| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 32.142857142857146, |
| "eval_steps": 50, |
| "global_step": 450, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.07142857142857142, |
| "grad_norm": 5.383044719696045, |
| "learning_rate": 0.0, |
| "loss": 1.5793, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.14285714285714285, |
| "grad_norm": 3.96494197845459, |
| "learning_rate": 1.2500000000000002e-07, |
| "loss": 1.62, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.21428571428571427, |
| "grad_norm": 4.022143363952637, |
| "learning_rate": 2.5000000000000004e-07, |
| "loss": 1.6035, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.2857142857142857, |
| "grad_norm": 6.808481216430664, |
| "learning_rate": 3.75e-07, |
| "loss": 1.6309, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.35714285714285715, |
| "grad_norm": 4.0089311599731445, |
| "learning_rate": 5.000000000000001e-07, |
| "loss": 1.596, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.42857142857142855, |
| "grad_norm": 6.727263450622559, |
| "learning_rate": 6.25e-07, |
| "loss": 1.6312, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 7.648512840270996, |
| "learning_rate": 7.5e-07, |
| "loss": 1.6206, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.5714285714285714, |
| "grad_norm": 11.115941047668457, |
| "learning_rate": 8.750000000000001e-07, |
| "loss": 1.6133, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.6428571428571429, |
| "grad_norm": 8.82610034942627, |
| "learning_rate": 1.0000000000000002e-06, |
| "loss": 1.6419, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.7142857142857143, |
| "grad_norm": 3.3373453617095947, |
| "learning_rate": 1.125e-06, |
| "loss": 1.5894, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.7857142857142857, |
| "grad_norm": 7.521600723266602, |
| "learning_rate": 1.25e-06, |
| "loss": 1.6188, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 5.497554302215576, |
| "learning_rate": 1.3750000000000002e-06, |
| "loss": 1.6221, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.9285714285714286, |
| "grad_norm": 3.5530192852020264, |
| "learning_rate": 1.5e-06, |
| "loss": 1.6112, |
| "step": 13 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 4.452667713165283, |
| "learning_rate": 1.6250000000000001e-06, |
| "loss": 1.5999, |
| "step": 14 |
| }, |
| { |
| "epoch": 1.0714285714285714, |
| "grad_norm": 2.4583044052124023, |
| "learning_rate": 1.7500000000000002e-06, |
| "loss": 1.5917, |
| "step": 15 |
| }, |
| { |
| "epoch": 1.1428571428571428, |
| "grad_norm": 9.272625923156738, |
| "learning_rate": 1.875e-06, |
| "loss": 1.6092, |
| "step": 16 |
| }, |
| { |
| "epoch": 1.2142857142857142, |
| "grad_norm": 4.451343059539795, |
| "learning_rate": 2.0000000000000003e-06, |
| "loss": 1.5686, |
| "step": 17 |
| }, |
| { |
| "epoch": 1.2857142857142856, |
| "grad_norm": 5.756788730621338, |
| "learning_rate": 2.1250000000000004e-06, |
| "loss": 1.6265, |
| "step": 18 |
| }, |
| { |
| "epoch": 1.3571428571428572, |
| "grad_norm": 4.12752103805542, |
| "learning_rate": 2.25e-06, |
| "loss": 1.6073, |
| "step": 19 |
| }, |
| { |
| "epoch": 1.4285714285714286, |
| "grad_norm": 2.2131590843200684, |
| "learning_rate": 2.375e-06, |
| "loss": 1.5929, |
| "step": 20 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 7.726499080657959, |
| "learning_rate": 2.5e-06, |
| "loss": 1.6232, |
| "step": 21 |
| }, |
| { |
| "epoch": 1.5714285714285714, |
| "grad_norm": 5.906811714172363, |
| "learning_rate": 2.625e-06, |
| "loss": 1.5777, |
| "step": 22 |
| }, |
| { |
| "epoch": 1.6428571428571428, |
| "grad_norm": 4.695427894592285, |
| "learning_rate": 2.7500000000000004e-06, |
| "loss": 1.591, |
| "step": 23 |
| }, |
| { |
| "epoch": 1.7142857142857144, |
| "grad_norm": 1.5481104850769043, |
| "learning_rate": 2.8750000000000004e-06, |
| "loss": 1.5743, |
| "step": 24 |
| }, |
| { |
| "epoch": 1.7857142857142856, |
| "grad_norm": 4.624354839324951, |
| "learning_rate": 3e-06, |
| "loss": 1.6065, |
| "step": 25 |
| }, |
| { |
| "epoch": 1.8571428571428572, |
| "grad_norm": 5.646456718444824, |
| "learning_rate": 3.125e-06, |
| "loss": 1.5975, |
| "step": 26 |
| }, |
| { |
| "epoch": 1.9285714285714286, |
| "grad_norm": 8.583253860473633, |
| "learning_rate": 3.2500000000000002e-06, |
| "loss": 1.5809, |
| "step": 27 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.7897872924804688, |
| "learning_rate": 3.3750000000000003e-06, |
| "loss": 1.5839, |
| "step": 28 |
| }, |
| { |
| "epoch": 2.0714285714285716, |
| "grad_norm": 2.509343385696411, |
| "learning_rate": 3.5000000000000004e-06, |
| "loss": 1.5725, |
| "step": 29 |
| }, |
| { |
| "epoch": 2.142857142857143, |
| "grad_norm": 3.86081600189209, |
| "learning_rate": 3.625e-06, |
| "loss": 1.5766, |
| "step": 30 |
| }, |
| { |
| "epoch": 2.2142857142857144, |
| "grad_norm": 4.29940128326416, |
| "learning_rate": 3.75e-06, |
| "loss": 1.5727, |
| "step": 31 |
| }, |
| { |
| "epoch": 2.2857142857142856, |
| "grad_norm": 3.209973096847534, |
| "learning_rate": 3.875e-06, |
| "loss": 1.5783, |
| "step": 32 |
| }, |
| { |
| "epoch": 2.357142857142857, |
| "grad_norm": 2.813647985458374, |
| "learning_rate": 4.000000000000001e-06, |
| "loss": 1.5687, |
| "step": 33 |
| }, |
| { |
| "epoch": 2.4285714285714284, |
| "grad_norm": 2.750006914138794, |
| "learning_rate": 4.125e-06, |
| "loss": 1.5739, |
| "step": 34 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 4.352213382720947, |
| "learning_rate": 4.250000000000001e-06, |
| "loss": 1.5781, |
| "step": 35 |
| }, |
| { |
| "epoch": 2.571428571428571, |
| "grad_norm": 2.743788957595825, |
| "learning_rate": 4.375e-06, |
| "loss": 1.5731, |
| "step": 36 |
| }, |
| { |
| "epoch": 2.642857142857143, |
| "grad_norm": 5.008453845977783, |
| "learning_rate": 4.5e-06, |
| "loss": 1.5828, |
| "step": 37 |
| }, |
| { |
| "epoch": 2.7142857142857144, |
| "grad_norm": 3.1128811836242676, |
| "learning_rate": 4.625e-06, |
| "loss": 1.5706, |
| "step": 38 |
| }, |
| { |
| "epoch": 2.7857142857142856, |
| "grad_norm": 5.385458946228027, |
| "learning_rate": 4.75e-06, |
| "loss": 1.5834, |
| "step": 39 |
| }, |
| { |
| "epoch": 2.857142857142857, |
| "grad_norm": 1.6079157590866089, |
| "learning_rate": 4.875000000000001e-06, |
| "loss": 1.5721, |
| "step": 40 |
| }, |
| { |
| "epoch": 2.928571428571429, |
| "grad_norm": 3.1016640663146973, |
| "learning_rate": 5e-06, |
| "loss": 1.5763, |
| "step": 41 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 4.650889873504639, |
| "learning_rate": 5.125e-06, |
| "loss": 1.5426, |
| "step": 42 |
| }, |
| { |
| "epoch": 3.0714285714285716, |
| "grad_norm": 1.5727084875106812, |
| "learning_rate": 5.25e-06, |
| "loss": 1.5432, |
| "step": 43 |
| }, |
| { |
| "epoch": 3.142857142857143, |
| "grad_norm": 2.637890577316284, |
| "learning_rate": 5.375e-06, |
| "loss": 1.5472, |
| "step": 44 |
| }, |
| { |
| "epoch": 3.2142857142857144, |
| "grad_norm": 1.8277631998062134, |
| "learning_rate": 5.500000000000001e-06, |
| "loss": 1.5363, |
| "step": 45 |
| }, |
| { |
| "epoch": 3.2857142857142856, |
| "grad_norm": 1.6959493160247803, |
| "learning_rate": 5.625e-06, |
| "loss": 1.5397, |
| "step": 46 |
| }, |
| { |
| "epoch": 3.357142857142857, |
| "grad_norm": 7.91141414642334, |
| "learning_rate": 5.750000000000001e-06, |
| "loss": 1.5392, |
| "step": 47 |
| }, |
| { |
| "epoch": 3.4285714285714284, |
| "grad_norm": 8.705608367919922, |
| "learning_rate": 5.875e-06, |
| "loss": 1.5499, |
| "step": 48 |
| }, |
| { |
| "epoch": 3.5, |
| "grad_norm": 1.4096094369888306, |
| "learning_rate": 6e-06, |
| "loss": 1.5554, |
| "step": 49 |
| }, |
| { |
| "epoch": 3.571428571428571, |
| "grad_norm": 1.2479711771011353, |
| "learning_rate": 6.125e-06, |
| "loss": 1.5537, |
| "step": 50 |
| }, |
| { |
| "epoch": 3.642857142857143, |
| "grad_norm": 1.7531940937042236, |
| "learning_rate": 6.25e-06, |
| "loss": 1.533, |
| "step": 51 |
| }, |
| { |
| "epoch": 3.7142857142857144, |
| "grad_norm": 3.1394143104553223, |
| "learning_rate": 6.375000000000001e-06, |
| "loss": 1.5223, |
| "step": 52 |
| }, |
| { |
| "epoch": 3.7857142857142856, |
| "grad_norm": 5.379184246063232, |
| "learning_rate": 6.5000000000000004e-06, |
| "loss": 1.5204, |
| "step": 53 |
| }, |
| { |
| "epoch": 3.857142857142857, |
| "grad_norm": 1.6642723083496094, |
| "learning_rate": 6.625000000000001e-06, |
| "loss": 1.5195, |
| "step": 54 |
| }, |
| { |
| "epoch": 3.928571428571429, |
| "grad_norm": 1.8299416303634644, |
| "learning_rate": 6.750000000000001e-06, |
| "loss": 1.5111, |
| "step": 55 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 3.942272901535034, |
| "learning_rate": 6.875000000000001e-06, |
| "loss": 1.5134, |
| "step": 56 |
| }, |
| { |
| "epoch": 4.071428571428571, |
| "grad_norm": 1.4287347793579102, |
| "learning_rate": 7.000000000000001e-06, |
| "loss": 1.5119, |
| "step": 57 |
| }, |
| { |
| "epoch": 4.142857142857143, |
| "grad_norm": 2.0702247619628906, |
| "learning_rate": 7.1249999999999995e-06, |
| "loss": 1.4962, |
| "step": 58 |
| }, |
| { |
| "epoch": 4.214285714285714, |
| "grad_norm": 2.863966703414917, |
| "learning_rate": 7.25e-06, |
| "loss": 1.5204, |
| "step": 59 |
| }, |
| { |
| "epoch": 4.285714285714286, |
| "grad_norm": 2.398355722427368, |
| "learning_rate": 7.375e-06, |
| "loss": 1.4955, |
| "step": 60 |
| }, |
| { |
| "epoch": 4.357142857142857, |
| "grad_norm": 1.4015003442764282, |
| "learning_rate": 7.5e-06, |
| "loss": 1.4942, |
| "step": 61 |
| }, |
| { |
| "epoch": 4.428571428571429, |
| "grad_norm": 1.6098459959030151, |
| "learning_rate": 7.625e-06, |
| "loss": 1.4712, |
| "step": 62 |
| }, |
| { |
| "epoch": 4.5, |
| "grad_norm": 4.722957611083984, |
| "learning_rate": 7.75e-06, |
| "loss": 1.5288, |
| "step": 63 |
| }, |
| { |
| "epoch": 4.571428571428571, |
| "grad_norm": 3.6340856552124023, |
| "learning_rate": 7.875e-06, |
| "loss": 1.524, |
| "step": 64 |
| }, |
| { |
| "epoch": 4.642857142857143, |
| "grad_norm": 2.6402580738067627, |
| "learning_rate": 8.000000000000001e-06, |
| "loss": 1.5417, |
| "step": 65 |
| }, |
| { |
| "epoch": 4.714285714285714, |
| "grad_norm": 2.7053163051605225, |
| "learning_rate": 8.125000000000001e-06, |
| "loss": 1.5101, |
| "step": 66 |
| }, |
| { |
| "epoch": 4.785714285714286, |
| "grad_norm": 4.306228160858154, |
| "learning_rate": 8.25e-06, |
| "loss": 1.5474, |
| "step": 67 |
| }, |
| { |
| "epoch": 4.857142857142857, |
| "grad_norm": 3.2443113327026367, |
| "learning_rate": 8.375e-06, |
| "loss": 1.5324, |
| "step": 68 |
| }, |
| { |
| "epoch": 4.928571428571429, |
| "grad_norm": 2.3825535774230957, |
| "learning_rate": 8.500000000000002e-06, |
| "loss": 1.522, |
| "step": 69 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 3.2580692768096924, |
| "learning_rate": 8.625e-06, |
| "loss": 1.5179, |
| "step": 70 |
| }, |
| { |
| "epoch": 5.071428571428571, |
| "grad_norm": 1.4018006324768066, |
| "learning_rate": 8.75e-06, |
| "loss": 1.4971, |
| "step": 71 |
| }, |
| { |
| "epoch": 5.142857142857143, |
| "grad_norm": 5.274407386779785, |
| "learning_rate": 8.875e-06, |
| "loss": 1.513, |
| "step": 72 |
| }, |
| { |
| "epoch": 5.214285714285714, |
| "grad_norm": 1.3473955392837524, |
| "learning_rate": 9e-06, |
| "loss": 1.4898, |
| "step": 73 |
| }, |
| { |
| "epoch": 5.285714285714286, |
| "grad_norm": 5.1161346435546875, |
| "learning_rate": 9.125e-06, |
| "loss": 1.5032, |
| "step": 74 |
| }, |
| { |
| "epoch": 5.357142857142857, |
| "grad_norm": 10.268465042114258, |
| "learning_rate": 9.25e-06, |
| "loss": 1.4926, |
| "step": 75 |
| }, |
| { |
| "epoch": 5.428571428571429, |
| "grad_norm": 1.676137089729309, |
| "learning_rate": 9.375000000000001e-06, |
| "loss": 1.4833, |
| "step": 76 |
| }, |
| { |
| "epoch": 5.5, |
| "grad_norm": 1.441231369972229, |
| "learning_rate": 9.5e-06, |
| "loss": 1.4934, |
| "step": 77 |
| }, |
| { |
| "epoch": 5.571428571428571, |
| "grad_norm": 3.917909860610962, |
| "learning_rate": 9.625e-06, |
| "loss": 1.5473, |
| "step": 78 |
| }, |
| { |
| "epoch": 5.642857142857143, |
| "grad_norm": 1.9755581617355347, |
| "learning_rate": 9.750000000000002e-06, |
| "loss": 1.5192, |
| "step": 79 |
| }, |
| { |
| "epoch": 5.714285714285714, |
| "grad_norm": 2.587038516998291, |
| "learning_rate": 9.875000000000001e-06, |
| "loss": 1.5212, |
| "step": 80 |
| }, |
| { |
| "epoch": 5.785714285714286, |
| "grad_norm": 2.155751943588257, |
| "learning_rate": 1e-05, |
| "loss": 1.5337, |
| "step": 81 |
| }, |
| { |
| "epoch": 5.857142857142857, |
| "grad_norm": 2.8446056842803955, |
| "learning_rate": 1.0125e-05, |
| "loss": 1.5174, |
| "step": 82 |
| }, |
| { |
| "epoch": 5.928571428571429, |
| "grad_norm": 2.9765541553497314, |
| "learning_rate": 1.025e-05, |
| "loss": 1.5267, |
| "step": 83 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 2.4112436771392822, |
| "learning_rate": 1.0375e-05, |
| "loss": 1.4897, |
| "step": 84 |
| }, |
| { |
| "epoch": 6.071428571428571, |
| "grad_norm": 3.6984405517578125, |
| "learning_rate": 1.05e-05, |
| "loss": 1.5304, |
| "step": 85 |
| }, |
| { |
| "epoch": 6.142857142857143, |
| "grad_norm": 3.8474667072296143, |
| "learning_rate": 1.0625e-05, |
| "loss": 1.5232, |
| "step": 86 |
| }, |
| { |
| "epoch": 6.214285714285714, |
| "grad_norm": 2.087263584136963, |
| "learning_rate": 1.075e-05, |
| "loss": 1.5285, |
| "step": 87 |
| }, |
| { |
| "epoch": 6.285714285714286, |
| "grad_norm": 2.1326711177825928, |
| "learning_rate": 1.0875e-05, |
| "loss": 1.5387, |
| "step": 88 |
| }, |
| { |
| "epoch": 6.357142857142857, |
| "grad_norm": 4.559895992279053, |
| "learning_rate": 1.1000000000000001e-05, |
| "loss": 1.524, |
| "step": 89 |
| }, |
| { |
| "epoch": 6.428571428571429, |
| "grad_norm": 1.8320558071136475, |
| "learning_rate": 1.1125000000000001e-05, |
| "loss": 1.5143, |
| "step": 90 |
| }, |
| { |
| "epoch": 6.5, |
| "grad_norm": 1.8108539581298828, |
| "learning_rate": 1.125e-05, |
| "loss": 1.4987, |
| "step": 91 |
| }, |
| { |
| "epoch": 6.571428571428571, |
| "grad_norm": 1.3095402717590332, |
| "learning_rate": 1.1375e-05, |
| "loss": 1.5025, |
| "step": 92 |
| }, |
| { |
| "epoch": 6.642857142857143, |
| "grad_norm": 4.299781799316406, |
| "learning_rate": 1.1500000000000002e-05, |
| "loss": 1.5251, |
| "step": 93 |
| }, |
| { |
| "epoch": 6.714285714285714, |
| "grad_norm": 1.1237058639526367, |
| "learning_rate": 1.1625000000000001e-05, |
| "loss": 1.5185, |
| "step": 94 |
| }, |
| { |
| "epoch": 6.785714285714286, |
| "grad_norm": 6.838643550872803, |
| "learning_rate": 1.175e-05, |
| "loss": 1.5032, |
| "step": 95 |
| }, |
| { |
| "epoch": 6.857142857142857, |
| "grad_norm": 1.886619210243225, |
| "learning_rate": 1.1875e-05, |
| "loss": 1.5394, |
| "step": 96 |
| }, |
| { |
| "epoch": 6.928571428571429, |
| "grad_norm": 2.4293034076690674, |
| "learning_rate": 1.2e-05, |
| "loss": 1.4958, |
| "step": 97 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 1.3641681671142578, |
| "learning_rate": 1.2125e-05, |
| "loss": 1.5182, |
| "step": 98 |
| }, |
| { |
| "epoch": 7.071428571428571, |
| "grad_norm": 1.6436591148376465, |
| "learning_rate": 1.225e-05, |
| "loss": 1.5007, |
| "step": 99 |
| }, |
| { |
| "epoch": 7.142857142857143, |
| "grad_norm": 1.4990618228912354, |
| "learning_rate": 1.2375000000000001e-05, |
| "loss": 1.4881, |
| "step": 100 |
| }, |
| { |
| "epoch": 7.214285714285714, |
| "grad_norm": 1.2271665334701538, |
| "learning_rate": 1.25e-05, |
| "loss": 1.5017, |
| "step": 101 |
| }, |
| { |
| "epoch": 7.285714285714286, |
| "grad_norm": 1.2291427850723267, |
| "learning_rate": 1.2625e-05, |
| "loss": 1.4979, |
| "step": 102 |
| }, |
| { |
| "epoch": 7.357142857142857, |
| "grad_norm": 5.997069358825684, |
| "learning_rate": 1.2750000000000002e-05, |
| "loss": 1.5238, |
| "step": 103 |
| }, |
| { |
| "epoch": 7.428571428571429, |
| "grad_norm": 1.4330472946166992, |
| "learning_rate": 1.2875000000000001e-05, |
| "loss": 1.5133, |
| "step": 104 |
| }, |
| { |
| "epoch": 7.5, |
| "grad_norm": 1.3329960107803345, |
| "learning_rate": 1.3000000000000001e-05, |
| "loss": 1.5044, |
| "step": 105 |
| }, |
| { |
| "epoch": 7.571428571428571, |
| "grad_norm": 2.7081637382507324, |
| "learning_rate": 1.3125e-05, |
| "loss": 1.4825, |
| "step": 106 |
| }, |
| { |
| "epoch": 7.642857142857143, |
| "grad_norm": 2.8797085285186768, |
| "learning_rate": 1.3250000000000002e-05, |
| "loss": 1.4711, |
| "step": 107 |
| }, |
| { |
| "epoch": 7.714285714285714, |
| "grad_norm": 4.054922580718994, |
| "learning_rate": 1.3375000000000002e-05, |
| "loss": 1.4982, |
| "step": 108 |
| }, |
| { |
| "epoch": 7.785714285714286, |
| "grad_norm": 7.4362311363220215, |
| "learning_rate": 1.3500000000000001e-05, |
| "loss": 1.4786, |
| "step": 109 |
| }, |
| { |
| "epoch": 7.857142857142857, |
| "grad_norm": 2.5698859691619873, |
| "learning_rate": 1.3625e-05, |
| "loss": 1.4572, |
| "step": 110 |
| }, |
| { |
| "epoch": 7.928571428571429, |
| "grad_norm": 2.9194092750549316, |
| "learning_rate": 1.3750000000000002e-05, |
| "loss": 1.4916, |
| "step": 111 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 2.044605016708374, |
| "learning_rate": 1.3875000000000002e-05, |
| "loss": 1.4879, |
| "step": 112 |
| }, |
| { |
| "epoch": 8.071428571428571, |
| "grad_norm": 1.3414772748947144, |
| "learning_rate": 1.4000000000000001e-05, |
| "loss": 1.4887, |
| "step": 113 |
| }, |
| { |
| "epoch": 8.142857142857142, |
| "grad_norm": 2.0897555351257324, |
| "learning_rate": 1.4125e-05, |
| "loss": 1.4908, |
| "step": 114 |
| }, |
| { |
| "epoch": 8.214285714285714, |
| "grad_norm": 3.721792697906494, |
| "learning_rate": 1.4249999999999999e-05, |
| "loss": 1.4978, |
| "step": 115 |
| }, |
| { |
| "epoch": 8.285714285714286, |
| "grad_norm": 1.3067519664764404, |
| "learning_rate": 1.4374999999999999e-05, |
| "loss": 1.4903, |
| "step": 116 |
| }, |
| { |
| "epoch": 8.357142857142858, |
| "grad_norm": 2.3188929557800293, |
| "learning_rate": 1.45e-05, |
| "loss": 1.475, |
| "step": 117 |
| }, |
| { |
| "epoch": 8.428571428571429, |
| "grad_norm": 2.131817102432251, |
| "learning_rate": 1.4625e-05, |
| "loss": 1.4877, |
| "step": 118 |
| }, |
| { |
| "epoch": 8.5, |
| "grad_norm": 5.011297225952148, |
| "learning_rate": 1.475e-05, |
| "loss": 1.4651, |
| "step": 119 |
| }, |
| { |
| "epoch": 8.571428571428571, |
| "grad_norm": 2.497302770614624, |
| "learning_rate": 1.4875e-05, |
| "loss": 1.4488, |
| "step": 120 |
| }, |
| { |
| "epoch": 8.642857142857142, |
| "grad_norm": 1.3582658767700195, |
| "learning_rate": 1.5e-05, |
| "loss": 1.4446, |
| "step": 121 |
| }, |
| { |
| "epoch": 8.714285714285714, |
| "grad_norm": 5.030256748199463, |
| "learning_rate": 1.5125e-05, |
| "loss": 1.4445, |
| "step": 122 |
| }, |
| { |
| "epoch": 8.785714285714286, |
| "grad_norm": 6.480685234069824, |
| "learning_rate": 1.525e-05, |
| "loss": 1.4853, |
| "step": 123 |
| }, |
| { |
| "epoch": 8.857142857142858, |
| "grad_norm": 9.358418464660645, |
| "learning_rate": 1.5375e-05, |
| "loss": 1.5026, |
| "step": 124 |
| }, |
| { |
| "epoch": 8.928571428571429, |
| "grad_norm": 8.761038780212402, |
| "learning_rate": 1.55e-05, |
| "loss": 1.4884, |
| "step": 125 |
| }, |
| { |
| "epoch": 9.0, |
| "grad_norm": 1.6179229021072388, |
| "learning_rate": 1.5625e-05, |
| "loss": 1.4705, |
| "step": 126 |
| }, |
| { |
| "epoch": 9.071428571428571, |
| "grad_norm": 2.0552783012390137, |
| "learning_rate": 1.575e-05, |
| "loss": 1.4869, |
| "step": 127 |
| }, |
| { |
| "epoch": 9.142857142857142, |
| "grad_norm": 6.032142162322998, |
| "learning_rate": 1.5875e-05, |
| "loss": 1.4461, |
| "step": 128 |
| }, |
| { |
| "epoch": 9.214285714285714, |
| "grad_norm": 2.0145914554595947, |
| "learning_rate": 1.6000000000000003e-05, |
| "loss": 1.4625, |
| "step": 129 |
| }, |
| { |
| "epoch": 9.285714285714286, |
| "grad_norm": 5.505954742431641, |
| "learning_rate": 1.6125000000000002e-05, |
| "loss": 1.4764, |
| "step": 130 |
| }, |
| { |
| "epoch": 9.357142857142858, |
| "grad_norm": 4.161317348480225, |
| "learning_rate": 1.6250000000000002e-05, |
| "loss": 1.4558, |
| "step": 131 |
| }, |
| { |
| "epoch": 9.428571428571429, |
| "grad_norm": 6.453343391418457, |
| "learning_rate": 1.6375e-05, |
| "loss": 1.4914, |
| "step": 132 |
| }, |
| { |
| "epoch": 9.5, |
| "grad_norm": 2.7060706615448, |
| "learning_rate": 1.65e-05, |
| "loss": 1.4479, |
| "step": 133 |
| }, |
| { |
| "epoch": 9.571428571428571, |
| "grad_norm": 6.37178897857666, |
| "learning_rate": 1.6625e-05, |
| "loss": 1.4951, |
| "step": 134 |
| }, |
| { |
| "epoch": 9.642857142857142, |
| "grad_norm": 1.644932746887207, |
| "learning_rate": 1.675e-05, |
| "loss": 1.4492, |
| "step": 135 |
| }, |
| { |
| "epoch": 9.714285714285714, |
| "grad_norm": 4.173709392547607, |
| "learning_rate": 1.6875000000000004e-05, |
| "loss": 1.4419, |
| "step": 136 |
| }, |
| { |
| "epoch": 9.785714285714286, |
| "grad_norm": 2.661614179611206, |
| "learning_rate": 1.7000000000000003e-05, |
| "loss": 1.4419, |
| "step": 137 |
| }, |
| { |
| "epoch": 9.857142857142858, |
| "grad_norm": 2.468867778778076, |
| "learning_rate": 1.7125000000000003e-05, |
| "loss": 1.416, |
| "step": 138 |
| }, |
| { |
| "epoch": 9.928571428571429, |
| "grad_norm": 1.0798838138580322, |
| "learning_rate": 1.725e-05, |
| "loss": 1.4473, |
| "step": 139 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 1.5816832780838013, |
| "learning_rate": 1.7375e-05, |
| "loss": 1.4399, |
| "step": 140 |
| }, |
| { |
| "epoch": 10.071428571428571, |
| "grad_norm": 2.063046932220459, |
| "learning_rate": 1.75e-05, |
| "loss": 1.4415, |
| "step": 141 |
| }, |
| { |
| "epoch": 10.142857142857142, |
| "grad_norm": 3.9278879165649414, |
| "learning_rate": 1.7625e-05, |
| "loss": 1.4463, |
| "step": 142 |
| }, |
| { |
| "epoch": 10.214285714285714, |
| "grad_norm": 1.3775101900100708, |
| "learning_rate": 1.775e-05, |
| "loss": 1.4019, |
| "step": 143 |
| }, |
| { |
| "epoch": 10.285714285714286, |
| "grad_norm": 6.007996082305908, |
| "learning_rate": 1.7875e-05, |
| "loss": 1.4282, |
| "step": 144 |
| }, |
| { |
| "epoch": 10.357142857142858, |
| "grad_norm": 4.488779067993164, |
| "learning_rate": 1.8e-05, |
| "loss": 1.451, |
| "step": 145 |
| }, |
| { |
| "epoch": 10.428571428571429, |
| "grad_norm": 5.015976905822754, |
| "learning_rate": 1.8125e-05, |
| "loss": 1.4089, |
| "step": 146 |
| }, |
| { |
| "epoch": 10.5, |
| "grad_norm": 1.4387730360031128, |
| "learning_rate": 1.825e-05, |
| "loss": 1.4249, |
| "step": 147 |
| }, |
| { |
| "epoch": 10.571428571428571, |
| "grad_norm": 2.4665675163269043, |
| "learning_rate": 1.8375e-05, |
| "loss": 1.4327, |
| "step": 148 |
| }, |
| { |
| "epoch": 10.642857142857142, |
| "grad_norm": 1.8020267486572266, |
| "learning_rate": 1.85e-05, |
| "loss": 1.3995, |
| "step": 149 |
| }, |
| { |
| "epoch": 10.714285714285714, |
| "grad_norm": 1.6808390617370605, |
| "learning_rate": 1.8625000000000002e-05, |
| "loss": 1.389, |
| "step": 150 |
| }, |
| { |
| "epoch": 10.785714285714286, |
| "grad_norm": 2.3376588821411133, |
| "learning_rate": 1.8750000000000002e-05, |
| "loss": 1.4748, |
| "step": 151 |
| }, |
| { |
| "epoch": 10.857142857142858, |
| "grad_norm": 2.5257272720336914, |
| "learning_rate": 1.8875e-05, |
| "loss": 1.4221, |
| "step": 152 |
| }, |
| { |
| "epoch": 10.928571428571429, |
| "grad_norm": 1.6355801820755005, |
| "learning_rate": 1.9e-05, |
| "loss": 1.3544, |
| "step": 153 |
| }, |
| { |
| "epoch": 11.0, |
| "grad_norm": 2.1227149963378906, |
| "learning_rate": 1.9125e-05, |
| "loss": 1.3493, |
| "step": 154 |
| }, |
| { |
| "epoch": 11.071428571428571, |
| "grad_norm": 4.7650370597839355, |
| "learning_rate": 1.925e-05, |
| "loss": 1.4708, |
| "step": 155 |
| }, |
| { |
| "epoch": 11.142857142857142, |
| "grad_norm": 2.83673095703125, |
| "learning_rate": 1.9375e-05, |
| "loss": 1.4064, |
| "step": 156 |
| }, |
| { |
| "epoch": 11.214285714285714, |
| "grad_norm": 2.4201812744140625, |
| "learning_rate": 1.9500000000000003e-05, |
| "loss": 1.3221, |
| "step": 157 |
| }, |
| { |
| "epoch": 11.285714285714286, |
| "grad_norm": 2.0276103019714355, |
| "learning_rate": 1.9625000000000003e-05, |
| "loss": 1.2859, |
| "step": 158 |
| }, |
| { |
| "epoch": 11.357142857142858, |
| "grad_norm": 7.189219951629639, |
| "learning_rate": 1.9750000000000002e-05, |
| "loss": 1.3526, |
| "step": 159 |
| }, |
| { |
| "epoch": 11.428571428571429, |
| "grad_norm": 2.9911937713623047, |
| "learning_rate": 1.9875000000000002e-05, |
| "loss": 1.319, |
| "step": 160 |
| }, |
| { |
| "epoch": 11.5, |
| "grad_norm": 7.41688346862793, |
| "learning_rate": 2e-05, |
| "loss": 1.2701, |
| "step": 161 |
| }, |
| { |
| "epoch": 11.571428571428571, |
| "grad_norm": 2.5221216678619385, |
| "learning_rate": 2.0125e-05, |
| "loss": 1.2342, |
| "step": 162 |
| }, |
| { |
| "epoch": 11.642857142857142, |
| "grad_norm": 3.1596360206604004, |
| "learning_rate": 2.025e-05, |
| "loss": 1.2818, |
| "step": 163 |
| }, |
| { |
| "epoch": 11.714285714285714, |
| "grad_norm": 4.858260631561279, |
| "learning_rate": 2.0375e-05, |
| "loss": 1.263, |
| "step": 164 |
| }, |
| { |
| "epoch": 11.785714285714286, |
| "grad_norm": 5.346636772155762, |
| "learning_rate": 2.05e-05, |
| "loss": 1.2936, |
| "step": 165 |
| }, |
| { |
| "epoch": 11.857142857142858, |
| "grad_norm": 4.936079502105713, |
| "learning_rate": 2.0625e-05, |
| "loss": 1.2727, |
| "step": 166 |
| }, |
| { |
| "epoch": 11.928571428571429, |
| "grad_norm": 10.273465156555176, |
| "learning_rate": 2.075e-05, |
| "loss": 1.2605, |
| "step": 167 |
| }, |
| { |
| "epoch": 12.0, |
| "grad_norm": 4.918086051940918, |
| "learning_rate": 2.0875e-05, |
| "loss": 1.2611, |
| "step": 168 |
| }, |
| { |
| "epoch": 12.071428571428571, |
| "grad_norm": 12.318052291870117, |
| "learning_rate": 2.1e-05, |
| "loss": 1.2175, |
| "step": 169 |
| }, |
| { |
| "epoch": 12.142857142857142, |
| "grad_norm": 4.972585201263428, |
| "learning_rate": 2.1125000000000002e-05, |
| "loss": 1.1154, |
| "step": 170 |
| }, |
| { |
| "epoch": 12.214285714285714, |
| "grad_norm": 4.9560441970825195, |
| "learning_rate": 2.125e-05, |
| "loss": 1.1845, |
| "step": 171 |
| }, |
| { |
| "epoch": 12.285714285714286, |
| "grad_norm": 5.026245594024658, |
| "learning_rate": 2.1375e-05, |
| "loss": 1.1234, |
| "step": 172 |
| }, |
| { |
| "epoch": 12.357142857142858, |
| "grad_norm": 7.455307483673096, |
| "learning_rate": 2.15e-05, |
| "loss": 1.1865, |
| "step": 173 |
| }, |
| { |
| "epoch": 12.428571428571429, |
| "grad_norm": 5.8239946365356445, |
| "learning_rate": 2.1625e-05, |
| "loss": 1.2061, |
| "step": 174 |
| }, |
| { |
| "epoch": 12.5, |
| "grad_norm": 7.225159168243408, |
| "learning_rate": 2.175e-05, |
| "loss": 1.2647, |
| "step": 175 |
| }, |
| { |
| "epoch": 12.571428571428571, |
| "grad_norm": 3.666829824447632, |
| "learning_rate": 2.1875e-05, |
| "loss": 1.1374, |
| "step": 176 |
| }, |
| { |
| "epoch": 12.642857142857142, |
| "grad_norm": 6.110956192016602, |
| "learning_rate": 2.2000000000000003e-05, |
| "loss": 1.1125, |
| "step": 177 |
| }, |
| { |
| "epoch": 12.714285714285714, |
| "grad_norm": 10.124720573425293, |
| "learning_rate": 2.2125000000000002e-05, |
| "loss": 1.1762, |
| "step": 178 |
| }, |
| { |
| "epoch": 12.785714285714286, |
| "grad_norm": 6.521277904510498, |
| "learning_rate": 2.2250000000000002e-05, |
| "loss": 1.2196, |
| "step": 179 |
| }, |
| { |
| "epoch": 12.857142857142858, |
| "grad_norm": 3.0751357078552246, |
| "learning_rate": 2.2375000000000002e-05, |
| "loss": 1.1462, |
| "step": 180 |
| }, |
| { |
| "epoch": 12.928571428571429, |
| "grad_norm": 4.803239345550537, |
| "learning_rate": 2.25e-05, |
| "loss": 1.0391, |
| "step": 181 |
| }, |
| { |
| "epoch": 13.0, |
| "grad_norm": 3.005739450454712, |
| "learning_rate": 2.2625e-05, |
| "loss": 1.0753, |
| "step": 182 |
| }, |
| { |
| "epoch": 13.071428571428571, |
| "grad_norm": 4.087983131408691, |
| "learning_rate": 2.275e-05, |
| "loss": 1.2163, |
| "step": 183 |
| }, |
| { |
| "epoch": 13.142857142857142, |
| "grad_norm": 3.475637674331665, |
| "learning_rate": 2.2875e-05, |
| "loss": 1.197, |
| "step": 184 |
| }, |
| { |
| "epoch": 13.214285714285714, |
| "grad_norm": 6.145781517028809, |
| "learning_rate": 2.3000000000000003e-05, |
| "loss": 1.1744, |
| "step": 185 |
| }, |
| { |
| "epoch": 13.285714285714286, |
| "grad_norm": 4.0928544998168945, |
| "learning_rate": 2.3125000000000003e-05, |
| "loss": 1.1189, |
| "step": 186 |
| }, |
| { |
| "epoch": 13.357142857142858, |
| "grad_norm": 9.976082801818848, |
| "learning_rate": 2.3250000000000003e-05, |
| "loss": 1.2488, |
| "step": 187 |
| }, |
| { |
| "epoch": 13.428571428571429, |
| "grad_norm": 6.045898914337158, |
| "learning_rate": 2.3375000000000002e-05, |
| "loss": 1.139, |
| "step": 188 |
| }, |
| { |
| "epoch": 13.5, |
| "grad_norm": 4.3854451179504395, |
| "learning_rate": 2.35e-05, |
| "loss": 1.0929, |
| "step": 189 |
| }, |
| { |
| "epoch": 13.571428571428571, |
| "grad_norm": 3.349039316177368, |
| "learning_rate": 2.3624999999999998e-05, |
| "loss": 1.0427, |
| "step": 190 |
| }, |
| { |
| "epoch": 13.642857142857142, |
| "grad_norm": 3.4710395336151123, |
| "learning_rate": 2.375e-05, |
| "loss": 1.1249, |
| "step": 191 |
| }, |
| { |
| "epoch": 13.714285714285714, |
| "grad_norm": 9.711404800415039, |
| "learning_rate": 2.3875e-05, |
| "loss": 1.2626, |
| "step": 192 |
| }, |
| { |
| "epoch": 13.785714285714286, |
| "grad_norm": 3.2990989685058594, |
| "learning_rate": 2.4e-05, |
| "loss": 1.1687, |
| "step": 193 |
| }, |
| { |
| "epoch": 13.857142857142858, |
| "grad_norm": 2.510340929031372, |
| "learning_rate": 2.4125e-05, |
| "loss": 1.0953, |
| "step": 194 |
| }, |
| { |
| "epoch": 13.928571428571429, |
| "grad_norm": 3.561298131942749, |
| "learning_rate": 2.425e-05, |
| "loss": 1.0411, |
| "step": 195 |
| }, |
| { |
| "epoch": 14.0, |
| "grad_norm": 4.719476699829102, |
| "learning_rate": 2.4375e-05, |
| "loss": 1.0803, |
| "step": 196 |
| }, |
| { |
| "epoch": 14.071428571428571, |
| "grad_norm": 3.3607053756713867, |
| "learning_rate": 2.45e-05, |
| "loss": 1.0709, |
| "step": 197 |
| }, |
| { |
| "epoch": 14.142857142857142, |
| "grad_norm": 4.803561687469482, |
| "learning_rate": 2.4625000000000002e-05, |
| "loss": 1.0335, |
| "step": 198 |
| }, |
| { |
| "epoch": 14.214285714285714, |
| "grad_norm": 3.8269684314727783, |
| "learning_rate": 2.4750000000000002e-05, |
| "loss": 1.0742, |
| "step": 199 |
| }, |
| { |
| "epoch": 14.285714285714286, |
| "grad_norm": 2.7520132064819336, |
| "learning_rate": 2.4875e-05, |
| "loss": 1.17, |
| "step": 200 |
| }, |
| { |
| "epoch": 14.357142857142858, |
| "grad_norm": 5.653459548950195, |
| "learning_rate": 2.5e-05, |
| "loss": 1.0918, |
| "step": 201 |
| }, |
| { |
| "epoch": 14.428571428571429, |
| "grad_norm": 2.774630546569824, |
| "learning_rate": 2.4999980961416097e-05, |
| "loss": 1.058, |
| "step": 202 |
| }, |
| { |
| "epoch": 14.5, |
| "grad_norm": 3.926753520965576, |
| "learning_rate": 2.499992384572238e-05, |
| "loss": 1.0942, |
| "step": 203 |
| }, |
| { |
| "epoch": 14.571428571428571, |
| "grad_norm": 2.9734742641448975, |
| "learning_rate": 2.4999828653092835e-05, |
| "loss": 1.007, |
| "step": 204 |
| }, |
| { |
| "epoch": 14.642857142857142, |
| "grad_norm": 2.9620471000671387, |
| "learning_rate": 2.4999695383817435e-05, |
| "loss": 1.0881, |
| "step": 205 |
| }, |
| { |
| "epoch": 14.714285714285714, |
| "grad_norm": 5.124941349029541, |
| "learning_rate": 2.499952403830214e-05, |
| "loss": 1.0806, |
| "step": 206 |
| }, |
| { |
| "epoch": 14.785714285714286, |
| "grad_norm": 3.0979349613189697, |
| "learning_rate": 2.4999314617068904e-05, |
| "loss": 1.0281, |
| "step": 207 |
| }, |
| { |
| "epoch": 14.857142857142858, |
| "grad_norm": 3.4293224811553955, |
| "learning_rate": 2.4999067120755652e-05, |
| "loss": 1.051, |
| "step": 208 |
| }, |
| { |
| "epoch": 14.928571428571429, |
| "grad_norm": 6.487819194793701, |
| "learning_rate": 2.4998781550116305e-05, |
| "loss": 1.0308, |
| "step": 209 |
| }, |
| { |
| "epoch": 15.0, |
| "grad_norm": 3.3049116134643555, |
| "learning_rate": 2.499845790602076e-05, |
| "loss": 0.9572, |
| "step": 210 |
| }, |
| { |
| "epoch": 15.071428571428571, |
| "grad_norm": 4.562008380889893, |
| "learning_rate": 2.4998096189454893e-05, |
| "loss": 1.0758, |
| "step": 211 |
| }, |
| { |
| "epoch": 15.142857142857142, |
| "grad_norm": 4.130517482757568, |
| "learning_rate": 2.4997696401520555e-05, |
| "loss": 1.0565, |
| "step": 212 |
| }, |
| { |
| "epoch": 15.214285714285714, |
| "grad_norm": 7.463335037231445, |
| "learning_rate": 2.499725854343557e-05, |
| "loss": 0.9738, |
| "step": 213 |
| }, |
| { |
| "epoch": 15.285714285714286, |
| "grad_norm": 5.149586200714111, |
| "learning_rate": 2.4996782616533732e-05, |
| "loss": 1.0234, |
| "step": 214 |
| }, |
| { |
| "epoch": 15.357142857142858, |
| "grad_norm": 5.69240140914917, |
| "learning_rate": 2.499626862226479e-05, |
| "loss": 1.1353, |
| "step": 215 |
| }, |
| { |
| "epoch": 15.428571428571429, |
| "grad_norm": 4.097994327545166, |
| "learning_rate": 2.4995716562194465e-05, |
| "loss": 1.0006, |
| "step": 216 |
| }, |
| { |
| "epoch": 15.5, |
| "grad_norm": 3.709420919418335, |
| "learning_rate": 2.499512643800443e-05, |
| "loss": 0.9626, |
| "step": 217 |
| }, |
| { |
| "epoch": 15.571428571428571, |
| "grad_norm": 6.93643856048584, |
| "learning_rate": 2.4994498251492302e-05, |
| "loss": 0.9727, |
| "step": 218 |
| }, |
| { |
| "epoch": 15.642857142857142, |
| "grad_norm": 2.359558582305908, |
| "learning_rate": 2.4993832004571646e-05, |
| "loss": 1.0169, |
| "step": 219 |
| }, |
| { |
| "epoch": 15.714285714285714, |
| "grad_norm": 3.7421205043792725, |
| "learning_rate": 2.4993127699271966e-05, |
| "loss": 1.0996, |
| "step": 220 |
| }, |
| { |
| "epoch": 15.785714285714286, |
| "grad_norm": 3.670799493789673, |
| "learning_rate": 2.49923853377387e-05, |
| "loss": 1.0078, |
| "step": 221 |
| }, |
| { |
| "epoch": 15.857142857142858, |
| "grad_norm": 3.7862744331359863, |
| "learning_rate": 2.4991604922233204e-05, |
| "loss": 0.9785, |
| "step": 222 |
| }, |
| { |
| "epoch": 15.928571428571429, |
| "grad_norm": 2.4504923820495605, |
| "learning_rate": 2.4990786455132764e-05, |
| "loss": 0.9815, |
| "step": 223 |
| }, |
| { |
| "epoch": 16.0, |
| "grad_norm": 4.063878536224365, |
| "learning_rate": 2.4989929938930576e-05, |
| "loss": 0.9463, |
| "step": 224 |
| }, |
| { |
| "epoch": 16.071428571428573, |
| "grad_norm": 2.9702260494232178, |
| "learning_rate": 2.498903537623573e-05, |
| "loss": 0.9618, |
| "step": 225 |
| }, |
| { |
| "epoch": 16.142857142857142, |
| "grad_norm": 2.672618865966797, |
| "learning_rate": 2.4988102769773227e-05, |
| "loss": 0.9338, |
| "step": 226 |
| }, |
| { |
| "epoch": 16.214285714285715, |
| "grad_norm": 6.381713390350342, |
| "learning_rate": 2.4987132122383936e-05, |
| "loss": 1.0038, |
| "step": 227 |
| }, |
| { |
| "epoch": 16.285714285714285, |
| "grad_norm": 2.369471549987793, |
| "learning_rate": 2.4986123437024627e-05, |
| "loss": 1.0239, |
| "step": 228 |
| }, |
| { |
| "epoch": 16.357142857142858, |
| "grad_norm": 2.5277535915374756, |
| "learning_rate": 2.4985076716767927e-05, |
| "loss": 0.9014, |
| "step": 229 |
| }, |
| { |
| "epoch": 16.428571428571427, |
| "grad_norm": 3.924433946609497, |
| "learning_rate": 2.4983991964802327e-05, |
| "loss": 0.966, |
| "step": 230 |
| }, |
| { |
| "epoch": 16.5, |
| "grad_norm": 3.6045427322387695, |
| "learning_rate": 2.4982869184432174e-05, |
| "loss": 0.9695, |
| "step": 231 |
| }, |
| { |
| "epoch": 16.571428571428573, |
| "grad_norm": 6.360043048858643, |
| "learning_rate": 2.498170837907765e-05, |
| "loss": 1.0349, |
| "step": 232 |
| }, |
| { |
| "epoch": 16.642857142857142, |
| "grad_norm": 4.736248970031738, |
| "learning_rate": 2.4980509552274765e-05, |
| "loss": 0.992, |
| "step": 233 |
| }, |
| { |
| "epoch": 16.714285714285715, |
| "grad_norm": 2.659874677658081, |
| "learning_rate": 2.4979272707675356e-05, |
| "loss": 0.9548, |
| "step": 234 |
| }, |
| { |
| "epoch": 16.785714285714285, |
| "grad_norm": 2.803654909133911, |
| "learning_rate": 2.497799784904707e-05, |
| "loss": 0.8971, |
| "step": 235 |
| }, |
| { |
| "epoch": 16.857142857142858, |
| "grad_norm": 4.661594867706299, |
| "learning_rate": 2.4976684980273338e-05, |
| "loss": 0.9161, |
| "step": 236 |
| }, |
| { |
| "epoch": 16.928571428571427, |
| "grad_norm": 3.7907001972198486, |
| "learning_rate": 2.4975334105353396e-05, |
| "loss": 0.8663, |
| "step": 237 |
| }, |
| { |
| "epoch": 17.0, |
| "grad_norm": 6.289381504058838, |
| "learning_rate": 2.497394522840224e-05, |
| "loss": 0.9157, |
| "step": 238 |
| }, |
| { |
| "epoch": 17.071428571428573, |
| "grad_norm": 3.7386388778686523, |
| "learning_rate": 2.4972518353650626e-05, |
| "loss": 0.9065, |
| "step": 239 |
| }, |
| { |
| "epoch": 17.142857142857142, |
| "grad_norm": 2.243206739425659, |
| "learning_rate": 2.497105348544507e-05, |
| "loss": 0.8066, |
| "step": 240 |
| }, |
| { |
| "epoch": 17.214285714285715, |
| "grad_norm": 3.9942452907562256, |
| "learning_rate": 2.4969550628247805e-05, |
| "loss": 0.9173, |
| "step": 241 |
| }, |
| { |
| "epoch": 17.285714285714285, |
| "grad_norm": 4.815189361572266, |
| "learning_rate": 2.49680097866368e-05, |
| "loss": 0.9629, |
| "step": 242 |
| }, |
| { |
| "epoch": 17.357142857142858, |
| "grad_norm": 5.594153881072998, |
| "learning_rate": 2.4966430965305727e-05, |
| "loss": 0.9673, |
| "step": 243 |
| }, |
| { |
| "epoch": 17.428571428571427, |
| "grad_norm": 4.274055004119873, |
| "learning_rate": 2.4964814169063948e-05, |
| "loss": 0.9222, |
| "step": 244 |
| }, |
| { |
| "epoch": 17.5, |
| "grad_norm": 3.3747873306274414, |
| "learning_rate": 2.4963159402836506e-05, |
| "loss": 0.8992, |
| "step": 245 |
| }, |
| { |
| "epoch": 17.571428571428573, |
| "grad_norm": 4.040309429168701, |
| "learning_rate": 2.49614666716641e-05, |
| "loss": 0.8257, |
| "step": 246 |
| }, |
| { |
| "epoch": 17.642857142857142, |
| "grad_norm": 3.115257740020752, |
| "learning_rate": 2.495973598070309e-05, |
| "loss": 0.9713, |
| "step": 247 |
| }, |
| { |
| "epoch": 17.714285714285715, |
| "grad_norm": 2.5480332374572754, |
| "learning_rate": 2.4957967335225456e-05, |
| "loss": 0.8687, |
| "step": 248 |
| }, |
| { |
| "epoch": 17.785714285714285, |
| "grad_norm": 3.8874218463897705, |
| "learning_rate": 2.4956160740618806e-05, |
| "loss": 0.9432, |
| "step": 249 |
| }, |
| { |
| "epoch": 17.857142857142858, |
| "grad_norm": 3.870405673980713, |
| "learning_rate": 2.495431620238633e-05, |
| "loss": 0.8584, |
| "step": 250 |
| }, |
| { |
| "epoch": 17.928571428571427, |
| "grad_norm": 2.534801483154297, |
| "learning_rate": 2.495243372614682e-05, |
| "loss": 0.9255, |
| "step": 251 |
| }, |
| { |
| "epoch": 18.0, |
| "grad_norm": 2.785574197769165, |
| "learning_rate": 2.495051331763462e-05, |
| "loss": 0.8727, |
| "step": 252 |
| }, |
| { |
| "epoch": 18.071428571428573, |
| "grad_norm": 4.180792331695557, |
| "learning_rate": 2.494855498269963e-05, |
| "loss": 0.9683, |
| "step": 253 |
| }, |
| { |
| "epoch": 18.142857142857142, |
| "grad_norm": 5.634334564208984, |
| "learning_rate": 2.4946558727307277e-05, |
| "loss": 0.9485, |
| "step": 254 |
| }, |
| { |
| "epoch": 18.214285714285715, |
| "grad_norm": 4.938521862030029, |
| "learning_rate": 2.4944524557538503e-05, |
| "loss": 0.878, |
| "step": 255 |
| }, |
| { |
| "epoch": 18.285714285714285, |
| "grad_norm": 3.9465174674987793, |
| "learning_rate": 2.4942452479589735e-05, |
| "loss": 0.9731, |
| "step": 256 |
| }, |
| { |
| "epoch": 18.357142857142858, |
| "grad_norm": 3.601600170135498, |
| "learning_rate": 2.494034249977289e-05, |
| "loss": 0.9057, |
| "step": 257 |
| }, |
| { |
| "epoch": 18.428571428571427, |
| "grad_norm": 3.65264892578125, |
| "learning_rate": 2.4938194624515333e-05, |
| "loss": 0.9126, |
| "step": 258 |
| }, |
| { |
| "epoch": 18.5, |
| "grad_norm": 4.7837677001953125, |
| "learning_rate": 2.4936008860359854e-05, |
| "loss": 0.9847, |
| "step": 259 |
| }, |
| { |
| "epoch": 18.571428571428573, |
| "grad_norm": 4.8172831535339355, |
| "learning_rate": 2.4933785213964677e-05, |
| "loss": 0.9202, |
| "step": 260 |
| }, |
| { |
| "epoch": 18.642857142857142, |
| "grad_norm": 4.068408012390137, |
| "learning_rate": 2.4931523692103418e-05, |
| "loss": 0.9072, |
| "step": 261 |
| }, |
| { |
| "epoch": 18.714285714285715, |
| "grad_norm": 4.106720924377441, |
| "learning_rate": 2.492922430166506e-05, |
| "loss": 0.8837, |
| "step": 262 |
| }, |
| { |
| "epoch": 18.785714285714285, |
| "grad_norm": 3.7029213905334473, |
| "learning_rate": 2.4926887049653943e-05, |
| "loss": 0.9464, |
| "step": 263 |
| }, |
| { |
| "epoch": 18.857142857142858, |
| "grad_norm": 3.3589251041412354, |
| "learning_rate": 2.492451194318975e-05, |
| "loss": 0.8956, |
| "step": 264 |
| }, |
| { |
| "epoch": 18.928571428571427, |
| "grad_norm": 3.4119558334350586, |
| "learning_rate": 2.4922098989507454e-05, |
| "loss": 0.9177, |
| "step": 265 |
| }, |
| { |
| "epoch": 19.0, |
| "grad_norm": 2.3078062534332275, |
| "learning_rate": 2.4919648195957344e-05, |
| "loss": 0.8086, |
| "step": 266 |
| }, |
| { |
| "epoch": 19.071428571428573, |
| "grad_norm": 2.213747978210449, |
| "learning_rate": 2.4917159570004954e-05, |
| "loss": 0.8417, |
| "step": 267 |
| }, |
| { |
| "epoch": 19.142857142857142, |
| "grad_norm": 2.629509687423706, |
| "learning_rate": 2.491463311923108e-05, |
| "loss": 0.9628, |
| "step": 268 |
| }, |
| { |
| "epoch": 19.214285714285715, |
| "grad_norm": 1.877100944519043, |
| "learning_rate": 2.491206885133171e-05, |
| "loss": 0.8174, |
| "step": 269 |
| }, |
| { |
| "epoch": 19.285714285714285, |
| "grad_norm": 2.653271436691284, |
| "learning_rate": 2.490946677411807e-05, |
| "loss": 0.8823, |
| "step": 270 |
| }, |
| { |
| "epoch": 19.357142857142858, |
| "grad_norm": 2.488171339035034, |
| "learning_rate": 2.4906826895516528e-05, |
| "loss": 0.888, |
| "step": 271 |
| }, |
| { |
| "epoch": 19.428571428571427, |
| "grad_norm": 2.5691208839416504, |
| "learning_rate": 2.490414922356861e-05, |
| "loss": 0.87, |
| "step": 272 |
| }, |
| { |
| "epoch": 19.5, |
| "grad_norm": 3.2644972801208496, |
| "learning_rate": 2.4901433766430975e-05, |
| "loss": 0.9013, |
| "step": 273 |
| }, |
| { |
| "epoch": 19.571428571428573, |
| "grad_norm": 5.025469779968262, |
| "learning_rate": 2.4898680532375374e-05, |
| "loss": 0.9013, |
| "step": 274 |
| }, |
| { |
| "epoch": 19.642857142857142, |
| "grad_norm": 3.607896566390991, |
| "learning_rate": 2.489588952978863e-05, |
| "loss": 0.8409, |
| "step": 275 |
| }, |
| { |
| "epoch": 19.714285714285715, |
| "grad_norm": 2.6993868350982666, |
| "learning_rate": 2.4893060767172632e-05, |
| "loss": 0.7845, |
| "step": 276 |
| }, |
| { |
| "epoch": 19.785714285714285, |
| "grad_norm": 2.4249427318573, |
| "learning_rate": 2.489019425314427e-05, |
| "loss": 0.8113, |
| "step": 277 |
| }, |
| { |
| "epoch": 19.857142857142858, |
| "grad_norm": 4.231166362762451, |
| "learning_rate": 2.4887289996435452e-05, |
| "loss": 0.8879, |
| "step": 278 |
| }, |
| { |
| "epoch": 19.928571428571427, |
| "grad_norm": 7.234214782714844, |
| "learning_rate": 2.4884348005893045e-05, |
| "loss": 0.8847, |
| "step": 279 |
| }, |
| { |
| "epoch": 20.0, |
| "grad_norm": 4.228268146514893, |
| "learning_rate": 2.488136829047886e-05, |
| "loss": 0.8504, |
| "step": 280 |
| }, |
| { |
| "epoch": 20.071428571428573, |
| "grad_norm": 2.944539785385132, |
| "learning_rate": 2.487835085926963e-05, |
| "loss": 0.845, |
| "step": 281 |
| }, |
| { |
| "epoch": 20.142857142857142, |
| "grad_norm": 2.7270450592041016, |
| "learning_rate": 2.487529572145697e-05, |
| "loss": 0.8335, |
| "step": 282 |
| }, |
| { |
| "epoch": 20.214285714285715, |
| "grad_norm": 2.7035233974456787, |
| "learning_rate": 2.4872202886347362e-05, |
| "loss": 0.815, |
| "step": 283 |
| }, |
| { |
| "epoch": 20.285714285714285, |
| "grad_norm": 3.4698843955993652, |
| "learning_rate": 2.486907236336212e-05, |
| "loss": 0.875, |
| "step": 284 |
| }, |
| { |
| "epoch": 20.357142857142858, |
| "grad_norm": 4.334729194641113, |
| "learning_rate": 2.4865904162037358e-05, |
| "loss": 0.8447, |
| "step": 285 |
| }, |
| { |
| "epoch": 20.428571428571427, |
| "grad_norm": 3.5744895935058594, |
| "learning_rate": 2.4862698292023963e-05, |
| "loss": 0.792, |
| "step": 286 |
| }, |
| { |
| "epoch": 20.5, |
| "grad_norm": 3.5507006645202637, |
| "learning_rate": 2.4859454763087577e-05, |
| "loss": 0.7948, |
| "step": 287 |
| }, |
| { |
| "epoch": 20.571428571428573, |
| "grad_norm": 2.2967231273651123, |
| "learning_rate": 2.4856173585108544e-05, |
| "loss": 0.7975, |
| "step": 288 |
| }, |
| { |
| "epoch": 20.642857142857142, |
| "grad_norm": 2.5421581268310547, |
| "learning_rate": 2.4852854768081912e-05, |
| "loss": 0.8074, |
| "step": 289 |
| }, |
| { |
| "epoch": 20.714285714285715, |
| "grad_norm": 2.4942188262939453, |
| "learning_rate": 2.4849498322117364e-05, |
| "loss": 0.7962, |
| "step": 290 |
| }, |
| { |
| "epoch": 20.785714285714285, |
| "grad_norm": 4.169875621795654, |
| "learning_rate": 2.4846104257439222e-05, |
| "loss": 0.8181, |
| "step": 291 |
| }, |
| { |
| "epoch": 20.857142857142858, |
| "grad_norm": 6.13093376159668, |
| "learning_rate": 2.4842672584386396e-05, |
| "loss": 0.8434, |
| "step": 292 |
| }, |
| { |
| "epoch": 20.928571428571427, |
| "grad_norm": 3.3157453536987305, |
| "learning_rate": 2.483920331341235e-05, |
| "loss": 0.8108, |
| "step": 293 |
| }, |
| { |
| "epoch": 21.0, |
| "grad_norm": 3.0505266189575195, |
| "learning_rate": 2.4835696455085093e-05, |
| "loss": 0.7412, |
| "step": 294 |
| }, |
| { |
| "epoch": 21.071428571428573, |
| "grad_norm": 5.5504350662231445, |
| "learning_rate": 2.483215202008712e-05, |
| "loss": 0.8437, |
| "step": 295 |
| }, |
| { |
| "epoch": 21.142857142857142, |
| "grad_norm": 6.492306709289551, |
| "learning_rate": 2.4828570019215396e-05, |
| "loss": 0.8533, |
| "step": 296 |
| }, |
| { |
| "epoch": 21.214285714285715, |
| "grad_norm": 3.0854108333587646, |
| "learning_rate": 2.4824950463381314e-05, |
| "loss": 0.7881, |
| "step": 297 |
| }, |
| { |
| "epoch": 21.285714285714285, |
| "grad_norm": 3.006676435470581, |
| "learning_rate": 2.482129336361067e-05, |
| "loss": 0.8156, |
| "step": 298 |
| }, |
| { |
| "epoch": 21.357142857142858, |
| "grad_norm": 3.312669515609741, |
| "learning_rate": 2.481759873104363e-05, |
| "loss": 0.8011, |
| "step": 299 |
| }, |
| { |
| "epoch": 21.428571428571427, |
| "grad_norm": 5.346591949462891, |
| "learning_rate": 2.4813866576934676e-05, |
| "loss": 0.7895, |
| "step": 300 |
| }, |
| { |
| "epoch": 21.5, |
| "grad_norm": 5.8296074867248535, |
| "learning_rate": 2.4810096912652604e-05, |
| "loss": 0.893, |
| "step": 301 |
| }, |
| { |
| "epoch": 21.571428571428573, |
| "grad_norm": 3.022014617919922, |
| "learning_rate": 2.480628974968046e-05, |
| "loss": 0.7875, |
| "step": 302 |
| }, |
| { |
| "epoch": 21.642857142857142, |
| "grad_norm": 3.372377395629883, |
| "learning_rate": 2.4802445099615525e-05, |
| "loss": 0.8107, |
| "step": 303 |
| }, |
| { |
| "epoch": 21.714285714285715, |
| "grad_norm": 4.238394260406494, |
| "learning_rate": 2.479856297416927e-05, |
| "loss": 0.7879, |
| "step": 304 |
| }, |
| { |
| "epoch": 21.785714285714285, |
| "grad_norm": 4.552757263183594, |
| "learning_rate": 2.4794643385167327e-05, |
| "loss": 0.8352, |
| "step": 305 |
| }, |
| { |
| "epoch": 21.857142857142858, |
| "grad_norm": 4.247888088226318, |
| "learning_rate": 2.4790686344549436e-05, |
| "loss": 0.7733, |
| "step": 306 |
| }, |
| { |
| "epoch": 21.928571428571427, |
| "grad_norm": 2.4774329662323, |
| "learning_rate": 2.478669186436943e-05, |
| "loss": 0.7338, |
| "step": 307 |
| }, |
| { |
| "epoch": 22.0, |
| "grad_norm": 2.973707675933838, |
| "learning_rate": 2.478265995679519e-05, |
| "loss": 0.7532, |
| "step": 308 |
| }, |
| { |
| "epoch": 22.071428571428573, |
| "grad_norm": 5.437268257141113, |
| "learning_rate": 2.4778590634108613e-05, |
| "loss": 0.8951, |
| "step": 309 |
| }, |
| { |
| "epoch": 22.142857142857142, |
| "grad_norm": 5.616093635559082, |
| "learning_rate": 2.4774483908705546e-05, |
| "loss": 0.8117, |
| "step": 310 |
| }, |
| { |
| "epoch": 22.214285714285715, |
| "grad_norm": 3.7091429233551025, |
| "learning_rate": 2.4770339793095802e-05, |
| "loss": 0.8166, |
| "step": 311 |
| }, |
| { |
| "epoch": 22.285714285714285, |
| "grad_norm": 3.873770236968994, |
| "learning_rate": 2.4766158299903062e-05, |
| "loss": 0.7786, |
| "step": 312 |
| }, |
| { |
| "epoch": 22.357142857142858, |
| "grad_norm": 3.692065477371216, |
| "learning_rate": 2.4761939441864895e-05, |
| "loss": 0.8148, |
| "step": 313 |
| }, |
| { |
| "epoch": 22.428571428571427, |
| "grad_norm": 4.50544548034668, |
| "learning_rate": 2.4757683231832662e-05, |
| "loss": 0.9163, |
| "step": 314 |
| }, |
| { |
| "epoch": 22.5, |
| "grad_norm": 4.464144229888916, |
| "learning_rate": 2.4753389682771523e-05, |
| "loss": 0.8223, |
| "step": 315 |
| }, |
| { |
| "epoch": 22.571428571428573, |
| "grad_norm": 3.729602575302124, |
| "learning_rate": 2.474905880776037e-05, |
| "loss": 0.8479, |
| "step": 316 |
| }, |
| { |
| "epoch": 22.642857142857142, |
| "grad_norm": 3.674757957458496, |
| "learning_rate": 2.47446906199918e-05, |
| "loss": 0.7319, |
| "step": 317 |
| }, |
| { |
| "epoch": 22.714285714285715, |
| "grad_norm": 4.361830711364746, |
| "learning_rate": 2.4740285132772072e-05, |
| "loss": 0.8054, |
| "step": 318 |
| }, |
| { |
| "epoch": 22.785714285714285, |
| "grad_norm": 4.365119934082031, |
| "learning_rate": 2.4735842359521064e-05, |
| "loss": 0.801, |
| "step": 319 |
| }, |
| { |
| "epoch": 22.857142857142858, |
| "grad_norm": 4.092256546020508, |
| "learning_rate": 2.4731362313772233e-05, |
| "loss": 0.8389, |
| "step": 320 |
| }, |
| { |
| "epoch": 22.928571428571427, |
| "grad_norm": 3.239090919494629, |
| "learning_rate": 2.4726845009172572e-05, |
| "loss": 0.7814, |
| "step": 321 |
| }, |
| { |
| "epoch": 23.0, |
| "grad_norm": 2.6484477519989014, |
| "learning_rate": 2.4722290459482578e-05, |
| "loss": 0.7468, |
| "step": 322 |
| }, |
| { |
| "epoch": 23.071428571428573, |
| "grad_norm": 3.275247573852539, |
| "learning_rate": 2.47176986785762e-05, |
| "loss": 0.827, |
| "step": 323 |
| }, |
| { |
| "epoch": 23.142857142857142, |
| "grad_norm": 3.040330410003662, |
| "learning_rate": 2.47130696804408e-05, |
| "loss": 0.7849, |
| "step": 324 |
| }, |
| { |
| "epoch": 23.214285714285715, |
| "grad_norm": 4.0800395011901855, |
| "learning_rate": 2.47084034791771e-05, |
| "loss": 0.856, |
| "step": 325 |
| }, |
| { |
| "epoch": 23.285714285714285, |
| "grad_norm": 3.5290443897247314, |
| "learning_rate": 2.4703700088999167e-05, |
| "loss": 0.825, |
| "step": 326 |
| }, |
| { |
| "epoch": 23.357142857142858, |
| "grad_norm": 3.670090436935425, |
| "learning_rate": 2.4698959524234346e-05, |
| "loss": 0.8061, |
| "step": 327 |
| }, |
| { |
| "epoch": 23.428571428571427, |
| "grad_norm": 1.9602779150009155, |
| "learning_rate": 2.4694181799323206e-05, |
| "loss": 0.7803, |
| "step": 328 |
| }, |
| { |
| "epoch": 23.5, |
| "grad_norm": 3.8044676780700684, |
| "learning_rate": 2.468936692881954e-05, |
| "loss": 0.767, |
| "step": 329 |
| }, |
| { |
| "epoch": 23.571428571428573, |
| "grad_norm": 3.5850207805633545, |
| "learning_rate": 2.4684514927390274e-05, |
| "loss": 0.7555, |
| "step": 330 |
| }, |
| { |
| "epoch": 23.642857142857142, |
| "grad_norm": 2.4394097328186035, |
| "learning_rate": 2.4679625809815443e-05, |
| "loss": 0.7911, |
| "step": 331 |
| }, |
| { |
| "epoch": 23.714285714285715, |
| "grad_norm": 3.6944406032562256, |
| "learning_rate": 2.467469959098815e-05, |
| "loss": 0.7708, |
| "step": 332 |
| }, |
| { |
| "epoch": 23.785714285714285, |
| "grad_norm": 3.9482243061065674, |
| "learning_rate": 2.4669736285914505e-05, |
| "loss": 0.8035, |
| "step": 333 |
| }, |
| { |
| "epoch": 23.857142857142858, |
| "grad_norm": 4.452454566955566, |
| "learning_rate": 2.4664735909713606e-05, |
| "loss": 0.7837, |
| "step": 334 |
| }, |
| { |
| "epoch": 23.928571428571427, |
| "grad_norm": 3.8159029483795166, |
| "learning_rate": 2.465969847761746e-05, |
| "loss": 0.8188, |
| "step": 335 |
| }, |
| { |
| "epoch": 24.0, |
| "grad_norm": 2.743255138397217, |
| "learning_rate": 2.4654624004970957e-05, |
| "loss": 0.7004, |
| "step": 336 |
| }, |
| { |
| "epoch": 24.071428571428573, |
| "grad_norm": 3.636960506439209, |
| "learning_rate": 2.464951250723183e-05, |
| "loss": 0.7809, |
| "step": 337 |
| }, |
| { |
| "epoch": 24.142857142857142, |
| "grad_norm": 3.9498450756073, |
| "learning_rate": 2.4644363999970576e-05, |
| "loss": 0.7903, |
| "step": 338 |
| }, |
| { |
| "epoch": 24.214285714285715, |
| "grad_norm": 3.0506386756896973, |
| "learning_rate": 2.4639178498870452e-05, |
| "loss": 0.8114, |
| "step": 339 |
| }, |
| { |
| "epoch": 24.285714285714285, |
| "grad_norm": 3.2507994174957275, |
| "learning_rate": 2.4633956019727385e-05, |
| "loss": 0.7556, |
| "step": 340 |
| }, |
| { |
| "epoch": 24.357142857142858, |
| "grad_norm": 5.005919933319092, |
| "learning_rate": 2.4628696578449956e-05, |
| "loss": 0.8456, |
| "step": 341 |
| }, |
| { |
| "epoch": 24.428571428571427, |
| "grad_norm": 3.5123672485351562, |
| "learning_rate": 2.4623400191059335e-05, |
| "loss": 0.7921, |
| "step": 342 |
| }, |
| { |
| "epoch": 24.5, |
| "grad_norm": 3.3725931644439697, |
| "learning_rate": 2.4618066873689238e-05, |
| "loss": 0.7986, |
| "step": 343 |
| }, |
| { |
| "epoch": 24.571428571428573, |
| "grad_norm": 2.6177029609680176, |
| "learning_rate": 2.461269664258587e-05, |
| "loss": 0.7449, |
| "step": 344 |
| }, |
| { |
| "epoch": 24.642857142857142, |
| "grad_norm": 5.391937732696533, |
| "learning_rate": 2.4607289514107888e-05, |
| "loss": 0.7433, |
| "step": 345 |
| }, |
| { |
| "epoch": 24.714285714285715, |
| "grad_norm": 2.888105630874634, |
| "learning_rate": 2.460184550472635e-05, |
| "loss": 0.7079, |
| "step": 346 |
| }, |
| { |
| "epoch": 24.785714285714285, |
| "grad_norm": 2.3010149002075195, |
| "learning_rate": 2.4596364631024643e-05, |
| "loss": 0.7376, |
| "step": 347 |
| }, |
| { |
| "epoch": 24.857142857142858, |
| "grad_norm": 3.590585470199585, |
| "learning_rate": 2.459084690969846e-05, |
| "loss": 0.7532, |
| "step": 348 |
| }, |
| { |
| "epoch": 24.928571428571427, |
| "grad_norm": 6.115037441253662, |
| "learning_rate": 2.4585292357555746e-05, |
| "loss": 0.7568, |
| "step": 349 |
| }, |
| { |
| "epoch": 25.0, |
| "grad_norm": 2.7985963821411133, |
| "learning_rate": 2.457970099151662e-05, |
| "loss": 0.7001, |
| "step": 350 |
| }, |
| { |
| "epoch": 25.071428571428573, |
| "grad_norm": 5.373832702636719, |
| "learning_rate": 2.4574072828613354e-05, |
| "loss": 0.8254, |
| "step": 351 |
| }, |
| { |
| "epoch": 25.142857142857142, |
| "grad_norm": 3.988442897796631, |
| "learning_rate": 2.4568407885990313e-05, |
| "loss": 0.7305, |
| "step": 352 |
| }, |
| { |
| "epoch": 25.214285714285715, |
| "grad_norm": 3.5456085205078125, |
| "learning_rate": 2.4562706180903894e-05, |
| "loss": 0.7546, |
| "step": 353 |
| }, |
| { |
| "epoch": 25.285714285714285, |
| "grad_norm": 3.8878087997436523, |
| "learning_rate": 2.4556967730722478e-05, |
| "loss": 0.794, |
| "step": 354 |
| }, |
| { |
| "epoch": 25.357142857142858, |
| "grad_norm": 3.997696876525879, |
| "learning_rate": 2.455119255292638e-05, |
| "loss": 0.7983, |
| "step": 355 |
| }, |
| { |
| "epoch": 25.428571428571427, |
| "grad_norm": 5.822238445281982, |
| "learning_rate": 2.4545380665107786e-05, |
| "loss": 0.899, |
| "step": 356 |
| }, |
| { |
| "epoch": 25.5, |
| "grad_norm": 3.2456319332122803, |
| "learning_rate": 2.453953208497073e-05, |
| "loss": 0.7284, |
| "step": 357 |
| }, |
| { |
| "epoch": 25.571428571428573, |
| "grad_norm": 3.3598856925964355, |
| "learning_rate": 2.4533646830330986e-05, |
| "loss": 0.8056, |
| "step": 358 |
| }, |
| { |
| "epoch": 25.642857142857142, |
| "grad_norm": 3.6620326042175293, |
| "learning_rate": 2.452772491911607e-05, |
| "loss": 0.7994, |
| "step": 359 |
| }, |
| { |
| "epoch": 25.714285714285715, |
| "grad_norm": 4.005545139312744, |
| "learning_rate": 2.4521766369365142e-05, |
| "loss": 0.7773, |
| "step": 360 |
| }, |
| { |
| "epoch": 25.785714285714285, |
| "grad_norm": 4.023702144622803, |
| "learning_rate": 2.4515771199228987e-05, |
| "loss": 0.7717, |
| "step": 361 |
| }, |
| { |
| "epoch": 25.857142857142858, |
| "grad_norm": 1.9900436401367188, |
| "learning_rate": 2.450973942696993e-05, |
| "loss": 0.7331, |
| "step": 362 |
| }, |
| { |
| "epoch": 25.928571428571427, |
| "grad_norm": 5.718785285949707, |
| "learning_rate": 2.450367107096179e-05, |
| "loss": 0.7818, |
| "step": 363 |
| }, |
| { |
| "epoch": 26.0, |
| "grad_norm": 3.5384066104888916, |
| "learning_rate": 2.449756614968984e-05, |
| "loss": 0.8165, |
| "step": 364 |
| }, |
| { |
| "epoch": 26.071428571428573, |
| "grad_norm": 2.2805933952331543, |
| "learning_rate": 2.449142468175072e-05, |
| "loss": 0.68, |
| "step": 365 |
| }, |
| { |
| "epoch": 26.142857142857142, |
| "grad_norm": 2.818986654281616, |
| "learning_rate": 2.4485246685852413e-05, |
| "loss": 0.6765, |
| "step": 366 |
| }, |
| { |
| "epoch": 26.214285714285715, |
| "grad_norm": 3.089205741882324, |
| "learning_rate": 2.4479032180814166e-05, |
| "loss": 0.6901, |
| "step": 367 |
| }, |
| { |
| "epoch": 26.285714285714285, |
| "grad_norm": 2.990636110305786, |
| "learning_rate": 2.447278118556644e-05, |
| "loss": 0.6839, |
| "step": 368 |
| }, |
| { |
| "epoch": 26.357142857142858, |
| "grad_norm": 2.0169079303741455, |
| "learning_rate": 2.446649371915084e-05, |
| "loss": 0.6499, |
| "step": 369 |
| }, |
| { |
| "epoch": 26.428571428571427, |
| "grad_norm": 2.073349714279175, |
| "learning_rate": 2.4460169800720095e-05, |
| "loss": 0.6312, |
| "step": 370 |
| }, |
| { |
| "epoch": 26.5, |
| "grad_norm": 5.544507026672363, |
| "learning_rate": 2.4453809449537947e-05, |
| "loss": 0.7886, |
| "step": 371 |
| }, |
| { |
| "epoch": 26.571428571428573, |
| "grad_norm": 3.550513505935669, |
| "learning_rate": 2.4447412684979127e-05, |
| "loss": 0.7682, |
| "step": 372 |
| }, |
| { |
| "epoch": 26.642857142857142, |
| "grad_norm": 4.850046634674072, |
| "learning_rate": 2.4440979526529295e-05, |
| "loss": 0.7891, |
| "step": 373 |
| }, |
| { |
| "epoch": 26.714285714285715, |
| "grad_norm": 5.702778339385986, |
| "learning_rate": 2.4434509993784972e-05, |
| "loss": 0.7478, |
| "step": 374 |
| }, |
| { |
| "epoch": 26.785714285714285, |
| "grad_norm": 2.3614633083343506, |
| "learning_rate": 2.4428004106453462e-05, |
| "loss": 0.6622, |
| "step": 375 |
| }, |
| { |
| "epoch": 26.857142857142858, |
| "grad_norm": 4.546297073364258, |
| "learning_rate": 2.4421461884352836e-05, |
| "loss": 0.7229, |
| "step": 376 |
| }, |
| { |
| "epoch": 26.928571428571427, |
| "grad_norm": 4.9091386795043945, |
| "learning_rate": 2.4414883347411836e-05, |
| "loss": 0.7186, |
| "step": 377 |
| }, |
| { |
| "epoch": 27.0, |
| "grad_norm": 3.949092388153076, |
| "learning_rate": 2.440826851566983e-05, |
| "loss": 0.7257, |
| "step": 378 |
| }, |
| { |
| "epoch": 27.071428571428573, |
| "grad_norm": 5.0857439041137695, |
| "learning_rate": 2.4401617409276735e-05, |
| "loss": 0.6953, |
| "step": 379 |
| }, |
| { |
| "epoch": 27.142857142857142, |
| "grad_norm": 4.325730323791504, |
| "learning_rate": 2.439493004849298e-05, |
| "loss": 0.6696, |
| "step": 380 |
| }, |
| { |
| "epoch": 27.214285714285715, |
| "grad_norm": 3.15690541267395, |
| "learning_rate": 2.438820645368942e-05, |
| "loss": 0.7532, |
| "step": 381 |
| }, |
| { |
| "epoch": 27.285714285714285, |
| "grad_norm": 4.08896017074585, |
| "learning_rate": 2.4381446645347297e-05, |
| "loss": 0.7039, |
| "step": 382 |
| }, |
| { |
| "epoch": 27.357142857142858, |
| "grad_norm": 3.996779441833496, |
| "learning_rate": 2.4374650644058156e-05, |
| "loss": 0.7489, |
| "step": 383 |
| }, |
| { |
| "epoch": 27.428571428571427, |
| "grad_norm": 4.629286766052246, |
| "learning_rate": 2.43678184705238e-05, |
| "loss": 0.7634, |
| "step": 384 |
| }, |
| { |
| "epoch": 27.5, |
| "grad_norm": 4.668010234832764, |
| "learning_rate": 2.4360950145556208e-05, |
| "loss": 0.7321, |
| "step": 385 |
| }, |
| { |
| "epoch": 27.571428571428573, |
| "grad_norm": 2.8770227432250977, |
| "learning_rate": 2.4354045690077492e-05, |
| "loss": 0.6417, |
| "step": 386 |
| }, |
| { |
| "epoch": 27.642857142857142, |
| "grad_norm": 4.506302356719971, |
| "learning_rate": 2.4347105125119824e-05, |
| "loss": 0.6698, |
| "step": 387 |
| }, |
| { |
| "epoch": 27.714285714285715, |
| "grad_norm": 2.6643428802490234, |
| "learning_rate": 2.4340128471825362e-05, |
| "loss": 0.6938, |
| "step": 388 |
| }, |
| { |
| "epoch": 27.785714285714285, |
| "grad_norm": 2.8056280612945557, |
| "learning_rate": 2.4333115751446208e-05, |
| "loss": 0.6743, |
| "step": 389 |
| }, |
| { |
| "epoch": 27.857142857142858, |
| "grad_norm": 2.7733471393585205, |
| "learning_rate": 2.4326066985344318e-05, |
| "loss": 0.6573, |
| "step": 390 |
| }, |
| { |
| "epoch": 27.928571428571427, |
| "grad_norm": 5.0971174240112305, |
| "learning_rate": 2.4318982194991463e-05, |
| "loss": 0.6754, |
| "step": 391 |
| }, |
| { |
| "epoch": 28.0, |
| "grad_norm": 6.083596706390381, |
| "learning_rate": 2.4311861401969138e-05, |
| "loss": 0.7646, |
| "step": 392 |
| }, |
| { |
| "epoch": 28.071428571428573, |
| "grad_norm": 4.150708198547363, |
| "learning_rate": 2.4304704627968515e-05, |
| "loss": 0.6958, |
| "step": 393 |
| }, |
| { |
| "epoch": 28.142857142857142, |
| "grad_norm": 3.6580939292907715, |
| "learning_rate": 2.429751189479037e-05, |
| "loss": 0.6848, |
| "step": 394 |
| }, |
| { |
| "epoch": 28.214285714285715, |
| "grad_norm": 2.1017942428588867, |
| "learning_rate": 2.429028322434501e-05, |
| "loss": 0.6576, |
| "step": 395 |
| }, |
| { |
| "epoch": 28.285714285714285, |
| "grad_norm": 1.8062525987625122, |
| "learning_rate": 2.4283018638652234e-05, |
| "loss": 0.6963, |
| "step": 396 |
| }, |
| { |
| "epoch": 28.357142857142858, |
| "grad_norm": 3.1975321769714355, |
| "learning_rate": 2.427571815984121e-05, |
| "loss": 0.6599, |
| "step": 397 |
| }, |
| { |
| "epoch": 28.428571428571427, |
| "grad_norm": 3.0259714126586914, |
| "learning_rate": 2.4268381810150474e-05, |
| "loss": 0.6959, |
| "step": 398 |
| }, |
| { |
| "epoch": 28.5, |
| "grad_norm": 5.031489849090576, |
| "learning_rate": 2.426100961192782e-05, |
| "loss": 0.7208, |
| "step": 399 |
| }, |
| { |
| "epoch": 28.571428571428573, |
| "grad_norm": 3.5637011528015137, |
| "learning_rate": 2.4253601587630236e-05, |
| "loss": 0.6896, |
| "step": 400 |
| }, |
| { |
| "epoch": 28.642857142857142, |
| "grad_norm": 3.333253860473633, |
| "learning_rate": 2.4246157759823855e-05, |
| "loss": 0.6815, |
| "step": 401 |
| }, |
| { |
| "epoch": 28.714285714285715, |
| "grad_norm": 3.422375440597534, |
| "learning_rate": 2.4238678151183863e-05, |
| "loss": 0.6606, |
| "step": 402 |
| }, |
| { |
| "epoch": 28.785714285714285, |
| "grad_norm": 2.458256721496582, |
| "learning_rate": 2.423116278449445e-05, |
| "loss": 0.6484, |
| "step": 403 |
| }, |
| { |
| "epoch": 28.857142857142858, |
| "grad_norm": 4.026810169219971, |
| "learning_rate": 2.4223611682648724e-05, |
| "loss": 0.7167, |
| "step": 404 |
| }, |
| { |
| "epoch": 28.928571428571427, |
| "grad_norm": 3.2683417797088623, |
| "learning_rate": 2.4216024868648644e-05, |
| "loss": 0.6899, |
| "step": 405 |
| }, |
| { |
| "epoch": 29.0, |
| "grad_norm": 3.54608416557312, |
| "learning_rate": 2.4208402365604972e-05, |
| "loss": 0.6843, |
| "step": 406 |
| }, |
| { |
| "epoch": 29.071428571428573, |
| "grad_norm": 4.817044734954834, |
| "learning_rate": 2.420074419673717e-05, |
| "loss": 0.6544, |
| "step": 407 |
| }, |
| { |
| "epoch": 29.142857142857142, |
| "grad_norm": 3.0443451404571533, |
| "learning_rate": 2.4193050385373344e-05, |
| "loss": 0.6134, |
| "step": 408 |
| }, |
| { |
| "epoch": 29.214285714285715, |
| "grad_norm": 6.059633731842041, |
| "learning_rate": 2.418532095495018e-05, |
| "loss": 0.6718, |
| "step": 409 |
| }, |
| { |
| "epoch": 29.285714285714285, |
| "grad_norm": 2.560316324234009, |
| "learning_rate": 2.417755592901287e-05, |
| "loss": 0.6134, |
| "step": 410 |
| }, |
| { |
| "epoch": 29.357142857142858, |
| "grad_norm": 3.06530499458313, |
| "learning_rate": 2.4169755331215023e-05, |
| "loss": 0.6158, |
| "step": 411 |
| }, |
| { |
| "epoch": 29.428571428571427, |
| "grad_norm": 2.439854621887207, |
| "learning_rate": 2.4161919185318617e-05, |
| "loss": 0.6428, |
| "step": 412 |
| }, |
| { |
| "epoch": 29.5, |
| "grad_norm": 4.74956750869751, |
| "learning_rate": 2.4154047515193904e-05, |
| "loss": 0.6847, |
| "step": 413 |
| }, |
| { |
| "epoch": 29.571428571428573, |
| "grad_norm": 3.4892289638519287, |
| "learning_rate": 2.4146140344819363e-05, |
| "loss": 0.6293, |
| "step": 414 |
| }, |
| { |
| "epoch": 29.642857142857142, |
| "grad_norm": 2.6727051734924316, |
| "learning_rate": 2.4138197698281606e-05, |
| "loss": 0.6429, |
| "step": 415 |
| }, |
| { |
| "epoch": 29.714285714285715, |
| "grad_norm": 4.598572254180908, |
| "learning_rate": 2.413021959977531e-05, |
| "loss": 0.6504, |
| "step": 416 |
| }, |
| { |
| "epoch": 29.785714285714285, |
| "grad_norm": 5.181699752807617, |
| "learning_rate": 2.4122206073603142e-05, |
| "loss": 0.6994, |
| "step": 417 |
| }, |
| { |
| "epoch": 29.857142857142858, |
| "grad_norm": 4.484119415283203, |
| "learning_rate": 2.4114157144175703e-05, |
| "loss": 0.7659, |
| "step": 418 |
| }, |
| { |
| "epoch": 29.928571428571427, |
| "grad_norm": 4.818487644195557, |
| "learning_rate": 2.4106072836011422e-05, |
| "loss": 0.7583, |
| "step": 419 |
| }, |
| { |
| "epoch": 30.0, |
| "grad_norm": 3.408137083053589, |
| "learning_rate": 2.40979531737365e-05, |
| "loss": 0.6579, |
| "step": 420 |
| }, |
| { |
| "epoch": 30.071428571428573, |
| "grad_norm": 4.11182975769043, |
| "learning_rate": 2.4089798182084845e-05, |
| "loss": 0.69, |
| "step": 421 |
| }, |
| { |
| "epoch": 30.142857142857142, |
| "grad_norm": 4.751646518707275, |
| "learning_rate": 2.4081607885897966e-05, |
| "loss": 0.7584, |
| "step": 422 |
| }, |
| { |
| "epoch": 30.214285714285715, |
| "grad_norm": 5.190206050872803, |
| "learning_rate": 2.407338231012494e-05, |
| "loss": 0.7468, |
| "step": 423 |
| }, |
| { |
| "epoch": 30.285714285714285, |
| "grad_norm": 3.750694513320923, |
| "learning_rate": 2.406512147982228e-05, |
| "loss": 0.743, |
| "step": 424 |
| }, |
| { |
| "epoch": 30.357142857142858, |
| "grad_norm": 3.4604289531707764, |
| "learning_rate": 2.4056825420153917e-05, |
| "loss": 0.7051, |
| "step": 425 |
| }, |
| { |
| "epoch": 30.428571428571427, |
| "grad_norm": 2.6612353324890137, |
| "learning_rate": 2.4048494156391087e-05, |
| "loss": 0.6852, |
| "step": 426 |
| }, |
| { |
| "epoch": 30.5, |
| "grad_norm": 4.154074192047119, |
| "learning_rate": 2.4040127713912264e-05, |
| "loss": 0.6733, |
| "step": 427 |
| }, |
| { |
| "epoch": 30.571428571428573, |
| "grad_norm": 3.900482654571533, |
| "learning_rate": 2.403172611820308e-05, |
| "loss": 0.6134, |
| "step": 428 |
| }, |
| { |
| "epoch": 30.642857142857142, |
| "grad_norm": 5.209915637969971, |
| "learning_rate": 2.4023289394856257e-05, |
| "loss": 0.7324, |
| "step": 429 |
| }, |
| { |
| "epoch": 30.714285714285715, |
| "grad_norm": 2.668140411376953, |
| "learning_rate": 2.401481756957152e-05, |
| "loss": 0.6655, |
| "step": 430 |
| }, |
| { |
| "epoch": 30.785714285714285, |
| "grad_norm": 3.480163812637329, |
| "learning_rate": 2.4006310668155508e-05, |
| "loss": 0.7295, |
| "step": 431 |
| }, |
| { |
| "epoch": 30.857142857142858, |
| "grad_norm": 2.9695725440979004, |
| "learning_rate": 2.3997768716521723e-05, |
| "loss": 0.6949, |
| "step": 432 |
| }, |
| { |
| "epoch": 30.928571428571427, |
| "grad_norm": 2.298661470413208, |
| "learning_rate": 2.398919174069043e-05, |
| "loss": 0.6687, |
| "step": 433 |
| }, |
| { |
| "epoch": 31.0, |
| "grad_norm": 3.2847728729248047, |
| "learning_rate": 2.398057976678859e-05, |
| "loss": 0.6779, |
| "step": 434 |
| }, |
| { |
| "epoch": 31.071428571428573, |
| "grad_norm": 4.464836120605469, |
| "learning_rate": 2.3971932821049765e-05, |
| "loss": 0.7223, |
| "step": 435 |
| }, |
| { |
| "epoch": 31.142857142857142, |
| "grad_norm": 2.811166763305664, |
| "learning_rate": 2.396325092981405e-05, |
| "loss": 0.6677, |
| "step": 436 |
| }, |
| { |
| "epoch": 31.214285714285715, |
| "grad_norm": 3.1231977939605713, |
| "learning_rate": 2.3954534119527996e-05, |
| "loss": 0.6745, |
| "step": 437 |
| }, |
| { |
| "epoch": 31.285714285714285, |
| "grad_norm": 1.7231147289276123, |
| "learning_rate": 2.3945782416744517e-05, |
| "loss": 0.5596, |
| "step": 438 |
| }, |
| { |
| "epoch": 31.357142857142858, |
| "grad_norm": 2.8085505962371826, |
| "learning_rate": 2.3936995848122812e-05, |
| "loss": 0.5903, |
| "step": 439 |
| }, |
| { |
| "epoch": 31.428571428571427, |
| "grad_norm": 5.913967132568359, |
| "learning_rate": 2.3928174440428297e-05, |
| "loss": 0.6269, |
| "step": 440 |
| }, |
| { |
| "epoch": 31.5, |
| "grad_norm": 3.01706862449646, |
| "learning_rate": 2.391931822053251e-05, |
| "loss": 0.6192, |
| "step": 441 |
| }, |
| { |
| "epoch": 31.571428571428573, |
| "grad_norm": 7.216449737548828, |
| "learning_rate": 2.3910427215413036e-05, |
| "loss": 0.6928, |
| "step": 442 |
| }, |
| { |
| "epoch": 31.642857142857142, |
| "grad_norm": 3.636003255844116, |
| "learning_rate": 2.390150145215341e-05, |
| "loss": 0.6932, |
| "step": 443 |
| }, |
| { |
| "epoch": 31.714285714285715, |
| "grad_norm": 2.978515625, |
| "learning_rate": 2.3892540957943067e-05, |
| "loss": 0.6355, |
| "step": 444 |
| }, |
| { |
| "epoch": 31.785714285714285, |
| "grad_norm": 2.5752205848693848, |
| "learning_rate": 2.3883545760077215e-05, |
| "loss": 0.6208, |
| "step": 445 |
| }, |
| { |
| "epoch": 31.857142857142858, |
| "grad_norm": 2.671752452850342, |
| "learning_rate": 2.3874515885956792e-05, |
| "loss": 0.6078, |
| "step": 446 |
| }, |
| { |
| "epoch": 31.928571428571427, |
| "grad_norm": 3.273820400238037, |
| "learning_rate": 2.386545136308836e-05, |
| "loss": 0.6566, |
| "step": 447 |
| }, |
| { |
| "epoch": 32.0, |
| "grad_norm": 3.485921859741211, |
| "learning_rate": 2.3856352219084024e-05, |
| "loss": 0.606, |
| "step": 448 |
| }, |
| { |
| "epoch": 32.07142857142857, |
| "grad_norm": 2.268240451812744, |
| "learning_rate": 2.384721848166136e-05, |
| "loss": 0.5724, |
| "step": 449 |
| }, |
| { |
| "epoch": 32.142857142857146, |
| "grad_norm": 4.0276689529418945, |
| "learning_rate": 2.3838050178643312e-05, |
| "loss": 0.5929, |
| "step": 450 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 2000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 143, |
| "save_steps": 200, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 64, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
