| { | |
| "best_metric": 1.1385369300842285, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-50", | |
| "epoch": 3.004016064257028, | |
| "eval_steps": 50, | |
| "global_step": 187, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01606425702811245, | |
| "grad_norm": 8.358396530151367, | |
| "learning_rate": 1e-05, | |
| "loss": 10.6499, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01606425702811245, | |
| "eval_loss": 2.549973487854004, | |
| "eval_runtime": 5.012, | |
| "eval_samples_per_second": 20.95, | |
| "eval_steps_per_second": 5.387, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0321285140562249, | |
| "grad_norm": 8.561280250549316, | |
| "learning_rate": 2e-05, | |
| "loss": 10.1041, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.04819277108433735, | |
| "grad_norm": 7.631792068481445, | |
| "learning_rate": 3e-05, | |
| "loss": 9.1649, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0642570281124498, | |
| "grad_norm": 8.794489860534668, | |
| "learning_rate": 4e-05, | |
| "loss": 10.3223, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08032128514056225, | |
| "grad_norm": 8.187786102294922, | |
| "learning_rate": 5e-05, | |
| "loss": 8.9209, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0963855421686747, | |
| "grad_norm": 9.42801284790039, | |
| "learning_rate": 6e-05, | |
| "loss": 8.9984, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11244979919678715, | |
| "grad_norm": 8.824488639831543, | |
| "learning_rate": 7e-05, | |
| "loss": 9.4921, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1285140562248996, | |
| "grad_norm": 7.4875640869140625, | |
| "learning_rate": 8e-05, | |
| "loss": 7.2293, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14457831325301204, | |
| "grad_norm": 5.650693416595459, | |
| "learning_rate": 9e-05, | |
| "loss": 6.9636, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1606425702811245, | |
| "grad_norm": 4.513960838317871, | |
| "learning_rate": 0.0001, | |
| "loss": 6.1581, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17670682730923695, | |
| "grad_norm": 4.6525983810424805, | |
| "learning_rate": 9.99921244331919e-05, | |
| "loss": 5.6187, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1927710843373494, | |
| "grad_norm": 4.981886863708496, | |
| "learning_rate": 9.996850021374968e-05, | |
| "loss": 5.8963, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20883534136546184, | |
| "grad_norm": 4.619255065917969, | |
| "learning_rate": 9.99291347838381e-05, | |
| "loss": 6.3916, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2248995983935743, | |
| "grad_norm": 42.78565979003906, | |
| "learning_rate": 9.987404054446008e-05, | |
| "loss": 5.2134, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.24096385542168675, | |
| "grad_norm": 4.6155290603637695, | |
| "learning_rate": 9.980323485155013e-05, | |
| "loss": 3.8579, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2570281124497992, | |
| "grad_norm": 4.810256004333496, | |
| "learning_rate": 9.971674001050686e-05, | |
| "loss": 6.0338, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.27309236947791166, | |
| "grad_norm": 4.097535610198975, | |
| "learning_rate": 9.961458326916624e-05, | |
| "loss": 5.6122, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2891566265060241, | |
| "grad_norm": 4.028859615325928, | |
| "learning_rate": 9.94967968092179e-05, | |
| "loss": 5.251, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.30522088353413657, | |
| "grad_norm": 3.8736844062805176, | |
| "learning_rate": 9.936341773606723e-05, | |
| "loss": 5.9792, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.321285140562249, | |
| "grad_norm": 3.783942937850952, | |
| "learning_rate": 9.921448806714631e-05, | |
| "loss": 5.6756, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3373493975903614, | |
| "grad_norm": 3.129595994949341, | |
| "learning_rate": 9.905005471867739e-05, | |
| "loss": 4.8383, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3534136546184739, | |
| "grad_norm": 3.2784130573272705, | |
| "learning_rate": 9.887016949089333e-05, | |
| "loss": 4.6347, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.36947791164658633, | |
| "grad_norm": 3.4912328720092773, | |
| "learning_rate": 9.867488905171933e-05, | |
| "loss": 4.8285, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3855421686746988, | |
| "grad_norm": 4.122783184051514, | |
| "learning_rate": 9.846427491892118e-05, | |
| "loss": 4.9993, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.40160642570281124, | |
| "grad_norm": 3.454385757446289, | |
| "learning_rate": 9.82383934407258e-05, | |
| "loss": 4.7978, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.41767068273092367, | |
| "grad_norm": 3.439769744873047, | |
| "learning_rate": 9.79973157749201e-05, | |
| "loss": 4.8947, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.43373493975903615, | |
| "grad_norm": 3.4278244972229004, | |
| "learning_rate": 9.77411178664346e-05, | |
| "loss": 4.5172, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4497991967871486, | |
| "grad_norm": 3.5005717277526855, | |
| "learning_rate": 9.746988042341906e-05, | |
| "loss": 4.7726, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.46586345381526106, | |
| "grad_norm": 3.3018925189971924, | |
| "learning_rate": 9.718368889181764e-05, | |
| "loss": 4.3067, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4819277108433735, | |
| "grad_norm": 3.5539093017578125, | |
| "learning_rate": 9.68826334284514e-05, | |
| "loss": 3.7747, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4979919678714859, | |
| "grad_norm": 3.4804933071136475, | |
| "learning_rate": 9.656680887261693e-05, | |
| "loss": 5.1528, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5140562248995983, | |
| "grad_norm": 3.2824409008026123, | |
| "learning_rate": 9.62363147162098e-05, | |
| "loss": 4.9029, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5301204819277109, | |
| "grad_norm": 3.1446473598480225, | |
| "learning_rate": 9.589125507238233e-05, | |
| "loss": 4.8181, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5461847389558233, | |
| "grad_norm": 3.476330041885376, | |
| "learning_rate": 9.553173864274567e-05, | |
| "loss": 5.1613, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5622489959839357, | |
| "grad_norm": 3.2551815509796143, | |
| "learning_rate": 9.515787868312619e-05, | |
| "loss": 4.7212, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5783132530120482, | |
| "grad_norm": 3.1978161334991455, | |
| "learning_rate": 9.476979296788747e-05, | |
| "loss": 4.7767, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5943775100401606, | |
| "grad_norm": 2.728731632232666, | |
| "learning_rate": 9.436760375282859e-05, | |
| "loss": 4.4009, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6104417670682731, | |
| "grad_norm": 2.96553373336792, | |
| "learning_rate": 9.395143773667088e-05, | |
| "loss": 5.6028, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6265060240963856, | |
| "grad_norm": 3.6579477787017822, | |
| "learning_rate": 9.352142602114486e-05, | |
| "loss": 4.9442, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.642570281124498, | |
| "grad_norm": 2.903290271759033, | |
| "learning_rate": 9.30777040696903e-05, | |
| "loss": 4.5713, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6586345381526104, | |
| "grad_norm": 3.4129068851470947, | |
| "learning_rate": 9.262041166478214e-05, | |
| "loss": 4.6581, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6746987951807228, | |
| "grad_norm": 2.9026341438293457, | |
| "learning_rate": 9.214969286389576e-05, | |
| "loss": 3.8248, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6907630522088354, | |
| "grad_norm": 2.8055315017700195, | |
| "learning_rate": 9.166569595412575e-05, | |
| "loss": 3.7736, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7068273092369478, | |
| "grad_norm": 3.1325647830963135, | |
| "learning_rate": 9.116857340547202e-05, | |
| "loss": 4.526, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7228915662650602, | |
| "grad_norm": 3.3222157955169678, | |
| "learning_rate": 9.065848182280833e-05, | |
| "loss": 3.4443, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7389558232931727, | |
| "grad_norm": 9.711248397827148, | |
| "learning_rate": 9.013558189654819e-05, | |
| "loss": 4.5834, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7550200803212851, | |
| "grad_norm": 3.783061981201172, | |
| "learning_rate": 8.96000383520237e-05, | |
| "loss": 4.9376, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7710843373493976, | |
| "grad_norm": 3.666478157043457, | |
| "learning_rate": 8.905201989759341e-05, | |
| "loss": 4.9693, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7871485943775101, | |
| "grad_norm": 2.7563154697418213, | |
| "learning_rate": 8.849169917149531e-05, | |
| "loss": 4.0881, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8032128514056225, | |
| "grad_norm": 3.036672353744507, | |
| "learning_rate": 8.791925268746193e-05, | |
| "loss": 4.4524, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8032128514056225, | |
| "eval_loss": 1.1385369300842285, | |
| "eval_runtime": 5.1628, | |
| "eval_samples_per_second": 20.338, | |
| "eval_steps_per_second": 5.23, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8192771084337349, | |
| "grad_norm": 2.9426279067993164, | |
| "learning_rate": 8.73348607791144e-05, | |
| "loss": 4.6989, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8353413654618473, | |
| "grad_norm": 3.0048654079437256, | |
| "learning_rate": 8.673870754315336e-05, | |
| "loss": 4.2927, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8514056224899599, | |
| "grad_norm": 3.0875372886657715, | |
| "learning_rate": 8.613098078136437e-05, | |
| "loss": 4.8515, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8674698795180723, | |
| "grad_norm": 3.7591512203216553, | |
| "learning_rate": 8.551187194145592e-05, | |
| "loss": 4.032, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8835341365461847, | |
| "grad_norm": 2.896423816680908, | |
| "learning_rate": 8.488157605674925e-05, | |
| "loss": 4.0854, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8995983935742972, | |
| "grad_norm": 2.782850503921509, | |
| "learning_rate": 8.424029168473829e-05, | |
| "loss": 4.6205, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9156626506024096, | |
| "grad_norm": 2.9316649436950684, | |
| "learning_rate": 8.358822084453965e-05, | |
| "loss": 3.9744, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9317269076305221, | |
| "grad_norm": 2.7768986225128174, | |
| "learning_rate": 8.292556895325194e-05, | |
| "loss": 4.2567, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9477911646586346, | |
| "grad_norm": 3.6404924392700195, | |
| "learning_rate": 8.225254476124478e-05, | |
| "loss": 4.2149, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.963855421686747, | |
| "grad_norm": 3.508265972137451, | |
| "learning_rate": 8.156936028639767e-05, | |
| "loss": 3.9642, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9799196787148594, | |
| "grad_norm": 3.133049726486206, | |
| "learning_rate": 8.08762307473096e-05, | |
| "loss": 4.7128, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9959839357429718, | |
| "grad_norm": 3.1079959869384766, | |
| "learning_rate": 8.01733744955002e-05, | |
| "loss": 4.2072, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0120481927710843, | |
| "grad_norm": 5.682511806488037, | |
| "learning_rate": 7.946101294662418e-05, | |
| "loss": 5.271, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0281124497991967, | |
| "grad_norm": 2.7712900638580322, | |
| "learning_rate": 7.873937051072035e-05, | |
| "loss": 4.2853, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0441767068273093, | |
| "grad_norm": 2.630441188812256, | |
| "learning_rate": 7.80086745215173e-05, | |
| "loss": 3.7277, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0602409638554218, | |
| "grad_norm": 2.798489570617676, | |
| "learning_rate": 7.726915516481824e-05, | |
| "loss": 4.2442, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0763052208835342, | |
| "grad_norm": 2.5500569343566895, | |
| "learning_rate": 7.652104540598712e-05, | |
| "loss": 3.378, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0923694779116466, | |
| "grad_norm": 2.5407848358154297, | |
| "learning_rate": 7.57645809165594e-05, | |
| "loss": 4.2423, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.108433734939759, | |
| "grad_norm": 2.9551448822021484, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.8987, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1244979919678715, | |
| "grad_norm": 2.51187801361084, | |
| "learning_rate": 7.422754351663252e-05, | |
| "loss": 3.8333, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.140562248995984, | |
| "grad_norm": 2.5731523036956787, | |
| "learning_rate": 7.344745480776257e-05, | |
| "loss": 3.8188, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1566265060240963, | |
| "grad_norm": 2.586987018585205, | |
| "learning_rate": 7.265997961901987e-05, | |
| "loss": 3.5334, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1726907630522088, | |
| "grad_norm": 2.6580755710601807, | |
| "learning_rate": 7.186536602294278e-05, | |
| "loss": 3.6089, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1887550200803212, | |
| "grad_norm": 2.7322309017181396, | |
| "learning_rate": 7.106386434082979e-05, | |
| "loss": 3.5912, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2048192771084336, | |
| "grad_norm": 2.6883673667907715, | |
| "learning_rate": 7.025572706388268e-05, | |
| "loss": 3.4219, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2208835341365463, | |
| "grad_norm": 3.3504085540771484, | |
| "learning_rate": 6.944120877366604e-05, | |
| "loss": 3.5701, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2369477911646587, | |
| "grad_norm": 2.829219102859497, | |
| "learning_rate": 6.86205660619083e-05, | |
| "loss": 2.6253, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2530120481927711, | |
| "grad_norm": 3.903244733810425, | |
| "learning_rate": 6.779405744966954e-05, | |
| "loss": 3.8675, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.2690763052208835, | |
| "grad_norm": 3.95448637008667, | |
| "learning_rate": 6.696194330590151e-05, | |
| "loss": 4.0529, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.285140562248996, | |
| "grad_norm": 2.8472986221313477, | |
| "learning_rate": 6.612448576542545e-05, | |
| "loss": 3.1193, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3012048192771084, | |
| "grad_norm": 3.7642226219177246, | |
| "learning_rate": 6.528194864635369e-05, | |
| "loss": 3.6889, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3172690763052208, | |
| "grad_norm": 3.163255214691162, | |
| "learning_rate": 6.443459736698105e-05, | |
| "loss": 3.8981, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 3.7267022132873535, | |
| "learning_rate": 6.358269886217194e-05, | |
| "loss": 3.7673, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3493975903614457, | |
| "grad_norm": 3.128511428833008, | |
| "learning_rate": 6.272652149926988e-05, | |
| "loss": 3.1023, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3654618473895583, | |
| "grad_norm": 3.0362024307250977, | |
| "learning_rate": 6.186633499355576e-05, | |
| "loss": 3.2074, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3815261044176708, | |
| "grad_norm": 3.2987911701202393, | |
| "learning_rate": 6.100241032328124e-05, | |
| "loss": 3.4199, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.3975903614457832, | |
| "grad_norm": 3.5199403762817383, | |
| "learning_rate": 6.013501964430468e-05, | |
| "loss": 3.5126, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.4136546184738956, | |
| "grad_norm": 3.5599286556243896, | |
| "learning_rate": 5.9264436204355724e-05, | |
| "loss": 3.6378, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.429718875502008, | |
| "grad_norm": 3.366896152496338, | |
| "learning_rate": 5.839093425695609e-05, | |
| "loss": 3.1565, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4457831325301205, | |
| "grad_norm": 3.9521143436431885, | |
| "learning_rate": 5.751478897502352e-05, | |
| "loss": 3.5207, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.461847389558233, | |
| "grad_norm": 3.8751003742218018, | |
| "learning_rate": 5.6636276364186105e-05, | |
| "loss": 3.0394, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4779116465863453, | |
| "grad_norm": 4.066254615783691, | |
| "learning_rate": 5.5755673175834145e-05, | |
| "loss": 3.2113, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4939759036144578, | |
| "grad_norm": 4.048666954040527, | |
| "learning_rate": 5.487325681993733e-05, | |
| "loss": 3.5889, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.5100401606425704, | |
| "grad_norm": 4.316978931427002, | |
| "learning_rate": 5.3989305277654156e-05, | |
| "loss": 3.9679, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.5261044176706826, | |
| "grad_norm": 3.883434534072876, | |
| "learning_rate": 5.31040970137617e-05, | |
| "loss": 3.3537, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5421686746987953, | |
| "grad_norm": 4.3469672203063965, | |
| "learning_rate": 5.221791088893282e-05, | |
| "loss": 4.7528, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5582329317269075, | |
| "grad_norm": 3.930917739868164, | |
| "learning_rate": 5.133102607188874e-05, | |
| "loss": 3.8965, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5742971887550201, | |
| "grad_norm": 3.873133420944214, | |
| "learning_rate": 5.044372195145455e-05, | |
| "loss": 3.3711, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5903614457831325, | |
| "grad_norm": 3.6565146446228027, | |
| "learning_rate": 4.955627804854545e-05, | |
| "loss": 3.6828, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.606425702811245, | |
| "grad_norm": 3.9164047241210938, | |
| "learning_rate": 4.866897392811126e-05, | |
| "loss": 4.1462, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.606425702811245, | |
| "eval_loss": 1.1504853963851929, | |
| "eval_runtime": 5.1386, | |
| "eval_samples_per_second": 20.434, | |
| "eval_steps_per_second": 5.254, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6224899598393574, | |
| "grad_norm": 4.047707557678223, | |
| "learning_rate": 4.7782089111067176e-05, | |
| "loss": 3.7656, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.6385542168674698, | |
| "grad_norm": 3.6224374771118164, | |
| "learning_rate": 4.6895902986238304e-05, | |
| "loss": 3.7808, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6546184738955825, | |
| "grad_norm": 3.4380745887756348, | |
| "learning_rate": 4.601069472234584e-05, | |
| "loss": 2.9976, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6706827309236947, | |
| "grad_norm": 3.690455436706543, | |
| "learning_rate": 4.512674318006268e-05, | |
| "loss": 2.969, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6867469879518073, | |
| "grad_norm": 3.4021284580230713, | |
| "learning_rate": 4.424432682416585e-05, | |
| "loss": 3.0345, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.7028112449799195, | |
| "grad_norm": 3.821765422821045, | |
| "learning_rate": 4.336372363581391e-05, | |
| "loss": 3.0546, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.7188755020080322, | |
| "grad_norm": 3.7544493675231934, | |
| "learning_rate": 4.2485211024976496e-05, | |
| "loss": 2.6392, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.7349397590361446, | |
| "grad_norm": 3.898599863052368, | |
| "learning_rate": 4.160906574304392e-05, | |
| "loss": 3.675, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.751004016064257, | |
| "grad_norm": 3.801405429840088, | |
| "learning_rate": 4.0735563795644294e-05, | |
| "loss": 3.8661, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7670682730923695, | |
| "grad_norm": 4.313454627990723, | |
| "learning_rate": 3.986498035569532e-05, | |
| "loss": 4.2648, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.783132530120482, | |
| "grad_norm": 3.6857495307922363, | |
| "learning_rate": 3.899758967671878e-05, | |
| "loss": 3.2822, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7991967871485943, | |
| "grad_norm": 3.5725905895233154, | |
| "learning_rate": 3.8133665006444255e-05, | |
| "loss": 3.4181, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.8152610441767068, | |
| "grad_norm": 3.708075761795044, | |
| "learning_rate": 3.727347850073012e-05, | |
| "loss": 3.0178, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8313253012048194, | |
| "grad_norm": 4.1876220703125, | |
| "learning_rate": 3.641730113782807e-05, | |
| "loss": 3.5986, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8473895582329316, | |
| "grad_norm": 3.863704204559326, | |
| "learning_rate": 3.556540263301896e-05, | |
| "loss": 3.5571, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8634538152610443, | |
| "grad_norm": 3.584493398666382, | |
| "learning_rate": 3.47180513536463e-05, | |
| "loss": 3.3766, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8795180722891565, | |
| "grad_norm": 3.944830894470215, | |
| "learning_rate": 3.3875514234574556e-05, | |
| "loss": 3.5299, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.895582329317269, | |
| "grad_norm": 3.59722900390625, | |
| "learning_rate": 3.303805669409848e-05, | |
| "loss": 3.1769, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.9116465863453815, | |
| "grad_norm": 3.9105989933013916, | |
| "learning_rate": 3.2205942550330456e-05, | |
| "loss": 3.3873, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.927710843373494, | |
| "grad_norm": 3.794104814529419, | |
| "learning_rate": 3.1379433938091696e-05, | |
| "loss": 3.3131, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9437751004016064, | |
| "grad_norm": 3.642608165740967, | |
| "learning_rate": 3.055879122633397e-05, | |
| "loss": 2.8884, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9598393574297188, | |
| "grad_norm": 4.019187927246094, | |
| "learning_rate": 2.9744272936117323e-05, | |
| "loss": 2.927, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9759036144578315, | |
| "grad_norm": 4.073461532592773, | |
| "learning_rate": 2.8936135659170216e-05, | |
| "loss": 3.2049, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9919678714859437, | |
| "grad_norm": 3.866515874862671, | |
| "learning_rate": 2.8134633977057235e-05, | |
| "loss": 3.0155, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.0080321285140563, | |
| "grad_norm": 6.784224033355713, | |
| "learning_rate": 2.7340020380980146e-05, | |
| "loss": 2.7591, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0240963855421685, | |
| "grad_norm": 3.676565408706665, | |
| "learning_rate": 2.655254519223746e-05, | |
| "loss": 3.2366, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.040160642570281, | |
| "grad_norm": 3.3340706825256348, | |
| "learning_rate": 2.5772456483367497e-05, | |
| "loss": 2.8017, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.0562248995983934, | |
| "grad_norm": 3.5175349712371826, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.9206, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.072289156626506, | |
| "grad_norm": 3.3527610301971436, | |
| "learning_rate": 2.4235419083440613e-05, | |
| "loss": 2.5938, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.0883534136546187, | |
| "grad_norm": 3.2494187355041504, | |
| "learning_rate": 2.347895459401288e-05, | |
| "loss": 2.2693, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.104417670682731, | |
| "grad_norm": 4.090683460235596, | |
| "learning_rate": 2.2730844835181757e-05, | |
| "loss": 2.9655, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.1204819277108435, | |
| "grad_norm": 3.746830701828003, | |
| "learning_rate": 2.1991325478482694e-05, | |
| "loss": 2.6929, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.1365461847389557, | |
| "grad_norm": 3.2926580905914307, | |
| "learning_rate": 2.126062948927966e-05, | |
| "loss": 2.395, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.1526104417670684, | |
| "grad_norm": 4.051050662994385, | |
| "learning_rate": 2.053898705337583e-05, | |
| "loss": 2.8629, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.1686746987951806, | |
| "grad_norm": 3.782912254333496, | |
| "learning_rate": 1.9826625504499806e-05, | |
| "loss": 2.7832, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.1847389558232932, | |
| "grad_norm": 3.7877721786499023, | |
| "learning_rate": 1.912376925269041e-05, | |
| "loss": 1.9438, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.2008032128514055, | |
| "grad_norm": 3.5854597091674805, | |
| "learning_rate": 1.8430639713602316e-05, | |
| "loss": 1.9395, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.216867469879518, | |
| "grad_norm": 4.118504524230957, | |
| "learning_rate": 1.7747455238755223e-05, | |
| "loss": 2.7666, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.2329317269076308, | |
| "grad_norm": 4.0326924324035645, | |
| "learning_rate": 1.7074431046748075e-05, | |
| "loss": 1.8387, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.248995983935743, | |
| "grad_norm": 4.268743991851807, | |
| "learning_rate": 1.641177915546036e-05, | |
| "loss": 2.5107, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.2650602409638556, | |
| "grad_norm": 4.364184856414795, | |
| "learning_rate": 1.5759708315261722e-05, | |
| "loss": 2.4838, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.281124497991968, | |
| "grad_norm": 4.482365131378174, | |
| "learning_rate": 1.5118423943250771e-05, | |
| "loss": 2.5102, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.2971887550200805, | |
| "grad_norm": 4.34283447265625, | |
| "learning_rate": 1.4488128058544098e-05, | |
| "loss": 2.5356, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.3132530120481927, | |
| "grad_norm": 4.787532329559326, | |
| "learning_rate": 1.3869019218635642e-05, | |
| "loss": 3.0075, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.3293172690763053, | |
| "grad_norm": 4.763615608215332, | |
| "learning_rate": 1.3261292456846647e-05, | |
| "loss": 2.6659, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.3453815261044175, | |
| "grad_norm": 4.660789489746094, | |
| "learning_rate": 1.2665139220885613e-05, | |
| "loss": 2.6328, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.36144578313253, | |
| "grad_norm": 4.864875793457031, | |
| "learning_rate": 1.2080747312538083e-05, | |
| "loss": 2.6357, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.3775100401606424, | |
| "grad_norm": 5.13457727432251, | |
| "learning_rate": 1.150830082850468e-05, | |
| "loss": 2.601, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.393574297188755, | |
| "grad_norm": 4.843163967132568, | |
| "learning_rate": 1.0947980102406596e-05, | |
| "loss": 2.1365, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.4096385542168672, | |
| "grad_norm": 4.978252410888672, | |
| "learning_rate": 1.0399961647976314e-05, | |
| "loss": 2.4167, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4096385542168672, | |
| "eval_loss": 1.3296469449996948, | |
| "eval_runtime": 5.1411, | |
| "eval_samples_per_second": 20.423, | |
| "eval_steps_per_second": 5.252, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.42570281124498, | |
| "grad_norm": 5.148977279663086, | |
| "learning_rate": 9.864418103451828e-06, | |
| "loss": 2.2449, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.4417670682730925, | |
| "grad_norm": 4.7471466064453125, | |
| "learning_rate": 9.34151817719166e-06, | |
| "loss": 1.7333, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.4578313253012047, | |
| "grad_norm": 5.056032180786133, | |
| "learning_rate": 8.831426594527975e-06, | |
| "loss": 1.7489, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.4738955823293174, | |
| "grad_norm": 5.261003017425537, | |
| "learning_rate": 8.334304045874247e-06, | |
| "loss": 1.6044, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.4899598393574296, | |
| "grad_norm": 5.535795211791992, | |
| "learning_rate": 7.850307136104247e-06, | |
| "loss": 2.2367, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.5060240963855422, | |
| "grad_norm": 5.856411933898926, | |
| "learning_rate": 7.379588335217874e-06, | |
| "loss": 3.0989, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.522088353413655, | |
| "grad_norm": 5.084694862365723, | |
| "learning_rate": 6.922295930309691e-06, | |
| "loss": 2.7171, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.538152610441767, | |
| "grad_norm": 6.212282657623291, | |
| "learning_rate": 6.478573978855146e-06, | |
| "loss": 2.8528, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.5542168674698793, | |
| "grad_norm": 4.591296195983887, | |
| "learning_rate": 6.048562263329138e-06, | |
| "loss": 2.0066, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.570281124497992, | |
| "grad_norm": 5.471847057342529, | |
| "learning_rate": 5.6323962471714286e-06, | |
| "loss": 2.6712, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5863453815261046, | |
| "grad_norm": 6.344399929046631, | |
| "learning_rate": 5.23020703211255e-06, | |
| "loss": 2.3738, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.602409638554217, | |
| "grad_norm": 4.858112335205078, | |
| "learning_rate": 4.842121316873821e-06, | |
| "loss": 2.2417, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.6184738955823295, | |
| "grad_norm": 5.603279113769531, | |
| "learning_rate": 4.468261357254339e-06, | |
| "loss": 2.6957, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.6345381526104417, | |
| "grad_norm": 5.545329570770264, | |
| "learning_rate": 4.108744927617669e-06, | |
| "loss": 2.2776, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.6506024096385543, | |
| "grad_norm": 4.7836222648620605, | |
| "learning_rate": 3.763685283790208e-06, | |
| "loss": 2.0821, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 4.492027282714844, | |
| "learning_rate": 3.4331911273830784e-06, | |
| "loss": 2.031, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.682730923694779, | |
| "grad_norm": 6.321730136871338, | |
| "learning_rate": 3.117366571548608e-06, | |
| "loss": 2.5016, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.6987951807228914, | |
| "grad_norm": 8.664602279663086, | |
| "learning_rate": 2.816311108182368e-06, | |
| "loss": 2.1106, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.714859437751004, | |
| "grad_norm": 5.606003284454346, | |
| "learning_rate": 2.530119576580936e-06, | |
| "loss": 2.185, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.7309236947791167, | |
| "grad_norm": 5.336299896240234, | |
| "learning_rate": 2.258882133565404e-06, | |
| "loss": 2.0902, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.746987951807229, | |
| "grad_norm": 5.943612575531006, | |
| "learning_rate": 2.0026842250799038e-06, | |
| "loss": 3.1001, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.7630522088353415, | |
| "grad_norm": 5.235433101654053, | |
| "learning_rate": 1.7616065592742038e-06, | |
| "loss": 2.7069, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.7791164658634537, | |
| "grad_norm": 4.886165142059326, | |
| "learning_rate": 1.5357250810788314e-06, | |
| "loss": 2.3469, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.7951807228915664, | |
| "grad_norm": 5.185632228851318, | |
| "learning_rate": 1.3251109482806666e-06, | |
| "loss": 2.7937, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.8112449799196786, | |
| "grad_norm": 5.091357231140137, | |
| "learning_rate": 1.1298305091066664e-06, | |
| "loss": 2.5354, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.8273092369477912, | |
| "grad_norm": 4.6926774978637695, | |
| "learning_rate": 9.499452813226284e-07, | |
| "loss": 2.2616, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.8433734939759034, | |
| "grad_norm": 4.782034397125244, | |
| "learning_rate": 7.855119328537109e-07, | |
| "loss": 2.0711, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.859437751004016, | |
| "grad_norm": 5.071019172668457, | |
| "learning_rate": 6.365822639327723e-07, | |
| "loss": 2.1635, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.8755020080321287, | |
| "grad_norm": 4.6255974769592285, | |
| "learning_rate": 5.032031907821089e-07, | |
| "loss": 2.1709, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.891566265060241, | |
| "grad_norm": 5.228384494781494, | |
| "learning_rate": 3.854167308337708e-07, | |
| "loss": 2.1617, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.907630522088353, | |
| "grad_norm": 5.127053260803223, | |
| "learning_rate": 2.8325998949314536e-07, | |
| "loss": 2.1208, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.923694779116466, | |
| "grad_norm": 4.4402055740356445, | |
| "learning_rate": 1.9676514844987337e-07, | |
| "loss": 1.9813, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.9397590361445785, | |
| "grad_norm": 6.084360122680664, | |
| "learning_rate": 1.2595945553992573e-07, | |
| "loss": 2.1314, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.9558232931726907, | |
| "grad_norm": 4.875538349151611, | |
| "learning_rate": 7.086521616190279e-08, | |
| "loss": 1.9403, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.9718875502008033, | |
| "grad_norm": 4.680423736572266, | |
| "learning_rate": 3.149978625032191e-08, | |
| "loss": 1.7237, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.9879518072289155, | |
| "grad_norm": 5.582780838012695, | |
| "learning_rate": 7.875566808107637e-09, | |
| "loss": 2.8136, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.004016064257028, | |
| "grad_norm": 5.710657596588135, | |
| "learning_rate": 0.0, | |
| "loss": 1.9331, | |
| "step": 187 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 187, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 2 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4076236595304858e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |