| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984, | |
| "eval_steps": 500, | |
| "global_step": 156, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 4.704057693481445, | |
| "learning_rate": 0.0, | |
| "loss": 3.1525, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 5.1675567626953125, | |
| "learning_rate": 3.125e-06, | |
| "loss": 3.0046, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 4.877377986907959, | |
| "learning_rate": 6.25e-06, | |
| "loss": 3.0847, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 3.2780067920684814, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 2.429, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 5.668214797973633, | |
| "learning_rate": 1.25e-05, | |
| "loss": 3.2744, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 3.768420934677124, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 2.6841, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 4.710444927215576, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 2.9005, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 4.031246662139893, | |
| "learning_rate": 2.1875e-05, | |
| "loss": 2.9428, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 5.702942848205566, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.152, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.755662441253662, | |
| "learning_rate": 2.8125000000000003e-05, | |
| "loss": 2.9298, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 5.924014568328857, | |
| "learning_rate": 3.125e-05, | |
| "loss": 2.9489, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.938117265701294, | |
| "learning_rate": 3.4375e-05, | |
| "loss": 2.7606, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 3.8817155361175537, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 2.7327, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 3.159653425216675, | |
| "learning_rate": 4.0625000000000005e-05, | |
| "loss": 2.3963, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 3.476162910461426, | |
| "learning_rate": 4.375e-05, | |
| "loss": 2.6892, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.5705180168151855, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 2.8817, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 4.219945430755615, | |
| "learning_rate": 5e-05, | |
| "loss": 2.2068, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.645051956176758, | |
| "learning_rate": 4.9993705873562665e-05, | |
| "loss": 2.7778, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 3.949077606201172, | |
| "learning_rate": 4.997482666353287e-05, | |
| "loss": 2.6612, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 3.4730210304260254, | |
| "learning_rate": 4.99433718761614e-05, | |
| "loss": 2.3016, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.7095143795013428, | |
| "learning_rate": 4.989935734988098e-05, | |
| "loss": 2.101, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.7890281677246094, | |
| "learning_rate": 4.984280524733107e-05, | |
| "loss": 2.7039, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 2.9922304153442383, | |
| "learning_rate": 4.977374404419837e-05, | |
| "loss": 2.2395, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.832943916320801, | |
| "learning_rate": 4.9692208514878444e-05, | |
| "loss": 1.8696, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 3.6049652099609375, | |
| "learning_rate": 4.959823971496574e-05, | |
| "loss": 2.3086, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 3.174640655517578, | |
| "learning_rate": 4.9491884960580894e-05, | |
| "loss": 2.3692, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 3.7016489505767822, | |
| "learning_rate": 4.937319780454559e-05, | |
| "loss": 2.0063, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 3.894601583480835, | |
| "learning_rate": 4.9242238009417175e-05, | |
| "loss": 2.7018, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 2.9560084342956543, | |
| "learning_rate": 4.909907151739633e-05, | |
| "loss": 2.1114, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 3.3287124633789062, | |
| "learning_rate": 4.894377041712326e-05, | |
| "loss": 2.2772, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 3.333181142807007, | |
| "learning_rate": 4.877641290737884e-05, | |
| "loss": 2.0911, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 3.5498669147491455, | |
| "learning_rate": 4.8597083257709194e-05, | |
| "loss": 2.4118, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 3.9147493839263916, | |
| "learning_rate": 4.8405871765993433e-05, | |
| "loss": 2.1987, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 3.2478370666503906, | |
| "learning_rate": 4.820287471297598e-05, | |
| "loss": 2.2038, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 3.533186435699463, | |
| "learning_rate": 4.7988194313786275e-05, | |
| "loss": 2.4584, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 3.903890609741211, | |
| "learning_rate": 4.7761938666470403e-05, | |
| "loss": 2.5773, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 3.5339035987854004, | |
| "learning_rate": 4.752422169756048e-05, | |
| "loss": 2.3672, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 3.715303897857666, | |
| "learning_rate": 4.72751631047092e-05, | |
| "loss": 2.6468, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 3.1525583267211914, | |
| "learning_rate": 4.701488829641845e-05, | |
| "loss": 2.0683, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 3.6886160373687744, | |
| "learning_rate": 4.674352832889239e-05, | |
| "loss": 2.0946, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 3.839430570602417, | |
| "learning_rate": 4.6461219840046654e-05, | |
| "loss": 2.1582, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 3.764770269393921, | |
| "learning_rate": 4.6168104980707107e-05, | |
| "loss": 2.2363, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 2.832568407058716, | |
| "learning_rate": 4.586433134303257e-05, | |
| "loss": 1.837, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 3.857421875, | |
| "learning_rate": 4.5550051886197754e-05, | |
| "loss": 2.4129, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 4.4628682136535645, | |
| "learning_rate": 4.522542485937369e-05, | |
| "loss": 2.2797, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 2.5777652263641357, | |
| "learning_rate": 4.489061372204453e-05, | |
| "loss": 2.0681, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 3.1532561779022217, | |
| "learning_rate": 4.454578706170075e-05, | |
| "loss": 1.953, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 2.7607555389404297, | |
| "learning_rate": 4.419111850895028e-05, | |
| "loss": 1.5028, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 2.608351945877075, | |
| "learning_rate": 4.382678665009028e-05, | |
| "loss": 2.0067, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 3.18412184715271, | |
| "learning_rate": 4.345297493718352e-05, | |
| "loss": 2.1055, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 2.8627803325653076, | |
| "learning_rate": 4.306987159568479e-05, | |
| "loss": 1.9178, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 4.6859965324401855, | |
| "learning_rate": 4.267766952966369e-05, | |
| "loss": 2.7699, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 3.795930862426758, | |
| "learning_rate": 4.227656622467162e-05, | |
| "loss": 2.0893, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 3.7162909507751465, | |
| "learning_rate": 4.186676364830186e-05, | |
| "loss": 1.8699, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 2.9527604579925537, | |
| "learning_rate": 4.144846814849282e-05, | |
| "loss": 1.6775, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 3.3146612644195557, | |
| "learning_rate": 4.10218903496256e-05, | |
| "loss": 2.1451, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 3.666790723800659, | |
| "learning_rate": 4.058724504646834e-05, | |
| "loss": 2.5592, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 3.4549150466918945, | |
| "learning_rate": 4.01447510960205e-05, | |
| "loss": 1.9437, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 3.1792612075805664, | |
| "learning_rate": 3.969463130731183e-05, | |
| "loss": 1.9685, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 3.402393341064453, | |
| "learning_rate": 3.92371123292113e-05, | |
| "loss": 2.4101, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 2.981243848800659, | |
| "learning_rate": 3.8772424536302564e-05, | |
| "loss": 1.618, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 3.2454357147216797, | |
| "learning_rate": 3.830080191288342e-05, | |
| "loss": 2.2354, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 2.588693618774414, | |
| "learning_rate": 3.782248193514766e-05, | |
| "loss": 1.9035, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 3.0176501274108887, | |
| "learning_rate": 3.7337705451608674e-05, | |
| "loss": 1.8453, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 2.457637071609497, | |
| "learning_rate": 3.6846716561824965e-05, | |
| "loss": 1.8545, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 3.0648324489593506, | |
| "learning_rate": 3.634976249348867e-05, | |
| "loss": 1.7224, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 3.6240382194519043, | |
| "learning_rate": 3.5847093477938956e-05, | |
| "loss": 2.0639, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 2.4212806224823, | |
| "learning_rate": 3.533896262416302e-05, | |
| "loss": 1.6711, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 2.8671603202819824, | |
| "learning_rate": 3.4825625791348096e-05, | |
| "loss": 1.8226, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 2.922027349472046, | |
| "learning_rate": 3.4307341460048633e-05, | |
| "loss": 1.4442, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 4.1288838386535645, | |
| "learning_rate": 3.378437060203357e-05, | |
| "loss": 2.022, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 3.45540189743042, | |
| "learning_rate": 3.3256976548879184e-05, | |
| "loss": 2.1066, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 2.9696531295776367, | |
| "learning_rate": 3.272542485937369e-05, | |
| "loss": 1.9677, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 2.752713680267334, | |
| "learning_rate": 3.218998318580043e-05, | |
| "loss": 1.537, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.8638994693756104, | |
| "learning_rate": 3.165092113916688e-05, | |
| "loss": 1.7642, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 2.7981038093566895, | |
| "learning_rate": 3.110851015344735e-05, | |
| "loss": 1.7863, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 3.0902326107025146, | |
| "learning_rate": 3.056302334890786e-05, | |
| "loss": 2.0919, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 2.8343629837036133, | |
| "learning_rate": 3.0014735394581823e-05, | |
| "loss": 2.2425, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 2.7442476749420166, | |
| "learning_rate": 2.9463922369965917e-05, | |
| "loss": 1.5119, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 3.2816710472106934, | |
| "learning_rate": 2.8910861626005776e-05, | |
| "loss": 2.4317, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 2.9412364959716797, | |
| "learning_rate": 2.8355831645441388e-05, | |
| "loss": 1.7199, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 2.4085612297058105, | |
| "learning_rate": 2.7799111902582696e-05, | |
| "loss": 1.6436, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 3.3494606018066406, | |
| "learning_rate": 2.724098272258584e-05, | |
| "loss": 1.9426, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 3.2767434120178223, | |
| "learning_rate": 2.6681725140300997e-05, | |
| "loss": 1.7712, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 3.9682724475860596, | |
| "learning_rate": 2.6121620758762877e-05, | |
| "loss": 2.0355, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 2.4859561920166016, | |
| "learning_rate": 2.556095160739513e-05, | |
| "loss": 1.7458, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 3.25398588180542, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.7167, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 2.7223641872406006, | |
| "learning_rate": 2.443904839260488e-05, | |
| "loss": 1.8083, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 3.3979554176330566, | |
| "learning_rate": 2.3878379241237136e-05, | |
| "loss": 1.8491, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 2.537713050842285, | |
| "learning_rate": 2.331827485969901e-05, | |
| "loss": 1.4982, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 2.481809616088867, | |
| "learning_rate": 2.2759017277414166e-05, | |
| "loss": 1.6553, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 2.704338312149048, | |
| "learning_rate": 2.2200888097417307e-05, | |
| "loss": 1.3204, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 3.5501091480255127, | |
| "learning_rate": 2.164416835455862e-05, | |
| "loss": 2.2412, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 2.8751282691955566, | |
| "learning_rate": 2.1089138373994223e-05, | |
| "loss": 1.6818, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 3.134270191192627, | |
| "learning_rate": 2.0536077630034086e-05, | |
| "loss": 1.7345, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 3.4039204120635986, | |
| "learning_rate": 1.9985264605418183e-05, | |
| "loss": 1.8876, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 2.858107566833496, | |
| "learning_rate": 1.9436976651092144e-05, | |
| "loss": 1.3758, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 3.3995652198791504, | |
| "learning_rate": 1.8891489846552646e-05, | |
| "loss": 1.7547, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 3.344923257827759, | |
| "learning_rate": 1.8349078860833123e-05, | |
| "loss": 1.6181, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.736323118209839, | |
| "learning_rate": 1.781001681419957e-05, | |
| "loss": 1.3949, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 3.303896427154541, | |
| "learning_rate": 1.7274575140626318e-05, | |
| "loss": 1.6077, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 2.998396158218384, | |
| "learning_rate": 1.6743023451120832e-05, | |
| "loss": 1.613, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 4.653803825378418, | |
| "learning_rate": 1.621562939796643e-05, | |
| "loss": 1.9022, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 2.9627840518951416, | |
| "learning_rate": 1.5692658539951372e-05, | |
| "loss": 1.6514, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 4.12007474899292, | |
| "learning_rate": 1.5174374208651912e-05, | |
| "loss": 2.0834, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 3.2830817699432373, | |
| "learning_rate": 1.466103737583699e-05, | |
| "loss": 1.7247, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 2.853792667388916, | |
| "learning_rate": 1.4152906522061048e-05, | |
| "loss": 1.6993, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 2.8175644874572754, | |
| "learning_rate": 1.3650237506511331e-05, | |
| "loss": 1.5805, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 2.7665114402770996, | |
| "learning_rate": 1.3153283438175034e-05, | |
| "loss": 1.4819, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 2.951678514480591, | |
| "learning_rate": 1.2662294548391328e-05, | |
| "loss": 1.4978, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 3.3238184452056885, | |
| "learning_rate": 1.217751806485235e-05, | |
| "loss": 1.8354, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 2.4784748554229736, | |
| "learning_rate": 1.1699198087116589e-05, | |
| "loss": 1.3185, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 3.491586923599243, | |
| "learning_rate": 1.122757546369744e-05, | |
| "loss": 2.1179, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 3.4382336139678955, | |
| "learning_rate": 1.0762887670788702e-05, | |
| "loss": 1.7461, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 3.145657777786255, | |
| "learning_rate": 1.0305368692688174e-05, | |
| "loss": 1.5797, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 3.600944757461548, | |
| "learning_rate": 9.855248903979506e-06, | |
| "loss": 1.5077, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 3.1408603191375732, | |
| "learning_rate": 9.412754953531663e-06, | |
| "loss": 1.6516, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 3.197890043258667, | |
| "learning_rate": 8.978109650374397e-06, | |
| "loss": 1.743, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 2.4245102405548096, | |
| "learning_rate": 8.551531851507186e-06, | |
| "loss": 1.443, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 2.798938512802124, | |
| "learning_rate": 8.133236351698143e-06, | |
| "loss": 1.3257, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 3.1017847061157227, | |
| "learning_rate": 7.723433775328384e-06, | |
| "loss": 1.6557, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 2.4948604106903076, | |
| "learning_rate": 7.3223304703363135e-06, | |
| "loss": 1.6317, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 2.6814234256744385, | |
| "learning_rate": 6.930128404315214e-06, | |
| "loss": 1.6922, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 3.2181453704833984, | |
| "learning_rate": 6.547025062816486e-06, | |
| "loss": 1.6152, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.9685909748077393, | |
| "learning_rate": 6.173213349909729e-06, | |
| "loss": 1.6251, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 3.1302011013031006, | |
| "learning_rate": 5.808881491049723e-06, | |
| "loss": 1.821, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 3.1760427951812744, | |
| "learning_rate": 5.454212938299255e-06, | |
| "loss": 1.9791, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 2.8934803009033203, | |
| "learning_rate": 5.1093862779554776e-06, | |
| "loss": 1.356, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 2.6532275676727295, | |
| "learning_rate": 4.7745751406263165e-06, | |
| "loss": 1.3942, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 2.6779580116271973, | |
| "learning_rate": 4.4499481138022544e-06, | |
| "loss": 1.2577, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 2.9538562297821045, | |
| "learning_rate": 4.135668656967434e-06, | |
| "loss": 1.5183, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 2.1692054271698, | |
| "learning_rate": 3.831895019292897e-06, | |
| "loss": 1.0417, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 3.9423463344573975, | |
| "learning_rate": 3.5387801599533475e-06, | |
| "loss": 2.0205, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 3.072523593902588, | |
| "learning_rate": 3.2564716711076167e-06, | |
| "loss": 1.2959, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 2.2884180545806885, | |
| "learning_rate": 2.98511170358155e-06, | |
| "loss": 1.2522, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 3.3118743896484375, | |
| "learning_rate": 2.7248368952908053e-06, | |
| "loss": 1.6146, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 3.355278968811035, | |
| "learning_rate": 2.475778302439524e-06, | |
| "loss": 1.6541, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 2.1355369091033936, | |
| "learning_rate": 2.2380613335296036e-06, | |
| "loss": 1.0463, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 2.4513282775878906, | |
| "learning_rate": 2.0118056862137357e-06, | |
| "loss": 1.1024, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 3.1772351264953613, | |
| "learning_rate": 1.7971252870240291e-06, | |
| "loss": 1.2631, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 2.525724411010742, | |
| "learning_rate": 1.59412823400657e-06, | |
| "loss": 1.2131, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 4.4134416580200195, | |
| "learning_rate": 1.4029167422908107e-06, | |
| "loss": 1.4394, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 2.139984607696533, | |
| "learning_rate": 1.2235870926211619e-06, | |
| "loss": 1.1603, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 2.7187249660491943, | |
| "learning_rate": 1.0562295828767387e-06, | |
| "loss": 1.283, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 4.029594898223877, | |
| "learning_rate": 9.009284826036691e-07, | |
| "loss": 1.855, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 2.671586036682129, | |
| "learning_rate": 7.577619905828282e-07, | |
| "loss": 1.3775, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 3.0629734992980957, | |
| "learning_rate": 6.268021954544096e-07, | |
| "loss": 1.4093, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 2.15995454788208, | |
| "learning_rate": 5.08115039419113e-07, | |
| "loss": 1.1758, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 2.610565423965454, | |
| "learning_rate": 4.0176028503425835e-07, | |
| "loss": 1.6003, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 3.3255887031555176, | |
| "learning_rate": 3.077914851215585e-07, | |
| "loss": 1.3682, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 3.8929100036621094, | |
| "learning_rate": 2.262559558016325e-07, | |
| "loss": 1.7135, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 2.5593698024749756, | |
| "learning_rate": 1.571947526689349e-07, | |
| "loss": 1.497, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 2.687330961227417, | |
| "learning_rate": 1.006426501190233e-07, | |
| "loss": 1.3305, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 2.625413656234741, | |
| "learning_rate": 5.662812383859795e-08, | |
| "loss": 1.3444, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 2.951202154159546, | |
| "learning_rate": 2.5173336467135267e-08, | |
| "loss": 1.3641, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 2.8701841831207275, | |
| "learning_rate": 6.294126437336734e-09, | |
| "loss": 1.561, | |
| "step": 156 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 156, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3416669053378560.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |