| { | |
| "best_metric": 0.37483757734298706, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 2.191780821917808, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010958904109589041, | |
| "grad_norm": 167.30447387695312, | |
| "learning_rate": 1e-05, | |
| "loss": 6.8803, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010958904109589041, | |
| "eval_loss": 1.3351311683654785, | |
| "eval_runtime": 1.9423, | |
| "eval_samples_per_second": 79.286, | |
| "eval_steps_per_second": 20.079, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.021917808219178082, | |
| "grad_norm": 96.19314575195312, | |
| "learning_rate": 2e-05, | |
| "loss": 4.1016, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03287671232876712, | |
| "grad_norm": 95.96031188964844, | |
| "learning_rate": 3e-05, | |
| "loss": 3.6096, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.043835616438356165, | |
| "grad_norm": 89.8621597290039, | |
| "learning_rate": 4e-05, | |
| "loss": 4.8139, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 141.93055725097656, | |
| "learning_rate": 5e-05, | |
| "loss": 4.9924, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06575342465753424, | |
| "grad_norm": 116.39649200439453, | |
| "learning_rate": 6e-05, | |
| "loss": 4.2838, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07671232876712329, | |
| "grad_norm": 109.42125701904297, | |
| "learning_rate": 7e-05, | |
| "loss": 3.6636, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08767123287671233, | |
| "grad_norm": 98.26384735107422, | |
| "learning_rate": 8e-05, | |
| "loss": 3.6923, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09863013698630137, | |
| "grad_norm": 96.4810562133789, | |
| "learning_rate": 9e-05, | |
| "loss": 3.2082, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 83.88074493408203, | |
| "learning_rate": 0.0001, | |
| "loss": 2.9755, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12054794520547946, | |
| "grad_norm": 79.13579559326172, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 3.4502, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.13150684931506848, | |
| "grad_norm": 98.4308090209961, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 3.0756, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14246575342465753, | |
| "grad_norm": 103.85986328125, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 3.5216, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15342465753424658, | |
| "grad_norm": 107.19340515136719, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 2.8189, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 104.6748275756836, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 3.6509, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17534246575342466, | |
| "grad_norm": 106.12808990478516, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 4.2083, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1863013698630137, | |
| "grad_norm": 119.818115234375, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 4.0229, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.19726027397260273, | |
| "grad_norm": 95.47771453857422, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 3.595, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20821917808219179, | |
| "grad_norm": 82.28040313720703, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 3.1048, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 68.5257339477539, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 2.9802, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23013698630136986, | |
| "grad_norm": 88.13601684570312, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 3.6063, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2410958904109589, | |
| "grad_norm": 128.11737060546875, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 4.7627, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.25205479452054796, | |
| "grad_norm": 161.507568359375, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 5.7056, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.26301369863013696, | |
| "grad_norm": 93.0267333984375, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 2.9235, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 60.84797286987305, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 2.7503, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.28493150684931506, | |
| "grad_norm": 118.22985076904297, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 3.2503, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2958904109589041, | |
| "grad_norm": 83.68474578857422, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 2.633, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.30684931506849317, | |
| "grad_norm": 74.43077087402344, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 2.8113, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3178082191780822, | |
| "grad_norm": 72.55152130126953, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.7201, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 42.331382751464844, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 2.0678, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33972602739726027, | |
| "grad_norm": 43.60592269897461, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 2.1782, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3506849315068493, | |
| "grad_norm": 62.631996154785156, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 2.4158, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.36164383561643837, | |
| "grad_norm": 39.32170104980469, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 1.9671, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3726027397260274, | |
| "grad_norm": 45.91651153564453, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 2.4934, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 45.4543571472168, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 2.1458, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.39452054794520547, | |
| "grad_norm": 39.656856536865234, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 2.0924, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4054794520547945, | |
| "grad_norm": 38.19993591308594, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 1.7267, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41643835616438357, | |
| "grad_norm": 41.50175476074219, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 1.964, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4273972602739726, | |
| "grad_norm": 54.87953567504883, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 2.4834, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 39.60464859008789, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 2.3171, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.44931506849315067, | |
| "grad_norm": 34.88728713989258, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 1.6985, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4602739726027397, | |
| "grad_norm": 34.25281524658203, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 1.5808, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4712328767123288, | |
| "grad_norm": 53.074317932128906, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 2.2194, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4821917808219178, | |
| "grad_norm": 61.571441650390625, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 3.0192, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 70.4915771484375, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 3.7234, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5041095890410959, | |
| "grad_norm": 45.693336486816406, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 1.9188, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5150684931506849, | |
| "grad_norm": 42.399410247802734, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 1.7099, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5260273972602739, | |
| "grad_norm": 47.592716217041016, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.764, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.536986301369863, | |
| "grad_norm": 66.78082275390625, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 2.0808, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 57.66780471801758, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 2.48, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "eval_loss": 0.4879509210586548, | |
| "eval_runtime": 1.9779, | |
| "eval_samples_per_second": 77.86, | |
| "eval_steps_per_second": 19.718, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5589041095890411, | |
| "grad_norm": 43.39772033691406, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 1.581, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5698630136986301, | |
| "grad_norm": 31.633132934570312, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 1.6626, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5808219178082191, | |
| "grad_norm": 46.863033294677734, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 1.766, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5917808219178082, | |
| "grad_norm": 53.44274139404297, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 1.7055, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 61.89936447143555, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 1.815, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6136986301369863, | |
| "grad_norm": 70.4791030883789, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 1.9802, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6246575342465753, | |
| "grad_norm": 68.4205551147461, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 1.5906, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6356164383561644, | |
| "grad_norm": 62.756404876708984, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 1.6494, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6465753424657534, | |
| "grad_norm": 61.685237884521484, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 1.5551, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 62.66923522949219, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 1.7507, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6684931506849315, | |
| "grad_norm": 87.8627700805664, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 1.641, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6794520547945205, | |
| "grad_norm": 85.07118225097656, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 2.1592, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6904109589041096, | |
| "grad_norm": 60.446292877197266, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 1.967, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.7013698630136986, | |
| "grad_norm": 48.728492736816406, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 1.6903, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 55.193328857421875, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 1.4474, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7232876712328767, | |
| "grad_norm": 83.90128326416016, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 2.3551, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7342465753424657, | |
| "grad_norm": 72.22972869873047, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.2786, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7452054794520548, | |
| "grad_norm": 71.7395248413086, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 2.1909, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7561643835616438, | |
| "grad_norm": 44.32119369506836, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 1.8315, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 38.783790588378906, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 1.6491, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7780821917808219, | |
| "grad_norm": 32.161075592041016, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 1.3311, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7890410958904109, | |
| "grad_norm": 48.930931091308594, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 1.9247, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 46.455318450927734, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 1.7578, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.810958904109589, | |
| "grad_norm": 48.81123733520508, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 1.738, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 42.357940673828125, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 1.5499, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8328767123287671, | |
| "grad_norm": 86.15032958984375, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 1.789, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8438356164383561, | |
| "grad_norm": 55.27198791503906, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 1.8107, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8547945205479452, | |
| "grad_norm": 73.23171997070312, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 2.21, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8657534246575342, | |
| "grad_norm": 37.8917350769043, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 1.6816, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 39.092281341552734, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 1.4816, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8876712328767123, | |
| "grad_norm": 39.103973388671875, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 1.5351, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8986301369863013, | |
| "grad_norm": 58.27376937866211, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 2.1233, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9095890410958904, | |
| "grad_norm": 76.3632583618164, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 2.2209, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9205479452054794, | |
| "grad_norm": 79.86170959472656, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 3.1822, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 39.175010681152344, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 1.7872, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9424657534246575, | |
| "grad_norm": 40.35182571411133, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.5002, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9534246575342465, | |
| "grad_norm": 37.38121795654297, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 1.4619, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9643835616438357, | |
| "grad_norm": 70.59363555908203, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 2.13, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9753424657534246, | |
| "grad_norm": 51.04860305786133, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 1.5764, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 34.680580139160156, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 1.5032, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9972602739726028, | |
| "grad_norm": 31.850120544433594, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 1.5399, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0082191780821919, | |
| "grad_norm": 71.03590393066406, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 2.3794, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0191780821917809, | |
| "grad_norm": 65.01827239990234, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 2.4749, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0301369863013699, | |
| "grad_norm": 34.10368728637695, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 1.4223, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 37.319000244140625, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 1.5575, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0520547945205478, | |
| "grad_norm": 31.34611701965332, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 1.4979, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.063013698630137, | |
| "grad_norm": 28.475486755371094, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 1.6096, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.073972602739726, | |
| "grad_norm": 34.53931427001953, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 1.6613, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.084931506849315, | |
| "grad_norm": 85.2432632446289, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 1.2195, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 23.506343841552734, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 1.062, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_loss": 0.400832861661911, | |
| "eval_runtime": 1.9406, | |
| "eval_samples_per_second": 79.357, | |
| "eval_steps_per_second": 20.097, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.106849315068493, | |
| "grad_norm": 28.613204956054688, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 1.3003, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1178082191780823, | |
| "grad_norm": 28.404552459716797, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 1.3235, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1287671232876713, | |
| "grad_norm": 31.971712112426758, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 1.5493, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1397260273972603, | |
| "grad_norm": 31.164323806762695, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 1.3944, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 27.70460319519043, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2497, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1616438356164385, | |
| "grad_norm": 42.503597259521484, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 1.1932, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1726027397260275, | |
| "grad_norm": 27.519493103027344, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 1.3152, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.1835616438356165, | |
| "grad_norm": 30.81327247619629, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 1.7175, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1945205479452055, | |
| "grad_norm": 36.185123443603516, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 1.9324, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 30.46416473388672, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 1.2068, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2164383561643834, | |
| "grad_norm": 22.46263885498047, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 1.1815, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2273972602739727, | |
| "grad_norm": 20.89829444885254, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 1.2924, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2383561643835617, | |
| "grad_norm": 43.927669525146484, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 1.9196, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2493150684931507, | |
| "grad_norm": 55.13964080810547, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 2.5384, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 71.3852310180664, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 1.8759, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2712328767123289, | |
| "grad_norm": 20.873424530029297, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 1.2816, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2821917808219179, | |
| "grad_norm": 30.306793212890625, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 1.3711, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.2931506849315069, | |
| "grad_norm": 27.79682159423828, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 1.426, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.3041095890410959, | |
| "grad_norm": 49.33466339111328, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 1.464, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 35.36948013305664, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 1.618, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3260273972602739, | |
| "grad_norm": 38.644927978515625, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 1.3514, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.336986301369863, | |
| "grad_norm": 29.98199462890625, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 1.3034, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.347945205479452, | |
| "grad_norm": 39.6750602722168, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 1.3634, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.358904109589041, | |
| "grad_norm": 46.256343841552734, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.5002, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 32.30176544189453, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 1.4205, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3808219178082193, | |
| "grad_norm": 32.818485260009766, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 1.2477, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3917808219178083, | |
| "grad_norm": 45.166534423828125, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 1.4078, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.4027397260273973, | |
| "grad_norm": 32.647891998291016, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 1.2282, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.4136986301369863, | |
| "grad_norm": 40.9683837890625, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 1.2926, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 51.406612396240234, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 1.9739, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4356164383561643, | |
| "grad_norm": 45.16895294189453, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 1.6762, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4465753424657535, | |
| "grad_norm": 44.308189392089844, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 1.3818, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4575342465753425, | |
| "grad_norm": 35.05088424682617, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 1.1228, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4684931506849315, | |
| "grad_norm": 40.73629379272461, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 1.2048, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 77.88436889648438, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 1.837, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.4904109589041097, | |
| "grad_norm": 125.36991882324219, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 2.7284, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.5013698630136987, | |
| "grad_norm": 57.473472595214844, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 1.4545, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.5123287671232877, | |
| "grad_norm": 48.656131744384766, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 1.3017, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5232876712328767, | |
| "grad_norm": 45.12188720703125, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 1.4474, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 69.8457260131836, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 1.3396, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5452054794520547, | |
| "grad_norm": 53.79582977294922, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 1.5312, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.5561643835616439, | |
| "grad_norm": 151.50416564941406, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 1.7986, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.5671232876712329, | |
| "grad_norm": 49.776119232177734, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.1307, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5780821917808219, | |
| "grad_norm": 46.405677795410156, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 1.4035, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 55.024227142333984, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 1.2362, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 54.06251907348633, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 1.1771, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.610958904109589, | |
| "grad_norm": 69.43602752685547, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 1.627, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.621917808219178, | |
| "grad_norm": 47.403690338134766, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 1.161, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.632876712328767, | |
| "grad_norm": 63.99250030517578, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 1.3082, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 64.2621078491211, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 1.3423, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "eval_loss": 0.37483757734298706, | |
| "eval_runtime": 1.9385, | |
| "eval_samples_per_second": 79.444, | |
| "eval_steps_per_second": 20.119, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.654794520547945, | |
| "grad_norm": 48.69328689575195, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 1.1261, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.6657534246575343, | |
| "grad_norm": 76.65846252441406, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 2.0332, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6767123287671233, | |
| "grad_norm": 93.13089752197266, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 1.4993, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6876712328767123, | |
| "grad_norm": 66.61369323730469, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 1.8902, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 39.77993392944336, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 1.1385, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7095890410958905, | |
| "grad_norm": 50.28946304321289, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 1.2971, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7205479452054795, | |
| "grad_norm": 100.63103485107422, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 1.7498, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7315068493150685, | |
| "grad_norm": 112.16035461425781, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 2.1322, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7424657534246575, | |
| "grad_norm": 112.50196075439453, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 2.7743, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 40.098724365234375, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 0.8506, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.7643835616438355, | |
| "grad_norm": 66.71056365966797, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 1.1509, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.7753424657534247, | |
| "grad_norm": 60.821983337402344, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.2879, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7863013698630137, | |
| "grad_norm": 81.99188232421875, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 1.1506, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7972602739726027, | |
| "grad_norm": 58.01636505126953, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 1.348, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 54.829917907714844, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 1.0759, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.819178082191781, | |
| "grad_norm": 63.361385345458984, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 1.1913, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.83013698630137, | |
| "grad_norm": 61.732547760009766, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 1.1642, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.841095890410959, | |
| "grad_norm": 93.66639709472656, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 1.1224, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8520547945205479, | |
| "grad_norm": 81.65064239501953, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 1.4895, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 51.923160552978516, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 1.2478, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.8739726027397259, | |
| "grad_norm": 45.89485168457031, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 1.3706, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.884931506849315, | |
| "grad_norm": 48.91796112060547, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 1.3465, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.895890410958904, | |
| "grad_norm": 50.174766540527344, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 1.348, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.9068493150684933, | |
| "grad_norm": 63.196136474609375, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 1.3328, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 74.82865905761719, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 1.4565, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9287671232876713, | |
| "grad_norm": 92.00929260253906, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 1.8394, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9397260273972603, | |
| "grad_norm": 42.00975036621094, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 1.0673, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9506849315068493, | |
| "grad_norm": 60.778114318847656, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 1.2479, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9616438356164383, | |
| "grad_norm": 81.8672103881836, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 1.3253, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 133.6543426513672, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 1.4278, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.9835616438356163, | |
| "grad_norm": 55.93128204345703, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.3203, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9945205479452055, | |
| "grad_norm": 90.5474624633789, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 1.5685, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.0054794520547947, | |
| "grad_norm": 67.73770904541016, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 1.3844, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.0164383561643837, | |
| "grad_norm": 161.63433837890625, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 3.5821, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 50.99284362792969, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 1.0625, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.0383561643835617, | |
| "grad_norm": 55.48767852783203, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 0.9959, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.0493150684931507, | |
| "grad_norm": 51.87129211425781, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 1.193, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0602739726027397, | |
| "grad_norm": 120.1318130493164, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 1.5371, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.0712328767123287, | |
| "grad_norm": 53.63728332519531, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 1.2595, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 85.54694366455078, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 1.6823, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.0931506849315067, | |
| "grad_norm": 47.97751998901367, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 1.126, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.1041095890410957, | |
| "grad_norm": 62.840545654296875, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 1.2285, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.115068493150685, | |
| "grad_norm": 51.649688720703125, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 1.0927, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.126027397260274, | |
| "grad_norm": 53.68096160888672, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 1.5724, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 56.60838317871094, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 1.3107, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.147945205479452, | |
| "grad_norm": 54.82625198364258, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 1.316, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.158904109589041, | |
| "grad_norm": 66.14631652832031, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 1.3025, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.16986301369863, | |
| "grad_norm": 48.34274673461914, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 1.215, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.180821917808219, | |
| "grad_norm": 95.66395568847656, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 1.7791, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 67.01224517822266, | |
| "learning_rate": 0.0, | |
| "loss": 1.7992, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_loss": 0.37968552112579346, | |
| "eval_runtime": 1.9377, | |
| "eval_samples_per_second": 79.474, | |
| "eval_steps_per_second": 20.127, | |
| "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": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5587016582365184e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |