| { | |
| "best_metric": 6.177871227264404, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.5029337803855826, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003352891869237217, | |
| "grad_norm": 10.310164451599121, | |
| "learning_rate": 1e-05, | |
| "loss": 10.559, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003352891869237217, | |
| "eval_loss": 10.350041389465332, | |
| "eval_runtime": 2.3515, | |
| "eval_samples_per_second": 213.902, | |
| "eval_steps_per_second": 53.582, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006705783738474434, | |
| "grad_norm": 10.899702072143555, | |
| "learning_rate": 2e-05, | |
| "loss": 10.5371, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010058675607711651, | |
| "grad_norm": 9.904793739318848, | |
| "learning_rate": 3e-05, | |
| "loss": 10.4231, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013411567476948869, | |
| "grad_norm": 10.616510391235352, | |
| "learning_rate": 4e-05, | |
| "loss": 10.2717, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.016764459346186086, | |
| "grad_norm": 10.927786827087402, | |
| "learning_rate": 5e-05, | |
| "loss": 10.4633, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.020117351215423303, | |
| "grad_norm": 11.313493728637695, | |
| "learning_rate": 6e-05, | |
| "loss": 10.213, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02347024308466052, | |
| "grad_norm": 8.426546096801758, | |
| "learning_rate": 7e-05, | |
| "loss": 9.9917, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.026823134953897737, | |
| "grad_norm": 8.101781845092773, | |
| "learning_rate": 8e-05, | |
| "loss": 9.9345, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.030176026823134954, | |
| "grad_norm": 7.540008544921875, | |
| "learning_rate": 9e-05, | |
| "loss": 9.8567, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03352891869237217, | |
| "grad_norm": 7.71835470199585, | |
| "learning_rate": 0.0001, | |
| "loss": 9.7241, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.036881810561609385, | |
| "grad_norm": 7.580140590667725, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 9.4494, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.040234702430846606, | |
| "grad_norm": 7.461150646209717, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 9.2959, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04358759430008382, | |
| "grad_norm": 6.163961410522461, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 9.3115, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04694048616932104, | |
| "grad_norm": 5.853614330291748, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.009, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.050293378038558254, | |
| "grad_norm": 6.250462532043457, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 8.8049, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.053646269907795474, | |
| "grad_norm": 5.09857177734375, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.669, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05699916177703269, | |
| "grad_norm": 4.981748104095459, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 8.6501, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06035205364626991, | |
| "grad_norm": 4.854025840759277, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 8.7071, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06370494551550712, | |
| "grad_norm": 4.858048439025879, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 8.4474, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06705783738474434, | |
| "grad_norm": 5.338009357452393, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 8.1648, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07041072925398156, | |
| "grad_norm": 4.889835834503174, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 8.2165, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07376362112321877, | |
| "grad_norm": 5.622138977050781, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 7.931, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07711651299245599, | |
| "grad_norm": 5.048961639404297, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 8.1645, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08046940486169321, | |
| "grad_norm": 4.849130153656006, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 7.924, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08382229673093043, | |
| "grad_norm": 4.658658981323242, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 7.9077, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08717518860016764, | |
| "grad_norm": 4.544646739959717, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 7.8783, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09052808046940486, | |
| "grad_norm": 4.6037797927856445, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 7.7723, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09388097233864208, | |
| "grad_norm": 4.962968349456787, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 7.5476, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0972338642078793, | |
| "grad_norm": 5.028897285461426, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 7.6711, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10058675607711651, | |
| "grad_norm": 4.4867634773254395, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 7.7857, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10393964794635373, | |
| "grad_norm": 5.30435037612915, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 7.2347, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10729253981559095, | |
| "grad_norm": 4.072600841522217, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 7.5166, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11064543168482817, | |
| "grad_norm": 4.204601287841797, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 7.3602, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11399832355406538, | |
| "grad_norm": 3.4331986904144287, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 7.3824, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1173512154233026, | |
| "grad_norm": 3.4345741271972656, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 7.606, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12070410729253982, | |
| "grad_norm": 3.057077407836914, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 7.4587, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12405699916177704, | |
| "grad_norm": 3.7867636680603027, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 7.2907, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.12740989103101424, | |
| "grad_norm": 3.5214974880218506, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 7.2044, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13076278290025148, | |
| "grad_norm": 3.451026678085327, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 6.8786, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13411567476948869, | |
| "grad_norm": 3.36582612991333, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 6.9517, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1374685666387259, | |
| "grad_norm": 3.5128426551818848, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 6.6585, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.14082145850796313, | |
| "grad_norm": 2.9634475708007812, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 6.7733, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.14417435037720033, | |
| "grad_norm": 3.0367186069488525, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.791, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.14752724224643754, | |
| "grad_norm": 3.497209072113037, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 7.0197, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.15088013411567477, | |
| "grad_norm": 3.1642208099365234, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 6.3949, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15423302598491198, | |
| "grad_norm": 3.0919415950775146, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 6.4779, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15758591785414922, | |
| "grad_norm": 3.063319206237793, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 6.0493, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.16093880972338642, | |
| "grad_norm": 4.066051483154297, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 6.6208, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.16429170159262363, | |
| "grad_norm": 3.7912440299987793, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 6.3157, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.16764459346186086, | |
| "grad_norm": 4.728034973144531, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 7.6777, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16764459346186086, | |
| "eval_loss": 6.940974235534668, | |
| "eval_runtime": 2.3354, | |
| "eval_samples_per_second": 215.384, | |
| "eval_steps_per_second": 53.953, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17099748533109807, | |
| "grad_norm": 5.036187648773193, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 8.0701, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.17435037720033528, | |
| "grad_norm": 5.0193657875061035, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 8.0449, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1777032690695725, | |
| "grad_norm": 5.033466339111328, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 7.7675, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18105616093880972, | |
| "grad_norm": 4.792846202850342, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 7.7466, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18440905280804695, | |
| "grad_norm": 4.175502300262451, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 7.6881, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18776194467728416, | |
| "grad_norm": 4.628329753875732, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 7.3475, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.19111483654652137, | |
| "grad_norm": 4.324792861938477, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 7.1017, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1944677284157586, | |
| "grad_norm": 2.8266139030456543, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 7.4095, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1978206202849958, | |
| "grad_norm": 2.7232511043548584, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 7.3605, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.20117351215423301, | |
| "grad_norm": 2.7311155796051025, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 7.3004, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20452640402347025, | |
| "grad_norm": 2.7659690380096436, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 7.2224, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.20787929589270746, | |
| "grad_norm": 3.115063428878784, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 7.0739, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2112321877619447, | |
| "grad_norm": 2.8544118404388428, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 6.9763, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2145850796311819, | |
| "grad_norm": 2.846555233001709, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 7.2025, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2179379715004191, | |
| "grad_norm": 2.393690824508667, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 6.8528, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22129086336965634, | |
| "grad_norm": 2.448185920715332, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 6.8611, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.22464375523889354, | |
| "grad_norm": 2.406033992767334, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 6.747, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.22799664710813075, | |
| "grad_norm": 2.2242228984832764, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 6.9588, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.23134953897736799, | |
| "grad_norm": 2.746422052383423, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 6.8364, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2347024308466052, | |
| "grad_norm": 3.0429162979125977, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 6.6605, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23805532271584243, | |
| "grad_norm": 2.6157066822052, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 6.3457, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.24140821458507963, | |
| "grad_norm": 2.5024068355560303, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 6.5167, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.24476110645431684, | |
| "grad_norm": 2.3516902923583984, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 6.7865, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.24811399832355407, | |
| "grad_norm": 2.800886869430542, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 6.4162, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2514668901927913, | |
| "grad_norm": 2.8395519256591797, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 6.3922, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2548197820620285, | |
| "grad_norm": 2.524817943572998, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 6.2891, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2581726739312657, | |
| "grad_norm": 2.717404365539551, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 6.1493, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.26152556580050296, | |
| "grad_norm": 2.3688623905181885, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 6.1955, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.26487845766974016, | |
| "grad_norm": 2.5166866779327393, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 6.3359, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.26823134953897737, | |
| "grad_norm": 2.320457696914673, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 6.174, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2715842414082146, | |
| "grad_norm": 2.4732210636138916, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 6.5524, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2749371332774518, | |
| "grad_norm": 2.4983346462249756, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 6.2446, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.27829002514668905, | |
| "grad_norm": 2.350950002670288, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 6.1318, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.28164291701592625, | |
| "grad_norm": 2.0726985931396484, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 5.9952, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.28499580888516346, | |
| "grad_norm": 2.2949085235595703, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 6.1386, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28834870075440067, | |
| "grad_norm": 2.433393716812134, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 5.9786, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2917015926236379, | |
| "grad_norm": 2.5826408863067627, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 5.7708, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2950544844928751, | |
| "grad_norm": 2.6289031505584717, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 5.8766, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.29840737636211234, | |
| "grad_norm": 2.296018362045288, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 5.8882, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.30176026823134955, | |
| "grad_norm": 2.409599542617798, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 6.0668, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.30511316010058676, | |
| "grad_norm": 2.3355817794799805, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 6.4358, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.30846605196982396, | |
| "grad_norm": 2.571406364440918, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 6.2783, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.31181894383906117, | |
| "grad_norm": 2.281155824661255, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 6.0217, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.31517183570829843, | |
| "grad_norm": 2.6151113510131836, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 6.0906, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.31852472757753564, | |
| "grad_norm": 2.3498058319091797, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 6.0867, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.32187761944677284, | |
| "grad_norm": 2.6332180500030518, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 6.0301, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.32523051131601005, | |
| "grad_norm": 2.7580060958862305, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 6.2039, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.32858340318524726, | |
| "grad_norm": 3.8314266204833984, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 6.1379, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3319362950544845, | |
| "grad_norm": 3.0416414737701416, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 6.0274, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3352891869237217, | |
| "grad_norm": 9.444997787475586, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 7.7854, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3352891869237217, | |
| "eval_loss": 6.350980281829834, | |
| "eval_runtime": 2.34, | |
| "eval_samples_per_second": 214.954, | |
| "eval_steps_per_second": 53.845, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.33864207879295893, | |
| "grad_norm": 4.4523773193359375, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 7.649, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.34199497066219614, | |
| "grad_norm": 3.549696922302246, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 7.4084, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.34534786253143335, | |
| "grad_norm": 3.322315216064453, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 7.2737, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.34870075440067055, | |
| "grad_norm": 3.443470001220703, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 7.045, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3520536462699078, | |
| "grad_norm": 3.4193520545959473, | |
| "learning_rate": 5e-05, | |
| "loss": 6.8108, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.355406538139145, | |
| "grad_norm": 3.9259274005889893, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 6.7975, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.35875943000838223, | |
| "grad_norm": 3.5500876903533936, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 6.6573, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.36211232187761944, | |
| "grad_norm": 2.9335997104644775, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 6.6699, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.36546521374685664, | |
| "grad_norm": 2.8951656818389893, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 6.8695, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3688181056160939, | |
| "grad_norm": 2.834850788116455, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 6.7373, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3721709974853311, | |
| "grad_norm": 2.6709818840026855, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 6.8468, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3755238893545683, | |
| "grad_norm": 2.5505619049072266, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 6.6848, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3788767812238055, | |
| "grad_norm": 2.616159200668335, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 6.7016, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.38222967309304273, | |
| "grad_norm": 2.5756642818450928, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 6.5948, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.38558256496227994, | |
| "grad_norm": 2.640015125274658, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 6.599, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3889354568315172, | |
| "grad_norm": 2.656965732574463, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 6.6489, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3922883487007544, | |
| "grad_norm": 2.2515344619750977, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 6.2407, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3956412405699916, | |
| "grad_norm": 2.502758502960205, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 6.0948, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3989941324392288, | |
| "grad_norm": 2.3483591079711914, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 6.3413, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.40234702430846603, | |
| "grad_norm": 2.403675079345703, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 6.3964, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4056999161777033, | |
| "grad_norm": 3.0116775035858154, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 6.2897, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4090528080469405, | |
| "grad_norm": 3.0004560947418213, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 5.937, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4124056999161777, | |
| "grad_norm": 2.1759047508239746, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 6.2358, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4157585917854149, | |
| "grad_norm": 2.386132001876831, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 5.9525, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4191114836546521, | |
| "grad_norm": 2.041156053543091, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 5.9755, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4224643755238894, | |
| "grad_norm": 2.2770140171051025, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 5.8259, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4258172673931266, | |
| "grad_norm": 2.0245702266693115, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 5.8435, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4291701592623638, | |
| "grad_norm": 2.377305269241333, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 5.8376, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.432523051131601, | |
| "grad_norm": 2.8456356525421143, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 6.233, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4358759430008382, | |
| "grad_norm": 2.39567494392395, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 5.8123, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4392288348700754, | |
| "grad_norm": 2.140490770339966, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 6.13, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4425817267393127, | |
| "grad_norm": 2.1710660457611084, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 5.9013, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4459346186085499, | |
| "grad_norm": 2.3693878650665283, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 6.1766, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4492875104777871, | |
| "grad_norm": 2.3662402629852295, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 5.7176, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4526404023470243, | |
| "grad_norm": 2.3599185943603516, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.7964, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4559932942162615, | |
| "grad_norm": 2.625692367553711, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 5.5996, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.45934618608549876, | |
| "grad_norm": 2.4023356437683105, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.57, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.46269907795473597, | |
| "grad_norm": 2.0743892192840576, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 5.7452, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4660519698239732, | |
| "grad_norm": 2.321120023727417, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 5.7244, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4694048616932104, | |
| "grad_norm": 1.9818140268325806, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 5.8637, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4727577535624476, | |
| "grad_norm": 2.0799217224121094, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 5.9869, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.47611064543168485, | |
| "grad_norm": 2.1854805946350098, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 5.8397, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.47946353730092206, | |
| "grad_norm": 2.3658244609832764, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 5.7513, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.48281642917015927, | |
| "grad_norm": 2.870225429534912, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 6.3676, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4861693210393965, | |
| "grad_norm": 2.556222915649414, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 5.7333, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4895222129086337, | |
| "grad_norm": 2.5183026790618896, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 5.7318, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4928751047778709, | |
| "grad_norm": 3.7576239109039307, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 5.3302, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.49622799664710815, | |
| "grad_norm": 3.4823052883148193, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 5.74, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.49958088851634536, | |
| "grad_norm": 4.952392578125, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 6.3364, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5029337803855826, | |
| "grad_norm": 4.829474449157715, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 6.0675, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5029337803855826, | |
| "eval_loss": 6.177871227264404, | |
| "eval_runtime": 2.3253, | |
| "eval_samples_per_second": 216.319, | |
| "eval_steps_per_second": 54.187, | |
| "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": 114471469056000.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |