| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 922, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0021691973969631237, | |
| "grad_norm": 0.18814353593860253, | |
| "learning_rate": 1.0752688172043011e-06, | |
| "loss": 0.7804, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004338394793926247, | |
| "grad_norm": 0.14783898318898897, | |
| "learning_rate": 2.1505376344086023e-06, | |
| "loss": 0.6624, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006507592190889371, | |
| "grad_norm": 0.19503814642198758, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.7092, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.008676789587852495, | |
| "grad_norm": 0.1618902610161204, | |
| "learning_rate": 4.3010752688172045e-06, | |
| "loss": 0.7077, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.010845986984815618, | |
| "grad_norm": 0.12013865404820073, | |
| "learning_rate": 5.376344086021506e-06, | |
| "loss": 0.6043, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.013015184381778741, | |
| "grad_norm": 0.1374084610184008, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.6217, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.015184381778741865, | |
| "grad_norm": 0.14626420092970213, | |
| "learning_rate": 7.526881720430108e-06, | |
| "loss": 0.6813, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01735357917570499, | |
| "grad_norm": 0.13817440123812896, | |
| "learning_rate": 8.602150537634409e-06, | |
| "loss": 0.5804, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.019522776572668113, | |
| "grad_norm": 0.15287513050585858, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.623, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.021691973969631236, | |
| "grad_norm": 0.1598803439887024, | |
| "learning_rate": 1.0752688172043012e-05, | |
| "loss": 0.6743, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02386117136659436, | |
| "grad_norm": 0.19385413161643053, | |
| "learning_rate": 1.1827956989247313e-05, | |
| "loss": 0.7609, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.026030368763557483, | |
| "grad_norm": 0.22754758020045657, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.6324, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.028199566160520606, | |
| "grad_norm": 0.17059715163003164, | |
| "learning_rate": 1.3978494623655914e-05, | |
| "loss": 0.6333, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03036876355748373, | |
| "grad_norm": 0.18873401138092946, | |
| "learning_rate": 1.5053763440860215e-05, | |
| "loss": 0.7387, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03253796095444685, | |
| "grad_norm": 0.18496670943100807, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.6586, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03470715835140998, | |
| "grad_norm": 0.17005766008579865, | |
| "learning_rate": 1.7204301075268818e-05, | |
| "loss": 0.6018, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0368763557483731, | |
| "grad_norm": 0.19850157253361694, | |
| "learning_rate": 1.827956989247312e-05, | |
| "loss": 0.7417, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.039045553145336226, | |
| "grad_norm": 0.16224315084212368, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.4922, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04121475054229935, | |
| "grad_norm": 0.17598848375010048, | |
| "learning_rate": 2.0430107526881722e-05, | |
| "loss": 0.6445, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04338394793926247, | |
| "grad_norm": 0.18904672614562715, | |
| "learning_rate": 2.1505376344086024e-05, | |
| "loss": 0.6504, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0455531453362256, | |
| "grad_norm": 0.17265224349161137, | |
| "learning_rate": 2.258064516129032e-05, | |
| "loss": 0.6245, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04772234273318872, | |
| "grad_norm": 0.2017634897229929, | |
| "learning_rate": 2.3655913978494626e-05, | |
| "loss": 0.6795, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.049891540130151846, | |
| "grad_norm": 0.18337107237337277, | |
| "learning_rate": 2.4731182795698928e-05, | |
| "loss": 0.5461, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.052060737527114966, | |
| "grad_norm": 0.1819879312384368, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.6217, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.05422993492407809, | |
| "grad_norm": 0.1796615077383316, | |
| "learning_rate": 2.6881720430107527e-05, | |
| "loss": 0.6185, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05639913232104121, | |
| "grad_norm": 0.154004978022556, | |
| "learning_rate": 2.7956989247311828e-05, | |
| "loss": 0.5814, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05856832971800434, | |
| "grad_norm": 0.15691846371505647, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.5873, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06073752711496746, | |
| "grad_norm": 0.13885516449257118, | |
| "learning_rate": 3.010752688172043e-05, | |
| "loss": 0.5345, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06290672451193059, | |
| "grad_norm": 0.15573868807503372, | |
| "learning_rate": 3.118279569892473e-05, | |
| "loss": 0.5894, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0650759219088937, | |
| "grad_norm": 0.14860751998326854, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.5983, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06724511930585683, | |
| "grad_norm": 0.12573553934841997, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.5542, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.06941431670281996, | |
| "grad_norm": 0.12949334468283782, | |
| "learning_rate": 3.4408602150537636e-05, | |
| "loss": 0.5369, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07158351409978309, | |
| "grad_norm": 0.1541384187233925, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.6545, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0737527114967462, | |
| "grad_norm": 0.1342436339209465, | |
| "learning_rate": 3.655913978494624e-05, | |
| "loss": 0.5753, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07592190889370933, | |
| "grad_norm": 0.13687890281351986, | |
| "learning_rate": 3.763440860215054e-05, | |
| "loss": 0.5251, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07809110629067245, | |
| "grad_norm": 0.14224847495035575, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.5494, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08026030368763558, | |
| "grad_norm": 0.1368222255571023, | |
| "learning_rate": 3.978494623655914e-05, | |
| "loss": 0.5793, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0824295010845987, | |
| "grad_norm": 0.14032242415063437, | |
| "learning_rate": 4.0860215053763444e-05, | |
| "loss": 0.5738, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08459869848156182, | |
| "grad_norm": 0.1077744610907224, | |
| "learning_rate": 4.1935483870967746e-05, | |
| "loss": 0.4712, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.08676789587852494, | |
| "grad_norm": 0.1269827092811396, | |
| "learning_rate": 4.301075268817205e-05, | |
| "loss": 0.5071, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08893709327548807, | |
| "grad_norm": 0.15098582104711297, | |
| "learning_rate": 4.408602150537635e-05, | |
| "loss": 0.5202, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0911062906724512, | |
| "grad_norm": 0.13316194280918697, | |
| "learning_rate": 4.516129032258064e-05, | |
| "loss": 0.4469, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09327548806941431, | |
| "grad_norm": 0.13385436180609817, | |
| "learning_rate": 4.6236559139784944e-05, | |
| "loss": 0.4936, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.09544468546637744, | |
| "grad_norm": 0.1456999546987616, | |
| "learning_rate": 4.731182795698925e-05, | |
| "loss": 0.4559, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.09761388286334056, | |
| "grad_norm": 0.13569281275379386, | |
| "learning_rate": 4.8387096774193554e-05, | |
| "loss": 0.4314, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.09978308026030369, | |
| "grad_norm": 0.15212176484932607, | |
| "learning_rate": 4.9462365591397855e-05, | |
| "loss": 0.4876, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1019522776572668, | |
| "grad_norm": 0.12839200001749063, | |
| "learning_rate": 5.053763440860215e-05, | |
| "loss": 0.4748, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.10412147505422993, | |
| "grad_norm": 0.12862351761485963, | |
| "learning_rate": 5.161290322580645e-05, | |
| "loss": 0.4125, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.10629067245119306, | |
| "grad_norm": 0.12676342192056295, | |
| "learning_rate": 5.268817204301075e-05, | |
| "loss": 0.425, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.10845986984815618, | |
| "grad_norm": 0.12433212919621899, | |
| "learning_rate": 5.3763440860215054e-05, | |
| "loss": 0.4082, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11062906724511931, | |
| "grad_norm": 0.13762864881467574, | |
| "learning_rate": 5.4838709677419355e-05, | |
| "loss": 0.4342, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11279826464208242, | |
| "grad_norm": 0.16375581387185012, | |
| "learning_rate": 5.5913978494623656e-05, | |
| "loss": 0.5276, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.11496746203904555, | |
| "grad_norm": 0.14903663671440864, | |
| "learning_rate": 5.6989247311827965e-05, | |
| "loss": 0.4966, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.11713665943600868, | |
| "grad_norm": 0.13085616750795798, | |
| "learning_rate": 5.8064516129032266e-05, | |
| "loss": 0.4177, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1193058568329718, | |
| "grad_norm": 0.1399410088881321, | |
| "learning_rate": 5.913978494623657e-05, | |
| "loss": 0.4759, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12147505422993492, | |
| "grad_norm": 0.17552603285175208, | |
| "learning_rate": 6.021505376344086e-05, | |
| "loss": 0.3852, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12364425162689804, | |
| "grad_norm": 0.150563045551466, | |
| "learning_rate": 6.129032258064517e-05, | |
| "loss": 0.4456, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.12581344902386118, | |
| "grad_norm": 0.15103853247096755, | |
| "learning_rate": 6.236559139784946e-05, | |
| "loss": 0.486, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1279826464208243, | |
| "grad_norm": 0.13897437214670214, | |
| "learning_rate": 6.344086021505376e-05, | |
| "loss": 0.393, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.1301518438177874, | |
| "grad_norm": 0.13998124057548186, | |
| "learning_rate": 6.451612903225807e-05, | |
| "loss": 0.3703, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13232104121475055, | |
| "grad_norm": 0.14910024036690483, | |
| "learning_rate": 6.559139784946236e-05, | |
| "loss": 0.4163, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.13449023861171366, | |
| "grad_norm": 0.1426571111554546, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.3021, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.13665943600867678, | |
| "grad_norm": 0.2255435027702441, | |
| "learning_rate": 6.774193548387096e-05, | |
| "loss": 0.348, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.13882863340563992, | |
| "grad_norm": 0.15326407709554848, | |
| "learning_rate": 6.881720430107527e-05, | |
| "loss": 0.4446, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14099783080260303, | |
| "grad_norm": 0.15096250596140368, | |
| "learning_rate": 6.989247311827958e-05, | |
| "loss": 0.4356, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14316702819956617, | |
| "grad_norm": 0.14931162214298302, | |
| "learning_rate": 7.096774193548388e-05, | |
| "loss": 0.3921, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.14533622559652928, | |
| "grad_norm": 0.13357955043603964, | |
| "learning_rate": 7.204301075268818e-05, | |
| "loss": 0.285, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1475054229934924, | |
| "grad_norm": 0.14467881960033757, | |
| "learning_rate": 7.311827956989248e-05, | |
| "loss": 0.3828, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.14967462039045554, | |
| "grad_norm": 0.13526505831717678, | |
| "learning_rate": 7.419354838709677e-05, | |
| "loss": 0.3523, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.15184381778741865, | |
| "grad_norm": 0.17267391484274627, | |
| "learning_rate": 7.526881720430108e-05, | |
| "loss": 0.4506, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1540130151843818, | |
| "grad_norm": 0.15979585190145928, | |
| "learning_rate": 7.634408602150538e-05, | |
| "loss": 0.3862, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1561822125813449, | |
| "grad_norm": 0.15434731006679933, | |
| "learning_rate": 7.741935483870968e-05, | |
| "loss": 0.3694, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.15835140997830802, | |
| "grad_norm": 0.14815314471149627, | |
| "learning_rate": 7.849462365591398e-05, | |
| "loss": 0.3644, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16052060737527116, | |
| "grad_norm": 0.15387792339587666, | |
| "learning_rate": 7.956989247311829e-05, | |
| "loss": 0.4095, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.16268980477223427, | |
| "grad_norm": 0.15135351623798715, | |
| "learning_rate": 8.064516129032258e-05, | |
| "loss": 0.4128, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1648590021691974, | |
| "grad_norm": 0.13006929468252665, | |
| "learning_rate": 8.172043010752689e-05, | |
| "loss": 0.2786, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.16702819956616052, | |
| "grad_norm": 0.17289441324627974, | |
| "learning_rate": 8.27956989247312e-05, | |
| "loss": 0.4148, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.16919739696312364, | |
| "grad_norm": 0.15172742838256395, | |
| "learning_rate": 8.387096774193549e-05, | |
| "loss": 0.3758, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17136659436008678, | |
| "grad_norm": 0.16805860911767084, | |
| "learning_rate": 8.494623655913979e-05, | |
| "loss": 0.3755, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1735357917570499, | |
| "grad_norm": 0.13731854156086779, | |
| "learning_rate": 8.60215053763441e-05, | |
| "loss": 0.3031, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.175704989154013, | |
| "grad_norm": 0.15898088745954764, | |
| "learning_rate": 8.709677419354839e-05, | |
| "loss": 0.3785, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.17787418655097614, | |
| "grad_norm": 0.14869088820217116, | |
| "learning_rate": 8.81720430107527e-05, | |
| "loss": 0.3831, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18004338394793926, | |
| "grad_norm": 0.15285942249523343, | |
| "learning_rate": 8.924731182795699e-05, | |
| "loss": 0.3302, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1822125813449024, | |
| "grad_norm": 0.15925128432243588, | |
| "learning_rate": 9.032258064516129e-05, | |
| "loss": 0.389, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1843817787418655, | |
| "grad_norm": 0.1538607349870438, | |
| "learning_rate": 9.13978494623656e-05, | |
| "loss": 0.366, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.18655097613882862, | |
| "grad_norm": 0.16016055121309444, | |
| "learning_rate": 9.247311827956989e-05, | |
| "loss": 0.3554, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.18872017353579176, | |
| "grad_norm": 0.14283942143425307, | |
| "learning_rate": 9.35483870967742e-05, | |
| "loss": 0.3336, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.19088937093275488, | |
| "grad_norm": 0.16354104148599777, | |
| "learning_rate": 9.46236559139785e-05, | |
| "loss": 0.4439, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.19305856832971802, | |
| "grad_norm": 0.18296179182683667, | |
| "learning_rate": 9.56989247311828e-05, | |
| "loss": 0.4064, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.19522776572668113, | |
| "grad_norm": 0.15460229153045282, | |
| "learning_rate": 9.677419354838711e-05, | |
| "loss": 0.3407, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19739696312364424, | |
| "grad_norm": 0.15215604875393682, | |
| "learning_rate": 9.78494623655914e-05, | |
| "loss": 0.3553, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.19956616052060738, | |
| "grad_norm": 0.1597213335615803, | |
| "learning_rate": 9.892473118279571e-05, | |
| "loss": 0.3458, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2017353579175705, | |
| "grad_norm": 0.15328246596239517, | |
| "learning_rate": 0.0001, | |
| "loss": 0.342, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2039045553145336, | |
| "grad_norm": 0.15921412490863757, | |
| "learning_rate": 9.999964097046781e-05, | |
| "loss": 0.3843, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.20607375271149675, | |
| "grad_norm": 0.15390306970937798, | |
| "learning_rate": 9.999856388702731e-05, | |
| "loss": 0.3442, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.20824295010845986, | |
| "grad_norm": 0.15618794488353185, | |
| "learning_rate": 9.999676876514667e-05, | |
| "loss": 0.3396, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.210412147505423, | |
| "grad_norm": 0.13701461086257688, | |
| "learning_rate": 9.999425563060602e-05, | |
| "loss": 0.2914, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.21258134490238612, | |
| "grad_norm": 0.17727970752759084, | |
| "learning_rate": 9.999102451949688e-05, | |
| "loss": 0.3815, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.21475054229934923, | |
| "grad_norm": 0.17606273323635682, | |
| "learning_rate": 9.998707547822186e-05, | |
| "loss": 0.377, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.21691973969631237, | |
| "grad_norm": 0.16285779873289172, | |
| "learning_rate": 9.998240856349383e-05, | |
| "loss": 0.3221, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21908893709327548, | |
| "grad_norm": 0.17385851069798777, | |
| "learning_rate": 9.997702384233523e-05, | |
| "loss": 0.3539, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.22125813449023862, | |
| "grad_norm": 0.15796293985116644, | |
| "learning_rate": 9.9970921392077e-05, | |
| "loss": 0.3407, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.22342733188720174, | |
| "grad_norm": 0.16210590012135875, | |
| "learning_rate": 9.996410130035751e-05, | |
| "loss": 0.3433, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.22559652928416485, | |
| "grad_norm": 0.1815484690495641, | |
| "learning_rate": 9.995656366512138e-05, | |
| "loss": 0.3576, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.227765726681128, | |
| "grad_norm": 0.16370570821991282, | |
| "learning_rate": 9.994830859461793e-05, | |
| "loss": 0.3713, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2299349240780911, | |
| "grad_norm": 0.16719534192377292, | |
| "learning_rate": 9.993933620739974e-05, | |
| "loss": 0.3868, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.23210412147505424, | |
| "grad_norm": 0.19714493676306685, | |
| "learning_rate": 9.992964663232086e-05, | |
| "loss": 0.3378, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.23427331887201736, | |
| "grad_norm": 0.14799736868617666, | |
| "learning_rate": 9.991924000853505e-05, | |
| "loss": 0.3252, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.23644251626898047, | |
| "grad_norm": 0.15459646771807045, | |
| "learning_rate": 9.990811648549374e-05, | |
| "loss": 0.349, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2386117136659436, | |
| "grad_norm": 0.15814773704274712, | |
| "learning_rate": 9.989627622294384e-05, | |
| "loss": 0.3459, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24078091106290672, | |
| "grad_norm": 0.1573452355191302, | |
| "learning_rate": 9.988371939092551e-05, | |
| "loss": 0.3436, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.24295010845986983, | |
| "grad_norm": 0.17857552638412025, | |
| "learning_rate": 9.987044616976969e-05, | |
| "loss": 0.3515, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.24511930585683298, | |
| "grad_norm": 0.17548675076361434, | |
| "learning_rate": 9.985645675009551e-05, | |
| "loss": 0.4118, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2472885032537961, | |
| "grad_norm": 0.16436582516347584, | |
| "learning_rate": 9.984175133280758e-05, | |
| "loss": 0.3623, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.24945770065075923, | |
| "grad_norm": 0.14320940017640504, | |
| "learning_rate": 9.982633012909304e-05, | |
| "loss": 0.296, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.25162689804772237, | |
| "grad_norm": 0.16287542910761935, | |
| "learning_rate": 9.981019336041861e-05, | |
| "loss": 0.296, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.25379609544468545, | |
| "grad_norm": 0.16861388898681715, | |
| "learning_rate": 9.979334125852735e-05, | |
| "loss": 0.337, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2559652928416486, | |
| "grad_norm": 0.15898794759349474, | |
| "learning_rate": 9.977577406543535e-05, | |
| "loss": 0.3329, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.25813449023861174, | |
| "grad_norm": 0.1734384500805255, | |
| "learning_rate": 9.975749203342823e-05, | |
| "loss": 0.3463, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2603036876355748, | |
| "grad_norm": 0.1687288542407359, | |
| "learning_rate": 9.97384954250576e-05, | |
| "loss": 0.3716, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26247288503253796, | |
| "grad_norm": 0.15571310481512463, | |
| "learning_rate": 9.971878451313719e-05, | |
| "loss": 0.3174, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2646420824295011, | |
| "grad_norm": 0.1519598125985657, | |
| "learning_rate": 9.969835958073897e-05, | |
| "loss": 0.2869, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2668112798264642, | |
| "grad_norm": 0.1631777318786355, | |
| "learning_rate": 9.967722092118909e-05, | |
| "loss": 0.3269, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.26898047722342733, | |
| "grad_norm": 0.15802300841876601, | |
| "learning_rate": 9.965536883806368e-05, | |
| "loss": 0.3369, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.27114967462039047, | |
| "grad_norm": 0.15777870559434587, | |
| "learning_rate": 9.963280364518448e-05, | |
| "loss": 0.3434, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27331887201735355, | |
| "grad_norm": 0.14460302054911775, | |
| "learning_rate": 9.96095256666143e-05, | |
| "loss": 0.2965, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2754880694143167, | |
| "grad_norm": 0.17074141051395292, | |
| "learning_rate": 9.958553523665242e-05, | |
| "loss": 0.3627, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.27765726681127983, | |
| "grad_norm": 0.16289437443944546, | |
| "learning_rate": 9.956083269982973e-05, | |
| "loss": 0.3269, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.279826464208243, | |
| "grad_norm": 0.19184991332619222, | |
| "learning_rate": 9.953541841090388e-05, | |
| "loss": 0.3729, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.28199566160520606, | |
| "grad_norm": 0.14220941041693624, | |
| "learning_rate": 9.950929273485404e-05, | |
| "loss": 0.3038, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2841648590021692, | |
| "grad_norm": 0.15132399774722008, | |
| "learning_rate": 9.948245604687581e-05, | |
| "loss": 0.3162, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.28633405639913234, | |
| "grad_norm": 0.1567696241360261, | |
| "learning_rate": 9.945490873237571e-05, | |
| "loss": 0.3041, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2885032537960954, | |
| "grad_norm": 0.14672088173624798, | |
| "learning_rate": 9.942665118696575e-05, | |
| "loss": 0.2985, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.29067245119305857, | |
| "grad_norm": 0.1769806338719645, | |
| "learning_rate": 9.939768381645762e-05, | |
| "loss": 0.3648, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2928416485900217, | |
| "grad_norm": 0.1600455826423261, | |
| "learning_rate": 9.9368007036857e-05, | |
| "loss": 0.3428, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2950108459869848, | |
| "grad_norm": 0.15482109589656579, | |
| "learning_rate": 9.933762127435751e-05, | |
| "loss": 0.3211, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.29718004338394793, | |
| "grad_norm": 0.1376974659275616, | |
| "learning_rate": 9.930652696533458e-05, | |
| "loss": 0.2655, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2993492407809111, | |
| "grad_norm": 0.1604514505800568, | |
| "learning_rate": 9.927472455633921e-05, | |
| "loss": 0.3424, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.30151843817787416, | |
| "grad_norm": 0.14831623076631825, | |
| "learning_rate": 9.92422145040916e-05, | |
| "loss": 0.2929, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3036876355748373, | |
| "grad_norm": 0.14497199913662426, | |
| "learning_rate": 9.920899727547446e-05, | |
| "loss": 0.3063, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30585683297180044, | |
| "grad_norm": 0.1686161432311634, | |
| "learning_rate": 9.917507334752647e-05, | |
| "loss": 0.3695, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3080260303687636, | |
| "grad_norm": 0.1415843035366969, | |
| "learning_rate": 9.914044320743527e-05, | |
| "loss": 0.2989, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.31019522776572667, | |
| "grad_norm": 0.1537154117560618, | |
| "learning_rate": 9.91051073525306e-05, | |
| "loss": 0.323, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3123644251626898, | |
| "grad_norm": 0.19421465105677843, | |
| "learning_rate": 9.90690662902771e-05, | |
| "loss": 0.3331, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.31453362255965295, | |
| "grad_norm": 0.15822998528244542, | |
| "learning_rate": 9.903232053826696e-05, | |
| "loss": 0.2934, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31670281995661603, | |
| "grad_norm": 0.15358991548479717, | |
| "learning_rate": 9.89948706242126e-05, | |
| "loss": 0.3206, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3188720173535792, | |
| "grad_norm": 0.14959200755836038, | |
| "learning_rate": 9.895671708593903e-05, | |
| "loss": 0.3114, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3210412147505423, | |
| "grad_norm": 0.16473377943898643, | |
| "learning_rate": 9.891786047137615e-05, | |
| "loss": 0.3312, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3232104121475054, | |
| "grad_norm": 0.1648174790366256, | |
| "learning_rate": 9.887830133855079e-05, | |
| "loss": 0.3597, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.32537960954446854, | |
| "grad_norm": 0.1887880278076816, | |
| "learning_rate": 9.883804025557888e-05, | |
| "loss": 0.3968, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3275488069414317, | |
| "grad_norm": 0.17273118800696421, | |
| "learning_rate": 9.879707780065712e-05, | |
| "loss": 0.3731, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3297180043383948, | |
| "grad_norm": 0.16253972479744438, | |
| "learning_rate": 9.875541456205473e-05, | |
| "loss": 0.3408, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3318872017353579, | |
| "grad_norm": 0.1583694488182022, | |
| "learning_rate": 9.871305113810505e-05, | |
| "loss": 0.3121, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.33405639913232105, | |
| "grad_norm": 0.16161186391084162, | |
| "learning_rate": 9.86699881371969e-05, | |
| "loss": 0.3331, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3362255965292842, | |
| "grad_norm": 0.15558997399940616, | |
| "learning_rate": 9.862622617776582e-05, | |
| "loss": 0.3136, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3383947939262473, | |
| "grad_norm": 0.14717176320577952, | |
| "learning_rate": 9.858176588828526e-05, | |
| "loss": 0.3227, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3405639913232104, | |
| "grad_norm": 0.15604482540053785, | |
| "learning_rate": 9.85366079072575e-05, | |
| "loss": 0.3093, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.34273318872017355, | |
| "grad_norm": 0.164872478623103, | |
| "learning_rate": 9.849075288320446e-05, | |
| "loss": 0.3231, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.34490238611713664, | |
| "grad_norm": 0.18011689706998146, | |
| "learning_rate": 9.84442014746585e-05, | |
| "loss": 0.3571, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3470715835140998, | |
| "grad_norm": 0.2281829326423908, | |
| "learning_rate": 9.839695435015279e-05, | |
| "loss": 0.4504, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3492407809110629, | |
| "grad_norm": 0.17155618609511838, | |
| "learning_rate": 9.83490121882119e-05, | |
| "loss": 0.3318, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.351409978308026, | |
| "grad_norm": 0.1451699212552566, | |
| "learning_rate": 9.830037567734187e-05, | |
| "loss": 0.27, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.35357917570498915, | |
| "grad_norm": 0.16172715854603054, | |
| "learning_rate": 9.825104551602047e-05, | |
| "loss": 0.2993, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3557483731019523, | |
| "grad_norm": 0.16625581355250432, | |
| "learning_rate": 9.820102241268708e-05, | |
| "loss": 0.3447, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3579175704989154, | |
| "grad_norm": 0.17497945522123493, | |
| "learning_rate": 9.815030708573256e-05, | |
| "loss": 0.3604, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3600867678958785, | |
| "grad_norm": 0.16549978577941174, | |
| "learning_rate": 9.809890026348891e-05, | |
| "loss": 0.3325, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.36225596529284165, | |
| "grad_norm": 0.15123204350737743, | |
| "learning_rate": 9.804680268421885e-05, | |
| "loss": 0.3047, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3644251626898048, | |
| "grad_norm": 0.1653117817038064, | |
| "learning_rate": 9.799401509610511e-05, | |
| "loss": 0.3336, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3665943600867679, | |
| "grad_norm": 0.15941088150160534, | |
| "learning_rate": 9.794053825723983e-05, | |
| "loss": 0.3415, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.368763557483731, | |
| "grad_norm": 0.1569489292887316, | |
| "learning_rate": 9.788637293561363e-05, | |
| "loss": 0.3086, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37093275488069416, | |
| "grad_norm": 0.15521077438298633, | |
| "learning_rate": 9.783151990910446e-05, | |
| "loss": 0.3129, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.37310195227765725, | |
| "grad_norm": 0.15853168180990132, | |
| "learning_rate": 9.777597996546661e-05, | |
| "loss": 0.3197, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3752711496746204, | |
| "grad_norm": 0.17021980724333097, | |
| "learning_rate": 9.771975390231927e-05, | |
| "loss": 0.3438, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3774403470715835, | |
| "grad_norm": 0.1467658403488705, | |
| "learning_rate": 9.766284252713511e-05, | |
| "loss": 0.2868, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3796095444685466, | |
| "grad_norm": 0.19325560102675848, | |
| "learning_rate": 9.760524665722874e-05, | |
| "loss": 0.4445, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38177874186550975, | |
| "grad_norm": 0.1676049985924069, | |
| "learning_rate": 9.754696711974486e-05, | |
| "loss": 0.3596, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3839479392624729, | |
| "grad_norm": 0.1517046172036335, | |
| "learning_rate": 9.748800475164648e-05, | |
| "loss": 0.2814, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.38611713665943603, | |
| "grad_norm": 0.1499206779996939, | |
| "learning_rate": 9.742836039970287e-05, | |
| "loss": 0.3078, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3882863340563991, | |
| "grad_norm": 0.16835317016600657, | |
| "learning_rate": 9.736803492047736e-05, | |
| "loss": 0.346, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.39045553145336226, | |
| "grad_norm": 0.1735161690202731, | |
| "learning_rate": 9.730702918031511e-05, | |
| "loss": 0.3339, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3926247288503254, | |
| "grad_norm": 0.178265268230418, | |
| "learning_rate": 9.724534405533061e-05, | |
| "loss": 0.3581, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3947939262472885, | |
| "grad_norm": 0.15928651246621914, | |
| "learning_rate": 9.718298043139513e-05, | |
| "loss": 0.3158, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3969631236442516, | |
| "grad_norm": 0.15680437561632446, | |
| "learning_rate": 9.711993920412395e-05, | |
| "loss": 0.3612, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.39913232104121477, | |
| "grad_norm": 0.16147834287905172, | |
| "learning_rate": 9.70562212788636e-05, | |
| "loss": 0.3213, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.40130151843817785, | |
| "grad_norm": 0.16810458922445323, | |
| "learning_rate": 9.699182757067875e-05, | |
| "loss": 0.3715, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.403470715835141, | |
| "grad_norm": 0.16273652493527896, | |
| "learning_rate": 9.69267590043391e-05, | |
| "loss": 0.2901, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.40563991323210413, | |
| "grad_norm": 0.17226018154021203, | |
| "learning_rate": 9.686101651430612e-05, | |
| "loss": 0.3774, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4078091106290672, | |
| "grad_norm": 0.18016919749286983, | |
| "learning_rate": 9.679460104471965e-05, | |
| "loss": 0.3602, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.40997830802603036, | |
| "grad_norm": 0.15507719486138638, | |
| "learning_rate": 9.672751354938429e-05, | |
| "loss": 0.3129, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4121475054229935, | |
| "grad_norm": 0.1582451132361843, | |
| "learning_rate": 9.66597549917557e-05, | |
| "loss": 0.3183, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.41431670281995664, | |
| "grad_norm": 0.14849683648502013, | |
| "learning_rate": 9.659132634492684e-05, | |
| "loss": 0.3103, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4164859002169197, | |
| "grad_norm": 0.24758668306536105, | |
| "learning_rate": 9.652222859161388e-05, | |
| "loss": 0.3157, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.41865509761388287, | |
| "grad_norm": 0.17938378050442594, | |
| "learning_rate": 9.645246272414221e-05, | |
| "loss": 0.3715, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.420824295010846, | |
| "grad_norm": 0.14292889306164755, | |
| "learning_rate": 9.63820297444321e-05, | |
| "loss": 0.281, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4229934924078091, | |
| "grad_norm": 0.15121757619907555, | |
| "learning_rate": 9.63109306639843e-05, | |
| "loss": 0.285, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42516268980477223, | |
| "grad_norm": 0.15479911159606385, | |
| "learning_rate": 9.623916650386564e-05, | |
| "loss": 0.333, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.42733188720173537, | |
| "grad_norm": 0.19245481852848167, | |
| "learning_rate": 9.61667382946942e-05, | |
| "loss": 0.3803, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.42950108459869846, | |
| "grad_norm": 0.16251333268925855, | |
| "learning_rate": 9.609364707662467e-05, | |
| "loss": 0.3326, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4316702819956616, | |
| "grad_norm": 0.15843432631003965, | |
| "learning_rate": 9.601989389933323e-05, | |
| "loss": 0.3151, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.43383947939262474, | |
| "grad_norm": 0.1804329476464318, | |
| "learning_rate": 9.594547982200266e-05, | |
| "loss": 0.3436, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43383947939262474, | |
| "eval_loss": 0.33424264192581177, | |
| "eval_runtime": 39.7662, | |
| "eval_samples_per_second": 0.478, | |
| "eval_steps_per_second": 0.126, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4360086767895879, | |
| "grad_norm": 0.19645387562310979, | |
| "learning_rate": 9.5870405913307e-05, | |
| "loss": 0.3159, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.43817787418655096, | |
| "grad_norm": 0.1644546677880722, | |
| "learning_rate": 9.579467325139627e-05, | |
| "loss": 0.3315, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4403470715835141, | |
| "grad_norm": 0.15812376696679412, | |
| "learning_rate": 9.571828292388096e-05, | |
| "loss": 0.3348, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.44251626898047725, | |
| "grad_norm": 0.1675206237509223, | |
| "learning_rate": 9.56412360278164e-05, | |
| "loss": 0.3721, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.44468546637744033, | |
| "grad_norm": 0.14308548515478076, | |
| "learning_rate": 9.556353366968705e-05, | |
| "loss": 0.2651, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44685466377440347, | |
| "grad_norm": 0.15386997755578125, | |
| "learning_rate": 9.548517696539054e-05, | |
| "loss": 0.2701, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.4490238611713666, | |
| "grad_norm": 0.14640343506659434, | |
| "learning_rate": 9.540616704022173e-05, | |
| "loss": 0.2496, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.4511930585683297, | |
| "grad_norm": 0.16521172350141197, | |
| "learning_rate": 9.532650502885646e-05, | |
| "loss": 0.3337, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.45336225596529284, | |
| "grad_norm": 0.1540772430165373, | |
| "learning_rate": 9.524619207533532e-05, | |
| "loss": 0.3043, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.455531453362256, | |
| "grad_norm": 0.16165589280753936, | |
| "learning_rate": 9.516522933304721e-05, | |
| "loss": 0.3276, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45770065075921906, | |
| "grad_norm": 0.1883035011589188, | |
| "learning_rate": 9.508361796471272e-05, | |
| "loss": 0.3584, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.4598698481561822, | |
| "grad_norm": 0.14934674223988761, | |
| "learning_rate": 9.500135914236755e-05, | |
| "loss": 0.2946, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.46203904555314534, | |
| "grad_norm": 0.16930578402110355, | |
| "learning_rate": 9.491845404734551e-05, | |
| "loss": 0.3423, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4642082429501085, | |
| "grad_norm": 0.16818902583571363, | |
| "learning_rate": 9.483490387026174e-05, | |
| "loss": 0.3245, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.46637744034707157, | |
| "grad_norm": 0.1646282158556431, | |
| "learning_rate": 9.475070981099545e-05, | |
| "loss": 0.3377, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4685466377440347, | |
| "grad_norm": 0.15601143737978804, | |
| "learning_rate": 9.466587307867281e-05, | |
| "loss": 0.3181, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.47071583514099785, | |
| "grad_norm": 0.16754840015499342, | |
| "learning_rate": 9.458039489164951e-05, | |
| "loss": 0.3682, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.47288503253796094, | |
| "grad_norm": 0.1426551523167691, | |
| "learning_rate": 9.449427647749328e-05, | |
| "loss": 0.267, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.4750542299349241, | |
| "grad_norm": 0.16599632915307785, | |
| "learning_rate": 9.440751907296628e-05, | |
| "loss": 0.3181, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.4772234273318872, | |
| "grad_norm": 0.1620746064207198, | |
| "learning_rate": 9.432012392400733e-05, | |
| "loss": 0.3676, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4793926247288503, | |
| "grad_norm": 0.15120817805826722, | |
| "learning_rate": 9.423209228571398e-05, | |
| "loss": 0.3158, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.48156182212581344, | |
| "grad_norm": 0.14803905174097284, | |
| "learning_rate": 9.414342542232462e-05, | |
| "loss": 0.2836, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4837310195227766, | |
| "grad_norm": 0.15220763579005372, | |
| "learning_rate": 9.405412460720006e-05, | |
| "loss": 0.2827, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.48590021691973967, | |
| "grad_norm": 0.15404544864962286, | |
| "learning_rate": 9.396419112280555e-05, | |
| "loss": 0.3025, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.4880694143167028, | |
| "grad_norm": 0.15073679537979376, | |
| "learning_rate": 9.387362626069216e-05, | |
| "loss": 0.3144, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49023861171366595, | |
| "grad_norm": 0.1610803970826281, | |
| "learning_rate": 9.378243132147825e-05, | |
| "loss": 0.2977, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.4924078091106291, | |
| "grad_norm": 0.16316823817190756, | |
| "learning_rate": 9.369060761483095e-05, | |
| "loss": 0.3128, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.4945770065075922, | |
| "grad_norm": 0.163300375716653, | |
| "learning_rate": 9.359815645944709e-05, | |
| "loss": 0.3438, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4967462039045553, | |
| "grad_norm": 0.14836269729887547, | |
| "learning_rate": 9.35050791830345e-05, | |
| "loss": 0.2525, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.49891540130151846, | |
| "grad_norm": 0.14711791523393897, | |
| "learning_rate": 9.341137712229282e-05, | |
| "loss": 0.2713, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5010845986984815, | |
| "grad_norm": 0.1563819160365463, | |
| "learning_rate": 9.331705162289433e-05, | |
| "loss": 0.2919, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5032537960954447, | |
| "grad_norm": 0.1723872547155802, | |
| "learning_rate": 9.322210403946461e-05, | |
| "loss": 0.3121, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5054229934924078, | |
| "grad_norm": 0.17043959065030484, | |
| "learning_rate": 9.312653573556316e-05, | |
| "loss": 0.3561, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5075921908893709, | |
| "grad_norm": 0.14967506341505113, | |
| "learning_rate": 9.303034808366367e-05, | |
| "loss": 0.282, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5097613882863341, | |
| "grad_norm": 0.3206125419158534, | |
| "learning_rate": 9.293354246513448e-05, | |
| "loss": 0.3589, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5119305856832972, | |
| "grad_norm": 0.15389716021946065, | |
| "learning_rate": 9.283612027021862e-05, | |
| "loss": 0.3168, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5140997830802603, | |
| "grad_norm": 0.16629395409267134, | |
| "learning_rate": 9.273808289801388e-05, | |
| "loss": 0.2989, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5162689804772235, | |
| "grad_norm": 0.17350303956302596, | |
| "learning_rate": 9.263943175645275e-05, | |
| "loss": 0.3515, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5184381778741866, | |
| "grad_norm": 0.16804219683380595, | |
| "learning_rate": 9.254016826228215e-05, | |
| "loss": 0.3296, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5206073752711496, | |
| "grad_norm": 0.15382387009648443, | |
| "learning_rate": 9.244029384104311e-05, | |
| "loss": 0.2929, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5227765726681128, | |
| "grad_norm": 0.1817152430149277, | |
| "learning_rate": 9.233980992705031e-05, | |
| "loss": 0.409, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5249457700650759, | |
| "grad_norm": 0.15284178391247238, | |
| "learning_rate": 9.223871796337147e-05, | |
| "loss": 0.29, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.527114967462039, | |
| "grad_norm": 0.17168700031201556, | |
| "learning_rate": 9.213701940180657e-05, | |
| "loss": 0.3596, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5292841648590022, | |
| "grad_norm": 0.16236461998196408, | |
| "learning_rate": 9.203471570286711e-05, | |
| "loss": 0.3531, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5314533622559653, | |
| "grad_norm": 0.17510095685795815, | |
| "learning_rate": 9.193180833575506e-05, | |
| "loss": 0.4172, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5336225596529284, | |
| "grad_norm": 0.1618295606935838, | |
| "learning_rate": 9.182829877834176e-05, | |
| "loss": 0.3086, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5357917570498916, | |
| "grad_norm": 0.16138627788900253, | |
| "learning_rate": 9.172418851714676e-05, | |
| "loss": 0.3409, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5379609544468547, | |
| "grad_norm": 0.15375456963212109, | |
| "learning_rate": 9.161947904731636e-05, | |
| "loss": 0.3306, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.5401301518438177, | |
| "grad_norm": 0.14093510508892232, | |
| "learning_rate": 9.151417187260226e-05, | |
| "loss": 0.2723, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.5422993492407809, | |
| "grad_norm": 0.1566994003310743, | |
| "learning_rate": 9.140826850533987e-05, | |
| "loss": 0.3047, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.544468546637744, | |
| "grad_norm": 0.16263764510479803, | |
| "learning_rate": 9.130177046642667e-05, | |
| "loss": 0.3313, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5466377440347071, | |
| "grad_norm": 0.2635396490290928, | |
| "learning_rate": 9.119467928530027e-05, | |
| "loss": 0.3137, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5488069414316703, | |
| "grad_norm": 0.19432404691982347, | |
| "learning_rate": 9.108699649991659e-05, | |
| "loss": 0.3271, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5509761388286334, | |
| "grad_norm": 0.16321335163395914, | |
| "learning_rate": 9.097872365672757e-05, | |
| "loss": 0.3464, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5531453362255966, | |
| "grad_norm": 0.1674780863277491, | |
| "learning_rate": 9.086986231065917e-05, | |
| "loss": 0.3479, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5553145336225597, | |
| "grad_norm": 0.16720487351831387, | |
| "learning_rate": 9.076041402508893e-05, | |
| "loss": 0.3765, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.5574837310195228, | |
| "grad_norm": 0.1768233975227357, | |
| "learning_rate": 9.06503803718235e-05, | |
| "loss": 0.3611, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.559652928416486, | |
| "grad_norm": 0.1658738115128469, | |
| "learning_rate": 9.053976293107612e-05, | |
| "loss": 0.3425, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.561822125813449, | |
| "grad_norm": 0.15075392628393477, | |
| "learning_rate": 9.042856329144393e-05, | |
| "loss": 0.2831, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5639913232104121, | |
| "grad_norm": 0.15462905264522137, | |
| "learning_rate": 9.031678304988509e-05, | |
| "loss": 0.3359, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5661605206073753, | |
| "grad_norm": 0.1486760035515965, | |
| "learning_rate": 9.020442381169593e-05, | |
| "loss": 0.3036, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5683297180043384, | |
| "grad_norm": 0.14672655464193096, | |
| "learning_rate": 9.009148719048785e-05, | |
| "loss": 0.2655, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5704989154013015, | |
| "grad_norm": 0.14525138583221342, | |
| "learning_rate": 8.99779748081641e-05, | |
| "loss": 0.2893, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5726681127982647, | |
| "grad_norm": 0.1649767278464621, | |
| "learning_rate": 8.986388829489663e-05, | |
| "loss": 0.3416, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5748373101952278, | |
| "grad_norm": 0.1486891920011246, | |
| "learning_rate": 8.97492292891025e-05, | |
| "loss": 0.3157, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5770065075921909, | |
| "grad_norm": 0.16844270718485851, | |
| "learning_rate": 8.96339994374205e-05, | |
| "loss": 0.3434, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.579175704989154, | |
| "grad_norm": 0.15180653954202514, | |
| "learning_rate": 8.951820039468741e-05, | |
| "loss": 0.2775, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5813449023861171, | |
| "grad_norm": 0.15671071191539224, | |
| "learning_rate": 8.940183382391429e-05, | |
| "loss": 0.2713, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.5835140997830802, | |
| "grad_norm": 0.16147849592493357, | |
| "learning_rate": 8.928490139626253e-05, | |
| "loss": 0.3173, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5856832971800434, | |
| "grad_norm": 0.19489581469202696, | |
| "learning_rate": 8.916740479101995e-05, | |
| "loss": 0.3289, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878524945770065, | |
| "grad_norm": 0.14998764666371173, | |
| "learning_rate": 8.90493456955766e-05, | |
| "loss": 0.2677, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.5900216919739696, | |
| "grad_norm": 0.15423637992388417, | |
| "learning_rate": 8.893072580540053e-05, | |
| "loss": 0.3233, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5921908893709328, | |
| "grad_norm": 0.1618307748258286, | |
| "learning_rate": 8.88115468240135e-05, | |
| "loss": 0.335, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5943600867678959, | |
| "grad_norm": 0.28889311411388907, | |
| "learning_rate": 8.869181046296647e-05, | |
| "loss": 0.3162, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.596529284164859, | |
| "grad_norm": 0.16811888898141328, | |
| "learning_rate": 8.857151844181502e-05, | |
| "loss": 0.3321, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5986984815618221, | |
| "grad_norm": 0.15974545714844673, | |
| "learning_rate": 8.845067248809469e-05, | |
| "loss": 0.3424, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6008676789587852, | |
| "grad_norm": 0.14627135337253672, | |
| "learning_rate": 8.83292743372961e-05, | |
| "loss": 0.2914, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6030368763557483, | |
| "grad_norm": 0.15447685130795957, | |
| "learning_rate": 8.820732573284012e-05, | |
| "loss": 0.3175, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6052060737527115, | |
| "grad_norm": 0.17049697629599625, | |
| "learning_rate": 8.808482842605277e-05, | |
| "loss": 0.3328, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6073752711496746, | |
| "grad_norm": 0.15776013624499943, | |
| "learning_rate": 8.796178417614007e-05, | |
| "loss": 0.3109, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6095444685466378, | |
| "grad_norm": 0.16222342988686625, | |
| "learning_rate": 8.783819475016282e-05, | |
| "loss": 0.3748, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6117136659436009, | |
| "grad_norm": 0.2331822347475535, | |
| "learning_rate": 8.771406192301113e-05, | |
| "loss": 0.3138, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.613882863340564, | |
| "grad_norm": 0.15710540429632355, | |
| "learning_rate": 8.758938747737909e-05, | |
| "loss": 0.3204, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6160520607375272, | |
| "grad_norm": 0.1696778017954279, | |
| "learning_rate": 8.746417320373896e-05, | |
| "loss": 0.3545, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6182212581344902, | |
| "grad_norm": 0.160133581870411, | |
| "learning_rate": 8.733842090031565e-05, | |
| "loss": 0.3079, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6203904555314533, | |
| "grad_norm": 0.15694259520801862, | |
| "learning_rate": 8.72121323730608e-05, | |
| "loss": 0.311, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6225596529284165, | |
| "grad_norm": 0.1585388671488109, | |
| "learning_rate": 8.708530943562683e-05, | |
| "loss": 0.3366, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6247288503253796, | |
| "grad_norm": 0.1637524110812479, | |
| "learning_rate": 8.695795390934094e-05, | |
| "loss": 0.3351, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.6268980477223427, | |
| "grad_norm": 0.17338170936952552, | |
| "learning_rate": 8.683006762317891e-05, | |
| "loss": 0.3683, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6290672451193059, | |
| "grad_norm": 0.15279053222915287, | |
| "learning_rate": 8.670165241373891e-05, | |
| "loss": 0.2839, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.631236442516269, | |
| "grad_norm": 0.15271124650902063, | |
| "learning_rate": 8.657271012521504e-05, | |
| "loss": 0.274, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.6334056399132321, | |
| "grad_norm": 0.20119865632692643, | |
| "learning_rate": 8.644324260937085e-05, | |
| "loss": 0.2766, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.6355748373101953, | |
| "grad_norm": 0.16358970886050292, | |
| "learning_rate": 8.631325172551284e-05, | |
| "loss": 0.3223, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.6377440347071583, | |
| "grad_norm": 0.16024548206523562, | |
| "learning_rate": 8.618273934046364e-05, | |
| "loss": 0.3255, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.6399132321041214, | |
| "grad_norm": 0.18933994539002422, | |
| "learning_rate": 8.60517073285353e-05, | |
| "loss": 0.4288, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6420824295010846, | |
| "grad_norm": 0.14871982406160117, | |
| "learning_rate": 8.592015757150225e-05, | |
| "loss": 0.3048, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.6442516268980477, | |
| "grad_norm": 0.1634546407796359, | |
| "learning_rate": 8.578809195857445e-05, | |
| "loss": 0.3775, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.6464208242950108, | |
| "grad_norm": 0.16293792940515708, | |
| "learning_rate": 8.565551238637006e-05, | |
| "loss": 0.3142, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.648590021691974, | |
| "grad_norm": 0.14930487066180928, | |
| "learning_rate": 8.552242075888838e-05, | |
| "loss": 0.2613, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.6507592190889371, | |
| "grad_norm": 0.14814023625501535, | |
| "learning_rate": 8.538881898748241e-05, | |
| "loss": 0.3157, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6529284164859002, | |
| "grad_norm": 0.15429106087854486, | |
| "learning_rate": 8.525470899083138e-05, | |
| "loss": 0.3023, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6550976138828634, | |
| "grad_norm": 0.1526889270326536, | |
| "learning_rate": 8.51200926949133e-05, | |
| "loss": 0.2622, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.6572668112798264, | |
| "grad_norm": 0.16836180192848554, | |
| "learning_rate": 8.498497203297716e-05, | |
| "loss": 0.3181, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6594360086767896, | |
| "grad_norm": 0.1568597185268113, | |
| "learning_rate": 8.48493489455153e-05, | |
| "loss": 0.3164, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6616052060737527, | |
| "grad_norm": 0.1717660325284911, | |
| "learning_rate": 8.47132253802355e-05, | |
| "loss": 0.3144, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6637744034707158, | |
| "grad_norm": 0.15864917205877277, | |
| "learning_rate": 8.457660329203289e-05, | |
| "loss": 0.3057, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.665943600867679, | |
| "grad_norm": 0.16569505505355422, | |
| "learning_rate": 8.443948464296211e-05, | |
| "loss": 0.3327, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.6681127982646421, | |
| "grad_norm": 0.16057628499246557, | |
| "learning_rate": 8.430187140220889e-05, | |
| "loss": 0.291, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.6702819956616052, | |
| "grad_norm": 0.14767031347275292, | |
| "learning_rate": 8.416376554606195e-05, | |
| "loss": 0.2795, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.6724511930585684, | |
| "grad_norm": 0.15854515513637216, | |
| "learning_rate": 8.402516905788455e-05, | |
| "loss": 0.3009, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6746203904555315, | |
| "grad_norm": 0.14713035448630263, | |
| "learning_rate": 8.388608392808593e-05, | |
| "loss": 0.2793, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.6767895878524945, | |
| "grad_norm": 0.16017456682784914, | |
| "learning_rate": 8.37465121540929e-05, | |
| "loss": 0.305, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.6789587852494577, | |
| "grad_norm": 0.15120391692718194, | |
| "learning_rate": 8.360645574032098e-05, | |
| "loss": 0.2749, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.6811279826464208, | |
| "grad_norm": 0.15560001063747786, | |
| "learning_rate": 8.346591669814572e-05, | |
| "loss": 0.3186, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.6832971800433839, | |
| "grad_norm": 0.1613679084226523, | |
| "learning_rate": 8.332489704587381e-05, | |
| "loss": 0.3328, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6854663774403471, | |
| "grad_norm": 0.15605720716726337, | |
| "learning_rate": 8.318339880871402e-05, | |
| "loss": 0.3077, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.6876355748373102, | |
| "grad_norm": 0.15488519210128188, | |
| "learning_rate": 8.304142401874818e-05, | |
| "loss": 0.3256, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.6898047722342733, | |
| "grad_norm": 0.1403558799176201, | |
| "learning_rate": 8.2898974714902e-05, | |
| "loss": 0.2849, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.6919739696312365, | |
| "grad_norm": 0.16073415458894003, | |
| "learning_rate": 8.275605294291576e-05, | |
| "loss": 0.3148, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.6941431670281996, | |
| "grad_norm": 0.14531860683004322, | |
| "learning_rate": 8.261266075531493e-05, | |
| "loss": 0.2949, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6963123644251626, | |
| "grad_norm": 0.153620854119112, | |
| "learning_rate": 8.24688002113807e-05, | |
| "loss": 0.3024, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.6984815618221258, | |
| "grad_norm": 0.1598435331827638, | |
| "learning_rate": 8.232447337712045e-05, | |
| "loss": 0.3116, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7006507592190889, | |
| "grad_norm": 0.16126012816422514, | |
| "learning_rate": 8.217968232523798e-05, | |
| "loss": 0.3302, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.702819956616052, | |
| "grad_norm": 0.167819401197193, | |
| "learning_rate": 8.203442913510386e-05, | |
| "loss": 0.2961, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7049891540130152, | |
| "grad_norm": 0.16091517631557192, | |
| "learning_rate": 8.188871589272547e-05, | |
| "loss": 0.3058, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7071583514099783, | |
| "grad_norm": 0.15072907511506445, | |
| "learning_rate": 8.174254469071711e-05, | |
| "loss": 0.2768, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7093275488069414, | |
| "grad_norm": 0.15463419748008012, | |
| "learning_rate": 8.15959176282699e-05, | |
| "loss": 0.2698, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7114967462039046, | |
| "grad_norm": 0.15765765627941916, | |
| "learning_rate": 8.144883681112168e-05, | |
| "loss": 0.3327, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7136659436008677, | |
| "grad_norm": 0.1625944955327596, | |
| "learning_rate": 8.130130435152671e-05, | |
| "loss": 0.3115, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.7158351409978309, | |
| "grad_norm": 0.14663689685956108, | |
| "learning_rate": 8.115332236822543e-05, | |
| "loss": 0.2813, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7180043383947939, | |
| "grad_norm": 0.14366366452382384, | |
| "learning_rate": 8.100489298641387e-05, | |
| "loss": 0.252, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.720173535791757, | |
| "grad_norm": 0.14062872637814394, | |
| "learning_rate": 8.085601833771332e-05, | |
| "loss": 0.2426, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.7223427331887202, | |
| "grad_norm": 0.151319434703271, | |
| "learning_rate": 8.070670056013963e-05, | |
| "loss": 0.2804, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.7245119305856833, | |
| "grad_norm": 0.17339443782515734, | |
| "learning_rate": 8.055694179807241e-05, | |
| "loss": 0.3386, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.7266811279826464, | |
| "grad_norm": 0.162057871765424, | |
| "learning_rate": 8.040674420222442e-05, | |
| "loss": 0.3059, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7288503253796096, | |
| "grad_norm": 0.17452086702897832, | |
| "learning_rate": 8.025610992961059e-05, | |
| "loss": 0.323, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.7310195227765727, | |
| "grad_norm": 0.15218002802937822, | |
| "learning_rate": 8.010504114351699e-05, | |
| "loss": 0.2892, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.7331887201735358, | |
| "grad_norm": 0.1786691433855208, | |
| "learning_rate": 7.995354001346984e-05, | |
| "loss": 0.2993, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.735357917570499, | |
| "grad_norm": 0.15572465398388327, | |
| "learning_rate": 7.980160871520434e-05, | |
| "loss": 0.2751, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.737527114967462, | |
| "grad_norm": 0.14944653365494096, | |
| "learning_rate": 7.964924943063341e-05, | |
| "loss": 0.295, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7396963123644251, | |
| "grad_norm": 0.15137015815284247, | |
| "learning_rate": 7.949646434781637e-05, | |
| "loss": 0.2928, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.7418655097613883, | |
| "grad_norm": 0.13520870208050645, | |
| "learning_rate": 7.934325566092749e-05, | |
| "loss": 0.2392, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.7440347071583514, | |
| "grad_norm": 0.17474668202456908, | |
| "learning_rate": 7.918962557022445e-05, | |
| "loss": 0.3343, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.7462039045553145, | |
| "grad_norm": 0.15920430785387085, | |
| "learning_rate": 7.903557628201689e-05, | |
| "loss": 0.3108, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.7483731019522777, | |
| "grad_norm": 0.14495723872575775, | |
| "learning_rate": 7.888111000863455e-05, | |
| "loss": 0.2723, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7505422993492408, | |
| "grad_norm": 0.16184390654404834, | |
| "learning_rate": 7.872622896839556e-05, | |
| "loss": 0.2729, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.7527114967462039, | |
| "grad_norm": 0.16644660892068602, | |
| "learning_rate": 7.857093538557465e-05, | |
| "loss": 0.3267, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.754880694143167, | |
| "grad_norm": 0.15541714907955761, | |
| "learning_rate": 7.841523149037109e-05, | |
| "loss": 0.3217, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.7570498915401301, | |
| "grad_norm": 0.15934934206985665, | |
| "learning_rate": 7.825911951887677e-05, | |
| "loss": 0.2913, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.7592190889370932, | |
| "grad_norm": 0.1578728957708445, | |
| "learning_rate": 7.810260171304399e-05, | |
| "loss": 0.3093, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7613882863340564, | |
| "grad_norm": 0.15490528067171122, | |
| "learning_rate": 7.794568032065337e-05, | |
| "loss": 0.3009, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.7635574837310195, | |
| "grad_norm": 0.16699421067292403, | |
| "learning_rate": 7.778835759528145e-05, | |
| "loss": 0.3656, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.7657266811279827, | |
| "grad_norm": 0.22044364766801572, | |
| "learning_rate": 7.763063579626839e-05, | |
| "loss": 0.334, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.7678958785249458, | |
| "grad_norm": 0.19643699768744521, | |
| "learning_rate": 7.747251718868557e-05, | |
| "loss": 0.3231, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.7700650759219089, | |
| "grad_norm": 0.20910555862615757, | |
| "learning_rate": 7.731400404330298e-05, | |
| "loss": 0.3161, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7722342733188721, | |
| "grad_norm": 0.16468231811761894, | |
| "learning_rate": 7.715509863655661e-05, | |
| "loss": 0.2979, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.7744034707158352, | |
| "grad_norm": 0.1367744772298653, | |
| "learning_rate": 7.699580325051583e-05, | |
| "loss": 0.2166, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.7765726681127982, | |
| "grad_norm": 0.16273122371163448, | |
| "learning_rate": 7.683612017285056e-05, | |
| "loss": 0.2847, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.7787418655097614, | |
| "grad_norm": 0.15669925061953327, | |
| "learning_rate": 7.667605169679842e-05, | |
| "loss": 0.2777, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.7809110629067245, | |
| "grad_norm": 0.17725284818408057, | |
| "learning_rate": 7.651560012113182e-05, | |
| "loss": 0.3337, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7830802603036876, | |
| "grad_norm": 0.17427480006189128, | |
| "learning_rate": 7.635476775012493e-05, | |
| "loss": 0.3665, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.7852494577006508, | |
| "grad_norm": 0.17414268214659534, | |
| "learning_rate": 7.619355689352056e-05, | |
| "loss": 0.3309, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.7874186550976139, | |
| "grad_norm": 0.17512481390885523, | |
| "learning_rate": 7.60319698664971e-05, | |
| "loss": 0.3328, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.789587852494577, | |
| "grad_norm": 0.15289988099891824, | |
| "learning_rate": 7.587000898963508e-05, | |
| "loss": 0.2515, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.7917570498915402, | |
| "grad_norm": 0.1624765535782596, | |
| "learning_rate": 7.570767658888405e-05, | |
| "loss": 0.2518, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7939262472885033, | |
| "grad_norm": 0.1617918820535666, | |
| "learning_rate": 7.554497499552902e-05, | |
| "loss": 0.3053, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.7960954446854663, | |
| "grad_norm": 0.16383299523189962, | |
| "learning_rate": 7.538190654615711e-05, | |
| "loss": 0.3033, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.7982646420824295, | |
| "grad_norm": 0.1740801344083908, | |
| "learning_rate": 7.521847358262384e-05, | |
| "loss": 0.3324, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8004338394793926, | |
| "grad_norm": 0.16040212743690085, | |
| "learning_rate": 7.505467845201965e-05, | |
| "loss": 0.3583, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8026030368763557, | |
| "grad_norm": 0.1581087352900414, | |
| "learning_rate": 7.48905235066361e-05, | |
| "loss": 0.3184, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8047722342733189, | |
| "grad_norm": 0.15619092562938275, | |
| "learning_rate": 7.472601110393212e-05, | |
| "loss": 0.3248, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.806941431670282, | |
| "grad_norm": 0.1595073883740819, | |
| "learning_rate": 7.456114360650015e-05, | |
| "loss": 0.3324, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8091106290672451, | |
| "grad_norm": 0.16937228457587902, | |
| "learning_rate": 7.439592338203221e-05, | |
| "loss": 0.272, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.8112798264642083, | |
| "grad_norm": 0.21955900610207899, | |
| "learning_rate": 7.423035280328589e-05, | |
| "loss": 0.2876, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.8134490238611713, | |
| "grad_norm": 0.1622542075335067, | |
| "learning_rate": 7.406443424805031e-05, | |
| "loss": 0.3292, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8156182212581344, | |
| "grad_norm": 0.1539483896012786, | |
| "learning_rate": 7.389817009911188e-05, | |
| "loss": 0.3103, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.8177874186550976, | |
| "grad_norm": 0.15631416958623467, | |
| "learning_rate": 7.373156274422022e-05, | |
| "loss": 0.2877, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.8199566160520607, | |
| "grad_norm": 0.16629989037855605, | |
| "learning_rate": 7.356461457605373e-05, | |
| "loss": 0.2955, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.8221258134490239, | |
| "grad_norm": 0.17821592166706385, | |
| "learning_rate": 7.339732799218535e-05, | |
| "loss": 0.3395, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.824295010845987, | |
| "grad_norm": 0.16495627039143948, | |
| "learning_rate": 7.322970539504802e-05, | |
| "loss": 0.2975, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8264642082429501, | |
| "grad_norm": 0.1658072464206595, | |
| "learning_rate": 7.306174919190025e-05, | |
| "loss": 0.3178, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.8286334056399133, | |
| "grad_norm": 0.15286362948439744, | |
| "learning_rate": 7.28934617947915e-05, | |
| "loss": 0.2405, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.8308026030368764, | |
| "grad_norm": 0.25352133804652655, | |
| "learning_rate": 7.272484562052762e-05, | |
| "loss": 0.3318, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.8329718004338394, | |
| "grad_norm": 0.16883132171707005, | |
| "learning_rate": 7.255590309063604e-05, | |
| "loss": 0.2697, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.8351409978308026, | |
| "grad_norm": 0.17506322037825633, | |
| "learning_rate": 7.238663663133108e-05, | |
| "loss": 0.2947, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8373101952277657, | |
| "grad_norm": 0.162905554800095, | |
| "learning_rate": 7.221704867347901e-05, | |
| "loss": 0.316, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.8394793926247288, | |
| "grad_norm": 0.16905944662023994, | |
| "learning_rate": 7.204714165256325e-05, | |
| "loss": 0.3287, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.841648590021692, | |
| "grad_norm": 0.15092794187286074, | |
| "learning_rate": 7.187691800864936e-05, | |
| "loss": 0.2997, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.8438177874186551, | |
| "grad_norm": 0.1355944525377151, | |
| "learning_rate": 7.170638018634993e-05, | |
| "loss": 0.2278, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.8459869848156182, | |
| "grad_norm": 0.17675290778005076, | |
| "learning_rate": 7.153553063478953e-05, | |
| "loss": 0.3531, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8481561822125814, | |
| "grad_norm": 0.1534412441507167, | |
| "learning_rate": 7.136437180756954e-05, | |
| "loss": 0.2781, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.8503253796095445, | |
| "grad_norm": 0.16137646593619714, | |
| "learning_rate": 7.119290616273294e-05, | |
| "loss": 0.2834, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.8524945770065075, | |
| "grad_norm": 0.15944482254384695, | |
| "learning_rate": 7.10211361627289e-05, | |
| "loss": 0.2979, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.8546637744034707, | |
| "grad_norm": 0.1656541893638849, | |
| "learning_rate": 7.084906427437757e-05, | |
| "loss": 0.2898, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.8568329718004338, | |
| "grad_norm": 0.17951549268207853, | |
| "learning_rate": 7.06766929688345e-05, | |
| "loss": 0.4042, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8590021691973969, | |
| "grad_norm": 0.1566199701101273, | |
| "learning_rate": 7.050402472155526e-05, | |
| "loss": 0.3079, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.8611713665943601, | |
| "grad_norm": 0.15458950588314033, | |
| "learning_rate": 7.03310620122599e-05, | |
| "loss": 0.2778, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.8633405639913232, | |
| "grad_norm": 0.17104556871315288, | |
| "learning_rate": 7.015780732489717e-05, | |
| "loss": 0.3175, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.8655097613882863, | |
| "grad_norm": 0.19835635840780777, | |
| "learning_rate": 6.99842631476091e-05, | |
| "loss": 0.3446, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.8676789587852495, | |
| "grad_norm": 0.17218889670168014, | |
| "learning_rate": 6.981043197269505e-05, | |
| "loss": 0.3048, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8676789587852495, | |
| "eval_loss": 0.31443339586257935, | |
| "eval_runtime": 39.7245, | |
| "eval_samples_per_second": 0.478, | |
| "eval_steps_per_second": 0.126, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8698481561822126, | |
| "grad_norm": 0.16106337443031235, | |
| "learning_rate": 6.963631629657606e-05, | |
| "loss": 0.282, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.8720173535791758, | |
| "grad_norm": 0.1549710087992951, | |
| "learning_rate": 6.946191861975888e-05, | |
| "loss": 0.2923, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.8741865509761388, | |
| "grad_norm": 0.17143963315653743, | |
| "learning_rate": 6.928724144680022e-05, | |
| "loss": 0.304, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.8763557483731019, | |
| "grad_norm": 0.16661780813221244, | |
| "learning_rate": 6.911228728627059e-05, | |
| "loss": 0.3294, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.8785249457700651, | |
| "grad_norm": 0.16403696305382504, | |
| "learning_rate": 6.893705865071842e-05, | |
| "loss": 0.3014, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8806941431670282, | |
| "grad_norm": 0.15319848299652775, | |
| "learning_rate": 6.876155805663389e-05, | |
| "loss": 0.2691, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.8828633405639913, | |
| "grad_norm": 0.17118141480932222, | |
| "learning_rate": 6.858578802441288e-05, | |
| "loss": 0.3132, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.8850325379609545, | |
| "grad_norm": 0.16178546049978892, | |
| "learning_rate": 6.840975107832067e-05, | |
| "loss": 0.2564, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.8872017353579176, | |
| "grad_norm": 0.16291335526409656, | |
| "learning_rate": 6.823344974645576e-05, | |
| "loss": 0.2873, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.8893709327548807, | |
| "grad_norm": 0.1609692593321932, | |
| "learning_rate": 6.805688656071354e-05, | |
| "loss": 0.305, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8915401301518439, | |
| "grad_norm": 0.14901262397966952, | |
| "learning_rate": 6.788006405674992e-05, | |
| "loss": 0.2673, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.8937093275488069, | |
| "grad_norm": 0.16659840958516206, | |
| "learning_rate": 6.770298477394495e-05, | |
| "loss": 0.3277, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.89587852494577, | |
| "grad_norm": 0.1713410771401561, | |
| "learning_rate": 6.75256512553663e-05, | |
| "loss": 0.3714, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.8980477223427332, | |
| "grad_norm": 0.1818967831157803, | |
| "learning_rate": 6.734806604773277e-05, | |
| "loss": 0.3811, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.9002169197396963, | |
| "grad_norm": 0.21925123588651396, | |
| "learning_rate": 6.717023170137774e-05, | |
| "loss": 0.3056, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9023861171366594, | |
| "grad_norm": 0.15314422603554104, | |
| "learning_rate": 6.69921507702125e-05, | |
| "loss": 0.2837, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.9045553145336226, | |
| "grad_norm": 0.16258221254945943, | |
| "learning_rate": 6.681382581168956e-05, | |
| "loss": 0.3085, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.9067245119305857, | |
| "grad_norm": 0.16032298987917648, | |
| "learning_rate": 6.663525938676603e-05, | |
| "loss": 0.3223, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.9088937093275488, | |
| "grad_norm": 0.16085131851237103, | |
| "learning_rate": 6.645645405986665e-05, | |
| "loss": 0.303, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.911062906724512, | |
| "grad_norm": 0.1650727655299475, | |
| "learning_rate": 6.627741239884716e-05, | |
| "loss": 0.3276, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.913232104121475, | |
| "grad_norm": 0.15360660494909031, | |
| "learning_rate": 6.609813697495731e-05, | |
| "loss": 0.2751, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.9154013015184381, | |
| "grad_norm": 0.15157055567722505, | |
| "learning_rate": 6.591863036280398e-05, | |
| "loss": 0.3085, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.9175704989154013, | |
| "grad_norm": 0.15803099826725212, | |
| "learning_rate": 6.573889514031415e-05, | |
| "loss": 0.2939, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.9197396963123644, | |
| "grad_norm": 0.1585787128334511, | |
| "learning_rate": 6.555893388869793e-05, | |
| "loss": 0.2503, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.9219088937093276, | |
| "grad_norm": 0.17931795418506688, | |
| "learning_rate": 6.537874919241149e-05, | |
| "loss": 0.3886, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9240780911062907, | |
| "grad_norm": 0.15343352858299092, | |
| "learning_rate": 6.519834363911992e-05, | |
| "loss": 0.2793, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.9262472885032538, | |
| "grad_norm": 0.15328728201248645, | |
| "learning_rate": 6.501771981966007e-05, | |
| "loss": 0.2644, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.928416485900217, | |
| "grad_norm": 0.16064411420877225, | |
| "learning_rate": 6.483688032800337e-05, | |
| "loss": 0.3188, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.93058568329718, | |
| "grad_norm": 0.1679425352918729, | |
| "learning_rate": 6.465582776121852e-05, | |
| "loss": 0.3602, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.9327548806941431, | |
| "grad_norm": 0.16379595982132292, | |
| "learning_rate": 6.447456471943427e-05, | |
| "loss": 0.2769, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9349240780911063, | |
| "grad_norm": 0.15943476584744723, | |
| "learning_rate": 6.429309380580202e-05, | |
| "loss": 0.2702, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.9370932754880694, | |
| "grad_norm": 0.15321548417679473, | |
| "learning_rate": 6.411141762645846e-05, | |
| "loss": 0.2463, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.9392624728850325, | |
| "grad_norm": 0.15564677916963437, | |
| "learning_rate": 6.392953879048813e-05, | |
| "loss": 0.2885, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.9414316702819957, | |
| "grad_norm": 0.1631290118518175, | |
| "learning_rate": 6.374745990988598e-05, | |
| "loss": 0.2748, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.9436008676789588, | |
| "grad_norm": 0.14619225672799796, | |
| "learning_rate": 6.356518359951982e-05, | |
| "loss": 0.249, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9457700650759219, | |
| "grad_norm": 0.1571277122359372, | |
| "learning_rate": 6.338271247709278e-05, | |
| "loss": 0.2533, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.9479392624728851, | |
| "grad_norm": 0.1550916491663443, | |
| "learning_rate": 6.320004916310573e-05, | |
| "loss": 0.2677, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.9501084598698482, | |
| "grad_norm": 0.16661022388778246, | |
| "learning_rate": 6.301719628081965e-05, | |
| "loss": 0.2878, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.9522776572668112, | |
| "grad_norm": 0.2040573789571519, | |
| "learning_rate": 6.283415645621791e-05, | |
| "loss": 0.2703, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.9544468546637744, | |
| "grad_norm": 0.16469631306397306, | |
| "learning_rate": 6.265093231796864e-05, | |
| "loss": 0.3037, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9566160520607375, | |
| "grad_norm": 0.15770760260143657, | |
| "learning_rate": 6.246752649738686e-05, | |
| "loss": 0.2657, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.9587852494577006, | |
| "grad_norm": 0.15983199407271448, | |
| "learning_rate": 6.228394162839686e-05, | |
| "loss": 0.2764, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.9609544468546638, | |
| "grad_norm": 0.16489136986239505, | |
| "learning_rate": 6.210018034749421e-05, | |
| "loss": 0.2955, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.9631236442516269, | |
| "grad_norm": 0.22721431990817345, | |
| "learning_rate": 6.191624529370796e-05, | |
| "loss": 0.3395, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.96529284164859, | |
| "grad_norm": 0.157681168570052, | |
| "learning_rate": 6.173213910856277e-05, | |
| "loss": 0.3018, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9674620390455532, | |
| "grad_norm": 0.1576772973443425, | |
| "learning_rate": 6.154786443604098e-05, | |
| "loss": 0.3013, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.9696312364425163, | |
| "grad_norm": 0.1579242415527196, | |
| "learning_rate": 6.13634239225445e-05, | |
| "loss": 0.294, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.9718004338394793, | |
| "grad_norm": 0.20593023409205996, | |
| "learning_rate": 6.117882021685704e-05, | |
| "loss": 0.2883, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.9739696312364425, | |
| "grad_norm": 0.16203552803868196, | |
| "learning_rate": 6.099405597010585e-05, | |
| "loss": 0.2957, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.9761388286334056, | |
| "grad_norm": 0.1614999286165849, | |
| "learning_rate": 6.0809133835723774e-05, | |
| "loss": 0.2885, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9783080260303688, | |
| "grad_norm": 0.16900731974324282, | |
| "learning_rate": 6.0624056469411125e-05, | |
| "loss": 0.301, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.9804772234273319, | |
| "grad_norm": 0.17129193051244268, | |
| "learning_rate": 6.043882652909752e-05, | |
| "loss": 0.2701, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.982646420824295, | |
| "grad_norm": 0.19758216454445776, | |
| "learning_rate": 6.025344667490369e-05, | |
| "loss": 0.3732, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.9848156182212582, | |
| "grad_norm": 0.15788414206460427, | |
| "learning_rate": 6.006791956910334e-05, | |
| "loss": 0.286, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.9869848156182213, | |
| "grad_norm": 0.1686049946630881, | |
| "learning_rate": 5.9882247876084865e-05, | |
| "loss": 0.3423, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9891540130151844, | |
| "grad_norm": 0.1623316957404726, | |
| "learning_rate": 5.969643426231309e-05, | |
| "loss": 0.2972, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.9913232104121475, | |
| "grad_norm": 0.16633964946287502, | |
| "learning_rate": 5.951048139629105e-05, | |
| "loss": 0.3207, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.9934924078091106, | |
| "grad_norm": 0.15578806657191704, | |
| "learning_rate": 5.932439194852153e-05, | |
| "loss": 0.2821, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.9956616052060737, | |
| "grad_norm": 0.141680355386348, | |
| "learning_rate": 5.9138168591468845e-05, | |
| "loss": 0.211, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.9978308026030369, | |
| "grad_norm": 0.1626311176710574, | |
| "learning_rate": 5.8951813999520375e-05, | |
| "loss": 0.2776, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.2168432501312755, | |
| "learning_rate": 5.876533084894821e-05, | |
| "loss": 0.3189, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.002169197396963, | |
| "grad_norm": 0.15195564980233445, | |
| "learning_rate": 5.8578721817870666e-05, | |
| "loss": 0.2746, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.0043383947939262, | |
| "grad_norm": 0.1491438360530865, | |
| "learning_rate": 5.839198958621388e-05, | |
| "loss": 0.2053, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.0065075921908895, | |
| "grad_norm": 0.15367262427759873, | |
| "learning_rate": 5.820513683567328e-05, | |
| "loss": 0.2371, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.0086767895878526, | |
| "grad_norm": 0.1511433155050122, | |
| "learning_rate": 5.801816624967509e-05, | |
| "loss": 0.2213, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0108459869848156, | |
| "grad_norm": 0.1663892898536712, | |
| "learning_rate": 5.783108051333779e-05, | |
| "loss": 0.2576, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.0130151843817787, | |
| "grad_norm": 0.17685832695456444, | |
| "learning_rate": 5.764388231343356e-05, | |
| "loss": 0.3009, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.0151843817787418, | |
| "grad_norm": 0.21787472552900225, | |
| "learning_rate": 5.745657433834968e-05, | |
| "loss": 0.2636, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.017353579175705, | |
| "grad_norm": 0.19123373552730527, | |
| "learning_rate": 5.726915927804995e-05, | |
| "loss": 0.2743, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.0195227765726682, | |
| "grad_norm": 0.17768041272481136, | |
| "learning_rate": 5.7081639824036e-05, | |
| "loss": 0.2623, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0216919739696313, | |
| "grad_norm": 0.1697916576481495, | |
| "learning_rate": 5.6894018669308735e-05, | |
| "loss": 0.2272, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.0238611713665944, | |
| "grad_norm": 0.17650211499609383, | |
| "learning_rate": 5.670629850832956e-05, | |
| "loss": 0.2761, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.0260303687635575, | |
| "grad_norm": 0.18214913995299273, | |
| "learning_rate": 5.6518482036981725e-05, | |
| "loss": 0.2459, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.0281995661605206, | |
| "grad_norm": 0.17835571006990567, | |
| "learning_rate": 5.633057195253164e-05, | |
| "loss": 0.2468, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.0303687635574836, | |
| "grad_norm": 0.2069539317038426, | |
| "learning_rate": 5.614257095359009e-05, | |
| "loss": 0.3213, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.032537960954447, | |
| "grad_norm": 0.19136911570719517, | |
| "learning_rate": 5.5954481740073505e-05, | |
| "loss": 0.2823, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.03470715835141, | |
| "grad_norm": 0.19262401192871317, | |
| "learning_rate": 5.5766307013165156e-05, | |
| "loss": 0.2595, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.0368763557483731, | |
| "grad_norm": 0.1860152304131639, | |
| "learning_rate": 5.557804947527645e-05, | |
| "loss": 0.2801, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.0390455531453362, | |
| "grad_norm": 0.16924664525627553, | |
| "learning_rate": 5.5389711830007984e-05, | |
| "loss": 0.2308, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.0412147505422993, | |
| "grad_norm": 0.1816528885351813, | |
| "learning_rate": 5.5201296782110845e-05, | |
| "loss": 0.265, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0433839479392624, | |
| "grad_norm": 0.16993844286408233, | |
| "learning_rate": 5.501280703744769e-05, | |
| "loss": 0.2538, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.0455531453362257, | |
| "grad_norm": 0.1949073398717587, | |
| "learning_rate": 5.48242453029539e-05, | |
| "loss": 0.3056, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.0477223427331888, | |
| "grad_norm": 0.16891171117764792, | |
| "learning_rate": 5.463561428659875e-05, | |
| "loss": 0.2197, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.0498915401301518, | |
| "grad_norm": 0.18215169314661425, | |
| "learning_rate": 5.444691669734643e-05, | |
| "loss": 0.2447, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.052060737527115, | |
| "grad_norm": 0.182026157244086, | |
| "learning_rate": 5.425815524511726e-05, | |
| "loss": 0.2978, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.054229934924078, | |
| "grad_norm": 0.17881791802607705, | |
| "learning_rate": 5.406933264074866e-05, | |
| "loss": 0.2469, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.056399132321041, | |
| "grad_norm": 0.18807032896417783, | |
| "learning_rate": 5.3880451595956294e-05, | |
| "loss": 0.297, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.0585683297180044, | |
| "grad_norm": 0.20875267489125912, | |
| "learning_rate": 5.369151482329506e-05, | |
| "loss": 0.282, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.0607375271149675, | |
| "grad_norm": 0.19549344502300633, | |
| "learning_rate": 5.350252503612024e-05, | |
| "loss": 0.3833, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.0629067245119306, | |
| "grad_norm": 0.17937531323085554, | |
| "learning_rate": 5.331348494854841e-05, | |
| "loss": 0.2622, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0650759219088937, | |
| "grad_norm": 0.1870147791370534, | |
| "learning_rate": 5.3124397275418524e-05, | |
| "loss": 0.2322, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.0672451193058567, | |
| "grad_norm": 0.1951898037972042, | |
| "learning_rate": 5.2935264732252965e-05, | |
| "loss": 0.2357, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.06941431670282, | |
| "grad_norm": 0.18975302549273446, | |
| "learning_rate": 5.274609003521846e-05, | |
| "loss": 0.2419, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.0715835140997831, | |
| "grad_norm": 0.20439478227045751, | |
| "learning_rate": 5.255687590108711e-05, | |
| "loss": 0.3275, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.0737527114967462, | |
| "grad_norm": 0.18609479985163507, | |
| "learning_rate": 5.236762504719742e-05, | |
| "loss": 0.2754, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0759219088937093, | |
| "grad_norm": 0.1830303728930595, | |
| "learning_rate": 5.217834019141521e-05, | |
| "loss": 0.2595, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.0780911062906724, | |
| "grad_norm": 0.20315831733475495, | |
| "learning_rate": 5.1989024052094605e-05, | |
| "loss": 0.3317, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.0802603036876355, | |
| "grad_norm": 0.18316883083338042, | |
| "learning_rate": 5.1799679348039e-05, | |
| "loss": 0.2458, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.0824295010845988, | |
| "grad_norm": 0.2041065318820468, | |
| "learning_rate": 5.1610308798462016e-05, | |
| "loss": 0.3169, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.0845986984815619, | |
| "grad_norm": 0.1944264750547026, | |
| "learning_rate": 5.142091512294844e-05, | |
| "loss": 0.2903, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.086767895878525, | |
| "grad_norm": 0.18918880781763017, | |
| "learning_rate": 5.123150104141521e-05, | |
| "loss": 0.2638, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.088937093275488, | |
| "grad_norm": 0.18983712841034528, | |
| "learning_rate": 5.104206927407225e-05, | |
| "loss": 0.2871, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.0911062906724511, | |
| "grad_norm": 0.20226737440215853, | |
| "learning_rate": 5.085262254138353e-05, | |
| "loss": 0.3285, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.0932754880694142, | |
| "grad_norm": 0.19118754054883752, | |
| "learning_rate": 5.0663163564027935e-05, | |
| "loss": 0.2594, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.0954446854663775, | |
| "grad_norm": 0.2138619141242844, | |
| "learning_rate": 5.047369506286017e-05, | |
| "loss": 0.2832, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0976138828633406, | |
| "grad_norm": 0.3269054916680425, | |
| "learning_rate": 5.028421975887173e-05, | |
| "loss": 0.2988, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.0997830802603037, | |
| "grad_norm": 0.19040645241202558, | |
| "learning_rate": 5.00947403731518e-05, | |
| "loss": 0.2712, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.1019522776572668, | |
| "grad_norm": 0.1967133503261764, | |
| "learning_rate": 4.99052596268482e-05, | |
| "loss": 0.2815, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.1041214750542299, | |
| "grad_norm": 0.1903074378491903, | |
| "learning_rate": 4.9715780241128286e-05, | |
| "loss": 0.2737, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.106290672451193, | |
| "grad_norm": 0.18119699874786585, | |
| "learning_rate": 4.952630493713984e-05, | |
| "loss": 0.269, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.1084598698481563, | |
| "grad_norm": 0.221788539906692, | |
| "learning_rate": 4.9336836435972076e-05, | |
| "loss": 0.2707, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.1106290672451193, | |
| "grad_norm": 0.1928601828912626, | |
| "learning_rate": 4.914737745861646e-05, | |
| "loss": 0.2226, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.1127982646420824, | |
| "grad_norm": 0.21645221349069002, | |
| "learning_rate": 4.895793072592776e-05, | |
| "loss": 0.2286, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.1149674620390455, | |
| "grad_norm": 0.1970144418295259, | |
| "learning_rate": 4.8768498958584795e-05, | |
| "loss": 0.2825, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.1171366594360086, | |
| "grad_norm": 0.1957160462190446, | |
| "learning_rate": 4.8579084877051565e-05, | |
| "loss": 0.2391, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.119305856832972, | |
| "grad_norm": 0.21402378244369188, | |
| "learning_rate": 4.838969120153798e-05, | |
| "loss": 0.2555, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.121475054229935, | |
| "grad_norm": 0.21877951289590283, | |
| "learning_rate": 4.820032065196101e-05, | |
| "loss": 0.334, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.123644251626898, | |
| "grad_norm": 0.21929824905146836, | |
| "learning_rate": 4.801097594790539e-05, | |
| "loss": 0.2973, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.1258134490238612, | |
| "grad_norm": 0.20619197296358086, | |
| "learning_rate": 4.78216598085848e-05, | |
| "loss": 0.3018, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.1279826464208242, | |
| "grad_norm": 0.19555135005105484, | |
| "learning_rate": 4.763237495280258e-05, | |
| "loss": 0.2462, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1301518438177873, | |
| "grad_norm": 0.21114282630013761, | |
| "learning_rate": 4.74431240989129e-05, | |
| "loss": 0.327, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.1323210412147506, | |
| "grad_norm": 0.20234364960770707, | |
| "learning_rate": 4.725390996478155e-05, | |
| "loss": 0.266, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.1344902386117137, | |
| "grad_norm": 0.19961583213636722, | |
| "learning_rate": 4.706473526774705e-05, | |
| "loss": 0.255, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.1366594360086768, | |
| "grad_norm": 0.19946499474596735, | |
| "learning_rate": 4.6875602724581474e-05, | |
| "loss": 0.2567, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.13882863340564, | |
| "grad_norm": 0.190031071960285, | |
| "learning_rate": 4.668651505145161e-05, | |
| "loss": 0.2536, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.140997830802603, | |
| "grad_norm": 0.19912223902629478, | |
| "learning_rate": 4.649747496387976e-05, | |
| "loss": 0.2814, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.1431670281995663, | |
| "grad_norm": 0.2029251712939136, | |
| "learning_rate": 4.630848517670495e-05, | |
| "loss": 0.2357, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.1453362255965294, | |
| "grad_norm": 0.1906141032119321, | |
| "learning_rate": 4.611954840404371e-05, | |
| "loss": 0.2814, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.1475054229934925, | |
| "grad_norm": 0.188787160150689, | |
| "learning_rate": 4.593066735925135e-05, | |
| "loss": 0.2696, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1496746203904555, | |
| "grad_norm": 0.2343991376514014, | |
| "learning_rate": 4.574184475488274e-05, | |
| "loss": 0.2429, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.1518438177874186, | |
| "grad_norm": 0.19283832286439154, | |
| "learning_rate": 4.5553083302653576e-05, | |
| "loss": 0.2433, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.1540130151843817, | |
| "grad_norm": 0.21911781620446846, | |
| "learning_rate": 4.5364385713401256e-05, | |
| "loss": 0.2821, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.1561822125813448, | |
| "grad_norm": 0.22969903578133888, | |
| "learning_rate": 4.517575469704611e-05, | |
| "loss": 0.2776, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.158351409978308, | |
| "grad_norm": 0.20410973608962502, | |
| "learning_rate": 4.498719296255231e-05, | |
| "loss": 0.2549, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.1605206073752712, | |
| "grad_norm": 0.2146495328698227, | |
| "learning_rate": 4.4798703217889166e-05, | |
| "loss": 0.2809, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1626898047722343, | |
| "grad_norm": 0.20839631573723869, | |
| "learning_rate": 4.461028816999203e-05, | |
| "loss": 0.2974, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.1648590021691974, | |
| "grad_norm": 0.19237241347004658, | |
| "learning_rate": 4.442195052472357e-05, | |
| "loss": 0.2636, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.1670281995661604, | |
| "grad_norm": 0.18275212969570778, | |
| "learning_rate": 4.423369298683485e-05, | |
| "loss": 0.2346, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.1691973969631237, | |
| "grad_norm": 0.21224147212685224, | |
| "learning_rate": 4.404551825992651e-05, | |
| "loss": 0.2942, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.1713665943600868, | |
| "grad_norm": 0.1947203315252727, | |
| "learning_rate": 4.385742904640993e-05, | |
| "loss": 0.239, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.17353579175705, | |
| "grad_norm": 0.2014217555241961, | |
| "learning_rate": 4.366942804746837e-05, | |
| "loss": 0.282, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.175704989154013, | |
| "grad_norm": 0.20005402924569843, | |
| "learning_rate": 4.3481517963018294e-05, | |
| "loss": 0.2567, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.177874186550976, | |
| "grad_norm": 0.19683374608842036, | |
| "learning_rate": 4.329370149167046e-05, | |
| "loss": 0.2422, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.1800433839479392, | |
| "grad_norm": 0.18295647045924832, | |
| "learning_rate": 4.310598133069128e-05, | |
| "loss": 0.2608, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.1822125813449025, | |
| "grad_norm": 0.25320255904534206, | |
| "learning_rate": 4.291836017596401e-05, | |
| "loss": 0.2522, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1843817787418656, | |
| "grad_norm": 0.25012095918625965, | |
| "learning_rate": 4.273084072195008e-05, | |
| "loss": 0.2392, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.1865509761388287, | |
| "grad_norm": 0.2042418325652416, | |
| "learning_rate": 4.2543425661650325e-05, | |
| "loss": 0.2632, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.1887201735357917, | |
| "grad_norm": 0.20971549718156754, | |
| "learning_rate": 4.2356117686566464e-05, | |
| "loss": 0.3056, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.1908893709327548, | |
| "grad_norm": 0.22494235538454008, | |
| "learning_rate": 4.2168919486662225e-05, | |
| "loss": 0.1999, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.1930585683297181, | |
| "grad_norm": 0.19594451325509984, | |
| "learning_rate": 4.1981833750324934e-05, | |
| "loss": 0.2237, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1952277657266812, | |
| "grad_norm": 0.20048311843021904, | |
| "learning_rate": 4.179486316432674e-05, | |
| "loss": 0.2457, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.1973969631236443, | |
| "grad_norm": 0.2133231404323406, | |
| "learning_rate": 4.1608010413786145e-05, | |
| "loss": 0.2428, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.1995661605206074, | |
| "grad_norm": 0.22391429220016193, | |
| "learning_rate": 4.1421278182129345e-05, | |
| "loss": 0.2835, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.2017353579175705, | |
| "grad_norm": 0.20513974766270474, | |
| "learning_rate": 4.1234669151051814e-05, | |
| "loss": 0.2837, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.2039045553145336, | |
| "grad_norm": 0.21910779105815384, | |
| "learning_rate": 4.104818600047963e-05, | |
| "loss": 0.2552, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.2060737527114966, | |
| "grad_norm": 0.20347018099490702, | |
| "learning_rate": 4.0861831408531174e-05, | |
| "loss": 0.2825, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.20824295010846, | |
| "grad_norm": 0.20019781967770364, | |
| "learning_rate": 4.067560805147848e-05, | |
| "loss": 0.2646, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.210412147505423, | |
| "grad_norm": 0.19985925625011156, | |
| "learning_rate": 4.048951860370897e-05, | |
| "loss": 0.2339, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.2125813449023861, | |
| "grad_norm": 0.20804567399279866, | |
| "learning_rate": 4.030356573768691e-05, | |
| "loss": 0.2685, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.2147505422993492, | |
| "grad_norm": 0.210586914513503, | |
| "learning_rate": 4.0117752123915166e-05, | |
| "loss": 0.2949, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2169197396963123, | |
| "grad_norm": 0.19395560738783052, | |
| "learning_rate": 3.9932080430896674e-05, | |
| "loss": 0.2581, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.2190889370932756, | |
| "grad_norm": 0.19335081571244175, | |
| "learning_rate": 3.974655332509632e-05, | |
| "loss": 0.2398, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.2212581344902387, | |
| "grad_norm": 0.21139167236935025, | |
| "learning_rate": 3.956117347090249e-05, | |
| "loss": 0.294, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.2234273318872018, | |
| "grad_norm": 0.2136565272145805, | |
| "learning_rate": 3.937594353058888e-05, | |
| "loss": 0.249, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.2255965292841648, | |
| "grad_norm": 0.20983577033903353, | |
| "learning_rate": 3.9190866164276224e-05, | |
| "loss": 0.277, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.227765726681128, | |
| "grad_norm": 0.22523284294717572, | |
| "learning_rate": 3.900594402989416e-05, | |
| "loss": 0.2965, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.229934924078091, | |
| "grad_norm": 0.19361498352831719, | |
| "learning_rate": 3.8821179783142976e-05, | |
| "loss": 0.2468, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.2321041214750543, | |
| "grad_norm": 0.20349907006204035, | |
| "learning_rate": 3.863657607745551e-05, | |
| "loss": 0.2707, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.2342733188720174, | |
| "grad_norm": 0.21437394856709535, | |
| "learning_rate": 3.8452135563959035e-05, | |
| "loss": 0.2482, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.2364425162689805, | |
| "grad_norm": 0.4418854079923846, | |
| "learning_rate": 3.8267860891437224e-05, | |
| "loss": 0.2526, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2386117136659436, | |
| "grad_norm": 0.20107219218854192, | |
| "learning_rate": 3.8083754706292044e-05, | |
| "loss": 0.2302, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.2407809110629067, | |
| "grad_norm": 0.22317690892260889, | |
| "learning_rate": 3.7899819652505805e-05, | |
| "loss": 0.3018, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.2429501084598698, | |
| "grad_norm": 0.24225532136972353, | |
| "learning_rate": 3.771605837160315e-05, | |
| "loss": 0.2945, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.245119305856833, | |
| "grad_norm": 0.21311877910331786, | |
| "learning_rate": 3.753247350261314e-05, | |
| "loss": 0.268, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.2472885032537961, | |
| "grad_norm": 0.21276496906364534, | |
| "learning_rate": 3.734906768203137e-05, | |
| "loss": 0.2796, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.2494577006507592, | |
| "grad_norm": 0.21666315750271123, | |
| "learning_rate": 3.7165843543782094e-05, | |
| "loss": 0.3048, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.2516268980477223, | |
| "grad_norm": 0.2117830975027958, | |
| "learning_rate": 3.698280371918035e-05, | |
| "loss": 0.3134, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.2537960954446854, | |
| "grad_norm": 0.21204614791084325, | |
| "learning_rate": 3.679995083689427e-05, | |
| "loss": 0.2583, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.2559652928416485, | |
| "grad_norm": 0.20266543030388112, | |
| "learning_rate": 3.6617287522907215e-05, | |
| "loss": 0.2187, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.2581344902386118, | |
| "grad_norm": 0.19799515726039077, | |
| "learning_rate": 3.643481640048019e-05, | |
| "loss": 0.2486, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2603036876355749, | |
| "grad_norm": 0.22965267318434665, | |
| "learning_rate": 3.6252540090114014e-05, | |
| "loss": 0.3141, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.262472885032538, | |
| "grad_norm": 0.20300927451116837, | |
| "learning_rate": 3.607046120951187e-05, | |
| "loss": 0.2481, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.264642082429501, | |
| "grad_norm": 0.19473972219878735, | |
| "learning_rate": 3.588858237354154e-05, | |
| "loss": 0.2425, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.2668112798264641, | |
| "grad_norm": 0.19460999115969896, | |
| "learning_rate": 3.5706906194197995e-05, | |
| "loss": 0.2695, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.2689804772234274, | |
| "grad_norm": 0.2088031103817509, | |
| "learning_rate": 3.552543528056573e-05, | |
| "loss": 0.2894, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2711496746203905, | |
| "grad_norm": 0.20225624597683012, | |
| "learning_rate": 3.534417223878149e-05, | |
| "loss": 0.2532, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.2733188720173536, | |
| "grad_norm": 0.1925296097866438, | |
| "learning_rate": 3.516311967199664e-05, | |
| "loss": 0.2544, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.2754880694143167, | |
| "grad_norm": 0.19510001512503033, | |
| "learning_rate": 3.498228018033994e-05, | |
| "loss": 0.262, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.2776572668112798, | |
| "grad_norm": 0.2049108757868876, | |
| "learning_rate": 3.4801656360880083e-05, | |
| "loss": 0.2562, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.2798264642082429, | |
| "grad_norm": 0.2266927606829423, | |
| "learning_rate": 3.4621250807588524e-05, | |
| "loss": 0.3551, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.281995661605206, | |
| "grad_norm": 0.21569872794030454, | |
| "learning_rate": 3.444106611130209e-05, | |
| "loss": 0.2774, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.2841648590021693, | |
| "grad_norm": 0.21023989849968328, | |
| "learning_rate": 3.4261104859685865e-05, | |
| "loss": 0.2622, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.2863340563991323, | |
| "grad_norm": 0.19106084312223887, | |
| "learning_rate": 3.408136963719605e-05, | |
| "loss": 0.2252, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.2885032537960954, | |
| "grad_norm": 0.19995458008751701, | |
| "learning_rate": 3.39018630250427e-05, | |
| "loss": 0.2372, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.2906724511930585, | |
| "grad_norm": 0.22200947357348208, | |
| "learning_rate": 3.3722587601152855e-05, | |
| "loss": 0.2321, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.2928416485900218, | |
| "grad_norm": 0.1950829833237249, | |
| "learning_rate": 3.354354594013337e-05, | |
| "loss": 0.2241, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.295010845986985, | |
| "grad_norm": 0.20036238676413232, | |
| "learning_rate": 3.336474061323399e-05, | |
| "loss": 0.2055, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.297180043383948, | |
| "grad_norm": 0.2107825415915727, | |
| "learning_rate": 3.318617418831044e-05, | |
| "loss": 0.256, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.299349240780911, | |
| "grad_norm": 0.21989801828531605, | |
| "learning_rate": 3.3007849229787516e-05, | |
| "loss": 0.2764, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.3015184381778742, | |
| "grad_norm": 0.21680377382599092, | |
| "learning_rate": 3.282976829862227e-05, | |
| "loss": 0.2566, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3015184381778742, | |
| "eval_loss": 0.3054097592830658, | |
| "eval_runtime": 39.7394, | |
| "eval_samples_per_second": 0.478, | |
| "eval_steps_per_second": 0.126, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3036876355748372, | |
| "grad_norm": 0.20308783149132958, | |
| "learning_rate": 3.2651933952267245e-05, | |
| "loss": 0.2112, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.3058568329718003, | |
| "grad_norm": 0.21690298502452043, | |
| "learning_rate": 3.247434874463372e-05, | |
| "loss": 0.2649, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.3080260303687636, | |
| "grad_norm": 0.2055181727108075, | |
| "learning_rate": 3.2297015226055076e-05, | |
| "loss": 0.2824, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.3101952277657267, | |
| "grad_norm": 0.2320338793221242, | |
| "learning_rate": 3.211993594325009e-05, | |
| "loss": 0.2456, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.3123644251626898, | |
| "grad_norm": 0.20493923957441781, | |
| "learning_rate": 3.194311343928649e-05, | |
| "loss": 0.2366, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.314533622559653, | |
| "grad_norm": 0.2030392074075536, | |
| "learning_rate": 3.176655025354425e-05, | |
| "loss": 0.2508, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.316702819956616, | |
| "grad_norm": 0.2110656534399716, | |
| "learning_rate": 3.159024892167935e-05, | |
| "loss": 0.2502, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.3188720173535793, | |
| "grad_norm": 0.25854440170290666, | |
| "learning_rate": 3.141421197558713e-05, | |
| "loss": 0.3055, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.3210412147505424, | |
| "grad_norm": 0.19420699816129983, | |
| "learning_rate": 3.123844194336613e-05, | |
| "loss": 0.2271, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.3232104121475055, | |
| "grad_norm": 0.21496315608117952, | |
| "learning_rate": 3.1062941349281594e-05, | |
| "loss": 0.2673, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3253796095444685, | |
| "grad_norm": 0.20106583560494895, | |
| "learning_rate": 3.0887712713729435e-05, | |
| "loss": 0.2365, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.3275488069414316, | |
| "grad_norm": 0.21120123263668, | |
| "learning_rate": 3.071275855319979e-05, | |
| "loss": 0.2554, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.3297180043383947, | |
| "grad_norm": 0.21665345900294192, | |
| "learning_rate": 3.053808138024113e-05, | |
| "loss": 0.2701, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.3318872017353578, | |
| "grad_norm": 0.21025986355262039, | |
| "learning_rate": 3.036368370342396e-05, | |
| "loss": 0.243, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.334056399132321, | |
| "grad_norm": 0.2060375691889601, | |
| "learning_rate": 3.018956802730497e-05, | |
| "loss": 0.275, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.3362255965292842, | |
| "grad_norm": 0.22136533659676028, | |
| "learning_rate": 3.0015736852390918e-05, | |
| "loss": 0.257, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.3383947939262473, | |
| "grad_norm": 0.21449871481305618, | |
| "learning_rate": 2.984219267510285e-05, | |
| "loss": 0.2704, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.3405639913232104, | |
| "grad_norm": 0.24014701061684474, | |
| "learning_rate": 2.966893798774012e-05, | |
| "loss": 0.2891, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.3427331887201737, | |
| "grad_norm": 0.2244269288798775, | |
| "learning_rate": 2.9495975278444743e-05, | |
| "loss": 0.2935, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.3449023861171367, | |
| "grad_norm": 0.21096761572427544, | |
| "learning_rate": 2.9323307031165503e-05, | |
| "loss": 0.2739, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3470715835140998, | |
| "grad_norm": 0.20262736058046335, | |
| "learning_rate": 2.9150935725622434e-05, | |
| "loss": 0.2499, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.349240780911063, | |
| "grad_norm": 0.19820653616274597, | |
| "learning_rate": 2.8978863837271096e-05, | |
| "loss": 0.2393, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.351409978308026, | |
| "grad_norm": 0.23042991309507332, | |
| "learning_rate": 2.8807093837267062e-05, | |
| "loss": 0.302, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.353579175704989, | |
| "grad_norm": 0.1909580869184299, | |
| "learning_rate": 2.8635628192430457e-05, | |
| "loss": 0.2492, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.3557483731019522, | |
| "grad_norm": 0.1954261314272082, | |
| "learning_rate": 2.8464469365210476e-05, | |
| "loss": 0.2369, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3579175704989155, | |
| "grad_norm": 0.21435896485747777, | |
| "learning_rate": 2.829361981365008e-05, | |
| "loss": 0.2825, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.3600867678958786, | |
| "grad_norm": 0.20366176793808735, | |
| "learning_rate": 2.812308199135064e-05, | |
| "loss": 0.2455, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.3622559652928417, | |
| "grad_norm": 0.220096792471076, | |
| "learning_rate": 2.795285834743674e-05, | |
| "loss": 0.2281, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.3644251626898047, | |
| "grad_norm": 0.23939270387667466, | |
| "learning_rate": 2.7782951326521e-05, | |
| "loss": 0.3235, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.3665943600867678, | |
| "grad_norm": 0.2501946465822018, | |
| "learning_rate": 2.7613363368668933e-05, | |
| "loss": 0.3561, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3687635574837311, | |
| "grad_norm": 0.19776494315538312, | |
| "learning_rate": 2.744409690936396e-05, | |
| "loss": 0.235, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.3709327548806942, | |
| "grad_norm": 0.22972123156189964, | |
| "learning_rate": 2.7275154379472383e-05, | |
| "loss": 0.2781, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.3731019522776573, | |
| "grad_norm": 0.22520549969079728, | |
| "learning_rate": 2.7106538205208503e-05, | |
| "loss": 0.2952, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.3752711496746204, | |
| "grad_norm": 0.2056228650880094, | |
| "learning_rate": 2.6938250808099765e-05, | |
| "loss": 0.2337, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.3774403470715835, | |
| "grad_norm": 0.2286918375083098, | |
| "learning_rate": 2.677029460495199e-05, | |
| "loss": 0.2896, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3796095444685466, | |
| "grad_norm": 0.19803374847789326, | |
| "learning_rate": 2.6602672007814657e-05, | |
| "loss": 0.2445, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.3817787418655096, | |
| "grad_norm": 0.21204934763741792, | |
| "learning_rate": 2.6435385423946268e-05, | |
| "loss": 0.2565, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.383947939262473, | |
| "grad_norm": 0.2623356517904308, | |
| "learning_rate": 2.6268437255779795e-05, | |
| "loss": 0.2574, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.386117136659436, | |
| "grad_norm": 0.22828545407934486, | |
| "learning_rate": 2.610182990088813e-05, | |
| "loss": 0.3083, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.3882863340563991, | |
| "grad_norm": 0.21316635288288857, | |
| "learning_rate": 2.5935565751949708e-05, | |
| "loss": 0.2884, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.3904555314533622, | |
| "grad_norm": 0.18200487727356005, | |
| "learning_rate": 2.5769647196714115e-05, | |
| "loss": 0.1971, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.3926247288503255, | |
| "grad_norm": 0.20394759124623543, | |
| "learning_rate": 2.5604076617967797e-05, | |
| "loss": 0.2531, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.3947939262472886, | |
| "grad_norm": 0.2211940525129031, | |
| "learning_rate": 2.5438856393499856e-05, | |
| "loss": 0.2642, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.3969631236442517, | |
| "grad_norm": 0.21677701478466002, | |
| "learning_rate": 2.527398889606789e-05, | |
| "loss": 0.2471, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.3991323210412148, | |
| "grad_norm": 0.19805770191187302, | |
| "learning_rate": 2.5109476493363903e-05, | |
| "loss": 0.2275, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.4013015184381779, | |
| "grad_norm": 0.370645367139852, | |
| "learning_rate": 2.494532154798036e-05, | |
| "loss": 0.2652, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.403470715835141, | |
| "grad_norm": 0.4096140214680738, | |
| "learning_rate": 2.4781526417376167e-05, | |
| "loss": 0.2214, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.405639913232104, | |
| "grad_norm": 0.20027276095982202, | |
| "learning_rate": 2.4618093453842904e-05, | |
| "loss": 0.2283, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.407809110629067, | |
| "grad_norm": 0.2042256862448447, | |
| "learning_rate": 2.4455025004470983e-05, | |
| "loss": 0.2317, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.4099783080260304, | |
| "grad_norm": 0.2170785919045779, | |
| "learning_rate": 2.4292323411115963e-05, | |
| "loss": 0.2582, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.4121475054229935, | |
| "grad_norm": 0.5111593541353036, | |
| "learning_rate": 2.412999101036493e-05, | |
| "loss": 0.3233, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.4143167028199566, | |
| "grad_norm": 0.22766256947955918, | |
| "learning_rate": 2.3968030133502916e-05, | |
| "loss": 0.2548, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.4164859002169197, | |
| "grad_norm": 0.21004376919072829, | |
| "learning_rate": 2.380644310647944e-05, | |
| "loss": 0.2463, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.418655097613883, | |
| "grad_norm": 0.22454612498741602, | |
| "learning_rate": 2.364523224987508e-05, | |
| "loss": 0.2186, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.420824295010846, | |
| "grad_norm": 0.21635926190963323, | |
| "learning_rate": 2.3484399878868185e-05, | |
| "loss": 0.2574, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4229934924078091, | |
| "grad_norm": 0.20472709985073356, | |
| "learning_rate": 2.3323948303201586e-05, | |
| "loss": 0.2453, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.4251626898047722, | |
| "grad_norm": 0.2204401348193557, | |
| "learning_rate": 2.3163879827149448e-05, | |
| "loss": 0.2507, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.4273318872017353, | |
| "grad_norm": 0.24175915571290157, | |
| "learning_rate": 2.300419674948418e-05, | |
| "loss": 0.2907, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.4295010845986984, | |
| "grad_norm": 0.22445756380382137, | |
| "learning_rate": 2.2844901363443404e-05, | |
| "loss": 0.2791, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.4316702819956615, | |
| "grad_norm": 0.22386370995786847, | |
| "learning_rate": 2.2685995956697038e-05, | |
| "loss": 0.2779, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.4338394793926248, | |
| "grad_norm": 0.23249928031701667, | |
| "learning_rate": 2.2527482811314437e-05, | |
| "loss": 0.311, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.4360086767895879, | |
| "grad_norm": 0.2342495773652486, | |
| "learning_rate": 2.2369364203731618e-05, | |
| "loss": 0.2679, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.438177874186551, | |
| "grad_norm": 0.20675103233277622, | |
| "learning_rate": 2.221164240471857e-05, | |
| "loss": 0.2313, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.440347071583514, | |
| "grad_norm": 0.20528823101739593, | |
| "learning_rate": 2.205431967934664e-05, | |
| "loss": 0.2492, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.4425162689804774, | |
| "grad_norm": 0.2128126904380516, | |
| "learning_rate": 2.1897398286956012e-05, | |
| "loss": 0.2777, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4446854663774404, | |
| "grad_norm": 0.21518099561214468, | |
| "learning_rate": 2.1740880481123238e-05, | |
| "loss": 0.2518, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.4468546637744035, | |
| "grad_norm": 0.2186083154676726, | |
| "learning_rate": 2.1584768509628922e-05, | |
| "loss": 0.2453, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.4490238611713666, | |
| "grad_norm": 0.25175528122891083, | |
| "learning_rate": 2.142906461442537e-05, | |
| "loss": 0.3037, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.4511930585683297, | |
| "grad_norm": 0.22763386297900118, | |
| "learning_rate": 2.1273771031604456e-05, | |
| "loss": 0.2644, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.4533622559652928, | |
| "grad_norm": 0.2295090408465574, | |
| "learning_rate": 2.1118889991365476e-05, | |
| "loss": 0.2686, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.4555314533622559, | |
| "grad_norm": 0.20891584135932628, | |
| "learning_rate": 2.096442371798313e-05, | |
| "loss": 0.2204, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.457700650759219, | |
| "grad_norm": 0.2254247187316148, | |
| "learning_rate": 2.0810374429775565e-05, | |
| "loss": 0.2913, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.4598698481561823, | |
| "grad_norm": 0.24216942402374153, | |
| "learning_rate": 2.0656744339072542e-05, | |
| "loss": 0.3349, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.4620390455531453, | |
| "grad_norm": 0.22222297204476354, | |
| "learning_rate": 2.0503535652183643e-05, | |
| "loss": 0.2793, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.4642082429501084, | |
| "grad_norm": 0.20870048389264126, | |
| "learning_rate": 2.035075056936659e-05, | |
| "loss": 0.2655, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4663774403470715, | |
| "grad_norm": 0.22841755003782033, | |
| "learning_rate": 2.0198391284795664e-05, | |
| "loss": 0.3155, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.4685466377440348, | |
| "grad_norm": 0.22269122758511214, | |
| "learning_rate": 2.004645998653017e-05, | |
| "loss": 0.2491, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.470715835140998, | |
| "grad_norm": 0.22332389169722103, | |
| "learning_rate": 1.9894958856483026e-05, | |
| "loss": 0.2648, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.472885032537961, | |
| "grad_norm": 0.214833703844543, | |
| "learning_rate": 1.974389007038942e-05, | |
| "loss": 0.2609, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.475054229934924, | |
| "grad_norm": 0.20682197278555356, | |
| "learning_rate": 1.9593255797775577e-05, | |
| "loss": 0.2328, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4772234273318872, | |
| "grad_norm": 0.23230767426539825, | |
| "learning_rate": 1.9443058201927588e-05, | |
| "loss": 0.2347, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.4793926247288502, | |
| "grad_norm": 0.2261449038214272, | |
| "learning_rate": 1.9293299439860396e-05, | |
| "loss": 0.2758, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.4815618221258133, | |
| "grad_norm": 0.3168804778663666, | |
| "learning_rate": 1.9143981662286665e-05, | |
| "loss": 0.273, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.4837310195227766, | |
| "grad_norm": 0.22898408446614574, | |
| "learning_rate": 1.8995107013586137e-05, | |
| "loss": 0.2713, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.4859002169197397, | |
| "grad_norm": 0.22912219172967715, | |
| "learning_rate": 1.8846677631774575e-05, | |
| "loss": 0.2937, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4880694143167028, | |
| "grad_norm": 0.20900331783691264, | |
| "learning_rate": 1.869869564847329e-05, | |
| "loss": 0.2269, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.490238611713666, | |
| "grad_norm": 0.21015606821149282, | |
| "learning_rate": 1.8551163188878313e-05, | |
| "loss": 0.2299, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.4924078091106292, | |
| "grad_norm": 0.2250692099687625, | |
| "learning_rate": 1.840408237173011e-05, | |
| "loss": 0.2892, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.4945770065075923, | |
| "grad_norm": 0.2145016212764728, | |
| "learning_rate": 1.8257455309282882e-05, | |
| "loss": 0.2476, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.4967462039045554, | |
| "grad_norm": 0.2236334148717297, | |
| "learning_rate": 1.811128410727454e-05, | |
| "loss": 0.2259, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4989154013015185, | |
| "grad_norm": 0.22580326511659854, | |
| "learning_rate": 1.7965570864896138e-05, | |
| "loss": 0.2611, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.5010845986984815, | |
| "grad_norm": 0.23566517211163768, | |
| "learning_rate": 1.7820317674762034e-05, | |
| "loss": 0.3256, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.5032537960954446, | |
| "grad_norm": 0.2387014338194217, | |
| "learning_rate": 1.767552662287955e-05, | |
| "loss": 0.2169, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.5054229934924077, | |
| "grad_norm": 0.2131558457240097, | |
| "learning_rate": 1.7531199788619305e-05, | |
| "loss": 0.2166, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.5075921908893708, | |
| "grad_norm": 0.21663448852517472, | |
| "learning_rate": 1.738733924468507e-05, | |
| "loss": 0.2473, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.509761388286334, | |
| "grad_norm": 0.2302024457559936, | |
| "learning_rate": 1.7243947057084252e-05, | |
| "loss": 0.2579, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.5119305856832972, | |
| "grad_norm": 0.21162459671820377, | |
| "learning_rate": 1.7101025285097988e-05, | |
| "loss": 0.2367, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.5140997830802603, | |
| "grad_norm": 0.22899234818180603, | |
| "learning_rate": 1.695857598125183e-05, | |
| "loss": 0.2487, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.5162689804772236, | |
| "grad_norm": 0.21180166251652038, | |
| "learning_rate": 1.681660119128598e-05, | |
| "loss": 0.2454, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.5184381778741867, | |
| "grad_norm": 0.21900430622153325, | |
| "learning_rate": 1.6675102954126204e-05, | |
| "loss": 0.2505, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5206073752711498, | |
| "grad_norm": 0.22388111292395452, | |
| "learning_rate": 1.6534083301854287e-05, | |
| "loss": 0.256, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.5227765726681128, | |
| "grad_norm": 0.19613947069940627, | |
| "learning_rate": 1.639354425967904e-05, | |
| "loss": 0.2029, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.524945770065076, | |
| "grad_norm": 0.20905100864213258, | |
| "learning_rate": 1.6253487845907122e-05, | |
| "loss": 0.2275, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.527114967462039, | |
| "grad_norm": 0.21880038965460166, | |
| "learning_rate": 1.6113916071914082e-05, | |
| "loss": 0.2617, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.529284164859002, | |
| "grad_norm": 0.21208440823468386, | |
| "learning_rate": 1.5974830942115472e-05, | |
| "loss": 0.242, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5314533622559652, | |
| "grad_norm": 0.23456256487858015, | |
| "learning_rate": 1.5836234453938054e-05, | |
| "loss": 0.2549, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.5336225596529283, | |
| "grad_norm": 0.2396806217154987, | |
| "learning_rate": 1.5698128597791122e-05, | |
| "loss": 0.3393, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.5357917570498916, | |
| "grad_norm": 0.2031182271876605, | |
| "learning_rate": 1.5560515357037898e-05, | |
| "loss": 0.2152, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.5379609544468547, | |
| "grad_norm": 0.22000222658603283, | |
| "learning_rate": 1.542339670796712e-05, | |
| "loss": 0.2648, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.5401301518438177, | |
| "grad_norm": 0.265809393717544, | |
| "learning_rate": 1.528677461976451e-05, | |
| "loss": 0.192, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.542299349240781, | |
| "grad_norm": 0.20877625917509104, | |
| "learning_rate": 1.5150651054484705e-05, | |
| "loss": 0.2257, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.5444685466377441, | |
| "grad_norm": 0.21683968747146132, | |
| "learning_rate": 1.5015027967022838e-05, | |
| "loss": 0.2702, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.5466377440347072, | |
| "grad_norm": 0.20896474897637557, | |
| "learning_rate": 1.4879907305086721e-05, | |
| "loss": 0.2459, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.5488069414316703, | |
| "grad_norm": 0.2615773501556193, | |
| "learning_rate": 1.4745291009168616e-05, | |
| "loss": 0.2782, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.5509761388286334, | |
| "grad_norm": 0.22282625503762615, | |
| "learning_rate": 1.461118101251761e-05, | |
| "loss": 0.2885, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.5531453362255965, | |
| "grad_norm": 0.21801422944358295, | |
| "learning_rate": 1.4477579241111616e-05, | |
| "loss": 0.2406, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.5553145336225596, | |
| "grad_norm": 0.22073898585184953, | |
| "learning_rate": 1.4344487613629958e-05, | |
| "loss": 0.2656, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.5574837310195226, | |
| "grad_norm": 0.2187048550685909, | |
| "learning_rate": 1.4211908041425565e-05, | |
| "loss": 0.2613, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.559652928416486, | |
| "grad_norm": 0.2101381320422253, | |
| "learning_rate": 1.4079842428497764e-05, | |
| "loss": 0.2333, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.561822125813449, | |
| "grad_norm": 0.20391298465277313, | |
| "learning_rate": 1.3948292671464708e-05, | |
| "loss": 0.2343, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5639913232104121, | |
| "grad_norm": 0.22749006672655192, | |
| "learning_rate": 1.3817260659536368e-05, | |
| "loss": 0.2563, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.5661605206073754, | |
| "grad_norm": 0.22567622662999204, | |
| "learning_rate": 1.368674827448716e-05, | |
| "loss": 0.3026, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.5683297180043385, | |
| "grad_norm": 0.22202307546172664, | |
| "learning_rate": 1.355675739062916e-05, | |
| "loss": 0.2365, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.5704989154013016, | |
| "grad_norm": 0.20324438958594626, | |
| "learning_rate": 1.3427289874784965e-05, | |
| "loss": 0.2113, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.5726681127982647, | |
| "grad_norm": 0.21573774200754933, | |
| "learning_rate": 1.3298347586261101e-05, | |
| "loss": 0.2256, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.5748373101952278, | |
| "grad_norm": 0.2107630411490319, | |
| "learning_rate": 1.3169932376821087e-05, | |
| "loss": 0.2243, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.5770065075921909, | |
| "grad_norm": 0.2250209366455123, | |
| "learning_rate": 1.3042046090659082e-05, | |
| "loss": 0.2311, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.579175704989154, | |
| "grad_norm": 0.21327462482312162, | |
| "learning_rate": 1.2914690564373172e-05, | |
| "loss": 0.2025, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.581344902386117, | |
| "grad_norm": 0.21048416380580592, | |
| "learning_rate": 1.278786762693921e-05, | |
| "loss": 0.2505, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.58351409978308, | |
| "grad_norm": 0.23805779361424165, | |
| "learning_rate": 1.2661579099684345e-05, | |
| "loss": 0.2993, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5856832971800434, | |
| "grad_norm": 0.2236199058892429, | |
| "learning_rate": 1.2535826796261058e-05, | |
| "loss": 0.2559, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.5878524945770065, | |
| "grad_norm": 0.2267692199206878, | |
| "learning_rate": 1.2410612522620923e-05, | |
| "loss": 0.2631, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.5900216919739696, | |
| "grad_norm": 0.22527317437513994, | |
| "learning_rate": 1.2285938076988879e-05, | |
| "loss": 0.2381, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.592190889370933, | |
| "grad_norm": 0.20643186489302753, | |
| "learning_rate": 1.2161805249837189e-05, | |
| "loss": 0.2261, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.594360086767896, | |
| "grad_norm": 0.2363410474128828, | |
| "learning_rate": 1.2038215823859944e-05, | |
| "loss": 0.2848, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.596529284164859, | |
| "grad_norm": 0.22064343187390295, | |
| "learning_rate": 1.1915171573947231e-05, | |
| "loss": 0.2114, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.5986984815618221, | |
| "grad_norm": 0.2241110750390057, | |
| "learning_rate": 1.179267426715988e-05, | |
| "loss": 0.273, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.6008676789587852, | |
| "grad_norm": 0.24065816206033924, | |
| "learning_rate": 1.1670725662703907e-05, | |
| "loss": 0.2667, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.6030368763557483, | |
| "grad_norm": 0.216540199511313, | |
| "learning_rate": 1.1549327511905322e-05, | |
| "loss": 0.2603, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.6052060737527114, | |
| "grad_norm": 0.22867224484504703, | |
| "learning_rate": 1.1428481558184984e-05, | |
| "loss": 0.2835, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.6073752711496745, | |
| "grad_norm": 0.21664360434007227, | |
| "learning_rate": 1.1308189537033532e-05, | |
| "loss": 0.2799, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.6095444685466378, | |
| "grad_norm": 0.21600464380836015, | |
| "learning_rate": 1.1188453175986502e-05, | |
| "loss": 0.2555, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.6117136659436009, | |
| "grad_norm": 0.22003089603782433, | |
| "learning_rate": 1.1069274194599477e-05, | |
| "loss": 0.2764, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.613882863340564, | |
| "grad_norm": 0.21966503118494438, | |
| "learning_rate": 1.0950654304423408e-05, | |
| "loss": 0.243, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.6160520607375273, | |
| "grad_norm": 0.2134291281940422, | |
| "learning_rate": 1.0832595208980052e-05, | |
| "loss": 0.2235, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.6182212581344904, | |
| "grad_norm": 0.22637670128987095, | |
| "learning_rate": 1.0715098603737473e-05, | |
| "loss": 0.2597, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.6203904555314534, | |
| "grad_norm": 0.21648611339436485, | |
| "learning_rate": 1.0598166176085722e-05, | |
| "loss": 0.2542, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.6225596529284165, | |
| "grad_norm": 0.23629777995443355, | |
| "learning_rate": 1.0481799605312598e-05, | |
| "loss": 0.3228, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.6247288503253796, | |
| "grad_norm": 0.21725498937866275, | |
| "learning_rate": 1.0366000562579509e-05, | |
| "loss": 0.2344, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.6268980477223427, | |
| "grad_norm": 0.21952448011527256, | |
| "learning_rate": 1.0250770710897512e-05, | |
| "loss": 0.3131, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6290672451193058, | |
| "grad_norm": 0.2011634686461227, | |
| "learning_rate": 1.0136111705103384e-05, | |
| "loss": 0.2069, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.6312364425162689, | |
| "grad_norm": 0.22463127569065536, | |
| "learning_rate": 1.0022025191835905e-05, | |
| "loss": 0.2281, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.633405639913232, | |
| "grad_norm": 0.2062264163167854, | |
| "learning_rate": 9.90851280951216e-06, | |
| "loss": 0.2179, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.6355748373101953, | |
| "grad_norm": 0.22691208712294778, | |
| "learning_rate": 9.795576188304068e-06, | |
| "loss": 0.3075, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.6377440347071583, | |
| "grad_norm": 0.21132869155040268, | |
| "learning_rate": 9.68321695011491e-06, | |
| "loss": 0.1903, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6399132321041214, | |
| "grad_norm": 0.21551342495874198, | |
| "learning_rate": 9.571436708556076e-06, | |
| "loss": 0.221, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.6420824295010847, | |
| "grad_norm": 0.2174975697740617, | |
| "learning_rate": 9.460237068923883e-06, | |
| "loss": 0.2587, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.6442516268980478, | |
| "grad_norm": 0.2160546965912458, | |
| "learning_rate": 9.34961962817652e-06, | |
| "loss": 0.2303, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.646420824295011, | |
| "grad_norm": 0.2376163843745843, | |
| "learning_rate": 9.239585974911074e-06, | |
| "loss": 0.3243, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.648590021691974, | |
| "grad_norm": 0.21539719738458052, | |
| "learning_rate": 9.130137689340839e-06, | |
| "loss": 0.2571, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.650759219088937, | |
| "grad_norm": 0.2410719861687306, | |
| "learning_rate": 9.021276343272434e-06, | |
| "loss": 0.2633, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.6529284164859002, | |
| "grad_norm": 0.2079469970240713, | |
| "learning_rate": 8.913003500083438e-06, | |
| "loss": 0.2379, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.6550976138828633, | |
| "grad_norm": 0.22772263664943265, | |
| "learning_rate": 8.805320714699728e-06, | |
| "loss": 0.2395, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.6572668112798263, | |
| "grad_norm": 0.20307660213818468, | |
| "learning_rate": 8.698229533573338e-06, | |
| "loss": 0.1864, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.6594360086767896, | |
| "grad_norm": 0.2148519135414743, | |
| "learning_rate": 8.591731494660132e-06, | |
| "loss": 0.2053, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6616052060737527, | |
| "grad_norm": 0.229814247011586, | |
| "learning_rate": 8.485828127397749e-06, | |
| "loss": 0.2875, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.6637744034707158, | |
| "grad_norm": 0.21643960802023465, | |
| "learning_rate": 8.380520952683646e-06, | |
| "loss": 0.2599, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.6659436008676791, | |
| "grad_norm": 0.22244934877456704, | |
| "learning_rate": 8.275811482853245e-06, | |
| "loss": 0.2598, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.6681127982646422, | |
| "grad_norm": 0.23343908563620214, | |
| "learning_rate": 8.17170122165824e-06, | |
| "loss": 0.3098, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.6702819956616053, | |
| "grad_norm": 0.19803997345115992, | |
| "learning_rate": 8.068191664244945e-06, | |
| "loss": 0.2029, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.6724511930585684, | |
| "grad_norm": 0.2190516176521641, | |
| "learning_rate": 7.965284297132896e-06, | |
| "loss": 0.241, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.6746203904555315, | |
| "grad_norm": 0.21290567846006564, | |
| "learning_rate": 7.862980598193442e-06, | |
| "loss": 0.2063, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.6767895878524945, | |
| "grad_norm": 0.23699837862628473, | |
| "learning_rate": 7.761282036628548e-06, | |
| "loss": 0.3082, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.6789587852494576, | |
| "grad_norm": 0.22947643986082064, | |
| "learning_rate": 7.660190072949692e-06, | |
| "loss": 0.2628, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.6811279826464207, | |
| "grad_norm": 0.23210641035367943, | |
| "learning_rate": 7.559706158956898e-06, | |
| "loss": 0.2715, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6832971800433838, | |
| "grad_norm": 0.2294766385634235, | |
| "learning_rate": 7.459831737717859e-06, | |
| "loss": 0.2749, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.685466377440347, | |
| "grad_norm": 0.20107115872856784, | |
| "learning_rate": 7.360568243547261e-06, | |
| "loss": 0.219, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.6876355748373102, | |
| "grad_norm": 0.20113176199889593, | |
| "learning_rate": 7.261917101986127e-06, | |
| "loss": 0.2074, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.6898047722342733, | |
| "grad_norm": 0.23609135019264865, | |
| "learning_rate": 7.163879729781392e-06, | |
| "loss": 0.2667, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.6919739696312366, | |
| "grad_norm": 0.2228940425973063, | |
| "learning_rate": 7.066457534865528e-06, | |
| "loss": 0.2643, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6941431670281997, | |
| "grad_norm": 0.23796760069402623, | |
| "learning_rate": 6.969651916336334e-06, | |
| "loss": 0.2249, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.6963123644251628, | |
| "grad_norm": 0.2327272256132086, | |
| "learning_rate": 6.8734642644368576e-06, | |
| "loss": 0.2543, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.6984815618221258, | |
| "grad_norm": 0.2472213046502843, | |
| "learning_rate": 6.7778959605353906e-06, | |
| "loss": 0.2585, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.700650759219089, | |
| "grad_norm": 0.22682131480226772, | |
| "learning_rate": 6.682948377105686e-06, | |
| "loss": 0.226, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.702819956616052, | |
| "grad_norm": 0.23469843862434467, | |
| "learning_rate": 6.588622877707196e-06, | |
| "loss": 0.2933, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.704989154013015, | |
| "grad_norm": 0.2244645457364394, | |
| "learning_rate": 6.49492081696551e-06, | |
| "loss": 0.2965, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.7071583514099782, | |
| "grad_norm": 0.20869135290493288, | |
| "learning_rate": 6.401843540552921e-06, | |
| "loss": 0.2295, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.7093275488069413, | |
| "grad_norm": 0.21494691194019463, | |
| "learning_rate": 6.309392385169066e-06, | |
| "loss": 0.2418, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.7114967462039046, | |
| "grad_norm": 0.2521558215114822, | |
| "learning_rate": 6.217568678521746e-06, | |
| "loss": 0.3596, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.7136659436008677, | |
| "grad_norm": 0.21627894705194553, | |
| "learning_rate": 6.126373739307856e-06, | |
| "loss": 0.2358, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.715835140997831, | |
| "grad_norm": 0.2520399170899315, | |
| "learning_rate": 6.035808877194454e-06, | |
| "loss": 0.3594, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.718004338394794, | |
| "grad_norm": 0.22790741465610556, | |
| "learning_rate": 5.945875392799944e-06, | |
| "loss": 0.3138, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.7201735357917571, | |
| "grad_norm": 0.21149040475910053, | |
| "learning_rate": 5.8565745776754e-06, | |
| "loss": 0.2251, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.7223427331887202, | |
| "grad_norm": 0.23452143004512307, | |
| "learning_rate": 5.7679077142860135e-06, | |
| "loss": 0.2652, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.7245119305856833, | |
| "grad_norm": 0.22627084576919643, | |
| "learning_rate": 5.679876075992685e-06, | |
| "loss": 0.2856, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7266811279826464, | |
| "grad_norm": 0.23065836475523283, | |
| "learning_rate": 5.592480927033733e-06, | |
| "loss": 0.2407, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.7288503253796095, | |
| "grad_norm": 0.21666121524543802, | |
| "learning_rate": 5.505723522506734e-06, | |
| "loss": 0.2533, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.7310195227765726, | |
| "grad_norm": 0.21851910440097064, | |
| "learning_rate": 5.419605108350501e-06, | |
| "loss": 0.1987, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.7331887201735356, | |
| "grad_norm": 0.22502982751463446, | |
| "learning_rate": 5.334126921327193e-06, | |
| "loss": 0.2522, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.735357917570499, | |
| "grad_norm": 0.24282939171015397, | |
| "learning_rate": 5.249290189004552e-06, | |
| "loss": 0.3036, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.735357917570499, | |
| "eval_loss": 0.2993714511394501, | |
| "eval_runtime": 39.7253, | |
| "eval_samples_per_second": 0.478, | |
| "eval_steps_per_second": 0.126, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.737527114967462, | |
| "grad_norm": 0.22684517743489496, | |
| "learning_rate": 5.165096129738267e-06, | |
| "loss": 0.2412, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.7396963123644251, | |
| "grad_norm": 0.21496799405491135, | |
| "learning_rate": 5.081545952654493e-06, | |
| "loss": 0.2373, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.7418655097613884, | |
| "grad_norm": 0.22310608285043565, | |
| "learning_rate": 4.998640857632464e-06, | |
| "loss": 0.2496, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.7440347071583515, | |
| "grad_norm": 0.2320457312319671, | |
| "learning_rate": 4.916382035287276e-06, | |
| "loss": 0.2582, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.7462039045553146, | |
| "grad_norm": 0.22615551698433095, | |
| "learning_rate": 4.8347706669527985e-06, | |
| "loss": 0.2548, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7483731019522777, | |
| "grad_norm": 0.23006147126943094, | |
| "learning_rate": 4.7538079246646825e-06, | |
| "loss": 0.2776, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.7505422993492408, | |
| "grad_norm": 0.22519218682482992, | |
| "learning_rate": 4.673494971143538e-06, | |
| "loss": 0.2457, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.7527114967462039, | |
| "grad_norm": 0.21484715264028661, | |
| "learning_rate": 4.5938329597782824e-06, | |
| "loss": 0.2163, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.754880694143167, | |
| "grad_norm": 0.22724812121756507, | |
| "learning_rate": 4.514823034609455e-06, | |
| "loss": 0.2568, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.75704989154013, | |
| "grad_norm": 0.2192304364191693, | |
| "learning_rate": 4.436466330312966e-06, | |
| "loss": 0.235, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.7592190889370931, | |
| "grad_norm": 0.1833973092649833, | |
| "learning_rate": 4.358763972183599e-06, | |
| "loss": 0.166, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.7613882863340564, | |
| "grad_norm": 0.22170941113081136, | |
| "learning_rate": 4.281717076119057e-06, | |
| "loss": 0.2543, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.7635574837310195, | |
| "grad_norm": 0.2181063063924005, | |
| "learning_rate": 4.205326748603744e-06, | |
| "loss": 0.2231, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.7657266811279828, | |
| "grad_norm": 0.23650416189819906, | |
| "learning_rate": 4.129594086693012e-06, | |
| "loss": 0.2647, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.767895878524946, | |
| "grad_norm": 0.23715186447220826, | |
| "learning_rate": 4.0545201779973564e-06, | |
| "loss": 0.2395, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.770065075921909, | |
| "grad_norm": 0.25382512329158124, | |
| "learning_rate": 3.980106100666781e-06, | |
| "loss": 0.2909, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.772234273318872, | |
| "grad_norm": 0.21476601477039556, | |
| "learning_rate": 3.906352923375345e-06, | |
| "loss": 0.2326, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.7744034707158352, | |
| "grad_norm": 0.31142043160518457, | |
| "learning_rate": 3.833261705305796e-06, | |
| "loss": 0.2506, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.7765726681127982, | |
| "grad_norm": 0.23754127992716897, | |
| "learning_rate": 3.7608334961343695e-06, | |
| "loss": 0.2753, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.7787418655097613, | |
| "grad_norm": 0.2606633745028959, | |
| "learning_rate": 3.6890693360157105e-06, | |
| "loss": 0.2735, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7809110629067244, | |
| "grad_norm": 0.2308719975444426, | |
| "learning_rate": 3.6179702555679264e-06, | |
| "loss": 0.2391, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.7830802603036875, | |
| "grad_norm": 0.20821914212005022, | |
| "learning_rate": 3.5475372758577984e-06, | |
| "loss": 0.2041, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.7852494577006508, | |
| "grad_norm": 0.27553949669527766, | |
| "learning_rate": 3.4777714083861268e-06, | |
| "loss": 0.2641, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.7874186550976139, | |
| "grad_norm": 0.2298326829683291, | |
| "learning_rate": 3.4086736550731747e-06, | |
| "loss": 0.2694, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.789587852494577, | |
| "grad_norm": 0.23725507049973954, | |
| "learning_rate": 3.340245008244308e-06, | |
| "loss": 0.2672, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7917570498915403, | |
| "grad_norm": 0.23615628058747723, | |
| "learning_rate": 3.272486450615725e-06, | |
| "loss": 0.298, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.7939262472885034, | |
| "grad_norm": 0.24396047621014613, | |
| "learning_rate": 3.2053989552803586e-06, | |
| "loss": 0.3171, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.7960954446854664, | |
| "grad_norm": 0.22614687587834664, | |
| "learning_rate": 3.1389834856938916e-06, | |
| "loss": 0.2379, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.7982646420824295, | |
| "grad_norm": 0.22756878563712518, | |
| "learning_rate": 3.0732409956609197e-06, | |
| "loss": 0.2636, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.8004338394793926, | |
| "grad_norm": 0.19317698277364906, | |
| "learning_rate": 3.0081724293212653e-06, | |
| "loss": 0.1733, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.8026030368763557, | |
| "grad_norm": 0.2290227505004182, | |
| "learning_rate": 2.943778721136403e-06, | |
| "loss": 0.281, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.8047722342733188, | |
| "grad_norm": 0.21862342101596383, | |
| "learning_rate": 2.8800607958760497e-06, | |
| "loss": 0.2335, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.8069414316702819, | |
| "grad_norm": 0.2951658359969376, | |
| "learning_rate": 2.8170195686048837e-06, | |
| "loss": 0.2809, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.809110629067245, | |
| "grad_norm": 0.24062215600970813, | |
| "learning_rate": 2.7546559446693876e-06, | |
| "loss": 0.2618, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.8112798264642083, | |
| "grad_norm": 0.222892598517785, | |
| "learning_rate": 2.692970819684898e-06, | |
| "loss": 0.2255, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.8134490238611713, | |
| "grad_norm": 0.22574094208163784, | |
| "learning_rate": 2.6319650795226414e-06, | |
| "loss": 0.2844, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.8156182212581344, | |
| "grad_norm": 0.2143823574732022, | |
| "learning_rate": 2.571639600297143e-06, | |
| "loss": 0.2179, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.8177874186550977, | |
| "grad_norm": 0.20810428440270837, | |
| "learning_rate": 2.5119952483535214e-06, | |
| "loss": 0.2273, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.8199566160520608, | |
| "grad_norm": 0.2219262917061002, | |
| "learning_rate": 2.4530328802551516e-06, | |
| "loss": 0.2469, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.822125813449024, | |
| "grad_norm": 0.22814813313597437, | |
| "learning_rate": 2.3947533427712666e-06, | |
| "loss": 0.2717, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.824295010845987, | |
| "grad_norm": 0.20474219388181666, | |
| "learning_rate": 2.3371574728648928e-06, | |
| "loss": 0.2206, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.82646420824295, | |
| "grad_norm": 0.2393953186831936, | |
| "learning_rate": 2.2802460976807384e-06, | |
| "loss": 0.2773, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.8286334056399132, | |
| "grad_norm": 0.20242210841493188, | |
| "learning_rate": 2.2240200345333972e-06, | |
| "loss": 0.241, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.8308026030368763, | |
| "grad_norm": 0.2479672969940867, | |
| "learning_rate": 2.1684800908955362e-06, | |
| "loss": 0.304, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.8329718004338393, | |
| "grad_norm": 0.2924875615305496, | |
| "learning_rate": 2.1136270643863807e-06, | |
| "loss": 0.1846, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8351409978308026, | |
| "grad_norm": 0.23730285407509868, | |
| "learning_rate": 2.0594617427601603e-06, | |
| "loss": 0.2475, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.8373101952277657, | |
| "grad_norm": 0.22663266708243113, | |
| "learning_rate": 2.0059849038949084e-06, | |
| "loss": 0.2662, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.8394793926247288, | |
| "grad_norm": 0.24163203609153988, | |
| "learning_rate": 1.953197315781169e-06, | |
| "loss": 0.2949, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.8416485900216921, | |
| "grad_norm": 0.22227338821415707, | |
| "learning_rate": 1.9010997365110971e-06, | |
| "loss": 0.2483, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.8438177874186552, | |
| "grad_norm": 0.20465862377896551, | |
| "learning_rate": 1.8496929142674424e-06, | |
| "loss": 0.2406, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8459869848156183, | |
| "grad_norm": 0.2364370583373017, | |
| "learning_rate": 1.7989775873129278e-06, | |
| "loss": 0.2902, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.8481561822125814, | |
| "grad_norm": 0.26218568934111874, | |
| "learning_rate": 1.7489544839795314e-06, | |
| "loss": 0.2492, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.8503253796095445, | |
| "grad_norm": 0.21511824786196265, | |
| "learning_rate": 1.6996243226581388e-06, | |
| "loss": 0.2368, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.8524945770065075, | |
| "grad_norm": 0.2198436062463957, | |
| "learning_rate": 1.6509878117881073e-06, | |
| "loss": 0.2208, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.8546637744034706, | |
| "grad_norm": 0.21644370095781967, | |
| "learning_rate": 1.6030456498472124e-06, | |
| "loss": 0.212, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.8568329718004337, | |
| "grad_norm": 0.22133422953240703, | |
| "learning_rate": 1.5557985253415119e-06, | |
| "loss": 0.2553, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.8590021691973968, | |
| "grad_norm": 0.22714457687645023, | |
| "learning_rate": 1.509247116795548e-06, | |
| "loss": 0.2462, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.86117136659436, | |
| "grad_norm": 0.222946120475819, | |
| "learning_rate": 1.463392092742516e-06, | |
| "loss": 0.2306, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.8633405639913232, | |
| "grad_norm": 0.23418079900923416, | |
| "learning_rate": 1.4182341117147501e-06, | |
| "loss": 0.2696, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.8655097613882863, | |
| "grad_norm": 0.24002442145224728, | |
| "learning_rate": 1.373773822234181e-06, | |
| "loss": 0.3045, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8676789587852496, | |
| "grad_norm": 0.23645720857751817, | |
| "learning_rate": 1.3300118628031044e-06, | |
| "loss": 0.2677, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.8698481561822127, | |
| "grad_norm": 0.20964939053220596, | |
| "learning_rate": 1.2869488618949488e-06, | |
| "loss": 0.2387, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.8720173535791758, | |
| "grad_norm": 0.23651470700477456, | |
| "learning_rate": 1.2445854379452726e-06, | |
| "loss": 0.2706, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.8741865509761388, | |
| "grad_norm": 0.22140918135138807, | |
| "learning_rate": 1.2029221993428873e-06, | |
| "loss": 0.2607, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.876355748373102, | |
| "grad_norm": 0.2348213137855099, | |
| "learning_rate": 1.1619597444211206e-06, | |
| "loss": 0.2672, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.878524945770065, | |
| "grad_norm": 0.21784895708257698, | |
| "learning_rate": 1.121698661449211e-06, | |
| "loss": 0.2285, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.880694143167028, | |
| "grad_norm": 0.22922875389410033, | |
| "learning_rate": 1.0821395286238656e-06, | |
| "loss": 0.2343, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.8828633405639912, | |
| "grad_norm": 0.23166059465595806, | |
| "learning_rate": 1.0432829140609723e-06, | |
| "loss": 0.2714, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.8850325379609545, | |
| "grad_norm": 0.2360632284131841, | |
| "learning_rate": 1.0051293757874002e-06, | |
| "loss": 0.2716, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.8872017353579176, | |
| "grad_norm": 0.2314940438713054, | |
| "learning_rate": 9.67679461733051e-07, | |
| "loss": 0.2835, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8893709327548807, | |
| "grad_norm": 0.21435494553958442, | |
| "learning_rate": 9.309337097229087e-07, | |
| "loss": 0.2131, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.891540130151844, | |
| "grad_norm": 0.2210305803057693, | |
| "learning_rate": 8.94892647469403e-07, | |
| "loss": 0.2575, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.893709327548807, | |
| "grad_norm": 0.24869986480220152, | |
| "learning_rate": 8.595567925647363e-07, | |
| "loss": 0.2738, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.8958785249457701, | |
| "grad_norm": 0.2305734211734528, | |
| "learning_rate": 8.249266524735455e-07, | |
| "loss": 0.2733, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.8980477223427332, | |
| "grad_norm": 0.2536697393369725, | |
| "learning_rate": 7.910027245255413e-07, | |
| "loss": 0.3261, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.9002169197396963, | |
| "grad_norm": 0.23661550044126228, | |
| "learning_rate": 7.577854959084085e-07, | |
| "loss": 0.3014, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.9023861171366594, | |
| "grad_norm": 0.22995644681345337, | |
| "learning_rate": 7.252754436607834e-07, | |
| "loss": 0.2957, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.9045553145336225, | |
| "grad_norm": 0.22988735331559226, | |
| "learning_rate": 6.934730346654316e-07, | |
| "loss": 0.2418, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.9067245119305856, | |
| "grad_norm": 0.22730994636228313, | |
| "learning_rate": 6.623787256424984e-07, | |
| "loss": 0.2746, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.9088937093275486, | |
| "grad_norm": 0.21461778114516813, | |
| "learning_rate": 6.319929631430077e-07, | |
| "loss": 0.2225, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.911062906724512, | |
| "grad_norm": 0.23234603239697021, | |
| "learning_rate": 6.023161835423896e-07, | |
| "loss": 0.2969, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.913232104121475, | |
| "grad_norm": 0.21550966631710464, | |
| "learning_rate": 5.733488130342635e-07, | |
| "loss": 0.2503, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.9154013015184381, | |
| "grad_norm": 0.2231153295751743, | |
| "learning_rate": 5.45091267624287e-07, | |
| "loss": 0.249, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.9175704989154014, | |
| "grad_norm": 0.2293111442665084, | |
| "learning_rate": 5.175439531241999e-07, | |
| "loss": 0.29, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.9197396963123645, | |
| "grad_norm": 0.2378040770485283, | |
| "learning_rate": 4.907072651459621e-07, | |
| "loss": 0.2891, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.9219088937093276, | |
| "grad_norm": 0.21536664945034864, | |
| "learning_rate": 4.645815890961358e-07, | |
| "loss": 0.22, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.9240780911062907, | |
| "grad_norm": 0.21799292627016253, | |
| "learning_rate": 4.391673001702734e-07, | |
| "loss": 0.2168, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.9262472885032538, | |
| "grad_norm": 0.2272046818418607, | |
| "learning_rate": 4.14464763347594e-07, | |
| "loss": 0.2555, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.9284164859002169, | |
| "grad_norm": 0.26748625735516185, | |
| "learning_rate": 3.904743333857097e-07, | |
| "loss": 0.2186, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.93058568329718, | |
| "grad_norm": 0.23511293905776262, | |
| "learning_rate": 3.671963548155244e-07, | |
| "loss": 0.2622, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.932754880694143, | |
| "grad_norm": 0.2132059321481103, | |
| "learning_rate": 3.446311619363207e-07, | |
| "loss": 0.2162, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.9349240780911063, | |
| "grad_norm": 0.21597717686523027, | |
| "learning_rate": 3.227790788109197e-07, | |
| "loss": 0.2307, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.9370932754880694, | |
| "grad_norm": 0.2978176072682262, | |
| "learning_rate": 3.0164041926104e-07, | |
| "loss": 0.2577, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.9392624728850325, | |
| "grad_norm": 0.23250354980532184, | |
| "learning_rate": 2.8121548686281805e-07, | |
| "loss": 0.2714, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.9414316702819958, | |
| "grad_norm": 0.23296800107183185, | |
| "learning_rate": 2.6150457494240057e-07, | |
| "loss": 0.2414, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.943600867678959, | |
| "grad_norm": 0.2136515878728428, | |
| "learning_rate": 2.4250796657177e-07, | |
| "loss": 0.2173, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.945770065075922, | |
| "grad_norm": 0.20347321610165708, | |
| "learning_rate": 2.2422593456466468e-07, | |
| "loss": 0.2152, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.947939262472885, | |
| "grad_norm": 0.23997716902494964, | |
| "learning_rate": 2.0665874147265395e-07, | |
| "loss": 0.2305, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.9501084598698482, | |
| "grad_norm": 0.22269717072616485, | |
| "learning_rate": 1.8980663958139133e-07, | |
| "loss": 0.2665, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.9522776572668112, | |
| "grad_norm": 0.21024780049164776, | |
| "learning_rate": 1.736698709069673e-07, | |
| "loss": 0.1967, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.9544468546637743, | |
| "grad_norm": 0.2267046711412435, | |
| "learning_rate": 1.5824866719243436e-07, | |
| "loss": 0.2445, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.9566160520607374, | |
| "grad_norm": 0.22809702055070344, | |
| "learning_rate": 1.4354324990449863e-07, | |
| "loss": 0.2569, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.9587852494577005, | |
| "grad_norm": 0.23382391297633118, | |
| "learning_rate": 1.295538302303223e-07, | |
| "loss": 0.2876, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.9609544468546638, | |
| "grad_norm": 0.20472909033825798, | |
| "learning_rate": 1.1628060907449834e-07, | |
| "loss": 0.1924, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.9631236442516269, | |
| "grad_norm": 0.2096943890587732, | |
| "learning_rate": 1.0372377705616942e-07, | |
| "loss": 0.233, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.96529284164859, | |
| "grad_norm": 0.211371475954759, | |
| "learning_rate": 9.188351450626353e-08, | |
| "loss": 0.2166, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.9674620390455533, | |
| "grad_norm": 0.21471581374101012, | |
| "learning_rate": 8.075999146494595e-08, | |
| "loss": 0.2282, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.9696312364425164, | |
| "grad_norm": 0.2660219035351609, | |
| "learning_rate": 7.035336767914346e-08, | |
| "loss": 0.2735, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.9718004338394794, | |
| "grad_norm": 0.22465441666056135, | |
| "learning_rate": 6.066379260026845e-08, | |
| "loss": 0.2765, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.9739696312364425, | |
| "grad_norm": 0.22816673191661355, | |
| "learning_rate": 5.169140538207051e-08, | |
| "loss": 0.2685, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.9761388286334056, | |
| "grad_norm": 0.24062260875061986, | |
| "learning_rate": 4.343633487861598e-08, | |
| "loss": 0.2479, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.9783080260303687, | |
| "grad_norm": 0.23204079070364966, | |
| "learning_rate": 3.589869964248371e-08, | |
| "loss": 0.2544, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.9804772234273318, | |
| "grad_norm": 0.2339821898793586, | |
| "learning_rate": 2.907860792301098e-08, | |
| "loss": 0.3132, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.9826464208242949, | |
| "grad_norm": 0.21123283163235765, | |
| "learning_rate": 2.297615766477801e-08, | |
| "loss": 0.2454, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.9848156182212582, | |
| "grad_norm": 0.23481457672784686, | |
| "learning_rate": 1.7591436506170235e-08, | |
| "loss": 0.2803, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9869848156182213, | |
| "grad_norm": 0.22666968506581125, | |
| "learning_rate": 1.2924521778151511e-08, | |
| "loss": 0.262, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.9891540130151844, | |
| "grad_norm": 0.22371033015850025, | |
| "learning_rate": 8.975480503126133e-09, | |
| "loss": 0.2253, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.9913232104121477, | |
| "grad_norm": 0.24195113149305236, | |
| "learning_rate": 5.74436939398959e-09, | |
| "loss": 0.2679, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.9934924078091107, | |
| "grad_norm": 0.22275331811796895, | |
| "learning_rate": 3.2312348533236614e-09, | |
| "loss": 0.3069, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.9956616052060738, | |
| "grad_norm": 0.2209663606112055, | |
| "learning_rate": 1.4361129727025191e-09, | |
| "loss": 0.2256, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.997830802603037, | |
| "grad_norm": 0.2356916076109314, | |
| "learning_rate": 3.590295321986847e-10, | |
| "loss": 0.2807, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.22113655920342676, | |
| "learning_rate": 0.0, | |
| "loss": 0.2328, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 922, | |
| "total_flos": 4515096717950976.0, | |
| "train_loss": 0.30544997376759503, | |
| "train_runtime": 16133.364, | |
| "train_samples_per_second": 0.228, | |
| "train_steps_per_second": 0.057 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 922, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4515096717950976.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |