| { | |
| "best_metric": 0.44719642400741577, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.29112081513828236, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0019408054342552159, | |
| "grad_norm": 1.7362804412841797, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5136, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0019408054342552159, | |
| "eval_loss": 1.874531865119934, | |
| "eval_runtime": 81.4006, | |
| "eval_samples_per_second": 10.663, | |
| "eval_steps_per_second": 2.666, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0038816108685104317, | |
| "grad_norm": 1.5140223503112793, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4853, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005822416302765648, | |
| "grad_norm": 1.4223251342773438, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4313, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0077632217370208634, | |
| "grad_norm": 1.4768247604370117, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4361, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009704027171276079, | |
| "grad_norm": 1.4820680618286133, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4635, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.011644832605531296, | |
| "grad_norm": 1.6700338125228882, | |
| "learning_rate": 6e-05, | |
| "loss": 0.4921, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.013585638039786511, | |
| "grad_norm": 1.621264934539795, | |
| "learning_rate": 7e-05, | |
| "loss": 0.4881, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.015526443474041727, | |
| "grad_norm": 2.4338958263397217, | |
| "learning_rate": 8e-05, | |
| "loss": 0.543, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.017467248908296942, | |
| "grad_norm": 4.253851413726807, | |
| "learning_rate": 9e-05, | |
| "loss": 0.5981, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.019408054342552158, | |
| "grad_norm": 1.9289650917053223, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5067, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.021348859776807377, | |
| "grad_norm": 2.416295051574707, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 0.6093, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.023289665211062592, | |
| "grad_norm": 2.7022767066955566, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 0.6354, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.025230470645317808, | |
| "grad_norm": 2.856457471847534, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 0.7435, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.027171276079573023, | |
| "grad_norm": 3.1738202571868896, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 0.515, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02911208151382824, | |
| "grad_norm": 4.503840923309326, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.674, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.031052886948083454, | |
| "grad_norm": 6.186589241027832, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.7711, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03299369238233867, | |
| "grad_norm": 3.400998115539551, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.752, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.034934497816593885, | |
| "grad_norm": 3.9549994468688965, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 0.7825, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.036875303250849104, | |
| "grad_norm": 3.5601806640625, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.7426, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.038816108685104316, | |
| "grad_norm": 4.084705352783203, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.7438, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.040756914119359534, | |
| "grad_norm": 4.018491744995117, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 0.6672, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04269771955361475, | |
| "grad_norm": 4.947319984436035, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 0.7293, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.044638524987869965, | |
| "grad_norm": 4.075100421905518, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 0.7317, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.046579330422125184, | |
| "grad_norm": 5.203902244567871, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 0.7481, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.048520135856380396, | |
| "grad_norm": 5.136916637420654, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.7474, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.050460941290635615, | |
| "grad_norm": 5.637122631072998, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 0.8763, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05240174672489083, | |
| "grad_norm": 7.640203952789307, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.7664, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.054342552159146046, | |
| "grad_norm": 5.054821968078613, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.7598, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.056283357593401265, | |
| "grad_norm": 5.661840915679932, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.7727, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05822416302765648, | |
| "grad_norm": 5.301083087921143, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.6411, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.060164968461911696, | |
| "grad_norm": 3.903432607650757, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 0.6496, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.06210577389616691, | |
| "grad_norm": 4.859428405761719, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.6607, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06404657933042213, | |
| "grad_norm": 5.144859790802002, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 0.7444, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06598738476467735, | |
| "grad_norm": 50.76873016357422, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 3.0572, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06792819019893255, | |
| "grad_norm": 55.62409591674805, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 2.7413, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06986899563318777, | |
| "grad_norm": 61.083866119384766, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 2.2786, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.07180980106744299, | |
| "grad_norm": 169.48739624023438, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 1.7695, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07375060650169821, | |
| "grad_norm": 49.281402587890625, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 1.1155, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07569141193595343, | |
| "grad_norm": 38.76823425292969, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.7584, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07763221737020863, | |
| "grad_norm": 32.327022552490234, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.5989, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07957302280446385, | |
| "grad_norm": 27.781795501708984, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.4766, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.08151382823871907, | |
| "grad_norm": 17.971399307250977, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.2487, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.08345463367297429, | |
| "grad_norm": 25.678585052490234, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.2989, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0853954391072295, | |
| "grad_norm": 16.72321891784668, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.1528, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08733624454148471, | |
| "grad_norm": 7.9136176109313965, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.0656, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08927704997573993, | |
| "grad_norm": 9.157156944274902, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.1174, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.09121785540999515, | |
| "grad_norm": 8.771122932434082, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.0721, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.09315866084425037, | |
| "grad_norm": 7.430817604064941, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.038, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.09509946627850557, | |
| "grad_norm": 6.458409309387207, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.0381, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09704027171276079, | |
| "grad_norm": 14.757686614990234, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.1181, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09704027171276079, | |
| "eval_loss": 2.9756832122802734, | |
| "eval_runtime": 82.6227, | |
| "eval_samples_per_second": 10.506, | |
| "eval_steps_per_second": 2.626, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09898107714701601, | |
| "grad_norm": 71.17930603027344, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 5.2893, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.10092188258127123, | |
| "grad_norm": 52.50294876098633, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 3.4814, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.10286268801552645, | |
| "grad_norm": 26.074068069458008, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 1.5954, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.10480349344978165, | |
| "grad_norm": 7.427390098571777, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 0.8787, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10674429888403687, | |
| "grad_norm": 3.58815860748291, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.6604, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10868510431829209, | |
| "grad_norm": 2.99554443359375, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 0.5315, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.11062590975254731, | |
| "grad_norm": 2.7059144973754883, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 0.5737, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.11256671518680253, | |
| "grad_norm": 1.8956794738769531, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 0.5386, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.11450752062105773, | |
| "grad_norm": 2.838517189025879, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 0.5438, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11644832605531295, | |
| "grad_norm": 2.675584316253662, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.6574, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11838913148956817, | |
| "grad_norm": 2.1755335330963135, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.5774, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.12032993692382339, | |
| "grad_norm": 2.3560686111450195, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.5783, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1222707423580786, | |
| "grad_norm": 1.8677828311920166, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.5582, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.12421154779233382, | |
| "grad_norm": 2.1840298175811768, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.645, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.12615235322658905, | |
| "grad_norm": 2.2143750190734863, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.5533, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.12809315866084425, | |
| "grad_norm": 2.0592188835144043, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.6437, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.13003396409509946, | |
| "grad_norm": 2.2719454765319824, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.6332, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1319747695293547, | |
| "grad_norm": 2.2744674682617188, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.5922, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1339155749636099, | |
| "grad_norm": 2.8700437545776367, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 0.6904, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1358563803978651, | |
| "grad_norm": 2.6853842735290527, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.6535, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13779718583212033, | |
| "grad_norm": 3.1052772998809814, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.7032, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13973799126637554, | |
| "grad_norm": 2.2774715423583984, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.7064, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.14167879670063077, | |
| "grad_norm": 3.7868194580078125, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 0.6476, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.14361960213488598, | |
| "grad_norm": 2.482116460800171, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.577, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.14556040756914118, | |
| "grad_norm": 3.1218602657318115, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.8259, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14750121300339641, | |
| "grad_norm": 4.078639030456543, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.8581, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.14944201843765162, | |
| "grad_norm": 4.382765769958496, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.8175, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.15138282387190685, | |
| "grad_norm": 3.7048816680908203, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.808, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.15332362930616206, | |
| "grad_norm": 4.033629417419434, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.7593, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.15526443474041726, | |
| "grad_norm": 4.56980037689209, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.672, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1572052401746725, | |
| "grad_norm": 4.158092975616455, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.6214, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.1591460456089277, | |
| "grad_norm": 4.216987133026123, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.6047, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.16108685104318293, | |
| "grad_norm": 3.5317301750183105, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.6012, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.16302765647743814, | |
| "grad_norm": 12.014446258544922, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.9439, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.16496846191169334, | |
| "grad_norm": 22.006229400634766, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 1.583, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.16690926734594858, | |
| "grad_norm": 37.996788024902344, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.3416, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.16885007278020378, | |
| "grad_norm": 20.80458641052246, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.9229, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.170790878214459, | |
| "grad_norm": 20.84687042236328, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.8207, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.17273168364871422, | |
| "grad_norm": 17.68490982055664, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.6995, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.17467248908296942, | |
| "grad_norm": 25.93991470336914, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.5536, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17661329451722466, | |
| "grad_norm": 16.228805541992188, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.2994, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.17855409995147986, | |
| "grad_norm": 8.50417423248291, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.1472, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1804949053857351, | |
| "grad_norm": 11.803483009338379, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.1267, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1824357108199903, | |
| "grad_norm": 6.2093424797058105, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.0646, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1843765162542455, | |
| "grad_norm": 17.27790069580078, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.07, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.18631732168850074, | |
| "grad_norm": 6.931070804595947, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.1174, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.18825812712275594, | |
| "grad_norm": 10.341382026672363, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.0852, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.19019893255701115, | |
| "grad_norm": 6.310524940490723, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.053, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.19213973799126638, | |
| "grad_norm": 3.348977565765381, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.0262, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.19408054342552158, | |
| "grad_norm": 4.910148620605469, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.0362, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19408054342552158, | |
| "eval_loss": 1.0220495462417603, | |
| "eval_runtime": 82.6786, | |
| "eval_samples_per_second": 10.498, | |
| "eval_steps_per_second": 2.625, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19602134885977682, | |
| "grad_norm": 26.032215118408203, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 1.586, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.19796215429403202, | |
| "grad_norm": 13.921479225158691, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 0.9164, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.19990295972828723, | |
| "grad_norm": 4.853932857513428, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 0.5651, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.20184376516254246, | |
| "grad_norm": 2.3462417125701904, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.524, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.20378457059679767, | |
| "grad_norm": 2.373321533203125, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4601, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2057253760310529, | |
| "grad_norm": 1.6338130235671997, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.4492, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2076661814653081, | |
| "grad_norm": 1.6441563367843628, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 0.3925, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2096069868995633, | |
| "grad_norm": 1.5351994037628174, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 0.5077, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.21154779233381854, | |
| "grad_norm": 1.5161257982254028, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 0.4823, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.21348859776807375, | |
| "grad_norm": 1.9085850715637207, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.5749, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.21542940320232898, | |
| "grad_norm": 1.5658867359161377, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.4475, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.21737020863658418, | |
| "grad_norm": 1.9136401414871216, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.5457, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2193110140708394, | |
| "grad_norm": 1.8787047863006592, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.5437, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.22125181950509462, | |
| "grad_norm": 1.8258041143417358, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.5642, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.22319262493934983, | |
| "grad_norm": 1.7421337366104126, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.5053, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.22513343037360506, | |
| "grad_norm": 2.2626960277557373, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.6171, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.22707423580786026, | |
| "grad_norm": 2.2537620067596436, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.6538, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.22901504124211547, | |
| "grad_norm": 2.510111093521118, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.5928, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2309558466763707, | |
| "grad_norm": 2.4367897510528564, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.6158, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2328966521106259, | |
| "grad_norm": 2.847177743911743, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.6439, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2348374575448811, | |
| "grad_norm": 2.9993271827697754, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.6262, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.23677826297913634, | |
| "grad_norm": 2.2285447120666504, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.677, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.23871906841339155, | |
| "grad_norm": 2.3862266540527344, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.6068, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.24065987384764678, | |
| "grad_norm": 2.6299383640289307, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.7526, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.242600679281902, | |
| "grad_norm": 2.758273124694824, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.6645, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2445414847161572, | |
| "grad_norm": 3.015491008758545, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.7075, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.24648229015041243, | |
| "grad_norm": 3.425689935684204, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.8658, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.24842309558466763, | |
| "grad_norm": 5.679259777069092, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.7661, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.25036390101892286, | |
| "grad_norm": 4.268316268920898, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.7086, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2523047064531781, | |
| "grad_norm": 4.242010116577148, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.6903, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2542455118874333, | |
| "grad_norm": 3.7318978309631348, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.6262, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.2561863173216885, | |
| "grad_norm": 4.692379474639893, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.6059, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.25812712275594374, | |
| "grad_norm": 4.109679222106934, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.5912, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2600679281901989, | |
| "grad_norm": 4.626393795013428, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.699, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.26200873362445415, | |
| "grad_norm": 30.26133155822754, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 1.0972, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2639495390587094, | |
| "grad_norm": 19.875703811645508, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 1.2316, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.26589034449296456, | |
| "grad_norm": 19.294471740722656, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 1.0234, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2678311499272198, | |
| "grad_norm": 29.8734130859375, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.9117, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.269771955361475, | |
| "grad_norm": 25.97123908996582, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.8581, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2717127607957302, | |
| "grad_norm": 25.970752716064453, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.8129, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.27365356622998543, | |
| "grad_norm": 14.080256462097168, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.7298, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.27559437166424067, | |
| "grad_norm": 10.821141242980957, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.6428, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.2775351770984959, | |
| "grad_norm": 18.16455841064453, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.6751, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2794759825327511, | |
| "grad_norm": 12.17893123626709, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.5146, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2814167879670063, | |
| "grad_norm": 10.612948417663574, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.4776, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.28335759340126154, | |
| "grad_norm": 19.925426483154297, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.4856, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2852983988355167, | |
| "grad_norm": 20.155916213989258, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.362, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.28723920426977195, | |
| "grad_norm": 9.823814392089844, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.2978, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2891800097040272, | |
| "grad_norm": 10.883167266845703, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.2979, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.29112081513828236, | |
| "grad_norm": 7.7256340980529785, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.1943, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.29112081513828236, | |
| "eval_loss": 0.44719642400741577, | |
| "eval_runtime": 82.6454, | |
| "eval_samples_per_second": 10.503, | |
| "eval_steps_per_second": 2.626, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.61787760607232e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |