| { | |
| "best_metric": 3.4029853343963623, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 2.962962962962963, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014814814814814815, | |
| "grad_norm": 68.03536224365234, | |
| "learning_rate": 1e-05, | |
| "loss": 15.6754, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.014814814814814815, | |
| "eval_loss": 4.450826644897461, | |
| "eval_runtime": 1.3264, | |
| "eval_samples_per_second": 85.944, | |
| "eval_steps_per_second": 21.863, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02962962962962963, | |
| "grad_norm": 62.42438507080078, | |
| "learning_rate": 2e-05, | |
| "loss": 16.7019, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 63.97747039794922, | |
| "learning_rate": 3e-05, | |
| "loss": 16.8566, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05925925925925926, | |
| "grad_norm": 60.66354751586914, | |
| "learning_rate": 4e-05, | |
| "loss": 16.9645, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 78.175537109375, | |
| "learning_rate": 5e-05, | |
| "loss": 17.1199, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 67.00360870361328, | |
| "learning_rate": 6e-05, | |
| "loss": 17.0548, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1037037037037037, | |
| "grad_norm": 66.37737274169922, | |
| "learning_rate": 7e-05, | |
| "loss": 16.9028, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 74.29869079589844, | |
| "learning_rate": 8e-05, | |
| "loss": 16.7044, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 66.08704376220703, | |
| "learning_rate": 9e-05, | |
| "loss": 17.2147, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 68.14957427978516, | |
| "learning_rate": 0.0001, | |
| "loss": 16.74, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16296296296296298, | |
| "grad_norm": 79.75003051757812, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 17.1835, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 84.72122192382812, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 17.0103, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1925925925925926, | |
| "grad_norm": 92.36153411865234, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 16.8602, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2074074074074074, | |
| "grad_norm": 101.77613830566406, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 17.5061, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 141.8746337890625, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 17.715, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 153.5200958251953, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 18.9296, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2518518518518518, | |
| "grad_norm": 95.71971130371094, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 15.4648, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 113.3729476928711, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 16.0014, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2814814814814815, | |
| "grad_norm": 154.21725463867188, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 15.9663, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 119.4923324584961, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 15.9603, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 103.814697265625, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 15.116, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.32592592592592595, | |
| "grad_norm": 111.0808334350586, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 15.9146, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.34074074074074073, | |
| "grad_norm": 94.6805191040039, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 15.0992, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 179.71383666992188, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 15.6517, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 122.9061050415039, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 15.9564, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3851851851851852, | |
| "grad_norm": 192.7732696533203, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 15.6552, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 135.9822998046875, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 16.2295, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4148148148148148, | |
| "grad_norm": 231.39137268066406, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 16.7316, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.42962962962962964, | |
| "grad_norm": 128.73583984375, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 16.5944, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 158.39361572265625, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 17.0088, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.45925925925925926, | |
| "grad_norm": 175.06552124023438, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 17.0239, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 205.9481964111328, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 18.026, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 115.01461029052734, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 14.85, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5037037037037037, | |
| "grad_norm": 110.00332641601562, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 15.3822, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 147.27503967285156, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 14.9821, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 116.0947494506836, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 15.5853, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5481481481481482, | |
| "grad_norm": 131.1149139404297, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 15.1955, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.562962962962963, | |
| "grad_norm": 148.51016235351562, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 15.3535, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 131.2111358642578, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 15.6877, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 118.9223403930664, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 15.2199, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6074074074074074, | |
| "grad_norm": 122.85724639892578, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 15.2264, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 120.46707916259766, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 15.5699, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6370370370370371, | |
| "grad_norm": 122.61256408691406, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 15.3393, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6518518518518519, | |
| "grad_norm": 163.58267211914062, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 15.6592, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 115.89888000488281, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 15.2963, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6814814814814815, | |
| "grad_norm": 103.28407287597656, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 14.984, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6962962962962963, | |
| "grad_norm": 111.25080108642578, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 16.1359, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 137.91490173339844, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 16.4392, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.725925925925926, | |
| "grad_norm": 76.70240783691406, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 14.1255, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 81.87256622314453, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 14.8406, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "eval_loss": 3.808467388153076, | |
| "eval_runtime": 1.3212, | |
| "eval_samples_per_second": 86.284, | |
| "eval_steps_per_second": 21.949, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 93.05923461914062, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 15.1534, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.7703703703703704, | |
| "grad_norm": 81.33842468261719, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 14.8638, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.7851851851851852, | |
| "grad_norm": 77.30476379394531, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 15.2695, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 74.18260955810547, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 14.8741, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 71.07466125488281, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 14.8219, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 72.3966293334961, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 14.7146, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 68.81806945800781, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 15.0473, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.8592592592592593, | |
| "grad_norm": 80.04654693603516, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 15.2216, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.8740740740740741, | |
| "grad_norm": 70.65907287597656, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 15.2407, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 92.8138198852539, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 14.8447, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9037037037037037, | |
| "grad_norm": 87.86979675292969, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 14.5299, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9185185185185185, | |
| "grad_norm": 62.56691360473633, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 15.0672, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 73.16510772705078, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 14.3267, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 65.85083770751953, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 14.5356, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 52.39100646972656, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 15.1857, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 47.188316345214844, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 14.8252, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.9925925925925926, | |
| "grad_norm": 46.87759780883789, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 14.9699, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0074074074074073, | |
| "grad_norm": 71.50855255126953, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 15.1472, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0222222222222221, | |
| "grad_norm": 40.31723403930664, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 14.2369, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 41.28828048706055, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 14.1751, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0518518518518518, | |
| "grad_norm": 36.52976608276367, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 14.2329, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 39.3419075012207, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 14.2825, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.0814814814814815, | |
| "grad_norm": 42.4677848815918, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 13.5302, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.0962962962962963, | |
| "grad_norm": 38.49083709716797, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 14.238, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 47.739566802978516, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 13.7855, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.125925925925926, | |
| "grad_norm": 43.2079963684082, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 14.6091, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1407407407407408, | |
| "grad_norm": 38.224483489990234, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 14.0073, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 41.32490539550781, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 14.2531, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.1703703703703703, | |
| "grad_norm": 42.44271469116211, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 13.957, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 48.5529899597168, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 13.9293, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 43.183502197265625, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 14.1451, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2148148148148148, | |
| "grad_norm": 60.8570442199707, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 14.62, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2296296296296296, | |
| "grad_norm": 48.83097839355469, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 14.5866, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 49.138729095458984, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 13.7515, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.2592592592592593, | |
| "grad_norm": 48.98204040527344, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 13.6181, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.2740740740740741, | |
| "grad_norm": 43.89327621459961, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 14.1007, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.2888888888888888, | |
| "grad_norm": 48.41681671142578, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 14.2193, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 46.7381477355957, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 14.5718, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3185185185185184, | |
| "grad_norm": 54.48136901855469, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 14.1603, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 50.096588134765625, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 14.2687, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.348148148148148, | |
| "grad_norm": 50.09853744506836, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 13.8544, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.362962962962963, | |
| "grad_norm": 49.0280876159668, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 13.4479, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.3777777777777778, | |
| "grad_norm": 46.293983459472656, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 14.1747, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.3925925925925926, | |
| "grad_norm": 44.48542785644531, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 13.9886, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4074074074074074, | |
| "grad_norm": 45.454872131347656, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 14.6054, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 44.1904296875, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 14.0452, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.4370370370370371, | |
| "grad_norm": 49.55455780029297, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 13.8419, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.4518518518518517, | |
| "grad_norm": 49.50934982299805, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 14.6953, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.4666666666666668, | |
| "grad_norm": 53.63299560546875, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 15.0987, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "grad_norm": 53.22015380859375, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 14.2955, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "eval_loss": 3.5386621952056885, | |
| "eval_runtime": 1.3166, | |
| "eval_samples_per_second": 86.585, | |
| "eval_steps_per_second": 22.026, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4962962962962962, | |
| "grad_norm": 44.37554931640625, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 14.2234, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 58.02170944213867, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 13.661, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.525925925925926, | |
| "grad_norm": 38.92431640625, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 13.5878, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 39.48297119140625, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 13.3939, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 42.054927825927734, | |
| "learning_rate": 5e-05, | |
| "loss": 13.3387, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.5703703703703704, | |
| "grad_norm": 47.19960021972656, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 13.5884, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.585185185185185, | |
| "grad_norm": 57.89037322998047, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 13.3828, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 43.95449447631836, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 13.6622, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.6148148148148147, | |
| "grad_norm": 41.22948455810547, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 13.7168, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.6296296296296298, | |
| "grad_norm": 43.31999588012695, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 13.8183, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.6444444444444444, | |
| "grad_norm": 39.958412170410156, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 14.0399, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 43.44700241088867, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 14.0554, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.674074074074074, | |
| "grad_norm": 45.25251007080078, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 13.2956, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 38.81556701660156, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 13.3152, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.7037037037037037, | |
| "grad_norm": 46.15458297729492, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 12.9937, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.7185185185185186, | |
| "grad_norm": 41.13903045654297, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 13.2159, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.7333333333333334, | |
| "grad_norm": 38.98662567138672, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 13.5577, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.748148148148148, | |
| "grad_norm": 38.35002899169922, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 14.135, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.762962962962963, | |
| "grad_norm": 43.60272216796875, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 13.452, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 36.646766662597656, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 14.3169, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.7925925925925927, | |
| "grad_norm": 36.85139083862305, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 13.3019, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.8074074074074074, | |
| "grad_norm": 38.35990524291992, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 14.3625, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.8222222222222222, | |
| "grad_norm": 35.13533401489258, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 13.4863, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.837037037037037, | |
| "grad_norm": 46.36710739135742, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 13.094, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 37.9239387512207, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 13.9984, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 36.332122802734375, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 13.5551, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.8814814814814815, | |
| "grad_norm": 36.7701416015625, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 13.9656, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 37.59199523925781, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 12.6013, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.911111111111111, | |
| "grad_norm": 39.84718704223633, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 13.8366, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.925925925925926, | |
| "grad_norm": 41.96089553833008, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 13.0529, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.9407407407407407, | |
| "grad_norm": 43.02454376220703, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 13.1329, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 38.45022201538086, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 13.6878, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.9703703703703703, | |
| "grad_norm": 41.045597076416016, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 13.1349, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.9851851851851852, | |
| "grad_norm": 35.30647659301758, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 13.4581, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 75.02015686035156, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 12.7503, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 39.12497329711914, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 13.0417, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.0296296296296297, | |
| "grad_norm": 35.86410140991211, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 13.306, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.0444444444444443, | |
| "grad_norm": 33.3465576171875, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 13.4527, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.0592592592592593, | |
| "grad_norm": 34.85940933227539, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 13.6003, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.074074074074074, | |
| "grad_norm": 40.27370071411133, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 13.3528, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.088888888888889, | |
| "grad_norm": 34.71219253540039, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 13.0197, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.1037037037037036, | |
| "grad_norm": 34.127620697021484, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 12.9773, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.1185185185185187, | |
| "grad_norm": 37.75566101074219, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 13.1995, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 32.003971099853516, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 13.1908, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.148148148148148, | |
| "grad_norm": 41.12171936035156, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 13.5615, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.162962962962963, | |
| "grad_norm": 36.161502838134766, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 12.8052, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.1777777777777776, | |
| "grad_norm": 34.585540771484375, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 13.0802, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.1925925925925926, | |
| "grad_norm": 42.922637939453125, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 13.7596, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.2074074074074073, | |
| "grad_norm": 39.281063079833984, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 13.3407, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 37.62614059448242, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 13.8796, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "eval_loss": 3.4405109882354736, | |
| "eval_runtime": 1.3208, | |
| "eval_samples_per_second": 86.309, | |
| "eval_steps_per_second": 21.956, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.237037037037037, | |
| "grad_norm": 44.44696807861328, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 13.9692, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 36.68385314941406, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 13.0204, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.2666666666666666, | |
| "grad_norm": 36.87557601928711, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 13.3115, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.2814814814814817, | |
| "grad_norm": 32.70282745361328, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 12.7057, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.2962962962962963, | |
| "grad_norm": 35.604366302490234, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 13.6825, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.311111111111111, | |
| "grad_norm": 31.27665138244629, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 13.0649, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.325925925925926, | |
| "grad_norm": 33.390132904052734, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 13.3277, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.3407407407407406, | |
| "grad_norm": 35.84580993652344, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 13.3949, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.3555555555555556, | |
| "grad_norm": 35.63726043701172, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 13.4783, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 34.74530792236328, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 12.5656, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.3851851851851853, | |
| "grad_norm": 36.04372787475586, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 12.7653, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 35.15157699584961, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 12.7899, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.414814814814815, | |
| "grad_norm": 31.54987335205078, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 13.592, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.4296296296296296, | |
| "grad_norm": 37.31626510620117, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 12.9408, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 39.413875579833984, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 13.6008, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.4592592592592593, | |
| "grad_norm": 41.51494598388672, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 13.8659, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.474074074074074, | |
| "grad_norm": 48.0342903137207, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 12.7716, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 32.830902099609375, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 12.7054, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.5037037037037035, | |
| "grad_norm": 31.159988403320312, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 13.135, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.5185185185185186, | |
| "grad_norm": 31.03050994873047, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 12.8175, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.533333333333333, | |
| "grad_norm": 32.80025863647461, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 13.3156, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.5481481481481483, | |
| "grad_norm": 32.93952178955078, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 12.9351, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.562962962962963, | |
| "grad_norm": 34.07162094116211, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 13.1636, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.5777777777777775, | |
| "grad_norm": 33.17424392700195, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 13.1144, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.5925925925925926, | |
| "grad_norm": 30.324411392211914, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 12.8677, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 33.38523483276367, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 13.1734, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.6222222222222222, | |
| "grad_norm": 36.65361022949219, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 12.7936, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.637037037037037, | |
| "grad_norm": 33.760528564453125, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 14.0332, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.651851851851852, | |
| "grad_norm": 33.93669128417969, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 12.844, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 36.08841323852539, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 12.6121, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.6814814814814816, | |
| "grad_norm": 37.48418426513672, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 13.1848, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.696296296296296, | |
| "grad_norm": 41.217803955078125, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 12.4734, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.7111111111111112, | |
| "grad_norm": 59.459205627441406, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 12.4785, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 29.268831253051758, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 12.4509, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.7407407407407405, | |
| "grad_norm": 32.732177734375, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 13.2912, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.7555555555555555, | |
| "grad_norm": 29.52249526977539, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 13.3678, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.7703703703703706, | |
| "grad_norm": 40.58732986450195, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 12.8701, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.785185185185185, | |
| "grad_norm": 30.74588966369629, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 12.9349, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 32.116024017333984, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 12.653, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.814814814814815, | |
| "grad_norm": 38.772377014160156, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 13.2693, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.8296296296296295, | |
| "grad_norm": 34.81737518310547, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 12.5551, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 34.92587661743164, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 13.1372, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.859259259259259, | |
| "grad_norm": 33.52151870727539, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 13.1456, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.8740740740740742, | |
| "grad_norm": 34.61021423339844, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 13.3978, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 36.74519348144531, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 13.3115, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.9037037037037035, | |
| "grad_norm": 38.88361358642578, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 12.649, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.9185185185185185, | |
| "grad_norm": 46.594017028808594, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 12.3873, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 40.79484176635742, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 13.7795, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.948148148148148, | |
| "grad_norm": 45.47343063354492, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 12.6251, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 35.297115325927734, | |
| "learning_rate": 0.0, | |
| "loss": 13.283, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "eval_loss": 3.4029853343963623, | |
| "eval_runtime": 1.3181, | |
| "eval_samples_per_second": 86.485, | |
| "eval_steps_per_second": 22.001, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4874493864378368e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |