| { | |
| "best_metric": 0.22850316762924194, | |
| "best_model_checkpoint": "miner_id_24_1/checkpoint-1000", | |
| "epoch": 0.5092297899427116, | |
| "eval_steps": 200, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0005092297899427117, | |
| "grad_norm": 7.810187816619873, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0307, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0010184595798854233, | |
| "grad_norm": 11.901244163513184, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2115, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.001527689369828135, | |
| "grad_norm": 26.66057777404785, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3876, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0020369191597708466, | |
| "grad_norm": 36.39434051513672, | |
| "learning_rate": 4e-05, | |
| "loss": 2.3747, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.002546148949713558, | |
| "grad_norm": 38.85515594482422, | |
| "learning_rate": 5e-05, | |
| "loss": 2.2585, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00305537873965627, | |
| "grad_norm": 31.20570945739746, | |
| "learning_rate": 6e-05, | |
| "loss": 1.8399, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0035646085295989813, | |
| "grad_norm": 53.85519027709961, | |
| "learning_rate": 7e-05, | |
| "loss": 1.7199, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.004073838319541693, | |
| "grad_norm": 39.146419525146484, | |
| "learning_rate": 8e-05, | |
| "loss": 2.0047, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.004583068109484405, | |
| "grad_norm": 67.70555877685547, | |
| "learning_rate": 9e-05, | |
| "loss": 1.7328, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.005092297899427116, | |
| "grad_norm": 43.93235778808594, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2339, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005601527689369828, | |
| "grad_norm": 28.690454483032227, | |
| "learning_rate": 9.999974825027756e-05, | |
| "loss": 0.9872, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00611075747931254, | |
| "grad_norm": 56.98477554321289, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 1.0972, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.006619987269255252, | |
| "grad_norm": 18.829212188720703, | |
| "learning_rate": 9.999773426770865e-05, | |
| "loss": 0.4863, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.007129217059197963, | |
| "grad_norm": 5.442621231079102, | |
| "learning_rate": 9.999597205514297e-05, | |
| "loss": 0.124, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0076384468491406746, | |
| "grad_norm": 19.687393188476562, | |
| "learning_rate": 9.999370638369377e-05, | |
| "loss": 0.2149, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.008147676639083386, | |
| "grad_norm": 6.360881805419922, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.1119, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.008656906429026098, | |
| "grad_norm": 4.7020978927612305, | |
| "learning_rate": 9.998766476047547e-05, | |
| "loss": 0.1823, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00916613621896881, | |
| "grad_norm": 2.705449342727661, | |
| "learning_rate": 9.998388886954547e-05, | |
| "loss": 0.1085, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.009675366008911522, | |
| "grad_norm": 2.5857841968536377, | |
| "learning_rate": 9.997960964140947e-05, | |
| "loss": 0.0976, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.010184595798854232, | |
| "grad_norm": 16.0050106048584, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.1568, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.010693825588796944, | |
| "grad_norm": 8.77568244934082, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.1151, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.011203055378739656, | |
| "grad_norm": 1.004607915878296, | |
| "learning_rate": 9.996375239002369e-05, | |
| "loss": 0.0802, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.011712285168682368, | |
| "grad_norm": 2.0350327491760254, | |
| "learning_rate": 9.995746029466071e-05, | |
| "loss": 0.0452, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01222151495862508, | |
| "grad_norm": 301.12799072265625, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 7.738, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.012730744748567792, | |
| "grad_norm": 391.00762939453125, | |
| "learning_rate": 9.99433669591504e-05, | |
| "loss": 6.659, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.013239974538510503, | |
| "grad_norm": 192.62831115722656, | |
| "learning_rate": 9.993556586092281e-05, | |
| "loss": 3.7739, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.013749204328453215, | |
| "grad_norm": 79.25320434570312, | |
| "learning_rate": 9.992726191210138e-05, | |
| "loss": 1.652, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.014258434118395925, | |
| "grad_norm": 54.88349914550781, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.9881, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.014767663908338637, | |
| "grad_norm": 44.994449615478516, | |
| "learning_rate": 9.990914580222257e-05, | |
| "loss": 0.6137, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.015276893698281349, | |
| "grad_norm": 26.1174259185791, | |
| "learning_rate": 9.989933382359422e-05, | |
| "loss": 0.5558, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01578612348822406, | |
| "grad_norm": 15.71074390411377, | |
| "learning_rate": 9.988901935922826e-05, | |
| "loss": 0.5383, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.016295353278166773, | |
| "grad_norm": 15.574135780334473, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.3819, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.016804583068109485, | |
| "grad_norm": 10.029973030090332, | |
| "learning_rate": 9.986688339380862e-05, | |
| "loss": 0.2771, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.017313812858052197, | |
| "grad_norm": 10.430713653564453, | |
| "learning_rate": 9.985506211566388e-05, | |
| "loss": 0.2751, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01782304264799491, | |
| "grad_norm": 11.25537109375, | |
| "learning_rate": 9.984273879759713e-05, | |
| "loss": 0.1899, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01833227243793762, | |
| "grad_norm": 9.658145904541016, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.3, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.018841502227880332, | |
| "grad_norm": 19.340084075927734, | |
| "learning_rate": 9.981658654313457e-05, | |
| "loss": 0.275, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.019350732017823044, | |
| "grad_norm": 16.586389541625977, | |
| "learning_rate": 9.98027578700917e-05, | |
| "loss": 0.2442, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.019859961807765756, | |
| "grad_norm": 7.003499984741211, | |
| "learning_rate": 9.978842768382998e-05, | |
| "loss": 0.11, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.020369191597708464, | |
| "grad_norm": 21.9989013671875, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.1143, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.020878421387651176, | |
| "grad_norm": 11.020797729492188, | |
| "learning_rate": 9.975826335391808e-05, | |
| "loss": 0.0857, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.021387651177593888, | |
| "grad_norm": 10.189589500427246, | |
| "learning_rate": 9.974242951402235e-05, | |
| "loss": 0.0433, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0218968809675366, | |
| "grad_norm": 9.400593757629395, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.1319, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.022406110757479312, | |
| "grad_norm": 38.592594146728516, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.1928, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.022915340547422024, | |
| "grad_norm": 35.85441589355469, | |
| "learning_rate": 9.969192322306271e-05, | |
| "loss": 0.0925, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.023424570337364736, | |
| "grad_norm": 42.03386688232422, | |
| "learning_rate": 9.967408676742751e-05, | |
| "loss": 0.0906, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.023933800127307447, | |
| "grad_norm": 11.04820728302002, | |
| "learning_rate": 9.965575009429006e-05, | |
| "loss": 0.0713, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.02444302991725016, | |
| "grad_norm": 14.304110527038574, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.1799, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.02495225970719287, | |
| "grad_norm": 12.146672248840332, | |
| "learning_rate": 9.961757683914406e-05, | |
| "loss": 0.0691, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.025461489497135583, | |
| "grad_norm": 20.32611846923828, | |
| "learning_rate": 9.959774064153977e-05, | |
| "loss": 0.2816, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.025970719287078295, | |
| "grad_norm": 7.227572441101074, | |
| "learning_rate": 9.957740499523787e-05, | |
| "loss": 1.0021, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.026479949077021007, | |
| "grad_norm": 10.347118377685547, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 1.0069, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.02698917886696372, | |
| "grad_norm": 15.916715621948242, | |
| "learning_rate": 9.953523618068749e-05, | |
| "loss": 0.9734, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.02749840865690643, | |
| "grad_norm": 17.439035415649414, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.2134, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.028007638446849142, | |
| "grad_norm": 28.1020450592041, | |
| "learning_rate": 9.949107209404665e-05, | |
| "loss": 1.4139, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.02851686823679185, | |
| "grad_norm": 12.133536338806152, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.5215, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.029026098026734563, | |
| "grad_norm": 20.322139739990234, | |
| "learning_rate": 9.944491451423828e-05, | |
| "loss": 0.8477, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.029535327816677275, | |
| "grad_norm": 18.471527099609375, | |
| "learning_rate": 9.942108874226811e-05, | |
| "loss": 0.9458, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.030044557606619986, | |
| "grad_norm": 18.556156158447266, | |
| "learning_rate": 9.939676530048301e-05, | |
| "loss": 0.9743, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.030553787396562698, | |
| "grad_norm": 14.058014869689941, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.6557, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03106301718650541, | |
| "grad_norm": 25.136165618896484, | |
| "learning_rate": 9.934662639222412e-05, | |
| "loss": 0.8277, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.03157224697644812, | |
| "grad_norm": 13.859735488891602, | |
| "learning_rate": 9.93208114306486e-05, | |
| "loss": 0.3034, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.03208147676639083, | |
| "grad_norm": 3.4193358421325684, | |
| "learning_rate": 9.929449980904952e-05, | |
| "loss": 0.0745, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.032590706556333546, | |
| "grad_norm": 6.933294773101807, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.1584, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.033099936346276254, | |
| "grad_norm": 8.660609245300293, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.2456, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03360916613621897, | |
| "grad_norm": 1.4555833339691162, | |
| "learning_rate": 9.921258765867919e-05, | |
| "loss": 0.0626, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.03411839592616168, | |
| "grad_norm": 2.2231998443603516, | |
| "learning_rate": 9.918429209653662e-05, | |
| "loss": 0.1195, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.03462762571610439, | |
| "grad_norm": 5.060727596282959, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.1108, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0351368555060471, | |
| "grad_norm": 6.343770503997803, | |
| "learning_rate": 9.912621540634887e-05, | |
| "loss": 0.1127, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.03564608529598982, | |
| "grad_norm": 4.716639518737793, | |
| "learning_rate": 9.909643486313533e-05, | |
| "loss": 0.0833, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.036155315085932525, | |
| "grad_norm": 7.187185764312744, | |
| "learning_rate": 9.90661599193678e-05, | |
| "loss": 0.1607, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.03666454487587524, | |
| "grad_norm": 2.778947353363037, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.1233, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.03717377466581795, | |
| "grad_norm": 1.5811668634414673, | |
| "learning_rate": 9.900412805461967e-05, | |
| "loss": 0.1392, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.037683004455760664, | |
| "grad_norm": 136.75079345703125, | |
| "learning_rate": 9.897237175829926e-05, | |
| "loss": 4.0886, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.03819223424570337, | |
| "grad_norm": 242.6491241455078, | |
| "learning_rate": 9.894012231073894e-05, | |
| "loss": 6.7782, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03870146403564609, | |
| "grad_norm": 100.27400207519531, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.7429, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.0392106938255888, | |
| "grad_norm": 49.41951370239258, | |
| "learning_rate": 9.887414526586763e-05, | |
| "loss": 0.6478, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.03971992361553151, | |
| "grad_norm": 15.152222633361816, | |
| "learning_rate": 9.884041833294476e-05, | |
| "loss": 0.3272, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.04022915340547422, | |
| "grad_norm": 11.572675704956055, | |
| "learning_rate": 9.880619957755151e-05, | |
| "loss": 0.2417, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.04073838319541693, | |
| "grad_norm": 6.667335033416748, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.2272, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.041247612985359644, | |
| "grad_norm": 11.849577903747559, | |
| "learning_rate": 9.873628798263296e-05, | |
| "loss": 0.2427, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.04175684277530235, | |
| "grad_norm": 9.412842750549316, | |
| "learning_rate": 9.870059584711668e-05, | |
| "loss": 0.2196, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.04226607256524507, | |
| "grad_norm": 9.481209754943848, | |
| "learning_rate": 9.866441329714088e-05, | |
| "loss": 0.1711, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.042775302355187776, | |
| "grad_norm": 8.62416934967041, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.2399, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.04328453214513049, | |
| "grad_norm": 13.821141242980957, | |
| "learning_rate": 9.859057841617709e-05, | |
| "loss": 0.1519, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0437937619350732, | |
| "grad_norm": 9.316679000854492, | |
| "learning_rate": 9.855292682870551e-05, | |
| "loss": 0.2179, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.044302991725015915, | |
| "grad_norm": 20.34062385559082, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.1978, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.044812221514958624, | |
| "grad_norm": 14.848787307739258, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.203, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.04532145130490134, | |
| "grad_norm": 9.669304847717285, | |
| "learning_rate": 9.843704004290392e-05, | |
| "loss": 0.1324, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.04583068109484405, | |
| "grad_norm": 5.988533020019531, | |
| "learning_rate": 9.839743506981782e-05, | |
| "loss": 0.0655, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.04633991088478676, | |
| "grad_norm": 7.444858074188232, | |
| "learning_rate": 9.835734273509786e-05, | |
| "loss": 0.0616, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.04684914067472947, | |
| "grad_norm": 19.514366149902344, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.0411, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.047358370464672186, | |
| "grad_norm": 6.495416164398193, | |
| "learning_rate": 9.827569760057755e-05, | |
| "loss": 0.0808, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.047867600254614895, | |
| "grad_norm": 9.067540168762207, | |
| "learning_rate": 9.82341456229428e-05, | |
| "loss": 0.1453, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0483768300445576, | |
| "grad_norm": 4.768366813659668, | |
| "learning_rate": 9.819210792799712e-05, | |
| "loss": 0.0876, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.04888605983450032, | |
| "grad_norm": 10.522812843322754, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.0634, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.04939528962444303, | |
| "grad_norm": 35.72486877441406, | |
| "learning_rate": 9.810657708433637e-05, | |
| "loss": 0.1568, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.04990451941438574, | |
| "grad_norm": 58.850852966308594, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0519, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.05041374920432845, | |
| "grad_norm": 10.575145721435547, | |
| "learning_rate": 9.801910851476523e-05, | |
| "loss": 0.1669, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.050922978994271166, | |
| "grad_norm": 13.279051780700684, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.1777, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.051432208784213874, | |
| "grad_norm": 9.112642288208008, | |
| "learning_rate": 9.792970574250493e-05, | |
| "loss": 0.7836, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.05194143857415659, | |
| "grad_norm": 10.229832649230957, | |
| "learning_rate": 9.788428015268027e-05, | |
| "loss": 1.2149, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.0524506683640993, | |
| "grad_norm": 12.976715087890625, | |
| "learning_rate": 9.783837236868609e-05, | |
| "loss": 1.1371, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.052959898154042014, | |
| "grad_norm": 6.877295970916748, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.2919, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.05346912794398472, | |
| "grad_norm": 12.871697425842285, | |
| "learning_rate": 9.77451120722037e-05, | |
| "loss": 1.0255, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.05397835773392744, | |
| "grad_norm": 19.823925018310547, | |
| "learning_rate": 9.769776049884563e-05, | |
| "loss": 1.4255, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.054487587523870146, | |
| "grad_norm": 14.80654525756836, | |
| "learning_rate": 9.764992860956889e-05, | |
| "loss": 0.8401, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.05499681731381286, | |
| "grad_norm": 24.64175033569336, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 1.0325, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.05550604710375557, | |
| "grad_norm": 26.200424194335938, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.4366, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.056015276893698285, | |
| "grad_norm": 17.368080139160156, | |
| "learning_rate": 9.750355588704727e-05, | |
| "loss": 1.0868, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05652450668364099, | |
| "grad_norm": 11.78357219696045, | |
| "learning_rate": 9.745380759905647e-05, | |
| "loss": 0.4192, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.0570337364735837, | |
| "grad_norm": 13.371063232421875, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.9413, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.05754296626352642, | |
| "grad_norm": 15.481206893920898, | |
| "learning_rate": 9.735287795090455e-05, | |
| "loss": 1.2106, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.058052196053469125, | |
| "grad_norm": 7.285238265991211, | |
| "learning_rate": 9.730169760710386e-05, | |
| "loss": 0.1473, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.05856142584341184, | |
| "grad_norm": 13.799015045166016, | |
| "learning_rate": 9.725004093573342e-05, | |
| "loss": 0.1718, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.05907065563335455, | |
| "grad_norm": 9.683958053588867, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.1202, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.059579885423297264, | |
| "grad_norm": 2.17852520942688, | |
| "learning_rate": 9.714530069580309e-05, | |
| "loss": 0.0936, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.06008911521323997, | |
| "grad_norm": 3.735697031021118, | |
| "learning_rate": 9.709221818197624e-05, | |
| "loss": 0.1458, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.06059834500318269, | |
| "grad_norm": 6.55530309677124, | |
| "learning_rate": 9.703866145003511e-05, | |
| "loss": 0.189, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.061107574793125397, | |
| "grad_norm": 5.885622978210449, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.1048, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06161680458306811, | |
| "grad_norm": 7.249486446380615, | |
| "learning_rate": 9.693012749384279e-05, | |
| "loss": 0.112, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.06212603437301082, | |
| "grad_norm": 7.16094970703125, | |
| "learning_rate": 9.687515136252731e-05, | |
| "loss": 0.1314, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.06263526416295354, | |
| "grad_norm": 13.14775276184082, | |
| "learning_rate": 9.681970319895803e-05, | |
| "loss": 0.15, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.06314449395289624, | |
| "grad_norm": 1.7535552978515625, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.1078, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.06365372374283895, | |
| "grad_norm": 157.29830932617188, | |
| "learning_rate": 9.670739301325534e-05, | |
| "loss": 1.5483, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06416295353278166, | |
| "grad_norm": 396.1761474609375, | |
| "learning_rate": 9.665053212208426e-05, | |
| "loss": 3.7025, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.06467218332272438, | |
| "grad_norm": 119.90931701660156, | |
| "learning_rate": 9.659320146057262e-05, | |
| "loss": 0.9201, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.06518141311266709, | |
| "grad_norm": 37.91402053833008, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.5631, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.0656906429026098, | |
| "grad_norm": 12.050881385803223, | |
| "learning_rate": 9.647713314052896e-05, | |
| "loss": 0.3231, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.06619987269255251, | |
| "grad_norm": 8.638886451721191, | |
| "learning_rate": 9.641839665080363e-05, | |
| "loss": 0.1641, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06670910248249523, | |
| "grad_norm": 6.96414041519165, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.1458, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.06721833227243794, | |
| "grad_norm": 9.606210708618164, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.1746, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.06772756206238065, | |
| "grad_norm": 10.889528274536133, | |
| "learning_rate": 9.623938497462646e-05, | |
| "loss": 0.1276, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.06823679185232336, | |
| "grad_norm": 10.934017181396484, | |
| "learning_rate": 9.617878234984055e-05, | |
| "loss": 0.2305, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.06874602164226608, | |
| "grad_norm": 10.173789978027344, | |
| "learning_rate": 9.611771470522908e-05, | |
| "loss": 0.2105, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.06925525143220879, | |
| "grad_norm": 8.497604370117188, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.1653, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.0697644812221515, | |
| "grad_norm": 11.240355491638184, | |
| "learning_rate": 9.599418682100793e-05, | |
| "loss": 0.1518, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.0702737110120942, | |
| "grad_norm": 23.68250274658203, | |
| "learning_rate": 9.593172782532268e-05, | |
| "loss": 0.1257, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.07078294080203693, | |
| "grad_norm": 12.878501892089844, | |
| "learning_rate": 9.586880629764817e-05, | |
| "loss": 0.0837, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.07129217059197963, | |
| "grad_norm": 7.515162467956543, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0495, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07180140038192234, | |
| "grad_norm": 8.767376899719238, | |
| "learning_rate": 9.574157818545901e-05, | |
| "loss": 0.0835, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.07231063017186505, | |
| "grad_norm": 13.098119735717773, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.0426, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.07281985996180776, | |
| "grad_norm": 5.534914970397949, | |
| "learning_rate": 9.561250760917027e-05, | |
| "loss": 0.0358, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.07332908975175048, | |
| "grad_norm": 6.893864154815674, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.0407, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.07383831954169319, | |
| "grad_norm": 9.333673477172852, | |
| "learning_rate": 9.548159976772592e-05, | |
| "loss": 0.0433, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.0743475493316359, | |
| "grad_norm": 10.31505012512207, | |
| "learning_rate": 9.541545851748186e-05, | |
| "loss": 0.0814, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.0748567791215786, | |
| "grad_norm": 4.145845413208008, | |
| "learning_rate": 9.534885993407474e-05, | |
| "loss": 0.0354, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.07536600891152133, | |
| "grad_norm": 7.609471321105957, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.0806, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.07587523870146404, | |
| "grad_norm": 6.148724555969238, | |
| "learning_rate": 9.521429345495787e-05, | |
| "loss": 0.0316, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.07638446849140675, | |
| "grad_norm": 14.903536796569824, | |
| "learning_rate": 9.514632691433107e-05, | |
| "loss": 0.1945, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07689369828134945, | |
| "grad_norm": 8.966485977172852, | |
| "learning_rate": 9.507790575069347e-05, | |
| "loss": 0.9285, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.07740292807129218, | |
| "grad_norm": 10.010855674743652, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.9562, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.07791215786123488, | |
| "grad_norm": 15.471893310546875, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.5357, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.0784213876511776, | |
| "grad_norm": 9.118985176086426, | |
| "learning_rate": 9.486992143456792e-05, | |
| "loss": 0.5833, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.0789306174411203, | |
| "grad_norm": 25.249879837036133, | |
| "learning_rate": 9.479968871456679e-05, | |
| "loss": 1.2881, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.07943984723106302, | |
| "grad_norm": 29.803438186645508, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 1.3695, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.07994907702100573, | |
| "grad_norm": 23.109800338745117, | |
| "learning_rate": 9.46578705892462e-05, | |
| "loss": 0.8352, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.08045830681094844, | |
| "grad_norm": 19.88869285583496, | |
| "learning_rate": 9.458628661203367e-05, | |
| "loss": 1.0865, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.08096753660089115, | |
| "grad_norm": 27.150493621826172, | |
| "learning_rate": 9.451425365140996e-05, | |
| "loss": 0.9212, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.08147676639083386, | |
| "grad_norm": 21.8201847076416, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 1.2252, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08198599618077658, | |
| "grad_norm": 14.624476432800293, | |
| "learning_rate": 9.43688436859274e-05, | |
| "loss": 0.8962, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.08249522597071929, | |
| "grad_norm": 21.079187393188477, | |
| "learning_rate": 9.429546814534529e-05, | |
| "loss": 0.8082, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.083004455760662, | |
| "grad_norm": 25.841928482055664, | |
| "learning_rate": 9.422164654989072e-05, | |
| "loss": 0.8788, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.0835136855506047, | |
| "grad_norm": 6.6645941734313965, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.2285, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.08402291534054743, | |
| "grad_norm": 9.64619255065918, | |
| "learning_rate": 9.407266817237911e-05, | |
| "loss": 0.1814, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.08453214513049014, | |
| "grad_norm": 2.2856459617614746, | |
| "learning_rate": 9.399751289053267e-05, | |
| "loss": 0.0883, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.08504137492043284, | |
| "grad_norm": 5.339344024658203, | |
| "learning_rate": 9.392191455421988e-05, | |
| "loss": 0.1506, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.08555060471037555, | |
| "grad_norm": 3.965207099914551, | |
| "learning_rate": 9.384587392471515e-05, | |
| "loss": 0.1156, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.08605983450031827, | |
| "grad_norm": 2.786485195159912, | |
| "learning_rate": 9.376939176774679e-05, | |
| "loss": 0.095, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.08656906429026098, | |
| "grad_norm": 2.5898613929748535, | |
| "learning_rate": 9.369246885348926e-05, | |
| "loss": 0.1513, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08707829408020369, | |
| "grad_norm": 3.1944878101348877, | |
| "learning_rate": 9.361510595655545e-05, | |
| "loss": 0.1361, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.0875875238701464, | |
| "grad_norm": 2.0986759662628174, | |
| "learning_rate": 9.353730385598887e-05, | |
| "loss": 0.112, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.08809675366008912, | |
| "grad_norm": 3.246211528778076, | |
| "learning_rate": 9.345906333525581e-05, | |
| "loss": 0.0693, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.08860598345003183, | |
| "grad_norm": 2.6745364665985107, | |
| "learning_rate": 9.338038518223747e-05, | |
| "loss": 0.1327, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.08911521323997454, | |
| "grad_norm": 97.89273071289062, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.9497, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.08962444302991725, | |
| "grad_norm": 20.47549819946289, | |
| "learning_rate": 9.322171915289635e-05, | |
| "loss": 0.3825, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.09013367281985996, | |
| "grad_norm": 10.546618461608887, | |
| "learning_rate": 9.314173287433873e-05, | |
| "loss": 0.3422, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.09064290260980268, | |
| "grad_norm": 8.488680839538574, | |
| "learning_rate": 9.306131215901003e-05, | |
| "loss": 0.4135, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.09115213239974539, | |
| "grad_norm": 9.975773811340332, | |
| "learning_rate": 9.298045781674596e-05, | |
| "loss": 0.2914, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.0916613621896881, | |
| "grad_norm": 9.072954177856445, | |
| "learning_rate": 9.289917066174886e-05, | |
| "loss": 0.2371, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.0921705919796308, | |
| "grad_norm": 7.301230430603027, | |
| "learning_rate": 9.281745151257946e-05, | |
| "loss": 0.1286, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.09267982176957353, | |
| "grad_norm": 5.187671184539795, | |
| "learning_rate": 9.273530119214868e-05, | |
| "loss": 0.1105, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.09318905155951623, | |
| "grad_norm": 10.456426620483398, | |
| "learning_rate": 9.265272052770936e-05, | |
| "loss": 0.1123, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.09369828134945894, | |
| "grad_norm": 7.074233531951904, | |
| "learning_rate": 9.256971035084785e-05, | |
| "loss": 0.174, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.09420751113940165, | |
| "grad_norm": 7.056065082550049, | |
| "learning_rate": 9.248627149747573e-05, | |
| "loss": 0.209, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.09471674092934437, | |
| "grad_norm": 7.530999183654785, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.1251, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.09522597071928708, | |
| "grad_norm": 11.38913631439209, | |
| "learning_rate": 9.231811112642121e-05, | |
| "loss": 0.2015, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.09573520050922979, | |
| "grad_norm": 6.872330188751221, | |
| "learning_rate": 9.223339130211192e-05, | |
| "loss": 0.1438, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.0962444302991725, | |
| "grad_norm": 10.519857406616211, | |
| "learning_rate": 9.214824618802109e-05, | |
| "loss": 0.1032, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.0967536600891152, | |
| "grad_norm": 8.731643676757812, | |
| "learning_rate": 9.206267664155907e-05, | |
| "loss": 0.0608, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09726288987905793, | |
| "grad_norm": 4.590799808502197, | |
| "learning_rate": 9.197668352441025e-05, | |
| "loss": 0.0585, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.09777211966900064, | |
| "grad_norm": 4.542694568634033, | |
| "learning_rate": 9.189026770252436e-05, | |
| "loss": 0.038, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.09828134945894335, | |
| "grad_norm": 22.831832885742188, | |
| "learning_rate": 9.18034300461078e-05, | |
| "loss": 0.0438, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.09879057924888605, | |
| "grad_norm": 21.84329605102539, | |
| "learning_rate": 9.171617142961477e-05, | |
| "loss": 0.1501, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.09929980903882878, | |
| "grad_norm": 8.787395477294922, | |
| "learning_rate": 9.162849273173857e-05, | |
| "loss": 0.0465, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.09980903882877148, | |
| "grad_norm": 3.925507068634033, | |
| "learning_rate": 9.154039483540273e-05, | |
| "loss": 0.0239, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1003182686187142, | |
| "grad_norm": 4.768054008483887, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.0463, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.1008274984086569, | |
| "grad_norm": 10.094438552856445, | |
| "learning_rate": 9.136294500014386e-05, | |
| "loss": 0.024, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.10133672819859962, | |
| "grad_norm": 5.6756463050842285, | |
| "learning_rate": 9.12735948481387e-05, | |
| "loss": 0.0279, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.10184595798854233, | |
| "grad_norm": 18.045185089111328, | |
| "learning_rate": 9.118382907149165e-05, | |
| "loss": 0.1826, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.10184595798854233, | |
| "eval_loss": 0.6390340328216553, | |
| "eval_runtime": 378.2914, | |
| "eval_samples_per_second": 8.745, | |
| "eval_steps_per_second": 2.186, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.10235518777848504, | |
| "grad_norm": 5.750209808349609, | |
| "learning_rate": 9.109364857414306e-05, | |
| "loss": 1.0072, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.10286441756842775, | |
| "grad_norm": 6.447965145111084, | |
| "learning_rate": 9.100305426420956e-05, | |
| "loss": 0.903, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.10337364735837047, | |
| "grad_norm": 9.898489952087402, | |
| "learning_rate": 9.091204705397484e-05, | |
| "loss": 1.0206, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.10388287714831318, | |
| "grad_norm": 6.285497188568115, | |
| "learning_rate": 9.082062785988049e-05, | |
| "loss": 0.3841, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.10439210693825589, | |
| "grad_norm": 17.02065658569336, | |
| "learning_rate": 9.072879760251679e-05, | |
| "loss": 1.5343, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1049013367281986, | |
| "grad_norm": 17.704208374023438, | |
| "learning_rate": 9.06365572066134e-05, | |
| "loss": 0.9827, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.1054105665181413, | |
| "grad_norm": 10.860851287841797, | |
| "learning_rate": 9.05439076010301e-05, | |
| "loss": 0.3846, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.10591979630808403, | |
| "grad_norm": 16.583065032958984, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.6791, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.10642902609802674, | |
| "grad_norm": 15.18968677520752, | |
| "learning_rate": 9.035738449685707e-05, | |
| "loss": 1.09, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.10693825588796944, | |
| "grad_norm": 16.335735321044922, | |
| "learning_rate": 9.026351287655294e-05, | |
| "loss": 1.1956, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.10744748567791215, | |
| "grad_norm": 17.723297119140625, | |
| "learning_rate": 9.016923580312113e-05, | |
| "loss": 0.4732, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.10795671546785487, | |
| "grad_norm": 20.851490020751953, | |
| "learning_rate": 9.007455422593077e-05, | |
| "loss": 0.821, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.10846594525779758, | |
| "grad_norm": 15.048613548278809, | |
| "learning_rate": 8.997946909842425e-05, | |
| "loss": 0.9223, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.10897517504774029, | |
| "grad_norm": 2.7917280197143555, | |
| "learning_rate": 8.988398137810777e-05, | |
| "loss": 0.1104, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.109484404837683, | |
| "grad_norm": 3.5661802291870117, | |
| "learning_rate": 8.978809202654162e-05, | |
| "loss": 0.1082, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.10999363462762572, | |
| "grad_norm": 4.870175838470459, | |
| "learning_rate": 8.969180200933047e-05, | |
| "loss": 0.1652, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.11050286441756843, | |
| "grad_norm": 2.00311541557312, | |
| "learning_rate": 8.959511229611376e-05, | |
| "loss": 0.1195, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.11101209420751114, | |
| "grad_norm": 3.949178695678711, | |
| "learning_rate": 8.949802386055581e-05, | |
| "loss": 0.1085, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.11152132399745385, | |
| "grad_norm": 1.9956891536712646, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.0933, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.11203055378739657, | |
| "grad_norm": 2.3544695377349854, | |
| "learning_rate": 8.930265473713938e-05, | |
| "loss": 0.121, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.11253978357733928, | |
| "grad_norm": 1.6948124170303345, | |
| "learning_rate": 8.92043760166458e-05, | |
| "loss": 0.0807, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.11304901336728199, | |
| "grad_norm": 3.766237735748291, | |
| "learning_rate": 8.910570250852097e-05, | |
| "loss": 0.2488, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.1135582431572247, | |
| "grad_norm": 1.9978468418121338, | |
| "learning_rate": 8.900663520640604e-05, | |
| "loss": 0.0865, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.1140674729471674, | |
| "grad_norm": 3.3670530319213867, | |
| "learning_rate": 8.890717510790763e-05, | |
| "loss": 0.1258, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.11457670273711013, | |
| "grad_norm": 135.1156463623047, | |
| "learning_rate": 8.880732321458784e-05, | |
| "loss": 1.1012, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.11508593252705283, | |
| "grad_norm": 78.30040740966797, | |
| "learning_rate": 8.870708053195413e-05, | |
| "loss": 0.6246, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.11559516231699554, | |
| "grad_norm": 9.838759422302246, | |
| "learning_rate": 8.860644806944918e-05, | |
| "loss": 0.1683, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.11610439210693825, | |
| "grad_norm": 11.655884742736816, | |
| "learning_rate": 8.850542684044078e-05, | |
| "loss": 0.191, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.11661362189688097, | |
| "grad_norm": 5.8414130210876465, | |
| "learning_rate": 8.840401786221159e-05, | |
| "loss": 0.2464, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.11712285168682368, | |
| "grad_norm": 7.483433723449707, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.2158, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11763208147676639, | |
| "grad_norm": 6.090114116668701, | |
| "learning_rate": 8.820004074673433e-05, | |
| "loss": 0.199, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.1181413112667091, | |
| "grad_norm": 7.228814125061035, | |
| "learning_rate": 8.809747466353356e-05, | |
| "loss": 0.1263, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.11865054105665182, | |
| "grad_norm": 7.769136428833008, | |
| "learning_rate": 8.799452493918585e-05, | |
| "loss": 0.1385, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.11915977084659453, | |
| "grad_norm": 7.88167667388916, | |
| "learning_rate": 8.789119261039385e-05, | |
| "loss": 0.1602, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.11966900063653724, | |
| "grad_norm": 5.976058483123779, | |
| "learning_rate": 8.778747871771292e-05, | |
| "loss": 0.1044, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.12017823042647995, | |
| "grad_norm": 7.186121463775635, | |
| "learning_rate": 8.768338430554082e-05, | |
| "loss": 0.1026, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.12068746021642265, | |
| "grad_norm": 7.408992767333984, | |
| "learning_rate": 8.757891042210714e-05, | |
| "loss": 0.1399, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.12119669000636538, | |
| "grad_norm": 9.789539337158203, | |
| "learning_rate": 8.74740581194627e-05, | |
| "loss": 0.082, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.12170591979630808, | |
| "grad_norm": 7.328338146209717, | |
| "learning_rate": 8.736882845346906e-05, | |
| "loss": 0.0742, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.12221514958625079, | |
| "grad_norm": 12.503471374511719, | |
| "learning_rate": 8.726322248378775e-05, | |
| "loss": 0.0757, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.1227243793761935, | |
| "grad_norm": 5.140643119812012, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.0174, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.12323360916613622, | |
| "grad_norm": 4.965789318084717, | |
| "learning_rate": 8.705088589094459e-05, | |
| "loss": 0.0311, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.12374283895607893, | |
| "grad_norm": 22.65897560119629, | |
| "learning_rate": 8.694415740600988e-05, | |
| "loss": 0.0395, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.12425206874602164, | |
| "grad_norm": 5.265443325042725, | |
| "learning_rate": 8.683705689382024e-05, | |
| "loss": 0.0543, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.12476129853596435, | |
| "grad_norm": 6.697726249694824, | |
| "learning_rate": 8.672958543287666e-05, | |
| "loss": 0.0549, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.12527052832590707, | |
| "grad_norm": 9.752626419067383, | |
| "learning_rate": 8.662174410541555e-05, | |
| "loss": 0.0395, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.12577975811584977, | |
| "grad_norm": 7.68189001083374, | |
| "learning_rate": 8.651353399739787e-05, | |
| "loss": 0.0678, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.1262889879057925, | |
| "grad_norm": 11.428817749023438, | |
| "learning_rate": 8.640495619849821e-05, | |
| "loss": 0.1091, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.1267982176957352, | |
| "grad_norm": 18.397584915161133, | |
| "learning_rate": 8.629601180209381e-05, | |
| "loss": 0.128, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.1273074474856779, | |
| "grad_norm": 7.687522888183594, | |
| "learning_rate": 8.618670190525352e-05, | |
| "loss": 0.0705, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.12781667727562063, | |
| "grad_norm": 6.8483805656433105, | |
| "learning_rate": 8.607702760872678e-05, | |
| "loss": 0.8196, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.12832590706556332, | |
| "grad_norm": 8.953585624694824, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.9635, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.12883513685550604, | |
| "grad_norm": 7.499317646026611, | |
| "learning_rate": 8.585659023794818e-05, | |
| "loss": 0.5593, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.12934436664544877, | |
| "grad_norm": 10.518817901611328, | |
| "learning_rate": 8.574582938349817e-05, | |
| "loss": 0.5622, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.12985359643539146, | |
| "grad_norm": 18.279752731323242, | |
| "learning_rate": 8.563470856894316e-05, | |
| "loss": 1.0561, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.13036282622533418, | |
| "grad_norm": 13.749988555908203, | |
| "learning_rate": 8.552322891326846e-05, | |
| "loss": 0.6541, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.1308720560152769, | |
| "grad_norm": 16.35479164123535, | |
| "learning_rate": 8.541139153907296e-05, | |
| "loss": 0.8946, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.1313812858052196, | |
| "grad_norm": 15.60219955444336, | |
| "learning_rate": 8.529919757255783e-05, | |
| "loss": 1.2002, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.13189051559516232, | |
| "grad_norm": 17.93536949157715, | |
| "learning_rate": 8.518664814351502e-05, | |
| "loss": 0.7515, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.13239974538510502, | |
| "grad_norm": 10.655067443847656, | |
| "learning_rate": 8.507374438531607e-05, | |
| "loss": 0.7691, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.13290897517504774, | |
| "grad_norm": 17.118453979492188, | |
| "learning_rate": 8.496048743490053e-05, | |
| "loss": 0.8141, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.13341820496499046, | |
| "grad_norm": 6.289425373077393, | |
| "learning_rate": 8.484687843276469e-05, | |
| "loss": 0.334, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.13392743475493316, | |
| "grad_norm": 7.217257976531982, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 0.2386, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.13443666454487588, | |
| "grad_norm": 3.586156129837036, | |
| "learning_rate": 8.461860885303114e-05, | |
| "loss": 0.1299, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.1349458943348186, | |
| "grad_norm": 3.3640799522399902, | |
| "learning_rate": 8.450395057410561e-05, | |
| "loss": 0.0798, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.1354551241247613, | |
| "grad_norm": 4.526169300079346, | |
| "learning_rate": 8.438894484078086e-05, | |
| "loss": 0.1487, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.13596435391470402, | |
| "grad_norm": 2.676795244216919, | |
| "learning_rate": 8.427359281116334e-05, | |
| "loss": 0.109, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.1364735837046467, | |
| "grad_norm": 0.9461959004402161, | |
| "learning_rate": 8.415789564684673e-05, | |
| "loss": 0.1007, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.13698281349458943, | |
| "grad_norm": 2.397101879119873, | |
| "learning_rate": 8.404185451290018e-05, | |
| "loss": 0.1238, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.13749204328453216, | |
| "grad_norm": 6.66584587097168, | |
| "learning_rate": 8.392547057785661e-05, | |
| "loss": 0.1607, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.13800127307447485, | |
| "grad_norm": 1.6675441265106201, | |
| "learning_rate": 8.380874501370097e-05, | |
| "loss": 0.0728, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.13851050286441757, | |
| "grad_norm": 1.8109138011932373, | |
| "learning_rate": 8.369167899585841e-05, | |
| "loss": 0.0844, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.13901973265436027, | |
| "grad_norm": 9.230074882507324, | |
| "learning_rate": 8.357427370318239e-05, | |
| "loss": 0.2044, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.139528962444303, | |
| "grad_norm": 10.465958595275879, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.2347, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.1400381922342457, | |
| "grad_norm": 14.745471000671387, | |
| "learning_rate": 8.333845002581458e-05, | |
| "loss": 0.254, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.1405474220241884, | |
| "grad_norm": 6.833587646484375, | |
| "learning_rate": 8.322003401586462e-05, | |
| "loss": 0.1399, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.14105665181413113, | |
| "grad_norm": 7.580754280090332, | |
| "learning_rate": 8.310128348054094e-05, | |
| "loss": 0.164, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.14156588160407385, | |
| "grad_norm": 10.550285339355469, | |
| "learning_rate": 8.298219961566009e-05, | |
| "loss": 0.132, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.14207511139401655, | |
| "grad_norm": 4.947827339172363, | |
| "learning_rate": 8.286278362039528e-05, | |
| "loss": 0.0824, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.14258434118395927, | |
| "grad_norm": 9.395261764526367, | |
| "learning_rate": 8.274303669726426e-05, | |
| "loss": 0.1049, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.14309357097390196, | |
| "grad_norm": 6.372900009155273, | |
| "learning_rate": 8.262296005211721e-05, | |
| "loss": 0.119, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.14360280076384468, | |
| "grad_norm": 10.334807395935059, | |
| "learning_rate": 8.250255489412463e-05, | |
| "loss": 0.1494, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.1441120305537874, | |
| "grad_norm": 7.954561710357666, | |
| "learning_rate": 8.238182243576512e-05, | |
| "loss": 0.1484, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.1446212603437301, | |
| "grad_norm": 6.839869022369385, | |
| "learning_rate": 8.226076389281316e-05, | |
| "loss": 0.1011, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.14513049013367282, | |
| "grad_norm": 8.359034538269043, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.1297, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.14563971992361552, | |
| "grad_norm": 5.433843612670898, | |
| "learning_rate": 8.201767343263612e-05, | |
| "loss": 0.1206, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.14614894971355824, | |
| "grad_norm": 8.870092391967773, | |
| "learning_rate": 8.189564396332928e-05, | |
| "loss": 0.0581, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.14665817950350096, | |
| "grad_norm": 10.579423904418945, | |
| "learning_rate": 8.177329330524182e-05, | |
| "loss": 0.0415, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.14716740929344366, | |
| "grad_norm": 4.38845157623291, | |
| "learning_rate": 8.165062269044353e-05, | |
| "loss": 0.0424, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.14767663908338638, | |
| "grad_norm": 3.134714126586914, | |
| "learning_rate": 8.152763335422613e-05, | |
| "loss": 0.0138, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1481858688733291, | |
| "grad_norm": 7.17431116104126, | |
| "learning_rate": 8.140432653509089e-05, | |
| "loss": 0.1052, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.1486950986632718, | |
| "grad_norm": 9.987316131591797, | |
| "learning_rate": 8.128070347473609e-05, | |
| "loss": 0.1305, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.14920432845321452, | |
| "grad_norm": 14.93751335144043, | |
| "learning_rate": 8.115676541804456e-05, | |
| "loss": 0.0562, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.1497135582431572, | |
| "grad_norm": 11.657471656799316, | |
| "learning_rate": 8.103251361307119e-05, | |
| "loss": 0.1214, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.15022278803309994, | |
| "grad_norm": 0.8434633612632751, | |
| "learning_rate": 8.090794931103026e-05, | |
| "loss": 0.0029, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.15073201782304266, | |
| "grad_norm": 3.910121440887451, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.0248, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.15124124761298535, | |
| "grad_norm": 4.701965808868408, | |
| "learning_rate": 8.065788823632451e-05, | |
| "loss": 0.0083, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.15175047740292807, | |
| "grad_norm": 5.969517707824707, | |
| "learning_rate": 8.053239398177191e-05, | |
| "loss": 0.0785, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.1522597071928708, | |
| "grad_norm": 9.37517261505127, | |
| "learning_rate": 8.04065922663509e-05, | |
| "loss": 0.0725, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.1527689369828135, | |
| "grad_norm": 22.21939468383789, | |
| "learning_rate": 8.028048435688333e-05, | |
| "loss": 0.0937, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1532781667727562, | |
| "grad_norm": 15.828902244567871, | |
| "learning_rate": 8.015407152327448e-05, | |
| "loss": 0.6293, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.1537873965626989, | |
| "grad_norm": 7.590906620025635, | |
| "learning_rate": 8.002735503850016e-05, | |
| "loss": 0.6786, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.15429662635264163, | |
| "grad_norm": 8.043692588806152, | |
| "learning_rate": 7.990033617859396e-05, | |
| "loss": 0.731, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.15480585614258435, | |
| "grad_norm": 9.872035026550293, | |
| "learning_rate": 7.97730162226344e-05, | |
| "loss": 0.2062, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.15531508593252705, | |
| "grad_norm": 18.133535385131836, | |
| "learning_rate": 7.964539645273204e-05, | |
| "loss": 1.1256, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.15582431572246977, | |
| "grad_norm": 17.49220848083496, | |
| "learning_rate": 7.95174781540165e-05, | |
| "loss": 1.0813, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.15633354551241246, | |
| "grad_norm": 19.740583419799805, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.5487, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.1568427753023552, | |
| "grad_norm": 19.193031311035156, | |
| "learning_rate": 7.926075112568259e-05, | |
| "loss": 1.0428, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.1573520050922979, | |
| "grad_norm": 14.3748197555542, | |
| "learning_rate": 7.913194498130252e-05, | |
| "loss": 0.901, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.1578612348822406, | |
| "grad_norm": 18.86087989807129, | |
| "learning_rate": 7.900284547855991e-05, | |
| "loss": 0.7603, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.15837046467218333, | |
| "grad_norm": 15.366621971130371, | |
| "learning_rate": 7.887345391748533e-05, | |
| "loss": 0.2624, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.15887969446212605, | |
| "grad_norm": 19.380908966064453, | |
| "learning_rate": 7.874377160105036e-05, | |
| "loss": 1.0304, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.15938892425206874, | |
| "grad_norm": 13.959190368652344, | |
| "learning_rate": 7.861379983515449e-05, | |
| "loss": 0.7812, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.15989815404201146, | |
| "grad_norm": 3.1605825424194336, | |
| "learning_rate": 7.848353992861195e-05, | |
| "loss": 0.0522, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.16040738383195416, | |
| "grad_norm": 3.092629909515381, | |
| "learning_rate": 7.835299319313853e-05, | |
| "loss": 0.1617, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.16091661362189688, | |
| "grad_norm": 3.086357355117798, | |
| "learning_rate": 7.822216094333847e-05, | |
| "loss": 0.2058, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.1614258434118396, | |
| "grad_norm": 3.3937745094299316, | |
| "learning_rate": 7.809104449669101e-05, | |
| "loss": 0.1046, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.1619350732017823, | |
| "grad_norm": 1.3770623207092285, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.1156, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.16244430299172502, | |
| "grad_norm": 1.293850302696228, | |
| "learning_rate": 7.78279642970672e-05, | |
| "loss": 0.0919, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.16295353278166771, | |
| "grad_norm": 4.897433757781982, | |
| "learning_rate": 7.769600319330552e-05, | |
| "loss": 0.1203, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.16346276257161044, | |
| "grad_norm": 0.744999885559082, | |
| "learning_rate": 7.756376319109917e-05, | |
| "loss": 0.0906, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.16397199236155316, | |
| "grad_norm": 2.635500192642212, | |
| "learning_rate": 7.74312456221035e-05, | |
| "loss": 0.1324, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.16448122215149585, | |
| "grad_norm": 3.7716286182403564, | |
| "learning_rate": 7.729845182076895e-05, | |
| "loss": 0.09, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.16499045194143858, | |
| "grad_norm": 1.3198482990264893, | |
| "learning_rate": 7.716538312432766e-05, | |
| "loss": 0.0999, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.1654996817313813, | |
| "grad_norm": 6.107431888580322, | |
| "learning_rate": 7.703204087277988e-05, | |
| "loss": 0.0661, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.166008911521324, | |
| "grad_norm": 15.861464500427246, | |
| "learning_rate": 7.689842640888063e-05, | |
| "loss": 0.1399, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.16651814131126672, | |
| "grad_norm": 8.088520050048828, | |
| "learning_rate": 7.676454107812607e-05, | |
| "loss": 0.2258, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.1670273711012094, | |
| "grad_norm": 10.10229206085205, | |
| "learning_rate": 7.663038622873999e-05, | |
| "loss": 0.206, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.16753660089115213, | |
| "grad_norm": 9.973199844360352, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 0.1449, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.16804583068109485, | |
| "grad_norm": 5.558497905731201, | |
| "learning_rate": 7.636127338052512e-05, | |
| "loss": 0.084, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.16855506047103755, | |
| "grad_norm": 6.0195722579956055, | |
| "learning_rate": 7.622631809165973e-05, | |
| "loss": 0.2007, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.16906429026098027, | |
| "grad_norm": 9.740483283996582, | |
| "learning_rate": 7.60910987040623e-05, | |
| "loss": 0.1116, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.16957352005092297, | |
| "grad_norm": 4.960144519805908, | |
| "learning_rate": 7.595561657939061e-05, | |
| "loss": 0.0648, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.1700827498408657, | |
| "grad_norm": 8.019902229309082, | |
| "learning_rate": 7.58198730819481e-05, | |
| "loss": 0.1621, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.1705919796308084, | |
| "grad_norm": 5.274026870727539, | |
| "learning_rate": 7.568386957867033e-05, | |
| "loss": 0.1088, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.1711012094207511, | |
| "grad_norm": 8.776835441589355, | |
| "learning_rate": 7.554760743911103e-05, | |
| "loss": 0.071, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.17161043921069383, | |
| "grad_norm": 4.958322525024414, | |
| "learning_rate": 7.541108803542846e-05, | |
| "loss": 0.1097, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.17211966900063655, | |
| "grad_norm": 8.615314483642578, | |
| "learning_rate": 7.52743127423715e-05, | |
| "loss": 0.114, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.17262889879057924, | |
| "grad_norm": 8.077454566955566, | |
| "learning_rate": 7.51372829372658e-05, | |
| "loss": 0.0687, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.17313812858052197, | |
| "grad_norm": 9.958930969238281, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0614, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.17364735837046466, | |
| "grad_norm": 7.942490100860596, | |
| "learning_rate": 7.486246531301177e-05, | |
| "loss": 0.0948, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.17415658816040738, | |
| "grad_norm": 3.127333164215088, | |
| "learning_rate": 7.472468026127385e-05, | |
| "loss": 0.0392, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.1746658179503501, | |
| "grad_norm": 8.586334228515625, | |
| "learning_rate": 7.45866462322802e-05, | |
| "loss": 0.1213, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.1751750477402928, | |
| "grad_norm": 13.56935977935791, | |
| "learning_rate": 7.444836461603195e-05, | |
| "loss": 0.2354, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.17568427753023552, | |
| "grad_norm": 4.473952770233154, | |
| "learning_rate": 7.430983680502344e-05, | |
| "loss": 0.0905, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.17619350732017824, | |
| "grad_norm": 4.765993118286133, | |
| "learning_rate": 7.417106419422819e-05, | |
| "loss": 0.0345, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.17670273711012094, | |
| "grad_norm": 17.153976440429688, | |
| "learning_rate": 7.403204818108487e-05, | |
| "loss": 0.1301, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.17721196690006366, | |
| "grad_norm": 6.228205680847168, | |
| "learning_rate": 7.389279016548316e-05, | |
| "loss": 0.0662, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.17772119669000636, | |
| "grad_norm": 6.544225692749023, | |
| "learning_rate": 7.375329154974975e-05, | |
| "loss": 0.0388, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.17823042647994908, | |
| "grad_norm": 9.20956039428711, | |
| "learning_rate": 7.361355373863414e-05, | |
| "loss": 0.0441, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1787396562698918, | |
| "grad_norm": 7.032777786254883, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 0.889, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.1792488860598345, | |
| "grad_norm": 6.679413795471191, | |
| "learning_rate": 7.333336616128369e-05, | |
| "loss": 0.8573, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.17975811584977722, | |
| "grad_norm": 30.212705612182617, | |
| "learning_rate": 7.319291921653464e-05, | |
| "loss": 0.5632, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.1802673456397199, | |
| "grad_norm": 11.796157836914062, | |
| "learning_rate": 7.305223871934657e-05, | |
| "loss": 0.9938, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.18077657542966263, | |
| "grad_norm": 10.97513484954834, | |
| "learning_rate": 7.291132608637052e-05, | |
| "loss": 0.6269, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.18128580521960536, | |
| "grad_norm": 31.140769958496094, | |
| "learning_rate": 7.277018273659517e-05, | |
| "loss": 0.7136, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.18179503500954805, | |
| "grad_norm": 12.688136100769043, | |
| "learning_rate": 7.262881009133242e-05, | |
| "loss": 0.9942, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.18230426479949077, | |
| "grad_norm": 14.422432899475098, | |
| "learning_rate": 7.24872095742033e-05, | |
| "loss": 1.2091, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.1828134945894335, | |
| "grad_norm": 14.67697525024414, | |
| "learning_rate": 7.23453826111234e-05, | |
| "loss": 0.8391, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.1833227243793762, | |
| "grad_norm": 19.619991302490234, | |
| "learning_rate": 7.220333063028872e-05, | |
| "loss": 0.714, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.1838319541693189, | |
| "grad_norm": 23.350360870361328, | |
| "learning_rate": 7.206105506216106e-05, | |
| "loss": 1.0458, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.1843411839592616, | |
| "grad_norm": 14.576053619384766, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.7657, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.18485041374920433, | |
| "grad_norm": 5.723056793212891, | |
| "learning_rate": 7.177583889711762e-05, | |
| "loss": 0.1924, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.18535964353914705, | |
| "grad_norm": 4.669350624084473, | |
| "learning_rate": 7.163290117232542e-05, | |
| "loss": 0.1258, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.18586887332908975, | |
| "grad_norm": 3.4824283123016357, | |
| "learning_rate": 7.148974560445859e-05, | |
| "loss": 0.1544, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.18637810311903247, | |
| "grad_norm": 2.7098116874694824, | |
| "learning_rate": 7.13463736350921e-05, | |
| "loss": 0.1359, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.18688733290897516, | |
| "grad_norm": 9.567126274108887, | |
| "learning_rate": 7.120278670798009e-05, | |
| "loss": 0.1792, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.18739656269891788, | |
| "grad_norm": 15.874075889587402, | |
| "learning_rate": 7.105898626904134e-05, | |
| "loss": 0.2289, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.1879057924888606, | |
| "grad_norm": 15.76208782196045, | |
| "learning_rate": 7.091497376634464e-05, | |
| "loss": 0.2294, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.1884150222788033, | |
| "grad_norm": 10.5492525100708, | |
| "learning_rate": 7.077075065009433e-05, | |
| "loss": 0.1747, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.18892425206874602, | |
| "grad_norm": 4.195391654968262, | |
| "learning_rate": 7.062631837261557e-05, | |
| "loss": 0.1075, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.18943348185868875, | |
| "grad_norm": 2.2104382514953613, | |
| "learning_rate": 7.048167838833977e-05, | |
| "loss": 0.1676, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.18994271164863144, | |
| "grad_norm": 2.322936534881592, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 0.1456, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.19045194143857416, | |
| "grad_norm": 9.757192611694336, | |
| "learning_rate": 7.019178112756624e-05, | |
| "loss": 0.1133, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.19096117122851686, | |
| "grad_norm": 8.49422550201416, | |
| "learning_rate": 7.004652677033068e-05, | |
| "loss": 0.1131, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.19147040101845958, | |
| "grad_norm": 8.021090507507324, | |
| "learning_rate": 6.990107054479312e-05, | |
| "loss": 0.184, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.1919796308084023, | |
| "grad_norm": 10.471813201904297, | |
| "learning_rate": 6.97554139156961e-05, | |
| "loss": 0.1842, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.192488860598345, | |
| "grad_norm": 7.779923915863037, | |
| "learning_rate": 6.960955834980028e-05, | |
| "loss": 0.1842, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.19299809038828772, | |
| "grad_norm": 5.776662826538086, | |
| "learning_rate": 6.946350531586959e-05, | |
| "loss": 0.0991, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.1935073201782304, | |
| "grad_norm": 6.379248142242432, | |
| "learning_rate": 6.931725628465643e-05, | |
| "loss": 0.1489, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.19401654996817314, | |
| "grad_norm": 7.362051010131836, | |
| "learning_rate": 6.917081272888697e-05, | |
| "loss": 0.2167, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.19452577975811586, | |
| "grad_norm": 4.724095821380615, | |
| "learning_rate": 6.902417612324615e-05, | |
| "loss": 0.1161, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.19503500954805855, | |
| "grad_norm": 5.9344563484191895, | |
| "learning_rate": 6.8877347944363e-05, | |
| "loss": 0.1312, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.19554423933800127, | |
| "grad_norm": 7.308444976806641, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.0874, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.196053469127944, | |
| "grad_norm": 4.611720561981201, | |
| "learning_rate": 6.858312278301637e-05, | |
| "loss": 0.1284, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.1965626989178867, | |
| "grad_norm": 4.12031364440918, | |
| "learning_rate": 6.843572876339705e-05, | |
| "loss": 0.0646, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.1970719287078294, | |
| "grad_norm": 9.122203826904297, | |
| "learning_rate": 6.828814909619373e-05, | |
| "loss": 0.1022, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.1975811584977721, | |
| "grad_norm": 11.14821720123291, | |
| "learning_rate": 6.814038526753205e-05, | |
| "loss": 0.1363, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.19809038828771483, | |
| "grad_norm": 9.928780555725098, | |
| "learning_rate": 6.799243876539212e-05, | |
| "loss": 0.0705, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.19859961807765755, | |
| "grad_norm": 3.8702805042266846, | |
| "learning_rate": 6.784431107959359e-05, | |
| "loss": 0.06, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.19910884786760025, | |
| "grad_norm": 8.202292442321777, | |
| "learning_rate": 6.769600370178059e-05, | |
| "loss": 0.1628, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.19961807765754297, | |
| "grad_norm": 7.946099758148193, | |
| "learning_rate": 6.754751812540679e-05, | |
| "loss": 0.0789, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.2001273074474857, | |
| "grad_norm": 7.631869316101074, | |
| "learning_rate": 6.739885584572026e-05, | |
| "loss": 0.0329, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.2006365372374284, | |
| "grad_norm": 6.246660232543945, | |
| "learning_rate": 6.725001835974853e-05, | |
| "loss": 0.1132, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.2011457670273711, | |
| "grad_norm": 4.672296524047852, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 0.0243, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2016549968173138, | |
| "grad_norm": 3.7364211082458496, | |
| "learning_rate": 6.695182376586603e-05, | |
| "loss": 0.0231, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.20216422660725653, | |
| "grad_norm": 14.668068885803223, | |
| "learning_rate": 6.680246966077151e-05, | |
| "loss": 0.0503, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.20267345639719925, | |
| "grad_norm": 10.774435043334961, | |
| "learning_rate": 6.665294635499404e-05, | |
| "loss": 0.0425, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.20318268618714194, | |
| "grad_norm": 5.576310634613037, | |
| "learning_rate": 6.650325535423167e-05, | |
| "loss": 0.0264, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.20369191597708466, | |
| "grad_norm": 6.781991958618164, | |
| "learning_rate": 6.635339816587109e-05, | |
| "loss": 0.0692, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.20369191597708466, | |
| "eval_loss": 0.2951599359512329, | |
| "eval_runtime": 378.4652, | |
| "eval_samples_per_second": 8.741, | |
| "eval_steps_per_second": 2.185, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.20420114576702736, | |
| "grad_norm": 5.35024881362915, | |
| "learning_rate": 6.620337629897254e-05, | |
| "loss": 0.8258, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.20471037555697008, | |
| "grad_norm": 6.47793436050415, | |
| "learning_rate": 6.605319126425454e-05, | |
| "loss": 0.722, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.2052196053469128, | |
| "grad_norm": 4.193264961242676, | |
| "learning_rate": 6.590284457407876e-05, | |
| "loss": 0.2432, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.2057288351368555, | |
| "grad_norm": 8.675887107849121, | |
| "learning_rate": 6.575233774243465e-05, | |
| "loss": 0.7769, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.20623806492679822, | |
| "grad_norm": 11.540266036987305, | |
| "learning_rate": 6.560167228492436e-05, | |
| "loss": 0.7733, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.20674729471674094, | |
| "grad_norm": 12.353882789611816, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.6889, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.20725652450668364, | |
| "grad_norm": 7.550869941711426, | |
| "learning_rate": 6.529987156268526e-05, | |
| "loss": 0.5014, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.20776575429662636, | |
| "grad_norm": 38.89470672607422, | |
| "learning_rate": 6.514873933708638e-05, | |
| "loss": 1.478, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.20827498408656905, | |
| "grad_norm": 12.111885070800781, | |
| "learning_rate": 6.499745456385054e-05, | |
| "loss": 1.0569, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.20878421387651178, | |
| "grad_norm": 16.79547691345215, | |
| "learning_rate": 6.484601876641375e-05, | |
| "loss": 0.9007, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2092934436664545, | |
| "grad_norm": 7.7194108963012695, | |
| "learning_rate": 6.46944334697328e-05, | |
| "loss": 0.5792, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.2098026734563972, | |
| "grad_norm": 10.509634017944336, | |
| "learning_rate": 6.454270020026995e-05, | |
| "loss": 1.0819, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.21031190324633992, | |
| "grad_norm": 8.156822204589844, | |
| "learning_rate": 6.439082048597755e-05, | |
| "loss": 0.5238, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.2108211330362826, | |
| "grad_norm": 6.081987380981445, | |
| "learning_rate": 6.423879585628261e-05, | |
| "loss": 0.0934, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.21133036282622533, | |
| "grad_norm": 3.127657413482666, | |
| "learning_rate": 6.408662784207149e-05, | |
| "loss": 0.1243, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.21183959261616805, | |
| "grad_norm": 2.4891467094421387, | |
| "learning_rate": 6.39343179756744e-05, | |
| "loss": 0.0989, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.21234882240611075, | |
| "grad_norm": 3.107302665710449, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 0.1264, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.21285805219605347, | |
| "grad_norm": 4.113406181335449, | |
| "learning_rate": 6.36292788227699e-05, | |
| "loss": 0.1366, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.2133672819859962, | |
| "grad_norm": 3.0692574977874756, | |
| "learning_rate": 6.34765526080034e-05, | |
| "loss": 0.1656, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.2138765117759389, | |
| "grad_norm": 2.2917490005493164, | |
| "learning_rate": 6.332369068450174e-05, | |
| "loss": 0.0729, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2143857415658816, | |
| "grad_norm": 3.4825942516326904, | |
| "learning_rate": 6.317069459158284e-05, | |
| "loss": 0.1348, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.2148949713558243, | |
| "grad_norm": 7.805583477020264, | |
| "learning_rate": 6.30175658699156e-05, | |
| "loss": 0.061, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.21540420114576703, | |
| "grad_norm": 4.299949645996094, | |
| "learning_rate": 6.286430606150459e-05, | |
| "loss": 0.1709, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.21591343093570975, | |
| "grad_norm": 5.390031814575195, | |
| "learning_rate": 6.271091670967436e-05, | |
| "loss": 0.1293, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.21642266072565244, | |
| "grad_norm": 10.182729721069336, | |
| "learning_rate": 6.255739935905396e-05, | |
| "loss": 0.2065, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.21693189051559517, | |
| "grad_norm": 8.289228439331055, | |
| "learning_rate": 6.240375555556145e-05, | |
| "loss": 0.17, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.21744112030553786, | |
| "grad_norm": 9.04494571685791, | |
| "learning_rate": 6.22499868463882e-05, | |
| "loss": 0.2247, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.21795035009548058, | |
| "grad_norm": 7.830061435699463, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.1934, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.2184595798854233, | |
| "grad_norm": 11.166521072387695, | |
| "learning_rate": 6.194208090603844e-05, | |
| "loss": 0.1904, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.218968809675366, | |
| "grad_norm": 5.363670825958252, | |
| "learning_rate": 6.178794677547137e-05, | |
| "loss": 0.1516, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.21947803946530872, | |
| "grad_norm": 4.773486614227295, | |
| "learning_rate": 6.163369394041111e-05, | |
| "loss": 0.1323, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.21998726925525144, | |
| "grad_norm": 4.628374099731445, | |
| "learning_rate": 6.147932395418205e-05, | |
| "loss": 0.1193, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.22049649904519414, | |
| "grad_norm": 4.363916397094727, | |
| "learning_rate": 6.132483837128823e-05, | |
| "loss": 0.134, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.22100572883513686, | |
| "grad_norm": 5.665075302124023, | |
| "learning_rate": 6.117023874739772e-05, | |
| "loss": 0.16, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.22151495862507956, | |
| "grad_norm": 6.854954242706299, | |
| "learning_rate": 6.1015526639327035e-05, | |
| "loss": 0.169, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.22202418841502228, | |
| "grad_norm": 7.282088756561279, | |
| "learning_rate": 6.0860703605025395e-05, | |
| "loss": 0.139, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.222533418204965, | |
| "grad_norm": 9.19260025024414, | |
| "learning_rate": 6.0705771203559024e-05, | |
| "loss": 0.1304, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.2230426479949077, | |
| "grad_norm": 6.518429279327393, | |
| "learning_rate": 6.05507309950955e-05, | |
| "loss": 0.1446, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.22355187778485042, | |
| "grad_norm": 4.916598320007324, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 0.025, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.22406110757479314, | |
| "grad_norm": 4.131253719329834, | |
| "learning_rate": 6.024033340325954e-05, | |
| "loss": 0.0445, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.22457033736473583, | |
| "grad_norm": 5.542179107666016, | |
| "learning_rate": 6.008497914558744e-05, | |
| "loss": 0.0124, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.22507956715467856, | |
| "grad_norm": 9.51887035369873, | |
| "learning_rate": 5.992952333228728e-05, | |
| "loss": 0.1175, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.22558879694462125, | |
| "grad_norm": 1.3054120540618896, | |
| "learning_rate": 5.9773967528797414e-05, | |
| "loss": 0.0157, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.22609802673456397, | |
| "grad_norm": 3.81061053276062, | |
| "learning_rate": 5.9618313301563055e-05, | |
| "loss": 0.0323, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.2266072565245067, | |
| "grad_norm": 6.797232151031494, | |
| "learning_rate": 5.946256221802051e-05, | |
| "loss": 0.0509, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.2271164863144494, | |
| "grad_norm": 12.037452697753906, | |
| "learning_rate": 5.9306715846581506e-05, | |
| "loss": 0.0986, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.2276257161043921, | |
| "grad_norm": 8.045632362365723, | |
| "learning_rate": 5.915077575661723e-05, | |
| "loss": 0.1348, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.2281349458943348, | |
| "grad_norm": 6.889299392700195, | |
| "learning_rate": 5.8994743518442694e-05, | |
| "loss": 0.0482, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.22864417568427753, | |
| "grad_norm": 7.197052955627441, | |
| "learning_rate": 5.8838620703300784e-05, | |
| "loss": 0.098, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.22915340547422025, | |
| "grad_norm": 4.620372295379639, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.0217, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.22966263526416295, | |
| "grad_norm": 8.069796562194824, | |
| "learning_rate": 5.85261096316312e-05, | |
| "loss": 0.7489, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.23017186505410567, | |
| "grad_norm": 6.844252109527588, | |
| "learning_rate": 5.836972452208654e-05, | |
| "loss": 0.591, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.2306810948440484, | |
| "grad_norm": 4.255478858947754, | |
| "learning_rate": 5.821325512950886e-05, | |
| "loss": 0.201, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.23119032463399108, | |
| "grad_norm": 13.691847801208496, | |
| "learning_rate": 5.805670302954321e-05, | |
| "loss": 0.9461, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.2316995544239338, | |
| "grad_norm": 13.480101585388184, | |
| "learning_rate": 5.79000697986675e-05, | |
| "loss": 0.8203, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.2322087842138765, | |
| "grad_norm": 10.721153259277344, | |
| "learning_rate": 5.7743357014176624e-05, | |
| "loss": 0.6616, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.23271801400381922, | |
| "grad_norm": 13.593427658081055, | |
| "learning_rate": 5.7586566254166583e-05, | |
| "loss": 1.0115, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.23322724379376195, | |
| "grad_norm": 16.89806365966797, | |
| "learning_rate": 5.7429699097518585e-05, | |
| "loss": 1.2747, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.23373647358370464, | |
| "grad_norm": 9.398344993591309, | |
| "learning_rate": 5.7272757123883184e-05, | |
| "loss": 0.8052, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.23424570337364736, | |
| "grad_norm": 9.867512702941895, | |
| "learning_rate": 5.7115741913664264e-05, | |
| "loss": 0.5866, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.23475493316359006, | |
| "grad_norm": 11.531373023986816, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 0.7859, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.23526416295353278, | |
| "grad_norm": 14.971497535705566, | |
| "learning_rate": 5.680149810876322e-05, | |
| "loss": 0.8722, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.2357733927434755, | |
| "grad_norm": 6.004801273345947, | |
| "learning_rate": 5.664427267851271e-05, | |
| "loss": 0.3771, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.2362826225334182, | |
| "grad_norm": 1.8728851079940796, | |
| "learning_rate": 5.6486980340510086e-05, | |
| "loss": 0.1001, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.23679185232336092, | |
| "grad_norm": 5.338820934295654, | |
| "learning_rate": 5.6329622678687463e-05, | |
| "loss": 0.1665, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.23730108211330364, | |
| "grad_norm": 1.342680811882019, | |
| "learning_rate": 5.617220127763474e-05, | |
| "loss": 0.1127, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.23781031190324634, | |
| "grad_norm": 0.8968675136566162, | |
| "learning_rate": 5.601471772258368e-05, | |
| "loss": 0.1042, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.23831954169318906, | |
| "grad_norm": 1.3605763912200928, | |
| "learning_rate": 5.585717359939192e-05, | |
| "loss": 0.0833, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.23882877148313175, | |
| "grad_norm": 1.4521266222000122, | |
| "learning_rate": 5.569957049452703e-05, | |
| "loss": 0.0471, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.23933800127307447, | |
| "grad_norm": 0.7509562969207764, | |
| "learning_rate": 5.5541909995050554e-05, | |
| "loss": 0.0411, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2398472310630172, | |
| "grad_norm": 4.2879743576049805, | |
| "learning_rate": 5.538419368860196e-05, | |
| "loss": 0.1356, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.2403564608529599, | |
| "grad_norm": 2.3868203163146973, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.0857, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.2408656906429026, | |
| "grad_norm": 5.234532833099365, | |
| "learning_rate": 5.506860000814017e-05, | |
| "loss": 0.2399, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.2413749204328453, | |
| "grad_norm": 5.502321243286133, | |
| "learning_rate": 5.4910725812151864e-05, | |
| "loss": 0.1112, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.24188415022278803, | |
| "grad_norm": 6.480447292327881, | |
| "learning_rate": 5.475280216520913e-05, | |
| "loss": 0.1651, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.24239338001273075, | |
| "grad_norm": 7.716391086578369, | |
| "learning_rate": 5.4594830657601384e-05, | |
| "loss": 0.18, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.24290260980267345, | |
| "grad_norm": 6.902336120605469, | |
| "learning_rate": 5.443681288009991e-05, | |
| "loss": 0.184, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.24341183959261617, | |
| "grad_norm": 9.32155704498291, | |
| "learning_rate": 5.427875042394199e-05, | |
| "loss": 0.238, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.2439210693825589, | |
| "grad_norm": 7.200125217437744, | |
| "learning_rate": 5.412064488081482e-05, | |
| "loss": 0.1355, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.24443029917250159, | |
| "grad_norm": 7.890231132507324, | |
| "learning_rate": 5.396249784283942e-05, | |
| "loss": 0.2136, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2449395289624443, | |
| "grad_norm": 3.1145589351654053, | |
| "learning_rate": 5.3804310902554754e-05, | |
| "loss": 0.1003, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.245448758752387, | |
| "grad_norm": 7.380165100097656, | |
| "learning_rate": 5.364608565290155e-05, | |
| "loss": 0.0856, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.24595798854232973, | |
| "grad_norm": 7.165464878082275, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 0.1174, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.24646721833227245, | |
| "grad_norm": 6.7587175369262695, | |
| "learning_rate": 5.3329526599165204e-05, | |
| "loss": 0.1159, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.24697644812221514, | |
| "grad_norm": 6.113307952880859, | |
| "learning_rate": 5.317119598282823e-05, | |
| "loss": 0.1381, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.24748567791215786, | |
| "grad_norm": 4.088643550872803, | |
| "learning_rate": 5.301283343258293e-05, | |
| "loss": 0.0852, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.2479949077021006, | |
| "grad_norm": 6.351160526275635, | |
| "learning_rate": 5.2854440543138406e-05, | |
| "loss": 0.1145, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.24850413749204328, | |
| "grad_norm": 6.1809539794921875, | |
| "learning_rate": 5.2696018909509306e-05, | |
| "loss": 0.1721, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.249013367281986, | |
| "grad_norm": 6.563344478607178, | |
| "learning_rate": 5.253757012699972e-05, | |
| "loss": 0.19, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.2495225970719287, | |
| "grad_norm": 9.084705352783203, | |
| "learning_rate": 5.2379095791187124e-05, | |
| "loss": 0.0925, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.2500318268618714, | |
| "grad_norm": 2.910612106323242, | |
| "learning_rate": 5.2220597497906307e-05, | |
| "loss": 0.0231, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.25054105665181414, | |
| "grad_norm": 7.0364556312561035, | |
| "learning_rate": 5.2062076843233366e-05, | |
| "loss": 0.0679, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.25105028644175686, | |
| "grad_norm": 2.0886571407318115, | |
| "learning_rate": 5.1903535423469505e-05, | |
| "loss": 0.0088, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.25155951623169953, | |
| "grad_norm": 6.553752422332764, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.0633, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.25206874602164225, | |
| "grad_norm": 6.098299503326416, | |
| "learning_rate": 5.158639667490339e-05, | |
| "loss": 0.0672, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.252577975811585, | |
| "grad_norm": 2.6952426433563232, | |
| "learning_rate": 5.142780253968481e-05, | |
| "loss": 0.0067, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.2530872056015277, | |
| "grad_norm": 4.000742435455322, | |
| "learning_rate": 5.126919402651052e-05, | |
| "loss": 0.0326, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.2535964353914704, | |
| "grad_norm": 3.690657615661621, | |
| "learning_rate": 5.1110572732566475e-05, | |
| "loss": 0.0185, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.2541056651814131, | |
| "grad_norm": 3.398660898208618, | |
| "learning_rate": 5.095194025516733e-05, | |
| "loss": 0.0101, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.2546148949713558, | |
| "grad_norm": 4.581557273864746, | |
| "learning_rate": 5.0793298191740404e-05, | |
| "loss": 0.0981, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.25512412476129853, | |
| "grad_norm": 7.636013984680176, | |
| "learning_rate": 5.063464813980948e-05, | |
| "loss": 0.9549, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.25563335455124125, | |
| "grad_norm": 7.066486358642578, | |
| "learning_rate": 5.047599169697884e-05, | |
| "loss": 0.728, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.256142584341184, | |
| "grad_norm": 7.615071773529053, | |
| "learning_rate": 5.03173304609171e-05, | |
| "loss": 0.2771, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.25665181413112664, | |
| "grad_norm": 12.108363151550293, | |
| "learning_rate": 5.015866602934112e-05, | |
| "loss": 0.8318, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.25716104392106937, | |
| "grad_norm": 11.92260456085205, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6557, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.2576702737110121, | |
| "grad_norm": 10.278399467468262, | |
| "learning_rate": 4.984133397065889e-05, | |
| "loss": 0.4904, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.2581795035009548, | |
| "grad_norm": 26.604215621948242, | |
| "learning_rate": 4.968266953908292e-05, | |
| "loss": 1.0672, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.25868873329089753, | |
| "grad_norm": 16.138303756713867, | |
| "learning_rate": 4.952400830302117e-05, | |
| "loss": 1.4055, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.25919796308084025, | |
| "grad_norm": 18.837907791137695, | |
| "learning_rate": 4.9365351860190526e-05, | |
| "loss": 0.9641, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.2597071928707829, | |
| "grad_norm": 27.194965362548828, | |
| "learning_rate": 4.92067018082596e-05, | |
| "loss": 0.7279, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.26021642266072564, | |
| "grad_norm": 12.584843635559082, | |
| "learning_rate": 4.9048059744832666e-05, | |
| "loss": 0.8671, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.26072565245066837, | |
| "grad_norm": 11.650589942932129, | |
| "learning_rate": 4.888942726743353e-05, | |
| "loss": 1.0271, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.2612348822406111, | |
| "grad_norm": 1.3385618925094604, | |
| "learning_rate": 4.8730805973489476e-05, | |
| "loss": 0.0686, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.2617441120305538, | |
| "grad_norm": 2.49957013130188, | |
| "learning_rate": 4.85721974603152e-05, | |
| "loss": 0.1366, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.2622533418204965, | |
| "grad_norm": 3.0434892177581787, | |
| "learning_rate": 4.841360332509663e-05, | |
| "loss": 0.1553, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.2627625716104392, | |
| "grad_norm": 3.903759717941284, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.1314, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.2632718014003819, | |
| "grad_norm": 1.2716883420944214, | |
| "learning_rate": 4.8096464576530507e-05, | |
| "loss": 0.0685, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.26378103119032464, | |
| "grad_norm": 1.2184836864471436, | |
| "learning_rate": 4.7937923156766646e-05, | |
| "loss": 0.1099, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.26429026098026737, | |
| "grad_norm": 5.961424350738525, | |
| "learning_rate": 4.77794025020937e-05, | |
| "loss": 0.1555, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.26479949077021003, | |
| "grad_norm": 4.472005844116211, | |
| "learning_rate": 4.762090420881289e-05, | |
| "loss": 0.1798, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.26530872056015276, | |
| "grad_norm": 1.6707897186279297, | |
| "learning_rate": 4.7462429873000295e-05, | |
| "loss": 0.0802, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.2658179503500955, | |
| "grad_norm": 1.6405971050262451, | |
| "learning_rate": 4.730398109049071e-05, | |
| "loss": 0.1076, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.2663271801400382, | |
| "grad_norm": 5.0963053703308105, | |
| "learning_rate": 4.71455594568616e-05, | |
| "loss": 0.137, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.2668364099299809, | |
| "grad_norm": 7.195677280426025, | |
| "learning_rate": 4.698716656741708e-05, | |
| "loss": 0.0832, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.2673456397199236, | |
| "grad_norm": 6.091291427612305, | |
| "learning_rate": 4.6828804017171776e-05, | |
| "loss": 0.0695, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2678548695098663, | |
| "grad_norm": 5.164731025695801, | |
| "learning_rate": 4.667047340083481e-05, | |
| "loss": 0.1888, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.26836409929980903, | |
| "grad_norm": 6.838079929351807, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 0.1742, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.26887332908975176, | |
| "grad_norm": 7.102747440338135, | |
| "learning_rate": 4.635391434709847e-05, | |
| "loss": 0.1379, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.2693825588796945, | |
| "grad_norm": 4.239371299743652, | |
| "learning_rate": 4.619568909744524e-05, | |
| "loss": 0.0998, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.2698917886696372, | |
| "grad_norm": 7.476315498352051, | |
| "learning_rate": 4.603750215716057e-05, | |
| "loss": 0.1463, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.27040101845957987, | |
| "grad_norm": 8.441755294799805, | |
| "learning_rate": 4.587935511918521e-05, | |
| "loss": 0.1281, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.2709102482495226, | |
| "grad_norm": 5.9278564453125, | |
| "learning_rate": 4.5721249576058027e-05, | |
| "loss": 0.0733, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.2714194780394653, | |
| "grad_norm": 3.918498992919922, | |
| "learning_rate": 4.5563187119900104e-05, | |
| "loss": 0.0616, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.27192870782940803, | |
| "grad_norm": 9.709593772888184, | |
| "learning_rate": 4.5405169342398634e-05, | |
| "loss": 0.1179, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.27243793761935076, | |
| "grad_norm": 5.810922145843506, | |
| "learning_rate": 4.5247197834790876e-05, | |
| "loss": 0.1212, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.2729471674092934, | |
| "grad_norm": 4.741395950317383, | |
| "learning_rate": 4.508927418784815e-05, | |
| "loss": 0.0867, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.27345639719923615, | |
| "grad_norm": 3.2584261894226074, | |
| "learning_rate": 4.493139999185983e-05, | |
| "loss": 0.0676, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.27396562698917887, | |
| "grad_norm": 3.621077060699463, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.0495, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.2744748567791216, | |
| "grad_norm": 3.1728439331054688, | |
| "learning_rate": 4.461580631139805e-05, | |
| "loss": 0.0396, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.2749840865690643, | |
| "grad_norm": 3.342538833618164, | |
| "learning_rate": 4.445809000494946e-05, | |
| "loss": 0.0535, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.275493316359007, | |
| "grad_norm": 9.337960243225098, | |
| "learning_rate": 4.4300429505472976e-05, | |
| "loss": 0.0665, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.2760025461489497, | |
| "grad_norm": 6.381386756896973, | |
| "learning_rate": 4.4142826400608086e-05, | |
| "loss": 0.0668, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.2765117759388924, | |
| "grad_norm": 2.0168566703796387, | |
| "learning_rate": 4.398528227741633e-05, | |
| "loss": 0.0073, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.27702100572883515, | |
| "grad_norm": 1.1732912063598633, | |
| "learning_rate": 4.3827798722365264e-05, | |
| "loss": 0.0031, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.27753023551877787, | |
| "grad_norm": 2.5314676761627197, | |
| "learning_rate": 4.3670377321312535e-05, | |
| "loss": 0.0079, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.27803946530872053, | |
| "grad_norm": 9.042901039123535, | |
| "learning_rate": 4.351301965948991e-05, | |
| "loss": 0.1532, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.27854869509866326, | |
| "grad_norm": 11.261495590209961, | |
| "learning_rate": 4.33557273214873e-05, | |
| "loss": 0.1224, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.279057924888606, | |
| "grad_norm": 2.12158465385437, | |
| "learning_rate": 4.3198501891236804e-05, | |
| "loss": 0.0167, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.2795671546785487, | |
| "grad_norm": 2.104827404022217, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 0.0119, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.2800763844684914, | |
| "grad_norm": 7.128782749176025, | |
| "learning_rate": 4.288425808633575e-05, | |
| "loss": 0.1068, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2805856142584341, | |
| "grad_norm": 4.171801567077637, | |
| "learning_rate": 4.272724287611684e-05, | |
| "loss": 0.7447, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.2810948440483768, | |
| "grad_norm": 5.680448532104492, | |
| "learning_rate": 4.2570300902481426e-05, | |
| "loss": 0.7541, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.28160407383831954, | |
| "grad_norm": 3.7267653942108154, | |
| "learning_rate": 4.241343374583343e-05, | |
| "loss": 0.2126, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.28211330362826226, | |
| "grad_norm": 12.336726188659668, | |
| "learning_rate": 4.2256642985823395e-05, | |
| "loss": 0.7713, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.282622533418205, | |
| "grad_norm": 14.465222358703613, | |
| "learning_rate": 4.20999302013325e-05, | |
| "loss": 0.8353, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.2831317632081477, | |
| "grad_norm": 15.680033683776855, | |
| "learning_rate": 4.19432969704568e-05, | |
| "loss": 0.7337, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.28364099299809037, | |
| "grad_norm": 19.08047866821289, | |
| "learning_rate": 4.178674487049116e-05, | |
| "loss": 0.6195, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.2841502227880331, | |
| "grad_norm": 15.031834602355957, | |
| "learning_rate": 4.163027547791347e-05, | |
| "loss": 1.3184, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.2846594525779758, | |
| "grad_norm": 15.743155479431152, | |
| "learning_rate": 4.147389036836881e-05, | |
| "loss": 1.1384, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.28516868236791854, | |
| "grad_norm": 8.661066055297852, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.6242, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.28567791215786126, | |
| "grad_norm": 13.872480392456055, | |
| "learning_rate": 4.116137929669921e-05, | |
| "loss": 0.8754, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.2861871419478039, | |
| "grad_norm": 10.3219575881958, | |
| "learning_rate": 4.100525648155731e-05, | |
| "loss": 0.6358, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.28669637173774665, | |
| "grad_norm": 11.022652626037598, | |
| "learning_rate": 4.084922424338277e-05, | |
| "loss": 0.2152, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.28720560152768937, | |
| "grad_norm": 9.58282470703125, | |
| "learning_rate": 4.06932841534185e-05, | |
| "loss": 0.1844, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.2877148313176321, | |
| "grad_norm": 6.9711503982543945, | |
| "learning_rate": 4.0537437781979506e-05, | |
| "loss": 0.1819, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.2882240611075748, | |
| "grad_norm": 2.8225958347320557, | |
| "learning_rate": 4.038168669843697e-05, | |
| "loss": 0.12, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.2887332908975175, | |
| "grad_norm": 3.3902416229248047, | |
| "learning_rate": 4.0226032471202604e-05, | |
| "loss": 0.1144, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.2892425206874602, | |
| "grad_norm": 2.9288244247436523, | |
| "learning_rate": 4.007047666771274e-05, | |
| "loss": 0.1365, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.2897517504774029, | |
| "grad_norm": 1.9439268112182617, | |
| "learning_rate": 3.991502085441259e-05, | |
| "loss": 0.1248, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.29026098026734565, | |
| "grad_norm": 3.1058437824249268, | |
| "learning_rate": 3.9759666596740476e-05, | |
| "loss": 0.1281, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.29077021005728837, | |
| "grad_norm": 1.4958165884017944, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 0.0681, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.29127943984723104, | |
| "grad_norm": 7.544384479522705, | |
| "learning_rate": 3.944926900490452e-05, | |
| "loss": 0.0738, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.29178866963717376, | |
| "grad_norm": 10.52266788482666, | |
| "learning_rate": 3.929422879644099e-05, | |
| "loss": 0.1693, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.2922978994271165, | |
| "grad_norm": 18.5694637298584, | |
| "learning_rate": 3.913929639497462e-05, | |
| "loss": 0.2253, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.2928071292170592, | |
| "grad_norm": 4.89534330368042, | |
| "learning_rate": 3.898447336067297e-05, | |
| "loss": 0.0848, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.2933163590070019, | |
| "grad_norm": 3.438270092010498, | |
| "learning_rate": 3.882976125260229e-05, | |
| "loss": 0.0828, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.29382558879694465, | |
| "grad_norm": 3.7927188873291016, | |
| "learning_rate": 3.8675161628711776e-05, | |
| "loss": 0.0915, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.2943348185868873, | |
| "grad_norm": 10.36831283569336, | |
| "learning_rate": 3.852067604581794e-05, | |
| "loss": 0.1148, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.29484404837683004, | |
| "grad_norm": 7.768423557281494, | |
| "learning_rate": 3.836630605958888e-05, | |
| "loss": 0.1411, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.29535327816677276, | |
| "grad_norm": 4.291931629180908, | |
| "learning_rate": 3.821205322452863e-05, | |
| "loss": 0.0666, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2958625079567155, | |
| "grad_norm": 5.690462589263916, | |
| "learning_rate": 3.8057919093961553e-05, | |
| "loss": 0.1047, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.2963717377466582, | |
| "grad_norm": 4.0527238845825195, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.1114, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.29688096753660087, | |
| "grad_norm": 5.845585346221924, | |
| "learning_rate": 3.775001315361183e-05, | |
| "loss": 0.1266, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.2973901973265436, | |
| "grad_norm": 3.8803865909576416, | |
| "learning_rate": 3.759624444443858e-05, | |
| "loss": 0.0934, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.2978994271164863, | |
| "grad_norm": 4.592535495758057, | |
| "learning_rate": 3.744260064094604e-05, | |
| "loss": 0.0689, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.29840865690642904, | |
| "grad_norm": 4.82613468170166, | |
| "learning_rate": 3.728908329032567e-05, | |
| "loss": 0.1099, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.29891788669637176, | |
| "grad_norm": 7.046941757202148, | |
| "learning_rate": 3.713569393849543e-05, | |
| "loss": 0.1384, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.2994271164863144, | |
| "grad_norm": 8.008650779724121, | |
| "learning_rate": 3.69824341300844e-05, | |
| "loss": 0.0777, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.29993634627625715, | |
| "grad_norm": 3.9791100025177, | |
| "learning_rate": 3.6829305408417166e-05, | |
| "loss": 0.0693, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.30044557606619987, | |
| "grad_norm": 6.2310590744018555, | |
| "learning_rate": 3.6676309315498256e-05, | |
| "loss": 0.0861, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3009548058561426, | |
| "grad_norm": 8.531147956848145, | |
| "learning_rate": 3.6523447391996614e-05, | |
| "loss": 0.0882, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.3014640356460853, | |
| "grad_norm": 3.792165994644165, | |
| "learning_rate": 3.6370721177230116e-05, | |
| "loss": 0.0685, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.301973265436028, | |
| "grad_norm": 4.004833221435547, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 0.0936, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.3024824952259707, | |
| "grad_norm": 15.85920238494873, | |
| "learning_rate": 3.606568202432562e-05, | |
| "loss": 0.2744, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.3029917250159134, | |
| "grad_norm": 0.3684404492378235, | |
| "learning_rate": 3.591337215792852e-05, | |
| "loss": 0.0018, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.30350095480585615, | |
| "grad_norm": 9.898280143737793, | |
| "learning_rate": 3.5761204143717385e-05, | |
| "loss": 0.0506, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.30401018459579887, | |
| "grad_norm": 1.1307116746902466, | |
| "learning_rate": 3.560917951402245e-05, | |
| "loss": 0.0033, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.3045194143857416, | |
| "grad_norm": 4.563685417175293, | |
| "learning_rate": 3.545729979973005e-05, | |
| "loss": 0.0299, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.30502864417568426, | |
| "grad_norm": 4.1302080154418945, | |
| "learning_rate": 3.530556653026721e-05, | |
| "loss": 0.0331, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.305537873965627, | |
| "grad_norm": 6.200318813323975, | |
| "learning_rate": 3.515398123358627e-05, | |
| "loss": 0.0764, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.305537873965627, | |
| "eval_loss": 0.2474220246076584, | |
| "eval_runtime": 377.4722, | |
| "eval_samples_per_second": 8.764, | |
| "eval_steps_per_second": 2.191, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3060471037555697, | |
| "grad_norm": 5.079921722412109, | |
| "learning_rate": 3.5002545436149474e-05, | |
| "loss": 0.8877, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.3065563335455124, | |
| "grad_norm": 5.308548927307129, | |
| "learning_rate": 3.485126066291364e-05, | |
| "loss": 0.8411, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.30706556333545515, | |
| "grad_norm": 7.06653356552124, | |
| "learning_rate": 3.470012843731476e-05, | |
| "loss": 0.8479, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.3075747931253978, | |
| "grad_norm": 2.4465208053588867, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.0402, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.30808402291534054, | |
| "grad_norm": 7.19598913192749, | |
| "learning_rate": 3.439832771507565e-05, | |
| "loss": 0.5183, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.30859325270528326, | |
| "grad_norm": 10.159460067749023, | |
| "learning_rate": 3.424766225756537e-05, | |
| "loss": 0.6912, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.309102482495226, | |
| "grad_norm": 6.36456298828125, | |
| "learning_rate": 3.4097155425921254e-05, | |
| "loss": 0.4891, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.3096117122851687, | |
| "grad_norm": 28.50994873046875, | |
| "learning_rate": 3.394680873574546e-05, | |
| "loss": 1.2461, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.31012094207511137, | |
| "grad_norm": 11.220614433288574, | |
| "learning_rate": 3.3796623701027476e-05, | |
| "loss": 0.8775, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.3106301718650541, | |
| "grad_norm": 9.175284385681152, | |
| "learning_rate": 3.364660183412892e-05, | |
| "loss": 0.8295, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.3111394016549968, | |
| "grad_norm": 13.259340286254883, | |
| "learning_rate": 3.349674464576834e-05, | |
| "loss": 0.8886, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.31164863144493954, | |
| "grad_norm": 10.061495780944824, | |
| "learning_rate": 3.334705364500596e-05, | |
| "loss": 0.7599, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.31215786123488226, | |
| "grad_norm": 4.028329849243164, | |
| "learning_rate": 3.3197530339228487e-05, | |
| "loss": 0.1656, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.31266709102482493, | |
| "grad_norm": 2.669538736343384, | |
| "learning_rate": 3.304817623413397e-05, | |
| "loss": 0.124, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.31317632081476765, | |
| "grad_norm": 3.3735191822052, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 0.1331, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.3136855506047104, | |
| "grad_norm": 1.7686545848846436, | |
| "learning_rate": 3.274998164025148e-05, | |
| "loss": 0.1367, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.3141947803946531, | |
| "grad_norm": 2.966285467147827, | |
| "learning_rate": 3.260114415427975e-05, | |
| "loss": 0.1389, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.3147040101845958, | |
| "grad_norm": 6.92048454284668, | |
| "learning_rate": 3.2452481874593234e-05, | |
| "loss": 0.1157, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.3152132399745385, | |
| "grad_norm": 8.68355655670166, | |
| "learning_rate": 3.230399629821942e-05, | |
| "loss": 0.1763, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.3157224697644812, | |
| "grad_norm": 3.1240475177764893, | |
| "learning_rate": 3.215568892040641e-05, | |
| "loss": 0.1004, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.31623169955442393, | |
| "grad_norm": 1.7762776613235474, | |
| "learning_rate": 3.200756123460788e-05, | |
| "loss": 0.0993, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.31674092934436665, | |
| "grad_norm": 2.550874710083008, | |
| "learning_rate": 3.1859614732467954e-05, | |
| "loss": 0.0514, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.3172501591343094, | |
| "grad_norm": 3.7702434062957764, | |
| "learning_rate": 3.171185090380628e-05, | |
| "loss": 0.0691, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.3177593889242521, | |
| "grad_norm": 4.985774517059326, | |
| "learning_rate": 3.156427123660297e-05, | |
| "loss": 0.0982, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.31826861871419476, | |
| "grad_norm": 9.22007942199707, | |
| "learning_rate": 3.141687721698363e-05, | |
| "loss": 0.123, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.3187778485041375, | |
| "grad_norm": 3.031754970550537, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.0507, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.3192870782940802, | |
| "grad_norm": 6.516983509063721, | |
| "learning_rate": 3.1122652055637015e-05, | |
| "loss": 0.0944, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.31979630808402293, | |
| "grad_norm": 4.626400470733643, | |
| "learning_rate": 3.097582387675385e-05, | |
| "loss": 0.1333, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.32030553787396565, | |
| "grad_norm": 5.953018665313721, | |
| "learning_rate": 3.082918727111304e-05, | |
| "loss": 0.1317, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.3208147676639083, | |
| "grad_norm": 4.307436466217041, | |
| "learning_rate": 3.0682743715343564e-05, | |
| "loss": 0.0859, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.32132399745385104, | |
| "grad_norm": 5.189202308654785, | |
| "learning_rate": 3.053649468413043e-05, | |
| "loss": 0.1205, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.32183322724379376, | |
| "grad_norm": 7.541786193847656, | |
| "learning_rate": 3.0390441650199724e-05, | |
| "loss": 0.1567, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.3223424570337365, | |
| "grad_norm": 5.107945442199707, | |
| "learning_rate": 3.0244586084303905e-05, | |
| "loss": 0.0836, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.3228516868236792, | |
| "grad_norm": 5.2220563888549805, | |
| "learning_rate": 3.0098929455206904e-05, | |
| "loss": 0.0487, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.3233609166136219, | |
| "grad_norm": 3.8281970024108887, | |
| "learning_rate": 2.9953473229669328e-05, | |
| "loss": 0.0783, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.3238701464035646, | |
| "grad_norm": 4.100976943969727, | |
| "learning_rate": 2.9808218872433767e-05, | |
| "loss": 0.1336, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.3243793761935073, | |
| "grad_norm": 4.79497766494751, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.0657, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.32488860598345004, | |
| "grad_norm": 5.689431190490723, | |
| "learning_rate": 2.9518321611660237e-05, | |
| "loss": 0.033, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.32539783577339276, | |
| "grad_norm": 7.910229206085205, | |
| "learning_rate": 2.9373681627384447e-05, | |
| "loss": 0.0893, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.32590706556333543, | |
| "grad_norm": 3.019529104232788, | |
| "learning_rate": 2.9229249349905684e-05, | |
| "loss": 0.0205, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.32641629535327815, | |
| "grad_norm": 2.8111181259155273, | |
| "learning_rate": 2.9085026233655365e-05, | |
| "loss": 0.0133, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.3269255251432209, | |
| "grad_norm": 4.1778974533081055, | |
| "learning_rate": 2.894101373095867e-05, | |
| "loss": 0.0162, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.3274347549331636, | |
| "grad_norm": 40.197940826416016, | |
| "learning_rate": 2.8797213292019926e-05, | |
| "loss": 0.0561, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.3279439847231063, | |
| "grad_norm": 5.798105239868164, | |
| "learning_rate": 2.8653626364907917e-05, | |
| "loss": 0.0468, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.32845321451304904, | |
| "grad_norm": 0.09037820249795914, | |
| "learning_rate": 2.851025439554142e-05, | |
| "loss": 0.0004, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3289624443029917, | |
| "grad_norm": 6.750631809234619, | |
| "learning_rate": 2.8367098827674578e-05, | |
| "loss": 0.12, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.32947167409293443, | |
| "grad_norm": 8.365553855895996, | |
| "learning_rate": 2.8224161102882397e-05, | |
| "loss": 0.0613, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.32998090388287715, | |
| "grad_norm": 8.15538501739502, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.1204, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.3304901336728199, | |
| "grad_norm": 8.706940650939941, | |
| "learning_rate": 2.7938944937838923e-05, | |
| "loss": 0.0566, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.3309993634627626, | |
| "grad_norm": 4.927478790283203, | |
| "learning_rate": 2.7796669369711294e-05, | |
| "loss": 0.1079, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.33150859325270526, | |
| "grad_norm": 4.741428852081299, | |
| "learning_rate": 2.7654617388876615e-05, | |
| "loss": 0.7197, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.332017823042648, | |
| "grad_norm": 4.9175944328308105, | |
| "learning_rate": 2.7512790425796718e-05, | |
| "loss": 0.5017, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.3325270528325907, | |
| "grad_norm": 3.583587884902954, | |
| "learning_rate": 2.7371189908667604e-05, | |
| "loss": 0.2062, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.33303628262253343, | |
| "grad_norm": 9.022587776184082, | |
| "learning_rate": 2.7229817263404866e-05, | |
| "loss": 0.7837, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.33354551241247615, | |
| "grad_norm": 9.906291007995605, | |
| "learning_rate": 2.708867391362948e-05, | |
| "loss": 0.6524, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.3340547422024188, | |
| "grad_norm": 5.258704662322998, | |
| "learning_rate": 2.694776128065345e-05, | |
| "loss": 0.3687, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.33456397199236154, | |
| "grad_norm": 7.302427768707275, | |
| "learning_rate": 2.6807080783465376e-05, | |
| "loss": 0.4876, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.33507320178230426, | |
| "grad_norm": 13.841546058654785, | |
| "learning_rate": 2.6666633838716314e-05, | |
| "loss": 1.4033, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.335582431572247, | |
| "grad_norm": 11.859368324279785, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 0.9077, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.3360916613621897, | |
| "grad_norm": 13.534501075744629, | |
| "learning_rate": 2.638644626136587e-05, | |
| "loss": 0.7607, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3366008911521324, | |
| "grad_norm": 7.9164958000183105, | |
| "learning_rate": 2.6246708450250256e-05, | |
| "loss": 0.599, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.3371101209420751, | |
| "grad_norm": 8.299764633178711, | |
| "learning_rate": 2.6107209834516854e-05, | |
| "loss": 1.0367, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.3376193507320178, | |
| "grad_norm": 4.36458683013916, | |
| "learning_rate": 2.596795181891514e-05, | |
| "loss": 0.4012, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.33812858052196054, | |
| "grad_norm": 2.5791258811950684, | |
| "learning_rate": 2.5828935805771802e-05, | |
| "loss": 0.1412, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.33863781031190326, | |
| "grad_norm": 4.219427585601807, | |
| "learning_rate": 2.5690163194976575e-05, | |
| "loss": 0.1824, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.33914704010184593, | |
| "grad_norm": 2.9232358932495117, | |
| "learning_rate": 2.5551635383968065e-05, | |
| "loss": 0.17, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.33965626989178865, | |
| "grad_norm": 2.1133458614349365, | |
| "learning_rate": 2.5413353767719805e-05, | |
| "loss": 0.1109, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.3401654996817314, | |
| "grad_norm": 0.8237698078155518, | |
| "learning_rate": 2.5275319738726165e-05, | |
| "loss": 0.0692, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.3406747294716741, | |
| "grad_norm": 2.2106404304504395, | |
| "learning_rate": 2.513753468698826e-05, | |
| "loss": 0.1241, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.3411839592616168, | |
| "grad_norm": 2.5900228023529053, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.1617, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.34169318905155954, | |
| "grad_norm": 0.9092320799827576, | |
| "learning_rate": 2.486271706273421e-05, | |
| "loss": 0.1002, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.3422024188415022, | |
| "grad_norm": 3.5675482749938965, | |
| "learning_rate": 2.4725687257628534e-05, | |
| "loss": 0.0889, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.34271164863144493, | |
| "grad_norm": 4.1621479988098145, | |
| "learning_rate": 2.4588911964571553e-05, | |
| "loss": 0.119, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.34322087842138765, | |
| "grad_norm": 1.428825855255127, | |
| "learning_rate": 2.4452392560888976e-05, | |
| "loss": 0.0871, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.3437301082113304, | |
| "grad_norm": 4.961071968078613, | |
| "learning_rate": 2.4316130421329697e-05, | |
| "loss": 0.0718, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.3442393380012731, | |
| "grad_norm": 5.432969570159912, | |
| "learning_rate": 2.418012691805191e-05, | |
| "loss": 0.0782, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.34474856779121577, | |
| "grad_norm": 4.57743501663208, | |
| "learning_rate": 2.4044383420609406e-05, | |
| "loss": 0.1596, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.3452577975811585, | |
| "grad_norm": 7.462673187255859, | |
| "learning_rate": 2.3908901295937713e-05, | |
| "loss": 0.1173, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.3457670273711012, | |
| "grad_norm": 4.732002258300781, | |
| "learning_rate": 2.3773681908340284e-05, | |
| "loss": 0.111, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.34627625716104393, | |
| "grad_norm": 5.835433483123779, | |
| "learning_rate": 2.363872661947488e-05, | |
| "loss": 0.1395, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.34678548695098665, | |
| "grad_norm": 4.240106105804443, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 0.0979, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.3472947167409293, | |
| "grad_norm": 6.862554550170898, | |
| "learning_rate": 2.336961377126001e-05, | |
| "loss": 0.1356, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.34780394653087204, | |
| "grad_norm": 5.05124568939209, | |
| "learning_rate": 2.3235458921873925e-05, | |
| "loss": 0.1112, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.34831317632081477, | |
| "grad_norm": 5.341157913208008, | |
| "learning_rate": 2.310157359111938e-05, | |
| "loss": 0.0908, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.3488224061107575, | |
| "grad_norm": 3.359665870666504, | |
| "learning_rate": 2.296795912722014e-05, | |
| "loss": 0.081, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.3493316359007002, | |
| "grad_norm": 4.865165710449219, | |
| "learning_rate": 2.283461687567236e-05, | |
| "loss": 0.1172, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.3498408656906429, | |
| "grad_norm": 8.436277389526367, | |
| "learning_rate": 2.2701548179231048e-05, | |
| "loss": 0.1962, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.3503500954805856, | |
| "grad_norm": 6.217014312744141, | |
| "learning_rate": 2.2568754377896516e-05, | |
| "loss": 0.0362, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.3508593252705283, | |
| "grad_norm": 19.41240119934082, | |
| "learning_rate": 2.2436236808900844e-05, | |
| "loss": 0.275, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.35136855506047104, | |
| "grad_norm": 6.9711456298828125, | |
| "learning_rate": 2.2303996806694488e-05, | |
| "loss": 0.0658, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.35187778485041377, | |
| "grad_norm": 6.803055763244629, | |
| "learning_rate": 2.2172035702932825e-05, | |
| "loss": 0.0384, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.3523870146403565, | |
| "grad_norm": 3.9196035861968994, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 0.0957, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.35289624443029916, | |
| "grad_norm": 3.145606756210327, | |
| "learning_rate": 2.1908955503308993e-05, | |
| "loss": 0.0623, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.3534054742202419, | |
| "grad_norm": 16.734477996826172, | |
| "learning_rate": 2.1777839056661554e-05, | |
| "loss": 0.0769, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.3539147040101846, | |
| "grad_norm": 9.658271789550781, | |
| "learning_rate": 2.164700680686147e-05, | |
| "loss": 0.0265, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3544239338001273, | |
| "grad_norm": 1.6071585416793823, | |
| "learning_rate": 2.1516460071388062e-05, | |
| "loss": 0.0115, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.35493316359007004, | |
| "grad_norm": 4.15781307220459, | |
| "learning_rate": 2.1386200164845526e-05, | |
| "loss": 0.0719, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.3554423933800127, | |
| "grad_norm": 1.0766063928604126, | |
| "learning_rate": 2.125622839894964e-05, | |
| "loss": 0.0062, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.35595162316995543, | |
| "grad_norm": 3.8494861125946045, | |
| "learning_rate": 2.1126546082514664e-05, | |
| "loss": 0.0104, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.35646085295989816, | |
| "grad_norm": 4.333930015563965, | |
| "learning_rate": 2.09971545214401e-05, | |
| "loss": 0.069, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3569700827498409, | |
| "grad_norm": 5.104018211364746, | |
| "learning_rate": 2.086805501869749e-05, | |
| "loss": 0.735, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.3574793125397836, | |
| "grad_norm": 4.622631072998047, | |
| "learning_rate": 2.073924887431744e-05, | |
| "loss": 0.6267, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.35798854232972627, | |
| "grad_norm": 2.708951234817505, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.1031, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.358497772119669, | |
| "grad_norm": 6.911785125732422, | |
| "learning_rate": 2.048252184598352e-05, | |
| "loss": 0.5889, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.3590070019096117, | |
| "grad_norm": 8.187259674072266, | |
| "learning_rate": 2.0354603547267985e-05, | |
| "loss": 0.5466, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.35951623169955443, | |
| "grad_norm": 6.681227207183838, | |
| "learning_rate": 2.0226983777365604e-05, | |
| "loss": 0.5335, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.36002546148949716, | |
| "grad_norm": 13.172379493713379, | |
| "learning_rate": 2.0099663821406056e-05, | |
| "loss": 0.6291, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.3605346912794398, | |
| "grad_norm": 10.422917366027832, | |
| "learning_rate": 1.9972644961499854e-05, | |
| "loss": 0.8471, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.36104392106938255, | |
| "grad_norm": 13.669676780700684, | |
| "learning_rate": 1.9845928476725524e-05, | |
| "loss": 1.0402, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.36155315085932527, | |
| "grad_norm": 6.9141716957092285, | |
| "learning_rate": 1.9719515643116674e-05, | |
| "loss": 0.6035, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.362062380649268, | |
| "grad_norm": 8.188703536987305, | |
| "learning_rate": 1.959340773364911e-05, | |
| "loss": 0.6699, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.3625716104392107, | |
| "grad_norm": 7.684673309326172, | |
| "learning_rate": 1.946760601822809e-05, | |
| "loss": 0.7657, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.3630808402291534, | |
| "grad_norm": 2.5300745964050293, | |
| "learning_rate": 1.9342111763675512e-05, | |
| "loss": 0.1569, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.3635900700190961, | |
| "grad_norm": 1.3801597356796265, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 0.081, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.3640992998090388, | |
| "grad_norm": 1.2587306499481201, | |
| "learning_rate": 1.9092050688969738e-05, | |
| "loss": 0.1421, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.36460852959898155, | |
| "grad_norm": 1.1372148990631104, | |
| "learning_rate": 1.8967486386928817e-05, | |
| "loss": 0.0961, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.36511775938892427, | |
| "grad_norm": 0.7794922590255737, | |
| "learning_rate": 1.8843234581955442e-05, | |
| "loss": 0.1008, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.365626989178867, | |
| "grad_norm": 1.1420189142227173, | |
| "learning_rate": 1.8719296525263922e-05, | |
| "loss": 0.1071, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.36613621896880966, | |
| "grad_norm": 1.6180529594421387, | |
| "learning_rate": 1.859567346490913e-05, | |
| "loss": 0.0875, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.3666454487587524, | |
| "grad_norm": 1.0748051404953003, | |
| "learning_rate": 1.847236664577389e-05, | |
| "loss": 0.1288, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3671546785486951, | |
| "grad_norm": 1.0483026504516602, | |
| "learning_rate": 1.8349377309556486e-05, | |
| "loss": 0.0774, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.3676639083386378, | |
| "grad_norm": 1.506415843963623, | |
| "learning_rate": 1.8226706694758195e-05, | |
| "loss": 0.1362, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.36817313812858055, | |
| "grad_norm": 1.0291399955749512, | |
| "learning_rate": 1.810435603667075e-05, | |
| "loss": 0.1071, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.3686823679185232, | |
| "grad_norm": 5.942626953125, | |
| "learning_rate": 1.7982326567363888e-05, | |
| "loss": 0.1614, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.36919159770846594, | |
| "grad_norm": 12.163902282714844, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 0.1458, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.36970082749840866, | |
| "grad_norm": 3.687087059020996, | |
| "learning_rate": 1.773923610718686e-05, | |
| "loss": 0.0919, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.3702100572883514, | |
| "grad_norm": 4.903958320617676, | |
| "learning_rate": 1.7618177564234905e-05, | |
| "loss": 0.1254, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.3707192870782941, | |
| "grad_norm": 4.808300495147705, | |
| "learning_rate": 1.7497445105875377e-05, | |
| "loss": 0.1473, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.37122851686823677, | |
| "grad_norm": 5.668433666229248, | |
| "learning_rate": 1.73770399478828e-05, | |
| "loss": 0.1524, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.3717377466581795, | |
| "grad_norm": 8.04811954498291, | |
| "learning_rate": 1.725696330273575e-05, | |
| "loss": 0.0919, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.3722469764481222, | |
| "grad_norm": 4.406056880950928, | |
| "learning_rate": 1.7137216379604727e-05, | |
| "loss": 0.1046, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.37275620623806494, | |
| "grad_norm": 7.855915546417236, | |
| "learning_rate": 1.7017800384339928e-05, | |
| "loss": 0.0709, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.37326543602800766, | |
| "grad_norm": 3.947338342666626, | |
| "learning_rate": 1.6898716519459074e-05, | |
| "loss": 0.072, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.3737746658179503, | |
| "grad_norm": 7.404916763305664, | |
| "learning_rate": 1.6779965984135377e-05, | |
| "loss": 0.1313, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.37428389560789305, | |
| "grad_norm": 4.375075817108154, | |
| "learning_rate": 1.6661549974185424e-05, | |
| "loss": 0.0829, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.37479312539783577, | |
| "grad_norm": 4.642003536224365, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.0602, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.3753023551877785, | |
| "grad_norm": 5.987782001495361, | |
| "learning_rate": 1.6425726296817633e-05, | |
| "loss": 0.0868, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.3758115849777212, | |
| "grad_norm": 10.39226245880127, | |
| "learning_rate": 1.6308321004141607e-05, | |
| "loss": 0.1273, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.37632081476766394, | |
| "grad_norm": 6.080224514007568, | |
| "learning_rate": 1.619125498629904e-05, | |
| "loss": 0.0465, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.3768300445576066, | |
| "grad_norm": 4.397584915161133, | |
| "learning_rate": 1.60745294221434e-05, | |
| "loss": 0.0625, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3773392743475493, | |
| "grad_norm": 8.622246742248535, | |
| "learning_rate": 1.595814548709983e-05, | |
| "loss": 0.0664, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.37784850413749205, | |
| "grad_norm": 8.313824653625488, | |
| "learning_rate": 1.5842104353153287e-05, | |
| "loss": 0.0271, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.37835773392743477, | |
| "grad_norm": 2.949397325515747, | |
| "learning_rate": 1.5726407188836673e-05, | |
| "loss": 0.0345, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.3788669637173775, | |
| "grad_norm": 3.27532696723938, | |
| "learning_rate": 1.5611055159219152e-05, | |
| "loss": 0.0561, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.37937619350732016, | |
| "grad_norm": 0.19333244860172272, | |
| "learning_rate": 1.549604942589441e-05, | |
| "loss": 0.0013, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.3798854232972629, | |
| "grad_norm": 4.911413192749023, | |
| "learning_rate": 1.5381391146968866e-05, | |
| "loss": 0.0187, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.3803946530872056, | |
| "grad_norm": 6.5160627365112305, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 0.0928, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.3809038828771483, | |
| "grad_norm": 5.048818111419678, | |
| "learning_rate": 1.5153121567235335e-05, | |
| "loss": 0.013, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.38141311266709105, | |
| "grad_norm": 4.852001667022705, | |
| "learning_rate": 1.5039512565099467e-05, | |
| "loss": 0.079, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.3819223424570337, | |
| "grad_norm": 3.08077335357666, | |
| "learning_rate": 1.4926255614683932e-05, | |
| "loss": 0.0403, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.38243157224697644, | |
| "grad_norm": 4.824317932128906, | |
| "learning_rate": 1.481335185648498e-05, | |
| "loss": 0.7268, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.38294080203691916, | |
| "grad_norm": 4.761933326721191, | |
| "learning_rate": 1.4700802427442179e-05, | |
| "loss": 0.5991, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.3834500318268619, | |
| "grad_norm": 4.185091495513916, | |
| "learning_rate": 1.458860846092705e-05, | |
| "loss": 0.2459, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.3839592616168046, | |
| "grad_norm": 10.620789527893066, | |
| "learning_rate": 1.4476771086731567e-05, | |
| "loss": 0.79, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.38446849140674727, | |
| "grad_norm": 11.214977264404297, | |
| "learning_rate": 1.4365291431056871e-05, | |
| "loss": 0.7926, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.38497772119669, | |
| "grad_norm": 5.3124895095825195, | |
| "learning_rate": 1.4254170616501827e-05, | |
| "loss": 0.4243, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.3854869509866327, | |
| "grad_norm": 11.870231628417969, | |
| "learning_rate": 1.414340976205183e-05, | |
| "loss": 0.6992, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.38599618077657544, | |
| "grad_norm": 9.170869827270508, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 0.8608, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.38650541056651816, | |
| "grad_norm": 10.806239128112793, | |
| "learning_rate": 1.3922972391273226e-05, | |
| "loss": 0.8629, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.3870146403564608, | |
| "grad_norm": 9.26331901550293, | |
| "learning_rate": 1.3813298094746491e-05, | |
| "loss": 0.5738, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.38752387014640355, | |
| "grad_norm": 10.236473083496094, | |
| "learning_rate": 1.3703988197906209e-05, | |
| "loss": 0.6671, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.38803309993634627, | |
| "grad_norm": 6.966268539428711, | |
| "learning_rate": 1.3595043801501794e-05, | |
| "loss": 0.7432, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.388542329726289, | |
| "grad_norm": 5.45960807800293, | |
| "learning_rate": 1.3486466002602133e-05, | |
| "loss": 0.5434, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.3890515595162317, | |
| "grad_norm": 1.5403192043304443, | |
| "learning_rate": 1.3378255894584463e-05, | |
| "loss": 0.1034, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.38956078930617444, | |
| "grad_norm": 1.8921170234680176, | |
| "learning_rate": 1.327041456712334e-05, | |
| "loss": 0.084, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.3900700190961171, | |
| "grad_norm": 4.200987815856934, | |
| "learning_rate": 1.3162943106179749e-05, | |
| "loss": 0.1189, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.3905792488860598, | |
| "grad_norm": 2.609253406524658, | |
| "learning_rate": 1.3055842593990131e-05, | |
| "loss": 0.2064, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.39108847867600255, | |
| "grad_norm": 2.868013858795166, | |
| "learning_rate": 1.2949114109055415e-05, | |
| "loss": 0.1238, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.39159770846594527, | |
| "grad_norm": 2.5417683124542236, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 0.0929, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.392106938255888, | |
| "grad_norm": 1.0600427389144897, | |
| "learning_rate": 1.2736777516212266e-05, | |
| "loss": 0.1323, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.39261616804583066, | |
| "grad_norm": 1.9886938333511353, | |
| "learning_rate": 1.2631171546530968e-05, | |
| "loss": 0.0904, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.3931253978357734, | |
| "grad_norm": 0.7915336489677429, | |
| "learning_rate": 1.2525941880537307e-05, | |
| "loss": 0.0964, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.3936346276257161, | |
| "grad_norm": 2.3213069438934326, | |
| "learning_rate": 1.2421089577892869e-05, | |
| "loss": 0.1248, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.3941438574156588, | |
| "grad_norm": 3.6811985969543457, | |
| "learning_rate": 1.2316615694459189e-05, | |
| "loss": 0.0861, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.39465308720560155, | |
| "grad_norm": 4.516244411468506, | |
| "learning_rate": 1.2212521282287092e-05, | |
| "loss": 0.1168, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.3951623169955442, | |
| "grad_norm": 4.109873294830322, | |
| "learning_rate": 1.2108807389606158e-05, | |
| "loss": 0.1002, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.39567154678548694, | |
| "grad_norm": 4.258279323577881, | |
| "learning_rate": 1.2005475060814159e-05, | |
| "loss": 0.1388, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.39618077657542966, | |
| "grad_norm": 3.9544007778167725, | |
| "learning_rate": 1.1902525336466464e-05, | |
| "loss": 0.1318, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.3966900063653724, | |
| "grad_norm": 5.926074981689453, | |
| "learning_rate": 1.1799959253265668e-05, | |
| "loss": 0.1125, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.3971992361553151, | |
| "grad_norm": 8.683218955993652, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.0995, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.39770846594525777, | |
| "grad_norm": 3.9073028564453125, | |
| "learning_rate": 1.1595982137788403e-05, | |
| "loss": 0.0903, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.3982176957352005, | |
| "grad_norm": 5.088838577270508, | |
| "learning_rate": 1.1494573159559213e-05, | |
| "loss": 0.1193, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.3987269255251432, | |
| "grad_norm": 5.782503128051758, | |
| "learning_rate": 1.1393551930550828e-05, | |
| "loss": 0.0823, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.39923615531508594, | |
| "grad_norm": 5.548112869262695, | |
| "learning_rate": 1.1292919468045877e-05, | |
| "loss": 0.0797, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.39974538510502866, | |
| "grad_norm": 6.514894962310791, | |
| "learning_rate": 1.1192676785412154e-05, | |
| "loss": 0.1448, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.4002546148949714, | |
| "grad_norm": 4.215356826782227, | |
| "learning_rate": 1.1092824892092373e-05, | |
| "loss": 0.0659, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.40076384468491405, | |
| "grad_norm": 7.318501949310303, | |
| "learning_rate": 1.099336479359398e-05, | |
| "loss": 0.1557, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.4012730744748568, | |
| "grad_norm": 4.508605003356934, | |
| "learning_rate": 1.0894297491479045e-05, | |
| "loss": 0.0383, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.4017823042647995, | |
| "grad_norm": 2.6559200286865234, | |
| "learning_rate": 1.0795623983354215e-05, | |
| "loss": 0.0212, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.4022915340547422, | |
| "grad_norm": 1.4912053346633911, | |
| "learning_rate": 1.0697345262860636e-05, | |
| "loss": 0.0104, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.40280076384468494, | |
| "grad_norm": 7.236735820770264, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 0.0706, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.4033099936346276, | |
| "grad_norm": 4.715954303741455, | |
| "learning_rate": 1.0501976139444191e-05, | |
| "loss": 0.048, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.40381922342457033, | |
| "grad_norm": 2.1985514163970947, | |
| "learning_rate": 1.0404887703886251e-05, | |
| "loss": 0.0125, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.40432845321451305, | |
| "grad_norm": 7.244698524475098, | |
| "learning_rate": 1.0308197990669538e-05, | |
| "loss": 0.0468, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.4048376830044558, | |
| "grad_norm": 1.7899662256240845, | |
| "learning_rate": 1.021190797345839e-05, | |
| "loss": 0.0033, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4053469127943985, | |
| "grad_norm": 7.417436122894287, | |
| "learning_rate": 1.0116018621892237e-05, | |
| "loss": 0.1219, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.40585614258434116, | |
| "grad_norm": 2.332493782043457, | |
| "learning_rate": 1.0020530901575754e-05, | |
| "loss": 0.0177, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.4063653723742839, | |
| "grad_norm": 6.933743000030518, | |
| "learning_rate": 9.92544577406923e-06, | |
| "loss": 0.0993, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.4068746021642266, | |
| "grad_norm": 6.604279518127441, | |
| "learning_rate": 9.830764196878872e-06, | |
| "loss": 0.0523, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.40738383195416933, | |
| "grad_norm": 1.6020522117614746, | |
| "learning_rate": 9.73648712344707e-06, | |
| "loss": 0.0222, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.40738383195416933, | |
| "eval_loss": 0.23512445390224457, | |
| "eval_runtime": 376.2064, | |
| "eval_samples_per_second": 8.793, | |
| "eval_steps_per_second": 2.198, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.40789306174411205, | |
| "grad_norm": 4.510726451873779, | |
| "learning_rate": 9.642615503142926e-06, | |
| "loss": 0.7091, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.4084022915340547, | |
| "grad_norm": 4.7646942138671875, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.4744, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.40891152132399744, | |
| "grad_norm": 6.916759014129639, | |
| "learning_rate": 9.456092398969902e-06, | |
| "loss": 0.6527, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.40942075111394016, | |
| "grad_norm": 4.222986221313477, | |
| "learning_rate": 9.363442793386606e-06, | |
| "loss": 0.3524, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.4099299809038829, | |
| "grad_norm": 8.524075508117676, | |
| "learning_rate": 9.271202397483215e-06, | |
| "loss": 0.9119, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.4104392106938256, | |
| "grad_norm": 12.714594841003418, | |
| "learning_rate": 9.179372140119525e-06, | |
| "loss": 0.4737, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.4109484404837683, | |
| "grad_norm": 8.219633102416992, | |
| "learning_rate": 9.087952946025175e-06, | |
| "loss": 0.4039, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.411457670273711, | |
| "grad_norm": 11.758675575256348, | |
| "learning_rate": 8.996945735790447e-06, | |
| "loss": 0.8365, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.4119669000636537, | |
| "grad_norm": 10.59483528137207, | |
| "learning_rate": 8.906351425856952e-06, | |
| "loss": 1.0241, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.41247612985359644, | |
| "grad_norm": 5.28472375869751, | |
| "learning_rate": 8.816170928508365e-06, | |
| "loss": 0.5595, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.41298535964353916, | |
| "grad_norm": 8.57630729675293, | |
| "learning_rate": 8.7264051518613e-06, | |
| "loss": 0.5879, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.4134945894334819, | |
| "grad_norm": 8.637681007385254, | |
| "learning_rate": 8.637054999856148e-06, | |
| "loss": 0.8988, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.41400381922342455, | |
| "grad_norm": 1.3487462997436523, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 0.1034, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.4145130490133673, | |
| "grad_norm": 0.9683374762535095, | |
| "learning_rate": 8.459605164597267e-06, | |
| "loss": 0.0647, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.41502227880331, | |
| "grad_norm": 0.556820273399353, | |
| "learning_rate": 8.371507268261437e-06, | |
| "loss": 0.0749, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.4155315085932527, | |
| "grad_norm": 2.4960896968841553, | |
| "learning_rate": 8.283828570385238e-06, | |
| "loss": 0.1691, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.41604073838319544, | |
| "grad_norm": 1.6859462261199951, | |
| "learning_rate": 8.196569953892202e-06, | |
| "loss": 0.1085, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.4165499681731381, | |
| "grad_norm": 3.0034708976745605, | |
| "learning_rate": 8.109732297475635e-06, | |
| "loss": 0.0917, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.41705919796308083, | |
| "grad_norm": 1.4022327661514282, | |
| "learning_rate": 8.023316475589754e-06, | |
| "loss": 0.1208, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.41756842775302355, | |
| "grad_norm": 0.8047292232513428, | |
| "learning_rate": 7.937323358440935e-06, | |
| "loss": 0.0582, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4180776575429663, | |
| "grad_norm": 1.763879656791687, | |
| "learning_rate": 7.851753811978924e-06, | |
| "loss": 0.1686, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.418586887332909, | |
| "grad_norm": 1.504451036453247, | |
| "learning_rate": 7.766608697888095e-06, | |
| "loss": 0.0718, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.41909611712285166, | |
| "grad_norm": 1.7512840032577515, | |
| "learning_rate": 7.681888873578786e-06, | |
| "loss": 0.1577, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.4196053469127944, | |
| "grad_norm": 10.731456756591797, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 0.1255, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.4201145767027371, | |
| "grad_norm": 4.191404819488525, | |
| "learning_rate": 7.513728502524286e-06, | |
| "loss": 0.1061, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.42062380649267983, | |
| "grad_norm": 6.975319862365723, | |
| "learning_rate": 7.430289649152156e-06, | |
| "loss": 0.1258, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.42113303628262255, | |
| "grad_norm": 5.663224697113037, | |
| "learning_rate": 7.347279472290647e-06, | |
| "loss": 0.1082, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.4216422660725652, | |
| "grad_norm": 6.514274597167969, | |
| "learning_rate": 7.264698807851328e-06, | |
| "loss": 0.1209, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.42215149586250794, | |
| "grad_norm": 5.257199764251709, | |
| "learning_rate": 7.182548487420554e-06, | |
| "loss": 0.0995, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.42266072565245066, | |
| "grad_norm": 3.9015791416168213, | |
| "learning_rate": 7.100829338251147e-06, | |
| "loss": 0.1605, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4231699554423934, | |
| "grad_norm": 5.840038776397705, | |
| "learning_rate": 7.019542183254046e-06, | |
| "loss": 0.1126, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.4236791852323361, | |
| "grad_norm": 11.773760795593262, | |
| "learning_rate": 6.9386878409899715e-06, | |
| "loss": 0.1449, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.42418841502227883, | |
| "grad_norm": 6.619400978088379, | |
| "learning_rate": 6.858267125661272e-06, | |
| "loss": 0.1207, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.4246976448122215, | |
| "grad_norm": 5.580018520355225, | |
| "learning_rate": 6.778280847103669e-06, | |
| "loss": 0.0902, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.4252068746021642, | |
| "grad_norm": 5.762646675109863, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.0773, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.42571610439210694, | |
| "grad_norm": 5.908243179321289, | |
| "learning_rate": 6.619614817762537e-06, | |
| "loss": 0.0891, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.42622533418204966, | |
| "grad_norm": 4.759620666503906, | |
| "learning_rate": 6.540936664744196e-06, | |
| "loss": 0.092, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.4267345639719924, | |
| "grad_norm": 3.2303318977355957, | |
| "learning_rate": 6.462696144011149e-06, | |
| "loss": 0.0409, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.42724379376193505, | |
| "grad_norm": 4.1941986083984375, | |
| "learning_rate": 6.384894043444567e-06, | |
| "loss": 0.1015, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.4277530235518778, | |
| "grad_norm": 2.3727965354919434, | |
| "learning_rate": 6.3075311465107535e-06, | |
| "loss": 0.0168, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4282622533418205, | |
| "grad_norm": 6.094393730163574, | |
| "learning_rate": 6.230608232253227e-06, | |
| "loss": 0.098, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.4287714831317632, | |
| "grad_norm": 6.916983127593994, | |
| "learning_rate": 6.154126075284855e-06, | |
| "loss": 0.0278, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.42928071292170594, | |
| "grad_norm": 1.7201404571533203, | |
| "learning_rate": 6.078085445780129e-06, | |
| "loss": 0.0032, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.4297899427116486, | |
| "grad_norm": 9.973702430725098, | |
| "learning_rate": 6.002487109467347e-06, | |
| "loss": 0.0374, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.43029917250159133, | |
| "grad_norm": 1.6437783241271973, | |
| "learning_rate": 5.927331827620903e-06, | |
| "loss": 0.0081, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.43080840229153405, | |
| "grad_norm": 4.856413841247559, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 0.0392, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.4313176320814768, | |
| "grad_norm": 5.264596939086914, | |
| "learning_rate": 5.778353450109286e-06, | |
| "loss": 0.0289, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.4318268618714195, | |
| "grad_norm": 4.01646089553833, | |
| "learning_rate": 5.704531854654721e-06, | |
| "loss": 0.0063, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.43233609166136217, | |
| "grad_norm": 5.703643798828125, | |
| "learning_rate": 5.631156314072605e-06, | |
| "loss": 0.0557, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.4328453214513049, | |
| "grad_norm": 9.367152214050293, | |
| "learning_rate": 5.558227567253832e-06, | |
| "loss": 0.1552, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4333545512412476, | |
| "grad_norm": 4.262320518493652, | |
| "learning_rate": 5.485746348590048e-06, | |
| "loss": 0.6515, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.43386378103119033, | |
| "grad_norm": 5.231907367706299, | |
| "learning_rate": 5.413713387966329e-06, | |
| "loss": 0.5992, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.43437301082113305, | |
| "grad_norm": 4.903708457946777, | |
| "learning_rate": 5.34212941075381e-06, | |
| "loss": 0.3601, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.4348822406110757, | |
| "grad_norm": 6.984781742095947, | |
| "learning_rate": 5.270995137802315e-06, | |
| "loss": 0.5169, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.43539147040101844, | |
| "grad_norm": 7.836650371551514, | |
| "learning_rate": 5.200311285433213e-06, | |
| "loss": 0.6677, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.43590070019096117, | |
| "grad_norm": 5.570544719696045, | |
| "learning_rate": 5.13007856543209e-06, | |
| "loss": 0.3495, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.4364099299809039, | |
| "grad_norm": 21.38435935974121, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 1.0934, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.4369191597708466, | |
| "grad_norm": 11.60948371887207, | |
| "learning_rate": 4.99096934695461e-06, | |
| "loss": 0.8771, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.43742838956078933, | |
| "grad_norm": 11.168967247009277, | |
| "learning_rate": 4.922094249306558e-06, | |
| "loss": 0.822, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.437937619350732, | |
| "grad_norm": 9.118983268737793, | |
| "learning_rate": 4.853673085668947e-06, | |
| "loss": 0.7252, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4384468491406747, | |
| "grad_norm": 8.641736030578613, | |
| "learning_rate": 4.78570654504214e-06, | |
| "loss": 0.8113, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.43895607893061744, | |
| "grad_norm": 7.092012882232666, | |
| "learning_rate": 4.7181953118484556e-06, | |
| "loss": 0.6584, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.43946530872056017, | |
| "grad_norm": 1.0570333003997803, | |
| "learning_rate": 4.651140065925269e-06, | |
| "loss": 0.1191, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.4399745385105029, | |
| "grad_norm": 2.1120429039001465, | |
| "learning_rate": 4.58454148251814e-06, | |
| "loss": 0.1825, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.44048376830044556, | |
| "grad_norm": 0.5101509094238281, | |
| "learning_rate": 4.5184002322740785e-06, | |
| "loss": 0.0776, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.4409929980903883, | |
| "grad_norm": 1.4004946947097778, | |
| "learning_rate": 4.452716981234744e-06, | |
| "loss": 0.1078, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.441502227880331, | |
| "grad_norm": 1.1376584768295288, | |
| "learning_rate": 4.387492390829734e-06, | |
| "loss": 0.0968, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.4420114576702737, | |
| "grad_norm": 1.968505620956421, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.1204, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.44252068746021644, | |
| "grad_norm": 0.5514800548553467, | |
| "learning_rate": 4.258421814540992e-06, | |
| "loss": 0.0586, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.4430299172501591, | |
| "grad_norm": 2.0910208225250244, | |
| "learning_rate": 4.19457712839652e-06, | |
| "loss": 0.1634, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.44353914704010183, | |
| "grad_norm": 3.2385759353637695, | |
| "learning_rate": 4.131193702351827e-06, | |
| "loss": 0.1482, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.44404837683004456, | |
| "grad_norm": 1.8257766962051392, | |
| "learning_rate": 4.068272174677335e-06, | |
| "loss": 0.1311, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.4445576066199873, | |
| "grad_norm": 1.4228204488754272, | |
| "learning_rate": 4.005813178992091e-06, | |
| "loss": 0.0839, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.44506683640993, | |
| "grad_norm": 14.156405448913574, | |
| "learning_rate": 3.9438173442575e-06, | |
| "loss": 0.1061, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.44557606619987267, | |
| "grad_norm": 7.2103047370910645, | |
| "learning_rate": 3.8822852947709375e-06, | |
| "loss": 0.1517, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.4460852959898154, | |
| "grad_norm": 4.403718948364258, | |
| "learning_rate": 3.821217650159453e-06, | |
| "loss": 0.1108, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.4465945257797581, | |
| "grad_norm": 4.354316711425781, | |
| "learning_rate": 3.760615025373543e-06, | |
| "loss": 0.1314, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.44710375556970083, | |
| "grad_norm": 3.967412233352661, | |
| "learning_rate": 3.700478030680987e-06, | |
| "loss": 0.1001, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.44761298535964356, | |
| "grad_norm": 4.008363246917725, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 0.0968, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.4481222151495863, | |
| "grad_norm": 6.099944114685059, | |
| "learning_rate": 3.581603349196372e-06, | |
| "loss": 0.1582, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.44863144493952894, | |
| "grad_norm": 5.96859073638916, | |
| "learning_rate": 3.522866859471047e-06, | |
| "loss": 0.1445, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.44914067472947167, | |
| "grad_norm": 3.3854691982269287, | |
| "learning_rate": 3.4645983939604496e-06, | |
| "loss": 0.0567, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.4496499045194144, | |
| "grad_norm": 3.6964199542999268, | |
| "learning_rate": 3.406798539427386e-06, | |
| "loss": 0.1331, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.4501591343093571, | |
| "grad_norm": 3.9392178058624268, | |
| "learning_rate": 3.349467877915746e-06, | |
| "loss": 0.1401, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.45066836409929983, | |
| "grad_norm": 8.413680076599121, | |
| "learning_rate": 3.2926069867446675e-06, | |
| "loss": 0.0784, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.4511775938892425, | |
| "grad_norm": 3.3428969383239746, | |
| "learning_rate": 3.2362164385026706e-06, | |
| "loss": 0.1658, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.4516868236791852, | |
| "grad_norm": 4.6566386222839355, | |
| "learning_rate": 3.180296801041971e-06, | |
| "loss": 0.1105, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.45219605346912795, | |
| "grad_norm": 4.742708683013916, | |
| "learning_rate": 3.1248486374726883e-06, | |
| "loss": 0.0988, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.45270528325907067, | |
| "grad_norm": 5.364091873168945, | |
| "learning_rate": 3.069872506157212e-06, | |
| "loss": 0.0267, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.4532145130490134, | |
| "grad_norm": 7.658147811889648, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.1993, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.45372374283895606, | |
| "grad_norm": 5.430888652801514, | |
| "learning_rate": 2.961338549964893e-06, | |
| "loss": 0.0843, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.4542329726288988, | |
| "grad_norm": 4.619664669036865, | |
| "learning_rate": 2.9077818180237693e-06, | |
| "loss": 0.0777, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.4547422024188415, | |
| "grad_norm": 0.4149484932422638, | |
| "learning_rate": 2.8546993041969173e-06, | |
| "loss": 0.0021, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.4552514322087842, | |
| "grad_norm": 3.0641720294952393, | |
| "learning_rate": 2.802091543024671e-06, | |
| "loss": 0.0088, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.45576066199872695, | |
| "grad_norm": 3.299743175506592, | |
| "learning_rate": 2.7499590642665774e-06, | |
| "loss": 0.0406, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.4562698917886696, | |
| "grad_norm": 5.375297546386719, | |
| "learning_rate": 2.6983023928961404e-06, | |
| "loss": 0.0383, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.45677912157861233, | |
| "grad_norm": 1.769721508026123, | |
| "learning_rate": 2.647122049095463e-06, | |
| "loss": 0.0077, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.45728835136855506, | |
| "grad_norm": 7.771859645843506, | |
| "learning_rate": 2.596418548250029e-06, | |
| "loss": 0.0876, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.4577975811584978, | |
| "grad_norm": 5.935695171356201, | |
| "learning_rate": 2.546192400943537e-06, | |
| "loss": 0.0387, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.4583068109484405, | |
| "grad_norm": 5.693915367126465, | |
| "learning_rate": 2.496444112952734e-06, | |
| "loss": 0.1489, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.45881604073838317, | |
| "grad_norm": 3.7777862548828125, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.6961, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.4593252705283259, | |
| "grad_norm": 5.661333084106445, | |
| "learning_rate": 2.3983831139599287e-06, | |
| "loss": 0.6676, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.4598345003182686, | |
| "grad_norm": 3.3415160179138184, | |
| "learning_rate": 2.3500713904311024e-06, | |
| "loss": 0.1848, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.46034373010821134, | |
| "grad_norm": 9.008618354797363, | |
| "learning_rate": 2.3022395011543686e-06, | |
| "loss": 0.8188, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.46085295989815406, | |
| "grad_norm": 6.555294036865234, | |
| "learning_rate": 2.2548879277963064e-06, | |
| "loss": 0.5172, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.4613621896880968, | |
| "grad_norm": 15.021227836608887, | |
| "learning_rate": 2.208017147186736e-06, | |
| "loss": 0.6723, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.46187141947803945, | |
| "grad_norm": 9.838711738586426, | |
| "learning_rate": 2.161627631313923e-06, | |
| "loss": 0.7112, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.46238064926798217, | |
| "grad_norm": 12.633716583251953, | |
| "learning_rate": 2.1157198473197414e-06, | |
| "loss": 0.9927, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.4628898790579249, | |
| "grad_norm": 7.977221488952637, | |
| "learning_rate": 2.070294257495081e-06, | |
| "loss": 0.558, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.4633991088478676, | |
| "grad_norm": 9.081966400146484, | |
| "learning_rate": 2.0253513192751373e-06, | |
| "loss": 0.5821, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.46390833863781034, | |
| "grad_norm": 8.361538887023926, | |
| "learning_rate": 1.9808914852347813e-06, | |
| "loss": 0.9417, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.464417568427753, | |
| "grad_norm": 2.8661348819732666, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 0.3311, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.4649267982176957, | |
| "grad_norm": 1.429822564125061, | |
| "learning_rate": 1.8934229156636452e-06, | |
| "loss": 0.1388, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.46543602800763845, | |
| "grad_norm": 1.2748867273330688, | |
| "learning_rate": 1.8504150609403858e-06, | |
| "loss": 0.1259, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.46594525779758117, | |
| "grad_norm": 0.8820096850395203, | |
| "learning_rate": 1.807892072002898e-06, | |
| "loss": 0.0854, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.4664544875875239, | |
| "grad_norm": 2.467775344848633, | |
| "learning_rate": 1.7658543770572189e-06, | |
| "loss": 0.0945, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.46696371737746656, | |
| "grad_norm": 4.9203877449035645, | |
| "learning_rate": 1.724302399422456e-06, | |
| "loss": 0.1121, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.4674729471674093, | |
| "grad_norm": 0.4738912284374237, | |
| "learning_rate": 1.6832365575265741e-06, | |
| "loss": 0.0745, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.467982176957352, | |
| "grad_norm": 1.508836269378662, | |
| "learning_rate": 1.6426572649021476e-06, | |
| "loss": 0.1326, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.4684914067472947, | |
| "grad_norm": 1.2223467826843262, | |
| "learning_rate": 1.6025649301821876e-06, | |
| "loss": 0.1413, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.46900063653723745, | |
| "grad_norm": 0.6743190288543701, | |
| "learning_rate": 1.5629599570960718e-06, | |
| "loss": 0.0968, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.4695098663271801, | |
| "grad_norm": 1.4873098134994507, | |
| "learning_rate": 1.523842744465437e-06, | |
| "loss": 0.102, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.47001909611712284, | |
| "grad_norm": 1.1552408933639526, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 0.0608, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.47052832590706556, | |
| "grad_norm": 5.109307289123535, | |
| "learning_rate": 1.4470731712944884e-06, | |
| "loss": 0.0851, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.4710375556970083, | |
| "grad_norm": 4.766353130340576, | |
| "learning_rate": 1.4094215838229176e-06, | |
| "loss": 0.1149, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.471546785486951, | |
| "grad_norm": 14.194299697875977, | |
| "learning_rate": 1.372259302936546e-06, | |
| "loss": 0.2078, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.4720560152768937, | |
| "grad_norm": 4.660562038421631, | |
| "learning_rate": 1.3355867028591208e-06, | |
| "loss": 0.0858, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.4725652450668364, | |
| "grad_norm": 10.358325004577637, | |
| "learning_rate": 1.2994041528833266e-06, | |
| "loss": 0.0983, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.4730744748567791, | |
| "grad_norm": 6.56409215927124, | |
| "learning_rate": 1.2637120173670358e-06, | |
| "loss": 0.0809, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.47358370464672184, | |
| "grad_norm": 4.377633094787598, | |
| "learning_rate": 1.2285106557296477e-06, | |
| "loss": 0.094, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.47409293443666456, | |
| "grad_norm": 5.809483051300049, | |
| "learning_rate": 1.1938004224484988e-06, | |
| "loss": 0.1196, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.4746021642266073, | |
| "grad_norm": 3.2537131309509277, | |
| "learning_rate": 1.1595816670552428e-06, | |
| "loss": 0.1031, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.47511139401654995, | |
| "grad_norm": 6.928438186645508, | |
| "learning_rate": 1.1258547341323699e-06, | |
| "loss": 0.068, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.47562062380649267, | |
| "grad_norm": 3.240041971206665, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.0739, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.4761298535964354, | |
| "grad_norm": 6.833902835845947, | |
| "learning_rate": 1.0598776892610685e-06, | |
| "loss": 0.0857, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.4766390833863781, | |
| "grad_norm": 5.259301662445068, | |
| "learning_rate": 1.02762824170074e-06, | |
| "loss": 0.1534, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.47714831317632084, | |
| "grad_norm": 11.154175758361816, | |
| "learning_rate": 9.958719453803278e-07, | |
| "loss": 0.1273, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.4776575429662635, | |
| "grad_norm": 2.975895404815674, | |
| "learning_rate": 9.646091200853802e-07, | |
| "loss": 0.015, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.4781667727562062, | |
| "grad_norm": 3.766637086868286, | |
| "learning_rate": 9.338400806321978e-07, | |
| "loss": 0.0202, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.47867600254614895, | |
| "grad_norm": 2.1836998462677, | |
| "learning_rate": 9.035651368646648e-07, | |
| "loss": 0.0426, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.47918523233609167, | |
| "grad_norm": 3.746504306793213, | |
| "learning_rate": 8.737845936511335e-07, | |
| "loss": 0.0458, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.4796944621260344, | |
| "grad_norm": 4.881604194641113, | |
| "learning_rate": 8.444987508813451e-07, | |
| "loss": 0.0482, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.48020369191597706, | |
| "grad_norm": 15.250972747802734, | |
| "learning_rate": 8.157079034633974e-07, | |
| "loss": 0.0486, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.4807129217059198, | |
| "grad_norm": 4.671970844268799, | |
| "learning_rate": 7.874123413208145e-07, | |
| "loss": 0.0408, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.4812221514958625, | |
| "grad_norm": 3.882068157196045, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 0.0702, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.4817313812858052, | |
| "grad_norm": 3.0033390522003174, | |
| "learning_rate": 7.323082076153509e-07, | |
| "loss": 0.0722, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.48224061107574795, | |
| "grad_norm": 16.160083770751953, | |
| "learning_rate": 7.055001909504755e-07, | |
| "loss": 0.0567, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.4827498408656906, | |
| "grad_norm": 5.195967674255371, | |
| "learning_rate": 6.791885693514133e-07, | |
| "loss": 0.0777, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.48325907065563334, | |
| "grad_norm": 15.034565925598145, | |
| "learning_rate": 6.533736077758868e-07, | |
| "loss": 0.073, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.48376830044557606, | |
| "grad_norm": 3.8854053020477295, | |
| "learning_rate": 6.280555661802856e-07, | |
| "loss": 0.0394, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.4842775302355188, | |
| "grad_norm": 3.7877042293548584, | |
| "learning_rate": 6.032346995169968e-07, | |
| "loss": 0.7431, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.4847867600254615, | |
| "grad_norm": 11.732136726379395, | |
| "learning_rate": 5.78911257731879e-07, | |
| "loss": 0.6649, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.4852959898154042, | |
| "grad_norm": 5.163148403167725, | |
| "learning_rate": 5.550854857617193e-07, | |
| "loss": 0.4255, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.4858052196053469, | |
| "grad_norm": 7.868194580078125, | |
| "learning_rate": 5.317576235317756e-07, | |
| "loss": 0.7605, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.4863144493952896, | |
| "grad_norm": 7.837756633758545, | |
| "learning_rate": 5.089279059533658e-07, | |
| "loss": 0.7603, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.48682367918523234, | |
| "grad_norm": 6.926958084106445, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 0.3583, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.48733290897517506, | |
| "grad_norm": 6.891408920288086, | |
| "learning_rate": 4.647638193125137e-07, | |
| "loss": 0.5063, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.4878421387651178, | |
| "grad_norm": 17.593307495117188, | |
| "learning_rate": 4.434298949819449e-07, | |
| "loss": 1.0916, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.48835136855506045, | |
| "grad_norm": 13.12249755859375, | |
| "learning_rate": 4.2259500476214407e-07, | |
| "loss": 1.178, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.48886059834500317, | |
| "grad_norm": 8.691963195800781, | |
| "learning_rate": 4.02259358460233e-07, | |
| "loss": 0.6026, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.4893698281349459, | |
| "grad_norm": 5.080837726593018, | |
| "learning_rate": 3.824231608559492e-07, | |
| "loss": 0.379, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.4898790579248886, | |
| "grad_norm": 10.101880073547363, | |
| "learning_rate": 3.630866116995757e-07, | |
| "loss": 0.8812, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.49038828771483134, | |
| "grad_norm": 2.2711408138275146, | |
| "learning_rate": 3.4424990570994797e-07, | |
| "loss": 0.1018, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.490897517504774, | |
| "grad_norm": 0.9632775187492371, | |
| "learning_rate": 3.2591323257248893e-07, | |
| "loss": 0.1066, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.49140674729471673, | |
| "grad_norm": 2.8695173263549805, | |
| "learning_rate": 3.080767769372939e-07, | |
| "loss": 0.1316, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.49191597708465945, | |
| "grad_norm": 1.3276772499084473, | |
| "learning_rate": 2.907407184172706e-07, | |
| "loss": 0.0816, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.4924252068746022, | |
| "grad_norm": 4.322078227996826, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 0.1199, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.4929344366645449, | |
| "grad_norm": 0.6163370013237, | |
| "learning_rate": 2.5757048597765396e-07, | |
| "loss": 0.0881, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.49344366645448756, | |
| "grad_norm": 0.8970826864242554, | |
| "learning_rate": 2.4173664608193593e-07, | |
| "loss": 0.095, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.4939528962444303, | |
| "grad_norm": 1.772135615348816, | |
| "learning_rate": 2.2640387134577058e-07, | |
| "loss": 0.0945, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.494462126034373, | |
| "grad_norm": 1.8468323945999146, | |
| "learning_rate": 2.1157231617002783e-07, | |
| "loss": 0.136, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.49497135582431573, | |
| "grad_norm": 2.6759517192840576, | |
| "learning_rate": 1.9724212990830938e-07, | |
| "loss": 0.138, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.49548058561425845, | |
| "grad_norm": 0.8388791680335999, | |
| "learning_rate": 1.8341345686543332e-07, | |
| "loss": 0.1131, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.4959898154042012, | |
| "grad_norm": 5.01349401473999, | |
| "learning_rate": 1.7008643629596866e-07, | |
| "loss": 0.104, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.49649904519414384, | |
| "grad_norm": 8.658512115478516, | |
| "learning_rate": 1.5726120240288634e-07, | |
| "loss": 0.088, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.49700827498408656, | |
| "grad_norm": 10.031734466552734, | |
| "learning_rate": 1.449378843361271e-07, | |
| "loss": 0.1412, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.4975175047740293, | |
| "grad_norm": 5.323798656463623, | |
| "learning_rate": 1.3311660619138578e-07, | |
| "loss": 0.0799, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.498026734563972, | |
| "grad_norm": 14.896018981933594, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 0.0547, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.49853596435391473, | |
| "grad_norm": 2.8598153591156006, | |
| "learning_rate": 1.109806407717462e-07, | |
| "loss": 0.057, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.4990451941438574, | |
| "grad_norm": 4.470668792724609, | |
| "learning_rate": 1.0066617640578368e-07, | |
| "loss": 0.1221, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.4995544239338001, | |
| "grad_norm": 6.321918964385986, | |
| "learning_rate": 9.085419777743465e-08, | |
| "loss": 0.1274, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.5000636537237428, | |
| "grad_norm": 4.125798225402832, | |
| "learning_rate": 8.15448036932176e-08, | |
| "loss": 0.0822, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.5005728835136856, | |
| "grad_norm": 5.414987564086914, | |
| "learning_rate": 7.273808789862724e-08, | |
| "loss": 0.0796, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.5010821133036283, | |
| "grad_norm": 5.301061153411865, | |
| "learning_rate": 6.443413907720186e-08, | |
| "loss": 0.1011, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.501591343093571, | |
| "grad_norm": 6.373870849609375, | |
| "learning_rate": 5.663304084960186e-08, | |
| "loss": 0.1335, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5021005728835137, | |
| "grad_norm": 3.59035062789917, | |
| "learning_rate": 4.933487177280482e-08, | |
| "loss": 0.06, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.5026098026734565, | |
| "grad_norm": 4.328834533691406, | |
| "learning_rate": 4.253970533929508e-08, | |
| "loss": 0.0565, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.5031190324633991, | |
| "grad_norm": 2.652255058288574, | |
| "learning_rate": 3.624760997631982e-08, | |
| "loss": 0.0428, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.5036282622533418, | |
| "grad_norm": 4.667551040649414, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 0.0206, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.5041374920432845, | |
| "grad_norm": 7.6786112785339355, | |
| "learning_rate": 2.5172880840745873e-08, | |
| "loss": 0.1095, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5046467218332272, | |
| "grad_norm": 3.08980131149292, | |
| "learning_rate": 2.0390358590538504e-08, | |
| "loss": 0.0337, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.50515595162317, | |
| "grad_norm": 2.030238389968872, | |
| "learning_rate": 1.6111130454543598e-08, | |
| "loss": 0.0064, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.5056651814131127, | |
| "grad_norm": 0.7045100331306458, | |
| "learning_rate": 1.2335239524541299e-08, | |
| "loss": 0.0105, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.5061744112030554, | |
| "grad_norm": 8.8118896484375, | |
| "learning_rate": 9.06272382371065e-09, | |
| "loss": 0.0789, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.5066836409929981, | |
| "grad_norm": 8.802488327026367, | |
| "learning_rate": 6.293616306246586e-09, | |
| "loss": 0.0811, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.5071928707829408, | |
| "grad_norm": 4.467785358428955, | |
| "learning_rate": 4.0279448570323954e-09, | |
| "loss": 0.0437, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.5077021005728836, | |
| "grad_norm": 8.837767601013184, | |
| "learning_rate": 2.265732291356626e-09, | |
| "loss": 0.0425, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.5082113303628262, | |
| "grad_norm": 7.292797565460205, | |
| "learning_rate": 1.0069963546743832e-09, | |
| "loss": 0.0808, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.5087205601527689, | |
| "grad_norm": 3.702165365219116, | |
| "learning_rate": 2.5174972244634833e-10, | |
| "loss": 0.0469, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.5092297899427116, | |
| "grad_norm": 8.40251636505127, | |
| "learning_rate": 0.0, | |
| "loss": 0.0725, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5092297899427116, | |
| "eval_loss": 0.22850316762924194, | |
| "eval_runtime": 378.274, | |
| "eval_samples_per_second": 8.745, | |
| "eval_steps_per_second": 2.186, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.0800300944866673e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |