eval_steps_per_second": 21.596, + "step": 457914 + }, + { + "epoch": 0.0, + "learning_rate": 4.249723034428901e-05, + "loss": 4.1816, + "step": 458240 + }, + { + "epoch": 0.0, + "learning_rate": 4.248884439677849e-05, + "loss": 4.1651, + "step": 458752 + }, + { + "epoch": 0.0, + "learning_rate": 4.248045844926797e-05, + "loss": 4.1576, + "step": 459264 + }, + { + "epoch": 0.0, + "learning_rate": 4.247207250175745e-05, + "loss": 4.1582, + "step": 459776 + }, + { + "epoch": 0.0, + "learning_rate": 4.246368655424693e-05, + "loss": 4.1521, + "step": 460288 + }, + { + "epoch": 0.0, + "learning_rate": 4.245530060673641e-05, + "loss": 4.1487, + "step": 460800 + }, + { + "epoch": 0.0, + "learning_rate": 4.244691465922589e-05, + "loss": 4.1503, + "step": 461312 + }, + { + "epoch": 0.0, + "learning_rate": 4.243852871171537e-05, + "loss": 4.1561, + "step": 461824 + }, + { + "epoch": 0.0, + "learning_rate": 4.243014276420485e-05, + "loss": 4.1586, + "step": 462336 + }, + { + "epoch": 0.0, + "learning_rate": 4.242175681669432e-05, + "loss": 4.1638, + "step": 462848 + }, + { + "epoch": 0.0, + "learning_rate": 4.241337086918381e-05, + "loss": 4.1624, + "step": 463360 + }, + { + "epoch": 0.0, + "learning_rate": 4.2405001300477024e-05, + "loss": 4.154, + "step": 463872 + }, + { + "epoch": 0.0, + "learning_rate": 4.23966153529665e-05, + "loss": 4.1687, + "step": 464384 + }, + { + "epoch": 0.0, + "learning_rate": 4.238822940545598e-05, + "loss": 4.1573, + "step": 464896 + }, + { + "epoch": 0.0, + "learning_rate": 4.237984345794546e-05, + "loss": 4.1593, + "step": 465408 + }, + { + "epoch": 0.0, + "learning_rate": 4.237145751043494e-05, + "loss": 4.1464, + "step": 465920 + }, + { + "epoch": 0.0, + "learning_rate": 4.236307156292442e-05, + "loss": 4.1449, + "step": 466432 + }, + { + "epoch": 0.0, + "learning_rate": 4.23546856154139e-05, + "loss": 4.1612, + "step": 466944 + }, + { + "epoch": 0.0, + "learning_rate": 4.234629966790338e-05, + "loss": 4.1551, + "step": 467456 + }, + { + "epoch": 0.0, + "learning_rate": 4.233791372039286e-05, + "loss": 4.1683, + "step": 467968 + }, + { + "epoch": 0.0, + "learning_rate": 4.232952777288234e-05, + "loss": 4.1527, + "step": 468480 + }, + { + "epoch": 0.0, + "learning_rate": 4.232114182537182e-05, + "loss": 4.163, + "step": 468992 + }, + { + "epoch": 0.0, + "learning_rate": 4.23127558778613e-05, + "loss": 4.1597, + "step": 469504 + }, + { + "epoch": 0.0, + "learning_rate": 4.230436993035078e-05, + "loss": 4.1576, + "step": 470016 + }, + { + "epoch": 0.0, + "learning_rate": 4.229598398284026e-05, + "loss": 4.1535, + "step": 470528 + }, + { + "epoch": 0.0, + "learning_rate": 4.228759803532974e-05, + "loss": 4.1529, + "step": 471040 + }, + { + "epoch": 0.0, + "learning_rate": 4.227921208781922e-05, + "loss": 4.1608, + "step": 471552 + }, + { + "epoch": 0.0, + "learning_rate": 4.227084251911243e-05, + "loss": 4.1442, + "step": 472064 + }, + { + "epoch": 0.0, + "learning_rate": 4.226245657160191e-05, + "loss": 4.1619, + "step": 472576 + }, + { + "epoch": 0.0, + "learning_rate": 4.2254070624091386e-05, + "loss": 4.1507, + "step": 473088 + }, + { + "epoch": 0.01, + "learning_rate": 4.2245684676580866e-05, + "loss": 4.1434, + "step": 473600 + }, + { + "epoch": 0.01, + "learning_rate": 4.223731510787408e-05, + "loss": 4.1557, + "step": 474112 + }, + { + "epoch": 0.01, + "learning_rate": 4.222892916036356e-05, + "loss": 4.157, + "step": 474624 + }, + { + "epoch": 0.01, + "learning_rate": 4.2220543212853035e-05, + "loss": 4.1613, + "step": 475136 + }, + { + "epoch": 0.01, + "learning_rate": 4.2212157265342515e-05, + "loss": 4.1468, + "step": 475648 + }, + { + "epoch": 0.01, + "learning_rate": 4.220378769663573e-05, + "loss": 4.1531, + "step": 476160 + }, + { + "epoch": 0.01, + "learning_rate": 4.219541812792894e-05, + "loss": 4.1535, + "step": 476672 + }, + { + "epoch": 0.01, + "learning_rate": 4.2187032180418426e-05, + "loss": 4.1662, + "step": 477184 + }, + { + "epoch": 0.01, + "learning_rate": 4.2178646232907906e-05, + "loss": 4.154, + "step": 477696 + }, + { + "epoch": 0.01, + "learning_rate": 4.2170260285397386e-05, + "loss": 4.1405, + "step": 478208 + }, + { + "epoch": 0.01, + "learning_rate": 4.216187433788686e-05, + "loss": 4.1411, + "step": 478720 + }, + { + "epoch": 0.01, + "learning_rate": 4.2153504769180075e-05, + "loss": 4.1395, + "step": 479232 + }, + { + "epoch": 0.01, + "learning_rate": 4.2145118821669555e-05, + "loss": 4.1574, + "step": 479744 + }, + { + "epoch": 0.01, + "learning_rate": 4.2136732874159035e-05, + "loss": 4.1476, + "step": 480256 + }, + { + "epoch": 0.01, + "learning_rate": 4.212834692664851e-05, + "loss": 4.1477, + "step": 480768 + }, + { + "epoch": 0.01, + "learning_rate": 4.211996097913799e-05, + "loss": 4.159, + "step": 481280 + }, + { + "epoch": 0.01, + "learning_rate": 4.211157503162747e-05, + "loss": 4.1471, + "step": 481792 + }, + { + "epoch": 0.01, + "learning_rate": 4.210318908411695e-05, + "loss": 4.1579, + "step": 482304 + }, + { + "epoch": 0.01, + "learning_rate": 4.209480313660643e-05, + "loss": 4.1398, + "step": 482816 + }, + { + "epoch": 0.01, + "learning_rate": 4.2086433567899644e-05, + "loss": 4.1344, + "step": 483328 + }, + { + "epoch": 0.01, + "learning_rate": 4.2078047620389124e-05, + "loss": 4.1448, + "step": 483840 + }, + { + "epoch": 0.01, + "learning_rate": 4.2069661672878604e-05, + "loss": 4.1414, + "step": 484352 + }, + { + "epoch": 0.01, + "learning_rate": 4.2061275725368084e-05, + "loss": 4.141, + "step": 484864 + }, + { + "epoch": 0.01, + "learning_rate": 4.205290615666129e-05, + "loss": 4.1488, + "step": 485376 + }, + { + "epoch": 0.01, + "learning_rate": 4.204453658795451e-05, + "loss": 4.1388, + "step": 485888 + }, + { + "epoch": 0.01, + "learning_rate": 4.203615064044398e-05, + "loss": 4.1342, + "step": 486400 + }, + { + "epoch": 0.01, + "learning_rate": 4.202776469293346e-05, + "loss": 4.1362, + "step": 486912 + }, + { + "epoch": 0.01, + "learning_rate": 4.201939512422668e-05, + "loss": 4.136, + "step": 487424 + }, + { + "epoch": 0.01, + "learning_rate": 4.201100917671615e-05, + "loss": 4.1486, + "step": 487936 + }, + { + "epoch": 0.01, + "learning_rate": 4.200262322920563e-05, + "loss": 4.1393, + "step": 488448 + }, + { + "epoch": 0.01, + "learning_rate": 4.199423728169512e-05, + "loss": 4.1345, + "step": 488960 + }, + { + "epoch": 0.01, + "learning_rate": 4.19858513341846e-05, + "loss": 4.1367, + "step": 489472 + }, + { + "epoch": 0.01, + "learning_rate": 4.197746538667408e-05, + "loss": 4.126, + "step": 489984 + }, + { + "epoch": 0.01, + "learning_rate": 4.196907943916356e-05, + "loss": 4.1537, + "step": 490496 + }, + { + "epoch": 0.01, + "learning_rate": 4.196069349165304e-05, + "loss": 4.1349, + "step": 491008 + }, + { + "epoch": 0.01, + "learning_rate": 4.195230754414252e-05, + "loss": 4.144, + "step": 491520 + }, + { + "epoch": 0.01, + "learning_rate": 4.1943921596632e-05, + "loss": 4.1306, + "step": 492032 + }, + { + "epoch": 0.01, + "learning_rate": 4.193553564912148e-05, + "loss": 4.1368, + "step": 492544 + }, + { + "epoch": 0.01, + "learning_rate": 4.192714970161096e-05, + "loss": 4.1388, + "step": 493056 + }, + { + "epoch": 0.01, + "learning_rate": 4.1918780132904167e-05, + "loss": 4.1325, + "step": 493568 + }, + { + "epoch": 0.01, + "learning_rate": 4.1910394185393647e-05, + "loss": 4.1447, + "step": 494080 + }, + { + "epoch": 0.01, + "learning_rate": 4.1902008237883126e-05, + "loss": 4.136, + "step": 494592 + }, + { + "epoch": 0.01, + "learning_rate": 4.1893622290372606e-05, + "loss": 4.1456, + "step": 495104 + }, + { + "epoch": 0.01, + "learning_rate": 4.1885252721665816e-05, + "loss": 4.1513, + "step": 495616 + }, + { + "epoch": 0.01, + "learning_rate": 4.18768667741553e-05, + "loss": 4.1486, + "step": 496128 + }, + { + "epoch": 0.01, + "learning_rate": 4.186849720544851e-05, + "loss": 4.1473, + "step": 496640 + }, + { + "epoch": 0.01, + "learning_rate": 4.186011125793799e-05, + "loss": 4.1346, + "step": 497152 + }, + { + "epoch": 0.01, + "learning_rate": 4.185172531042747e-05, + "loss": 4.1289, + "step": 497664 + }, + { + "epoch": 0.01, + "learning_rate": 4.184333936291695e-05, + "loss": 4.1379, + "step": 498176 + }, + { + "epoch": 0.01, + "learning_rate": 4.183495341540643e-05, + "loss": 4.1483, + "step": 498688 + }, + { + "epoch": 0.01, + "learning_rate": 4.182656746789591e-05, + "loss": 4.1433, + "step": 499200 + }, + { + "epoch": 0.01, + "learning_rate": 4.181818152038539e-05, + "loss": 4.1359, + "step": 499712 + }, + { + "epoch": 0.01, + "learning_rate": 4.18098119516786e-05, + "loss": 4.1211, + "step": 500224 + }, + { + "epoch": 0.01, + "learning_rate": 4.180142600416808e-05, + "loss": 4.1501, + "step": 500736 + }, + { + "epoch": 0.01, + "learning_rate": 4.179304005665756e-05, + "loss": 4.1162, + "step": 501248 + }, + { + "epoch": 0.01, + "learning_rate": 4.178467048795077e-05, + "loss": 4.1354, + "step": 501760 + }, + { + "epoch": 0.01, + "learning_rate": 4.177628454044025e-05, + "loss": 4.1383, + "step": 502272 + }, + { + "epoch": 0.01, + "learning_rate": 4.1767898592929736e-05, + "loss": 4.1344, + "step": 502784 + }, + { + "epoch": 0.01, + "learning_rate": 4.1759512645419216e-05, + "loss": 4.1408, + "step": 503296 + }, + { + "epoch": 0.02, + "learning_rate": 4.1751126697908696e-05, + "loss": 4.1471, + "step": 503808 + }, + { + "epoch": 0.02, + "learning_rate": 4.1742757129201905e-05, + "loss": 4.1207, + "step": 504320 + }, + { + "epoch": 0.02, + "learning_rate": 4.1734371181691385e-05, + "loss": 4.1357, + "step": 504832 + }, + { + "epoch": 0.02, + "learning_rate": 4.1725985234180865e-05, + "loss": 4.1376, + "step": 505344 + }, + { + "epoch": 0.02, + "learning_rate": 4.171759928667034e-05, + "loss": 4.1297, + "step": 505856 + }, + { + "epoch": 0.02, + "learning_rate": 4.1709229717963554e-05, + "loss": 4.1227, + "step": 506368 + }, + { + "epoch": 0.02, + "learning_rate": 4.1700843770453034e-05, + "loss": 4.1439, + "step": 506880 + }, + { + "epoch": 0.02, + "learning_rate": 4.1692457822942514e-05, + "loss": 4.1381, + "step": 507392 + }, + { + "epoch": 0.02, + "learning_rate": 4.168407187543199e-05, + "loss": 4.1471, + "step": 507904 + }, + { + "epoch": 0.02, + "learning_rate": 4.1675685927921474e-05, + "loss": 4.1297, + "step": 508416 + }, + { + "epoch": 0.02, + "learning_rate": 4.166731635921469e-05, + "loss": 4.1379, + "step": 508928 + }, + { + "epoch": 0.02, + "learning_rate": 4.165893041170417e-05, + "loss": 4.1234, + "step": 509440 + }, + { + "epoch": 0.02, + "learning_rate": 4.165054446419364e-05, + "loss": 4.1404, + "step": 509952 + }, + { + "epoch": 0.02, + "learning_rate": 4.164215851668312e-05, + "loss": 4.1431, + "step": 510464 + }, + { + "epoch": 0.02, + "learning_rate": 4.16337725691726e-05, + "loss": 4.136, + "step": 510976 + }, + { + "epoch": 0.02, + "learning_rate": 4.162538662166208e-05, + "loss": 4.1407, + "step": 511488 + }, + { + "epoch": 0.02, + "learning_rate": 4.161700067415156e-05, + "loss": 4.1143, + "step": 512000 + }, + { + "epoch": 0.02, + "learning_rate": 4.160863110544477e-05, + "loss": 4.1171, + "step": 512512 + }, + { + "epoch": 0.02, + "learning_rate": 4.160024515793425e-05, + "loss": 4.1183, + "step": 513024 + }, + { + "epoch": 0.02, + "learning_rate": 4.159185921042373e-05, + "loss": 4.1294, + "step": 513536 + }, + { + "epoch": 0.02, + "learning_rate": 4.158347326291321e-05, + "loss": 4.1383, + "step": 514048 + }, + { + "epoch": 0.02, + "learning_rate": 4.157510369420643e-05, + "loss": 4.1371, + "step": 514560 + }, + { + "epoch": 0.02, + "learning_rate": 4.156671774669591e-05, + "loss": 4.1093, + "step": 515072 + }, + { + "epoch": 0.02, + "learning_rate": 4.155833179918539e-05, + "loss": 4.1286, + "step": 515584 + }, + { + "epoch": 0.02, + "learning_rate": 4.154994585167487e-05, + "loss": 4.1383, + "step": 516096 + }, + { + "epoch": 0.02, + "learning_rate": 4.154155990416435e-05, + "loss": 4.1375, + "step": 516608 + }, + { + "epoch": 0.02, + "learning_rate": 4.1533190335457556e-05, + "loss": 4.1346, + "step": 517120 + }, + { + "epoch": 0.02, + "learning_rate": 4.1524804387947036e-05, + "loss": 4.1263, + "step": 517632 + }, + { + "epoch": 0.02, + "learning_rate": 4.1516418440436516e-05, + "loss": 4.1276, + "step": 518144 + }, + { + "epoch": 0.02, + "learning_rate": 4.1508032492925996e-05, + "loss": 4.1374, + "step": 518656 + }, + { + "epoch": 0.02, + "learning_rate": 4.1499662924219205e-05, + "loss": 4.129, + "step": 519168 + }, + { + "epoch": 0.02, + "learning_rate": 4.1491276976708685e-05, + "loss": 4.1251, + "step": 519680 + }, + { + "epoch": 0.02, + "learning_rate": 4.1482891029198165e-05, + "loss": 4.1244, + "step": 520192 + }, + { + "epoch": 0.02, + "learning_rate": 4.1474505081687645e-05, + "loss": 4.1187, + "step": 520704 + }, + { + "epoch": 0.02, + "learning_rate": 4.1466119134177125e-05, + "loss": 4.1312, + "step": 521216 + }, + { + "epoch": 0.02, + "learning_rate": 4.145773318666661e-05, + "loss": 4.131, + "step": 521728 + }, + { + "epoch": 0.02, + "learning_rate": 4.144934723915609e-05, + "loss": 4.1317, + "step": 522240 + }, + { + "epoch": 0.02, + "learning_rate": 4.144096129164557e-05, + "loss": 4.1239, + "step": 522752 + }, + { + "epoch": 0.02, + "learning_rate": 4.143259172293878e-05, + "loss": 4.1193, + "step": 523264 + }, + { + "epoch": 0.02, + "learning_rate": 4.142422215423199e-05, + "loss": 4.1386, + "step": 523776 + }, + { + "epoch": 0.02, + "learning_rate": 4.141583620672147e-05, + "loss": 4.121, + "step": 524288 + }, + { + "epoch": 0.02, + "learning_rate": 4.140745025921095e-05, + "loss": 4.1345, + "step": 524800 + }, + { + "epoch": 0.02, + "learning_rate": 4.139906431170043e-05, + "loss": 4.1379, + "step": 525312 + }, + { + "epoch": 0.02, + "learning_rate": 4.139067836418991e-05, + "loss": 4.1351, + "step": 525824 + }, + { + "epoch": 0.02, + "learning_rate": 4.138229241667939e-05, + "loss": 4.1361, + "step": 526336 + }, + { + "epoch": 0.02, + "learning_rate": 4.137390646916887e-05, + "loss": 4.1266, + "step": 526848 + }, + { + "epoch": 0.02, + "learning_rate": 4.136552052165835e-05, + "loss": 4.1103, + "step": 527360 + }, + { + "epoch": 0.02, + "learning_rate": 4.1357150952951565e-05, + "loss": 4.1212, + "step": 527872 + }, + { + "epoch": 0.02, + "learning_rate": 4.1348781384244775e-05, + "loss": 4.1259, + "step": 528384 + }, + { + "epoch": 0.02, + "learning_rate": 4.1340395436734254e-05, + "loss": 4.1404, + "step": 528896 + }, + { + "epoch": 0.02, + "learning_rate": 4.1332009489223734e-05, + "loss": 4.1289, + "step": 529408 + }, + { + "epoch": 0.02, + "learning_rate": 4.1323623541713214e-05, + "loss": 4.1322, + "step": 529920 + }, + { + "epoch": 0.02, + "learning_rate": 4.1315237594202694e-05, + "loss": 4.1245, + "step": 530432 + }, + { + "epoch": 0.02, + "learning_rate": 4.1306851646692174e-05, + "loss": 4.1263, + "step": 530944 + }, + { + "epoch": 0.02, + "learning_rate": 4.129846569918165e-05, + "loss": 4.1138, + "step": 531456 + }, + { + "epoch": 0.02, + "learning_rate": 4.129009613047486e-05, + "loss": 4.1239, + "step": 531968 + }, + { + "epoch": 0.02, + "learning_rate": 4.128171018296434e-05, + "loss": 4.136, + "step": 532480 + }, + { + "epoch": 0.02, + "learning_rate": 4.127332423545382e-05, + "loss": 4.1377, + "step": 532992 + }, + { + "epoch": 0.02, + "learning_rate": 4.12649382879433e-05, + "loss": 4.1134, + "step": 533504 + }, + { + "epoch": 0.02, + "learning_rate": 4.125656871923652e-05, + "loss": 4.1224, + "step": 534016 + }, + { + "epoch": 0.03, + "eval_loss": 4.129609107971191, + "eval_runtime": 546.8154, + "eval_samples_per_second": 697.842, + "eval_steps_per_second": 21.808, + "step": 534233 + }, + { + "epoch": 0.0, + "learning_rate": 4.1248182771726e-05, + "loss": 4.1447, + "step": 534528 + }, + { + "epoch": 0.0, + "learning_rate": 4.123979682421548e-05, + "loss": 4.1253, + "step": 535040 + }, + { + "epoch": 0.0, + "learning_rate": 4.123141087670495e-05, + "loss": 4.1198, + "step": 535552 + }, + { + "epoch": 0.0, + "learning_rate": 4.122302492919443e-05, + "loss": 4.1222, + "step": 536064 + }, + { + "epoch": 0.0, + "learning_rate": 4.121463898168391e-05, + "loss": 4.1102, + "step": 536576 + }, + { + "epoch": 0.0, + "learning_rate": 4.120625303417339e-05, + "loss": 4.1154, + "step": 537088 + }, + { + "epoch": 0.0, + "learning_rate": 4.119786708666287e-05, + "loss": 4.1073, + "step": 537600 + }, + { + "epoch": 0.0, + "learning_rate": 4.118948113915235e-05, + "loss": 4.1209, + "step": 538112 + }, + { + "epoch": 0.0, + "learning_rate": 4.118109519164183e-05, + "loss": 4.117, + "step": 538624 + }, + { + "epoch": 0.0, + "learning_rate": 4.117270924413131e-05, + "loss": 4.1261, + "step": 539136 + }, + { + "epoch": 0.0, + "learning_rate": 4.116432329662079e-05, + "loss": 4.1228, + "step": 539648 + }, + { + "epoch": 0.0, + "learning_rate": 4.115593734911027e-05, + "loss": 4.1196, + "step": 540160 + }, + { + "epoch": 0.0, + "learning_rate": 4.114755140159975e-05, + "loss": 4.1311, + "step": 540672 + }, + { + "epoch": 0.0, + "learning_rate": 4.113918183289297e-05, + "loss": 4.1171, + "step": 541184 + }, + { + "epoch": 0.0, + "learning_rate": 4.113079588538245e-05, + "loss": 4.1267, + "step": 541696 + }, + { + "epoch": 0.0, + "learning_rate": 4.112240993787193e-05, + "loss": 4.1051, + "step": 542208 + }, + { + "epoch": 0.0, + "learning_rate": 4.111402399036141e-05, + "loss": 4.1108, + "step": 542720 + }, + { + "epoch": 0.0, + "learning_rate": 4.110563804285089e-05, + "loss": 4.1225, + "step": 543232 + }, + { + "epoch": 0.0, + "learning_rate": 4.109725209534036e-05, + "loss": 4.1166, + "step": 543744 + }, + { + "epoch": 0.0, + "learning_rate": 4.108886614782984e-05, + "loss": 4.1321, + "step": 544256 + }, + { + "epoch": 0.0, + "learning_rate": 4.108048020031932e-05, + "loss": 4.1174, + "step": 544768 + }, + { + "epoch": 0.0, + "learning_rate": 4.10720942528088e-05, + "loss": 4.1228, + "step": 545280 + }, + { + "epoch": 0.0, + "learning_rate": 4.106372468410201e-05, + "loss": 4.123, + "step": 545792 + }, + { + "epoch": 0.0, + "learning_rate": 4.1055355115395225e-05, + "loss": 4.1162, + "step": 546304 + }, + { + "epoch": 0.0, + "learning_rate": 4.1046969167884705e-05, + "loss": 4.1208, + "step": 546816 + }, + { + "epoch": 0.0, + "learning_rate": 4.1038583220374185e-05, + "loss": 4.1121, + "step": 547328 + }, + { + "epoch": 0.0, + "learning_rate": 4.1030197272863665e-05, + "loss": 4.1227, + "step": 547840 + }, + { + "epoch": 0.0, + "learning_rate": 4.1021811325353145e-05, + "loss": 4.1136, + "step": 548352 + }, + { + "epoch": 0.0, + "learning_rate": 4.1013425377842625e-05, + "loss": 4.1232, + "step": 548864 + }, + { + "epoch": 0.0, + "learning_rate": 4.1005039430332105e-05, + "loss": 4.1103, + "step": 549376 + }, + { + "epoch": 0.01, + "learning_rate": 4.0996653482821585e-05, + "loss": 4.1098, + "step": 549888 + }, + { + "epoch": 0.01, + "learning_rate": 4.0988283914114794e-05, + "loss": 4.1202, + "step": 550400 + }, + { + "epoch": 0.01, + "learning_rate": 4.0979897966604274e-05, + "loss": 4.1179, + "step": 550912 + }, + { + "epoch": 0.01, + "learning_rate": 4.0971512019093754e-05, + "loss": 4.1221, + "step": 551424 + }, + { + "epoch": 0.01, + "learning_rate": 4.0963126071583234e-05, + "loss": 4.118, + "step": 551936 + }, + { + "epoch": 0.01, + "learning_rate": 4.0954740124072714e-05, + "loss": 4.1108, + "step": 552448 + }, + { + "epoch": 0.01, + "learning_rate": 4.094637055536592e-05, + "loss": 4.1207, + "step": 552960 + }, + { + "epoch": 0.01, + "learning_rate": 4.09379846078554e-05, + "loss": 4.1276, + "step": 553472 + }, + { + "epoch": 0.01, + "learning_rate": 4.092959866034489e-05, + "loss": 4.1188, + "step": 553984 + }, + { + "epoch": 0.01, + "learning_rate": 4.092121271283437e-05, + "loss": 4.1081, + "step": 554496 + }, + { + "epoch": 0.01, + "learning_rate": 4.091284314412758e-05, + "loss": 4.102, + "step": 555008 + }, + { + "epoch": 0.01, + "learning_rate": 4.090445719661706e-05, + "loss": 4.1054, + "step": 555520 + }, + { + "epoch": 0.01, + "learning_rate": 4.089607124910654e-05, + "loss": 4.121, + "step": 556032 + }, + { + "epoch": 0.01, + "learning_rate": 4.088768530159602e-05, + "loss": 4.1125, + "step": 556544 + }, + { + "epoch": 0.01, + "learning_rate": 4.08792993540855e-05, + "loss": 4.1084, + "step": 557056 + }, + { + "epoch": 0.01, + "learning_rate": 4.087091340657498e-05, + "loss": 4.1265, + "step": 557568 + }, + { + "epoch": 0.01, + "learning_rate": 4.086254383786819e-05, + "loss": 4.1099, + "step": 558080 + }, + { + "epoch": 0.01, + "learning_rate": 4.085415789035767e-05, + "loss": 4.1222, + "step": 558592 + }, + { + "epoch": 0.01, + "learning_rate": 4.084577194284715e-05, + "loss": 4.1051, + "step": 559104 + }, + { + "epoch": 0.01, + "learning_rate": 4.083738599533663e-05, + "loss": 4.0974, + "step": 559616 + }, + { + "epoch": 0.01, + "learning_rate": 4.082900004782611e-05, + "loss": 4.1076, + "step": 560128 + }, + { + "epoch": 0.01, + "learning_rate": 4.082061410031559e-05, + "loss": 4.1089, + "step": 560640 + }, + { + "epoch": 0.01, + "learning_rate": 4.0812228152805074e-05, + "loss": 4.1015, + "step": 561152 + }, + { + "epoch": 0.01, + "learning_rate": 4.0803858584098283e-05, + "loss": 4.1156, + "step": 561664 + }, + { + "epoch": 0.01, + "learning_rate": 4.079547263658776e-05, + "loss": 4.1059, + "step": 562176 + }, + { + "epoch": 0.01, + "learning_rate": 4.078708668907724e-05, + "loss": 4.097, + "step": 562688 + }, + { + "epoch": 0.01, + "learning_rate": 4.077870074156672e-05, + "loss": 4.1056, + "step": 563200 + }, + { + "epoch": 0.01, + "learning_rate": 4.0770314794056196e-05, + "loss": 4.0959, + "step": 563712 + }, + { + "epoch": 0.01, + "learning_rate": 4.076194522534941e-05, + "loss": 4.1106, + "step": 564224 + }, + { + "epoch": 0.01, + "learning_rate": 4.075355927783889e-05, + "loss": 4.1039, + "step": 564736 + }, + { + "epoch": 0.01, + "learning_rate": 4.074517333032837e-05, + "loss": 4.102, + "step": 565248 + }, + { + "epoch": 0.01, + "learning_rate": 4.0736787382817845e-05, + "loss": 4.1018, + "step": 565760 + }, + { + "epoch": 0.01, + "learning_rate": 4.0728401435307325e-05, + "loss": 4.0917, + "step": 566272 + }, + { + "epoch": 0.01, + "learning_rate": 4.072001548779681e-05, + "loss": 4.117, + "step": 566784 + }, + { + "epoch": 0.01, + "learning_rate": 4.071162954028629e-05, + "loss": 4.0981, + "step": 567296 + }, + { + "epoch": 0.01, + "learning_rate": 4.070324359277577e-05, + "loss": 4.1091, + "step": 567808 + }, + { + "epoch": 0.01, + "learning_rate": 4.069487402406898e-05, + "loss": 4.0972, + "step": 568320 + }, + { + "epoch": 0.01, + "learning_rate": 4.068648807655846e-05, + "loss": 4.1022, + "step": 568832 + }, + { + "epoch": 0.01, + "learning_rate": 4.067810212904794e-05, + "loss": 4.1001, + "step": 569344 + }, + { + "epoch": 0.01, + "learning_rate": 4.066971618153742e-05, + "loss": 4.1031, + "step": 569856 + }, + { + "epoch": 0.01, + "learning_rate": 4.06613302340269e-05, + "loss": 4.1069, + "step": 570368 + }, + { + "epoch": 0.01, + "learning_rate": 4.065294428651638e-05, + "loss": 4.1018, + "step": 570880 + }, + { + "epoch": 0.01, + "learning_rate": 4.064455833900586e-05, + "loss": 4.1078, + "step": 571392 + }, + { + "epoch": 0.01, + "learning_rate": 4.063617239149534e-05, + "loss": 4.1167, + "step": 571904 + }, + { + "epoch": 0.01, + "learning_rate": 4.062781920159228e-05, + "loss": 4.116, + "step": 572416 + }, + { + "epoch": 0.01, + "learning_rate": 4.0619433254081766e-05, + "loss": 4.1081, + "step": 572928 + }, + { + "epoch": 0.01, + "learning_rate": 4.0611047306571246e-05, + "loss": 4.1063, + "step": 573440 + }, + { + "epoch": 0.01, + "learning_rate": 4.0602661359060726e-05, + "loss": 4.0881, + "step": 573952 + }, + { + "epoch": 0.01, + "learning_rate": 4.0594275411550206e-05, + "loss": 4.1041, + "step": 574464 + }, + { + "epoch": 0.01, + "learning_rate": 4.0585889464039686e-05, + "loss": 4.1164, + "step": 574976 + }, + { + "epoch": 0.01, + "learning_rate": 4.0577503516529166e-05, + "loss": 4.1072, + "step": 575488 + }, + { + "epoch": 0.01, + "learning_rate": 4.0569117569018645e-05, + "loss": 4.107, + "step": 576000 + }, + { + "epoch": 0.01, + "learning_rate": 4.0560731621508125e-05, + "loss": 4.0882, + "step": 576512 + }, + { + "epoch": 0.01, + "learning_rate": 4.0552362052801335e-05, + "loss": 4.1109, + "step": 577024 + }, + { + "epoch": 0.01, + "learning_rate": 4.0543992484094544e-05, + "loss": 4.0866, + "step": 577536 + }, + { + "epoch": 0.01, + "learning_rate": 4.0535606536584024e-05, + "loss": 4.0994, + "step": 578048 + }, + { + "epoch": 0.01, + "learning_rate": 4.0527220589073504e-05, + "loss": 4.0989, + "step": 578560 + }, + { + "epoch": 0.01, + "learning_rate": 4.0518834641562984e-05, + "loss": 4.1032, + "step": 579072 + }, + { + "epoch": 0.01, + "learning_rate": 4.0510448694052463e-05, + "loss": 4.1082, + "step": 579584 + }, + { + "epoch": 0.02, + "learning_rate": 4.0502062746541943e-05, + "loss": 4.108, + "step": 580096 + }, + { + "epoch": 0.02, + "learning_rate": 4.049369317783516e-05, + "loss": 4.0902, + "step": 580608 + }, + { + "epoch": 0.02, + "learning_rate": 4.048530723032464e-05, + "loss": 4.1072, + "step": 581120 + }, + { + "epoch": 0.02, + "learning_rate": 4.047692128281412e-05, + "loss": 4.1015, + "step": 581632 + }, + { + "epoch": 0.02, + "learning_rate": 4.04685353353036e-05, + "loss": 4.0956, + "step": 582144 + }, + { + "epoch": 0.02, + "learning_rate": 4.046014938779308e-05, + "loss": 4.087, + "step": 582656 + }, + { + "epoch": 0.02, + "learning_rate": 4.045176344028256e-05, + "loss": 4.1082, + "step": 583168 + }, + { + "epoch": 0.02, + "learning_rate": 4.044337749277203e-05, + "loss": 4.105, + "step": 583680 + }, + { + "epoch": 0.02, + "learning_rate": 4.043499154526151e-05, + "loss": 4.1122, + "step": 584192 + }, + { + "epoch": 0.02, + "learning_rate": 4.042660559775099e-05, + "loss": 4.1005, + "step": 584704 + }, + { + "epoch": 0.02, + "learning_rate": 4.041823602904421e-05, + "loss": 4.1041, + "step": 585216 + }, + { + "epoch": 0.02, + "learning_rate": 4.040985008153368e-05, + "loss": 4.0881, + "step": 585728 + }, + { + "epoch": 0.02, + "learning_rate": 4.040146413402317e-05, + "loss": 4.1107, + "step": 586240 + }, + { + "epoch": 0.02, + "learning_rate": 4.039307818651265e-05, + "loss": 4.106, + "step": 586752 + }, + { + "epoch": 0.02, + "learning_rate": 4.038470861780586e-05, + "loss": 4.1036, + "step": 587264 + }, + { + "epoch": 0.02, + "learning_rate": 4.037632267029534e-05, + "loss": 4.1035, + "step": 587776 + }, + { + "epoch": 0.02, + "learning_rate": 4.036793672278482e-05, + "loss": 4.0849, + "step": 588288 + }, + { + "epoch": 0.02, + "learning_rate": 4.03595507752743e-05, + "loss": 4.0829, + "step": 588800 + }, + { + "epoch": 0.02, + "learning_rate": 4.0351181206567506e-05, + "loss": 4.0867, + "step": 589312 + }, + { + "epoch": 0.02, + "learning_rate": 4.0342795259056986e-05, + "loss": 4.0941, + "step": 589824 + }, + { + "epoch": 0.02, + "learning_rate": 4.0334409311546466e-05, + "loss": 4.1065, + "step": 590336 + }, + { + "epoch": 0.02, + "learning_rate": 4.0326023364035946e-05, + "loss": 4.1022, + "step": 590848 + }, + { + "epoch": 0.02, + "learning_rate": 4.0317653795329155e-05, + "loss": 4.0734, + "step": 591360 + }, + { + "epoch": 0.02, + "learning_rate": 4.0309267847818635e-05, + "loss": 4.099, + "step": 591872 + }, + { + "epoch": 0.02, + "learning_rate": 4.030088190030812e-05, + "loss": 4.1055, + "step": 592384 + }, + { + "epoch": 0.02, + "learning_rate": 4.029251233160133e-05, + "loss": 4.1006, + "step": 592896 + }, + { + "epoch": 0.02, + "learning_rate": 4.028412638409081e-05, + "loss": 4.1023, + "step": 593408 + }, + { + "epoch": 0.02, + "learning_rate": 4.027574043658029e-05, + "loss": 4.0955, + "step": 593920 + }, + { + "epoch": 0.02, + "learning_rate": 4.026735448906977e-05, + "loss": 4.0886, + "step": 594432 + }, + { + "epoch": 0.02, + "learning_rate": 4.025898492036298e-05, + "loss": 4.105, + "step": 594944 + }, + { + "epoch": 0.02, + "learning_rate": 4.0250615351656196e-05, + "loss": 4.0977, + "step": 595456 + }, + { + "epoch": 0.02, + "learning_rate": 4.0242229404145675e-05, + "loss": 4.092, + "step": 595968 + }, + { + "epoch": 0.02, + "learning_rate": 4.0233843456635155e-05, + "loss": 4.087, + "step": 596480 + }, + { + "epoch": 0.02, + "learning_rate": 4.022545750912463e-05, + "loss": 4.0896, + "step": 596992 + }, + { + "epoch": 0.02, + "learning_rate": 4.021707156161411e-05, + "loss": 4.0953, + "step": 597504 + }, + { + "epoch": 0.02, + "learning_rate": 4.020868561410359e-05, + "loss": 4.0954, + "step": 598016 + }, + { + "epoch": 0.02, + "learning_rate": 4.0200299666593075e-05, + "loss": 4.1026, + "step": 598528 + }, + { + "epoch": 0.02, + "learning_rate": 4.0191913719082555e-05, + "loss": 4.0892, + "step": 599040 + }, + { + "epoch": 0.02, + "learning_rate": 4.0183527771572035e-05, + "loss": 4.0864, + "step": 599552 + }, + { + "epoch": 0.02, + "learning_rate": 4.0175141824061515e-05, + "loss": 4.1063, + "step": 600064 + }, + { + "epoch": 0.02, + "learning_rate": 4.0166755876550995e-05, + "loss": 4.086, + "step": 600576 + }, + { + "epoch": 0.02, + "learning_rate": 4.0158369929040475e-05, + "loss": 4.1017, + "step": 601088 + }, + { + "epoch": 0.02, + "learning_rate": 4.0150000360333684e-05, + "loss": 4.1061, + "step": 601600 + }, + { + "epoch": 0.02, + "learning_rate": 4.0141614412823164e-05, + "loss": 4.1006, + "step": 602112 + }, + { + "epoch": 0.02, + "learning_rate": 4.0133228465312644e-05, + "loss": 4.1038, + "step": 602624 + }, + { + "epoch": 0.02, + "learning_rate": 4.0124842517802124e-05, + "loss": 4.0946, + "step": 603136 + }, + { + "epoch": 0.02, + "learning_rate": 4.0116456570291604e-05, + "loss": 4.082, + "step": 603648 + }, + { + "epoch": 0.02, + "learning_rate": 4.0108070622781084e-05, + "loss": 4.0859, + "step": 604160 + }, + { + "epoch": 0.02, + "learning_rate": 4.009971743287803e-05, + "loss": 4.0961, + "step": 604672 + }, + { + "epoch": 0.02, + "learning_rate": 4.009133148536751e-05, + "loss": 4.1015, + "step": 605184 + }, + { + "epoch": 0.02, + "learning_rate": 4.008294553785699e-05, + "loss": 4.1004, + "step": 605696 + }, + { + "epoch": 0.02, + "learning_rate": 4.00745759691502e-05, + "loss": 4.0992, + "step": 606208 + }, + { + "epoch": 0.02, + "learning_rate": 4.006619002163968e-05, + "loss": 4.0909, + "step": 606720 + }, + { + "epoch": 0.02, + "learning_rate": 4.005780407412916e-05, + "loss": 4.0937, + "step": 607232 + }, + { + "epoch": 0.02, + "learning_rate": 4.004941812661864e-05, + "loss": 4.0811, + "step": 607744 + }, + { + "epoch": 0.02, + "learning_rate": 4.004103217910812e-05, + "loss": 4.094, + "step": 608256 + }, + { + "epoch": 0.02, + "learning_rate": 4.00326462315976e-05, + "loss": 4.1018, + "step": 608768 + }, + { + "epoch": 0.02, + "learning_rate": 4.002426028408708e-05, + "loss": 4.106, + "step": 609280 + }, + { + "epoch": 0.02, + "learning_rate": 4.001587433657656e-05, + "loss": 4.0783, + "step": 609792 + }, + { + "epoch": 0.02, + "learning_rate": 4.000750476786977e-05, + "loss": 4.0928, + "step": 610304 + }, + { + "epoch": 0.03, + "eval_loss": 4.105195999145508, + "eval_runtime": 543.7497, + "eval_samples_per_second": 701.777, + "eval_steps_per_second": 21.931, + "step": 610552 + }, + { + "epoch": 1.0, + "learning_rate": 3.999911882035925e-05, + "loss": 4.1015, + "step": 610816 + }, + { + "epoch": 1.0, + "learning_rate": 3.999073287284873e-05, + "loss": 4.0976, + "step": 611328 + }, + { + "epoch": 1.0, + "learning_rate": 3.9982346925338207e-05, + "loss": 4.0886, + "step": 611840 + }, + { + "epoch": 1.0, + "learning_rate": 3.997396097782769e-05, + "loss": 4.0907, + "step": 612352 + }, + { + "epoch": 1.0, + "learning_rate": 3.99655914091209e-05, + "loss": 4.078, + "step": 612864 + }, + { + "epoch": 1.0, + "learning_rate": 3.995720546161038e-05, + "loss": 4.082, + "step": 613376 + }, + { + "epoch": 1.0, + "learning_rate": 3.994881951409986e-05, + "loss": 4.0812, + "step": 613888 + }, + { + "epoch": 1.0, + "learning_rate": 3.994043356658934e-05, + "loss": 4.0831, + "step": 614400 + }, + { + "epoch": 1.0, + "learning_rate": 3.9932047619078815e-05, + "loss": 4.0848, + "step": 614912 + }, + { + "epoch": 1.0, + "learning_rate": 3.9923661671568295e-05, + "loss": 4.0933, + "step": 615424 + }, + { + "epoch": 1.0, + "learning_rate": 3.991529210286151e-05, + "loss": 4.0914, + "step": 615936 + }, + { + "epoch": 1.0, + "learning_rate": 3.9906906155350984e-05, + "loss": 4.0896, + "step": 616448 + }, + { + "epoch": 1.0, + "learning_rate": 3.9898520207840464e-05, + "loss": 4.0996, + "step": 616960 + }, + { + "epoch": 1.0, + "learning_rate": 3.9890134260329944e-05, + "loss": 4.0811, + "step": 617472 + }, + { + "epoch": 1.0, + "learning_rate": 3.988174831281943e-05, + "loss": 4.0961, + "step": 617984 + }, + { + "epoch": 1.0, + "learning_rate": 3.987336236530891e-05, + "loss": 4.0757, + "step": 618496 + }, + { + "epoch": 1.0, + "learning_rate": 3.986497641779839e-05, + "loss": 4.0775, + "step": 619008 + }, + { + "epoch": 1.0, + "learning_rate": 3.98566068490916e-05, + "loss": 4.0904, + "step": 619520 + }, + { + "epoch": 1.0, + "learning_rate": 3.984822090158108e-05, + "loss": 4.0876, + "step": 620032 + }, + { + "epoch": 1.0, + "learning_rate": 3.983983495407056e-05, + "loss": 4.0983, + "step": 620544 + }, + { + "epoch": 1.0, + "learning_rate": 3.983144900656004e-05, + "loss": 4.0889, + "step": 621056 + }, + { + "epoch": 1.0, + "learning_rate": 3.982306305904952e-05, + "loss": 4.0904, + "step": 621568 + }, + { + "epoch": 1.0, + "learning_rate": 3.9814677111539e-05, + "loss": 4.0917, + "step": 622080 + }, + { + "epoch": 1.0, + "learning_rate": 3.980629116402848e-05, + "loss": 4.0846, + "step": 622592 + }, + { + "epoch": 1.0, + "learning_rate": 3.979790521651796e-05, + "loss": 4.0917, + "step": 623104 + }, + { + "epoch": 1.0, + "learning_rate": 3.97895520266149e-05, + "loss": 4.0809, + "step": 623616 + }, + { + "epoch": 1.0, + "learning_rate": 3.9781166079104385e-05, + "loss": 4.0918, + "step": 624128 + }, + { + "epoch": 1.0, + "learning_rate": 3.9772780131593865e-05, + "loss": 4.0813, + "step": 624640 + }, + { + "epoch": 1.0, + "learning_rate": 3.9764394184083345e-05, + "loss": 4.097, + "step": 625152 + }, + { + "epoch": 1.0, + "learning_rate": 3.9756008236572825e-05, + "loss": 4.0773, + "step": 625664 + }, + { + "epoch": 1.01, + "learning_rate": 3.9747622289062305e-05, + "loss": 4.0784, + "step": 626176 + }, + { + "epoch": 1.01, + "learning_rate": 3.9739236341551785e-05, + "loss": 4.0897, + "step": 626688 + }, + { + "epoch": 1.01, + "learning_rate": 3.9730850394041265e-05, + "loss": 4.0876, + "step": 627200 + }, + { + "epoch": 1.01, + "learning_rate": 3.97224972041382e-05, + "loss": 4.0921, + "step": 627712 + }, + { + "epoch": 1.01, + "learning_rate": 3.971411125662768e-05, + "loss": 4.0842, + "step": 628224 + }, + { + "epoch": 1.01, + "learning_rate": 3.970572530911716e-05, + "loss": 4.086, + "step": 628736 + }, + { + "epoch": 1.01, + "learning_rate": 3.969733936160664e-05, + "loss": 4.0871, + "step": 629248 + }, + { + "epoch": 1.01, + "learning_rate": 3.968895341409612e-05, + "loss": 4.0954, + "step": 629760 + }, + { + "epoch": 1.01, + "learning_rate": 3.968058384538934e-05, + "loss": 4.0893, + "step": 630272 + }, + { + "epoch": 1.01, + "learning_rate": 3.967219789787882e-05, + "loss": 4.0761, + "step": 630784 + }, + { + "epoch": 1.01, + "learning_rate": 3.96638119503683e-05, + "loss": 4.0755, + "step": 631296 + }, + { + "epoch": 1.01, + "learning_rate": 3.965542600285778e-05, + "loss": 4.0738, + "step": 631808 + }, + { + "epoch": 1.01, + "learning_rate": 3.964704005534726e-05, + "loss": 4.0889, + "step": 632320 + }, + { + "epoch": 1.01, + "learning_rate": 3.963865410783674e-05, + "loss": 4.0858, + "step": 632832 + }, + { + "epoch": 1.01, + "learning_rate": 3.963026816032622e-05, + "loss": 4.0719, + "step": 633344 + }, + { + "epoch": 1.01, + "learning_rate": 3.96218822128157e-05, + "loss": 4.0965, + "step": 633856 + }, + { + "epoch": 1.01, + "learning_rate": 3.961351264410891e-05, + "loss": 4.0771, + "step": 634368 + }, + { + "epoch": 1.01, + "learning_rate": 3.960512669659839e-05, + "loss": 4.0917, + "step": 634880 + }, + { + "epoch": 1.01, + "learning_rate": 3.959674074908787e-05, + "loss": 4.0764, + "step": 635392 + }, + { + "epoch": 1.01, + "learning_rate": 3.9588371180381076e-05, + "loss": 4.0707, + "step": 635904 + }, + { + "epoch": 1.01, + "learning_rate": 3.9579985232870556e-05, + "loss": 4.0771, + "step": 636416 + }, + { + "epoch": 1.01, + "learning_rate": 3.9571599285360036e-05, + "loss": 4.0757, + "step": 636928 + }, + { + "epoch": 1.01, + "learning_rate": 3.956321333784952e-05, + "loss": 4.0786, + "step": 637440 + }, + { + "epoch": 1.01, + "learning_rate": 3.9554827390339e-05, + "loss": 4.0789, + "step": 637952 + }, + { + "epoch": 1.01, + "learning_rate": 3.9546441442828476e-05, + "loss": 4.0783, + "step": 638464 + }, + { + "epoch": 1.01, + "learning_rate": 3.953807187412169e-05, + "loss": 4.0646, + "step": 638976 + }, + { + "epoch": 1.01, + "learning_rate": 3.952968592661117e-05, + "loss": 4.0759, + "step": 639488 + }, + { + "epoch": 1.01, + "learning_rate": 3.952129997910065e-05, + "loss": 4.0678, + "step": 640000 + }, + { + "epoch": 1.01, + "learning_rate": 3.9512914031590125e-05, + "loss": 4.0799, + "step": 640512 + }, + { + "epoch": 1.01, + "learning_rate": 3.950454446288334e-05, + "loss": 4.0712, + "step": 641024 + }, + { + "epoch": 1.01, + "learning_rate": 3.949615851537282e-05, + "loss": 4.0759, + "step": 641536 + }, + { + "epoch": 1.01, + "learning_rate": 3.9487772567862294e-05, + "loss": 4.0733, + "step": 642048 + }, + { + "epoch": 1.01, + "learning_rate": 3.9479386620351774e-05, + "loss": 4.064, + "step": 642560 + }, + { + "epoch": 1.01, + "learning_rate": 3.947100067284126e-05, + "loss": 4.084, + "step": 643072 + }, + { + "epoch": 1.01, + "learning_rate": 3.946261472533074e-05, + "loss": 4.068, + "step": 643584 + }, + { + "epoch": 1.01, + "learning_rate": 3.945422877782022e-05, + "loss": 4.0807, + "step": 644096 + }, + { + "epoch": 1.01, + "learning_rate": 3.94458428303097e-05, + "loss": 4.071, + "step": 644608 + }, + { + "epoch": 1.01, + "learning_rate": 3.943747326160291e-05, + "loss": 4.0705, + "step": 645120 + }, + { + "epoch": 1.01, + "learning_rate": 3.942908731409239e-05, + "loss": 4.0713, + "step": 645632 + }, + { + "epoch": 1.01, + "learning_rate": 3.942070136658187e-05, + "loss": 4.0739, + "step": 646144 + }, + { + "epoch": 1.01, + "learning_rate": 3.941231541907135e-05, + "loss": 4.0767, + "step": 646656 + }, + { + "epoch": 1.01, + "learning_rate": 3.940394585036456e-05, + "loss": 4.0739, + "step": 647168 + }, + { + "epoch": 1.01, + "learning_rate": 3.939557628165777e-05, + "loss": 4.0786, + "step": 647680 + }, + { + "epoch": 1.01, + "learning_rate": 3.938719033414725e-05, + "loss": 4.0934, + "step": 648192 + }, + { + "epoch": 1.01, + "learning_rate": 3.9378820765440464e-05, + "loss": 4.0826, + "step": 648704 + }, + { + "epoch": 1.01, + "learning_rate": 3.9370434817929943e-05, + "loss": 4.0776, + "step": 649216 + }, + { + "epoch": 1.01, + "learning_rate": 3.9362048870419423e-05, + "loss": 4.0771, + "step": 649728 + }, + { + "epoch": 1.01, + "learning_rate": 3.9353662922908903e-05, + "loss": 4.0617, + "step": 650240 + }, + { + "epoch": 1.01, + "learning_rate": 3.934527697539838e-05, + "loss": 4.073, + "step": 650752 + }, + { + "epoch": 1.01, + "learning_rate": 3.933689102788786e-05, + "loss": 4.0859, + "step": 651264 + }, + { + "epoch": 1.01, + "learning_rate": 3.932852145918107e-05, + "loss": 4.0837, + "step": 651776 + }, + { + "epoch": 1.01, + "learning_rate": 3.932013551167055e-05, + "loss": 4.0747, + "step": 652288 + }, + { + "epoch": 1.01, + "learning_rate": 3.931174956416003e-05, + "loss": 4.0604, + "step": 652800 + }, + { + "epoch": 1.01, + "learning_rate": 3.930336361664951e-05, + "loss": 4.0795, + "step": 653312 + }, + { + "epoch": 1.01, + "learning_rate": 3.929497766913899e-05, + "loss": 4.0577, + "step": 653824 + }, + { + "epoch": 1.01, + "learning_rate": 3.92866081004322e-05, + "loss": 4.0737, + "step": 654336 + }, + { + "epoch": 1.01, + "learning_rate": 3.927822215292168e-05, + "loss": 4.0718, + "step": 654848 + }, + { + "epoch": 1.01, + "learning_rate": 3.926983620541116e-05, + "loss": 4.0718, + "step": 655360 + }, + { + "epoch": 1.01, + "learning_rate": 3.926145025790065e-05, + "loss": 4.0815, + "step": 655872 + }, + { + "epoch": 1.02, + "learning_rate": 3.925306431039013e-05, + "loss": 4.0824, + "step": 656384 + }, + { + "epoch": 1.02, + "learning_rate": 3.924467836287961e-05, + "loss": 4.0639, + "step": 656896 + }, + { + "epoch": 1.02, + "learning_rate": 3.923629241536909e-05, + "loss": 4.071, + "step": 657408 + }, + { + "epoch": 1.02, + "learning_rate": 3.922790646785857e-05, + "loss": 4.0738, + "step": 657920 + }, + { + "epoch": 1.02, + "learning_rate": 3.921952052034805e-05, + "loss": 4.066, + "step": 658432 + }, + { + "epoch": 1.02, + "learning_rate": 3.921113457283753e-05, + "loss": 4.0637, + "step": 658944 + }, + { + "epoch": 1.02, + "learning_rate": 3.920276500413074e-05, + "loss": 4.0753, + "step": 659456 + }, + { + "epoch": 1.02, + "learning_rate": 3.919437905662022e-05, + "loss": 4.0796, + "step": 659968 + }, + { + "epoch": 1.02, + "learning_rate": 3.91859931091097e-05, + "loss": 4.0853, + "step": 660480 + }, + { + "epoch": 1.02, + "learning_rate": 3.917760716159918e-05, + "loss": 4.0702, + "step": 660992 + }, + { + "epoch": 1.02, + "learning_rate": 3.916922121408866e-05, + "loss": 4.0758, + "step": 661504 + }, + { + "epoch": 1.02, + "learning_rate": 3.9160851645381866e-05, + "loss": 4.0659, + "step": 662016 + }, + { + "epoch": 1.02, + "learning_rate": 3.9152465697871346e-05, + "loss": 4.0756, + "step": 662528 + }, + { + "epoch": 1.02, + "learning_rate": 3.914407975036083e-05, + "loss": 4.0829, + "step": 663040 + }, + { + "epoch": 1.02, + "learning_rate": 3.913569380285031e-05, + "loss": 4.0749, + "step": 663552 + }, + { + "epoch": 1.02, + "learning_rate": 3.9127307855339786e-05, + "loss": 4.0783, + "step": 664064 + }, + { + "epoch": 1.02, + "learning_rate": 3.9118921907829266e-05, + "loss": 4.0555, + "step": 664576 + }, + { + "epoch": 1.02, + "learning_rate": 3.9110535960318745e-05, + "loss": 4.0567, + "step": 665088 + }, + { + "epoch": 1.02, + "learning_rate": 3.9102166391611955e-05, + "loss": 4.0606, + "step": 665600 + }, + { + "epoch": 1.02, + "learning_rate": 3.9093780444101435e-05, + "loss": 4.0693, + "step": 666112 + }, + { + "epoch": 1.02, + "learning_rate": 3.9085394496590914e-05, + "loss": 4.0736, + "step": 666624 + }, + { + "epoch": 1.02, + "learning_rate": 3.9077008549080394e-05, + "loss": 4.0755, + "step": 667136 + }, + { + "epoch": 1.02, + "learning_rate": 3.9068622601569874e-05, + "loss": 4.0463, + "step": 667648 + }, + { + "epoch": 1.02, + "learning_rate": 3.9060236654059354e-05, + "loss": 4.0715, + "step": 668160 + }, + { + "epoch": 1.02, + "learning_rate": 3.9051850706548834e-05, + "loss": 4.0793, + "step": 668672 + }, + { + "epoch": 1.02, + "learning_rate": 3.9043464759038314e-05, + "loss": 4.0713, + "step": 669184 + }, + { + "epoch": 1.02, + "learning_rate": 3.90350788115278e-05, + "loss": 4.0739, + "step": 669696 + }, + { + "epoch": 1.02, + "learning_rate": 3.902670924282101e-05, + "loss": 4.0707, + "step": 670208 + }, + { + "epoch": 1.02, + "learning_rate": 3.901833967411422e-05, + "loss": 4.0582, + "step": 670720 + }, + { + "epoch": 1.02, + "learning_rate": 3.90099537266037e-05, + "loss": 4.076, + "step": 671232 + }, + { + "epoch": 1.02, + "learning_rate": 3.900156777909318e-05, + "loss": 4.0785, + "step": 671744 + }, + { + "epoch": 1.02, + "learning_rate": 3.899318183158266e-05, + "loss": 4.0629, + "step": 672256 + }, + { + "epoch": 1.02, + "learning_rate": 3.898479588407214e-05, + "loss": 4.0621, + "step": 672768 + }, + { + "epoch": 1.02, + "learning_rate": 3.897642631536535e-05, + "loss": 4.0608, + "step": 673280 + }, + { + "epoch": 1.02, + "learning_rate": 3.896804036785483e-05, + "loss": 4.0681, + "step": 673792 + }, + { + "epoch": 1.02, + "learning_rate": 3.895965442034431e-05, + "loss": 4.0664, + "step": 674304 + }, + { + "epoch": 1.02, + "learning_rate": 3.895126847283379e-05, + "loss": 4.0733, + "step": 674816 + }, + { + "epoch": 1.02, + "learning_rate": 3.894288252532327e-05, + "loss": 4.0671, + "step": 675328 + }, + { + "epoch": 1.02, + "learning_rate": 3.8934496577812755e-05, + "loss": 4.0572, + "step": 675840 + }, + { + "epoch": 1.02, + "learning_rate": 3.8926110630302235e-05, + "loss": 4.0802, + "step": 676352 + }, + { + "epoch": 1.02, + "learning_rate": 3.8917724682791715e-05, + "loss": 4.0597, + "step": 676864 + }, + { + "epoch": 1.02, + "learning_rate": 3.8909338735281195e-05, + "loss": 4.0744, + "step": 677376 + }, + { + "epoch": 1.02, + "learning_rate": 3.8900969166574404e-05, + "loss": 4.0808, + "step": 677888 + }, + { + "epoch": 1.02, + "learning_rate": 3.8892583219063884e-05, + "loss": 4.0729, + "step": 678400 + }, + { + "epoch": 1.02, + "learning_rate": 3.8884197271553364e-05, + "loss": 4.0777, + "step": 678912 + }, + { + "epoch": 1.02, + "learning_rate": 3.8875811324042843e-05, + "loss": 4.0679, + "step": 679424 + }, + { + "epoch": 1.02, + "learning_rate": 3.886745813413978e-05, + "loss": 4.0567, + "step": 679936 + }, + { + "epoch": 1.02, + "learning_rate": 3.885907218662926e-05, + "loss": 4.0638, + "step": 680448 + }, + { + "epoch": 1.02, + "learning_rate": 3.885068623911874e-05, + "loss": 4.0657, + "step": 680960 + }, + { + "epoch": 1.02, + "learning_rate": 3.884230029160822e-05, + "loss": 4.0789, + "step": 681472 + }, + { + "epoch": 1.02, + "learning_rate": 3.88339143440977e-05, + "loss": 4.0742, + "step": 681984 + }, + { + "epoch": 1.02, + "learning_rate": 3.882552839658719e-05, + "loss": 4.0703, + "step": 682496 + }, + { + "epoch": 1.02, + "learning_rate": 3.88171588278804e-05, + "loss": 4.0686, + "step": 683008 + }, + { + "epoch": 1.02, + "learning_rate": 3.880877288036988e-05, + "loss": 4.064, + "step": 683520 + }, + { + "epoch": 1.02, + "learning_rate": 3.880038693285936e-05, + "loss": 4.055, + "step": 684032 + }, + { + "epoch": 1.02, + "learning_rate": 3.879200098534884e-05, + "loss": 4.0674, + "step": 684544 + }, + { + "epoch": 1.02, + "learning_rate": 3.878361503783832e-05, + "loss": 4.0757, + "step": 685056 + }, + { + "epoch": 1.02, + "learning_rate": 3.877522909032779e-05, + "loss": 4.0808, + "step": 685568 + }, + { + "epoch": 1.02, + "learning_rate": 3.876684314281727e-05, + "loss": 4.0568, + "step": 686080 + }, + { + "epoch": 1.02, + "learning_rate": 3.875845719530675e-05, + "loss": 4.0627, + "step": 686592 + }, + { + "epoch": 1.03, + "eval_loss": 4.086565971374512, + "eval_runtime": 543.9251, + "eval_samples_per_second": 701.551, + "eval_steps_per_second": 21.924, + "step": 686871 + }, + { + "epoch": 0.0, + "learning_rate": 3.875007124779623e-05, + "loss": 4.0849, + "step": 687104 + }, + { + "epoch": 0.0, + "learning_rate": 3.874168530028571e-05, + "loss": 4.069, + "step": 687616 + }, + { + "epoch": 0.0, + "learning_rate": 3.873329935277519e-05, + "loss": 4.0652, + "step": 688128 + }, + { + "epoch": 0.0, + "learning_rate": 3.872491340526467e-05, + "loss": 4.0623, + "step": 688640 + }, + { + "epoch": 0.0, + "learning_rate": 3.871652745775416e-05, + "loss": 4.0575, + "step": 689152 + }, + { + "epoch": 0.0, + "learning_rate": 3.870814151024364e-05, + "loss": 4.0547, + "step": 689664 + }, + { + "epoch": 0.0, + "learning_rate": 3.869975556273312e-05, + "loss": 4.0528, + "step": 690176 + }, + { + "epoch": 0.0, + "learning_rate": 3.86913696152226e-05, + "loss": 4.0593, + "step": 690688 + }, + { + "epoch": 0.0, + "learning_rate": 3.868298366771208e-05, + "loss": 4.0562, + "step": 691200 + }, + { + "epoch": 0.0, + "learning_rate": 3.8674597720201557e-05, + "loss": 4.0706, + "step": 691712 + }, + { + "epoch": 0.0, + "learning_rate": 3.866621177269103e-05, + "loss": 4.062, + "step": 692224 + }, + { + "epoch": 0.0, + "learning_rate": 3.865782582518051e-05, + "loss": 4.0628, + "step": 692736 + }, + { + "epoch": 0.0, + "learning_rate": 3.864943987766999e-05, + "loss": 4.074, + "step": 693248 + }, + { + "epoch": 0.0, + "learning_rate": 3.864105393015947e-05, + "loss": 4.058, + "step": 693760 + }, + { + "epoch": 0.0, + "learning_rate": 3.863266798264895e-05, + "loss": 4.0657, + "step": 694272 + }, + { + "epoch": 0.0, + "learning_rate": 3.862429841394216e-05, + "loss": 4.0497, + "step": 694784 + }, + { + "epoch": 0.0, + "learning_rate": 3.861591246643164e-05, + "loss": 4.0524, + "step": 695296 + }, + { + "epoch": 0.0, + "learning_rate": 3.8607526518921125e-05, + "loss": 4.0671, + "step": 695808 + }, + { + "epoch": 0.0, + "learning_rate": 3.8599140571410605e-05, + "loss": 4.0606, + "step": 696320 + }, + { + "epoch": 0.0, + "learning_rate": 3.8590754623900085e-05, + "loss": 4.0745, + "step": 696832 + }, + { + "epoch": 0.0, + "learning_rate": 3.8582368676389565e-05, + "loss": 4.0584, + "step": 697344 + }, + { + "epoch": 0.0, + "learning_rate": 3.8573982728879045e-05, + "loss": 4.0736, + "step": 697856 + }, + { + "epoch": 0.0, + "learning_rate": 3.8565596781368525e-05, + "loss": 4.0601, + "step": 698368 + }, + { + "epoch": 0.0, + "learning_rate": 3.8557210833858005e-05, + "loss": 4.0596, + "step": 698880 + }, + { + "epoch": 0.0, + "learning_rate": 3.8548841265151214e-05, + "loss": 4.0643, + "step": 699392 + }, + { + "epoch": 0.0, + "learning_rate": 3.8540455317640694e-05, + "loss": 4.0609, + "step": 699904 + }, + { + "epoch": 0.0, + "learning_rate": 3.8532069370130174e-05, + "loss": 4.0647, + "step": 700416 + }, + { + "epoch": 0.0, + "learning_rate": 3.8523683422619654e-05, + "loss": 4.0559, + "step": 700928 + }, + { + "epoch": 0.0, + "learning_rate": 3.8515297475109134e-05, + "loss": 4.0697, + "step": 701440 + }, + { + "epoch": 0.0, + "learning_rate": 3.850692790640234e-05, + "loss": 4.0563, + "step": 701952 + }, + { + "epoch": 0.01, + "learning_rate": 3.849854195889182e-05, + "loss": 4.0465, + "step": 702464 + }, + { + "epoch": 0.01, + "learning_rate": 3.849015601138131e-05, + "loss": 4.0678, + "step": 702976 + }, + { + "epoch": 0.01, + "learning_rate": 3.848177006387079e-05, + "loss": 4.0615, + "step": 703488 + }, + { + "epoch": 0.01, + "learning_rate": 3.847338411636027e-05, + "loss": 4.0655, + "step": 704000 + }, + { + "epoch": 0.01, + "learning_rate": 3.846501454765348e-05, + "loss": 4.0592, + "step": 704512 + }, + { + "epoch": 0.01, + "learning_rate": 3.845662860014296e-05, + "loss": 4.0573, + "step": 705024 + }, + { + "epoch": 0.01, + "learning_rate": 3.844824265263244e-05, + "loss": 4.063, + "step": 705536 + }, + { + "epoch": 0.01, + "learning_rate": 3.843985670512192e-05, + "loss": 4.0661, + "step": 706048 + }, + { + "epoch": 0.01, + "learning_rate": 3.843147075761139e-05, + "loss": 4.0692, + "step": 706560 + }, + { + "epoch": 0.01, + "learning_rate": 3.842308481010087e-05, + "loss": 4.0459, + "step": 707072 + }, + { + "epoch": 0.01, + "learning_rate": 3.841469886259035e-05, + "loss": 4.0527, + "step": 707584 + }, + { + "epoch": 0.01, + "learning_rate": 3.840631291507983e-05, + "loss": 4.0505, + "step": 708096 + }, + { + "epoch": 0.01, + "learning_rate": 3.839794334637305e-05, + "loss": 4.0576, + "step": 708608 + }, + { + "epoch": 0.01, + "learning_rate": 3.838955739886253e-05, + "loss": 4.0635, + "step": 709120 + }, + { + "epoch": 0.01, + "learning_rate": 3.838117145135201e-05, + "loss": 4.0488, + "step": 709632 + }, + { + "epoch": 0.01, + "learning_rate": 3.837278550384149e-05, + "loss": 4.0691, + "step": 710144 + }, + { + "epoch": 0.01, + "learning_rate": 3.836439955633097e-05, + "loss": 4.0501, + "step": 710656 + }, + { + "epoch": 0.01, + "learning_rate": 3.8356029987624177e-05, + "loss": 4.0689, + "step": 711168 + }, + { + "epoch": 0.01, + "learning_rate": 3.8347644040113656e-05, + "loss": 4.0531, + "step": 711680 + }, + { + "epoch": 0.01, + "learning_rate": 3.8339258092603136e-05, + "loss": 4.0465, + "step": 712192 + }, + { + "epoch": 0.01, + "learning_rate": 3.8330888523896346e-05, + "loss": 4.0551, + "step": 712704 + }, + { + "epoch": 0.01, + "learning_rate": 3.8322502576385826e-05, + "loss": 4.0449, + "step": 713216 + }, + { + "epoch": 0.01, + "learning_rate": 3.8314116628875305e-05, + "loss": 4.0588, + "step": 713728 + }, + { + "epoch": 0.01, + "learning_rate": 3.8305730681364785e-05, + "loss": 4.0502, + "step": 714240 + }, + { + "epoch": 0.01, + "learning_rate": 3.8297344733854265e-05, + "loss": 4.0555, + "step": 714752 + }, + { + "epoch": 0.01, + "learning_rate": 3.8288958786343745e-05, + "loss": 4.0357, + "step": 715264 + }, + { + "epoch": 0.01, + "learning_rate": 3.8280572838833225e-05, + "loss": 4.0531, + "step": 715776 + }, + { + "epoch": 0.01, + "learning_rate": 3.827218689132271e-05, + "loss": 4.0399, + "step": 716288 + }, + { + "epoch": 0.01, + "learning_rate": 3.826380094381219e-05, + "loss": 4.0572, + "step": 716800 + }, + { + "epoch": 0.01, + "learning_rate": 3.82554313751054e-05, + "loss": 4.0474, + "step": 717312 + }, + { + "epoch": 0.01, + "learning_rate": 3.824704542759488e-05, + "loss": 4.0512, + "step": 717824 + }, + { + "epoch": 0.01, + "learning_rate": 3.823865948008436e-05, + "loss": 4.0475, + "step": 718336 + }, + { + "epoch": 0.01, + "learning_rate": 3.823027353257384e-05, + "loss": 4.0385, + "step": 718848 + }, + { + "epoch": 0.01, + "learning_rate": 3.822190396386705e-05, + "loss": 4.0587, + "step": 719360 + }, + { + "epoch": 0.01, + "learning_rate": 3.821351801635653e-05, + "loss": 4.0465, + "step": 719872 + }, + { + "epoch": 0.01, + "learning_rate": 3.820513206884601e-05, + "loss": 4.0589, + "step": 720384 + }, + { + "epoch": 0.01, + "learning_rate": 3.819674612133549e-05, + "loss": 4.0415, + "step": 720896 + }, + { + "epoch": 0.01, + "learning_rate": 3.818836017382497e-05, + "loss": 4.0494, + "step": 721408 + }, + { + "epoch": 0.01, + "learning_rate": 3.817997422631445e-05, + "loss": 4.0454, + "step": 721920 + }, + { + "epoch": 0.01, + "learning_rate": 3.817158827880393e-05, + "loss": 4.0502, + "step": 722432 + }, + { + "epoch": 0.01, + "learning_rate": 3.8163218710097146e-05, + "loss": 4.0479, + "step": 722944 + }, + { + "epoch": 0.01, + "learning_rate": 3.8154832762586626e-05, + "loss": 4.0503, + "step": 723456 + }, + { + "epoch": 0.01, + "learning_rate": 3.8146446815076106e-05, + "loss": 4.0535, + "step": 723968 + }, + { + "epoch": 0.01, + "learning_rate": 3.813806086756558e-05, + "loss": 4.0691, + "step": 724480 + }, + { + "epoch": 0.01, + "learning_rate": 3.812967492005506e-05, + "loss": 4.055, + "step": 724992 + }, + { + "epoch": 0.01, + "learning_rate": 3.812128897254454e-05, + "loss": 4.0551, + "step": 725504 + }, + { + "epoch": 0.01, + "learning_rate": 3.811290302503402e-05, + "loss": 4.0527, + "step": 726016 + }, + { + "epoch": 0.01, + "learning_rate": 3.81045170775235e-05, + "loss": 4.0412, + "step": 726528 + }, + { + "epoch": 0.01, + "learning_rate": 3.809614750881671e-05, + "loss": 4.0432, + "step": 727040 + }, + { + "epoch": 0.01, + "learning_rate": 3.808776156130619e-05, + "loss": 4.0577, + "step": 727552 + }, + { + "epoch": 0.01, + "learning_rate": 3.807937561379567e-05, + "loss": 4.062, + "step": 728064 + }, + { + "epoch": 0.01, + "learning_rate": 3.807098966628515e-05, + "loss": 4.0533, + "step": 728576 + }, + { + "epoch": 0.01, + "learning_rate": 3.806262009757836e-05, + "loss": 4.0368, + "step": 729088 + }, + { + "epoch": 0.01, + "learning_rate": 3.805423415006784e-05, + "loss": 4.0497, + "step": 729600 + }, + { + "epoch": 0.01, + "learning_rate": 3.804584820255732e-05, + "loss": 4.0357, + "step": 730112 + }, + { + "epoch": 0.01, + "learning_rate": 3.803747863385053e-05, + "loss": 4.0463, + "step": 730624 + }, + { + "epoch": 0.01, + "learning_rate": 3.802909268634001e-05, + "loss": 4.048, + "step": 731136 + }, + { + "epoch": 0.01, + "learning_rate": 3.802070673882949e-05, + "loss": 4.0465, + "step": 731648 + }, + { + "epoch": 0.01, + "learning_rate": 3.801232079131897e-05, + "loss": 4.0572, + "step": 732160 + }, + { + "epoch": 0.02, + "learning_rate": 3.800393484380845e-05, + "loss": 4.0535, + "step": 732672 + }, + { + "epoch": 0.02, + "learning_rate": 3.799554889629793e-05, + "loss": 4.0421, + "step": 733184 + }, + { + "epoch": 0.02, + "learning_rate": 3.798716294878741e-05, + "loss": 4.0475, + "step": 733696 + }, + { + "epoch": 0.02, + "learning_rate": 3.797877700127689e-05, + "loss": 4.0515, + "step": 734208 + }, + { + "epoch": 0.02, + "learning_rate": 3.797042381137384e-05, + "loss": 4.0399, + "step": 734720 + }, + { + "epoch": 0.02, + "learning_rate": 3.796203786386332e-05, + "loss": 4.0414, + "step": 735232 + }, + { + "epoch": 0.02, + "learning_rate": 3.79536519163528e-05, + "loss": 4.0504, + "step": 735744 + }, + { + "epoch": 0.02, + "learning_rate": 3.794526596884228e-05, + "loss": 4.0548, + "step": 736256 + }, + { + "epoch": 0.02, + "learning_rate": 3.793688002133176e-05, + "loss": 4.0605, + "step": 736768 + }, + { + "epoch": 0.02, + "learning_rate": 3.792849407382124e-05, + "loss": 4.0438, + "step": 737280 + }, + { + "epoch": 0.02, + "learning_rate": 3.792010812631072e-05, + "loss": 4.0527, + "step": 737792 + }, + { + "epoch": 0.02, + "learning_rate": 3.7911738557603926e-05, + "loss": 4.0427, + "step": 738304 + }, + { + "epoch": 0.02, + "learning_rate": 3.7903352610093406e-05, + "loss": 4.0519, + "step": 738816 + }, + { + "epoch": 0.02, + "learning_rate": 3.7894966662582886e-05, + "loss": 4.0514, + "step": 739328 + }, + { + "epoch": 0.02, + "learning_rate": 3.7886580715072366e-05, + "loss": 4.0517, + "step": 739840 + }, + { + "epoch": 0.02, + "learning_rate": 3.7878194767561846e-05, + "loss": 4.0554, + "step": 740352 + }, + { + "epoch": 0.02, + "learning_rate": 3.7869825198855055e-05, + "loss": 4.0321, + "step": 740864 + }, + { + "epoch": 0.02, + "learning_rate": 3.7861439251344535e-05, + "loss": 4.0341, + "step": 741376 + }, + { + "epoch": 0.02, + "learning_rate": 3.785305330383402e-05, + "loss": 4.0354, + "step": 741888 + }, + { + "epoch": 0.02, + "learning_rate": 3.78446673563235e-05, + "loss": 4.0395, + "step": 742400 + }, + { + "epoch": 0.02, + "learning_rate": 3.783628140881298e-05, + "loss": 4.0494, + "step": 742912 + }, + { + "epoch": 0.02, + "learning_rate": 3.782789546130246e-05, + "loss": 4.0505, + "step": 743424 + }, + { + "epoch": 0.02, + "learning_rate": 3.781950951379194e-05, + "loss": 4.0279, + "step": 743936 + }, + { + "epoch": 0.02, + "learning_rate": 3.7811123566281415e-05, + "loss": 4.047, + "step": 744448 + }, + { + "epoch": 0.02, + "learning_rate": 3.780275399757463e-05, + "loss": 4.0544, + "step": 744960 + }, + { + "epoch": 0.02, + "learning_rate": 3.779436805006411e-05, + "loss": 4.0483, + "step": 745472 + }, + { + "epoch": 0.02, + "learning_rate": 3.778598210255359e-05, + "loss": 4.0509, + "step": 745984 + }, + { + "epoch": 0.02, + "learning_rate": 3.77776125338468e-05, + "loss": 4.0434, + "step": 746496 + }, + { + "epoch": 0.02, + "learning_rate": 3.776922658633628e-05, + "loss": 4.0385, + "step": 747008 + }, + { + "epoch": 0.02, + "learning_rate": 3.776084063882576e-05, + "loss": 4.0478, + "step": 747520 + }, + { + "epoch": 0.02, + "learning_rate": 3.775245469131524e-05, + "loss": 4.054, + "step": 748032 + }, + { + "epoch": 0.02, + "learning_rate": 3.774406874380472e-05, + "loss": 4.0441, + "step": 748544 + }, + { + "epoch": 0.02, + "learning_rate": 3.77356827962942e-05, + "loss": 4.0325, + "step": 749056 + }, + { + "epoch": 0.02, + "learning_rate": 3.7727313227587415e-05, + "loss": 4.0395, + "step": 749568 + }, + { + "epoch": 0.02, + "learning_rate": 3.771892728007689e-05, + "loss": 4.0437, + "step": 750080 + }, + { + "epoch": 0.02, + "learning_rate": 3.771054133256637e-05, + "loss": 4.0472, + "step": 750592 + }, + { + "epoch": 0.02, + "learning_rate": 3.770215538505585e-05, + "loss": 4.044, + "step": 751104 + }, + { + "epoch": 0.02, + "learning_rate": 3.769380219515279e-05, + "loss": 4.0494, + "step": 751616 + }, + { + "epoch": 0.02, + "learning_rate": 3.768541624764227e-05, + "loss": 4.0337, + "step": 752128 + }, + { + "epoch": 0.02, + "learning_rate": 3.767703030013175e-05, + "loss": 4.0559, + "step": 752640 + }, + { + "epoch": 0.02, + "learning_rate": 3.766864435262123e-05, + "loss": 4.0398, + "step": 753152 + }, + { + "epoch": 0.02, + "learning_rate": 3.766025840511071e-05, + "loss": 4.0501, + "step": 753664 + }, + { + "epoch": 0.02, + "learning_rate": 3.765187245760019e-05, + "loss": 4.0565, + "step": 754176 + }, + { + "epoch": 0.02, + "learning_rate": 3.764348651008967e-05, + "loss": 4.0491, + "step": 754688 + }, + { + "epoch": 0.02, + "learning_rate": 3.763510056257915e-05, + "loss": 4.0536, + "step": 755200 + }, + { + "epoch": 0.02, + "learning_rate": 3.762671461506863e-05, + "loss": 4.048, + "step": 755712 + }, + { + "epoch": 0.02, + "learning_rate": 3.761832866755811e-05, + "loss": 4.0335, + "step": 756224 + }, + { + "epoch": 0.02, + "learning_rate": 3.760994272004759e-05, + "loss": 4.0408, + "step": 756736 + }, + { + "epoch": 0.02, + "learning_rate": 3.760155677253707e-05, + "loss": 4.0376, + "step": 757248 + }, + { + "epoch": 0.02, + "learning_rate": 3.759320358263401e-05, + "loss": 4.0562, + "step": 757760 + }, + { + "epoch": 0.02, + "learning_rate": 3.758481763512349e-05, + "loss": 4.0501, + "step": 758272 + }, + { + "epoch": 0.02, + "learning_rate": 3.757643168761297e-05, + "loss": 4.0482, + "step": 758784 + }, + { + "epoch": 0.02, + "learning_rate": 3.756804574010245e-05, + "loss": 4.0498, + "step": 759296 + }, + { + "epoch": 0.02, + "learning_rate": 3.755965979259193e-05, + "loss": 4.0403, + "step": 759808 + }, + { + "epoch": 0.02, + "learning_rate": 3.755127384508141e-05, + "loss": 4.0289, + "step": 760320 + }, + { + "epoch": 0.02, + "learning_rate": 3.75428878975709e-05, + "loss": 4.0447, + "step": 760832 + }, + { + "epoch": 0.02, + "learning_rate": 3.753450195006038e-05, + "loss": 4.0535, + "step": 761344 + }, + { + "epoch": 0.02, + "learning_rate": 3.7526132381353586e-05, + "loss": 4.0529, + "step": 761856 + }, + { + "epoch": 0.02, + "learning_rate": 3.7517746433843066e-05, + "loss": 4.0324, + "step": 762368 + }, + { + "epoch": 0.02, + "learning_rate": 3.7509360486332546e-05, + "loss": 4.0409, + "step": 762880 + }, + { + "epoch": 0.03, + "eval_loss": 4.070580005645752, + "eval_runtime": 563.8279, + "eval_samples_per_second": 676.786, + "eval_steps_per_second": 21.15, + "step": 763190 + }, + { + "epoch": 1.0, + "learning_rate": 3.7500974538822026e-05, + "loss": 4.0571, + "step": 763392 + }, + { + "epoch": 1.0, + "learning_rate": 3.7492588591311506e-05, + "loss": 4.0416, + "step": 763904 + }, + { + "epoch": 1.0, + "learning_rate": 3.7484202643800986e-05, + "loss": 4.0453, + "step": 764416 + }, + { + "epoch": 1.0, + "learning_rate": 3.7475833075094195e-05, + "loss": 4.0361, + "step": 764928 + }, + { + "epoch": 1.0, + "learning_rate": 3.7467447127583675e-05, + "loss": 4.0406, + "step": 765440 + }, + { + "epoch": 1.0, + "learning_rate": 3.7459061180073155e-05, + "loss": 4.0302, + "step": 765952 + }, + { + "epoch": 1.0, + "learning_rate": 3.7450675232562635e-05, + "loss": 4.0338, + "step": 766464 + }, + { + "epoch": 1.0, + "learning_rate": 3.744230566385585e-05, + "loss": 4.0307, + "step": 766976 + }, + { + "epoch": 1.0, + "learning_rate": 3.743391971634533e-05, + "loss": 4.0358, + "step": 767488 + }, + { + "epoch": 1.0, + "learning_rate": 3.742556652644227e-05, + "loss": 4.0453, + "step": 768000 + }, + { + "epoch": 1.0, + "learning_rate": 3.741718057893175e-05, + "loss": 4.0442, + "step": 768512 + }, + { + "epoch": 1.0, + "learning_rate": 3.740879463142123e-05, + "loss": 4.0397, + "step": 769024 + }, + { + "epoch": 1.0, + "learning_rate": 3.740040868391071e-05, + "loss": 4.0484, + "step": 769536 + }, + { + "epoch": 1.0, + "learning_rate": 3.739202273640019e-05, + "loss": 4.0351, + "step": 770048 + }, + { + "epoch": 1.0, + "learning_rate": 3.738363678888967e-05, + "loss": 4.0468, + "step": 770560 + }, + { + "epoch": 1.0, + "learning_rate": 3.737525084137915e-05, + "loss": 4.0278, + "step": 771072 + }, + { + "epoch": 1.0, + "learning_rate": 3.736686489386863e-05, + "loss": 4.0259, + "step": 771584 + }, + { + "epoch": 1.0, + "learning_rate": 3.735847894635811e-05, + "loss": 4.0469, + "step": 772096 + }, + { + "epoch": 1.0, + "learning_rate": 3.735009299884759e-05, + "loss": 4.0379, + "step": 772608 + }, + { + "epoch": 1.0, + "learning_rate": 3.734170705133707e-05, + "loss": 4.0521, + "step": 773120 + }, + { + "epoch": 1.0, + "learning_rate": 3.733332110382655e-05, + "loss": 4.0325, + "step": 773632 + }, + { + "epoch": 1.0, + "learning_rate": 3.7324951535119765e-05, + "loss": 4.0534, + "step": 774144 + }, + { + "epoch": 1.0, + "learning_rate": 3.7316581966412974e-05, + "loss": 4.0397, + "step": 774656 + }, + { + "epoch": 1.0, + "learning_rate": 3.7308196018902454e-05, + "loss": 4.0393, + "step": 775168 + }, + { + "epoch": 1.0, + "learning_rate": 3.7299810071391934e-05, + "loss": 4.0407, + "step": 775680 + }, + { + "epoch": 1.0, + "learning_rate": 3.7291424123881414e-05, + "loss": 4.0383, + "step": 776192 + }, + { + "epoch": 1.0, + "learning_rate": 3.7283038176370894e-05, + "loss": 4.0422, + "step": 776704 + }, + { + "epoch": 1.0, + "learning_rate": 3.7274652228860374e-05, + "loss": 4.0311, + "step": 777216 + }, + { + "epoch": 1.0, + "learning_rate": 3.726626628134985e-05, + "loss": 4.046, + "step": 777728 + }, + { + "epoch": 1.0, + "learning_rate": 3.725788033383933e-05, + "loss": 4.0388, + "step": 778240 + }, + { + "epoch": 1.01, + "learning_rate": 3.724949438632881e-05, + "loss": 4.0236, + "step": 778752 + }, + { + "epoch": 1.01, + "learning_rate": 3.724112481762202e-05, + "loss": 4.0437, + "step": 779264 + }, + { + "epoch": 1.01, + "learning_rate": 3.72327388701115e-05, + "loss": 4.0399, + "step": 779776 + }, + { + "epoch": 1.01, + "learning_rate": 3.722436930140472e-05, + "loss": 4.0432, + "step": 780288 + }, + { + "epoch": 1.01, + "learning_rate": 3.72159833538942e-05, + "loss": 4.0351, + "step": 780800 + }, + { + "epoch": 1.01, + "learning_rate": 3.720759740638367e-05, + "loss": 4.0395, + "step": 781312 + }, + { + "epoch": 1.01, + "learning_rate": 3.719921145887315e-05, + "loss": 4.0398, + "step": 781824 + }, + { + "epoch": 1.01, + "learning_rate": 3.719082551136263e-05, + "loss": 4.0429, + "step": 782336 + }, + { + "epoch": 1.01, + "learning_rate": 3.718243956385211e-05, + "loss": 4.0459, + "step": 782848 + }, + { + "epoch": 1.01, + "learning_rate": 3.717406999514532e-05, + "loss": 4.0274, + "step": 783360 + }, + { + "epoch": 1.01, + "learning_rate": 3.71656840476348e-05, + "loss": 4.0306, + "step": 783872 + }, + { + "epoch": 1.01, + "learning_rate": 3.715729810012428e-05, + "loss": 4.0281, + "step": 784384 + }, + { + "epoch": 1.01, + "learning_rate": 3.714891215261376e-05, + "loss": 4.0381, + "step": 784896 + }, + { + "epoch": 1.01, + "learning_rate": 3.714052620510324e-05, + "loss": 4.0395, + "step": 785408 + }, + { + "epoch": 1.01, + "learning_rate": 3.713214025759272e-05, + "loss": 4.0245, + "step": 785920 + }, + { + "epoch": 1.01, + "learning_rate": 3.712375431008221e-05, + "loss": 4.0482, + "step": 786432 + }, + { + "epoch": 1.01, + "learning_rate": 3.711536836257169e-05, + "loss": 4.0299, + "step": 786944 + }, + { + "epoch": 1.01, + "learning_rate": 3.7106998793864896e-05, + "loss": 4.0462, + "step": 787456 + }, + { + "epoch": 1.01, + "learning_rate": 3.7098612846354376e-05, + "loss": 4.0379, + "step": 787968 + }, + { + "epoch": 1.01, + "learning_rate": 3.7090243277647585e-05, + "loss": 4.0191, + "step": 788480 + }, + { + "epoch": 1.01, + "learning_rate": 3.7081857330137065e-05, + "loss": 4.0313, + "step": 788992 + }, + { + "epoch": 1.01, + "learning_rate": 3.7073471382626545e-05, + "loss": 4.0251, + "step": 789504 + }, + { + "epoch": 1.01, + "learning_rate": 3.7065085435116025e-05, + "loss": 4.0371, + "step": 790016 + }, + { + "epoch": 1.01, + "learning_rate": 3.7056699487605505e-05, + "loss": 4.0292, + "step": 790528 + }, + { + "epoch": 1.01, + "learning_rate": 3.7048313540094985e-05, + "loss": 4.0305, + "step": 791040 + }, + { + "epoch": 1.01, + "learning_rate": 3.7039943971388194e-05, + "loss": 4.0136, + "step": 791552 + }, + { + "epoch": 1.01, + "learning_rate": 3.7031558023877674e-05, + "loss": 4.0321, + "step": 792064 + }, + { + "epoch": 1.01, + "learning_rate": 3.702317207636716e-05, + "loss": 4.0206, + "step": 792576 + }, + { + "epoch": 1.01, + "learning_rate": 3.701478612885664e-05, + "loss": 4.0361, + "step": 793088 + }, + { + "epoch": 1.01, + "learning_rate": 3.700640018134612e-05, + "loss": 4.0267, + "step": 793600 + }, + { + "epoch": 1.01, + "learning_rate": 3.69980142338356e-05, + "loss": 4.0273, + "step": 794112 + }, + { + "epoch": 1.01, + "learning_rate": 3.698962828632508e-05, + "loss": 4.0275, + "step": 794624 + }, + { + "epoch": 1.01, + "learning_rate": 3.698124233881456e-05, + "loss": 4.0151, + "step": 795136 + }, + { + "epoch": 1.01, + "learning_rate": 3.697287277010777e-05, + "loss": 4.0343, + "step": 795648 + }, + { + "epoch": 1.01, + "learning_rate": 3.696448682259725e-05, + "loss": 4.0256, + "step": 796160 + }, + { + "epoch": 1.01, + "learning_rate": 3.695610087508673e-05, + "loss": 4.0392, + "step": 796672 + }, + { + "epoch": 1.01, + "learning_rate": 3.69477149275762e-05, + "loss": 4.0212, + "step": 797184 + }, + { + "epoch": 1.01, + "learning_rate": 3.693934535886942e-05, + "loss": 4.0271, + "step": 797696 + }, + { + "epoch": 1.01, + "learning_rate": 3.69309594113589e-05, + "loss": 4.0233, + "step": 798208 + }, + { + "epoch": 1.01, + "learning_rate": 3.692257346384838e-05, + "loss": 4.0274, + "step": 798720 + }, + { + "epoch": 1.01, + "learning_rate": 3.691418751633786e-05, + "loss": 4.0258, + "step": 799232 + }, + { + "epoch": 1.01, + "learning_rate": 3.6905817947631074e-05, + "loss": 4.0315, + "step": 799744 + }, + { + "epoch": 1.01, + "learning_rate": 3.6897432000120554e-05, + "loss": 4.0284, + "step": 800256 + }, + { + "epoch": 1.01, + "learning_rate": 3.6889046052610034e-05, + "loss": 4.0486, + "step": 800768 + }, + { + "epoch": 1.01, + "learning_rate": 3.688066010509951e-05, + "loss": 4.0387, + "step": 801280 + }, + { + "epoch": 1.01, + "learning_rate": 3.687227415758899e-05, + "loss": 4.0323, + "step": 801792 + }, + { + "epoch": 1.01, + "learning_rate": 3.68639045888822e-05, + "loss": 4.0335, + "step": 802304 + }, + { + "epoch": 1.01, + "learning_rate": 3.685555139897914e-05, + "loss": 4.0178, + "step": 802816 + }, + { + "epoch": 1.01, + "learning_rate": 3.684716545146862e-05, + "loss": 4.0293, + "step": 803328 + }, + { + "epoch": 1.01, + "learning_rate": 3.68387795039581e-05, + "loss": 4.0307, + "step": 803840 + }, + { + "epoch": 1.01, + "learning_rate": 3.683039355644758e-05, + "loss": 4.0408, + "step": 804352 + }, + { + "epoch": 1.01, + "learning_rate": 3.682200760893707e-05, + "loss": 4.0338, + "step": 804864 + }, + { + "epoch": 1.01, + "learning_rate": 3.681362166142655e-05, + "loss": 4.0186, + "step": 805376 + }, + { + "epoch": 1.01, + "learning_rate": 3.680523571391603e-05, + "loss": 4.0265, + "step": 805888 + }, + { + "epoch": 1.01, + "learning_rate": 3.679684976640551e-05, + "loss": 4.0167, + "step": 806400 + }, + { + "epoch": 1.01, + "learning_rate": 3.678848019769872e-05, + "loss": 4.027, + "step": 806912 + }, + { + "epoch": 1.01, + "learning_rate": 3.67800942501882e-05, + "loss": 4.0234, + "step": 807424 + }, + { + "epoch": 1.01, + "learning_rate": 3.677170830267768e-05, + "loss": 4.0301, + "step": 807936 + }, + { + "epoch": 1.01, + "learning_rate": 3.676332235516715e-05, + "loss": 4.0374, + "step": 808448 + }, + { + "epoch": 1.01, + "learning_rate": 3.675493640765663e-05, + "loss": 4.0345, + "step": 808960 + }, + { + "epoch": 1.02, + "learning_rate": 3.6746566838949846e-05, + "loss": 4.0189, + "step": 809472 + }, + { + "epoch": 1.02, + "learning_rate": 3.6738180891439326e-05, + "loss": 4.0277, + "step": 809984 + }, + { + "epoch": 1.02, + "learning_rate": 3.6729794943928806e-05, + "loss": 4.0298, + "step": 810496 + }, + { + "epoch": 1.02, + "learning_rate": 3.6721408996418286e-05, + "loss": 4.0192, + "step": 811008 + }, + { + "epoch": 1.02, + "learning_rate": 3.6713023048907766e-05, + "loss": 4.0197, + "step": 811520 + }, + { + "epoch": 1.02, + "learning_rate": 3.6704637101397246e-05, + "loss": 4.0289, + "step": 812032 + }, + { + "epoch": 1.02, + "learning_rate": 3.6696251153886726e-05, + "loss": 4.0379, + "step": 812544 + }, + { + "epoch": 1.02, + "learning_rate": 3.6687881585179935e-05, + "loss": 4.0378, + "step": 813056 + }, + { + "epoch": 1.02, + "learning_rate": 3.6679495637669415e-05, + "loss": 4.0261, + "step": 813568 + }, + { + "epoch": 1.02, + "learning_rate": 3.6671109690158895e-05, + "loss": 4.0335, + "step": 814080 + }, + { + "epoch": 1.02, + "learning_rate": 3.6662723742648375e-05, + "loss": 4.023, + "step": 814592 + }, + { + "epoch": 1.02, + "learning_rate": 3.6654337795137854e-05, + "loss": 4.032, + "step": 815104 + }, + { + "epoch": 1.02, + "learning_rate": 3.6645951847627334e-05, + "loss": 4.0266, + "step": 815616 + }, + { + "epoch": 1.02, + "learning_rate": 3.6637565900116814e-05, + "loss": 4.0371, + "step": 816128 + }, + { + "epoch": 1.02, + "learning_rate": 3.6629179952606294e-05, + "loss": 4.033, + "step": 816640 + }, + { + "epoch": 1.02, + "learning_rate": 3.6620794005095774e-05, + "loss": 4.0107, + "step": 817152 + }, + { + "epoch": 1.02, + "learning_rate": 3.6612424436388983e-05, + "loss": 4.0145, + "step": 817664 + }, + { + "epoch": 1.02, + "learning_rate": 3.660403848887847e-05, + "loss": 4.0146, + "step": 818176 + }, + { + "epoch": 1.02, + "learning_rate": 3.659565254136795e-05, + "loss": 4.0207, + "step": 818688 + }, + { + "epoch": 1.02, + "learning_rate": 3.658726659385743e-05, + "loss": 4.028, + "step": 819200 + }, + { + "epoch": 1.02, + "learning_rate": 3.657889702515064e-05, + "loss": 4.0324, + "step": 819712 + }, + { + "epoch": 1.02, + "learning_rate": 3.657052745644385e-05, + "loss": 4.0084, + "step": 820224 + }, + { + "epoch": 1.02, + "learning_rate": 3.656214150893333e-05, + "loss": 4.0226, + "step": 820736 + }, + { + "epoch": 1.02, + "learning_rate": 3.655375556142281e-05, + "loss": 4.034, + "step": 821248 + }, + { + "epoch": 1.02, + "learning_rate": 3.654536961391229e-05, + "loss": 4.0291, + "step": 821760 + }, + { + "epoch": 1.02, + "learning_rate": 3.653698366640177e-05, + "loss": 4.0322, + "step": 822272 + }, + { + "epoch": 1.02, + "learning_rate": 3.652859771889125e-05, + "loss": 4.0247, + "step": 822784 + }, + { + "epoch": 1.02, + "learning_rate": 3.652021177138073e-05, + "loss": 4.0121, + "step": 823296 + }, + { + "epoch": 1.02, + "learning_rate": 3.651184220267394e-05, + "loss": 4.0279, + "step": 823808 + }, + { + "epoch": 1.02, + "learning_rate": 3.6503456255163424e-05, + "loss": 4.0358, + "step": 824320 + }, + { + "epoch": 1.02, + "learning_rate": 3.6495070307652904e-05, + "loss": 4.0213, + "step": 824832 + }, + { + "epoch": 1.02, + "learning_rate": 3.6486684360142384e-05, + "loss": 4.0139, + "step": 825344 + }, + { + "epoch": 1.02, + "learning_rate": 3.6478298412631864e-05, + "loss": 4.0196, + "step": 825856 + }, + { + "epoch": 1.02, + "learning_rate": 3.6469912465121344e-05, + "loss": 4.0212, + "step": 826368 + }, + { + "epoch": 1.02, + "learning_rate": 3.646152651761082e-05, + "loss": 4.0276, + "step": 826880 + }, + { + "epoch": 1.02, + "learning_rate": 3.64531405701003e-05, + "loss": 4.0215, + "step": 827392 + }, + { + "epoch": 1.02, + "learning_rate": 3.644477100139351e-05, + "loss": 4.0333, + "step": 827904 + }, + { + "epoch": 1.02, + "learning_rate": 3.6436385053882986e-05, + "loss": 4.0098, + "step": 828416 + }, + { + "epoch": 1.02, + "learning_rate": 3.64280154851762e-05, + "loss": 4.0337, + "step": 828928 + }, + { + "epoch": 1.02, + "learning_rate": 3.641962953766568e-05, + "loss": 4.0199, + "step": 829440 + }, + { + "epoch": 1.02, + "learning_rate": 3.641124359015516e-05, + "loss": 4.0313, + "step": 829952 + }, + { + "epoch": 1.02, + "learning_rate": 3.640285764264464e-05, + "loss": 4.0418, + "step": 830464 + }, + { + "epoch": 1.02, + "learning_rate": 3.639447169513412e-05, + "loss": 4.0246, + "step": 830976 + }, + { + "epoch": 1.02, + "learning_rate": 3.63860857476236e-05, + "loss": 4.0305, + "step": 831488 + }, + { + "epoch": 1.02, + "learning_rate": 3.637771617891681e-05, + "loss": 4.0295, + "step": 832000 + }, + { + "epoch": 1.02, + "learning_rate": 3.636933023140629e-05, + "loss": 4.0138, + "step": 832512 + }, + { + "epoch": 1.02, + "learning_rate": 3.636094428389577e-05, + "loss": 4.0192, + "step": 833024 + }, + { + "epoch": 1.02, + "learning_rate": 3.635255833638525e-05, + "loss": 4.0193, + "step": 833536 + }, + { + "epoch": 1.02, + "learning_rate": 3.634417238887473e-05, + "loss": 4.0314, + "step": 834048 + }, + { + "epoch": 1.02, + "learning_rate": 3.633578644136421e-05, + "loss": 4.0322, + "step": 834560 + }, + { + "epoch": 1.02, + "learning_rate": 3.632741687265742e-05, + "loss": 4.0299, + "step": 835072 + }, + { + "epoch": 1.02, + "learning_rate": 3.6319047303950635e-05, + "loss": 4.0264, + "step": 835584 + }, + { + "epoch": 1.02, + "learning_rate": 3.6310661356440115e-05, + "loss": 4.0189, + "step": 836096 + }, + { + "epoch": 1.02, + "learning_rate": 3.6302275408929595e-05, + "loss": 4.01, + "step": 836608 + }, + { + "epoch": 1.02, + "learning_rate": 3.6293889461419075e-05, + "loss": 4.025, + "step": 837120 + }, + { + "epoch": 1.02, + "learning_rate": 3.6285503513908555e-05, + "loss": 4.0317, + "step": 837632 + }, + { + "epoch": 1.02, + "learning_rate": 3.6277117566398035e-05, + "loss": 4.0337, + "step": 838144 + }, + { + "epoch": 1.02, + "learning_rate": 3.6268731618887515e-05, + "loss": 4.0159, + "step": 838656 + }, + { + "epoch": 1.02, + "learning_rate": 3.6260345671376995e-05, + "loss": 4.0209, + "step": 839168 + }, + { + "epoch": 1.03, + "eval_loss": 4.05756139755249, + "eval_runtime": 562.3468, + "eval_samples_per_second": 678.569, + "eval_steps_per_second": 21.206, + "step": 839509 + }, + { + "epoch": 0.0, + "learning_rate": 3.6251976102670204e-05, + "loss": 4.0473, + "step": 839680 + }, + { + "epoch": 0.0, + "learning_rate": 3.6243590155159684e-05, + "loss": 4.0262, + "step": 840192 + }, + { + "epoch": 0.0, + "learning_rate": 3.6235204207649164e-05, + "loss": 4.0227, + "step": 840704 + }, + { + "epoch": 0.0, + "learning_rate": 3.6226818260138644e-05, + "loss": 4.0178, + "step": 841216 + }, + { + "epoch": 0.0, + "learning_rate": 3.6218432312628124e-05, + "loss": 4.0225, + "step": 841728 + }, + { + "epoch": 0.0, + "learning_rate": 3.6210046365117604e-05, + "loss": 4.0097, + "step": 842240 + }, + { + "epoch": 0.0, + "learning_rate": 3.6201660417607084e-05, + "loss": 4.0117, + "step": 842752 + }, + { + "epoch": 0.0, + "learning_rate": 3.6193274470096564e-05, + "loss": 4.0128, + "step": 843264 + }, + { + "epoch": 0.0, + "learning_rate": 3.6184888522586044e-05, + "loss": 4.0222, + "step": 843776 + }, + { + "epoch": 0.0, + "learning_rate": 3.617650257507553e-05, + "loss": 4.021, + "step": 844288 + }, + { + "epoch": 0.0, + "learning_rate": 3.6168116627565004e-05, + "loss": 4.0192, + "step": 844800 + }, + { + "epoch": 0.0, + "learning_rate": 3.6159730680054484e-05, + "loss": 4.0222, + "step": 845312 + }, + { + "epoch": 0.0, + "learning_rate": 3.6151344732543964e-05, + "loss": 4.0273, + "step": 845824 + }, + { + "epoch": 0.0, + "learning_rate": 3.6142958785033444e-05, + "loss": 4.0169, + "step": 846336 + }, + { + "epoch": 0.0, + "learning_rate": 3.6134572837522923e-05, + "loss": 4.028, + "step": 846848 + }, + { + "epoch": 0.0, + "learning_rate": 3.6126186890012403e-05, + "loss": 4.0082, + "step": 847360 + }, + { + "epoch": 0.0, + "learning_rate": 3.611780094250188e-05, + "loss": 4.0058, + "step": 847872 + }, + { + "epoch": 0.0, + "learning_rate": 3.610941499499136e-05, + "loss": 4.0283, + "step": 848384 + }, + { + "epoch": 0.0, + "learning_rate": 3.610104542628457e-05, + "loss": 4.0197, + "step": 848896 + }, + { + "epoch": 0.0, + "learning_rate": 3.609265947877405e-05, + "loss": 4.0281, + "step": 849408 + }, + { + "epoch": 0.0, + "learning_rate": 3.608427353126353e-05, + "loss": 4.0164, + "step": 849920 + }, + { + "epoch": 0.0, + "learning_rate": 3.607588758375301e-05, + "loss": 4.0301, + "step": 850432 + }, + { + "epoch": 0.0, + "learning_rate": 3.606751801504623e-05, + "loss": 4.0186, + "step": 850944 + }, + { + "epoch": 0.0, + "learning_rate": 3.605913206753571e-05, + "loss": 4.0196, + "step": 851456 + }, + { + "epoch": 0.0, + "learning_rate": 3.605074612002519e-05, + "loss": 4.0194, + "step": 851968 + }, + { + "epoch": 0.0, + "learning_rate": 3.604236017251467e-05, + "loss": 4.021, + "step": 852480 + }, + { + "epoch": 0.0, + "learning_rate": 3.603397422500415e-05, + "loss": 4.0191, + "step": 852992 + }, + { + "epoch": 0.0, + "learning_rate": 3.602558827749363e-05, + "loss": 4.0115, + "step": 853504 + }, + { + "epoch": 0.0, + "learning_rate": 3.601721870878684e-05, + "loss": 4.0264, + "step": 854016 + }, + { + "epoch": 0.0, + "learning_rate": 3.600883276127632e-05, + "loss": 4.016, + "step": 854528 + }, + { + "epoch": 0.01, + "learning_rate": 3.60004468137658e-05, + "loss": 4.0096, + "step": 855040 + }, + { + "epoch": 0.01, + "learning_rate": 3.599206086625528e-05, + "loss": 4.0186, + "step": 855552 + }, + { + "epoch": 0.01, + "learning_rate": 3.598367491874476e-05, + "loss": 4.0187, + "step": 856064 + }, + { + "epoch": 0.01, + "learning_rate": 3.597528897123424e-05, + "loss": 4.0265, + "step": 856576 + }, + { + "epoch": 0.01, + "learning_rate": 3.596690302372372e-05, + "loss": 4.0138, + "step": 857088 + }, + { + "epoch": 0.01, + "learning_rate": 3.595853345501693e-05, + "loss": 4.0192, + "step": 857600 + }, + { + "epoch": 0.01, + "learning_rate": 3.595014750750641e-05, + "loss": 4.0213, + "step": 858112 + }, + { + "epoch": 0.01, + "learning_rate": 3.594176155999589e-05, + "loss": 4.0215, + "step": 858624 + }, + { + "epoch": 0.01, + "learning_rate": 3.5933375612485366e-05, + "loss": 4.028, + "step": 859136 + }, + { + "epoch": 0.01, + "learning_rate": 3.5924989664974846e-05, + "loss": 4.0066, + "step": 859648 + }, + { + "epoch": 0.01, + "learning_rate": 3.5916603717464326e-05, + "loss": 4.0133, + "step": 860160 + }, + { + "epoch": 0.01, + "learning_rate": 3.5908217769953806e-05, + "loss": 4.0088, + "step": 860672 + }, + { + "epoch": 0.01, + "learning_rate": 3.5899848201247015e-05, + "loss": 4.0136, + "step": 861184 + }, + { + "epoch": 0.01, + "learning_rate": 3.5891462253736495e-05, + "loss": 4.02, + "step": 861696 + }, + { + "epoch": 0.01, + "learning_rate": 3.5883076306225975e-05, + "loss": 4.0075, + "step": 862208 + }, + { + "epoch": 0.01, + "learning_rate": 3.5874690358715455e-05, + "loss": 4.0258, + "step": 862720 + }, + { + "epoch": 0.01, + "learning_rate": 3.5866304411204935e-05, + "loss": 4.0119, + "step": 863232 + }, + { + "epoch": 0.01, + "learning_rate": 3.5857918463694414e-05, + "loss": 4.0274, + "step": 863744 + }, + { + "epoch": 0.01, + "learning_rate": 3.58495325161839e-05, + "loss": 4.018, + "step": 864256 + }, + { + "epoch": 0.01, + "learning_rate": 3.584114656867338e-05, + "loss": 3.9986, + "step": 864768 + }, + { + "epoch": 0.01, + "learning_rate": 3.583277699996659e-05, + "loss": 4.0099, + "step": 865280 + }, + { + "epoch": 0.01, + "learning_rate": 3.582439105245607e-05, + "loss": 4.0063, + "step": 865792 + }, + { + "epoch": 0.01, + "learning_rate": 3.581600510494555e-05, + "loss": 4.0163, + "step": 866304 + }, + { + "epoch": 0.01, + "learning_rate": 3.580763553623876e-05, + "loss": 4.0076, + "step": 866816 + }, + { + "epoch": 0.01, + "learning_rate": 3.579924958872824e-05, + "loss": 4.0129, + "step": 867328 + }, + { + "epoch": 0.01, + "learning_rate": 3.579086364121772e-05, + "loss": 3.9963, + "step": 867840 + }, + { + "epoch": 0.01, + "learning_rate": 3.57824776937072e-05, + "loss": 4.0119, + "step": 868352 + }, + { + "epoch": 0.01, + "learning_rate": 3.577409174619668e-05, + "loss": 4.0058, + "step": 868864 + }, + { + "epoch": 0.01, + "learning_rate": 3.576570579868616e-05, + "loss": 4.0161, + "step": 869376 + }, + { + "epoch": 0.01, + "learning_rate": 3.575731985117564e-05, + "loss": 3.9997, + "step": 869888 + }, + { + "epoch": 0.01, + "learning_rate": 3.5748950282468855e-05, + "loss": 4.0143, + "step": 870400 + }, + { + "epoch": 0.01, + "learning_rate": 3.5740564334958335e-05, + "loss": 4.0061, + "step": 870912 + }, + { + "epoch": 0.01, + "learning_rate": 3.5732178387447815e-05, + "loss": 3.9993, + "step": 871424 + }, + { + "epoch": 0.01, + "learning_rate": 3.5723792439937295e-05, + "loss": 4.0135, + "step": 871936 + }, + { + "epoch": 0.01, + "learning_rate": 3.5715406492426775e-05, + "loss": 4.0102, + "step": 872448 + }, + { + "epoch": 0.01, + "learning_rate": 3.570702054491625e-05, + "loss": 4.018, + "step": 872960 + }, + { + "epoch": 0.01, + "learning_rate": 3.569863459740573e-05, + "loss": 4.0015, + "step": 873472 + }, + { + "epoch": 0.01, + "learning_rate": 3.569024864989521e-05, + "loss": 4.0119, + "step": 873984 + }, + { + "epoch": 0.01, + "learning_rate": 3.5681879081188424e-05, + "loss": 4.0034, + "step": 874496 + }, + { + "epoch": 0.01, + "learning_rate": 3.56734931336779e-05, + "loss": 4.0095, + "step": 875008 + }, + { + "epoch": 0.01, + "learning_rate": 3.566510718616738e-05, + "loss": 4.0052, + "step": 875520 + }, + { + "epoch": 0.01, + "learning_rate": 3.565672123865686e-05, + "loss": 4.0133, + "step": 876032 + }, + { + "epoch": 0.01, + "learning_rate": 3.564835166995007e-05, + "loss": 4.0073, + "step": 876544 + }, + { + "epoch": 0.01, + "learning_rate": 3.563996572243955e-05, + "loss": 4.0299, + "step": 877056 + }, + { + "epoch": 0.01, + "learning_rate": 3.563159615373277e-05, + "loss": 4.0217, + "step": 877568 + }, + { + "epoch": 0.01, + "learning_rate": 3.562321020622225e-05, + "loss": 4.0106, + "step": 878080 + }, + { + "epoch": 0.01, + "learning_rate": 3.561482425871172e-05, + "loss": 4.0117, + "step": 878592 + }, + { + "epoch": 0.01, + "learning_rate": 3.56064383112012e-05, + "loss": 4.0036, + "step": 879104 + }, + { + "epoch": 0.01, + "learning_rate": 3.559805236369068e-05, + "loss": 4.0055, + "step": 879616 + }, + { + "epoch": 0.01, + "learning_rate": 3.558966641618016e-05, + "loss": 4.0146, + "step": 880128 + }, + { + "epoch": 0.01, + "learning_rate": 3.558128046866964e-05, + "loss": 4.0176, + "step": 880640 + }, + { + "epoch": 0.01, + "learning_rate": 3.557291089996285e-05, + "loss": 4.021, + "step": 881152 + }, + { + "epoch": 0.01, + "learning_rate": 3.556452495245233e-05, + "loss": 3.9997, + "step": 881664 + }, + { + "epoch": 0.01, + "learning_rate": 3.5556155383745546e-05, + "loss": 4.0057, + "step": 882176 + }, + { + "epoch": 0.01, + "learning_rate": 3.5547769436235026e-05, + "loss": 3.9958, + "step": 882688 + }, + { + "epoch": 0.01, + "learning_rate": 3.5539383488724506e-05, + "loss": 4.0134, + "step": 883200 + }, + { + "epoch": 0.01, + "learning_rate": 3.5530997541213986e-05, + "loss": 4.0033, + "step": 883712 + }, + { + "epoch": 0.01, + "learning_rate": 3.5522611593703466e-05, + "loss": 4.0107, + "step": 884224 + }, + { + "epoch": 0.01, + "learning_rate": 3.5514225646192946e-05, + "loss": 4.0174, + "step": 884736 + }, + { + "epoch": 0.01, + "learning_rate": 3.5505856077486155e-05, + "loss": 4.017, + "step": 885248 + }, + { + "epoch": 0.02, + "learning_rate": 3.5497470129975635e-05, + "loss": 3.9999, + "step": 885760 + }, + { + "epoch": 0.02, + "learning_rate": 3.5489084182465115e-05, + "loss": 4.0052, + "step": 886272 + }, + { + "epoch": 0.02, + "learning_rate": 3.5480698234954595e-05, + "loss": 4.0145, + "step": 886784 + }, + { + "epoch": 0.02, + "learning_rate": 3.5472312287444075e-05, + "loss": 4.0027, + "step": 887296 + }, + { + "epoch": 0.02, + "learning_rate": 3.5463926339933555e-05, + "loss": 4.0005, + "step": 887808 + }, + { + "epoch": 0.02, + "learning_rate": 3.5455540392423035e-05, + "loss": 4.0135, + "step": 888320 + }, + { + "epoch": 0.02, + "learning_rate": 3.5447154444912515e-05, + "loss": 4.0176, + "step": 888832 + }, + { + "epoch": 0.02, + "learning_rate": 3.5438784876205724e-05, + "loss": 4.0208, + "step": 889344 + }, + { + "epoch": 0.02, + "learning_rate": 3.543039892869521e-05, + "loss": 4.011, + "step": 889856 + }, + { + "epoch": 0.02, + "learning_rate": 3.542201298118469e-05, + "loss": 4.0122, + "step": 890368 + }, + { + "epoch": 0.02, + "learning_rate": 3.541362703367417e-05, + "loss": 4.002, + "step": 890880 + }, + { + "epoch": 0.02, + "learning_rate": 3.540524108616365e-05, + "loss": 4.0127, + "step": 891392 + }, + { + "epoch": 0.02, + "learning_rate": 3.539687151745686e-05, + "loss": 4.0099, + "step": 891904 + }, + { + "epoch": 0.02, + "learning_rate": 3.538848556994634e-05, + "loss": 4.019, + "step": 892416 + }, + { + "epoch": 0.02, + "learning_rate": 3.538009962243582e-05, + "loss": 4.0093, + "step": 892928 + }, + { + "epoch": 0.02, + "learning_rate": 3.53717136749253e-05, + "loss": 3.9958, + "step": 893440 + }, + { + "epoch": 0.02, + "learning_rate": 3.536332772741478e-05, + "loss": 3.9973, + "step": 893952 + }, + { + "epoch": 0.02, + "learning_rate": 3.535495815870799e-05, + "loss": 3.9912, + "step": 894464 + }, + { + "epoch": 0.02, + "learning_rate": 3.534657221119747e-05, + "loss": 4.0042, + "step": 894976 + }, + { + "epoch": 0.02, + "learning_rate": 3.533818626368695e-05, + "loss": 4.012, + "step": 895488 + }, + { + "epoch": 0.02, + "learning_rate": 3.532980031617643e-05, + "loss": 4.0103, + "step": 896000 + }, + { + "epoch": 0.02, + "learning_rate": 3.532141436866591e-05, + "loss": 3.9932, + "step": 896512 + }, + { + "epoch": 0.02, + "learning_rate": 3.531302842115539e-05, + "loss": 4.001, + "step": 897024 + }, + { + "epoch": 0.02, + "learning_rate": 3.530464247364487e-05, + "loss": 4.015, + "step": 897536 + }, + { + "epoch": 0.02, + "learning_rate": 3.529625652613435e-05, + "loss": 4.0139, + "step": 898048 + }, + { + "epoch": 0.02, + "learning_rate": 3.528788695742756e-05, + "loss": 4.016, + "step": 898560 + }, + { + "epoch": 0.02, + "learning_rate": 3.527950100991704e-05, + "loss": 4.0053, + "step": 899072 + }, + { + "epoch": 0.02, + "learning_rate": 3.527111506240652e-05, + "loss": 3.9984, + "step": 899584 + }, + { + "epoch": 0.02, + "learning_rate": 3.5262729114896e-05, + "loss": 4.0085, + "step": 900096 + }, + { + "epoch": 0.02, + "learning_rate": 3.5254359546189206e-05, + "loss": 4.0142, + "step": 900608 + }, + { + "epoch": 0.02, + "learning_rate": 3.524598997748242e-05, + "loss": 4.005, + "step": 901120 + }, + { + "epoch": 0.02, + "learning_rate": 3.52376040299719e-05, + "loss": 3.9951, + "step": 901632 + }, + { + "epoch": 0.02, + "learning_rate": 3.522921808246138e-05, + "loss": 4.0012, + "step": 902144 + }, + { + "epoch": 0.02, + "learning_rate": 3.522083213495086e-05, + "loss": 4.0065, + "step": 902656 + }, + { + "epoch": 0.02, + "learning_rate": 3.521244618744034e-05, + "loss": 4.0098, + "step": 903168 + }, + { + "epoch": 0.02, + "learning_rate": 3.520406023992982e-05, + "loss": 4.003, + "step": 903680 + }, + { + "epoch": 0.02, + "learning_rate": 3.51956742924193e-05, + "loss": 4.0162, + "step": 904192 + }, + { + "epoch": 0.02, + "learning_rate": 3.518730472371251e-05, + "loss": 3.9942, + "step": 904704 + }, + { + "epoch": 0.02, + "learning_rate": 3.517891877620199e-05, + "loss": 4.0162, + "step": 905216 + }, + { + "epoch": 0.02, + "learning_rate": 3.517053282869147e-05, + "loss": 4.0025, + "step": 905728 + }, + { + "epoch": 0.02, + "learning_rate": 3.516214688118095e-05, + "loss": 4.0148, + "step": 906240 + }, + { + "epoch": 0.02, + "learning_rate": 3.515377731247416e-05, + "loss": 4.0209, + "step": 906752 + }, + { + "epoch": 0.02, + "learning_rate": 3.514539136496364e-05, + "loss": 4.0125, + "step": 907264 + }, + { + "epoch": 0.02, + "learning_rate": 3.513700541745312e-05, + "loss": 4.0087, + "step": 907776 + }, + { + "epoch": 0.02, + "learning_rate": 3.51286194699426e-05, + "loss": 4.0143, + "step": 908288 + }, + { + "epoch": 0.02, + "learning_rate": 3.512023352243209e-05, + "loss": 3.9932, + "step": 908800 + }, + { + "epoch": 0.02, + "learning_rate": 3.5111863953725296e-05, + "loss": 4.0028, + "step": 909312 + }, + { + "epoch": 0.02, + "learning_rate": 3.5103478006214776e-05, + "loss": 3.999, + "step": 909824 + }, + { + "epoch": 0.02, + "learning_rate": 3.5095092058704256e-05, + "loss": 4.0172, + "step": 910336 + }, + { + "epoch": 0.02, + "learning_rate": 3.5086706111193736e-05, + "loss": 4.0142, + "step": 910848 + }, + { + "epoch": 0.02, + "learning_rate": 3.5078320163683216e-05, + "loss": 4.0167, + "step": 911360 + }, + { + "epoch": 0.02, + "learning_rate": 3.5069950594976425e-05, + "loss": 4.0099, + "step": 911872 + }, + { + "epoch": 0.02, + "learning_rate": 3.5061564647465905e-05, + "loss": 4.0005, + "step": 912384 + }, + { + "epoch": 0.02, + "learning_rate": 3.5053178699955385e-05, + "loss": 3.9941, + "step": 912896 + }, + { + "epoch": 0.02, + "learning_rate": 3.5044792752444865e-05, + "loss": 4.0084, + "step": 913408 + }, + { + "epoch": 0.02, + "learning_rate": 3.5036423183738074e-05, + "loss": 4.0124, + "step": 913920 + }, + { + "epoch": 0.02, + "learning_rate": 3.5028037236227554e-05, + "loss": 4.0196, + "step": 914432 + }, + { + "epoch": 0.02, + "learning_rate": 3.501965128871704e-05, + "loss": 3.9953, + "step": 914944 + }, + { + "epoch": 0.02, + "learning_rate": 3.501126534120652e-05, + "loss": 4.0041, + "step": 915456 + }, + { + "epoch": 0.03, + "eval_loss": 4.046953201293945, + "eval_runtime": 553.0572, + "eval_samples_per_second": 689.967, + "eval_steps_per_second": 21.562, + "step": 915828 + }, + { + "epoch": 0.0, + "learning_rate": 3.5002879393696e-05, + "loss": 4.0287, + "step": 915968 + }, + { + "epoch": 0.0, + "learning_rate": 3.499449344618548e-05, + "loss": 4.0096, + "step": 916480 + }, + { + "epoch": 0.0, + "learning_rate": 3.498610749867496e-05, + "loss": 4.0052, + "step": 916992 + }, + { + "epoch": 0.0, + "learning_rate": 3.497772155116444e-05, + "loss": 4.0028, + "step": 917504 + }, + { + "epoch": 0.0, + "learning_rate": 3.496933560365392e-05, + "loss": 4.0059, + "step": 918016 + }, + { + "epoch": 0.0, + "learning_rate": 3.496094965614339e-05, + "loss": 3.9932, + "step": 918528 + }, + { + "epoch": 0.0, + "learning_rate": 3.495256370863287e-05, + "loss": 3.9964, + "step": 919040 + }, + { + "epoch": 0.0, + "learning_rate": 3.494417776112235e-05, + "loss": 3.9907, + "step": 919552 + }, + { + "epoch": 0.0, + "learning_rate": 3.493579181361183e-05, + "loss": 4.009, + "step": 920064 + }, + { + "epoch": 0.0, + "learning_rate": 3.492740586610131e-05, + "loss": 4.0005, + "step": 920576 + }, + { + "epoch": 0.0, + "learning_rate": 3.491901991859079e-05, + "loss": 4.0034, + "step": 921088 + }, + { + "epoch": 0.0, + "learning_rate": 3.491063397108027e-05, + "loss": 4.0053, + "step": 921600 + }, + { + "epoch": 0.0, + "learning_rate": 3.490224802356975e-05, + "loss": 4.0072, + "step": 922112 + }, + { + "epoch": 0.0, + "learning_rate": 3.489386207605923e-05, + "loss": 4.0029, + "step": 922624 + }, + { + "epoch": 0.0, + "learning_rate": 3.488547612854872e-05, + "loss": 4.0093, + "step": 923136 + }, + { + "epoch": 0.0, + "learning_rate": 3.48770901810382e-05, + "loss": 3.9937, + "step": 923648 + }, + { + "epoch": 0.0, + "learning_rate": 3.486870423352768e-05, + "loss": 3.9875, + "step": 924160 + }, + { + "epoch": 0.0, + "learning_rate": 3.486031828601716e-05, + "loss": 4.0085, + "step": 924672 + }, + { + "epoch": 0.0, + "learning_rate": 3.485194871731037e-05, + "loss": 4.001, + "step": 925184 + }, + { + "epoch": 0.0, + "learning_rate": 3.484356276979985e-05, + "loss": 4.0141, + "step": 925696 + }, + { + "epoch": 0.0, + "learning_rate": 3.483517682228933e-05, + "loss": 3.9965, + "step": 926208 + }, + { + "epoch": 0.0, + "learning_rate": 3.482679087477881e-05, + "loss": 4.0142, + "step": 926720 + }, + { + "epoch": 0.0, + "learning_rate": 3.481840492726828e-05, + "loss": 4.0032, + "step": 927232 + }, + { + "epoch": 0.0, + "learning_rate": 3.481001897975776e-05, + "loss": 3.9991, + "step": 927744 + }, + { + "epoch": 0.0, + "learning_rate": 3.480164941105098e-05, + "loss": 4.0038, + "step": 928256 + }, + { + "epoch": 0.0, + "learning_rate": 3.479326346354046e-05, + "loss": 4.003, + "step": 928768 + }, + { + "epoch": 0.0, + "learning_rate": 3.478487751602994e-05, + "loss": 4.0071, + "step": 929280 + }, + { + "epoch": 0.0, + "learning_rate": 3.477649156851942e-05, + "loss": 3.999, + "step": 929792 + }, + { + "epoch": 0.0, + "learning_rate": 3.47681056210089e-05, + "loss": 4.0027, + "step": 930304 + }, + { + "epoch": 0.0, + "learning_rate": 3.475971967349838e-05, + "loss": 4.0055, + "step": 930816 + }, + { + "epoch": 0.01, + "learning_rate": 3.475133372598786e-05, + "loss": 3.9904, + "step": 931328 + }, + { + "epoch": 0.01, + "learning_rate": 3.474294777847734e-05, + "loss": 3.9988, + "step": 931840 + }, + { + "epoch": 0.01, + "learning_rate": 3.473456183096682e-05, + "loss": 4.0027, + "step": 932352 + }, + { + "epoch": 0.01, + "learning_rate": 3.4726192262260026e-05, + "loss": 4.0125, + "step": 932864 + }, + { + "epoch": 0.01, + "learning_rate": 3.4717806314749506e-05, + "loss": 3.9968, + "step": 933376 + }, + { + "epoch": 0.01, + "learning_rate": 3.4709420367238986e-05, + "loss": 4.0006, + "step": 933888 + }, + { + "epoch": 0.01, + "learning_rate": 3.4701050798532195e-05, + "loss": 4.0085, + "step": 934400 + }, + { + "epoch": 0.01, + "learning_rate": 3.4692664851021675e-05, + "loss": 4.0021, + "step": 934912 + }, + { + "epoch": 0.01, + "learning_rate": 3.4684278903511155e-05, + "loss": 4.0149, + "step": 935424 + }, + { + "epoch": 0.01, + "learning_rate": 3.467589295600064e-05, + "loss": 3.9906, + "step": 935936 + }, + { + "epoch": 0.01, + "learning_rate": 3.466750700849012e-05, + "loss": 3.9987, + "step": 936448 + }, + { + "epoch": 0.01, + "learning_rate": 3.46591210609796e-05, + "loss": 3.9917, + "step": 936960 + }, + { + "epoch": 0.01, + "learning_rate": 3.465075149227281e-05, + "loss": 3.9991, + "step": 937472 + }, + { + "epoch": 0.01, + "learning_rate": 3.464236554476229e-05, + "loss": 4.0044, + "step": 937984 + }, + { + "epoch": 0.01, + "learning_rate": 3.463397959725177e-05, + "loss": 3.9941, + "step": 938496 + }, + { + "epoch": 0.01, + "learning_rate": 3.462559364974125e-05, + "loss": 4.0059, + "step": 939008 + }, + { + "epoch": 0.01, + "learning_rate": 3.461720770223073e-05, + "loss": 3.997, + "step": 939520 + }, + { + "epoch": 0.01, + "learning_rate": 3.460882175472021e-05, + "loss": 4.0118, + "step": 940032 + }, + { + "epoch": 0.01, + "learning_rate": 3.460043580720969e-05, + "loss": 4.0021, + "step": 940544 + }, + { + "epoch": 0.01, + "learning_rate": 3.459204985969917e-05, + "loss": 3.9777, + "step": 941056 + }, + { + "epoch": 0.01, + "learning_rate": 3.4583663912188644e-05, + "loss": 3.9943, + "step": 941568 + }, + { + "epoch": 0.01, + "learning_rate": 3.4575310722285595e-05, + "loss": 3.9904, + "step": 942080 + }, + { + "epoch": 0.01, + "learning_rate": 3.4566924774775075e-05, + "loss": 4.0008, + "step": 942592 + }, + { + "epoch": 0.01, + "learning_rate": 3.4558538827264555e-05, + "loss": 3.9919, + "step": 943104 + }, + { + "epoch": 0.01, + "learning_rate": 3.4550152879754035e-05, + "loss": 3.9979, + "step": 943616 + }, + { + "epoch": 0.01, + "learning_rate": 3.4541766932243515e-05, + "loss": 3.9781, + "step": 944128 + }, + { + "epoch": 0.01, + "learning_rate": 3.4533380984732995e-05, + "loss": 3.9962, + "step": 944640 + }, + { + "epoch": 0.01, + "learning_rate": 3.452499503722247e-05, + "loss": 3.99, + "step": 945152 + }, + { + "epoch": 0.01, + "learning_rate": 3.451660908971195e-05, + "loss": 4.0012, + "step": 945664 + }, + { + "epoch": 0.01, + "learning_rate": 3.4508239521005164e-05, + "loss": 3.9831, + "step": 946176 + }, + { + "epoch": 0.01, + "learning_rate": 3.449986995229837e-05, + "loss": 3.9918, + "step": 946688 + }, + { + "epoch": 0.01, + "learning_rate": 3.449148400478785e-05, + "loss": 3.9932, + "step": 947200 + }, + { + "epoch": 0.01, + "learning_rate": 3.448309805727733e-05, + "loss": 3.9858, + "step": 947712 + }, + { + "epoch": 0.01, + "learning_rate": 3.447471210976681e-05, + "loss": 3.9938, + "step": 948224 + }, + { + "epoch": 0.01, + "learning_rate": 3.446632616225629e-05, + "loss": 3.9963, + "step": 948736 + }, + { + "epoch": 0.01, + "learning_rate": 3.445795659354951e-05, + "loss": 4.0007, + "step": 949248 + }, + { + "epoch": 0.01, + "learning_rate": 3.444957064603899e-05, + "loss": 3.9866, + "step": 949760 + }, + { + "epoch": 0.01, + "learning_rate": 3.444118469852847e-05, + "loss": 3.9926, + "step": 950272 + }, + { + "epoch": 0.01, + "learning_rate": 3.443279875101794e-05, + "loss": 3.9888, + "step": 950784 + }, + { + "epoch": 0.01, + "learning_rate": 3.442441280350742e-05, + "loss": 3.9905, + "step": 951296 + }, + { + "epoch": 0.01, + "learning_rate": 3.44160268559969e-05, + "loss": 3.9899, + "step": 951808 + }, + { + "epoch": 0.01, + "learning_rate": 3.440764090848638e-05, + "loss": 4.0, + "step": 952320 + }, + { + "epoch": 0.01, + "learning_rate": 3.439925496097586e-05, + "loss": 3.9907, + "step": 952832 + }, + { + "epoch": 0.01, + "learning_rate": 3.439088539226907e-05, + "loss": 4.0128, + "step": 953344 + }, + { + "epoch": 0.01, + "learning_rate": 3.438249944475855e-05, + "loss": 4.0056, + "step": 953856 + }, + { + "epoch": 0.01, + "learning_rate": 3.437411349724803e-05, + "loss": 3.9976, + "step": 954368 + }, + { + "epoch": 0.01, + "learning_rate": 3.436572754973751e-05, + "loss": 3.9959, + "step": 954880 + }, + { + "epoch": 0.01, + "learning_rate": 3.435735798103073e-05, + "loss": 3.9895, + "step": 955392 + }, + { + "epoch": 0.01, + "learning_rate": 3.434897203352021e-05, + "loss": 3.9849, + "step": 955904 + }, + { + "epoch": 0.01, + "learning_rate": 3.434058608600969e-05, + "loss": 4.0019, + "step": 956416 + }, + { + "epoch": 0.01, + "learning_rate": 3.4332216517302896e-05, + "loss": 4.005, + "step": 956928 + }, + { + "epoch": 0.01, + "learning_rate": 3.4323830569792376e-05, + "loss": 4.0001, + "step": 957440 + }, + { + "epoch": 0.01, + "learning_rate": 3.4315444622281856e-05, + "loss": 3.9855, + "step": 957952 + }, + { + "epoch": 0.01, + "learning_rate": 3.4307058674771336e-05, + "loss": 3.9927, + "step": 958464 + }, + { + "epoch": 0.01, + "learning_rate": 3.4298672727260816e-05, + "loss": 3.9815, + "step": 958976 + }, + { + "epoch": 0.01, + "learning_rate": 3.4290286779750296e-05, + "loss": 3.9964, + "step": 959488 + }, + { + "epoch": 0.01, + "learning_rate": 3.4281917211043505e-05, + "loss": 3.9859, + "step": 960000 + }, + { + "epoch": 0.01, + "learning_rate": 3.4273531263532985e-05, + "loss": 3.9933, + "step": 960512 + }, + { + "epoch": 0.01, + "learning_rate": 3.4265145316022465e-05, + "loss": 4.0024, + "step": 961024 + }, + { + "epoch": 0.01, + "learning_rate": 3.425675936851195e-05, + "loss": 4.0008, + "step": 961536 + }, + { + "epoch": 0.02, + "learning_rate": 3.424837342100143e-05, + "loss": 3.9836, + "step": 962048 + }, + { + "epoch": 0.02, + "learning_rate": 3.423998747349091e-05, + "loss": 3.992, + "step": 962560 + }, + { + "epoch": 0.02, + "learning_rate": 3.423160152598039e-05, + "loss": 3.9995, + "step": 963072 + }, + { + "epoch": 0.02, + "learning_rate": 3.422321557846987e-05, + "loss": 3.9841, + "step": 963584 + }, + { + "epoch": 0.02, + "learning_rate": 3.421482963095935e-05, + "loss": 3.9822, + "step": 964096 + }, + { + "epoch": 0.02, + "learning_rate": 3.420646006225256e-05, + "loss": 3.9949, + "step": 964608 + }, + { + "epoch": 0.02, + "learning_rate": 3.419807411474204e-05, + "loss": 3.9999, + "step": 965120 + }, + { + "epoch": 0.02, + "learning_rate": 3.418968816723152e-05, + "loss": 4.0044, + "step": 965632 + }, + { + "epoch": 0.02, + "learning_rate": 3.4181302219721e-05, + "loss": 3.9955, + "step": 966144 + }, + { + "epoch": 0.02, + "learning_rate": 3.417293265101421e-05, + "loss": 4.0001, + "step": 966656 + }, + { + "epoch": 0.02, + "learning_rate": 3.416454670350369e-05, + "loss": 3.9839, + "step": 967168 + }, + { + "epoch": 0.02, + "learning_rate": 3.415616075599317e-05, + "loss": 3.9962, + "step": 967680 + }, + { + "epoch": 0.02, + "learning_rate": 3.414777480848265e-05, + "loss": 3.996, + "step": 968192 + }, + { + "epoch": 0.02, + "learning_rate": 3.4139405239775865e-05, + "loss": 4.0068, + "step": 968704 + }, + { + "epoch": 0.02, + "learning_rate": 3.4131035671069074e-05, + "loss": 3.992, + "step": 969216 + }, + { + "epoch": 0.02, + "learning_rate": 3.4122649723558554e-05, + "loss": 3.9869, + "step": 969728 + }, + { + "epoch": 0.02, + "learning_rate": 3.4114263776048034e-05, + "loss": 3.9787, + "step": 970240 + }, + { + "epoch": 0.02, + "learning_rate": 3.4105877828537514e-05, + "loss": 3.9785, + "step": 970752 + }, + { + "epoch": 0.02, + "learning_rate": 3.4097491881026994e-05, + "loss": 3.9853, + "step": 971264 + }, + { + "epoch": 0.02, + "learning_rate": 3.4089105933516474e-05, + "loss": 3.9993, + "step": 971776 + }, + { + "epoch": 0.02, + "learning_rate": 3.408071998600595e-05, + "loss": 3.9937, + "step": 972288 + }, + { + "epoch": 0.02, + "learning_rate": 3.407233403849543e-05, + "loss": 3.9784, + "step": 972800 + }, + { + "epoch": 0.02, + "learning_rate": 3.406394809098491e-05, + "loss": 3.9864, + "step": 973312 + }, + { + "epoch": 0.02, + "learning_rate": 3.405557852227812e-05, + "loss": 3.9983, + "step": 973824 + }, + { + "epoch": 0.02, + "learning_rate": 3.40471925747676e-05, + "loss": 3.9996, + "step": 974336 + }, + { + "epoch": 0.02, + "learning_rate": 3.403880662725708e-05, + "loss": 3.9988, + "step": 974848 + }, + { + "epoch": 0.02, + "learning_rate": 3.403042067974656e-05, + "loss": 3.9924, + "step": 975360 + }, + { + "epoch": 0.02, + "learning_rate": 3.402205111103978e-05, + "loss": 3.9841, + "step": 975872 + }, + { + "epoch": 0.02, + "learning_rate": 3.401366516352925e-05, + "loss": 3.9921, + "step": 976384 + }, + { + "epoch": 0.02, + "learning_rate": 3.400527921601873e-05, + "loss": 3.9989, + "step": 976896 + }, + { + "epoch": 0.02, + "learning_rate": 3.399690964731195e-05, + "loss": 3.9907, + "step": 977408 + }, + { + "epoch": 0.02, + "learning_rate": 3.398852369980142e-05, + "loss": 3.9783, + "step": 977920 + }, + { + "epoch": 0.02, + "learning_rate": 3.39801377522909e-05, + "loss": 3.9884, + "step": 978432 + }, + { + "epoch": 0.02, + "learning_rate": 3.397175180478038e-05, + "loss": 3.9858, + "step": 978944 + }, + { + "epoch": 0.02, + "learning_rate": 3.3963382236073596e-05, + "loss": 3.9954, + "step": 979456 + }, + { + "epoch": 0.02, + "learning_rate": 3.3954996288563076e-05, + "loss": 3.9884, + "step": 979968 + }, + { + "epoch": 0.02, + "learning_rate": 3.3946610341052556e-05, + "loss": 3.9982, + "step": 980480 + }, + { + "epoch": 0.02, + "learning_rate": 3.3938224393542036e-05, + "loss": 3.9813, + "step": 980992 + }, + { + "epoch": 0.02, + "learning_rate": 3.3929838446031516e-05, + "loss": 4.0015, + "step": 981504 + }, + { + "epoch": 0.02, + "learning_rate": 3.3921452498520996e-05, + "loss": 3.9866, + "step": 982016 + }, + { + "epoch": 0.02, + "learning_rate": 3.3913066551010476e-05, + "loss": 3.9986, + "step": 982528 + }, + { + "epoch": 0.02, + "learning_rate": 3.3904680603499956e-05, + "loss": 4.0009, + "step": 983040 + }, + { + "epoch": 0.02, + "learning_rate": 3.3896294655989436e-05, + "loss": 4.0009, + "step": 983552 + }, + { + "epoch": 0.02, + "learning_rate": 3.3887908708478916e-05, + "loss": 3.9946, + "step": 984064 + }, + { + "epoch": 0.02, + "learning_rate": 3.3879522760968396e-05, + "loss": 3.9997, + "step": 984576 + }, + { + "epoch": 0.02, + "learning_rate": 3.3871136813457876e-05, + "loss": 3.9802, + "step": 985088 + }, + { + "epoch": 0.02, + "learning_rate": 3.3862767244751085e-05, + "loss": 3.9829, + "step": 985600 + }, + { + "epoch": 0.02, + "learning_rate": 3.3854381297240565e-05, + "loss": 3.9853, + "step": 986112 + }, + { + "epoch": 0.02, + "learning_rate": 3.3846011728533774e-05, + "loss": 3.9966, + "step": 986624 + }, + { + "epoch": 0.02, + "learning_rate": 3.383762578102326e-05, + "loss": 4.003, + "step": 987136 + }, + { + "epoch": 0.02, + "learning_rate": 3.382923983351274e-05, + "loss": 3.9978, + "step": 987648 + }, + { + "epoch": 0.02, + "learning_rate": 3.382085388600222e-05, + "loss": 3.9968, + "step": 988160 + }, + { + "epoch": 0.02, + "learning_rate": 3.38124679384917e-05, + "loss": 3.9831, + "step": 988672 + }, + { + "epoch": 0.02, + "learning_rate": 3.380408199098118e-05, + "loss": 3.9775, + "step": 989184 + }, + { + "epoch": 0.02, + "learning_rate": 3.379569604347066e-05, + "loss": 3.9921, + "step": 989696 + }, + { + "epoch": 0.02, + "learning_rate": 3.378731009596014e-05, + "loss": 3.997, + "step": 990208 + }, + { + "epoch": 0.02, + "learning_rate": 3.3778924148449614e-05, + "loss": 4.005, + "step": 990720 + }, + { + "epoch": 0.02, + "learning_rate": 3.377055457974283e-05, + "loss": 3.9834, + "step": 991232 + }, + { + "epoch": 0.02, + "learning_rate": 3.376216863223231e-05, + "loss": 3.9838, + "step": 991744 + }, + { + "epoch": 0.03, + "eval_loss": 4.03767728805542, + "eval_runtime": 540.9831, + "eval_samples_per_second": 705.366, + "eval_steps_per_second": 22.043, + "step": 992147 + }, + { + "epoch": 0.0, + "learning_rate": 3.375379906352552e-05, + "loss": 3.9709, + "step": 992256 + }, + { + "epoch": 0.0, + "learning_rate": 3.3745413116015e-05, + "loss": 4.0075, + "step": 992768 + }, + { + "epoch": 0.0, + "learning_rate": 3.373702716850448e-05, + "loss": 3.9834, + "step": 993280 + }, + { + "epoch": 0.0, + "learning_rate": 3.372864122099396e-05, + "loss": 3.9716, + "step": 993792 + }, + { + "epoch": 0.0, + "learning_rate": 3.372025527348344e-05, + "loss": 3.9949, + "step": 994304 + }, + { + "epoch": 0.0, + "learning_rate": 3.371186932597292e-05, + "loss": 3.9867, + "step": 994816 + }, + { + "epoch": 0.0, + "learning_rate": 3.37034833784624e-05, + "loss": 3.9916, + "step": 995328 + }, + { + "epoch": 0.0, + "learning_rate": 3.369509743095188e-05, + "loss": 3.9813, + "step": 995840 + }, + { + "epoch": 0.0, + "learning_rate": 3.368671148344136e-05, + "loss": 3.9862, + "step": 996352 + }, + { + "epoch": 0.0, + "learning_rate": 3.367832553593084e-05, + "loss": 3.9707, + "step": 996864 + }, + { + "epoch": 0.0, + "learning_rate": 3.366993958842032e-05, + "loss": 4.0025, + "step": 997376 + }, + { + "epoch": 0.0, + "learning_rate": 3.36615536409098e-05, + "loss": 3.9834, + "step": 997888 + }, + { + "epoch": 0.0, + "learning_rate": 3.365316769339928e-05, + "loss": 3.9916, + "step": 998400 + }, + { + "epoch": 0.0, + "learning_rate": 3.364478174588876e-05, + "loss": 3.9972, + "step": 998912 + }, + { + "epoch": 0.0, + "learning_rate": 3.363639579837824e-05, + "loss": 3.9873, + "step": 999424 + }, + { + "epoch": 0.0, + "learning_rate": 3.362800985086772e-05, + "loss": 3.9775, + "step": 999936 + }, + { + "epoch": 0.0, + "learning_rate": 3.36196239033572e-05, + "loss": 3.9974, + "step": 1000448 + }, + { + "epoch": 0.0, + "learning_rate": 3.3611254334650414e-05, + "loss": 3.9894, + "step": 1000960 + }, + { + "epoch": 0.0, + "learning_rate": 3.3602868387139894e-05, + "loss": 3.9817, + "step": 1001472 + }, + { + "epoch": 0.0, + "learning_rate": 3.3594482439629374e-05, + "loss": 4.0012, + "step": 1001984 + }, + { + "epoch": 0.0, + "learning_rate": 3.3586096492118854e-05, + "loss": 4.0094, + "step": 1002496 + }, + { + "epoch": 0.0, + "learning_rate": 3.357771054460833e-05, + "loss": 3.991, + "step": 1003008 + }, + { + "epoch": 0.0, + "learning_rate": 3.356934097590154e-05, + "loss": 3.9815, + "step": 1003520 + }, + { + "epoch": 0.0, + "learning_rate": 3.356095502839102e-05, + "loss": 3.9918, + "step": 1004032 + }, + { + "epoch": 0.0, + "learning_rate": 3.35525690808805e-05, + "loss": 3.9836, + "step": 1004544 + }, + { + "epoch": 0.0, + "learning_rate": 3.3544183133369976e-05, + "loss": 3.9827, + "step": 1005056 + }, + { + "epoch": 0.0, + "learning_rate": 3.353581356466319e-05, + "loss": 3.9881, + "step": 1005568 + }, + { + "epoch": 0.0, + "learning_rate": 3.352742761715267e-05, + "loss": 3.9813, + "step": 1006080 + }, + { + "epoch": 0.0, + "learning_rate": 3.351904166964215e-05, + "loss": 3.9861, + "step": 1006592 + }, + { + "epoch": 0.0, + "learning_rate": 3.351065572213163e-05, + "loss": 3.9933, + "step": 1007104 + }, + { + "epoch": 0.01, + "learning_rate": 3.350226977462111e-05, + "loss": 3.9933, + "step": 1007616 + }, + { + "epoch": 0.01, + "learning_rate": 3.349388382711059e-05, + "loss": 3.9876, + "step": 1008128 + }, + { + "epoch": 0.01, + "learning_rate": 3.34855142584038e-05, + "loss": 3.9884, + "step": 1008640 + }, + { + "epoch": 0.01, + "learning_rate": 3.347712831089328e-05, + "loss": 3.9996, + "step": 1009152 + }, + { + "epoch": 0.01, + "learning_rate": 3.346874236338276e-05, + "loss": 3.9786, + "step": 1009664 + }, + { + "epoch": 0.01, + "learning_rate": 3.346035641587224e-05, + "loss": 3.9876, + "step": 1010176 + }, + { + "epoch": 0.01, + "learning_rate": 3.345197046836172e-05, + "loss": 3.9829, + "step": 1010688 + }, + { + "epoch": 0.01, + "learning_rate": 3.34435845208512e-05, + "loss": 3.9976, + "step": 1011200 + }, + { + "epoch": 0.01, + "learning_rate": 3.343519857334068e-05, + "loss": 3.9975, + "step": 1011712 + }, + { + "epoch": 0.01, + "learning_rate": 3.342681262583016e-05, + "loss": 3.9996, + "step": 1012224 + }, + { + "epoch": 0.01, + "learning_rate": 3.341842667831964e-05, + "loss": 3.993, + "step": 1012736 + }, + { + "epoch": 0.01, + "learning_rate": 3.341004073080912e-05, + "loss": 3.9925, + "step": 1013248 + }, + { + "epoch": 0.01, + "learning_rate": 3.34016547832986e-05, + "loss": 3.9815, + "step": 1013760 + }, + { + "epoch": 0.01, + "learning_rate": 3.339326883578808e-05, + "loss": 3.9799, + "step": 1014272 + }, + { + "epoch": 0.01, + "learning_rate": 3.3384899267081296e-05, + "loss": 3.9743, + "step": 1014784 + }, + { + "epoch": 0.01, + "learning_rate": 3.3376513319570776e-05, + "loss": 3.9721, + "step": 1015296 + }, + { + "epoch": 0.01, + "learning_rate": 3.3368127372060256e-05, + "loss": 3.9801, + "step": 1015808 + }, + { + "epoch": 0.01, + "learning_rate": 3.3359741424549736e-05, + "loss": 3.9844, + "step": 1016320 + }, + { + "epoch": 0.01, + "learning_rate": 3.3351371855842945e-05, + "loss": 3.9841, + "step": 1016832 + }, + { + "epoch": 0.01, + "learning_rate": 3.3343002287136154e-05, + "loss": 3.9855, + "step": 1017344 + }, + { + "epoch": 0.01, + "learning_rate": 3.3334616339625634e-05, + "loss": 3.9717, + "step": 1017856 + }, + { + "epoch": 0.01, + "learning_rate": 3.3326230392115114e-05, + "loss": 3.9846, + "step": 1018368 + }, + { + "epoch": 0.01, + "learning_rate": 3.3317844444604594e-05, + "loss": 3.9857, + "step": 1018880 + }, + { + "epoch": 0.01, + "learning_rate": 3.3309458497094074e-05, + "loss": 3.9688, + "step": 1019392 + }, + { + "epoch": 0.01, + "learning_rate": 3.3301072549583554e-05, + "loss": 3.9726, + "step": 1019904 + }, + { + "epoch": 0.01, + "learning_rate": 3.3292686602073034e-05, + "loss": 3.9812, + "step": 1020416 + }, + { + "epoch": 0.01, + "learning_rate": 3.3284300654562514e-05, + "loss": 3.9948, + "step": 1020928 + }, + { + "epoch": 0.01, + "learning_rate": 3.3275914707051994e-05, + "loss": 3.9659, + "step": 1021440 + }, + { + "epoch": 0.01, + "learning_rate": 3.326754513834521e-05, + "loss": 3.9863, + "step": 1021952 + }, + { + "epoch": 0.01, + "learning_rate": 3.325915919083469e-05, + "loss": 3.9878, + "step": 1022464 + }, + { + "epoch": 0.01, + "learning_rate": 3.325077324332416e-05, + "loss": 3.9739, + "step": 1022976 + }, + { + "epoch": 0.01, + "learning_rate": 3.324240367461738e-05, + "loss": 3.9718, + "step": 1023488 + }, + { + "epoch": 0.01, + "learning_rate": 3.323401772710686e-05, + "loss": 3.9678, + "step": 1024000 + }, + { + "epoch": 0.01, + "learning_rate": 3.322563177959633e-05, + "loss": 3.9712, + "step": 1024512 + }, + { + "epoch": 0.01, + "learning_rate": 3.321724583208581e-05, + "loss": 3.9926, + "step": 1025024 + }, + { + "epoch": 0.01, + "learning_rate": 3.320885988457529e-05, + "loss": 3.994, + "step": 1025536 + }, + { + "epoch": 0.01, + "learning_rate": 3.320047393706477e-05, + "loss": 3.9766, + "step": 1026048 + }, + { + "epoch": 0.01, + "learning_rate": 3.319208798955425e-05, + "loss": 3.983, + "step": 1026560 + }, + { + "epoch": 0.01, + "learning_rate": 3.318370204204374e-05, + "loss": 3.9886, + "step": 1027072 + }, + { + "epoch": 0.01, + "learning_rate": 3.317533247333695e-05, + "loss": 3.976, + "step": 1027584 + }, + { + "epoch": 0.01, + "learning_rate": 3.316694652582643e-05, + "loss": 3.9898, + "step": 1028096 + }, + { + "epoch": 0.01, + "learning_rate": 3.315856057831591e-05, + "loss": 3.9799, + "step": 1028608 + }, + { + "epoch": 0.01, + "learning_rate": 3.315017463080539e-05, + "loss": 3.9805, + "step": 1029120 + }, + { + "epoch": 0.01, + "learning_rate": 3.314178868329487e-05, + "loss": 3.9757, + "step": 1029632 + }, + { + "epoch": 0.01, + "learning_rate": 3.3133419114588076e-05, + "loss": 3.9926, + "step": 1030144 + }, + { + "epoch": 0.01, + "learning_rate": 3.3125049545881285e-05, + "loss": 3.9801, + "step": 1030656 + }, + { + "epoch": 0.01, + "learning_rate": 3.3116663598370765e-05, + "loss": 3.99, + "step": 1031168 + }, + { + "epoch": 0.01, + "learning_rate": 3.3108277650860245e-05, + "loss": 3.9852, + "step": 1031680 + }, + { + "epoch": 0.01, + "learning_rate": 3.3099891703349725e-05, + "loss": 3.984, + "step": 1032192 + }, + { + "epoch": 0.01, + "learning_rate": 3.3091505755839205e-05, + "loss": 3.9782, + "step": 1032704 + }, + { + "epoch": 0.01, + "learning_rate": 3.308311980832869e-05, + "loss": 3.984, + "step": 1033216 + }, + { + "epoch": 0.01, + "learning_rate": 3.307473386081817e-05, + "loss": 3.9751, + "step": 1033728 + }, + { + "epoch": 0.01, + "learning_rate": 3.306636429211138e-05, + "loss": 3.9748, + "step": 1034240 + }, + { + "epoch": 0.01, + "learning_rate": 3.305797834460086e-05, + "loss": 3.9778, + "step": 1034752 + }, + { + "epoch": 0.01, + "learning_rate": 3.304959239709034e-05, + "loss": 3.9703, + "step": 1035264 + }, + { + "epoch": 0.01, + "learning_rate": 3.304120644957982e-05, + "loss": 3.9846, + "step": 1035776 + }, + { + "epoch": 0.01, + "learning_rate": 3.30328205020693e-05, + "loss": 3.9711, + "step": 1036288 + }, + { + "epoch": 0.01, + "learning_rate": 3.302443455455878e-05, + "loss": 3.9805, + "step": 1036800 + }, + { + "epoch": 0.01, + "learning_rate": 3.301604860704826e-05, + "loss": 3.9672, + "step": 1037312 + }, + { + "epoch": 0.01, + "learning_rate": 3.300766265953774e-05, + "loss": 3.9872, + "step": 1037824 + }, + { + "epoch": 0.02, + "learning_rate": 3.299927671202722e-05, + "loss": 3.9873, + "step": 1038336 + }, + { + "epoch": 0.02, + "learning_rate": 3.299090714332043e-05, + "loss": 3.9889, + "step": 1038848 + }, + { + "epoch": 0.02, + "learning_rate": 3.298252119580991e-05, + "loss": 3.9852, + "step": 1039360 + }, + { + "epoch": 0.02, + "learning_rate": 3.297413524829939e-05, + "loss": 3.973, + "step": 1039872 + }, + { + "epoch": 0.02, + "learning_rate": 3.296574930078887e-05, + "loss": 3.9834, + "step": 1040384 + }, + { + "epoch": 0.02, + "learning_rate": 3.295736335327835e-05, + "loss": 3.9783, + "step": 1040896 + }, + { + "epoch": 0.02, + "learning_rate": 3.2948993784571565e-05, + "loss": 3.9774, + "step": 1041408 + }, + { + "epoch": 0.02, + "learning_rate": 3.2940624215864775e-05, + "loss": 3.9683, + "step": 1041920 + }, + { + "epoch": 0.02, + "learning_rate": 3.2932238268354254e-05, + "loss": 3.973, + "step": 1042432 + }, + { + "epoch": 0.02, + "learning_rate": 3.2923852320843734e-05, + "loss": 3.9888, + "step": 1042944 + }, + { + "epoch": 0.02, + "learning_rate": 3.2915466373333214e-05, + "loss": 3.9801, + "step": 1043456 + }, + { + "epoch": 0.02, + "learning_rate": 3.2907080425822694e-05, + "loss": 3.9879, + "step": 1043968 + }, + { + "epoch": 0.02, + "learning_rate": 3.289869447831217e-05, + "loss": 3.9793, + "step": 1044480 + }, + { + "epoch": 0.02, + "learning_rate": 3.289030853080165e-05, + "loss": 3.9793, + "step": 1044992 + }, + { + "epoch": 0.02, + "learning_rate": 3.288192258329113e-05, + "loss": 3.9695, + "step": 1045504 + }, + { + "epoch": 0.02, + "learning_rate": 3.287355301458434e-05, + "loss": 3.9731, + "step": 1046016 + }, + { + "epoch": 0.02, + "learning_rate": 3.286516706707382e-05, + "loss": 3.9854, + "step": 1046528 + }, + { + "epoch": 0.02, + "learning_rate": 3.28567811195633e-05, + "loss": 3.9848, + "step": 1047040 + }, + { + "epoch": 0.02, + "learning_rate": 3.284839517205278e-05, + "loss": 3.973, + "step": 1047552 + }, + { + "epoch": 0.02, + "learning_rate": 3.284000922454226e-05, + "loss": 3.966, + "step": 1048064 + }, + { + "epoch": 0.02, + "learning_rate": 3.283162327703174e-05, + "loss": 3.98, + "step": 1048576 + }, + { + "epoch": 0.02, + "learning_rate": 3.282323732952122e-05, + "loss": 3.984, + "step": 1049088 + }, + { + "epoch": 0.02, + "learning_rate": 3.28148513820107e-05, + "loss": 3.9662, + "step": 1049600 + }, + { + "epoch": 0.02, + "learning_rate": 3.280648181330391e-05, + "loss": 3.9796, + "step": 1050112 + }, + { + "epoch": 0.02, + "learning_rate": 3.279811224459712e-05, + "loss": 3.9718, + "step": 1050624 + }, + { + "epoch": 0.02, + "learning_rate": 3.27897262970866e-05, + "loss": 3.9734, + "step": 1051136 + }, + { + "epoch": 0.02, + "learning_rate": 3.278134034957608e-05, + "loss": 3.9681, + "step": 1051648 + }, + { + "epoch": 0.02, + "learning_rate": 3.277295440206556e-05, + "loss": 3.9799, + "step": 1052160 + }, + { + "epoch": 0.02, + "learning_rate": 3.276456845455505e-05, + "loss": 3.9786, + "step": 1052672 + }, + { + "epoch": 0.02, + "learning_rate": 3.275618250704453e-05, + "loss": 3.9757, + "step": 1053184 + }, + { + "epoch": 0.02, + "learning_rate": 3.274779655953401e-05, + "loss": 3.9865, + "step": 1053696 + }, + { + "epoch": 0.02, + "learning_rate": 3.273941061202349e-05, + "loss": 3.9937, + "step": 1054208 + }, + { + "epoch": 0.02, + "learning_rate": 3.27310410433167e-05, + "loss": 3.9775, + "step": 1054720 + }, + { + "epoch": 0.02, + "learning_rate": 3.272265509580618e-05, + "loss": 3.9811, + "step": 1055232 + }, + { + "epoch": 0.02, + "learning_rate": 3.2714285527099386e-05, + "loss": 3.9783, + "step": 1055744 + }, + { + "epoch": 0.02, + "learning_rate": 3.2705899579588866e-05, + "loss": 3.981, + "step": 1056256 + }, + { + "epoch": 0.02, + "learning_rate": 3.2697513632078346e-05, + "loss": 3.9788, + "step": 1056768 + }, + { + "epoch": 0.02, + "learning_rate": 3.2689127684567826e-05, + "loss": 3.9737, + "step": 1057280 + }, + { + "epoch": 0.02, + "learning_rate": 3.2680741737057306e-05, + "loss": 3.9782, + "step": 1057792 + }, + { + "epoch": 0.02, + "learning_rate": 3.2672355789546786e-05, + "loss": 3.9753, + "step": 1058304 + }, + { + "epoch": 0.02, + "learning_rate": 3.266398622084e-05, + "loss": 3.9914, + "step": 1058816 + }, + { + "epoch": 0.02, + "learning_rate": 3.265560027332948e-05, + "loss": 3.9759, + "step": 1059328 + }, + { + "epoch": 0.02, + "learning_rate": 3.264721432581896e-05, + "loss": 3.9909, + "step": 1059840 + }, + { + "epoch": 0.02, + "learning_rate": 3.263882837830844e-05, + "loss": 3.9764, + "step": 1060352 + }, + { + "epoch": 0.02, + "learning_rate": 3.263044243079792e-05, + "loss": 3.9846, + "step": 1060864 + }, + { + "epoch": 0.02, + "learning_rate": 3.26220564832874e-05, + "loss": 3.9781, + "step": 1061376 + }, + { + "epoch": 0.02, + "learning_rate": 3.261367053577688e-05, + "loss": 3.9829, + "step": 1061888 + }, + { + "epoch": 0.02, + "learning_rate": 3.260530096707009e-05, + "loss": 3.9893, + "step": 1062400 + }, + { + "epoch": 0.02, + "learning_rate": 3.259691501955957e-05, + "loss": 3.9745, + "step": 1062912 + }, + { + "epoch": 0.02, + "learning_rate": 3.258852907204905e-05, + "loss": 3.9807, + "step": 1063424 + }, + { + "epoch": 0.02, + "learning_rate": 3.258014312453853e-05, + "loss": 3.9843, + "step": 1063936 + }, + { + "epoch": 0.02, + "learning_rate": 3.2571757177028e-05, + "loss": 3.9735, + "step": 1064448 + }, + { + "epoch": 0.02, + "learning_rate": 3.256337122951748e-05, + "loss": 3.9757, + "step": 1064960 + }, + { + "epoch": 0.02, + "learning_rate": 3.255498528200697e-05, + "loss": 3.9836, + "step": 1065472 + }, + { + "epoch": 0.02, + "learning_rate": 3.254659933449645e-05, + "loss": 3.9799, + "step": 1065984 + }, + { + "epoch": 0.02, + "learning_rate": 3.253822976578966e-05, + "loss": 3.9809, + "step": 1066496 + }, + { + "epoch": 0.02, + "learning_rate": 3.252984381827914e-05, + "loss": 3.9799, + "step": 1067008 + }, + { + "epoch": 0.02, + "learning_rate": 3.252145787076862e-05, + "loss": 3.9752, + "step": 1067520 + }, + { + "epoch": 0.02, + "learning_rate": 3.25130719232581e-05, + "loss": 3.9707, + "step": 1068032 + }, + { + "epoch": 0.03, + "eval_loss": 4.028290271759033, + "eval_runtime": 293.3806, + "eval_samples_per_second": 1300.669, + "eval_steps_per_second": 40.647, + "step": 1068467 + }, + { + "epoch": 1.0, + "learning_rate": 3.250468597574758e-05, + "loss": 3.9771, + "step": 1068544 + }, + { + "epoch": 1.0, + "learning_rate": 3.249631640704079e-05, + "loss": 3.9926, + "step": 1069056 + }, + { + "epoch": 1.0, + "learning_rate": 3.2487946838334004e-05, + "loss": 3.9735, + "step": 1069568 + }, + { + "epoch": 1.0, + "learning_rate": 3.247956089082348e-05, + "loss": 3.9582, + "step": 1070080 + }, + { + "epoch": 1.0, + "learning_rate": 3.247117494331296e-05, + "loss": 3.9766, + "step": 1070592 + }, + { + "epoch": 1.0, + "learning_rate": 3.246278899580244e-05, + "loss": 3.9758, + "step": 1071104 + }, + { + "epoch": 1.0, + "learning_rate": 3.2454403048291924e-05, + "loss": 3.9773, + "step": 1071616 + }, + { + "epoch": 1.0, + "learning_rate": 3.2446017100781404e-05, + "loss": 3.9681, + "step": 1072128 + }, + { + "epoch": 1.0, + "learning_rate": 3.2437631153270884e-05, + "loss": 3.9681, + "step": 1072640 + }, + { + "epoch": 1.0, + "learning_rate": 3.2429245205760364e-05, + "loss": 3.964, + "step": 1073152 + }, + { + "epoch": 1.0, + "learning_rate": 3.242087563705357e-05, + "loss": 3.9851, + "step": 1073664 + }, + { + "epoch": 1.0, + "learning_rate": 3.241248968954305e-05, + "loss": 3.9702, + "step": 1074176 + }, + { + "epoch": 1.0, + "learning_rate": 3.240410374203253e-05, + "loss": 3.9795, + "step": 1074688 + }, + { + "epoch": 1.0, + "learning_rate": 3.239571779452201e-05, + "loss": 3.9823, + "step": 1075200 + }, + { + "epoch": 1.0, + "learning_rate": 3.238734822581522e-05, + "loss": 3.978, + "step": 1075712 + }, + { + "epoch": 1.0, + "learning_rate": 3.23789622783047e-05, + "loss": 3.9636, + "step": 1076224 + }, + { + "epoch": 1.0, + "learning_rate": 3.237059270959791e-05, + "loss": 3.9793, + "step": 1076736 + }, + { + "epoch": 1.0, + "learning_rate": 3.236220676208739e-05, + "loss": 3.9765, + "step": 1077248 + }, + { + "epoch": 1.0, + "learning_rate": 3.235382081457688e-05, + "loss": 3.9654, + "step": 1077760 + }, + { + "epoch": 1.0, + "learning_rate": 3.234543486706636e-05, + "loss": 3.9866, + "step": 1078272 + }, + { + "epoch": 1.0, + "learning_rate": 3.233704891955584e-05, + "loss": 3.9971, + "step": 1078784 + }, + { + "epoch": 1.0, + "learning_rate": 3.232866297204532e-05, + "loss": 3.9829, + "step": 1079296 + }, + { + "epoch": 1.0, + "learning_rate": 3.23202770245348e-05, + "loss": 3.9673, + "step": 1079808 + }, + { + "epoch": 1.0, + "learning_rate": 3.231189107702428e-05, + "loss": 3.9741, + "step": 1080320 + }, + { + "epoch": 1.0, + "learning_rate": 3.2303521508317486e-05, + "loss": 3.9703, + "step": 1080832 + }, + { + "epoch": 1.0, + "learning_rate": 3.2295135560806966e-05, + "loss": 3.9714, + "step": 1081344 + }, + { + "epoch": 1.0, + "learning_rate": 3.2286782370903904e-05, + "loss": 3.9767, + "step": 1081856 + }, + { + "epoch": 1.0, + "learning_rate": 3.2278396423393384e-05, + "loss": 3.9663, + "step": 1082368 + }, + { + "epoch": 1.0, + "learning_rate": 3.2270010475882864e-05, + "loss": 3.9692, + "step": 1082880 + }, + { + "epoch": 1.0, + "learning_rate": 3.2261624528372344e-05, + "loss": 3.9824, + "step": 1083392 + }, + { + "epoch": 1.01, + "learning_rate": 3.2253238580861824e-05, + "loss": 3.9764, + "step": 1083904 + }, + { + "epoch": 1.01, + "learning_rate": 3.224485263335131e-05, + "loss": 3.9774, + "step": 1084416 + }, + { + "epoch": 1.01, + "learning_rate": 3.223648306464452e-05, + "loss": 3.9743, + "step": 1084928 + }, + { + "epoch": 1.01, + "learning_rate": 3.2228097117134e-05, + "loss": 3.9839, + "step": 1085440 + }, + { + "epoch": 1.01, + "learning_rate": 3.221971116962348e-05, + "loss": 3.9631, + "step": 1085952 + }, + { + "epoch": 1.01, + "learning_rate": 3.221132522211296e-05, + "loss": 3.9749, + "step": 1086464 + }, + { + "epoch": 1.01, + "learning_rate": 3.220293927460244e-05, + "loss": 3.9665, + "step": 1086976 + }, + { + "epoch": 1.01, + "learning_rate": 3.219455332709192e-05, + "loss": 3.9888, + "step": 1087488 + }, + { + "epoch": 1.01, + "learning_rate": 3.21861673795814e-05, + "loss": 3.9793, + "step": 1088000 + }, + { + "epoch": 1.01, + "learning_rate": 3.217778143207088e-05, + "loss": 3.9883, + "step": 1088512 + }, + { + "epoch": 1.01, + "learning_rate": 3.216939548456036e-05, + "loss": 3.9823, + "step": 1089024 + }, + { + "epoch": 1.01, + "learning_rate": 3.216102591585357e-05, + "loss": 3.9765, + "step": 1089536 + }, + { + "epoch": 1.01, + "learning_rate": 3.215263996834305e-05, + "loss": 3.9701, + "step": 1090048 + }, + { + "epoch": 1.01, + "learning_rate": 3.214425402083253e-05, + "loss": 3.9639, + "step": 1090560 + }, + { + "epoch": 1.01, + "learning_rate": 3.213586807332201e-05, + "loss": 3.9662, + "step": 1091072 + }, + { + "epoch": 1.01, + "learning_rate": 3.212748212581149e-05, + "loss": 3.9595, + "step": 1091584 + }, + { + "epoch": 1.01, + "learning_rate": 3.211909617830097e-05, + "loss": 3.9694, + "step": 1092096 + }, + { + "epoch": 1.01, + "learning_rate": 3.211071023079045e-05, + "loss": 3.9712, + "step": 1092608 + }, + { + "epoch": 1.01, + "learning_rate": 3.210232428327993e-05, + "loss": 3.9704, + "step": 1093120 + }, + { + "epoch": 1.01, + "learning_rate": 3.209393833576941e-05, + "loss": 3.969, + "step": 1093632 + }, + { + "epoch": 1.01, + "learning_rate": 3.208556876706262e-05, + "loss": 3.962, + "step": 1094144 + }, + { + "epoch": 1.01, + "learning_rate": 3.20771828195521e-05, + "loss": 3.9692, + "step": 1094656 + }, + { + "epoch": 1.01, + "learning_rate": 3.206879687204158e-05, + "loss": 3.9773, + "step": 1095168 + }, + { + "epoch": 1.01, + "learning_rate": 3.206041092453106e-05, + "loss": 3.9474, + "step": 1095680 + }, + { + "epoch": 1.01, + "learning_rate": 3.2052041355824267e-05, + "loss": 3.9651, + "step": 1096192 + }, + { + "epoch": 1.01, + "learning_rate": 3.204367178711748e-05, + "loss": 3.9628, + "step": 1096704 + }, + { + "epoch": 1.01, + "learning_rate": 3.203528583960696e-05, + "loss": 3.982, + "step": 1097216 + }, + { + "epoch": 1.01, + "learning_rate": 3.202689989209644e-05, + "loss": 3.9508, + "step": 1097728 + }, + { + "epoch": 1.01, + "learning_rate": 3.201851394458592e-05, + "loss": 3.9757, + "step": 1098240 + }, + { + "epoch": 1.01, + "learning_rate": 3.20101279970754e-05, + "loss": 3.9686, + "step": 1098752 + }, + { + "epoch": 1.01, + "learning_rate": 3.200174204956488e-05, + "loss": 3.9644, + "step": 1099264 + }, + { + "epoch": 1.01, + "learning_rate": 3.199337248085809e-05, + "loss": 3.9634, + "step": 1099776 + }, + { + "epoch": 1.01, + "learning_rate": 3.198498653334757e-05, + "loss": 3.9553, + "step": 1100288 + }, + { + "epoch": 1.01, + "learning_rate": 3.197660058583705e-05, + "loss": 3.9559, + "step": 1100800 + }, + { + "epoch": 1.01, + "learning_rate": 3.196821463832653e-05, + "loss": 3.9795, + "step": 1101312 + }, + { + "epoch": 1.01, + "learning_rate": 3.195982869081601e-05, + "loss": 3.9761, + "step": 1101824 + }, + { + "epoch": 1.01, + "learning_rate": 3.195144274330549e-05, + "loss": 3.9638, + "step": 1102336 + }, + { + "epoch": 1.01, + "learning_rate": 3.194305679579497e-05, + "loss": 3.972, + "step": 1102848 + }, + { + "epoch": 1.01, + "learning_rate": 3.193467084828445e-05, + "loss": 3.9731, + "step": 1103360 + }, + { + "epoch": 1.01, + "learning_rate": 3.192630127957767e-05, + "loss": 3.9665, + "step": 1103872 + }, + { + "epoch": 1.01, + "learning_rate": 3.191791533206715e-05, + "loss": 3.9726, + "step": 1104384 + }, + { + "epoch": 1.01, + "learning_rate": 3.190952938455663e-05, + "loss": 3.9699, + "step": 1104896 + }, + { + "epoch": 1.01, + "learning_rate": 3.190114343704611e-05, + "loss": 3.9634, + "step": 1105408 + }, + { + "epoch": 1.01, + "learning_rate": 3.1892773868339316e-05, + "loss": 3.9648, + "step": 1105920 + }, + { + "epoch": 1.01, + "learning_rate": 3.1884387920828796e-05, + "loss": 3.9785, + "step": 1106432 + }, + { + "epoch": 1.01, + "learning_rate": 3.1876001973318276e-05, + "loss": 3.9695, + "step": 1106944 + }, + { + "epoch": 1.01, + "learning_rate": 3.1867616025807756e-05, + "loss": 3.9789, + "step": 1107456 + }, + { + "epoch": 1.01, + "learning_rate": 3.1859246457100965e-05, + "loss": 3.9741, + "step": 1107968 + }, + { + "epoch": 1.01, + "learning_rate": 3.1850860509590445e-05, + "loss": 3.9683, + "step": 1108480 + }, + { + "epoch": 1.01, + "learning_rate": 3.1842474562079925e-05, + "loss": 3.9651, + "step": 1108992 + }, + { + "epoch": 1.01, + "learning_rate": 3.1834088614569405e-05, + "loss": 3.9722, + "step": 1109504 + }, + { + "epoch": 1.01, + "learning_rate": 3.182571904586262e-05, + "loss": 3.9639, + "step": 1110016 + }, + { + "epoch": 1.01, + "learning_rate": 3.18173330983521e-05, + "loss": 3.9615, + "step": 1110528 + }, + { + "epoch": 1.01, + "learning_rate": 3.180894715084158e-05, + "loss": 3.9651, + "step": 1111040 + }, + { + "epoch": 1.01, + "learning_rate": 3.180056120333106e-05, + "loss": 3.9577, + "step": 1111552 + }, + { + "epoch": 1.01, + "learning_rate": 3.179217525582054e-05, + "loss": 3.9682, + "step": 1112064 + }, + { + "epoch": 1.01, + "learning_rate": 3.178378930831002e-05, + "loss": 3.961, + "step": 1112576 + }, + { + "epoch": 1.01, + "learning_rate": 3.177541973960323e-05, + "loss": 3.9686, + "step": 1113088 + }, + { + "epoch": 1.01, + "learning_rate": 3.176703379209271e-05, + "loss": 3.953, + "step": 1113600 + }, + { + "epoch": 1.01, + "learning_rate": 3.175864784458219e-05, + "loss": 3.9714, + "step": 1114112 + }, + { + "epoch": 1.02, + "learning_rate": 3.175026189707167e-05, + "loss": 3.9759, + "step": 1114624 + }, + { + "epoch": 1.02, + "learning_rate": 3.174189232836488e-05, + "loss": 3.9746, + "step": 1115136 + }, + { + "epoch": 1.02, + "learning_rate": 3.173350638085436e-05, + "loss": 3.9719, + "step": 1115648 + }, + { + "epoch": 1.02, + "learning_rate": 3.172512043334384e-05, + "loss": 3.9637, + "step": 1116160 + }, + { + "epoch": 1.02, + "learning_rate": 3.171673448583332e-05, + "loss": 3.9645, + "step": 1116672 + }, + { + "epoch": 1.02, + "learning_rate": 3.17083485383228e-05, + "loss": 3.9684, + "step": 1117184 + }, + { + "epoch": 1.02, + "learning_rate": 3.1699978969616014e-05, + "loss": 3.9737, + "step": 1117696 + }, + { + "epoch": 1.02, + "learning_rate": 3.169160940090922e-05, + "loss": 3.953, + "step": 1118208 + }, + { + "epoch": 1.02, + "learning_rate": 3.16832234533987e-05, + "loss": 3.96, + "step": 1118720 + }, + { + "epoch": 1.02, + "learning_rate": 3.167483750588818e-05, + "loss": 3.9769, + "step": 1119232 + }, + { + "epoch": 1.02, + "learning_rate": 3.166645155837766e-05, + "loss": 3.9637, + "step": 1119744 + }, + { + "epoch": 1.02, + "learning_rate": 3.165806561086714e-05, + "loss": 3.9799, + "step": 1120256 + }, + { + "epoch": 1.02, + "learning_rate": 3.164969604216035e-05, + "loss": 3.9628, + "step": 1120768 + }, + { + "epoch": 1.02, + "learning_rate": 3.164131009464983e-05, + "loss": 3.9724, + "step": 1121280 + }, + { + "epoch": 1.02, + "learning_rate": 3.163292414713931e-05, + "loss": 3.9536, + "step": 1121792 + }, + { + "epoch": 1.02, + "learning_rate": 3.162453819962879e-05, + "loss": 3.96, + "step": 1122304 + }, + { + "epoch": 1.02, + "learning_rate": 3.161615225211827e-05, + "loss": 3.972, + "step": 1122816 + }, + { + "epoch": 1.02, + "learning_rate": 3.160776630460775e-05, + "loss": 3.9726, + "step": 1123328 + }, + { + "epoch": 1.02, + "learning_rate": 3.159938035709723e-05, + "loss": 3.962, + "step": 1123840 + }, + { + "epoch": 1.02, + "learning_rate": 3.159099440958671e-05, + "loss": 3.9505, + "step": 1124352 + }, + { + "epoch": 1.02, + "learning_rate": 3.158262484087992e-05, + "loss": 3.9679, + "step": 1124864 + }, + { + "epoch": 1.02, + "learning_rate": 3.15742388933694e-05, + "loss": 3.9694, + "step": 1125376 + }, + { + "epoch": 1.02, + "learning_rate": 3.156585294585888e-05, + "loss": 3.9592, + "step": 1125888 + }, + { + "epoch": 1.02, + "learning_rate": 3.155746699834836e-05, + "loss": 3.9651, + "step": 1126400 + }, + { + "epoch": 1.02, + "learning_rate": 3.154908105083784e-05, + "loss": 3.9578, + "step": 1126912 + }, + { + "epoch": 1.02, + "learning_rate": 3.154069510332732e-05, + "loss": 3.962, + "step": 1127424 + }, + { + "epoch": 1.02, + "learning_rate": 3.15323091558168e-05, + "loss": 3.9602, + "step": 1127936 + }, + { + "epoch": 1.02, + "learning_rate": 3.152392320830628e-05, + "loss": 3.9662, + "step": 1128448 + }, + { + "epoch": 1.02, + "learning_rate": 3.1515553639599496e-05, + "loss": 3.9649, + "step": 1128960 + }, + { + "epoch": 1.02, + "learning_rate": 3.1507184070892705e-05, + "loss": 3.9643, + "step": 1129472 + }, + { + "epoch": 1.02, + "learning_rate": 3.1498798123382185e-05, + "loss": 3.97, + "step": 1129984 + }, + { + "epoch": 1.02, + "learning_rate": 3.1490412175871665e-05, + "loss": 3.9859, + "step": 1130496 + }, + { + "epoch": 1.02, + "learning_rate": 3.1482026228361145e-05, + "loss": 3.9622, + "step": 1131008 + }, + { + "epoch": 1.02, + "learning_rate": 3.1473640280850625e-05, + "loss": 3.9699, + "step": 1131520 + }, + { + "epoch": 1.02, + "learning_rate": 3.1465270712143834e-05, + "loss": 3.9674, + "step": 1132032 + }, + { + "epoch": 1.02, + "learning_rate": 3.1456884764633314e-05, + "loss": 3.9638, + "step": 1132544 + }, + { + "epoch": 1.02, + "learning_rate": 3.1448498817122794e-05, + "loss": 3.9664, + "step": 1133056 + }, + { + "epoch": 1.02, + "learning_rate": 3.1440112869612274e-05, + "loss": 3.9606, + "step": 1133568 + }, + { + "epoch": 1.02, + "learning_rate": 3.1431726922101754e-05, + "loss": 3.97, + "step": 1134080 + }, + { + "epoch": 1.02, + "learning_rate": 3.1423340974591234e-05, + "loss": 3.9567, + "step": 1134592 + }, + { + "epoch": 1.02, + "learning_rate": 3.1414955027080714e-05, + "loss": 3.9822, + "step": 1135104 + }, + { + "epoch": 1.02, + "learning_rate": 3.1406569079570194e-05, + "loss": 3.9668, + "step": 1135616 + }, + { + "epoch": 1.02, + "learning_rate": 3.139821588966714e-05, + "loss": 3.9761, + "step": 1136128 + }, + { + "epoch": 1.02, + "learning_rate": 3.138982994215662e-05, + "loss": 3.9659, + "step": 1136640 + }, + { + "epoch": 1.02, + "learning_rate": 3.13814439946461e-05, + "loss": 3.9734, + "step": 1137152 + }, + { + "epoch": 1.02, + "learning_rate": 3.137305804713558e-05, + "loss": 3.9643, + "step": 1137664 + }, + { + "epoch": 1.02, + "learning_rate": 3.136468847842879e-05, + "loss": 3.9707, + "step": 1138176 + }, + { + "epoch": 1.02, + "learning_rate": 3.135630253091827e-05, + "loss": 3.9774, + "step": 1138688 + }, + { + "epoch": 1.02, + "learning_rate": 3.134791658340775e-05, + "loss": 3.9613, + "step": 1139200 + }, + { + "epoch": 1.02, + "learning_rate": 3.133953063589723e-05, + "loss": 3.9706, + "step": 1139712 + }, + { + "epoch": 1.02, + "learning_rate": 3.133114468838671e-05, + "loss": 3.971, + "step": 1140224 + }, + { + "epoch": 1.02, + "learning_rate": 3.132275874087619e-05, + "loss": 3.9647, + "step": 1140736 + }, + { + "epoch": 1.02, + "learning_rate": 3.131437279336567e-05, + "loss": 3.9604, + "step": 1141248 + }, + { + "epoch": 1.02, + "learning_rate": 3.130598684585515e-05, + "loss": 3.9717, + "step": 1141760 + }, + { + "epoch": 1.02, + "learning_rate": 3.1297600898344634e-05, + "loss": 3.9691, + "step": 1142272 + }, + { + "epoch": 1.02, + "learning_rate": 3.1289231329637844e-05, + "loss": 3.9672, + "step": 1142784 + }, + { + "epoch": 1.02, + "learning_rate": 3.1280845382127324e-05, + "loss": 3.9667, + "step": 1143296 + }, + { + "epoch": 1.02, + "learning_rate": 3.1272459434616803e-05, + "loss": 3.9687, + "step": 1143808 + }, + { + "epoch": 1.02, + "learning_rate": 3.1264073487106283e-05, + "loss": 3.96, + "step": 1144320 + }, + { + "epoch": 1.03, + "eval_loss": 4.021714210510254, + "eval_runtime": 293.5152, + "eval_samples_per_second": 1300.072, + "eval_steps_per_second": 40.628, + "step": 1144787 + }, + { + "epoch": 0.0, + "learning_rate": 3.125568753959576e-05, + "loss": 3.9269, + "step": 1144832 + }, + { + "epoch": 0.0, + "learning_rate": 3.1247301592085237e-05, + "loss": 3.96, + "step": 1145344 + }, + { + "epoch": 0.0, + "learning_rate": 3.1238915644574717e-05, + "loss": 3.9556, + "step": 1145856 + }, + { + "epoch": 0.0, + "learning_rate": 3.1230529697064196e-05, + "loss": 3.9752, + "step": 1146368 + }, + { + "epoch": 0.0, + "learning_rate": 3.1222143749553676e-05, + "loss": 3.9662, + "step": 1146880 + }, + { + "epoch": 0.0, + "learning_rate": 3.1213757802043156e-05, + "loss": 3.9695, + "step": 1147392 + }, + { + "epoch": 0.0, + "learning_rate": 3.1205371854532636e-05, + "loss": 3.9614, + "step": 1147904 + }, + { + "epoch": 0.0, + "learning_rate": 3.1196985907022116e-05, + "loss": 3.9644, + "step": 1148416 + }, + { + "epoch": 0.0, + "learning_rate": 3.1188599959511596e-05, + "loss": 3.9541, + "step": 1148928 + }, + { + "epoch": 0.0, + "learning_rate": 3.118021401200108e-05, + "loss": 3.9608, + "step": 1149440 + }, + { + "epoch": 0.0, + "learning_rate": 3.117182806449056e-05, + "loss": 3.9638, + "step": 1149952 + }, + { + "epoch": 0.0, + "learning_rate": 3.116344211698004e-05, + "loss": 3.9794, + "step": 1150464 + }, + { + "epoch": 0.0, + "learning_rate": 3.115507254827325e-05, + "loss": 3.9725, + "step": 1150976 + }, + { + "epoch": 0.0, + "learning_rate": 3.114668660076273e-05, + "loss": 3.9456, + "step": 1151488 + }, + { + "epoch": 0.0, + "learning_rate": 3.113830065325221e-05, + "loss": 3.9624, + "step": 1152000 + }, + { + "epoch": 0.0, + "learning_rate": 3.112991470574169e-05, + "loss": 3.954, + "step": 1152512 + }, + { + "epoch": 0.0, + "learning_rate": 3.1121528758231165e-05, + "loss": 3.9652, + "step": 1153024 + }, + { + "epoch": 0.0, + "learning_rate": 3.1113142810720645e-05, + "loss": 3.958, + "step": 1153536 + }, + { + "epoch": 0.0, + "learning_rate": 3.1104756863210125e-05, + "loss": 3.9643, + "step": 1154048 + }, + { + "epoch": 0.0, + "learning_rate": 3.1096370915699605e-05, + "loss": 3.9589, + "step": 1154560 + }, + { + "epoch": 0.0, + "learning_rate": 3.1087984968189085e-05, + "loss": 3.9769, + "step": 1155072 + }, + { + "epoch": 0.0, + "learning_rate": 3.1079599020678565e-05, + "loss": 3.9659, + "step": 1155584 + }, + { + "epoch": 0.0, + "learning_rate": 3.107121307316805e-05, + "loss": 3.9641, + "step": 1156096 + }, + { + "epoch": 0.0, + "learning_rate": 3.106282712565753e-05, + "loss": 3.9591, + "step": 1156608 + }, + { + "epoch": 0.0, + "learning_rate": 3.105445755695074e-05, + "loss": 3.9734, + "step": 1157120 + }, + { + "epoch": 0.0, + "learning_rate": 3.104607160944022e-05, + "loss": 3.9612, + "step": 1157632 + }, + { + "epoch": 0.0, + "learning_rate": 3.10376856619297e-05, + "loss": 3.9526, + "step": 1158144 + }, + { + "epoch": 0.0, + "learning_rate": 3.102929971441918e-05, + "loss": 3.963, + "step": 1158656 + }, + { + "epoch": 0.0, + "learning_rate": 3.102091376690866e-05, + "loss": 3.9584, + "step": 1159168 + }, + { + "epoch": 0.0, + "learning_rate": 3.101254419820187e-05, + "loss": 3.9566, + "step": 1159680 + }, + { + "epoch": 0.01, + "learning_rate": 3.100415825069135e-05, + "loss": 3.9633, + "step": 1160192 + }, + { + "epoch": 0.01, + "learning_rate": 3.099577230318083e-05, + "loss": 3.9569, + "step": 1160704 + }, + { + "epoch": 0.01, + "learning_rate": 3.098740273447404e-05, + "loss": 3.9685, + "step": 1161216 + }, + { + "epoch": 0.01, + "learning_rate": 3.097901678696352e-05, + "loss": 3.967, + "step": 1161728 + }, + { + "epoch": 0.01, + "learning_rate": 3.0970630839453005e-05, + "loss": 3.9686, + "step": 1162240 + }, + { + "epoch": 0.01, + "learning_rate": 3.0962261270746214e-05, + "loss": 3.9636, + "step": 1162752 + }, + { + "epoch": 0.01, + "learning_rate": 3.0953875323235694e-05, + "loss": 3.9715, + "step": 1163264 + }, + { + "epoch": 0.01, + "learning_rate": 3.0945489375725174e-05, + "loss": 3.9523, + "step": 1163776 + }, + { + "epoch": 0.01, + "learning_rate": 3.0937103428214654e-05, + "loss": 3.9581, + "step": 1164288 + }, + { + "epoch": 0.01, + "learning_rate": 3.0928717480704134e-05, + "loss": 3.9592, + "step": 1164800 + }, + { + "epoch": 0.01, + "learning_rate": 3.0920331533193614e-05, + "loss": 3.9505, + "step": 1165312 + }, + { + "epoch": 0.01, + "learning_rate": 3.0911945585683094e-05, + "loss": 3.9632, + "step": 1165824 + }, + { + "epoch": 0.01, + "learning_rate": 3.0903559638172574e-05, + "loss": 3.9619, + "step": 1166336 + }, + { + "epoch": 0.01, + "learning_rate": 3.0895173690662054e-05, + "loss": 3.9535, + "step": 1166848 + }, + { + "epoch": 0.01, + "learning_rate": 3.088678774315153e-05, + "loss": 3.9639, + "step": 1167360 + }, + { + "epoch": 0.01, + "learning_rate": 3.087840179564101e-05, + "loss": 3.9656, + "step": 1167872 + }, + { + "epoch": 0.01, + "learning_rate": 3.087001584813049e-05, + "loss": 3.9641, + "step": 1168384 + }, + { + "epoch": 0.01, + "learning_rate": 3.086166265822744e-05, + "loss": 3.9627, + "step": 1168896 + }, + { + "epoch": 0.01, + "learning_rate": 3.085327671071692e-05, + "loss": 3.9432, + "step": 1169408 + }, + { + "epoch": 0.01, + "learning_rate": 3.08448907632064e-05, + "loss": 3.9617, + "step": 1169920 + }, + { + "epoch": 0.01, + "learning_rate": 3.083650481569588e-05, + "loss": 3.9646, + "step": 1170432 + }, + { + "epoch": 0.01, + "learning_rate": 3.082813524698909e-05, + "loss": 3.9624, + "step": 1170944 + }, + { + "epoch": 0.01, + "learning_rate": 3.081974929947857e-05, + "loss": 3.9551, + "step": 1171456 + }, + { + "epoch": 0.01, + "learning_rate": 3.081136335196805e-05, + "loss": 3.9539, + "step": 1171968 + }, + { + "epoch": 0.01, + "learning_rate": 3.080297740445753e-05, + "loss": 3.9486, + "step": 1172480 + }, + { + "epoch": 0.01, + "learning_rate": 3.0794591456947e-05, + "loss": 3.9502, + "step": 1172992 + }, + { + "epoch": 0.01, + "learning_rate": 3.078620550943648e-05, + "loss": 3.9592, + "step": 1173504 + }, + { + "epoch": 0.01, + "learning_rate": 3.077781956192596e-05, + "loss": 3.9441, + "step": 1174016 + }, + { + "epoch": 0.01, + "learning_rate": 3.076943361441544e-05, + "loss": 3.9653, + "step": 1174528 + }, + { + "epoch": 0.01, + "learning_rate": 3.0761064045708657e-05, + "loss": 3.9642, + "step": 1175040 + }, + { + "epoch": 0.01, + "learning_rate": 3.0752678098198137e-05, + "loss": 3.9482, + "step": 1175552 + }, + { + "epoch": 0.01, + "learning_rate": 3.0744292150687617e-05, + "loss": 3.9682, + "step": 1176064 + }, + { + "epoch": 0.01, + "learning_rate": 3.0735906203177096e-05, + "loss": 3.9449, + "step": 1176576 + }, + { + "epoch": 0.01, + "learning_rate": 3.0727520255666576e-05, + "loss": 3.9498, + "step": 1177088 + }, + { + "epoch": 0.01, + "learning_rate": 3.0719150686959786e-05, + "loss": 3.9492, + "step": 1177600 + }, + { + "epoch": 0.01, + "learning_rate": 3.0710764739449265e-05, + "loss": 3.9617, + "step": 1178112 + }, + { + "epoch": 0.01, + "learning_rate": 3.0702378791938745e-05, + "loss": 3.9477, + "step": 1178624 + }, + { + "epoch": 0.01, + "learning_rate": 3.0693992844428225e-05, + "loss": 3.9636, + "step": 1179136 + }, + { + "epoch": 0.01, + "learning_rate": 3.0685623275721434e-05, + "loss": 3.9389, + "step": 1179648 + }, + { + "epoch": 0.01, + "learning_rate": 3.0677237328210914e-05, + "loss": 3.9499, + "step": 1180160 + }, + { + "epoch": 0.01, + "learning_rate": 3.066886775950413e-05, + "loss": 3.9663, + "step": 1180672 + }, + { + "epoch": 0.01, + "learning_rate": 3.066048181199361e-05, + "loss": 3.9572, + "step": 1181184 + }, + { + "epoch": 0.01, + "learning_rate": 3.065209586448309e-05, + "loss": 3.962, + "step": 1181696 + }, + { + "epoch": 0.01, + "learning_rate": 3.064370991697257e-05, + "loss": 3.9632, + "step": 1182208 + }, + { + "epoch": 0.01, + "learning_rate": 3.063532396946205e-05, + "loss": 3.9728, + "step": 1182720 + }, + { + "epoch": 0.01, + "learning_rate": 3.062693802195153e-05, + "loss": 3.9521, + "step": 1183232 + }, + { + "epoch": 0.01, + "learning_rate": 3.061855207444101e-05, + "loss": 3.9694, + "step": 1183744 + }, + { + "epoch": 0.01, + "learning_rate": 3.061016612693049e-05, + "loss": 3.9493, + "step": 1184256 + }, + { + "epoch": 0.01, + "learning_rate": 3.060181293702743e-05, + "loss": 3.9543, + "step": 1184768 + }, + { + "epoch": 0.01, + "learning_rate": 3.059342698951691e-05, + "loss": 3.9562, + "step": 1185280 + }, + { + "epoch": 0.01, + "learning_rate": 3.058504104200639e-05, + "loss": 3.9631, + "step": 1185792 + }, + { + "epoch": 0.01, + "learning_rate": 3.057665509449587e-05, + "loss": 3.9623, + "step": 1186304 + }, + { + "epoch": 0.01, + "learning_rate": 3.056826914698535e-05, + "loss": 3.9687, + "step": 1186816 + }, + { + "epoch": 0.01, + "learning_rate": 3.055988319947483e-05, + "loss": 3.9492, + "step": 1187328 + }, + { + "epoch": 0.01, + "learning_rate": 3.0551497251964315e-05, + "loss": 3.9452, + "step": 1187840 + }, + { + "epoch": 0.01, + "learning_rate": 3.0543111304453795e-05, + "loss": 3.9472, + "step": 1188352 + }, + { + "epoch": 0.01, + "learning_rate": 3.0534741735747004e-05, + "loss": 3.9589, + "step": 1188864 + }, + { + "epoch": 0.01, + "learning_rate": 3.052637216704021e-05, + "loss": 3.9522, + "step": 1189376 + }, + { + "epoch": 0.01, + "learning_rate": 3.051798621952969e-05, + "loss": 3.9525, + "step": 1189888 + }, + { + "epoch": 0.01, + "learning_rate": 3.0509600272019173e-05, + "loss": 3.959, + "step": 1190400 + }, + { + "epoch": 0.02, + "learning_rate": 3.0501214324508653e-05, + "loss": 3.9523, + "step": 1190912 + }, + { + "epoch": 0.02, + "learning_rate": 3.0492828376998133e-05, + "loss": 3.9548, + "step": 1191424 + }, + { + "epoch": 0.02, + "learning_rate": 3.0484442429487613e-05, + "loss": 3.951, + "step": 1191936 + }, + { + "epoch": 0.02, + "learning_rate": 3.0476056481977096e-05, + "loss": 3.9639, + "step": 1192448 + }, + { + "epoch": 0.02, + "learning_rate": 3.0467670534466576e-05, + "loss": 3.9676, + "step": 1192960 + }, + { + "epoch": 0.02, + "learning_rate": 3.0459284586956056e-05, + "loss": 3.9628, + "step": 1193472 + }, + { + "epoch": 0.02, + "learning_rate": 3.0450915018249265e-05, + "loss": 3.9421, + "step": 1193984 + }, + { + "epoch": 0.02, + "learning_rate": 3.0442545449542474e-05, + "loss": 3.954, + "step": 1194496 + }, + { + "epoch": 0.02, + "learning_rate": 3.0434159502031957e-05, + "loss": 3.9522, + "step": 1195008 + }, + { + "epoch": 0.02, + "learning_rate": 3.0425773554521437e-05, + "loss": 3.9598, + "step": 1195520 + }, + { + "epoch": 0.02, + "learning_rate": 3.0417387607010917e-05, + "loss": 3.9648, + "step": 1196032 + }, + { + "epoch": 0.02, + "learning_rate": 3.0409001659500397e-05, + "loss": 3.9481, + "step": 1196544 + }, + { + "epoch": 0.02, + "learning_rate": 3.0400615711989877e-05, + "loss": 3.9582, + "step": 1197056 + }, + { + "epoch": 0.02, + "learning_rate": 3.0392229764479357e-05, + "loss": 3.9601, + "step": 1197568 + }, + { + "epoch": 0.02, + "learning_rate": 3.0383843816968837e-05, + "loss": 3.9565, + "step": 1198080 + }, + { + "epoch": 0.02, + "learning_rate": 3.037547424826205e-05, + "loss": 3.9494, + "step": 1198592 + }, + { + "epoch": 0.02, + "learning_rate": 3.036710467955526e-05, + "loss": 3.9516, + "step": 1199104 + }, + { + "epoch": 0.02, + "learning_rate": 3.035871873204474e-05, + "loss": 3.9445, + "step": 1199616 + }, + { + "epoch": 0.02, + "learning_rate": 3.035033278453422e-05, + "loss": 3.955, + "step": 1200128 + }, + { + "epoch": 0.02, + "learning_rate": 3.03419468370237e-05, + "loss": 3.949, + "step": 1200640 + }, + { + "epoch": 0.02, + "learning_rate": 3.033357726831691e-05, + "loss": 3.949, + "step": 1201152 + }, + { + "epoch": 0.02, + "learning_rate": 3.032519132080639e-05, + "loss": 3.9492, + "step": 1201664 + }, + { + "epoch": 0.02, + "learning_rate": 3.031680537329587e-05, + "loss": 3.9575, + "step": 1202176 + }, + { + "epoch": 0.02, + "learning_rate": 3.030841942578535e-05, + "loss": 3.9491, + "step": 1202688 + }, + { + "epoch": 0.02, + "learning_rate": 3.030003347827483e-05, + "loss": 3.9544, + "step": 1203200 + }, + { + "epoch": 0.02, + "learning_rate": 3.029164753076431e-05, + "loss": 3.9653, + "step": 1203712 + }, + { + "epoch": 0.02, + "learning_rate": 3.0283261583253787e-05, + "loss": 3.9608, + "step": 1204224 + }, + { + "epoch": 0.02, + "learning_rate": 3.0274875635743267e-05, + "loss": 3.9613, + "step": 1204736 + }, + { + "epoch": 0.02, + "learning_rate": 3.0266506067036483e-05, + "loss": 3.9502, + "step": 1205248 + }, + { + "epoch": 0.02, + "learning_rate": 3.0258120119525963e-05, + "loss": 3.9559, + "step": 1205760 + }, + { + "epoch": 0.02, + "learning_rate": 3.0249734172015436e-05, + "loss": 3.9539, + "step": 1206272 + }, + { + "epoch": 0.02, + "learning_rate": 3.0241364603308652e-05, + "loss": 3.945, + "step": 1206784 + }, + { + "epoch": 0.02, + "learning_rate": 3.0232978655798132e-05, + "loss": 3.9567, + "step": 1207296 + }, + { + "epoch": 0.02, + "learning_rate": 3.022459270828761e-05, + "loss": 3.9492, + "step": 1207808 + }, + { + "epoch": 0.02, + "learning_rate": 3.021620676077709e-05, + "loss": 3.9589, + "step": 1208320 + }, + { + "epoch": 0.02, + "learning_rate": 3.020782081326657e-05, + "loss": 3.9514, + "step": 1208832 + }, + { + "epoch": 0.02, + "learning_rate": 3.019943486575605e-05, + "loss": 3.9579, + "step": 1209344 + }, + { + "epoch": 0.02, + "learning_rate": 3.019104891824553e-05, + "loss": 3.9515, + "step": 1209856 + }, + { + "epoch": 0.02, + "learning_rate": 3.018266297073501e-05, + "loss": 3.9503, + "step": 1210368 + }, + { + "epoch": 0.02, + "learning_rate": 3.017429340202822e-05, + "loss": 3.9581, + "step": 1210880 + }, + { + "epoch": 0.02, + "learning_rate": 3.01659074545177e-05, + "loss": 3.9594, + "step": 1211392 + }, + { + "epoch": 0.02, + "learning_rate": 3.015752150700718e-05, + "loss": 3.9633, + "step": 1211904 + }, + { + "epoch": 0.02, + "learning_rate": 3.014913555949666e-05, + "loss": 3.9544, + "step": 1212416 + }, + { + "epoch": 0.02, + "learning_rate": 3.014076599078987e-05, + "loss": 3.9593, + "step": 1212928 + }, + { + "epoch": 0.02, + "learning_rate": 3.013238004327935e-05, + "loss": 3.9589, + "step": 1213440 + }, + { + "epoch": 0.02, + "learning_rate": 3.012399409576883e-05, + "loss": 3.9597, + "step": 1213952 + }, + { + "epoch": 0.02, + "learning_rate": 3.0115608148258313e-05, + "loss": 3.9582, + "step": 1214464 + }, + { + "epoch": 0.02, + "learning_rate": 3.0107222200747793e-05, + "loss": 3.9542, + "step": 1214976 + }, + { + "epoch": 0.02, + "learning_rate": 3.0098836253237273e-05, + "loss": 3.9546, + "step": 1215488 + }, + { + "epoch": 0.02, + "learning_rate": 3.0090466684530482e-05, + "loss": 3.9633, + "step": 1216000 + }, + { + "epoch": 0.02, + "learning_rate": 3.008209711582369e-05, + "loss": 3.9516, + "step": 1216512 + }, + { + "epoch": 0.02, + "learning_rate": 3.0073711168313175e-05, + "loss": 3.9498, + "step": 1217024 + }, + { + "epoch": 0.02, + "learning_rate": 3.0065325220802655e-05, + "loss": 3.9566, + "step": 1217536 + }, + { + "epoch": 0.02, + "learning_rate": 3.0056939273292135e-05, + "loss": 3.9545, + "step": 1218048 + }, + { + "epoch": 0.02, + "learning_rate": 3.0048553325781615e-05, + "loss": 3.9529, + "step": 1218560 + }, + { + "epoch": 0.02, + "learning_rate": 3.0040167378271095e-05, + "loss": 3.9586, + "step": 1219072 + }, + { + "epoch": 0.02, + "learning_rate": 3.0031781430760575e-05, + "loss": 3.9521, + "step": 1219584 + }, + { + "epoch": 0.02, + "learning_rate": 3.0023395483250055e-05, + "loss": 3.9659, + "step": 1220096 + }, + { + "epoch": 0.02, + "learning_rate": 3.0015009535739534e-05, + "loss": 3.9578, + "step": 1220608 + }, + { + "epoch": 0.03, + "eval_loss": 4.01444149017334, + "eval_runtime": 342.8971, + "eval_samples_per_second": 1112.844, + "eval_steps_per_second": 34.777, + "step": 1221107 + }, + { + "epoch": 1.0, + "learning_rate": 3.0006623588229014e-05, + "loss": 3.9518, + "step": 1221120 + }, + { + "epoch": 1.0, + "learning_rate": 2.9998237640718498e-05, + "loss": 3.9453, + "step": 1221632 + }, + { + "epoch": 1.0, + "learning_rate": 2.9989868072011707e-05, + "loss": 3.9451, + "step": 1222144 + }, + { + "epoch": 1.0, + "learning_rate": 2.9981482124501187e-05, + "loss": 3.9626, + "step": 1222656 + }, + { + "epoch": 1.0, + "learning_rate": 2.9973096176990667e-05, + "loss": 3.9592, + "step": 1223168 + }, + { + "epoch": 1.0, + "learning_rate": 2.9964710229480147e-05, + "loss": 3.9549, + "step": 1223680 + }, + { + "epoch": 1.0, + "learning_rate": 2.9956324281969623e-05, + "loss": 3.9507, + "step": 1224192 + }, + { + "epoch": 1.0, + "learning_rate": 2.994795471326284e-05, + "loss": 3.9537, + "step": 1224704 + }, + { + "epoch": 1.0, + "learning_rate": 2.993956876575232e-05, + "loss": 3.94, + "step": 1225216 + }, + { + "epoch": 1.0, + "learning_rate": 2.9931182818241792e-05, + "loss": 3.9523, + "step": 1225728 + }, + { + "epoch": 1.0, + "learning_rate": 2.9922796870731272e-05, + "loss": 3.9506, + "step": 1226240 + }, + { + "epoch": 1.0, + "learning_rate": 2.9914427302024488e-05, + "loss": 3.9706, + "step": 1226752 + }, + { + "epoch": 1.0, + "learning_rate": 2.99060577333177e-05, + "loss": 3.9642, + "step": 1227264 + }, + { + "epoch": 1.0, + "learning_rate": 2.989767178580718e-05, + "loss": 3.9335, + "step": 1227776 + }, + { + "epoch": 1.0, + "learning_rate": 2.988928583829666e-05, + "loss": 3.9489, + "step": 1228288 + }, + { + "epoch": 1.0, + "learning_rate": 2.988089989078614e-05, + "loss": 3.9464, + "step": 1228800 + }, + { + "epoch": 1.0, + "learning_rate": 2.987251394327562e-05, + "loss": 3.9472, + "step": 1229312 + }, + { + "epoch": 1.0, + "learning_rate": 2.9864127995765097e-05, + "loss": 3.9525, + "step": 1229824 + }, + { + "epoch": 1.0, + "learning_rate": 2.9855758427058313e-05, + "loss": 3.9507, + "step": 1230336 + }, + { + "epoch": 1.0, + "learning_rate": 2.9847372479547793e-05, + "loss": 3.9454, + "step": 1230848 + }, + { + "epoch": 1.0, + "learning_rate": 2.9838986532037266e-05, + "loss": 3.9681, + "step": 1231360 + }, + { + "epoch": 1.0, + "learning_rate": 2.9830600584526746e-05, + "loss": 3.9559, + "step": 1231872 + }, + { + "epoch": 1.0, + "learning_rate": 2.9822214637016226e-05, + "loss": 3.9502, + "step": 1232384 + }, + { + "epoch": 1.0, + "learning_rate": 2.9813828689505706e-05, + "loss": 3.9462, + "step": 1232896 + }, + { + "epoch": 1.0, + "learning_rate": 2.980544274199519e-05, + "loss": 3.9658, + "step": 1233408 + }, + { + "epoch": 1.0, + "learning_rate": 2.979705679448467e-05, + "loss": 3.9498, + "step": 1233920 + }, + { + "epoch": 1.0, + "learning_rate": 2.978867084697415e-05, + "loss": 3.9408, + "step": 1234432 + }, + { + "epoch": 1.0, + "learning_rate": 2.978028489946363e-05, + "loss": 3.9507, + "step": 1234944 + }, + { + "epoch": 1.0, + "learning_rate": 2.977189895195311e-05, + "loss": 3.9442, + "step": 1235456 + }, + { + "epoch": 1.0, + "learning_rate": 2.9763529383246318e-05, + "loss": 3.9476, + "step": 1235968 + }, + { + "epoch": 1.01, + "learning_rate": 2.9755143435735798e-05, + "loss": 3.9496, + "step": 1236480 + }, + { + "epoch": 1.01, + "learning_rate": 2.974675748822528e-05, + "loss": 3.9473, + "step": 1236992 + }, + { + "epoch": 1.01, + "learning_rate": 2.973837154071476e-05, + "loss": 3.9564, + "step": 1237504 + }, + { + "epoch": 1.01, + "learning_rate": 2.973000197200797e-05, + "loss": 3.9518, + "step": 1238016 + }, + { + "epoch": 1.01, + "learning_rate": 2.972161602449745e-05, + "loss": 3.9608, + "step": 1238528 + }, + { + "epoch": 1.01, + "learning_rate": 2.971324645579066e-05, + "loss": 3.9549, + "step": 1239040 + }, + { + "epoch": 1.01, + "learning_rate": 2.9704860508280143e-05, + "loss": 3.9591, + "step": 1239552 + }, + { + "epoch": 1.01, + "learning_rate": 2.9696474560769623e-05, + "loss": 3.939, + "step": 1240064 + }, + { + "epoch": 1.01, + "learning_rate": 2.9688088613259103e-05, + "loss": 3.9485, + "step": 1240576 + }, + { + "epoch": 1.01, + "learning_rate": 2.9679702665748583e-05, + "loss": 3.9416, + "step": 1241088 + }, + { + "epoch": 1.01, + "learning_rate": 2.9671316718238063e-05, + "loss": 3.9468, + "step": 1241600 + }, + { + "epoch": 1.01, + "learning_rate": 2.9662930770727543e-05, + "loss": 3.9461, + "step": 1242112 + }, + { + "epoch": 1.01, + "learning_rate": 2.9654544823217023e-05, + "loss": 3.9527, + "step": 1242624 + }, + { + "epoch": 1.01, + "learning_rate": 2.9646175254510232e-05, + "loss": 3.9432, + "step": 1243136 + }, + { + "epoch": 1.01, + "learning_rate": 2.9637805685803444e-05, + "loss": 3.9552, + "step": 1243648 + }, + { + "epoch": 1.01, + "learning_rate": 2.9629419738292924e-05, + "loss": 3.9505, + "step": 1244160 + }, + { + "epoch": 1.01, + "learning_rate": 2.9621033790782404e-05, + "loss": 3.956, + "step": 1244672 + }, + { + "epoch": 1.01, + "learning_rate": 2.9612647843271884e-05, + "loss": 3.9517, + "step": 1245184 + }, + { + "epoch": 1.01, + "learning_rate": 2.9604261895761364e-05, + "loss": 3.9324, + "step": 1245696 + }, + { + "epoch": 1.01, + "learning_rate": 2.9595875948250844e-05, + "loss": 3.9475, + "step": 1246208 + }, + { + "epoch": 1.01, + "learning_rate": 2.9587490000740324e-05, + "loss": 3.9582, + "step": 1246720 + }, + { + "epoch": 1.01, + "learning_rate": 2.9579104053229807e-05, + "loss": 3.953, + "step": 1247232 + }, + { + "epoch": 1.01, + "learning_rate": 2.9570734484523016e-05, + "loss": 3.9416, + "step": 1247744 + }, + { + "epoch": 1.01, + "learning_rate": 2.9562348537012496e-05, + "loss": 3.9419, + "step": 1248256 + }, + { + "epoch": 1.01, + "learning_rate": 2.9553962589501976e-05, + "loss": 3.9405, + "step": 1248768 + }, + { + "epoch": 1.01, + "learning_rate": 2.9545593020795185e-05, + "loss": 3.936, + "step": 1249280 + }, + { + "epoch": 1.01, + "learning_rate": 2.953720707328467e-05, + "loss": 3.9538, + "step": 1249792 + }, + { + "epoch": 1.01, + "learning_rate": 2.952882112577415e-05, + "loss": 3.9318, + "step": 1250304 + }, + { + "epoch": 1.01, + "learning_rate": 2.952043517826363e-05, + "loss": 3.9528, + "step": 1250816 + }, + { + "epoch": 1.01, + "learning_rate": 2.9512049230753102e-05, + "loss": 3.9518, + "step": 1251328 + }, + { + "epoch": 1.01, + "learning_rate": 2.9503663283242582e-05, + "loss": 3.937, + "step": 1251840 + }, + { + "epoch": 1.01, + "learning_rate": 2.9495277335732062e-05, + "loss": 3.9572, + "step": 1252352 + }, + { + "epoch": 1.01, + "learning_rate": 2.9486891388221545e-05, + "loss": 3.9356, + "step": 1252864 + }, + { + "epoch": 1.01, + "learning_rate": 2.9478505440711025e-05, + "loss": 3.9424, + "step": 1253376 + }, + { + "epoch": 1.01, + "learning_rate": 2.9470135872004234e-05, + "loss": 3.934, + "step": 1253888 + }, + { + "epoch": 1.01, + "learning_rate": 2.9461749924493714e-05, + "loss": 3.9547, + "step": 1254400 + }, + { + "epoch": 1.01, + "learning_rate": 2.9453363976983194e-05, + "loss": 3.9397, + "step": 1254912 + }, + { + "epoch": 1.01, + "learning_rate": 2.9444978029472674e-05, + "loss": 3.9476, + "step": 1255424 + }, + { + "epoch": 1.01, + "learning_rate": 2.9436592081962154e-05, + "loss": 3.9308, + "step": 1255936 + }, + { + "epoch": 1.01, + "learning_rate": 2.9428222513255366e-05, + "loss": 3.9378, + "step": 1256448 + }, + { + "epoch": 1.01, + "learning_rate": 2.9419836565744846e-05, + "loss": 3.9534, + "step": 1256960 + }, + { + "epoch": 1.01, + "learning_rate": 2.9411450618234326e-05, + "loss": 3.9509, + "step": 1257472 + }, + { + "epoch": 1.01, + "learning_rate": 2.9403064670723806e-05, + "loss": 3.9448, + "step": 1257984 + }, + { + "epoch": 1.01, + "learning_rate": 2.9394678723213286e-05, + "loss": 3.9526, + "step": 1258496 + }, + { + "epoch": 1.01, + "learning_rate": 2.9386292775702766e-05, + "loss": 3.9612, + "step": 1259008 + }, + { + "epoch": 1.01, + "learning_rate": 2.937792320699598e-05, + "loss": 3.9413, + "step": 1259520 + }, + { + "epoch": 1.01, + "learning_rate": 2.936953725948546e-05, + "loss": 3.9594, + "step": 1260032 + }, + { + "epoch": 1.01, + "learning_rate": 2.936115131197494e-05, + "loss": 3.9366, + "step": 1260544 + }, + { + "epoch": 1.01, + "learning_rate": 2.935276536446442e-05, + "loss": 3.9454, + "step": 1261056 + }, + { + "epoch": 1.01, + "learning_rate": 2.93443794169539e-05, + "loss": 3.9449, + "step": 1261568 + }, + { + "epoch": 1.01, + "learning_rate": 2.9336009848247108e-05, + "loss": 3.951, + "step": 1262080 + }, + { + "epoch": 1.01, + "learning_rate": 2.932762390073659e-05, + "loss": 3.9501, + "step": 1262592 + }, + { + "epoch": 1.01, + "learning_rate": 2.931923795322607e-05, + "loss": 3.9599, + "step": 1263104 + }, + { + "epoch": 1.01, + "learning_rate": 2.931086838451928e-05, + "loss": 3.9437, + "step": 1263616 + }, + { + "epoch": 1.01, + "learning_rate": 2.930248243700876e-05, + "loss": 3.928, + "step": 1264128 + }, + { + "epoch": 1.01, + "learning_rate": 2.929409648949824e-05, + "loss": 3.9366, + "step": 1264640 + }, + { + "epoch": 1.01, + "learning_rate": 2.928571054198772e-05, + "loss": 3.9474, + "step": 1265152 + }, + { + "epoch": 1.01, + "learning_rate": 2.9277340973280932e-05, + "loss": 3.9398, + "step": 1265664 + }, + { + "epoch": 1.01, + "learning_rate": 2.9268955025770412e-05, + "loss": 3.9429, + "step": 1266176 + }, + { + "epoch": 1.01, + "learning_rate": 2.9260569078259892e-05, + "loss": 3.9486, + "step": 1266688 + }, + { + "epoch": 1.02, + "learning_rate": 2.9252183130749372e-05, + "loss": 3.939, + "step": 1267200 + }, + { + "epoch": 1.02, + "learning_rate": 2.9243797183238852e-05, + "loss": 3.9435, + "step": 1267712 + }, + { + "epoch": 1.02, + "learning_rate": 2.9235411235728332e-05, + "loss": 3.9387, + "step": 1268224 + }, + { + "epoch": 1.02, + "learning_rate": 2.9227025288217812e-05, + "loss": 3.9566, + "step": 1268736 + }, + { + "epoch": 1.02, + "learning_rate": 2.921863934070729e-05, + "loss": 3.9551, + "step": 1269248 + }, + { + "epoch": 1.02, + "learning_rate": 2.9210269772000505e-05, + "loss": 3.9508, + "step": 1269760 + }, + { + "epoch": 1.02, + "learning_rate": 2.9201883824489984e-05, + "loss": 3.9344, + "step": 1270272 + }, + { + "epoch": 1.02, + "learning_rate": 2.9193497876979464e-05, + "loss": 3.9441, + "step": 1270784 + }, + { + "epoch": 1.02, + "learning_rate": 2.9185111929468938e-05, + "loss": 3.9455, + "step": 1271296 + }, + { + "epoch": 1.02, + "learning_rate": 2.917672598195842e-05, + "loss": 3.9455, + "step": 1271808 + }, + { + "epoch": 1.02, + "learning_rate": 2.9168356413251637e-05, + "loss": 3.9538, + "step": 1272320 + }, + { + "epoch": 1.02, + "learning_rate": 2.9159970465741117e-05, + "loss": 3.9389, + "step": 1272832 + }, + { + "epoch": 1.02, + "learning_rate": 2.915158451823059e-05, + "loss": 3.9451, + "step": 1273344 + }, + { + "epoch": 1.02, + "learning_rate": 2.914319857072007e-05, + "loss": 3.9479, + "step": 1273856 + }, + { + "epoch": 1.02, + "learning_rate": 2.9134829002013286e-05, + "loss": 3.95, + "step": 1274368 + }, + { + "epoch": 1.02, + "learning_rate": 2.9126443054502762e-05, + "loss": 3.935, + "step": 1274880 + }, + { + "epoch": 1.02, + "learning_rate": 2.9118073485795978e-05, + "loss": 3.9432, + "step": 1275392 + }, + { + "epoch": 1.02, + "learning_rate": 2.9109687538285458e-05, + "loss": 3.934, + "step": 1275904 + }, + { + "epoch": 1.02, + "learning_rate": 2.9101301590774938e-05, + "loss": 3.9377, + "step": 1276416 + }, + { + "epoch": 1.02, + "learning_rate": 2.909291564326441e-05, + "loss": 3.94, + "step": 1276928 + }, + { + "epoch": 1.02, + "learning_rate": 2.908452969575389e-05, + "loss": 3.9402, + "step": 1277440 + }, + { + "epoch": 1.02, + "learning_rate": 2.9076143748243375e-05, + "loss": 3.9338, + "step": 1277952 + }, + { + "epoch": 1.02, + "learning_rate": 2.9067757800732855e-05, + "loss": 3.9467, + "step": 1278464 + }, + { + "epoch": 1.02, + "learning_rate": 2.9059371853222335e-05, + "loss": 3.94, + "step": 1278976 + }, + { + "epoch": 1.02, + "learning_rate": 2.9051002284515544e-05, + "loss": 3.9411, + "step": 1279488 + }, + { + "epoch": 1.02, + "learning_rate": 2.9042616337005024e-05, + "loss": 3.953, + "step": 1280000 + }, + { + "epoch": 1.02, + "learning_rate": 2.9034230389494504e-05, + "loss": 3.9549, + "step": 1280512 + }, + { + "epoch": 1.02, + "learning_rate": 2.9025844441983984e-05, + "loss": 3.9509, + "step": 1281024 + }, + { + "epoch": 1.02, + "learning_rate": 2.9017474873277196e-05, + "loss": 3.9414, + "step": 1281536 + }, + { + "epoch": 1.02, + "learning_rate": 2.9009088925766676e-05, + "loss": 3.9451, + "step": 1282048 + }, + { + "epoch": 1.02, + "learning_rate": 2.9000719357059885e-05, + "loss": 3.9433, + "step": 1282560 + }, + { + "epoch": 1.02, + "learning_rate": 2.89923497883531e-05, + "loss": 3.9364, + "step": 1283072 + }, + { + "epoch": 1.02, + "learning_rate": 2.898396384084258e-05, + "loss": 3.9424, + "step": 1283584 + }, + { + "epoch": 1.02, + "learning_rate": 2.897557789333206e-05, + "loss": 3.9392, + "step": 1284096 + }, + { + "epoch": 1.02, + "learning_rate": 2.8967191945821537e-05, + "loss": 3.9468, + "step": 1284608 + }, + { + "epoch": 1.02, + "learning_rate": 2.8958805998311017e-05, + "loss": 3.942, + "step": 1285120 + }, + { + "epoch": 1.02, + "learning_rate": 2.8950420050800497e-05, + "loss": 3.9482, + "step": 1285632 + }, + { + "epoch": 1.02, + "learning_rate": 2.8942034103289977e-05, + "loss": 3.9454, + "step": 1286144 + }, + { + "epoch": 1.02, + "learning_rate": 2.8933648155779457e-05, + "loss": 3.9384, + "step": 1286656 + }, + { + "epoch": 1.02, + "learning_rate": 2.8925262208268937e-05, + "loss": 3.9371, + "step": 1287168 + }, + { + "epoch": 1.02, + "learning_rate": 2.891689263956215e-05, + "loss": 3.9562, + "step": 1287680 + }, + { + "epoch": 1.02, + "learning_rate": 2.890850669205163e-05, + "loss": 3.9477, + "step": 1288192 + }, + { + "epoch": 1.02, + "learning_rate": 2.890012074454111e-05, + "loss": 3.9481, + "step": 1288704 + }, + { + "epoch": 1.02, + "learning_rate": 2.889173479703059e-05, + "loss": 3.9467, + "step": 1289216 + }, + { + "epoch": 1.02, + "learning_rate": 2.888334884952007e-05, + "loss": 3.95, + "step": 1289728 + }, + { + "epoch": 1.02, + "learning_rate": 2.887496290200955e-05, + "loss": 3.9509, + "step": 1290240 + }, + { + "epoch": 1.02, + "learning_rate": 2.886657695449903e-05, + "loss": 3.95, + "step": 1290752 + }, + { + "epoch": 1.02, + "learning_rate": 2.885819100698851e-05, + "loss": 3.9417, + "step": 1291264 + }, + { + "epoch": 1.02, + "learning_rate": 2.8849821438281722e-05, + "loss": 3.9476, + "step": 1291776 + }, + { + "epoch": 1.02, + "learning_rate": 2.8841435490771202e-05, + "loss": 3.9523, + "step": 1292288 + }, + { + "epoch": 1.02, + "learning_rate": 2.883306592206441e-05, + "loss": 3.9413, + "step": 1292800 + }, + { + "epoch": 1.02, + "learning_rate": 2.882467997455389e-05, + "loss": 3.9373, + "step": 1293312 + }, + { + "epoch": 1.02, + "learning_rate": 2.881629402704337e-05, + "loss": 3.9433, + "step": 1293824 + }, + { + "epoch": 1.02, + "learning_rate": 2.8807908079532854e-05, + "loss": 3.9499, + "step": 1294336 + }, + { + "epoch": 1.02, + "learning_rate": 2.8799522132022334e-05, + "loss": 3.937, + "step": 1294848 + }, + { + "epoch": 1.02, + "learning_rate": 2.8791136184511814e-05, + "loss": 3.9492, + "step": 1295360 + }, + { + "epoch": 1.02, + "learning_rate": 2.8782750237001294e-05, + "loss": 3.9422, + "step": 1295872 + }, + { + "epoch": 1.02, + "learning_rate": 2.8774380668294503e-05, + "loss": 3.952, + "step": 1296384 + }, + { + "epoch": 1.02, + "learning_rate": 2.8765994720783983e-05, + "loss": 3.949, + "step": 1296896 + }, + { + "epoch": 1.02, + "learning_rate": 2.8757608773273463e-05, + "loss": 3.9374, + "step": 1297408 + }, + { + "epoch": 1.03, + "eval_loss": 4.009064674377441, + "eval_runtime": 310.5769, + "eval_samples_per_second": 1228.652, + "eval_steps_per_second": 38.396, + "step": 1297427 + }, + { + "epoch": 0.0, + "learning_rate": 2.8749222825762946e-05, + "loss": 3.9344, + "step": 1297920 + }, + { + "epoch": 0.0, + "learning_rate": 2.8740836878252426e-05, + "loss": 3.9359, + "step": 1298432 + }, + { + "epoch": 0.0, + "learning_rate": 2.87324509307419e-05, + "loss": 3.9525, + "step": 1298944 + }, + { + "epoch": 0.0, + "learning_rate": 2.872406498323138e-05, + "loss": 3.9472, + "step": 1299456 + }, + { + "epoch": 0.0, + "learning_rate": 2.871567903572086e-05, + "loss": 3.9472, + "step": 1299968 + }, + { + "epoch": 0.0, + "learning_rate": 2.870729308821034e-05, + "loss": 3.9449, + "step": 1300480 + }, + { + "epoch": 0.0, + "learning_rate": 2.8698907140699823e-05, + "loss": 3.9383, + "step": 1300992 + }, + { + "epoch": 0.0, + "learning_rate": 2.8690521193189303e-05, + "loss": 3.9316, + "step": 1301504 + }, + { + "epoch": 0.0, + "learning_rate": 2.8682135245678783e-05, + "loss": 3.946, + "step": 1302016 + }, + { + "epoch": 0.0, + "learning_rate": 2.8673749298168263e-05, + "loss": 3.9347, + "step": 1302528 + }, + { + "epoch": 0.0, + "learning_rate": 2.8665363350657743e-05, + "loss": 3.9628, + "step": 1303040 + }, + { + "epoch": 0.0, + "learning_rate": 2.8656977403147222e-05, + "loss": 3.9502, + "step": 1303552 + }, + { + "epoch": 0.0, + "learning_rate": 2.8648591455636702e-05, + "loss": 3.9305, + "step": 1304064 + }, + { + "epoch": 0.0, + "learning_rate": 2.8640221886929915e-05, + "loss": 3.9358, + "step": 1304576 + }, + { + "epoch": 0.0, + "learning_rate": 2.8631835939419395e-05, + "loss": 3.9369, + "step": 1305088 + }, + { + "epoch": 0.0, + "learning_rate": 2.8623449991908875e-05, + "loss": 3.9353, + "step": 1305600 + }, + { + "epoch": 0.0, + "learning_rate": 2.8615064044398355e-05, + "loss": 3.9425, + "step": 1306112 + }, + { + "epoch": 0.0, + "learning_rate": 2.8606678096887835e-05, + "loss": 3.9455, + "step": 1306624 + }, + { + "epoch": 0.0, + "learning_rate": 2.8598292149377308e-05, + "loss": 3.9309, + "step": 1307136 + }, + { + "epoch": 0.0, + "learning_rate": 2.858990620186679e-05, + "loss": 3.9618, + "step": 1307648 + }, + { + "epoch": 0.0, + "learning_rate": 2.858152025435627e-05, + "loss": 3.9386, + "step": 1308160 + }, + { + "epoch": 0.0, + "learning_rate": 2.857313430684575e-05, + "loss": 3.946, + "step": 1308672 + }, + { + "epoch": 0.0, + "learning_rate": 2.856476473813896e-05, + "loss": 3.9379, + "step": 1309184 + }, + { + "epoch": 0.0, + "learning_rate": 2.855637879062844e-05, + "loss": 3.9571, + "step": 1309696 + }, + { + "epoch": 0.0, + "learning_rate": 2.854799284311792e-05, + "loss": 3.9374, + "step": 1310208 + }, + { + "epoch": 0.0, + "learning_rate": 2.85396068956074e-05, + "loss": 3.939, + "step": 1310720 + }, + { + "epoch": 0.0, + "learning_rate": 2.853122094809688e-05, + "loss": 3.9366, + "step": 1311232 + }, + { + "epoch": 0.0, + "learning_rate": 2.8522835000586363e-05, + "loss": 3.9373, + "step": 1311744 + }, + { + "epoch": 0.0, + "learning_rate": 2.8514449053075843e-05, + "loss": 3.9382, + "step": 1312256 + }, + { + "epoch": 0.01, + "learning_rate": 2.8506063105565323e-05, + "loss": 3.9394, + "step": 1312768 + }, + { + "epoch": 0.01, + "learning_rate": 2.8497693536858532e-05, + "loss": 3.9409, + "step": 1313280 + }, + { + "epoch": 0.01, + "learning_rate": 2.8489307589348012e-05, + "loss": 3.9421, + "step": 1313792 + }, + { + "epoch": 0.01, + "learning_rate": 2.8480921641837492e-05, + "loss": 3.9448, + "step": 1314304 + }, + { + "epoch": 0.01, + "learning_rate": 2.8472535694326972e-05, + "loss": 3.9496, + "step": 1314816 + }, + { + "epoch": 0.01, + "learning_rate": 2.8464166125620185e-05, + "loss": 3.9423, + "step": 1315328 + }, + { + "epoch": 0.01, + "learning_rate": 2.8455780178109665e-05, + "loss": 3.9468, + "step": 1315840 + }, + { + "epoch": 0.01, + "learning_rate": 2.8447394230599145e-05, + "loss": 3.9344, + "step": 1316352 + }, + { + "epoch": 0.01, + "learning_rate": 2.8439024661892354e-05, + "loss": 3.9383, + "step": 1316864 + }, + { + "epoch": 0.01, + "learning_rate": 2.8430638714381834e-05, + "loss": 3.9321, + "step": 1317376 + }, + { + "epoch": 0.01, + "learning_rate": 2.8422252766871317e-05, + "loss": 3.9398, + "step": 1317888 + }, + { + "epoch": 0.01, + "learning_rate": 2.8413866819360797e-05, + "loss": 3.9358, + "step": 1318400 + }, + { + "epoch": 0.01, + "learning_rate": 2.8405480871850277e-05, + "loss": 3.9414, + "step": 1318912 + }, + { + "epoch": 0.01, + "learning_rate": 2.8397094924339757e-05, + "loss": 3.933, + "step": 1319424 + }, + { + "epoch": 0.01, + "learning_rate": 2.8388708976829237e-05, + "loss": 3.9468, + "step": 1319936 + }, + { + "epoch": 0.01, + "learning_rate": 2.8380323029318717e-05, + "loss": 3.9396, + "step": 1320448 + }, + { + "epoch": 0.01, + "learning_rate": 2.8371937081808197e-05, + "loss": 3.9451, + "step": 1320960 + }, + { + "epoch": 0.01, + "learning_rate": 2.8363551134297673e-05, + "loss": 3.9415, + "step": 1321472 + }, + { + "epoch": 0.01, + "learning_rate": 2.8355165186787153e-05, + "loss": 3.9259, + "step": 1321984 + }, + { + "epoch": 0.01, + "learning_rate": 2.8346779239276633e-05, + "loss": 3.9328, + "step": 1322496 + }, + { + "epoch": 0.01, + "learning_rate": 2.833840967056985e-05, + "loss": 3.9511, + "step": 1323008 + }, + { + "epoch": 0.01, + "learning_rate": 2.8330023723059322e-05, + "loss": 3.9392, + "step": 1323520 + }, + { + "epoch": 0.01, + "learning_rate": 2.8321637775548802e-05, + "loss": 3.9324, + "step": 1324032 + }, + { + "epoch": 0.01, + "learning_rate": 2.8313251828038286e-05, + "loss": 3.9284, + "step": 1324544 + }, + { + "epoch": 0.01, + "learning_rate": 2.8304865880527766e-05, + "loss": 3.9377, + "step": 1325056 + }, + { + "epoch": 0.01, + "learning_rate": 2.8296496311820975e-05, + "loss": 3.9194, + "step": 1325568 + }, + { + "epoch": 0.01, + "learning_rate": 2.8288110364310455e-05, + "loss": 3.9489, + "step": 1326080 + }, + { + "epoch": 0.01, + "learning_rate": 2.8279724416799935e-05, + "loss": 3.919, + "step": 1326592 + }, + { + "epoch": 0.01, + "learning_rate": 2.8271338469289415e-05, + "loss": 3.944, + "step": 1327104 + }, + { + "epoch": 0.01, + "learning_rate": 2.8262968900582627e-05, + "loss": 3.9443, + "step": 1327616 + }, + { + "epoch": 0.01, + "learning_rate": 2.8254582953072107e-05, + "loss": 3.9265, + "step": 1328128 + }, + { + "epoch": 0.01, + "learning_rate": 2.8246213384365323e-05, + "loss": 3.9423, + "step": 1328640 + }, + { + "epoch": 0.01, + "learning_rate": 2.8237827436854796e-05, + "loss": 3.9301, + "step": 1329152 + }, + { + "epoch": 0.01, + "learning_rate": 2.8229441489344276e-05, + "loss": 3.9325, + "step": 1329664 + }, + { + "epoch": 0.01, + "learning_rate": 2.8221055541833756e-05, + "loss": 3.9232, + "step": 1330176 + }, + { + "epoch": 0.01, + "learning_rate": 2.821266959432324e-05, + "loss": 3.9454, + "step": 1330688 + }, + { + "epoch": 0.01, + "learning_rate": 2.820430002561645e-05, + "loss": 3.929, + "step": 1331200 + }, + { + "epoch": 0.01, + "learning_rate": 2.819591407810593e-05, + "loss": 3.9351, + "step": 1331712 + }, + { + "epoch": 0.01, + "learning_rate": 2.818752813059541e-05, + "loss": 3.9214, + "step": 1332224 + }, + { + "epoch": 0.01, + "learning_rate": 2.8179142183084888e-05, + "loss": 3.9303, + "step": 1332736 + }, + { + "epoch": 0.01, + "learning_rate": 2.8170756235574368e-05, + "loss": 3.9394, + "step": 1333248 + }, + { + "epoch": 0.01, + "learning_rate": 2.8162370288063848e-05, + "loss": 3.9423, + "step": 1333760 + }, + { + "epoch": 0.01, + "learning_rate": 2.815398434055333e-05, + "loss": 3.9389, + "step": 1334272 + }, + { + "epoch": 0.01, + "learning_rate": 2.814559839304281e-05, + "loss": 3.9429, + "step": 1334784 + }, + { + "epoch": 0.01, + "learning_rate": 2.813721244553229e-05, + "loss": 3.9448, + "step": 1335296 + }, + { + "epoch": 0.01, + "learning_rate": 2.81288428768255e-05, + "loss": 3.9347, + "step": 1335808 + }, + { + "epoch": 0.01, + "learning_rate": 2.812045692931498e-05, + "loss": 3.9469, + "step": 1336320 + }, + { + "epoch": 0.01, + "learning_rate": 2.8112087360608193e-05, + "loss": 3.9309, + "step": 1336832 + }, + { + "epoch": 0.01, + "learning_rate": 2.8103701413097673e-05, + "loss": 3.9362, + "step": 1337344 + }, + { + "epoch": 0.01, + "learning_rate": 2.8095315465587153e-05, + "loss": 3.9321, + "step": 1337856 + }, + { + "epoch": 0.01, + "learning_rate": 2.8086929518076633e-05, + "loss": 3.9422, + "step": 1338368 + }, + { + "epoch": 0.01, + "learning_rate": 2.8078543570566113e-05, + "loss": 3.9416, + "step": 1338880 + }, + { + "epoch": 0.01, + "learning_rate": 2.8070174001859322e-05, + "loss": 3.9493, + "step": 1339392 + }, + { + "epoch": 0.01, + "learning_rate": 2.8061788054348802e-05, + "loss": 3.9398, + "step": 1339904 + }, + { + "epoch": 0.01, + "learning_rate": 2.8053402106838282e-05, + "loss": 3.9174, + "step": 1340416 + }, + { + "epoch": 0.01, + "learning_rate": 2.8045016159327765e-05, + "loss": 3.9242, + "step": 1340928 + }, + { + "epoch": 0.01, + "learning_rate": 2.8036630211817245e-05, + "loss": 3.9378, + "step": 1341440 + }, + { + "epoch": 0.01, + "learning_rate": 2.8028244264306725e-05, + "loss": 3.9322, + "step": 1341952 + }, + { + "epoch": 0.01, + "learning_rate": 2.8019858316796205e-05, + "loss": 3.9323, + "step": 1342464 + }, + { + "epoch": 0.01, + "learning_rate": 2.8011472369285685e-05, + "loss": 3.94, + "step": 1342976 + }, + { + "epoch": 0.02, + "learning_rate": 2.8003119179382627e-05, + "loss": 3.928, + "step": 1343488 + }, + { + "epoch": 0.02, + "learning_rate": 2.7994733231872107e-05, + "loss": 3.9354, + "step": 1344000 + }, + { + "epoch": 0.02, + "learning_rate": 2.7986347284361587e-05, + "loss": 3.9336, + "step": 1344512 + }, + { + "epoch": 0.02, + "learning_rate": 2.7977961336851066e-05, + "loss": 3.9424, + "step": 1345024 + }, + { + "epoch": 0.02, + "learning_rate": 2.7969575389340546e-05, + "loss": 3.9419, + "step": 1345536 + }, + { + "epoch": 0.02, + "learning_rate": 2.7961189441830026e-05, + "loss": 3.9474, + "step": 1346048 + }, + { + "epoch": 0.02, + "learning_rate": 2.7952803494319506e-05, + "loss": 3.9203, + "step": 1346560 + }, + { + "epoch": 0.02, + "learning_rate": 2.7944417546808983e-05, + "loss": 3.9359, + "step": 1347072 + }, + { + "epoch": 0.02, + "learning_rate": 2.79360479781022e-05, + "loss": 3.9332, + "step": 1347584 + }, + { + "epoch": 0.02, + "learning_rate": 2.792766203059168e-05, + "loss": 3.9361, + "step": 1348096 + }, + { + "epoch": 0.02, + "learning_rate": 2.791927608308116e-05, + "loss": 3.9494, + "step": 1348608 + }, + { + "epoch": 0.02, + "learning_rate": 2.7910890135570632e-05, + "loss": 3.9237, + "step": 1349120 + }, + { + "epoch": 0.02, + "learning_rate": 2.7902520566863848e-05, + "loss": 3.9339, + "step": 1349632 + }, + { + "epoch": 0.02, + "learning_rate": 2.7894134619353328e-05, + "loss": 3.9389, + "step": 1350144 + }, + { + "epoch": 0.02, + "learning_rate": 2.788576505064654e-05, + "loss": 3.9403, + "step": 1350656 + }, + { + "epoch": 0.02, + "learning_rate": 2.787737910313602e-05, + "loss": 3.9249, + "step": 1351168 + }, + { + "epoch": 0.02, + "learning_rate": 2.78689931556255e-05, + "loss": 3.9316, + "step": 1351680 + }, + { + "epoch": 0.02, + "learning_rate": 2.786060720811498e-05, + "loss": 3.9258, + "step": 1352192 + }, + { + "epoch": 0.02, + "learning_rate": 2.7852221260604457e-05, + "loss": 3.928, + "step": 1352704 + }, + { + "epoch": 0.02, + "learning_rate": 2.7843835313093937e-05, + "loss": 3.9305, + "step": 1353216 + }, + { + "epoch": 0.02, + "learning_rate": 2.7835449365583417e-05, + "loss": 3.9335, + "step": 1353728 + }, + { + "epoch": 0.02, + "learning_rate": 2.7827063418072897e-05, + "loss": 3.9242, + "step": 1354240 + }, + { + "epoch": 0.02, + "learning_rate": 2.7818677470562376e-05, + "loss": 3.9354, + "step": 1354752 + }, + { + "epoch": 0.02, + "learning_rate": 2.7810307901855586e-05, + "loss": 3.9283, + "step": 1355264 + }, + { + "epoch": 0.02, + "learning_rate": 2.7801921954345066e-05, + "loss": 3.933, + "step": 1355776 + }, + { + "epoch": 0.02, + "learning_rate": 2.7793552385638278e-05, + "loss": 3.948, + "step": 1356288 + }, + { + "epoch": 0.02, + "learning_rate": 2.7785166438127758e-05, + "loss": 3.9422, + "step": 1356800 + }, + { + "epoch": 0.02, + "learning_rate": 2.7776780490617238e-05, + "loss": 3.936, + "step": 1357312 + }, + { + "epoch": 0.02, + "learning_rate": 2.7768394543106718e-05, + "loss": 3.933, + "step": 1357824 + }, + { + "epoch": 0.02, + "learning_rate": 2.7760008595596198e-05, + "loss": 3.9347, + "step": 1358336 + }, + { + "epoch": 0.02, + "learning_rate": 2.7751622648085678e-05, + "loss": 3.934, + "step": 1358848 + }, + { + "epoch": 0.02, + "learning_rate": 2.7743236700575158e-05, + "loss": 3.925, + "step": 1359360 + }, + { + "epoch": 0.02, + "learning_rate": 2.773485075306464e-05, + "loss": 3.933, + "step": 1359872 + }, + { + "epoch": 0.02, + "learning_rate": 2.772648118435785e-05, + "loss": 3.9337, + "step": 1360384 + }, + { + "epoch": 0.02, + "learning_rate": 2.771811161565106e-05, + "loss": 3.9364, + "step": 1360896 + }, + { + "epoch": 0.02, + "learning_rate": 2.770972566814054e-05, + "loss": 3.9348, + "step": 1361408 + }, + { + "epoch": 0.02, + "learning_rate": 2.770133972063002e-05, + "loss": 3.9367, + "step": 1361920 + }, + { + "epoch": 0.02, + "learning_rate": 2.7692953773119503e-05, + "loss": 3.9322, + "step": 1362432 + }, + { + "epoch": 0.02, + "learning_rate": 2.7684567825608982e-05, + "loss": 3.9332, + "step": 1362944 + }, + { + "epoch": 0.02, + "learning_rate": 2.7676181878098462e-05, + "loss": 3.9254, + "step": 1363456 + }, + { + "epoch": 0.02, + "learning_rate": 2.766781230939167e-05, + "loss": 3.9492, + "step": 1363968 + }, + { + "epoch": 0.02, + "learning_rate": 2.765942636188115e-05, + "loss": 3.94, + "step": 1364480 + }, + { + "epoch": 0.02, + "learning_rate": 2.765104041437063e-05, + "loss": 3.9388, + "step": 1364992 + }, + { + "epoch": 0.02, + "learning_rate": 2.764265446686011e-05, + "loss": 3.9355, + "step": 1365504 + }, + { + "epoch": 0.02, + "learning_rate": 2.7634268519349595e-05, + "loss": 3.9378, + "step": 1366016 + }, + { + "epoch": 0.02, + "learning_rate": 2.7625898950642804e-05, + "loss": 3.9415, + "step": 1366528 + }, + { + "epoch": 0.02, + "learning_rate": 2.7617513003132284e-05, + "loss": 3.9438, + "step": 1367040 + }, + { + "epoch": 0.02, + "learning_rate": 2.7609127055621764e-05, + "loss": 3.9325, + "step": 1367552 + }, + { + "epoch": 0.02, + "learning_rate": 2.7600741108111244e-05, + "loss": 3.9371, + "step": 1368064 + }, + { + "epoch": 0.02, + "learning_rate": 2.7592355160600724e-05, + "loss": 3.9355, + "step": 1368576 + }, + { + "epoch": 0.02, + "learning_rate": 2.7583969213090204e-05, + "loss": 3.9343, + "step": 1369088 + }, + { + "epoch": 0.02, + "learning_rate": 2.7575599644383416e-05, + "loss": 3.9317, + "step": 1369600 + }, + { + "epoch": 0.02, + "learning_rate": 2.7567213696872896e-05, + "loss": 3.934, + "step": 1370112 + }, + { + "epoch": 0.02, + "learning_rate": 2.7558827749362376e-05, + "loss": 3.9395, + "step": 1370624 + }, + { + "epoch": 0.02, + "learning_rate": 2.7550441801851856e-05, + "loss": 3.9259, + "step": 1371136 + }, + { + "epoch": 0.02, + "learning_rate": 2.7542055854341336e-05, + "loss": 3.9424, + "step": 1371648 + }, + { + "epoch": 0.02, + "learning_rate": 2.7533669906830816e-05, + "loss": 3.9346, + "step": 1372160 + }, + { + "epoch": 0.02, + "learning_rate": 2.7525283959320292e-05, + "loss": 3.9384, + "step": 1372672 + }, + { + "epoch": 0.02, + "learning_rate": 2.7516898011809772e-05, + "loss": 3.9381, + "step": 1373184 + }, + { + "epoch": 0.02, + "learning_rate": 2.7508512064299252e-05, + "loss": 3.9288, + "step": 1373696 + }, + { + "epoch": 0.03, + "eval_loss": 4.004461288452148, + "eval_runtime": 295.3464, + "eval_samples_per_second": 1292.012, + "eval_steps_per_second": 40.376, + "step": 1373747 + }, + { + "epoch": 1.0, + "learning_rate": 2.750014249559246e-05, + "loss": 3.9285, + "step": 1374208 + }, + { + "epoch": 1.0, + "learning_rate": 2.749175654808194e-05, + "loss": 3.9246, + "step": 1374720 + }, + { + "epoch": 1.0, + "learning_rate": 2.7483370600571425e-05, + "loss": 3.9447, + "step": 1375232 + }, + { + "epoch": 1.0, + "learning_rate": 2.7474984653060905e-05, + "loss": 3.9323, + "step": 1375744 + }, + { + "epoch": 1.0, + "learning_rate": 2.7466615084354114e-05, + "loss": 3.9431, + "step": 1376256 + }, + { + "epoch": 1.0, + "learning_rate": 2.7458229136843594e-05, + "loss": 3.9302, + "step": 1376768 + }, + { + "epoch": 1.0, + "learning_rate": 2.7449843189333074e-05, + "loss": 3.9304, + "step": 1377280 + }, + { + "epoch": 1.0, + "learning_rate": 2.744147362062629e-05, + "loss": 3.9199, + "step": 1377792 + }, + { + "epoch": 1.0, + "learning_rate": 2.7433087673115766e-05, + "loss": 3.9351, + "step": 1378304 + }, + { + "epoch": 1.0, + "learning_rate": 2.7424701725605246e-05, + "loss": 3.9317, + "step": 1378816 + }, + { + "epoch": 1.0, + "learning_rate": 2.7416315778094726e-05, + "loss": 3.9509, + "step": 1379328 + }, + { + "epoch": 1.0, + "learning_rate": 2.7407929830584206e-05, + "loss": 3.9389, + "step": 1379840 + }, + { + "epoch": 1.0, + "learning_rate": 2.739957664068115e-05, + "loss": 3.9212, + "step": 1380352 + }, + { + "epoch": 1.0, + "learning_rate": 2.739119069317063e-05, + "loss": 3.9263, + "step": 1380864 + }, + { + "epoch": 1.0, + "learning_rate": 2.738280474566011e-05, + "loss": 3.9361, + "step": 1381376 + }, + { + "epoch": 1.0, + "learning_rate": 2.7374418798149588e-05, + "loss": 3.9158, + "step": 1381888 + }, + { + "epoch": 1.0, + "learning_rate": 2.7366032850639067e-05, + "loss": 3.9366, + "step": 1382400 + }, + { + "epoch": 1.0, + "learning_rate": 2.7357646903128547e-05, + "loss": 3.9319, + "step": 1382912 + }, + { + "epoch": 1.0, + "learning_rate": 2.7349260955618027e-05, + "loss": 3.9232, + "step": 1383424 + }, + { + "epoch": 1.0, + "learning_rate": 2.7340875008107507e-05, + "loss": 3.9533, + "step": 1383936 + }, + { + "epoch": 1.0, + "learning_rate": 2.733250543940072e-05, + "loss": 3.9303, + "step": 1384448 + }, + { + "epoch": 1.0, + "learning_rate": 2.73241194918902e-05, + "loss": 3.9371, + "step": 1384960 + }, + { + "epoch": 1.0, + "learning_rate": 2.731573354437968e-05, + "loss": 3.9301, + "step": 1385472 + }, + { + "epoch": 1.0, + "learning_rate": 2.730734759686916e-05, + "loss": 3.944, + "step": 1385984 + }, + { + "epoch": 1.0, + "learning_rate": 2.729896164935864e-05, + "loss": 3.9232, + "step": 1386496 + }, + { + "epoch": 1.0, + "learning_rate": 2.729057570184812e-05, + "loss": 3.9319, + "step": 1387008 + }, + { + "epoch": 1.0, + "learning_rate": 2.72821897543376e-05, + "loss": 3.9238, + "step": 1387520 + }, + { + "epoch": 1.0, + "learning_rate": 2.727380380682708e-05, + "loss": 3.9318, + "step": 1388032 + }, + { + "epoch": 1.0, + "learning_rate": 2.726541785931656e-05, + "loss": 3.9305, + "step": 1388544 + }, + { + "epoch": 1.01, + "learning_rate": 2.7257048290609772e-05, + "loss": 3.93, + "step": 1389056 + }, + { + "epoch": 1.01, + "learning_rate": 2.724867872190298e-05, + "loss": 3.9324, + "step": 1389568 + }, + { + "epoch": 1.01, + "learning_rate": 2.724029277439246e-05, + "loss": 3.9358, + "step": 1390080 + }, + { + "epoch": 1.01, + "learning_rate": 2.723190682688194e-05, + "loss": 3.9288, + "step": 1390592 + }, + { + "epoch": 1.01, + "learning_rate": 2.722352087937142e-05, + "loss": 3.944, + "step": 1391104 + }, + { + "epoch": 1.01, + "learning_rate": 2.7215167689468363e-05, + "loss": 3.9311, + "step": 1391616 + }, + { + "epoch": 1.01, + "learning_rate": 2.7206781741957842e-05, + "loss": 3.9342, + "step": 1392128 + }, + { + "epoch": 1.01, + "learning_rate": 2.7198395794447322e-05, + "loss": 3.9272, + "step": 1392640 + }, + { + "epoch": 1.01, + "learning_rate": 2.7190009846936802e-05, + "loss": 3.9294, + "step": 1393152 + }, + { + "epoch": 1.01, + "learning_rate": 2.7181623899426282e-05, + "loss": 3.9232, + "step": 1393664 + }, + { + "epoch": 1.01, + "learning_rate": 2.7173237951915766e-05, + "loss": 3.9294, + "step": 1394176 + }, + { + "epoch": 1.01, + "learning_rate": 2.7164852004405246e-05, + "loss": 3.9258, + "step": 1394688 + }, + { + "epoch": 1.01, + "learning_rate": 2.7156466056894726e-05, + "loss": 3.9374, + "step": 1395200 + }, + { + "epoch": 1.01, + "learning_rate": 2.7148080109384206e-05, + "loss": 3.9225, + "step": 1395712 + }, + { + "epoch": 1.01, + "learning_rate": 2.7139694161873686e-05, + "loss": 3.9342, + "step": 1396224 + }, + { + "epoch": 1.01, + "learning_rate": 2.7131308214363165e-05, + "loss": 3.9343, + "step": 1396736 + }, + { + "epoch": 1.01, + "learning_rate": 2.7122922266852645e-05, + "loss": 3.9331, + "step": 1397248 + }, + { + "epoch": 1.01, + "learning_rate": 2.7114569076949587e-05, + "loss": 3.9344, + "step": 1397760 + }, + { + "epoch": 1.01, + "learning_rate": 2.7106183129439067e-05, + "loss": 3.9153, + "step": 1398272 + }, + { + "epoch": 1.01, + "learning_rate": 2.7097797181928547e-05, + "loss": 3.9241, + "step": 1398784 + }, + { + "epoch": 1.01, + "learning_rate": 2.7089411234418027e-05, + "loss": 3.9388, + "step": 1399296 + }, + { + "epoch": 1.01, + "learning_rate": 2.7081025286907507e-05, + "loss": 3.9297, + "step": 1399808 + }, + { + "epoch": 1.01, + "learning_rate": 2.7072639339396987e-05, + "loss": 3.9275, + "step": 1400320 + }, + { + "epoch": 1.01, + "learning_rate": 2.7064253391886467e-05, + "loss": 3.9161, + "step": 1400832 + }, + { + "epoch": 1.01, + "learning_rate": 2.705588382317968e-05, + "loss": 3.9307, + "step": 1401344 + }, + { + "epoch": 1.01, + "learning_rate": 2.704749787566916e-05, + "loss": 3.9089, + "step": 1401856 + }, + { + "epoch": 1.01, + "learning_rate": 2.703911192815864e-05, + "loss": 3.9421, + "step": 1402368 + }, + { + "epoch": 1.01, + "learning_rate": 2.703072598064812e-05, + "loss": 3.9076, + "step": 1402880 + }, + { + "epoch": 1.01, + "learning_rate": 2.7022356411941328e-05, + "loss": 3.9398, + "step": 1403392 + }, + { + "epoch": 1.01, + "learning_rate": 2.701397046443081e-05, + "loss": 3.9291, + "step": 1403904 + }, + { + "epoch": 1.01, + "learning_rate": 2.700558451692029e-05, + "loss": 3.9204, + "step": 1404416 + }, + { + "epoch": 1.01, + "learning_rate": 2.699719856940977e-05, + "loss": 3.9343, + "step": 1404928 + }, + { + "epoch": 1.01, + "learning_rate": 2.6988812621899245e-05, + "loss": 3.9217, + "step": 1405440 + }, + { + "epoch": 1.01, + "learning_rate": 2.6980426674388725e-05, + "loss": 3.9214, + "step": 1405952 + }, + { + "epoch": 1.01, + "learning_rate": 2.6972040726878205e-05, + "loss": 3.9159, + "step": 1406464 + }, + { + "epoch": 1.01, + "learning_rate": 2.6963654779367688e-05, + "loss": 3.9376, + "step": 1406976 + }, + { + "epoch": 1.01, + "learning_rate": 2.6955268831857168e-05, + "loss": 3.92, + "step": 1407488 + }, + { + "epoch": 1.01, + "learning_rate": 2.6946882884346648e-05, + "loss": 3.9289, + "step": 1408000 + }, + { + "epoch": 1.01, + "learning_rate": 2.6938496936836128e-05, + "loss": 3.9106, + "step": 1408512 + }, + { + "epoch": 1.01, + "learning_rate": 2.6930110989325608e-05, + "loss": 3.9196, + "step": 1409024 + }, + { + "epoch": 1.01, + "learning_rate": 2.692175779942255e-05, + "loss": 3.9305, + "step": 1409536 + }, + { + "epoch": 1.01, + "learning_rate": 2.691337185191203e-05, + "loss": 3.9349, + "step": 1410048 + }, + { + "epoch": 1.01, + "learning_rate": 2.690498590440151e-05, + "loss": 3.9257, + "step": 1410560 + }, + { + "epoch": 1.01, + "learning_rate": 2.689659995689099e-05, + "loss": 3.935, + "step": 1411072 + }, + { + "epoch": 1.01, + "learning_rate": 2.688821400938047e-05, + "loss": 3.9351, + "step": 1411584 + }, + { + "epoch": 1.01, + "learning_rate": 2.687982806186995e-05, + "loss": 3.9292, + "step": 1412096 + }, + { + "epoch": 1.01, + "learning_rate": 2.6871458493163158e-05, + "loss": 3.9333, + "step": 1412608 + }, + { + "epoch": 1.01, + "learning_rate": 2.686307254565264e-05, + "loss": 3.9262, + "step": 1413120 + }, + { + "epoch": 1.01, + "learning_rate": 2.685468659814212e-05, + "loss": 3.9256, + "step": 1413632 + }, + { + "epoch": 1.01, + "learning_rate": 2.68463006506316e-05, + "loss": 3.9253, + "step": 1414144 + }, + { + "epoch": 1.01, + "learning_rate": 2.683791470312108e-05, + "loss": 3.9314, + "step": 1414656 + }, + { + "epoch": 1.01, + "learning_rate": 2.682954513441429e-05, + "loss": 3.9331, + "step": 1415168 + }, + { + "epoch": 1.01, + "learning_rate": 2.682115918690377e-05, + "loss": 3.9424, + "step": 1415680 + }, + { + "epoch": 1.01, + "learning_rate": 2.681277323939325e-05, + "loss": 3.9284, + "step": 1416192 + }, + { + "epoch": 1.01, + "learning_rate": 2.680438729188273e-05, + "loss": 3.9079, + "step": 1416704 + }, + { + "epoch": 1.01, + "learning_rate": 2.6796001344372214e-05, + "loss": 3.9144, + "step": 1417216 + }, + { + "epoch": 1.01, + "learning_rate": 2.6787615396861694e-05, + "loss": 3.9315, + "step": 1417728 + }, + { + "epoch": 1.01, + "learning_rate": 2.6779245828154903e-05, + "loss": 3.9231, + "step": 1418240 + }, + { + "epoch": 1.01, + "learning_rate": 2.6770859880644383e-05, + "loss": 3.9235, + "step": 1418752 + }, + { + "epoch": 1.01, + "learning_rate": 2.6762473933133863e-05, + "loss": 3.9294, + "step": 1419264 + }, + { + "epoch": 1.02, + "learning_rate": 2.6754087985623343e-05, + "loss": 3.9229, + "step": 1419776 + }, + { + "epoch": 1.02, + "learning_rate": 2.6745702038112823e-05, + "loss": 3.9192, + "step": 1420288 + }, + { + "epoch": 1.02, + "learning_rate": 2.6737316090602306e-05, + "loss": 3.9245, + "step": 1420800 + }, + { + "epoch": 1.02, + "learning_rate": 2.6728930143091786e-05, + "loss": 3.9402, + "step": 1421312 + }, + { + "epoch": 1.02, + "learning_rate": 2.672054419558126e-05, + "loss": 3.9309, + "step": 1421824 + }, + { + "epoch": 1.02, + "learning_rate": 2.671215824807074e-05, + "loss": 3.9364, + "step": 1422336 + }, + { + "epoch": 1.02, + "learning_rate": 2.6703788679363955e-05, + "loss": 3.9178, + "step": 1422848 + }, + { + "epoch": 1.02, + "learning_rate": 2.669540273185343e-05, + "loss": 3.9271, + "step": 1423360 + }, + { + "epoch": 1.02, + "learning_rate": 2.668701678434291e-05, + "loss": 3.9265, + "step": 1423872 + }, + { + "epoch": 1.02, + "learning_rate": 2.6678647215636127e-05, + "loss": 3.9273, + "step": 1424384 + }, + { + "epoch": 1.02, + "learning_rate": 2.6670261268125607e-05, + "loss": 3.9387, + "step": 1424896 + }, + { + "epoch": 1.02, + "learning_rate": 2.666187532061508e-05, + "loss": 3.9193, + "step": 1425408 + }, + { + "epoch": 1.02, + "learning_rate": 2.665348937310456e-05, + "loss": 3.9236, + "step": 1425920 + }, + { + "epoch": 1.02, + "learning_rate": 2.6645103425594044e-05, + "loss": 3.9311, + "step": 1426432 + }, + { + "epoch": 1.02, + "learning_rate": 2.6636717478083524e-05, + "loss": 3.9331, + "step": 1426944 + }, + { + "epoch": 1.02, + "learning_rate": 2.6628347909376733e-05, + "loss": 3.9159, + "step": 1427456 + }, + { + "epoch": 1.02, + "learning_rate": 2.661997834066995e-05, + "loss": 3.9233, + "step": 1427968 + }, + { + "epoch": 1.02, + "learning_rate": 2.661159239315943e-05, + "loss": 3.9209, + "step": 1428480 + }, + { + "epoch": 1.02, + "learning_rate": 2.6603206445648905e-05, + "loss": 3.9216, + "step": 1428992 + }, + { + "epoch": 1.02, + "learning_rate": 2.6594820498138385e-05, + "loss": 3.9187, + "step": 1429504 + }, + { + "epoch": 1.02, + "learning_rate": 2.6586434550627865e-05, + "loss": 3.9218, + "step": 1430016 + }, + { + "epoch": 1.02, + "learning_rate": 2.6578048603117345e-05, + "loss": 3.9199, + "step": 1430528 + }, + { + "epoch": 1.02, + "learning_rate": 2.6569662655606825e-05, + "loss": 3.9279, + "step": 1431040 + }, + { + "epoch": 1.02, + "learning_rate": 2.6561276708096305e-05, + "loss": 3.9169, + "step": 1431552 + }, + { + "epoch": 1.02, + "learning_rate": 2.6552890760585785e-05, + "loss": 3.9239, + "step": 1432064 + }, + { + "epoch": 1.02, + "learning_rate": 2.6544504813075265e-05, + "loss": 3.9388, + "step": 1432576 + }, + { + "epoch": 1.02, + "learning_rate": 2.6536118865564745e-05, + "loss": 3.9355, + "step": 1433088 + }, + { + "epoch": 1.02, + "learning_rate": 2.6527749296857957e-05, + "loss": 3.9292, + "step": 1433600 + }, + { + "epoch": 1.02, + "learning_rate": 2.6519363349347437e-05, + "loss": 3.9284, + "step": 1434112 + }, + { + "epoch": 1.02, + "learning_rate": 2.6510977401836917e-05, + "loss": 3.9253, + "step": 1434624 + }, + { + "epoch": 1.02, + "learning_rate": 2.6502591454326397e-05, + "loss": 3.9232, + "step": 1435136 + }, + { + "epoch": 1.02, + "learning_rate": 2.6494205506815877e-05, + "loss": 3.9212, + "step": 1435648 + }, + { + "epoch": 1.02, + "learning_rate": 2.648583593810909e-05, + "loss": 3.92, + "step": 1436160 + }, + { + "epoch": 1.02, + "learning_rate": 2.647744999059857e-05, + "loss": 3.9216, + "step": 1436672 + }, + { + "epoch": 1.02, + "learning_rate": 2.646906404308805e-05, + "loss": 3.9327, + "step": 1437184 + }, + { + "epoch": 1.02, + "learning_rate": 2.646067809557753e-05, + "loss": 3.9243, + "step": 1437696 + }, + { + "epoch": 1.02, + "learning_rate": 2.645229214806701e-05, + "loss": 3.9262, + "step": 1438208 + }, + { + "epoch": 1.02, + "learning_rate": 2.644393895816395e-05, + "loss": 3.9269, + "step": 1438720 + }, + { + "epoch": 1.02, + "learning_rate": 2.643555301065343e-05, + "loss": 3.9204, + "step": 1439232 + }, + { + "epoch": 1.02, + "learning_rate": 2.642716706314291e-05, + "loss": 3.9172, + "step": 1439744 + }, + { + "epoch": 1.02, + "learning_rate": 2.641878111563239e-05, + "loss": 3.9414, + "step": 1440256 + }, + { + "epoch": 1.02, + "learning_rate": 2.641039516812187e-05, + "loss": 3.934, + "step": 1440768 + }, + { + "epoch": 1.02, + "learning_rate": 2.640200922061135e-05, + "loss": 3.9331, + "step": 1441280 + }, + { + "epoch": 1.02, + "learning_rate": 2.639363965190456e-05, + "loss": 3.9258, + "step": 1441792 + }, + { + "epoch": 1.02, + "learning_rate": 2.6385253704394043e-05, + "loss": 3.9257, + "step": 1442304 + }, + { + "epoch": 1.02, + "learning_rate": 2.6376867756883523e-05, + "loss": 3.9338, + "step": 1442816 + }, + { + "epoch": 1.02, + "learning_rate": 2.6368481809373003e-05, + "loss": 3.9398, + "step": 1443328 + }, + { + "epoch": 1.02, + "learning_rate": 2.6360095861862483e-05, + "loss": 3.9272, + "step": 1443840 + }, + { + "epoch": 1.02, + "learning_rate": 2.6351709914351963e-05, + "loss": 3.9231, + "step": 1444352 + }, + { + "epoch": 1.02, + "learning_rate": 2.6343323966841443e-05, + "loss": 3.9292, + "step": 1444864 + }, + { + "epoch": 1.02, + "learning_rate": 2.633493801933092e-05, + "loss": 3.9263, + "step": 1445376 + }, + { + "epoch": 1.02, + "learning_rate": 2.63265520718204e-05, + "loss": 3.9234, + "step": 1445888 + }, + { + "epoch": 1.02, + "learning_rate": 2.631816612430988e-05, + "loss": 3.9261, + "step": 1446400 + }, + { + "epoch": 1.02, + "learning_rate": 2.630978017679936e-05, + "loss": 3.932, + "step": 1446912 + }, + { + "epoch": 1.02, + "learning_rate": 2.630141060809257e-05, + "loss": 3.9181, + "step": 1447424 + }, + { + "epoch": 1.02, + "learning_rate": 2.6293041039385785e-05, + "loss": 3.9329, + "step": 1447936 + }, + { + "epoch": 1.02, + "learning_rate": 2.6284655091875264e-05, + "loss": 3.9293, + "step": 1448448 + }, + { + "epoch": 1.02, + "learning_rate": 2.627626914436474e-05, + "loss": 3.927, + "step": 1448960 + }, + { + "epoch": 1.02, + "learning_rate": 2.626788319685422e-05, + "loss": 3.9321, + "step": 1449472 + }, + { + "epoch": 1.02, + "learning_rate": 2.62594972493437e-05, + "loss": 3.9178, + "step": 1449984 + }, + { + "epoch": 1.03, + "eval_loss": 4.000810623168945, + "eval_runtime": 303.8438, + "eval_samples_per_second": 1255.879, + "eval_steps_per_second": 39.247, + "step": 1450067 + }, + { + "epoch": 0.0, + "learning_rate": 2.625111130183318e-05, + "loss": 3.9179, + "step": 1450496 + }, + { + "epoch": 0.0, + "learning_rate": 2.624272535432266e-05, + "loss": 3.9156, + "step": 1451008 + }, + { + "epoch": 0.0, + "learning_rate": 2.623433940681214e-05, + "loss": 3.9378, + "step": 1451520 + }, + { + "epoch": 0.0, + "learning_rate": 2.622595345930162e-05, + "loss": 3.9199, + "step": 1452032 + }, + { + "epoch": 0.0, + "learning_rate": 2.62175675117911e-05, + "loss": 3.9379, + "step": 1452544 + }, + { + "epoch": 0.0, + "learning_rate": 2.6209181564280584e-05, + "loss": 3.9211, + "step": 1453056 + }, + { + "epoch": 0.0, + "learning_rate": 2.6200795616770064e-05, + "loss": 3.9253, + "step": 1453568 + }, + { + "epoch": 0.0, + "learning_rate": 2.6192409669259544e-05, + "loss": 3.914, + "step": 1454080 + }, + { + "epoch": 0.0, + "learning_rate": 2.6184023721749024e-05, + "loss": 3.9309, + "step": 1454592 + }, + { + "epoch": 0.0, + "learning_rate": 2.6175637774238504e-05, + "loss": 3.9182, + "step": 1455104 + }, + { + "epoch": 0.0, + "learning_rate": 2.6167251826727977e-05, + "loss": 3.9427, + "step": 1455616 + }, + { + "epoch": 0.0, + "learning_rate": 2.615886587921746e-05, + "loss": 3.9297, + "step": 1456128 + }, + { + "epoch": 0.0, + "learning_rate": 2.6150496310510676e-05, + "loss": 3.9109, + "step": 1456640 + }, + { + "epoch": 0.0, + "learning_rate": 2.6142110363000156e-05, + "loss": 3.9167, + "step": 1457152 + }, + { + "epoch": 0.0, + "learning_rate": 2.613372441548963e-05, + "loss": 3.9283, + "step": 1457664 + }, + { + "epoch": 0.0, + "learning_rate": 2.612533846797911e-05, + "loss": 3.9125, + "step": 1458176 + }, + { + "epoch": 0.0, + "learning_rate": 2.611695252046859e-05, + "loss": 3.9259, + "step": 1458688 + }, + { + "epoch": 0.0, + "learning_rate": 2.610856657295807e-05, + "loss": 3.923, + "step": 1459200 + }, + { + "epoch": 0.0, + "learning_rate": 2.6100180625447553e-05, + "loss": 3.9179, + "step": 1459712 + }, + { + "epoch": 0.0, + "learning_rate": 2.6091794677937033e-05, + "loss": 3.9479, + "step": 1460224 + }, + { + "epoch": 0.0, + "learning_rate": 2.6083408730426513e-05, + "loss": 3.9244, + "step": 1460736 + }, + { + "epoch": 0.0, + "learning_rate": 2.607503916171972e-05, + "loss": 3.9274, + "step": 1461248 + }, + { + "epoch": 0.0, + "learning_rate": 2.60666532142092e-05, + "loss": 3.9231, + "step": 1461760 + }, + { + "epoch": 0.0, + "learning_rate": 2.605826726669868e-05, + "loss": 3.9383, + "step": 1462272 + }, + { + "epoch": 0.0, + "learning_rate": 2.604988131918816e-05, + "loss": 3.9165, + "step": 1462784 + }, + { + "epoch": 0.0, + "learning_rate": 2.6041511750481374e-05, + "loss": 3.9198, + "step": 1463296 + }, + { + "epoch": 0.0, + "learning_rate": 2.6033125802970854e-05, + "loss": 3.9188, + "step": 1463808 + }, + { + "epoch": 0.0, + "learning_rate": 2.6024739855460334e-05, + "loss": 3.9239, + "step": 1464320 + }, + { + "epoch": 0.0, + "learning_rate": 2.6016353907949814e-05, + "loss": 3.9202, + "step": 1464832 + }, + { + "epoch": 0.01, + "learning_rate": 2.6007967960439294e-05, + "loss": 3.9194, + "step": 1465344 + }, + { + "epoch": 0.01, + "learning_rate": 2.5999582012928774e-05, + "loss": 3.9262, + "step": 1465856 + }, + { + "epoch": 0.01, + "learning_rate": 2.5991196065418254e-05, + "loss": 3.9241, + "step": 1466368 + }, + { + "epoch": 0.01, + "learning_rate": 2.5982810117907737e-05, + "loss": 3.9238, + "step": 1466880 + }, + { + "epoch": 0.01, + "learning_rate": 2.5974456928004675e-05, + "loss": 3.9344, + "step": 1467392 + }, + { + "epoch": 0.01, + "learning_rate": 2.5966087359297884e-05, + "loss": 3.9235, + "step": 1467904 + }, + { + "epoch": 0.01, + "learning_rate": 2.5957701411787368e-05, + "loss": 3.9265, + "step": 1468416 + }, + { + "epoch": 0.01, + "learning_rate": 2.5949315464276848e-05, + "loss": 3.9139, + "step": 1468928 + }, + { + "epoch": 0.01, + "learning_rate": 2.5940929516766328e-05, + "loss": 3.926, + "step": 1469440 + }, + { + "epoch": 0.01, + "learning_rate": 2.5932543569255808e-05, + "loss": 3.9125, + "step": 1469952 + }, + { + "epoch": 0.01, + "learning_rate": 2.5924157621745288e-05, + "loss": 3.9227, + "step": 1470464 + }, + { + "epoch": 0.01, + "learning_rate": 2.5915771674234768e-05, + "loss": 3.9196, + "step": 1470976 + }, + { + "epoch": 0.01, + "learning_rate": 2.5907385726724247e-05, + "loss": 3.9302, + "step": 1471488 + }, + { + "epoch": 0.01, + "learning_rate": 2.5898999779213727e-05, + "loss": 3.9123, + "step": 1472000 + }, + { + "epoch": 0.01, + "learning_rate": 2.5890613831703207e-05, + "loss": 3.9235, + "step": 1472512 + }, + { + "epoch": 0.01, + "learning_rate": 2.588222788419269e-05, + "loss": 3.9293, + "step": 1473024 + }, + { + "epoch": 0.01, + "learning_rate": 2.587384193668217e-05, + "loss": 3.9242, + "step": 1473536 + }, + { + "epoch": 0.01, + "learning_rate": 2.586547236797538e-05, + "loss": 3.9296, + "step": 1474048 + }, + { + "epoch": 0.01, + "learning_rate": 2.585708642046486e-05, + "loss": 3.9071, + "step": 1474560 + }, + { + "epoch": 0.01, + "learning_rate": 2.584870047295434e-05, + "loss": 3.9139, + "step": 1475072 + }, + { + "epoch": 0.01, + "learning_rate": 2.5840314525443816e-05, + "loss": 3.929, + "step": 1475584 + }, + { + "epoch": 0.01, + "learning_rate": 2.5831944956737032e-05, + "loss": 3.9201, + "step": 1476096 + }, + { + "epoch": 0.01, + "learning_rate": 2.5823559009226512e-05, + "loss": 3.9193, + "step": 1476608 + }, + { + "epoch": 0.01, + "learning_rate": 2.581518944051972e-05, + "loss": 3.9102, + "step": 1477120 + }, + { + "epoch": 0.01, + "learning_rate": 2.58068034930092e-05, + "loss": 3.9238, + "step": 1477632 + }, + { + "epoch": 0.01, + "learning_rate": 2.579841754549868e-05, + "loss": 3.9003, + "step": 1478144 + }, + { + "epoch": 0.01, + "learning_rate": 2.579003159798816e-05, + "loss": 3.9276, + "step": 1478656 + }, + { + "epoch": 0.01, + "learning_rate": 2.5781645650477638e-05, + "loss": 3.9069, + "step": 1479168 + }, + { + "epoch": 0.01, + "learning_rate": 2.5773276081770854e-05, + "loss": 3.9231, + "step": 1479680 + }, + { + "epoch": 0.01, + "learning_rate": 2.5764890134260333e-05, + "loss": 3.9246, + "step": 1480192 + }, + { + "epoch": 0.01, + "learning_rate": 2.5756504186749813e-05, + "loss": 3.9118, + "step": 1480704 + }, + { + "epoch": 0.01, + "learning_rate": 2.574811823923929e-05, + "loss": 3.9264, + "step": 1481216 + }, + { + "epoch": 0.01, + "learning_rate": 2.573973229172877e-05, + "loss": 3.9151, + "step": 1481728 + }, + { + "epoch": 0.01, + "learning_rate": 2.573134634421825e-05, + "loss": 3.91, + "step": 1482240 + }, + { + "epoch": 0.01, + "learning_rate": 2.572296039670773e-05, + "loss": 3.9085, + "step": 1482752 + }, + { + "epoch": 0.01, + "learning_rate": 2.571459082800094e-05, + "loss": 3.9261, + "step": 1483264 + }, + { + "epoch": 0.01, + "learning_rate": 2.570620488049042e-05, + "loss": 3.9139, + "step": 1483776 + }, + { + "epoch": 0.01, + "learning_rate": 2.56978189329799e-05, + "loss": 3.919, + "step": 1484288 + }, + { + "epoch": 0.01, + "learning_rate": 2.568943298546938e-05, + "loss": 3.905, + "step": 1484800 + }, + { + "epoch": 0.01, + "learning_rate": 2.5681047037958862e-05, + "loss": 3.9111, + "step": 1485312 + }, + { + "epoch": 0.01, + "learning_rate": 2.5672661090448342e-05, + "loss": 3.9215, + "step": 1485824 + }, + { + "epoch": 0.01, + "learning_rate": 2.5664275142937822e-05, + "loss": 3.9271, + "step": 1486336 + }, + { + "epoch": 0.01, + "learning_rate": 2.5655889195427302e-05, + "loss": 3.913, + "step": 1486848 + }, + { + "epoch": 0.01, + "learning_rate": 2.564751962672051e-05, + "loss": 3.9324, + "step": 1487360 + }, + { + "epoch": 0.01, + "learning_rate": 2.5639150058013724e-05, + "loss": 3.9264, + "step": 1487872 + }, + { + "epoch": 0.01, + "learning_rate": 2.5630764110503204e-05, + "loss": 3.9198, + "step": 1488384 + }, + { + "epoch": 0.01, + "learning_rate": 2.5622394541796413e-05, + "loss": 3.9236, + "step": 1488896 + }, + { + "epoch": 0.01, + "learning_rate": 2.5614008594285893e-05, + "loss": 3.9182, + "step": 1489408 + }, + { + "epoch": 0.01, + "learning_rate": 2.5605622646775373e-05, + "loss": 3.9198, + "step": 1489920 + }, + { + "epoch": 0.01, + "learning_rate": 2.5597236699264853e-05, + "loss": 3.9175, + "step": 1490432 + }, + { + "epoch": 0.01, + "learning_rate": 2.5588850751754332e-05, + "loss": 3.9221, + "step": 1490944 + }, + { + "epoch": 0.01, + "learning_rate": 2.5580464804243816e-05, + "loss": 3.9284, + "step": 1491456 + }, + { + "epoch": 0.01, + "learning_rate": 2.5572078856733296e-05, + "loss": 3.9322, + "step": 1491968 + }, + { + "epoch": 0.01, + "learning_rate": 2.5563692909222776e-05, + "loss": 3.9214, + "step": 1492480 + }, + { + "epoch": 0.01, + "learning_rate": 2.5555306961712256e-05, + "loss": 3.8994, + "step": 1492992 + }, + { + "epoch": 0.01, + "learning_rate": 2.5546921014201736e-05, + "loss": 3.9112, + "step": 1493504 + }, + { + "epoch": 0.01, + "learning_rate": 2.5538551445494945e-05, + "loss": 3.9162, + "step": 1494016 + }, + { + "epoch": 0.01, + "learning_rate": 2.5530165497984425e-05, + "loss": 3.9195, + "step": 1494528 + }, + { + "epoch": 0.01, + "learning_rate": 2.5521779550473908e-05, + "loss": 3.9138, + "step": 1495040 + }, + { + "epoch": 0.01, + "learning_rate": 2.5513393602963388e-05, + "loss": 3.9216, + "step": 1495552 + }, + { + "epoch": 0.02, + "learning_rate": 2.5505007655452868e-05, + "loss": 3.9161, + "step": 1496064 + }, + { + "epoch": 0.02, + "learning_rate": 2.5496621707942348e-05, + "loss": 3.9118, + "step": 1496576 + }, + { + "epoch": 0.02, + "learning_rate": 2.5488235760431828e-05, + "loss": 3.919, + "step": 1497088 + }, + { + "epoch": 0.02, + "learning_rate": 2.5479866191725037e-05, + "loss": 3.9252, + "step": 1497600 + }, + { + "epoch": 0.02, + "learning_rate": 2.5471480244214517e-05, + "loss": 3.9218, + "step": 1498112 + }, + { + "epoch": 0.02, + "learning_rate": 2.5463094296704e-05, + "loss": 3.9349, + "step": 1498624 + }, + { + "epoch": 0.02, + "learning_rate": 2.5454708349193473e-05, + "loss": 3.9114, + "step": 1499136 + }, + { + "epoch": 0.02, + "learning_rate": 2.5446322401682953e-05, + "loss": 3.9173, + "step": 1499648 + }, + { + "epoch": 0.02, + "learning_rate": 2.5437936454172433e-05, + "loss": 3.918, + "step": 1500160 + }, + { + "epoch": 0.02, + "learning_rate": 2.5429550506661913e-05, + "loss": 3.9212, + "step": 1500672 + }, + { + "epoch": 0.02, + "learning_rate": 2.5421164559151393e-05, + "loss": 3.9295, + "step": 1501184 + }, + { + "epoch": 0.02, + "learning_rate": 2.5412794990444606e-05, + "loss": 3.9124, + "step": 1501696 + }, + { + "epoch": 0.02, + "learning_rate": 2.5404409042934086e-05, + "loss": 3.9184, + "step": 1502208 + }, + { + "epoch": 0.02, + "learning_rate": 2.5396023095423566e-05, + "loss": 3.9205, + "step": 1502720 + }, + { + "epoch": 0.02, + "learning_rate": 2.5387637147913046e-05, + "loss": 3.925, + "step": 1503232 + }, + { + "epoch": 0.02, + "learning_rate": 2.5379267579206255e-05, + "loss": 3.9057, + "step": 1503744 + }, + { + "epoch": 0.02, + "learning_rate": 2.5370881631695738e-05, + "loss": 3.917, + "step": 1504256 + }, + { + "epoch": 0.02, + "learning_rate": 2.5362495684185218e-05, + "loss": 3.9165, + "step": 1504768 + }, + { + "epoch": 0.02, + "learning_rate": 2.5354109736674698e-05, + "loss": 3.91, + "step": 1505280 + }, + { + "epoch": 0.02, + "learning_rate": 2.5345723789164178e-05, + "loss": 3.9093, + "step": 1505792 + }, + { + "epoch": 0.02, + "learning_rate": 2.5337337841653658e-05, + "loss": 3.9149, + "step": 1506304 + }, + { + "epoch": 0.02, + "learning_rate": 2.5328968272946867e-05, + "loss": 3.9178, + "step": 1506816 + }, + { + "epoch": 0.02, + "learning_rate": 2.5320582325436347e-05, + "loss": 3.915, + "step": 1507328 + }, + { + "epoch": 0.02, + "learning_rate": 2.531219637792583e-05, + "loss": 3.9102, + "step": 1507840 + }, + { + "epoch": 0.02, + "learning_rate": 2.530381043041531e-05, + "loss": 3.9154, + "step": 1508352 + }, + { + "epoch": 0.02, + "learning_rate": 2.529542448290479e-05, + "loss": 3.9309, + "step": 1508864 + }, + { + "epoch": 0.02, + "learning_rate": 2.528703853539427e-05, + "loss": 3.9266, + "step": 1509376 + }, + { + "epoch": 0.02, + "learning_rate": 2.527865258788375e-05, + "loss": 3.9206, + "step": 1509888 + }, + { + "epoch": 0.02, + "learning_rate": 2.527026664037323e-05, + "loss": 3.9172, + "step": 1510400 + }, + { + "epoch": 0.02, + "learning_rate": 2.526191345047017e-05, + "loss": 3.9179, + "step": 1510912 + }, + { + "epoch": 0.02, + "learning_rate": 2.525352750295965e-05, + "loss": 3.9188, + "step": 1511424 + }, + { + "epoch": 0.02, + "learning_rate": 2.524514155544913e-05, + "loss": 3.9094, + "step": 1511936 + }, + { + "epoch": 0.02, + "learning_rate": 2.523675560793861e-05, + "loss": 3.9128, + "step": 1512448 + }, + { + "epoch": 0.02, + "learning_rate": 2.522836966042809e-05, + "loss": 3.9131, + "step": 1512960 + }, + { + "epoch": 0.02, + "learning_rate": 2.521998371291757e-05, + "loss": 3.9249, + "step": 1513472 + }, + { + "epoch": 0.02, + "learning_rate": 2.521159776540705e-05, + "loss": 3.9159, + "step": 1513984 + }, + { + "epoch": 0.02, + "learning_rate": 2.520321181789653e-05, + "loss": 3.923, + "step": 1514496 + }, + { + "epoch": 0.02, + "learning_rate": 2.5194842249189744e-05, + "loss": 3.9153, + "step": 1515008 + }, + { + "epoch": 0.02, + "learning_rate": 2.5186456301679224e-05, + "loss": 3.9124, + "step": 1515520 + }, + { + "epoch": 0.02, + "learning_rate": 2.5178070354168704e-05, + "loss": 3.911, + "step": 1516032 + }, + { + "epoch": 0.02, + "learning_rate": 2.5169684406658184e-05, + "loss": 3.9314, + "step": 1516544 + }, + { + "epoch": 0.02, + "learning_rate": 2.5161314837951393e-05, + "loss": 3.9303, + "step": 1517056 + }, + { + "epoch": 0.02, + "learning_rate": 2.5152928890440873e-05, + "loss": 3.9203, + "step": 1517568 + }, + { + "epoch": 0.02, + "learning_rate": 2.5144542942930356e-05, + "loss": 3.9183, + "step": 1518080 + }, + { + "epoch": 0.02, + "learning_rate": 2.5136156995419836e-05, + "loss": 3.92, + "step": 1518592 + }, + { + "epoch": 0.02, + "learning_rate": 2.512777104790931e-05, + "loss": 3.9214, + "step": 1519104 + }, + { + "epoch": 0.02, + "learning_rate": 2.511938510039879e-05, + "loss": 3.9338, + "step": 1519616 + }, + { + "epoch": 0.02, + "learning_rate": 2.511099915288827e-05, + "loss": 3.9173, + "step": 1520128 + }, + { + "epoch": 0.02, + "learning_rate": 2.510261320537775e-05, + "loss": 3.9194, + "step": 1520640 + }, + { + "epoch": 0.02, + "learning_rate": 2.509424363667096e-05, + "loss": 3.9157, + "step": 1521152 + }, + { + "epoch": 0.02, + "learning_rate": 2.5085874067964177e-05, + "loss": 3.9219, + "step": 1521664 + }, + { + "epoch": 0.02, + "learning_rate": 2.5077488120453657e-05, + "loss": 3.9153, + "step": 1522176 + }, + { + "epoch": 0.02, + "learning_rate": 2.506910217294313e-05, + "loss": 3.918, + "step": 1522688 + }, + { + "epoch": 0.02, + "learning_rate": 2.506071622543261e-05, + "loss": 3.9249, + "step": 1523200 + }, + { + "epoch": 0.02, + "learning_rate": 2.5052330277922094e-05, + "loss": 3.914, + "step": 1523712 + }, + { + "epoch": 0.02, + "learning_rate": 2.504396070921531e-05, + "loss": 3.923, + "step": 1524224 + }, + { + "epoch": 0.02, + "learning_rate": 2.5035574761704783e-05, + "loss": 3.927, + "step": 1524736 + }, + { + "epoch": 0.02, + "learning_rate": 2.5027188814194263e-05, + "loss": 3.9162, + "step": 1525248 + }, + { + "epoch": 0.02, + "learning_rate": 2.5018802866683743e-05, + "loss": 3.9222, + "step": 1525760 + }, + { + "epoch": 0.02, + "learning_rate": 2.5010416919173223e-05, + "loss": 3.9111, + "step": 1526272 + }, + { + "epoch": 0.03, + "eval_loss": 3.997249126434326, + "eval_runtime": 295.1895, + "eval_samples_per_second": 1292.698, + "eval_steps_per_second": 40.398, + "step": 1526387 + }, + { + "epoch": 1.0, + "learning_rate": 2.5002047350466435e-05, + "loss": 3.9165, + "step": 1526784 + }, + { + "epoch": 1.0, + "learning_rate": 2.499366140295592e-05, + "loss": 3.9074, + "step": 1527296 + }, + { + "epoch": 1.0, + "learning_rate": 2.49852754554454e-05, + "loss": 3.9269, + "step": 1527808 + }, + { + "epoch": 1.0, + "learning_rate": 2.4976889507934875e-05, + "loss": 3.9157, + "step": 1528320 + }, + { + "epoch": 1.0, + "learning_rate": 2.4968503560424355e-05, + "loss": 3.9333, + "step": 1528832 + }, + { + "epoch": 1.0, + "learning_rate": 2.4960117612913835e-05, + "loss": 3.9084, + "step": 1529344 + }, + { + "epoch": 1.0, + "learning_rate": 2.4951731665403315e-05, + "loss": 3.9198, + "step": 1529856 + }, + { + "epoch": 1.0, + "learning_rate": 2.4943345717892795e-05, + "loss": 3.904, + "step": 1530368 + }, + { + "epoch": 1.0, + "learning_rate": 2.4934976149186007e-05, + "loss": 3.9219, + "step": 1530880 + }, + { + "epoch": 1.0, + "learning_rate": 2.4926590201675487e-05, + "loss": 3.9063, + "step": 1531392 + }, + { + "epoch": 1.0, + "learning_rate": 2.4918204254164967e-05, + "loss": 3.9328, + "step": 1531904 + }, + { + "epoch": 1.0, + "learning_rate": 2.4909818306654447e-05, + "loss": 3.9269, + "step": 1532416 + }, + { + "epoch": 1.0, + "learning_rate": 2.4901448737947656e-05, + "loss": 3.9014, + "step": 1532928 + }, + { + "epoch": 1.0, + "learning_rate": 2.489306279043714e-05, + "loss": 3.9118, + "step": 1533440 + }, + { + "epoch": 1.0, + "learning_rate": 2.488467684292662e-05, + "loss": 3.9175, + "step": 1533952 + }, + { + "epoch": 1.0, + "learning_rate": 2.48762908954161e-05, + "loss": 3.9074, + "step": 1534464 + }, + { + "epoch": 1.0, + "learning_rate": 2.486790494790558e-05, + "loss": 3.9145, + "step": 1534976 + }, + { + "epoch": 1.0, + "learning_rate": 2.4859519000395056e-05, + "loss": 3.9154, + "step": 1535488 + }, + { + "epoch": 1.0, + "learning_rate": 2.4851133052884536e-05, + "loss": 3.9121, + "step": 1536000 + }, + { + "epoch": 1.0, + "learning_rate": 2.4842747105374016e-05, + "loss": 3.9361, + "step": 1536512 + }, + { + "epoch": 1.0, + "learning_rate": 2.4834361157863496e-05, + "loss": 3.9158, + "step": 1537024 + }, + { + "epoch": 1.0, + "learning_rate": 2.482597521035298e-05, + "loss": 3.917, + "step": 1537536 + }, + { + "epoch": 1.0, + "learning_rate": 2.481758926284246e-05, + "loss": 3.9187, + "step": 1538048 + }, + { + "epoch": 1.0, + "learning_rate": 2.480921969413567e-05, + "loss": 3.9257, + "step": 1538560 + }, + { + "epoch": 1.0, + "learning_rate": 2.480083374662515e-05, + "loss": 3.9109, + "step": 1539072 + }, + { + "epoch": 1.0, + "learning_rate": 2.479244779911463e-05, + "loss": 3.9205, + "step": 1539584 + }, + { + "epoch": 1.0, + "learning_rate": 2.478406185160411e-05, + "loss": 3.9063, + "step": 1540096 + }, + { + "epoch": 1.0, + "learning_rate": 2.477567590409359e-05, + "loss": 3.9135, + "step": 1540608 + }, + { + "epoch": 1.0, + "learning_rate": 2.4767289956583068e-05, + "loss": 3.9133, + "step": 1541120 + }, + { + "epoch": 1.0, + "learning_rate": 2.4758904009072548e-05, + "loss": 3.9151, + "step": 1541632 + }, + { + "epoch": 1.01, + "learning_rate": 2.4750518061562028e-05, + "loss": 3.9121, + "step": 1542144 + }, + { + "epoch": 1.01, + "learning_rate": 2.4742148492855237e-05, + "loss": 3.9204, + "step": 1542656 + }, + { + "epoch": 1.01, + "learning_rate": 2.4733762545344717e-05, + "loss": 3.9205, + "step": 1543168 + }, + { + "epoch": 1.01, + "learning_rate": 2.47253765978342e-05, + "loss": 3.9251, + "step": 1543680 + }, + { + "epoch": 1.01, + "learning_rate": 2.4717007029127413e-05, + "loss": 3.9124, + "step": 1544192 + }, + { + "epoch": 1.01, + "learning_rate": 2.470862108161689e-05, + "loss": 3.9222, + "step": 1544704 + }, + { + "epoch": 1.01, + "learning_rate": 2.470023513410637e-05, + "loss": 3.9104, + "step": 1545216 + }, + { + "epoch": 1.01, + "learning_rate": 2.469184918659585e-05, + "loss": 3.9181, + "step": 1545728 + }, + { + "epoch": 1.01, + "learning_rate": 2.468346323908533e-05, + "loss": 3.9055, + "step": 1546240 + }, + { + "epoch": 1.01, + "learning_rate": 2.4675093670378542e-05, + "loss": 3.9135, + "step": 1546752 + }, + { + "epoch": 1.01, + "learning_rate": 2.4666707722868022e-05, + "loss": 3.9095, + "step": 1547264 + }, + { + "epoch": 1.01, + "learning_rate": 2.4658321775357502e-05, + "loss": 3.9235, + "step": 1547776 + }, + { + "epoch": 1.01, + "learning_rate": 2.4649935827846982e-05, + "loss": 3.9102, + "step": 1548288 + }, + { + "epoch": 1.01, + "learning_rate": 2.4641549880336462e-05, + "loss": 3.9121, + "step": 1548800 + }, + { + "epoch": 1.01, + "learning_rate": 2.4633163932825942e-05, + "loss": 3.9241, + "step": 1549312 + }, + { + "epoch": 1.01, + "learning_rate": 2.462477798531542e-05, + "loss": 3.9171, + "step": 1549824 + }, + { + "epoch": 1.01, + "learning_rate": 2.4616408416608634e-05, + "loss": 3.9224, + "step": 1550336 + }, + { + "epoch": 1.01, + "learning_rate": 2.4608022469098114e-05, + "loss": 3.9042, + "step": 1550848 + }, + { + "epoch": 1.01, + "learning_rate": 2.459963652158759e-05, + "loss": 3.903, + "step": 1551360 + }, + { + "epoch": 1.01, + "learning_rate": 2.459125057407707e-05, + "loss": 3.9207, + "step": 1551872 + }, + { + "epoch": 1.01, + "learning_rate": 2.4582881005370283e-05, + "loss": 3.9154, + "step": 1552384 + }, + { + "epoch": 1.01, + "learning_rate": 2.4574495057859763e-05, + "loss": 3.9081, + "step": 1552896 + }, + { + "epoch": 1.01, + "learning_rate": 2.4566109110349243e-05, + "loss": 3.9057, + "step": 1553408 + }, + { + "epoch": 1.01, + "learning_rate": 2.4557723162838723e-05, + "loss": 3.9164, + "step": 1553920 + }, + { + "epoch": 1.01, + "learning_rate": 2.4549337215328203e-05, + "loss": 3.895, + "step": 1554432 + }, + { + "epoch": 1.01, + "learning_rate": 2.4540951267817683e-05, + "loss": 3.9172, + "step": 1554944 + }, + { + "epoch": 1.01, + "learning_rate": 2.4532581699110892e-05, + "loss": 3.8995, + "step": 1555456 + }, + { + "epoch": 1.01, + "learning_rate": 2.4524195751600372e-05, + "loss": 3.9113, + "step": 1555968 + }, + { + "epoch": 1.01, + "learning_rate": 2.4515809804089855e-05, + "loss": 3.9198, + "step": 1556480 + }, + { + "epoch": 1.01, + "learning_rate": 2.4507423856579335e-05, + "loss": 3.9064, + "step": 1556992 + }, + { + "epoch": 1.01, + "learning_rate": 2.4499037909068815e-05, + "loss": 3.9152, + "step": 1557504 + }, + { + "epoch": 1.01, + "learning_rate": 2.4490668340362024e-05, + "loss": 3.9099, + "step": 1558016 + }, + { + "epoch": 1.01, + "learning_rate": 2.4482282392851504e-05, + "loss": 3.9034, + "step": 1558528 + }, + { + "epoch": 1.01, + "learning_rate": 2.4473896445340984e-05, + "loss": 3.9077, + "step": 1559040 + }, + { + "epoch": 1.01, + "learning_rate": 2.4465510497830464e-05, + "loss": 3.9135, + "step": 1559552 + }, + { + "epoch": 1.01, + "learning_rate": 2.4457124550319948e-05, + "loss": 3.9068, + "step": 1560064 + }, + { + "epoch": 1.01, + "learning_rate": 2.4448738602809424e-05, + "loss": 3.9109, + "step": 1560576 + }, + { + "epoch": 1.01, + "learning_rate": 2.4440369034102637e-05, + "loss": 3.9013, + "step": 1561088 + }, + { + "epoch": 1.01, + "learning_rate": 2.4431983086592117e-05, + "loss": 3.9013, + "step": 1561600 + }, + { + "epoch": 1.01, + "learning_rate": 2.4423597139081597e-05, + "loss": 3.9148, + "step": 1562112 + }, + { + "epoch": 1.01, + "learning_rate": 2.4415211191571073e-05, + "loss": 3.921, + "step": 1562624 + }, + { + "epoch": 1.01, + "learning_rate": 2.4406825244060556e-05, + "loss": 3.9066, + "step": 1563136 + }, + { + "epoch": 1.01, + "learning_rate": 2.4398439296550036e-05, + "loss": 3.9219, + "step": 1563648 + }, + { + "epoch": 1.01, + "learning_rate": 2.4390069727843245e-05, + "loss": 3.9178, + "step": 1564160 + }, + { + "epoch": 1.01, + "learning_rate": 2.4381683780332725e-05, + "loss": 3.9109, + "step": 1564672 + }, + { + "epoch": 1.01, + "learning_rate": 2.4373297832822205e-05, + "loss": 3.9133, + "step": 1565184 + }, + { + "epoch": 1.01, + "learning_rate": 2.4364911885311685e-05, + "loss": 3.918, + "step": 1565696 + }, + { + "epoch": 1.01, + "learning_rate": 2.4356525937801165e-05, + "loss": 3.9105, + "step": 1566208 + }, + { + "epoch": 1.01, + "learning_rate": 2.434813999029065e-05, + "loss": 3.9118, + "step": 1566720 + }, + { + "epoch": 1.01, + "learning_rate": 2.433975404278013e-05, + "loss": 3.9167, + "step": 1567232 + }, + { + "epoch": 1.01, + "learning_rate": 2.4331368095269605e-05, + "loss": 3.9176, + "step": 1567744 + }, + { + "epoch": 1.01, + "learning_rate": 2.4322998526562818e-05, + "loss": 3.9194, + "step": 1568256 + }, + { + "epoch": 1.01, + "learning_rate": 2.4314628957856027e-05, + "loss": 3.9165, + "step": 1568768 + }, + { + "epoch": 1.01, + "learning_rate": 2.430624301034551e-05, + "loss": 3.8984, + "step": 1569280 + }, + { + "epoch": 1.01, + "learning_rate": 2.429785706283499e-05, + "loss": 3.8987, + "step": 1569792 + }, + { + "epoch": 1.01, + "learning_rate": 2.428947111532447e-05, + "loss": 3.9115, + "step": 1570304 + }, + { + "epoch": 1.01, + "learning_rate": 2.428108516781395e-05, + "loss": 3.9136, + "step": 1570816 + }, + { + "epoch": 1.01, + "learning_rate": 2.4272699220303427e-05, + "loss": 3.9093, + "step": 1571328 + }, + { + "epoch": 1.01, + "learning_rate": 2.426432965159664e-05, + "loss": 3.9128, + "step": 1571840 + }, + { + "epoch": 1.02, + "learning_rate": 2.425594370408612e-05, + "loss": 3.9089, + "step": 1572352 + }, + { + "epoch": 1.02, + "learning_rate": 2.4247557756575602e-05, + "loss": 3.9062, + "step": 1572864 + }, + { + "epoch": 1.02, + "learning_rate": 2.423917180906508e-05, + "loss": 3.912, + "step": 1573376 + }, + { + "epoch": 1.02, + "learning_rate": 2.423078586155456e-05, + "loss": 3.9129, + "step": 1573888 + }, + { + "epoch": 1.02, + "learning_rate": 2.422239991404404e-05, + "loss": 3.9182, + "step": 1574400 + }, + { + "epoch": 1.02, + "learning_rate": 2.421403034533725e-05, + "loss": 3.9185, + "step": 1574912 + }, + { + "epoch": 1.02, + "learning_rate": 2.4205644397826728e-05, + "loss": 3.9148, + "step": 1575424 + }, + { + "epoch": 1.02, + "learning_rate": 2.419725845031621e-05, + "loss": 3.9031, + "step": 1575936 + }, + { + "epoch": 1.02, + "learning_rate": 2.418887250280569e-05, + "loss": 3.9143, + "step": 1576448 + }, + { + "epoch": 1.02, + "learning_rate": 2.418048655529517e-05, + "loss": 3.9117, + "step": 1576960 + }, + { + "epoch": 1.02, + "learning_rate": 2.417210060778465e-05, + "loss": 3.9218, + "step": 1577472 + }, + { + "epoch": 1.02, + "learning_rate": 2.416371466027413e-05, + "loss": 3.9078, + "step": 1577984 + }, + { + "epoch": 1.02, + "learning_rate": 2.4155328712763608e-05, + "loss": 3.9059, + "step": 1578496 + }, + { + "epoch": 1.02, + "learning_rate": 2.414695914405682e-05, + "loss": 3.9152, + "step": 1579008 + }, + { + "epoch": 1.02, + "learning_rate": 2.4138573196546303e-05, + "loss": 3.9173, + "step": 1579520 + }, + { + "epoch": 1.02, + "learning_rate": 2.4130187249035783e-05, + "loss": 3.9051, + "step": 1580032 + }, + { + "epoch": 1.02, + "learning_rate": 2.4121817680328992e-05, + "loss": 3.91, + "step": 1580544 + }, + { + "epoch": 1.02, + "learning_rate": 2.4113431732818472e-05, + "loss": 3.908, + "step": 1581056 + }, + { + "epoch": 1.02, + "learning_rate": 2.4105045785307952e-05, + "loss": 3.8998, + "step": 1581568 + }, + { + "epoch": 1.02, + "learning_rate": 2.4096659837797432e-05, + "loss": 3.9003, + "step": 1582080 + }, + { + "epoch": 1.02, + "learning_rate": 2.4088273890286912e-05, + "loss": 3.9133, + "step": 1582592 + }, + { + "epoch": 1.02, + "learning_rate": 2.4079887942776392e-05, + "loss": 3.9073, + "step": 1583104 + }, + { + "epoch": 1.02, + "learning_rate": 2.4071518374069605e-05, + "loss": 3.9076, + "step": 1583616 + }, + { + "epoch": 1.02, + "learning_rate": 2.406313242655908e-05, + "loss": 3.9032, + "step": 1584128 + }, + { + "epoch": 1.02, + "learning_rate": 2.405474647904856e-05, + "loss": 3.9078, + "step": 1584640 + }, + { + "epoch": 1.02, + "learning_rate": 2.404636053153804e-05, + "loss": 3.9238, + "step": 1585152 + }, + { + "epoch": 1.02, + "learning_rate": 2.403797458402752e-05, + "loss": 3.9252, + "step": 1585664 + }, + { + "epoch": 1.02, + "learning_rate": 2.4029588636517004e-05, + "loss": 3.9121, + "step": 1586176 + }, + { + "epoch": 1.02, + "learning_rate": 2.4021202689006484e-05, + "loss": 3.9132, + "step": 1586688 + }, + { + "epoch": 1.02, + "learning_rate": 2.4012816741495964e-05, + "loss": 3.9077, + "step": 1587200 + }, + { + "epoch": 1.02, + "learning_rate": 2.4004447172789174e-05, + "loss": 3.9094, + "step": 1587712 + }, + { + "epoch": 1.02, + "learning_rate": 2.3996077604082383e-05, + "loss": 3.9038, + "step": 1588224 + }, + { + "epoch": 1.02, + "learning_rate": 2.3987691656571866e-05, + "loss": 3.9078, + "step": 1588736 + }, + { + "epoch": 1.02, + "learning_rate": 2.3979305709061346e-05, + "loss": 3.9092, + "step": 1589248 + }, + { + "epoch": 1.02, + "learning_rate": 2.3970919761550826e-05, + "loss": 3.9162, + "step": 1589760 + }, + { + "epoch": 1.02, + "learning_rate": 2.3962533814040306e-05, + "loss": 3.9079, + "step": 1590272 + }, + { + "epoch": 1.02, + "learning_rate": 2.3954147866529786e-05, + "loss": 3.9162, + "step": 1590784 + }, + { + "epoch": 1.02, + "learning_rate": 2.3945761919019262e-05, + "loss": 3.9071, + "step": 1591296 + }, + { + "epoch": 1.02, + "learning_rate": 2.3937375971508742e-05, + "loss": 3.9058, + "step": 1591808 + }, + { + "epoch": 1.02, + "learning_rate": 2.3928990023998226e-05, + "loss": 3.9025, + "step": 1592320 + }, + { + "epoch": 1.02, + "learning_rate": 2.3920620455291438e-05, + "loss": 3.9221, + "step": 1592832 + }, + { + "epoch": 1.02, + "learning_rate": 2.3912234507780915e-05, + "loss": 3.9225, + "step": 1593344 + }, + { + "epoch": 1.02, + "learning_rate": 2.3903864939074127e-05, + "loss": 3.9186, + "step": 1593856 + }, + { + "epoch": 1.02, + "learning_rate": 2.3895478991563607e-05, + "loss": 3.9051, + "step": 1594368 + }, + { + "epoch": 1.02, + "learning_rate": 2.3887093044053087e-05, + "loss": 3.9157, + "step": 1594880 + }, + { + "epoch": 1.02, + "learning_rate": 2.3878707096542567e-05, + "loss": 3.9128, + "step": 1595392 + }, + { + "epoch": 1.02, + "learning_rate": 2.3870321149032047e-05, + "loss": 3.9244, + "step": 1595904 + }, + { + "epoch": 1.02, + "learning_rate": 2.3861935201521527e-05, + "loss": 3.9087, + "step": 1596416 + }, + { + "epoch": 1.02, + "learning_rate": 2.3853549254011007e-05, + "loss": 3.9108, + "step": 1596928 + }, + { + "epoch": 1.02, + "learning_rate": 2.3845163306500487e-05, + "loss": 3.9147, + "step": 1597440 + }, + { + "epoch": 1.02, + "learning_rate": 2.383681011659743e-05, + "loss": 3.9175, + "step": 1597952 + }, + { + "epoch": 1.02, + "learning_rate": 2.3828424169086912e-05, + "loss": 3.9055, + "step": 1598464 + }, + { + "epoch": 1.02, + "learning_rate": 2.382003822157639e-05, + "loss": 3.9084, + "step": 1598976 + }, + { + "epoch": 1.02, + "learning_rate": 2.381165227406587e-05, + "loss": 3.9154, + "step": 1599488 + }, + { + "epoch": 1.02, + "learning_rate": 2.380326632655535e-05, + "loss": 3.9068, + "step": 1600000 + }, + { + "epoch": 1.02, + "learning_rate": 2.3794896757848557e-05, + "loss": 3.9146, + "step": 1600512 + }, + { + "epoch": 1.02, + "learning_rate": 2.378651081033804e-05, + "loss": 3.9193, + "step": 1601024 + }, + { + "epoch": 1.02, + "learning_rate": 2.377812486282752e-05, + "loss": 3.911, + "step": 1601536 + }, + { + "epoch": 1.02, + "learning_rate": 2.3769738915317e-05, + "loss": 3.9109, + "step": 1602048 + }, + { + "epoch": 1.02, + "learning_rate": 2.376135296780648e-05, + "loss": 3.9123, + "step": 1602560 + }, + { + "epoch": 1.03, + "eval_loss": 3.9945480823516846, + "eval_runtime": 294.3847, + "eval_samples_per_second": 1296.232, + "eval_steps_per_second": 40.508, + "step": 1602707 + }, + { + "epoch": 0.0, + "learning_rate": 2.375296702029596e-05, + "loss": 3.9016, + "step": 1603072 + }, + { + "epoch": 0.0, + "learning_rate": 2.374458107278544e-05, + "loss": 3.9003, + "step": 1603584 + }, + { + "epoch": 0.0, + "learning_rate": 2.3736195125274917e-05, + "loss": 3.9198, + "step": 1604096 + }, + { + "epoch": 0.0, + "learning_rate": 2.3727809177764397e-05, + "loss": 3.9082, + "step": 1604608 + }, + { + "epoch": 0.0, + "learning_rate": 2.371942323025388e-05, + "loss": 3.9214, + "step": 1605120 + }, + { + "epoch": 0.0, + "learning_rate": 2.371103728274336e-05, + "loss": 3.9093, + "step": 1605632 + }, + { + "epoch": 0.0, + "learning_rate": 2.370265133523284e-05, + "loss": 3.9117, + "step": 1606144 + }, + { + "epoch": 0.0, + "learning_rate": 2.369426538772232e-05, + "loss": 3.8982, + "step": 1606656 + }, + { + "epoch": 0.0, + "learning_rate": 2.36858794402118e-05, + "loss": 3.9117, + "step": 1607168 + }, + { + "epoch": 0.0, + "learning_rate": 2.3677493492701277e-05, + "loss": 3.9036, + "step": 1607680 + }, + { + "epoch": 0.0, + "learning_rate": 2.3669107545190757e-05, + "loss": 3.9199, + "step": 1608192 + }, + { + "epoch": 0.0, + "learning_rate": 2.3660721597680237e-05, + "loss": 3.9188, + "step": 1608704 + }, + { + "epoch": 0.0, + "learning_rate": 2.365235202897345e-05, + "loss": 3.8941, + "step": 1609216 + }, + { + "epoch": 0.0, + "learning_rate": 2.364396608146293e-05, + "loss": 3.9069, + "step": 1609728 + }, + { + "epoch": 0.0, + "learning_rate": 2.363558013395241e-05, + "loss": 3.9155, + "step": 1610240 + }, + { + "epoch": 0.0, + "learning_rate": 2.362719418644189e-05, + "loss": 3.8981, + "step": 1610752 + }, + { + "epoch": 0.0, + "learning_rate": 2.361880823893137e-05, + "loss": 3.9081, + "step": 1611264 + }, + { + "epoch": 0.0, + "learning_rate": 2.361042229142085e-05, + "loss": 3.9096, + "step": 1611776 + }, + { + "epoch": 0.0, + "learning_rate": 2.360203634391033e-05, + "loss": 3.9067, + "step": 1612288 + }, + { + "epoch": 0.0, + "learning_rate": 2.359365039639981e-05, + "loss": 3.9252, + "step": 1612800 + }, + { + "epoch": 0.0, + "learning_rate": 2.358526444888929e-05, + "loss": 3.9102, + "step": 1613312 + }, + { + "epoch": 0.0, + "learning_rate": 2.357687850137877e-05, + "loss": 3.9102, + "step": 1613824 + }, + { + "epoch": 0.0, + "learning_rate": 2.356849255386825e-05, + "loss": 3.9151, + "step": 1614336 + }, + { + "epoch": 0.0, + "learning_rate": 2.356010660635773e-05, + "loss": 3.9133, + "step": 1614848 + }, + { + "epoch": 0.0, + "learning_rate": 2.3551737037650938e-05, + "loss": 3.9063, + "step": 1615360 + }, + { + "epoch": 0.0, + "learning_rate": 2.354335109014042e-05, + "loss": 3.9119, + "step": 1615872 + }, + { + "epoch": 0.0, + "learning_rate": 2.35349651426299e-05, + "loss": 3.9028, + "step": 1616384 + }, + { + "epoch": 0.0, + "learning_rate": 2.352657919511938e-05, + "loss": 3.9051, + "step": 1616896 + }, + { + "epoch": 0.0, + "learning_rate": 2.351819324760886e-05, + "loss": 3.9052, + "step": 1617408 + }, + { + "epoch": 0.0, + "learning_rate": 2.350982367890207e-05, + "loss": 3.9106, + "step": 1617920 + }, + { + "epoch": 0.01, + "learning_rate": 2.350143773139155e-05, + "loss": 3.9058, + "step": 1618432 + }, + { + "epoch": 0.01, + "learning_rate": 2.349305178388103e-05, + "loss": 3.912, + "step": 1618944 + }, + { + "epoch": 0.01, + "learning_rate": 2.3484665836370513e-05, + "loss": 3.9127, + "step": 1619456 + }, + { + "epoch": 0.01, + "learning_rate": 2.3476296267663722e-05, + "loss": 3.9175, + "step": 1619968 + }, + { + "epoch": 0.01, + "learning_rate": 2.346792669895693e-05, + "loss": 3.9077, + "step": 1620480 + }, + { + "epoch": 0.01, + "learning_rate": 2.345954075144641e-05, + "loss": 3.9123, + "step": 1620992 + }, + { + "epoch": 0.01, + "learning_rate": 2.345115480393589e-05, + "loss": 3.9057, + "step": 1621504 + }, + { + "epoch": 0.01, + "learning_rate": 2.3442768856425375e-05, + "loss": 3.9069, + "step": 1622016 + }, + { + "epoch": 0.01, + "learning_rate": 2.3434382908914855e-05, + "loss": 3.8978, + "step": 1622528 + }, + { + "epoch": 0.01, + "learning_rate": 2.3425996961404335e-05, + "loss": 3.9052, + "step": 1623040 + }, + { + "epoch": 0.01, + "learning_rate": 2.341761101389381e-05, + "loss": 3.905, + "step": 1623552 + }, + { + "epoch": 0.01, + "learning_rate": 2.340922506638329e-05, + "loss": 3.915, + "step": 1624064 + }, + { + "epoch": 0.01, + "learning_rate": 2.340083911887277e-05, + "loss": 3.9013, + "step": 1624576 + }, + { + "epoch": 0.01, + "learning_rate": 2.339245317136225e-05, + "loss": 3.9064, + "step": 1625088 + }, + { + "epoch": 0.01, + "learning_rate": 2.338406722385173e-05, + "loss": 3.9181, + "step": 1625600 + }, + { + "epoch": 0.01, + "learning_rate": 2.3375681276341214e-05, + "loss": 3.9101, + "step": 1626112 + }, + { + "epoch": 0.01, + "learning_rate": 2.3367311707634424e-05, + "loss": 3.9131, + "step": 1626624 + }, + { + "epoch": 0.01, + "learning_rate": 2.3358925760123903e-05, + "loss": 3.9001, + "step": 1627136 + }, + { + "epoch": 0.01, + "learning_rate": 2.3350539812613383e-05, + "loss": 3.8908, + "step": 1627648 + }, + { + "epoch": 0.01, + "learning_rate": 2.3342153865102863e-05, + "loss": 3.9178, + "step": 1628160 + }, + { + "epoch": 0.01, + "learning_rate": 2.3333767917592343e-05, + "loss": 3.9068, + "step": 1628672 + }, + { + "epoch": 0.01, + "learning_rate": 2.3325414727689285e-05, + "loss": 3.9017, + "step": 1629184 + }, + { + "epoch": 0.01, + "learning_rate": 2.3317028780178765e-05, + "loss": 3.9033, + "step": 1629696 + }, + { + "epoch": 0.01, + "learning_rate": 2.3308642832668245e-05, + "loss": 3.9116, + "step": 1630208 + }, + { + "epoch": 0.01, + "learning_rate": 2.3300256885157725e-05, + "loss": 3.8868, + "step": 1630720 + }, + { + "epoch": 0.01, + "learning_rate": 2.3291870937647205e-05, + "loss": 3.913, + "step": 1631232 + }, + { + "epoch": 0.01, + "learning_rate": 2.3283484990136685e-05, + "loss": 3.8912, + "step": 1631744 + }, + { + "epoch": 0.01, + "learning_rate": 2.3275099042626168e-05, + "loss": 3.9065, + "step": 1632256 + }, + { + "epoch": 0.01, + "learning_rate": 2.3266713095115645e-05, + "loss": 3.9085, + "step": 1632768 + }, + { + "epoch": 0.01, + "learning_rate": 2.3258343526408857e-05, + "loss": 3.9067, + "step": 1633280 + }, + { + "epoch": 0.01, + "learning_rate": 2.3249973957702066e-05, + "loss": 3.9038, + "step": 1633792 + }, + { + "epoch": 0.01, + "learning_rate": 2.3241588010191546e-05, + "loss": 3.9071, + "step": 1634304 + }, + { + "epoch": 0.01, + "learning_rate": 2.323320206268103e-05, + "loss": 3.8919, + "step": 1634816 + }, + { + "epoch": 0.01, + "learning_rate": 2.322481611517051e-05, + "loss": 3.901, + "step": 1635328 + }, + { + "epoch": 0.01, + "learning_rate": 2.321643016765999e-05, + "loss": 3.906, + "step": 1635840 + }, + { + "epoch": 0.01, + "learning_rate": 2.3208044220149466e-05, + "loss": 3.9043, + "step": 1636352 + }, + { + "epoch": 0.01, + "learning_rate": 2.3199658272638946e-05, + "loss": 3.9012, + "step": 1636864 + }, + { + "epoch": 0.01, + "learning_rate": 2.3191272325128426e-05, + "loss": 3.8942, + "step": 1637376 + }, + { + "epoch": 0.01, + "learning_rate": 2.318290275642164e-05, + "loss": 3.895, + "step": 1637888 + }, + { + "epoch": 0.01, + "learning_rate": 2.317451680891112e-05, + "loss": 3.9047, + "step": 1638400 + }, + { + "epoch": 0.01, + "learning_rate": 2.31661308614006e-05, + "loss": 3.9141, + "step": 1638912 + }, + { + "epoch": 0.01, + "learning_rate": 2.315776129269381e-05, + "loss": 3.8989, + "step": 1639424 + }, + { + "epoch": 0.01, + "learning_rate": 2.3149375345183287e-05, + "loss": 3.917, + "step": 1639936 + }, + { + "epoch": 0.01, + "learning_rate": 2.3140989397672767e-05, + "loss": 3.9147, + "step": 1640448 + }, + { + "epoch": 0.01, + "learning_rate": 2.3132619828965983e-05, + "loss": 3.9072, + "step": 1640960 + }, + { + "epoch": 0.01, + "learning_rate": 2.3124233881455463e-05, + "loss": 3.9046, + "step": 1641472 + }, + { + "epoch": 0.01, + "learning_rate": 2.311584793394494e-05, + "loss": 3.9101, + "step": 1641984 + }, + { + "epoch": 0.01, + "learning_rate": 2.310746198643442e-05, + "loss": 3.9049, + "step": 1642496 + }, + { + "epoch": 0.01, + "learning_rate": 2.30990760389239e-05, + "loss": 3.9042, + "step": 1643008 + }, + { + "epoch": 0.01, + "learning_rate": 2.309069009141338e-05, + "loss": 3.91, + "step": 1643520 + }, + { + "epoch": 0.01, + "learning_rate": 2.308230414390286e-05, + "loss": 3.9104, + "step": 1644032 + }, + { + "epoch": 0.01, + "learning_rate": 2.307391819639234e-05, + "loss": 3.9169, + "step": 1644544 + }, + { + "epoch": 0.01, + "learning_rate": 2.3065532248881823e-05, + "loss": 3.9101, + "step": 1645056 + }, + { + "epoch": 0.01, + "learning_rate": 2.30571463013713e-05, + "loss": 3.8921, + "step": 1645568 + }, + { + "epoch": 0.01, + "learning_rate": 2.304876035386078e-05, + "loss": 3.8943, + "step": 1646080 + }, + { + "epoch": 0.01, + "learning_rate": 2.3040390785153992e-05, + "loss": 3.8988, + "step": 1646592 + }, + { + "epoch": 0.01, + "learning_rate": 2.303200483764347e-05, + "loss": 3.9077, + "step": 1647104 + }, + { + "epoch": 0.01, + "learning_rate": 2.3023618890132952e-05, + "loss": 3.9041, + "step": 1647616 + }, + { + "epoch": 0.01, + "learning_rate": 2.3015232942622432e-05, + "loss": 3.9045, + "step": 1648128 + }, + { + "epoch": 0.02, + "learning_rate": 2.3006846995111912e-05, + "loss": 3.9005, + "step": 1648640 + }, + { + "epoch": 0.02, + "learning_rate": 2.299846104760139e-05, + "loss": 3.9004, + "step": 1649152 + }, + { + "epoch": 0.02, + "learning_rate": 2.299007510009087e-05, + "loss": 3.9055, + "step": 1649664 + }, + { + "epoch": 0.02, + "learning_rate": 2.298170553138408e-05, + "loss": 3.9019, + "step": 1650176 + }, + { + "epoch": 0.02, + "learning_rate": 2.297331958387356e-05, + "loss": 3.9119, + "step": 1650688 + }, + { + "epoch": 0.02, + "learning_rate": 2.2964950015166773e-05, + "loss": 3.9139, + "step": 1651200 + }, + { + "epoch": 0.02, + "learning_rate": 2.2956564067656253e-05, + "loss": 3.91, + "step": 1651712 + }, + { + "epoch": 0.02, + "learning_rate": 2.2948178120145733e-05, + "loss": 3.8953, + "step": 1652224 + }, + { + "epoch": 0.02, + "learning_rate": 2.2939792172635213e-05, + "loss": 3.9054, + "step": 1652736 + }, + { + "epoch": 0.02, + "learning_rate": 2.2931406225124693e-05, + "loss": 3.906, + "step": 1653248 + }, + { + "epoch": 0.02, + "learning_rate": 2.2923020277614173e-05, + "loss": 3.9173, + "step": 1653760 + }, + { + "epoch": 0.02, + "learning_rate": 2.2914634330103653e-05, + "loss": 3.9014, + "step": 1654272 + }, + { + "epoch": 0.02, + "learning_rate": 2.2906248382593133e-05, + "loss": 3.8967, + "step": 1654784 + }, + { + "epoch": 0.02, + "learning_rate": 2.2897862435082613e-05, + "loss": 3.916, + "step": 1655296 + }, + { + "epoch": 0.02, + "learning_rate": 2.2889492866375825e-05, + "loss": 3.9038, + "step": 1655808 + }, + { + "epoch": 0.02, + "learning_rate": 2.2881106918865302e-05, + "loss": 3.9008, + "step": 1656320 + }, + { + "epoch": 0.02, + "learning_rate": 2.2872720971354782e-05, + "loss": 3.9026, + "step": 1656832 + }, + { + "epoch": 0.02, + "learning_rate": 2.2864351402647998e-05, + "loss": 3.9011, + "step": 1657344 + }, + { + "epoch": 0.02, + "learning_rate": 2.2855965455137474e-05, + "loss": 3.8924, + "step": 1657856 + }, + { + "epoch": 0.02, + "learning_rate": 2.2847579507626954e-05, + "loss": 3.8905, + "step": 1658368 + }, + { + "epoch": 0.02, + "learning_rate": 2.2839193560116434e-05, + "loss": 3.9073, + "step": 1658880 + }, + { + "epoch": 0.02, + "learning_rate": 2.2830807612605914e-05, + "loss": 3.9032, + "step": 1659392 + }, + { + "epoch": 0.02, + "learning_rate": 2.2822438043899123e-05, + "loss": 3.8999, + "step": 1659904 + }, + { + "epoch": 0.02, + "learning_rate": 2.2814052096388607e-05, + "loss": 3.8963, + "step": 1660416 + }, + { + "epoch": 0.02, + "learning_rate": 2.2805666148878086e-05, + "loss": 3.8967, + "step": 1660928 + }, + { + "epoch": 0.02, + "learning_rate": 2.2797280201367566e-05, + "loss": 3.9171, + "step": 1661440 + }, + { + "epoch": 0.02, + "learning_rate": 2.2788894253857046e-05, + "loss": 3.9163, + "step": 1661952 + }, + { + "epoch": 0.02, + "learning_rate": 2.2780508306346526e-05, + "loss": 3.9106, + "step": 1662464 + }, + { + "epoch": 0.02, + "learning_rate": 2.2772122358836006e-05, + "loss": 3.9063, + "step": 1662976 + }, + { + "epoch": 0.02, + "learning_rate": 2.2763752790129215e-05, + "loss": 3.8967, + "step": 1663488 + }, + { + "epoch": 0.02, + "learning_rate": 2.27553668426187e-05, + "loss": 3.9011, + "step": 1664000 + }, + { + "epoch": 0.02, + "learning_rate": 2.274698089510818e-05, + "loss": 3.8991, + "step": 1664512 + }, + { + "epoch": 0.02, + "learning_rate": 2.2738594947597655e-05, + "loss": 3.9004, + "step": 1665024 + }, + { + "epoch": 0.02, + "learning_rate": 2.2730209000087135e-05, + "loss": 3.901, + "step": 1665536 + }, + { + "epoch": 0.02, + "learning_rate": 2.2721823052576615e-05, + "loss": 3.9068, + "step": 1666048 + }, + { + "epoch": 0.02, + "learning_rate": 2.2713437105066095e-05, + "loss": 3.9042, + "step": 1666560 + }, + { + "epoch": 0.02, + "learning_rate": 2.2705067536359308e-05, + "loss": 3.9047, + "step": 1667072 + }, + { + "epoch": 0.02, + "learning_rate": 2.2696681588848788e-05, + "loss": 3.9014, + "step": 1667584 + }, + { + "epoch": 0.02, + "learning_rate": 2.2688295641338268e-05, + "loss": 3.9008, + "step": 1668096 + }, + { + "epoch": 0.02, + "learning_rate": 2.2679909693827748e-05, + "loss": 3.8992, + "step": 1668608 + }, + { + "epoch": 0.02, + "learning_rate": 2.2671540125120957e-05, + "loss": 3.9104, + "step": 1669120 + }, + { + "epoch": 0.02, + "learning_rate": 2.266317055641417e-05, + "loss": 3.9162, + "step": 1669632 + }, + { + "epoch": 0.02, + "learning_rate": 2.2654784608903652e-05, + "loss": 3.9128, + "step": 1670144 + }, + { + "epoch": 0.02, + "learning_rate": 2.264639866139313e-05, + "loss": 3.9002, + "step": 1670656 + }, + { + "epoch": 0.02, + "learning_rate": 2.263801271388261e-05, + "loss": 3.9121, + "step": 1671168 + }, + { + "epoch": 0.02, + "learning_rate": 2.262962676637209e-05, + "loss": 3.9043, + "step": 1671680 + }, + { + "epoch": 0.02, + "learning_rate": 2.262124081886157e-05, + "loss": 3.9164, + "step": 1672192 + }, + { + "epoch": 0.02, + "learning_rate": 2.261285487135105e-05, + "loss": 3.9062, + "step": 1672704 + }, + { + "epoch": 0.02, + "learning_rate": 2.260446892384053e-05, + "loss": 3.9034, + "step": 1673216 + }, + { + "epoch": 0.02, + "learning_rate": 2.259609935513374e-05, + "loss": 3.9045, + "step": 1673728 + }, + { + "epoch": 0.02, + "learning_rate": 2.258771340762322e-05, + "loss": 3.9102, + "step": 1674240 + }, + { + "epoch": 0.02, + "learning_rate": 2.25793274601127e-05, + "loss": 3.9025, + "step": 1674752 + }, + { + "epoch": 0.02, + "learning_rate": 2.257094151260218e-05, + "loss": 3.902, + "step": 1675264 + }, + { + "epoch": 0.02, + "learning_rate": 2.256255556509166e-05, + "loss": 3.9016, + "step": 1675776 + }, + { + "epoch": 0.02, + "learning_rate": 2.2554169617581138e-05, + "loss": 3.9079, + "step": 1676288 + }, + { + "epoch": 0.02, + "learning_rate": 2.254578367007062e-05, + "loss": 3.9078, + "step": 1676800 + }, + { + "epoch": 0.02, + "learning_rate": 2.25373977225601e-05, + "loss": 3.9073, + "step": 1677312 + }, + { + "epoch": 0.02, + "learning_rate": 2.252902815385331e-05, + "loss": 3.9095, + "step": 1677824 + }, + { + "epoch": 0.02, + "learning_rate": 2.252064220634279e-05, + "loss": 3.9022, + "step": 1678336 + }, + { + "epoch": 0.02, + "learning_rate": 2.2512272637636002e-05, + "loss": 3.9072, + "step": 1678848 + }, + { + "epoch": 0.03, + "eval_loss": 3.9921131134033203, + "eval_runtime": 315.453, + "eval_samples_per_second": 1209.66, + "eval_steps_per_second": 37.803, + "step": 1679027 + }, + { + "epoch": 1.0, + "learning_rate": 2.2503886690125482e-05, + "loss": 3.9052, + "step": 1679360 + }, + { + "epoch": 1.0, + "learning_rate": 2.2495500742614962e-05, + "loss": 3.8935, + "step": 1679872 + }, + { + "epoch": 1.0, + "learning_rate": 2.2487114795104442e-05, + "loss": 3.91, + "step": 1680384 + }, + { + "epoch": 1.0, + "learning_rate": 2.2478728847593922e-05, + "loss": 3.9068, + "step": 1680896 + }, + { + "epoch": 1.0, + "learning_rate": 2.2470342900083402e-05, + "loss": 3.9168, + "step": 1681408 + }, + { + "epoch": 1.0, + "learning_rate": 2.2461956952572882e-05, + "loss": 3.9005, + "step": 1681920 + }, + { + "epoch": 1.0, + "learning_rate": 2.2453571005062362e-05, + "loss": 3.8988, + "step": 1682432 + }, + { + "epoch": 1.0, + "learning_rate": 2.2445185057551842e-05, + "loss": 3.8995, + "step": 1682944 + }, + { + "epoch": 1.0, + "learning_rate": 2.2436815488845055e-05, + "loss": 3.8999, + "step": 1683456 + }, + { + "epoch": 1.0, + "learning_rate": 2.2428429541334535e-05, + "loss": 3.8987, + "step": 1683968 + }, + { + "epoch": 1.0, + "learning_rate": 2.2420043593824015e-05, + "loss": 3.9132, + "step": 1684480 + }, + { + "epoch": 1.0, + "learning_rate": 2.241165764631349e-05, + "loss": 3.9125, + "step": 1684992 + }, + { + "epoch": 1.0, + "learning_rate": 2.2403304456410433e-05, + "loss": 3.8913, + "step": 1685504 + }, + { + "epoch": 1.0, + "learning_rate": 2.2394918508899916e-05, + "loss": 3.8984, + "step": 1686016 + }, + { + "epoch": 1.0, + "learning_rate": 2.2386532561389396e-05, + "loss": 3.9083, + "step": 1686528 + }, + { + "epoch": 1.0, + "learning_rate": 2.2378146613878876e-05, + "loss": 3.8903, + "step": 1687040 + }, + { + "epoch": 1.0, + "learning_rate": 2.2369760666368356e-05, + "loss": 3.9, + "step": 1687552 + }, + { + "epoch": 1.0, + "learning_rate": 2.2361374718857836e-05, + "loss": 3.9004, + "step": 1688064 + }, + { + "epoch": 1.0, + "learning_rate": 2.2352988771347316e-05, + "loss": 3.8998, + "step": 1688576 + }, + { + "epoch": 1.0, + "learning_rate": 2.2344602823836792e-05, + "loss": 3.9183, + "step": 1689088 + }, + { + "epoch": 1.0, + "learning_rate": 2.2336216876326276e-05, + "loss": 3.9047, + "step": 1689600 + }, + { + "epoch": 1.0, + "learning_rate": 2.2327830928815756e-05, + "loss": 3.907, + "step": 1690112 + }, + { + "epoch": 1.0, + "learning_rate": 2.2319444981305236e-05, + "loss": 3.905, + "step": 1690624 + }, + { + "epoch": 1.0, + "learning_rate": 2.2311059033794716e-05, + "loss": 3.9074, + "step": 1691136 + }, + { + "epoch": 1.0, + "learning_rate": 2.2302689465087925e-05, + "loss": 3.9013, + "step": 1691648 + }, + { + "epoch": 1.0, + "learning_rate": 2.2294319896381137e-05, + "loss": 3.9038, + "step": 1692160 + }, + { + "epoch": 1.0, + "learning_rate": 2.2285933948870617e-05, + "loss": 3.8987, + "step": 1692672 + }, + { + "epoch": 1.0, + "learning_rate": 2.227756438016383e-05, + "loss": 3.893, + "step": 1693184 + }, + { + "epoch": 1.0, + "learning_rate": 2.226917843265331e-05, + "loss": 3.8988, + "step": 1693696 + }, + { + "epoch": 1.0, + "learning_rate": 2.226079248514279e-05, + "loss": 3.9015, + "step": 1694208 + }, + { + "epoch": 1.01, + "learning_rate": 2.2252406537632266e-05, + "loss": 3.905, + "step": 1694720 + }, + { + "epoch": 1.01, + "learning_rate": 2.2244020590121746e-05, + "loss": 3.9033, + "step": 1695232 + }, + { + "epoch": 1.01, + "learning_rate": 2.223563464261123e-05, + "loss": 3.909, + "step": 1695744 + }, + { + "epoch": 1.01, + "learning_rate": 2.222724869510071e-05, + "loss": 3.9104, + "step": 1696256 + }, + { + "epoch": 1.01, + "learning_rate": 2.221887912639392e-05, + "loss": 3.8998, + "step": 1696768 + }, + { + "epoch": 1.01, + "learning_rate": 2.22104931788834e-05, + "loss": 3.9031, + "step": 1697280 + }, + { + "epoch": 1.01, + "learning_rate": 2.220210723137288e-05, + "loss": 3.9029, + "step": 1697792 + }, + { + "epoch": 1.01, + "learning_rate": 2.219372128386236e-05, + "loss": 3.9004, + "step": 1698304 + }, + { + "epoch": 1.01, + "learning_rate": 2.2185335336351838e-05, + "loss": 3.892, + "step": 1698816 + }, + { + "epoch": 1.01, + "learning_rate": 2.2176949388841318e-05, + "loss": 3.8975, + "step": 1699328 + }, + { + "epoch": 1.01, + "learning_rate": 2.2168563441330798e-05, + "loss": 3.8994, + "step": 1699840 + }, + { + "epoch": 1.01, + "learning_rate": 2.2160177493820278e-05, + "loss": 3.9075, + "step": 1700352 + }, + { + "epoch": 1.01, + "learning_rate": 2.2151791546309758e-05, + "loss": 3.8958, + "step": 1700864 + }, + { + "epoch": 1.01, + "learning_rate": 2.2143421977602967e-05, + "loss": 3.9001, + "step": 1701376 + }, + { + "epoch": 1.01, + "learning_rate": 2.2135036030092447e-05, + "loss": 3.9156, + "step": 1701888 + }, + { + "epoch": 1.01, + "learning_rate": 2.212665008258193e-05, + "loss": 3.9005, + "step": 1702400 + }, + { + "epoch": 1.01, + "learning_rate": 2.211826413507141e-05, + "loss": 3.9058, + "step": 1702912 + }, + { + "epoch": 1.01, + "learning_rate": 2.210987818756089e-05, + "loss": 3.8942, + "step": 1703424 + }, + { + "epoch": 1.01, + "learning_rate": 2.210149224005037e-05, + "loss": 3.8854, + "step": 1703936 + }, + { + "epoch": 1.01, + "learning_rate": 2.209310629253985e-05, + "loss": 3.911, + "step": 1704448 + }, + { + "epoch": 1.01, + "learning_rate": 2.2084720345029327e-05, + "loss": 3.8994, + "step": 1704960 + }, + { + "epoch": 1.01, + "learning_rate": 2.2076334397518807e-05, + "loss": 3.8927, + "step": 1705472 + }, + { + "epoch": 1.01, + "learning_rate": 2.2067948450008287e-05, + "loss": 3.8954, + "step": 1705984 + }, + { + "epoch": 1.01, + "learning_rate": 2.205956250249777e-05, + "loss": 3.9074, + "step": 1706496 + }, + { + "epoch": 1.01, + "learning_rate": 2.2051209312594712e-05, + "loss": 3.8796, + "step": 1707008 + }, + { + "epoch": 1.01, + "learning_rate": 2.2042823365084192e-05, + "loss": 3.8975, + "step": 1707520 + }, + { + "epoch": 1.01, + "learning_rate": 2.203443741757367e-05, + "loss": 3.8925, + "step": 1708032 + }, + { + "epoch": 1.01, + "learning_rate": 2.2026051470063148e-05, + "loss": 3.8945, + "step": 1708544 + }, + { + "epoch": 1.01, + "learning_rate": 2.201766552255263e-05, + "loss": 3.91, + "step": 1709056 + }, + { + "epoch": 1.01, + "learning_rate": 2.200927957504211e-05, + "loss": 3.8965, + "step": 1709568 + }, + { + "epoch": 1.01, + "learning_rate": 2.200089362753159e-05, + "loss": 3.8943, + "step": 1710080 + }, + { + "epoch": 1.01, + "learning_rate": 2.199250768002107e-05, + "loss": 3.9037, + "step": 1710592 + }, + { + "epoch": 1.01, + "learning_rate": 2.198413811131428e-05, + "loss": 3.8863, + "step": 1711104 + }, + { + "epoch": 1.01, + "learning_rate": 2.197575216380376e-05, + "loss": 3.8912, + "step": 1711616 + }, + { + "epoch": 1.01, + "learning_rate": 2.196736621629324e-05, + "loss": 3.8975, + "step": 1712128 + }, + { + "epoch": 1.01, + "learning_rate": 2.1958980268782724e-05, + "loss": 3.8967, + "step": 1712640 + }, + { + "epoch": 1.01, + "learning_rate": 2.1950610700075933e-05, + "loss": 3.8991, + "step": 1713152 + }, + { + "epoch": 1.01, + "learning_rate": 2.1942224752565413e-05, + "loss": 3.8894, + "step": 1713664 + }, + { + "epoch": 1.01, + "learning_rate": 2.1933838805054893e-05, + "loss": 3.8829, + "step": 1714176 + }, + { + "epoch": 1.01, + "learning_rate": 2.1925452857544373e-05, + "loss": 3.8947, + "step": 1714688 + }, + { + "epoch": 1.01, + "learning_rate": 2.1917066910033853e-05, + "loss": 3.9107, + "step": 1715200 + }, + { + "epoch": 1.01, + "learning_rate": 2.1908697341327065e-05, + "loss": 3.8936, + "step": 1715712 + }, + { + "epoch": 1.01, + "learning_rate": 2.1900311393816545e-05, + "loss": 3.9124, + "step": 1716224 + }, + { + "epoch": 1.01, + "learning_rate": 2.1891925446306025e-05, + "loss": 3.9084, + "step": 1716736 + }, + { + "epoch": 1.01, + "learning_rate": 2.1883539498795505e-05, + "loss": 3.9017, + "step": 1717248 + }, + { + "epoch": 1.01, + "learning_rate": 2.1875169930088714e-05, + "loss": 3.8906, + "step": 1717760 + }, + { + "epoch": 1.01, + "learning_rate": 2.1866783982578194e-05, + "loss": 3.9083, + "step": 1718272 + }, + { + "epoch": 1.01, + "learning_rate": 2.1858398035067677e-05, + "loss": 3.8962, + "step": 1718784 + }, + { + "epoch": 1.01, + "learning_rate": 2.1850012087557157e-05, + "loss": 3.8949, + "step": 1719296 + }, + { + "epoch": 1.01, + "learning_rate": 2.1841642518850367e-05, + "loss": 3.9043, + "step": 1719808 + }, + { + "epoch": 1.01, + "learning_rate": 2.1833272950143576e-05, + "loss": 3.902, + "step": 1720320 + }, + { + "epoch": 1.01, + "learning_rate": 2.1824887002633056e-05, + "loss": 3.9106, + "step": 1720832 + }, + { + "epoch": 1.01, + "learning_rate": 2.181650105512254e-05, + "loss": 3.9012, + "step": 1721344 + }, + { + "epoch": 1.01, + "learning_rate": 2.180811510761202e-05, + "loss": 3.8849, + "step": 1721856 + }, + { + "epoch": 1.01, + "learning_rate": 2.1799745538905228e-05, + "loss": 3.8852, + "step": 1722368 + }, + { + "epoch": 1.01, + "learning_rate": 2.1791359591394708e-05, + "loss": 3.8936, + "step": 1722880 + }, + { + "epoch": 1.01, + "learning_rate": 2.1782973643884188e-05, + "loss": 3.9017, + "step": 1723392 + }, + { + "epoch": 1.01, + "learning_rate": 2.1774587696373668e-05, + "loss": 3.8969, + "step": 1723904 + }, + { + "epoch": 1.01, + "learning_rate": 2.1766201748863148e-05, + "loss": 3.8963, + "step": 1724416 + }, + { + "epoch": 1.02, + "learning_rate": 2.175781580135263e-05, + "loss": 3.8987, + "step": 1724928 + }, + { + "epoch": 1.02, + "learning_rate": 2.1749429853842108e-05, + "loss": 3.8908, + "step": 1725440 + }, + { + "epoch": 1.02, + "learning_rate": 2.1741043906331588e-05, + "loss": 3.8973, + "step": 1725952 + }, + { + "epoch": 1.02, + "learning_rate": 2.1732657958821068e-05, + "loss": 3.8929, + "step": 1726464 + }, + { + "epoch": 1.02, + "learning_rate": 2.1724272011310548e-05, + "loss": 3.9122, + "step": 1726976 + }, + { + "epoch": 1.02, + "learning_rate": 2.1715886063800028e-05, + "loss": 3.9058, + "step": 1727488 + }, + { + "epoch": 1.02, + "learning_rate": 2.1707500116289507e-05, + "loss": 3.902, + "step": 1728000 + }, + { + "epoch": 1.02, + "learning_rate": 2.169913054758272e-05, + "loss": 3.8921, + "step": 1728512 + }, + { + "epoch": 1.02, + "learning_rate": 2.16907446000722e-05, + "loss": 3.8991, + "step": 1729024 + }, + { + "epoch": 1.02, + "learning_rate": 2.168235865256168e-05, + "loss": 3.8951, + "step": 1729536 + }, + { + "epoch": 1.02, + "learning_rate": 2.167397270505116e-05, + "loss": 3.9111, + "step": 1730048 + }, + { + "epoch": 1.02, + "learning_rate": 2.166560313634437e-05, + "loss": 3.896, + "step": 1730560 + }, + { + "epoch": 1.02, + "learning_rate": 2.165721718883385e-05, + "loss": 3.8912, + "step": 1731072 + }, + { + "epoch": 1.02, + "learning_rate": 2.1648831241323332e-05, + "loss": 3.9044, + "step": 1731584 + }, + { + "epoch": 1.02, + "learning_rate": 2.164044529381281e-05, + "loss": 3.9001, + "step": 1732096 + }, + { + "epoch": 1.02, + "learning_rate": 2.163207572510602e-05, + "loss": 3.8925, + "step": 1732608 + }, + { + "epoch": 1.02, + "learning_rate": 2.162370615639923e-05, + "loss": 3.8985, + "step": 1733120 + }, + { + "epoch": 1.02, + "learning_rate": 2.161532020888871e-05, + "loss": 3.8905, + "step": 1733632 + }, + { + "epoch": 1.02, + "learning_rate": 2.1606934261378194e-05, + "loss": 3.8885, + "step": 1734144 + }, + { + "epoch": 1.02, + "learning_rate": 2.1598548313867674e-05, + "loss": 3.8827, + "step": 1734656 + }, + { + "epoch": 1.02, + "learning_rate": 2.1590162366357154e-05, + "loss": 3.9019, + "step": 1735168 + }, + { + "epoch": 1.02, + "learning_rate": 2.1581776418846634e-05, + "loss": 3.8941, + "step": 1735680 + }, + { + "epoch": 1.02, + "learning_rate": 2.157339047133611e-05, + "loss": 3.8942, + "step": 1736192 + }, + { + "epoch": 1.02, + "learning_rate": 2.156500452382559e-05, + "loss": 3.8935, + "step": 1736704 + }, + { + "epoch": 1.02, + "learning_rate": 2.155661857631507e-05, + "loss": 3.8891, + "step": 1737216 + }, + { + "epoch": 1.02, + "learning_rate": 2.1548249007608283e-05, + "loss": 3.9076, + "step": 1737728 + }, + { + "epoch": 1.02, + "learning_rate": 2.1539863060097762e-05, + "loss": 3.9103, + "step": 1738240 + }, + { + "epoch": 1.02, + "learning_rate": 2.1531477112587242e-05, + "loss": 3.9053, + "step": 1738752 + }, + { + "epoch": 1.02, + "learning_rate": 2.1523091165076722e-05, + "loss": 3.8977, + "step": 1739264 + }, + { + "epoch": 1.02, + "learning_rate": 2.151472159636993e-05, + "loss": 3.8921, + "step": 1739776 + }, + { + "epoch": 1.02, + "learning_rate": 2.1506352027663147e-05, + "loss": 3.8961, + "step": 1740288 + }, + { + "epoch": 1.02, + "learning_rate": 2.1497966080152627e-05, + "loss": 3.8884, + "step": 1740800 + }, + { + "epoch": 1.02, + "learning_rate": 2.1489580132642107e-05, + "loss": 3.8963, + "step": 1741312 + }, + { + "epoch": 1.02, + "learning_rate": 2.1481194185131584e-05, + "loss": 3.8951, + "step": 1741824 + }, + { + "epoch": 1.02, + "learning_rate": 2.1472808237621064e-05, + "loss": 3.8977, + "step": 1742336 + }, + { + "epoch": 1.02, + "learning_rate": 2.1464438668914276e-05, + "loss": 3.9028, + "step": 1742848 + }, + { + "epoch": 1.02, + "learning_rate": 2.1456052721403756e-05, + "loss": 3.8934, + "step": 1743360 + }, + { + "epoch": 1.02, + "learning_rate": 2.1447666773893236e-05, + "loss": 3.9011, + "step": 1743872 + }, + { + "epoch": 1.02, + "learning_rate": 2.1439280826382716e-05, + "loss": 3.892, + "step": 1744384 + }, + { + "epoch": 1.02, + "learning_rate": 2.1430894878872196e-05, + "loss": 3.8948, + "step": 1744896 + }, + { + "epoch": 1.02, + "learning_rate": 2.1422525310165405e-05, + "loss": 3.8983, + "step": 1745408 + }, + { + "epoch": 1.02, + "learning_rate": 2.1414139362654885e-05, + "loss": 3.9109, + "step": 1745920 + }, + { + "epoch": 1.02, + "learning_rate": 2.1405753415144365e-05, + "loss": 3.9061, + "step": 1746432 + }, + { + "epoch": 1.02, + "learning_rate": 2.139736746763385e-05, + "loss": 3.8956, + "step": 1746944 + }, + { + "epoch": 1.02, + "learning_rate": 2.138898152012333e-05, + "loss": 3.9039, + "step": 1747456 + }, + { + "epoch": 1.02, + "learning_rate": 2.138059557261281e-05, + "loss": 3.9005, + "step": 1747968 + }, + { + "epoch": 1.02, + "learning_rate": 2.137220962510229e-05, + "loss": 3.9082, + "step": 1748480 + }, + { + "epoch": 1.02, + "learning_rate": 2.1363823677591765e-05, + "loss": 3.9008, + "step": 1748992 + }, + { + "epoch": 1.02, + "learning_rate": 2.1355454108884977e-05, + "loss": 3.9016, + "step": 1749504 + }, + { + "epoch": 1.02, + "learning_rate": 2.1347068161374457e-05, + "loss": 3.8958, + "step": 1750016 + }, + { + "epoch": 1.02, + "learning_rate": 2.1338682213863937e-05, + "loss": 3.9025, + "step": 1750528 + }, + { + "epoch": 1.02, + "learning_rate": 2.1330296266353417e-05, + "loss": 3.8937, + "step": 1751040 + }, + { + "epoch": 1.02, + "learning_rate": 2.1321910318842897e-05, + "loss": 3.8958, + "step": 1751552 + }, + { + "epoch": 1.02, + "learning_rate": 2.131354075013611e-05, + "loss": 3.8988, + "step": 1752064 + }, + { + "epoch": 1.02, + "learning_rate": 2.1305154802625586e-05, + "loss": 3.8993, + "step": 1752576 + }, + { + "epoch": 1.02, + "learning_rate": 2.129676885511507e-05, + "loss": 3.898, + "step": 1753088 + }, + { + "epoch": 1.02, + "learning_rate": 2.128838290760455e-05, + "loss": 3.8987, + "step": 1753600 + }, + { + "epoch": 1.02, + "learning_rate": 2.1280013338897762e-05, + "loss": 3.9061, + "step": 1754112 + }, + { + "epoch": 1.02, + "learning_rate": 2.127162739138724e-05, + "loss": 3.8938, + "step": 1754624 + }, + { + "epoch": 1.02, + "learning_rate": 2.126324144387672e-05, + "loss": 3.9062, + "step": 1755136 + }, + { + "epoch": 1.03, + "eval_loss": 3.9889142513275146, + "eval_runtime": 292.5433, + "eval_samples_per_second": 1304.391, + "eval_steps_per_second": 40.763, + "step": 1755347 + }, + { + "epoch": 0.0, + "learning_rate": 2.12548554963662e-05, + "loss": 3.893, + "step": 1755648 + }, + { + "epoch": 0.0, + "learning_rate": 2.124646954885568e-05, + "loss": 3.8852, + "step": 1756160 + }, + { + "epoch": 0.0, + "learning_rate": 2.123808360134516e-05, + "loss": 3.9032, + "step": 1756672 + }, + { + "epoch": 0.0, + "learning_rate": 2.1229697653834642e-05, + "loss": 3.8984, + "step": 1757184 + }, + { + "epoch": 0.0, + "learning_rate": 2.122131170632412e-05, + "loss": 3.9059, + "step": 1757696 + }, + { + "epoch": 0.0, + "learning_rate": 2.1212925758813598e-05, + "loss": 3.8942, + "step": 1758208 + }, + { + "epoch": 0.0, + "learning_rate": 2.1204539811303078e-05, + "loss": 3.8933, + "step": 1758720 + }, + { + "epoch": 0.0, + "learning_rate": 2.1196153863792558e-05, + "loss": 3.893, + "step": 1759232 + }, + { + "epoch": 0.0, + "learning_rate": 2.1187767916282038e-05, + "loss": 3.8976, + "step": 1759744 + }, + { + "epoch": 0.0, + "learning_rate": 2.1179381968771518e-05, + "loss": 3.8893, + "step": 1760256 + }, + { + "epoch": 0.0, + "learning_rate": 2.1170996021261e-05, + "loss": 3.9093, + "step": 1760768 + }, + { + "epoch": 0.0, + "learning_rate": 2.1162610073750478e-05, + "loss": 3.9058, + "step": 1761280 + }, + { + "epoch": 0.0, + "learning_rate": 2.115424050504369e-05, + "loss": 3.8882, + "step": 1761792 + }, + { + "epoch": 0.0, + "learning_rate": 2.114585455753317e-05, + "loss": 3.8892, + "step": 1762304 + }, + { + "epoch": 0.0, + "learning_rate": 2.113746861002265e-05, + "loss": 3.9012, + "step": 1762816 + }, + { + "epoch": 0.0, + "learning_rate": 2.1129082662512127e-05, + "loss": 3.8856, + "step": 1763328 + }, + { + "epoch": 0.0, + "learning_rate": 2.112069671500161e-05, + "loss": 3.8933, + "step": 1763840 + }, + { + "epoch": 0.0, + "learning_rate": 2.111231076749109e-05, + "loss": 3.8908, + "step": 1764352 + }, + { + "epoch": 0.0, + "learning_rate": 2.110392481998057e-05, + "loss": 3.8984, + "step": 1764864 + }, + { + "epoch": 0.0, + "learning_rate": 2.109553887247005e-05, + "loss": 3.9105, + "step": 1765376 + }, + { + "epoch": 0.0, + "learning_rate": 2.108715292495953e-05, + "loss": 3.9029, + "step": 1765888 + }, + { + "epoch": 0.0, + "learning_rate": 2.1078766977449007e-05, + "loss": 3.8994, + "step": 1766400 + }, + { + "epoch": 0.0, + "learning_rate": 2.1070381029938487e-05, + "loss": 3.8964, + "step": 1766912 + }, + { + "epoch": 0.0, + "learning_rate": 2.1062011461231703e-05, + "loss": 3.9007, + "step": 1767424 + }, + { + "epoch": 0.0, + "learning_rate": 2.1053625513721183e-05, + "loss": 3.8907, + "step": 1767936 + }, + { + "epoch": 0.0, + "learning_rate": 2.104523956621066e-05, + "loss": 3.9006, + "step": 1768448 + }, + { + "epoch": 0.0, + "learning_rate": 2.103685361870014e-05, + "loss": 3.8902, + "step": 1768960 + }, + { + "epoch": 0.0, + "learning_rate": 2.102848404999335e-05, + "loss": 3.8867, + "step": 1769472 + }, + { + "epoch": 0.0, + "learning_rate": 2.1020098102482828e-05, + "loss": 3.8948, + "step": 1769984 + }, + { + "epoch": 0.0, + "learning_rate": 2.101171215497231e-05, + "loss": 3.8989, + "step": 1770496 + }, + { + "epoch": 0.01, + "learning_rate": 2.100332620746179e-05, + "loss": 3.894, + "step": 1771008 + }, + { + "epoch": 0.01, + "learning_rate": 2.099494025995127e-05, + "loss": 3.8983, + "step": 1771520 + }, + { + "epoch": 0.01, + "learning_rate": 2.098655431244075e-05, + "loss": 3.9016, + "step": 1772032 + }, + { + "epoch": 0.01, + "learning_rate": 2.097816836493023e-05, + "loss": 3.9059, + "step": 1772544 + }, + { + "epoch": 0.01, + "learning_rate": 2.0969815175027173e-05, + "loss": 3.8896, + "step": 1773056 + }, + { + "epoch": 0.01, + "learning_rate": 2.0961429227516656e-05, + "loss": 3.8984, + "step": 1773568 + }, + { + "epoch": 0.01, + "learning_rate": 2.0953043280006133e-05, + "loss": 3.901, + "step": 1774080 + }, + { + "epoch": 0.01, + "learning_rate": 2.0944657332495613e-05, + "loss": 3.89, + "step": 1774592 + }, + { + "epoch": 0.01, + "learning_rate": 2.0936271384985093e-05, + "loss": 3.8847, + "step": 1775104 + }, + { + "epoch": 0.01, + "learning_rate": 2.0927885437474573e-05, + "loss": 3.8902, + "step": 1775616 + }, + { + "epoch": 0.01, + "learning_rate": 2.0919499489964053e-05, + "loss": 3.8944, + "step": 1776128 + }, + { + "epoch": 0.01, + "learning_rate": 2.0911113542453533e-05, + "loss": 3.9019, + "step": 1776640 + }, + { + "epoch": 0.01, + "learning_rate": 2.0902727594943013e-05, + "loss": 3.8872, + "step": 1777152 + }, + { + "epoch": 0.01, + "learning_rate": 2.0894341647432492e-05, + "loss": 3.8908, + "step": 1777664 + }, + { + "epoch": 0.01, + "learning_rate": 2.0885988457529434e-05, + "loss": 3.9104, + "step": 1778176 + }, + { + "epoch": 0.01, + "learning_rate": 2.0877602510018914e-05, + "loss": 3.8973, + "step": 1778688 + }, + { + "epoch": 0.01, + "learning_rate": 2.0869216562508394e-05, + "loss": 3.9032, + "step": 1779200 + }, + { + "epoch": 0.01, + "learning_rate": 2.0860830614997874e-05, + "loss": 3.8831, + "step": 1779712 + }, + { + "epoch": 0.01, + "learning_rate": 2.0852444667487357e-05, + "loss": 3.8778, + "step": 1780224 + }, + { + "epoch": 0.01, + "learning_rate": 2.0844058719976837e-05, + "loss": 3.9076, + "step": 1780736 + }, + { + "epoch": 0.01, + "learning_rate": 2.0835689151270046e-05, + "loss": 3.8919, + "step": 1781248 + }, + { + "epoch": 0.01, + "learning_rate": 2.0827303203759526e-05, + "loss": 3.8878, + "step": 1781760 + }, + { + "epoch": 0.01, + "learning_rate": 2.0818917256249006e-05, + "loss": 3.8927, + "step": 1782272 + }, + { + "epoch": 0.01, + "learning_rate": 2.0810531308738486e-05, + "loss": 3.8937, + "step": 1782784 + }, + { + "epoch": 0.01, + "learning_rate": 2.0802145361227966e-05, + "loss": 3.8794, + "step": 1783296 + }, + { + "epoch": 0.01, + "learning_rate": 2.0793759413717446e-05, + "loss": 3.8879, + "step": 1783808 + }, + { + "epoch": 0.01, + "learning_rate": 2.0785373466206926e-05, + "loss": 3.8863, + "step": 1784320 + }, + { + "epoch": 0.01, + "learning_rate": 2.0776987518696406e-05, + "loss": 3.8882, + "step": 1784832 + }, + { + "epoch": 0.01, + "learning_rate": 2.0768601571185886e-05, + "loss": 3.9019, + "step": 1785344 + }, + { + "epoch": 0.01, + "learning_rate": 2.0760215623675366e-05, + "loss": 3.8904, + "step": 1785856 + }, + { + "epoch": 0.01, + "learning_rate": 2.0751829676164843e-05, + "loss": 3.8867, + "step": 1786368 + }, + { + "epoch": 0.01, + "learning_rate": 2.0743443728654326e-05, + "loss": 3.9003, + "step": 1786880 + }, + { + "epoch": 0.01, + "learning_rate": 2.073507415994754e-05, + "loss": 3.8794, + "step": 1787392 + }, + { + "epoch": 0.01, + "learning_rate": 2.072668821243702e-05, + "loss": 3.8892, + "step": 1787904 + }, + { + "epoch": 0.01, + "learning_rate": 2.0718302264926495e-05, + "loss": 3.8904, + "step": 1788416 + }, + { + "epoch": 0.01, + "learning_rate": 2.0709916317415975e-05, + "loss": 3.8897, + "step": 1788928 + }, + { + "epoch": 0.01, + "learning_rate": 2.070156312751292e-05, + "loss": 3.8913, + "step": 1789440 + }, + { + "epoch": 0.01, + "learning_rate": 2.06931771800024e-05, + "loss": 3.8905, + "step": 1789952 + }, + { + "epoch": 0.01, + "learning_rate": 2.068479123249188e-05, + "loss": 3.8736, + "step": 1790464 + }, + { + "epoch": 0.01, + "learning_rate": 2.067640528498136e-05, + "loss": 3.8872, + "step": 1790976 + }, + { + "epoch": 0.01, + "learning_rate": 2.066801933747084e-05, + "loss": 3.9043, + "step": 1791488 + }, + { + "epoch": 0.01, + "learning_rate": 2.0659633389960316e-05, + "loss": 3.8828, + "step": 1792000 + }, + { + "epoch": 0.01, + "learning_rate": 2.0651247442449796e-05, + "loss": 3.909, + "step": 1792512 + }, + { + "epoch": 0.01, + "learning_rate": 2.064286149493928e-05, + "loss": 3.8982, + "step": 1793024 + }, + { + "epoch": 0.01, + "learning_rate": 2.0634491926232492e-05, + "loss": 3.8993, + "step": 1793536 + }, + { + "epoch": 0.01, + "learning_rate": 2.062610597872197e-05, + "loss": 3.8841, + "step": 1794048 + }, + { + "epoch": 0.01, + "learning_rate": 2.061773641001518e-05, + "loss": 3.8978, + "step": 1794560 + }, + { + "epoch": 0.01, + "learning_rate": 2.060935046250466e-05, + "loss": 3.8977, + "step": 1795072 + }, + { + "epoch": 0.01, + "learning_rate": 2.060096451499414e-05, + "loss": 3.8826, + "step": 1795584 + }, + { + "epoch": 0.01, + "learning_rate": 2.059257856748362e-05, + "loss": 3.9026, + "step": 1796096 + }, + { + "epoch": 0.01, + "learning_rate": 2.0584208998776833e-05, + "loss": 3.897, + "step": 1796608 + }, + { + "epoch": 0.01, + "learning_rate": 2.0575823051266313e-05, + "loss": 3.9013, + "step": 1797120 + }, + { + "epoch": 0.01, + "learning_rate": 2.056743710375579e-05, + "loss": 3.8982, + "step": 1797632 + }, + { + "epoch": 0.01, + "learning_rate": 2.055905115624527e-05, + "loss": 3.8781, + "step": 1798144 + }, + { + "epoch": 0.01, + "learning_rate": 2.055066520873475e-05, + "loss": 3.8791, + "step": 1798656 + }, + { + "epoch": 0.01, + "learning_rate": 2.0542295640027966e-05, + "loss": 3.8831, + "step": 1799168 + }, + { + "epoch": 0.01, + "learning_rate": 2.0533909692517442e-05, + "loss": 3.8962, + "step": 1799680 + }, + { + "epoch": 0.01, + "learning_rate": 2.0525523745006922e-05, + "loss": 3.8925, + "step": 1800192 + }, + { + "epoch": 0.01, + "learning_rate": 2.0517137797496402e-05, + "loss": 3.8897, + "step": 1800704 + }, + { + "epoch": 0.02, + "learning_rate": 2.0508751849985882e-05, + "loss": 3.8909, + "step": 1801216 + }, + { + "epoch": 0.02, + "learning_rate": 2.0500365902475362e-05, + "loss": 3.8884, + "step": 1801728 + }, + { + "epoch": 0.02, + "learning_rate": 2.0491996333768575e-05, + "loss": 3.8884, + "step": 1802240 + }, + { + "epoch": 0.02, + "learning_rate": 2.0483610386258055e-05, + "loss": 3.8891, + "step": 1802752 + }, + { + "epoch": 0.02, + "learning_rate": 2.0475224438747535e-05, + "loss": 3.9036, + "step": 1803264 + }, + { + "epoch": 0.02, + "learning_rate": 2.0466838491237014e-05, + "loss": 3.9062, + "step": 1803776 + }, + { + "epoch": 0.02, + "learning_rate": 2.0458468922530224e-05, + "loss": 3.8904, + "step": 1804288 + }, + { + "epoch": 0.02, + "learning_rate": 2.0450082975019704e-05, + "loss": 3.8863, + "step": 1804800 + }, + { + "epoch": 0.02, + "learning_rate": 2.0441697027509183e-05, + "loss": 3.8913, + "step": 1805312 + }, + { + "epoch": 0.02, + "learning_rate": 2.0433311079998667e-05, + "loss": 3.8906, + "step": 1805824 + }, + { + "epoch": 0.02, + "learning_rate": 2.0424925132488143e-05, + "loss": 3.9048, + "step": 1806336 + }, + { + "epoch": 0.02, + "learning_rate": 2.0416555563781356e-05, + "loss": 3.8902, + "step": 1806848 + }, + { + "epoch": 0.02, + "learning_rate": 2.0408169616270836e-05, + "loss": 3.8884, + "step": 1807360 + }, + { + "epoch": 0.02, + "learning_rate": 2.0399783668760316e-05, + "loss": 3.897, + "step": 1807872 + }, + { + "epoch": 0.02, + "learning_rate": 2.0391397721249796e-05, + "loss": 3.8964, + "step": 1808384 + }, + { + "epoch": 0.02, + "learning_rate": 2.0383011773739276e-05, + "loss": 3.8816, + "step": 1808896 + }, + { + "epoch": 0.02, + "learning_rate": 2.0374625826228756e-05, + "loss": 3.892, + "step": 1809408 + }, + { + "epoch": 0.02, + "learning_rate": 2.0366256257521968e-05, + "loss": 3.8884, + "step": 1809920 + }, + { + "epoch": 0.02, + "learning_rate": 2.0357870310011445e-05, + "loss": 3.8797, + "step": 1810432 + }, + { + "epoch": 0.02, + "learning_rate": 2.0349484362500925e-05, + "loss": 3.884, + "step": 1810944 + }, + { + "epoch": 0.02, + "learning_rate": 2.0341098414990405e-05, + "loss": 3.8925, + "step": 1811456 + }, + { + "epoch": 0.02, + "learning_rate": 2.0332712467479888e-05, + "loss": 3.8911, + "step": 1811968 + }, + { + "epoch": 0.02, + "learning_rate": 2.0324326519969368e-05, + "loss": 3.8847, + "step": 1812480 + }, + { + "epoch": 0.02, + "learning_rate": 2.0315940572458848e-05, + "loss": 3.8875, + "step": 1812992 + }, + { + "epoch": 0.02, + "learning_rate": 2.0307554624948324e-05, + "loss": 3.889, + "step": 1813504 + }, + { + "epoch": 0.02, + "learning_rate": 2.0299185056241537e-05, + "loss": 3.8977, + "step": 1814016 + }, + { + "epoch": 0.02, + "learning_rate": 2.0290799108731017e-05, + "loss": 3.9062, + "step": 1814528 + }, + { + "epoch": 0.02, + "learning_rate": 2.0282413161220497e-05, + "loss": 3.8987, + "step": 1815040 + }, + { + "epoch": 0.02, + "learning_rate": 2.0274027213709977e-05, + "loss": 3.8907, + "step": 1815552 + }, + { + "epoch": 0.02, + "learning_rate": 2.026565764500319e-05, + "loss": 3.8858, + "step": 1816064 + }, + { + "epoch": 0.02, + "learning_rate": 2.025727169749267e-05, + "loss": 3.8953, + "step": 1816576 + }, + { + "epoch": 0.02, + "learning_rate": 2.024888574998215e-05, + "loss": 3.8806, + "step": 1817088 + }, + { + "epoch": 0.02, + "learning_rate": 2.0240499802471626e-05, + "loss": 3.8926, + "step": 1817600 + }, + { + "epoch": 0.02, + "learning_rate": 2.023213023376484e-05, + "loss": 3.8915, + "step": 1818112 + }, + { + "epoch": 0.02, + "learning_rate": 2.022374428625432e-05, + "loss": 3.8891, + "step": 1818624 + }, + { + "epoch": 0.02, + "learning_rate": 2.0215358338743798e-05, + "loss": 3.8936, + "step": 1819136 + }, + { + "epoch": 0.02, + "learning_rate": 2.0206972391233278e-05, + "loss": 3.8914, + "step": 1819648 + }, + { + "epoch": 0.02, + "learning_rate": 2.019860282252649e-05, + "loss": 3.8976, + "step": 1820160 + }, + { + "epoch": 0.02, + "learning_rate": 2.019021687501597e-05, + "loss": 3.8884, + "step": 1820672 + }, + { + "epoch": 0.02, + "learning_rate": 2.018183092750545e-05, + "loss": 3.8874, + "step": 1821184 + }, + { + "epoch": 0.02, + "learning_rate": 2.017344497999493e-05, + "loss": 3.8943, + "step": 1821696 + }, + { + "epoch": 0.02, + "learning_rate": 2.0165075411288143e-05, + "loss": 3.8994, + "step": 1822208 + }, + { + "epoch": 0.02, + "learning_rate": 2.0156689463777623e-05, + "loss": 3.904, + "step": 1822720 + }, + { + "epoch": 0.02, + "learning_rate": 2.01483035162671e-05, + "loss": 3.8887, + "step": 1823232 + }, + { + "epoch": 0.02, + "learning_rate": 2.013991756875658e-05, + "loss": 3.8973, + "step": 1823744 + }, + { + "epoch": 0.02, + "learning_rate": 2.0131548000049792e-05, + "loss": 3.8958, + "step": 1824256 + }, + { + "epoch": 0.02, + "learning_rate": 2.0123178431343004e-05, + "loss": 3.9003, + "step": 1824768 + }, + { + "epoch": 0.02, + "learning_rate": 2.0114792483832484e-05, + "loss": 3.8912, + "step": 1825280 + }, + { + "epoch": 0.02, + "learning_rate": 2.0106406536321964e-05, + "loss": 3.8953, + "step": 1825792 + }, + { + "epoch": 0.02, + "learning_rate": 2.0098020588811444e-05, + "loss": 3.8887, + "step": 1826304 + }, + { + "epoch": 0.02, + "learning_rate": 2.008963464130092e-05, + "loss": 3.9024, + "step": 1826816 + }, + { + "epoch": 0.02, + "learning_rate": 2.0081248693790404e-05, + "loss": 3.8842, + "step": 1827328 + }, + { + "epoch": 0.02, + "learning_rate": 2.0072862746279884e-05, + "loss": 3.8925, + "step": 1827840 + }, + { + "epoch": 0.02, + "learning_rate": 2.0064476798769364e-05, + "loss": 3.8931, + "step": 1828352 + }, + { + "epoch": 0.02, + "learning_rate": 2.0056107230062573e-05, + "loss": 3.8954, + "step": 1828864 + }, + { + "epoch": 0.02, + "learning_rate": 2.0047737661355786e-05, + "loss": 3.89, + "step": 1829376 + }, + { + "epoch": 0.02, + "learning_rate": 2.0039351713845266e-05, + "loss": 3.8923, + "step": 1829888 + }, + { + "epoch": 0.02, + "learning_rate": 2.0030965766334746e-05, + "loss": 3.9016, + "step": 1830400 + }, + { + "epoch": 0.02, + "learning_rate": 2.0022579818824226e-05, + "loss": 3.8859, + "step": 1830912 + }, + { + "epoch": 0.02, + "learning_rate": 2.0014193871313705e-05, + "loss": 3.9022, + "step": 1831424 + }, + { + "epoch": 0.03, + "eval_loss": 3.9869699478149414, + "eval_runtime": 297.3932, + "eval_samples_per_second": 1283.119, + "eval_steps_per_second": 40.098, + "step": 1831667 + }, + { + "epoch": 0.0, + "learning_rate": 2.0005807923803185e-05, + "loss": 3.8848, + "step": 1831936 + }, + { + "epoch": 0.0, + "learning_rate": 1.9997421976292665e-05, + "loss": 3.8821, + "step": 1832448 + }, + { + "epoch": 0.0, + "learning_rate": 1.9989036028782145e-05, + "loss": 3.8934, + "step": 1832960 + }, + { + "epoch": 0.0, + "learning_rate": 1.9980650081271625e-05, + "loss": 3.8934, + "step": 1833472 + }, + { + "epoch": 0.0, + "learning_rate": 1.9972264133761105e-05, + "loss": 3.9046, + "step": 1833984 + }, + { + "epoch": 0.0, + "learning_rate": 1.9963878186250585e-05, + "loss": 3.8868, + "step": 1834496 + }, + { + "epoch": 0.0, + "learning_rate": 1.9955492238740065e-05, + "loss": 3.8925, + "step": 1835008 + }, + { + "epoch": 0.0, + "learning_rate": 1.9947106291229545e-05, + "loss": 3.8887, + "step": 1835520 + }, + { + "epoch": 0.0, + "learning_rate": 1.9938720343719025e-05, + "loss": 3.8875, + "step": 1836032 + }, + { + "epoch": 0.0, + "learning_rate": 1.9930334396208505e-05, + "loss": 3.8885, + "step": 1836544 + }, + { + "epoch": 0.0, + "learning_rate": 1.9921948448697985e-05, + "loss": 3.8973, + "step": 1837056 + }, + { + "epoch": 0.0, + "learning_rate": 1.991356250118746e-05, + "loss": 3.8987, + "step": 1837568 + }, + { + "epoch": 0.0, + "learning_rate": 1.9905192932480677e-05, + "loss": 3.8836, + "step": 1838080 + }, + { + "epoch": 0.0, + "learning_rate": 1.9896806984970157e-05, + "loss": 3.8808, + "step": 1838592 + }, + { + "epoch": 0.0, + "learning_rate": 1.9888421037459634e-05, + "loss": 3.8979, + "step": 1839104 + }, + { + "epoch": 0.0, + "learning_rate": 1.9880035089949114e-05, + "loss": 3.879, + "step": 1839616 + }, + { + "epoch": 0.0, + "learning_rate": 1.9871649142438594e-05, + "loss": 3.8913, + "step": 1840128 + }, + { + "epoch": 0.0, + "learning_rate": 1.9863263194928074e-05, + "loss": 3.8868, + "step": 1840640 + }, + { + "epoch": 0.0, + "learning_rate": 1.9854877247417554e-05, + "loss": 3.8927, + "step": 1841152 + }, + { + "epoch": 0.0, + "learning_rate": 1.9846491299907037e-05, + "loss": 3.899, + "step": 1841664 + }, + { + "epoch": 0.0, + "learning_rate": 1.9838105352396517e-05, + "loss": 3.897, + "step": 1842176 + }, + { + "epoch": 0.0, + "learning_rate": 1.9829719404885994e-05, + "loss": 3.8942, + "step": 1842688 + }, + { + "epoch": 0.0, + "learning_rate": 1.9821333457375474e-05, + "loss": 3.8902, + "step": 1843200 + }, + { + "epoch": 0.0, + "learning_rate": 1.9812963888668686e-05, + "loss": 3.8908, + "step": 1843712 + }, + { + "epoch": 0.0, + "learning_rate": 1.9804577941158166e-05, + "loss": 3.8891, + "step": 1844224 + }, + { + "epoch": 0.0, + "learning_rate": 1.9796191993647646e-05, + "loss": 3.8969, + "step": 1844736 + }, + { + "epoch": 0.0, + "learning_rate": 1.9787806046137126e-05, + "loss": 3.8835, + "step": 1845248 + }, + { + "epoch": 0.0, + "learning_rate": 1.9779420098626606e-05, + "loss": 3.882, + "step": 1845760 + }, + { + "epoch": 0.0, + "learning_rate": 1.9771034151116086e-05, + "loss": 3.8927, + "step": 1846272 + }, + { + "epoch": 0.0, + "learning_rate": 1.9762648203605566e-05, + "loss": 3.889, + "step": 1846784 + }, + { + "epoch": 0.01, + "learning_rate": 1.9754262256095046e-05, + "loss": 3.887, + "step": 1847296 + }, + { + "epoch": 0.01, + "learning_rate": 1.9745892687388258e-05, + "loss": 3.8961, + "step": 1847808 + }, + { + "epoch": 0.01, + "learning_rate": 1.9737506739877738e-05, + "loss": 3.8979, + "step": 1848320 + }, + { + "epoch": 0.01, + "learning_rate": 1.9729120792367218e-05, + "loss": 3.8927, + "step": 1848832 + }, + { + "epoch": 0.01, + "learning_rate": 1.9720751223660427e-05, + "loss": 3.886, + "step": 1849344 + }, + { + "epoch": 0.01, + "learning_rate": 1.9712365276149907e-05, + "loss": 3.8972, + "step": 1849856 + }, + { + "epoch": 0.01, + "learning_rate": 1.9703979328639387e-05, + "loss": 3.8943, + "step": 1850368 + }, + { + "epoch": 0.01, + "learning_rate": 1.9695593381128867e-05, + "loss": 3.8848, + "step": 1850880 + }, + { + "epoch": 0.01, + "learning_rate": 1.9687207433618347e-05, + "loss": 3.8806, + "step": 1851392 + }, + { + "epoch": 0.01, + "learning_rate": 1.9678821486107827e-05, + "loss": 3.8831, + "step": 1851904 + }, + { + "epoch": 0.01, + "learning_rate": 1.9670435538597307e-05, + "loss": 3.8863, + "step": 1852416 + }, + { + "epoch": 0.01, + "learning_rate": 1.9662049591086787e-05, + "loss": 3.8981, + "step": 1852928 + }, + { + "epoch": 0.01, + "learning_rate": 1.9653663643576267e-05, + "loss": 3.8807, + "step": 1853440 + }, + { + "epoch": 0.01, + "learning_rate": 1.9645277696065747e-05, + "loss": 3.8853, + "step": 1853952 + }, + { + "epoch": 0.01, + "learning_rate": 1.963690812735896e-05, + "loss": 3.9043, + "step": 1854464 + }, + { + "epoch": 0.01, + "learning_rate": 1.962852217984844e-05, + "loss": 3.8867, + "step": 1854976 + }, + { + "epoch": 0.01, + "learning_rate": 1.962013623233792e-05, + "loss": 3.8999, + "step": 1855488 + }, + { + "epoch": 0.01, + "learning_rate": 1.96117502848274e-05, + "loss": 3.8826, + "step": 1856000 + }, + { + "epoch": 0.01, + "learning_rate": 1.960338071612061e-05, + "loss": 3.8637, + "step": 1856512 + }, + { + "epoch": 0.01, + "learning_rate": 1.9594994768610088e-05, + "loss": 3.9038, + "step": 1857024 + }, + { + "epoch": 0.01, + "learning_rate": 1.95866251999033e-05, + "loss": 3.8897, + "step": 1857536 + }, + { + "epoch": 0.01, + "learning_rate": 1.957823925239278e-05, + "loss": 3.8846, + "step": 1858048 + }, + { + "epoch": 0.01, + "learning_rate": 1.956985330488226e-05, + "loss": 3.887, + "step": 1858560 + }, + { + "epoch": 0.01, + "learning_rate": 1.956146735737174e-05, + "loss": 3.8848, + "step": 1859072 + }, + { + "epoch": 0.01, + "learning_rate": 1.955308140986122e-05, + "loss": 3.8769, + "step": 1859584 + }, + { + "epoch": 0.01, + "learning_rate": 1.95446954623507e-05, + "loss": 3.8803, + "step": 1860096 + }, + { + "epoch": 0.01, + "learning_rate": 1.9536309514840177e-05, + "loss": 3.8822, + "step": 1860608 + }, + { + "epoch": 0.01, + "learning_rate": 1.9527939946133393e-05, + "loss": 3.8835, + "step": 1861120 + }, + { + "epoch": 0.01, + "learning_rate": 1.9519553998622873e-05, + "loss": 3.8959, + "step": 1861632 + }, + { + "epoch": 0.01, + "learning_rate": 1.9511168051112353e-05, + "loss": 3.8857, + "step": 1862144 + }, + { + "epoch": 0.01, + "learning_rate": 1.950278210360183e-05, + "loss": 3.8841, + "step": 1862656 + }, + { + "epoch": 0.01, + "learning_rate": 1.949439615609131e-05, + "loss": 3.8971, + "step": 1863168 + }, + { + "epoch": 0.01, + "learning_rate": 1.948601020858079e-05, + "loss": 3.8708, + "step": 1863680 + }, + { + "epoch": 0.01, + "learning_rate": 1.947762426107027e-05, + "loss": 3.8823, + "step": 1864192 + }, + { + "epoch": 0.01, + "learning_rate": 1.9469254692363482e-05, + "loss": 3.8834, + "step": 1864704 + }, + { + "epoch": 0.01, + "learning_rate": 1.9460868744852962e-05, + "loss": 3.8871, + "step": 1865216 + }, + { + "epoch": 0.01, + "learning_rate": 1.9452482797342442e-05, + "loss": 3.885, + "step": 1865728 + }, + { + "epoch": 0.01, + "learning_rate": 1.944409684983192e-05, + "loss": 3.8866, + "step": 1866240 + }, + { + "epoch": 0.01, + "learning_rate": 1.94357109023214e-05, + "loss": 3.8624, + "step": 1866752 + }, + { + "epoch": 0.01, + "learning_rate": 1.942732495481088e-05, + "loss": 3.8847, + "step": 1867264 + }, + { + "epoch": 0.01, + "learning_rate": 1.9418955386104094e-05, + "loss": 3.8959, + "step": 1867776 + }, + { + "epoch": 0.01, + "learning_rate": 1.9410569438593574e-05, + "loss": 3.8792, + "step": 1868288 + }, + { + "epoch": 0.01, + "learning_rate": 1.9402183491083054e-05, + "loss": 3.9008, + "step": 1868800 + }, + { + "epoch": 0.01, + "learning_rate": 1.9393797543572534e-05, + "loss": 3.8929, + "step": 1869312 + }, + { + "epoch": 0.01, + "learning_rate": 1.938541159606201e-05, + "loss": 3.9002, + "step": 1869824 + }, + { + "epoch": 0.01, + "learning_rate": 1.9377042027355223e-05, + "loss": 3.8753, + "step": 1870336 + }, + { + "epoch": 0.01, + "learning_rate": 1.9368656079844706e-05, + "loss": 3.8922, + "step": 1870848 + }, + { + "epoch": 0.01, + "learning_rate": 1.9360270132334183e-05, + "loss": 3.8915, + "step": 1871360 + }, + { + "epoch": 0.01, + "learning_rate": 1.9351884184823663e-05, + "loss": 3.8776, + "step": 1871872 + }, + { + "epoch": 0.01, + "learning_rate": 1.9343498237313143e-05, + "loss": 3.8982, + "step": 1872384 + }, + { + "epoch": 0.01, + "learning_rate": 1.9335112289802623e-05, + "loss": 3.8863, + "step": 1872896 + }, + { + "epoch": 0.01, + "learning_rate": 1.9326726342292103e-05, + "loss": 3.8972, + "step": 1873408 + }, + { + "epoch": 0.01, + "learning_rate": 1.9318356773585315e-05, + "loss": 3.8941, + "step": 1873920 + }, + { + "epoch": 0.01, + "learning_rate": 1.9309970826074795e-05, + "loss": 3.8706, + "step": 1874432 + }, + { + "epoch": 0.01, + "learning_rate": 1.9301584878564275e-05, + "loss": 3.8766, + "step": 1874944 + }, + { + "epoch": 0.01, + "learning_rate": 1.9293198931053755e-05, + "loss": 3.881, + "step": 1875456 + }, + { + "epoch": 0.01, + "learning_rate": 1.9284812983543235e-05, + "loss": 3.8906, + "step": 1875968 + }, + { + "epoch": 0.01, + "learning_rate": 1.927642703603271e-05, + "loss": 3.8847, + "step": 1876480 + }, + { + "epoch": 0.01, + "learning_rate": 1.926804108852219e-05, + "loss": 3.8823, + "step": 1876992 + }, + { + "epoch": 0.02, + "learning_rate": 1.9259655141011675e-05, + "loss": 3.8826, + "step": 1877504 + }, + { + "epoch": 0.02, + "learning_rate": 1.9251269193501155e-05, + "loss": 3.8853, + "step": 1878016 + }, + { + "epoch": 0.02, + "learning_rate": 1.9242899624794364e-05, + "loss": 3.8839, + "step": 1878528 + }, + { + "epoch": 0.02, + "learning_rate": 1.9234530056087576e-05, + "loss": 3.8793, + "step": 1879040 + }, + { + "epoch": 0.02, + "learning_rate": 1.9226144108577056e-05, + "loss": 3.9021, + "step": 1879552 + }, + { + "epoch": 0.02, + "learning_rate": 1.9217758161066536e-05, + "loss": 3.8975, + "step": 1880064 + }, + { + "epoch": 0.02, + "learning_rate": 1.9209372213556016e-05, + "loss": 3.8903, + "step": 1880576 + }, + { + "epoch": 0.02, + "learning_rate": 1.9200986266045496e-05, + "loss": 3.879, + "step": 1881088 + }, + { + "epoch": 0.02, + "learning_rate": 1.9192600318534976e-05, + "loss": 3.8878, + "step": 1881600 + }, + { + "epoch": 0.02, + "learning_rate": 1.9184214371024456e-05, + "loss": 3.8826, + "step": 1882112 + }, + { + "epoch": 0.02, + "learning_rate": 1.9175828423513936e-05, + "loss": 3.9005, + "step": 1882624 + }, + { + "epoch": 0.02, + "learning_rate": 1.9167442476003416e-05, + "loss": 3.8838, + "step": 1883136 + }, + { + "epoch": 0.02, + "learning_rate": 1.9159072907296625e-05, + "loss": 3.8843, + "step": 1883648 + }, + { + "epoch": 0.02, + "learning_rate": 1.915068695978611e-05, + "loss": 3.8912, + "step": 1884160 + }, + { + "epoch": 0.02, + "learning_rate": 1.914230101227559e-05, + "loss": 3.8906, + "step": 1884672 + }, + { + "epoch": 0.02, + "learning_rate": 1.913391506476507e-05, + "loss": 3.8757, + "step": 1885184 + }, + { + "epoch": 0.02, + "learning_rate": 1.9125545496058278e-05, + "loss": 3.8841, + "step": 1885696 + }, + { + "epoch": 0.02, + "learning_rate": 1.911717592735149e-05, + "loss": 3.8816, + "step": 1886208 + }, + { + "epoch": 0.02, + "learning_rate": 1.910878997984097e-05, + "loss": 3.8771, + "step": 1886720 + }, + { + "epoch": 0.02, + "learning_rate": 1.910040403233045e-05, + "loss": 3.8768, + "step": 1887232 + }, + { + "epoch": 0.02, + "learning_rate": 1.909201808481993e-05, + "loss": 3.8837, + "step": 1887744 + }, + { + "epoch": 0.02, + "learning_rate": 1.908363213730941e-05, + "loss": 3.8852, + "step": 1888256 + }, + { + "epoch": 0.02, + "learning_rate": 1.907524618979889e-05, + "loss": 3.8841, + "step": 1888768 + }, + { + "epoch": 0.02, + "learning_rate": 1.9066860242288366e-05, + "loss": 3.8838, + "step": 1889280 + }, + { + "epoch": 0.02, + "learning_rate": 1.9058474294777846e-05, + "loss": 3.8796, + "step": 1889792 + }, + { + "epoch": 0.02, + "learning_rate": 1.9050104726071062e-05, + "loss": 3.8917, + "step": 1890304 + }, + { + "epoch": 0.02, + "learning_rate": 1.9041718778560542e-05, + "loss": 3.9007, + "step": 1890816 + }, + { + "epoch": 0.02, + "learning_rate": 1.903334920985375e-05, + "loss": 3.8914, + "step": 1891328 + }, + { + "epoch": 0.02, + "learning_rate": 1.902496326234323e-05, + "loss": 3.8891, + "step": 1891840 + }, + { + "epoch": 0.02, + "learning_rate": 1.901657731483271e-05, + "loss": 3.8747, + "step": 1892352 + }, + { + "epoch": 0.02, + "learning_rate": 1.900819136732219e-05, + "loss": 3.8881, + "step": 1892864 + }, + { + "epoch": 0.02, + "learning_rate": 1.899980541981167e-05, + "loss": 3.8816, + "step": 1893376 + }, + { + "epoch": 0.02, + "learning_rate": 1.8991435851104884e-05, + "loss": 3.8851, + "step": 1893888 + }, + { + "epoch": 0.02, + "learning_rate": 1.8983049903594363e-05, + "loss": 3.8789, + "step": 1894400 + }, + { + "epoch": 0.02, + "learning_rate": 1.897466395608384e-05, + "loss": 3.8833, + "step": 1894912 + }, + { + "epoch": 0.02, + "learning_rate": 1.896627800857332e-05, + "loss": 3.8882, + "step": 1895424 + }, + { + "epoch": 0.02, + "learning_rate": 1.89578920610628e-05, + "loss": 3.8844, + "step": 1895936 + }, + { + "epoch": 0.02, + "learning_rate": 1.8949522492356016e-05, + "loss": 3.8924, + "step": 1896448 + }, + { + "epoch": 0.02, + "learning_rate": 1.8941136544845492e-05, + "loss": 3.8871, + "step": 1896960 + }, + { + "epoch": 0.02, + "learning_rate": 1.8932750597334972e-05, + "loss": 3.8781, + "step": 1897472 + }, + { + "epoch": 0.02, + "learning_rate": 1.8924364649824452e-05, + "loss": 3.8928, + "step": 1897984 + }, + { + "epoch": 0.02, + "learning_rate": 1.8915978702313932e-05, + "loss": 3.8923, + "step": 1898496 + }, + { + "epoch": 0.02, + "learning_rate": 1.8907592754803412e-05, + "loss": 3.9006, + "step": 1899008 + }, + { + "epoch": 0.02, + "learning_rate": 1.8899206807292892e-05, + "loss": 3.8816, + "step": 1899520 + }, + { + "epoch": 0.02, + "learning_rate": 1.8890820859782372e-05, + "loss": 3.8959, + "step": 1900032 + }, + { + "epoch": 0.02, + "learning_rate": 1.8882451291075585e-05, + "loss": 3.8876, + "step": 1900544 + }, + { + "epoch": 0.02, + "learning_rate": 1.8874065343565065e-05, + "loss": 3.8931, + "step": 1901056 + }, + { + "epoch": 0.02, + "learning_rate": 1.8865695774858274e-05, + "loss": 3.8904, + "step": 1901568 + }, + { + "epoch": 0.02, + "learning_rate": 1.8857309827347754e-05, + "loss": 3.8875, + "step": 1902080 + }, + { + "epoch": 0.02, + "learning_rate": 1.8848923879837234e-05, + "loss": 3.8818, + "step": 1902592 + }, + { + "epoch": 0.02, + "learning_rate": 1.8840537932326717e-05, + "loss": 3.8972, + "step": 1903104 + }, + { + "epoch": 0.02, + "learning_rate": 1.8832151984816197e-05, + "loss": 3.878, + "step": 1903616 + }, + { + "epoch": 0.02, + "learning_rate": 1.8823766037305673e-05, + "loss": 3.8851, + "step": 1904128 + }, + { + "epoch": 0.02, + "learning_rate": 1.8815380089795153e-05, + "loss": 3.8896, + "step": 1904640 + }, + { + "epoch": 0.02, + "learning_rate": 1.8807010521088366e-05, + "loss": 3.8942, + "step": 1905152 + }, + { + "epoch": 0.02, + "learning_rate": 1.8798624573577846e-05, + "loss": 3.8794, + "step": 1905664 + }, + { + "epoch": 0.02, + "learning_rate": 1.8790238626067326e-05, + "loss": 3.8872, + "step": 1906176 + }, + { + "epoch": 0.02, + "learning_rate": 1.8781852678556806e-05, + "loss": 3.8936, + "step": 1906688 + }, + { + "epoch": 0.02, + "learning_rate": 1.8773466731046286e-05, + "loss": 3.8843, + "step": 1907200 + }, + { + "epoch": 0.02, + "learning_rate": 1.8765080783535766e-05, + "loss": 3.8944, + "step": 1907712 + }, + { + "epoch": 0.03, + "eval_loss": 3.985036611557007, + "eval_runtime": 294.0646, + "eval_samples_per_second": 1297.643, + "eval_steps_per_second": 40.552, + "step": 1907987 + }, + { + "epoch": 1.0, + "learning_rate": 1.8756694836025246e-05, + "loss": 3.8776, + "step": 1908224 + }, + { + "epoch": 1.0, + "learning_rate": 1.8748325267318455e-05, + "loss": 3.8841, + "step": 1908736 + }, + { + "epoch": 1.0, + "learning_rate": 1.8739939319807938e-05, + "loss": 3.8861, + "step": 1909248 + }, + { + "epoch": 1.0, + "learning_rate": 1.8731553372297418e-05, + "loss": 3.8897, + "step": 1909760 + }, + { + "epoch": 1.0, + "learning_rate": 1.8723167424786898e-05, + "loss": 3.8973, + "step": 1910272 + }, + { + "epoch": 1.0, + "learning_rate": 1.8714781477276378e-05, + "loss": 3.8811, + "step": 1910784 + }, + { + "epoch": 1.0, + "learning_rate": 1.8706395529765855e-05, + "loss": 3.8885, + "step": 1911296 + }, + { + "epoch": 1.0, + "learning_rate": 1.8698009582255334e-05, + "loss": 3.8829, + "step": 1911808 + }, + { + "epoch": 1.0, + "learning_rate": 1.8689623634744814e-05, + "loss": 3.879, + "step": 1912320 + }, + { + "epoch": 1.0, + "learning_rate": 1.8681254066038027e-05, + "loss": 3.8831, + "step": 1912832 + }, + { + "epoch": 1.0, + "learning_rate": 1.8672868118527507e-05, + "loss": 3.8921, + "step": 1913344 + }, + { + "epoch": 1.0, + "learning_rate": 1.8664482171016987e-05, + "loss": 3.8914, + "step": 1913856 + }, + { + "epoch": 1.0, + "learning_rate": 1.86561126023102e-05, + "loss": 3.8855, + "step": 1914368 + }, + { + "epoch": 1.0, + "learning_rate": 1.8647726654799676e-05, + "loss": 3.8732, + "step": 1914880 + }, + { + "epoch": 1.0, + "learning_rate": 1.8639340707289156e-05, + "loss": 3.8879, + "step": 1915392 + }, + { + "epoch": 1.0, + "learning_rate": 1.863095475977864e-05, + "loss": 3.875, + "step": 1915904 + }, + { + "epoch": 1.0, + "learning_rate": 1.862256881226812e-05, + "loss": 3.8841, + "step": 1916416 + }, + { + "epoch": 1.0, + "learning_rate": 1.86141828647576e-05, + "loss": 3.8833, + "step": 1916928 + }, + { + "epoch": 1.0, + "learning_rate": 1.860579691724708e-05, + "loss": 3.8835, + "step": 1917440 + }, + { + "epoch": 1.0, + "learning_rate": 1.859741096973656e-05, + "loss": 3.8917, + "step": 1917952 + }, + { + "epoch": 1.0, + "learning_rate": 1.8589025022226036e-05, + "loss": 3.8946, + "step": 1918464 + }, + { + "epoch": 1.0, + "learning_rate": 1.8580639074715516e-05, + "loss": 3.8912, + "step": 1918976 + }, + { + "epoch": 1.0, + "learning_rate": 1.8572253127204995e-05, + "loss": 3.8848, + "step": 1919488 + }, + { + "epoch": 1.0, + "learning_rate": 1.856386717969448e-05, + "loss": 3.8819, + "step": 1920000 + }, + { + "epoch": 1.0, + "learning_rate": 1.8555497610987688e-05, + "loss": 3.8884, + "step": 1920512 + }, + { + "epoch": 1.0, + "learning_rate": 1.8547111663477168e-05, + "loss": 3.8843, + "step": 1921024 + }, + { + "epoch": 1.0, + "learning_rate": 1.8538725715966648e-05, + "loss": 3.8823, + "step": 1921536 + }, + { + "epoch": 1.0, + "learning_rate": 1.8530356147259857e-05, + "loss": 3.8811, + "step": 1922048 + }, + { + "epoch": 1.0, + "learning_rate": 1.852197019974934e-05, + "loss": 3.8822, + "step": 1922560 + }, + { + "epoch": 1.0, + "learning_rate": 1.851358425223882e-05, + "loss": 3.8865, + "step": 1923072 + }, + { + "epoch": 1.01, + "learning_rate": 1.85051983047283e-05, + "loss": 3.8769, + "step": 1923584 + }, + { + "epoch": 1.01, + "learning_rate": 1.849682873602151e-05, + "loss": 3.8928, + "step": 1924096 + }, + { + "epoch": 1.01, + "learning_rate": 1.848844278851099e-05, + "loss": 3.8908, + "step": 1924608 + }, + { + "epoch": 1.01, + "learning_rate": 1.848005684100047e-05, + "loss": 3.8895, + "step": 1925120 + }, + { + "epoch": 1.01, + "learning_rate": 1.847167089348995e-05, + "loss": 3.8811, + "step": 1925632 + }, + { + "epoch": 1.01, + "learning_rate": 1.846330132478316e-05, + "loss": 3.8912, + "step": 1926144 + }, + { + "epoch": 1.01, + "learning_rate": 1.845491537727264e-05, + "loss": 3.8877, + "step": 1926656 + }, + { + "epoch": 1.01, + "learning_rate": 1.844652942976212e-05, + "loss": 3.881, + "step": 1927168 + }, + { + "epoch": 1.01, + "learning_rate": 1.84381434822516e-05, + "loss": 3.8762, + "step": 1927680 + }, + { + "epoch": 1.01, + "learning_rate": 1.842975753474108e-05, + "loss": 3.8781, + "step": 1928192 + }, + { + "epoch": 1.01, + "learning_rate": 1.842137158723056e-05, + "loss": 3.8804, + "step": 1928704 + }, + { + "epoch": 1.01, + "learning_rate": 1.841298563972004e-05, + "loss": 3.8894, + "step": 1929216 + }, + { + "epoch": 1.01, + "learning_rate": 1.840459969220952e-05, + "loss": 3.8769, + "step": 1929728 + }, + { + "epoch": 1.01, + "learning_rate": 1.8396213744699e-05, + "loss": 3.8811, + "step": 1930240 + }, + { + "epoch": 1.01, + "learning_rate": 1.838782779718848e-05, + "loss": 3.8956, + "step": 1930752 + }, + { + "epoch": 1.01, + "learning_rate": 1.837944184967796e-05, + "loss": 3.8833, + "step": 1931264 + }, + { + "epoch": 1.01, + "learning_rate": 1.837105590216744e-05, + "loss": 3.8941, + "step": 1931776 + }, + { + "epoch": 1.01, + "learning_rate": 1.836268633346065e-05, + "loss": 3.8767, + "step": 1932288 + }, + { + "epoch": 1.01, + "learning_rate": 1.8354316764753863e-05, + "loss": 3.86, + "step": 1932800 + }, + { + "epoch": 1.01, + "learning_rate": 1.8345930817243343e-05, + "loss": 3.8995, + "step": 1933312 + }, + { + "epoch": 1.01, + "learning_rate": 1.8337544869732823e-05, + "loss": 3.8811, + "step": 1933824 + }, + { + "epoch": 1.01, + "learning_rate": 1.8329158922222303e-05, + "loss": 3.8842, + "step": 1934336 + }, + { + "epoch": 1.01, + "learning_rate": 1.8320772974711783e-05, + "loss": 3.8792, + "step": 1934848 + }, + { + "epoch": 1.01, + "learning_rate": 1.8312387027201262e-05, + "loss": 3.883, + "step": 1935360 + }, + { + "epoch": 1.01, + "learning_rate": 1.8304001079690742e-05, + "loss": 3.8716, + "step": 1935872 + }, + { + "epoch": 1.01, + "learning_rate": 1.8295631510983955e-05, + "loss": 3.8734, + "step": 1936384 + }, + { + "epoch": 1.01, + "learning_rate": 1.8287245563473435e-05, + "loss": 3.8748, + "step": 1936896 + }, + { + "epoch": 1.01, + "learning_rate": 1.8278859615962915e-05, + "loss": 3.8759, + "step": 1937408 + }, + { + "epoch": 1.01, + "learning_rate": 1.8270473668452395e-05, + "loss": 3.8907, + "step": 1937920 + }, + { + "epoch": 1.01, + "learning_rate": 1.826208772094187e-05, + "loss": 3.8799, + "step": 1938432 + }, + { + "epoch": 1.01, + "learning_rate": 1.8253701773431355e-05, + "loss": 3.8761, + "step": 1938944 + }, + { + "epoch": 1.01, + "learning_rate": 1.8245315825920835e-05, + "loss": 3.897, + "step": 1939456 + }, + { + "epoch": 1.01, + "learning_rate": 1.8236946257214044e-05, + "loss": 3.8626, + "step": 1939968 + }, + { + "epoch": 1.01, + "learning_rate": 1.8228560309703524e-05, + "loss": 3.8725, + "step": 1940480 + }, + { + "epoch": 1.01, + "learning_rate": 1.8220174362193004e-05, + "loss": 3.8826, + "step": 1940992 + }, + { + "epoch": 1.01, + "learning_rate": 1.8211788414682484e-05, + "loss": 3.8825, + "step": 1941504 + }, + { + "epoch": 1.01, + "learning_rate": 1.8203402467171964e-05, + "loss": 3.8786, + "step": 1942016 + }, + { + "epoch": 1.01, + "learning_rate": 1.8195016519661447e-05, + "loss": 3.8822, + "step": 1942528 + }, + { + "epoch": 1.01, + "learning_rate": 1.8186630572150927e-05, + "loss": 3.8604, + "step": 1943040 + }, + { + "epoch": 1.01, + "learning_rate": 1.8178244624640403e-05, + "loss": 3.8742, + "step": 1943552 + }, + { + "epoch": 1.01, + "learning_rate": 1.8169875055933616e-05, + "loss": 3.8952, + "step": 1944064 + }, + { + "epoch": 1.01, + "learning_rate": 1.8161489108423096e-05, + "loss": 3.8742, + "step": 1944576 + }, + { + "epoch": 1.01, + "learning_rate": 1.815311953971631e-05, + "loss": 3.8927, + "step": 1945088 + }, + { + "epoch": 1.01, + "learning_rate": 1.814473359220579e-05, + "loss": 3.8886, + "step": 1945600 + }, + { + "epoch": 1.01, + "learning_rate": 1.8136347644695268e-05, + "loss": 3.8961, + "step": 1946112 + }, + { + "epoch": 1.01, + "learning_rate": 1.8127961697184748e-05, + "loss": 3.8717, + "step": 1946624 + }, + { + "epoch": 1.01, + "learning_rate": 1.8119575749674225e-05, + "loss": 3.8862, + "step": 1947136 + }, + { + "epoch": 1.01, + "learning_rate": 1.8111206180967437e-05, + "loss": 3.8857, + "step": 1947648 + }, + { + "epoch": 1.01, + "learning_rate": 1.810283661226065e-05, + "loss": 3.8711, + "step": 1948160 + }, + { + "epoch": 1.01, + "learning_rate": 1.809445066475013e-05, + "loss": 3.8923, + "step": 1948672 + }, + { + "epoch": 1.01, + "learning_rate": 1.808606471723961e-05, + "loss": 3.884, + "step": 1949184 + }, + { + "epoch": 1.01, + "learning_rate": 1.807767876972909e-05, + "loss": 3.8876, + "step": 1949696 + }, + { + "epoch": 1.01, + "learning_rate": 1.806929282221857e-05, + "loss": 3.8918, + "step": 1950208 + }, + { + "epoch": 1.01, + "learning_rate": 1.8060906874708046e-05, + "loss": 3.8683, + "step": 1950720 + }, + { + "epoch": 1.01, + "learning_rate": 1.8052520927197526e-05, + "loss": 3.8715, + "step": 1951232 + }, + { + "epoch": 1.01, + "learning_rate": 1.804413497968701e-05, + "loss": 3.8748, + "step": 1951744 + }, + { + "epoch": 1.01, + "learning_rate": 1.803574903217649e-05, + "loss": 3.8831, + "step": 1952256 + }, + { + "epoch": 1.01, + "learning_rate": 1.802736308466597e-05, + "loss": 3.8828, + "step": 1952768 + }, + { + "epoch": 1.01, + "learning_rate": 1.801899351595918e-05, + "loss": 3.8762, + "step": 1953280 + }, + { + "epoch": 1.02, + "learning_rate": 1.801060756844866e-05, + "loss": 3.8768, + "step": 1953792 + }, + { + "epoch": 1.02, + "learning_rate": 1.800222162093814e-05, + "loss": 3.8851, + "step": 1954304 + }, + { + "epoch": 1.02, + "learning_rate": 1.799383567342762e-05, + "loss": 3.8736, + "step": 1954816 + }, + { + "epoch": 1.02, + "learning_rate": 1.798546610472083e-05, + "loss": 3.876, + "step": 1955328 + }, + { + "epoch": 1.02, + "learning_rate": 1.797708015721031e-05, + "loss": 3.8988, + "step": 1955840 + }, + { + "epoch": 1.02, + "learning_rate": 1.796869420969979e-05, + "loss": 3.8905, + "step": 1956352 + }, + { + "epoch": 1.02, + "learning_rate": 1.796030826218927e-05, + "loss": 3.8872, + "step": 1956864 + }, + { + "epoch": 1.02, + "learning_rate": 1.795192231467875e-05, + "loss": 3.8726, + "step": 1957376 + }, + { + "epoch": 1.02, + "learning_rate": 1.7943536367168227e-05, + "loss": 3.8821, + "step": 1957888 + }, + { + "epoch": 1.02, + "learning_rate": 1.793515041965771e-05, + "loss": 3.8776, + "step": 1958400 + }, + { + "epoch": 1.02, + "learning_rate": 1.792676447214719e-05, + "loss": 3.8952, + "step": 1958912 + }, + { + "epoch": 1.02, + "learning_rate": 1.7918427661047865e-05, + "loss": 3.8819, + "step": 1959424 + }, + { + "epoch": 1.02, + "learning_rate": 1.7910041713537345e-05, + "loss": 3.8766, + "step": 1959936 + }, + { + "epoch": 1.02, + "learning_rate": 1.7901655766026825e-05, + "loss": 3.883, + "step": 1960448 + }, + { + "epoch": 1.02, + "learning_rate": 1.7893269818516305e-05, + "loss": 3.8897, + "step": 1960960 + }, + { + "epoch": 1.02, + "learning_rate": 1.7884883871005784e-05, + "loss": 3.8697, + "step": 1961472 + }, + { + "epoch": 1.02, + "learning_rate": 1.7876497923495264e-05, + "loss": 3.8831, + "step": 1961984 + }, + { + "epoch": 1.02, + "learning_rate": 1.7868111975984744e-05, + "loss": 3.8703, + "step": 1962496 + }, + { + "epoch": 1.02, + "learning_rate": 1.7859726028474224e-05, + "loss": 3.8721, + "step": 1963008 + }, + { + "epoch": 1.02, + "learning_rate": 1.78513400809637e-05, + "loss": 3.8735, + "step": 1963520 + }, + { + "epoch": 1.02, + "learning_rate": 1.784295413345318e-05, + "loss": 3.8778, + "step": 1964032 + }, + { + "epoch": 1.02, + "learning_rate": 1.7834568185942664e-05, + "loss": 3.8794, + "step": 1964544 + }, + { + "epoch": 1.02, + "learning_rate": 1.7826182238432144e-05, + "loss": 3.8846, + "step": 1965056 + }, + { + "epoch": 1.02, + "learning_rate": 1.7817812669725353e-05, + "loss": 3.8739, + "step": 1965568 + }, + { + "epoch": 1.02, + "learning_rate": 1.7809443101018566e-05, + "loss": 3.8755, + "step": 1966080 + }, + { + "epoch": 1.02, + "learning_rate": 1.7801057153508046e-05, + "loss": 3.8872, + "step": 1966592 + }, + { + "epoch": 1.02, + "learning_rate": 1.7792671205997526e-05, + "loss": 3.8956, + "step": 1967104 + }, + { + "epoch": 1.02, + "learning_rate": 1.7784285258487006e-05, + "loss": 3.8842, + "step": 1967616 + }, + { + "epoch": 1.02, + "learning_rate": 1.7775899310976486e-05, + "loss": 3.8873, + "step": 1968128 + }, + { + "epoch": 1.02, + "learning_rate": 1.7767529742269698e-05, + "loss": 3.8721, + "step": 1968640 + }, + { + "epoch": 1.02, + "learning_rate": 1.7759143794759175e-05, + "loss": 3.8809, + "step": 1969152 + }, + { + "epoch": 1.02, + "learning_rate": 1.7750757847248655e-05, + "loss": 3.8803, + "step": 1969664 + }, + { + "epoch": 1.02, + "learning_rate": 1.7742371899738135e-05, + "loss": 3.8739, + "step": 1970176 + }, + { + "epoch": 1.02, + "learning_rate": 1.7733985952227618e-05, + "loss": 3.8773, + "step": 1970688 + }, + { + "epoch": 1.02, + "learning_rate": 1.7725600004717098e-05, + "loss": 3.8795, + "step": 1971200 + }, + { + "epoch": 1.02, + "learning_rate": 1.7717214057206578e-05, + "loss": 3.8824, + "step": 1971712 + }, + { + "epoch": 1.02, + "learning_rate": 1.7708828109696058e-05, + "loss": 3.8787, + "step": 1972224 + }, + { + "epoch": 1.02, + "learning_rate": 1.7700458540989267e-05, + "loss": 3.8845, + "step": 1972736 + }, + { + "epoch": 1.02, + "learning_rate": 1.7692072593478747e-05, + "loss": 3.8823, + "step": 1973248 + }, + { + "epoch": 1.02, + "learning_rate": 1.7683686645968227e-05, + "loss": 3.8743, + "step": 1973760 + }, + { + "epoch": 1.02, + "learning_rate": 1.767530069845771e-05, + "loss": 3.8836, + "step": 1974272 + }, + { + "epoch": 1.02, + "learning_rate": 1.766693112975092e-05, + "loss": 3.8912, + "step": 1974784 + }, + { + "epoch": 1.02, + "learning_rate": 1.765856156104413e-05, + "loss": 3.889, + "step": 1975296 + }, + { + "epoch": 1.02, + "learning_rate": 1.7650175613533608e-05, + "loss": 3.8781, + "step": 1975808 + }, + { + "epoch": 1.02, + "learning_rate": 1.7641789666023088e-05, + "loss": 3.8928, + "step": 1976320 + }, + { + "epoch": 1.02, + "learning_rate": 1.763340371851257e-05, + "loss": 3.8865, + "step": 1976832 + }, + { + "epoch": 1.02, + "learning_rate": 1.762501777100205e-05, + "loss": 3.8881, + "step": 1977344 + }, + { + "epoch": 1.02, + "learning_rate": 1.761663182349153e-05, + "loss": 3.8841, + "step": 1977856 + }, + { + "epoch": 1.02, + "learning_rate": 1.7608245875981008e-05, + "loss": 3.88, + "step": 1978368 + }, + { + "epoch": 1.02, + "learning_rate": 1.7599859928470488e-05, + "loss": 3.8792, + "step": 1978880 + }, + { + "epoch": 1.02, + "learning_rate": 1.7591473980959968e-05, + "loss": 3.8899, + "step": 1979392 + }, + { + "epoch": 1.02, + "learning_rate": 1.758310441225318e-05, + "loss": 3.8732, + "step": 1979904 + }, + { + "epoch": 1.02, + "learning_rate": 1.757471846474266e-05, + "loss": 3.8814, + "step": 1980416 + }, + { + "epoch": 1.02, + "learning_rate": 1.756633251723214e-05, + "loss": 3.8857, + "step": 1980928 + }, + { + "epoch": 1.02, + "learning_rate": 1.755794656972162e-05, + "loss": 3.8871, + "step": 1981440 + }, + { + "epoch": 1.02, + "learning_rate": 1.754957700101483e-05, + "loss": 3.8765, + "step": 1981952 + }, + { + "epoch": 1.02, + "learning_rate": 1.7541207432308042e-05, + "loss": 3.8796, + "step": 1982464 + }, + { + "epoch": 1.02, + "learning_rate": 1.7532821484797525e-05, + "loss": 3.8872, + "step": 1982976 + }, + { + "epoch": 1.02, + "learning_rate": 1.7524435537287005e-05, + "loss": 3.8822, + "step": 1983488 + }, + { + "epoch": 1.02, + "learning_rate": 1.7516049589776482e-05, + "loss": 3.8882, + "step": 1984000 + }, + { + "epoch": 1.03, + "eval_loss": 3.9835500717163086, + "eval_runtime": 295.0853, + "eval_samples_per_second": 1293.155, + "eval_steps_per_second": 40.412, + "step": 1984307 + }, + { + "epoch": 0.0, + "learning_rate": 1.7507663642265962e-05, + "loss": 3.863, + "step": 1984512 + }, + { + "epoch": 0.0, + "learning_rate": 1.749927769475544e-05, + "loss": 3.8764, + "step": 1985024 + }, + { + "epoch": 0.0, + "learning_rate": 1.749089174724492e-05, + "loss": 3.8845, + "step": 1985536 + }, + { + "epoch": 0.0, + "learning_rate": 1.74825057997344e-05, + "loss": 3.882, + "step": 1986048 + }, + { + "epoch": 0.0, + "learning_rate": 1.747411985222388e-05, + "loss": 3.8938, + "step": 1986560 + }, + { + "epoch": 0.0, + "learning_rate": 1.746573390471336e-05, + "loss": 3.8758, + "step": 1987072 + }, + { + "epoch": 0.0, + "learning_rate": 1.745734795720284e-05, + "loss": 3.8844, + "step": 1987584 + }, + { + "epoch": 0.0, + "learning_rate": 1.744896200969232e-05, + "loss": 3.8791, + "step": 1988096 + }, + { + "epoch": 0.0, + "learning_rate": 1.74405760621818e-05, + "loss": 3.8737, + "step": 1988608 + }, + { + "epoch": 0.0, + "learning_rate": 1.743219011467128e-05, + "loss": 3.8788, + "step": 1989120 + }, + { + "epoch": 0.0, + "learning_rate": 1.742380416716076e-05, + "loss": 3.8867, + "step": 1989632 + }, + { + "epoch": 0.0, + "learning_rate": 1.741541821965024e-05, + "loss": 3.8878, + "step": 1990144 + }, + { + "epoch": 0.0, + "learning_rate": 1.7407048650943454e-05, + "loss": 3.8864, + "step": 1990656 + }, + { + "epoch": 0.0, + "learning_rate": 1.7398662703432934e-05, + "loss": 3.8618, + "step": 1991168 + }, + { + "epoch": 0.0, + "learning_rate": 1.7390276755922414e-05, + "loss": 3.8831, + "step": 1991680 + }, + { + "epoch": 0.0, + "learning_rate": 1.7381890808411894e-05, + "loss": 3.8677, + "step": 1992192 + }, + { + "epoch": 0.0, + "learning_rate": 1.737350486090137e-05, + "loss": 3.8788, + "step": 1992704 + }, + { + "epoch": 0.0, + "learning_rate": 1.736511891339085e-05, + "loss": 3.8759, + "step": 1993216 + }, + { + "epoch": 0.0, + "learning_rate": 1.7356732965880333e-05, + "loss": 3.8789, + "step": 1993728 + }, + { + "epoch": 0.0, + "learning_rate": 1.7348363397173543e-05, + "loss": 3.8846, + "step": 1994240 + }, + { + "epoch": 0.0, + "learning_rate": 1.7339977449663022e-05, + "loss": 3.8939, + "step": 1994752 + }, + { + "epoch": 0.0, + "learning_rate": 1.7331591502152502e-05, + "loss": 3.8854, + "step": 1995264 + }, + { + "epoch": 0.0, + "learning_rate": 1.7323205554641982e-05, + "loss": 3.8799, + "step": 1995776 + }, + { + "epoch": 0.0, + "learning_rate": 1.7314819607131462e-05, + "loss": 3.8752, + "step": 1996288 + }, + { + "epoch": 0.0, + "learning_rate": 1.7306433659620942e-05, + "loss": 3.8849, + "step": 1996800 + }, + { + "epoch": 0.0, + "learning_rate": 1.7298047712110426e-05, + "loss": 3.8828, + "step": 1997312 + }, + { + "epoch": 0.0, + "learning_rate": 1.7289661764599902e-05, + "loss": 3.8766, + "step": 1997824 + }, + { + "epoch": 0.0, + "learning_rate": 1.7281275817089382e-05, + "loss": 3.8739, + "step": 1998336 + }, + { + "epoch": 0.0, + "learning_rate": 1.7272889869578862e-05, + "loss": 3.8784, + "step": 1998848 + }, + { + "epoch": 0.0, + "learning_rate": 1.7264503922068342e-05, + "loss": 3.8787, + "step": 1999360 + }, + { + "epoch": 0.01, + "learning_rate": 1.725613435336155e-05, + "loss": 3.8693, + "step": 1999872 + }, + { + "epoch": 0.01, + "learning_rate": 1.7247748405851035e-05, + "loss": 3.891, + "step": 2000384 + }, + { + "epoch": 0.01, + "learning_rate": 1.7239378837144247e-05, + "loss": 3.8879, + "step": 2000896 + }, + { + "epoch": 0.01, + "learning_rate": 1.7230992889633724e-05, + "loss": 3.8823, + "step": 2001408 + }, + { + "epoch": 0.01, + "learning_rate": 1.7222606942123204e-05, + "loss": 3.8796, + "step": 2001920 + }, + { + "epoch": 0.01, + "learning_rate": 1.7214237373416416e-05, + "loss": 3.8788, + "step": 2002432 + }, + { + "epoch": 0.01, + "learning_rate": 1.7205851425905896e-05, + "loss": 3.8859, + "step": 2002944 + }, + { + "epoch": 0.01, + "learning_rate": 1.7197465478395376e-05, + "loss": 3.8768, + "step": 2003456 + }, + { + "epoch": 0.01, + "learning_rate": 1.7189079530884856e-05, + "loss": 3.873, + "step": 2003968 + }, + { + "epoch": 0.01, + "learning_rate": 1.7180693583374336e-05, + "loss": 3.8735, + "step": 2004480 + }, + { + "epoch": 0.01, + "learning_rate": 1.7172307635863816e-05, + "loss": 3.8781, + "step": 2004992 + }, + { + "epoch": 0.01, + "learning_rate": 1.7163921688353296e-05, + "loss": 3.8828, + "step": 2005504 + }, + { + "epoch": 0.01, + "learning_rate": 1.7155535740842776e-05, + "loss": 3.8723, + "step": 2006016 + }, + { + "epoch": 0.01, + "learning_rate": 1.7147149793332256e-05, + "loss": 3.8779, + "step": 2006528 + }, + { + "epoch": 0.01, + "learning_rate": 1.7138763845821736e-05, + "loss": 3.8878, + "step": 2007040 + }, + { + "epoch": 0.01, + "learning_rate": 1.7130377898311216e-05, + "loss": 3.8783, + "step": 2007552 + }, + { + "epoch": 0.01, + "learning_rate": 1.7121991950800696e-05, + "loss": 3.885, + "step": 2008064 + }, + { + "epoch": 0.01, + "learning_rate": 1.7113606003290175e-05, + "loss": 3.8784, + "step": 2008576 + }, + { + "epoch": 0.01, + "learning_rate": 1.7105220055779655e-05, + "loss": 3.8575, + "step": 2009088 + }, + { + "epoch": 0.01, + "learning_rate": 1.7096834108269135e-05, + "loss": 3.8901, + "step": 2009600 + }, + { + "epoch": 0.01, + "learning_rate": 1.7088464539562344e-05, + "loss": 3.8778, + "step": 2010112 + }, + { + "epoch": 0.01, + "learning_rate": 1.7080078592051828e-05, + "loss": 3.8785, + "step": 2010624 + }, + { + "epoch": 0.01, + "learning_rate": 1.7071692644541308e-05, + "loss": 3.8777, + "step": 2011136 + }, + { + "epoch": 0.01, + "learning_rate": 1.7063306697030788e-05, + "loss": 3.8759, + "step": 2011648 + }, + { + "epoch": 0.01, + "learning_rate": 1.7054920749520264e-05, + "loss": 3.8632, + "step": 2012160 + }, + { + "epoch": 0.01, + "learning_rate": 1.7046534802009744e-05, + "loss": 3.8709, + "step": 2012672 + }, + { + "epoch": 0.01, + "learning_rate": 1.7038148854499224e-05, + "loss": 3.8698, + "step": 2013184 + }, + { + "epoch": 0.01, + "learning_rate": 1.7029762906988704e-05, + "loss": 3.8715, + "step": 2013696 + }, + { + "epoch": 0.01, + "learning_rate": 1.7021393338281917e-05, + "loss": 3.8866, + "step": 2014208 + }, + { + "epoch": 0.01, + "learning_rate": 1.701302376957513e-05, + "loss": 3.8757, + "step": 2014720 + }, + { + "epoch": 0.01, + "learning_rate": 1.700463782206461e-05, + "loss": 3.8725, + "step": 2015232 + }, + { + "epoch": 0.01, + "learning_rate": 1.6996251874554086e-05, + "loss": 3.8906, + "step": 2015744 + }, + { + "epoch": 0.01, + "learning_rate": 1.6987865927043566e-05, + "loss": 3.8604, + "step": 2016256 + }, + { + "epoch": 0.01, + "learning_rate": 1.6979479979533046e-05, + "loss": 3.8663, + "step": 2016768 + }, + { + "epoch": 0.01, + "learning_rate": 1.697109403202253e-05, + "loss": 3.8807, + "step": 2017280 + }, + { + "epoch": 0.01, + "learning_rate": 1.696270808451201e-05, + "loss": 3.8738, + "step": 2017792 + }, + { + "epoch": 0.01, + "learning_rate": 1.695432213700149e-05, + "loss": 3.873, + "step": 2018304 + }, + { + "epoch": 0.01, + "learning_rate": 1.6945952568294698e-05, + "loss": 3.8784, + "step": 2018816 + }, + { + "epoch": 0.01, + "learning_rate": 1.6937566620784178e-05, + "loss": 3.8609, + "step": 2019328 + }, + { + "epoch": 0.01, + "learning_rate": 1.6929180673273658e-05, + "loss": 3.8704, + "step": 2019840 + }, + { + "epoch": 0.01, + "learning_rate": 1.6920794725763138e-05, + "loss": 3.8852, + "step": 2020352 + }, + { + "epoch": 0.01, + "learning_rate": 1.691242515705635e-05, + "loss": 3.8718, + "step": 2020864 + }, + { + "epoch": 0.01, + "learning_rate": 1.690405558834956e-05, + "loss": 3.8849, + "step": 2021376 + }, + { + "epoch": 0.01, + "learning_rate": 1.689566964083904e-05, + "loss": 3.8783, + "step": 2021888 + }, + { + "epoch": 0.01, + "learning_rate": 1.688728369332852e-05, + "loss": 3.8958, + "step": 2022400 + }, + { + "epoch": 0.01, + "learning_rate": 1.6878897745818e-05, + "loss": 3.8625, + "step": 2022912 + }, + { + "epoch": 0.01, + "learning_rate": 1.6870511798307483e-05, + "loss": 3.8848, + "step": 2023424 + }, + { + "epoch": 0.01, + "learning_rate": 1.6862142229600692e-05, + "loss": 3.8805, + "step": 2023936 + }, + { + "epoch": 0.01, + "learning_rate": 1.685375628209017e-05, + "loss": 3.8643, + "step": 2024448 + }, + { + "epoch": 0.01, + "learning_rate": 1.684537033457965e-05, + "loss": 3.8885, + "step": 2024960 + }, + { + "epoch": 0.01, + "learning_rate": 1.683698438706913e-05, + "loss": 3.8798, + "step": 2025472 + }, + { + "epoch": 0.01, + "learning_rate": 1.682859843955861e-05, + "loss": 3.8806, + "step": 2025984 + }, + { + "epoch": 0.01, + "learning_rate": 1.6820228870851824e-05, + "loss": 3.8866, + "step": 2026496 + }, + { + "epoch": 0.01, + "learning_rate": 1.6811842923341304e-05, + "loss": 3.8651, + "step": 2027008 + }, + { + "epoch": 0.01, + "learning_rate": 1.6803456975830784e-05, + "loss": 3.8655, + "step": 2027520 + }, + { + "epoch": 0.01, + "learning_rate": 1.6795071028320264e-05, + "loss": 3.8674, + "step": 2028032 + }, + { + "epoch": 0.01, + "learning_rate": 1.678668508080974e-05, + "loss": 3.8799, + "step": 2028544 + }, + { + "epoch": 0.01, + "learning_rate": 1.677829913329922e-05, + "loss": 3.8742, + "step": 2029056 + }, + { + "epoch": 0.01, + "learning_rate": 1.6769913185788704e-05, + "loss": 3.8719, + "step": 2029568 + }, + { + "epoch": 0.01, + "learning_rate": 1.6761527238278184e-05, + "loss": 3.8727, + "step": 2030080 + }, + { + "epoch": 0.02, + "learning_rate": 1.6753141290767664e-05, + "loss": 3.881, + "step": 2030592 + }, + { + "epoch": 0.02, + "learning_rate": 1.6744771722060873e-05, + "loss": 3.8734, + "step": 2031104 + }, + { + "epoch": 0.02, + "learning_rate": 1.6736385774550353e-05, + "loss": 3.87, + "step": 2031616 + }, + { + "epoch": 0.02, + "learning_rate": 1.6727999827039833e-05, + "loss": 3.8914, + "step": 2032128 + }, + { + "epoch": 0.02, + "learning_rate": 1.6719613879529313e-05, + "loss": 3.8842, + "step": 2032640 + }, + { + "epoch": 0.02, + "learning_rate": 1.6711244310822525e-05, + "loss": 3.8834, + "step": 2033152 + }, + { + "epoch": 0.02, + "learning_rate": 1.6702874742115738e-05, + "loss": 3.8659, + "step": 2033664 + }, + { + "epoch": 0.02, + "learning_rate": 1.6694488794605214e-05, + "loss": 3.8772, + "step": 2034176 + }, + { + "epoch": 0.02, + "learning_rate": 1.6686102847094694e-05, + "loss": 3.871, + "step": 2034688 + }, + { + "epoch": 0.02, + "learning_rate": 1.6677716899584174e-05, + "loss": 3.8889, + "step": 2035200 + }, + { + "epoch": 0.02, + "learning_rate": 1.6669330952073654e-05, + "loss": 3.8772, + "step": 2035712 + }, + { + "epoch": 0.02, + "learning_rate": 1.6660945004563137e-05, + "loss": 3.8735, + "step": 2036224 + }, + { + "epoch": 0.02, + "learning_rate": 1.6652575435856346e-05, + "loss": 3.8796, + "step": 2036736 + }, + { + "epoch": 0.02, + "learning_rate": 1.6644189488345826e-05, + "loss": 3.8819, + "step": 2037248 + }, + { + "epoch": 0.02, + "learning_rate": 1.6635803540835306e-05, + "loss": 3.8717, + "step": 2037760 + }, + { + "epoch": 0.02, + "learning_rate": 1.6627417593324786e-05, + "loss": 3.8772, + "step": 2038272 + }, + { + "epoch": 0.02, + "learning_rate": 1.6619031645814266e-05, + "loss": 3.8652, + "step": 2038784 + }, + { + "epoch": 0.02, + "learning_rate": 1.6610645698303746e-05, + "loss": 3.8718, + "step": 2039296 + }, + { + "epoch": 0.02, + "learning_rate": 1.660227612959696e-05, + "loss": 3.8679, + "step": 2039808 + }, + { + "epoch": 0.02, + "learning_rate": 1.659389018208644e-05, + "loss": 3.8681, + "step": 2040320 + }, + { + "epoch": 0.02, + "learning_rate": 1.658550423457592e-05, + "loss": 3.8748, + "step": 2040832 + }, + { + "epoch": 0.02, + "learning_rate": 1.6577118287065395e-05, + "loss": 3.8756, + "step": 2041344 + }, + { + "epoch": 0.02, + "learning_rate": 1.6568732339554875e-05, + "loss": 3.8727, + "step": 2041856 + }, + { + "epoch": 0.02, + "learning_rate": 1.656034639204436e-05, + "loss": 3.8688, + "step": 2042368 + }, + { + "epoch": 0.02, + "learning_rate": 1.655197682333757e-05, + "loss": 3.8835, + "step": 2042880 + }, + { + "epoch": 0.02, + "learning_rate": 1.6543590875827048e-05, + "loss": 3.8886, + "step": 2043392 + }, + { + "epoch": 0.02, + "learning_rate": 1.653522130712026e-05, + "loss": 3.878, + "step": 2043904 + }, + { + "epoch": 0.02, + "learning_rate": 1.652683535960974e-05, + "loss": 3.8818, + "step": 2044416 + }, + { + "epoch": 0.02, + "learning_rate": 1.651844941209922e-05, + "loss": 3.868, + "step": 2044928 + }, + { + "epoch": 0.02, + "learning_rate": 1.65100634645887e-05, + "loss": 3.8784, + "step": 2045440 + }, + { + "epoch": 0.02, + "learning_rate": 1.650167751707818e-05, + "loss": 3.8761, + "step": 2045952 + }, + { + "epoch": 0.02, + "learning_rate": 1.649329156956766e-05, + "loss": 3.8657, + "step": 2046464 + }, + { + "epoch": 0.02, + "learning_rate": 1.648490562205714e-05, + "loss": 3.8759, + "step": 2046976 + }, + { + "epoch": 0.02, + "learning_rate": 1.647651967454662e-05, + "loss": 3.876, + "step": 2047488 + }, + { + "epoch": 0.02, + "learning_rate": 1.64681337270361e-05, + "loss": 3.8804, + "step": 2048000 + }, + { + "epoch": 0.02, + "learning_rate": 1.6459747779525576e-05, + "loss": 3.8754, + "step": 2048512 + }, + { + "epoch": 0.02, + "learning_rate": 1.645136183201506e-05, + "loss": 3.8778, + "step": 2049024 + }, + { + "epoch": 0.02, + "learning_rate": 1.644297588450454e-05, + "loss": 3.8769, + "step": 2049536 + }, + { + "epoch": 0.02, + "learning_rate": 1.6434606315797752e-05, + "loss": 3.8688, + "step": 2050048 + }, + { + "epoch": 0.02, + "learning_rate": 1.642622036828723e-05, + "loss": 3.8802, + "step": 2050560 + }, + { + "epoch": 0.02, + "learning_rate": 1.641785079958044e-05, + "loss": 3.8839, + "step": 2051072 + }, + { + "epoch": 0.02, + "learning_rate": 1.640946485206992e-05, + "loss": 3.8834, + "step": 2051584 + }, + { + "epoch": 0.02, + "learning_rate": 1.64010789045594e-05, + "loss": 3.8728, + "step": 2052096 + }, + { + "epoch": 0.02, + "learning_rate": 1.639269295704888e-05, + "loss": 3.8904, + "step": 2052608 + }, + { + "epoch": 0.02, + "learning_rate": 1.638430700953836e-05, + "loss": 3.8798, + "step": 2053120 + }, + { + "epoch": 0.02, + "learning_rate": 1.637592106202784e-05, + "loss": 3.8852, + "step": 2053632 + }, + { + "epoch": 0.02, + "learning_rate": 1.636753511451732e-05, + "loss": 3.884, + "step": 2054144 + }, + { + "epoch": 0.02, + "learning_rate": 1.63591491670068e-05, + "loss": 3.8708, + "step": 2054656 + }, + { + "epoch": 0.02, + "learning_rate": 1.6350779598300013e-05, + "loss": 3.8747, + "step": 2055168 + }, + { + "epoch": 0.02, + "learning_rate": 1.6342393650789493e-05, + "loss": 3.8897, + "step": 2055680 + }, + { + "epoch": 0.02, + "learning_rate": 1.6334024082082702e-05, + "loss": 3.8649, + "step": 2056192 + }, + { + "epoch": 0.02, + "learning_rate": 1.6325638134572182e-05, + "loss": 3.8764, + "step": 2056704 + }, + { + "epoch": 0.02, + "learning_rate": 1.6317252187061662e-05, + "loss": 3.8814, + "step": 2057216 + }, + { + "epoch": 0.02, + "learning_rate": 1.6308866239551142e-05, + "loss": 3.8802, + "step": 2057728 + }, + { + "epoch": 0.02, + "learning_rate": 1.6300480292040622e-05, + "loss": 3.8711, + "step": 2058240 + }, + { + "epoch": 0.02, + "learning_rate": 1.6292094344530105e-05, + "loss": 3.8772, + "step": 2058752 + }, + { + "epoch": 0.02, + "learning_rate": 1.6283708397019582e-05, + "loss": 3.881, + "step": 2059264 + }, + { + "epoch": 0.02, + "learning_rate": 1.6275322449509062e-05, + "loss": 3.8826, + "step": 2059776 + }, + { + "epoch": 0.02, + "learning_rate": 1.6266952880802274e-05, + "loss": 3.883, + "step": 2060288 + }, + { + "epoch": 0.03, + "eval_loss": 3.9817380905151367, + "eval_runtime": 298.9818, + "eval_samples_per_second": 1276.302, + "eval_steps_per_second": 39.885, + "step": 2060627 + }, + { + "epoch": 1.0, + "learning_rate": 1.6258583312095484e-05, + "loss": 3.8664, + "step": 2060800 + }, + { + "epoch": 1.0, + "learning_rate": 1.6250197364584967e-05, + "loss": 3.8734, + "step": 2061312 + }, + { + "epoch": 1.0, + "learning_rate": 1.6241811417074447e-05, + "loss": 3.8747, + "step": 2061824 + }, + { + "epoch": 1.0, + "learning_rate": 1.6233425469563927e-05, + "loss": 3.8773, + "step": 2062336 + }, + { + "epoch": 1.0, + "learning_rate": 1.6225039522053403e-05, + "loss": 3.8909, + "step": 2062848 + }, + { + "epoch": 1.0, + "learning_rate": 1.6216653574542883e-05, + "loss": 3.8758, + "step": 2063360 + }, + { + "epoch": 1.0, + "learning_rate": 1.6208267627032363e-05, + "loss": 3.8759, + "step": 2063872 + }, + { + "epoch": 1.0, + "learning_rate": 1.6199881679521843e-05, + "loss": 3.8767, + "step": 2064384 + }, + { + "epoch": 1.0, + "learning_rate": 1.6191512110815056e-05, + "loss": 3.8704, + "step": 2064896 + }, + { + "epoch": 1.0, + "learning_rate": 1.6183126163304536e-05, + "loss": 3.8702, + "step": 2065408 + }, + { + "epoch": 1.0, + "learning_rate": 1.6174740215794016e-05, + "loss": 3.88, + "step": 2065920 + }, + { + "epoch": 1.0, + "learning_rate": 1.6166354268283496e-05, + "loss": 3.8878, + "step": 2066432 + }, + { + "epoch": 1.0, + "learning_rate": 1.6157984699576705e-05, + "loss": 3.8785, + "step": 2066944 + }, + { + "epoch": 1.0, + "learning_rate": 1.6149598752066185e-05, + "loss": 3.8602, + "step": 2067456 + }, + { + "epoch": 1.0, + "learning_rate": 1.6141212804555668e-05, + "loss": 3.8786, + "step": 2067968 + }, + { + "epoch": 1.0, + "learning_rate": 1.6132826857045148e-05, + "loss": 3.8629, + "step": 2068480 + }, + { + "epoch": 1.0, + "learning_rate": 1.6124440909534628e-05, + "loss": 3.8747, + "step": 2068992 + }, + { + "epoch": 1.0, + "learning_rate": 1.6116054962024108e-05, + "loss": 3.8721, + "step": 2069504 + }, + { + "epoch": 1.0, + "learning_rate": 1.6107669014513584e-05, + "loss": 3.8755, + "step": 2070016 + }, + { + "epoch": 1.0, + "learning_rate": 1.6099283067003064e-05, + "loss": 3.8777, + "step": 2070528 + }, + { + "epoch": 1.0, + "learning_rate": 1.6090913498296277e-05, + "loss": 3.8867, + "step": 2071040 + }, + { + "epoch": 1.0, + "learning_rate": 1.608252755078576e-05, + "loss": 3.8808, + "step": 2071552 + }, + { + "epoch": 1.0, + "learning_rate": 1.6074141603275237e-05, + "loss": 3.8786, + "step": 2072064 + }, + { + "epoch": 1.0, + "learning_rate": 1.6065755655764717e-05, + "loss": 3.8689, + "step": 2072576 + }, + { + "epoch": 1.0, + "learning_rate": 1.605738608705793e-05, + "loss": 3.8809, + "step": 2073088 + }, + { + "epoch": 1.0, + "learning_rate": 1.604900013954741e-05, + "loss": 3.8769, + "step": 2073600 + }, + { + "epoch": 1.0, + "learning_rate": 1.6040614192036886e-05, + "loss": 3.8701, + "step": 2074112 + }, + { + "epoch": 1.0, + "learning_rate": 1.603222824452637e-05, + "loss": 3.8744, + "step": 2074624 + }, + { + "epoch": 1.0, + "learning_rate": 1.602384229701585e-05, + "loss": 3.871, + "step": 2075136 + }, + { + "epoch": 1.0, + "learning_rate": 1.601545634950533e-05, + "loss": 3.8761, + "step": 2075648 + }, + { + "epoch": 1.01, + "learning_rate": 1.600707040199481e-05, + "loss": 3.8699, + "step": 2076160 + }, + { + "epoch": 1.01, + "learning_rate": 1.599868445448429e-05, + "loss": 3.8782, + "step": 2076672 + }, + { + "epoch": 1.01, + "learning_rate": 1.5990298506973765e-05, + "loss": 3.8847, + "step": 2077184 + }, + { + "epoch": 1.01, + "learning_rate": 1.5981928938266978e-05, + "loss": 3.8774, + "step": 2077696 + }, + { + "epoch": 1.01, + "learning_rate": 1.597354299075646e-05, + "loss": 3.877, + "step": 2078208 + }, + { + "epoch": 1.01, + "learning_rate": 1.596517342204967e-05, + "loss": 3.8745, + "step": 2078720 + }, + { + "epoch": 1.01, + "learning_rate": 1.595678747453915e-05, + "loss": 3.8837, + "step": 2079232 + }, + { + "epoch": 1.01, + "learning_rate": 1.594840152702863e-05, + "loss": 3.8695, + "step": 2079744 + }, + { + "epoch": 1.01, + "learning_rate": 1.594001557951811e-05, + "loss": 3.866, + "step": 2080256 + }, + { + "epoch": 1.01, + "learning_rate": 1.5931646010811323e-05, + "loss": 3.8734, + "step": 2080768 + }, + { + "epoch": 1.01, + "learning_rate": 1.5923260063300803e-05, + "loss": 3.8683, + "step": 2081280 + }, + { + "epoch": 1.01, + "learning_rate": 1.5914874115790283e-05, + "loss": 3.8758, + "step": 2081792 + }, + { + "epoch": 1.01, + "learning_rate": 1.5906488168279763e-05, + "loss": 3.8692, + "step": 2082304 + }, + { + "epoch": 1.01, + "learning_rate": 1.589810222076924e-05, + "loss": 3.874, + "step": 2082816 + }, + { + "epoch": 1.01, + "learning_rate": 1.5889732652062452e-05, + "loss": 3.8795, + "step": 2083328 + }, + { + "epoch": 1.01, + "learning_rate": 1.588134670455193e-05, + "loss": 3.8787, + "step": 2083840 + }, + { + "epoch": 1.01, + "learning_rate": 1.5872960757041415e-05, + "loss": 3.8814, + "step": 2084352 + }, + { + "epoch": 1.01, + "learning_rate": 1.586457480953089e-05, + "loss": 3.8707, + "step": 2084864 + }, + { + "epoch": 1.01, + "learning_rate": 1.585618886202037e-05, + "loss": 3.8499, + "step": 2085376 + }, + { + "epoch": 1.01, + "learning_rate": 1.584780291450985e-05, + "loss": 3.8836, + "step": 2085888 + }, + { + "epoch": 1.01, + "learning_rate": 1.583941696699933e-05, + "loss": 3.8715, + "step": 2086400 + }, + { + "epoch": 1.01, + "learning_rate": 1.583103101948881e-05, + "loss": 3.8759, + "step": 2086912 + }, + { + "epoch": 1.01, + "learning_rate": 1.5822677829585756e-05, + "loss": 3.8735, + "step": 2087424 + }, + { + "epoch": 1.01, + "learning_rate": 1.5814291882075236e-05, + "loss": 3.872, + "step": 2087936 + }, + { + "epoch": 1.01, + "learning_rate": 1.5805905934564713e-05, + "loss": 3.8599, + "step": 2088448 + }, + { + "epoch": 1.01, + "learning_rate": 1.5797519987054193e-05, + "loss": 3.862, + "step": 2088960 + }, + { + "epoch": 1.01, + "learning_rate": 1.5789134039543673e-05, + "loss": 3.8683, + "step": 2089472 + }, + { + "epoch": 1.01, + "learning_rate": 1.5780748092033153e-05, + "loss": 3.8635, + "step": 2089984 + }, + { + "epoch": 1.01, + "learning_rate": 1.5772362144522633e-05, + "loss": 3.886, + "step": 2090496 + }, + { + "epoch": 1.01, + "learning_rate": 1.5763976197012116e-05, + "loss": 3.8724, + "step": 2091008 + }, + { + "epoch": 1.01, + "learning_rate": 1.5755590249501596e-05, + "loss": 3.8628, + "step": 2091520 + }, + { + "epoch": 1.01, + "learning_rate": 1.5747220680794805e-05, + "loss": 3.8852, + "step": 2092032 + }, + { + "epoch": 1.01, + "learning_rate": 1.5738834733284285e-05, + "loss": 3.8612, + "step": 2092544 + }, + { + "epoch": 1.01, + "learning_rate": 1.5730448785773765e-05, + "loss": 3.8616, + "step": 2093056 + }, + { + "epoch": 1.01, + "learning_rate": 1.5722079217066978e-05, + "loss": 3.8776, + "step": 2093568 + }, + { + "epoch": 1.01, + "learning_rate": 1.5713693269556457e-05, + "loss": 3.8678, + "step": 2094080 + }, + { + "epoch": 1.01, + "learning_rate": 1.5705307322045937e-05, + "loss": 3.8665, + "step": 2094592 + }, + { + "epoch": 1.01, + "learning_rate": 1.5696921374535417e-05, + "loss": 3.8734, + "step": 2095104 + }, + { + "epoch": 1.01, + "learning_rate": 1.5688535427024894e-05, + "loss": 3.8552, + "step": 2095616 + }, + { + "epoch": 1.01, + "learning_rate": 1.5680149479514374e-05, + "loss": 3.8665, + "step": 2096128 + }, + { + "epoch": 1.01, + "learning_rate": 1.5671763532003854e-05, + "loss": 3.88, + "step": 2096640 + }, + { + "epoch": 1.01, + "learning_rate": 1.5663377584493337e-05, + "loss": 3.8707, + "step": 2097152 + }, + { + "epoch": 1.01, + "learning_rate": 1.5654991636982817e-05, + "loss": 3.882, + "step": 2097664 + }, + { + "epoch": 1.01, + "learning_rate": 1.5646622068276026e-05, + "loss": 3.8732, + "step": 2098176 + }, + { + "epoch": 1.01, + "learning_rate": 1.5638236120765506e-05, + "loss": 3.8887, + "step": 2098688 + }, + { + "epoch": 1.01, + "learning_rate": 1.5629850173254986e-05, + "loss": 3.8591, + "step": 2099200 + }, + { + "epoch": 1.01, + "learning_rate": 1.5621464225744466e-05, + "loss": 3.8805, + "step": 2099712 + }, + { + "epoch": 1.01, + "learning_rate": 1.561309465703768e-05, + "loss": 3.877, + "step": 2100224 + }, + { + "epoch": 1.01, + "learning_rate": 1.560470870952716e-05, + "loss": 3.8601, + "step": 2100736 + }, + { + "epoch": 1.01, + "learning_rate": 1.559632276201664e-05, + "loss": 3.8793, + "step": 2101248 + }, + { + "epoch": 1.01, + "learning_rate": 1.558793681450612e-05, + "loss": 3.8773, + "step": 2101760 + }, + { + "epoch": 1.01, + "learning_rate": 1.55795508669956e-05, + "loss": 3.8795, + "step": 2102272 + }, + { + "epoch": 1.01, + "learning_rate": 1.557119767709254e-05, + "loss": 3.8821, + "step": 2102784 + }, + { + "epoch": 1.01, + "learning_rate": 1.556281172958202e-05, + "loss": 3.8623, + "step": 2103296 + }, + { + "epoch": 1.01, + "learning_rate": 1.55544257820715e-05, + "loss": 3.855, + "step": 2103808 + }, + { + "epoch": 1.01, + "learning_rate": 1.554603983456098e-05, + "loss": 3.8656, + "step": 2104320 + }, + { + "epoch": 1.01, + "learning_rate": 1.553765388705046e-05, + "loss": 3.8743, + "step": 2104832 + }, + { + "epoch": 1.01, + "learning_rate": 1.552926793953994e-05, + "loss": 3.8708, + "step": 2105344 + }, + { + "epoch": 1.01, + "learning_rate": 1.552088199202942e-05, + "loss": 3.8646, + "step": 2105856 + }, + { + "epoch": 1.01, + "learning_rate": 1.55124960445189e-05, + "loss": 3.8665, + "step": 2106368 + }, + { + "epoch": 1.02, + "learning_rate": 1.5504126475812112e-05, + "loss": 3.8736, + "step": 2106880 + }, + { + "epoch": 1.02, + "learning_rate": 1.5495740528301592e-05, + "loss": 3.8706, + "step": 2107392 + }, + { + "epoch": 1.02, + "learning_rate": 1.5487354580791072e-05, + "loss": 3.8709, + "step": 2107904 + }, + { + "epoch": 1.02, + "learning_rate": 1.547896863328055e-05, + "loss": 3.8815, + "step": 2108416 + }, + { + "epoch": 1.02, + "learning_rate": 1.547058268577003e-05, + "loss": 3.885, + "step": 2108928 + }, + { + "epoch": 1.02, + "learning_rate": 1.546219673825951e-05, + "loss": 3.8744, + "step": 2109440 + }, + { + "epoch": 1.02, + "learning_rate": 1.5453810790748992e-05, + "loss": 3.867, + "step": 2109952 + }, + { + "epoch": 1.02, + "learning_rate": 1.5445424843238472e-05, + "loss": 3.8681, + "step": 2110464 + }, + { + "epoch": 1.02, + "learning_rate": 1.543705527453168e-05, + "loss": 3.8663, + "step": 2110976 + }, + { + "epoch": 1.02, + "learning_rate": 1.542866932702116e-05, + "loss": 3.8799, + "step": 2111488 + }, + { + "epoch": 1.02, + "learning_rate": 1.542028337951064e-05, + "loss": 3.8746, + "step": 2112000 + }, + { + "epoch": 1.02, + "learning_rate": 1.541189743200012e-05, + "loss": 3.8695, + "step": 2112512 + }, + { + "epoch": 1.02, + "learning_rate": 1.54035114844896e-05, + "loss": 3.8723, + "step": 2113024 + }, + { + "epoch": 1.02, + "learning_rate": 1.5395158294586546e-05, + "loss": 3.8751, + "step": 2113536 + }, + { + "epoch": 1.02, + "learning_rate": 1.5386788725879755e-05, + "loss": 3.8717, + "step": 2114048 + }, + { + "epoch": 1.02, + "learning_rate": 1.5378402778369235e-05, + "loss": 3.869, + "step": 2114560 + }, + { + "epoch": 1.02, + "learning_rate": 1.5370033209662447e-05, + "loss": 3.8636, + "step": 2115072 + }, + { + "epoch": 1.02, + "learning_rate": 1.5361647262151927e-05, + "loss": 3.8617, + "step": 2115584 + }, + { + "epoch": 1.02, + "learning_rate": 1.5353261314641407e-05, + "loss": 3.8688, + "step": 2116096 + }, + { + "epoch": 1.02, + "learning_rate": 1.5344875367130887e-05, + "loss": 3.8623, + "step": 2116608 + }, + { + "epoch": 1.02, + "learning_rate": 1.5336489419620367e-05, + "loss": 3.8698, + "step": 2117120 + }, + { + "epoch": 1.02, + "learning_rate": 1.5328103472109844e-05, + "loss": 3.868, + "step": 2117632 + }, + { + "epoch": 1.02, + "learning_rate": 1.5319717524599324e-05, + "loss": 3.8673, + "step": 2118144 + }, + { + "epoch": 1.02, + "learning_rate": 1.5311331577088807e-05, + "loss": 3.8637, + "step": 2118656 + }, + { + "epoch": 1.02, + "learning_rate": 1.5302945629578287e-05, + "loss": 3.883, + "step": 2119168 + }, + { + "epoch": 1.02, + "learning_rate": 1.5294559682067767e-05, + "loss": 3.8869, + "step": 2119680 + }, + { + "epoch": 1.02, + "learning_rate": 1.5286173734557247e-05, + "loss": 3.8723, + "step": 2120192 + }, + { + "epoch": 1.02, + "learning_rate": 1.5277787787046727e-05, + "loss": 3.8784, + "step": 2120704 + }, + { + "epoch": 1.02, + "learning_rate": 1.5269401839536203e-05, + "loss": 3.8674, + "step": 2121216 + }, + { + "epoch": 1.02, + "learning_rate": 1.5261015892025683e-05, + "loss": 3.8723, + "step": 2121728 + }, + { + "epoch": 1.02, + "learning_rate": 1.5252629944515165e-05, + "loss": 3.8696, + "step": 2122240 + }, + { + "epoch": 1.02, + "learning_rate": 1.5244260375808378e-05, + "loss": 3.8618, + "step": 2122752 + }, + { + "epoch": 1.02, + "learning_rate": 1.5235874428297856e-05, + "loss": 3.868, + "step": 2123264 + }, + { + "epoch": 1.02, + "learning_rate": 1.5227488480787336e-05, + "loss": 3.871, + "step": 2123776 + }, + { + "epoch": 1.02, + "learning_rate": 1.5219102533276816e-05, + "loss": 3.8787, + "step": 2124288 + }, + { + "epoch": 1.02, + "learning_rate": 1.5210716585766296e-05, + "loss": 3.872, + "step": 2124800 + }, + { + "epoch": 1.02, + "learning_rate": 1.5202347017059506e-05, + "loss": 3.874, + "step": 2125312 + }, + { + "epoch": 1.02, + "learning_rate": 1.5193961069548986e-05, + "loss": 3.8715, + "step": 2125824 + }, + { + "epoch": 1.02, + "learning_rate": 1.5185575122038468e-05, + "loss": 3.8643, + "step": 2126336 + }, + { + "epoch": 1.02, + "learning_rate": 1.5177189174527948e-05, + "loss": 3.8745, + "step": 2126848 + }, + { + "epoch": 1.02, + "learning_rate": 1.5168803227017428e-05, + "loss": 3.8799, + "step": 2127360 + }, + { + "epoch": 1.02, + "learning_rate": 1.5160417279506908e-05, + "loss": 3.8816, + "step": 2127872 + }, + { + "epoch": 1.02, + "learning_rate": 1.5152031331996386e-05, + "loss": 3.8655, + "step": 2128384 + }, + { + "epoch": 1.02, + "learning_rate": 1.5143645384485866e-05, + "loss": 3.8882, + "step": 2128896 + }, + { + "epoch": 1.02, + "learning_rate": 1.5135259436975346e-05, + "loss": 3.8699, + "step": 2129408 + }, + { + "epoch": 1.02, + "learning_rate": 1.5126889868268557e-05, + "loss": 3.8829, + "step": 2129920 + }, + { + "epoch": 1.02, + "learning_rate": 1.5118503920758037e-05, + "loss": 3.8741, + "step": 2130432 + }, + { + "epoch": 1.02, + "learning_rate": 1.5110117973247517e-05, + "loss": 3.8694, + "step": 2130944 + }, + { + "epoch": 1.02, + "learning_rate": 1.5101732025736997e-05, + "loss": 3.8679, + "step": 2131456 + }, + { + "epoch": 1.02, + "learning_rate": 1.5093346078226478e-05, + "loss": 3.8871, + "step": 2131968 + }, + { + "epoch": 1.02, + "learning_rate": 1.5084976509519688e-05, + "loss": 3.861, + "step": 2132480 + }, + { + "epoch": 1.02, + "learning_rate": 1.507659056200917e-05, + "loss": 3.8736, + "step": 2132992 + }, + { + "epoch": 1.02, + "learning_rate": 1.5068204614498649e-05, + "loss": 3.8742, + "step": 2133504 + }, + { + "epoch": 1.02, + "learning_rate": 1.5059818666988129e-05, + "loss": 3.8751, + "step": 2134016 + }, + { + "epoch": 1.02, + "learning_rate": 1.505144909828134e-05, + "loss": 3.8687, + "step": 2134528 + }, + { + "epoch": 1.02, + "learning_rate": 1.5043079529574552e-05, + "loss": 3.8719, + "step": 2135040 + }, + { + "epoch": 1.02, + "learning_rate": 1.503469358206403e-05, + "loss": 3.8737, + "step": 2135552 + }, + { + "epoch": 1.02, + "learning_rate": 1.502630763455351e-05, + "loss": 3.8848, + "step": 2136064 + }, + { + "epoch": 1.02, + "learning_rate": 1.501792168704299e-05, + "loss": 3.8755, + "step": 2136576 + }, + { + "epoch": 1.03, + "eval_loss": 3.980809211730957, + "eval_runtime": 303.474, + "eval_samples_per_second": 1257.409, + "eval_steps_per_second": 39.295, + "step": 2136947 + }, + { + "epoch": 0.0, + "learning_rate": 1.500953573953247e-05, + "loss": 3.8547, + "step": 2137088 + }, + { + "epoch": 0.0, + "learning_rate": 1.500114979202195e-05, + "loss": 3.8675, + "step": 2137600 + }, + { + "epoch": 0.0, + "learning_rate": 1.4992763844511432e-05, + "loss": 3.8701, + "step": 2138112 + }, + { + "epoch": 0.0, + "learning_rate": 1.4984377897000912e-05, + "loss": 3.8743, + "step": 2138624 + }, + { + "epoch": 0.0, + "learning_rate": 1.4975991949490389e-05, + "loss": 3.8824, + "step": 2139136 + }, + { + "epoch": 0.0, + "learning_rate": 1.496760600197987e-05, + "loss": 3.8678, + "step": 2139648 + }, + { + "epoch": 0.0, + "learning_rate": 1.495922005446935e-05, + "loss": 3.8769, + "step": 2140160 + }, + { + "epoch": 0.0, + "learning_rate": 1.495083410695883e-05, + "loss": 3.8718, + "step": 2140672 + }, + { + "epoch": 0.0, + "learning_rate": 1.494244815944831e-05, + "loss": 3.8647, + "step": 2141184 + }, + { + "epoch": 0.0, + "learning_rate": 1.4934062211937792e-05, + "loss": 3.8608, + "step": 2141696 + }, + { + "epoch": 0.0, + "learning_rate": 1.4925676264427272e-05, + "loss": 3.8803, + "step": 2142208 + }, + { + "epoch": 0.0, + "learning_rate": 1.4917290316916748e-05, + "loss": 3.8836, + "step": 2142720 + }, + { + "epoch": 0.0, + "learning_rate": 1.4908920748209963e-05, + "loss": 3.8744, + "step": 2143232 + }, + { + "epoch": 0.0, + "learning_rate": 1.4900534800699442e-05, + "loss": 3.8553, + "step": 2143744 + }, + { + "epoch": 0.0, + "learning_rate": 1.4892148853188919e-05, + "loss": 3.8738, + "step": 2144256 + }, + { + "epoch": 0.0, + "learning_rate": 1.48837629056784e-05, + "loss": 3.8584, + "step": 2144768 + }, + { + "epoch": 0.0, + "learning_rate": 1.487537695816788e-05, + "loss": 3.8761, + "step": 2145280 + }, + { + "epoch": 0.0, + "learning_rate": 1.486699101065736e-05, + "loss": 3.8696, + "step": 2145792 + }, + { + "epoch": 0.0, + "learning_rate": 1.485860506314684e-05, + "loss": 3.8682, + "step": 2146304 + }, + { + "epoch": 0.0, + "learning_rate": 1.485021911563632e-05, + "loss": 3.873, + "step": 2146816 + }, + { + "epoch": 0.0, + "learning_rate": 1.4841833168125802e-05, + "loss": 3.8811, + "step": 2147328 + }, + { + "epoch": 0.0, + "learning_rate": 1.4833447220615279e-05, + "loss": 3.8736, + "step": 2147840 + }, + { + "epoch": 0.0, + "learning_rate": 1.4825061273104759e-05, + "loss": 3.8776, + "step": 2148352 + }, + { + "epoch": 0.0, + "learning_rate": 1.481667532559424e-05, + "loss": 3.8692, + "step": 2148864 + }, + { + "epoch": 0.0, + "learning_rate": 1.4808305756887453e-05, + "loss": 3.874, + "step": 2149376 + }, + { + "epoch": 0.0, + "learning_rate": 1.4799919809376931e-05, + "loss": 3.875, + "step": 2149888 + }, + { + "epoch": 0.0, + "learning_rate": 1.4791533861866411e-05, + "loss": 3.8608, + "step": 2150400 + }, + { + "epoch": 0.0, + "learning_rate": 1.4783147914355891e-05, + "loss": 3.869, + "step": 2150912 + }, + { + "epoch": 0.0, + "learning_rate": 1.4774761966845371e-05, + "loss": 3.8696, + "step": 2151424 + }, + { + "epoch": 0.0, + "learning_rate": 1.4766376019334851e-05, + "loss": 3.8694, + "step": 2151936 + }, + { + "epoch": 0.01, + "learning_rate": 1.4757990071824333e-05, + "loss": 3.867, + "step": 2152448 + }, + { + "epoch": 0.01, + "learning_rate": 1.4749604124313809e-05, + "loss": 3.875, + "step": 2152960 + }, + { + "epoch": 0.01, + "learning_rate": 1.4741234555607023e-05, + "loss": 3.8743, + "step": 2153472 + }, + { + "epoch": 0.01, + "learning_rate": 1.4732848608096503e-05, + "loss": 3.8788, + "step": 2153984 + }, + { + "epoch": 0.01, + "learning_rate": 1.4724462660585983e-05, + "loss": 3.8705, + "step": 2154496 + }, + { + "epoch": 0.01, + "learning_rate": 1.4716093091879194e-05, + "loss": 3.8707, + "step": 2155008 + }, + { + "epoch": 0.01, + "learning_rate": 1.4707707144368674e-05, + "loss": 3.8818, + "step": 2155520 + }, + { + "epoch": 0.01, + "learning_rate": 1.4699321196858154e-05, + "loss": 3.8649, + "step": 2156032 + }, + { + "epoch": 0.01, + "learning_rate": 1.4690935249347634e-05, + "loss": 3.8632, + "step": 2156544 + }, + { + "epoch": 0.01, + "learning_rate": 1.4682549301837112e-05, + "loss": 3.8672, + "step": 2157056 + }, + { + "epoch": 0.01, + "learning_rate": 1.4674163354326592e-05, + "loss": 3.8641, + "step": 2157568 + }, + { + "epoch": 0.01, + "learning_rate": 1.4665777406816072e-05, + "loss": 3.87, + "step": 2158080 + }, + { + "epoch": 0.01, + "learning_rate": 1.4657391459305552e-05, + "loss": 3.8626, + "step": 2158592 + }, + { + "epoch": 0.01, + "learning_rate": 1.4649021890598763e-05, + "loss": 3.8707, + "step": 2159104 + }, + { + "epoch": 0.01, + "learning_rate": 1.4640635943088243e-05, + "loss": 3.8736, + "step": 2159616 + }, + { + "epoch": 0.01, + "learning_rate": 1.4632249995577724e-05, + "loss": 3.8756, + "step": 2160128 + }, + { + "epoch": 0.01, + "learning_rate": 1.4623864048067204e-05, + "loss": 3.875, + "step": 2160640 + }, + { + "epoch": 0.01, + "learning_rate": 1.4615494479360413e-05, + "loss": 3.8728, + "step": 2161152 + }, + { + "epoch": 0.01, + "learning_rate": 1.4607108531849895e-05, + "loss": 3.8479, + "step": 2161664 + }, + { + "epoch": 0.01, + "learning_rate": 1.4598722584339375e-05, + "loss": 3.8803, + "step": 2162176 + }, + { + "epoch": 0.01, + "learning_rate": 1.4590336636828855e-05, + "loss": 3.8648, + "step": 2162688 + }, + { + "epoch": 0.01, + "learning_rate": 1.4581967068122066e-05, + "loss": 3.8703, + "step": 2163200 + }, + { + "epoch": 0.01, + "learning_rate": 1.4573581120611546e-05, + "loss": 3.8669, + "step": 2163712 + }, + { + "epoch": 0.01, + "learning_rate": 1.4565211551904757e-05, + "loss": 3.8666, + "step": 2164224 + }, + { + "epoch": 0.01, + "learning_rate": 1.4556825604394236e-05, + "loss": 3.8569, + "step": 2164736 + }, + { + "epoch": 0.01, + "learning_rate": 1.4548439656883716e-05, + "loss": 3.8614, + "step": 2165248 + }, + { + "epoch": 0.01, + "learning_rate": 1.4540053709373196e-05, + "loss": 3.8642, + "step": 2165760 + }, + { + "epoch": 0.01, + "learning_rate": 1.4531667761862678e-05, + "loss": 3.8547, + "step": 2166272 + }, + { + "epoch": 0.01, + "learning_rate": 1.4523298193155887e-05, + "loss": 3.8799, + "step": 2166784 + }, + { + "epoch": 0.01, + "learning_rate": 1.4514912245645367e-05, + "loss": 3.8703, + "step": 2167296 + }, + { + "epoch": 0.01, + "learning_rate": 1.4506526298134849e-05, + "loss": 3.8581, + "step": 2167808 + }, + { + "epoch": 0.01, + "learning_rate": 1.4498140350624329e-05, + "loss": 3.8799, + "step": 2168320 + }, + { + "epoch": 0.01, + "learning_rate": 1.4489754403113809e-05, + "loss": 3.857, + "step": 2168832 + }, + { + "epoch": 0.01, + "learning_rate": 1.4481368455603287e-05, + "loss": 3.8578, + "step": 2169344 + }, + { + "epoch": 0.01, + "learning_rate": 1.4472982508092767e-05, + "loss": 3.8693, + "step": 2169856 + }, + { + "epoch": 0.01, + "learning_rate": 1.4464596560582247e-05, + "loss": 3.8647, + "step": 2170368 + }, + { + "epoch": 0.01, + "learning_rate": 1.445622699187546e-05, + "loss": 3.8601, + "step": 2170880 + }, + { + "epoch": 0.01, + "learning_rate": 1.4447841044364938e-05, + "loss": 3.867, + "step": 2171392 + }, + { + "epoch": 0.01, + "learning_rate": 1.4439455096854418e-05, + "loss": 3.8566, + "step": 2171904 + }, + { + "epoch": 0.01, + "learning_rate": 1.4431069149343897e-05, + "loss": 3.8567, + "step": 2172416 + }, + { + "epoch": 0.01, + "learning_rate": 1.4422683201833379e-05, + "loss": 3.8768, + "step": 2172928 + }, + { + "epoch": 0.01, + "learning_rate": 1.4414297254322859e-05, + "loss": 3.868, + "step": 2173440 + }, + { + "epoch": 0.01, + "learning_rate": 1.4405911306812339e-05, + "loss": 3.8734, + "step": 2173952 + }, + { + "epoch": 0.01, + "learning_rate": 1.4397525359301819e-05, + "loss": 3.8718, + "step": 2174464 + }, + { + "epoch": 0.01, + "learning_rate": 1.438915579059503e-05, + "loss": 3.8849, + "step": 2174976 + }, + { + "epoch": 0.01, + "learning_rate": 1.438078622188824e-05, + "loss": 3.8566, + "step": 2175488 + }, + { + "epoch": 0.01, + "learning_rate": 1.437240027437772e-05, + "loss": 3.8766, + "step": 2176000 + }, + { + "epoch": 0.01, + "learning_rate": 1.43640143268672e-05, + "loss": 3.8677, + "step": 2176512 + }, + { + "epoch": 0.01, + "learning_rate": 1.435562837935668e-05, + "loss": 3.8609, + "step": 2177024 + }, + { + "epoch": 0.01, + "learning_rate": 1.434724243184616e-05, + "loss": 3.8705, + "step": 2177536 + }, + { + "epoch": 0.01, + "learning_rate": 1.4338856484335642e-05, + "loss": 3.8775, + "step": 2178048 + }, + { + "epoch": 0.01, + "learning_rate": 1.4330470536825119e-05, + "loss": 3.8752, + "step": 2178560 + }, + { + "epoch": 0.01, + "learning_rate": 1.4322084589314599e-05, + "loss": 3.882, + "step": 2179072 + }, + { + "epoch": 0.01, + "learning_rate": 1.431369864180408e-05, + "loss": 3.8596, + "step": 2179584 + }, + { + "epoch": 0.01, + "learning_rate": 1.4305329073097293e-05, + "loss": 3.8499, + "step": 2180096 + }, + { + "epoch": 0.01, + "learning_rate": 1.4296943125586771e-05, + "loss": 3.8603, + "step": 2180608 + }, + { + "epoch": 0.01, + "learning_rate": 1.4288557178076251e-05, + "loss": 3.8729, + "step": 2181120 + }, + { + "epoch": 0.01, + "learning_rate": 1.4280171230565731e-05, + "loss": 3.8684, + "step": 2181632 + }, + { + "epoch": 0.01, + "learning_rate": 1.4271801661858942e-05, + "loss": 3.8629, + "step": 2182144 + }, + { + "epoch": 0.01, + "learning_rate": 1.4263415714348422e-05, + "loss": 3.8625, + "step": 2182656 + }, + { + "epoch": 0.02, + "learning_rate": 1.4255029766837902e-05, + "loss": 3.8711, + "step": 2183168 + }, + { + "epoch": 0.02, + "learning_rate": 1.4246643819327382e-05, + "loss": 3.863, + "step": 2183680 + }, + { + "epoch": 0.02, + "learning_rate": 1.4238257871816863e-05, + "loss": 3.8661, + "step": 2184192 + }, + { + "epoch": 0.02, + "learning_rate": 1.4229888303110072e-05, + "loss": 3.8757, + "step": 2184704 + }, + { + "epoch": 0.02, + "learning_rate": 1.4221502355599552e-05, + "loss": 3.8817, + "step": 2185216 + }, + { + "epoch": 0.02, + "learning_rate": 1.4213116408089034e-05, + "loss": 3.8716, + "step": 2185728 + }, + { + "epoch": 0.02, + "learning_rate": 1.4204730460578514e-05, + "loss": 3.8626, + "step": 2186240 + }, + { + "epoch": 0.02, + "learning_rate": 1.4196360891871725e-05, + "loss": 3.8649, + "step": 2186752 + }, + { + "epoch": 0.02, + "learning_rate": 1.4187974944361205e-05, + "loss": 3.8631, + "step": 2187264 + }, + { + "epoch": 0.02, + "learning_rate": 1.4179588996850685e-05, + "loss": 3.8761, + "step": 2187776 + }, + { + "epoch": 0.02, + "learning_rate": 1.4171203049340165e-05, + "loss": 3.8734, + "step": 2188288 + }, + { + "epoch": 0.02, + "learning_rate": 1.4162817101829644e-05, + "loss": 3.8657, + "step": 2188800 + }, + { + "epoch": 0.02, + "learning_rate": 1.4154431154319123e-05, + "loss": 3.8703, + "step": 2189312 + }, + { + "epoch": 0.02, + "learning_rate": 1.4146045206808603e-05, + "loss": 3.8663, + "step": 2189824 + }, + { + "epoch": 0.02, + "learning_rate": 1.4137675638101817e-05, + "loss": 3.8684, + "step": 2190336 + }, + { + "epoch": 0.02, + "learning_rate": 1.4129289690591297e-05, + "loss": 3.8627, + "step": 2190848 + }, + { + "epoch": 0.02, + "learning_rate": 1.4120903743080773e-05, + "loss": 3.8634, + "step": 2191360 + }, + { + "epoch": 0.02, + "learning_rate": 1.4112517795570255e-05, + "loss": 3.8557, + "step": 2191872 + }, + { + "epoch": 0.02, + "learning_rate": 1.4104131848059735e-05, + "loss": 3.8631, + "step": 2192384 + }, + { + "epoch": 0.02, + "learning_rate": 1.4095745900549215e-05, + "loss": 3.8657, + "step": 2192896 + }, + { + "epoch": 0.02, + "learning_rate": 1.4087359953038695e-05, + "loss": 3.8622, + "step": 2193408 + }, + { + "epoch": 0.02, + "learning_rate": 1.4078974005528175e-05, + "loss": 3.8632, + "step": 2193920 + }, + { + "epoch": 0.02, + "learning_rate": 1.4070604436821386e-05, + "loss": 3.8641, + "step": 2194432 + }, + { + "epoch": 0.02, + "learning_rate": 1.4062218489310866e-05, + "loss": 3.8588, + "step": 2194944 + }, + { + "epoch": 0.02, + "learning_rate": 1.4053848920604076e-05, + "loss": 3.8767, + "step": 2195456 + }, + { + "epoch": 0.02, + "learning_rate": 1.4045462973093556e-05, + "loss": 3.8804, + "step": 2195968 + }, + { + "epoch": 0.02, + "learning_rate": 1.4037077025583036e-05, + "loss": 3.8705, + "step": 2196480 + }, + { + "epoch": 0.02, + "learning_rate": 1.4028691078072518e-05, + "loss": 3.8705, + "step": 2196992 + }, + { + "epoch": 0.02, + "learning_rate": 1.4020305130561998e-05, + "loss": 3.8663, + "step": 2197504 + }, + { + "epoch": 0.02, + "learning_rate": 1.4011935561855207e-05, + "loss": 3.867, + "step": 2198016 + }, + { + "epoch": 0.02, + "learning_rate": 1.4003549614344689e-05, + "loss": 3.8711, + "step": 2198528 + }, + { + "epoch": 0.02, + "learning_rate": 1.3995163666834169e-05, + "loss": 3.8549, + "step": 2199040 + }, + { + "epoch": 0.02, + "learning_rate": 1.3986777719323649e-05, + "loss": 3.8671, + "step": 2199552 + }, + { + "epoch": 0.02, + "learning_rate": 1.3978391771813127e-05, + "loss": 3.8665, + "step": 2200064 + }, + { + "epoch": 0.02, + "learning_rate": 1.3970005824302607e-05, + "loss": 3.8738, + "step": 2200576 + }, + { + "epoch": 0.02, + "learning_rate": 1.3961619876792087e-05, + "loss": 3.8661, + "step": 2201088 + }, + { + "epoch": 0.02, + "learning_rate": 1.39532503080853e-05, + "loss": 3.8704, + "step": 2201600 + }, + { + "epoch": 0.02, + "learning_rate": 1.3944864360574777e-05, + "loss": 3.8675, + "step": 2202112 + }, + { + "epoch": 0.02, + "learning_rate": 1.3936478413064257e-05, + "loss": 3.8596, + "step": 2202624 + }, + { + "epoch": 0.02, + "learning_rate": 1.3928092465553737e-05, + "loss": 3.8717, + "step": 2203136 + }, + { + "epoch": 0.02, + "learning_rate": 1.3919722896846952e-05, + "loss": 3.8723, + "step": 2203648 + }, + { + "epoch": 0.02, + "learning_rate": 1.3911336949336428e-05, + "loss": 3.876, + "step": 2204160 + }, + { + "epoch": 0.02, + "learning_rate": 1.390295100182591e-05, + "loss": 3.8653, + "step": 2204672 + }, + { + "epoch": 0.02, + "learning_rate": 1.389456505431539e-05, + "loss": 3.8814, + "step": 2205184 + }, + { + "epoch": 0.02, + "learning_rate": 1.388617910680487e-05, + "loss": 3.869, + "step": 2205696 + }, + { + "epoch": 0.02, + "learning_rate": 1.387780953809808e-05, + "loss": 3.8755, + "step": 2206208 + }, + { + "epoch": 0.02, + "learning_rate": 1.386942359058756e-05, + "loss": 3.8741, + "step": 2206720 + }, + { + "epoch": 0.02, + "learning_rate": 1.386103764307704e-05, + "loss": 3.8667, + "step": 2207232 + }, + { + "epoch": 0.02, + "learning_rate": 1.385265169556652e-05, + "loss": 3.8636, + "step": 2207744 + }, + { + "epoch": 0.02, + "learning_rate": 1.3844265748056002e-05, + "loss": 3.8825, + "step": 2208256 + }, + { + "epoch": 0.02, + "learning_rate": 1.3835879800545482e-05, + "loss": 3.8591, + "step": 2208768 + }, + { + "epoch": 0.02, + "learning_rate": 1.3827493853034959e-05, + "loss": 3.8662, + "step": 2209280 + }, + { + "epoch": 0.02, + "learning_rate": 1.381910790552444e-05, + "loss": 3.872, + "step": 2209792 + }, + { + "epoch": 0.02, + "learning_rate": 1.381072195801392e-05, + "loss": 3.873, + "step": 2210304 + }, + { + "epoch": 0.02, + "learning_rate": 1.3802352389307133e-05, + "loss": 3.8662, + "step": 2210816 + }, + { + "epoch": 0.02, + "learning_rate": 1.3793966441796611e-05, + "loss": 3.868, + "step": 2211328 + }, + { + "epoch": 0.02, + "learning_rate": 1.378558049428609e-05, + "loss": 3.8695, + "step": 2211840 + }, + { + "epoch": 0.02, + "learning_rate": 1.3777210925579303e-05, + "loss": 3.8762, + "step": 2212352 + }, + { + "epoch": 0.02, + "learning_rate": 1.3768824978068782e-05, + "loss": 3.8735, + "step": 2212864 + }, + { + "epoch": 0.03, + "eval_loss": 3.9795851707458496, + "eval_runtime": 310.6169, + "eval_samples_per_second": 1228.494, + "eval_steps_per_second": 38.391, + "step": 2213267 + }, + { + "epoch": 1.0, + "learning_rate": 1.3760455409361994e-05, + "loss": 3.8622, + "step": 2213376 + }, + { + "epoch": 1.0, + "learning_rate": 1.3752069461851474e-05, + "loss": 3.8564, + "step": 2213888 + }, + { + "epoch": 1.0, + "learning_rate": 1.3743683514340954e-05, + "loss": 3.8659, + "step": 2214400 + }, + { + "epoch": 1.0, + "learning_rate": 1.3735297566830432e-05, + "loss": 3.8766, + "step": 2214912 + }, + { + "epoch": 1.0, + "learning_rate": 1.3726911619319912e-05, + "loss": 3.8732, + "step": 2215424 + }, + { + "epoch": 1.0, + "learning_rate": 1.3718525671809392e-05, + "loss": 3.8676, + "step": 2215936 + }, + { + "epoch": 1.0, + "learning_rate": 1.3710139724298874e-05, + "loss": 3.8707, + "step": 2216448 + }, + { + "epoch": 1.0, + "learning_rate": 1.3701753776788354e-05, + "loss": 3.8693, + "step": 2216960 + }, + { + "epoch": 1.0, + "learning_rate": 1.3693367829277834e-05, + "loss": 3.859, + "step": 2217472 + }, + { + "epoch": 1.0, + "learning_rate": 1.3684998260571045e-05, + "loss": 3.8594, + "step": 2217984 + }, + { + "epoch": 1.0, + "learning_rate": 1.3676612313060524e-05, + "loss": 3.8738, + "step": 2218496 + }, + { + "epoch": 1.0, + "learning_rate": 1.3668226365550004e-05, + "loss": 3.874, + "step": 2219008 + }, + { + "epoch": 1.0, + "learning_rate": 1.3659856796843215e-05, + "loss": 3.878, + "step": 2219520 + }, + { + "epoch": 1.0, + "learning_rate": 1.3651470849332695e-05, + "loss": 3.8468, + "step": 2220032 + }, + { + "epoch": 1.0, + "learning_rate": 1.3643084901822175e-05, + "loss": 3.8679, + "step": 2220544 + }, + { + "epoch": 1.0, + "learning_rate": 1.3634698954311657e-05, + "loss": 3.8566, + "step": 2221056 + }, + { + "epoch": 1.0, + "learning_rate": 1.3626313006801137e-05, + "loss": 3.8691, + "step": 2221568 + }, + { + "epoch": 1.0, + "learning_rate": 1.3617927059290613e-05, + "loss": 3.8661, + "step": 2222080 + }, + { + "epoch": 1.0, + "learning_rate": 1.3609541111780095e-05, + "loss": 3.8639, + "step": 2222592 + }, + { + "epoch": 1.0, + "learning_rate": 1.3601155164269575e-05, + "loss": 3.8663, + "step": 2223104 + }, + { + "epoch": 1.0, + "learning_rate": 1.3592769216759055e-05, + "loss": 3.8831, + "step": 2223616 + }, + { + "epoch": 1.0, + "learning_rate": 1.3584399648052266e-05, + "loss": 3.8676, + "step": 2224128 + }, + { + "epoch": 1.0, + "learning_rate": 1.3576013700541746e-05, + "loss": 3.8703, + "step": 2224640 + }, + { + "epoch": 1.0, + "learning_rate": 1.3567627753031226e-05, + "loss": 3.8631, + "step": 2225152 + }, + { + "epoch": 1.0, + "learning_rate": 1.3559241805520706e-05, + "loss": 3.8739, + "step": 2225664 + }, + { + "epoch": 1.0, + "learning_rate": 1.3550872236813916e-05, + "loss": 3.8678, + "step": 2226176 + }, + { + "epoch": 1.0, + "learning_rate": 1.3542486289303396e-05, + "loss": 3.8578, + "step": 2226688 + }, + { + "epoch": 1.0, + "learning_rate": 1.3534100341792876e-05, + "loss": 3.8666, + "step": 2227200 + }, + { + "epoch": 1.0, + "learning_rate": 1.3525714394282358e-05, + "loss": 3.863, + "step": 2227712 + }, + { + "epoch": 1.0, + "learning_rate": 1.3517328446771838e-05, + "loss": 3.8621, + "step": 2228224 + }, + { + "epoch": 1.01, + "learning_rate": 1.3508958878065049e-05, + "loss": 3.867, + "step": 2228736 + }, + { + "epoch": 1.01, + "learning_rate": 1.3500572930554529e-05, + "loss": 3.8692, + "step": 2229248 + }, + { + "epoch": 1.01, + "learning_rate": 1.3492186983044009e-05, + "loss": 3.8705, + "step": 2229760 + }, + { + "epoch": 1.01, + "learning_rate": 1.3483801035533488e-05, + "loss": 3.8703, + "step": 2230272 + }, + { + "epoch": 1.01, + "learning_rate": 1.3475415088022968e-05, + "loss": 3.8699, + "step": 2230784 + }, + { + "epoch": 1.01, + "learning_rate": 1.346704551931618e-05, + "loss": 3.8668, + "step": 2231296 + }, + { + "epoch": 1.01, + "learning_rate": 1.345865957180566e-05, + "loss": 3.8739, + "step": 2231808 + }, + { + "epoch": 1.01, + "learning_rate": 1.3450273624295139e-05, + "loss": 3.8577, + "step": 2232320 + }, + { + "epoch": 1.01, + "learning_rate": 1.3441887676784617e-05, + "loss": 3.8582, + "step": 2232832 + }, + { + "epoch": 1.01, + "learning_rate": 1.3433501729274097e-05, + "loss": 3.8657, + "step": 2233344 + }, + { + "epoch": 1.01, + "learning_rate": 1.3425115781763577e-05, + "loss": 3.8592, + "step": 2233856 + }, + { + "epoch": 1.01, + "learning_rate": 1.3416729834253059e-05, + "loss": 3.8662, + "step": 2234368 + }, + { + "epoch": 1.01, + "learning_rate": 1.3408343886742539e-05, + "loss": 3.8617, + "step": 2234880 + }, + { + "epoch": 1.01, + "learning_rate": 1.339997431803575e-05, + "loss": 3.8675, + "step": 2235392 + }, + { + "epoch": 1.01, + "learning_rate": 1.339158837052523e-05, + "loss": 3.8715, + "step": 2235904 + }, + { + "epoch": 1.01, + "learning_rate": 1.338320242301471e-05, + "loss": 3.869, + "step": 2236416 + }, + { + "epoch": 1.01, + "learning_rate": 1.337481647550419e-05, + "loss": 3.8703, + "step": 2236928 + }, + { + "epoch": 1.01, + "learning_rate": 1.33664469067974e-05, + "loss": 3.8669, + "step": 2237440 + }, + { + "epoch": 1.01, + "learning_rate": 1.335806095928688e-05, + "loss": 3.8451, + "step": 2237952 + }, + { + "epoch": 1.01, + "learning_rate": 1.334967501177636e-05, + "loss": 3.8713, + "step": 2238464 + }, + { + "epoch": 1.01, + "learning_rate": 1.3341305443069571e-05, + "loss": 3.8639, + "step": 2238976 + }, + { + "epoch": 1.01, + "learning_rate": 1.3332919495559051e-05, + "loss": 3.8679, + "step": 2239488 + }, + { + "epoch": 1.01, + "learning_rate": 1.3324533548048531e-05, + "loss": 3.8624, + "step": 2240000 + }, + { + "epoch": 1.01, + "learning_rate": 1.3316147600538013e-05, + "loss": 3.859, + "step": 2240512 + }, + { + "epoch": 1.01, + "learning_rate": 1.3307761653027493e-05, + "loss": 3.8542, + "step": 2241024 + }, + { + "epoch": 1.01, + "learning_rate": 1.3299375705516973e-05, + "loss": 3.8566, + "step": 2241536 + }, + { + "epoch": 1.01, + "learning_rate": 1.329098975800645e-05, + "loss": 3.8631, + "step": 2242048 + }, + { + "epoch": 1.01, + "learning_rate": 1.3282620189299663e-05, + "loss": 3.8505, + "step": 2242560 + }, + { + "epoch": 1.01, + "learning_rate": 1.3274234241789143e-05, + "loss": 3.877, + "step": 2243072 + }, + { + "epoch": 1.01, + "learning_rate": 1.3265848294278621e-05, + "loss": 3.8682, + "step": 2243584 + }, + { + "epoch": 1.01, + "learning_rate": 1.3257462346768101e-05, + "loss": 3.8486, + "step": 2244096 + }, + { + "epoch": 1.01, + "learning_rate": 1.3249076399257581e-05, + "loss": 3.8769, + "step": 2244608 + }, + { + "epoch": 1.01, + "learning_rate": 1.3240690451747061e-05, + "loss": 3.8534, + "step": 2245120 + }, + { + "epoch": 1.01, + "learning_rate": 1.3232304504236543e-05, + "loss": 3.8522, + "step": 2245632 + }, + { + "epoch": 1.01, + "learning_rate": 1.3223918556726023e-05, + "loss": 3.8647, + "step": 2246144 + }, + { + "epoch": 1.01, + "learning_rate": 1.3215548988019234e-05, + "loss": 3.8658, + "step": 2246656 + }, + { + "epoch": 1.01, + "learning_rate": 1.3207163040508714e-05, + "loss": 3.8545, + "step": 2247168 + }, + { + "epoch": 1.01, + "learning_rate": 1.3198777092998194e-05, + "loss": 3.8695, + "step": 2247680 + }, + { + "epoch": 1.01, + "learning_rate": 1.3190391145487674e-05, + "loss": 3.8473, + "step": 2248192 + }, + { + "epoch": 1.01, + "learning_rate": 1.3182005197977154e-05, + "loss": 3.8536, + "step": 2248704 + }, + { + "epoch": 1.01, + "learning_rate": 1.3173619250466632e-05, + "loss": 3.8732, + "step": 2249216 + }, + { + "epoch": 1.01, + "learning_rate": 1.3165249681759844e-05, + "loss": 3.8591, + "step": 2249728 + }, + { + "epoch": 1.01, + "learning_rate": 1.3156863734249324e-05, + "loss": 3.8703, + "step": 2250240 + }, + { + "epoch": 1.01, + "learning_rate": 1.3148477786738803e-05, + "loss": 3.8633, + "step": 2250752 + }, + { + "epoch": 1.01, + "learning_rate": 1.3140091839228283e-05, + "loss": 3.8816, + "step": 2251264 + }, + { + "epoch": 1.01, + "learning_rate": 1.3131705891717762e-05, + "loss": 3.8554, + "step": 2251776 + }, + { + "epoch": 1.01, + "learning_rate": 1.3123336323010977e-05, + "loss": 3.8708, + "step": 2252288 + }, + { + "epoch": 1.01, + "learning_rate": 1.3114950375500453e-05, + "loss": 3.8621, + "step": 2252800 + }, + { + "epoch": 1.01, + "learning_rate": 1.3106564427989935e-05, + "loss": 3.8592, + "step": 2253312 + }, + { + "epoch": 1.01, + "learning_rate": 1.3098178480479415e-05, + "loss": 3.8655, + "step": 2253824 + }, + { + "epoch": 1.01, + "learning_rate": 1.3089792532968895e-05, + "loss": 3.8767, + "step": 2254336 + }, + { + "epoch": 1.01, + "learning_rate": 1.3081406585458375e-05, + "loss": 3.8636, + "step": 2254848 + }, + { + "epoch": 1.01, + "learning_rate": 1.3073020637947855e-05, + "loss": 3.8789, + "step": 2255360 + }, + { + "epoch": 1.01, + "learning_rate": 1.3064634690437336e-05, + "loss": 3.8517, + "step": 2255872 + }, + { + "epoch": 1.01, + "learning_rate": 1.3056265121730545e-05, + "loss": 3.851, + "step": 2256384 + }, + { + "epoch": 1.01, + "learning_rate": 1.3047879174220027e-05, + "loss": 3.8561, + "step": 2256896 + }, + { + "epoch": 1.01, + "learning_rate": 1.3039509605513236e-05, + "loss": 3.8625, + "step": 2257408 + }, + { + "epoch": 1.01, + "learning_rate": 1.3031123658002716e-05, + "loss": 3.8639, + "step": 2257920 + }, + { + "epoch": 1.01, + "learning_rate": 1.3022737710492198e-05, + "loss": 3.8616, + "step": 2258432 + }, + { + "epoch": 1.01, + "learning_rate": 1.3014351762981678e-05, + "loss": 3.8589, + "step": 2258944 + }, + { + "epoch": 1.02, + "learning_rate": 1.3005965815471158e-05, + "loss": 3.8632, + "step": 2259456 + }, + { + "epoch": 1.02, + "learning_rate": 1.2997579867960636e-05, + "loss": 3.8617, + "step": 2259968 + }, + { + "epoch": 1.02, + "learning_rate": 1.2989193920450116e-05, + "loss": 3.8594, + "step": 2260480 + }, + { + "epoch": 1.02, + "learning_rate": 1.2980824351743328e-05, + "loss": 3.8733, + "step": 2260992 + }, + { + "epoch": 1.02, + "learning_rate": 1.2972438404232808e-05, + "loss": 3.8743, + "step": 2261504 + }, + { + "epoch": 1.02, + "learning_rate": 1.2964052456722287e-05, + "loss": 3.8717, + "step": 2262016 + }, + { + "epoch": 1.02, + "learning_rate": 1.2955666509211767e-05, + "loss": 3.8551, + "step": 2262528 + }, + { + "epoch": 1.02, + "learning_rate": 1.2947280561701247e-05, + "loss": 3.8613, + "step": 2263040 + }, + { + "epoch": 1.02, + "learning_rate": 1.2938894614190728e-05, + "loss": 3.8615, + "step": 2263552 + }, + { + "epoch": 1.02, + "learning_rate": 1.2930508666680208e-05, + "loss": 3.8689, + "step": 2264064 + }, + { + "epoch": 1.02, + "learning_rate": 1.2922122719169688e-05, + "loss": 3.8687, + "step": 2264576 + }, + { + "epoch": 1.02, + "learning_rate": 1.2913753150462899e-05, + "loss": 3.8613, + "step": 2265088 + }, + { + "epoch": 1.02, + "learning_rate": 1.2905367202952379e-05, + "loss": 3.8646, + "step": 2265600 + }, + { + "epoch": 1.02, + "learning_rate": 1.2896981255441859e-05, + "loss": 3.8665, + "step": 2266112 + }, + { + "epoch": 1.02, + "learning_rate": 1.2888595307931339e-05, + "loss": 3.8647, + "step": 2266624 + }, + { + "epoch": 1.02, + "learning_rate": 1.288022573922455e-05, + "loss": 3.8573, + "step": 2267136 + }, + { + "epoch": 1.02, + "learning_rate": 1.287185617051776e-05, + "loss": 3.8592, + "step": 2267648 + }, + { + "epoch": 1.02, + "learning_rate": 1.286347022300724e-05, + "loss": 3.8492, + "step": 2268160 + }, + { + "epoch": 1.02, + "learning_rate": 1.285508427549672e-05, + "loss": 3.8614, + "step": 2268672 + }, + { + "epoch": 1.02, + "learning_rate": 1.28466983279862e-05, + "loss": 3.8583, + "step": 2269184 + }, + { + "epoch": 1.02, + "learning_rate": 1.2838328759279411e-05, + "loss": 3.8596, + "step": 2269696 + }, + { + "epoch": 1.02, + "learning_rate": 1.2829942811768891e-05, + "loss": 3.8596, + "step": 2270208 + }, + { + "epoch": 1.02, + "learning_rate": 1.2821556864258371e-05, + "loss": 3.8631, + "step": 2270720 + }, + { + "epoch": 1.02, + "learning_rate": 1.2813170916747853e-05, + "loss": 3.8531, + "step": 2271232 + }, + { + "epoch": 1.02, + "learning_rate": 1.2804784969237332e-05, + "loss": 3.8729, + "step": 2271744 + }, + { + "epoch": 1.02, + "learning_rate": 1.2796399021726812e-05, + "loss": 3.8739, + "step": 2272256 + }, + { + "epoch": 1.02, + "learning_rate": 1.278801307421629e-05, + "loss": 3.8677, + "step": 2272768 + }, + { + "epoch": 1.02, + "learning_rate": 1.277962712670577e-05, + "loss": 3.8675, + "step": 2273280 + }, + { + "epoch": 1.02, + "learning_rate": 1.2771257557998983e-05, + "loss": 3.8632, + "step": 2273792 + }, + { + "epoch": 1.02, + "learning_rate": 1.2762871610488461e-05, + "loss": 3.861, + "step": 2274304 + }, + { + "epoch": 1.02, + "learning_rate": 1.2754485662977941e-05, + "loss": 3.8672, + "step": 2274816 + }, + { + "epoch": 1.02, + "learning_rate": 1.2746099715467421e-05, + "loss": 3.8492, + "step": 2275328 + }, + { + "epoch": 1.02, + "learning_rate": 1.2737713767956901e-05, + "loss": 3.8657, + "step": 2275840 + }, + { + "epoch": 1.02, + "learning_rate": 1.2729344199250112e-05, + "loss": 3.8558, + "step": 2276352 + }, + { + "epoch": 1.02, + "learning_rate": 1.2720958251739592e-05, + "loss": 3.873, + "step": 2276864 + }, + { + "epoch": 1.02, + "learning_rate": 1.2712588683032806e-05, + "loss": 3.865, + "step": 2277376 + }, + { + "epoch": 1.02, + "learning_rate": 1.2704202735522286e-05, + "loss": 3.8658, + "step": 2277888 + }, + { + "epoch": 1.02, + "learning_rate": 1.2695816788011763e-05, + "loss": 3.859, + "step": 2278400 + }, + { + "epoch": 1.02, + "learning_rate": 1.2687430840501244e-05, + "loss": 3.8564, + "step": 2278912 + }, + { + "epoch": 1.02, + "learning_rate": 1.2679044892990724e-05, + "loss": 3.8662, + "step": 2279424 + }, + { + "epoch": 1.02, + "learning_rate": 1.2670658945480204e-05, + "loss": 3.8714, + "step": 2279936 + }, + { + "epoch": 1.02, + "learning_rate": 1.2662272997969684e-05, + "loss": 3.8749, + "step": 2280448 + }, + { + "epoch": 1.02, + "learning_rate": 1.2653887050459164e-05, + "loss": 3.8591, + "step": 2280960 + }, + { + "epoch": 1.02, + "learning_rate": 1.2645517481752375e-05, + "loss": 3.8765, + "step": 2281472 + }, + { + "epoch": 1.02, + "learning_rate": 1.2637131534241855e-05, + "loss": 3.8642, + "step": 2281984 + }, + { + "epoch": 1.02, + "learning_rate": 1.2628745586731337e-05, + "loss": 3.8721, + "step": 2282496 + }, + { + "epoch": 1.02, + "learning_rate": 1.2620359639220817e-05, + "loss": 3.8712, + "step": 2283008 + }, + { + "epoch": 1.02, + "learning_rate": 1.2611990070514026e-05, + "loss": 3.8619, + "step": 2283520 + }, + { + "epoch": 1.02, + "learning_rate": 1.2603604123003507e-05, + "loss": 3.8591, + "step": 2284032 + }, + { + "epoch": 1.02, + "learning_rate": 1.2595218175492987e-05, + "loss": 3.8775, + "step": 2284544 + }, + { + "epoch": 1.02, + "learning_rate": 1.2586832227982467e-05, + "loss": 3.8576, + "step": 2285056 + }, + { + "epoch": 1.02, + "learning_rate": 1.2578446280471945e-05, + "loss": 3.8586, + "step": 2285568 + }, + { + "epoch": 1.02, + "learning_rate": 1.2570076711765158e-05, + "loss": 3.8713, + "step": 2286080 + }, + { + "epoch": 1.02, + "learning_rate": 1.2561690764254638e-05, + "loss": 3.8691, + "step": 2286592 + }, + { + "epoch": 1.02, + "learning_rate": 1.2553304816744116e-05, + "loss": 3.8616, + "step": 2287104 + }, + { + "epoch": 1.02, + "learning_rate": 1.2544918869233596e-05, + "loss": 3.8643, + "step": 2287616 + }, + { + "epoch": 1.02, + "learning_rate": 1.2536532921723076e-05, + "loss": 3.8644, + "step": 2288128 + }, + { + "epoch": 1.02, + "learning_rate": 1.252816335301629e-05, + "loss": 3.8727, + "step": 2288640 + }, + { + "epoch": 1.02, + "learning_rate": 1.2519777405505767e-05, + "loss": 3.8684, + "step": 2289152 + }, + { + "epoch": 1.03, + "eval_loss": 3.9781768321990967, + "eval_runtime": 311.7696, + "eval_samples_per_second": 1223.952, + "eval_steps_per_second": 38.249, + "step": 2289587 + } + ], + "logging_steps": 512, + "max_steps": 3052726, + "num_train_epochs": 9223372036854775807, + "save_steps": 10, + "total_flos": 9.455002375787199e+17, + "trial_name": null, + "trial_params": null +}